United States
Environmental Protection
Agency
Office of Water
Regulations and Standards
Washington, D.C. 20460
EPA 440/5-81-010
Water
&EPA RESTORATION OF LAKES
AND INLAND WATERS
-------
REVIEW NOTICE
This report has been reviewed by the Office of
Water Regulations and Standards, EPA, and approved
for publication. Approval does not signify that the
contents necessarily reflect the views and policies of
the Environmental Protection Agency, nor does
mention of trade names or commercial products
constitute endorsement or recommendations for use.
EPA 440/5-81-010
-------
RESTORATION OF LAKES
AND INLAND WATERS
International Symposium
on Inland Waters and
Lake Restoration
September 8-12, 1980
Portland, Maine
U.S.ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF WATER REGULATIONS AND STANDARDS
WASHINGTON. D. C.
December 1980
-------
FOREWORD
This second biennial report on the protection and restoration
of our Nation's freshwater lakes comes at a particularly pro-
pitious time. Not only our own country, but other nations are
beginning to demonstrate significant progress in meeting the
environmental challenges to restore and protect freshwater
lakes.
Section 304(j) of the 1977 Amendments to the Clean Water
Act requires the U.S. Environmental Protection Agency (EPA)
to publish a biennial report on "methods, procedures, and pro-
cesses as may be appropriate to restore and enhance the quality
of the Nation's publicly-owned freshwater lakes." To fulfill that
legislative requirement EPA joined with the Organization for
Economic and Cooperative Development (OECD), which just
completed a decade-long study of eutrophication, to sponsor
an International Symposium on Inland Waters and Lake
Restoration.
Scientists and project managers involved with freshwater
lakes projects throughout the world presented the results of
their investigations to this Symposium. Ninety-one presenta-
tions are published in these proceedings.
Seven hundred fifty (750) people attended from 35 countries
and 46 States to hear and discuss the state-of-the-art. This
demonstrates the intense interest in lake restoration. Two years
ago, 460 attended the conference in Minneapolis.
We are learning to understand what creates problems in our
lakes. Lake restoration and protection is a developing science. It
is a challenge both to seasoned scientists who have worked
with environmental problems for many years and to their
younger counterparts. Both came to this Symposium, and
shared the podium to explain innovative investigations that
have produced methods and procedures that are working.
Freshwater lakes are being protected and restored in the
United States and throughout the world. We expect the
momentum behind this effort to grow stronger and become
even more effective over the next few years, as the States take
over the responsibility for the restoration and protection of their
publicly-owned freshwater lakes from the Federal government.
Steven Schatzow
Deputy Assistant Administrator
Office of Water Regulations and
Standards
-------
INTRODUCTION
The U.S. Environmental Protection Agency's Clean Lakes
Program has demonstrated that principal causes of our lake
problem can be identified, and comprehensive, cost-effective
solutions can be developed and successfully implemented for
fnost of our lake problems.
The symptoms of lake eutrophication are obvious to the
public. Aquatic weeds interfere with swimming and boating.
Observers notice that the once clear water has become increas-
ingly dark and murky. Fishermen rely on their memories to
describe the excitement of the sport fishery now diminished.
Eutrophication is not the only threat to our enjoyment and
use of our Nation's lakes. Heightened public awareness of the
dangers and widespread occurrence of toxic pollutants has
stimulated concern for the presence of toxics in lakes. Heavy
metals and synthetic organic compounds resulting from ex-
panding urbanization and industrialization may be interfering
with the natural systems of our lakes. Toxics impact the ability
of lake users to enjoy lake recreational activities and the quality
of their drinking water and fish harvested from the lakes.
Numerous techniques have been developed to reduce the
availability of nutrients and slow down the eutrophication pro-
cess. Simply treating the problems, however, is not the most
effective solution. The Clean Lakes Program has demonstrated
that controlling the cause of the problem is the best approach.
Point source and nonpoint source loadings in the lake water-
shed must be reduced to an acceptable level to achieve the
desired improvement in lake water quality.
In many instances, nonpoint source control is the most im-
portant aspect of lake restoration. The complexity and exten-
siveness of nonpoint sources make their control difficult. In cer-
tain lake watersheds, thousands of acres of forests, agricultural
land, and urbanized areas must be studied and subjected to ef-
fective management strategies. Public awareness of the con-
nection between watershed activities and lake water quality is
essential.
Lake watershed management strategies have used contour
plowing, manure handling systems, timber harvesting prac-
tices, street sweeping, and stormwater and sedimentation
basins to successfully prevent these nonpoint source pollutants
from entering the lake.
The science of lake protection and restoration has pro-
gressed over the past few years. It is essential to continue these
efforts to improve our knowledge of problems and our abilities
to solve them. The future of our lakes is bright, provided we
continue to:
• Improve this Nation's capability to predict which pollutants
have the greatest impact on our lakes so that lakes can be
preserved and restored in the most cost-effective way.
• Improve the effectiveness of pollution control and treat-
ment methods, while at the same time, reduce their costs and
use of energy.
• Find constructive uses for water materials such as those
dredged from lake bottoms and the byproducts of farming
(manure) and industry (toxics).
• Improve our understanding of the eutrophication process
and nutrient recycling within lakes—mechanisms, importance,
and impact on the lake systems.
• Evaluate lake restoration to improve the technology,
quality, and duration of lake pollution control.
• Develop innovative pollution control and treatment
technology addressing toxics.
Learning from others' experience is a basic truism of human
existence, and it certainly applies to the rapidly growing science
of limnology. Lakes throughout the world share similar prob-
lems. Various techniques to solve lake problems, whether
developed from farm ponds in Australia or reservoirs in Ger-
many, may be applicable to similar situations wherever they
exist.
Not only scientists, but laymen must learn the fundamentals
of lake ecosystems and their protection and restoration. Lakes'
problems are not solely the province of limnologists, nor of
governmental officials, but of the people who use and are af-
fected by those lakes. A Clean Lakes Program is by its very
nature a "grassroots program." That's what makes the Clean
Lakes Program work.
Government will continue to play an essential role in restoring
lakes, but it will be Government closest to those whose lakes
need help. State and local Governments are best able to set
their own priorities for the restoration and/or protection of lakes
and integrate lakes into their total water quality management
program.
-------
CONTENTS
Foreword iii
Introduction iv
OPENING SESSION
Welcome
Leslie Carothers
Whatever Became of Shagawa Lake? 67
David Larsen
A Retrospective Look at the Effects
of Phosphorus Removal in Lakes 73
Val Smith
Significance of Sediments in Lake Nutrient Balance .... 78
H. L Golterman
The U.S. EPA Clean Lakes Program 2
Steven Schatzow
Local Commitment to Lake Restoration: the
Cobbossee Watershed Example 4
Thomas Gordon
The Eutrophication Story Since Madison, 1967 10
A. F. Bartsch
North American OECD Eutrophication Project:
the United States Study 17
Walter Rast
Monitoring of Inland Waters: the Nordic Project 19
Sven-Olof Ryding
OECD Eutrophication Program Regional Project:
Alpine Lakes 21
Hansjorg Fricker
The Shallow Lakes and Reservoirs Project 23
Jurgen Clasen
Background and Summary Results of the OECD Cooperative
Program on Eutrophication 25
Vollen welder/Kerekes
FACTORS INFLUENCING THE DYNAMICS OF
EUTROPHICATION
Present Knowledge on Limiting Nutrients
Curt Forsberg
Non-nutrient Factors Influencing
the Dynamics of Eutrophication .
Dieter Imboden
Dynamics of Nutrient Enrichment in
Large Lakes: the Lake Michigan Case
Claire L Schelske
37
38
41
Modeling the Response of the Nuisance Alga,
Cladophora glomerata, to Reductions in
Phosphorus Loading 47
Martin Auer
DREDGING AND BIOMANIPULATION AS RESTORA-
TION TECHNIQUES
Predicting Dredging Depths to Minimize
Internal Nutrient Recycling in Shallow Lakes 79
Heinz G, Stefan
Dredging Activities in Wisconsin's Lake
Renewal Program
Russell C. Dunst
86
Nutting Lake Restoration Project: a Case Study 89
David D. Worth. Jr.
Mercury Speciation and Distribution in a Polluted River-Lake
System as Related to the Problem of Lake Restoration . 93
Togwell A. Jackson
Simplified Ecosystem Modeling for Assessing
Alternative Biomanipulation Strategies 102
Mark L Hutchins
Response of Zooplankton in Precambrian Shield Lakes to
Whole-Lake Chemical Modifications Causing pH Change 108
Diane F. Ma/ley
Sediment Treatment for Phosphorus Inactivation 115
Guy Barroin
Two Examples of Urban Stormwater Impoundment for
Aesthetics and for Protection of Receiving Waters 119
Thomas G. Brydges
AERATION/MIXING AND AQUATIC PLANT
HARVESTING AS RESTORATION TECHNIQUES
Review of Aeration/Circulation for Lake Management 124
Robert Pastorok
Predicting the Algal Response to Destratification 134
Bruce Forsberg
Reservoir Mixing Techniques: Recent Experience
in the UK
D. Johnson
,140
Roles of Materials Exported into Rivers and Reservoirs
in the Nutrition of Cladoceran Zooplankton 53
G. Richard Marzolf
Case Studies of Aquatic Plant Management for Lake
Preservation and Restoration in British Columbia,
Canada 146
Peter R. Newrow
NUTRIENT LOADING/TROPHIC RESPONSE
Methods of Assessing Nutrient Loading
Hansjorg Fricker
Quantification of Phosphorus Input to Lakes
and Its Impact on Trophic Conditions
Riaz Ahmed
56
61
German Experience in Reservoir Management
and Control 153
Jurgen Clasen
The Efficacy of Weed Harvesting for Lake Restoration . 158
Darrell L. King
-------
PUBLIC BENEFIT AND INSTITUTIONAL PROBLEMS
Lake Restoration - a Historical Perspective 162
Kenneth M. Mackenthun
Benefits and Problems of Eutrophication Control 166
D. J. Gregor
The Politics of Benefit Estimation 172
David J. A/lee
Clean Lakes Estimation System 177
Neils B. Christiansen
USDA Soil Conservation Service Standards
for Livestock Manure Management Practices
Charles E. Fogg
260
Agricultural Nonpoint Source Control of Phosphorus as a
Remedy to Eutrophication of a Drinking Water Supply 265
Mark P. Brown
Reservoir Protection by In-river Nutrient Reduction .... 272
Heinz Bernhardt
Agricultural Pollution Control in the Netherlands 278
H. L. Golterman
Impacts of Lake Protection on a Small Urban Community 182
Nico/aas W. Bouwes, Sr.
Lake Management and Cost-Benefit Analysis in Ontario 187
Peter A. Victor
The Leman Commission 192
Guy Barroin
Structure, Aims and Activities of the International Alpine
Commissions in Europe 195
Oscar Ravera
Institutional Arrangements for Shoreland
Protection and Lake Management in Wisconsin 197
Douglas A. Yanggen
SPECIAL PROJECTS ANDTOPICS FOR ASSESSING THE
TROPHIC STATE
Sampling Strategies for Estimating Chlorophyll
Standing Crops in Stratified Lakes 203
Robert Stauffer
The Influence of Nutrient Enrichment
on Freshwater Zooplankton
Oscar Ravera
Using Trophic State Indices to Examine the
Dynamics of Eutrophication
Robert £. Carlson
210
218
Regression Analysis of Reservoir Water Quality
Parameters with Digital Satellite Reflectance Data 222
Herbert.!. Grimshaw
URBAN AND POINT SOURCE POLLUTION CONTROL
TECHNOLOGY
Urban Stormwater/Combined Sewage
Management and Pollution Abatement Alternatives ... 279
Richard P. Traver
The Great Lakes: an Experiment in Technological
Innovation and Institutional Cooperation 290
Madonna F. McGrath
Design of Storage/Sedimentation Facilities to Control
Urban Runoff and Combined Sewer Overflows 294
W. Michael Stallard
Swedish Experience of Nutrient Removal
from Wastewater
Curt Forsberg
298
Stormwater Pollution Controls for Lake Management . 304
William C. Pisano
An Example of Urban Watershed Management
for Improving Lake Water Quality 307
Martin P. Wanielista
Lake Restoration by Effluents Diversion in France 312
Guy Barroin
MODELING AND ASSESSMENT OF THE TROPHIC STATE
Phosphorus Balance and Predictions:
Lake Constance, Obersee 316
G. Wagner
Lake Assessment in Preparation for a
Multiphase Restoration Treatment . ..
William H. Funk
226
The Continuing Dilution of Moses Lake, Washington .. 238
Eugene B. Welch
Managing Aquatic Plants with Fiberglas Screens 245
Michael A. Perkins
RURAL WATERSHED POLLUTION CONTROL
Relationships Between Agricultural Practices
and Receiving Water Quality 249
Frank J. Humenik
Source Control of Animal Wastes for Lake Watersheds 257
Lynn R. Schuyler
Prediction of Total Nitrogen in Lakes and Reservoirs
Roger W. Bachmann
320
An Incremental Phosphorus Loading Change
Approach for Prediction Error Reduction 325
Kenneth H. Reckhow
Application of Phosphorus Loading Models to
River-run Lakes and Other Incompletely Mixed Systems 329
Steven C. Chapra
The Application of the Lake Eutrophication Game SSWIMS
to the Management of Lake George, New York 335
Jay Bloomfield
Variability of Trophic State Indicators in Reservoirs ... 344
William W. Walker. Jr.
Reservoir Water Quality Sampling Design
Kent W. Thornton
349
-------
HEALTH-RELATED PROBLEMS
Health Aspects of Eutrophication 356
Michael J. Suess
General Impacts of Eutrophication on
Potable Water Preparation 359
Heinz Bernhardt
Organic Contaminants in the Great Lakes 364
David E. Weininger
Organochlorinated Compounds in Drinking Water
as a Result of Eutrophication 373
Gerard Dorin
The Impact of Toxic Trace Elements on Inland Waters
with Emphasis on Lead in Lake Michigan 379
Alan W. Elzerman
Waterborne Giardiasis 386
Edwin C. Lippy
Residential Well Water Quality in Wisconsin Inland
Lake Communities 390
George R. Gibson. Jr.
NUTRIENT PREVENTION AND INACTIVATION
Phosphorus Inactivation: a Summary of
Knowledge and Research Needs 395
G. Dennis Cooke
Control of Toxic Blue-Green Algae in Farm Dams 400
Valerie May
Aluminum Sulfate Dose Determination
and Application Techniques 405
Robert H. Kennedy
A Comparison of Two Alum Treated Lakes in Wisconsin 412
Doug/as R. Knauer
Hypolimnetic Aluminum Treatment of
Softwater Annabessacook Lake 417
David R. Dominie, II
Medical Lake Improvement Project: a Success Story .. 424
A. F. Gasperino
Detergent Modification: Scandinavian Experiences — 429
Curt Forsberg
Responses of Fishes to Acidification of
Streams and Lakes in Eastern North America
Terry A. Haines
Future Trends in Acid Precipitation and
Possible Programs
James R. Kramer
Mutual Relationship pH/Eutrophication-Acid Rain
H. L Go/terman
SPECIAL TOPICS
An Evaluation of Methods for Measuring
the Groundwater Contribution to Perch Lake
David R. Lee
467
474
... 479
Rehabilitation Project for a Quebec Lake:
Waterloo Lake, Near Montreal
Francois Guimont
480
485
Quantification of Allochthonous Organic Input to Cherokee
Reservoir: Implications for Hypolimnetic
Oxygen Depletions 489
Richard C. Young
Lake Restoration Methods Developed and
Used in Sweden
Wiihelm Ripl
495
APPENDIXES
A: Summary of Clean Lakes Project 501
B: Symposium Participants 520
C: Symposium Attendees 525
THE ACID RAIN PROBLEM: MECHANISM AND EFFECTS
The Long Range Transport of Air Pollution
and Acid Rain Formation 432
Brynjulf Ottar
Effects of Acid Precipitation on Aquatic
and Terrestrial Ecosystems 438
A me To/Ian
Changing pH and Metal Levels in Streams and Lakes in the
Eastern United States Caused by Acidic Precipitation .. 446
James N. Galloway
Variations in the Degree of Acidification of
River Waters Observed in Atlantic Canada
Mary Thompson
Responses of Freshwater Plants and
Invertebrates to Acidification
George Hendrey
453
457
VII
-------
WELCOME
LESLIE A. CAROTHERS
Deputy Regional Administrator
Region I, U.S. Environmental Protection Agency
Boston, Massachusetts
On behalf of EPA's New England regional office, I
want .to welcome you to New England and the
International Symposium on Lake Restoration. I am
standing in today for Bill Adams who is vacationing in
our western national parks. In fact, his itinerary for
today places him at the Great Salt Lake, and I know he
sends his greetings to all of you. The Clean Lakes
Program in close to his heart. When Bill Adams served
as commissioner of the Maine Department of Environ-
mental Protection before coming to EPA, he initiated
several of the first federally supported lake restoration
projects. The State of Maine continues to be a
pacesetter in lakes protection under Henry Warren's
aggressive leadership.
We are proud of the success of the Clean Lakes
Program throughout New England. To date, we have
approximately 40 operating Clean Lake projects
totaling approximately $10 million. Early results of
these projects are encouraging. Recreational uses have
been partially restored to Morses Pond and Nutting
Lake in Massachusetts, Lake Bomoseen in Vermont,
and Annabessacook Lake in Maine. I understand that
you will be learning more details of these projects
during this conference.
We are particularly pleased to have recently awarded
the first lake protection grant to the Cobbossee
watershed district in Maine. (One of the things you
learn at this conference is how to pronounce Maine's
Indian names!) This grant will provide financial
assistance for remedial and preventive activities to
protect 15 lakes and ponds in the Cobbossee
watershed in central Maine. A previous planning study
indicated that over 70 percent of the annual nutrient
loading to these lakes comes from both nonpoint
agricultural runoff and stormwater runoff. The impact
of the projected land use changes and population
growth will result in phosphorus loading increases
sufficient to trigger nuisance algae blooms in 11 of 15
lakes in the district.
With the positive action to improve land use
management, we expect that the existing high quality
lakes in the area can be maintained. We are
encouraged to see other areas and States pursuing
similar programs.
In addition, Region I has taken every opportunity to
use other EPA programs to benefit our recreational
lakes. These include Federal grants for construction of
municipal waste treatment facilities and the Rural
Clean Water Program. The restoration of Lake
Winnisquam in New Hampshire and Rangely Lake in
Maine are just two examples of the successful use of
the construction grant program to reduce pollution
adversely affecting important lake resources.
The RCWP is a multi-agency water quality improve-
ment program directed at abating pollution from
farming practices. It provides for planning and
implementation funds to correct activities which are
adversely affecting stream and lake quality. In Region I,
the St. Alban's Bay watershed in Vermont has been
selected for FY 1980 funding. The combined efforts of
the ongoing municipal construction program for the
City of St. Alban's and the Rural Clean Water Program,
tackling the basin's nonpoint source pollution, will
greatly clean up St. Alban's Bay and restore long
impaired recreational uses in Lake Champlain.
I see from the conference agenda that you are
beginning four days of intensive review of the latest
scientific research in your field. Although your work is
complex and difficult, I think you are fortunate because
the results of your efforts provide benefits that are seen
and appreciated by people who love our lakes, even if
they do not know what limnology means. It must surely
be rewarding to help to preserve and enhance
resources of economic and recreational value and
places of beauty and peace for our people to enjoy. I
wish all of you well in your important work.
-------
THE U.S. EPA CLEAN LAKES PROGRAM
STEVE SCHATZOW
Deputy Assistant Administrator
Office of Water Regulations and Standards
U.S. Environmental Protection Agency
Washington, D.C.
The enthusiasm for this symposium mirrors the
enthusiasm we see in this country for the Clean Lakes
Program. Granted, it is one of the few EPA programs
that is non-regulatory in nature. We are not forcing
people to clean up their lakes — we are helping them.
They are even digging into their own pockets and using
their own muscles to make the clean lakes projects
work for them.
And, as a result, the Clean Lakes Program is doing
more than protecting our lakes' resources — it is
rekindling the grass roots involvement that is the key to
this country's political system. Our citizens are solving
their lakes problems at the local level: we at the Federal
level are helping — and together, we're doing
something no other Federal program is doing — we're
turning a profit!
A recent study showed that every Federal dollar
invested in clean lakes projects is returning $8.30 in
benefits! That's something to brag about!
And it is but one indication of the success our Clean
Lakes Program is achieving. As they say in our
television commercials, we have come a long way in
the 8 years since the senior Senator for Minnesota —
you know him now as Vice President Mondale —
collaborated with Senator Quentin Burdick and
Congressman Donald Fraser to maneuver the Clean
Lakes Act through Congress. Together, they laid the
cornerstone of the Clean Lakes Program.
In that law, Congress declared that our Nation's
publicly owned freshwater lakes should be protected
and restored. It gave that job to the Environmental
Protection Agency. We took the responsibility very
seriously because we were concerned with a number
of problems we saw with that mandate.
In the first place, how could we start a national
program when we knew so little about the size of the
problem? We still don't know for sure how many lakes
we have — there's an argument about that number
every time we publish a book!
But there was another problem — we believed that
other pollution controls — the permits limiting pollution
discharge, for example — would be enough to protect
lake quality. And we also questioned whether
technology was adequate to deal with lake eutrophica-
tion problems. Finally, we were very concerned with
the cost-effectiveness of lake restorative measures.
So we proceeded deliberately, working out those
initial doubts, using our first $4 million appropriation in
1975 to initiate pilot lakes projects.
Today, we have no doubts, only proof that a national
Clean Lakes Program is needed and is working. We
believe that about 10,000 of our publicly owned lakes
need pollution control and restoration — that's a
sizable national problem that would cost us nearly $5
billion.
We have also discovered that those other pollution
control programs cannot alone solve the lakes'
problems. Granting permits for discharge helps, but
nonpoint source pollution and watershed management
are just as important. We have also discovered that the
technology to restore and protect lakes not only exists
but is rapidly becoming more sophisticated — you are
going to spend this entire week discussing that! And as
for cost-effectiveness, we've already told you that the
program in its short lifetime is returning benefits of
more than 8 to 1.
So our initial task of discovering and testing that
technology has proved successful. Where do we go
from here?
I just quoted you a $5 billion figure to clean up those
10,000 lakes we believe need help. Realistically, we
can't afford that — our Nation's financial pie just
doesn't cut into that big a slice for cleaning up our
lakes. What do we do? We look at the problem
realistically and we come up with a solution. A realistic
solution that we call our 5-year strategy. A solution
that makes the most of our limited funds to benefit the
most people.
We have set a goal to protect or restore at least one
lake with water quality suitable for contact recreation
within 25 miles of every major population center. This
goal will take at least $150 million in Federal funds but
it will serve almost all of our population. Frankly, our
current rate of budget support will not enable us to
meet the goal in 5 years. We may have to stretch it out
somewhat.
The goal, though, is within our reach. One of the
facts we have learned in these first years of the
program is that 99 percent of us live within 50 miles of
a publicly owned freshwater lake. And one-third of us
live 5 miles or less from a lake. So our goal is right on
target, particularly today when the price of gasoline is
soaring — Americans want recreation close to home.
To meet this goal, our strategy has five objectives
that must be considered when any clean lakes project
is approved. The first is that the project must maximize
public and environmental benefits. It must also
coordinate with other Federal and State programs. It
must emphasize pollution controls, particularly for
nonpoint sources before it uses any in-lake devices.
The project must also emphasize the Federal-State
partnership: and, as a final objective, the State must
-------
OPENING SESSION
continually evaluate the program to maintain high
quality pollution control practices.
We have developed a technical guidance manual for
State use and a citizen's guide on how to use the Clean
Lakes Program. Both will be available within the next
few weeks. The manual can be reviewed at the EPA
exhibit here.
These publications and our 5-year strategy all
communicate two overall objectives of our Clean Lakes
Program. In all our projects, we encourage best
management practices and pollution control in lake
watersheds. We will not make an award unless
pollution controls required by the Clean Water Act are
in place or are progressing — and we also require
control of nonpoint sources of pollution according to
procedures developed under the Act.
Our second overall objective stems from the first. We
insist on program integration. You may remember I
mentioned that as one of the strategy objectives. Many
Federal, State, and local programs have some impact
on water quality.
If used wisely, a combination of these programs can
meet a common goal at lower cost, and certainly avoid
duplication of effort. We have worked with the Heritage
Conservation and Recreation Service of the Depart-
ment of the Interior to restore 59th Street Pond in New
York's Central Park — and a current project at
Broadway Lake in South Carolina is using resources
from the Department of Agriculture and a State agency
to supplement the clean lakes award.
But the most essential element of the Clean Lakes
Program is still citizen participation. To date we have
spent nearly $60 million for over 200 projects in 46
States. But these are not just government handouts.
These are matching funds — somebody besides the
Federal Government must come up with 50 percent of
the cost of a lakes project. That could be the State.
More often, it is the local government because it was
the lakeside community that first saw those weeds
choking their waterway, and smelled the rotting fish,
and decided to do something about it.
So once their project is approved, they come up with
the money. Only sometimes it isn't money. Remember I
mentioned muscle earlier in this speech? The people of
Scotia, N. Y. rounded up privately owned tow trucks,
hooked them to tree trunks that had to be removed from
their lake, and pushed while the trucks pulled. That's
really working for your lakel
And that's probably why the Clean Lakes Program
has been so successful in this country. The public can
actually see and smell their problem and they can do
something about it. Once they've put their money and
muscles into it, they're careful to maintain their lakes,
and they look around to see how else they can improve
their communities. One town built parks: another
rebuilt decaying neighborhoods.
Congress just thought they put the responsibilty for
protecting and restoring lakes on the Environmental
Protection Agency.
They really put it on the people of this country, and
Americans have accepted the challenge to a point that
EPA would not have dreamed possible 8 years ago. We
are truly proud of our program.
And now let me turn to the more immediate goals of
this symposium. We have an impressive array of
experts on the program who will provide us with the
most up to date information on the science of lake
restoration. I want to extend an especially warm
welcome to our colleagues from the many countries
outside of our borders who are participating in this
symposium. We are especially appreciative of the
cooperation of the Organization for Economic Coopera-
tion and Development in sponsoring the symposium.
The program is impressive. The topics to be
presented indicate we do know a lot about the causes
of lake eutrophication and the techniques of restora-
tion.
Too many times scientists are too modest. This
conference is going to give us the opportunity to spread
their knowledge across the world.
We are rather proud in this country of the progress
made in a few short years. I discussed some of the
results of our national program a few minutes ago. We
know that lake restoration techniques work. You will
hear about them in more detail during the week. Our
technical efforts have been increasingly focused on
watershed control technology. Since many of the lakes
we want to restore are in or near cities, urban runoff
control technology is high on our list of the problems
which demand attention. We look forward to this
symposium to provide some of the answers.
We join with OECD to host you at this symposium. I
believe this is a great opportunity to exchange
information on lake restoration that may not present
itself for years to come. The discussions that take place
here can provide a springboard for. a worldwide
restoration of our lake resources during the next
decade.
*Mr. Schatzow's comments which suggest a Federal financial commitment in
future years do not reflect the current Federal position. Decisions made in the FY
1982 budget resulted in the elimination of Clean Lakes Program funds due to higher
environmental priorities. It is anticipated that local communities and States will
assume full responsibility for lake cleanup in an appreciable number of projects.
-------
LOCAL COMMITMENT TO LAKE RESTORATION:
COBBOSSEE WATERSHED EXAMPLE
THE
THOMAS U. GORDON
Executive Director
Cobbossee Watershed District
Winthrop, Maine
ABSTRACT
Successful lake management requires a strong local commitment to lake restoration and
protection. The Cobbossee Watershed District has integrated Federal, State, local, and private
resources to restore eutrophic lakes. The Cobbossee Watershed District is the State of Maine's
only local unit of government devoted exclusively to lake management. The District covers a
watershed of 622 square kilometers, including four eutrophic and 11 mesotrophic lakes of 30 to
2,259 hectare in size. The District's primary objective has been the restoration of 575-hectare
Annabessacook Lake. Utilizing a Clean Lakes grant, the Cobbossee Watershed District established
a cost-sharing program to construct agricultural waste management facilities in the
Annabessacook, Cobbossee, and Pleasant Pond drainages. A hypolimnetic application of
aluminum sulfate and sodium aluminate also was used on Annabessacook Lake for nutrient
inactivation. Completion of both phases of the project required coordination of Federal and State
agencies, as well as agricultural groups and lakeshore property owners' associations. Follow-up
monitoring of Annabessacook Lake has found significant reductions in phosphorus and
chlorophyll, and improvements in visibility. Similar responses are anticipated in the two
downstream lakes.
INTRODUCTION
Water pollution control has long been a function of
government, with specific responsibilities divided
among the national, State, and local levels. The Federal
Water Pollution Control Act Amendments of 1972
provided the first national initiative for lakes restoration
through section 314 of the law, the Clean Lakes
Program. Prior to 1972, many States, most notably
those in the Great Lakes Basin, had developed their own
programs for protecting and improving their publicly
owned lakes and ponds.
Today, State agencies have the primary responsibility
under the Federal Clean Water Act for diagnosing and
treating lake water quality problems. New EPA
regulations, which channel all lake restoration grants
through the State water pollution agencies, have
reinforced the States' central position in the Clean
Lakes Program.
Lake management activities have also been
conducted at the sub-state level by a variety of public
agencies, citizens groups, and private enterprise. The
diversity of American local government — cities,
counties, towns, special districts— greatly complicates
any summation of local roles in lake restoration.
Generally, successful implementation of restoration
projects and protection strategies for most lakes will
require a strong local concern for water quality.
Furthermore, effective lake management efforts will
require careful integration of limited Federal, State,
local, and private resources. The Cobbossee Watershed
District has provided one example of local leadership in
lake restoration.
DESCRIPTION OF THE COBBOSSEE
WATERSHED
The Cobbossee Watershed, a sub-basin of the
Kennebec River Basin, is located approximately 80
kilometers north of Portland, Maine. The 622 square
kilometer watershed consists of a chain of 24 lakes and
ponds, ranging in size from 30 to 2,259 hectares. Four
of these Jakes — Annabessacook, Cobbossee, Little
Cobbossee, and Pleasant — are culturally eutrophic.
Glacial till predominates in the surficial geology of the
watershed.
Approximately 25,000 people reside in the water-
shed on a permanent basis. During the summer
months, the population increases by 60 percent,
reflecting the significant tourist and second-home
economy of the lakes region. The lakes also serve as a
recreational resource for 47,000 people residing in
communities peripheral to the watershed. The water-
shed remains predominantly forested (approximately
75 percent of the land area), with agricultural (12
percent) and residential (8 percent) land uses also
significant.
HISTORY OF WATER QUALITY
PROBLEMS
Historically, water quality concerns in the Cobbossee
Watershed have focused on Annabessacook Lake. For
more than 150 years, the tributaries of Annabessacook
Lake served as conduits for municipal and industrial
effluent. The earliest reports of serious water quality
-------
OPENING SESSION
Figure 1 —The Cobbossee Stream watershed.
degradation occurred in the late 1930's. Despite years
of public complaints, no intensive evaluation of the
problem was made until the mid-1960's (Smith and
DeWick, 1965). A study of the lake by the Maine Water
Improvement Commission found violations of water
quality classifications and standards. Late summer
Secchi disk visibilities ranged from 0.76 to 1.83 meters,
caused by intense blooms of Anabaena and Aphani-
zomenon. The annual phosphorus loading to the lake
from municipal and industrial sources was estimated at
14,000 kilograms. Furthermore, Cobbossee Lake,
located immediately downstream from Annabessacook
and once famous for its salmon and other fisheries,
also experienced a significant decline in water quality.
As a result of these findings, an 18-kilometer
trunkline sewer was constructed to divert all point
sources from the Annabessacook watershed to
treatment facilities on the Kennebec River in the city of
Augusta. Thus, wastewater discharges from the town
of Winthrop terminated in 1972; the remaining
discharges from the town of Monmouth continued
through 1976, when a 10-kilometer extension of the
trunkline sewer was completed. Elimination of these
discharges reduced phosphorus loading to Annabessa-
cook Lake by approximately 90 percent (Sage and
Moran, 1977). However, the lake's ambient concentra-
tions of total phosphorus remained above 15 parts per
billion, and nuisance algal blooms persisted.
Despite the efforts being made to control nutrient
inputs in the early 1970's, there was a strong public
perception that not enough was being done to restore
Annabessacook and Cobbossee Lakes and to protect
the other lakes in the watershed from similar problems.
From 1964 through 1972 lakeshore property owners
on Annabessacook Lake took their own remedial action
by treating the lake with algacides. Over 8 years,
approximately 30 tons of copper sulfate were applied,
with diminishing effectiveness as copper-resistant
types of algae began to predominate. A proposed
sodium arsenate application was prohibited by State
health officials. Two small-scale attempts to aerate the
lake failed to produce any noticeable change in water
quality.
Frustrated by 30 years of failure in improving lake
water quality, lakeshore property owners on Anna-
oessacook and Cobbossee began working with munici-
pal officials of the Southern Kennebec Valley Regional
Planning Commission to develop a comprehensive
strategy for lake restoration in the 1970's. Given the
highly experimental nature of lake restoration techno-
logy (Imhoff, 1971), efforts concentrated on creation of
an institution to develop and implement appropriate
restoration techniques. The Federal and State Gov-
ernments were perceived as the primary sources for
funding the necessary research on the lakes. Since the
lakes and streams of the Cobbossee Watershed fall
under the jurisdiction of as many as 16 separate
municipalities, 2 counties, 4 water districts, and 5
sanitary districts, no one unit of local government could
Table 1. —
Annabessacook
Cobbossee
Pleasant Pd:
Morphometry
Surface area
Mean depth
Maximum depth
Total drainage area
Direct drainage
Flushes per year
Land Use Characteristics
(Direct drainages in percentages)
Forest and reverting fields
Developed
Agriculture
active, fields
active, tilled
Other
575 ha.
5.3 m
14.9m
85 mi2
21.8 mi2
4.5
69
12
16
1
2
2,244 ha.
8.07m
30.48 m
131.4 mi2
46.7
1.2
65
11
20
0
4
mi"
237 ha.
2.68m
7.9 m
217 mi2
23.6 mi2
5.6
73
8
16
1
2
-------
RESTORATION OF LAKES AND INLAND WATERS
provide a comprehensive approach to lake manage-
ment.
Instead, a quasi-municipal, special-purpose district
was proposed to assume the tasks of lake research,
restoration, and protection for the Cobbossee Water-
shed. In addition to providing a single jurisdictional unit
for watershed planning, the lake district often has
independent powers of taxation, the ability to focus all
its resources on a single issue, and a clearly defined
group of constituents who can support the district's
efforts (Gordon, 1977). Lake protection and rehabilita-
tion districts have subsequently found widespread
acceptance in Wisconsin and other States.
The Cobbossee Watershed District was authorized by
the Maine Legislature in 1971. The District's legislative
charter called for establishing a Board of Trustees of up
to 17 members, appointed by 10 municipalities and 3
water districts designated as members of the District.
To become operational, the District had to be ratified by
public referenda in each of the municipalities. In
November 1972, 8 of the 10 municipalities voted to join
the District. More than 80 percent of the voters
Figure 2. — Lakeshed and town boundaries.
supported creation of the District, a particularly
significant figure given the uncertainties about the
agency's tax assessment potential.
The District's legislated purposes are to protect,
improve, and conserve the lakes, ponds, and streams of
the Cobbossee Watershed for the public health and
welfare and for the benefit of residents and property
adjacent to these waters. To do so, the District is
authorized to do any and all things necessary to
improve water quality. The District is also authorized to
own and operate the 22 small dams in the watershed.
Specific powers of the District include eminent domain,
taxing all property within member municipalities,
bonding for major capital expenses, and the authority
to pass rules and regulations. To date, the District has
not acquired dams or passed new water quality
regulations. Instead, the District has emphasized a
voluntary, cooperative approach to watershed man-
agement problems.
The operating budget for the District is approved
annually by registered voters of the member munici-
palities at a public budget meeting. Once approved by
the voters, each municipality must pay a share based
on the proportionate value of its shoreland property.
The operating budget for 1973 was $25,000, derived
totally from local taxes. In 1980, the District's budget is
approximately $90,000, with 60 percent of the funds
derived from Federal and State grants. The consistent
support of local taxpayers has allowed the District to
maintain a stable program and permanent staffing.
DIAGNOSTIC STUDIES
The District's initial efforts at water quality man-
agement centered on individual subsurface waste-
water disposal systems. In 1974 and 1975 a
comprehensive sanitary survey was conducted of
approximately 1,200 lakeshore residences. Information
was gathered on the design, location, age, usage, and
maintenance of wastewater disposal systems. Less
than 5 percent of the systems surveyed were found to
be discharging effluent to the lakes or ground surface.
More than 50 percent of the systems, however, were
classified as inadequate, according to the upgraded
standards of the 1974 Maine State Plumbing Code
(Freedman, et al. 1977). At the time of the survey, no
definitive relationship between subsurface wastewater
disposal and lake water quality had been established.
Subsequent literature review and research (Sage and
Moran, 1977; Beals, 1980) has led the District away
from subsurface wastewater disposal as a significant
source of phosphorus loading to its lakes.
Regular water quality monitoring of the Cobbossee
Watershed lakes intensified in 1 976 with funding of an
Areawide Water Quality Management Plan by EPA,
pursuant to section 208 of the Federal Water Pollution
Control Act Amendments of 1 972. As a part of the 208
planning program, the District conducted intensive lake
studies on 11 lakes in the area. These studies
attempted to define the sources of phosphorus loading
to the lakes. In the Annabessacook Lake watershed,
phosphorus concentrations and stream flows were
monitored on five major tributaries and the lake outlet.
In-lake monitoring was also conducted, with particular
emphasis on spring and fall overturn. The water quality
data were then used in the Dillon-Rigler model to
Table 2. — Phosphorus runoff rates for the Cobbossee
Watershed.
Source
Forests
Clearcut forests
Reverting fields
Cultivated crops
Manured fields
Village (storm sewers)
Residential
nearshore
remote
Septic systems
Precipitation
(per ha lake surface)
Phosphorus Runoff
(kg/ha)
.03 ± .01
.30 ± .10
.03 ± .01
1.0 ±.5
1.6 ±.4
1.1 ±.2
.9 ±.3
.45 ± .15
0-20% annual input
.1 ±.041
1 Annual input = 1.5 kg/cap-yr for permanent residences;
.5 kg/cap-yr for seasonal.;
-------
OPENING SESSION
estimate phosphorus loading in the lake. Finally,
existing land uses in the lake's watershed were
examined. By applying appropriate phosphorus runoff
rates to the acreage in each land use category, the
significance of various land uses in phosphorus
enrichment of the lake could be estimated.
The completed lake studies established priorities for
phosphorus loading reductions on Annabessacook
Lake, Cobbossee Lake, and Pleasant Pond (see Table 3).
The primary watershed source of phosphorus loading
to these lakes, and almost all lakes in the District, was
found to be agricultural runoff, principally from animal
waste spread on frozen or snow-covered ground during
the winter. Only one farm of 26 surveyed in the
watershed had winter manure storage facilities (Sage,
1977c.). Recycling of phosphorus from bottom sedi-
ments in Annabessacook Lake was also a significant
source of loading to that lake. These diagnostic studies,
funded by the 208 planning program, provided the
basis for a Clean Lakes grant application to EPA in
March 1977.
AGRICULTURAL WASTE MANAGEMENT
Controlling agricultural nonpoint sources of phos-
phorus in the watershed presented several challenges.
Appropriate designs for animal waste management
practices had to be developed to meet the varying site
conditions, types of animal wastes, and existing farm
management practices. Financing costly waste man-
agement facilities required establishing a cost-sharing
program, using EPA Clean Lakes funds. Finally,
working relationships with existing agricultural service
agencies had to be defined.
Effective containment of animal waste for the
duration of Maine's winters requires storage capacity
for at least 6 months. This can reduce phosphorus
runoff by as much as 70 percent (Porter, 1975), thereby
minimizing water quality impacts and conserving
fertilizer for food production during the summer
months. The typical facility for daily manure storage is
a concrete box, with capacities ranging from 155 m3 to
1,130 m3, depending on the number of animals served.
The facilities are often roofed to eliminate excess
capacity otherwise required for precipitation. Another
typical facility for poultry and dairy manure storage is
an impervious pad (either asphalt or concrete) with
earth berm walls.
Both facilities are generally intended for solid or
semi-solid animal wastes. These storage facilities also
require transfer systems to move manure into the
containment area during the winter and to facilitate
cleaning in late spring for field application. In addition
to storage facilities for animal waste, runoff diversion
structures are often necessary in barnyard areas to
reduce phosphorus transport. Given the variability of
site topography and layout of barns, each farm requires
an individually designed manure management system.
In March 1977, the Cobbossee Watershed District
estimated the costs of agricultural waste management
practices necessary to restore Annabessacook and
Cobbossee Lakes and Pleasant Pond. The estimated
cost for 41 farms totaled $282,500. Of this amount, 23
major farms would require $266,000 worth of
facilities, an average cost of more than $11,500 per
farm. Furthermore, construction costs were expected
to increase substantially each year because of inflation.
Table 3. — Phosphorus
Annabessacook Lake
'Lake bottom sediments
"Agricultural runoff
Upstream lakes
Urban runoff
Forest runoff
Precipitation
Septic leachate
Cobbossee Lake
'Upstream lakes
"Agricultural runoff
Urban runoff
Precipitation
Forest runoff
Septic leachate
loading control priorities.
Kilograms Percent
1,500
1,000
1,000
450
100
60
30
4,200
4,900
2,600
850
225
200
100
36
24
24
11
2
1
1
56
29
9
2
2
1
Pleasant Pond
'Agricultural runoff
Urban runoff
Forest runoff
Precipitation
Septic leachate
8,900
1,500
250
125
25
25
1,930
78
13
7
1
1
* - Priority sources for lake restoration.
Given the precarious economic position of many small
farms in Maine, regulatory requirements for agri-
cultural nonpoint source control did not seem feasible.
Financial aid for constructing agricultural pollution
co'ntrols was limited. The U.S. Agricultural Stabilization
and Conservation Service (ASCS) provided a maximum
of $2,500 per year (now $3,500) for agricultural
conservation practices on each farm. Low interest
loans and tax relief measures were also available
(Moore, 1979), but generally had not been used for
high-cost practices such as manure storage. Thus, the
District established a new program of agricultural cost-
sharing as part of its lakes restoration effort.
The District offered 50 percent cost-sharing for
manure management systems on target farms. The
EPA Clean Lakes grant provided the District's cost-
sharing funds. The remaining half of construction
costs, paid by the participating farmers, provided the 50
percent non-Federal match required by EPA. By using
their own labor and materials, participating farmers
were able to reduce their actual cash outlays even
further. The District's cost-sharing program had no pre-
set ceiling on funds available per farm. Rather, detailed
cost estimates for the recommended agricultural waste
management plans were developed as the basis for
cost-sharing agreements. Thus, individual farms could
receive $25,000 in a single year for a $50,000 facility if
necessary. Also, certain equipment not usually funded
through the traditional ASCS cost-sharing program
could be included in the District's program.
Despite the vastly improved cost-sharing ratio,
participating farmers were being asked to make
significant personal investments in agricultural pollu-
tion control. Substantial assistance from the District
and various agricultural service agencies was required
to achieve voluntary participation. The Cobbossee
Watershed District and two county soil and water
conservation districts (SWCD's) presented numerous
design options to interested farmers. The Kennebec
-------
8
RESTORATION OF LAKES AND INLAND WATERS
County SWCD sponsored tours of existing manure
management facilities for interested farmers. The
Cooperative Extension Service provided data on the
economic benefits of conserving manure for use as
fertilizer. Personnel of the Watershed District and the
SWCD's.also assisted with obtaining construction bids
from local contractors, information on ASCS, Farmers
Home Administration, and Small Business Administra-
tion financial assistance, and information on State and
Federal tax requirements relating to cost-sharing and
pollution control investments. The U.S. Soil Conserva-
tion Service (SCS) finalized blueprints and manage-
ment plans, and inspected projects under construction.
As with any experimental program, unforseen delays
developed. The District's cost-sharing program was
originally intended to be administered by ASCS, which
has many years experience in agricultural cost-
sharing. However, Federal regulations made trans-
ferring Clean Lakes funds to ASCS extremely difficult.
Thus, the District had to establish its own administra-
tive procedures for cost-sharing. And, instead of
developing a single manure management plan for each
farm, the District and SCS had to produce three to five
detailed alternatives before farmers agreed to partici-
pate. These problems, as well as administrative grant
requirements, construction scheduling, and other
factors, extended the project period from 2 to31/2 years.
The extended period has had the significant benefit of
allowing a greater number of farmers to participate in
the program.
To date, 30 separate agricultural waste management
facilites have been constructed. These facilities provide
manure storage for approximately 80 percent of the
animal units in the watersheds of the three lakes. The
cost of these facilities is estimated to be $622,000
(original estimates were revised through grant amend-
ments in 1978). The Cobbossee Watershed District is
presently monitoring water quality to determine the
reductions in phosphorus resulting from these con-
trols.
NUTRIENT INACTIVATION TREATMENT
Effective restoration of Annabessacook Lake was
determined to require reduction of phosphorus re-
cycling from anoxic lake bottom sediments. Various in-
lake restoration techniques were considered prior to
selecting hypolimnetic treatment with aluminum. The
primary concern was controlling phosphorus release
from sediments in a 150-hectare area between 7 and
14 meters in depth. Because of the depth and area,
dredging, hypolimnetic aeration, and physically sealing
the lake bottom were found to be either impractical or
prohibitively expensive. Furthermore, flushing of
nutrient-rich water from Annabessacook would in-
crease the difficulty of restoring eutrophic Cobbossee
Lake, immediately downstream (Gordon et al., 1977).
Prior to implementing a nutrient inactivation treat-
ment, a detailed feasibility study of the project was
made (Dominie, 1978). This study attempted to further
define the area of bottom anoxia, optimum aluminum
application rates, and potential impacts on aquatic life.
A detailed discussion of this study and the subsequent
treatment is presented elsewhere in this symposium
(Dominie, 1980).
Based on the recommended application rates,
approximately 227,000 liters of aluminum sulfate and
142,000 liters of sodium aluminate, with a combined
weight of approximately 450 metric tons would be
needed to treat the lake. Chemical treatment on this
scale presented significant logistical problems, partic-
ularly with a limited project budget. Because of limited
funds, the cost of the treatment was limited to
$65,000. As with the agricultural construction, local
contributions of in-kind services were used to
maximize the Federal funding applied to the project.
The Maine Department of Environmental Protection
contributed manpower for lake monitoring, develop-
ment and evaluation of the treatment methods, and
laboratory analysis of water samples. Lakeshore
property owners affiliated with the Annabessacook
Lake Improvement Association and the Cobbossee
Yacht Club, provided financial assistance and partici-
pated in the lake treatment itself. The Maine National
Guard transported a barge for the treatment from
Portland to Annabessacook Lake. Boats.for the project
were donated by the Maine Department of Environ-
mental Protection and lakeshore property owners.
The nutrient inactivation treatment for Annabessa-
cook Lake was completed during August 1978. Despite
Table 4. — Annabessacook Lake visibility.
Secchi disk depth
0 — 0.9 meters
1—1.9 meters
2 — 2.9 meters
3 — 3.9 meters
4 — 4.9 meters
5 + meters
Public perception of water quality
gross pollution; lake is totally unusable
for recreation
algae blooms still evident; quality is
unacceptable for most uses •
some complaints of declining water
quality; some impairment of water use
satisfactory quality; no impairment of
water use
excellent water quality; a positive factor
encouraging lake use
exceptional quality for this lake
Total:
Days
(June
1972
43
63
0
0
0
0
106
at given visibility
1 — September 19)
1977
10
67
28
28
0
0
106
1979
0
0
30
30
35
1
106
notes on selected years:
1972 — prior to full diversion of municipal/industrial wastewater
1977 — prior to lakes restoration project
1979 — after agricultural waste controls and nutrient inactivation treatment
-------
OPENING SESSION
frequent delays caused by equipment problems,
approximately 95 percent of the designated project
area received treatment. A total of 179,334 liters of
aluminum sulfate and 121,039 liters of sodium
aluminate were applied to the lake.
PROGRESS ON LAKE IMPROVEMENT
Monitoring of water quality improvements resulting
from the agricultural waste management and nutrient
inactivation treatment projects will continue through-
out 1981, when a final report will be submitted to EPA.
To date, the three lakes affected by the project have not
had adequate time to fully respond to nutrient loading
reductions it produced. However, preliminary results
show progress, with Annabessacook Lake exhibiting
the most significant improvement thus far. Phosphorus
loading has been reduced by approximately 50 percent,
producing a marked improvement in Secchi disk
visibility (Table 4). Preliminary data for the summer of
1980 parallel the 1979 results. Reduced phosphorus
concentrations in the lake's hypolimnion indicate some
effectiveness of nutrient inactivation.
Water quality data for 1979 indicate little improve-
ment in Cobbossee Lake and Pleasant Pond. Tributary
sampling indicated continuing phosphorus runoff from
farms without proper animal waste management
systems (King, 1980). Subsequent construction of
manure storage facilities on most of these farms
should reduce phosphorus loadings and improve lake
quality in 1980-81.
The three lakes are expected to remain sensitive to
increased phosphorus loadings even after the full
effects of the restoration project are realized. Increas-
ing development in the watersheds of these lakes,
perhaps caused in part by their improved water quality,
is likely to result in additional stormwater runoff and
phosphorus loading. By 1995, the three lakes are
projected to once again exceed their phosphorus
loading limits, unless additional preventive measures
are taken (Gordon, et al. 1980). The Cobbossee
Watershed District is planning to concentrate its efforts
during the 1980's on preventing any significant
deterioration in the quality of its restored lakes, as well
as all other major lakes and ponds within its
jurisdiction.
CONCLUSIONS
The success of the Cobbossee Watershed District in
implementing a lake restoration project can be
attributed to integrating technical and financial
resources from many sources. Development of lake
restoration technology is not enough, if institutional
mechanisms to finance and use it are inadequate. The
District's efforts through the 1970's illustrate both the
opportunities and challenges presented in making the
Clean Lakes Program work on the local level.
To control sources of pollution, the District has used
non-regulatory approaches whenever possible, at-
tempting to develop a sense of local responsibility for
pollution control and lakes management. Public
concern about lake water quality led to the creation of
the Cobbossee Watershed District. Active citizen
involvement in the District's programs has been
essential to its success. This public participation will be
even more vital in the perpetual struggle to preserve
lake quality for future generations.
REFERENCES
Seals, L. M. 1980. Application of computer simulation of
phosphorus movement through soils. Master's thesis.
University of New Hampshire, Durham.
Dillon, P. J., and F. H. Rigler. 1974. A test of a simple nutrient
budget model predicting the phosphorus concentration in
lakewater. Jour. Fish. Res. Board Can. 31:1771.
Dominie, D. R. 1978. Aluminum application feasibility study
for Annabessacook Lake. Cobbossee Watershed Dist.,
Winthrop, Maine.
1980. Hypolimnetic aluminum treatment on
Annabessacook Lake. In Proc. Symp. for Inland Waters and
Lake Restoration, Portland, Maine. U.S. Environ. Prot.
Agency.
Freedman, S. J., et al. 1977. Non-sewered areas wastewater
disposal problems: phase III. Southern Kennebec Valley
Regional Plan. Comm., Augusta, Maine.
Gordon, T. U. 1977. Implementation of a regional water
resources plan in the Cobbossee Watershed District.
Cobbossee Watershed Dist., Winthrop, Maine.
Gordon, T. U., et al. 1977. Cobbossee Watershed District
lakes restoration project. Cobbossee Watershed Dist.,
Winthrop, Maine.
1979. Cobbossee Watershed District lakes
protection project. Cobbossee Watershed Dist., Winthrop,
Maine.
Imhoff, E. A., ed. 1971. Workshop conference on reclama-
tion of Maine's dying lakes. Water Resour. Center,
University of Maine, Orono.
King, W. L. 1980. Potters Brook phosphorus loading: 1979
spring runoff. Cobbossee Watershed Dist., Winthrop,
Maine.
Moore, I. C., et al. 1979. Financial incentives to control
agricultural nonpoint source pollution. Jour. Soil Water
Conserv. 34:60.
Porter, K.. S., ed. 1975. Nitrogen and phosphorus: food
production, waste, and the environment. Ann Arbor Science
Publishers, Ann Arbor, Mich.
Sage, K. J. 1977a. Cobbossee Lake study. Cobbossee
Watershed Dist., Winthrop, Maine.
-. 1977b. Pleasant Pond study. Cobbossee Water-
shed Dist., Winthrop, Maine.
1977c. Factors contributing to phosphorus export
from agricultural lands and alternatives for reduction.
Cobbossee Watershed Dist., Winthrop, Maine.
Sage, K. J., and E. K. Moran. 1977. Annabessacook Lake
study. Cobbossee Watershed Dist., Winthrop, Maine.
Smith, R. H., and S. C. DeWick. 1967. Annabessacook Lake:
Eutrophication and fertilization. Maine Water Improvement
Comm., Augusta, Maine.
U.S. Environmental Protection Agency. 1980. Cooperative
agreements for protecting and restoring publicly owned
freshwater lakes. Fed. Reg. 45:7792.
-------
10
THE EUTROPHICATION STORY SINCE MADISON, 1967
A. F BARTSCH
U.S. Environmental Protection Agency, Retired
Corvallis, Oregon
ABSTRACT
The International Symposium on Eutrophication at Madison, Wis. in 1967 summarized current
knowledge and provided recommendations for future action. Many of these recommendations
have given direction to recent research. This paper examines notable accomplishments during the
subsequent 13 years and emphasizes the need to enunciate a new challenge. Accomplishments
examined fall in three major categories: (1) Understanding the eutrophication process; (2)
developing methods to impede eutrophication; and (3) establishing laws, regulations, and
programs to help restore and protect lakes. In the first category, research has focused on: (a)
Critical nutrients and the question of carbon significance; (b) nutrient loading and new knowledge
derived from the OECD North American Project and the EPA National Eutrophication Survey; and
(c) the utility of algal assays in understanding phytoplankton dynamics. To impede eutrophication,
methods being tested include: (a) Nutrient manipulations such as diversion, waste treatment,
product modification, nutrient inactivation, dilution and flushing, and plant harvesting; (b) physical
actions such as aeration, dredging, and hypolimnetic withdrawal; and (c) symptomatic treatments
and biological controls. Several laws, regulations, and international agreements have been
adopted at various governmental levels to turn back the eutrophication clock. The Great Lakes
Agreement of 1972 is one of them. Section 314 of the Federal Water Pollution Control Act provides
financial incentives and other mechanisms for lake improvement projects.
INTRODUCTION
Just a few weeks ago, the population of the United
States passed the 222 million mark. Today there are 42
million more people than there were when we
gathered in Madison in 1967 to share what we knew
about eutrophication and to plot new, exciting paths to
follow. This past year, 1 billion acres of land were taken
up for urban development to meet the needs of
population growth. As our Nation moved in these
directions, more lakes were caught in the urban fringe
as city growth engulfed them; many lakes were newly
impacted by sewage effluents in the face of growing
demands for recreational use by urbanites.
Are there more eutrophic lakes today than in 1 967? I
expect so. Are lakes becoming more eutrophic than in
the past? Undoubtedly some are, but on the average we
don't really know. We do know this — in the lower 48
States there are some 12,000 to 15,000 lakes larger
than 40 hectares. They are susceptible. Perhaps 10 to
20 percent are eutrophic, especially ones near urban
development that have been sullied by human
indifference. Many are well known, and they stimulate
and strengthen our concern for the eutrophication
problem.
The International Symposium on Eutrophication at
Madison, Wis. in 1967 (Natl. Acad. Sci. 1969)
assembled the workers and coalesced existing knowl-
edge on eutrophication processes and controls. The
document, "Eutrophication — A Review" (Steward and
Rohlich, 1967) also appeared in the same year. These
developments, and others perhaps less well known,
mark 1967 as a most notable reference point for this
subject.
In looking at the Eutrophication Story Since Madison,
1967, I am encouraged for several reasons: Public
awareness of the problem has grown, many remedial
programs unthinkable 13 years ago are underway,
anti-eutrophication laws have been passed, and
significant new knowledge, ideas and tools, help us
probe the eutrophication process. But, I am also
disappointed. The Environmental Protection Agency's
National Eutrophication Research Program that once
was an energetic and moving force, has withered away
and no longer exists. We are losing our best
opportunity to learn how a lake responds when heroic
sewage treatment cuts off phosphorus input. Of
course, I'm referring to Minnesota's Shagawa Lake
where studies have been terminated before the
answers were obtained. Remedial measures available
for demonstration in today's restoration programs are
the same ones we talked about 13 years ago. I am
moved to ask: Where are the new, novel, and exciting
ideas that we need now to carry us forward again?
Perhaps they will come from what you do here this
week.
It is not possible for me to discuss or even cite all
important developments since 1967. I have therefore
selected examples that are indicative of research
trends, control technology currently being used, and
regulatory actions that seem characteristic of the past
13 years. Some are the product of research that began
much longer ago; others the result of recent
beginnings. In checking these examples with the
research recommendations that issued from the
Madison conference, we can truly say that the
conference was a strong inspiration for the years that
followed.
There are three main areas of emphasis as I trace
this brief history: (1) Understanding the eutrophication
process; (2) developing methods to impede eutrophica-
-------
OPENING SESSION
11
tion; and (3) establishing laws, regulations, and
programs to help restore and protect lakes.
UNDERSTANDING THE
EUTROPHICATION PROCESS
Nutrients
Today, as at Madison, one can still ask: What causes
eutrophication? The answer is fragmentary because
many interacting factors contribute to the overall
process, and how they do so is not always known.
Productivity depends on a complex interplay of solar
radiation, temperature, lake basin morphology, water
retention time, biotic interactions, and perhaps more
important, the availability of adequate nutrients. It is
generally agreed that algae and higher aquatic plants
require many different nutrients for growth, including
large amounts of carbon, nitrogen, hydrogen, phos-
phorus, and smaller amounts of approximately 25
others.
Obviously, rational control depends on somehow
interfering with the free action of one or more of these
factors. If we seek to starve the system, nitrogen and
phosphorus claim special interest because oligotrophic
lakes frequently are phosphorus limited; whereas, in
lakes enriched by urban sewage, the newly supplied
abundant phosphorus often leads to exhaustion of
nitrogen. The crucial question is not whether a
eutrophic lake is momentarily phosphorus-limited but
whether it can be made so through controlling
phosphorus input. Over the past 13 years, this has
come to be generally recognized.
Then, in the early 1970's, a major controversy arose
concerning the relative importance of carbon, nitrogen,
and phosphorus. One view contended that carbon is
really the regulator of algal production in many waters
— a view implying that control of phosphorus input is
falsely-based and doomed to failure. The opposing side
saw phosphorus as a critical nutrient, the most logical
one to be controlled. The heat of the controversy
appeared to be fueled by proposals to remove
phosphorus from detergents as a step in slowing down
cultural eutrophication. In February 1971 a symposium
on Nutrients and Eutrophication: The Limiting Nutrient
Controversy, was sponsored by the American Society
of Limnology and Oceanography. There was free and
lively debate in an effort to provide to the public a clear
scientific statement on the relative importance of
various regulating or limiting nutrients. The sympo-
sium ended in apparent general agreement that
phosphorus is the critical limiting nutrient in most
North American lakes and is the logical focal point for
management programs (Likens, et al. 1971; Likens,
1972). Today, the so-called "carbon controversy"
seems to have faded away.
Algal Assays
For many years scientists used assay procedures of
their own design to estimate the phytoplankton
production capacity of lakes and to seek guidance for
control procedures. By manipulating their tests in
various ways they could also identify critical nutrients
representative of the sample at the time. These assays
were valuable tools, but unfortunately the findings of
one worker could not be compared with another, nor
one lake with another, one test with another — and
there was little agreement on how to correct this
dilemma.
Less than a year following the Madison conference, a
small group concerned with this problem met in
Chicago under sponsorship of the Joint Industry
Government Task Force on Eutrophication (Anon.
1969). Their purpose was to jointly develop a research
plan to produce an algal assay procedure that would be
acceptable in North America and Europe and hopefully
worldwide (U.S. EPA, 1971). Nine organizations
participated in the task — four universities, four
industries, and EPA. That goal now seems to have been
reached. The bottle procedure, using Selenastrum
capricornutum as the test alga, has received broad
acceptance here and in 41 other countries.
Field Studies
Several large scale field programs have contributed
substantially to improved understanding of the eu-
trophication process. Almost 10 years ago, the
Organization for Economic Cooperation and Develop-
ment (OECD) initiated a study to formulate the
relationships between nutrient loadings to lakes and
their trophic response. The deliberations were based
largely on data available from European lakes. From
this effort came the early Vollenweider model
concerning nitrogen and phosphorus as factors in
eutrophication (Vollenweider, 1968). With time the
program broadened in both geography and scope. It
began, to collect comparable data on the degree and
extent to which nutrient loading correlates with
eutrophic state, and to measure the rate of eutrophica-
tion growth. A major element of the program was the
North American Project with specific objectives to: (1)
Develop detailed phosphorus and nitrogen budgets for
a number of water bodies; (2) assess their chemical,
physical, and biological characteristics; (3) relate their
trophic states to the nutrient budgets and to
limnological and environmental factors; and (4)
synthesize an optimal strategy to control the rate of
eutrophication. In the U.S. effort, 22 water bodies were
studied, and a final report has been published for each
(U.S. EPA, 1977). A summary analysis of these reports
(Rast and Lee, 1978) gives several points of im-
portance: (1) The Vollenweider nutrient load relation-
ships correlate well with assigned trophic states; (2)
good correlation exists between phosphorus loading,
normalized as to hydraulic residence time and mean
depth, and the average chlorophyll and water clarity;
and (3) these relationships can be used to help predict
improvement to be expected from controlling phos-
phorus when that is the critical nutrient.
Simultaneously, a National Eutrophication Survey
was underway by EPA to compile information on
nutrient sources, inputs, and impacts on selected lakes
and reservoirs, especially ones receiving municipal
sewage. For over 5 years, the survey sampled and
studied 'J12 water bodies. For each one, the trophic
state was estimated, and the sources and magnitudes
of nitrogen and phosphorus inputs established as a
step toward judging if reduction in phosphorus loading
would be a promising remedial approach. Each lake
was sampled three times during the growing season at
multiple sites and depths. The 15 to 20 analyses done
-------
12
RESTORATION OF LAKES AND INLAND WATERS
on each sample included nutrient concentrations, algal
types and numbers, algal assay data on productivity
potential, and limiting nutrient. A separate report is
available for each lake, primarily for local officials to
use as a starting point for lake restoration program
planning. Data summaries are available in four
volumes. (U.S. EPA.1978)
Never before has there been such a broad base of
similarly-collected data on so many lakes as that now
provided by the North American Project and the
National Eutrophication Survey (U.S. EPA, 1978). It is
noot surprising that these data have been sought and
manipulated so avidly in efforts to find the true
meaning they may contain. Fear that such interpreta-
tions may exceed the validity of the data caused the
Ecology Advisory Committee of EPA's Science Advisory
Board to take a critical look at the survey and its
products. Toward strengthening the credibility of the
study the Committee recommended that survey data
and evaluation techniques be used to compare well-
studied lakes with corresponding ones sampled in the
survey. While comparable non-survey data are sparse,
especially for tributary point and nonpoint nutrient
loads, some limited comparisons could be made. They
helped show that the survey data are surprisingly good
and certainly adequate to: (1) Assess trophic condition,
(2) infer the limiting nutrient, and (3) provide tributary
nutrient loads with acceptable accuracy (Allum, et al.
1977).
It has long been recognized that an acceptable
nutrient budget is needed for a sound control program.
Yet, because of the cost and time required, very few
U.S. lakes are even now characterized by such vital
information. This will soon change to some extent
because lakes can qualify for cost-sharing in the
national Clean Lakes Program only if they have a
nutrient budget. Major sources of nutrient input such
as sewage traditionally received most attention
because they were so obvious and easy to quantify.
Lesser ones were often ignored or at best only
estimated. Today, it is usually recognized that all
nutrient inputs are additive and may contribute
ultimately to the supply used by plants. Many
phosphorus sources now receiving attention include
precipitation, droppings of migratory birds, burned
gasoline, boats, undisturbed lands, urban land, agri-
cultural land, ground water, industries, and municipal
sewage (Bartsch, 1972). There is more emphasis than
ever in looking at the total watershed as a significant
nutrient source, and the lake and its watershed are
increasingly viewed together as intimately connected
elements in the management scheme. Maps have been
prepared for the United States (Omernik, 1977) that
provide a broad overview of nonpoint source stream-
nutrient level relationships. While there is no real
substitute for a measured nutrient budget, there are
now at least ways to provide quick and relatively
accurate predictions of nonpoint source stream
concentrations of nutrients. With obvious limitations
they can serve where more detailed information is not
available or resources for specific sampling are not at
hand.
The significance of internal phosphorus loading and
its impact on lake response to intense recovery efforts
has become more sharply appreciated through the
studies on Shagawa Lake (Malueg, et al. 1975). This
eutrophic lake has been impacted by municipal sewage
from the city of Ely, Minn, since 1901. First it was
discharged raw, then from 1912 to 1973 with various
degrees of treatment. Early in 1973, the input was
decreased by about 80 percent through tertiary sewage
treatment that yielded an effluent with only 1 /20 mg/l
of phosphorus. Although the lake improved visually and
responded with a prompt and persistent reduction in
phosphorus concentration, phosphorus has not de-
clined to the levels expected. The phosphorus
residence time model projected an equilibrium con-
centration of about 12 ug/l within 1.5 years, but it
reached only 51. This discrepancy was attributed to
feedback from the sediments, primarily during sum-
mer.
Modeling
At the Madison conference, the application of
modeling techniques to eutrophication was not a
prominent discussion subject. In fact, it was hardly'
mentioned, although research recommendation num-
ber 7 urged that "Ecosystem analysis and research on
models for simulating trends in eutrophication should
be strengthened." Two years later a workshop on
Modeling the Eutrophication Process was held at St.
Petersburg, Fla. (Anon. 1969), and this was followed by
a second at Logan, Utah in 1973 (Anon. 1973). Since
then, modeling effort has intensified until a growing
array of modifications, adjustments, and substitutions
have been made to the nutrient input-output and
critical loading models introduced by Vollenweider
(1968). Today, models and modeling are routinely used
in efforts to better understand the internal workings of
specific lakes, to guide regulatory actions, and to
anticipate results. Efforts continue on developing more
useful models not only for small lakes but for the Great
Lakes as well (Thoman, et al. 1979; Ditoro and
Matystik, 1978).
METHODS TO IMPEDE OR CONTROL
EUTROPHICATION
The Madison conference identified research needed
to facilitate control of eutrophication. In particular,
research recommendation number 4 urged investiga-
tors to seek ways to: (1) Limit nutrient input, (2)
accelerate nutrient outgo, (3) impair nutrient availabil-
ity, (4) reduce the volume of water participating in
production of plant material, (5) alter stratification, and
(6) modify ecological systems to provide for accelerated
consumption of plant material by an appropriate array
of animal populations.
Limit Nutrient Input
Because the role of nutrients in eutrophication has
been appreciated for a long time, the concept of control
through shutting off the supply was a natural step. At
Madison, where municipal sewage was discharged to
the chain of lakes for many years, the strong voice of
the people caused several diversions to take place: First
from Lake Mendota in 1899, from Lake Monona in
1936, and from Lake Waubesa in 1958. This is perhaps
the longest community struggle with eutrophication in
the United States. At Seattle, for the same reasons,
effluents from 11 sewage treatment plants were
-------
OPENING SESSION
13
diverted from Lake Washington between 1963 and
1968, and followup studies showed this to be an
effective remedial tool. Diversion has been used more
recently at Lake Sammamish, Wash., and Twin Lakes,
Ohio.
Within the past 13 years, point source nutrient
control strategies have shifted to phosphorus rather
than whole sewage. Reasons for this are obvious and
need no further delineation (Vallentyne, 1970). Two
principal phosphorus control strategies have emerged,
one concerned with sewage, the other with detergents.
Advanced waste treatment to strip phosphorus from
sewage was discussed at Madison (Rohlich, 1969), but
modification of detergent formulas to reduce phos-
phorus input to sewage was not on the agenda. The
idea of advanced treatment, and the technology for it,
are well developed and in many places are keyed to
meeting established effluent standards. Modification of
detergent formulas to decrease or eliminate their
phosphorus content is now required by several
jurisdictions. So today, with the focus on phosphorus,
several things can be said about standards, advanced
waste treatment, and detergents.
Several States, counties, and cities have implement-
ed blanket effluent standards to control point sources
of phosphorus. The United States and Canada have
jointly adopted standards to protect the Great Lakes.
Usually such standards require that treated sewage
contain not more than 1 mg/l of total phosphorus.
While one can sympathize with the managerial desire
for a standard that is technically reachable, financially
tolerable, and simple to administer, it ignores the fact
that each lake is unique and will respond in its own
way. Shagawa Lake helps make this point. Even after
an 80 percent reduction of total phosphorus input,
improvement has been disappointing. This is because
the feedback of phosphorus from the sediments, which
has persisted since phosphorus input was first
curtailed (Larsen, etal. 1979), is greater than expected.
This reaction of Shagawa Lake to an experimental
remedial program too costly for practical use, with an
effluent standard 20 times more stringent than legally
established ones in force, raised a very pertinent
question: How appropriate is a 1 mg/l standard? One
answer comes from the National Eutrophication
Survey. Several of the input-output and trophic state
models were applied to data for 225 survey lakes to find
how many would benefit under an effluent standard. It
was estimated that a 1 mg/l standard would favorably
impact only 22 percent of the selected lakes — a zero
standard no more than 28 percent (Gakstatter, Bartsch,
and Callahan, 1978). This must mean that restoration
cannot be accomplished by simply limiting phosphorus
input. If phosphorus recycling from bottom sediments
is a major factor in the nutrient system, actions to
minimize the result would seem to be required.
Added to this concern is the question of how well
advanced waste treatment plants remove phosphorus
to satisfy the standard. Give or take a little, untreated
municipal sewage contains an annual phosphorus load
of about 1.4 kilograms per capita. Conventional waste
treatment processes reduce this amount by about 36
percent, while phosphorus removal processes can
bring it down by about 68 percent or more. Effluents of
809 sewage treatment plants were sampled during the
National Eutrophication Survey. Of 33 plants using
phosphorus removal processes, the median effluent
concentration of total phosphorus was found to be 1.8
mg/l — nearly twice the usual effluent standard
(Gakstatter, et al. 1978).
Since the end of World War II, about half the
phosphorus in municipal sewage has come from
detergents. Decreasing or eliminating this source has
been found to be almost as effective as advanced waste
treatment in reducing effluent phosphorus. At Onon-
daga Lake, N.Y. for example, a detergent law limiting
phosphorus to 8.7 percent was followed by a 54
percent decrease in inorganic phosphate in treated
sewage discharged to the lake. Average concentrations
in the lake decreased by 57 percent, and Aphani-
zomenon disappeared during the first growing season.
Accelerate Nutrient Outgo
Harvesting a lake's production to help curb eutrophi-
cation through retrieval of nutrients has emphasized
macrophytes because effective weed harvesting e-
quipment has been available for many years. Recent
attempts to harvest planktonic algae in California have
not proved practical. Today, most weed cutting is still a
manicuring exercise with beneficial effects sometimes
persisting the following year (Kimbel and Carpenter,
1979). Obviously, cutting weeds removes some
measure of nutrients because aquatic plants contain
some minimal amounts. But, as a method to control
eutrophication by limiting nutrients, the real accom-
plishment is not impressive. Removal of 428,000
kilograms of plants from Sallie Lake, Minn, retrieved
less than 1.5 percent of the phosphorus entering the
lake (Peterson, Smith, and Malueg, 1974). Recent
experiences elsewhere (Burton, King, and Ervin, 1979)
have also shown that even the greatest potential
harvest will not remove sufficient nitrogen and
phosphorus to offset moderate to heavy loading. The
outlook might differ if harvesting were an adjunct to
cutting off phosphorus input. For whatever reasons,
mechanical plant removal is used in'only five of 102
U.S. lakes now being restored in the federally funded
Clean Lakes Program.
In some ways dredging is an extension of harvesting;
one of its goals is to remove nutrient-laden sediments
to prevent recycle of their nutrients to the overlying
waters. Another frequent goal is to deepen the lake
basin to control macrophytes and improve freedom of
boat movement. Many U.S. experiences with dredging
in recent years seem to have given favorable results
but high cost impedes its wider use. Nevertheless,
there is a growing interest in dredging as a restoration
technique, and more than half the lakes scheduled for
restoration in the national Clean Lakes Program will
use dredging alone or in conjunction with other
procedures.
There are two well known successful examples of
using dilution and flushing to cope with the symptoms
of eutrophication. At Green Lake at Seattle, Wash., the
first introduction of nutrient-poor water from the city's
domestic water supply in 1962 (Ogelsby, 1969)
produced striking improvement. The program has
continued since that beginning. Success here led to a
similar test program at Moses Lake, Wash, where
dilution water from the Columbia River was introduced
-------
14
RESTORATION OF LAKES AND INLAND WATERS
in spring and summer of 1977 and 1978 (Welch, 1979)
and several times since. Impressive reductions in total
phosphorus and chlorophyll a concentrations resulted
and Secchi disk depths increased strikingly. Obviously,
this simple approach to curbing eutrophication is
exceedingly attractive and is being further tested in
four other U.S. lakes. The practical barriers are lack of
large supply of high quality water, absence of physical
structures to introduce it, and need for sympathetic
people residing downstream who are not affronted by
the prospect of receiving the "flushings."
At Snake Lake, Wis. nutrient-rich water was once
pumped out to permit the seepage inflow of higher
quality ground water. This novel approach has not
attracted much attention and, to my knowledge, has
not been used elsewhere.
Where the physical setting permits, hypolimnetic
withdrawal can be used to accelerate nutrient outgo
and improve dissolved oxygen conditions near the
bottom. In reservoirs, where selective depth with-
drawal controls may be available, deep withdrawal may
be a choice approach. In lakes, equipped with only
surface exits, nutrient-rich water must be removed by
pumping or siphoning from the point of maximum
depth. This is not a well-known technique but has been
used in several States.
Impairing Nutrient Availability
Two approaches that reduce the availability of
nutrients have been used with some success. The first
involves chemical treatment of lake water in situ to
precipitate phosphorus — a nutrient inactivation
approach apparently first used at Langsjon, Sweden in
1968. Aluminum sulfate or other aluminum com-
pounds has since been used in many bodies of water
ranging upward in size from Cline's Pond, Ore. at 0.4
hectares to Liberty Lake, Wash, at 277 hectares (Funk
and Gibbons, 1979). With few exceptions the treat-
ments have reduced phosphorus concentration, limited
nuisance algae, and helped maintain adequate oxygen.
At least nine U.S. lakes are being treated by chemical
nutrient removal. Research since 1967 has empha-
sized improving procedures and equipment and
searching for more effective inactivating agents,
including small field tests with zirconium and
lanthanum compounds.
The second approach seeks to immobilize nutrients
through aeration of hypolimnetic water where large
reservoirs of phosphorus reside. Equipment and
procedures have been developed to permit aerating
only the hypolimnetic water without destratifying the
lake. This can be accomplished by injecting air or pure
oxygen or by mechanical means (Fast, 1979). As a
result, nutrient upwelling is minimized and suitable
temperature preserved for cold water fisheries. When
the method is designed to destratify, the lake becomes
isothermal with oxygen available to the bottom, and
other chemical conditions are fairly uniform. Both types
of aeration have been used in Europe and North
America but are not currently popular in the national
Clean Lakes Program.
Reducing the Volume of Water Participating in
Production of Plant Material
During the past 13 years, learning to reduce the
volume of water that participates in plant production
has been largely ignored. New ideas have not emerged.
One or two historical trials come to mind. In one,
decreasing the volume of the photic zone was
attempted by treating two Arizona ponds with the dye
nigrosine (Eicher, 1947). The reduction in light
penetration impaired growth of semi-emergents for a
few years. In 1977 analine dyes were used success-
fully in Nebraska farm pontfs (Buglewicz and Hergen-
rader, 1977). In another trial, weed-choked Deer Lake,
N.J., was treated with commercial fertilizer to stimulate
increased production of phytoplankton (Surber, 1948).
When sufficiently dense they served as a sun shield
and proved successful in curtailing plant growth. With
that purpose accomplished, the lake was drawn down
to dispose of the enriched water. Even with this
success, it is doubtful either one could stand the rigors
of today's environmental impact scrutiny.
Accelerate Consumption of Plant Material
Manipulating biological interactions to benefit lakes
is best known in fishery management. Its use in
alleviating symptoms of eutrophication was mentioned
at Madison and expanded research suggested. But
biomanipulation can only mature as our basic
knowledge of biota and biological interactions becomes
more complete. For now we have few triumphs to
exhibit. We can only point to a few herbivores with
voracious appetites that drive them to attack specific
plants; for example, the flea beetle that devours
alligator weed, another insect that eats water hyacinth,
and the grass carp. Crayfish, snails, swans, and
manatee that were once viewed as promising
candidates, do very little.
Not much has been done to control algal populations
through biological means. Microorganisms that destroy
blue-green algae were isolated a long time ago.
Although promising in laboratory tests, no full scale
lake treatments have been tried in the United States.
Interest in the so-called blue-green algal viruses
appears to be swinging upwards again. Recently,
Shapiro (1979), in championing biomanipulation, urged
us to not be so hypnotized by the easy use of
phosphorus loading models that lake biology is ignored.
Certainly, the admonition is worth your consideration.
LAWS, REGULATIONS, AND
PROGRAMS
It is safe to say that in the United States the past 13
years have produced more eutrophication superlatives
than all preceding history: (1) More dollars spent to
study the subject, (2) more dollars devoted to more
lakes for restoration, (3) more laws and regulations to
expedite correction.
Adoption and current upgrading of a phosphorus-
control plan established under the Great Lakes Water
Quality Agreement between the United States and
Canada is one of the two most important and far-
reaching milestones since 1967. The other is the
passage of the U.S. Clean Lakes Program legislation in
1972 and startup of the program in 1975.
Ten years ago the International Joint Commission
alerted the governments of Canada and the United
States to the accelerating eutrophication in Lakes Erie
and Ontario and cited the danger of permitting
unabated nutrient inputs. The two countries responded
-------
OPENING SESSION
15
by signing the Great Lakes Quality Agreement on April
15, 1972 to jointly implement programs to reduce
phosphorus loads entering the Great Lakes System.
These programs, which have made an impressive start
in these 8 short years, focus mostly on point sources
such as municipal sewage, industries, animal hus-
bandry operations, and detergents. Recognizing the
need to focus on diffuse sources as well and to attain
more stringent phosphorus load targets led to a new
agreement in 1978. A recent draft report (Phosphorus
Manage. Strat. Task Force, 1980) now outlines a
proposed updated plan for phosphorus management in
the Great Lakes. The plan is currently under
consideration by both countries.
Amendments to the Clean Water Act (P.L. 92-500)
passed in 1972 set the stage for a massive national
effort to protect and restore lakes. The resulting Clean
Lakes Program seeks to remedy in-lake problems and
control nonpoint source pollution in the tributary
watersheds. Local interest has been intense, largely
because matching funds are available to help cover the
cost of lake restoration projects. The program thus sets
the stage to demonstrate and evaluate a wide array of
remedial technologies. Unfortunately, the technologies
currently contemplated do not reflect a new giant step
forward since 1967. They include such familiar
approaches as hypolimnetic destratification, bottom
sealing, biofiltration, biomanipulation, chemical nutri-
ent removal, and flushing. Projects are now underway
or imminent in 102 impacted lakes located in 28 States
at a total cost of about $90 million. Each project must
be given a followup evaluation to record success or
failure, but 12 lakes are receiving in-depth study over a
period of years to sense lake response, durability of
improvement, and to gain a better understanding of
why the lakes responded as they did.
As part of this effort, States are required by law to
classify all publicly owned lakes as to their trophic
condition and to identify causal factors. Only a few
States have completed the task but this has stimulated
study of earlier classification schemes and develop-
ment of new alternates. Unfortunately, there is not yet
a uniform scheme for trophic classification.
These actions to protect precious lake resources
were built upon a legislative beginning which, like so
many lacustrine developments, began at Madison
many years ago. One of this country's earliest pieces of
legislation to address the nutrient input problem is
Wisconsin's Lewis Bill. Enacted just before the end of
World War II, it was carefully designed to prevent
Madison's effluent from reaching the chain of Madison
lakes. It was never enforced because it was judged
legally void in 1949. Nevertheless, its purpose was
ultimately attained, and the lakes are now protected.
Jurisdictions throughout the country have since passed
laws constraining nutrient loadings. Most either
specify maximum allowable amounts of phosphorus in
treated sewage that reaches susceptible waters, or
they set maximum amounts of phosphorus for
detergents.
In Minnesota, if a discharge from a sewage
treatment plant enters a lake directly, the phosphorus
content must not exceed 1.0 mg/l; if it reaches a lake
via a river, up to 2.0 mg/l are allowed. Illinois has an
effluent standard of 1.0 mg/l for discharges that flow
to Lake Michigan. Other States have similar standards.
Laws regulating detergent phosphorus were passed
in New York and Indiana in 1971. Both required total
elimination of phosphorus from detergents by specified
dates in 1973. Laws of the same intent were also
passed in Florida, Maine, Michigan, Minnesota,
Connecticut, and Oregon, as well as Chicago, Akron,
and Dade County, Fla.
A 1971 Iowa law required mandatory soil conserva-
tion, viewing as a nuisance soil erosion that causes
siltation damage. That erosion damage is not in the lost
soil alone but often in the silted lake as well is now
more appreciated. Funding of National Soil Conserva-
tion Programs is guided by this fact.
CONCLUSION
In conclusion, I wish to leave three points with you:
First, as I look at the lake protection and restoration
technology in use today, I am convinced there is room
for substantial improvement. I hope the discussions
you enjoy here this week will identify many new ideas
that can be pursued.
Second, scientific curiosity will continue to provide
better answers to the question: "What causes
eutrophication?" As the new answers emerge, so will
the prospects to develop the improved technology
needed for more effective lake management in the
years ahead.
Third, human attitudes, often of people in powerful
places, must be changed if lakes generally are to be
protected or restored. Four years ago, several col-
leagues and I used the following words to introduce a
paper on the status of eutrophication in the United
States (Bartsch, et al. 1978). They were spoken by
Chief Seattle of the Suquamish tribe in Washington
Territory in 1854 when agreeing to transfer Indian land
to Federal ownership.
"This shining water that moves in the streams and
rivers is not just water but the blood of our ancestors. If
we sell you land, you must remember that it is sacred,
and you must teach your children that it is sacred, and
that each ghostly reflection in the clear water of the
lakes tells of events and memories in the life of my
people. The water's murmur is the voice of my father's
father.
"The rivers are our brothers, they quench our thirst.
The rivers carry our canoes, and feed our children. If we
sell you our land you must remember, and teach your
children, that the rivers are our brothers, and yours,
and you must henceforth give the rivers the kindness
you would give any brother.
". . .The earth does not belong to man; man
belongs to the earth."
In light of the historical record, these pleading words
are even more timely today than when they were first
spoken 126 years ago.
REFERENCES
Allum, M. O., R. E. Glessner, and J. H. Gakstatter. 1977. An
evaluation of the National Eutrophication Survey data.
Working Pap. 900. U.S. Environ. Prot. Agency.
Anonymous. 1969. Modeling the eutrophication process.
Prpc. Workshop, St. Petersburg, Fla. Dep. Environ. Eng.,
University of Florida.
-------
16
RESTORATION OF LAKES AND INLAND WATERS
Anonymous 1973. Modeling the eutrophication process.
Proc. Workshop. College of Eng., Utah State University.
Bartsch, A. F. 1972. Role of phosphorus in eutrophication.
EPA-R3-72-001. U.S. Environ. Prot. Agency.
Bartsch, A. F., et al. 1978. Eutrophication in the United States
— past, present, future. Proc. 1st, 2nd USA USSR
Symposia on the effects of pollutants upon aquatic
ecosystems. EPA 600/3-78-076. U.S. Environ. Prot.
Agency.
Buglewicz, E. G., and G. L. Hergenrader. 1977. The impact of
artificial reduction of light on a eutrophic farm pond. Trans
Neb. Acad. Sci. 4:23
Burton, T. M., D. L. King, and J. L. Ervin. 1979. Aquatic plant
harvesting as a lake restoration technique. Pages 177-185
in Lake restoration. EPA 440/5-79-001. U.S. Environ. Prot.
Agency, Washington, D.C.
Ditoro, D. M., and W. F. Matystik, Jr. In preparation.
Mathematical models of water quality in large lakes. 1. Lake
Huron and Saginaw Bay — model development, verification
and limitations. Environ. Res. Lab., U.S. Environ. Prot.
Agency. Duluth, Minn.
Eicher, G. J. 1947. Analinedye in aquatic weed control. Jour.
Wildl. Manage. 11:193.
Fast, A. W. 1979. Artificial aeration as a lake restoration
technique. Pages 121-131 in Lake restoration. EPA 400/5-
79-001. U.S. Environ. Prot. Agency, Washington, D.C.
Funk, W. H., and H. L. Gibbons. 1979. Lake restoration by
nutrient inactivation. Pages 141-151 in Lake restoration.
EPA 440/5-79-001. U.S. Environ. Prot. Agency, Washing-
ton, D.C.
Gakstatter, J. H., A. F. Bartsch, and C. A. Callahan. 1978. The
impact of broadly applied effluent phosphorus standards on
eutrophication control. Water Resour. Res. 14:1155.
Gakstatter, J. H., et al. 1978. A survey of phosphorus and
nitrogen levels in treated municipal waste-water. Jour.
Water Pollut. Control Fed. 50:718.
Joint Industry Government Task Force on Eutrophication.
1969. Provisional algal assay procedure.
Kimbel, J. C., and S. R. Carpenter. 1979. The dynamics of
Myriophyllum spicatum biomass following harvest. Pages
43-49 in Aquatic plants, lake management, and ecosystem
consequences of lake harvesting. University of Wisconsin.
Larsen, D P., et al. 1979. The effect of wastewater
phosphorus removal on Shagawa Lake, Minnesota: Phos-
phorus supplied, lake phosphorus and chlorophyll a. Water
Res. 13:1259.
Likens, G. E., ed. 1972. Nutrients and eutrophication: the
limiting-nutrient controversy. Proc. Symp. at Michigan State
Univ. Feb. 11-12, 1971. Allen Press, Inc. Lawrence, Kan.
Likens, G. E , et al.
Science 172:873.
1971. Nutrients and eutrophication.
Malueg, K. W., et al. 1975. A 6-year water, phosphorus, and
nitrogen budget for Shagawa Lake, Minnesota. Jour.
Environ. Qual. 4:236.
National Academy of Sciences. 1969. Eutrophication:
Causes, consequences, correctives. Washington, D.C,
Ogelsby, R. T. 1 969. Effects of controlled nutrient dilution on
the eutrophication of a lake. Pages 483-493 in Eutrophica-
tion: Causes, consequences, correctives. Natl. Acad. Sci.,
Washington, D.C.
Omernik, J. M. 1977. Nonpoint source — stream nutrient
level relationships: a nationwide study. EPA 600/3-77-105.
U.S. Environ. Prot. Agency, Washington, D.C.
Peterson, S. A., W. Smith, and K. W. Malueg 1974. Full-scale
harvest of aquatic plants: nutrient removal from a eutrophic
lake. Jour. Water Pollut. Control Fed. 46:697.
Phosphorus Management Strategies Task Force. 1980.
Phosphorus management for the Great Lakes Int. Joint
Comm., Windsor, Ontario April 30.
Rast, W., and G. F. Lee. 1978. Summary analysis of the North
American (U.S. portion) OECD eutrophication project:
Nutrient loading — lake response relationships and trophic
state indices. EPA 600/3-78-008. U.S. Environ. Prot.
Agency, Washington, D.C.
Rohlich, G. A. 1969. Engineering aspects of nutrient removal.
Pages 371-382 in Eutrophication: Causes, consequences,
correctives. Natl. Acad. Sci., Washington, D.C.
Shapiro, J. 1979. The need for more biology in lake
restoration. Pages 161 -167 in Lake restoration. EPA 400/5-
70-001. U.S. Environ. Prot. Agency, Washington, D.C.
Stewart, K. M., and G. A. Rohlich. 1967. Eutrophication —a
review. Calif. State Water Qual. Control Board 34:1.
Surber, E. W. 1948. Fertilization of a recreational lake to
control submerged plants — effects of fertilization program
upon bathing, boating, fishing. Prog. Fish Cult. 10:53.
Thoman, R. V., R. P. Winfield, and J. J. Segma. 1979.
Verification analysis of Lake Ontario and Rochester
embayment three dimensional eutrophication models. EPA
600/3-79-094. U.S. Environ. Prot. Agency, Washington,
D.C.
U.S. Environmental Protection Agency. 1977. North Ameri-
can Project — a study of U.S. water bodies. EPA 600/3-77-
086.
U.S. Environmental Protection Agency. 1971 • Algal assay
procedure — bottle test.
U.S. Environmental Protection Agency.1978. A compendium
of lake and reservoir data collected by the National
Eutrophication Survey. Working Pap. 475.
Vallentyne, J. R. 1970. Phosphorus and control of eutroph-
ication. Can. Res. Dev. 3:36
Vollenweider, R. A. 1968. Scientific fundamentals of the
eutrophication of lakes and flowing waters, with particular
reference to nitrogen and phosphorus as factors in
eutrophication. Tech. Rep. DAS/CSI/68.27 Organ. Econ.
Coop. Dev., Pans.
Welch, E. B. 1979. Lake restoration by dilution. Pages 133-
139 in Lake restoration. EPA 440/5-79-001. U.S. Environ.
Prot. Agency, Washington, D.C.
-------
17
NORTH AMERICAN OECD EUTROPHICATION PROJECT:
THE UNITED STATES STUDY
W. PAST
International Joint Commission
Washington, D.C.
G. F. LEE
Colorado State University
Fort Collins, Colorado
ABSTRACT
The U.S. portion of the North American Project included 34 water bodies ranging from ultra
oligotrophic to hypereutrophic. The U.S. OECD study consisted of gathering, analyzing, and
synthesizing existing eutrophication-related water quality data from water bodies which had been
intensively studied, rather than conducting new field studies. It was determined that the
Vollenweider nutrient loading relationship correlated well with the trophic states identified by the
investigators. After the study, approximately 40 additional water bodies were found to exhibit the
same basic phosphorus load-response relationships. A summary of the original U.S. OECD study
as well as subsequent studies is presented.
The United States portion of the North American
Project consisted of 34 water bodies located primarily
in the north central and northeastern United States. In
contrast to the other Projects, the U.S. OECD water
bodies were not of one specific type, but rather
exhibited a range of trophic character from ultra-
oligotrophic to hypereutrophic. Further, the United
States participation consisted of gathering, analyzing
and synthesizing existing water quality and other data
related to eutrophication from water bodies which had
already been extensively studied, rather than conduct-
ing new field studies as was done in the other OECD
Projects. The U.S. OECD Study was completed before
the other Projects.
A summary analysis on the U.S. portion of the North
American Project was prepared by Past and Lee
(Summary Analysis of the North American (U.S.
Portion) OECD Eutrophication Project: Nutrient Loading
— Lake Response Relationships and Trophic State
Indices, Ecological Research Series, EPA-600/3-78-
008, 1978). The individual lake studies were reported
in a standardized format and were compiled by Seyb
and Randolph (North American Project: A Study of U.S.
Water Bodies, Ecological Research Series, EPA-600/3-
77-086, 1977).
The U.S. Environmental Protection Agency was the
lead agency for the study in the United States.
Emphasis was on the development and use of
quantitative lake management models for assessing
eutrophication and the effects of phosphorus control
programs, using the statistical nutrient loading models
of Vollenweider as an initial focus.
The 34 water bodies in the U.S. study included 24
lakes, nine impoundments and one estuary. When sub-
basins of these water bodies were considered, there
were 37 distinct water bodies in the U.S. study. The
principal investigators for the individual water bodies
classified 25 as eutrophic, five as mesotrophic, and
seven as oligotrophic as of the completion of the study.
Twenty-eight water bodies had mean depths less than
10 meters (range - 1.7 to 313 m), while 16 water
bodies had surface areas greater than 1,000 hectares
(range = 47 to 1.7x107 ha). Twenty water bodies had
hydraulic residence times greater than 1 year (range =
0.08 to 700 yr), while 28 had Secchi depths less than 3
meters (range = 0.6 to 28 meters). The general
morphometric, hydrologic, chemical, and biological
characteristics of the U.S. OECD water bodies, as well
as other pertinent data, are summarized in the original
Rast and Lee report.
Components of the summary analysis of the U.S.
study included an examination of analytical procedures
for major biological or chemical water quality para-
meters, determination of the limiting nutrient, and
evaluation of methods for the identification of major
nutrient sources and for the calculation of nutrient
loads. The nutrient load estimates provided by the U.S.
OECD investigators were compared with estimates
derived on the basis of the Vollenweider model relating
influent and in-lake phosphorus concentrations, and
with estimates based on nutrient export coefficients
and land use patterns within the U.S. OECD water body
watersheds.
In general, it was found that the phosphorus and
nitrogen load estimates for the water bodies were
within a factor of ±2 of the load predicted on the basis
of the Vollenweider approach and the nutrient export
coefficients. Possible reasons for any anomalous load
estimates that were encountered were investigated.
Phosphorus residence times were also investigated
and were generally found to be shorter than the
hydraulic residence times, usually by several-fold,
-------
18
RESTORATION OF LAKES AND INLAND WATERS
being shortest as the degree of eutrophy increased.
It was determined that the results of the Vollen-
weider nutrient loading diagram relating between
annual areal phosphorus load and hydraulic load
correlated well with the trophic states identified by the
individual investigators for the U.S. OECD water
bodies. The similar positions of the water bodies on a
nitrogen loading diagram as on the phosphorus loading
diagram indicated a relative constant ratio of nitrogen
to phosphorus loading to the water bodies.
Using the phosphorus load chlorophyll model of
Vollenweider as a guide, a statistical correlation was
developed between phosphorus loading, normalized by
mea.n depth and residence time, and chlorophyll a
concentrations in the U.S. OECD water bodies. The
chlorophyll Secchi depth relationship in water bodies
was also examined and used to derive a direct
relationship between normalized phosphorus load and
Secchi depth. A statistical correlation was also
developed which directly related normalized phos-
phorus load and hypolimnetic oxygen depletion rate.
These models are presented in graphic form in the
summary analysis of the U.S. study.
Several trophic status indices were also compared
using the U.S. OECD water bodies as a data base, and
were found to predict relatively identical results. A
trophic status index was also developed using the
Vollenweider diagram relating annual areal phos-
phorus load and hydraulic load, thereby relating trophic
status to critical phosphorus loading levels. A large
number of correlations between nutrient loads and/or
various in-lake chemical, biological, and physical
parameters in the U.S. OECD water bodies were also
examined. The use of different analytical and sampling
methodologies and the varying number of data sets for
a given correlation, however, limit the general
usefulness of these correlations based solely on data
from the U.S. study.
Overall, the statistical models developed in the U.S.
study can be used to predict the changes in water
quality related to eutrophication that will result from
changes in phosphorus loads to water bodies for which
phosphorus is the key element limiting planktonic algal
growth. These models relate the normalized phos-
phorus load of phosphorus-limited water bodies to
several commonly used water quality parameters. The
U.S. study indicated the validity of the basic Vollen-
weider approach for determining the critical phos-
phorus loading level and associated overall degree of
fertility of water bodies. The models developed during
the U.S. OECD study offer simple, practical, and
quantitative methodologies for assessing the expected
effects on water quality of eutrophication control
programs based on (1) phosphorus removal from
domestic wastewaters, and (2) other phosphorus
controls.
Following the completion of the U.S. study, approxi-
mately 40 additional load-response relationships in
water bodies were evaluated and found to exhibit the
same basic phosphorus load-response relationships.
The basic approach has also been extended in the
development of a correlation between normalized
phosphorus loads and overall fish yield. Further, the
predictive capability of this statistical modeling ap-
proach has been demonstrated by comparing measure
changes in water quality response parameters which
occurred after phosphorus load reductions to about 18
water bodies, with the changes predicted by the models
developed in the U.S. study. A detailed manual on the
practical use of the U.S. OECD models has also been
prepared. A summary of these subsequent related
studies is available from the authors.
-------
19
MONITORING OF INLAND WATERS: THE NORDIC
PROJECT
SVEN-OLOF RYDING
University of Uppsala
Institute of Limnology
Uppsala, Sweden
INTRODUCTION
Combating eutrophicaton requires a combination of
knowledge and money. The eutrophication problem has
been of central interest for OECD, the Organization for
Economic Cooperation and Development. In 1966
OECD supported a study of existing literature on
eutrophication with special reference to the roles
played by nitrogen and phosphorus in the process. This
report emphasized that relevant measurement data
were insufficient to permit more precise guidelines and
advice for the control of eutrophication.
In 1973 the OECD Water Management Group
initiated a 4-year cooperative program to monitor
inland waters. The program was subdivided into three
regional projects: the Alpine, Nordic, and North
American, plus a non-regional reservoir project. This
paper presents results from the Nordic Project in a
condensed form. The full report including recommen-
dations to improve and optimize lake management
programs and the outcome of using predictive lake
models has been published by the project coordinator,
the Nordic Cooperative Organization for Applied
Research — NORDFORSK (Ryding, 1980).
BACKGROUND DATA
The participation from the Nordic countries consisted
of research data from 10 lakes. The lakes differed a lot
regarding climate, morphometry, hydrology, and load-
ing conditions. The following ranges for some
important background data may be noted:
Height above sea level (m) 0.3 103
Catchment area (km2) 84 26,480
Ice coverage (days) 60 150
Lake surface (km2) 2.7 1,912
Volume (km3) 0.02 74
Average depth (m) 3.1 153
Outflow (m3-s~1) 0.8 320
Hydraulic residence time (year) 0.2 57
Nitrogen supply (g-m~2Yr~') 1.8 101
Phosphorus supply (g-rrf2Yr~') 0.1 3,6
As a consequence of different land-use patterns of
the drainage basins and the different morphometrical
and hydrological conditions of the water bodies, water
quality varied greatly. A high transparency was found
in lakes low in P and algae (chlorophyll) and vice versa.
N- and P-concentrations often maintained the same
relation to each other whether total concentrations or
soluble inorganic fractions (NhU + NO2 NO3) - N and
PCU -P were considered. Primary production and
chlorophyll were closely related. The hypolimnetic
oxygen depletion rates in the Nordic lakes, however,
did not seem to correlate to primary production, but the
lack of data regarding these parameters in some lakes
makes a straight comparison difficult.
The annual nitrogen load was found to be less
correlated to the nitrogen concentration in the lake
waters compared to the corresponding relationships for
phosphorus. The supply of P and its concentration in
the lake waters were more strongly correlated to the
trophic state of the water body than N, indicating that P
can be regarded as a key chemical element limiting
planktonic algal growth. P concentrations in lake water
were closely related to chlorophyll, based on different
annual and seasonal calculations. As a measure of
biological response predicted from the nutrient load the
parameter chlorophyll a was found to be superior to
primary production.
The very strong correlation between annual maxi-
mum and summer average or annual average
concentrations of chlorophyll reveals that a certain
basic level of chlorophyll is a prerequisite for peak
values to occur.
Adoption of the Nordic data to lake models based on
the phosphorus load versus mean depth relationship
was somewhat misleading, particularly for lakes with a
high flushing rate. Later modifications also taking into
account the hydraulic residence time and phosphorus
retention predicted the trophic states about equally
well for the majority of the Nordic lakes if compared
with that subjectively chosen by each project leader.
Phosphorus loading diagrams transferred to nitrogen
overestimated the trophic states of the Nordic lakes as
either a result of a too low conversion factor or the
unsuitability of applying data from P-limited lakes into
nitrogen-load models.
Predictive models based on nutrient load-lake
response relationships are valuable tools for assessing
the expected effects of phosphorus reduction, e.g.,
from domestic wastewater by advanced wastewater
treatment or sewage diversion as illustrated for three
of the Nordic lakes, on eutrophication control pro-
grams, and as a base for establishing phosphorus load
and water quality criteria.
-------
20
RESTORATION OF UXKES AND INLAND WATERS
SPECIAL HIGHLIGHTS
The treatment of the Nordic data was performed in
two ways — "handmade," sometimes excluding
outliers before calculations of correlations and load-
response relationship, and purely computerized on the
complete data set. The approach of treating the whole
data set totally computerized did not reveal the overall
relationships verified from the other sub-projects of the
OECD eutrophication program including the handmade
treatment of the Nordic data presented here. In a
research program carried out in a diverse group of
lakes, it is therefore necessary that data treatment and
assessment of the results are made using "biological
know-how." Treating biological research data using
only a statistical-technical approach may be hazardous.
Using algal assay, the algal growth potential, the
"free capital of nutrients," was generally found to
increase with a higher trophic state. Phosphorus was
generally the most limiting nutrient if the total nitrogen
to total phosphorus ratio exceeded 13. In waters with
lower values nitrogen played a major part regulating
algal growth. The corresponding figure if the ratio is
calculated for the dissolved inorganic fractions was 1 2.
Trace elements, iron and/or a chelating agent (EDTA)
were found to stimulate algal growth in some of the
Nordic lakes. Using information on the growth-limiting
role of nitrogen and phosphorus obtained by perform-
ing algal assays a stronger correlation for the nutrient
load-lake response relationships was obtained by
adding the growth effects of these nutrients together
and an expression for the "load of algal growth-limiting
nutrients."
The results from the Nordic project permit a
composite model, predicting the summer average and
annual maximum concentrations of chlorophyll in a
phosphorus limited lake derived from simple empirical
findings on phosphorus load and phosphorus con-
centration in lake water. It is important that the data
collected in the OECD study on monitoring of inland
waters are used also for evaluation and assessment of
the validity of the existing models in lake management.
The contribution from the Nordic project to improve the
predictive power of lake models is a list of different
aspects regarding sampling procedure, loading calcu-
lations, the limiting nutrient concept and phosphorus-
chlorophyll relationships that ought to be fulfilled to
optimize the outcome of using the models. These
aspects, graphically illustrated in Figure 1, should be
considered as a first-cut analysis to be done before
interpreting the outcome from a comprehensive data
set being used in the models.
Accurate loading figure* may
not be obtained if:
•- eitimatei are made bated on
the land Die pattern in the
drainage bairn or between
lubbaiini with open boun-
-- the in-and outlet are lo-
cated close to each other
-- the imported material con-
uru of eaiy-iettled mate-
nil
-- existence of internal sour-
CORRELATIONS
The outcome Uling a deicribed
relationihip for P may not be
reliable if
-- other nutnenu or factor!
limit algal growth
- algae, with a tpecific P-
requirament, are abundant
-- the applied data let de-
rive! from another climatic
region
LIMITING NUTRIENT
Before uung a predictive
model it may be valuable to
nutrient.
For N and/or P thu can be
done by
comparing N/P i
outlet
the i
- regarding N/P in lake water
- algal allays
If thit 19 not possible do not
apply data from a lake to a P
model If the P concentration
> 100 mg/m3
A lake model can be no more
reliable than ib data base.
therefore
- a comparatively frequent lamp-
ling 11 neceiury
- if average valuei are to be
uMd the lamplmg ought to be
evenly distributed for that
Figure 1. — Different aspects that ought to be fulfilled for
optimizing the outcome when using predictive lake models.
ecosystem models may not be able to replace the
simple parameters as chlorophyll and phosphorus.
Furthermore, using even a complex model based on
various interactions among the components of the
aquatic ecosystem it may be found that some lakes do
not fit the model because of special or local conditions.
Improved modeling techniques for large-scale lake
management schemes must be developed in conjunc-
tion with sound methods for making routine measure-
ments of sensitive environmental variables.
REFERENCES
Rydmg, S.-O. 1980. Monitoring of inland waters. OECD
Eutrophication Programme. The Nordic Project. Nordic
Cooperative Organization for Applied Research, Secretariat
for Environmental Sciences, Publ. No. 1980:2.
DISCUSSION
It is difficult to define the complex interactions that
occur in a body of water to the point where detailed
accurate assessments can be made of the impact of a
point or nonpomt wastewater source discharge on a
lake
Over the years, two basic approaches have evolved
for use by agencies in making decisions on the
limitations of nutrients into aquatic systems, i.e.,
ecosystem models and P-load models. Useful as they
might be, for practical and economical reasons complex
-------
21
OECD EUTROPHICATION PROGRAM REGIONAL
PROJECT: ALPINE LAKES
HANSJORG FRICKER
Swiss Federal Institute for Water Research and Water Pollution Control (EAWAG)
Dubendorf, Switzerland
ABSTRACT
In a coordinated international program, the correlation between nutrient loading of lakes and their
biological-chemical (trophic) response was examined in four partial projects. The Alpine project
which is described here, dealt with five countries: Austria, the Federal Republic of Germany,
France, Italy, and Switzerland. A critical analysis is made of the quantitative assessment of
nutrient load. Trophic classification was defined by means of a probability approach. The overturn
value of phosphate phosphorus and the maximum chlorophyll concentration were the most
significant trophic level indicators. The annual and the maximum daily primary production could be
associated with the spring overturn value of phosphorus by a hyperbolic estimate. Apart from
simple correlation techniques, empirical phosphorus loading models (elementary mass balance
concepts; mixed reactor theory) and modifications of the steady-state conditions for a conservative
compound with an additive time-variable term were used. No striking differences were observed
among the correlations. On the basis of the correlation found, a lake's reaction to a change in
phosphorus load can be predicted to a certain degree. The limits of the applied concept are
discussed.
The Alpine project that is described here, is part of
the OECD International Investigation Program on Lake
Eutrophication. The total program covered approxi-
mately 200 natural and artificial lakes, spread around
the world. They were grouped into projects according to
technical and geographical criteria:
• Alpine project
• Shallow lakes and reservoir project
• Scandinavian project
• US/Canada project
The results of each project are published in a
comprehensive report.
The Alpine project dealt with five countries: Austria,
the Federal Republic of Germany, France, Italy, and
Switzerland. Data on 28 Alpine lakes or lake basins
were obtained by voluntary cooperation. Most of the
data were calculated or adapted to the purpose of the
study by using a unit process. Several ringtests were
made to establish parallels between the results of each
laboratory, thus refining the method (detailed results
are given in the report).
The lakes of this entire region are strongly influenced
by their mountainous surrounding, topographically and
climatically. Basic criteria for classifying a lake as
Alpine are as follows:
• Complex mineralogy: limestones, dolomite, granite
etc.
• V-shaped or with rocky, steep side slopes, except
those lakes lying in the Swiss midlands and similar flat
valleys in Germany (Bavaria) and Italy (Brianza lakes).
• Relatively deep (100 and more meters), and
because of this, a special stratification behavior (if the
wind exposition of the valley is good (Urnersee), then
these lakes can mix fully. Consequently, they have high
tolerance level for phosphorus. But in other cases the
mountains prevent full circulation (Kreuztrichter, Lago
di Lugano) and these lakes tend to become anaerobic in
the hypolimnion.) A significant phosphorus input from
the sediments is a further consequence.
The final selection of the lakes for the OECD program
was influenced by the following points:
• Various trophic states, hydrological residence time,
and mixing regime.
• A monitoring program already in operation.
SUMMARY OF THE RESULTS
A critical analysis was made of the quantitative
evaluation of the nutrient load. Although the limnologi-
cal behavior of many lakes with respect to their
biological-chemical reaction has become well known,
loading measurements have often been neglected. But
in the last 5 to 10 years, increasing attention has been
given to this problem. A main step forward was made in
the OECD Eutrophication program. A deeper evaluation
of the assessment of nutrient loading in the Alpine part
makes it clear that the influxes had not been
sufficiently investigated. Especially water flow and
highwater surveillance have been insufficiently mea-
sured. (In a special chapter of these Proceedings
guidelines are given to encourage limnologists to
measure nutrient loading directly to acquire accurate
data base).
An important question during the OECD study was
the classification of trophic state. An attempt was made
to define a trophic state as a system of probability
approaches. The overturn value of phosphate phos-
phorus and the maximum chlorophyll concentration
were most significant trophic level indicators.
The main concept of the study was to describe the
average behavior of lakes in response to available data
on nutrient loads based on simple statistical approach-
-------
22
RESTORATION OF LAKES AND INLAND WATERS
es. The ratio between nitrogen and phosphorus
confirms that in most cases phosphorus is in fact the
limiting factor. To a certain extent it is possible to
calculate the phosphorus concentration in the lakes by
the annual inflowing concentration or by the more
complex Vollenweider model. A similar correlation can
be achieved by the somewhat different Schindler
approach.
Primary productivity measurements were unfortu-
nately not given adequate attention, even though this is
the only parameter which directly influences the
trophic level. Nevertheless, to some extent it is possible
to mathematically describe the relationship between
primary production and load by using statistical
approaches. The annual and the maximum daily
primary production could be associated with the spring
overturn value of phosphorus in a hyperbolic estimate.
The nutrient loading characteristics determined in
the program make it possible, by using simple
correlation techniques, to generalize to some extent on
ths average statistical behavior of lakes in response to
their nutrient loading. Statistically, it is possible to
predict, with a certain degree of reliability, the average
lake concentration of phosphorus and chlorophyll, as
well as the average annual primary production.
In general, the connections between load and lake
parameters are plausible, although we have to take
notice of the fact that they are based on statistics, with
a certain deviation and probability. Therefore they may
not be applied uncritically for practical purposes and
predictions. They need an interpretation in which the
peculiarities of the single lake in question must be
considered.
Due to technical reasons it was possible to measure
only the external input of phosphorus, while the
internal load, which is also an essential parameter in
lake eutrophication but extremely difficult to deter-
mine, was not taken into consideration. For this reason,
direct application of the results (e.g. for therapeutical
purposes) is not always possible.
It is an established fact that all the technical
measures for lake recovery and conservation aim at
reducing the phosphorus load. The lake's tolerance
load can be calculated by means of this study. In
numerous cases, it will be possible from now on to
reach this limit of tolerance by external measures
(wastewater treatment, ring trunk sewers); in others,
however, this will not be possible. Additional protection
measures will then have to be taken to decrease or
interrupt the internal phosphorus supply (aeration,
destratification, discharge of hypolimnetic waters). To
quantify these measures, a sound knowledge of the
complexity of the internal nutrient cycles is essential,
knowledge which the present study is not able to
supply. The complex strategy resulting needs planning
which is only possible on the basis of a time dependent
(dynamic) model. The statistical models of this report
are a suitable tool for decisions mentioned here and for
political decisions on whether a dynamic modeling is
necessary (which is rather time consuming and costly).
In any case the statistical model's application is the
indispensable basis or the first step to the following
decisions.
Some non-scientific but nevertheless most important
facts should be considered:
• The friendly cooperation between the laboratories
has been an extremely fruitful experience, has created
personal friendship and solid mutual confidence.
• By common work, in particular by the calibration
tests, the quality of the data has increased significantly.
• The necessities of the OECD program produced
enough pressure to set new analytical developments.
• The OECD program has been important enough to
set up new research laboratories which today are of
great value in the water protection networks of the
respective countries.
Abstracted from:
Fricker, Hj. 1980. OECD Eutrophication Program: Regional
Project Alpine Lakes. Swiss Fed. Board Environ. Prot.
(Bundesamt fur Umweltschutz). CH-3003 Bern, Switzer-
land.
-------
23
THE SHALLOW LAKES AND RESERVOIRS PROJECT
JURGEN CLASEN
Wahnbachtalsperrenverband
Siegburg, Federal Republic of Germany
The OECD regional project "Shallow Lakes and
Reservoirs" included a large number of extremely
varying waterbodies which, for the purpose of the
project, were divided into two main groups: natural
basins and artificial basins. The natural basins were
mainly shallow waterbodies (located in Ireland, Japan,
Netherlands and the United Kingdom). The group of
artificial basins included pump storage reservoirs,
which have been created by construction of ring-
shaped barriers (located in Netherlands and United
Kingdom), and semi-artificial reservoirs created by
impounding natural valleys (located in Australia,
Germany, Netherlands, Spain and United Kingdom).
Several reservoirs in the United States portion of the
North American Project were also included in this
project.
Initial data analysis consisted of a statistical survey
of the collected data. This survey showed that the
values were not normally distributed. Logarithmic
scales were suitable for most of the correlation graphs.
As expected from the title of this project, the average of
all the mean depths was low (approximately 9 meters).
If one considers shallow lakes to be defined as lakes in
which stratification never occurs, or in which it occurs
for only very short periods, then 47 percent of the lakes
in this project are shallow. The average retention time
of all the lakes under study was approximately 6
months, which is remarkably short. The majority of the
lakes (approximately 70 percent) were eutrophic.
If a model is to be developed which describes the
relation between algal biomass and nutrient input, it is
of fundamental importance to determine whether
nitrogen and phosphorus is the limiting nutrient. If the
N:P ratio is calculated for every lake or reservoir in the
project and compared with the N:P ratio considered to
be ideal for algal growth (N:P — 15:1),then the limiting
nutrient can be established. The results obtained show
that almost no lake or reservoir in this project can be
considered nitrogen-limited. Thus, it was possible to
apply models decribing the relation between phos-
phorus supply and trophic state to these waterbodies.
For this purpose Vollenweider's well-known formula
for phosphorus loading was generalized, and the
coefficients were recalcualted by iteration. This lead to
a slightly different relationship, which showed that
phosphorus retention in the lakes and reservoirs
examined was greater than that calculated by
Vollenweider's original formula. This deviation seems
to be independent of the lake type in this project since
natural lakes and pumped storage reservoirs did not
suit the original formula better than semi-artificial
reservoirs, to which the "chain of reactors" theory
could be applied. This theory assumes that the long and
narrow semi-artificial reservoirs can be regarded as a
cascade of reactors in which phosphorus is more
effectively retained than in one large reactor.
In further analysis of this deviation, it seemed best to
first examine the extent to which the phosphorus
retention depends on the inflow concentration,
disregarding the water residence time. This is done
simply by plotting average in-lake phosphorus con-
centration against average inflow concentration, and
correlating the two parameters. A simple power curve
was used for regression.
Although water residence time was not taken into
account, the correlation was remarkably good. It is
significant that in the equation obtained, the coefficient
was clearly less than 1. This means that, independent
of retention time, a high inflow concentration is
generally reduced in a lake to a greater degree than a
low inflow concentration. Thus, phosphorus retention
is of more importance in eutrophic lakes than in those
which are oligotrophic. This is contrary to the opinion
that it is in eutrophic lakes most of the phosphorus
which reaches the bottom as a result of sedimentation
is released again. One should, however, consider the
fact that phosphorus is probably more effectly used in
oligotrophic lakes, i.e., the algae in these lakes contain
less phosphorus than those in eutrophic lakes. In
eutrophic lakes, however, algae take up phosphorus in
excess of their requirements (luxury uptake), which has
long been known to limnologists.
A parameter describing phytoplankton density is
chlorophyll a, the primary assimilation pigment of all
algae. Not only is it easy to determine the chlorophyll
content of algae in a relatively simple way, but the
technique is also exceptionally sensitive, which means
that even low plankton densities in oligotrophic lakes
can be determined. In some reservoirs, the total
phosphorus-chlorophyll relationship was much lower
than expected. This was the case in the reservoirs,
"Honderd en Dertig" and "Petrusplaat", and in the
Australian reservoir, 'Mount Bold." In these reservoirs,
it was not phosphorus, but rather light which was
limiting primary production. Therefore, these reservoirs
were excluded from calculations of the relationship
between nutrient supply and algal.
For determining the correlationship between the
chlorophyll concentration and the total phosophorus
concentration in the euphotic zone, both Mitscherlich's
saturation function and a simple power curve were
applied. Very similar results are obtained with both
methods. The fact that the power coefficient was less
than 1 indicates that the ratio betwen chlorophyll and
-------
24 RESTORATION OF LAKES AND INLAND WATERS
phosophorus decreases with increasing phosphorus
concentrations. It was not possible to show that
chlorophyll concentrations depended on total nitrogen
concentrations, which was to be expected.
In general, the investigator's evaluations of trophic
state was quite consistent with the previously valid
border lines in the phosphorus loading models, as well
with the recent statistical data for phosphorus and
chlorophyll. On the basis of phosphorus only, the
Australian reservoir. Mount Bold, deviated consider-
ably from the general border lines defining predicted
trophic state. It had been classified as "mesotrophic"
by the investigator, whereas on the basis of its
phosphorus content, it should be considered " eu-
trophic" or even "hypereutrophic."
It is interesting to note that Mount Bold can be
classified as being mesotrophic or oligotrophic with
almost the same probability as if using chlorophyll
concentration as the criterion. This is probably because
a considerable amount of the phosphorus is not fixed in
planktonic algae but instead is in the silt. The Queen
Elizabeth-ll Reservoir in the United Kingdom did not fit
the picture at all. This is because since the total depth
was high compared to the euphotic depth, it was
possible to keep algal density low by means of artificial
circulation.
-------
25
BACKGROUND AND SUMMARY RESULTS OF THE OECD
COOPERATIVE PROGRAM ON EUTROPHICATION
R. A. VOLLENWEIDER
National Water Research Institute
Canada Centre for Inland Waters
Burlington, Ontario, Canada
J. J. KEREKES
Canadian Wildlife Service
Biology Department
Dalhousie University
Halifax, Nova Scotia, Canada
THE PROBLEM OF EUTROPHICATION
Early in the 1960 decade, it became obvious that a
large number of lakes and reservoirs were rapidly
changing their trophic characteristics due to the
addition of plant nutrients originating largely from
human activities. The main nutrient sources identified
were municipal and industrial wastewater and agri-
cultural and urban runoffs.
Eutrophication is the response to this over-
enrichment by nutrients (primarily phosphorus and
nitrogen) and can occur under natural or manmade
conditions. "Manmade" eutrophication, in the absence
of control measures, proceeds at an accelerated rate
compared to the natural phenomenon. A recent survey
(cf. Vollenweider 1979) has shown that eutrophication
is one of the main forms of water pollution reported in
countries throughout the world. The resultant increase
in fertility in affected lakes, reservoirs, slow-flowing
rivers and certain coastal waters causes symptoms
such as algal blooms, heavy growth of certain rooted
aquatic plants, algal mats, deoxygenation and, in some
cases, unpleasant odor, which often affects most of the
vital uses of the water, such as water supply, fisheries,
recreation or aesthetics. In short, manmade eutrophi-
cation of inland bodies of water becomes synonymous
with the deterioration of water quality and as such
frequently causes considerable extra economic costs.
Manmade accelerated eutrophication can, in princi-
ple, be reversed by the elimination or reduction of the
nutrient supply from such as municipal and industrial
wastewaters, agricultural wastes and fertilizers. In
most cases, however, it is not possible to eliminate all
sources of nutrient supply. Thus, it is important to
understand the qualitative and quantitative relation-
ships which exist between nutrient supply and the
degree of eutrophication in order to be able to develop
sound lake management strategies to control eutrophi-
cation at minimum costs.
HISTORY OF OECD ACTIVITIES IN
EUTROPHICATION
In 1967 a group of experts under the chairmanship of
Professor O. Jaag (EAWAG, Zurich) recommended to
the OECD that a comprehensive survey be made of the
existing literature on eutrophication processes. This led
to the publication of a report, "Scientific Fundamentals
of the Eutrophication of Lakes and Flowing Waters with
Particular Reference to Nitrogen and Phosphorus as
Factors in Eutrophication" by Vollenweider (1968). This
report introduced the concentration of nutrient loading
and lake response but also stressed the inadequacy of
limnological data for broad generalizations and for
producing precise guidelines for eutrophication control.
Further, a symposium on "Eutrophication in Large
Lakes and Impoundments" was held in Uppsala,
Sweden, and the resulting report was published by the
OECD in 1970.
In spite of the advances achieved in eutrophication
control, many basic questions concerning eutrophica-
tion remained unanswered, and it became obvious that
a broader limnological data base was required for inter-
comparison between bodies of water and assessment
of the status of lake eutrophication. The nutrient
loading concept and the related concept of loading
tolerance had been consolidated and accepted by a
large segment of the international scientific com-
munity, but controversies whether carbon and other
growth factors rather than phosphorus or nitrogen limit
algal growth in lakes continued for some time.
In 1971 the OECD established a Steering Group on
Eutrophication and in February 1973, approved and
adopted an "Agreed Program on Evaluation of
Eutrophication Control" and charged the Steering
Group on Eutrophication Control with the responsibility
for developing and coordinating the agreed program,
bringing into account its effectiveness, cost and
feasibility. Four ad hoc expert groups carried out the
program:
1. Expert Group on Detergents (published 1973);
-------
26
RESTORATION OF LAKES AND INLAND WATERS
2. Expert Group on Impact of Fertilizers and
Agricultural Waste Products on the Quality of Waters
(published 1973);
3. Expert Group on Wastewater Treatment Processes
for Phosphorus and Nitrogen Removal (published
1974);
4. Planning Group on Measurements and Monitoring
(published 1973).
The three expert groups and the planning group
completed their reports in 1972. The planning report
"Summary Report of the Agreed Monitoring Projects
on Eutrophication of Waters" (published 1973) gave a
common system of agreed measurements, guidelines
von background data and comments on existing
methods of sampling. It also outlined the basis for an
international program of measurements and monitor-
ing of waters being undertaken by interested OECD
member countries. This program came to a closure in
1980 and has resulted in a Synthesis Report and four
Regional Reports already being published. A fifth Test
Case Report is presently in its final stage.
nutrients potentially available, and on the transfer
function, i.e. the amount of nutrients released per unit
of time and unit of surface.
THE THREE LEVELS DETERMINING THE PRODUCTIVITY OF BODIES OF WATER
[^ -PHYSICAL COMPLEX ^l^ _^^^^^^.:_^^»-
«. ^ rtFOT-MFMICAL COMPLEX --^-^-
(
t
}
CLIMATIC PROPERTIES SEASONALLY LIGHT.
TEMPERATURE HYDROLOGY
GEOLOGICAL PROPERTIES HOCK FORMATIONS
CRYSTALLINE -SEDIMENTARY
-^
^-~. . 1 ] URBANIZATION L-_/ A I
f A V*^ LAND use VEGETATION - SOILS ; INDUSTRIALIZATION i\?_J
FLOW REGIME LIGHT-TEMPERATURE SILT-CLAY
REGIME TURBIDm
II 1
HUWC MACROCHEMISTRY NUTRIENTS P
SUBSTANCES TDS N.elc
— ^i^jL©-1 ^L^
MORPHOMETRY STRATIFICATION MIXING
WATER RENEWAL
PROCXJCTIVITY
® BASIN PROPERTIES © WATER PROPERTIES © LIMNOLOGICAL PROPERTIES
(A) ANTHROPOGENIC ALTERATIONS
THE CONCEPTUAL BACKGROUND
Scientifically speaking, eutrophication is but a
special aspect of water productivity. Seen in this
perspective, studies on eutrophication have to respond
to the same conceptual references as productivity
studies in general. Productivity is the expression of the
external physiographic complexes of the system as a
whole, as well as of its internal physico-chemical and
biological dynamics. Accordingly, the trophic properties
of bodies of water, lakes, estuaries, sea coasts or
running waters, have to be considered as the resultant
of a sequential nexus of geographic, geochemical,
climatic, hydrological and other factors.
In applying this concept to eutrophication studies on
lakes and reservoirs, the scheme expressed in Figure 1
proposes a quasi deductive procedure to derive the
cause effect relationships which determine any
observed specific limnological situation from the
characteristics of the catchment system by progression
from the general properties of the system to the
specific conditions of the water body considered. In the
progression from level to level, the degree of freedom
for the next level is narrowed down, i.e. the specific
properties of the physico-geochemical complex at the
top controls the hydrologic and qualitative properties
and characteristics of the water deflux level, which in
turn determines the limnological level in its connota-
tion "productivity."
In order to bring this concept into perspective, at
least one specific transfer compartment and two
transfer functions have to be singled out:
A. The vegetation-soil complex acts as an inter-
mediary between the physico-chemical complex and
the water property level. Under natural conditions this
compartment is practically the only source compart-
ment in terms of nutrients, yet — due to man's
intervention — has been substantially altered over the
centuries. The historical and modern development in
land use, urbanization and industrialization has had
effects on both the size of this compartment in terms of
Figure 1. — Principal components and relationships
determining the productivity of bodies of water.
The transfer function from the basin to the receiving
waters is expressed in terms of export coefficients (e.g.
kg/km2 year) for each source. Point sources*, in
general, are expressed in terms of unit load, yet in
principle, they can also be expressed in terms of export
coefficients at the condition that their density
distribution can be established. The specific values of
these export coefficients vary considerably from
situation to situation, depending on the general
geographic, climatological, hydrological and other
conditions, as well as on the specific land use, urban
and industrial development, etc. Export coefficients for
phosphorus vary from less than 5 kg/km2 y to over 500
kg/km2 y and for nitrogen from less than 50 kg/km2 y
to more than 3000 kg/km2 y.
In spite of this large variability, it is at times possible
for specific geographic regions to apply lump values as
has been shown by Vollenweider (1968, 1978) for
average European conditions, and by Rast and Lee
(1 978) for U.S. conditions. However, uncritical transfer
of such coefficients to unknown regions can lead to
gross error.
B. The nutrient loading concept, as distinct from the
transfer function refers to the receiving water body and
in most genera/ form means the intensity of supply to a
given body of water of any chemical factor necessary
for plant growth; in our context, however, its meaning
has been restricted to nitrogen and phosphorus.
* It is now customary to distinguish between diffused and
point sources. However, such a distinction has primarily
operational meaning: point sources, as opposed to diffused
sources, in general, offer less difficulty for quantification, and
at the same time are more amenable for technological control.
-------
27
OPENING SESSION
The theoretical limnology for decades has ignored
this aspect, or at least neglected it, despite the early
announcement made by e.g. Naumann (1932), Aberg
and Rodhe (1942), Ohle (1955) a.o. Accelerated
eutrophication of bodies of water over the last two or
three decades has brought this problem into the open.
The nutrient loading concept as defined here implies
the connotation of a quantifiable property called
"external load" which establishes the functional
relationship between the basin and the trophic
conditions of the receiving waterbody, and as such, is
fundamental to the understanding of the total system.
From the methodological point of view, the quantifi-
cation of the load-response relationship remains not
without certain perplexities. Part of these relate to the
question regarding the most appropriate way to
express the load. Advantages and disadvantages of
various options (e.g. absolute total amounts, specific
loading per unit of surface or volume over a selected
time-space, average inflow concentrations, etc.) are
still a matter of discussion.
More important, however, the loading-trophic reac-
tion relationship cannot be dealt with adequately
without due consideration being given also to the fate
of the various load components of a given substance
within the lake system itself. An improvement over
consideration of sole totals could be achieved by
distinguishing at least two principal components and
corresponding pathways, i.e. one component which
enters the internal cycle via an "autotrophic" pathway
— and which becomes immediately available to
primary producers, and a component which enters into
the internal cycle via a "heterotrophic" pathway of a
more refractory nature (cf. Figure 2). In part, this aspect
relates also to the question of what fraction is, or is not
biologically available. In practice, the analytical
distinction of these components is only partially
possible, yet in order to understand the full array of
reactions of different bodies of water to a given (total)
load, a clarification of this problem is not without
importance. Also, in many cases the internal loading
cannot be neglected, though in many lakes this internal
component remains far below the importance of the
external loading. The exact quantification and de-
pendency on external load is not entirely solved as yet,
although essential progress has been made (cf. e.g.
Golterman 1980).
However, important in this context is the basic idea
that the in-lake bioproduction and recycling machinery
(to use a more engineeristic analogy) is fed and driven
by the external loading, and maintains itself depending
on this external load in a repetitive cyclic steady state
as long as no (unidirectional) alteration of the external
supply occurs. On the other hand, any (unidirectional)
change in the supply function will have as a
consequence an alteration of the internal responses of
the machinery speeding up or decreasing the velocities
of exchange between the compartments, and corre-
spondingly producing a change in size of each
compartment.
In pursuing this concept, the question is posed as to
how far we can go at present to quantify the postulated
relationships. This implies the necessity to establish
"AUTOTROPHY"
(AUTOTROPHIC
PATHWAY)
•ALLOTROPHY"
(HETEROTROPHIC
PATHWAY)
i i
I PRIMARY I I SECONDARY I
PRODUCTION | »• PRODUCTION .
I POOL I I POOL I
I II I
T
EXCHANGEABLE
PERMANENTLY BURIED
SEDIMENTS
Figure 2. — Relationships between the principal lake-internal
compartments and pathways of the external and internal
loading components.
and to elucidate the function of those parameters
which primarily govern the relationship between the
external load and the reaction of the body of water.
From an applied point of view, such an understanding
of the various relationships — always expressed in
quantitative terms — would provide the scientific basis
to develop criteria to manage the system; in particular,
it would provide the basis to estimate the nutrient
supply reduction required for lakes which in terms of
preset water quality standards, appear to be over-
fertilized.
The far reaching, practical, i.e. economical, implica-
tions of solving these questions have been recognized
by OECD and have provided the motivations for the
OECD Cooperative Program on Eutrophication which
is the main theme of the following expose.
APPROACHES TAKEN IN THE OECD
COOPERATIVE PROGRAM ON
EUTROPHICATION
We realize that we have oversimplified the problem
considerably, yet this has been done with the intention
of bringing the problem into focus. Also, in speaking
further on about the OECD Program, its scope, outcome
and results, much oversimplification will be necessary,
which does not exclude that the single collaborators, as
well as the members of the Steering Committee, are
well aware of the many difficulties arising in specific
cases in applying a simplified approach.
-------
28
RESTORATION OF LAKES AND INLAND WATERS
To introduce the rationale for the OECD Cooperative
Program on Eutrophication, it is necessary to recall the
situation regarding the level of understanding of the
nutrient load-trophic reaction relationship, particularly
in regard to nitrogen and phosphorus, some 15 years
ago. At that time, only a few reliable data on nitrogen
and phosphorus loadings existed in the whole applied
and theoretical limnological literature, and much of the
data were no more than crude estimates which hardly
permitted any founded generalization. Nonetheless,
when the first author proposed in 1968 (cf. Vollen-
weider, OECD Technical Report) that in principle it was
feasible to distinguish between "acceptable" and
"excessive' loading, this proposal was welcomed in
the scientific community, and immediately had sub-
stantial influence on practical decisions as well as
stimulating a plethora of follow-up research.
It rapidly became clear through a number of
meetings organized by OECD that only through
international cooperation would it become possible to
arrive at a sufficiently large amount of comparable data
to derive valid quantitative relationships. Therefore, in
about 1972 it was decided to launch a major
cooperative program involving a majority of the OECD
member countries. Some 18 countries, including more
than 50 research centers covering between 100 to 200
lakes, have adhered to this program. It was conceived
to tap into and make use of ongoing research but also
to initiate new research. Accordingly, a full uniformity
in approach could not expect to be achieved, yet this
shortcoming was hoped to be counterbalanced by the
large variety of individual lake situations covered by the
program.
How did we develop this program? It was quite clear
from the beginning that the focus would be on nitrogen
and phosphorus, but that this aspect would have to be
related to the particular geographic and limnological
conditions of each lake individually studied. Further, it
was necessary to develop a common language, to
screen particular techniques and methods as to
suitability and reliability, and to select those study
items which appeared to be both pertinent to the
success of the program, and logistically feasible, i.e.
accessible for most cooperating centers involved. With
evolving results, serious thought had to be given to
data elaboration and exploration of the most useful way
to correlate them.
In order to account for geographic variability, as well
as for logistic considerations, we organized the
program into four main projects:
1. An Alpine Project
2. A Northern Project
3. A Reservoir and Shallow Lake Project
4. A lump project for North America
Each project was headed by a regional coordination
center, regional chairmen, plus some consultants
forming a Technical Bureau for overall coordination.
The first author has had the pleasure of chairing this
committee over the last few years, and wishes to
acknowledge the cooperation he enjoyed from his
colleagues, particularly Drs. Ambuhl, Bernhardt, Fors-
berg, Golterman, Lee, Loffler, Maloney, Ravera and
others, for steering this program, and the consultants
responsible for synthesizing the material into report
form (Drs. Kerekes, Clasen, Fricker, Lonholt, Ryding
and Rast).
Table 1 provides an illustration of the kind of
approach taken in developing a common language to
identify parameters to be measured, or thought to be
necessary to collate the information gathered into a
consistent picture. It was understood that not
necessarily all parameters would be measured in each
individual case; some have been singled out as
absolutely essential, whereas others have been left to
the choice of the individual centers, in accordance with
their capabilities and expertise.
In contrast to an approach of studying but a few
examples only in depth, the chosen approach permits
covering a wide spectrum of individual cases in an
extensive way. Hence, our attention was not primarily
focused on specific mechanisms, but on information
that is amenable to statistical analysis. We wish to
state this explicitly because at times the philosophy of
the program has been misunderstood, particularly by
those who expected a kind of material which could be
used for dynamic modelling. From the very beginning,
elaboration of the data at the basis of correlation and
other comparative techniques thought to be meaning-
ful, were envisaged. From this we expected to
determine the cause-effect relationship in the sense of
what we may call statistical behavior," examples of
which shall be given.
What kind of results have we obtained from this
program? The program has covered a wide variety of
limnological situations, including almost every type of
lake and impoundment of the temperate region, and a
few subtropical lakes and reservoirs, as well as some
estuarine situations. Although the majority of lakes
well studied fall into the meso- to eutrophic categories,
a sufficient number of lakes representing oligo- and
ultra-oligtrophic types have been included.
It is not the place here to discuss the whole array of
results, conclusions and implications for practical
management. These aspects are covered in the
regional reports, in site specific reports and scientific
papers, and in the final synthesis report. The
integration of the available data has proven to be a
worthwhile though difficult task. Such difficulties refer
to both conceptual aspects as well as to straight-
forward problems with data screening, selection and
appropriate interpretation. In many cases, it is not
immediately available whether a data point represents
a particular situation, a general uncertainty, or some
unintentional mistake. Frequency of sampling, e.g., is a
major factor in determining the reliability of a reported
system itself. Superimposed on these problems are
problems connected with calculation procedures; the
choice of which of the various alternatives to use often
remains a matter of taste rather than a matter of
objective judgement.
As an example, loading figures represent a key
parameter in the whole study, yet do we know little
about the inherent uncertainty of any specific value
reported. In our judgment, it is unlikely that individual
year specific loading figures in most cases are better
than ±25 percent. The natural year to year variability
in loadings, in addition, is found to be in the same order
-------
OPENING SESSION
29
Table 1. — Categorization of parameters for measuring and monitoring eutrophication.
Ergodic (Resultant) Variables
A. Short Term Variability: High B. Short Term Variability: Moderate to Low
Causative Variables
- Phytoplankton biomass
- Major algal groups and
dominant species
- Chlorophyll a and other
phytopigments
- Paniculate organic carbon
and nitrogen
- Daily primary production rates
- Secchi disk visibility
- Zooplankton standing crop
- Bottom fauna standing crop
- Epilimnetic A P, A N, A Si
(A = difference between winter and
summer concentrations)
- Hypolimnetic Oa and A Oi
- Annual primary production
- Nutrient Loadings
- Total Phosphorus
- Ortho phosphates
- Total Nitrogen
- Mineral Nitrogen (NO3 + NH3)
- Kjeldahl Nitrogen
- Nutrient Concentrations
- Same as above
- Reactive Silica
- Others (e.g. Microelements)
- Morphometric parameters of lake and
catchment area
- Flushing regime
- Geological and climatic parameters
• Land use
• Urbanization and industrialization
- Main nutrient sources
Related Descriptive Parameters
Temperature and mixing regime
- Conductivity, pH, alkalinity
- Major ion spectra
- Insolation and optical properties
- Others as deemed necessary
of magnitude (in some cases also considerably higher),
so that representative loading estimates have a built-in
uncertainty of at least ±35 percent. In-lake parameters
such as biomass, chlorophyll, nutrient parameters, etc.,
are affected by similar uncertainties that have to be
taken into account in data interpretation and correla-
tion.
In its final ouput, the OECD Program has paid
attention to the following aspects:
a. the qualitative assessment of the trophic state of
bodies of water in terms of a few easily measurable
parameters;
b. the dependence of this state on nutritional
conditions and nutrient load;
c. translation of these results to the needs of
eutrophication control for management.
One of the recurring problems we have run into
during the study was the question of how to relate the
classical trophic terminology— which is qualitative in
nature — with the quantitative information provided in
regard to selected parameters. In other words, the
question arose of how far it is possible to quantify, in an
objective way, the qualitatively defined trophic cate-
gories.*
Though apparently of academic interest, this ques-
tion is not without meaning, in two ways. First, it
relates to what has previously been stated relative to
the need of a common language between limnologists
themselves. Second, it relates to how the limnological
terminology applies to practical management. From the
practical point of view, there is no unequivocal
relationship between the main trophic limnological
* The pressing need for clarification in this context becomes
apparent if one recalls such examples as Lake Erie, which in
the early sixties was "dead," then became "eutrophic," and
finally is now considered, at least in regard to the main body of
the lake, as mesotrophic.
categories and water usage. The relationship depends
on specific use requirements. A categorization of
bodies of water for fishery purposes need not
necessarily correspond to the one for recreational
purposes or to the one for domestic water supply, and
none can entirely be matched with the limnological
categories. Generally speaking, however, one can say
that, proceeding from oligotrophy to eutrophy, multiple
use of any water progressively becomes adversely
affected with increasing trophy (cf. Table 2). Given this
inherent ambiguity, therefore, it is important to attach
quantitative meaning to the limnologically defined
categories as the basic reference independent of their
specific application. The OECD study has led to some
interesting and not necessarily anticipated results.
The quantitative information given by the single
contributors, together with their subjective judgment,
were combined into a 4 x 5 matrix and for each block,
mean and standard deviations have been calculated. A
log-transformation of the original data was found to be
necessary; the results are given in Table 3.
Table 2. — Trophic characterization of lakes impairment of various
uses.
Limnological
Characterization
Oligolrophic Mesotrophic Eutrophic
General level of production
Biomass
Green and/or blue-green algae
fractions
Hypolimnetic oxygen content
Impairment of multi-purpose
use of lake
• low
low
low
high
little
medium
medium
variable
variable
variable
high
high
high
low
great
-------
30
RESTORATION OF LUKES AND INLAND WATERS
Table 3. — Preliminary classification of trophic state in theOECD Eutrophication Program. Trophic status is assigned based on
the opinion of the investigator of each lake. The geometric mean (based on log 10 transformation) was calculated after removing
values < or > x 2 SD obtained (where applicable) in the first calculation.
Variable (Annual
Mean Values)
Total
Phosphorus
mg/m3
Total
Nitrogen
mg/m3
Chlorophyll a
mg/m3
Chlorophyll a
Peak Value
mg/m3
Secchi
Depth m
x = geometric mean
SD = standard deviation
( ) = value in bracket refers
x
x + 1 SD
x + 2 SD
Range
n
x
x± 1 SD
x± 2 SD
Range
n
x
x + 1 SD
x ± 2 SD
Range
n
x
x ± 1 SD
x ± 2 SD
Range
n
x
x ± 1 SD
x ± 2 SD
Range
n
to the number
Oligotrophic
8.0
4.85-13.3
2.9 -22.1
3.0 -17 7
21
661
371-1180
208-2103
307-1630
11
1.7
.8-3.4
.4-7.1
0.3-4.5
22
4.2
2.6- 7.6
1.5-13
1.3-10.6
16
9.9
4.9-16.5
3.6-27.5
5.4-28.3
13
of variables (n) employed
Mesotrophic
26.7
14.5-49
7.9-90.8
10.9-95.6
19(21)
753
485-1170
313-1816
361-1387
8
4.7
3. 7.4
1.9-11.6
3. -11
16(17)
16.1
8.9-29
4.9-52.5
4.9-49.5
12
4.2
2.4- 7.4
1.4-13
1.5- 8.1
20
in the first calculation.
Eutrophic
84.4
48 -189
16.8-424
16.2-386
71(72)
1875
861-4081
395-8913
393-6100
37(38)
14.3
6.7-31
3.1-66
2.7-78
70(72)
42.6
16.9-107
6.7-270
9.5-275
46
2.45
1.5-4.0
.9-6.7
.8-7.0
70(72)
Hypereutrophic
750-1200
2
100-150
2
0.4-0.5
2
Clearly, most investigators consider a lake to be
oligotrophic when the annual mean total phosphorus
concentration is <10 mg P/m3 It is noteworthy,
however, that a few lakes with <10 mg P/m3 were
classified as either mesotrophic or eutrophic. Careful
examination of the data revealed that in these cases
the lakes have received an increased nutrient load in
recent years, and as a consequence, have undergone
some perturbation and change in trophic response.
This may be in the form of a noticeable growth of
attached filamentous algae along the shore near
nutrient inflows, often accompanied by the appearance
of a nuisance algae not observed before, however,
without producing fundamental repercussions in the
overall metabolism of the lake, noticeably its hypo-
limnetic oxygen conditions. At the other extreme, lakes
with an annual total phosphorus concentration >30
mg P/m3 and as high as 80 mg P/m3 were assessed as
mesotrophic by some investigators. In these cases, a
variety of reasons, e.g. short water residence time or
high turbidity, a high rate of grazing by zooplankton in
the absence of fish, a.o., prevented the development of
a high standing stock of phytoplankton, and hence, the
lakes did not exhibit eutrophic charcteristics.
In regard to nitrogen, no consistent picture evolved.
In particular, it was impossible to separate oligotrophic
from mesotrophic lakes, although as a general rule,
lakes of more eutrophic characteristics tend to have
higher nitrogen concentrations.
A somewhat clearer delineation of trophic categories
resulted, however, when allocation was based on
chlorophyll a concentrations. In general, lakes were
assessed as oligotrophic, mesotrophic or eutrophic
when annual mean chlorophyll a concentrations were
<2.5 to 10, or > 10 mg chl a/m3, respectively. No lake
was classified as eutrophic with an annual mean
concentration of chlorophyll a< 2 mg/m3 In regard to
"worst case" situations, i.e. peak chlorophyll values,
lakes are considered to be oligotrophic, mesotrophic
and eutrophic when annual peak chlorophyll a
concentrations are around 5, 16 and > 25 mg/m3,
respectively.
What emerged from the assessment of all informa-
tion available, however, led to the conclusion that there
is no possibility of defining strict boundary values
between trophic categories. While the progression
from oligo- to eutrophy is a gliding one — as has been
stressed many times in the past — any one
combination of trophic factors, in terms of trophic
category allocation, can only be used in a probabilistic
sense. The probability distribution for the two single
factors, yearly average phosphorus and chlorophyll, for
the three main categories (oligo, meso, eutrophy) p|us
the two boundary categories (ultra-oligo and hyper-
-------
OPENING SESSION
31
trophic) is exemplified in Figures 3 and 4. e.g. the
probability of classification of a body of water having a
total phosphorus concentration of 10 mg/m3, respec-
tively, would be as follows:
ultra-oligotrophic
oligotrophic
mestrophic
eutrophic
hypertrophic
Phosphorus Chlorophyll
10% 6%
63% 49%
26% 42%
1% 3%
0% 0%
In judgment terms, then, such a water body is best
classified as oligotrophic with a certain tendency
toward mesotrophy. However, exceptionally, such a
body of water may have excellent ultra-oligotrophic
characteristics, or to the contrary, may show signs of
grave deterioration, as is the case with Lake Mjosa.
Probability Distrfcution for Trophic Categories
0.5-
mg
Figure 3. — Probability distribution of trophic categories
relative to average phosphorus concentrations.
Evidently, this way of looking at trophic categoriza-
tion has considerable management implication. If, in a
given case (e.g. a drinking water reservoir), it is
important that certain water quality characteristics are
maintained, then the management objectives must be
set at some level slightly lower than would be required
for maintaining average conditions.
To manage a lake with a certain objective in mind, we
need knowledge of (i) which of the nutritional factors
controls the system and (ii) what the relationship is
between nutrient loading and the trophic reaction of
the lake.
In regard to the first aspect, one of the primary
results evolving from the data collected is confirmation
that in at least 80 percent of the cases studied
phosphorus was found to be the production-controlling
factor; some cases remained inconclusive and the rest
were identified as nitrogen limited, or controlled by
some other factor.
In regard to the second aspect, the relationships
between trophic characteristics, such as nutrient
concentrations, mean annual chlorophyll, peak chloro-
phyll and loading, have been shown to be amenable to
quantification; in accordance with the program objec-
tives, these relationships are statistically not deter-
ministically defined. Hence, if these relationships are to
be used for prediction, the built-in uncertainty has to be
taken appropriately into account.
Restricting the discussion to phosphorus, the results
are based on the following methodology:
It has been obvious for some time now that simple
relationship between areal or volumetric loading and
lake phosphorus levels cannot be established without
consideration of sedimentation and flushing (cf. e.g.
Vollenweider 1969, 1975, 1976; Dillon 1975, Kerekes
1975). Basically, this relationship has to be thought of
as follows:
In the most simple way, this scheme can be
expressed mathematically as
Probability Distribution for Trophic Categories
0.5-
10
mg[Chll/m3
d[P]«
dt
Eq. 1
where
[P]A = average total lake concentration (which includes
both dissolved and particulate phosphorus
components)
[P]x = average inflow concentration of total
phosphorus
rp = average residence time of phosphorus
TW = average residence time of water
Figure 4. — Probability distribution of trophic categories
relative to yearly average chlorophyll concentrations.
-------
32
RESTORATION OF LUKES AND INLAND WATERS
The righthand terms represent the average rate of
supply to and the average rate of loss of total
phosphorus from the lake, respectively, and the
lefthand terms the corresponding temporal variations
of the average lake concentration. Note that in this
formulation no specific assumptions are made as to the
mechanism of loss.
Several possibilities are open to deal with the above
equation 1, yet in principle, it reduces to evaluating
statistically the quotient rp/Tw as a function of
parameters controlling the system such as mean depth,
epi-hypolimnion ratio, hydraulic load, length of stratifi-
cation, etc., assuming steady state conditions.*
From the various attempts made to analyse these
relationships, the fact evolved that mean depth and
hydraulie Joad are the most important factors, and that
TP/7w can be approximated by a function of
the form
OECD LAKES
1,000 p
100 -
1000
= [PUP]* «
a-qsbZc
Eq. 2
Figure 5. — Relationship between flushing corrected average
inflow concentrations and average lake concentrations of
phosphorus.
Approximate values fora, b andc were found to be 1,
-.5 and +.5 so that equation 2 reduces to
•= [P] \ I
1
1
1 +-
(Vollenweider 1976). These findings correspond to
results of similar approaches made by Larsen and
Mercier (1975), Dillon (1974), Kirchner and Dillon
(1975), Chapra (1975), Chapra and Tarapchak (1976),
Reckhow (1978), a.o. which all are variations of the
same theme.
Accordingly, mean lake phosphorus should be
predictable from load, in principle, by
IP] x =
Figure 5 shows that this indeed is the case yet (2b)
slightly underestimates concentrations at low levels,
and overestimates concentrations at higher levels.
*The term "steady state" is referred to in this context as
"repetitive state over time" for which I ± d[P]/dt = 0. Time
resolution is 1 year.
All these formulations, and their variations, contain
the underlying assumption that lakes can be treated as
mixed reactors in steady state. This is not true for most
lakes. It is therefore surprising that simplified
relationships of this sort provide a workable basis,
which in principle means that a large spectrum of lakes
(governed by phosphorus) behave statistically in a
similar way. The relationships derived describe the
average statistical relation pattern of lakes between
phosphorus load and phosphorus concentration.
Used as a diagnostic criterion, these relationships
also provide a tool to identify "outlayers." The term
"outlayer," as used here, refers to both statistical and
functional variability. Indeed, outlayers from the rule
may indicate simple data uncertainty as Rast and Lee
(1978) have shown to be the case for lakes for which
the load has been either under- or over-estimated.
However, outlayer lakes have also been identified
which behave functionally differently; either the
assumption of steady state does not apply, their
sedimentation quota is above or below normal, or an
internal or external disturbance of the system exists.
Examples for each possibility could be listed, yet more
importantly, in most cases it was possible to identify
the reason for deviation.
This experience shows that it would be wrong to
discard equation I simply because a given data set
would not fit it. Strong deviations from this relationship
can be used as a diagnostic indication for a particular
situation which requires further attention. Conversely
it would be wrong to blindly apply this relationship for
-------
OPENING SESSION
33
predictive purposes, regardless of special limnological
conditions.
The next step in the sequence was to establish the
relationship between chlorophyll (yearly and peak
values) and nutrient concentrations. Without entering
into detail for the present review, it may be said that, on
average, the yearly mean chlorophyll concentration
was found to be between 25 to 30 percent of the
average total phosphorus (cf. Figure 6A). Peak
chlorophyll values (which are of particular importance
for practical considerations) on the other hand, resulted
as roughly three times average chlorophyll, but
exceptionally can be considerably higher (cf. Figure
6B).
OECD LAKES
1000 q
OECD LAKES
100 =-
10 =-
r "i r~li07
[Chlqj = .58[PJX
r=.88 n = 5l p = 0 - / /
, '. V' '
/••,/&/
/
' /
. /
/
y/
'*>%'•/'
' .//< /,
//
/ /J o excluded from
/ // •/ regression
'//•/I i
I I I 1 I I III I I I 1 I I Ml I I I I I I
10
100
1000
Figure 6. — Relationship between average lake phosphorus
concentrations and chlorophyll. A. Yearly average chlorophyll.
B. Peak chlorophyll observed.
Interestingly, chlorophyll apparently also resulted in
being correlated to nitrogen in many cases, yet
statistical discrimination tests have shown that this is
primarily due to coupling of nitrogen with phosphorus.
In particular cases, however, the dependence of
bioproductivity on nitrogen, as well as on other factors,
has been found to be unquestionable. The interaction
between phosphorus and nitrogen has been identified
as an area which requires further research.
In the light of what has been said thus far, a close
relationship between phytoplankton biomass (as mea-
sured by chlorophyll) and phosphorus load can be
expected. The findings are illustrated with Figures 7A
and 7B. In regard to statistical variability, the same
applies regarding the relationship between phosphorus
loading and concentration. However, it is to be stressed
that chlorophyll is but a crude parameter to estimate
biomass. Indeed, cases did come to light indicating that
the biomass/chlorophyll ratio can vary by a factor of up
to 3 and is therefore a major contributor to the
scattering observed.
Nevertheless, the biomass (chlorophyll)/loading
relationships are perhaps the most important results of
the OECD study thus far. Within the range of the
identified uncertainties, they permit estimation of the
phosphorus reduction necessary to reduce eutrophica-
tion to any preset level of biomass. The main
conclusion which one can draw from the OECD results
is the fact that the production level of any given water
body, in principle, is proportioned to its nutrient load,
and therefore that load reduction will have effects
proportional to the reduction achieved.
In the long run and with consideration that
exceptions from this rule exist, it is desirable to base
such judgments not solely on standing crop but also on
related dynamic parameters. Unfortunately, the OECD
study has not permitted convincing establishment of
relationships between loading and dynamic para-
meters, such as primary production and hypolimnetic
oxygen depletion rates, etc. This is due, in part at least,
to the dearth of usable data points and in part to
considerable difficulties in measuring such parameters
uniformly. The problem of hypolimnetic oxygen
conditions is further compounded by conceptual
uncertainties (e.g. oxygen depletion rates versus
apparent or potential oxygen deficit).
The following is a short account of the present state
of the art.
The relationship between primary production and
phosphorus deviates structurally from the chlorophyll
[P] relationship by its non-linearity. This is due to the
self-shading effect of the biomass with increasing
levels of productivity which can be dealt with by
introducing a generalized primary production model.
This model assumes that the annual primary produc-
tion can be expressed with a hyperbolic function
similar U that of a daily photosynthesis integral (cf.
Vollenweider 1970) i.e.
[chl]
1C (g/m2-y)=K--
fw+ n [chl]
-------
34
RESTORATION OF LAKES AND INLAND WATERS
where [chl] is the average yearly chloropohyll con-
centration of the euphotic zone, £w a characteristic
average extinction coefficient (1/m) which includes
turbidity, humic substances and other colored sub-
stances, and ri the specific vertical extinction coef-
ficient per unit of chlorophyll.
In order to establish the relationship to nutritional
conditions, the chlorophyll term in this equation can be
substituted by the corresponding relationships, and the
remaining parameters calculated from measured data
OECD LAKES
HOOp
o>
E
.c
o.
1000
OECD LAKES
1000
I •< I II Kll I I I I 111 ll I I I 111
1000
Figure 7. — Relationship between flushing corrected average
inflow concentrations and chlorophyll. A. Yearly average
cholorophyll. B. Peak chlorophyll observed.
by least square techniques. Correspondingly, tne
hypolimnetic oxygen depletion rates should be predict-
able from primary production. This hypothesis further
implies that the relationship between oxygen depletion
rates (expressed as areal hypolimnetic oxygen deple-
tion rates) and nutritional conditions, should parallel
those for primary production.
Our preliminary results show that this is indeed the
case. Yet, the much larger scattering of the data points
also shows that factors other than those taken into
account in our analysis are involved in determining
primary production and hypolimnetic oxygen condi-
tions. The higher uncertainty, e.g., in linking
hypolimnetic oxygen depletion rates with loadings, as
found in our study, depends undoubtedly on the
complex interactions between the epilimnetic and
hypolimnetic regime in each individual case. The
underlying factors relate to specific lake morphology,
length and type of thermal stratification, vertical
entrainment and oxygen transfer, and interactions
between sediments and overlying waters. It is evident
that, in order to reduce the uncertainties, much
additional work is required.
HOW FAR DID THESE PRELIMINARY
RESULTS MEET OUR EXPECTATIONS?
Considering the large variety of lakes examined, and
considering also the unavoidable inequality in the data
collected, the results achieved to date probably exceed
by far what could be expected from this program.
Admittedly, some of the correlations of factors thought
to be interrelated, in part, were found to be poor, yet at
least some of the more important correlations turned
out to be highly significant (cf. relationship between
phosphorus load and in-lake phosphorus concentra-
tions, between this latter and chlorophyll, and between
loading and chlorophyll).
Generally speaking, what has been achieved in
terms of understanding lake behavior, lies say, half way
between the historic position that each lake is an entity
which has to be understood on its own, and solely on
its own, and an advanced but not yet attained level of
insight which would make it possible to deduce the
reaction of bodies of water with a high degree of
precision from a few parameters.
The program, seen in its totality, has provided a
unique opportunity to study limnology in a comparative
sense. In this respect, it can be considered as a
milestone in national and international cooperation,
the prospects of which are manyfold and leading into
the direction of what Elster outlined as the future of
limnology in his memorable 1956 conference (cf. Elster
1958).
However, the program would have failed if it had not
also provided the basis from which it is possible to
establish improved loading criteria for practically
combating eutrophication of lakes. A synthesis of such
criteria is given in Figure 8. These criteria are in logical
sequence of the criteria proposed in previous papers by
-------
OPENING SESSION
35
Vollenweider (1968, 1975, 1976), linking average
inflow concentrations for phosphorus with expected
average lake concentration and average chlorophyll
concentrations as a function of the flushing regime of
lakes. Division between the main trophic categories is
based on the 50 percent probability of belonging to the
indicated .class, and the vertical arrows may be read in
the sense of "belonging to or better as" the indicated
class. With this, management has a tool to establish
whatever goal is thought to be desirable to reach, or
conversely, to anticipate the level of improvement
which can be expected from an established reduction
program. How this should be done in practice, and with
what level of uncertainty one has to reckon with, is
discussed in the Synthesis Report.
Besides this positive note, however, it must be
underlined that many questions remain open, and that
a blind and uncritical application of the OECO results
can lead to gross error. Limnology, and its application
for practical purposes, was and is a complex science,
and remains a matter of skill and experience. The
establishment of group behavior of lakes, as was the
main objective of the OECD Program, does not
necessarily mean that each single case can be
subordinated to one single rule.
Indeed, a more detailed elaboration of the OECD data
— a work which still requires considerable time —
already indicates that a more selective grouping of
lakes having similar limnological properties would
reduce some of the uncertainties resulting from an
indiscriminate pooling of all data. From here on, one
has to find out what the discriminative parameters are
for group differences. Factors which lend themselves
for further consideration are: type and length of
stratification, epi- hypolimnetic ratio, mixing depth, ice
coverage, humic substances, N/P ratio, zooplankton
and fish population, etc.
An improved approach to discrimination analysis of
trophic conditions of lakes is underway by Chapra and
Reckhow (1979) who try to avoid some of the pitfalls of
the hitherto used prediction models by applying the
uncertainty theory. Schaffner and Oglesby (1978) and
Oglesby and Schaffner (1978) introduce in their
modifications some of the factors mentioned.
Last but not least, the next step in the endeavour will
be a concerted effort to link experimental with
theoretical limnology. Over the last decade or so,
theoretical limnologists have made much progress and
brought into the open many of the uncertainties in our
understanding. This throws the ball back to experi-
mental limnologists who will have to rethink many of
their programs. The extended experimental work of
Schindler and his colleagues in the Experimental Lakes
Area studies — which cannot be referred to in detail in
this review — provides further guides to understanding
the complex relationship between nutrient loading and
lake reaction.
1000
T(w) Years
1000
Figure 8. — Synthesis of the OECD information: Group relationships between average
inflow concentrations and average lake concentrations of phosphorus, average yearly
chlorophyll concentrations, and trophic categories relative to the water residence time of lakes.
-------
36 RESTORATION OF LAKES AND INLAND WATERS
REFERENCES
OECD Eutrophciation Program Regional Reports:
— Alpine Lakes, prepared by Hj. Fricker, Swiss Federal
Board for Environmental Protection (Bundesamt fur Um-
weltschutz), CH-3003 Bern, Switzerland. 1980.
— The Nordic Project, prepared by S-O. Ryding. Nordforsk.
Nordic Cooperative Organization for Applied Research.
Secretariat of Environmental Sciences. Folkskolegatan 10A,
SF-00100 Helsingfors 10, Finland. 1980.
— Shallow Lakes And Reservoirs, prepared by J. Clasen.
The Water Research Centre, Medmenham Laboratory, P.O.
Box 16, Medmenham, Marlow, Bucks., England. 1980.
— Summary Analysis of the North American OECD Project
(U.S. Portion), prepared by W. Rast and F. Lee. U.S. EPA-
600/3-78-008. Ecol. Res. Ser. 1978.
— A Test Case Study of the OECD Program on
Eutrophication (Canadian Portion), prepared by R. A.
Vollenweider and L. Janus, (in preparation). IWD-National
Water Research Institute, CCIW, Burlington, Ontario. 1981.
OECD Eutrophication Program
— Synthesis Report, prepared by R. A. Vollenweider and J.
Kerekes, and members of the Technical Bureau. OECD
Secretariat, Environment Directorate, 2, rue Andre Pascal,
75775 Paris Cedex 16, France. 1981.
— Eutrophication Control. Conclusions of the OECD
Cooperative Program on Eutrophication, prepared by R. A.
Vollenweider and the members of the Technical Bureau.
(published in UNESCO Nature and Resources 16, 3, 1980).
ACKNOWLEDGEMENTS
The OECD Cooperative Program on Eutrophication would
not have been possible without the efforts and generous
contributions made by all collaborators of this program. It is
impossible to list names individually. However, as principal
author of this paper and Chairman of the Technical Bureau,
I wish to express my thanks and those of the members of the
Technical Bureau to all colleagues, advisers, helpers,
governmental and other agencies, who made this unique
collaborative program possible.
-------
37
PRESENT KNOWLEDGE OF LIMITING NUTRIENTS
CURT FORSBERG
Institute of Limnology
University of Uppsala
Uppsala, Sweden
ABSTRACT
To develop an effective control program, it is important to know which factor limits maximum
biomass. Because peak algal mass often appears during a short period of time, studying the
standing crop limiting nutrient is probably more convenient than developing a water sampling
program to cover irregular peak situations. Nitrogen, phosphorus, and chlorophyll a are discussed.
For developing effective control programs it is
important to know which factor controls or limits the
maximum biomass developed during, for example, the
summer period. Analyzing the limiting nutrients during
a limited period of time, e.g., 1,2, or 3 weeks, when the
phytoplankton development is rapid and the absolute
concentrations of available nutrients change rapidly,
can be looked upon as a study of production or
productivity limiting nutrients. The evaluation of the
limiting roles of, for example, nitrogen and phosphorus,
is then based on the amounts of the available forms,
Nos-Ni, NhU-N, and PO4-P. If these are present in
excess they can't limit the biomass development. To be
considered limiting the concentrations have to be close
to zero.
As peak algal mass often appears during a short
period of time it can be difficult to have a water
sampling program covering just this situation. Peaks
can also appear irregularly from year to year,
depending on climatic conditions. Therefore, methods
analyzing limiting nutrients during a short and
chemically-biologically intensive period of time may be
difficult to use for lake water management. Included
here must also be the problems with "luxury"
consumption. Algae can, as is well known, assimilate
and store P for later use. This means that in spite of
PO.4-P concentrations in the surrounding water being
close to zero, phosphorus may not be the limiting
nutrient.
For water management, it is probably more conven-
ient to study the standing crop limiting nutrient. This
concept can be developed by looking at nutrient and
biomass levels without taking special notice about
assimilable forms of nutrients or the physiological
processes developing this specific biomass level. The
amounts of total-N and total-P and the ratios of N to P
in relation to, for example, chlorophyll or transparency
can be studied. This approach is easier to handle for
water management.
Comprehensive results demonstrate linear correla-
tion between summer averages for total-P and
chlorophyll a up to a P concentration level correspond-
ing to 100 mg/m3. Below this concentration level, P
can be considering as limiting phytoplankton standing
crop. Above, nitrogen will take over in relation to
phosphorus. Nz-fixing blue-greens will often help a
nitrogen stressed situation, which means that nitrogen
has less chance to act as a limiting nutrient. As a guide
for indicating their roles, the following ratios of N to P
can be used:
Total-N
Total-P
>17
10 — 17
<10
Limiting
nutrient
P
N and/or P
N
Chlorophyll a
level, mg/m3
<20
20-70
>70
These values are based on summer average values.
As recent studies demonstrate very strong correlation
between summer average and summer maximum
values, reliable average values will also give good
information of the worst situation; this knowledge is
essential for water management and physical plan-
ning.
-------
38
NON-NUTRIENT FACTORS INFLUENCING
THE DYNAMICS OF EUTROPHICATION
D. M. IMBODEN
Federal Institute for Water Resources and Water Pollution Control (EAWAG)
Dubendorf, Switzerland
ABSTRACT
Annual primary production — chosen as trophic state index of a lake — is determined by the
following non-nutrient factors: Morphology, hydraulics, meteorology, internal mixing, non-
nutrient water chemistry, and input of inorganic particles. As a consequence of various
anthropogenic activities, in many lakes some of these factors are influenced by man. Internal
mixing processes are considered the key to understanding natural and artificial non-nutrient
eutrophication factors since they control the vertical transport of nutrients from the sediments into
the hypolimnion and across the thermocline of lakes. During the stratification period, vertical
mixing — although slow but quasi-continuous — may cause an enormous internal loading to the
trophic zone of shallow or medium deep eutrophic lakes. In winter, single meteorological events
can be responsible for the intensity of deep water renewal and thus for the chemical dynamics in
the hypolimnion during the subsequent stagnation period. Minor physical perturbations can
significantly change an existing mixing pattern and thus the trophic evolution of the lake. On the
other hand, physical alterations may intentionally be combined with external measures to control
lake eutrophication. However, this will not be unproblematical unless a better understanding of
the relation between external physical forces and internal mixing processes has been achieved.
Annual primary production — chosen to measure the
dynamics of eutrophication in lakes— is determined by
various external and internal factors among which the
input of nutrients has been identified as the most
important cause of eutrophication. However, because
of non-nutrient factors lakes vary greatly in their
response to increasing nutrient input. Non-nutrient
factors are:
1. Morphology (lake size, mean depth, shape of lake
basin)
2. Hydraulics (water residence time, type of inlets
and outlets)
3. Meteorology (solar radiation, temperature, wind)
4. Internal mixing
5. Non-nutrient water chemistry
6. Input of particles and sedimentation patterns.
Analyzing lakes from different climatic zones of the
earth, Brylinsky and Mann (1973) have found that the
input of solar radiation is the most important factor
regulating primary productivity of a lake. For the more
homogeneous climate zones of Central Europe and
North America, where the problem of lake eutrophica-
tion is most urgent, nutrient input represents the
dominant influence on trophic state.
As a consequence of various anthropogenic activ-
ities, in many lakes not only nutrient input, but also
some of the non-nutrient factors, are (intentionally or
inadvertently) influenced by man. This is mainly the
case with respect to hydraulic properties of lakes (for
instance, flood control by diverting rivers through
natural lakes or by dams) whereas morphology and
meterology are still beyond man's technical ability (at
least for larger lakes). Non-nutrient water chemistry,
especially the input of salt, may influence the internal
mixing properties of a lake by affecting density. Particle
loading affects the transparency of the water column
giving rise to a different structure of primary production
and vertical temperature distribution because of
different adsorption of solar radiation.
Most of these phenomena affect directly or indirectly
the internal mixing processes which are considered to
be the key to understanding natural and artificial non-
nutrient eutrophication factors. Internal mixing con-
trols the vertical transport of nutrients from the
sediments into the hypolimnion and across the
thermocline (internal loading). A growing tendency
exists for man to use lakes in ways which change their
mixing pattern. Examples are the use of lakes for
hydropower and irrigation, the input of waste heat, the
export of heat for heat pumps, and the use of natural
lakes for pumped storage power operation (Imboden
1979, 1980). Physical alterations may also intention-
ally be applied in combination with external measures
to control lake eutrophication (artificial mixing, hypo-
limnion drainage).
Recently, Imboden and Gachter (1979) have ex-
tensively discussed the impact of physical processes on
the dynamics of eutrophication. Since in most lakes
phosphorus has been found to be the controlling input,
they analyze the relationship between annual P-
loading (Lp) and primary productivity (z: P) to identify
factors other than P-loading which influence pro-
ductivity.
The data used for this analysis (Figure 1) originates
from lakes of the relatively homogeneous climate of
Europe and North America where the dominant
influence of solar radiation mentioned earlier is less
important.
Another factor has been brought forth by Vollen-
weider (1968), who found an increase of nutrient
-------
FACTORS INFLUENCING THE DYNAMICS OF EUTROPHICATION
39
tolerance with increasing lake depth. Indeed, the data
in Figure 1 show a slightly higher productivity for
shallow lakes, a tendency mainly associated with
eutrophic lakes. A multidimensional regression analy-
sis results in the equation (z: mean depth)
log IP = 2.6 - 0.24 log z + 0.66 log LP
eq. 1
(correlation coefficient: 0.76)
but again the scattering is far too large to be acceptable
for a unique theory of primary productivity.
The existence of an LP-dependent lower limit for IP,
a salient feature of Figure 1, suggests that IP be
divided into the two components
IP = IP0 + IP1 eq. 2
IP is the productivity in a hypothetical isolated
system having the size of the trophogenic layer, i.e., a
system which is neither in contact with the sediments
nor influenced by turbulent nutrient transport across
the thermocline. It can be approximated by a simple
steady-state one-box model for primary production (see
Imboden and Gachter (1979) for details).
P1 is the contribution from internal nutrient
recirculation. It is this part of IP which is expected to
be sensitive to the mixing regime of the lake, its
morphometry and redox conditions at the sediment-
water interface.
(gCnfV1)
1000 r-
too
EP=200(LP)
0.70
o
ac.
a.
a:
oc.
a.
10
0.01 0.1 1 10
PHOSPHORUS LOADING LP
Figure 1 . — Correlation between annual P-loading per surface
area(LP)and primary productivity(IP)for various European and
North American lakes. Shaded area: Results from steady-state
productivity model for epilimnic residence times TE of water
between 0.06^ yr (curve A) and 1 5 yr (curve B). Shallow lakes
(mean depth z<10 m) show a slight tendency toward higher
productivities (from Imboden and Gachter, 1979).
Imboden and Gachter summarize their analysis of
primary productivity data as follows:
1. Productivity mainly depends on P-input, but there
exists no unique connection between IP and Lp neither
in the measured data nor from theoretical considera-
tions.
2. The minimum productivity IP0 is reasonably
approximated by a simple steady-state one-box model.
3. At high P-loading, the same model also reproduces
the saturation effect on P around 400 g C m'2 yr ]
4. The "internal fraction" IP1, i.e., the difference
between the measured productivities and the cor-
responding IP0, is largest for medium LP values
(around 1 g P m"2 yr"1).
5. P' exhibits a tendency to be larger for shallow
lakes, indicating the role of internal mixing. However,
mean depth alone cannot explain the magnitude of P1;
probably other factors such as morphometry of the
lake, exposure to wind, vertical mixing intensity and the
redox conditions at the sediment-water interface are of
equal importance.
6. As exemplified by the curves A and B in Figure 1,
hydraulic loading is only of limited influence on P
The statistical approach provides some basis for
speculation on the possible influence of vertical mixing
on the trophic state, but our present knowledge of lake
mixing mechanisms does not permit quantifying these
effects in a general way. The mechanisms of how
kinetic energy, entering the lake by sheer forces of
wind stress and by rivers, is transformed into
turbulence and finally dissipated into heat by viscosity
is not fully understood. Lakes can be even less
classified by simple schemes with respect to their
physical characteristics and biological phenomena.
At this time, case studies are most suitable to reveal
the physical processes which may interfere with
trophic conditions. In the highly eutrophic Greifensee
(Switzerland), Imboden and Emerson (1978) have
determined the vertical eddy diffusivity from the
distribution of radon-222, a natural radioactive isotope,
in the water column. Together with measured vertical
phosphorus gradients, they estimate internal loading in
this lake to be between 60 (June to September) and
100 percent (October to November) relative to external
P input.
The following general statements can be derived
from the case study of Greifensee:
1. Mineralization of organic material in the hypo-
limnion and at the sediment-water interface during
stagnation leads to vertical concentration differences
between epilimnion and hypolimnion. Qualitatively, the
gradients are inversely related to the hypolimnic/
epilimnic volume ratio, i.e., shallow lakes generally
have larger gradients.
2. In the case of phosphate, the mean depth
dependency of vertical gradients is further enhanced by
the factor of hypolimnic 02 depletion. Under anaerobic
conditions released phosphate at the sediment surface
is not bound by surface absorption mechanisms. The
occurrence of anaerobic conditions is more probable in
shallow lakes since, given a certain production of
biomass at the surface, it is the size of the hypolimnic
oxygen reserves (i.e., of the hypolimnic volume) which
determines at what time of the summer anaerobic
conditions occur.
3. As a consequence of (1) and (2), the trophic level of
shallow lakes should show a larger sensitivity with
respect to external nutrient input than deep lakes,
which is in accordance with the early findings by
Vollenweider (1968).
4. One effect partially counteracting these above
statements consists in the hidden correlation between
lake mean depth and surface area. Large lakes (which
-------
40
RESTORAHON OF LAKES AND INLAND WATERS
are often the deeper ones) generally have a higher
vertical mixing intensity which would favor transport
from the sediments through the thermocline into the
trophogenic layer. However, this effect seems to be
less pronounced than the other ones.
As mentioned before, the influence of the mixing
processes on primary productivity cannot, in general,
be quantified. But it is possible, at least, to predict
whether productivity would increase or decrease with
vertical mixing intensity.
In Table 1, a summary of the relevant mechanisms
and their sensitivity with respect to vertical mixing is
given. In most cases, only a careful weighing of
opposite effects is necessary to predict the behavior of
the system, a task possible only by using numerical
lake mixing models. A treatment of these models lies
beyond the intention and possibility of this contribu-
tion.
Imboden, D. M., and S. Emerson. 1978. Natural radon and
phosphorus as limnologic tracers: horizontal and vertical
eddy diffusion in Greifensee. Limnol. Oceanogr. 23:77.
Imboden, D. M.,and R. Gachter. 1978. A dynamic lake model
for trophic state prediction. Ecol. Model. 4:77.
1979. The impact of physical processes on the
trophic state of a lake. Pages 93-110 in O. Ravera, ed.
Biological aspects of freshwater pollution. Comm. Eur.
Commun. Pergamon Press, London.
Vollenweider, R. A. 1968. Water management research. Rep.
68.27. Organ. Econ. Coop. Develop. Paris.
Table 1. — A summary of mixing processes and their influence on
primary productivity.
Under the influence of the following changes, primary productivity
would
Increase Decrease
A. Deepening of the thermocline
Higher pool of nutrients
Lower rate constant for pro-
ductivity due to lower tem-
perature
Less favorable ratio between
zone of respiration and pro-
duction
B. Increase of vertical mixing in the thermocline and the hypolimnion
Increase of internal nu-
trient loading
Decrease of nutrient flux
from the sediments since the
sediment surface remains
aerobic during a longer period
Decrease of biomass density
by dilution due to mixing
C Increase of the probability that the lake undergoes total turn-over
during winter (of importance only for meromictic lakes)
Recycling of hypolimnic nu-
trient pool leading to lar-
ger initial concentrations
in spring
Decrease of sediment boundary
flux due to higher hypolimnic
oxygen concentrations
Turbulence decreases nu-
trient retention (lower se-
dimentation velocity and/or
resuspension of sediments)
REFERENCES
Brylinsky, M., and K. H. Mann. 1973. An analysis of factors
governing productivity in lakes and reservoirs. Limnol.
Oceanogr. 18:1.
Imboden, D. M. 1979. Modelling of vertical temperature
distribution and its implication on biological processes in
lakes. Pages 545-560 in S.E. Jorgensen, ed. State-of-the-
art of ecological modelling. Int. Soc. Ecol. Model.
In press. The impact of pumped storage operation
on the vertical temperature structure in a deep lake: a
mathematical model. In J. P. Clugston, ed. Proc. 5th Pumped
Storage Workshop, Clemson, S. C. May 1979.
-------
41
DYNAMICS OF NUTRIENT ENRICHMENT IN LARGE
LAKES: THE LAKE MICHIGAN CASE
CLAIRE L SCHELSKE
Great Lakes Research Division
University of Michigan
Ann Arbor, Michigan
ABSTRACT
Lake Michigan is an interesting case to consider in discussing the dynamics of eutrophication.
Although many environmental changes indicate that accelerated eutrophication has occurred, the
main body of the lake is primarily mesotrophic. For example, open lake chlorophyll a
concentrations do not exceed 3 to4/ug/liter during the spring bloom and concentrations on a lake-
wide basis average 8 fjg P/liter. Results of nutrient enrichment experiments with natural
phytoplankton assemblages show that phytoplankton growth can be increased with small
phosphorus additions and that effects of phosphorus are greater when water is enriched
simultaneously with phosphorus and trace constituents (EOTA, trace metals, and vitamins). Effects
of eutrophication are manifested to the greatest extent in nearshore areas where, if localized
nutrient sources are adequate, phytoplankton standing crops may be many times greater than in
the offshore waters. Data on nutrient inputs show that the effects should be localized because
inputs are not uniform over the lake basin.
INTRODUCTION
The dynamics of eutrophication in large lakes must
be considered from a different perspective than that for
small lakes. Data for Lake Michigan will be used to
illustrate the point that size is an important considera-
tion. Lake Michigan is a large lake, having a water
surface area of 56,500 km2 and a water volume of
4,800 km3 (Table 1). The main axis runs north-south
from 42-46°N so one would expect latitudinal
influences on lake processes. The main outflow is to
the north through the Straits of Mackinac and the main
sources for nutrient loading are in the southern part of
the lake. This means, therefore, that nutrients must be
transported from south to north to be removed from the
lake with the outflow.
The lake also may be divided into a nearshore zone
and an offshore zone. The water quality of the
nearshore zone, as will be shown in this paper, is very
distinct from that of the offshore waters. The nearshore
zone differs from the offshore zone in that it receives
higher nutrient loading from tributaries. In addition,
physical processes in the nearshore are distinct from
the offshore, currents are stronger, and effects of
waves and currents, particularly relative to interactions
with the sediments, are greater.
In this paper the nearshore zone has been defined
arbitrarily as that area lying within the 30-meter
contour line. This is roughly the average coastal or
nearshore zone suggested by Mortimer (1975) who
stated that the nearshore strip, about 10 kilometers
wide, is the "scene of the main transfer of energy from
wind to total basin motion and contains the greater part
of the lake's kinetic energy."
The need to examine the dynamics of large lakes
from a different perspective than that for smaller lakes
Table 1. — Comparison of nearshore and offshore
morphometric characteristics of Lake Michigan, excluding
Green Bay.
Depth range (m)
Water surface (km2)
Water surface (%)
Volume (km3)
Volume (%)
Mean depth (m)
Nearshore
0-30
11300
20
220
4.6
19.1
Offshore
30-275
45200
80
4580
95.5
101 '
Lake
0-275
56500
4800
85
is to a great extent a function of physical factors of
these large systems. Lake Michigan has a long
residence time: the volume divided by the outflow is
approximately 100 years. It has a mixing time of about
180 days (Boyce, 1974) which is long relative to time
scales for phytoplankton growth. Hypsographic rela-
tionships show that 20 percent of the area but only 4 or
5 percent of the volume is contained within the 30-
meter contour (Table 1). Finally, because of the large
size and thus, great variations in depth, the water
surface warms differentially. The shallower nearshore
waters warm more rapidly in the spring causing .a
thermal bar to form that separates the nearshore from
the offshore (Mortimer, 1975). The thermal bar
produces sharp nutrient gradients and has a pronounc-
ed effect on the distribution and abundance of
phytoplankton (Stoermer, et al. 1968; Davis, et al. In
press).
In this paper I will show that the nearshore zone
receives much greater phosphorous loads, produces
greater standing crops of algae, and contains different
species composition of phytoplankton than the offshore
waters. The nearshore phytoplankton differences can
-------
42
RESTORATION OF LAKES AND INLAND WATERS
be attributed to the combined effects of anthropogenic
materials, including phosphorus. In total these results
are significant for lake protection because the major
uses of water are in the nearshore, whereas most
research is directed at offshore problems.
PHOSPHORUS LOADING
The four major sources of phosphorus loading are
tributary flows, atmospheric inputs, direct municipal
discharges, and shoreline erosion (Table 2). Estimated
total and source inputs by different investigators vary
considerably; however, given the size of this particular
lake and its drainage basin and the limited study on
inputs these variations are not unexpected. All
estimates agree in that the tributary loading is the
greatest. Large discrepancies exist for estimates of
shoreline erosion, ranging from 1.35 to 3.7 x 106 kg/yr,
and for atmospheric inputs, ranging from 1.0 to 1.69 x
106 kg/yr. The main reason for the larger number of
1.69 106 kg/yr (Eisenreich, et al. 1977) for atmospheric
inputs is that dry fallout was not included in the
estimated input of 1.0 x 106 kg/yr (Murphy and Doskey,
1976). Absolute values for different sources are not
critical for the points emphasized in this paper.
In addition to uncertainties about the magnitude of
the loads there is, of course, the well-recognized
problem of the proportion of any given load that is
available to phytoplankton for growth. No attempt will
be made in this paper to address this because unequal
load distribution can be shown without addressing the
question of availability.
Disproportionate loading of phosphorus from trib-
utaries to different shoreline zones has been recogniz-
ed for some time. Schlelske (1975) pointed out that as
much as40 percent of the tributary phosphorus loading
to Lake Michigan could be attributed to inputs of the
Grand, St. Joseph and Kalamazoo rivers. All are located
within 120 kilometers of shoreline on the southeastern
part of the lake. More recently Sonzogni, et al. (1978)
reported that the same tributaries in 1975 and 1976
supplied 48 and 46 percent of the total tributary
phosphorus loading to Lake Michigan including Green
Bay (Figure 1). Roughly half of the tributary phosphorus
loading therefore was concentrated within a 120-
kilometer length of shoreline; nearly 25 percent of the
total came from one tributary, the Grand River.
Table 2. — Sources of total phosphorus input to Lake
Michigan for three time periods. All inputs are 106 kg/yr.
Direct industrial discharge
Direct municipal discharge
Tributary inputs
Municipal point
Industrial point
Shoreline erosion
Atmospheric inputs
19741
0.05
1.09
4.97
1.35
1.00
8.46
1975'
0.06
1.07
4.23
3.7
1.69
10.75
1975-762
3.39
(1.04)
(0.22)
Loadings from direct municipal discharges to the lake
are also disproportionate. This is readily illustrated by
some reports that as much as half of the total
phosphorus loading from direct municipal discharges
originates from the City of Milwaukee. That direct
municipal discharge of approximately 500 metric tons
is roughly half the municipal discharges to tributaries
of 1,191 metric tons/yr (Sonzogni, et al. 1978). It is
also a phosphorus load equivalent to that from the Fox
River and larger than any tributary load other than the
Grand River.
1Eisenreich, et al. 1977.
^averages of data for 1975 and 1976 are from Sonzogni, et al. 1978.
Figure 1. — Magnitude of total phosphorus tributary loadings to
Lake Michigan. Data in metric tons/yr from Sonzogni et al.
1978.
Atmospheric loadings also are not uniform over the
lake basin, although the variation is not as great as that
for tributary loading or direct municipal discharges.
According to Eisenreich, et al. (1977) atmospheric
loading to the southern basin is roughly twice as large
as to the northern basin. Although atmospheric
loadings comprise a relatively large part of the total
phosphorus loading to the lake, as much as 20 percent
by some estimates (Table 2), the relative effect in the
lake probably differs from that of either municipal or
tributary inputs. Atmospheric inputs are distributed
over the entire lake surface whereas loadings from
tributaries and municipal sources are primarily to the
nearshore zone.
Nearshore phosphorus inputs are distinguished
functionally from atmospheric inputs in that phos-
phorus loaded to the nearshore zone must be
transported through the nearshore waters prior to
being mixed with the offshore waters of the lake.
Phosphorus transported through the nearshore zone is
acted on by biological processes before it is diluted with
the open lake waters, whereas most of the atmospheric
input is transported directly to the open lake.
-------
FACTORS INFLUENCING THE DYNAMICS OF .EUTROPHICATION
43
Using, (data on phosphorus loading (Table 2) and
HiQisphometrie data (fable 1) I have calculated area I and
volumetric phpsphorus loads to the nearshore and
offshore zones. These data clearly show that loading in
the nearshore zone is disproportionate to that in the
offshore waters. Tributary loadings are 20 times
greater in the nearshore than in the offshore zone on a
volumetric basis and 5 times greater on an areal basis
(Figure 2).
It should be obvious that these average loads do not
reflect the absolute range in loadings that occur within
the lake. For example, because 50 percent of the
tributary loadings result from the three tributaries in
the southeastern part of the lake these loads represent
greater than average nearshore loading (Figure 1).
Likewise, some nearshore areas in the northern part of
the lake receive relatively small phosphorus loads,
resulting in smaller than average loading.
In summary, of the three areas where phosphorus
loading to the nearshore zone is greater than the
600
500
*r_ 400
300
200
100
531
513
152
18
18
ALL SOURCES
TRIBUTARY
ATMOSPHERIC
40
35
25
20
15
10
5
35.6
1.79
1 15
NEARSHORE
WHOLE LAKE
Figure 2. — Areal (upper) and volumetric (lower) total
phosphorus loading to Lake Michigan. Loads are calculated for
the nearshore zone and whole lake. Nearshore is set, at 25
percent of the surface area. See Table T for other morphometric
data.
average, only two are in the main body of Lake
Michigan: (1) The 120-kilometer length of shoreline in
the south where the Grand, St. Joseph, and Kalamazoo
rivers drain to the southeastern part of the lake; (2) the
area on the west shore directly across the lake from the
Grand River which receive municipal input from the
city of Milwaukee; (3) the southern end of Green Bay
where the Fox River is the major source of input.
Nutrient effects on biological processes in this area are
most evident in the southern part of Green Bay and
become less evident with distance north from the
mouth of Fox River.
BIOLOGICAL CONSEQUENCES OF
NEARSHORE LOADING
Given the disproportionate loadings to the nearshore
zone, the biological manifestations of nutrient enrich-
ment are most pronounced in this area of the lake.
Chlorophyll standing crops are several times larger on
the average in the nearshore zone than in the offshore
zone (Table 3), both along the southeastern shoreline
where tributary inputs are the greatest, and also on the
southwestern shoreline where tributary inputs are not
as large a factor (Figure 3).
Not only is there a difference in standing crop but
there is also a pronounced difference in species
composition between the nearshore zone and the
offshore zone. A number of'species characteristic of
highly enriched or polluted waters have been found in
enriched nearshore zones (Stoermer and Yang, 1970).
It has been found that species of Melosira tolerant of
nutrient enrichment dominated the nearshore zone
and were replaced by species less tolerant of
enrichment in offshore waters (Holland, 1968).
Recently we have completed studies on the
distribution of nutrients and their relationships to
species composition and standing crop of phytoplank-
ton in the nearshore zone (Schelske, et al. 1980). These
studies showed that, as expected, standing crops of
phytoplankton were greater and species composition
different in nearshore waters than in offshore waters.
In addition, we showed that the nutrient input from
rivers influenced the biological characteristics over a
considerable distance offshore from the mouth of the
tributary. In areas affected by tributary inputs phyto-
plankton composition was dominated by species
supplied with the tributary inflow and varied from river
to river and with the season of the year. Inshore-
offshore differences were less pronounced in the
northern part of the lake where tributary phosphorus
loading (Figure 4) was relatively small compared to the
southern basin.
Phytoplankton species composition in tributary
inputs obviously differed front that in nearshore waters
so that the major dominants in the tributaries could be
used as biological tracers for river inputs (Schelske, et
al. 1980), The plume of the Grand River with its
characteristic phytoplankton, for example, extended
1.6 kilometers offshore, but beyond this point the
nearshore phytoplankton were characteristic of en-
riched areas such as the transects offshore from
Milwaukee and the Kalamazoo River where there was
no large tributary input. Based on phytoplankton
species composition and chlorophyll standing crops,
-------
44
RESTORATION OF LAKES AND INLAND WATERS
the nearshore zone extended offshore at least 6.4
kilometers and on some transects 13 kilometers
offshore.
Table 3. — Comparison of chlorophyll and phosphorus
concentrations in nearshore and offshore waters of Lake
Michigan. Values for the lake are based on volume-weighted
averages for the nearshore and offshore waters. See Table 1
for volumetric data.
Nearshore
Average total phosphorus1
(fjg P/liter) 15.2
Average chlorophyll1
(//g/liter) 4.7
Maximum spring chlorophyll'
(//g/liter) 14.0
Average spring chlorophyll2
dug/liter) 12.0
Offshore Lake
8.1 8.4
2.2 2.3
4.4
2.3
'Rousar (1973). Nearshore, 4.8km from Milwaukee.
2Ladewski and Stoermer (1971). See Fig. 3.
May
July
Sept
Figure 3. — Chlorophyll concentration in 1971 averaged by
depth range and month. Key: Dotted line shows meanvaluefor
stations between 10 m and 40 m deep and solid line shows
mean value for stations deeper than 40 m. For each cruise
there are nominally 12 stations shallower than 10 m, 16
between 10 and 40 m deep and 13 deeper than 40 m. Error
flags show the standard error of the mean. (Ladewski and
Stoermer, 1973).
Greater standing crops of phytoplankton in the
nearshore zone on first analysis appear to be directly
attributable to tributary nutrient loading, particularly
phosphorus. Effects of enrichment, however, appear to
ring at least the southern basin of Lake Michigan, and
therefore the effect may not be due only to tributary
loading because it has already been shown that
tributary nutrient loading on the western shore is less
than that for the eastern shore of the southern basin
(Figure 1). It has also been reported that chlorophyll
standing crops in the nearshore zone off Milwaukee
were several times greater than those in the offshore
waters, a difference attributed to municipal phos-
phorus discharges at Milwaukee (Rousar, 1973).
Further analysis of inshore-offshore differences are
caused not only by greater loading of nutrients to the
nearshore zone but also by physical factors and
biological and chemical processes that are not clearly
understood (Beeton and Edmondson, 1972).
One important factor in considering phosphorus
loadings to the nearshore is the time response of
phytoplankton relative to such loadings. Time re-
sponses for phytoplankton growth vary depending on
the physiological state of phytoplankon. If phyto-
plankton respond immediately without a lag phase,
effects of added nutrients might be expected to occur
within a 4 or 5-day period. If it is assumed that
phytoplankton cells divide at a rate of about one per day
or slightly less, then within this 4 or 5-day period
standing crops could increase by a factor of 10. If the
response lag to enrichment were 1 to 3 days this time
would be extended to a 4 to 8-day period. Given
average coastal currents in the nearshore environment
of .5 km/hr one would then expect the phosphorus to
be transported no more than 50 to 100 kilometers from
the source before it was used by phytoplankton in the
coastal zone. Since current reversals are frequent in
the nearshore zone one would not expect the affected
area to extend as far as 50 to 100 kilometers from the
source very frequently.
The preceding calculations consider only effects of
phosphorus on growth of phytoplankton in the
nearshore zone and neglect any recycling of phos-
phorus or transport of phosphorus in phytoplankton to
other areas where it can be recycled and used again for
phytoplankton growth. Considering the long time
constants for physical transport and mixing of
phosphorus throughout the lake basin relative to the
time constants for uptake and growth by phytoplank-
ton, phosphorus is either recycled many times through
the plankton community or carried within the lake by
mechanisms other than simple mixing and diffusion.
NUTRIENT ENRICHMENT
EXPERIMENTS
Experimental work with the effects of nutrients on
growth and species composition of natural phyto-
plankton can provide insight into why species
composition differs between the nearshore and the
offshore zones. These experiments show that growth
rates of offshore natural phytoplankton assemblages
can be increased by adding phosphorus alone and that
greater growth rates can be obtained if vitamins, trace
metals, and a chelating agent are combined with
-------
FACTORS INFLUENCING THE DYNAMICS OF EUTROPHICATION
45
phosphorus additions (Figure 4). This effect has been
substantiated in experiments with water from Lake
Michigan (Schelske, et al. 1974), Lake Superior
(Schelske, et al. 1972), and Lake Huron (Lin and
Schelske, 1979). The specific agent that caused greater
phytoplankton responses was identified in one factorial
experiment. Analysis of variance showed that the
chelating agent, EDTA, produced a statistically signifi-
cant main effect and that vitamin and trace metal
additions did not (Schelske, et al. 1978). Higher growth
rates with additions of vitamins, trace metals, and
chelating agents indicate that the system may be
"nutrient saturated" with this treatment. Nutrient
saturated will be used in the paper to describe the
experimental conditions under which maximum
growth rates were obtained even though this condition
has not been verified experimentally.
In our experiments phytoplankton are sensitive to
phosphorus enrichment and are phosphorus saturated
at a relatively low level, at concentrations ranging from
5 to 15 jug P/liter. The absolute concentration is not
important because this would undoubtedly change
depending on experimental conditions. However, it is
important to note that with the addition of trace
constituents and phosphorus, growth rate is no longer
limited at low phosphorus levels and increases with
increasing phosphorus concentrations (Figure 4).
The presence of trace constituents and phosphorus
increases not only the growth rate but also the yield or
standing crop (Figure 4). Because yield is ultimately
affected by availability and quantity of nutrients it
seems obvious that a greater yield would result from
1
10 20 30
Phoiphorul addition! (/JO P/llttr)
Figure 4. — Relationship between phosphorus additions with
and without trace constituents (TC) and growth rate in
doublings/day. Bar graphs indicate final yield for enrichments
of 5,10,20 and 40 fjg P, solid portion represents increases due
to trace constituents.
nutrient-saturated growth than from nutrient-limited
growth.
Perhaps the most important result of these experi-
ments is the demonstration that trace constituents
added with phosphorus change the species composi-
tion in the phytoplankton assemblage. Stoermer, etal.
(1978) showed that the phytoplankton species suc-
cession resulting from phosphorus additions alone
closely paralleled that which occurred in Grand
Traverse Bay from which the water had been obtained
originally. Fragilaria crotonensis was the major
dominant in Grand Traverse Bay during and after the
experiment. This species was also the major dominant
in laboratory experiments which received only phos-
phorus enrichments. Under nutrient-saturated growth
conditions resulting from trace constituent enrichment,
species succession was different. The major dominant
shifted to Stephanodiscus subtilis (Stoermer and Yang,
1970) along with other nutrient tolerant species of
Stephanodiscus, including S. tenuis and S. minuteus
(Schelske, et al 1980).
DISCUSSION
Data show that from the standpoint of eutrophication
processes large lakes the size of Lake Michigan should
be considered as two separate systems, a nearshore
and an offshore area. The offshore volume is much
larger than the nearshore so average conditions in the
lake are mainly determined by offshore properties even
though conditions are greatly different in the near-
shore, (Table 3). These data indicate that empirical
models of the Vollenweider type are adequate to
address the relationship between average standing
crop and chlorophyll concentrations for the open lake
water mass. Such models, however, were not designed
and should not be used to evaluate water quality in the
nearshore zone where conditions in time and space are
highly variable.
Disproportionate loading of phosphorus (Figure 2) is
obviously one of the factors contributing to greater
standing crops of phytoplankton in the nearshore
(Table 3). Phytoplankton assemblages characteristic of
enriched nearshore zones apparently cannot be
attributed only to phosphorus enrichment because data
have been obtained that show succession of species in
assemblages is affected by trace materials (EDTA,
vitamins, and trace metals) associated with algal
nutrition (Stoermer, et al. 1978; Lin and Schelske,
1979). In addition, the presence of these minor
constituents stimulate growth rates to greater levels
than those realized from phosphorus additions alone
(Figure 4). Anthropogenic inputs to the nearshore
would be expected to enrich this area with vitamins,
trace metals, and chemical compounds with chelating
capacities.
In large lakes, symptoms of nutrient enrichment or
eutrophication are first evident in the nearshore and
later may be present in the entire lake basin as was the
case with Lake Erie (Beeton and Edmondson, 1972). In
Lake Michigan enrichment effects presently are most
evident in the nearshore. Whether this ring around the
lake will or could eventually cover the lake surface as it
did in Lake Erie is open to conjecture. Greater nutrient
loads would be required in Lake Michigan to cause an
-------
46
RESTORATION OF LAKES AND INLAND WATERS
effect comparable to Lake Erie because nearshore
waters are a relatively smaller proportion of the total
water mass and Lake Michigan is deeper.
Managing phosphorus inputs to Lake Michigan is
still the most critical problem in controlling eutrophica-
tion. To date, little attention has been given to critical
nearshore effects. Instead, considerable effort has
been devoted to calculating total phosphorus budgets
which are then used in lake basin models. This
approach is necessary for determining phosphorus
contributions to downstream lakes and to evaluating
long-term trends in open lake water quality, but it is not
directly applicable to the problem of nearshore water
quality. Any efforts to model nearshore water quality
will be hampered by the lack of a long-term data base
from which the model could be verified.
To be useful a model of nearshore water quality
would have to incorporate several features of the
model that was developed for Saginaw Bay, including
limitation of phosphorus, silica, and nitrogen for
phytoplankton growth and capability to adjust phyto-
plankton forms to nutrient conditions (Bierman, et al. In
press). Although it is possible to model succession of
phytoplankton forms, i.e., from diatoms to non-diatoms
as the result of silica limitation or to nitrogen-fixing
blue-greens as the result of nitrogen depletion, there
are at present no experimental data on which to base
the more complex modeling that would be required for
models of succession in multi-species assemblages.
Natural phytoplankton assemblages should be used
to determine effects of nutrient enrichment in
oligotrophic or mesotrophic waters because the quality
of phytoplankton in many cases is as important as the
quantity. Effects of perturbations of any type on species
succession and dominance can be studied only with
natural assemblages or possibly with several cultured
species. Data presented in this paper indicate that Pma«
and other kinetic parameters may vary with the trace
constituents in natural water or with trace nutrient
conditions in artificial media. That trace constituents
(EDTA, vitamins, and trace metals) increased growth
rates markedly (Figure 4) points to the problem of
simulating natural conditions m artificial media. These
considerations also point to the need to determine a
true Aima, (maximum growth rate) that would occur
under nutrient saturated conditions in experiments.
Possibly this growth rate should be based on calculated
quantum photosynthetic yields which would represent
a true maximum and provide a basis on which other
rates could be compared. Because physical, chemical,
and biological components of natural systems are
dynamic, experiments to determine the responses of
natural phytoplankton assemblages must be conducted
frequently so the influence of these changing
conditions on biological processes can be evaluated
(Lin and Schelske, 1979).
REFERENCES
Beetori, A. M., and W T. Edmondson. 1 972. The eutrophica-
tion problem. Jour. Fish. Res. Board Can. 29:673.
Bierman, V. J Jr., et al. In press. A development and
calibration of a spatially simplified multiclass phytoplankton
model for Saginaw Bay, Lake Huron. Ecol Res Ser. U.S.
Environ Prot Agency, Duluth, Minn.
Boyce, F. M 1974. Some aspects of Great Lakes physics of
importance to biological and chemical processes Jour. Fish.
Res Board Can 31:689
Davis, C. 0., C. L. Schelske, and R. G. Kreis, Jr. In press.
Influences of spring nearshore thermal bar. Pages 140-1 bo
in C. L. Schelske, R. A. Moll, and M. S. Simmons.
Limnological conditions in southern Lake Huron, 1974 and
1975. Ecol. Res. Ser. U.S. Environ. Prot. Agency, Duluth,
Minn.
Eisenreich, S. J., P. J. Emmling, and A. M. Beeton. 1977.
Atmospheric loading of phosphorus and other chemicals to
Lake Michigan. Jour. Great Lakes Res. 3:291.
Holland, R. E. 1968. Correlation of Melosira species with
trophic conditions in Lake Michigan. Limnol. Oceanogr.
13:555.
Ladewski, T. B., and E. F. Stoermer. 1973. Water
transparency in southern Lake Michigan in 1971 and 1972.
Proc. 16th Conf. Great Lakes Res. 791. Int. Assoc. Great
Lakes Res.
Lin, C. K., and C. L. Schelske. 1979. Effects of nutrient
enrichments, light intensity and temperature on growth of
phytoplankton from Lake Huron. EPA-600/3-79-049. U.S.
Environ. Prot. Agency, Duluth, Minn.
Mortimer, C. H 1975. Physical characteristics of Lake
Michigan and its response to applied forces. Pages 1 -102 in
Environmental status of the Lake Michigan region. Vol. 2.
ANL/ES-40. Argonne Natl. Lab., Argonne, III.
Murphy, T. J., and P. V. Doskey. 1976. Inputs of phosphorus
from precipitation to Lake Michigan. Jour. Great Lakes Res.
2:60.
Rousar, D. C. 1973. Seasonal and spatial changes in primary
production and nutrients in Lake Michigan. Water Air Soil
Pollut. 2:497.
Schelske, C. L. 1 975. Silica and nitrate depletion as related to
rate of eutrophication in Lakes Michigan, Huron and
Superior. Pages 277-298 in A. D. Hasler, ed Coupling of
land and water systems. Sprmger-Verlag, New York.
Schelske, C. L., L E. Feldt, and M S. Simmons. 1980.
Phytoplankton and physical-chemical conditions in selected
rivers and the coastal zone of Lake Michigan, 1972. Univ.
Michigan, Great Lakes Res. Div. Publ. 19.
Schelske, C. L., E. D. Rothman, and M. S. Simmons. 1978.
Comparison of bioassay procedures for growth-limiting
nutrients in the Laurentian Great Lakes. Mitt. Int. Verein.
Limnol. 21:65.
Schelske, C. L. et al. 1972. Nutrient enrichment and its effect
on phytoplankton production and species composition in
Lake Superior. Proc. 15th Conf. Great Lakes Res 149. Int.
Assoc. Great Lakes Res.
1974. Responses of phosphorus limited Lake
Michigan phytoplankton to factorial enrichments with
nitrogen and phosphorus. Limnol. Oceanogr. 19:409.
Sonzogni, W. C., et al. 1978. United States Great Lakes
tributary loadings — study on Great Lakes pollution from
land use activities Submitted to U.S. Environ. Prot. Agency.
Stoermer, E. F. 1968. Nearshore phytoplankton populations
in the Grand Haven, Michigan vicinity during thermal bar
conditions Proc. 11th Conf. Great Lakes Res. 137. Int.
Assoc. Great Lakes Res
Stoermer, E. F., and J.J.Yang 1970 Distribution and relative
abundance of dominant plankton diatoms m Lake Michigan.
Great Lakes Res. Div. Publ. 16. University of Michigan.
Stoermer, E. F., B. G. Ladewski, and C. L. Schelske 1978.
Population responses of Lake Michigan phytoplankton to
nitrogen and phosphorus enrichment. Hydrobioloaia
57:249. a
ACKNOWLEDGEMENTS
Support for research discussed in this paper was obtained
from the Department of Energy (COO-2003-39), and from
the U S Environmental Protection Agency, Grant numbers
R-804503 and R-806294. The author wishes to acknowl-
edge Mark Haibach for his collaboration in obtaining the
unpublished data in Figure 4.
-------
47
MODELING THE RESPONSE OF THE NUISANCE ALGA,
CLADOPHORA GLOMERATA, TO REDUCTIONS IN
PHOSPHORUS LOADING
M. T. AUER
R. P. CANALE
Y. MATSUOKA
H. C. GRUNDLER
Department of Civil Engineering
The University of Michigan
Ann Arbor, Michigan
ABSTRACT
A mathematical model was developed to evaluate the impact of various phosphorus management
strategies on nuisance growths of the filamentous alga Cladophora glomerata. The model was
supported by intensive ecological studies and an extensive field monitoring program. The results of
simulating spatial and seasonal variation in algal biomass and associated nutrient parameters
agree well with field observations. The calibrated model is used to predict the response of the
system under study to a demonstration phosphorus removal program. Implications to large-scale
phosphorus management strategies are discussed.
INTRODUCTION
Cladophora glomerata, an attached filamentous
green alga(Chlorophyceae), is a recognized nuisance in
the littoral region of the Laurentian Great Lakes and
many smaller inland lakes. Nuisance growths of this
organism in small inland lakes are typified by those
observed in the Madison (Wisconsin) lakes, particularly
Lake Mendota. In the Great Lakes, water quality is most
severely impacted by this alga in Lakes Erie and
Ontario (Shear and Konasewich, 1975). In these lower
Great Lakes massive accumulations of rotting algal
material has resulted in closed beaches, decreased
lakeshore property values, and reduced utility of the
environment as a recreational resource. The prolifera-
tion of C. glomerata in Lakes Erie and Ontario is
thought to be related to lakewide nutrient enrichment
rather than simply point discharges of nutrients. Site
specific occurrences of the alga have been reported
from Lakes Huron, Michigan, and Superior (Niel and
Owen, 1964; Lin, 1977; and Herbst, 1969). The
offshore waters of these lakes cannot support
significant growth of Cladophora because of low
phosphorus levels.
The presence of abundant growths of C. glomerata
may indicate an overall or local reduction in water
quality. In small inland lakes septic tank drainage may
encourage local growth of plant material or may
elevate lakewide nutrient levels so that plant growth is
prolific throughout the littoral region. C. glomerata is
naturally present in rivers, streams, and many inland
lakes. Its increase to nuisance proportions, however,
reflects a serious perturbation of the quality of our
inland waters. As such, amelioration of nuisance
conditions by reducing phosphorus loading rates would
reflect well upon our commitment toward the lessening
of man's impact on the environment.
In 1978, the University of Michigan, in cooperation
with the EPA Large Lakes Research Station at Grosse
Me, Mich., began a 3-year program to examine the
potential for reducing nuisance growths of Cladophora
in the Great Lakes. The key function of the project is to
develop a mathematical model relating the production
of Cladophora- biomass to phosphorus loadings to the
Great Lakes. Such a model would be useful in
evaluating the impact of various phosphorus manage-
ment strategies in controlling Cladophora growth. The
model integrates available information on the alga as
well as indicating areas where new basic studies on
the ecology of the organism are warranted. These
topics serve as subjects for special investigations
designed directly to support the model. A field
monitoring program at a site known for nuisance
Cladophora growth provides data for calibration of the
model. Observations following a demonstration phos-
phorus removal program at the site verify the utility of
the model. Projections regarding the impact of various
phosphorus management strategies at this and other
sites may be examined by using the proper set of
loading rates and boundary conditions associated with
that location.
FIELD SITE
A field site with a known Cladophora problem was
selected from which to gather data on the growth of the
alga for use in calibrating the mathematical model. A
site was chosen which was perturbed by a single major
nutrient source, isolated from the complexities of
whole-lake growth forcing conditions, e.g., Lake Erie. A
-------
48
RESTORATION OF LAKES AND INLAND WATERS
site was selected at Harbor Beach, Mich, on Lake
Huron. The water quality of Lake Huron is such that
significant growth of C. glomerata cannot be supported
by offshore waters.
Harbor Beach has a population of approximately
2,000 and is an agricultural and light industrial
community. The Harbor Beach wastewater treatment
plant is a high-rate trickling filter operation with a
design flow of 350,000 gallons/day. The soluble
reactive phosphorus loading from the plant for 1979
was 2.163 metric tons (5.93 kg/day). The plant
discharges to Spring Creek, a small stream with good
upstream water quality. The wastewater treatment
plant phosphorus loading results in an annual mean
soluble reactive phosphorus concentration of approxi-
mately 700 /jgP/l at the mouth of Spring Creek.
Nuisance growths of C. glomerata occur in a
symmetrical pattern about the mouth of Spring Creek.
the region of perturbation extends 2 kilometers south
of the discharge and 0.5 kilometers north where the
area suitable for growth is truncated by the presence of
a dredged manmade harbor. The presence of extensive
cobble shoals (substrate required for attachment) and
quiescent bays (reduced mixing with offshore waters)
contribute to the potential for algal problems.
The isolation of the site from whole-lake or multiple
source nutrient perturbations is confirmed by chemical
and biological data. Dissolved phosphorus levels and
that stored in algal cells (internal phosphorus) decline
with increasing distance from the loading point,
eventually reaching background or boundary levels.
Cladophora biomass decreases as well, with no
observable growth at the station most removed from
the nutrient source (1.8 km). The relationships between
chemical and biological parameters and distance from
the nutrient source have been described in detail in an
earlier publication (Auer and Canale, 1980).
MODELING FORMAT
The mathematical model used in this project was
developed to fully use support available from the
specialty studies and monitoring program. The model is
composed of a fluid transport and a kinetic submodel.
The function of the former is to relate nutrient loading
and advective and dispersive transport so that the
distribution of dissolved phosphorus in the study area
may be predicted. The resultant soluble reactive
phosphorus concentrations become the forcing par-
ameter for nutrient uptake rates, a component of the
kinetic submodel. Equation 1 summarizes the primary
factors regulating the growth of Cladophora at the
study site on Lake Huron. These factors are the
components of the kinetic submodel.
A,
Eq. 1
fj = (i (fi(I) ' f2(T) ' f3(Q) * f-t(X) ) - R - S
where: fj : specific growth rate
fr : maximum specific growth rate
fi(l) function relating growth to light intensity
f2(T) function relating growth to temperature
fs(Q) :function relating growth to internal
phosphorus
f4(X) function relating growth to carrying
capacity
R : respiration rate
S :sloughing loss
TRANSPORT SUBMODEL
Spatial resolution for the model is achieved through
establishing completely mixed cells with flows reflect-
ing wind-driven and wave-induced current regimes.
The mcdel cells are oriented in two layers parallel to
the shoreline. Current regimes are calculated with
classical momentum equations which include the
effects on non-linearities, wind, bottom friction, and
wave action. Several intensive chloride grids as well as
weekly and daily chloride measurements at selected
stations were also combined with daily wind observa-
tions to gain an understanding of nearshore current
regimes.
Nutrient loading from the wastewater treatment
plant was monitored twice weekly. Soluble reactive
phosphorus and total phosphorus concentrations were
measured weekly at 25 nearshore stations represent-
ing offshore and longshore boundary conditions.
Sampling was conducted from ice-out through Novem-
ber, with loading measurements continuing through
the winter. The transport submodel then considers
loading data and current regimes as well as algal
uptake in establishing soluble reactive phosphorus
concentrations throughout the study area. These
values may be compared with weekly monitoring data
from lake stations to calibrate the model.
KINETICS SUBMODEL
The kinetics submodel considers each factor thought
to contribute importantly to the growth of Cladophora
glomerata at the study site on Lake Huron. Three
growth terms are considered: Light, temperature,
internal phosphorus level, and carrying capacity. The
latter is dynamically related to dissolved phosphorus
levels through nutrient uptake kinetics. Loss terms
include respiration and sloughing. Sloughing refers to
the separation of algal filaments from the substrate,
leading to shoreline deposition of algal material. The
model calculates the specific growth rate and
ultimately the algal biomass throughout the study site
by relating these factors in the fashion described
previously in Equation 1. It is useful to examine the
derivation and data base associated with the major
components of the kinetic submodel.
LIGHT
An experiment conducted at the BIOTRON facility at
the University of Wisconsin examined the relationship
between growth rate and light (and temperature).
Isolates of C. glomerata obtained from Lake Huron and
two small inland lakes were cultured over a matrix of
light and temperature levels in a crossed-gradients
room. Carbon uptake was measured over a range of
light intensities (60 to 1,000/jE/m2 sec) and the
resultant specific growth rate was calculated. These
measurements correspond to net photosynthesis
Gross photosynthesis was calculated by adding to the
net photosynthesis data a factor representing the
respiration rate as derived from the literature (Jackson
1966). The results of this experiment are presented in
Figure 1. These data show growth to be a hyperbolic
function of light intensity, with saturation at high light
levels. Independent measurements were made to
-------
FACTORS INFLUENCING THE DYNAMICS OF EUTROPHICATION
49
determine the maximum specific growth rate. The data
of Figure 1 were then modified to reflect this
information. The range of light intensities used in this
experiment are representative of those observed in the
field. All field and laboratory measurements recorded
photosynthetically-active radiation (PAR) using a
quantum meter.
% refers to phosphorus
removal efficiency
0.35 0 0.25 0.50 0.75 1.00
DISTANCE FROM NUTRIENT SOURCE (km)
Figure 1. — The relationship between specific growth rate
and light intensity for C. glomerata.
The light function may be used to calculate daily
specific growth rates at depth if input data in the form
of light at depth are provided. The level of light at
specific depths was calculated on a daily basis. A series
of measurements was made to establish a relationship
between the extinction coefficient for light and Secchi
disk transparency. Depths across the study area were
obtained through extensive mapping and daily water
level measurements. Daily estimates of incident solar
radiation were obtained from the literature and
corroborated on site. These estimates were used with
the extinction coefficients resulting from daily Secchi
disk readings to calculate light at depth. This value
served as input to the hyperbolic light function, through
which daily values for growth rate as a function of light
intensity could be calculated. The results of these
calculations indicated that for this site on Lake Huron,
light limits the growth of C. glomerata below a depth of
approximately 1.25 meters.
TEMPERATURE
Temperature has been considered an important
factor in regulating growth of this nuisance alga. Data
from the BIOTRON experiments were used to describe
the function relating growth rate to temperature.
Again, carbon uptake rates were measured at 2, 5,10,
15, 20, 25, 30, and 35°C. A value for gross
photosynthesis was calculated by adding the curve for
respiration as a function of temperature to the curve for
net photosynthesis. The results of this experiment are
presented in Figure 2. The growth rate was observed to
increase in an approximately linear fashion with
temperature to an optimum range of 20 to 30°C.
Severe inhibition was noted above 30°C. The im-
portance of temperature in the nearshore Lake Huron
environment is seen most clearly in the spring and fall
cold periods. Temperature inhibition is not observed as
water temperatures seldom exceed 23°C. Water
temperatures in the May to August growing season are
generally within the optimum temperature range of the
alga.
The actual calculations of growth rate as a function
of temperature are made possible through the
availability of daily temperature data at the study site.
Temperature also affects respiration. This relationship
is discussed in a later section.
Q.
3?
O 0.6
-------
50
RESTORATION OF LAKES AND INLAND WATERS
mentally to increase with increasing external phos-
phorus levels in the classic Michaelis-Menten fashion.
Saturation is reached at very high substrate levels
relative to that observed for phytoplankton. Increases in
internal phosphorus concentrations reduce phosphor-
us uptake rates through negative feedback. This
process is a form of enzyme inhibition and has been
discussed by Rhee (1978). A kinetic structure relating
the substrate saturation process and feedback inhibi-
tion has been derived experimentally for C. glomerata
and included in the model. These experiments will be
described in detail in a later publication.
The model may be calibrated by comparing output
with measured values for internal phosphorus levels
obtained from stations throughout the study area.
Internal phosphorus levels are measured at 28 stations
at weekly intervals throughout the growing season.
These data confirm the longshore symmetry of internal
phosphorus concentration and define phosphorus as
the element limiting growth in the shallow regions of
the nearshore. Very little variation in the level of
internal phosphorus was noted for distance offshore at
a given station. This indicates that light is the important
factor controlling growth in the offshore direction.
O.t 0.2 0.3 0.4 0.5 06
INTERNAL PHOSPHORUS, 0(%P)
0.7
Figure 3. — The relationship between relative specific
growth rate and internal phosphorus concentration for C.
glomerata.
CARRYING CAPACITY
A carrying capacity term has been included in the
model to reflect spatial limitations on the substrate as
well as self-shading effects. An empirically derived
value for the maximum attainable biomass (X ) of
600 gDW/m2 is employed. As calculated biomass
approaches this level, the model term reduces the
growth rate through negative feedback.
RESPIRATION
Respiration represents the most important continu-
ous loss term for the mathematical model of
Cladophora growth. Experiments currently in progress
at the BIOTRON facility will carefully define the
relationship between respiration, temperature, and
overall metabolic activity. At the present time Jack-
son's data (1966) are used to derive the relationship
between respiration and gross and net photosynthesis.
A linear relation between respiration and temperature
is used in the model. This function is described in
Equation 2. Calculated values for respiration rate are
input to the model to obtain the net specific growth rate
as a function of temperature.
R = FT (T/20) Eq. 2
where: R : respiration at T°C
R* : respiration at 20°C
T : measured temperature
SLOUGHING
In that Cladophora is not grazed to any extent, the
only other loss term is sloughing. Sloughing occurs
when severe mechanical disruption causes the algal
filaments to become unattached from their substrate
and float free in the water column. Shoreline
deposition and nuisance accumulation generally result
from this process. Although overall physiological
condition is thought to bear importantly on sloughing,
we have been quite successful in relating sloughing
events directly and solely to severe storm (high wind)
events. For the current model, sloughing is related
empirically to the occurrence of storm events.
Experiments in progress with in situ algal populations
will better relate those storm events, standing crop,
and magnitude of sloughing loss.
BIOMASS
The end product of the model is the prediction of
standing crop of Cladophora biomass. Biomass is
accumulated in the model as the product of the growth
rate and the current standing crop. The standing crop of
Cladophora biomass at 14 stations at the Lake Huron
study site is measured weekly by harvesting repre-
sentative samples of the alga and substrate. Density of
coverage and areal distribution are measured as well.
The results are expressed as grams of oven-dried algal
material per square meter of substrate (gDW/m2). The
most important calibration of the model involves
comparison of model generated biomass data with that
observed through the growing season.
RESULTS OF MODEL CALIBRATION
An understanding of the growth dynamics of
Cladophora entails both spatial and temporal resolu-
tion. Generally, temporal or seasonal dynamics are
more difficult to simulate. Figures4 through 6 compare
model output of spatial variation for soluble reactive
phosphorus, internal phosphorus, and biomass with
observed annual average values. In the case of
biomass, the midsummer mean value is used to better
reflect the maximum standing crop. Model agreement
for the two phosphorus components is quite good,
accurately reflecting the reduction in dissolved and
stored phosphorus with increasing distance from the
loading point. Biomass simulation is also quite good,
especially considering the heterogeneity of the near-
shore substrate (mixed sand, gravel, and cobbles).
Figures 7 and 8 compare model output with
monitoring data describing the seasonal variation in
biomass and internal phosphorus for stations near the
-------
FACTORS INFLUENCING THE DYNAMICS OF EUTROPHICATION
51
nutrient source. The mean for each sample date for six
stations is plotted. Internal phosphorus levels are
relatively constant throughout the year for these
stations because of continuous loading from the
wastewater treatment plant. The biomass data illus-
trate temperature limitation in the spring and fall as
well as the rise to a maximum standing crop in June
through August. Superimposed upon the smooth curve
for maximum standing crop is the impact of sloughing
loss. The sharp declines in biomass are major
sloughing periods and have been empirically asso-
ciated with storm events. In most cases, biomass levels
return rapidly to their pre-slough levels as the
restrictions of carrying capacity and substrate/light
competition are relaxed following sloughing. This
excellent match between model calculated values for
biomass and the phosphorus components and observ-
ed data allow projections to be made regarding the
impact of various levels of phosphorus removal on
nuisance growth of C/adophora at the site.
300
250
~E200
^
Q
0>
— 150
(n
co
<
Q 100
m
50
• Data
o Model
0.25 0 0.25 0.50 0.75 1.00
DISTANCE FROM NUTRIENT SOURCE (km)
Figure 4. — Comparison of model output and observed data
for the distribution of soluble reactive phosphorus about the
nutrient source. Values are annual means.
0.40
I Q3°
"3.
£ 0.20
£ 0.10
! °
f 125 250 375 500 625 750 875 1000
kT = 52.4 LIGHT, I (uE/m2-sec)
,L
R = 0.144 day"
Figure 5. — Comparison of model output and observed data
for the distribution of internal phosphorus levels about the
nutrient source. Values are annual means.
CO
§
100
80
I- en
00
20
• Data
o Model
0.25 0 0.25 0.50 0.75
DISTANCE FROM NUTRIENT SOURCE (km)
1.00
Figure 6. — Comparison ot model output and observed data
for the distribution of C/adophora biomass about the
nutrient source. Values are midsummer means.
£0.50
o
J A
1979
Figure 7. — Comparison of model output and observed
values for the seasonal variation in C/adophora biomass.
Data are mean values for six stations near the nutrient
source.
300 -
-200 -
J A
1979
Figure 8. — Comparison of model output and observed
values for the seasonal variation in internal phosphorus
levels. Data are mean values for six stations near the
nutrient source.
DEMONSTRATION PROGRAM AND
MODEL PROJECTIONS
In February of 1980 a demonstration phosphorus
removal program was instituted at the Harbor Beach
wastewater treatment plant. Phosphorus was removed
by alum precipitation with a polyelectrolyte coagulant
aid. Initial results indicate that an 80 percent reduction
in soluble reactive phosphorus at the mouth of Spring
Creek may be anticipated. The calibrated model is used
-------
52
RESTORATION OF LAKES AND INLAND WATERS
to project the impact of various levels of phosphorus
reduction on the standing crop of Cladophora at the
site. In this manner, the accuracy of the calibrated
model may be verified. Figure 9 describes the spatial
distribution of midsummer biomass levels associated
with several degrees of phosphorus removal. The
impact is most noticeable at points remote from the
nutrient source. Such a result is consistent with the
relationships presented in Figures 3 and 5. Algal
material distant from the nutrient source has lower
levels of internal phosphorus ( 0.10 percent) and lies,
therefore, in the most sensitive part of the growth
response curve. Additionally reductions in biomass in
close proximity to the source is much less dramatic. At
these locations, internal phosphorus levels are high
(approx. 40 percent). Reductions in phosphorus loading
will alter the internal phosphorus levels of the algae,
but in most cases, pool levels will remain on the
saturated (insensitive) part of the growth response
curve (see Figure 3).
We have learned from examining this relatively small
environmental perturbation at Harbor Beach that such
events may drastically affect the capacity of the
organism to respond to modest incremental improve-
ments in water quality. From model projections it can
be established that significant reductions in biomass
may require almost complete removal of phosphorus,
particularly near the source. Loading reductions of the
magnitude necessary for a return to unperturbed
conditions may approach the limit of cost/benefit
analysis feasibility.
Jackson, D. F. 1966. Photosynthetic rates of Cladophora
fracta from two sites in Lake Ontario under natural and
laboratory conditions. Pages 44-50 in Proc. Ninth Conf.
Great Lakes Res., Publ. 15. Great Lakes Res. Div. University
of Michigan, Ann Arbor.
Lin C. K. 1971. Availability of phosphorus for Cladophora
growth in Lake Michigan. Pages 39-43 in Proc. 14th Conf.
Great Lakes Res. Int. Assoc. Great Lakes Res.
Neil, J. H., and G. E. Owen. 1964. Distribution, environmental
requirements and significance of Cladophora in the Great
Lakes. Pages 113-121 in Proc. Seventh Conf. Great Lakes
Res. Publ. 11. Great Lakes Res. Div. University of Michigan,
Ann Arbor.
Senft, W. H. 1978. Dependence of light-saturated rates of
algal photosynthesis on intracellular concentrations of
phosphorus. Limnol. Oceanogr. 23:709.
Shear, H., and D. E. Konasewich. 1975. Cladophora in the
Great Lakes. Int. Joint Comm. Windsor, Ontario.
Rhee, G. 1978. Effects of N:P atomic ratios and nitrate
limitations on algal growth, cell composition and nitrate
uptake. Limnol. Oceanogr. 23:10.
ACKNOWLEDGEMENTS
The support and encouragement of Nelson Thomas of the
EPA LLRS at Grosse lie, Mich, is gratefully acknowledged as
is the assistance of the EPA Project Officer, Dave Dolan,
Byron P. Lane, Thomas Bugliosi, and Joyce Mechling. Dr.
James M. Graham and Dr. James Hoffman of the University
of Wisconsin contributed to the BIOTRON studies. This
research was supported by EPA Grant R806600010.
I.O
15 20 25
TEMPERATURE (°C)
35
Figure 9. — Model projections of Cladophora biomass at
various locations at the study site related to several levels of
phosphorus removal.
REFERENCES
Auer, M. T., and R. P. Canale. 1980. Phosphorus uptake
dynamics related to the mathematical modeling of
Cladophora at a site on Lake Huron. Jour. Great Lakes Res.
6:1.
Droop, M. R. 1973. Some thoughts on nutrient limitation in
algae. Jour. Phycol. 9:264.
Herbst, R. P. 1969. Ecological factors and the distribution of
Cladophora glomerata in the Great Lakes. Am. Mid. Nat.
82:90.
-------
53
ROLES OF MATERIALS EXPORTED BY RIVERS INTO
RESERVOIRS IN THE NUTRITION OF
CLADOCERAN ZOOPLANKTON
G. R. MARZOLF
J. A. ARRUDA
Division of Biology
Kansas State University
Manhattan, Kansas
ABSTRACT
The differences between natural lakes and reservoirs are variously significant to the regulation of
limnetic processes. Large suspended sediment loads 0.1 to 0.5 gms per liter are common in rivers
of agricultural landscapes of the Great Plains of the United States. This material often reduces
light penetration to the extent that photosynthetic production is inhibited, despite high nutrient
loads that are also common in the Central Plains States. Such bodies of water do not fit most
phosphorous-chlorophyll regression models; they are "low and to the right." Nevertheless,
zooplankton populations do not seem to be reduced under such conditions. They are able to filter
small particles and clays and use the bacteria and/or the dissolved organic matter associated with
them. This implies that the river and its drainage basin continue to be the driving variable for
secondary production in a reservoir in a fashion distinctly separate from natural lakes where
zooplankton phenomena have been investigated more fully.
The differences between natural lakes and reservoirs
are variously significant to the regulation of limnetic
processes. (Neel, 1962; Baxter, 1977; Marzolf, 1980).
Large loads of suspended sediments (0.1 to 0.5 gm per
liter) are common in rivers of agricultural landscapes of
the Great Plains of the United States. This material
reduces light penetratiqn, often to the extent that
photosynthetic production is inhibited despite high
nutrient loads also common in the Central Plains
States, (Marzolf and Osborne, 1971). Such bodies of
water do not fit most phosphorus-chlorophyll regres-
sion models; they are "low to the right" (Jones and
Bach man, 1978). Nevertheless, zooplankton popula-
tions do not seem to be reduced under such conditions.
This implies that the river and its drainage basin
continue to be the driving variable for secondary
production even under the lake-like conditions of the
reservoir. The distribution pattern is distinct from
natural lakes where zooplankton have been investi-
gated more fully.
In this presentation we address the question: In the
face of reduced photosynthetic production by algae
because of silt and clay turbidity, what is the food
resource of the filter-feeding limnetic zooplankton?
The alternative candidates for zooplankton foods are
organic matter produced upstream in the watershed or
in the river itself, and/or the bacteria that are
decomposing this allochthonous material. Organic
detritus enters the reservoir from the river as both
dissolved and particulate fractions; the dissolved
fraction is usually of greater mass, often by as much as
20 times. This has been reported from natural lakes
and streams (Wetzel and Rich, 1973), from the oceans
(Durrsma, 1960) and from a few rivers (Weber and
Moore, 1967). The particulate fraction is directly
available to filter-feeders; the dissolved fraction is not.
Marzolf (1980) demonstrated in a preliminary way that
dissolved organic matter can be rendered available to
filter-feeders through adsorption onto clays, i.e., a
dissolved amino acid adsorbed on clay was desorbed
and retained by Daphnia pulex upon being allowed to
filter such a suspension. It is not clear that dissolved
organic matter generally can provide the nutrition to
maintain zooplankton metabolism, growth, and repro-
duction by this mechanism since the nutritional
qualities are likely to be variable, and in some cases,
inadequate. Bacterial use of dissolved organic sub-
strates, on the other hand, can render dissolved
fractions particulate; with the incorporation of in-
organic nutrients from ambient water the quality of the
particulate organic matter as zooplankton food will
increase. The details of that process remain to be
demonstrated, but it is clear that the presence of clay
particles enhances the activity of bacteria (Jannasch
and Pritchard, 1972). The association of silt and clay
particles, dissolved organic matter, and microorgan-
isms offers a usable food resource.
We are not prepared to discount the continued use of
algal cells as cladoceran food in turbid reservoirs but
we consider their importance to be reduced for two
reasons: (1) The largest concentrations of chlorophyll-
bearing cells are found in the inflowing river water
along with the highest concentrations of silt and clay
particles. We show here that high clay concentration
inhibits the feeding rate on algal cells. (2) Algal density
and the rate of photosynthetic production are reduced
with increasing distance from the inflow (Marzolf and
Osborne, 1971); thus, just as the inhibitory effect of
clay particles on filter-feeding is removed the avail-
ability of algal cells in the resource is diminished.
As part of an investigation into the roles of
suspended silts and clays in zooplankton nutrition, we
have made several measurements in the laboratory to
document the ingestion of inorganic silt and clay
particles by Daphnia sp. and the inhibition of algal
ingestion in the presence of inert particles. The
following describes this evidence.
-------
54
RESTORATION OF LAKES AND INLAND WATERS
METHODS
The first experiment reported here estimates the
clearance and ingestion rate of clay particles by the
cladoceran Daphnia pulex cultured from Tuttle Creek
Reservoir. The experimental methods have been
reviewed by Rigler (1971). The animals grazed, for at
least 1 hour, in a suspension of non-radioactive
paniculate food to acclimate to the experimental
conditions. The animals then were transferred to an
identical, but radioactively labeled feeding suspension
for 5 minutes, during which radioactive particles were
ingested. This length of time was short enough that
radioactive feces are not likely to be produced and long
enough to minimize any effect of transferring the
animals. After measuring the radioactivity of feeding
"suspensions and animals, clearance and ingestion
rates were calculated (Rigler, 1971).
The coarse clay mineral particles (mean diameter is
4.65 micrometers) used in the clay ingestion experi-
ment were processed from natural lake sediments. The
dominant mineral was montmorillonite; illite was also
present. After air-drying the sediments, they were
roller-milled then pin milled into a dry, textured ground
mineral composed of variously sized particles. Pre-
liminary size fractionation of the milled sediments in a
Bahco Micro-Particle Classifier (Harry Dietert Co.) was
followed by wet fractionation with centrifugation. The
suspended clay particles were labeled with the
radionuclide Zn-65. This divalent cation adsorbs to the
surface of the clay particles (Bachman, 1961), thus
making possible the estimates of clearance and
ingestion. In this experiment, clearance and ingestion
rates were measured over a range of clay particle
concentrations from 103 to 106 particles/ml. Fouradult
and four juvenile Daphnia pulex were used in each
treatment: 25 ml of the desired concentrations of clay
particles in DM2 medium (D'Agostino and Provasoli,
1972). The adult Daphnia were individually counted,
while the juveniles were paired. Animals and filtered
aliquots of suspensions were directly counted (Beck-
man 4000 Gamma Counter).
The second experiment examined the interference of
algal ingestion by Daphnia pulex when suspended
sediments were present. The alga, Ankistrodesmus
falcatus var. acicularis was labeled by incubation with
C-14 sodium bicarbonate. Feeding suspensions of
labeled and unlabeled algae were prepared by adding
the cells and the appropriate amount of washed lake
sediments to the DM2 medium to produce the desired
final concentration of both algae and sediments. These
data are part of a larger experiment in which the
concentrations of algae and sediments were varied.
Four levels of algae (1.65 X 103 to4.46l(T04 cells/ml)
and six levels of sediments (0.0 to 160.0 mg/l) were
used. Four adult Daphnia pulex were used in each
treatment combination in 25 ml of feeding suspension.
Prior to liquid scintillation counting, the animals were
allowed to dry in open scintillation vials before adding a
tissue-dissolving agent.
RESULT
Over the range of clay concentrations used in the
clay feeding experiment, ingestion rates (Fig. 1) of both
adult and juvenile Daphnia pulex increased with
particle concentration (treatment effect of particle
concentration on log ingestion rates: P > F 0.0001 for
both sizes). Clearance rates also declined as particle
concentration increased (P > F 0.0001 for adults, P > F
0.0012 for juveniles). The slopes of the adult and
juvenile lines within each parameter differed (log
clearance rate: P > F 0.0001; log ingestion rate: P > F
0.0001) This means that the adult and juvenile animals
used in this experiment differed in their response to
increasing particle concentration. The adults ingested
more particles as particle concentration increased and
their clearance rates decreased less rapidly than did
the juveniles.
The linearity of the ingestion rate function demon-
strates that up to 1.0 X 106 coarse clay particles/ml,
the capacity of these Daphnia to ingest these clay
particles is not limited. Clearance rates are generally
thought to be constant and maximal when particle
concentration is below some saturating level (Hall, et
al. 1976). This experiment suggests that clearance
rates decline as ingestion rates increase in response to
increasing particle concentration.
Ingestion of Ankistrodesmus cells by Daphnia pulex
(Figure 2) is decreased by the presence of suspended
sediments (treatment effect of sediment concentration
on log ingestion rate: P > F 0.0001) At a sediment
concentration of 160 mg/l ingestion rates are reduced
to about 6 percent of the rates in the treatments lacking
sediments. Clearance rates also decline ( P >F 0.0001)
• O ADULT
JUVENILE
CLAY CONCENTRATION IparllcloB/ ml I
Figure 1. — Clearance and ingestion rates of Daphnia
pulex in suspensions of coarse clay particles. Each point is
the mean of 4 observations (adults) or 2 observations
(juveniles), with standard error bars.
as sediment concentration increases. Clearance rates
represent the minimum volume of water that must be
filtered to produce the observed radioactive disintegra-
tions in each animal. The measure probably under-
estimates true filtering rates and it says little about the
efficiency of filtration of algal cells or sediment
particles as they are affected by sediment concentra-
tions.
DISCUSSION
In this conference session devoted to the factors
influencing the dynamics of eutrophication where most
attention is given to nutrient responses of limnetic flora
we are hesitant to divert too much attention toward
-------
FACTORS INFLUENCING THE DYNAMICS OF EUTROPHICATION
55
processes centering on secondary production in
reservoirs or suspended sediments. These subjects
may be trivial in comparison to nutrient regulation of
primary productivity in lakes and we are not inclined to
say much to the contrary. Lake restoration projects,
however, are often developed for bodies of water that
are: (l) Artificial impoundments, (2) identified as turbid
(often wrongly or simplistically associated with the
eutrophic condition), and (3) not essentially regulated
by the loading of the plant nutrients. It seems
appropriate to discuss other biological phenomena that
occur in lakes that need restoration whether they are
eutrophic in the classic limnological sense or not. Our
point of view is that the water quality in lakes,
reservoirs, and streams is controlled to a large degree
• • INOESTIOH
A * CLEARANCE
• 10 20 30 40 Si 10 70 10
CLAY CONCENTRATION ( mg / l|l*r >
Figure 2. — Clearance and ingestion rates of Daphnia
pulex in suspensions of Ankistrodesmus and clay
sediments. Each point is the mean of 16 animals, 4 from
each of 4 treatment levels of algal cell concentration (from
1.65 X 103 to 4.46 x 104 cells/ml), with standard error
bars.
by biological processes. Further, that the organisms
whose metabolism and activities are central to the
control adapt to the physical and chemical environment
influenced by that control.
We have considered the concentration and quality of
zooplankton resources to be of prime importance in
explaining zooplankton activity. This is parallel to the
perspective of our colleagues on this panel as they
have considered nutrients and phytoplankton activity.
The only difference is that we are constrained to focus
on paniculate materials because the resources of these
filter-feeders do not include dissolved materials. In fact,
a great unknown in evaluating the filter-feeding
process is the lower size limit of usable particles. Is
there a sharp threshold identified by the geometry of
zoopianKtonic filtering appendages? Do species of
zooplankton differ in their capacity to use paniculate
resources at the small end of the size spectrum? Do
zooplankters adapt to changing size frequency distribu-
tions of their resources by altering their filtering
behavior?
It is our thesis that different species of cladoceran
filter feeders respond differently to such changes in
resource size frequency.
If this is true then we should expect to find some
species at a competitive advantage in an environment
where the dominant size frequency categories are
small, say, less than 2 micrometers. We further
suggest that the mechanism for resource availability
that is related to the available surface area for the
adsorbtion of dissolved organic matter will reward
filter-feeders for their capacity to filter smaller
particles. That is, the surface area per "gut full" of
particles increases geometrically with decreasing
particle size (Arruda, 1980).
SUMMARY
I.The impoundment of rivers to form reservoirs
provides habitat for zooplankton in regions where
natural lakes are rare.
2. The river and its drainage export silts and clays
into the reservoir that regulate the trophic patterns in
reservoirs by establishing gradients in turbidity and
thus photosynthetic production.
3. Clay particles interfere with the filtration of algal
cells by Daphnia and decrease ingestion of them as
clay concentration increases.
4. Clay particles are ingested by filter-feeding
zooplankton and may be usable as a- food resource
because of adsorbed organics and bacteria associated
with them.
REFERENCES
Arruda, J.A. 1980. Some effects of suspended silts and
clays on the feeding behavior of Daphnia spp. from Tuttle
Creek Reservoir. Ph.D. thesis. Kansas State University.
(In prep.).
D'Agostino, A.S., and L. Provasoli. 1972. Dixenic culture of
Daphnia magna Straus. Biol. Bull., 139:485.
Baxter, P.M. 1977. Environmental effects of dams and
impoundments. Ann. Rev. Ecol. System., 8:255.
Duursma, E.K. 1960. Dissolved organic carbon, nitrogen,
and phosphorous in the sea. Neth. Jour. Sea Res., 1:1.
Hall, D.J., et al. 1976. The size efficiency hypothesis and
the size structure of zooplankton communities. Ann. Rev.
Ecol. Sys., 7:177.
Jannasch, H.W., and P.M. Pritchard. 1972. The role of inert
paniculate matter in the activity of aquatic micro-
organisms, in Proc. 1 BP-UNESCO Symp. on Detritus and
its Role in Aquatic Ecosystems. Pallanza. Mem. Inst. Ital.
Idrobiol. 29 Suppl: 289-308.
Jones, J.R., and R.W. Bachmann. 1978. Phosphorous
removal by sedimentation in some Iowa reservoirs. Verh.
Int. Verein. Limnol. 20:1576.
Marzolf, G.R. 1980. Some aspects of zooplankton existence
in surface water impoundments, in H. Stefan, ed. Proc.
Symp. Surface Water Impoundments. Am. Soc. Civil Eng.
Marzolf, G.R., and J.A. Osborne. 1971. Primary production
in a Great Plains Reservoir. Verh. Int. Verein. Limnol.
18:126.
Neel, J.K. 1964. Impact of Reservoirs. Pages 575-593. in
D.G. Frey, ed. Limnology in North America. University of
Wisconsin Press, Madison.
Rigler, F.H. 1971. Feeding rates: zooplankton. Pages 228-
255. in A manual on methods for the assessment of
secondary production in freshwaters. Int. Biol. Prog.
Handbook No. 17. Blackwell.
Saunders, G.W. 1969. Some aspects of feeding in
zooplankton. Pages 556-573. in Eutrophication: Causes,
consequences, correctives. Natl. Acad. Sci., Washington,
D.C.
Weber, C.I., and D.R. Moore. 1967. Phytoplankton seston
and dissolved organic carbon in the Little Miami River at
Cincinnati, Ohio. Limnol. Oceanogr., 12:311.
Wetzel, R.C., et al. 1972. Metabolism of dissolved and
paniculate detrital carbon in a temperature hard water
lake. Mem. Inst. Ital. Idrobiol., 29 Suppl: 185.
-------
56
METHODS OF ASSESSING NUTRIENT LOADING
HANSJORG FRICKER
Swiss Federal Institute for Water Research
and Water Pollution Control (EAWAG)
Dubendorf, Switzerland
ABSTRACT
The practical application of measuring techniques published by various authors which quantify the
nutrient load of a lake via influx and settlements have not been given sufficient attention. The
investigation of the influx, although time-consuming and costly, is essential for lake budgeting and
lake restoration studies. Experience acquired in connection with the OECD Eutrophication
Program is passed on for future practical use. Along with a short literature review and an
introduction to the theoretical correlation between waterflux and substance-concentration,
various possibilities of measuring the probable load are stated. In addition to continuous samples
proportional to the water flow in the nutrient-rich influxes (more than 15 percent of total annual
load), random sampling of other influxes (more than 5 percent of total annual load) is
recommended. The waterflux of the most important in- and outflow must be registered
automatically. High water demands special attention. Those areas not covered by direct
measurements can be assessed by statistical methods. The problem of nutrient export coefficients
is briefly handled. At the end, recommendations for practical application are given.
INTRODUCTION
Traditionally, limnological research is based on in-
lake investigation, and loading measurements have
often been neglected. However, in the past 5 to 10
years, increasing attention has been given to this
problem. Vollenweider (1968) was the first to establish
a quantitative correlation between nutrient loading and
the trophic state of a lake, thereby enabling us to
quantify the biological response of lakes. Application of
such models requires, however, a sound knowledge of
the component balance, which can only be achieved by
an admittedly more complex inlet and outlet investiga-
tion.
The fundamental objective of the recently completed
OECD Eutrophication Program (Fricker, 1980), was to
study the correlation between nutrient loading of lakes
and their biological-chemical (trophic) response. An-
other aim was to determine a critical loading value, i.e.
a level which can be tolerated by the lake without
changing its trophic character.
A preliminary evaluation showed that in many cases,
the influxes (e.g. measurement of water flow, flood
surveillance) had not been sufficiently investigated as
initially planned.
The sum of all nutrient inputs is the decisive factor
controlling the trophic state of a lake. Quantitative
evaluation of the nutrient sources is of prime
importance for planning purposes, also for assessing
the effectiveness of water pollution control measures
in the catchment area and in the lake itself.
The following are the most important sources of
pollution:
• Point sources:
input via influents
input via domestic wastewater from sewer
systems
wastewater treatment plant effluents
• Diffuse sources:
input via precipitation
urban and agricultural runoff
groundwater
lake sediments
When planning an influx investigation, it is essential
to assess in a preliminary study, the loading fraction of
each influx while always taking into account the water
flow; furthermore, the measurement must be limited to
the most important inflows. These should constitute at
least 80 to 90 percent of the total load. Here, it must be
taken into consideration that an investigation of a
minor influx requires the same technical and analytical
sampling procedure as a major influx.
There have been only a few examples of an analytical
classification of the pollutant loads in lakes: Ambuhl
(1979), Lohri (1977), Treunert, et al. (1974), Unger
(1970), Wagner (1969), Wagner, et al. (1976). On the
other hand, several studies have been carried out on
the monitoring of river quality (Dandy and Moore,
1979; Davis and Zobrist, 1978; Manczak, 1968;
Manczak and Florczyk, 1971; Sanders and Adrian
1978; Stevens and Smith, 1978) which deal with the
problems of outlet and nutrient concentration mea-
surements, as well as with the assessment of nutrient
loads.
-------
NUTRIENT LOADING/TROPHIC RESPONSE
57
The aim of this paper is to give a brief review of the
theoretical background between water flow and
nutrient concentration, and to discuss the different
ways to measure nutrient load, including the most
important factors involved.
RELATIONSHIPS BETWEEN WATER
FLOW AND NUTRIENT
CONCENTRATION
Because of the special conditions prevailing in the
catchment area, a fairly close correlation exists in every
water body between water flow and the simultaneously
measured concentration of a pollutant. This correlation
can also be used in determining pollutant loads
(Manczak, 1968; Manczak and Florczyk, 1971). Accord-
ing to Bernhardt, et al. (1974), such a correlation does
not seem to exist in small to middle-sized rivers (e.g.
Wahnbach approx. 1 mVs) with low water flows.
However, they obtained similar results as Manczak and
Florczyk (1971) with higher water flows. These authors
proposed three types of correlation (Figure 1) based on
two overlapping processes: (a) dilution, a low concen-
tration with an increasing water quantity, and (b)
erosion, an increased concentration with a higher
water flow. The shapes of the curves in Figure 1
depend on many factors. The most important are:
1. Chemical components.
2. Degree of pollution in the river.
3. Its hydrotechnical and hydrological characteristics.
4. Its self-purification capacity (temperature depen-
dent).
5. Distance of the monitoring station from any source
of pollution.
6.Quality and quantity variation of the discharged
loads.
In heavily polluted rivers (Type I), the main factor
influencing the shape of the curve is the dilution of
wastewater with increasing water runoff. In clean
rivers (Type II), the concentration increases with
increasing water runoff as a result of greater
resuspension and transport of river sediments. In
intermediately polluted rivers with a low flow, the
dilution of wastes is the predominant factor, so that the
curve is similar to Type I. At higher flows, the dilution
effect disappears and the influence of resuspension of
bottom deposits and washout of the drainage area
dominates. Thus, with flows which are higher than the
annual average, pollution increases with a higher flow.
The concentration can decrease again when diluted
with even higher water quantities. Based on this
model, analytical classification concepts of the total
load into different sources (loading from wastewater,
runoff, . . . .) are presented by various authors
(Dandy and Moore, 1979; Davis and Zobrist, 1978;
Liebetrau, 1979; Manczak and Florczyk, 1971; Smith,
1977; Zobrist, et al. 1977).
To start a lake balance investigation and loading
model, the annual load should first be determined.
However, for further evaluation of water protection
measures in a lake and catchment area or for
developing specific lake restoration strategies, it is
essential to quantify the different sources of pollution.
In practice, variation of the analytical values of the
correlation between water flow and concentration is
best compensated by the function of a higher degree.
According to Wagner (1969), polynomials proved to be
suitable, in particular the equation Y = a/x + b +cx + dx2
+ex3, as they often supply the least squares sum, and
are also comparatively easy to calculate. Polynomials
are often derivatives of the common type:
3
Y = I(BrX')
i = -1
Y = material cone.; x = amount of flow >
Bi = regression coefficient
i = exponents and indices of the regression
coefficents
(according to Wagner, et al. 1976)
Depending on the sample taking technique (random,
collective samples), an adapted polynomial function of
concentration and water flow neglects the sea.sonal
and/or daily concentration variations. According to
Davis (1980), the often observed large differences in
concentrations at the same water flow are not
necessarily caused by deviation, but rather by seasonal
fluctuation. For example, in the case of nitrate (or
phosphate) it is mainly the water temperature and
resulting biological activity.
McMichael and Hunter (1972) and Thomann (1967)
have introduced a cosine function to account for both
annual and weekly cycles in the flow model. Buhrer
(in preparation) combines the polynomial of the
regression computation with a Fourier series over time.
Schweingruber (1980) has tried to establish a linear
combination between the annual cycles (sine function)
and the water flow dependent (polynomial) terms, but
optimization of all coefficients proved to be difficult.
Basically, the calculation procedure for assessing loads
must be improved. Until new models are tested on
different lakes, the polynomial remains the most
suitable function for practical application. Above all, no
correlation has yet been established between the
computed results from integrated collective samples
and the ones obtained from random samples of the
same river. Besides, the nutrient loads registered
during 1 year may not be transferred so readily from 1
year to another. Particularly the varying climatic data
should be taken into consideration.
PRACTICAL CONSIDERATIONS
An influx investigation program should cover the
concentration of water substances over the entire
range of the water flow in order to statistically
guarantee the load calculation.
Numerous individual analyses are technically always
possible in low and normal water levels. In high water,
however, where individual .results can fluctuate
considerably, special measures should be undertaken.
According to Keller (1970), the total phosphorus load
remains small up to approximately five times the mean
annual flow. It increases only with high waters which
contain at least 5 to 10 times the mean annual flow,
thereby increasing particulate organic phosphorus,
while dissolved phosphorus remains small. Unger's
investigation (1970) of the Argen (mean water flow
18.6 mVs) shows that, in only 9 days, 10 percent of the
-------
58
RESTORATION OF UVKES AND INLAND WATERS
annual water volume, 25 percent of the total
phosphorus, and 15 percent of the phosphate load
flowed off. In general, not only do high waters
significantly increase substance flow off, but also
momentary water level increases after dry periods.
Therefore, as expected, greater component volumes
are discharged with increasing water levels than with
decreasing water levels directly after peak water levels.
How often do high waters occur? Should data of the
mean daily flow values exist, the occurrence probability
may then be assessed by a frequency distribution.
Depending on the sampling method used, additional
samples must be taken after and during abundant
rainfalls. A critical value of 15 mm rainfall/day
(Wagner, et el. 1976) was registered for Lake
Constance. Nevertheless, it will never be possible to
determine the nutrient load of the high water
maximum for each river. Already the requirement, to
determine as often as possible, unforseeable high
water flux, demands numerous personnel. Generally,
the extrapolation range between the extreme high-
water situation, where no concentration measure-
ments exist, and the range in which concentration
measurements exist must be minimized. Polynomes
used as fitting-curves tend to deviate too strongly
beyond the plotted values. This can cause large errors
in the load calculations.
MEASUREMENT OF WATER FLOW
The accuracy of nutrient load budgeting depends
upon the accuracy with which the water flow of the
influx is measured. According to Bernhardt, et al.
(1 974), continuous water flow measurement is a factor
two to four times more accurate than intermittent
measurement. Consequently, it is somewhat absurd to
determine a substance concentration with great
precision and only estimate the water flow roughly. The
exact determination of the water volume requires a
calibrated sampling spot in the flowing water, with a
solid installation safe from high waters, and equipped
with a water-level-registering instrument (limnigraph,
water gauging station). The water volume can then be
calculated from the water level by two methods:
A. A water-velocity-profile is taken of the stream. The
velocity, multiplied by the cross-section area, gives the
water volume per unit time. When this is done for
various water levels, probable relations between water
volume and water level can be calculated.
B. In the literature (Weyrauch, 1915), empirical
relations are described for measuring weirs with
known profiles but apply only for slow-moving water
currents.
MEASUREMENT OF CONCENTRATION
Different methods of sampling running waters exist
(Wagner, 1980; Ambuhl, 1973): Continuous sampling
(automated) over several days, 24-hour mixed samples,
a combination of several random samples to 24-hour
mixed samples and single samples. These methods
must be applied according to the loading importance of
the different flowing waters, which can be estimated by
the mentioned pre-study. An example here is the Lake
of Sempach study (EAWAG, 1979) in which all influxes
were walked off, once in wet weather and once in dry
weather. The annual load was approximated on the
basis of the relative frequency of the weather condition
in that the dry weather study was weighted with a
factor 5 and the wet weather study with the factor 1.
All influxes whose yearly load supply more than 5
percent of the calculated total load are considered
important.
SINGLE SAMPLES
Waters whose load constitute 5 to 1 5 percent of the
total load of a substance can be accurately measured
by random samples, providing they are not subjected to
systematic fluctuations and have a relatively constant
water flow, even at times of precipitation. To minimize
the high costs of river quality monitoring, Sanders and
Adrian (1978) have developed special statistical criteria
based on the standard deviation formula. According to
Bernhardt, et al. (1974) the standard deviation of the
load, determined from a theoretical reference value, is
20 percent for a sampling frequency of 28 days, and 5
to 10 percent for 14 days. The sampling days must be
distributed over the investigation period according to a
specific time plan. For example, the influxes of Lake
Constance (Bodensee) are investigated regularly all 18
days (not a factor of 7); therefore on each weekday the
influxes are investigated three times during the year
(Wagner, et al. 1976). The sampling must also include
all hydrological conditions and be supplemented with
sampling during high water. It is best when the water
volume is registered continuously or at least daily. This
way the random sample method requires the least
effort, and the load error can be determined statis-
tically. Diurnal fluctuations, depending upon the model
function, can be taken into account. In sewage-loaded
streams, the day-night rhythm of sewage-originated
substance concentration must be included in the
calculation. This can be taken into account by using a
complementary investigation program (EAWAG,1979):
First, the sampling times of all random samples must
be known. Secondly, during a period of normal water
flow, a 24-hour continuous sample is taken, and during
the same time, many single probes are taken in short
intervals. From the ratio of the single probe concentra-
tion and the 24-hour mean concentration, factors are
calculated to correct the measured random sample
concentrations to a mean daily value.
24-HOUR CONTINUOUS SAMPLES
In addition to the regular random samples, 24-hour
continuous samples are recommended. Instruments
for this use are already on the market. They take water
samples proportional to the water level over a longer
period of time (hours to days). The sampler of Quantum
Science Ltd., England is basically a plastic cylinder
which is anchored directly in the flowing water. The
amount of water entering the cylinder is regulated by
an adjustable valve which controls the amount of air
escaping from the cylinder. The analytical data
obtained in this manner can be treated statistically the
same as the random samples. The measured concen-
tration is set in relation to the mean water flow over the
same time period (limnigraph). In this manner, the
problem of diurnal fluctuations can easily be avoided
-------
NUTRIENT LOADING/TROPHIC RESPONSE
59
WATER FLOW-PROPORTIONAL
PERMANENT SAMPLING
An influx whose nutrient load constitutes an
important part (more than 15 percent) of the total load,
must be taken with permanent sampling devices.
Mountain streams, which can overflow rapidly during
storms and thereby transport a significant portion of
the total load into the lake within a short time also
belong to this category of influx. Fraction samples are
taken which are proportional to a previously deter-
mined water volume, i.e., the greater the water flow,
the greater the frequency with which fractional
samples are taken. Optimal sampling results when the
sampling is regulated by a water gauge (limnigraph).
This implies that the relationship between water level
and water flow must be accurately known in advance.
Correct operation of the installation depends on the
exactness of this relationship. Depending on the
amount of samples taken, problems can arise in high
water situations, so that, in spite of increasing water
flow, the sampler cannot operate quicker. For this
reason, the automat must provide for at least two
ranges of water flow, each of which applies for a
certain ratio between fraction sample and total water
volume. For each range a separate sampling container
is necessary. The easiest method of conserving the
probes is in a refrigerator. The accuracy of determining
the nitrite and ammonium concentration is lessened
with the duration of stay. The weekly analyzed sample
concentrations are multiplied by the corresponding
water flow and added up week by week for the total
load.
This method is surely the most reliable for
determining the total load and compared with the
random sampling method, mathematically much less
complicated to handle. The division of the total load into
its origin is not possible by mathematical techniques;
here additional random samples are necessary.
OTHER SOURCES
Sewage treatment plants: The concentration and
load of nutrients at the outlet of a sewage treatment
plant can be described only poorly, even with the help
of complicated functions. The total daily load can be
measured accurately by using a suitable collector
device; several 24-hour integrated samples would be
sufficient. Weekly and seasonal variation should be
taken into account. Synchronous measurement of the
flow is essential. In the evaluation, differences caused
by rain water overflow are to be taken into account.
Diffuse sources: Most authors (e.g. Duncan and
Rzoska, 1979; Uttomark, et al. 1974) agree that
increasing land use and the substantial outflow of
paniculate material from the drainage area results in
serious deterioration of many water bodies. The kinetic
energy of flowing water is the primary transport
mechanism of these materials. Other influencing
factors are: General topography, contour, soil proper-
ties, vegetative cover, agricultural practices and
livestock, and precipitation. While further research on
the cycle of nutrients is certainly necessary (c.f. MAB
project No. 5: Mechanisms of land use impacts on
inland waters), effective means of reducing agricultural
input into receiving waters are already known but
often, at least from an economical point of view,
difficult to realize. Origin of nutrient input which
cannot be analytically determined (e.g. single housing
on the lakeshore, ground water from slopes, areas
outside of the drainage area of investigated flowing
waters (statistical area) must be estimated. This can be
done with the help of so-called nutrient export
coefficients found in the literature. Another possibility
is to set the unknown nutrient export of an area
proportional to that of a neighboring area whose export
has been analytically determined.
NUTRIENT EXPORT COEFFICIENTS
It is quite probable that enough basic data are
available for some catchment areas to calculate the
nutrient load. In general, applying nutrient export
coefficients from literature may be sufficient to plan an
investigation program, even though, with regard to
accuracy, their values are not exact. There is a
substantial quantity of literature (see review in
Uttormark, 1974) for estimating the nutrient input into
lakes. It is apparent from the data that considerable
variation exists in the quantity of nutrients that are
exported from similar areas devoted to the same use.
Latest research from Greifensee studies has shown
that compared to the values found in the literature
(calculations on the basis of seven test areas; Gachter
and Furrer, 1972) the phosphorus export coefficients
can be up to twice that amount. In practice, we
recommend (in accordance with MAB 5) that in every
large influx study, a small test area be investigated.
Here, land use should be defined in some detail as to
type of agriculture, forestry, recreation, and tourism,
and the intensity of usage should be quantified (e.g.,
fertilization rates). The nutrient export from this area
must be intensively studied. This will produce
representative data so indispensable to the calculation
of load in the so-called statistical drainage area (area
with no loading measurements).
CONCLUSIONS AND
RECOMMENDATIONS
APPLICATION
FOR PRACTICAL
Studies of the influx are vital for lake restoration
programs. A pre-study can reduce the investigation
program to the most important (nutrient rich) influxes.
The waterflux must be measured at least once per day,
or better even, continuously registered with a
limnigraph. The influx with a nutrient load of more than
15 percent of the total load should be investigated with
a waterflux proportional sampling automat (7 days a
week). ~o be accurate, the relationship betwen
waterflux and water level must be given special
attention by repeated measurings. For influx with loads
up to 15 percent, a 1-day continuous sample or a
random sample (e.g., once every 18 days) is sufficient.
To obtain a statistically sufficient distribution of
samples over the entire range of water levels, including
high waters on call, special planning is required. To
-------
60
RESTORATION OF LAKES AND INLAND WATERS
evaluate random samples, a polynominal regression is
recommended, even though the systematical concen-
tration fluctuations are not taken into account by this
method. Improvements in this direction must be
attentively followed! Normally, for sewage treatment
plants, several 24-hour continuous samples are
sufficient. Depending on the conditions, precipitation
analysis may or may not be disregarded. The nutrient
export for parts of the drainage areas which have not
been measured, can be calculated by export coefficient
values found in literature. Better yet is the application
of export coefficients acquired from special test areas.
c
o>
o
c
o
o
c
a
Type I
Type I
Water Flow
Figure 1. — Basic types of curves representing the correlation between concentration of pollutants and rate of flow
(Manczak and Florczyk, 1971; Wagner, et al. 1976).
REFERENCES
Ambuhl, H. 1979. Die situation der Seen: Gewasserschutz
vor ungelosten Aufgaben. In Jahresbericht der EAWAG pro
1978.
Dandy, G. C., and S. F. Moore. 1979. Water quality sampling
programs in rivers. Jour. Environ. Eng. Div. Proc. Am. Soc
Civil. Eng. 105:695.
Davis, J. S. 1980. Jahreszeitlich bedingtes Verhalten der
Fliessgewasser. Gas-Wasser-Abwasser. 9. (in print).
Davis, J. S., and J. Zobrist. 1978. The interrelationship among
chemical parameter in rivers — analyzing the effect of
natural and anthropogenic sources. Progr. Water Technol.
10:65.
Duncan, N., and J. Rzoska, eds. 1980. Land use impacts on
lake and reservoir ecosystems. Proc. MAS Project 5
Workshop, Warsaw, Poland, 1978. Facultas Verlag Wien.
EAWAG. 1979. Gutachten uber die Sanierungsmoglich-
keiten fiir den Sempachersee. Dubendorf, Switzerland.
Fricker, Hj. 1980. OECD Eutrophication Program: Regional
Project Alpine Lakes. Bundesamt fur Umweltschutz, Bern
Schweiz).
Gachter, R., and 0. J. Furrer. 1972. Der Beitrag der
Landwirtschaft zur Eutrophierung der Gewasser in der
Schweiz. Schweiz. Z. Hydrol. 34:41.
Keller, H. M. 1970. Der Chemismus kleiner Bache in
teilweise bewaldeten Einzugsgebieten in der Flyschzone
eines Voralpentales. Mitt. Schweiz. Anst. f. forstl. Ver-
suchswesen 46:113.
Liebetrau, A. M. 1979. Water quality sampling: Some
stochastical consideration. Water Resour. Res. 15:1717.
Lohn, F. 1977. Untersuchung der Zuflusse des Baldeg-
gersees. Bericht z.H. des Schweiz. Bundes fur Naturschutz
(SBN) und des Kant. Amtes fur Gewasserschutz Luzern.
Manczak, H., 1968. Ueber die Auswertung von Gewassergute-
Untersuchungen. Vom Wasser 35:237.
Manczak, H. and H. Florczyk. 1971. Interpretation of results
from studies of pollution of surface flowing waters Water
Res. 5:575.
McMichael, F. C., and J. S. Hunter. 1972 Stochastic
modelling of temperature and flow in rivers. Water Resour
Res. 8:87.
Schweingruber, M. R. 1980. Der Bielersee 1973-1978: Ein
Beitrag zum Problem der Modellierung chemischer Proz-
esse in naturlichen Gewassern. Dissertation. University of
Bern.
Sanders, T. G. and D. D. Adrian. 1978. Sampling frequency
for river quality monitoring. Water Resour. Res. 14:569.
Stevens, R. J., and R. V. Smith. 1978. A comparison of
discrete and intensive sampling for measuring the loads of
nitrogen and phosphorus in the river Main, County Antrim
Water Res. 12:823.
Smith, R. V. 1977. Domestic and agricultural contributions to
the inputs of phosphorus and nitrogen to Louqh Neaqh
Water Res. 11:453.
Thomann, R. V. 1967. Time series analysis of water quality
data. Jour. San. Eng. Div. Am. Soc. Civil Eng. 93:1.
Treunert, E., A. Wilhelms, and H. Bernhardt. 1974. Einfluss
der Probenahme-Haufigkeit auf die Ermittlung der Jahres-
Phosphor-Frachtwerte mittlerer Bache. Hydrochem. Hydro-
geol. Mitt. (Munchen) 1:175.
Linger, U. 1970. Berechnung von Stoff-Frachten in Flussen
durch wenige Einzelanalysen im Vergleich zu kontinuier-
lichen einjahrigen chemischen Untersuchungen: Gezeigt
am Beispiel Argen. Schweiz. Z. Hydrol. 32:453.
Uttormark, P. D., J. D. Chapin, and K. M. Green. 1974.
Estimating nutrient loadings from non-point sources. Ecol
Res. Ser. EPA-660/3-74-020. U.S. Environ. Prot. Agency.
Wagner, G. 1969. Kenngleichungen zur Ermittlung der
Belastung von Flussen mit Phosphor- und Stickstoffver-
bindungen. GWF 110:93.
Wagner, G. 1980. Discussion and application of guidelines for
water quality management. Pages 209-222 in Hj. Fricker,
ed. OECD Eutrophication Programme: Regional Project
Alpine Lakes, Bundesamt fur Umweltschutz, Bern (Schweiz).
Wagner, G., H. Buhrer, and H. Ambuhl. 1976. Die Belastung
des Bondensees mit Phosphor-Stickstoff- und organischen
Verbindungen im Seejahr 1971/72. Bericht Mr. 17 der
Internationalen Gewasserschutzkommission fur den Bo-
densee: 55 S.
Weyrauch, R. 1915. Hydraulisches Rechnen Verlan K
Wittwer, Stuttgart, 255 S. u N'
Zobrist, J., J. Davis, and H. R. Hegi. 1977. Charakterisierun
des chemischen Zustandes von Fliessgewassern r=.
Wasser-Abwasser 57:402. ' ^as-
-------
61
QUANTIFICATION OF PHOSPHORUS INPUT TO LAKES
AND ITS IMPACT ON TROPHIC CONDITIONS
RIAZ AHMED
RICHARD SCHILLER
The Center for the Environment and Man, Inc.
Hartford, Connecticut
ABSTRACT
A simple model for quantification of nonpoint sources of pollution for a watershed is presented.
Analyses of sampled data from the Lake Waramaug and Still River watersheds are used to validate
the accuracy of the model. It can also be applied to develop a cost-effective watershed control
management plan. A discussion of the role of decision analysis techniques in lake management is
presented in a conceptual form.
Under the Clean Water Act, Congress set an interim
(July 1983) national goal of achieving, wherever
attainable, water quality which provides for the
protection and propagation of fish, shellfish, and
wildlife, and provides for recreation in and on the
water. The 1977 Amendments of the Act emphasized
restoring lakes with potentials for significant recre-
ational usage. A recent report by the Council on
Environmental Quality (1979) indicates that 67 percent
of the lakes in the Nation may have serious water
quality problems. Under the Clean Lakes Program of
the Act, Federal assistance is available to develop and
implement lake managemant plans.
To provide an effective lake management plan, a
nutrient budget must be produced. Phosphorus, being
the most manageable, is usually the nutrient con-
sidered in lake management planning studies. To be
useful for planning purposes, the phosphorus budget
needs not only to detail the net inflows and outflows,
but also to identify sources, their location and
magnitude. These details are needed to develop a cost-
effective lake watershed management plan. The Clean
Lakes Program fully recognizes this need. The
diagnostic/feasibility studies that must be conducted
prior to the implementation of lake restoration
measures, require, in addition to other information, a
description of land use and an assessment of the role of
point and nonpoint sources of water pollution within
the watershed.
While quantification of point sources is a routine
procedure, the inherent temporal and spatial variability
in nonpoint sources presents serious obstacles in their
quantification. Difficulty in quantifying nonpoint
sources is probably one of the primary reasons for
failure of all but a few of the 208 studies to effectively
deal with the problem of nonpoint sources.
Sampling programs aimed at obtaining reliable
estimates of nonpoint sources could be very expensive,
especially if one is interested in identifying the location
and quantity of major sources within the watershed.
More sophisticated computer based models often
require significant quantities of data for calibration and
verification purposes. Using them on a routine basis is
often beyond the means and resources of a planning
agency. On the other extreme, simple areal loads, i.e.,
an average emission rate for each type of land use, do
not allow for adjustments reflecting the variability in
site specific values of causative factors in the
watershed.
Over the last few years, The Center for the
Environment and Man, Inc. has formulated a model for
quantification of phosphorus and other pollutants from
nonpoint sources to lakes and streams. To date, the
model has been used, in one form or other, in
developing lake management plans for 16 lakes in
Massachusetts and Connecticut. In addition, the model
is being used by the State of Connecticut to assess the
impact of nonpoint sources in 94 watersheds.
MODEL DESCRIPTION
The model developed by CEM for Computing Loading
Estimates from Nonpoint Sources in a watershed
(CLENS) is a simple model with data requirements
limited to those which are readily available, at least for
most of the Eastern States. The model differs from
areal load models as it allows for including site specific
information. In contrast to more sophisticated nonpoint
source models, its resource requirement is modest. The
model can be considered more comprehensive as it
includes computation of nonpoint pollution from
sources other than erosion. In all, seven specific
sources of nonpoint pollution are considered in CLENS.
These are:
LWashoff from urban areas.
2. Erosion from other areas.
S.Washoff from barnyards and feedlots.
4.Leachate from landfills.
S.Washoff from roadways.
6. Leachate from septic systems.
7. Wet and dry fallout.
-------
62
RESTORATION OF LAKES AND INLAND WATERS
The basic formulations for each of these sources
have been derived from several EPA sponsored studies
(Midwest Res. Inst. 1975; Heaney, et al. 1977;
Shaheen, 1975; U.S. EPA, 1976, 1 977) and studies by
the Agricultural Research Service (Wischmeier, 1960,
1976; and Soil Conserv. Serv. 1976). Specific
formulations and equations for each of the seven
sources have been described in detail elsewhere
(Ahmed, 1979; Ahmed and Schiller, 1980). CLENS is a
management type model; in that respect it does not
follow the standard protocol of modeling, i.e., calibra-
tion and verification prior to simulation. Instead, the
values of the parameters used m the model are
obtained from available data, reports, and studies.
The parameters utilized in the model can be placed
into two categories — parameters whose values are
valid for a region (one or more watersheds) and
parameters whose values are chosen for a specific
location in the watershed. Examples of regional
parameters are crop management factors, leachate
characteristics, and pollutant loadings at roadways.
Examples of parameters with site specific values are
population density, soil type, and traffic density. A
favorable comparison of nonpomt source loads com-
puted by CLENS and those estimated from field
programs at specific locations (as described later) in
Connecticut indicate that such an approach is feasible.
The overall purpose of the CLENS model is to provide
water quality and land use planners with a tool to
construct preliminary nonpoint source pollutant bud-
gets. Because the nonpoint source loads computed by
CLENS can be associated with specific areas of the
watershed, they provide the needed versatility to allow
planners to create cost-effective management plans.
In addition to quantification of nutrients, organics,
and sediments, CLENS can also be used for:
• Analysis of tradeoff between advanced waste-
water treatment and nonpoint source control.
• Evaluating the effectiveness of best management
practices.
• Locating and quantifying nonpoint sources from
"hot spots."
• Designing an efficient sampling program.
ASSESSMENT OF TROPHIC
CONDITIONS
An estimate of the expected quantities of phosphorus
loads alone does not indicate the severity of the water
quality problems in a lake. Development of lake
management plans must be based upon consideration
of the impact of phosphorus loads into a lake. Hence,
the estimate of annual phosphorus loads is translated
into expected average in-lake phosphorus concentra-
tions through use of lake models such as the one
proposed by Dillon and Rigler (1 974). The Dillon-Rigler
model can be expressed as:
The phosphorus retention coefficient can be com-
puted using the relationship developed by Dillon and
Kirchner (1975) which is:
V
RD=
1 RP
in which:
P = total annual phosphorus load to lake (g/yr).
Q - total annual outflow from the lake (m/yr).
[P] mean annual outflow phosphorus concentra-
tion (g/m).
RP = phosphorus retention coefficient.
(V + qs)
in which:
V = net settling velocity (m/yr).
qs = area water load (m/yr).
Based upon computed phosphorus concentration,
lakes may be categorized by anticipated trophic status.
APPLICATION OF CLENS TO LAKE WARA-
MAUG AND THE STILL RIVER BASIN
So far, CLENS has been used to study 16 lakes and
over 90 river watersheds. This paper presents the
results using it on Lake Waramaug and the Still River
Basin, both in Connecticut. These two examples have
been selected for presentation because a modest
amount of field data exists to assess the predictive
accuracy of the model.
For application of the CLENS model to watersheds in
Connecticut, the values of regional parameters were
determined principally from a review of soil loss studies
by the Soil Conservation Service and data available
from the lake quality monitoring program of the
Connecticut Department of Environmental Protection.
The regional parameters once developed for Con-
necticut were applied to the Lake Waramaug water-
shed and the Still River Basin without any further
modification. Watershed specific data such as popula-
tion density land use, soil type, slope, traffic density,
and other were obtained from available maps, reports
and studies. It is noteworthy that the application of
CLENS to large watersheds requires that watersheds
be subdivided into subunits of homogenous land use
and topographic conditions.
Lake Waramaug
Lake Waramaug is the second largest natural lake in
Connecticut. The lake has a surface area of 2.7 square
kilometers and a drainage area of 36.6 square
kilometers, land usage within the basin is distributed
as follows:
Agriculture 10 percent
Forest = 66 percent
Pasture = 4 percent
Urban = 5 percent
Wetlands/water bodies 13 percent
Recreation = 2 percent
Figure 1 shows Lake Waramaug and its watershed.
CLENS was used to develop estimates of phosphorus
loads from all sources within the watershed under
pristine, current, and year 2000 conditions. The results
are shown in Table 1.
From March 1977 through April 1978, the U.S.
Geological Survey in cooperation with the Lake
Waramaug Task Force and other local and State
agencies conducted an extensive watershed and in-
lake sampling program. The sampling data were
analyzed to create flow-flux curves which in turn were
used to develop estimates of net phosphorus and
sediment export from the watershed. To truly compare
the results obtained by the model and the sampling
program, loads from other sources, such as septic
tanks, were added to the loads estimated from th
sampling program. Another estimate of the total
-------
NUTRIENT LOADING/TROPHIC RESPONSE
63
1
Table 1. — Estimated annual phosphorus loads — Lake Waramaug.
Time Frame
and Source
Pristine Conditions
Erosion-Related
Atmosphere
Total
Subbasin
A
204
204
B C
142 95
142 95
Consequent Lake Condition2 =
Current Conditions
Erosion-Related
Atmosphere
Septic Systems
Landfills
Livestock
Motor Vehicles
Point Sources
Total
Year 2000 Conditions
Erosion-Related
Atmosphere
Septic Systems
Landfills
Livestock
Motor Vehicles
Point Sources
Total
445
5
31
7
488
Consequent
511
92"
31
7
641
Consequent
234 260
25 683
1 2
260 330
Lake Condition
326 416
32 1213
1 2
359 539
Lake Condition
D E
43 47
43 47
= High Oligotrophic
128 60
20
148 60
F Lake
84
130
84 130
(Condition 1.8)
260
267
21
29 405
1
31 1 307
Total Percent
615
130
745
1,387
267
139
0
100
11
0
1,904
83
17
100
73
14
7
0
5
1
0
100
= Mid-Eutrophic (Condition 3.1)
143 71
30
173 71
382
267
45
29 405
1
457 307
1,849
267
320
0
100
11
0
2,547
73
11
12
0
3
1
0
100
= Mid-Eutrophic (Condition 3.4)
'As computed by CEM composite land use analysis.
2Lake Condition: 1.0 - Very Oligotrophic
2.0 - Mesotrophic
3.0 - Eutrophic
4.0 - Hypereutrophic
Includes 7 kg/yr from The Casino.
'Includes 87 kg/yr from Hopkins Inn and The Inn.
5Represents estimated annual P load from waterfowl.
Table 2. — Comparison of phosphorus loads — Lake
Waramaug.
Computation Model
Computed Load
(kg P/yr)
USGS Sampling Program
Dillon-Rigler Model
CLENS Model
1,700
1,648
1,904
Table 3. — Comparison of phosphorus loads — Still River.
Study
Computed Load
(kg P/yr)
Frink's Study
NES Study
CLENS Model
61,820
70,000
66,770
phosphorus load was obtained from using the observed
outflow phosphorus concentration data in the Dillon-
Rigler model. Hydraulic characteristics of Lake Wara-
maug were used to determine the expected settling
velocity in the lake. Table 2 presents the loading
estimates for the three separate analyses.
The three loads are considered to compare favorably.
It is noted that the CLENS model estimates the long-
term average phosphorus loads while the other two
estimates provide an estimate specific to the year of
sampling.
Still River Basin
The Still River Basin is located in western
Connecticut. The total area of the basin is approxi-
mately 184 square kilometers. The watershed (Figure
2) is rather steep and at the same time is heavily
urbanized, with a population of 68,000 in the basin.
The land uses within the basin are as follows:
• Residential 32.5 percent
• Commercial = 3.1 percent
• Industrial = 4.5 percent
• Agricultural 7.4 percent
• Forest - 36.6 percent
• Wetlands/water bodies = 15.0 percent
• Other 0.9 percent
As part of the Connecticut 208 Program, CLENS was
applied to the Still River Basin to compute annual
nutrient and organic loadings.
The Still River is a tributary of the Housatonic River,
entering it at Lake Lillinonah. A phosphorus budget for
the Still River has been prepared by the U.S. EPA
National Eutrophication Survey. In addition, Dr. Charles
Frink of the Connecticut Agricultural Experiment
Station developed a phosphorus budget for the Still
River on the basis of data collected over a period of 1
year.
The total annual phosphorus loads for three separate
analyses are presented in Table 3. The three values are
comparable.
-------
64
RESTORATION OF LAKES AND INLAND WATERS
FORMULATION OF THE LAKE WATER-
SHED MANAGEMENT PLAN
CLENS can be used very successfully to analyze the
cost-effectiveness of various phosphorus control
measures in a lake watershed. The model has been
applied in this mode to several lakes in Berkshire
County, Mass. As an example, cost-effectiveness
analyses of several phosphorus control alternatives for
Lake Onota are shown in Table 4. Further details in
cost-effectiveness analysis are available in the Upper
Housatonic 208 Water Quality Management Plan
(Berkshire County, 1977).
ROLE OF DECISION ANALYSIS IN LAKE
MANAGEMENT PLANNING
To date, several lake models have been developed
and several of them have been reviewed by Reckhow
(1979). Most of the simple lake models share one basic
characteristic — they are empirical models derived
from statistical analyses of data for several lakes. While
the models have a statistical basis, they are determi-
nistic in the sense that they predict a fixed value of in-
lake concentration.
Most recently, emphasis has been placed on the
quantification of uncertainty associated with the
predictions of in-lake phosphorus concentration (Reck-
how, 1979; Reckhow and Chapra, 1979; Chapra and
Reckhow, 1979). Reckhow (1979) considers the error
term associated with phosphorus prediction to consist
of model error, parameter error, and loading error.
Reckhow and Chapra (1979) rightly point out that the
non-negligible portion of the sum of the model error
and parameter error may be due to phosphorus loading
uncertainty in the model development data. Chapra and
Reckhow (1979) analyzed the data from north
temperate lakes to establish probability curves for a
phosphorus prediction to fall within a particulartrophic
class.
From a planning point of view, the uncertainty in the
loading estimates and the variation of loading from
year to year are important. Review of sediment yield
studies would show that the year-to-year variation
could be significant. Hence, a decision about controls to
improve the quality of the lake on the basis of a year's
observation could be erroneous. Over the life of the
project, the quality of lake water may substantially
deviate from that expected in the planning analysis.
This paper presents an approach for handling the
variation loads over time.
For all practical purposes most of the lake models can
be expressed as follows:
1
(P)=—
V
in which:
/\ = hydraulic detention time (yr)..
V lake volume (m3).
For the sake of simplicity, if one assumes that the
values of and V are constant over time, then a
knowledge of the distribution of P could yield the
distribution of [P]. However, in almost all cases, it
would be difficult to identify the distribution of P.
unless one entered into a multi-year sampling
program. P, however, can also be expressed as follows:
in which:
Q, = annual inflow (m3/yr)
Pi = influent concentration (g/mj).
While not available directly, the distribution of P and
Q can be identified based upon analyzing data from
surrounding similar watersheds. It is noted that CLENS
provides a long-term average value of P If P and Q
distributions are known, the distribution of P can be
computed as follows:
F(P) =
rr r
J-oo I J-o
Pi
p,'
-)dPil dP
Once the distribution of P is defined, the trophic
status of the lake can also be defined in probabilistic
terms.
If Q is a time dependent variable then the
assumption of both and V as constant over time is
contradictory. In reality, /\ is a variable; however,
consideration of as a variable poses difficulties in
obtaining an analytical solution. Such a problem is,
however, amenable to simulation. A simple simulation
routine can be used to identify the distribution of the
lake's trophic status.
APPLICATION TO LAKE MANAGEMENT
Improving lake quality enhances its use and benefits.
It is reasonable to assume that the functional
relationship between utility and the trophic status of a
lake is not linear. In an actual planning setting, the
concerned parties (lake associations or lake study
committees) may produce this utility curve based on
the resident's specific needs and desires. A hypo-
thetical curve relating benefits and trophic index is
shown in Figure 3. Having defined the distributions of
the load and the trophic status, and the utility curve,
decision-aid techniques, such as a payoff or decision
tree, can be used to develop an optimum control plan.
For illustrative purposes, a payoff table is shown in
Table 5.
In Table 5, the value of benefits (b's) will be derived
from a utility function. Based on the preceding analysis,
a plan which maximizes the expected payoff should be
expected.
SUMMARY AND CONCLUSIONS
The results of the detailed sampling program in the
Lake Waramaug and Still River Basin watersheds
validate the accuracy of the CLENS model. In addition
to quantification of nonpoint sources, the model has
been applied to develop cost-effective watershed
management control alternatives. The model in its
initial application can also be used to establish an
effective field monitoring program.
Though the use of CLENS does not require a
computer, the computer program developed at CEM
substantially enhances the efficiency of its application
While additional research is needed, it is apparent
that the application of a statistical decision theory is
quite appropriate to lake management planning and
could provide support for the selection of n|a
providing maximum benefits over the long run
-------
NUTRIENT LOADING/TROPHIC RESPONSE
65
Table 4. — Cost effectiveness analysis of lake watershed control measures.
Lake
On ota
Control Measure
1. Do nothing
2. Manage manure
3. Maintain catch basins
4. Manage crops per SCS
5. Sewer
6. Control construction practices
7. Build detention pond in Subbasin P
Effectiveness
(kg P/yr)
—
56
35
17
236
30
34
Cost
($1,000)
—
1.4
1.2
1.5
62.0
18.9
30.0
C/E
($/kg P/yr)
—
25
33
8B
260
630
880
Remaining P
(kg P/yr)
1,269
1,213
1,178
1,161
925
895
861
Lake
Condition
2.7
2.6
2.6
2.5
2.3
2.3
2.2
Recommended Program
408
115.0
280
861
2.2
Alternative Onota Program (without sewers):
1. Do nothing
2. Use nonphosphorus detergents
3. Manage manure
4. Maintain catch basins
5. Manage crops per SCS
6. Manage septic systems
7. Control construction practices
8. Build detention pond in Subbasin P
Alternative Program
'9. Sewer
—
84
56
35
17
93
30
34
349
59
—
0.8
1.4
1.2
1.5
8.9
18.9
30.0
62.7
62.0
—
10
25
33
88
96
630
880
180
1,050
1,269
1,185
1,129
1,094
1,077
984
954
920
920
2.7
2.6
2.6
2.5
2.5
2.4
2.3
2.3
2.3
* Not recomended because of poor cost effectiveness or insignificant change in lake condition.
Notesl- Each control measure is evaluated as if the measures above it were operating. For example, the effectiveness of non-recommended sewering
in this table is greatly reduced because much of the phosphorus will have been removed by the use of nonphosphorus detergents and the
management of septic systems.
2. Lake Conditions: 1.0-1.9 = oligotrophic; 2.0-2.9 = mesotrophic; 3.0 and higher = eutrophic.
Table 5. — Payoff table for lake management plans.
Lake Status
Phosphorus
Contamination
<(Pi)
(P.HPa)
(PsMPs)
>(P3)
Expected Payoff
Probability
P,
Pi
Ps
P.
Control Plan
1
b,,
b2i
b3,
b«
ZP.hi
Control Plan
2
b,2
b22
b32
b42
ZP,b,2
Control Plan
3
b,3
b23
b33
D43
IP,b,3
STILL
RIVER
BASIN
Figure 1. — Lake Waramaug watershed.
Figure 2. — Still River watershed.
-------
66 RESTORATION OF LAKES AND INLAND WATERS
REFERENCES
Ahmed, R. 1979. A manual for assessment of nonpoint
sources of water pollution. CEM Rep. No. 4256-644. The
Center for the Environment and Man, Inc., Hartford, Conn.
Ahmed, R., and R. Schiller. 1980. Nonpoint source
quantification and its role in lake and stream water quality
planning. Proc. 10th Int. Conf. Int. Assoc. Water Pollut. Res.
Pergamon Press, New York.
Berkshire County Regional Planning Commission. 1977.
Upper Housatonic 208 water quality management plan.
Pittsfield, Mass.
Chapra, S. C., and K. H. Reckhow. 1979. Expressing the
phosphorus loading concept in probabilistic terms. Jour.
Fish. Res. Board Can. 36.
Council on Environmental Quality. 1979. Environmental
quality—1979. 10th annual report. U.S. Government
Printing Office, Washington, D.C.
Dillon, P. J. and F H. Rigler. 1974. A test of a model for
calculating the concentration of phosphorus in lake water.
Jour. Fish. Res Board Can. 31.
Heaney, J. P., et al. 1977. Nationwide evaluation of combined
sewer overflows and urban stormwater discharges. EPA
Rep. No. 600/2-77-064. U.S. Environ. Prot. Agency,
Cincinnati, Ohio.
Midwest Research Institute. 1976. Loading functions for
assessment of water pollution from nonpoint sources. EPA
Rep. No. 600/2-76-151. U.S. Environ. Prot. Agency,
Washington, D.C.
Reckhow, K. H. 1979. Quantitative techniques for the
assessment of lake quality EPA Rep. No. 440/5-78. U.S.
Environ. Prot. Agency. Washington, D.C.
Reckhow, K. H., and S. C. Chapra. 1979. A note on error
analysis for a phosphorus retention model. Water Resour.
Res. 15.
Shaheen, D. B. 1975. Contributions of urban roadway usage
to water pollution. EPA Rep. No. 600/2-74-004. U.S.
Environ. Prot. Agency, Washington, D.C.
Soil Conservation Service. 1976. Erosion and sediment
control handbook. Storrs, Conn.
U.S. Environmental Protection Agency. 1976. Areawide
assessment procedures manual. Municipal Environ. Res.
Lab. Cincinnati, Ohio.
1977. 1976 needs survey. EPA Rep. No. MCD-
48E. Washington, D.C.
Wischmeier, W. H. 1976. Use and misuse of the universal soil
loss equation. Jour. Soil Water Conserv.
Wischmeier, W. H., and D. D. Smith. 1965. Predicting rainfall-
erosion losses from cropland east of the Rocky Mountains.
Agric. Handbook No. 252. Agric. Res. Serv.
ACKNOWLEDGMENTS
The authors wish to acknowledge the contribution of
William V. McGumness, Jr., Adjunct Senior Research
Scientist at The Center for the Environment and Man, Inc.,
in the initial formulation of this model and its application to
lake management planning.
-------
67
WHATEVER BECAME OF SHAGAWA LAKE?
DAVID P. LARSEN
KENNETH W. MALUEG
Corvallis Environmental Research Laboratory
U.S. Environmental Protection Agency
Corvallis, Oregon
ABSTRACT
The response of Shagawa Lake, Minn, to an 80 percent reduction in external phosphorus loading,
initiated in 1973 when a tertiary waste treatment plant began operation, is summarized. Total
phosphorus concentration of the treatment plant effluent was 50 /jg/\ from 1973-1977. In
November 1977 the level was raised to 400 fjg/\ This change produced a300 /ug/lyr increase in
total phosphorus loading to the lake from the treatment plant. Wastewater loading of total
phosphorus accounts for 20 to 30 percent of the total external loading to the lake, decreased from
80 percent during pre-treatment years. Through 1978 the post-treatment response of the lake was
stable with total phosphorus concentrations (average whole lake) reduced to about 60 percent of
the pre-treatment levels (a reduction of ~ 40 percent), chlorophyll a (epilimnion) only slightly
reduced, and Secchi disk depth increased slightly. The internal phosphorus loading phenomenon
which has prevented complete recovery of the lake might be diminishing. Relationships between
chlorophyll a and total phosphorus and between Secchi disk depth and chlorophyll a in Shagawa
Lake are consistent with those established for other lakes.
INTRODUCTION
The response of Shagawa Lake, Minn, to a
significant reduction in external phosphorus (P) loading
provides a clear example of how internal P supplies can
dely the recovery of heavily eutrophied lakes. Examples
exist to document the predictable recovery of lakes
when external P is reduced, based on predictions of P
washout models of the type described by Sonzogni, et
al. (1976), for example, Edmondson (1977) for Lake
Washington, and Dillon, et al. (1978) for Gravenhurst
Bay, Ontario. However, Shagawa Lake's response
provides a case study in which these simple models are
not adequate unless they are modified to account for
internal loading.
Lake Norrviken, Sweden (Ahlgren, 1977), provides
an example of an intermediate response — internal
loading is important but baseline P levels decreased in
accordance with washout model projections. The
whole lake fertilization experiments which Schindler
and co-workers (1974, and Schindler and Fee, 1975)
conducted show no evidence of significant internal
loading sojf fertilization were halted, these lakes would
be expected to respond as predicted by P washout
models. Thus, the study of the long-term response of
Shagawa Lake provides insight into the characteristic
response of a lake toward one end of a continuum, the
other end being demonstrated by lakes that respond as
predicted by P washout models.
This paper will: (1) Extend observations covering
Shagawa's response through 1976; and (2) compare
the relationships between chlorophyll a, total phos-
phorus (TP), and Secchi disk depth in Shagawa Lake
with those established for other relevant lake studies.
BACKGROUND
Bradbury (1978) described the development of
European settlements around Shagawa Lake begin-
ning in the late 1800's with the advent of mining and
lumbering in this region of northeastern Minnesota. He
showed how the sediments of the lake recorded this
development through changes in chemical character-
istics (especially P and iron), pollen types, and diatom
and cladoceran remains. These changes document the
lake's transformation from a relatively unproductive
system to one of high productivity as population
pressure increased in the watershed. Malueg, et al.
(1975) demonstrated that wastewater phosphorus
loading from the nearby community of Ely contributed
80 percent of the external TP loading in the late 1960's
and early 1970's. Others have documented the high
levels of algal productivity and biomass in the lake
resulting from wastewater enrichment (Megard and
Smith, 1974; Larsen, et al. 1975).
In 1973, a treatment plant which eliminated
essentially all the wastewater P flowing to the lake
began operating in Ely. The plant reduced the total
external input from 6,200 to 7,200 kg/yr to 900 to
1,500 kg/yr, sufficient to reduce the average influent P
concentration from 60 - 100/ug/liter to less than 20
/ug/l. Since Shagawa's water retention time is short
(less than 1 year) and the pre-treatment P retention
time is even shorter (Vi year), the lake could be
expected to respond rapidly to reduced external P
loading.
Larsen, et al. (1979) documented that the lake did
indeed respond rapidly but this response was tempered
by a resurgence of P from the sediments, especially
during July and August. When this resurgence pattern
-------
68
RESTORATION OF LAKES AND INLAND WATERS
was incorporated into a P mass balance model, the
model tracked the temporal pattern well and mimicked
the average values closely, but the average TP levels in
the lake were well above those predicted from a simple
TP washout model. One conclusion based on these
simulations was that the TP resurgence had not
diminished (through 1976) since the wastewaster
treatment process began. About 2,000 to 2,500 kg of
TP are released during the 2-month July through
August interval.
METHODS
The sampling methods used for the lake and
tributaries through September 1978 and the analytical
procedures have been described elsewhere (Malueg, et
al. 1975). From October 1976 to October 1978, the U.S.
Forest Service at Ely collected and analyzed the
samples under the direction of the U.S. Environmental
Protection Agency. Only the central station, Brisson's
Point South was sampled routinely. Samples were
taken bi-weekly through June, weekly during July and
August, then bi-weekly. All averages were time
weighted. During 1979, there was no routine sampling
program; however, Secchi disk data were obtained
from S. Kliest (pers. comm. Vermilion Community
College). During 1980, a limited sampling program
(conducted by Vermilion Community College) included
approximately weekly Secchi disk measurements at
three stations and collection of an integrated water
sample using a 5 m length of tygon tubing. Water
samples were preserved with HgCb (40 mg/l final
concentration) and sent to the EPA's laboratory in
Corvallis, Ore. for TP analysis. These observations
cover the period from mid-May through the end of
August.
For the Secchi disk, chlorophyll a, and TP regres-
sions, we used data from the upper 5 m at Brisson's
Point South for the period May-September, because
this station provides the longest continuous record of
values for these variables. For consistency with other
studies, we chose the May-September interval as
representative of summer conditions (e.g., Dillon and
Rigler, 1974; Dillon, et al. 1978; Ahlgren, 1980).
Routine weekly tributary sampling ended in Sep-
tember 1978, so from that time, natural loadings were
estimated by multiplying approximate total inflow by
the average inflow TP concentration of all natural
sources for the years 1969-1978 (see Larseh, et al.
1979). The representative value is 15.6 M9 - TP/I-
Estimated inflow came from a regression of measured
annual inflows (1969-1977) against measured precipi-
tation at the Winton Power Dam weather station, 8 km
east of Shagawa Lake. The regression equation is: F =
1.39 PR - 18.8(r2 =0.77), where F is flow (x 106 mVyr)
and PR is precipitation (cm/yr). To obtain wastewater
loadings, treatment plant operators measured plant
flows daily and phosphorus concentrations approxi-
mately every other day; loadings were calculated from
these values (Jackson and Lindroos, 1980).
RESULTS
1. Changes in TP Supply — During 1977, the
external input of TP into Shagawa Lake remained
similar to that for the post-treatment years, 1974-1976
(Table 1). In November 1977, the chemical treatment
process for P removal within the plant was modified for
economic reasons to produce an effluent concentration
of 400 g/liter, an increase from 50 jug/lforthe interval
from April 1973 to November 1977. This change in
treatment increased the wastewater TP input by about
300 kg/yr to 400 kg/yr. In 1978, wastewater supplied
330 kg TP through September (Table 1) and an
additional 95 kg from October through December
(Jackson, pers. comm.). During 1979, wastewater
supplied 330 kg (Table 1). A small unmeasured amount
of wastewater bypassed the plant during 1978-1979.
In previous years, this bypass has contributed less than
20 kg/yr TP to Shagawa Lake, so we expect that bypass
was not significant during 1978 and 1979.
Since the annual natural supply ranged from 850 to
1,760 kg/yr, wastewater now accounts for 20 to 30
percent of the total supply of P to the lake. From 1977-
1978, average inflow TP concentrations to the lake
incorporating all sources were similar to the values for
the 1974-1976 interval. These results indicate that for
the period 1977-1979, TP loading to Shagawa Lake
was similar to that for the 1974-1976 interval. The
increase in annual TP loading associated with the
revised wastewater effluent standard is masked by the
natural variation in TP loadings.
Table 1. — Total phosphorus supplies to Shagawa Lake, Minn., 1977-1979.
Natural Inflow
Year
1969-1972
1973-1976
1977
1978
(through
Sept.)
1979
Flow x
106 m3/yr
61.1-125.1
41.5-102.0
87.7
65.5
+71.4
Cone.
(Aig/i)
14.1-17.3
14.7-20.5
13.4
12.6
+15.6
Supply
(kg)
1,060-1,760
850-1,630
1,180
830
1,110
Wastewater Inflow
Flow x
106 m3/yr
1.2-2.1
1.3-1.7
1.5
1.10
1.2
Cone.
(/"9/I)
2,600-4,200
'29-80
70
340
290
Supply
(kg)
5,040-5,460
'40-130
110
330
350
Combined Inflow
Flow x
106 m3/yr
62.3-127.2
42.8-103.7
89.2
66.6
72.6
Cone.
(A<9/l)
58-100
'15.3-20.7
14.4
17.4
20.0
Supply
(kg)
6,230-7 200
'890-1 490
1 290
1 160
1,460
' Excludes average values for 1973 when wastewater treatment plant was not in operation for full year.
+• Estimated as described in text
-------
NUTRIENT LOADING/TROPHIC RESPONSE
69
2. Changes in TP, chlorophyll a, Secchi disk depth
— The stable summer pattern of lake TP which
developed in response to wastewater P reduction
(Larsen, et al. 1979) continued through September
1978, when routine sampling of the lake was
terminated (Figure 1). The internal loading event,
identified previously, continued through 1978 (Larsen,
et al. 1979, 1980). During 1977, the average TP
concentration of the lake increased by approximately
50 HQ/\ from late May through August, corresponding
to an increase of 2,650 kg TP within the lake supplied
from internal sources. Similar increases were seen in
previous years. The 1977 pulse apparently occurred
about 1 month earlier than usual.
In other years, P concentration began to increase
throughout the lake in late June and early July, then
declined in late August and early September. In 1977,
P concentration began to increase in late May and early
June and declined in August. During 1978, average TP
increased by approximately 30 jug/I corresponding to a
lakewide increase of 1,600 kg. This is substantially
lower than that seen in previous years. The average TP
concentration in Shagawa Lake during 1977 and 1978
has remained similar to that measured during 1974-
1976, indicating a reduction of approximately 40
percent over the pre-treatment period. (Table 2). This
response is also seen in the immediate reduction and
subsequent stability in average epilimnetic TP at
Brisson's Point South for the May-September interval
(Figure 2).
During June-August 1980, TP concentrations in-
creased in the upper 5 m from 15.6 /ug/l in early June
to about 45 /ug/l in August, corresponding to an
increase of 1,150 kg (Figure 3). This increase is not
directly comparable with the whole-lake increases
reported previously; however, its magnitude is similar
to that in 1978. The magnitude of these pulses in 1978
and 1980 is lower than that seen previously and may
signal a reduction in the internal loading in Shagawa
Lake. However, during both years, temperature profiles
indicated that the lake was not severely stratified;
temperatures in July and August near the bottom were
only slightly lower than surface values. Oxygen
depletion occurred, but its extent over the profundal
zone was probably minimal due to the vertical mixing
inferred from temperature profiles. Thus meterological
conditions may have modified the release pattern seen
in these years, while the potential for release may not
yet have diminished.
In Shagawa Lake, a summer algal bloom typically
develops in late June or early July and terminates in
late August or early September, although some
variation occurs in the timing. In the years since
treatment began, the duration of the bloom has been
shorter. During 1977 and 1978, this general pattern
continued to occur (Figure 4). In July and August,
chlorophyll a was similar to that seen in 1974-1976,
nearly reaching pre-treatment concentrations (Table 2).
There has been a substantial decrease in chlorophyll a
during post-treatment years, but the impact is not as
significant or as stable as that seen in the post-
treatment TP pattern (Figure 2).
Changes in mean Secchi disk depth for May and
June, July and August, and May-September at
Brisson's Point South are summarized in Table 3. Most
evident is the increased clarity during May and June
after treatment began, corresponding to the associated
reduction in algal biomass. Based on Secchi depth, the
mean transparency increased by 0.8 m as compared to
the pre-treatment period. When the 1978 value, which
is similar to pretreatment values, is excluded, clarity
increased 1 • m in May and June. Later (July and
August) no post-treatment improvement in Secchi
depth is seen because algal biomass was not reduced
significantly during this time period. Over the May-
September interval, transparency increased by 0.5 m
(0.7 m if the 1978 value is excluded) (Table 3 and
Figure 2).
< 20
"I 1 I T
I J I A I
TIME (monlhs)
Figure 1 . — Average total phosphorus concentrations in
Shagawa Lake, 1977 and 1978, for the ice-free season. The
values are whole lake, volume weighted.
Table 2. —Summary of changes in total phosphorus (whole lake) and chlorophyll a (upper 5m) in Shagawa Lake, Minn. Numbers
in parentheses are the ratios of the average concentrations of any 1 year to the mean of the 1971 and 1972 average
concentrations.
Year
•1971-1973
•1974-1976
1977
1978
(through
Sept.)
Annual
Average
47.4-54.1
29.3-31.4
33.6(0.66)
28.3(0.56)
Total Phosphorus (jjg/\)
Ice-Covered
Interval
36.5-51.4
19.4-24.6
23.8(0.60)
19.6(0.49)
Ice-Free
Interval
50.8-60.9
34.6-35.7
38.8(0.67)
33.4(0.58)
Chlorophyll
May-June
14.1-16.1
6.4-11.4
7.2(0.46)
14.0(0.90)
a (/ug/l)
July-August
23.0-32.9
15.5-33.4
20.7(0.74)
26.0(0.93)
'Values are ranges over the years shown from Larsen, et al. (1979).
-------
70
RESTORATION OF LAKES AND INLAND WATERS
5
8
LJ
CO
20-
10
°l 30
_l
i!"
i
o
»
0-
Figure 5. — Chlorophyll a — (Figure 4b) relationships for the
Shagawa Lake — Burntside Lake combination. Darkened
circles are Burntside Lake, open circles are Shagawa Lake.
1971 1972 1973
1974 1975
YEAR
1976 1977 1978
Figure 2. — Average summer (mid-way to mid-September) tola I
Figure 5a.
10 20 30 40 50 60
TOTAL PHOSPHORUS (pg/l)
the central station, Brisson's Point South. Values are water
column averages over the top 5m. 07
„ 06
no
\ BU
CP
a.
*•"*
co 60
cr
o
0- 40
CO
(-)
I
Q_
20
_l
<
t—
P 0
1 !
A
o EEDH
A BPS
.
811
0
- 0 o J A A
* A A Q A A
— A * A O
°
1 1
0.5
' — •
1
E
" 0.4
-*-
u
UJ 0.3
co
H ° M ' J 1 J A x.
TIME (month) —
0.2
Figure 3. — Total phosphorus concentrations at three stations
in Shagawa Lake, 1980. Values are for the upper 5 m.
0, 1
60
^ 50
CT
3.
1 ^0
°l
i! 30
OROPHY
o
5! 10
U
O
| | | j | | - j —
4
/I .
— , ' A —
A ' ' v~\
- ^ /Iki,
"x;>Nj>^ ^^^^
1 1 1 1 1
B
o SHAGAWA -,,
_ • BURNTSIDE 78' P _
o /
,/
,s
71 s
— s° —
^r 73
/
jS O
o s 75
o
— .
— ,. _
•^
1 /SD 0 141 + O.OI67CA
- _
1 III
0 5 10 15 20 25 3
CHLOROPHYLL o_ (ug/l)
Figure 5b.
RELATIONSHIPS AMONG SD, CA, AND
TP
TIME (months)
Figure 4. — Average chlorophyll a concentrations in Shagawa
Lake, 1977 and 1978, for the ice-free season. The values are
for the epilimnion (top 5 me), volume weighted
riyuie o buillii idi izeb UMIUI upiiyn a, i r, ana beCCnl
depth chlorophyll a relationships for the Shagawa
Lake - Burntside Lake combination. Burntside Lake lies
10 river km upstream of Shagawa Lake. Its outlet the
Burntside River, accounts for 60 to 70 percent of
Shagawa's water inflow. We chose to include the
Burntside Lake TP, chlorophyll a, and Secchi depth
-------
NUTRIENT LOADING/TROPHIC RESPONSE
71
values because we believe that further declines in
Shagawa's TP concentration will be accompanied by
changes in chlorophyll a and Secchi depth which follow
the regression lines. The regressions display signifi-
cant linear relationships among the variables (Table 4,
~ 0.05). The chlorophyll a TP relationship suggests
that 1 fjg of TP produces 0.5 /.ig of chlorophyll a,
regardless of whether Burntside Lake values are
included. The intercepts do not differ from zero. The
Secchi depth chlorophyll a relationship for the
Shagawa Burntside Lake combination implies a
background SD of 7.1 m. For Shagawa Lake alone, the
value is slightly lower and is not statistically different
from zero (a = 0.05).
Reciprocal Secchi depth is chosen as an independent
variable to be consistent with the arguments developed
in Lorenzen (1980) and Megard, et al. (1980).
Reciprocal Secchi depth can be expected to change as a
linear function of chlorophyll a as:
1 /SD = a + /3 CA
where a is the reciprocal of background Secchi depth
and /* is related to the partial extinction of light by
chlorophyll a:/3 = Kc/ln (lo/lz)where I0 and lz are surface
irradiance and irradiance at the Secchi disk depth, and
Kc is the partial attenuation of irradiance due to
chlorophyll a.
DISCUSSION
Post-treatment chlorophyll a and TP data obtained
since 1976 in Shagawa Lake are consistent with those
obtained for the years immediately after P loading was
reduced. The post-treatment TP data show that
Shagawa Lake reached an equilibrium in response to
reduced P loading rapidly and that the lake has
remained at this level through 1978, as indicated by
the low year to year differences in mean (Figure 2). The
responses of chlorophyll a and Secchi depth have not
been as clear, although post-treatment values are
lower than those for pre-treatment over the same
intervals (Tables 2 and 3).
Average TP in Shagawa declined in response to
loading reduction, but not to the extent predicted from
P washout models because internal sources have
supplied significant amounts of P (Larsen, et al. 1979,
1980). This internal source is probably the sediments of
the profundal plain and deep-hole areas of the lake
which release P after anaerobic conditions have
developed (Armstrong and Stauffer, 1980). The
magnitude of this internal supply does not appear to
have diminished over the post-treatment period
through 1977. This post-treatment pattern contrasts
with that seen in Norrviken in which internal P loading
declined significantly through time (Ahlgren, 1977).
The data obtained during 1978 and 1980 indicate that
the magnitude of internal loading may be declining.
Although Shagawa Lake has not responded accord-
ing to projections from TP washout models, relation-
ships among Secchi depth, chlorophyll a, and TP are
consistent with those seen for other lakes which have
shown significant recoveries after extended TP loading
reduction, as demonstrated bv data from Lakes
Washington and Norrviken, and Gravenhurst Bay.
Table 5 summarizes the relationships for these studies.
These cases were selected for comparison with
Shagawa Lake because P loading was reduced sharply
and because many years' data are available which
document the pre-treatment conditions and the
recovery patterns for each lake. The data for
Washington were obtained from Edmondson (1977)
and Smith and Shapiro (1980), for Gravenhurst Bay
from Dillon, et al. (1978), and for Norrviken from
Ahlgren (1980, and pers. comm.). The 1971 chlorophyll
a TP data pair for Norrviken was excluded from the
chlorophyll a TP regression because chlorophyll a is
an obvious outlier (see Ahlgren, 1978).
The relationships summarized in Table 4 for
Shagawa Lake and Table 5 for Lakes Washington and
Norrviken and Gravenhurst Bay indicate that Shag-
awa's response falls within the range of values
characteristic of these relationships for the other
cases. The slope of the chlorophyll a - TP relationship
for Shagawa Lake falls between that for Lakes
Washington and Norrviken while the slope of the
Secchi disk depth chlorophyll a relationship is lower
than those for Lake Washington and Gravenhurst Bay
but higher than that for Norrviken. This indicates that
although the Shagawa Lake TP response to loading
reduction was unique, the control exhibited by TP on
chlorophyll a on Secchi depth is similar to that for other
lakes and that further declines in lake TP can be
expected to produce further declines in chlorophyll a
and increases in transparency consistent with that
seen at other sites.
In summary, Shagawa Lake continues to display a
stable pattern in total phosphorus concentration which
was reached rapidly after external phosphorus input
was reduced by wastewater treatment. The average
total phosphorus concentration and seasonal patterns
continue to be controlled by internal loading during
summer months. Chlorophyll a and Secchi disk
transparency have changed only slightly since treat-
ment began. We now know that, in Shagawa Lake,
chlorophyll a responds to total phosphorus and Secchi
disk depth to chlorophyll a in a manner similar to that
seen for other lakes. Thus, further changes in algal
biomass and transparency are expected only if total
phosphorus declines further.
Table 3. — Summary of Secchi disk depth (meters) in
Shagawa Lake, Minn, for the period 1971-1980 at Brisson's
Point South. Numbers in parentheses are the ratios of the
average values for a particular year to the mean of the 1971-
1972 average values..
May-June
1971
1972
1973
1971-1973
1974
1975
1976
1977
1978
1979
1980
1974-1980
2.08
1.98
2.28
mean 2.11
2.74
3.19
3.31
3.24
2.07
2.94
mean 2.92
(1.02)
(0.98)
(1.12)
(1.04)
(1.35)
(1.57)
(1.63)
(1.60)
(1.02)
(1.45)
(1.44)
July-August
2.
1.
1.
1.
2.
1.
1.
2.
1.
1.
1.
1.
13
,44
,97
,85
,70
,77
,71
,09
36
,61
,62
,88
(1.20)
(0.81)
(1.11)
(1.04)
(1.52)
(1.10)
(0.96)
(1.17)
(0.76)
(0.91)
(0.91)
(1.05)
May-September
1.95
1.63
2.05
1.88
2.81
2.50
2.42
2.63
1.68
2.40
(1.09)
(0.91)
(1.14)
(1.05)
(1.57)
(1.40)
(1.35)
(1.45)
(0.93)
(1.34)
-------
72
RESTORATION OF LAKES AND INLAND WATERS
Table 4. — Relationships among Secchi disk depth chlorophyll a and total phosphorus in Shagawa and Burntside Lakes, Minn.
Slopes and intercepts are given with 95 percent confidence limits, r2 is the linear correlation coefficient, and n is the number of
data pairs.
Chlorophyll-a vs TP
Shagawa -
Shagawa
Secchi disk
Shagawa -
Shagawa
Burntside
depth vs chlorophyll a
- Burntside
-1.74 ± 3.58
1.17 ± 10.12
0.141 ± 0.035
0.185+ 0.223
0.529 +0.121
0.460 ±0.100
0.0167 + 0.002
0.0146 + 0.011
13
8
13
8
0.89
0.44
0.92
0.63
Table 5. — Relationships among Secchi disk depth chlorophyll a and total phosphorus in Lakes Washington and Norrviken and
in Gravenhurst Bay. Slopes and intercepts are given with 95 percent confidence limits, r2 is the linear correlation coefficient, and
n is the number of data pairs.
Chlorophyll a vs TP
Gravenhurst Bay
Norrviken
Washington
Secchi disk depth vs chlorophyll a
Gravenhurst
Norrviken
Washington
-1.61 ± 6.98
13.30 + 22.60
-4.37 ± 3.92
0.187+ 0.137
0.505 + 0.321
0.204 + 0.073
0.267 ±0.141
0.406 +0.138
0.597 ±0.110
0.0217 + 0.016
0.0107 + 0.004
0.0235 + 0.004
7
10
15
7
10
18
0.72
0.85
0.91
0.71
0.79
0.91
REFERENCES
Ahlgren, I. 1977. Role of sediments in the process of recovery
of a eutrophicated lake. Pages 372-377 in H. L. Golterman,
ed. Interactions between sediments and fresh water. Dr. W.
Junk B. V. Publishers.
1978. Response of Lake Norrviken to reduced
nutrient loading. Verh. Int. Verein. Limnol. 20:846.
1980. A dilution model applied to a system of
shallow eutrophic lakes after diversion of sewage effluents.
Arch. Hydrobiol. (In Press.)
Armstrong, D. E., and R. E. Stauffer. 1980. Internal loading in
Shagawa Lake. Ecol. Res. Ser. Rep. U.S. Environ. Prot.
Agency, Corvallis, Ore. (In prep.)
Bradbury, J. P. 1978. A paleolimnological comparison of
Burntside and Shagawa Lakes, northeastern Minnesota.
EPA-600/3-78-004. U.S. Environ. Prot. Agency, Corvallis,
Ore.
Dillon, P. J., and F. H. Rigler. 1974. The phosphorus-
chlorophyll relationship in lakes. Limnol. Oceanogr. 19:767.
Dillon, P. J., K. H. Nicholls, and G. W. Robinson. 1978.
Phosphorus removal at Gravenhurst Bay, Ontario. An 8-
year study on water quality changes. Verh. Int. Verein.
Limnol. 20:263.
Edmondson, W. T. 1977. Trophic equilibrium of Lake
Washington. EPA-600/3-77-087. U.S. Environ. Prot. Agen-
cy, Corvallis, Ore.
Jackson, T. C., and G. Lindroos. 1980. 1979 annual report:
Operation and performance of wastewater treatment
facility, Ely, Minnesota. SERCO Lab., Ely, Minn.
Larsen, D. P., D. W. Schults, and K. W. Malueg. 1980.
Summer internal phosphorus supplies in Shagawa Lake,
Minnesota. (Manuscript).
Larsen, D. P., et al. 1975. Response of eutrophic Shagawa
Lake, Minnesota, U.S.A., to point source, phosphorus
reduction. Verh. Int. Verein. Limnol. 19:884.
1979. The effect of wastewater phosphorus
removal on Shagawa Lake, Minnesota: Phosphorus sup-
plies, lake phosphorus and chlorophyll a. Water Res
13:1259.
Malueg, K. W., et al. 1975. A six-year water, phosphorus, and
nitrogen budget for Shagawa Lake, Minnesota. Jour.
Environ. Qual. 4:236.
Megard, R. O., and P. D. Smith. 1974. Mechanisms that
regulate growth rates of phytoplankton in Shagawa Lake,
Minnesota. Limnol. Oceanogr. 19:279.
Megard, R. O., et al. 1980. Light, Secchi disks, and trophic
states. Limnol. Oceanogr. 25:373.
Schindler, D. W. 1974. Eutrophication and recovery in
experimental lakes: Implications for lake management.
Science 184:897.
Schindler, D. W., and E. J. Fee. 1974. Experimental lakes and
whole-lake experiments in eutrophication. Jour. Fish. Res.
Board Can. 31:937.
Schindler, D. W., E. J. Fee, and T. Ruszczyniski. 1978.
Phosphorus input and its consequences for phytoplankton
standing crop and production in the Experimental Lakes
Area and in similar lakes. Jour. Fish. Res. Board Can.
35:190.
Smith, V. H., and J. Shapiro. 1980. Chlorophyll-phosphorus
relations in individual lakes: Their importance to lake
restoration strategies. (Manuscript.)
Sonzogni, W. G., P. C. Uttormark, and G. F. Lee. 1976. A
phosphorus residence time model: Theory and application.
Water Res. 10:429.
Lorenzen, M. W. 1980. Use of chlorophyll-Secchi disk
relationships. Limnol. Oceanogr. 25:371.
-------
NUTRIENT LOADING/TROPHIC RESPONSE
73
A RETROSPECTIVE LOOK AT THE EFFECTS OF
PHOSPHORUS REMOVAL IN LAKES
VAL H. SMITH
JOSEPH SHAPIRO
Limnological Research Center
University of Minnesota
Minneapolis, Minnesota
ABSTRACT
A retrospective look at 16 north temperate lakes which have undergone restoration shows that the
reductions in chlorophyll a which accompanied phosphorus removal were typically immediate and
continuous. However, the exact response of each lake to P removal was unique. It is suggested
that these differences in response result to a large extent from changes in TN:TP which accompany
restoration. A variable chlorophyll yield model, which depends explicitly on the TN:TP ratio, is
presented and tested using data from Lake Norrviken (Sweden). The new model appears to greatly
reduce chlorophyll prediction error in lakes which are undergoing restoration.
INTRODUCTION
The prediction of algal biomass in lakes undergoing
changes in nutrient loading is a topic of great concern
for lake management, and during the last decade a
number of eutrophication models have been developed
for this purpose. In this regard, the Dillon-Rigler (1975)
and Vollenweider (1976) models have been widely
used to justify phosphorus control as a lake restoration
measure. However, with notable exceptions (e.g.
Dunst, et al., 1974; Ryding and Forsberg, 1975; Born,
1979), few systematic evaluations have been made of
the response of a large number of lakes to restoration
measures. With this in mind. Smith and Shapiro (1980)
have analyzed the response of algal biomass to
successful phosphorus reduction in 16 lakes, using
data from the literature. The purpose of this paper is to
summarize the important features of that analysis, and
to present a model which helps explain a major portion
of the variance associated with chlorophyll-phosphorus
regressions noted in that and other studies.
VARIABILITY IN THE RESPONSE OF
LAKES TO PHOSPHORUS REMOVAL
We have analyzed the changes in algal biomass in 16
north temperate lakes where nutrient abatement, or
natural variation in phosphorus loading, has led to
measurable reductions in concentrations of total
phosphorus in the lake. Growing season mean values
of chlorophyll a (c), total P (TP), and total N (TN) for
these lakes are summarized in Smith and Shapiro
(1980).
When the data were examined using standard
regression and correlation techniques (Steel and
Torrie, 1960), significant regressions between (c) and
TP were found for nine of the 16 lakes. The chlorophyll-
phosphorus relationships for two of these lakes are
shown in Figure 1, in which a significant regression is
evident for Norrviken (Figure 1b), but not for
Oxundasjon (Figure 1a). An important aspect of the
analysis thus isjhat the response of individual lakes to
a reduction inTP is unique. Some lakes do show a good
relationship and some do not. Even among those that
do, however, a statistical comparison of their chloro-
phyll-phosphorus relationships shows that significant
differences (,P < 0.05') exist between the slopes and
intercepts of their regressions (Table 1). The variability
in response of these lakes is shown in Figure 2, in
which the cloud of data is compared to the nine
individual regression lines.
Two important points emerge from Figure 2. First, it
is clear that the difference in response of'individual
lakes to changes in total P can account for a major
proportion of the variance commonly noted in "global"
chlorophyll-phosphorus regressions (e.g. Dillon and
Rigler, 1974; Jones and Bachman, 1976; Nicholls and
Dillon, 1978). Second, the unique response of each
lake raises questions regarding an assumption made by
many users of current global eutrophication models
(e.g. the models of Dillon and Rigler, 1975; Lee, et al.
1978; and Vollenweider, 1976)—the assumption that
individual lakes will respond in a similar fashion to a
given change in total phosphorus.
THE IMPORTANCE OFTN:TP RATIOS TO
CHLOROPHYLL YIELD
If we are to develop eutrophication models that more
accurately predict the response of lakes to changes in
phosphorus loading, it is important that we understand
the sources of variability which generate the scatter
typically observed in chlorophyll-phosphorus regres-
sions (e.g. Figure 2a). In their recent review Nicholls
and Dillon (1978) discussed several reasons for the
scatter, including methodological variation, the relative
biological availability of phosphorus in different lake
-------
74
RESTORATION OF LAKES AND INLAND WATERS
waters, and variations in the chlorophyll/algal cell
volume ratio. However, these authors did not consider
a major factor pointed out by Sakamoto (1966): The
yield of chlorophyll at a given concentration of total P is
sensitive to variations in the TN :TP ratio. In his
original analysis, Sakamoto considered nitrogen to
limit algal biomass in lakes where TN:TP < 10;
similarly, he considered phosphorus to be limiting
where TN :TP > 17. JHe alsojelt that chlorophyll was
proportional to either TN or.TP in lakes where 10 <
TN :TP < 17. Thus, nitrogen availability should
influence the chlorophyll response to phosphorus over
a broad range of TN :Tp ratios. The importance of the
N:P ratio has since been commented upon by
Chiaudani and Vighi (1974), Porcella, et al. (1974),
Schindler (1977), Allan and Kenney (1978), Forsberg,
et al. (1978a) and Allan (1980).
This influence of the TN:TP ratio on chlorophyll
yield during lake restoration can be seen in Norrviken
and Oxundasjon (Sweden) (Figure 1). In the case of
Norrviken (Figure 1 b), diversion of wastewater led to
consistent declines in both TP and c (Ahlgren, 1978,
and 1980.) The only exception occurred in 1970, when
the algal biomass appeared to be N-limited (TN:TP =
8), and may have been regulated by intense
zooplankton grazing as well (Shapiro, 1979). In
Oxundasjon, the effect of changes in nutrient limitation
is also evident (Figure 1a): during the 6 years for which
data are available, the TN :TP indicated N-limitation in
100 150 200 230
NORRVIKEN
C-0.27 TP*2
r'- 0.46
I20
TP,
Figure 1 . — Phosphorus dependence of chlorop.iyll a in (A)
Lake Oxundass'jon 1970-1975, and (B) Lake Norrviken
1969-1978. Confidence limits for the slope and intercept in
(B)are m±0.44 and b± 1 .01 .Circled points denote years of
probable N-limitation (TN .IP < 10). Modified from Smith
and Shapiro (1980).
1971 TN :TP=6:4), and either N-or P-Mmitation in
other years TN :TP = 9.8 13.9.). As would be
predicted from Sakamoto's (1966) analysis, only when
the data from 1971 are excluded is a marked
relationship evident between TN and c.
Table 1. — A. Lower left — logarithmic regressions; upper
right — arithmetic regressions. Letter designates significant
(P< 0.05) difference between two slopes.
Lake W C
W
C
T
Gr a
G
L
B a
N
S
T Gr G L
a
a
a
a
a a
a
B
a
a
a
a
a
N S
a
a
Table 1. — B. Same format as above, except letter designates
significant differences between two intercepts.
Lake W
Gr
L B
W
c
T
Gr a
G
L
B a
N
S
a
a
a a
a a
a a
E
lo
TP, mg m"
TP, mg m'1
Figure 2. — (A) Phosphorus dependence of chlorophyll a in
nine north temperate lakes in which TN'TP <10 Each point
represents a single growing season mean. (B) Same as (A),
except each line represents the regression line for an
individual lake. E-lake Ekoln; GB-Gravenhurst Bay; other
symbols as in Table 1.
A VARIABLE YIELD CHLOROPHYLL-
PHOSPHORUS MODEL
As can be seen in Table 2, various lake restoration
measures lead not only to reductions in total P, but also
lead to changes in the TN :TP ratio. In fact, Table 2
suggests that the TN :TP ratio typically increased in
these lakes, regardless of the type of restoration
-------
NUTRIENT LOADING/TROPHIC RESPONSE
75
Table 2
Lake
Washington
Norrviken
Edssjon
Oxundasjon
Gravenhurst
Ekoln
Boren
Ramsjbn
Ryssbysjb'n
Cline's Pond
Years Range of TN:TP Trend*
1957-1975
1969-1978
1970-1975
1970-1975
Bay 1969- 1975
1972-1975
1973-1976
1972-1974
1973-1974
1970-1971
8.8-25.2
8.1-17.9
7.5-11.3
6.4-13.9
10.7-28.9
14.6-71.8
10.2-38.9
3.6-7.3
3.8-5.2
9.0-18.5
increase
increase
increase
increase
increase
increase
increase
increase
increase
increase
Restoration method
wastewater diversion
wastewater diversion
wastewater diversion
wastewater diversion
wastewater P removal
wastewater P removal
wastewater P removal
wastewater P removal
wastewater P removal
nutrient precipitation
References
W.T. Edmondson, pers. comm.
I. Ahlgren, pers. comm.
I. Ahlgren, pers. comm.
I. Ahlgren, pers. comm.
Dillon, et al. 1978
Forsberg, et al. 1978b
Forsberg, et al. 1978b
Ryding and Forsberg, 1975
Ryding and Forsberg, 1975
Funk and Gibbons, 1979
method used. It thus appears that variations in total
nitrogen, as well as changes in total phosphorus, must
now be considered in restoration efforts. A mechan-
istic, variable chlorophyll yield model which explicitly
considers variations in the (TN :TP ratio has been
developed for this purpose by Smith (1980). In general,
the model predicts a family of parallel chlorophyll-
phosphorus curves (Figure 3) described by the
following:
log c = 1.55 logTP — b,
where the y-intercept, b, is a function of the TN :TP
ratio:
b= 1.55 log
6.404
0.0204(TN :TP )+0.334
The derivation and assumptions of this variable yield
model are discussed by Smith (1980).
One feature of the model which is evident in Figure 3
is that many trajectories of change in chlorophyll a are
possible for a given reduction in total phosphorus. For
example, it appears from Figure 3 tharchlorophyll may
actually increase with a reduction in TP if the TN : TP
ratio increases sufficiently during restoration. Such a
trend was actually observed in Oxundasjon between
the years 1971 and 1972 (Figure 1a) and in Norrviken
between 1970 and 1971 (Figure 1b). Furthermore,
Figure 3 shows that a marked reduction in total P may
also lead to no change in chlorophyll a if there is a
modest change inTN :TP. This pattern was observed
in Lake Norrviken (1974-75), when TP dropped from
158 to 98 mg m-3, but TN:TP rose from 12.3 to 17.9.
As a result, the concentration of chlorophyll a remained
essentially constant (67 to 68 mg m-3).
The variable yield model thus makes general
predictions which appear to be confirmed in actual
restoration experiences. However, a detailed compari-
son of the variable yield model with the Dillon-Rigler
(1974) model emphasizes the_ greater accuracy of
chlorophyll prediction when TN : TP ratios £re con-
sidered. An analysis of 20 lakes for which TN : TP,
TR and c were known was made (Smith 1980), in
which the changes in chlorophyll a were predicted
using the Dillon-Rigler (1974) model and the variable
yield model. The predictions (cPred), which were based
on observed changes in TP during restoration, were
then compared to the actual changes in chlorophyll
tooo
0.1
1000
TP , mg.m"
Figure 3. — Graphical display of the variable yield model,
showing four potential trajectories of change in chlorophyll
a following reduction of TP from 100 to 60 mg m
which occurred in the lakes (cobs). The results of the
analysis for Norrviken (Figure 4) are typical of the
pattern noted for the remaining lakes. With the
exception of the last 4 years of restoration, the Dillon-
Rigler model consistently overestimates the concentra-
tions of chlorophyll actually observed in Norrviken
(Figure 4a). The variable yield model, however, much
more closely predicts the changes in c (Figure 4b). The
improved accuracy of the variable yield model is clearly
shown by a 90 percent reduction in the total prediction
error, estimated here as the sum of squares (SS):
SS = (Cpred — Cobs)2
-------
76
RESTORATION OF LAKES AND INLAND WATERS
When all the values of chlorophyll a predicted from
Equation_1_and 2 are regressed on the measured
values of TPfor Norrviken, the slopes and intercepts of
the regressions:
Cpred = 0.570TP — 24.3, r2 = 0.85
m ±0.195, b±35.3
log Cp,ed = 1-277f~P — 1.009, r2 = 0.92
m ± 0.297, b ± 0.653
are not significantly different from those actually
observed (Smith and Shapiro, 1980) (cf. Figure 1 b).
It should be pointed out that this comparison
considers 3 years (1970-1972) during which the
TN :Tp~ ratio was 12. Dillon and Rigler(1974) point
out, however, that their model should not be used in
such cases. Nonetheless, even when these years are
excluded, the variable yield model generates 75
percent less prediction error (SS = 4169) than does the
Dillon-Rigler model (SS 15551).
200
•8
,ua
100
0
200
•o
u
55=60039
55=6277
-3
100
150
Figure 4. — Comparison of observed concentrations of
chlorophyll s(cobs) in Lake Norrviken with (A) the predictions
made by the Dillon and Rigler (1974) model, and (B) the
predictions made by the variable yield model. (SS = sum of
squared deviations of predicted values from observed
concentrations of chlorophyll; see text.)
CONCLUSIONS
A retrospective look at 16 north temperate lakes
which have undergone restoration has provided
evidence that the reductions in chlorophyll a which
accompanied successful phosphorus reduction were
almost always immediate and continuous (Smith and
Shapiro, 1980). However, the individual lakes behaved
uniquely in their response to nutrient removal. The
TN :TP ratio also showed marked long-term changes
in the cases where nitrogen data were available, as
well. We believe that these changes in TN : TP
modified the quantitative response of the algae to the
declines in total P, and were, to a large extent,
responsible for the significant differences noted in the
slopes and intercepts of the chlorophyll-phosphorus
regressions for the individual lakes (Table 1; Figure 2b).
Because the TN:TP ratio typically increases over
the course of restoration (Table 2) the chlorophyll yield
per unit total P in restored lakes can also be expected to
increase and may tend to offset the potential benefits of
phosphorus removal. Although the majority_of lakes do
appear to be P-limited on the basis of the TN : TP ratio
(Jones and Bachmann, 1978; Wejjss, 1979; Smith,
unpubl.), lakes having a low TN : TP are not
uncommon in many regions (e.g. Florida, R. E. Carlson,
pers. commun.; Denmark, Lastein and Gargas, 1978;
Sweden, Ahlgren, 1980). Management strategies in
these regions should take this fact into account. We
believe, with Dillon and Rigler (1974), and Allen (1980),
that the use of current eutrophication models is
inappropriate for these lakes, and we hope that the
model presented here and in Smith (1980) will help
predict conditions in such lakes following restoration.
REFERENCES
Ahlgren, I. 1978. Response of Lake Norrviken to reduced
nutrient loading. Int. Ver. Theor. Angew. Limnol. Verh.
20:702.
1980. A dilution model applied to a system of
shallow eutrophic lakes after diversion of sewage
effluents. Arch. Hydrobiol. 89:17.
Allan, R. J. 1980. The inadequacy of existing chlorophyll
a/phosphorus concentration correlations for assessing
remedial measures for hypereutrophic lakes. Environ.
Pollut. 1(B) (in press.)
Allan, R. J., and B. C. Kenney. 1978. Rehabilitation of
eutrophic prairie lakes in Canada. Int. Ver. Theor. Angew.
Limnol. Verh. 20:214.
Born, S. M. 1979. Lake rehabilitation: a status report.
Environ. Manage. 3:145.
Carlson, R. 1980. Personal communication. Kent State
University, Kent, Ohio.
Chiandani, G., and M. Vighi. 1974. The N:P ratio and tests
with Selenastrum to predict eutrophication in lakes.
Water Res. 8:1063.
Dillon, P. J., and F. H. Rigler. 1974. The chloropohyll-
phosphorus relationship in lakes. Limnol. Oceanogr.
.— 1975. A simple method for predicting the
capacity of a lake for development based on lake trophic
status. Jour. Fish. Res. Board Can. 32:1519.
Dillon, P. J., K. H. Nicholls, and G. W. Robinson. 1978.
Phosphorus removal at Gravenhurst Bay, Ontario: An 8
year study on water quality changes. Int. Ver Theor
Angew. Limnol. Verh. 20:263.
Dunst, R. C., et al. 1974. Survey of lake rehabilitation
techniques and experiences. Wis. Dep. Nat Rp«nnr
Tech. Bull. No. 75. ' nBS>our'
Edmondson, W. T. 1980. Personal communication
University of Washington, Seattle.
-------
NUTRIENT LOADING/TROPHIC RESPONSE
77
Funk, W. H., and H. L. Gibbons. 1979. Lake restoration by
nutrient inactivation. Pages 141-152 in Lake restoration.
EPA-440/5-79-001. Off. Water Planning Standards, U.S.
Environ. Prot. Agency, Washington, D.C.
Forsberg, C., et al. 1978a. Water chemical analyses
and/or algal assay? Sewage effluent and polluted lake
water studies. Mitt. Int. Ver. Theor. Angew. Limnol.
21:352.
1978b. Research on recovery of polluted lakes.
I. Improved water quality in Lake Boren and Lake Ekoln
after nutrient reduction. Int. Ver. Theor. Angew. Limnol.
Verh. 20:825.
Jones, J. R., and R. W. Bachmann. 1976. Prediction of
phosphorus and chlorophyll levels in lakes. Jour. Water
Pollut. Control Fed. 48:2176.
Lastein, E., and E. Gargas. 1978. Relationship between
phytoplankton photosynthesis and light, temperature and
nutrients in shallow lakes. Int. Ver. Theor. Angew.
Limnol. Verh. 20:678.
Lee, G. F., W. Rast, and R. A. Jones. 1978. Eutrophication
of water bodies: Insights for an age-old problem. Environ.
Sci. Technol. 12:900.
Nicholls, K. H., and P. J. Dillon. 1978. An evaluation of
phosphorus-chlorophyll-phytoplankton relationships in
lakes. Int. Rev. Gesamten Hydrobiol. 63:141.
Porcella, D. B., et al. 1974. Comprehensive management
of phosphorus water pollution. EPA-600/5-74-010. Off.
Res. Develop., U.S. Environ. Prot. Agency, Washington,
D.C.
Ryding, S. O., and C. Forsberg. 1976. Six polluted lakes: a
preliminary evaluation of the treatment and recovery
processes. Ambio 5:151.
Sakamoto, M. 1966. Primary production by phytoplankton
community in some Japanese lakes and its dependence
on lake depth. Arch. Hydrobiol. 62:1.
Schindler, D. W. 1977. Evolution of phosphorus limitation
in lakes. Science 195:260.
Shapiro, J. 1979. The importance of trophic level
interactions to the abundance and species composition of
algae in lakes. Proc. SIL Workshop on Hypereutrophic
Systems, Vaxjo, Sweden, Sept. 10-14.
Smith, V. H. 1980. A variable yield chlorophyll-phosphorus
model for lakes. Unpubl. manuscript to be submitted to
Environ. Sci. Technol.
Smith, V. H., and J. Shapiro. 1980. Chlorophyll-
phosphorus relations in individual lakes: their importance
to lake-restoration strategies. Under review by Environ.
Sci. Technol.
Steel, R. G., and J. H. Torrie. 1960. Principles and
procedures of statistics. McGraw-Hill, New York.
Vollenweider, R. A. 1976. Advances in defining critical
leading levels for phosphorus in lake eutrophication.
Mem. Inst. Ital. Idrobiol. 33:53.
Weiss, C. M. 1979. Trophic indices and their use in trophic
classification of lakes and reservoirs of North Carolina.
Pages 141-211 in T.E. Malone, ed. Lake reservoir
classification systems. EPA-600/3-79-074. Environ.
Res. Lab. U.S. Environ. Prot. Agency, Corvallis, Ore.
ACKNOWLEDGEMENTS
We thank W.T. Edmondson, D.P. Larsen, and I. Ahlgren,
respectively, for the use of unpublished data for Lake
Washington, Shagawa Lake, and Lakes Norrviken, Edssjon,
and Oxundasjon. We also thank R.J. Allan for providing a
preprint of his study of N-limited Canadian prairie lakes. This
work was supported by National Science Foundation Grant
DEB77-15069 and by NIH Research Service Award
T32GM07323. Contribution Number 226 from the
Limnological Research Center.
-------
78
SIGNIFICANCE OF SEDIMENTS IN LAKE
NUTRIENT BALANCE
H. L GOLTERMAN
Biology Station
La Tour du Valat le Sambuc
Aries, France
ABSTRACT
A considerable part of phosphate entering a lake will enter the sediment; the concentration of the
phosphate in the lake will therefore be lower than when calculated as a conservative compound. It
has been suggested that the amount in sediments is a proportion of the phosphate entering the
lake, or that the phosphate in sediments is a constant fraction of the concentration.
Mathematically, it can be shown that for lakes in a steady state these assumptions are identical.
Recently it has been shown that better results could be obtained for some lakes on the assumption
that the amount which is in sediments is controlled by adsorption on the sediments; adsorption
isotherm can be used to describe this process. If this is the case, the amounts of phosphate in the
sediments are controlled by the phosphate concentration in the lake and by the total sediment load
of the lake. Variation in the sediment load can probably explain a large part of the scatter in
statistical (stochastic) phosphate models. There is some indication that due to sedimentation the
water retention time controls the amount of phosphate which is retained in the lake. It seems likely
that phosphate profiles in lake sediments can give semi-quantitative rapid information of the
loading history.
This paper has been published in Hydrobiologia 72:61 (1980).
For the complete paper, please contact Dr. Golterman at the following address'
Dr. H. L. Golterman
Station Biologique D
La Tour du Valat le Sambuc
F-13200 Aries, France
Phone: (90) 98. 90. 13
-------
79
PREDICTING DREDGING DEPTHS TO MINIMIZE
INTERNAL NUTRIENT RECYCLING IN SHALLOW LAKES
H. G. STEFAN
M. J. HANSON
St. Anthony Falls Hydraulic Laboratory
University of Minnesota
Minneapolis, Minnesota
ABSTRACT
In shallow eutrophic lakes alternating periods of temperature stratification and wind-induced
turnover events can produce a discontinuous but significant flow of phosphorus released from the
sediments to the photic zone. The result can be sequences of weather-dependent algal blooms.
The phosphorus release is typically associated with oxygen depletion of the water near the bed.
The mixing events are caused by strong winds. A method has been developed to predict vertical
temperature structures and multiple turnover events in shallow lakes in response to wind forces
and heat transfer from the atmosphere. The method is an extension of the Minnesota Lake
Temperature Model and based on integral energy transfer. It has been verified against field
measurements of stratification structures for a time scale of 12 hours. With several years of
receded weather data as input, the sequence and number of turnover events in the Fairmont Lakes
in southern Minnesota have been determined for selective alternative dredging depths. It was
possible to determine a relationship between the number of midsummer turnover events, the
number of stratification periods of 5 days or more, and the dredged lake depth. It was therefore
possible to estimate which dredged conditions would reduce summer fertilization of the photic
zone by phosphorus recycled from the lake bed.
INTRODUCTION
Dredging is one of several methods to restore
shallow and eutrophic lakes. Some dredging tech-
niques and case studies are summarized in Dunst, et
al. 1974, and U.S. EPA, 1979. This paper describes a
method to determine the depth to which a lake may be
dredged to prevent phosphorus recirculation from the
sediments. The method of computation is for shallow
lakes. It will be illustrated by using the Fairmont Lakes
in southern Minnesota.
CONCEPT
It has been found (Stefan and Hanson, 1979, 1980)
that in very shallow, eutrophic lakes much of the
phosphorus necessary for the growth of phytoplankton,
including nuisance blooms of blue-green algae in the
summer, can be recycled from the bottom sediments.
Several mechanisms will release phosphorus from the
sediments to the water: (a) Chemical release when the
hypolimnetic waters become anaerobic during stratifi-
cation; (b) uptake by the roots of macrophytes and
release through remineralization; and (c) release
through the digestive tract of bottom feeders. It is
therefore not always true that only phosphorus loading
from runoff produces nuisance blooms in shallow
lakes.
Dredging of a shallow lake usually does not remove
all phosphorus-containing materials from a lake bed.
Often newer layers of deposit are removed, exposing
older layers. If the benthic material is organic,
phosphorus will still be present in the sediments after
dredging. The release processes of phosphorus also
will not be significantly altered by dredging. The
success of a dredging program must therefore not be
related to the availability of phosphorus as a nutrient
but to other factors:
1. Dredging changes summer stratification and
vertical mixing characteristics by increasing depth. This
is illustrated in Figure 1, which displays the simulated
summer isotherms of the same lake under the same
weather conditions for three different depths. Deep-
ening the mixed layer or complete overturns brings the
phosphorus released on the lake bottom to the photic
zone near the lake surface, where it can be used by
phytoplankton. Greater depth reduces the frequency of
summer overturns in very shallow lakes.
2. The greater depth provides a larger volume of
hypolimnetic water which in turn contains a larger
quantity of oxygen. Given identical rates of benthic
oxygen uptake per unit area, the hypolimnion of a
deeper lake will take longer to become anaerobic than
the hypolimnion of a more shallow lake. Phosphorus
release thus will be delayed in the deeper lake.
3. A third and minor effect of dredging is to reduce
water temperature by increasing lake volume. The
water temperature depression increases oxygen solu-
bility and decreases biological kinetic rates; it thereby
delays oxygen depletion in the hypolimnion and slows
growth rates of algae.
The general concept is that shallow eutrophic lakes
can be dredged to such a depth that phosphorus
released from the sediments into the hypolimnion is
not recycled to the photic zone by lake overturns. This
-------
80
RESTORATION OF LAKES AND INLAND WATERS
will reduce the standing crop of algae. A method to
determine the required dredging depth will be
presented and illustrated.
BUDD LRKE 1977
SERSONflL ISOTHERMS I C I '
MfiY JUN JUL HUG
fin
BUDD LflKE 1977
SERSONfiL ISOTHERM 1 Cl '
JUN . JUL , HUC
BUDD LflKE 1977
SE930NFIL ISOTHERMS I • C I
JUN . JUL , RUG
Figure 1. — Simulated isotherms, Budd Lake, 1977.
FAIRMONT LAKES STRATIFICATION
PHOSPHORUS RECYCLING, AND
MIXING STUDY
Problem Description
The City of Fairmont in southern Minnesota has used
several different strategies to reduce algae blooms in
its chain of five very shallow city lakes. Treatment with
copper sulfate as well as diversion and treatment of
municipal sewage effluent have not solved the problem
permanently. Since 1966, the city has been pursuing a
dredging program.
The original basins of the Fairmont Lakes were
formed by melting of ice blocks in the postglacial
period. They have been filled with as much as 12 to 15
meters of lake-derived organic materials. Area versus
depth curves for undredged and anticipated dredged
conditions of one of the lakes are shown in Figure 2.
Conditions for the other lakes are similar.
The lakes have surface areas ranging from 0.34 to
2.25 km2 and mean depths from 2.1 to 3.7 meters.
Water budgets for the years 1973, 1974, and 1975
showed that hydraulic residence times varied with
weather from 0.2 years to 3.1 years. Much of the runoff
occurs during snowmelt.
Primary productivity in the shallow, eutrophic
Fairmont Lakes appears alternately limited by light and
by phosphorus availability. Phosphorus is the basic
material prerequisite. Light availability is often the
dynamic regulatory parameter, and is dependent on
solar radiation intensity, light attenuation in the water,
and mixed layer depth. Attenuation in turn depends on
the color and the suspended material content of the
water.
Phosphorus budgets for the Fairmont Lakes have
been presented by Barr (1974), Knoll and Megard
(1973), and Stefan and Hanson (1979, 1980). They
offer strong evidence that phosphorus loading by
surface runoff from rainfall or snowmelt or from
municipal waste water cannot account for the total
summer phosphorus used by the algae.
Observed phosphorus and chlorophyll a data mea-
sured in the summer of 1 979 are shown in Figure 3. A
line for phosphorus limitation has been added.
Area in Hectares
0 10 20 30 40 50 60 70 80 90
80 120 I60
Area in Acres
200
Figure 2. — Depth/area relationships for Budd Lake
-------
DREDGING AND BIOMANIPULATION AS RESTORATION TECHNIQUES
81
200
.c
o.
o
O SISSETON LAKE
°00
50
IOO 150 200
Phosphorus (mg/m3)
250
Figure 3. — Relationship between chlorophyll a and total
phosphorus in 1979. (Budd Lake treated with copper sulfate.)
Computer Model Simulations of Stratification
Structures and Vertical Mixing
The summer stratification dynamics and the effects
of dredging on vertical mixing were simulated in a
mathematical model.
The temperature stratification and the vertical mixing
in a shallow and small lake (large lakes are usually not
considered for dredging because of cost) are usually
the result of intensive air-water interaction. Transfer of
heat and wind energy is of great importance. Recent
advances (Ford and Stefan, 1980; Harleman and
Octavio, 1977; Stefan and Ford, 1975; Tucker and
Green, 1977} in the quantitative description of these
interactions make it possible to predict a lake's
temperature structure with reasonable accuracy on a
daily or even shorter time scale, provided that the
forcing weather parameters are known.
In the Minnesota Lake Temperature Model (MLTM)
(Ford and Stefan, 1980; Stefan and Ford, 1975), the
lake is considered as a stack of horizontal layers of
equal thickness z (e.g., 0.5 m) and variable horizontal
area A(z). Each of the layers is considered to be
homogeneously mixed and isothermal. A density
difference between layers resulting from thermal
differences restores horizontal stratification if the
system is brought out of equilibrium. Water tempera-
ture is considered variable with depth and time.
Temperatures T(i, k)for specific layers (i) at a particular
time (k) are computed by the model. The discrete
element approach gives the vertical temperature
distribution in the form of a step function. The
meteorological variables (air temperatures, dew point
temperature, wind, direction, solar radiation, and wind
speed) are input data.
It has been observed that the turbulence generated
in a lake by wind or natural convection mixes
homogeneously the upper layers to a depth called the
mixed layer depth. The mixed layer depth can range
from a few centimeters to the total depth of the lake.
The mixed layer depth and the temperatures of the
layers are found in the MLTM by applying internal
(thermal) energy balances and mechanical energy
balances for each time step. For each step the heat
energy input and the internal heat energy budget are
calculated and applied to give a particular temperature
profile. If the water is thermally (density) stratified,
lifting work is required to mix a lower layer with the
layer above it. The amount of energy required to lift a
layer depends on the sharpness of the temperature
gradient which determines the density differentials and
the distance between the center of mass of the upper
well-mixed layer and the center of mass of the layer to
be mixed. The energy required to do the lifting work is
derived from wind shear on the water surface. To
determine if the energy supplied by the wind is
adequate to mix an additional layer or if it is dissipated
by viscosity with no additional entrainment, an energy
ratio is used. When the ratio is greater than critical
(Ford and Stefan, 1980; Stefan and Ford, 1975), i.e.,
the energy provided by the wind per unit volume of
mixed layer is greater than the work needed to lift the
layer below the mixed layer, then additional deepening
of the mixed layer will occur.
To determine the heat input to the lake, net long
wave (mostly atmospheric radiation which is absorbed
at the water surface), net short wave radiation (mostly
solar radiation, which is absorbed exponentially with
depth with a specified attenuation coefficient), back-
radiation, heat losses at the water surface by
evaporation (condensation), and heat transfer by
convection are all considered.
To adjust the incoming radiation at the water surface
for reflection, an albedo of .06 was used. The cooling
which occurs by evaporation at the lake surface is
calculated using a relationship similar to that used by
Brady, et al. 1969. The energy input by the wind is
calculated from the shear stress at the water surface by
a relationship proposed by Wu, 1969.
To model the Fairmont Lakes, the MLTM was
modified so that weather data input and computations
were carried out at a time step of 12 hours rather than
24 hours. The Fairmont Lakes are very shallow and
respond more rapidly to meterological conditions than
deeper lakes. The night cooling in these weakly
stratified lakes induces mixing by natural convection,
which plays a significant role. In the model the heat
input is applied first, and then the wind energy input.
The 12-hour time periods used were from 6 a.m. to 6
p.m. to 6 a.m. All the solar radiation was considered to
occur during the 6 a.m. to 6 p.m. time period.
Solar radiation and wind velocity are the two most
important weather variables in the MLTM model. Both
may vary strongly from day to day and from year to
-------
82
RESTORATION OF LAKES AND INLAND WATERS
year, and the response of the lake in terms of water
surface temperature and vertical mixing is therefore
very dynamic.
Simulations with the modified MLTM were made for
five of the Fairmont Lakes under three observed
summer weather sequences (1974, 1976, and 1977).
The summer of 1974 was wet and cool and the
summers of 1976 and 1977 were dry and hot. The
simulation required weather data which were obtained
from the Fairmont Airport, the Fairmont Municipal
Water Filtration Plant, and the University of Minnesota
Agricultural Experiment Stations at Lamberton and
Waseca.
The simulations yielded vertical temperature profiles
at 12-hour intervals. The results were represented
graphically as mixed layer depths, as daily surface and
bottom temperature, and as seasonal isotherm con-
BUDD LfOE I 977
'ED LRrER DEPTHS
JUL RUG
BUDD LflKE 1977
12 HOUR MIXED LRTER DEPTHS
JUN . JUL , RUG
BUOO LflKE 1977
tours. Samples are shown in Figures 4, 5, and 1,
respectively.
Prior to its application, the model was calibrated to
minimize the standard error between measured and
predicted water temperatures. The model calibration
provided the constant reduction coefficient by which
wind velocity data from the Fairmont Municipal Airport
had to be adjusted to optimize agreement between
measurements and predictions. That unusual pro-
cedure was adopted because the only available wind
data were from an anemometer operated on top of an
airport hangar, hence requiring adjustment.
After calibration, 19 of 22 identifiable midsummer
lake overturns in 1974, 1976, and 1977 were
accurately predicted.
An independent model verification was made after
calibration bv comparing predicted and measured
BUDD LflKE 1977
5UHFREE TD BOTTOM TEMPERHIURE RRNCE ftFTER NIUHT MI/INC
JUN JUL RUG
BUDD LflKE 1977
SORFHEE TO BOTTOM TEMPEHHTURE flflNCE HFTER NIOHT Ml XII
JUL
RUG
SEP
BUDD LflKE 1977
SURFRCE TO BOTTOM TEMPERATURE RRNOE flFTER NIOHT MIXING
JUN JUL
Figure 4. — Simulated mixed layer depths, Budd Lake, 1977.
Maximum depth =5.0 m (top), 6.75 m (center and 8.0 m
(bottom).
Figure 5. — Simulated surface and bottom temperatures Budd
Lake, 1977. Maximum depth =5.0 m (top), 6.75 m (center and
8.0 m (bottom).
-------
DREDGING AND BIOMANIPULATION AS RESTORATION TECHNIQUES
83
water temperatures at a water withdrawal site (water
treatment plant). The arithmetic mean of the tempera-
ture differences between prediction and measure-
ments for a 3-year period was -0.41 °C; the standard
deviation was 0.95°C.
Dredging Effects on Stratification
Frequency of Mid-Summer Overturns
and
Lake temperature structures and mid-summer over-
turns were simulated for several post-dredging depths
to determine the frequency of mid-summer overturns
and the minimum depth for stable seasonal summer
stratification (no mid-summer overturn). Model simula-
tions were again made with the weather conditions
encountered in 1974, 1976, and 1977. Different lake
depth contours were used to simulate the effects of
dredging.
Model results showed that under present conditions
the lakes experienced several stratification periods
separated by overturns during each of the three
summers. Figure 4 provides detailed information on the
simulated mixed layer depths in the morning and the
evening of each day for Budd Lake under 1977
weather. By increasing the depth to 8 meters overturns
can be prevented, as shown by the simulated plots of
Figures 1-and 4. A seasonal water temperature
stratification lasting from the beginning of June
through the end of August was required.
A summary of simulated frequencies of summer
overturns and of stratification periods of 5 days or more
is given in Figure 6 for the year 1974. A 5-day
stratification period was chosen because field mea-
surements of hypolimnetic dissolved oxygen concen-
trations showed that anaerobic conditions usually
developed within 2 to 4 days, so that phosphorus
release from sediments in the Fairmont Lakes can be
expected to occur at the very latest after 5 days of
stratification (Stefan and Hanson, 1979, 1980).
Undesirable intermittent stratification and multiple
summer overturns which are a prerequisite for
phosphorus recycling, were found most frequently in
the depth range from 4 to 6 meters maximum lake
depth. One of the lakes (Sisseton) mixed to the bottom
in mid-June 1974, even when dredged to a maximum
depth of 8 meters. A maximum depth of 9 meters was
required to achieve a stable summer stratification in
that lake.
Dredging to greater depths also decreases bottom
water temperatures. Simulations of bottom water
temperatures from Budd Lake during July 1974 were
25°C at 5 meters, 21 °C at 7 meters, and 18°C at 8.5
meters maximum depth. The deeper lake conditions
would therefore favor growth of game fish (i.e., walleye
or northern pike) which are strained by low oxygen and
temperatures above 22 to 23°C.
Recommendations for Required
Dredging Depths
The information obtained by model simulations can
be used to make recommendations for required
dredging depths of the Fairmont Lakes based on mixing
criteria:
1. Deepening the lakes to a maximum of 4 meters
would be disadvantageous because it would increase
the potential for sedimentary phosphorus release
(increased number of periods of anaerobic conditions)
and for transport of that phosphorus to the photic zone
by overturns.
2. Dredging to maximum depths between 4 and 6.5
meters may be effective if the larger hypolimnetic
water volume contains enough oxygen to prevent
anaerobic conditions. From a mixing point of view, a
maximum depth of 6.5 meters is not better than 4
meters.
3. A maximum dredged depth of 8 meters will
effectively reduce phosphorus transport from the
bottom to the photic zone (9 meters in Sisseton). Table
1 summarizes the anticipated improvements at
different dredging depths.
The cost of dredging the Fairmont Lakes has been 72
cents per m3 in 1978 dollars. This figure ranks among
the lowest quoted by U.S. EPA (1979). The anticipated
total dredging costs for the Fairmont Lakes are given in
Table 2. These costs are significant, but other
restoration techniques such as inflow treatment,
nutrient inactivation, and aeration would require
continuous expenses and monitoring.
Compared to other lake improvement alternatives,
dredging has a very lasting effect. Results of carbon-14
Max. Lake Depth (m)
- i
x a
X CD > CD
§9-1.8
I lud I
k*—•->-
o
ro
o
o.
iti
4
01
Stable I Intermittent Stratification
Summer Stratification
I Wei I Mixed
Figure 6. — Predicted number of days with overturns (complete
vertical and mixing) and periods of stratification of 5 days or
longer as function of maximum dredged depths from June 1 —
August 31, 1974.
-------
84
RESTORATION OF LAKES AND INLAND WATERS
sediment dating suggest that over the last 9,000 years
sediments accumulated in Hall Lake at the rate of 0.12
cm/yr. At this rate it would take about 420 years to
refill 0.5 meters of dredged material.
35
30-
25-
20-
15-
10
SISSETON LAKE
•— Surface
»-- Bottom
10 15 20 25
June
5 10 15 20 25 5 10 15 20
July
August
Figure 7. — 1979 surface and bottom temperatures in Sisseton
Lake.
1979 Observations of Stratification
Dynamics and Effects
In the summer of 1979, Budd and Sisseton Lakes
were monitored two or three times a week. By a
fortunate coincidence the lakes showed a seasonal
stratification for the first time. They stratified toward
the end of .June and remained so until the middle of
August, whereas in preceding summers several
midsummer overturns had occurred. Unusually cold
weather with substantial winds in the earlier part of
June caused this occurrence. It was therefore possible
to observe the phenomena which dredging is expected
to produce annually.
Figure 7 illustrates the observed development and
strength of the vertical temperature gradient. Figure 8
gives the observed dynamics of the mixed layer and the
thermocline. Figure 9 illustrates the rise in ortho-
phosphorus in the hvpolimnion after its development
and the absence of ortho-phosphate in the epilimnion.
Associated chlorophyll a levels are shown in Figure 10.
A significant drop in chlorophyll a occurred after the
onset of stratification. There were no mid-summer
blooms.
Seasonal stratification as experienced in 1979
maintained dissolved phosphorus and surface chloro-
phyll a at lower levels than in previous years.
The data shown in Figures 9 and 12 for Sisseton
Lake and similar measurements in Budd Lake (Stefan
and Hanson, 1979) agree with the hypothesis of
phosphorus release and recycling from the sediments
and the anticipated effects of dredging.
Table 1 — Budd Lake dredging diagnosis, 1974.
Maximum
dredged
depth
(meters)
(1)
5.2
6.75
7.5
8.0
8.5
Added
depth
(meters)
(2)
0
1.5
1.3
2.8
3.3
Potential
for
anaerobic
conditions
and
phosphorus
release
(3)
High
Low
Number
of mid-
summer
overturns
(4)
5
4
2
0
0
Hypolimnetic water
temperature in July
(°C)
(5)
25-26
21-24
18-20
17-19
16-18
Potential
for algal
blooms
(6)
High
Low
Water
quality
(7)
poor
better
Table 2. — Predicted cost of dredging Fairmont Lakes to different depths.
Lake
(D
Amber
Hall
Budd
Sisseton
Surface area
(km2)
0.73
2.25
0.90
0.54
Present
maximum
depth
(meters)
(2)
4.5
3.4
5.2
6.0
Present
Proposed
average maximum dredged
depth
(meters)
(3)
3.6
2.1
3.7
3.5
depth
(meters)
(4)
6.75
8.0
4.25
6.75
8.0
6.75
8.0
8.0
Proposed
average depth
after
dredging
(meters)
(5)
5.0
5.7
3.9
5.6
6.1
4.8
6.0
5.8
Material
removed
(meters)
(6)
1.056x106
1.492X106
4.042X106
7.863X106
8.982X106
1.036x106
2.069x106
0.844x106
Percent
volume
increase
(7)
40
57
85
166
190
30
62
43
Cost at
$.72/meter2
(8)
$ 760,000
1,080,000
2,908,000
5,657,000
6,500,000
745,000
1,497,000
611,000
-------
DREDGING AND BIOMANIPULATION AS RESTORATION TECHNIQUES
85
10
2 -
3 •
4 •
5-
6 -
June
20
10
July
20
i
August
10
• Mixed Loyer - Thermocline
Figure 8. — 1979 mixed layer and thermocline depth in
Sisseton Lake.
1400
iooo-
5OO-
SISSETON LAKE
•— Surfoce
o-- Bottom
.» ."
f k-.i
10 15 20 25
June
5 10 15 20 25
July
5 10 15 20
August
REFERENCES
Barr Engineering. 1974. Fairmont Title II study: Nutrient
budget and nutrient sources. Tech. Phase I Memo. April.
Minneapolis, Mipn.
Brady, D. K., W. L Graves, and J. C. Geyer. 1969. Surface
exchange at power plant cooling lakes. Rep. 5, Cooling
Water Stud. Edison Electric Inst. EEI Publ. No. 69-901. The
Johns Hopkins University, Baltimore, Md.
Dunst, R. C., et al. 1974. Survey of lake rehabilitation
techniques and experiences. Tech. Bull. 75. Dep. Nat.
Resour. Madison, Wis.
Ford, D. E., and H. Stefan. 1980. Thermal predictions using
an integral energy model. Am. Soc. Civil Eng. Jour. Hydraul.
Div. 106:39.
Harleman, D. R. F., and K. H. Octavio. 1979. Heat transport
mechanisms in lakes and reservoirs. Proc. Congr. Int. Assoc.
Hydraul. Res. Baden-Baden.
Knoll, S., and R. 0. Megard. 1973. Algal photosynthesis, algal
abundance, and chemistry of lake water at Fairmont,
Minnesota. Interim Rep. 8, Limnol. Res. Center, University
of Minnesota.
Stefan, H., and D. E. Ford. 1975. Temperature dynamics in
dimictic lakes. Am. Soc. Civil Eng. Jour. Hydraul. Div.
101:97.
Stefan, H., and M. D. Hanson. 1979. Fairmont Lakes study:
Relationship between stratification, phosphorus recycling
and dredging. Proj. Rep. 183, St. Anthony Falls Hydraul.
Lab. University of Minnesota, Minneapolis.
1980. Phosphorus recycling and loading in five
shallow Minnesota lakes. Submitted for publ. to Am. Soc.
Civil Eng. Jour. Environ. Eng. Div.
Tucker, W. A. and A. W. Green. 1977. A time-dependent
model of the lake-averaged, vertical temperature distribu-
tion of lakes. Limnol. Oceanogr. 22:581.
U.S. Environmental Protection Agency. 1979. Lake restora-
tion. Proc. Nat. Conf., August 22-24, 1978, Minneapolis,
Minn. Off. Water Plan. Stand., Washington, D.C.
Wu, J. 1969. Wind stress and surface roughness at air-sea
interface. Jour. Geophys. Res. 74:444.
Figure 9. — 1979 surface and bottom orth-phosphorus
concentrations in Sisseton Lake.
250
200
\
a. 100-
§
50-
SISSETON LAKE
10
20
June
10
20
July
10
August
Figure 10. — 1979 surface chlorophyll a concentrations in
Sisseton lake.
-------
86
DREDGING ACTIVITIES IN WISCONSIN'S
LAKE RENEWAL PROGRAM
RUSSELL C. DUNST
Office of Inland Lake Renewal
Wisconsin Department of Natural Resources
Madison, Wisconsin
ABSTRACT
Dredging has been the technique most often used in Wisconsin's lake renewal program. The
program includes both natural and manmade lakes, with lake size and sediment removal up to 205
hectares and 1,720,250 cubic meters, respectively. A wide array of sediment type and disposal
methodologies have been involved. Removal costs have ranged from 37 cents to $1.96 per cubic
meter. During the dredging of Lilly Lake, water levels were lowered about 1.5 meters, resulting in a
temporary reversal of groundwater flow. Dissolved oxygen levels were unchanged in the lake.
However, there were increases in chlorophyll a, gross primary productivity, and some zoo-
plankton species. Reductions occurred in water clarity and macrophyte biomass. Initial post-dredg-
ing monitoring indicates that groundwater inflow has increased greatly.
INTRODUCTION
There are nearly 15,000 lakes in Wisconsin, with a
combined area of over400,000 hectares. They form the
foundation of the tourism/recreation economy, the
third largest industry in the State. Citizen demand for
better environmental protection of these lakes resulted
in the creation of the Office of Inland Lake Renewal
within the Department of Natural Resources in 1974.
Subsequent rehabilitation projects have used various
techniques such as aeration, dredging, drawdown,
storm sewer diversion, aluminum treatment, aquatic
macrophyte harvesting, improved animal manure
handling, streambank erosion control, and several
upland conservation methods. However, dredging has
been the primary technique. This paper will describe:
(1) The dredging program now underway; and (2) the
Lilly Lake project.
DREDGING PROGRAM
Twelve projects are now in the program. Four are
completed, two are currently underway, and six are
finalizing plan proposals. Most of these lakes are
manmade, originally created by dam construction. The
lakes range in size from 4 to 205 hectares, with
watersheds of 1.5 to 1,425 square kilometers. In some,
the infilling rate exceeds 3 centimeters per year.
Sediment removal varies from 26,760 to 1,720,250
cubic meters. Hydraulic dredging is the usual method
of removal, but in four cases drawdown has been
combined with lake bed excavation.
The sediment characteristics are diverse. In some
cases the materials are dense, primarily sand, with a
solids content of 70 to 80 percent. This is the usual
situation when the lake is located on a major river
system. At the other extreme, natural lake sediments
are low density and organic. The solids content may be
as low as 1 to 5 percent. Chemical composition is also
variable, subject to previous lake and watershed usage.
Financial assistance to any project is dependent
upon reasonable assurance of environmental im-
provement and permanency. Data collection is followed
by predictive modeling and professional judgment to
provide a basis for implementing a project (see Dunst,
1980 for further discussion). Removal costs have been
higher for dryland excavation ($1.69 to 1.76/m3)
versus hydraulic dredging ($1.12 to 1.29/m3) in
projects of similar size (109,300 to 191,100 m3).
However, one large-scale hydraulic dredging project
(683,900 m3) has been undertaken to date with a per
unit cost of only 37 cents.
Disposal site costs have been dictated by location,
ownership, and usage. Projects have used settling
basins (with or without allowance for return of carriage
waters), low level diking, spreading on agricultural
land, and spray irrigation. Wherever appropriate,
erosion control practices have been applied in the
watershed. These have involved primarily riprap,
porous plastics, fencing, grassed waterways, diversion
channels, contour strips, and conservation tillage.
Ongoing research activities include assessing: (1)
The value of lake sediments on agricultural production;
(2) the effect of dredging on a lake and associated
groundwater system; and (3) alternative methods of
lake deepening (organic sediments).
LILLY LAKE PROJECT
Lilly Lake is a natural, seepage lake located in
southeastern Wisconsin possessing no surface inlets
or outlets. The lake covers 37 hectares and in 1 977 narj
a mean depth of 1.4 meters. Maximum water depth
-------
DREDGING AND BIOMANIPULATION AS RESTORATION TECHNIQUES
87
was 1.8 meters over more than 10.7 meters of organic
sediments. The water content of the sediments ranged
from 90 to 98 percent. The lake bottom and water
column were filled with dense, rooted weed growth.
Winter fishkills were common, and recreational
opportunities severely restricted. Some fishing, boat-
ing, and swimming was possible in limited areas, but
the recreational value was considered of poor quality.
In-lake production and deposition were causing an
infilling rate of 0.5 centimeter per year (e.g.,
radiometric dating using the Pb-210 method).
A project was undertaken to remove 683,900 cubic
meters of sediment, increasing the maximum depth to
6.6 meters (Table 1). Dredging was initiated in July
1978 and continued until November. It commenced
again in May 1979 and was completed by September.
During 1978, a hydraulic dredge pumped approxi-
mately 382,000 cubic meters of sediment (or 798,200
m3 of sediment/lake water mixture) through a 30-
centimeter diameter polyethelene pipe a distance of
almost 3 kilometers to a settling basin. In 1979 the
sediment was also applied to 15 hectares of
agricultural land. A grant from the U.S. EPA has
provided for monitoring the disposal sites and
evaluating the effect of dredging on the lake.
Evaluations include algae, macrophytes, invertebrates,
fish, water quality, sediments, and groundwater.
Investigations began in 1976 and will continue into
1982.
Table 1. — Depth — Water storage relationship before and
after dredging; Lilly Lake.
Before
Depth
(m)
0
0.6
1.2
1.8
2.4
3.0
3.6
4.2
4.8
5.4
6.0
6.6
Area
(ha)
37
32
29
14
Accumulated
volume
storage (m3)
532,300
320,500
132,200
0
After
Area
(ha)
37
35
33
31
30
28
9
6
5
4
1
.1
Accumulated
volume
storage (m3)
1,216,200
995,800
787,100
590,000
403,700
228,600
115,700
70,900
40,300
15,400
2,300
0
Measurements are being taken at least monthly
during the summer. More frequent measurements
were taken when the dredge was in operation. In-lake
conditions before and during dredging are compared
for the July/September period in Table 2. Dredging
corresponded with increased chlorophyll a, gross
primary productivity, ammonia -N, paniculate phos-
phorus, conductivity, total alkalinity, turbidity, B.O.D.
(5-day), Bosmina, and Chydorus; and decreased water
clarity, soluble organic phosphorus, Ceriodaphnia, and
macrophyte biomass. Also, as a result of sediment/
water removal, the lake level was lowered about 1.5
meters. The groundwater system responded with
increased flow into the lake around the entire
perimeter (including reversal of previous outflow
regions). This effect was verified by a network of
monitoring wells and in-lake seepage meters.
Table 2. — In-lake conditions pre- and during dredging
(July/September average).
Parameter
1976 1977 1978 1979
Chlorophyll a Cug/l) 2.5 3.3 18.5 9.5
Gross Primary Productivity (mgC/mVday) 185 140 1005
Secchi disk (m) est. 5 - 1.3
Macrophyte biomass (g/ma; dry weight) 685 335
Bosmina longirostris (#/1) 56 274
Chydorus sphaericus (#/1) 2 12
Ceriodaphnia sp. (#/1) 20 11
Dissolved oxygen (mg/l) 9.7 7.8 8.8 7.6
B.O.D. (5 day; mg/l) 1.6 3.6
Conductivity (/u mhos/cm at 25°C 247 317 433
pH (standard units) 8.3 8.0 8.1
Total alkalinity (mg/l as CaCOs) 107 142 196
Turbidity (Formazin units) 1.2 3.0 4.1
Nitrite/nitrate -N (mg/l) .02 .04 .05
Ammonia-N (mg/l) .03 1.12 1.44
Organic-N (mg/l) 1.5 1.8 1.4
Soluble reactive phosphorus (/ug/l) 444
Soluble organic phosphorus (pg/l) 16 6 5
Particulate phosphorus (^g/l) 17 30 19
* Biomass was nearly eliminated during dredging; based on visual
examination.
The actual in-lake chlorophyll a levels in 1977 were
closely predicted using Vollenweider (1976) and
Sakamoto (1966) (Tables 2 and 3). Precipitation and
groundwater inflow were much greater in 1978 (pre-
July) versus 1977, and therefore these models were
used to estimate the impact of changed climatic
conditions alone on lake limnology. Mean phosphorus
Table 3. —Predicted chlorophyll a levels using Vollenweider
(1976) and Sakamoto (1966).
'Condition; 12345
Water loading (cfs.)
Precipitation
Groundwater
TOTAL
Phosphorus loading (kg.)
"Overland runoff
Precipitation
Groundwater
TOTAL
0.31 0.47 0.34 0.34 0.34
0.07 0.31 0.09 0.09 0.31
0.38 0.78 0.43 0.43 0.65
7.3 7.3 7.3 7.3 7.3
8.0 10.1 8.4 8.4 8.4
0.4 1.7 0.5 0.5 1.7
15.7 19.1 16.2 16.2 17.4
Predicted chlorophyll a (/ug/l) 5.8 3.5 5.3 3.7 2.7
Hydraulic residencetime (years) 1.5 0.7 1.4 3.1 2.0
"1. Actual measurements, 1977. Lake volume = 532,300 m3.
Precipitation = 78 cm.
2. Actual measurements, 1978 (pre-July). Lake volume =
532,300 m3. Precipitation = 119 cm.
3. Theoretical normal precipitation year before dredging.
Lake volume = 532,300 m3. Precipitation = 85 cm.
4. Theoretical normal precipitation year after dredging and
assuming no change in groundwater inflow. Lake volume
= 1,216,200 m3 Precipitation = 85 cm.
5. Theoretical normal precipitation year after dredging with
expected increase in groundwater inflow. Lake volume =
1,216,200 m3. Precipitation = 85 cm.
"Based on published phosphorus loss coefficients for
urban and forested lands.
-------
88
RESTORATION OF LAKES AND INLAND WATERS
levels in the areas of groundwater inflow were 6//g/l.
Phosphorus values for direct precipitation (15 A/g/l) and
dry fallout (108 /jg/ha/yr) were obtained from recent
studies at a nearby location (Andren and Stolzenburg,
1978). Despite the increased water and phosphorus
loading in 1978, chlorophyll a concentrations were
predicted to decrease slightly. This provides further
credence to the conclusion that dredging was
responsible for the 1978 lake water quality.
In addition, sediment cores were collected in June
1978, and interstitial waters were analyzed for total
dissolved phosphorus, ammonia -N, and nitrite/nitrate
-N (Table 4). The sediments were obviously a major,
potential source of ammonia -N and dissolved
phosphorus. Earlier work with nutrient regeneration
chambers on Lilly Lake in 1977 had also demonstrated
an ammonia -N release rate of 21 to 32 mg/m2 of lake
bottom/day. Dredging would cause an increased
impact of in-sediment conditions on lake water quality
through physical disturbance of the sediments and
greater transport in groundwater seepage, especially in
the reversal region. These mechanisms apparently
caused the higher in-lake ammonia -N levels. Some
dissolved phosphorus was also being transmitted into
the water column/biota; however, the impact was less
pronounced. This probably resulted from oxidizing
conditions present in the water column.
Table 4. — Sediment characteristics and interstitial water
chemistry of a core from Lilly Lake.
Sediment
depth
interval
(cm)
0-8
8-15
15-23
23-30
61-69
91-99
Total
dissolved
phosphorus
(W3/I)
105
128
118
198
202
202
Ammonia-N
(mg/l)
2.1
4.7
6.7
7.0
12.0
18.0
Nitrite/nitrate-N
(mg/l)
<0.1
<0.1
<0.1
<0.1
<0.1
<0.1
* These sediments contained a solids content of 3% by weight.
Investigations are continuing into the post-dredging
phase of this project. The lake water quality should
revert to pre-dredging conditions. Early 1980 informa-
tion indicates a permanent increase in groundwater
inflow as a result of dredging. The lake's water storage
capacity was increased by 128 percent while the
hydraulic residence time appears to have risen by only
43 percent. This is a result of removal of low
permeability lake sediments and an increase in the
area of the lake bed (Beauheim, 1980). Although this
impact will not be highly significant at Lilly Lake, the
increased influx of low phosphorus groundwaters
should have a beneficial effect on water quality (Table
3).
REFERENCES
Beauheim, R. 1980. The effects of dredging on groundwater-
lake interactions at Lilly Lake, Wis. M.S. Thesis. Dep. Geol.
Geophys., University of Wisconsin, Madison.
Dunst, R. C. In press. Sediment problems and lake restoration
in Wisconsin. In Proc. Manage, of Bottom Sediments
Containing Toxic Substances, New Orleans, La. Nov. 26-28,
1979. U.S. Environ. Prot. Agency, Corvallis, Ore.
Sakamoto, M. 1966. Primary production by phytoplankton
community in some Japanese lakes and its dependence on
lake depth. Arch. Hydrobiol. Bd. 62:1.
Vollenweider, R. 1976 . In P. Uttormark and M. L. Hutchins.
1978. Input/output models as decision criteria for lake
restoration. Tech. Rep. WRC 79-03. Water Resour. Center,
University of Wisconsin, Madison.
Financial support for the Lilly Lake investigation is being
provided through a grant from the U.S. Environmental
Protection Agency Corvallis Research Laboratory.
Andern.A. and T. Stolzenburg. 1978. Atmospheric deposition
of lead and phosphorus on the Menomonee River
watershed. Water Resour. Center, University of Wisconsin,
Madison.
-------
89
NUTTING LAKE RESTORATION PROJECT:
A CASE STUDY
DAVID D. WORTH, JR.
Perkins/Jordan, Consulting Engineers
Reading, Massachusetts
ABSTRACT
This paper is a case study of a 279,079 cubic meter lake dredging and watershed management
program for Nutting Lake in Billerica, Mass. The restoration program, a 3 1 /2 year effort, focused
upon (1) lake dredging to deepen the lake to prevent reemergence of nuisance aquatic plants; and,
(2) a watershed management program to reduce nutrient contribution to the lake from overland
runoff. Funding was through EPA's 314 Clean Lakes Program, the Massachusetts Water
Resources Commission's Research and Demonstration Program, and through cash and in-kind
contributions from the town of Billerica. A consulting engineering firm, Purcell Associates of
Hartford, Conn, was retained to refine the basic dredging concept, expand upon existing water
quality data, design the containment facilities, conduct on-going water quality analysis, and
evaluate the results. Two non-continuous operating containment basins with total capacity of
approximately 91,752 cubic meters of dredged material were designed. The solids settled out
while the remaining supernatant liquid decanted into small flocculation basins and then
discharged into an outflow stream. This paper assesses the efficaciousness of various methods of
treatment of the supernatant and compares the projected operating costs and dredged material
production rates with those actually encountered during the first 2 years of operation.
INTRODUCTION
Nutting Lake is a 32-hectare lake located in a small,
densely developed watershed in Billerica, Mass.,
situated in the larger Concord River watershed. Nutting
Lake was a fashionable resort area for many
Bostonians around the turn of the century. Today, one
finds these summer cottages converted to year round
housing to accommodate the suburban growth around
Boston. This dense development, combined with year
round use of seasonal houses, small, inadequate septic
systems, and until recently, unpaved roads, has
reduced Nutting Lake to little more than a large
sediment basin. Because of the highly eutrophied state
of the lake, boating, fishing, and swimming were
virtually nonexistent when this project got underway in
1977.
Physically, chemically, and biologically, Nutting Lake
has many characteristics of an urban and eutrophying
lake: Dense watershed development; unused or
underused recreational potential; high nutrient con-
centrations from septic and overland runoff; and
submergent and emergent macrophyte growth. The
lake itself is bisected by the Middlesex Turnpike,
creating two basins; one is 11 hectares, the other 20
hectares. Both basins, connected by a 1.5 meter
culvert, have the same mean depth of 1.3 meters, and
maximum depths not exceeding 2.1 meters, which
makes for a total volume of 41 hectare-meters. With its
high surface-to-volume ratio and its small size and
shallowness, the lake is highly productive.
There is one major inlet and one outlet to Nutting
Lake. The inlet in the northeast corner of the east basin
has very low flows, is dry during the summer months,
and is supplemented by seepage from several swampy
areas that surround the lake. The outlet, at the west
end of the west basin flows into Mill Brook, and
eventually into the Concord River. U.S. Route 3, a four
lane divided highway, passes almost directly over the
outflow as it leaves the lake.
The lake occupies a shallow depression in bedrock
that is overlain with a thin layer of glacial till and
debris. Depth to bedrock is generally 1/2 to 3 meters,
with frequent rock out-croppings. The soils are well
drained, and are not conducive to supporting on-lot
septic system's.
Until 1975 these septic systems, often of inadequate
size, were the sole means of sewage disposal. In 1975,
interceptor sewers were provided. By 1980, the town
required connection to the system. As recently as
1979, 45 percent of the residential dwellings and 32
percent of the dwellings that front directly on the lake
were unsewered. While there are several plausible
arguments that suggest greater compliance than these
numbers show, it remains that the septic system
problems have been a major contributor to declining
water quality.
Major discharge points into the lake along its north
shore drain an area of approximately 8 hectares, or 3
percent of watershed. In addition to direct stormwater
discharge at these four points, uncollected runoff from
streets and lots accounts for an unquantified but
significant source of additional input.
WATER QUALITY
Water quality information was gathered by the
Massachusetts Division of Water Pollution Control in
1975; baseline conditions were further analyzed before
the dredging program began. Similar analytical
-------
90
RESTORATION OF LAKES AND INLAND WATERS
methods and sampling locations were used in the most
recent baseline program so that consistent information
is developed by which water quality changes can be
monitored.
The water chemistry of Nutting Lake observed in
1975 and 1978 indicated the lake's eutrophic state.
Total phosphate concentrations ranged from 0.02 to
0.39 milligrams per liter (mg/l) as P; higher concentra-
tions were measured during the winter and early
spring periods. Ammonia nitrogen concentrations from
the various sampling locations ranged from a low of
0.00 mg/l in the 1975 study to 1.30 mg/l in the 1978
study. Color values were high, as expected. Secchi disk
visibility, during high algae growth periods, was limited
to approximately less than 1/3 meter.
Bacteriological quality information pointed to the
contributions of both stormwater and direct discharge.
Elevated coliform levels during storms were measured
in the spring and fall of 1977. Fecal coliform levels in
the stormwater, when measured with respect to time,
dropped to 12 percent of their initial value 1 hour after
the onset of the storm, leading to the conclusion that
some sewage was present in the stormwater.
Furthermore, field observations and aerial photographs
in 1977 indicated that raw sewage was reaching the
lake, particularly near its inlet in the east basin.
The greatest problem in Nutting Lake was, and
continues to be the presence of algal blooms and
floating and rooted macrophytes. Submergent and
floating macrophytes were most abundant throughout
the lake.
The submergent group, represented by bladderwort
and pondweed, were concentrated in the west end of
the west basin, near the outlet, though they were
present in fewer numbers throughout the lake. Floating
macrophytes, also concentrated near the outlet of the
west basin, were abundant elsewhere in the lake,
particularly along Middlesex Turnpike. Represented
primarily by watershield and yellow lilies, they
accumulated along the shores of the south coves
suggesting they were strongly influenced by the
prevailing westerly wind, allowing proliferation in the
more protected areas.
Emergent macrophytes were not significant, either in
species represented or areal extent, as the emergent or
floating variety. This is mainly because of the extensive
shoreline development.
During the initial phase of the restoration program,
prior to dredging, macrophytes were identified first
from a boat, then using an aerial high speed black and
white and color infra-red imagery. They were then
mapped for use in designing a dredging program.
Plankton samples were also taken during the initial
phases of the restoration program. Employing standard
methods of sampling and identification, enumeration
was done qualitatively as well as quantitatively. Both
the east and west basins exhibited high algal counts as
would be expected with high nutrient loadings. Two
blue-greens, Aphanizomenon sp. and Anabaena sp.
dominated in the east basin during the baseline survey
period, while Aphanizomenon sp. was the only
dominant blue-green in the west basin.
Subsequent sampling of the basins in the following
years coupled with aerial reconnaissance flights has
pointed to an interesting phenomenon: algal blooms
were occuring first in one basin, then in the other basin
the following year. This pattern, which has been
observed for 4 years, will be discussed later in this
report.
PROGRAM DEVELOPMENT
In 1977 the Billerica Conservation Commission, with
the assistance of the Northern Middlesex Regional
Planning Commission, sought a 314 Clean Lake Grant
from the U.S. Environmental Protection Agency. The
goal was to remove nutrient-laden sediment and
macrophyte growth. They also hoped to deepen the
lake by dredging. Coupled with this in-lake program
was a watershed management program. It was
primarily aimed at controlling nutrient input from
runoff by street-sweeping and limiting further water-
shed development by land acquisition. At the initial
stages, approximately 164,389 cubic meters of mate-
rial was to be removed from the lake by using a Mudcat
hydraulic dredge, and deposited in a basin on town-
owned land on the opposite side of U.S. Route 3.
With the basic program concept of dredging and
watershed management approved by EPA , the town
sought financial assistance from the Massachusetts
Division of Water Pollution Control and the Massa-
chusetts Water Resources Commission. Agreeing to
participate in the project and fund the demonstration of
dredge material disposal, the Water Resources Com-
mission contracted with Purcell Associates of Hartford,
Conn, to refine the dredging program, conduct baseline
water quality analyses, design a dredged material
containment area, and evaluate the entire restoration
project on completion.
At the same time, the town raised funds at a town
meeting to match the EPA and State contributions.
More importantly, the town agreed to supply dredge
operators for the duration of the project as an in-kind
service.
In the summer of 1977, baseline water quality
analysis, vegetation and biological analysis, dredging
area refinement and preliminary design of the dredge
material disposal areas were all undertaken by the
consultants. A Phase 1 report was then submitted to
the State, the EPA, and the town, with recommenda-
tions on dredging areas and dredged spoil containment
and disposal methods.
DREDGING AREAS
The Phase 1 report concluded with a recommenda-
tion of expanding the amount of bottom sediment to be
removed from 164,389 cubic meters to 279,079 cubic
meters. It also recommended expanding the initial 2
year program to 31/2 years to accommodate the
increased volume of material to be dredged. There
were several reasons for the decision to expand the
dredging program:
1. Analysis indicated that the very fine bottom
sediments suspended in the water column by the wind
drag further contributed to color and turbidity prob-
lems;
2. Removal of the existing macrophytes, including
their root systems;
3. Increased removal of the benthic nutrient store1
-------
DREDGING AND BIOMANIPULATION AS RESTORATION TECHNIQUES
91
4. Reduction of the substantial source of oxygen-
demanding organics; and
5. Enhancement of lake aesthetics.
Based on this expanded program, a 5-day, 10-
hour/day dredging program was designed, so that
dredging could be accomplished in the suggested
period. This would result in deepening Nutting Lake
from an average 1.3 to 2 meters. The dredging, which
was to be carried out by town workers under auspices
of the Conservation Commission, was to focus initially
on the dense macrophyte beds in the outflow region of
the west basin. Dredging was to be done to prescribed
depths at this area, so as to investigate macrophyte
regrowth as a function of lake depth. The preliminary
results of this experiment are discussed later in this
report.
DREDGED MATERIAL DISPOSAL
One of the most critical operational components of
any lake dredging project is the proper disposal — both
from an environmental and an engineering perspective
— of the dredged spoil. There are certain elements that
must be considered:
1. Selection of a suitable spoil area within pumping
capacity of the equipment;
2. The containment area must be designed and
operated so that the dredged material can be received
and dewatered at as high a rate as possible, and so that
the supernatant can be discharged at an acceptable
level of quality; and
3. The containment area must be reusable once the
project has been completed.
At Nutting Lake there was a suitable site available
within pumping distance and outside the watershed.
Designing a dredged material containment as well as
establishing operation procedures for the area is a
function of the amount of material to be placed in the
containment area during a dredging season, the dredge
production rates, the settling and bulking character-
istics of the dredged material, and the environmental
restrictions placed on the quality of the discharged
effluent. By using manufacturer-supplied information
and field observations of past projects, dredge
production rates from the Mudcat were projected and
related to the 10-hour work day chosen by the town.
Because the town chose to work only 10-hour days, a
noncontinuous mode of operation at the containment
area was employed, to allow for a quiescent settling
period prior to discharging the supernatant.
To size the containment basins, core samples of the
lake bottom sediment were obtained from both basins,
and their settling characteristcs were simulated in the
laboratory. This was done by mixing the sediment with
lake water (achieving a solids content approximately
the magnitude of that pumped by the Mudcat) and
pumping the slurry into a 1.8 meter column tube and
allowing it to settle. The rate of settlement for the
solid/liquid interface was charted, and a 90 percent
settlement of the suspended material was achieved
after 6 hours and 40 minutes. Thereafter, there was no
significant increase in settlement. Based on the lab
results, a containment area detention time of 7 hours
was established.
Further settlement column tests were run at
different heights to estimate sediment consolidation
caused by self-weight stresses. Sediment volume and
water content were first measured in the core tubes,
and then subsequent to sedimentation in the column
testing tubes. From these two measurements the
bulking factor (here defined as the ratio of a given
volume of the same amount of solids on the lake
bottom) was determined for the sediment from each
basin. For both basins the bulking factor, as calculated
from these tests, ranged from 1.2 to 1.6, depending
upon the effective stresses. Obviously, as more solids
are deposited into the basin the self-weight stresses
increase, further consolidating the material, and
resulting in a lower bulking factor. Based upon these
tests, a final solids height (approximately 30 percent of
the initial slurry height) was used as a design
parameter.
Based upon the town's dredging schedule and a 7-
hour quiescent settling period, preliminary design of
the containment area was initiated using the non-
continuous mode of operation. While the Billerica
Conservation Commission has shortened the detention
times at no loss in water quality, the original design
called for a 7-hour settling period and, prior to initiation
of dredging the following day, draw-off of the
supernatant using an adjustable outflow device.
The disposal area site made available by the town
was a 7-hectare wooded parcel west of Route 3,
approximately 152 meters downstream of the outflow
of Nutting Lake, adjacent to Mill Brook. The topography
at the site was such that, while it was somewhat
limiting, the slopes could be used to advantage in
designing the containment area. Several designs were
developed by the consultants. The one selected
consisted of a two basin design, one of 65,000 cubic
yard capacity and one of 63,000 cubic yard capacity.
These had capacity sufficient for a season's dredging.
At the end of each dredge season the material is
removed and initial capacity is restored.
Though no specific EPA effluent standards existed at
the time, treatment of the supernatant prior to
discharge to Mill Brook was mandatory. The column
settling tests indicated that settling periods of up to 12
hours produced no appreciable improvements in
supernatant turbidity beyond the planned 7-hour
detention time. These turbidity levels were 100
nephelometric turbidity units (NTU) at both 6- and 12-
hour intervals. Settlement times in excess of 3 days
only reduced turbidity to 50 NTU. While the 50 NTU
level met the agreed-upon guideline, the 3-day settling
period was unacceptable. As a result, several treat-
ment alternatives were investigated for cost as well as
their effectiveness in meeting the standard. One
alternative consisted of pumping the supernatant into a
fabric filter basin and draining it through the sides of
the basin into a perimeter swale for discharge into Mill
Brook. This method proved ineffective in improving
supernatant quality, although several weaves of fabric
were used in these tests.
A second treatment alternative consisted of dis-
charging the supernatant into the wetlands bordering
the lake in the northwest portion of the west basin. A
detailed site investigation revealed a typical Massa-
-------
92
RESTORATION OF LAKES AND INLAND WATERS
chusetts wetland with low pH values ranging from 4.1
to 5.3.
Although sedimentation and filtration in the wetland
would facilitate the removal of supernatant turbidity,
concern existed that the low pH might release various
metals that might otherwise remain in a bound state, or
that the pH of the wetland might be altered. Therefore,
this treatment alternative was rejected.
A third alternative involved decanting the super-
natant from the containment area to a flocculation
basin. A low molecular weight cationic polymer would
then be introduced by metering pump into the
discharge pipelines at the outflow from the contain-
ment area. The supernatant would then be held in a
quiescent state in the flocculation basin prior to its
discharge into Mill Brook.
Jar testing was done, using polymers at various
concentrations to determine the optimal concentration
to achieve the desired clarity. A range of 8 to 10 parts
per million was found to be sufficient to achieve
removal levels that permitted direct discharge into Mill
Brook. It was this method of treatment that was
chosen, a system that is functioning adequately after
21/2 years of dredging. It is interesting to note that
turbidity measurements at the flocculation basin
outflow average 2 NTU, and compare favorably with a
turbidity measurement of 9 NTU in Mill Brook and with
the 1 to 5 NTU's for Nutting Lake itself.
PROGRAM COSTS
Before discussing water quality results and the
efficacy of this program, it is useful to review the costs,
on a per unit basis. In a project such as this, two of the
three biggest cost items (purchase or lease of a dredge
and construction of a disposal area) are fixed costs.
Their unit value is inversely proportional to the quantity
of dredged material removed. Labor costs, the third
high-cost item, are variable, and generally proportional
to the quantity of material removed. Based on removal
of 279,079 cubic meters during a 31/2 year period,
estimated costs for the dredge, containment areas,
operation and maintenance and labor come to
approximately $522,000, or approximately $1.87
cu/meter. By most standards $1.87 cu/meter ($1.45
cu/yd) for removal and disposal is an excellent price
and compares most favorably with costs encountered
on other dredging projects. Parenthetically, it is
important to mention two things: Estimates of removal
rates are based upon information supplied by the
dredge manufacturer, as well as field observation, and
the town may be a little behind schedule which would
increase labor and operation costs. Secondly, labor
costs are not artificially depressed; they are based upon
hourly rates paid Department of Public Works
employees by the town.
Perhaps the element that appears most promising is
the fact that contractors are willing to pay for the
dredged material. As part of the Phase 1 report, a
detailed chemical analysis of the sediment was done to
determine if it has reuse value. Reuse of the material,
primarily as a soil conditioner, had been planned from
the beginning of the project, and earlier this year the
Conservation Commission let a contract for the
purchase of 152,920 cubic meters (200,000 cubic
yards) of material at a price of $1.40 cu/meter ($1.15
cu/yd). This would provide the project with a revenue
stream of $215,000 which, projected over the life of
the project, would lower the unit cost well below $1 per
unit measurement.
CONCLUSION
Greatly improved water quality has not resulted to
date. Before the wrong conclusion is drawn, however,
it should be said that sufficient dredging has not yet
occurred to produce dramatic water quality improve-
ment. We are optimistic that with sufficient sediment
removal (in the west basin dredging has taken place
down to a gravel bottom) the annual cycle of self-
fertilization will have been broken, and one of the
primary sources of nutrients will have been removed.
Of concern is the possible shift from a lake
dominated by macrophytes to one dominated by algae.
As alluded to earlier, an interesting algae situation
exists in Nutting Lake, which has been monitored for
the last 4 years. In 1977 and 1978 physical evidence
and observation indicated the dominance of blue-
greens (Aphanizomenon sp.) in the east basin with an
average areal biomassof 18,855 ASU/ml versus6,770
ASU/ml in the west basin. However, 1979 data
contradict the earlier observations and measurements
by the Division of Water Pollution Control that indicated
that areal biomass in the west basin is approximately
four times greater (3,845 ASU/ml versus 13,795
ASU/ml) in the east basin. There are several possible
explanations. The reversal may be because autumn
blooms occur at separate times in each basin, and
perhaps are dynamically related, or that sampling was
conducted at different stages of bloom development.
Since significant changes appear to be occurring
during September and October increased sampling
frequency will be employed this year in an attempt to
deduce the cause.
From an operations perspective the project has been
very successful. Once the initial bugs were worked out
of the system, dredging proceeded smoothly. There are
several small design changes that should be incor-
porated into similar projects, particularly in the areas of
weir design and embankment stabilization. The
Conservation Commission's experience with effluent
treatment and shortened detention times is evidence
that smaller basins could be employed, thereby
reducing construction costs and the land area required
for disposal.
Operator efficiency, a concern whenever labor is
employed, has improved dramatically with experience.
Down time has been reduced and dredge production
rates equal or exceed the rates supplied by the
manufacturer. With additional water quality informa-
tion still forthcoming, (information that will be critical
to the final evaluation of the success of dredging as a
restoration technique) it can be said that the viability of
dredging as a restoration has been at least partially
demonstrated and the Nutting Lake project can serve
as a model for municipalities and/or lake restoration
practitioners who are contemplating similar projects
-------
93
MERCURY SPECIATION AND DISTRIBUTION IN A
POLLUTED RIVER-LAKE SYSTEM AS RELATED
TO THE PROBLEM OF LAKE RESTORATION
TOGWELL A. JACKSON
ROBERT N. WOYCHUK
Department of the Environment
Freshwater Institute
Winnipeg, Manitoba, Canada
ABSTRACT
Available techniques for preventing mercury pollution and restoring mercury-polluted inland
waters are reviewed, and the feasibility of applying them to the Wabigoon River system of
Northwestern Ontario (Canada) is considered. The Wabigoon River and an associated chain of
lakes are polluted with mercury from a chlor-alkali plant/paper mill complex. Methyl mercury
(CHaHg*) levels in surficial bottom sediments depend on environmental factors but are unrelated
to total Hg concentrations. CHsHg* content is related to pH, nutrient supply, and microbial
methionine biosynthesis in wood-chip deposits near the source of pollution but appears to be a
function of sorption, complexation, and precipitation phenomena involving Fe and Mn oxides,
cation exchange sites, humic matter, and sulfide in clay-silt muds further downstream.
Sedimentary CHaHg* levels are relatively high throughout the system despite an exponential
decrease in total Hg downstream from the industrial complex. In the wood-chip sediments CH3Hg*
production is most intense at the sediment-water interface even though total Hg is most abundant
below the interface. The river water, and therefore surface waters of lakes through which the river
flows, are continually being contaminated by CHsHg*released from these deposits. CHaHg'isalso
released into hypolimnion water from Hg-contaminated lake mud. Pelagic fish (e.g., walleye) in the
lakes are probably contaminated principally by CH3Hg+ introduced by the river. Only bottom-
feeding animals (suckers and crayfish) seem to be strongly affected by CHaHg* formed locally in
the lake sediments. Consequently, Hg in the riverbed above the lakes must be removed or
immobilzed (e.g., by dredging or by accumulation and treatment in settling ponds) to reduce Hg
concentrations in the fish species of primary importance to fishermen. Attempts to ameliorate the
lakes without taking the constant influx of contaminated river water into account would probably
be unsuccessful.
INTRODUCTION
Pollution of natural waters by mercury (Hg) can be
prevented by removing Hg from effluents and res-
tricting its use. However, there have been virtually no
attempts to restore bodies of water already contam-
inated with Hg, although potentially useful procedures
have been tested in laboratory and field experiments.
In any attempt to reclaim a contaminated system, the
chemical speciation as well as the total quantity and
distribution of the Hg must be considered. Although Hg
in bodies of water is mostly in the form of Hgz+ ions
bound to sediments, the harmful effects of the Hg are
due principally to formation of monomethyl mercury
(CH3Hg ) from the Hg2* by free-living microorganisms
(Fagerstrom and Jernelov, 1972). This water-soluble,
yet fat-soluble, compound is released into the water
and is readily accumulated by fish, whose meat may
thereby be rendered poisonous to human consumers.
Rates of synthesis and release of CH3Hg* are
determined by environmental variables such as pH, Eh,
nutrient supply, and the abundance of sulfide and other
Hg-binding agents.
This report reviews the available pollution control
techniques and shows how some of them might be
applied in a specific problem region: The Wabigoon-
English-Winnipeg River system of Canada.
METHODS FOR EFFLUENT
PURIFICATION AND RESTORATION OF
LAKES AND RIVERS
Effluent Purification
Heavy metals can be removed from effluents at their
source (Bell, 1976) or at regional treatment centers, as
in Switzerland (Anonymous, 1976), and then recycled
or stored. Methods suitable for Hg include the
following:
1. Clarification, whereby paniculate Hg is allowed to
settle out of turbid wastewaters.
2. Chemical treatment, whereby dissolved Hg is
precipitated or is scavenged by adsorbants: Organic
and inorganic sulfides (e.g. S^, FeS, pyrite (FeS ), alkyl
thiols, and wool fibres) are particularly effective (Feick,
et al. 1972; Suggs, et al. 1972; Tratnyek, 1972;
Jernelov and Lann, 1973; Reimers and Krenkel, 1974;
Chow and Buksak, 1975; Brown, et al. 1979). Other
binding agents include elemental sulfur (Suggs, et al,
1972), peat (Feick, et al. 1972), Fe and Mn oxides
(Lockwood and Chen, 1973; Kinniburgh and Jackson,
1978), and nitrogenous polysaccharides and complex
amines (Moore, 1972; Snyder and Vigo, 1974).
Recovery of Hg2* as Hg° by reaction with Al° (Maag and
Hecker, 1972) and co-precipitation with wastes
-------
94
RESTORATION OF LUKES AND INLAND WATERS
flocculated by aluminum sulfate (Jernelov and Lann,
1973) have also been proposed.
3. Accumulation of Hg by batch cultures of algae or
other organisms followed by filtration or other
procedures for harvesting the cultures (Filip, et al.
1979).
A method employing sewage for production of both
algal blooms and biogenic sulfide in settling ponds has
been proposed for removal of Hg and other heavy
metals from effluents (Jackson, 1978).
Lake-river restoration
Restoration procedures must be selected to suit the
individual environment. At best, these methods entail
problems such as high cost, harmful side effects,
technical difficulties, and limited effectiveness. Thus,
the benefits and disadvantages must be carefully
weighed. The following techniques have been consid-
ered:
1. Physical removal of contaminated sediments by
dredging or pumping, and other mechanical operations
such as ploughing bottom sediments (to dilute
contaminated sediment with underlying "clean' sedi-
ment), river diversion, and manipulation of river flows
to flush contaminated sediments into holding ponds
(D'ltri, 1972; Feick, et al. 1972: Jernelov and Lann,
1973; Jernelov, et al. 1975; Parks, et al. 1980; Wilkins,
1980; Wilkins and Irwin, 1980). Difficulties include
high cost, possible resuspension of sediment (leading
to wider dissemination of Hg and acceleration of
methylation rates), and the problem of providing for
safe disposal of dredge spoil (to avoid contamination of
other environments).
2. Chemical treatments and biomanipulation: (a)
Maintenance of mildly alkaline to neutral pH levels in
the water and sediments (e.g. by adding lime or
reducing industrial SOa emissions to control "acid
rain") to favor production of dimethyl mercury
((CH ) Hg) rather than the more pernicious CHsHg*
(Jernelov and Lann, 1973; (b) maintenance of reducing
conditions and high rates of sulfide genesis in bottom
sediments, and eutrophic conditions in the water
column (e.g., by fertilization of lakes), to suppress
formation and release of CHaHg* and also to dilute the
CHaHg* with a large, rapidly growing biological
community (D'ltri, 1972; Jernelov and Lann, 1973;
Jernelov, et al. 1975); (c) addition of selenium
compounds to the water to detoxify Hg and inhibit Hg
accumulation by fish (Ganther, et al. 1972; Rudd, et al.
1980); (d) miscellaneous biomanipulations such as
promoting growth of demethylating bacteria, suppres-
sing formation ofCH3-cobalamin (a factor in microbial
Hg methylation), removing Hg from fish protein,
fostering bacterial conversion of Hg(ll) to Hg°, and
using batches of Hg-scavenging organisms (Wood,
1971; D'ltri, 1972; Snangler, et al. 1973). At best, none
of these methods would be fully effective, and some
could be harmful. Method (b) would involve undesirable
side effects, while method (c) requires further research
and extreme caution, as selenium itself can be toxic.
The procedures listed under (d) are merely hypothetical
possibilities.
3. Covering comammatea sediments with a layer of
sand, gravel, silt, clay, silica, nontoxic mine tailings,
sulfide minerals, elemental sulfur, or even sheets of
manmade materials such as plastic to inhibit methyla-
tion and retard the release of Hg species into the water
(Bongers and Khattak, 1972; D'ltri, 1972; Feick, et al.
1972; Suggs, et al. 1972; Widman and Epstein, 1972;
Jernelov and Lann, 1972; Langley, 1973). Possible
difficulties include disruption of protective sediment
layers by gas bubbles, current action, or benthic
animals. The use of plastic sheets would be costly and
could have harmful ecological effects.
3. Purification of Hg-polluted water bodies by natural
processes: Once the input of anthropogenic Hg is
halted, the Hg in the sediments is eventually sealed off
by layers of uncontaminated sediment or dispersed by
mechanisms such as current action and evaporation of
volatile species. In most cases, however, it would take
an excessively long time — several decades at least,
and possibly centuries — for a system to be fully
restored by natural processes alone (Langley, 1973;
Jernelov, et .;\. 1975).
MERCURY POLLUTION IN THE
WABIGOON-ENGLISH-WINNIPEG RIVER
SYSTEM
Introduction
The Wabigoon-English-Winnipeg River system flows
northwestward over a distance of 420 kilometers
through a chain of lakes extending from Wabigoon
Lake (Ontario) to Lake Winnipeg (Manitoba). The
system is situated in a sparsely populated region of
boreal forest, low relief, and Precambrian rocks
overlain by patches of Pleistocene deposits. This report
principally concerns the Wabigoon River together with
Wainwright Reservoir, Clay Lake, and Ball Lake,
through which the river flows, in that order, after
passing the town of Dryden, where the pollution
orginated (Figure 1).
Figure 1. - Map ot the Waoigoon-Englisn system from
Wabigoon Lake to Ball Lake.
Between 1962 and 1970 uncontrolled quantities of
inorganic Hg amounting to about 10 metric tons were
released into the Wabigoon River from the chlor-alkali
plant of a pulp and paper company (known at different
times as Dryden Paper, Reed, and Great Lakes Forest
-------
DREDGING AND BIOMANIPULATION AS RESTORATION TECHNIQUES
95
Products, Ltd.) at Dryden (Armstrong and Hamilton,
1973). Large volumes of wood fragments and other
paper mill wastes, as well as treated sewage, have also
been discharged into the river (German, 1969). Undei
pressure from the Ontario government, Hg discharges
were reduced in 1970 and supposedly halted in 1975
(Bishop and Neary, 1976) but have, to an appreciable
extent, continued to the present day (Parks, et al.
1980). Anomalously high Hg levels have been found in
the top 5 cm of sediments as far as 240 kilometers
downstream from Dryden (Parks, 1976), and fish at
least 270 kilometers downstream have Hg concentra-
tions above the background levels found in nearby
unpolluted lakes, and well above the 0.5 ppm legal limit
for edible fish marketed in Canada (Fimreite and
Reynolds, 1973; Bishop and Neary, 1976).
"fhe chief human victims of the Hg pollution are the
Ojibway Indians of the Grassy Narrows and White Dog
bands. They have suffered economic hardship because
commercial fishing in the region had to be banned, and
some of them have high blood levels of Hg owing to
consumption of locally caught fish. Neurological
abnormalities which may well be due to mild methyl
mercury poisoning have been detected in several
tribesmen (Troyer, 1977; Wheatley, 1979).
The present study concerns the biogeochemistry and
distribution of Hg species in the Wabigoon River
system and implications for restoration of the system.
This as yet unfinished work is reported in greater detail
elsewhere (Jackson and Woychuk, 1980); it has been
incorporated into a more extensive group project
undertaken jointly by the governments of Canada and
Ontario (Jackson, 1980).
Methods and materials
Water samples and grab samples or cores of bottom
sediment collected in the spring and summer of 1978
were analyzed for CH3Hg*by gas chromatography after
extraction with benzene or toulene, respectively, and
NaBr/H2SO4+CuSO4. Total Hg was determined by
flameless atomic absorption after digestion with
H2SO4/HNO3jt 160°C (in the case of sediments) or
KmnOVhfeSCU, KaSaOa, and ultraviolet radiation (in the
case of water). The sediments were analyzed for pH, E ,
free and "bound" (nonvolatile, 6 N HCI-soluble) sulfide
(S* ), organic carbon (org. C), nitrogen (N), iron (Fe),
manganese (Mn), NHZOH-HCI/HNO3 and citrate/bi-
carbonate/dithionite (CBD)-extractable Fe and Mn, and
amino acids (following acid hydrolysis). Following
centrifugation and rinsing with nitrogen-purged water,
the sediments were extracted with various nitrogen-
purged solvents such as Ca acetate and NH2OH-HCI/
HNO3 to isolate different operationally defined Hg
species. Data for sediments were based on oven-dry
(105°C.) weight.
Results
Total Hg concentrations in the surficial bottom
sediments of the Wabigoon River system decrease
exponentially with distance downstream from Dryden,
owing to progressive attenuation of Hg-contaminated
detritus (Figure 2, A & B). Similarly, these sediments
show a downstream decrease in organic C and N
accompanied by an increase in the N/org. C ratio and
an increase in pH (within the range 4.40 to 7.30),
reflecting a gradational change from deposits of
putrefying wood fragments near Dryden to clay-silt
mud associated with humic matter and Fe-Mn oxides
further down the system (Figure 2C). In contrast,
Figure 2. — Variation of bottom-sediment geochemistry
with distance downstream from the industrial complex at
Dryden A. Total and methly mercury data for the Wabigoon
River between Dryden and Quibell. B & C. Mean values of
total and methyl mercury, "bound" sulfide, and organic
carbon content, and nitrogen/carbon ratio, for principal
depositional basis from Dryden to Ball Lake.
3.0-
o>
X
O
JD
&
Z.5-
1.5-
1.0-
I I
10-
9-
8-
7-
6-
5
TOTAL Hg
r =-0.946
p< 0.001
O
O
CH3Hg+
r- -0.594
p>0.l
—i—
10
—i—
20
~1—
30
—I—
40
—I—
50
O
—r~
0 10 20 30 40 50 60
DISTANCE DOWNSTREAM FROM REED PLANT (Km)
Figure 2A
30-
o>
X 20-
_J
P
1 —
£ 10-
0.
Q.
o—
20-
x"
X° 10-
o
J3
Q.
CU
0
WAINWRIGHT
RESERVOIR
TOTAL Hq
CLAY LAKE
EAST WEST
BASIN BASIN
(BALL LAKE
SOUTH NORTH
BASIN BASIN
•_
1 ' ' i^ 1
CH3Hg+ I
III
n
1
1 ' ' ^< 1
50 100 150
DISTANCE DOWNSTREAM FROM REED PLANT (Km)
Figure 2B.
-------
96
RESTORATION OF LAKES AND INLAND WATERS
CH3Hg levels show no significant long-range variation
with distance (Figure 2, A & B). The slight depression of
CH3HgT levels in Clay Lake can be attributed to
anomalously high sulfide concentrations (Figure 2C)
resulting from eutrophication caused by the influx of
riverborne nutrients. CH3Hg+production in the sedi-
ments is evidently limited by environmental factors
rather than total Hg supply.
30-
20-
O
O>
O
jS 10-
0 —
WAINWRIGHT
RESERVOIR _
orq. C
CLAY LAKE
EAST WEST BALL LAKE
BASIN BASIN SOUTH NORTH
« BASIN BASIN
| i 1 l T
O ID-
5-
CO
o 3°H
z
O 20-
co
E
N
org. C
BOUND
0 50 IOO ISO
DISTANCE DOWNSTREAM FROM REED PLANT (Km)
Figure 2C.
Cores from Wainwright Reservoir and the riverbed
above it showed that CH3Hg* concentrations were
generally highest at the sediment-water interface
(Figure 3), presumably reflecting particularly favorable
conditions for microbial CH3Hg+production, despite the
tendency of Hg-enriched deposits laid down during
1962 to 1970 to be buried by post-1970 sediments of
lower total Hg content. Production of CH3Hg* at the
sediment-water interface leads to continual release of
CHsHg^into the overlying water, followed by fluvial
transport into the surface waters of Clay Lake and
probably Ball Lake. This process is illustrated by
CH3Hg+data for lake surface water in late July (Figure
4A): CH3Hg+concentration is greatest at the inflow to
Clay Lake a'nd decreases progressively downstream to
Ball Lake, reflecting the input of contaminated river
water. In Clay Lake this pattern of variation is especially
prevalent in the summer and winter (Parks, et al.
1980). In addition, CH3Hg+ is apparently released from
lake mud into the deeper waters of the lakes, as
indicated by parallels in the distribution patterns of
CH Hg concentrations in bottom muds and overlying
hypolimnetic waters (Figure 4A).
ppm TOTAL Hg ppbCH3Hg+
0 10 20 3043 5060 70 0 1020304050
NO 7
NO 6
NO. 8
NO 9
NO. 10
NO II
NO 13
NO 12
Figure 3. — Profiles of total and methyl mercury in cores
from Wainwright Reservoir (no. 6-10) and the Wabigoon
River upstream from it (no. 11 -13). "Org." refers to wood-
chip deposits overlying natural clay sediment of pre-
industrial riverbed.
A crucial question from the standpoint of lake
ecology and restoration is to what extent Hg
accumulation by fish is due to "allochthonous"
riverborne CH3Hg+ as opposed to "autochthonous"
CH3Hg+generated locally within the lakes. Comparison
of our data for sediments and water with other
workers' data for fish and crayfish provides some
helpful clues (Figure 4, A & B). Mean Hg concentrations
in different species of pelagic fish (walleye, pike, cisco,
whitefish, and sauger) decrease from Clay Lake to Ball
Lake (and this trend continues at least as far as Tetu
Lake near the Manitoba border), paralleling the trend
shown by allochthonous CH3Hg"" In contrast, the mean
Hg concentrations of bottom-feeding animals increase
(as in the case of suckers), paralleling the tendency
shown by sedimentary CH3Hg+ content, or show no
significant change (as in the case of crayfish). The
results suggest that Hg contamination of pelagic fish
(the species of particular importance to fishermen) is
due primarily to CH3Hg+ loadings from the river, while
bottom-feeding animals are contaminated to an equal
or greater extent by CH3Hg+generated in local bottom
sediments.
The factors controlling the concentrations of CH3Hg+
and other Hg species in bottom sediments varied
-------
DREDGING AND BIOMANIPULATION AS RESTORATION TECHNIQUES
97
greatly down the river-lake system reflecting the
observed gradient in sediment composition.
CH3Hg* concentration in the putrefying wood-chip
deposits between Dryden and Clay Lake is positively
correlated with the abundance of methionine relative
to certain other amino acids and with total N
concentration (as well as the product of the N and org.
C concentrations) (cf. Langley, 1973), and tends to
increase with decreasing pH (Figure 5, A-C). The data
suggest that methyl mercury production in this
biodegradable organic medium is simply a function of
the growth and metabolic activity of microorganisms
which decompose organic nutrients (Langley, 1973)
and employ the methionine biosynthetic pathway for
methylation of Hg (Landner, 1971; Wood, 1971). The
relationship with pH is consistent with the well-known
preferential formation of CH3Hg+ with respect to
(CHa2)Hg under acidic conditions (Fagerstrom and
Jernelov, 1972). CH3Hg* biosynthesis is probably
Figure 4. — Variation of (A) total and methyl mercury in
water and sediments and (B) total mercury in fish and
crayfish, as functions of distance downstream from
Dryden. Each point on each graph represents the mean of
multiple replicate samples. Sediment and water data
came from the present study; fish and crayfish data were
furnished by Fimreite and Reynolds (1973), Bishop and
Neary (1976), B. P. Neary (unpublished data, personal
communication), Armstrong and Hamilton (1973), and G.
McRae and A. Hamilton (unpublished data, personal
communication). Symbols in Figure 4B indicate year of
sample collection:
Walleye: * , 1970; O , 1975; D , 1977.
Suckers: • , 1970; y , 1972.
Crayfish: O , 1971; • , 1974; A , 1976.
15
3 10
LJ
i
S 5
o>
co
rr
o
l_ WALLEYE
WALLEYE MUSCLE
O
°0
I- O
O
CRAYFISH
MUSCLE
o
°70
i i
i
/. 4
CO
tr
o
CO
80 90 ICO MO I20 130 140
DISTANCE DOWNSTREAM FROM DRYDEN (Km)
Figure 4B.
3.0
rr 2.0
Ul
I
o>
O
I.O
O
10
CH3Hg+ IN
SURFACE WATER
10
E
CL
Q.
Q
ID
o>
o
70 80 90 100 110 120 130 140
DISTANCE FROM DRYDEN (Km)
Figure 5. — Variation of methyl mercury concentration
with respect to (a) methionine/threonine ratio(r = 0.873; p
0.001-0.01) and (B) total nitrogen concentration (r =
0.745; p = 0.01 -0.02) in Wainwright Reservoir woodchip
sediments (cores 7 (•), 8 (•), a nd 10 (A) from west side of
reservoir), and (C) pH in wood-chip sediments in
Wainwright Reservoir (•) and the Wabigoon River
between the industrial complex and the reservoir (•) (r = -
0.596; p = 0.001 -0.01).
50-
40-
10
X
o
20-
10-
Figure 4A.
-0.6
Figure 5A.
-0.5
log
-0.4
METHIONINE
THREONINE
-0.3
-0.2
-------
98
RESTORATION OF LAKES AND INLAND WATERS
I 7-
I 6-
I 5-
I 4-
£ 1.3-
ro
O i 2_
_Q
Q.
O.
o. ' '-
^
1.0-
09-
0.8-
07-
7
N/mg
Figure 5B.
Figure 5C.
stimulated both by microbial activity per se and by
acidic metabolic wastes resulting from it. Both are
maximized at the sediment-water interface; this
presumably reflects the importance of oxygen availabil-
ity (cf. Vonk and Sijpesteijn, 1973), even though
sediment Eh values throughout the system ranged from
-420 to -50 mV, indicating anaerobic conditions.
Downstream from the inflow to Clay Lake, these
effects linked directly to microbial nutrient metabolism
appear to be increasingly obscured by sorption-
desorption phenomena, complexation, and precipita-
tion involving colloidal minerals, humic matter, and
sulfide, Clay Lake representing a transition zone. At the
east end of Clay Lake (the mouth of the inflowing river),
CHsHg* levels again correlate with methionine levels
but also (and to an equal degree) with the relative
abundances of CBD-extractable Fe and NH2OH-HCI/
HNOa-extractable Mn (Figure 6), suggesting that the
nature of the hydrated oxides affects microbial
methylating activity (possibly owing to preferential
fixation of Hg2+ ions by the more "amorphous" Mn
oxide, leading to inhibition of CH3Hg+ production).
Further downstream, the apparent role of methionine-
synthesizing microorganisms and riverborne organic
nutrients becomes insignificant. From the center of
Clay Lake's east basin to the south basin of Ball Lake,
sedimentaryCHsHg* levels gave a negative correlation
with the total or interstitial N X org. C concentration
product (Figure 7), suggesting that the humified
organic matter in this part of the river system, unlike
the biodegradable wood chips (cf. Figure 5B), inhibits
methylation — perhaps by complexing Hg ions. In
both lakes, CHsHg* abundance also varies with
different parameters representing the composition or
physical state of the hydrated Fe and Mn oxides
(Jackson and Woychuk, 1980), except in the west basin
of Clay Lake, where methylation is probably limited by
sulfides (Figure 2, B & C).
13-
12-
+
o>
I
O
10-
a 9-
8-
7-
0.63 0.64 0.65 0.66 0.67 0.68 0.69 0.70 0.7I 0.72
log
CBD-EXTRACTABLE Fe
NH2OH-HCI-EXTRACTABLE Mn
Figure 6. — Variation of methyl mercury content with the
ratio of CBD-extractable iron NH OH.HCI/HN03
extractable manganese (r = 0.867; p = 0.01 -0.02) at the
east end of Clay Lake.
20-
i
o
I.8
log (Nxorg.C ]
Figure?. — Relationshipbetween methyl mercury content
and the product of the total nitrogen and organic carbon
concentrations in sediments from the center of the east
basin of Clay Lake (r = -0.914; p = 0.001 -0.01).
-------
DREDGING AND BIOMANIPULATION AS RESTORATION TECHNIQUES
99
Extraction of sediments with mild reagents such as
1 N Ca acetate revealed a progressive increase in the
abundance of the more weakly bound, or "exchange-
able," Hg species relative to total Hg as the wood-chip
sediments grade into clay-silt mud with increasing
distance from the source of pollution (Figure 8A). This
gradient in the binding characteristics of Hg may have
important implications for CH Hg production. The clay-
silt muds extending from Clay Lake to Ball Lake gave a
strong positive correlation between mean CH Hg
content and the mean ratio of Ca acetate-extract-
able (exchangeable) to NH OH HCI/HNO -extractable
(amorphous-oxide bound) Hg species (Figure 8B). A
reasonable interpretation of this relationship would be
O.I 0.5 I SO 50
ppm TOTAL Hq
Figure 8A. — Variation of calcium acetate-extractable
mercury with respect to total mercury in sediment samples
from Wainwright Reservoir (wood chips, X; clay, f ), the
Wabigoon River upstream from the reservoir(woodchips, u;
clay, A ) Clay Lake (east basin, O ; west basin, D ), the
Wabigoon River downstream from Clay Lake (A.), and Ball
Lake (south basin,* ; north basin, • ).
I
O
2.0
0.5 I.O 1.5
Co ACETATE - EXTRACTABLE Hg
NH2OH-HCI/HN03-EXTRACTABLE Hg
Figure 8B. — Relationship between mean methyl mercury
concentration and the mean ratio of Ca acetate-
extractable mercury to NH OH.HCI/HNO -extractable
mercury in clay-silt bottom sediments of Clay Lake (east
end of east basin, O ; center of east basin.V; west basin,
O), the Wabigoon River downstream from Clay Lake (A),
and Ball Lake (south basin: CH3Hg+-rich, •; CH3Hg+-poor,
A; north basin, •)) (r = 0.917; p = 0.001 -0.01).
that Hg ions on exchange sites are more readily
available for bipmethylation than Hg ions chemi-
sorbed, compl'exed, or occluded by amorphous hydrated
oxides and associated humic matter.
Finally, note that CHaHg* levels are nearly the same
in mud from the south basin of Ball Lake as in the
wood-chip deposits near Dryden (Figure 2B) but for
very different reasons. Despite relatively strong binding
of Hg by the wood-chip sediments, (implying low
availability for methylation), CH3Hg* production near
Dryden is fostered by the abundance of organic
nutrients, acidity, and other conditions which stimulate
growth of CH3Hg2-synthesizing microorganisms, as
well as by the paucity of sulfide. In Ball Lake, however,
CHaHgz production is fostered by the relatively weak
sorption of Hg by the abundant clay and silt despite
much lower concentrations of organic nutrients, higher
sulfide levels, and higher pH.
Discussion: Implications for Restoration of the
River-Lake System.
Any program to restore the river-lake system (that is,
to lower the Hg content of its fish to background levels)
would have to include (1) removal or immobilization of
the sedimentary Hg in the approximately 85 km
stretch of the Wabigoon River between Dryden and the
inflow to Clay Lake (an estimated 2 to 3 metric tons of
Hg) (Jackson and Woychuk, 1980)); and (2) prevention
of further Hg discharges from the industrial complex at
Dryden. Unless these steps were taken, any attempt to
ameliorate the lakes individually would be doomed to
failure by the constant influx of bio-available Hg from
the river above Clay Lake. Halting the fluvial transport
of CHsHg* and other Hg species from sources between
Dryden and Clay Lake would not solve the Hg problem
completely, as surface sediments far beyond this
segment of the system are themselves secondary
sources of bio-available Hg. Nevertheless, it could bring
about a rapid, substantial decrease in the Hg levels in
pelagic fish species. Such action is feasible, whereas
decontamination of the surface sediments in the entire
river-lake system would probably not be financially or
technically practical, considering the vastness, ir-
regularity, and poor accessibility of the system.
The following methods for ameliorating the system
have been receiving serious consideration:
1. Dredging. Removal of contaminated sediment by
dredging the riverbed between Dryden and Clay Lake
could be the most effective procedure, but it would be
extremely expensive and time-consuming, besides
entailing the problems of dredge-spoil disposal,
incomplete removal of contaminated sediment, and
resuspension of fine particles. At an estimated rate of
$10 to $50 /cu. yd., the project would probably
cost between $40,000,000 and $200,000,000 (plus
$1,000,000 for access roads) and could take up to 35
years (Wilkins, 1980; Wilkins and Irwin, 1980).
2. Accumulation and immobilization of Hg in a chain
of holding ponds followed by eventual removal or
burial. This method (Jackson and Woychuk, 1980;
Parks, et al. 1980) would require establishment of a
chain of ponds by damming the river at different points
between Dryden and Clay Lake. Contaminated sedi-
ments hydraulically excavated from the riverbed
-------
100
RESTORATION OF LAKES AND INLAND WATERS
(perhaps by manipulation of flow velocities) would
accumulate in the ponds, and treatment with chem-
icals and adsorbants could be used to immobilize
sedimentary Hg, scavenge Hg species from the water,
and inhibitCHaHgaformation. Possibly native sulfur, an
abundant and as yet unwanted byproduct of petroleum
refining in western Canada, could play a useful role.
Eventually the deposits in the ponds could be either
dredged out or sealed off with a thick layer of sand and
gravel. This method might be less expensive and more
feasible than dredging the river, although a cost/-
benefit analysis has not yet been done.
3. Prevention of further Hg discharges from the
industrial complex. Release of Hg from the paper
company is expected to be reduced by forthcoming (a)
modernization of the plant, involving replacement of
suspected sources of Hg such as old sewers; and (b)
installation of a primary clarifier and retention lagoons
for secondary treatment of effluent in compliance with
an Ontario government control order. However, the
secondary treatment (aeration and biodegradation of
organic refuse) might increase the rate ofCHsHgs,
production in the effluent.
4. Controlled addition of selenium compounds to the
river system. This ingenious but controversial approach
is still in the experimental stage (Rudd, et al. 1980).
Possible toxic effects of the selenium would have to be
investigated exhaustively before the method could be
recommended.
Additional restoration techniques were examined but
were judged to be unsuitable. These included (a) burial
of polluted sediments with uncontaminated clay and
silt (Rudd, et al. 1980; Wilkins and Irwin, 1980); (b)
ploughing of lake bottom sediments (Wilkins, 1980);
and (c) diversion of the Wabigoon River from its present
channel (Wilkins and Irwin, 1980).
The option of simply leaving the river-lake system
alone was also considered. Hg levels in fish have been
declining steadily since 1970, when uncontrolled
discharges ceased; but the slowness of the decline and
the presence of huge Hg accumulations undergoing
methylation in the riverbed above Clay Lake lead to the
conclusion that restoration by natural processes would
take many decades, if not centuries (Jackson and
Woychuk, 1980; Parks, et al. 1980).
In conclusion, there is hope that the river-lake
system can be ameliorated, although any program for
achieving this would have some limitations and
uncertainties. From the standpoint of ending the
human misery caused by the Hg pollution, the most
satisfactory solution might be resettlement of the
Grassy Narrows and White Dog communities.
The poisoning of the Wabigoon-English-Winnipeg
River system demonstrates the urgent need for
rigorous enforcement of strict legislation banning toxic
substances from effluents. Clearly, prevention is the
best cure.
ACKNOWLEDGEMENTS
The research was supported by the Federal Government of
Canada (Department of the Environment, Inland Waters
Branch). The amino acid analyses were done by P. Mills. R.
McNeely, M.' Mawhinney, and L. Cha assisted in the field. We
also thank B. Lamm for transportation to and from Ball Lake,
B, P Neary and G. McRae for unpublished data on mercury
levels in fish and crayfish, respectively, and D. P. Scott for
helpful discussions on statistical procedures and fish biology.
REFERENCES
Anonymous. 1976. Detoxifying industrial wastewaters.
Environ. Sci. Technol. 10:127.
Armstrong, F. A. J., and A. L. Hamilton. 1973. Pathways of
mercury in a polluted Northwestern Ontario Lake. Pages
131 -1 55 in P. C. Singer, ed. Trace metals and metal-organic
interactions in natural waters. Ann Arbor Sci. Publishers,
Ann Arbor, Mich.
Bell, A. V. 1976. Waste controls at base metal mines.
Environ. Sci. Technol. 10:130.
Bishop, J. N., and B. P. Neary. 1976. Mercury levels in fish
from Northwestern Ontario, 1970-1975. Rep. Ontario
Minist. Environ.
Bongers, L. H., and M. N. Khattak. 1972. Sand and gravel
overlay for control of mercury in sediments. Rep. 16080
HVA 01/72. Water Pollut. Control Res. Ser. U.S. Environ.
Prot. Agency, Washington, D.C.
Brown, J. R., et al. 1979. Mercury removal from water by iron
sulfide minerals. An electron spectroscopy for chemical
analysis (ESCA) study. Environ. Sci. Technol. 13:1142.
Chow, A., and D. Busak. 1975. The extraction of mercury and
methylmercury by diphenylthiocarbazone-treated polyure-
thane foams. Can. Jour. Chem. 53:1373.
D'ltri, F. M. 1972. The environmental mercury problem.
Chemical Rubber Co. Press, Cleveland, Ohio.
Fagerstrom, T., and A. Jernelov. 1972. Some aspects of the
quantitative ecology of mercury. Water Res. 6:1193.
Feick, G., E. E. Johanson, and D. S. Yeaple. 1972. Control of
mercury contamination in freshwater sediments. EPA-R2-
72-77. Environ. Prot. Technol. Ser. Off. Res. Monitor. U.S.
Environ. Prot. Agency, Washington, D.C.
Filip, D. S., et al. 1979. Residual heavy metal removal by an
algae-intermittent sand filtration system. Water Res.
13:305.
Fimreite, N., and L. M. Reynolds. 1973. Mercury contamina-
tion of fish in Northwestern Ontario. Jour. Wildl. Manage.
37:62.
Ganther, H. E., et al. 1972. Selenium: Relation to decreased
toxicity of methylmercury added to diets containing tuna.
Science 175:1122.
German, M. J. 1969. A water pollution survey of the
Wabigoon River. Rep. Ontario Water Resour. Comm.
Jackson, T. A. 1978. The biogeochemistry of heavy metals in
polluted lakes and streams of Flin Flon, Canada, and a
proposed method for limiting heavy metal pollution of
natural waters. Environ. Geol. 2:173.
Jackson, T. A., ed. 1980. Mercury pollution in the Wabigoon-
English River system of Northwestern Ontario, and possible
remedial measures. Progr. rep.: Rep. Wabigoon-English
River Mercury Study Steering Committee. Dep. Environ and
Ontario Minist. Environ.
Jackson, T. A., and R. N. Woychuk. 1980. The geochemistry
and distribution of mercury in the Wabigoon River system.
In T. A. Jackson, ed. Mercury pollution in the Wabigoon-
English River system of Northwestern Ontario, and possible
remedial measures. Progr. rep.: Rep. Wabigoon-English
River Mercury Study Steering Committee. Dep. Environ, and
Ontario Minist. Environ.
-------
DREDGING AND BIOMANIPULATION AS RESTORATION TECHNIQUES
101
Jernelov, A., L. Landner, and T. Larsson. 1975. Swedish
perspectives on mercury pollution. Jour. Water Pollut.
Control. Fed. 47:810.
Jernelov, A., and H. Lann. 1973. Studies in Sweden on
feasibility of some methods for restoration of mercury-
contaminated bodies of water. Environ. Sci. Technol. 7:712.
Kinniburgh, D. G., and M. L. Jackson. 1978. Adsorption of
mercury (II) by iron hydrous oxide gel. Soil Sci. Soc. Am.
Jour. 42:45.
Landner, L. 1971. Biochemical model of the biological
methylatipn of mercury suggested from methylation studies
in vivo with Neurospora crassa. Nature 230:452.
Langley, D. G. 1973. Mercury methylation in an aquatic
environment. Jour. Water Pollut. Control Fed. 45:44.
Lockwood, R. A., and K. Y. Chen. 1973. Adsorption of Hg (II)
by hydrous manganese oxides. Environ. Sci. Technol.
7:1028.
Maag, G. W., and R. J. Hecker. 1972. Recovery of mercury in
solution. Jour. Environ. Qua I. 1:192.
Moore, F. L. 1972. Solvent extraction of mercury from brine
solutions with high-molecular-weight amines. Environ. Sci.
Technol. 6:525.
Parks, J. W. 1976. Mercury in sediment and water in the
Wabigoon-English River system 1970-1975. Rep. Ontario
Minist. Environ.
Parks, J. W., J. D. Hollinger, and P. M. Almost. 1980. The
transport and dynamics of total and methyl mercury in the
Wabigoon River and Clay Lake. In T. A. Jackson, ed. Mercury
pollution in the Wabigoon-English River system of
Northwestern Ontario, and possible remedial measures.
Progr. rep.: Rep. Wabigoon-English River Mercury Study
Steering Committee. Oep. Environ, and Ontario Minist.
Environ.
Reimers, R. S., and P. A. Krenkel. 1974. Kinetics of mercury
adsorption and desorption in sediments. Jour. Water Pollut.
Control Fed. 46:352.
Rudd, J. W. M., et al. 1980. Mechanisms of movement of
mercury into aquatic biota and a preliminary examination of
some amelioration procedures. In T. A. Jackson, ed.
Mercury pollution in the Wabigoon-English River system of
Northwestern Ontario, and possible remedial measures.
Progr. rep.: Rep. Wabigoon-English River Mercury Study
Steering Committee. Dep. Environ, and Ontario Minist.
Environ.
Snyder, S. L., andT. L Vigo. 1974. Removal of mercury from
aqueous solutions by N-(2-aminoethyl) aminodeoxycellu-
lose cotton. Environ. Sci. Technol. 8:944.
Spangler, W. J., et al. 1973. Methylmercury: Bacterial
degradation in lake sediments. Science 180:192.
Suggs, J. D., D. H. Petersen, and J. B. Middlebrook, Jr. 1972.
Mercury pollution control in stream and lake sediments.
Rep. 16080 HTD 03/72. Off. Res. Monitor. U.S. Environ.
Prot. Agency, Washington, D.C.
Tratnyek, J. P. 1972. Waste wool as a scavenger for mercury
pollution in waters. Rep. 16080 HUB 04/72. Water Pollut.
Control Res. Ser. Off. Res. Monitor. U.S. Environ. Prot.
Agency, Washington, D.C.
Troyer, W. 1977. No safe place. Clarke and Irwin Ltd., Toronto
and Vancouver, Canada.
Vonk, J. W., and A. K. Sijpesteijn. 1973. Studies on the
methylation of mercuric chloride by pure cultures of
bacteria and fungi. Antonie van Leeuwenhoek 39:505.
Wheatley, B. 1979. Methylmercury in Canada. Rep. Med.
Serv. Branch, Dep. Nat. Health Welfare, Ottawa, Canada.
Widman, M. U., and M. M. Epstein. 1972. Polymer film overlay
system for mercury contaminated sludge — Phase I. Rep.
16080 HTZ 05/72. Water Pollut. Control Res. Ser. U.S.
Environ. Prot. Agency, Washington, D.C.
Wilkins, W. D. 1980. Engineering and economic evaluation of
selected remedial measures. In T. A. Jackson, ed. Mercury
pollution in the Wabigoon-English River system of
Northwestern Ontario, and possible remedial measures.
Progr. rep.: Rep. Wabigoon-English River Mercury Study
Steering Committee. Dep. Environ, and Ontario Minist.
Environ.
Wilkins, W. D., and S. Irwin. 1980. Comments on the
Wabigoon River dredging proposal. In T. A. Jackson, ed.
Mercury pollution in the Wabigoon-English River system of
Northwestern Ontario, and possible remedial measures.
Progr. rep.: Rep. Wabigoon-English River Mercury Study
Steering Committee. Dep. Environ, and Ontario Minist.
Environ.
Wood, J. M. 1971. Environmental pollution by mercury. In J.
N. Pitts, Jr., and R. L. Metcalf, eds. Advances in
environmental science and technology. Vol. 2. Wiley-
Interscience, New York, London, Toronto, Sydney.
-------
102
SIMPLIFIED ECOSYSTEM MODELING FOR ASSESSING
ALTERNATIVE BIOMANIPULATION STRATEGIES
MARK L. HUTCHINS
Land and Water Resources Center
University of Maine at Orono
Orono, Maine
ABSTRACT
The technique of "loop analysis" was applied to a variety of hypothetical lake ecosystems in an
attempt to assess qualitatively the potential of biomanipulation as a lake restoration technique.
Loop analysis is particularly well suited for evaluating complex ecosystems, where in many cases
system interactions can be specified only qualitatively. The technique is based on the equivalence
of a set of linear differential equations at or near equilibrium and their matrices and loop diagrams.
The only information required is that the number of system components and their direct
interactions (in terms of positive, negative, or no impact) be specified. In general, results support
current ecosystem theory and several recent field studies. In simple chain systems, perturbations
typically produce alternating effects, with perturbations from the top having more impact than
perturbations from the bottom. Similarly, planktivorous fish appear to play an important role in
structuring lower trophic levels. However, in more complex food webs, results are not nearly so
predictable, and in some cases are completely opposite to what might be expected. For example, in
certain systems an increase in the density of game fish leads to an eventual decrease in the game
fish; however, algae also experience a decrease. On the other hand, a decrease in game fish leads
to an increase in algae, a result which may have serious fisheries management implications. It is
conceivable that, in some lakes, changes in water quality may be directly attributable to fishing
pressure rather than to a change in nutrient status.
INTRODUCTION
Biomanipulation as a potential lake restoration tech-
nique has been discussed for a number of years. In
particular, Patten (1973) and Shapiro, et al. (1975)
pointed out the relative ease with which upper trophic
levels could be manipulated to control lower levels.
Recently, more and more experimental evidence is
surfacing to support the contention that biomanipula-
tion may be a viable lake restoration technique
(Andersson, et al. 1978; Briand and McCauley, 1978;
Gliwicz, 1975; Haertel, 1977; Henrikson, et al. 1980;
LeBrasseur, et al. 1978; Lynch, 1979; Molotkov, et al.
1978; Porter, 1977; Roman, 1978; Smyly, 1978; von
Ende, 1979). In fact, in some situations it may be the
only feasible technique.
Biomanipulation certainly is not a new concept; it
has been a standard tool of fisheries management for
many years. Understandably, the emphasis has been
placed almost exclusively on the upper trophic level
fisheries. Unfortunately, there has been little regard for
the rest of the ecosystem. What is presently needed is a
more holistic ecosystem management approach to
direct system productivity not only toward desirable
upper level components but also away from un-
desirable components such as algae.
In the past, many lake restoration activities have
focused on nutrient abatement using the simple
input/output models of Dillon and Rigler (1974) and
Vollenweider (1975, 1976) to predict resulting water
quality improvements. Unfortunately, no similar tech-
nique has been developed to assess potential impacts
of various biomanipulation strategies; as a result,
ecosystem management to improve water quality is
essentially untested on a whole-lake scale, and
application will likely remain limited to isolated case
studies until a suitable model is developed. The
purpose of this paper is to present a simple modeling
technique called "loop analysis," which may satisfy
some of these needs.
MODEL DESCRIPTION
Loop analysis was developed by Levins (1974, 1975)
to qualitatively evaluate ecosystem stability and
interaction. The technique has been applied to both
terrestrial and aquatic ecosystems (Levins, 1975),
model plankton communities (Lane and Levins, 1977),
and a hypothetical aquatic food web (Briand and
McCauley, 1978). This paper extends past work by
examining a variety of hypothetical biomanipulation
strategies.
Mathematically, loop analysis is founded in matrix
algebra and is based on the equivalence of linear
differential equations at or near equilibrium. The
equations are of the form
dX,
dt
, X2t X,,
X,,)
where x's are the system variables. Variables are
usually species or trophic level abundance, but can also
be nutrient levels or even such factors as toxic
byproducts or predation pressures. At or near
-------
DREDGING AND BIOMANIPULATION AS RESTORATION TECHNIQUES
103
equilibrium, the behavior of the system depends on the
properties of the community matrix.
811
an
811
812
812
813
313
813
A =
3m
3m
3nn
where the elements ay are given by
9F,
3X,
and are simply the
coefficients of the X| in Eq. (1) for dxi/dt, evaluated at
equilibrium. For example, an is the coefficient used to
describe the impact of Xi upon itself; the coefficient a
describes the impact of Xz on Xi, etc. The technique of
loop analysis is basically a shorthand method for
solving the community matrix.
Schematically the systems may be represented by
directed diagrams, as in Figure 1. Here the variables
Xi, Xz, and Xa are the vertices of the diagram, and the
interactions between the variables are represented by
oriented links. In most ecosystems, the actual values of
the matrix coefficients are not known; however, the
direction and signs of interactions can usually be
specified. In the above system, arrows indicate positive
links and circles are negative links. For example, Xi
might be an algal species, Xz an herbivorous
zooplanktore, and Xa an omnivorous zooplanktore. Thus
Xz feeds upon Xi. and Xa preys upon both Xi and Xz.
It should also be noted that variables may affect
themselves, either positively or negatively. In Figure 1,
Xi has a negative impact on itself which is termed self-
damping. In general, all resources which are not self-
reproducing, such as mineral nutrients, organic matter,
or detritus, are self-damped. If the resource is not
specifically included in the diagram, then the organism
which requires that resource incorporates the self-
damping, as in Figure 1. Thus, all ecosystems are
limited or self-damped by their ultimate food source.
32
Figure 1. — Loop diagram and community matrix of a three-
variable ecosystem.
required to perform loop analysis, application of the
technique to selected lake ecosystems, and potential
significance and limitations of results. For a formal
treatment of loop analysis theory and methodology, the
interested reader should consult Levins (1975).
The information necessary to perform loop analysis
is minimal; all that is required is that variables be
selected which adequately describe the system, and
that the direct variable interactions be specified
qualitatively. For example, Figure 2a depicts a system
containing a nutrient, an algal species, and zoo-
plankton species. It can be noted by the directed links
between variables that the nutrient is self-damped and
that it benefits the algae. The algae deplete the nutrient
but enhance zooplankton. The zooplankton feed on the
algae, as indicated by the negative link. It is important
to note at this point that the diagrams are not
descriptive of whole ecosystems. Numerous other
inputs and outputs may be involved (sunlight,
temperature, losses due to non-predatory mortality);
these are not shown. The diagrams attempt to portray
only system variables; all other components are
assumed to be constant, or their rates of reaction are
constant.
In Figure 2b, two species of algae compete for a
single nutrient. Zooplankton feed on algae Az, but A
have no direct impact on Ai.
Figure 2c is a bit more complicated. There are two
nutrients, two algal species, and a single zooplankton
species. Algae Ai require both nutrients, but A 2 need
only Nz. Zooplankton feed on both algal species.
The sample diagrams are not intended to represent
real lake ecosystems; rather, they are meant to
illustrate the flexibility of diagram construction. It is
Perturbation Effect on Level of
H A Z
+Z
N ..... •*• 0 +
A ..... 0 0 +
+ 0
(a)
Perturbation Effect on Level of
(b)
Perturbat Jon Effect on Level of
7 ' + 7
+ 7 ? +
7 0 7
+ ' 0' 7
7+770
Because the theoretical development and method-
ology required for actually solving loop analysis
problems is complicated and quite lengthy, this author
will not attempt to elaborate on it in this paper. Rather,
the following discussion will focus on information
(c)
Figure 2. — Loop diagrams and perturbation analyses of
partially specified lake ecosystems.
-------
104
RESTORATION OF L^KES AND INLAND WATERS
quite easy to see that the range of combinations and
permutations of different variables and interactions is
limited only by imagination.
The formal analysis of loop diagrams may have
several end results, but for this paper only that aspect
which deals with the effects of biomanipulation will be
emphasized. In particular, we will examine how a
system assumed to be at or near equilibrium adjusts to
a new equilibrium as a result of changes imposed on
the system. These changes or perturbations may be
step changes, such as an increase in phosphorus input,
or evolutionary changes which enhance a biological
variable's ability to survive in the system. Recently the
technique has been expanded to include transient and
periodic perturbations (Flake, 1980), but this will not be
considered in this paper.
In Figure 2, perturbation analyses are illustrated for
sample systems. The specific perturbation is indicated
by a +N, +A, etc., and the predicted effect of each
perturbation on the levels of all system variables is
shown by the matrix-type diagram* To avoid re-
dundancy, only positive perturbations are illustrated;
the effects of negative changes can be derived simply
by reversing the signs.
In Figure 2a, one would predict that an increase in
the nutrient supply would increase both the nutrient
and the zooplankton levels, but would not affect the
level of algae in the system. However, it is important to
note that while the level of algae is unaffected, its rates
of interaction with other variables may change. Since
the level of nutrient has increased, algae would
presumably respond with an increased growth rate.
However, zooplankton predation on algae increases
because zooplankton themselves increase. Thus, the
increase in algae productivity is passed directly to the
zooplankton without increasing algal abundance.
The second perturbation in Figure 2a is that of
enhancing the algae component. This results in
increasing zooplankton abundance with no changes in
either the nutrient or the algae. Finally, the enhance-
ment of zooplankton results in an increase in nutrient
levels, a decrease in algal biomass, and no change in
zooplankton.
In Figure 2b, results of a similar nature are portrayed.
It can be seen that nutrient levels are affected only by
perturbations of Ai and that levels of t\z are affected
only by changes in zooplankton. Another item of
interest is the effect of changes in Ai on zooplankton
levels. The double negative sign indicates that two
viable pathways through which Ai may react with
zooplankton, both of which have negative impact. It is
important to remember, however, that since this
technique is entirely qualitative a double negative does
not necessarily have any more quantitative signifi-
cance than does a single negative. Nevertheless, it is
interesting qualitatively, and for that reason it is
included in the results.
* This matrix is in no way related to the community matrix described previously.
It is simply a convenient formatter portraying results.
Finally, in Figure 2c it may be noted that the matrix is
largely filled with question marks. This results when a
perturbation affects a variable through two or more
viable pathways, the signs of which are opposite.
Because the magnitude of the impacts is unknown, the
net effect cannot be resolved; thus, the results are
ambiguous. Increases in system complexity— particu-
larly in foodweb situations — quite often lead to
increases in the number of ambiguous results. This
factor very possibly could be the most critical limitation
in applying loop analysis to complex ecosystems.
APPLICATION
In the following pages, the author will present three
hypothetical lake ecosystems and perturbation analy-
ses. The attempt is to portray a range of "real" systems
with the ultimate objective of reducing algae levels
through manipulation of other biological components.
Biological components are specified by functional
group rather than at the species level in an attempt to
avoid unnecessary duplication. Thus, for example, the
forage fish component may be composed of many
species of minnows, smelts, and young game fish.
Similarly, algae may be grouped by size and edibility
instead of by the more traditional green/blue —
green/diatoms grouping. The intent is that items in a
group act and react similarly, irrespective of their
specific taxonomies.
In Figure 3a, a simple chain system is illustrated
which contains phosphorus (P), algae (A), zooplankton
(Z), forage fish (Sm) (predominately smelts), and game
fish (Sa) (predominately landlocked salmon). The
coldwater fishery was selected because of its signifi-
cance to Maine. A similar diagram may describe many
warmwater fisheries. It should be pointed out that in
this system — and in those following — the game fish
are considered to be self-damping. In this case, self-
damping is included to account for the influence of man
on the ecosystem. Most lakes support sport fisheries
which have a direct negative impact on the game fish
and an indirect impact on other system components.
The end result is that the game fish level is no longer
totally responsive to the rest of the system; the
population is damped by man's influence, regardless of
what the rest of the system is doing.
Results of the perturbation analysis are illustrated in
Figure 3a. An increase in phosphorus input leads to an
increase in the levels of all system components, which
corresponds with many field observations and also
with common sense. Within limits, all biological
components should benefit from an increase in fertility.
Increases in the other components produce variable
responses. For example, an enhancement of zooplank-
ton reduces algae, increases phosphorus, and in-
creases smelt and salmon levels. In general, it can be
seen that components below the perturbed level
respond with alternating effects, while those above
experience identical responses. It is also interesting to
note that the highest trophic level in the chain
responds to perturbations in all levels, and vice versa,
perturbations in the highest level result in changes in
all other levels. That upper trophic level components
should inherently have considerable influence over
-------
DREDGING AND BOMANIPULATION AS RESTORATION TECHNIQUES
105
ecosystem structure and response has been stated
previously (Patten, 1973).
Finally, the impact of perturbations on the algae
component can be examined. It can be seen that
enhancement of zooplankton and salmon should
produce decreases in algae levels, and increases in the
phosphorus and smelt components. Enhancement of
the algae itself produces no change in algae, because
of a reduction of its own food supply and an increase in
predation.
The results presented in the previous example are for
the most part not surprising; the structure and
dynamics of simple food chains have been known for
some time. However, minor aberrations in the chain
structure may have drastic impacts on system
dynamics, and in more complex food webs many
results are not at all predictable.
The system illustrated in Figure 3a probably
represents the past ecosystem for Echo Lake, Maine.
Salmon production was not satisfactory, and in an
attempt to improve productivity, landlocked alewives
were introduced as an additional forage source. This
was a reasonable move from a fisheries management
perspective. From Figure 3a, loop analysis would
predict that enhancing the forage fish should increase
salmon biomass. Unfortunately, to date the salmon
refused to feed significantly on the alewives.
The modified ecosystem is illustrated in Figure 3b.
Schematically, the only change in the system is the
addition of the alewife component connected with the
zooplankton component; however, the results of the
perturbation analysis dramatically differ from the
previous example. The most obvious difference is the
lack of response of most levels to most perturbations. In
particular, increasing phosphorus input increases
algae and alewives with no changes in the other
components. Similarly, enhancement of the salmon
increases alewives, decreases smelts, but produces no
other changes. However, enhancement of the alewife
component decreases all other components except
algae, which increase, and alewives, which remain the
same. By the same token, alewives respond positively
to enhancement of all other components except smelt.
Perturbat ion Effect on Level of
P A Z Sm Sa
(a)
Perturbation
+P ...
+A ...
+Z ...
+A1 •••
+Sm • • •
+Sa ...
El
f
• • 0
.. 0
• - 0
•• 0
Ffect
A Z
+ 0
0 0
0 0
+
0 0
0 0
on Level of
Al Sm Sa
+ 00
+ 00
+ 00
0
0 +
+ 0
(b)
which decrease alewives because they compete
directly for food.
Interpretation of these results indicates that in-
troduction of alewives into Echo Lake has effectively
short-circuited the existing food chain, benefiting the
alewives and, importantly, also algae. Furthermore,
since alewives are neither harvested by man nor
preyed upon by game fish in this system, they are free
to act as a system buffer, readily absorbing changes
while other components — with the possible exception
of algae — remain largely unaffected. Fortunately,
Echo Lake is oligotrophic, and nuisance algae
conditions have not beome a problem. Nevertheless, a
large and unutilized alewife population presently
exists; salmon productivity has not improved.
+ 7 + + + 7
0 --++ + +--
7 0 -- 7 7 «+
7 7 0 » + 7
+ -- + 0 + 7
+ — + +0
7 ++ — — 7 0
Figure 3. — Loopdiagramsand perturbation analyses of past(a)
and present (b) simplified ecosystems for Echo Lake, Maine.
Figure 4. — Loop diagram and perturbation analysis of the
eight-variable ecosystem for Hermon Pond, Maine.
The final system, presented in Figure 4, is at least in
part representative of another Maine lake, Hermon
Pond. (Ecosystem data are still being collected.) System
components include a nutrient source, large and small
algae, large and small herbivorous zooplankton, an
aquatic insect group, forage fish, and game fish. The
insect level is composed primarily of Chaorborus
species, with minor amounts of predaceous zooplank-
ton. The forage fish level is predominately smelt, yellow
perch, and small white perch. The game fish are mainly
large white perch, smallmouth bass, and a few chain
pickerel. The large algae component contains most of
the undesirable species and is therefore the level to be
minimized.
Results of the perturbation analysis are mixed —
some responses are similar to those for chain systems,
and others are quite different. Conflicting pathways
have produced ambiguous results, but in general they
do not appear to limit seriously the utility of the
analysis. As in the chain system, an increase in
phosphorus input increases most other components.
However, enhancement of the game fish decreases all
components except small algae and insects; as
illustrated by this system, a general conclusion is that
the alternating impacts apparent in chain systems are
invalid in web systems.
It may be noted that only one perturbation produces
an unambiguous negative effect on the large algae —
that of enhancing game fish. Even increasing large
zooplankton does not definitely reduce large algae. This
is because the impact of the direct path — large
zooplankton to large algae — is at least in part negated
by the impact of the indirect path — large zooplankton-
forage fish-insects-small zooplankton-small algae-
phosphorus-large algae. Thus, it appears that a
management strategy for this pond might be to
-------
106
RESTORATION OF LAKES AND INLAND WATERS
increase game fish, perhaps through stocking or
through harvest reduction (or to decrease forage fish).
Though a decrease in small zooplankton should also
have a negative impact on large algae, man's ability to
directly manipulate zooplankton levels is probably quite
minimal.
DISCUSSION AND SUMMARY
For biomanipulation to become a reality in lake
restoration efforts, a technique must be available to
predict the entire community response to that
manipulation. As a first-cut qualitative approach, loop
analysis satisfies some of these needs. The preceding
examples demonstrate the generality and versatility of
the technique. The variety of ecosystem structures and
interactions that can be examined is in theory infinite,
and application should be possible for almost any
aquatic ecosystem imaginable. Results from the
perturbation analyses often support ecological theory
and field observations, but are sometimes surprising in
that some are entirely opposite to what one would
intuitively expect. It is in these counter-intuitive results
that the real strength of loop analysis lies — in firm
partnership with biological knowledge, the insight
gained from loop analysis can help unravel complex
system interactions which have long frustrated
ecological researchers. And when a holistic approach
is adopted, it becomes apparent that even results
which seem unreasonable on the surface are in fact
quite reasonable when the system as a whole is
considered.
Loop analysis is not, however, without faults. As with
any mathematical model, underlying assumptions
must be kept foremost in mind to prevent misusing the
technique. First, and perhaps most important, it is
assumed that ecosystems can be described by linear
equations. In fact, linearity in real systems is probably
the exception rather than the rule. Nevertheless, the
analysis attempts only to identify qualitative trends in
ecosystem dynamics — the magnitude of those trends
cannot be deduced (nor should it be inferred). In this
light, it is assumed that a linear approximation is
adequate.
Second, it is assumed that the ecosystem com-
ponents and interactions have been adquately de-
scribed. It has been shown that seemingly minor
changes in system structure may result in major
changes in system response to perturbations. In
applying loop analysis to real lake systems, it becomes
imperative that the structure of the specific ecosystem
be known in detail.
Third, it is assumed that the systems are in
equilibrium or at least in moving equilibrium. While
this assumption is common with many types of models,
it still warrants emphasis since many culturally
eutrophic lakes are probably not in equilibrium,
particularly ecologically.
Ultimately, improvement in managing community
structure in lake ecosystems will result primarily from
field experimentation rather than from mathematical
modeling. Lake systems are too complex to expect
otherwise. However, field experimentation, unless
guided by theoretical knowledge of subsystem interac-
tions, involves testing a multitude of management
possibilities and sorting out their ecosystem impacts
from the effects of all uncontrolled, natural variations
occurring simultaneously. Loop analysis can play a
much-needed intermediate role by providing the
insight necessary to guide future experimentation.
REFERENCES
Andersson, G., et al. 1978. Effects of planktivorous and
benthivorous fish on organisms and water chemistry in
eutrophic lakes. Hydrobiolgia 59:9.
Briand, F., and E. McCauley. 1978. Cybernetic mechanisms in
lake plankton systems: and how to control undesirable
algae. Nature 273:228.
Dillon, P. J., and F. H. Rigler. 1974. A test of a simple nutrient
budget model predicting the phosphorus concentrations in
lake water. Jour. Fish. Res. Board Can. 31:1771.
Flake, R. H. 1980. Extension of Levins' loop analysis to
transient and periodic disturbances. Ecol. Model. 9:83.
Gliwicz, Z. M. 1975. Effect of zooplankton grazing on
photosynthetic activity and composition of phytoplankton.
Verh. Verein. Limnol. 19:1490.
Haertel, L. 1977. Effects of zooplankton grazing on nuisance
algal blooms. Proj. A-047-SDAK. South Dakota State
University; Brookings.
Hennkson, L. etal. 1980. Trophic changes, without changes
in the external nutrient loading, hydrobiologia 68:257.
Lane, P., and R. Levins. 1977. The dynamics of aquatic
systems. 2. The effects of nutrient enrichment on model
plankton communities. Limnol. Oceanogr. 22:454.
LeBrasseur, R. J., et al. 1978. Enhancement of sockeye
salmon (Oncorhynchus nerka) by lake fertilization in Great
Central Lake: summary report. Journ. Fish. Res. Board Can.
35:1580.
Levins, R. 1974. The qualitative analysis of partially specified
systems. Ann. N.Y. Acad. Sep. 231:123
1975. Evolution in communites near equilibrium.
Pages 16-50 M. Cody and J. Diamond, & eds. Ecology and
evolution of communities. Belknap.
Lynch, M. 1979. Predation, competition, and zooplankton
community structure: an experimental study. Limnol.
Oceanogr. 24:253.
Molotov, V. E., V. G. Svirskiy, and N. A. Dmitriyenko. 1978.
Phyto- and zooplankton of fish ponds of the maritime
province. Hydrobiol. Jour. 6:96.
Patten, B. E. 1973. Need for an ecosystem perspective in
eutrophication modeling. Pages 83-87 E.J. Middlebrooks,
D.H. Falkenborg, and J. E. Maloney, & eds, Modeling the
eutrophication process. Utah Water Res. Lab. College Eng.
Utah State University, Logan.
Porter, K. G. 1977. The plant-animal interface in freshwater
ecosystems. Am. Sci. 65:159.
Roman, M. R. 1978. Ingestion of the blue-green alga
Trichodesmium by the harpactoid copepod, Macrosetella
gracillis. Limnol. Oceanogr. 23:1245.
Shapiro, J., V. Lamarr and M. Lynch. 1975. Biomanipulation
— an ecosystem approach to lake restoration. Manuscript.
Contro. No. 143. Limnol. Res. Center, University of
Minnesota, Minneapolis.
Smyly, W. J. P. 1978. Further observations on limnetic
zooplankton in a small lake and two enclosures containing
fish. Fresh. Bio. 8:491. a
-------
DREDGING AND BIOMANIPULATION AS RESTORATION TECHNIQUES
Vollenweider, R. A. 1975. Input-output models with special
reference to the phosphorus loading concept in limnology.
Schweiz. Z. Hydrol. 37:53.
1976. Advances in defining critical loading levels
for phosphorus in lake eutrophication. Me. 1st. Ital. Idrobiol.
33:53.
von Ende, C. N. 1979. Fish predation, interspecific predation,
and the distribution of two Chaoborus species. Ecology 60.
-------
108
RESPONSE OF ZOOPLANKTON IN PRECAMBRIAN
SHIELD LAKES TO WHOLE-LAKE CHEMICAL
MODIFICATIONS CAUSING pH CHANGE
D. F MALLEY
P S. S. CHANG
Department of Fisheries and Oceans
Freshwater Institute
Winnipeg, Manitoba, Canada
ABSTRACT
Two lakes, 227 and 223, in the Experimental Lakes Area of northwestern Ontario, have been
subjected, respectively, to whole-lake addition of fertilizer, nitrogen and phosphorus, and to
addition of sulf uric acid. Effects on their zooplankton populations are believed to be brought about
largely by changes in the pH. The low endogenous concentrations of dissolved inorganic carbon
render these lakes prone to extreme pH change. Phosphorus input to lake 227 was increased 10-
fold each year from 1969 to 1 974 by adding N:P in 15:1 ratio. Maximum mid-summer biomass of
cladocerans and calanoids declined each year after fertilization reaching very low levels by 1 972.
Cyclopoid biomass was only moderately reduced following fertilization. Rotifer biomass increased
manyfold m 1970 and 1971 but declined to very low levels by 1974. Mid-summer epilimnion pH
levels recorded were frequently above 10.0. Enhanced decomposition caused the anoxic zone on
the lake bottom to deepen significantly. Changes in crustacean biomass are thought to be due to
loss of oxygenated, near-neutral habitat within the water column A relationship between rate of
phosphorus loading, endogenous dissolved inorganic carbon, and epilimnion pH from 6.7 to 5.84
and severely reduced the dissolved inorganic carbon. By 1979 neither total zooplankton biomass
nor species diversity has changed appreciably. Nevertheless, the species composition changed
somewhat Rotifers showed various responses. No predictive relationships were evident between
species' tolerance of high or low pH.
INTRODUCTION
The Experimental Lakes Area (ELA) in the Pre-
cambrian Shield of Ontario was selected in the 1960's
for whole-lake experimental study of eutrophication.
One reason these lakes were suitable was the low
conductivity of the waters, making them amenable to
change to new chemical states by controlled additions
of substances (Johnson and Vallentyne, 1971). The
lakes are poorly buffered and thus are susceptible to
changes in pH. Alkalinity results principally from the
bicarbonate system derived by carbonic acid weather-
ing of the alumino-silicate bedrock (Brunskill, et al.
1971). Bicarbonate concentration in ELA lake surface
waters averages 4.1 mg HCO3 I"1 (67 p moles I"1) ,
among the lowest reported in the world, ranking with
lakes in the Adirondack Mountains, N.Y., the Cairn-
gorm area, Scotland, (Armstrong and Schindler, 1971),
and parts of Scandinavia (Wright, et al. 1976).
Changes in pH of ELA lakes have been brought about
experimentally in two ways. Adding phosphate and
nitrate fertilizer caused pH to increase. Enhanced
photosynthesis in Lake 227 drained the total epi-
limnion CO2 pool to such an extent that bicarbonate
was negligible. The hydroxyl ion generated in CO? and
nitrogen uptake by plankton dominated the alkalinity,so
that pH of 10.0 or higher was frequently recorded in the
epilimnion. In another experiment, the pH of Lake 223
was reduced by adding sulfuric acid. pH was lowered
0.25 to 0.50 pH units per year to quantify the rate of
acidification and the biological and chemical effects
resulting from the addition of known amounts of acid
(Schindler, et al. 1980).
FERTILIZATION OF LAKE 227
Lake 227 is a small, oligotrophic lake, 5.0 hectares in
area, with a mean depth of 4.4 m and maximum depth
of 10.0 m (Schindler, et al. 1971). It was selected in
1969 for its unusually low levels of dissolved inorganic
carbon (DIG) (70 /j moles l~1 average in epilimnion in
1969 prior to fertilization) to test whether carbon
shortage limited eutrophication.
In 1969 the lake received 0.34 g rrf2of P as Na2HP04
and 5.04 g m 2 of N as NaNO3 in 17 equal weekly
additions starting in late June. During May to October
of each year from 1970 to 1974, 21 weekly additions
were make of P asH3PO4andNas NaNO3for an annual
total of 0.48 g rrf2 and 6.29 g rrf2 N, and a N:P ratio of
13:1 by weight. The fertilization regime was changed in
1 975 to 1978 so that N:P ratio was about 5:0 by weight.
During the latter years 20 weekly additions were make
for annual loadings of 0.46 gm~2P as H3PC>4 and 2.25 9
rrf 2Nas NaNO3. The lake, methods of fertilization, and
the changes in chemistry and phytoplankton following
fertilization are described by Schindler, et al 1971-
1972; 1973; Schindler and Fee, 1974; Findlay and
Kling, 1975; Schindler, 1975; 1977; and Findlay, 1978
-------
DREDGING AND BOMANIPULATION AS RESTORATION TECHNIQUES
109
RESPONSE OF LAKE 227 TO
FERTILIZATION: 1969-1974 PERIOD
Both primary production and standing algal biomass
increased in Lake 227 following the addition of P and N.
Algal biomass increased several-fold after fertilization
in 1969 (Schindler, et al. 1971) and reached a
maximum of about 20 times the pre-fertilization values
in late July 1 972. Algal species composition during the
ice-free season changed with addition of N and P in
13:1 ratio; chlorophytes and cyanophytes replaced
cryptophytes and chrysophytes as dominants (Schin-
dler, et al. 1973). Edible species of algae were
abundantly available to the zooplankton during 1969 to
1974 (Kling, pers. comm.).
The increased primary production in Lake 227 was
associated with mid- or late morning pH values above
10.0 in the epilimnion on a number of sampling dates
in 1970, 1972, and 1973 (Schindler, et al. 1973). pH
fluctuated little diurnally (Schindler and Fee, 1973).
Photosynthetic activity draws from the free CO2 pool
which in turn is replenished from carbonate alkalinity
or by invasion of CO2 from the atmosphere. CO2
invasion was not sufficiently rapid to supply all the
required CO2 for photosynthesis on sunny mid-summer
days in Lake 227 (Schindler and Fee, 1973). The
removal of CO2 by the algae was sufficiently great to
drain the free-CCbpool and to elevate the pH reactions
such as:
(King, 197Q).
The anoxic bottom layer thickened in the years
following fertilization reaching a maximum thickness in
1972. In mid-July 1972 several extreme conditions
occurred together. The epilimnion, 0 to 2 m, was at pH
above 10.0, but was well-oxygenated. Below 2 m, pH
dropped to about 7, but Oz at 3 m and below was less
than 2 mg I"1
RESPONSE OF ZOOPLANKTON IN LAKE
227
Seasonal changes in abundance of species of
crustaceans and rotifers in Lake 227 from 1969 to
1974 are described by Malley, et al. (In prep. a).
Average number of individuals of crustacean species
during May to September in a column of water under 1
m2 of lake surface at the center of the lake for 1 969 to
1978 are reported in Tables 1 to 3. Typically, the
epilimnion, metalimnion, and hypolimnion were sampl-
ed separately and the numbers in each stratum per m2
were weighted according to the volume of the stratum
and summed together to give the number per m2. Data
for 1975 and 1976 are omitted for the tables because
only the uppermost 2 m of the water column were
sampled in those years. Total biomass of groups of
zooplankton including rotifers is shown in Figure 1 for
1969 to 1974. Dry weight biomass for individuals of
each species was calculated from simple geometric
shapes approximating the size and shape of each life
stage or size category (Lawrence, et al. In prep).
Zooplankton sampling methods are described by
Chang, et al. (1980).
Populations of the cladocerans Bosmina longirostris,
Diaphanosoma brachyurum, Daphnia retrocurva, and
Holopedium gibber urn declined on the average with
fertilization in the summers of 1970 and 1971
compared with abundances in the first year of
fertilization, 1969 (Table 1). All four species were
severely reduced in numbers or not recorded at all in
1972 and 1973. All were recorded again in 1974 but
mostly at low densities except for an unusually high
density of D. brachyurum on one date. By 1978
numbers of D. brachyurum were very similar to those
in 1969. B. longirostris was abundant on one sampling
date in 1978, resulting in a seasonal average as high as
in 1969, but the seasonal pattern was very different. In
1969 the species was well represented throughout
May to September but in 1978 was recorded only
during June and July. D. retrocurva and H. gibberum
failed to recover by 1978 to densities seen in 1969.
Reflecting these abundances, total dry weight biomass
of cladocerans was lower in 1970 and 1971 than in
1969 and very low in 1972 and 1973 (Figure 1). The
large biomass on one date in 1974 is due to the high
density of D. brachyurum. Rare species of cladocerans
in the Lake 227 samples included Ceriodaphnia sp.,
Chydorus sphaericus, and Alona sp. (Chang, et al.
1980).
The calanoids, originally dominated by Diaptomus
minutus. declined in population sizes of adults, nauplii,
and copepodids from 1969 to 1972. Minimum number
were found from August 1972 through 1973. Popula-
tions increased slightly in 1974 and after, but by 1978
were far below 1969 abundances (Table 2). Epischura
lacustris was the dominant calanoid in Lake 227 in
1974, 1977, and 1978. Diaptomus leptopus was found
occasionally in Lake 227. Total dry weight biomass of
calanoids declined to a minimum after July 1972 and
remained low throughout 1973 and 1974 (Figure 1).
tv, 350.0-
E 315.0-
g1 280.0-
CO 245.0-
^ 210.0-
Q 175.0-
m
|_ 140.0-
O 105.0-
5 7O.O -
> 35.0-
Q 0.0-
v_> i_ M L/ w \j t. rv M
^
1 1 1
j\JL
1969 1970 1971 1972 1973 1974
YEAR
Figure 1. — Dry weight biomass (mg rrr2) of four groups of
zooplankton in L227 during 1969 to 1974.
Cyclopoids were represented by three species (Table
3) which shifted in dominance from 1969 to 1971. M.
edax dominated in summer 1969; T. prasinus in 1970,
and C. bicuspidatus in 1971. Each species was present
in low numbers or not recorded in 1972, 1973, and
1974 except for M. edax. All four species were present
in 1977 and 1978. Populations of cyclopoid nauplii and
copepodids did not decline from 1969 to 1971, but
-------
no
RESTORATION OF LAKES AND INLAND WATERS
Table 1. — Numberm 2of cladocerans in Lake 227 averaged over the May to September period during the years 1969 to 1974 and
1977 and 1978.
Species
Bosmina longirostris
Diaphanosoma brachyurum
Daphnia retrocurva
Ho/opedium gibberum
Table 2. — Number rrf 2 of calanoid
1969
34,450
8,400
4,100
2,800
copepods in
1970
19,150
3,750
1,600
1,700
Lake 227
1971
24,350
3,700
1,800
800
1972
1,800
700
200
0
1973
350
500
0
0
averaged over the May to September
1974 and 1977 and
Species
Diaptomus minutus
aduljs
Epischura lacustris
adults
Calanoid nauplii
N,-NB
Caalanoid copepodids
Ci-Cs
Table 3. — Number rrf2 of cyclopoid
1969
32,650
0
61,400
110,500
copepods in
1970
16,400
0
66,300
57,450
Lake 227
1971
10,650
0
72,750
25,850
1978.
1972
2,800
0
22,000
2,350
1973
25
0
1,750
0
averaged over the May to September
1974 and 1977 and
Species
Cyclops bicuspidatus thorn.
adults
Mesocy/ops edax
adults
Tropocyc/ops prasinus mex.
adults
Cyclopoid nauplii
N,-N6
Cyclopoid copepodids
C,-C5
1969
0
2,550
1,700
55,600
59,750
1970
50
350
700
95,550
16,600
1971
2,000
400
0
109,900
54,350
1978.
1972
100
100
0
45,200
14,150
1973
50
0
0
49,200
9,150
1974
4,050
22,650
700
150
1977
450
4,450
0
0
period during the years
1974
0
150
1,950
1,250
1977
0
1,250
6,500
2,300
period during the years
1974
0
550
0
57,850
21,800
1977
300
550
150
1978
37,900
11,900
1,000
0
1969to
1978
7
1,050
11,400
4,800
1969 to
1978
350
950
100
16,650 42,700
4,000 15,100
were smaller in 1972 and 1973. Overall, cyclopoids did
not decline in numbers with fertilization as dra-
matically as did calanoids and cladocerans. This is
illustrated further for dry weight biomass in Figure 1.
Nevertheless, during the 1969 to 1974 period standing
cyclopoid biomass was at a (summer) minimum in
August 1972.
Individual species of rotifers responded variously to
fertilization from 1969 to 1974. Nevertheless, numbers
of a species were generally higher in 1970 and 1971
than during 1969 and most species were reduced in
1 974 to below their abundances in 1 969. These results
are shown in Figure 1 for total rotifer dry weight
biomass. The increases in biomass in 1970 and 1971
reflect larger population sizes of Keratella cochlearis,
Polyarthra vulgaris, and Anuraeopsis fissa.
CAUSES OF ZOOPLANKTON DECLINE
IN LAKE 227
The decline of crustacean zooplankton, particularly
herbivores, with fertilization of Lake 227 was an
unexpected result. Eutrophication usually increases
the standing crop of herbivores (Brooks, 1969; Smith,
1969; Hillbricht-llkowska and Weglendska, 1970;
LeBrasseur and Kennedy (1 972) or produces no change
(Briand and McCauley, 1978). The most plausible
explanation for the decline of the zooplankton in Lake
227 is that the combination of high pH in the
epilimnion and low 02 below, reaching extremes in
mid-July 1972, diminished conditions suitable for
zooplankton reproduction and survival within the water
column.
Published information on effects of high pH on
zooplankton is sparse. Davis and Ozburn (1969) report
that the maximum pH at which Daphnia pulex survived
was 10.4 in water of 70 mg HCOs but reproduction
occurred only up to pH 8.7. Limits in water of 10 mg
HCO3 I 1 were narrower, 10.3 for survival and 8.2 for
reproduction. O'Brien and DeNoylles (1972) report that
10.8 was acutely lethal to Ceriodaphnia reticulata in
the laboratory. Mid-morning pH of 10.6 in ponds was
associated with the disappearance of this species.
Given the very soft water of Lake 227, it is likely that pH
of 10.0 or above reflects conditions lethal for at least
some of the crustaceans. Escape from high pH by
remaining in the metalimnion would expose the
zooplankton to lowO2. Some species of cyclopoids are
reported to be able to withstand low O2 or anoxic
conditions (von Brand, 1944; Chaston, 1976). Cyclo-
poids appear to tolerate high pH better than clado-
cerans and calanoids judging from the pH limits given
by Lowndes (1 952). Resistance of cyclopoids to lowC>2
and/or high pH may account for their better overall
survival compared with cladocerans and calanoids
An alternate hypothesis that predation by Chaoborus
(Diptera:Chaoboridae) caused the decline of clado-
cerans and calanoids is discussed by Malley et al (In
-------
DREDGING AND BIOMANIPULATION AS RESTORATION TECHNIQUES
111
prep a.). Fertilization may have greatly enhanced the
numbers of Chaoborus in Lake 227 but no quantitative
data are available. Nevertheless, it is considered
unlikely that predation would be the major cause of
decline in these zooplankton species to such extremely
low levels. Starvation and predation by fish are not
considered to be important factors causing the decline.
Edible algal species were abundant. Zooplankton was a
minor food item of the fish in Lake 227 (fathead
minnow, pearl dace, redbelly dace, finescale dace).
RESPONSE OF LAKE 223 TO
ACIDIFICATION
Lake 223 is a small, oligotrophic lake with surface
area of 27.3 hectares and maximum depth 14.4 meters
Schindler, et al. 1980 and Schindler, 1980, describe
the lake, the acidification scheme, and chemical and
biological results from 1976 to 1979. Table 4 gives data
on the pH and DIG content of the epilimnion of Lake
223 for 2 pre-acidification years and for the years 1976
to 1979 when sulfuric acid was added. DIG was greatly
reduced in 1976, without lowering the pH. With the
buffering capacity depleted, pH declined in response to
acid addition in 1977 to 1979. Concentrations of P and
N were not affected by acidification (Schindler, 1980).
Phytoplankton biomass (Findlay and Saesura, 1980)
and production (Schindler, 1980) increased during the
first 4 years of acidification.
RESPONSE OF ZOOPLANKTON OF LAKE
223 TO ACIDIFICATION
The progressive acidification of Lake 223 from pH 6.8
in 1976 to an average of 5.60 during the ice-free
season of 1979 has changed the abundance of certain
zooplankton, particularly cladocerans and rotifers.
Year-to-year variation in abundance of zooplankton
species in the Experimental Lakes Area is not yet well-
documented. Therefore, until trends reported here can
be confirmed in 1980 or 1981, the conclusions are
somewhat tentative. Seasonal abundances of species
of crustaceans and rotifers in Lake 223 are described
by Malley et al. (In prep. b.).
D. galeata mendotae declined during 1977 to 1979;
on the other hand, D. brachyurum and H. gibberum
were more abundant with acidification. B. longirostris
remained relatively constant in numbers with acidifi-
cation (Table 5).
The dominant calanoid, Diaptomus minutus, showed
no change in numbers with acidification, whereas the
minor species D. sicilis, disappeared in 1978. Epis-
chura lacustris disappeared by 1979; E. nevadensis,
not recorded in 1974, appeared in low numbers in
1977 and 1978 during early acidification but was not
recorded in 1979. Overall, populations of calanoid
nauplii and copepodids were relatively constant with
acidification up to 1979 (Table 6). No effects of
acidification on cyclopoids are evident (Table 7).
Total number of rotifers was higher in 1977, 1978,
and 1979 than in the pre-acidification year, 1974. Most
marked changes were increases in numbers of
Polyarthra vulgaris, P. remata, Keratella taurocephala,
and Kellicottia longispina.
Although species composition changed during 1977
to 1979, overall there has been no significant change
in the biomass of crustacean zooplankton (Malley,
unpubl. data).
Dramatic effects of acidification were observed on
the population of the opposum shrimp, Mysisrelicta, in
Lake 223. This important fish food species in all but
very deep lakes is found near the bottom by day and is
planktonic at night, migrating vertically diurnally within
the hypolimnion (Beeton, 1960). Population size in
Lake 223 was estimated for the first time in summer
1978 as about 5.5 x 106 individuals. By summer 1979,
the population was reduced to 10 percent of the 1978
numbers or less. In early 1978 the population tolerated
a time when the maximum pH of their habitat was 5.9.
In early 1979, the maximum pH which they found
within theik environment was 5.6. Thus the pH limit for
survival of this species appears to be between 5.9 and
5.6 (Nero, unpubl. data)
Table 4. — Mean pH and DIG concentrations in the epilimnion
of Lake 223 in the ice-free seasons of 1974 to 1979.
Ice-free season
epiliminion Range of
mean pH DIG, epilimnion
Year
Pre-acidification
1974
1975
During acidification
1976
1977
1978
1979
(range)
6.64
(6.4-7.0)
6.61
(6.5-7.0)
6.79
(6.5-7.2)
6.08
(5.6-6.3)
5.84
(5.4-6.2)
5.6
(5.4-5.8)
/u moles f1
100-150
100-150
40-100
25-30
20-30
20-25
Table 5. — Number rrf2 of cladocerans in Lake 223 averaged
over the May to September period during 1974 and 1977 to
1979.
Species
1974 1977 1978 1979
Bosmina longirostris
Daphnia galeata mendotae
Holopedium gibberum
Diaphanosoma brachyurum
10,850 10,800 12,200 4,450
5,250 1,400 800 250
100 700 1,650 800
50 5,250 2,250 4,200
Table 6. — Number m~2 of calanoid copepods in L223 averaged over
the May to September period during 1974 and 1977 to 1979.
Species
Diaptomus minutus
adults
Diaptomus sicilis
adults
Epischura lacustris
adults
Epischura nevadensis
adults
Calanoid nauplii
Ni-N.
Calanoid copepodids
C,-C5
1974
4,320
650
400
0
54,050
71,950
1977
8,950
100
1,000
2,00
67,100
102,000
1978
800
0
750
550
46,100
85,650
1979
7,950
0
0
0
48,600
64,350
-------
112
RESTORATION OF LAKES AND INLAND WATERS
Table 7. — Number m 2 of cyclopoid copepods in L223 averaged over
the May to September period during 1974 and 1977 to 1979.
Species
1974
1977
1978 1979
Cylops bicuspidatus thorn. 3,650 11,500 7,900 8,700
adults
Jropocyclops prasinus mex. 1,350 5,050 6,550 1,250
adults
Mesocyclops edax 1,000 3,650 1,200 1,150
adults
Cyclopoid nauplii 51,850 102,000 155,750 58,350
NI-NB
Cyclopoid copepodids 44,450 128,700 61,700 58,200
Ci-C5
EFFECTS OF LOW pH ON BENTHIC AND
PLANKTONIC CRUSTACEANS
A consistent result of surveys of zooplankton
communities in lakes with a range of pH is that the
number of species of crustaceans declines below pH
5.5 or 5.0 (Sprules, 1975; Leivestad, et al. 1976; Roff
and Kwiatkowski, 1977). Daphnids are among the first
to disappear (Aimer, et al. 1974; Sprules, 1975).
Raddum, et al. (1980) report fewer species of
zooplankton in acidic lakes than in less acid lakes in
Norway, with cladocerans suffering greater reduction
than copepods and rotifers.
A number of benthic crustaceans, important as fish-
food organisms, are sensitive to pH below 6.0. The
amphipod Gammarus lacustris is absent from Nor-
wegian lakes with pH below 6.0. The branchiopod
Lepidurus arcticus is absent from these lakes below pH
6.1. In the laboratory, pH of 5.0 and below caused high
mortality in adult G. lacustris. L. arcticus was affected
at pH of 5.5 and below. Early life stages either did not
survive or were delayed in molting (Borgstrom and
Hendrey, 1 976). Okland (1980) reports that G. lacustris
tolerates pH down to 6.0 in colder mountain lakes but
only down to 6.6 in warmer lowland lakes. The isopod
Asellus aquaticus is rare below pH 5.6 in Norwegian
lakes.
The amphipod Gammarus pu/ex in laboratory studies
avoids water of below 6.2, or below 6.4 for young
(Costa, 1967).
The sensitivity of these benthic crustaceans, and
daphnids and Mysis, to the earlier stages of acidifica-
tion, pH 5.5 and above, leads us to expect that
acidification will have noticeable effects on fish species
ecologically through the food supply as well as by direct
physiological effects. The disappearance of these sig-
nificant fish-food crustaceans from acidifying fresh-
water systems is thus an important early biological
indicator of damage to the system.
The shifts in zooplankton species composition with
acidification from large daphnids to smaller Bosmina
and Diaptomus minutus leads Van and Strus(ln press)
to conclude tentatively that filtering rates are lower in
acidic than in non-acidic lakes. Raddum, et al. (1980)
note that filtering zooplankton were proportionately
more reduced in acidified lakes than were seizers.
Acidification may thus affect the efficiency of energy
transfer from primary to secondary trophic levels.
How low pH affects crustaceans is poorly known.
Low pH may interfere with ion uptake (Sutcliffe and
Carrick, 1 973), particularly Ca++ uptake during postmolt
(Borgstrom and Hendrey, 1976). Work on effects of low
pH on Ca++balance of postmolt crayfish Orconectes
virilis was initiated at ELA to provide hypotheses
concerning physiological effects of low pH on crus-
taceans which then could be tested on smaller
planktonic and benthic species. All crustaceans molt
periodically for growth and development and take up
Ca++ from the environment after molting for re-
calcification of the new, soft exoskeleton (Greenaway,
1974, for crayfish; Marshall, et al. 1964, and Porcella,
et al. 1967, for Daphnia magna). Postmolt O. virilis
were found to be more susceptible to mortality at pH
3.0 and 4.0 than were non-molting individuals (Malley,
1980). Ca -+ uptake during postmolt was progressively
inhibited by pH below 5.75 and completely inhibited
below pH 4.0 in crayfish in the laboratory in a known
volume of lake water.
Although O. virilis in Lake 223 in 1980 at mean pH of
5.3 maintained pre-acidification population size and
recruitment rate (Davis, unpubl. data), the exoskeletons
of individuals from Lake 223 were not as hard in early
fall 1980, as they were in control crayfish. Recalcifi-
cation following molting in Lake 223 apparently is
occurring at a rate slower than normal for these lakes.
Rate of Ca++uptake in postmolt 0. virilis depends upon
the concentration of HCOa in the experimental
medium, declining as the dissolved inorganic carbon
equilibrium shifts away from HCOa below pH 6.0
(Malley, 1980) and becoming greater asHCOs is added
to the medium of a crayfish or as the pH is
experimentally raised to 8.0 or 9.0. (Malley, unpubl.
data). Greenaway (1974) suggests that Ca is taken up
partly in exchange for H+ and partly accompanied by
HCO3 for electrical balance.
Crustaceans vary in their tolerance of low pH, some
disappearing when environmental pH reaches only 6.0,
others such as Diaptomus minutus and Bosmina
longirostris existing in lakes at pH 3.8 (Sprules, 1975)
or at 3.5 (Mesocyclops edax, Cyclops bicuspidatus
thorn., Bosmina longirostris, De Costa, 1 975). Thus the
pH range of 6.0 to 5.0 in which sensitive species of
crustaceans decline in numbers and which noticeably
affects the hardness of crayfish correlates with the
decrease inHCO-j, as well as with an increase in H+,
both of which play a role in Ca++uptake. It is interesting
to speculate that in the acid-resistant crustacean
species, ionic regulation, and particularly Ca++uptake,
depends upon exchanges with ions other than hTand
HCOa
CALANOIDA
I97I ' I972
YEAR
I973
Figure 2. —
-------
DREDGING AND BIOMANIPULATION AS RESTORATION TECHNIQUES
113
N 175.0-
E 157.0-
1*140.0-
$ 122.0-
^ 105.0-
I- 7O.O-
52 52.0-
5 35.0-
E 17.0-
CYCLOPOIDA
1969
1970
1971 1972
YEAR
1973
1974
Figure 3. —
ROTIFERA
1969
1970
1971 1972
YEAR
1973
1974
Figure 4. —
SUMMARY
Low alkalinity of these Experimental Lakes renders
them prone to pH change from moderate levels of
human activities such as nutrient addition or acid rain.
Fertilization of Lake 227 at levels which increased P
and N only five times above natural inputs, created
adverse and unstable conditions for survival of
zooplankton, including high pH and increased anoxia in
the lake. The zooplankton community has not been able
to recover pre-fertilization biomass or composition after
10 years of fertilization. Increased acidity to pH 5.6 has
altered zooplankton species composition to a small
extent but rate of loss of species is expected to be
higher as the pH falls below 5.0 (Sprules, 1975; Roff
and Kwiatowski, 1977). Until the acidification of Lake
223 progresses further, little can be said about the
relationship between a species' ability to tolerate acid
conditions and its ability to tolerate the high pH/lowOa
conditions. Daphnia was sensitive to a reduction in pH
to values below 6.0 and also declined in Lake 227 with
elevated pH but the species were different in the two
cases. Cyclopoids, tolerant of conditions in Lake 227
also were not affected by acidification in Lake 223.
REFERENCES
Aimer, B., et al. 1974. Effects of acidification on Swedish
lakes. Ambio 3:30.
Armstrong, F. A. J., and D. W. Schindler. 1971. Preliminary
chemical characterization of waters in the Experimental
Lakes Area, northwestern Ontario. Jour. Fish. Res. Board
Can. 28:171.
Beeton, A. M. 1960. The vertical migration of Mysis relicta in
Lakes Huron and Michigan. Jour. Fish. Res. Board Can.
17:517.
Borgstrom, R., and G. R. Hendrey. 1976. pH tolerance of the
first larval stages of Lepidurus arcticus (Pallas) and adult
Gammarus lacustris G. O. Sars. Intern. Rep. Norwegian
Inst. Water Res., Oslo, Norway.
Briand, F., and E. McCauley. 1978. Cybernetic mechanisms in
lake plankton systems: how to control undesirable algae.
Nature (Lond.) 273:228.
Brooks, J. L. 1969. Eutrophication and changes in the
composition of the zooplankton. Pages 236-255 in
Eutrophication: causes, consequences, correctives. Natl.
Acad. Sci., Washington, D.C.
Brunskill, G. J., et al. 1971. Chemistry of surface sediments
of sixteen lakes in the Experimental Lakes Area, north-
western Ontario. Jour. Fish. Res. Board Can. 28:277.
Chang, P. S. S., et al. 1980. Species composition and
seasonal abundance of zo^plankton in Lake 227, Experi-
mental Lakes Area, northwestern Ontario, 1969-1978. Can.
Data Rep. Fish Aquat. Sci. 182.
Chaston, I. 1969. Anaerobiosis in Cyclops varicans. Limnol.
Oceanogr. 14:298.
Costa, H. H. 1967. Responses of Gammarus pulex (L.) to
modified environment. II. Reactions to abnormal hydrogen
ion concentrations. Crustaceana 13:1.
Davis I. Unpubl. data. Freshw. Inst., Canada.
Davis, P., and G. W. Ozburn. 1969. The pH tolerance of
Daphnia pulex (Leydig, emend., Richard). Can. Jour. Zool.
47:1173.
DeCfista. J. 1975. The crustacean plankton of an acid
reservoir. Int. Ver. Theor. Angew. Limnol. Verh. 19:3221.
Findlay, D. L. 1978. Seasonal successions of phytoplankton
in seven lake basins in the Experimental Lakes Area,
northwestern Ontario, following artificial eutrophication.
Data from 1974 to 1976. Can. Fish. Mar. Serv. MS Rep.
1466.
Findlay, D. L., and H. R. Kling. 1975. Seasonal successions of
phytoplankton in seven lake basins in the Experimental
Lakes Area, northwestern Ontario following artificial
eutrophication. Can. Fish. Mar. Serv. Tech. Rep. 513.
Findlay, D. L., and G. Saesura. 1980. Effects on phyto-
plankton biomass, succession and composition in Lake 223
as a result of lowering of pH levels from 7.0 to 5.6: Data
from 1974 to 1979. Can. Mar. Rep. Fish. Aquat. Sci. (in
press).
Greenaway, P. 1974. Calcium balance at the postmoult stage
of the freshwater crayfish Austropotamobious pallipes
(Lereboullet). Jour Exp. Biol. 61:35.
Hillbricht-llkowska, A., and T. Weglenska. 1970. Some
relations between production and zooplankton structure of
two lakes of a varying trophy. Pol. Arch. Hydrobiol. 17: 233.
Johnson, W. E. and J. R. Vallentyne. 1971. Rationale,
background, and development of experimental lake studies
in northwestern Ontario. Jour Fish. Res. Board Can. 28:123.
King, D. L. 1970. The role of carbon in eutrophication. J.
Water Pollut. Control Fed. 42:2035-2051.
Kling, H. pers. comm. Freshw. Inst., Canada.
-------
114
RESTORATION OF LAKES AND INLAND WATERS
Lawrence, S. G., et al. Determination of dry weight biomass
of zooplankton species by estimation of volume. Can. Tech.
Rep. Fish. Aquat. Sci. In prep.
LeBrasseur, R.J., and 0. D. Kennedy. 1972. The fertilization
of Great Central Lake II. Zooplankton standing stock. U. S.
Natl. Mar. Fish. Serv. Fish. Bull. 70:25.
Leivestad, H., et al. 1976. Effects of acid precipitation of
freshwater organisms. In F. H. Braekke, (ed.) Impact of acid
precipitation on forest and freshwater ecosystems in
Norway. Rep. 6, SNSF-Project, NISK, 1432 Aas-NLH,
Norway.
Lowndes, A. G. 1952. Hydrogen-ion concentration and the
distribution of freshwater Entromostraca. Ann. Mag. Nat.
Ser. 12 5:58.
Ma'lley. D. F. 1980. Decreased survival and calcium uptake by
the crayfish Orconectes virilis in low pH. Can. Jour. Fish.
Aquat. Sci. 37:364.
Malley, D. F. unpubl. data. Freshwater Institute, Canada.
Malley, D. F., P. S. S. Chang and D. W. Schindler. Decline of
zooplankton populations following eutrophication of Lake
227, Experimental Lakes Area, Ontario; 1969-1974. Can.
Jour. Fish. Aquat. Sci. (a) In prep.
Malley, D. F., et al. Response of zooplankton to the
experimental acidification of Lake 223, Experimental Lakes
Area, northwestern Ontario: 1974-1979. Can. Jour. Fish.
Aquat. Sci. (b.) In prep.
Marshall, J. S., A. M. Beeton, and D. C. Chandler. 1964. Role
of zooplankton in the freshwater strontium cycle and
influence of dissolved salts. Int. Ver. Theor. Angew. Limnol.
Verh. 15:665.
Nero, R. unpubl. data. Freshwater Institute, Canada.
O'Brien, W. J., and F. DeNoyelles, Jr. 1972. Photo-
synthetically elevated pH as a factor in zooplankton
mortality in nutrient enriched ponds. Ecology 53:605,
Oakland, K. A. 1980. Some fish-food organisms in Norway
(Asellus aquaticus), Gammarus lacstris (Crustacea) and
smalLmussels(Sphaeriidae): Ecology and distribution. Page
133 //^Abstracts of voluntary contributions to Int. Conf.
Ecological Impact of Acid Precipitation. Sandefjord, Norway,
March 11-14, 1980. Volume II. SNSF-Project, Box61, 1432
Aas-NLH, Norway.
Porcella, D. B., C. E. Rixford, and J. V. slater. 1969. Molting
and calcification in Daphnia Magna. Physiol. Zool. 42:148.
Raddum, G. G., etal. 1980. Phytoplankton and zooplankton in
water with varying H* concentration. Pgs. 138-139. in
Abstracts of voluntary contributions to Int. Conf. Ecological
Impact of Acid Precipitation. Vol. II. Sandefjord, Norway,
March 11-14, 1980. SNSF-Project, Box 61, 1432 Aas-NLH,
Norway.
Roff, J. C. and R. E. Kwiatkowski. 1977. Zooplankton and
zoobenthos communities of selected northern Ontario lakes
of different acidities. Can. Jour. Zool. 55: 899.
Schindler, D. W. 1975. Whole-lake eutrophication experi-
ments with phosphorus, nitrogen and carbon. Int. Ver.
Theor. Agnew. Limnol. Verh. 19:3221.
1977. Evolution of phosphorus limitation in lakes.
Science. 195:260.
1980. Experimental acidification of a whole lake: a
test of the oligotrophication hypothesis. Pro. Int. Conf.
Ecological Impact of Acid Precipitation, Sandefjord, Norway
March 11-14, 1980. (In press).
Schindler, D. W., and E. J. Fee. 1973. Diurnal variation of
dissolved inorganic carbon and its use in estimating primary
production and CO2 invasion in Lake 227. Jour. Fish Res.
Board Can. 30: 1501.
1974. The Experimental Lakes Area: whole-lake
experiments in eutrophication. Jour. Fish Res. Board Can
31:937.
Schindler, D. W., et al. 1971. Eutrophication of Lake 227,
Experimental Lakes Area, northwestern Ontario by addition
of phosphate and nitrate. Jour. Fish Res. Board Can.
28:1763.
1972. Atmospheric carbon dioxide: its role in
maintaining phytoplankton standing crops. Science
177:1192.
1973. Eutrophication of Lake 227 by addition of
phosphate and nitrate: the second, third and fourth years of
enrichment, 1970, 1971 and 1972. Jour. Fish Res. Board
Can. 30:1415.
1980. Experimental acidification of Lake 223,
Experimental Lakes Area: background data and the first
three years of acidification. Can. Jour Fish. Aquat. Sci.
37:342.
Smith, M. W. 1969. Changes in environment and biota of a
natural lake after fertilization. Jour. Fish. Res. Board Can.
26:3103.
Sprules, W. G. 1975. Midsummer crustacean zooplankton
communities in acid-stressed lakes. Jour. Fish Res. Board
Can. 32:389.
Sutcliffe, D. W., andT. R. Carrick. 1973. Studies on mountain
streams in the English lake district. I. pH, calcium and the
distribution of invertebrates in the River Dudden. Freshw.
Biol. 3:437.
von Brand, T. 1944. Occurrence of anaerobiosis among
invertebrates. Biodynamica 4:185.
Wright, R. F., et al. 1976. Impact of acid preparation on
freshwater ecosystems in Norway. Water Air Soil Pollut.
6:483.
Yan.N. D., and R. Strus. 1980. Crustacean zooplankton
communities of acidic metal-contaminated lakes near
Sudbury, Ontario. Jour. Fish. Res. Board Can. (In press.).
ACKNOWLEDGEMENTS
P. Campbell, E. Fee, R. Hecky, R. Hesslein and D. Schindler
contributed useful comments to these papers.
-------
115
SEDIMENT TREATMENT FOR
PHOSPHORUS INACTIVATION
GUY BARROIN
Station d' Hydrobiologie Lacustre
Institut National de la Recherche Agronomique
Thonon les Bains, France
ABSTRACT
In situ sediment treatment has been studied to restore nutritionally polluted lakes. This method
can alleviate many of the economic and environmental obstacles associated with dredging,
bottom-sealing, or sediment consolidation via desiccation. A French experiment in progress since
1973 is based on improving the sediment sorptive capacity to adsorb phosphorus by injecting
aluminum sulfate in its top layer. A prototype device designed for this purpose is described. The
results indicate that the treatment significantly reduced the phosphorus in the lake even under
anoxic conditions and during at least a 4-year period. No adverse long-term effects were observed.
A Swedish experiment using nitrates, iron, and lime for sediment oxidation and phosphorus
inactivation is also mentioned.
INTRODUCTION
As reported by Dunst and coworkers (1974) it is
generally agreed that the most desirable approach to
lake restoration is "to restrict the quantities of
nutrients which reach the photic zone in a biologically
available form at a time when they can contribute to
the undesirable growth of aquatic plants." Curbing
excessive nutrient inputs from the watershed is the
most ecological and desirable long-term solution, but
under certain circumstances the fertilizing power of the
sediments is likely to delay or prevent lake restoration.
Different techniques are now available to reduce the
sediment contribution to lake fertility (Theis, 1979). A
new one is proposed which "defertilizes" sediments in
the same way that agricultural techniques fertilize
soils. Because phosphorus is a major eutrophicant and
the easiest to render inactive (Vallentyne, 1974), a
treatment was experimentally applied in summer 1973
to increase the phosphorus-binding capacity of the
sediments, even under anoxic conditions (Barroin,
1976). This report summarizes the field experiment and
its results.
10
iS TREATED X«EA : 1,800 m2
TOTAL AREA: 3,500 m2
MAXIMUM DEPTH: 5.5m
VOLUME: 9,500m3
Figure 1. — Bathymetric map and physical characteristics of
the experimental lake.
EXPERIMENTAL LAKE DESCRIPTION
Lake Morillon is a doline lake located at 460 meters
above sea level in the calcareous Chablais mountains
bordering the south shore of Lake Geneva. Its main
physical characteristics appear in Figure 1. Lateral
inputs are diffused from the watershed which is mainly
covered in gardens, lawns, and deciduous forest. Dead
leaves that fall on the surface provide the lake with its
mixotrophic characteristics such as yellow-brown
water and loose organic sediments. There is no
punctual outlet and the surface level is in equilibrium
with the water table. Water and sediment chemistry,
before treatment, is summarized in Tables 1 and 2.
Phytoplankton was dominated by Dinophyta. The
almost permanent presence of sulfides below a depth
of 2 meters restricted the planktonic and benthic fauna
to Chaoborus flavicans and the fish fauna to Carassius
auratus.
METHOD OF LAKE RESTORATION
Laboratory studies on sediments sampled at different
water depths indicated that only those below 2 meters
increased the fertility of epilimnetic water when mixed
with it. The treatment was thus restricted to 1,800 m2
and designed to affect the upper 15 centimeters of
sediments, considered to be a thick enough barrier to
-------
116
RESTORATION OF LAKES AND INLAND WATERS
Table 1 — Water chemistry (before treatment).
Surface
Parameters
Temperature
Conductivity
pH
02
s
NO3
NO2
NH4
PCvP
Tot-P
Mg
Ca
Na
K
Cl
SiO2
SO4
Transparency
Units
°C
fj S/cm (25°C)
mg Oz/l
mg S/l
mg N/l
mg N/l
mg N/l
/jg P/l
A3 18H2OJ was used
because of the sorptive capacity of its hydroxide under
anoxic and slightly acid conditions, its lack of acute
toxicity, and the low price of the commercial product.
Laboratory studies revealed that an alum injection on
the basis of 400 g*(m2)21 would limit phosphorus
release to an undetectable level. Figure 2 gives a
schematic representation of the treatment equipment
especially designed and constructed to inject alum into
the sediment, minimizing perturbation of the water
column stratification.
Epilimnetic water is pumped using a 12.5 HP pump,
continuously receiving at the strainer level a 400 g(m2)"1
alum stock solution. A strong firehose conducts
this diluted mixture to the ploughshare. This part of
the equipment is made of a V-shaped iron tube
fitted with regularly-spaced holes, the diameters of
which are calculated so that the ejection pressure
(c.a. 4kg(cm2)"') is the same for all. Pumps and
reagent tanks are placed on a pontoon towed from the
shore.
The treatment was applied during August 1973 in
three phases: (1) The sediment was first ploughed,
without alum, to degasify and prevent any subsequent
lifting of flocculated materials by entrapped gas
bubbles; then (2) 750 kilograms of alum were injected
in the sediments; and (3) finally, 200 kilograms poured
out on the whole surface for precipitation sediment
particles previously mobilized, thus performing an
epilimnetic inactivation.
Figure 2. — Schematic representation of the treatment
equipment.
RESULTS
Transparency
For a few days after treatment, Secchi disk readings
increased by up to 3 meters; later they decreased again
because of the rising and dispersion of a few sediment
floes. Final result: There has been a slight increase of
the mean value and a greater amplitude of the
variations.
WATER CHEMISTRY
POrP and Tot-P concentrations drastically de-
creased at the sediment interface (Figure 3) as well as
-------
'BlOMMUPtltATIOf* AS RESTORATION TECHNIQUES
117
pg. I
2QOQ
100O.
Figure 3. —• Concentrations of Tot-P, PO^Pand A1 in water at
sediment interface.
AMJJASONDJFMAONDJFMAMJJASO FMAMJJASONDJFMAMJJASOND
JJAS
-1972
-1973-
-197*
-1977-
I1OO-500
CDo-IO CH10-1OO IBlOO-SOO •• >500
Figure 4. — Distribution of POrP (/u.l~1) over time and depth.
0-1OO EI3100-1SO Bi 15P-50O BI500-1OOOH > 10OO
Figure 5. — Distribution of Tot-P Gu.f') over time and depth,
in the .whole lake (Figures 4,5) until 1977. Simultan-
eously, aluminum concentrations increased to detect-
able levels; likewise, sulfides' concentrations in bot-
tom layers rose to 110 mgSI"1 in 1974, falling to 57
57:mgsr1 in 1977. No other significant change was
noticed concerning water chemistry.
WATER BIOLOGY
During 1974, phytoplankton biovolume showed a 50
percent reduction accompanied with a specific shift
from Dinophyta and Chilorophyta to Cryptophyta and
Diatoms (Figure 6). During: 1977-1978 Chrysophyta
and Diatoms were dominant. No adverse effects were
observed concerning the originally scarce fish and
plankton fauna.
Chlorophyta
Gyanophy ta
Dinophyta
Chrysophyta
hryptophyta
Diatoms
A M J J A S O N D J F M A O N D J FMAMJJASO
1972 1973 1974
Figure 6. — Quantitative and qualitative evolution the
phytoplankton.
SEDIMENT CHEMISTRY
For a few days after treatment, many aluminum
hydroxide floes were observed at the sediment
interface probably resulting from the epilimnetic
inactivation. By studying cores sampled a few weeks
after treatment, no visual evidence of sediment mixing
could be detected and no significant change in the
distribution patterns of studied elements, among them
aluminum, could be measured.
MACROPHYTES
During the year after treatment, some nenuphars
showed irregular foliation but in 1975 no anomaly was
observed.
DISCUSSION
To properly evaluate the importance of treatment
efficiency and duration it is necessary first to evaluate
that of the part played by epilimnetic inactivation. In
fact, the introduction of 200 kilograms of alum in a
9,500 m3 volume, produces a final aluminum con-
centration of I.SJmgf1 , which represents about a
tenth of the amount indicated in the literature (Funk
and Gibbons, 1979). Therefore, it may be assumed that
the success in lowering phosphorus content of the
whole lake during several years are due chiefly to
sediment treatment. The phosphorus-binding capacity
of the treated sediment seems to show a long-term
saturation as indicated from phosphorus concentration
increases during 1978, perhaps because of constant
input including, for example, untreated littoral sedi-
ments or dead leaves.
Owing to the important buffer capacity of the water
and the sediment, no pH modification was observed
after alum had been injected, the reaction of which is
acid. The slight increase of the dissolved aluminum
concentration indicates that this element was totally in
-------
118
RESTORATION OF LAKES AND INLAND WATERS
a flocculated form. It is surprising that the analysis of
the cores did not show any significant increase.
Several explanations may be suggested: the reagent
was injected in a higher thickness than planned and
therefore was more diluted, or the sediment aluminum
content was high and variable enough to preclude any
observation of change or to make the sampling
inadequate.
The sulfates introduced in the sediments were
reduced to sulfides by the sulfato-reducing bacteria
and then migrated mostly to the hypolimnion increas-
ing its sulfide content. The slight oligotrophication
indicated by phytoplankton changes cannot be only
interpreted as resulting from the treatment because of
the possible interference of additional environmental
factors; but the nenuphar disease is due to rhizome
deteriorations and perhaps nutrient deficiency directly
connected with manipulation of the lake bottom.
FURTHER DEVELOPMENTS
The Morillon experiment having been relatively
successful, further research was done to construct
more elaborate equipment, as autonomous as possible.
An intermediary prototype will be tested during
summer 1980. Other research is being conducted to
investigate the efficiency of different chemicals for
increasing the phosphorus-binding capacity of sedi-
ments directly, or through oxidization using peroxides.
Meanwhile, in 1975, a Swedish experiment was
conducted on a larger scale, using iron for phosphorus
fixation and nitrates for sediment oxidization after pH
adjustment with lime (Ripl, 1 976; Bjork, 1 978; Bjork, et
al. 1978). The harrow especially designed for this
purpose lifted the sediment using compressed air,
chemicals being simultaneously distributed at the rear
of the device. After addition of nitrates, denitrification
processes took place, producing a vigorous release of
nitrogen bubbles. As Dr. Bjork said, "thanks to the
treatment Lake Lillesjon was converted to a lake with
normal ecosystem functions."
REFERENCES
Barroin, G. 1976. La regeneration des lacs; ne pourrait-on
pas "trailer" les sediments. In La mecanique des fluides et
I'environment: prevision et maitrise de la qualite de I'eau et
de I'air. Compte-reridu des 14emes journees de I'hydraul-
ique. Soc. Hydrotechnique de France, tome 1, question 3,
rapp. 11. Paris Sept. 7-9.
Bjork, S. 1978. Restoration of degraded lake ecosystems.
Institut of Limnology, University of Lund.
Bjork, S., et al. 1978. Lake management: studies and results
at the Institute of Limnology in Lund. University of Lund.
Dunst, R. C., et al. 1974. Survey of lake rehabilitation:
techniques and experiences. Wis. Dep. Nat. Resour,
Madison. Tech. Bull. 75.
Funk, W. H., and H. L. Gibbons. 1979. Lake restoration by
nutrient inactivation. Pages 141-151 in Lake restoration:
Proc. Nat. Conf. August 22-24, 1978, Minneapolis, Minn.
Off. Water Plan. Stand. U.S. Environ. Prot. Agency,
Washington, D.C.
Ripl, W. 1976. Biochemical oxidation of polluted lake
sediment with nitrate: a new lake restoration method.
Ambio 5:132.
Theis, T. L. 1979. Physical and chemical treatment of lake
sediment. Pages 115-120 in Lake restoration: Proc. Nat.
Conf. August 22-24, 1978, Minneapolis, Minn. Off. Water
Plan. Stand. U.S. Environ. Prot. Agency, Washington, D.C.
Vallentyne, J. R. 1974. The algal bowl: lakes and man. Dep.
Environ. Fish. Mar. Serv. Misc. Spec. Publ. 22. Ottawa, Can.
ACKNOWLEDGEMENTS
Thanks are due to the Ministere de la Culture et de
I'Environment which provided financial support, and to M.
Colon for his technical collaboration.
CONCLUSION
Sediment treatment methods open up an important
field for applications which make it possible to control
sediment conditions and therefore the state of the
entire lake. This control concerns not only phosphorus
fixation with or without oxidization but also every
phenomenon occurring in the sediments. The same
Swedish researchers, in collaboration with the private
firm Atlas-Copco, are now developing the "Contracid
method" to counteract lake acidification (Bjork, 1978).
The harrow used for injecting an alkaline sodium
solution has a capacity of about 10 times that of the
prototype used for sediment manipulation in Lake
Lillesjon.
It is easy to imagine using such techniques for
solving problems of oil pollution by injecting cultures of
"trained" bacteria, or of macrophyte proliferation by
injecting selected biocides. But sediment treatment is
only at its neolithic stage and requires not only more
technological research but also more limnological
knowledge.
-------
119
TWO EXAMPLES OF URBAN STORMWATER
IMPOUNDMENT FOR AESTHETICS AND FOR
PROTECTION OF RECEIVING WATERS
THOMAS BRYDGES
GLENN ROBINSON
Water Resources Branch
Ontario Ministry of the Environment
Rexdale, Ontario
ABSTRACT
Stormwater impoundments in urban areas can improve the quality of runoff prior to discharge to
the receiving waters, and at the same time, aesthetically improve the urban environment. Two
manmade lakes of contrasting design that are fed only by Stormwater runoff are examined. Lake
Aquitaine has a small drainage basin and a sophisticated sediment removal system; Lake
Wabukayne has a large catchment area and only a gabion wall for sediment removal. The long-
term average retention time of the Lake Aquitaine sedimentation basin, 17 days, is similar to the
retention time of the whole of Lake Wabukayne, 13 days. Suspended solids are reduced by 69 to
90 percent across Lake Aquitaine, while the solids reduction across Lake Wabukayne (29 to 33
percent) is similar to the reduction across the sedimentation basin in Lake Aquitaine (4 to 25
percent). Cladophora has become the dominant alga in Lake Aquitaine while in Lake Wabukayne
aquatic macrophytes, Cladophora, phytoplankton and floating algae are prevalent. Both lakes have
reduced dissolved oxygen levels at all depths following heavy rainfalls. Lake Aquitaine wasanoxic
at 1 m above the bottom in August 1980. The retention time of the lakes and sedimentation basin
appears to be the main factor controlling the reduction of suspended solids. Eutrophication
problems may require further control measures to maintain the aesthetic value of the lakes.
INTRODUCTION
The Province of Ontario has an estimated 500,000
lakes providing a huge potential for water based
recreation. The high quality lakes on the Precambrian
Shield support a large tourist industry and have been
the recreational playground for many residents living in
urban areas such as Toronto and Hamilton. The
recreational value and water quality of the smaller
number of lakes closer to the urban areas in Southern
Ontario have not been studied as much. However, in
recent years an increased awareness and concern for
the quality of the urban environment has produced-
demand for high quality lakes close to and within the
urban areas. This demand is reflected in two ways: A
need to protect and upgrade water quality in existing
urban waterways and a desire to create new lakes
within city limits as aesthetic improvements to the
urban environment.
Urbanization itself causes deterioration of receiving
water quality by increasing the total amount and peak
loading of runoff as well as adding a wide range of
contaminants to the runoff (Weibel, 1969).
Using impoundments to improve the quality of the
runoff before it enters the receiving waters is an
attractive concept; it is logical to design such
impoundments to meet some aesthetic demands as
well. This type of lake is inexpensive to maintain
compared to an equivalent amount of parkland (Proj.
Plan. Associates, 1976) and does not constitute a
safety hazard provided it is well designed and properly
maintained, with certain use restrictions.
This paper discusses two impoundment lakes of
contrasting design.
Study Lakes
Lake Aquitaine is located in the Meadowvale "new
town" in Mississauga about 20 kilometers west of
Toronto. Construction began in the 1970's and this
agricultural land is still being developed for housing,
shopping, and light industry. The town covers 1,200
hectares and is expected to have a population of
65,000 when complete in 1985. Its proximity to the
Credit River and the impoundment lake have been
strong selling features used by the developers.
Lake Aquitaine was constructed by excavating a farm
field to a depth of 5 meters with a bank slope of 4:1 at
the shore. A concrete sedimentation basin, with energy
dissipation system, surface skimming weir, and
perforated spillway, were constructed at the inlet. A
"morning glory' spillway including bottom draw
facilities, controls water levels in the lake and an
emergency spillway and drainage channel have been
provided as a precaution in the event of a hurricane-
like storm. An aerial view is shown in Figure 1 and lake
characteristics are shown in Figure 3 and Table 1.
Discharge is to the Credit River system. The main inlet
is a single 2 m * 3 m storm drain. There is a
supplementary water supply from a 10 centimeter (4-
inch) municipal water main to maintain water level in
-------
120
RESTORATION OF LAKES AND INLAND WATERS
Figure 1. — Aerial photograph of Lake Aquitaine (eastward
view).
Figure 2. — Aerial photograph of Lake Wabukayne (westward
view).
Emergency Horning Glory
SptlliMr Spt Duty
-9
Figure 3. — Diagrammatic representation of longitudinal
section through Lake Aquitaine (not to scale).
Figure 4. — Diagrammatic representation of longitudinal
section through Lake Wabukayne (not to scale).
the summer if necessary. Since the lake was entirely
manmade, it was designed to maximize both water
quality improvement and aesthetic value. Model
predictions were made on the sedimentation basin and
lake characteristics (Murrey and Ganczarczk, 1977).
Construction was completed during the winter of
1976-1977 and the maximum water level was reached
in mid-August 1977. At the same time sodding and tree
planting were completed on the surrounding parkland.
The area immediately around the park is currently
under development.
Regulations prohibit activities requiring primary
contact (e.g. swimming, bathing, wind-surfing etc.(and
power boats are not permitted. However, canoeing,
sailing, and paddleboating are encouraged.
Lake Aquitaine was stocked with 3,300 rainbow
trout Salmo gairdneri(10 to 30 cm) in September 1977
and again in September 1979 with =500 rainbow
trout. Huge schools of small minnows have been
observed near the marina, and a large population of
pumpkinseeds Lepomis gibbosus has become estab-
lished in the lake. Fishing is permissible subject to
Ministry of Natural Resources fishing regulations for
Peel Region. In September 1979 a fishing derby was
organized at Lake Aquitaine.
Lake Wabukayne is located in the Erin Mills
development of Mississauga about 1 km from Lake
Aquitaine. Prior to development, a farm pond had been
built in the steep-sided valley of Wabukayne Creek. The
topography greatly limited possible design changes for
the lake. The farm pond was drained and lined with
clay. An aerial view is shown in Figure 2 and the lake
characteristics are shown in Figure 4 and Table 1.
Table 1 — Characteristics of the impoundment lakes.
Lake area (ha)
Lake volume (m3)
Mean depth (m)
Maximum depth (m)
Sedimentation basin area (ha)
Long term average yearly flow (m3)1
Retention time of the lake (days)
Retention time of the sedimentation
basin (days)
Drainage basin area (ha)
developed (ha)
undeveloped (ha)
Aquitaine
4.7
1.8x 105
3.8
50
0.38
2.0 x 105
3290
170
48.0
59.0-'
Wabukayne
2.0
3.2 x 10"
1.6
29
0.27
9.3 x 10s
13.0
0.55
466.0
263.0
203. 03
Average (or 10 years period (or fully developed watershed (3)
'34 ha being developed in 1980
'184 ha being developed in 1980
A large concrete dam and spillway replaced tne
original earth dam and the lake was re-filled by the fall
of 1976. The lake receives inputs from several storm
sewers and Wabukayne Creek; all are channeled
through twin 30m diameter storm sewers and a single
1.6 m sewer. A 6.3 I/sec supplementary supply of
water is pumped to the lake from groundwater sources
to help maintain lake levels. Discharge is to the Credit
River. The sedimentation basin in Lake Wabukayne
was separated from the main lake by a submerged
gabion weir near the inlet end of the lake and an access
ramp for sediment removal was constructed. In 1980,
another row of gabions was added to bring the weir
above the waterline. A 3 m gap was left at each end of
-------
DREDGING AND BIOMANIPULATION AS RESTORATION TECHNIQUES
121
the weir to discourage children from playing on it. A
small park with some naturally wooded areas borders
the shoreline; most of the land is sodded, but erosion
does occur in the steep-banked areas (slope 2:1). The
area adjacent to the park is completely developed, but
the remainder of the watershed is a mixture of
developed, developing, and rural land. No primary
contact recreation or boating is permitted. Fish stocking
has not been undertaken but natural populations of
sticklebacks (Gasterosteidae) have been observed.
Tadpoles, frogs, crayfish, and leeches also seem to
thrive.
Methods
Samples are collected weekly from May to Septem-
ber and less frequently in October.
Sampling at mid-lake stations included Secchi disk
depths, temperature profiles at the deepest point in
each lake, and dissolved oxygen at two depths.
Chemical samples were collected as composites
through the euphotic zone or to 1 m above bottom
using a plastic hose or a weighted sampling can. A
separate sample was collected at 1 m above bottom if
the euphotic zone did not penetrate to this depth.
Water samples were collected at the inlets and
outlets of the lake and sedimentation basin of
Aquitaine whenever flow was sufficient and at the inlet
and outlet of Wabukayne. It was not possible to sample
at the outlet of the Wabukayne sedimentation basin
since it was defined only by a submerged barrier.
Fish samples for heavy metal analysis were collected
from both lakes by the Ontario Ministry of Natural
Resources in September 1979.
All analyses were performed at the Ontario Ministry
of the Environment Laboratory according to their
standard methods (Outlines of Analytical Methods,
1975).
RESULTS AND DISCUSSION
The watersheds are not yet fully developed and the
construction activity will continue to influence the
quality of the runoff so the data represent a transition
stage for the lakes. However, a number of observations
can be made regarding the ability of the lakes to
achieve their two main objectives: Aesthetics and
receiving water protection.
The main contrasting feature of the lakes is the
retention times. The long-term average retention time
of the Aquitaine sedimentation basin, 17 days, is
similar to the retention time of the entire Lake
Wabukayne, 13 days (Table 2).
Mineral Chemistry
The major ion content of both lakes is shown in Table
2 for early June 1979.
With the exception of sodium and chloride, the major
ions are similar, reflecting similar soil conditions.
These values are typical of the sampling period for all 3
years.
Sodium chloride from road de-icing dominates the
ion content of Lake Aquitaine. Thousands of tons of salt
are used each year by the city of Mississauga and these
Table 2. — Major ion content of Lakes Aquitaine and
Wabukayne on June 14, 1979. Units are in meq/l except
conductivity (in /umhos/cm at 25°C) and pH.
Ca+
Mg*
Na*
K*
cr
SO/
Alkalinity (HCOs
Cond.
PH
Aquitaine
2.1
0.67
5.4
0.11
6.2
0.89
1.7
1000.0
8.24
Wabukayne
2.1
1.0
1.6
0.17
1.8
1.0
2.3
530.0
8.06
watersheds no doubt receive a share along with
contributions from householders treating driveways
and sidewalks. It is not clear why salt concentrations in
Lake Wabukayne are much lower since the same
percentage of the watershed is developed.
The chloride concentrations and conductivity de-
crease during the summer in both lakes although salt
content has generally increased, particularly in
Aquitaine, Figure 5.
Suspended Solids and Turbidity
One of the prime functions of the impoundments is to
reduce the solids loading to the receiving water. Table
3 shows the performance of both lakes in this respect.
In Lake Aquitaine, very good reduction of solids is
occurring across the whole lake; the sedimentation
basin itself is not so effective. However, about 20
centimeters of black sludge has accumulated in the
basin since it was built.
Lake Wabukayne is achieving an overall reduction in
solids similar to the Aquitaine sedimentation basin;
this is not surprising since they have similar average
retention times.
The model projected a fivefold increase in suspended
solids discharge from the Aquitaine watershed follow-
ing development (Murrey and Ganczarczyk, 1977). It
further projected that the lake and sedimentation basin
combined would give a 93 percent reduction. While it is
not possible to draw a comparison between the
projected loadings and the observed concentrations, it
would appear that Lake Aquitaine is very close to
meeting its suspended solids objective since the
reductions in solids concentrations ranged from 69 to
90 percent.
Results for August 24 and 31 were left out of the
1978 data set averaged for Table 3. A prolonged dry
spell followed by some rain produced high suspended
solids of 261 and 1,367 mg/l, respectively, in the low
inlet flows. These values are one to two orders of
magnitude greater than recorded for any other
sampling date. Including them in the yearly average
distorts the impression of the effectiveness of solids
removal.
Turbidity data generally paralleled the suspended
solids results. The data for 1979 are shown in Figure 6
along with the total rainfall since the previous sampling
date. Lake Wabukayne is always more turbid than
Aquitaine with a greater response to rainfall. The
-------
122
RESTORATION OF LAKES AND INLAND WATERS
Lake Aquitaine
Suspended
Solids
Nitrogen
Total
phosphorus
Lake Wabukayne
Suspended
Solids
Nitrogen
Total
Phosphorus
* only 3 data sets
•* 2 data sets left out.
Table
Year
1977
1978
1979
1977
1978
1979
1977
1978
1979
1977
1978"
1979
1977
1978**
1979
1977
1978"
1979
see text
3. — Effects
Inlet
57.0
11.0
10.0
2.41
2.37
1.79
0.20
0.17
0.10
16.0
13.0
15.0
2.23
1.42
1.90
0.060
0.050
0.89
M
of the impoundments on water quality (in mg/l).
Overflow from
sedimentation
basin
48.0
8.0
9.8
2.31
2.33
1.98
0.20
0.16
0.10
-
Percent Lake
reduction
15.0 10.0
25.0 3.8
4.0 2.0
1.84
0.93
1.01
0.039
0.036
0.023
14.0
13.0
11.0
2.21
1.12
1.53
0.071
0.037
0.068
Outlet
5.8*
3.4
1.5
1.84
0.98
1.05
0.045*
0.032
0.022
12
8.5
11
2.24
1.21
1.43
0.070
0.044
0.065
% reduction
from the
inlet
90
69
85
25
58
39
78
81
79
29
33
29
0
15
25
12
27
results reflect the retention time of the lakes. On a
number of sampling dates, the turbidity at the
Wabukayne outlet was nearly the same as at the inlet
from the storm sewers while turbidity at the Aquitaine
outlet was consistently lower than at the inlet. For
example, on June 14, 1 979 the inlet and outlet values
for Aquitaine were 16 and 2 Formazin Turbidity Units
(F.T.U.), respectively, and for Wabukayne they were 84
and 86 F.T.U. respectively.
Lake Wabukayne would have to be increased in
volume by 25 times to give a retention time equal to
Aquitaine; this may be necessary to achieve an
effective suspended solids control. However, such a
large increase in size would not be practical in this
particular case.
Nutrients and Eutrophication
Although nutrient levels in both lakes have been high
enough to produce nuisance growths of algae, Lake
Aquitaine has never had a serious phytoplankton
bloom, and Lake Wabukayne only began to exhibit high
chlorophyll a concentrations in 1979 (Tables 3 and 4).
Chlorophyll concentrations were not reported for 1977
because of an analytical problem.
In Lake Aquitaine the macrophyte, Potamogeton
foliosus, was first noticed in August 1978. At the same
time small colonies of attached algae (Oedogonium and
Cladophora) began to develop. By 1979 Cladophora
occupied almost all available substrate; only the finer
gravel remained devoid of the alga. A band of
Cladophora currently extends around the entire
shoreline of the lake. Raking has become necessary to
remove the alga to prevent odors from decomposition.
Other algal types identified near the inlet end include
Closterium, Spirogyra, Synedra, and Rhizoclonium.
Attached algae have flourished, and by successfully
competing for available nutrients, may have caused a
decrease in free-floating algae in 1979, as measured
by chlorophyll a.
In Lake Wabukayne a wide variety of plant life was
established as early as June 1977. Polygonum sp.,
Typha sp., and P. zosteriformis began to develop in an
area of the north shore just below the sedimentation
basin. By July 1 977 Alisma plantago-aquatica occupied
most of the shoreline including the sedimentation
basin. The following summer dense mats of floating
algae (Spirogyra and Oscillatoria) began to develop over
the macrophyte beds. Oil and floating debris tended to
collect in these mats, further detracting from the
appearance of the lake. Cladophora growths were
observed on the gabions and on all concrete surfaces
near the outlet. Conditions continued to deteriorate
when a relatively wet spring and early summer in 1 979
resulted in highly turbid conditions and poor water
clarity. Macrophyte growth was thus restricted, but
phytoplankton levels began to increase, producing
chlorophyll a levels as high as 50 /jg/l (July 10). By late
July drier weather prevailed and water clarity improved
as turbidity and suspended solids levels decreased. At
this time many large Daphnia sp. were observed which
could have contributed to the decrease in Chlorophyll a
-------
DREDGING AND BIOMANIPULATION AS RESTORATION TECHNIQUES
123
that occurred (2.2 jug/l,. August 8). As in 1978, the
improved light conditions promoted the growth of algal
mats over the macrophyte beds. Cladophora continued
to grow where suitable substrate was available but
lacked the "healthy" appearance of the Cladophora in
Lake Aquitaine.
The combination of solids removal and plant growth
is substantially reducing phosphorus in Lake Aquitaine
but having a minimal effect in Lake Wabukayne, Table
3.
The steadily deteriorating conditions in Lake Wa-
bukayne prompted a public meeting in the fall of 1979
which resulted in the clearing of the sedimentation
basin and the modifications to the gabion wall. A
regular program of surveillance and debris removal
was also initiated.
Dissolved Oxygen
Dissolved oxygen levels in the surface waters of both
lakes have generally been adequate although there
have been brief periods of reduced oxygen conditions in
the bottom waters of both lakes. In Lake Aquitaine the
bottom water dissolved oxygen levels have progres-
sively deteriorated from 1977 through 1979. By August
of 1980 anoxic conditions were measured at 1 m above
bottom and hydrogen sulfide was present.
Both lakes suffer from short-lived reduced oxygen
conditions at all depths following heavy rainfall. The
effect is most pronounced in Lake Wabukayne where
higher sediment loads produce higher oxygen de-
mands. Lowest observed oxygen concentrations in the
surface waters have been 5.4 mg/l and 3.7 mg/l in
Aquitaine and Wabukayne, respectively.
Fishery
Nine samples or rainbow trout from Lake Aquitaine
and seven samples of sticklebacks from Lake Wa-
bukayne were analyzed for PCB's, Mirex, pesticides,
mercury, copper, nickel, zinc, lead, cadmium, chromi-
um, arsenic, selenium, and iron. The rainbow trout
were in the 30 to 46 cm (12 to 18 in.) size range and in
all cases were acceptable for unrestricted consump-
tion. The Lake Wabukayne sticklebacks were all less
than 15 cm (6 in.) and, although unlikely to be
consumed by humans, were similarly low in con-
taminants.
Angling for rainbow trout has been a popular event in
Lake Aquitaine since the first stocking of fish. The
fishing derby in 1979 was a great success with
numerous fishermen taking part, again emphasizing
the recreational potential and aesthetic value of the
lake.
Conclusions
Properly designed stormwater impoundments can
effectively protect the quality of receiving waters and
provide aesthetic value to the urban environment.
Good control of suspended solids has been achieved
at an average retention time of 329 days while a 13-
day average retention time gives a very limited control.
Problems related to eutrophication of the jmpound-
ments seem to be increasing with time and may require
control measures in the future; otherwise, the
aesthetic value of the lakes may be reduced.
J J A S
Figure 5. — Summary of monthly mean chloride and
conductivity levels for 3 years (1977-79). Solid line represents
Lake Aquitaine; broken line represents Lake Wabukayne.
•50 Rainfall
August Sept.
Figure 6. — Summary of turbidity and precipitation levels from
May to September, 1979. Black triangles indicate the total
amount of rainfall in the previous week.
Table 4. — Chlorophyll a concentrations in Lakes Aquitaine
and Wabukayne in 1978 and 1979 in iig/\.
Lake Aquitaine
Lake Wabukayne
1978
1979
1978
1979
Mean
6.0
3.4
6.8
23.0
Range
1.0-16.6
1.0- 8.2
0.7-34.4
1.8-50.0
REFERENCES
Murrey, P. H., and J. J. Ganczarczyk. 1977. Storage for storm
water quality control — Meadowvale Test Site Study.
Environ. Can. Res. Rep. 63. Proj. 75-8-36.
Outlines of Analytical Methods. 1975. Ontario Minist.
Environ. Lab. Serv. Branch, Rexdale, Ontario.
Project Planning Associates Limited. 1976. Report on
maintenance and Hability aspects Meadowvale West Lake,
City of Mississauga. M5R 3K1. Toronto, Ontario.
Weibel, S. E. 1969. Urban drainage as a factor in
eutrophication. Pages 383-403 in Eutrophication: Causes,.
consequences, correctives. Natl. Acad. Sci. Washington
D.C.
-------
124
REVIEW OF AERATION/CIRCULATION FOR
LAKE MANAGEMENT
ROBERT A. PASTOROK
THOMAS C. GINN
MARC W. LORENZEN
Tetra Tech
Bellevue, Washington
ABSTRACT
Artificial circulation is a management technique for oxygenating eutrophic lakes subject to water
quality problems, algal blooms, and fish kills. Whole lake mixing may reduce regeneration of
nutrients from profundal sediments, while often controlling blue-green algal blooms. Models
predict that overall algal biomass will decrease in deeper lakes when light limitation is induced by
mixing. If destratification elevates epilimnetic C02 levels and causes a sufficient drop in pH,
dominance in the algal community will likely shift from a nuisance blue-green species to a mixed
assemblage of green algae. This more edible resource combined with an expansion of habitat
leads to more abundant zooplankton and with provisioning of a hypolimnetic refuge, invasion of
large-bodied daphnids. Habitat expansion and shifts in community structure of benthic
macroinvertebrates potentially elevates the abundance of fish food organisms. Although short-
term increases in fish growth and yield have been attributed to improvements of food and habitat
resources, documentation of long-term changes is lacking. In southern regions, artificial
circulation provides benefits for warmwater fishes only.
INTRODUCTION
Artificial aeration or circulation of lakes is commonly
used for managing the ecological consequences of
eutrophication. By inducing dramatic changes in
species abundance and distribution, diversity, and
trophic structure, the technique has potential useful-
ness in controlling algal blooms and improving
fisheries. This paper examines artificial circulation
techniques; i.e., those that mix the whole lake and
provide aeration without attempting to preserve the
normal thermal structure. Hypolimnetic aeration,
which maintains aerobic conditions without disrupting
thermal stratification, is covered elsewhere (Fast and
Lorenzen 1976; Pastorok, et al. in press).
EFFECTS OF ARTIFICIAL CIRCULATION
ON WATER QUALITY
Chemical Parameters
In most cases, artificial destratification increases the
concentration of dissolved oxygen in bottom waters
immediately (e.g., Hooper, et al. 1953; Lackey, 1972;
Haynes, 1973). Dissolved oxygen in the former
epilimnion may show a corresponding decrease due to
reduced photosynthesis (Haynes, 1973) or mixing of
hypolimnetic waters with low dissolved oxygen and
high BOD into the surface layer (Ridley, et al. 1966;
Thomas, 1966). Over a period of several weeks, the
oxygen content of the whole lake increases (Pastorok,
et al. in press). Under some circumstances, oxygen
depletion cannot be prevented by normal levels of
artificial aeration and massive fish kills result (R. S.
Kerr Res. Center, 1970; McNall, 1971).
Oxygen levels influence redox reactions involving Fe,
Mn, and Al; in turn, these elements and their
complexes partly determine the availability of nitrogen
and phosphorus compounds through release processes
occurring at the surface of profundal sediments
(Mortimer, 1941, 1942; Holdren and Armstrong, 1980).
As hypolimnetic waters are brought to the lake's
surface, excess gases such asCO2, H2S, and NH3 are
released to the atmosphere (R. S. Kerr Res. Center,
1970; Toetz, et al. 1972; Haynes, 1973). Along with
oxygen and other chemical species, these gases
become isochemical with depth (Toetz, et al. 1972).
Transparency
Artificial circulation has varied effects on water
transparency, depending on the intensity of mixing and
the contribution of phytoplankton to turbidity levels
before treatment. When mixing is induced during a
surface bloom of blue-green algae, transparency will
increase immediately due to distribution of the algae
throughout a greater water volume (Haynes, 1973).
Thereafter, water clarity may be enhanced by de-
struction of the bloom through light limitation in deep
lakes (Lorenzen and Mitchell, 1975) or through a
change in some other environmental factor in shallow
lakes (Malueg, et al. 1971).
A decrease in transparency after mixing generally
correlates with a rise in total seston (Carton, 1978;
Carton, et al. 1978), which may be caused by surface
algal blooms (Hooper, et al. 1953; Drury, et al. 1975) or
resuspension of sediments (Fast, 1971 a). Most de-
stratification devices have been undersized with
respect to the scaling rule suggested by Lorenzen and
-------
AERATION/MIXING AND AQUATIC PLANT HARVESTING
125
Fast (1977, i.e., 9.2 mVmin per 106 m2 lake surface.
When more thermal energy is absorbed at the lake's
surface than the circulation device can distribute, then
microthermal stratification of 2 to 3°C provides algal
populations a surface refuge with high light levels (e.g.,
Fast, 1973a; Drury, et al. 1975).
EFFECTS OF ARTIFICIAL CIRCULATION
ON PHYTOPLANKTON
In 40 cases of complete destratification, only 65
percent (=26 experiments) led to any significant change
in algal concentrations; of these, about 30 percent
resulted in more algae. Table 1 summarizes the
responses of phytoplankton to artificial circulation for
each lake. When more than one experiment was
conducted in a lake, the predominant response is given
unless the data are too variable to indicate an overall
trend; then, the responses for individual experiments
are given. Where mixing was complete, aeration
decreased algal density or biomass in 13 of 23 lakes. In
three lakes, the amount of phytoplankton remained
about the same, and in seven lakes it increased or the
overall response was unclear. Where mixing was
incomplete, algal density generally stayed the same or
Table 1. — Responses of phytoplankton to artificial circulation8
Lake
Complete Mixing
Cline's Pond
Parvin Lake
Section 4 Lake
Boltz Lake
University Lake
Kezar Lake
King George VI
Indian Brook"
Prompton Lake"
Cox Hollow''
Stewart Lake
U.K. Reservoir"
Reference
Malueg, et al. 1971
Lackey, 1973a
Fast, 1971 a
Fast, et al. 1973
Symons, et al. 1967, 1970
Robinson, et al. 1969
Weiss and Breedlove, 1973
Haynes, 1973
N H WS PC C 1971
Lorenzen and Mitchell, 1975
Ridley, et al. 1966
Riddick, 1957
McCullough, 1974
Wirth and Dunst, 1967
Wirth et al. 1970
Barnes and Griswold, 1975
Ridley, 1970
Algal Mean
Algal Standing Chlorophyll- a Green Blue-green Ratio
Densityc Biomass" Concentration Algae Algae Gr:BI-gr
0 +
0" 0
-f
+
0 + +
+
0 +
+
Wahnbach Reservoir Bernhardt, 1967
Queen Elizabeth II
Lake Roberts
Falmouth Lake
Test Res. II
Buchanan Lake
Ham's Lake1
Test Res. 1
Mirror Lake
4 Lakes"
Starodworskie Lake'
Incomplete Mixing
Casitas Res."
Hyrum Res.
West Lost Lake
Pfaffikersee
Waco Res."
Lake Maarsseveen"
Lake Catharine
El Capitan1
Arbuckle Lake
Lake Calhoun
McNall, 1971
R.S. Kerr Res. Cen., 1970
Symons, et al. 1967, 1970
Robinson, el al. 1969
Knoppert, et al. 1970
Brown, et al. 1971
Steichen, et al. 1974
Toetz, 1977a, b
Garton, 1978
Knoppert, et al. 1970
Smith, et al. 1975
Irwin, et al. 1966
Lossow, et al. 1975
Barnett, 1975
Drury, et al. 1975
Hooper, et al. 1953
Thomas, 1966
Biederman and Fulton, 1971
Knoppert, et al. 1970
Kothandaraman, et al. 1979
Fast, 1973a
Toetz, 19773, 1979
Shapiro and Pfannkuch, 1973
+
+
+
+ +
+ +
0
0+ 0+
0s 0"
0
+ +
+ +
+
0
0
0
+?
0
+
+
+
+ +
0+0
+ + +
0000
0+ 0-
og
+'
+
+ +
+
+
0
0
0
0
0 0
+ 0 +
" + = decrease, 0 = no significant change
" qualitative information only
c cells or colonies per liter; weighted mean for water column unless noted
" weight per square meter of lake surface
* increase observed, but control year was unusual
1 samples were taken near lake surface
8 increase observed, but it was correlated with large input of allochthonous nutrients
h Stewart Hollow Lake, Caldwell Lake, Pine Lake, Vesuvius Lake
-------
126
RESTORATION OF LAKES AND INLAND WATERS
increased following treatment (Table 1). Although
artificial circulation usually has a negative influence on
blue-green algae, its effect on green algae is
ambiguous.
Physical Mechanisms
In lakes where algal production is potentially limited
by light, several models predict a decrease in net
photosynthesis and a reduction in standing crop of
algae as depth of the mixed layer increases (e.g.,
Lorenzen and Mitchell, 1975; Oskam, 1978). If algae
are limited by nutrients before circulation, however, a
slight increase in mixing depth could cause an
••elevation of standing crop (e.g., point A to point B in
Figure 1). If mixing shifts the controlling mechanism
from nutrient limitation to light limitation, a moderate
increase in mixed depth can cause a substantial rise of
peak algal biomass or at best only a slight decline (A to
C or B to C, respectively, in Figure 1). With large
increases in mixed depth, the imposition of light
limitation might cause substantial decreases in water
column algal biomass (B to D in Figure 1). When algal
biomass decreases with increased mixed depth the
concentration of algae will decrease dramatically
because less biomass is distributed in a much larger
water volume. Finally, because of differences in growth
parameters among algal species, a major shift in
species composition could generate a change in peak
quantity of algae apart from the effects of mixed depth.
In oligotrophic lakes, artificial destratification usually
produces little change in cell concentrations (Knoppert,
et al. 1970; Biedermanand Fulton, 1971;Toetz, 1977a,
b; but see Fast, 1971 a). Sometimes, standing stock
increases due to change in mixing depth, although the
change is small in cases of incomplete destratification.
Since the slope of the ascending curve in Figure 1
equals the peak nutrient-limited concentration of algae
(Lorenzen and Mitchell, 1975), the slope will be
smallest for oligotrophic lakes. Hence, any given
change in mixed depth over the range of nutrient-
limited biomasses will result in only small displace-
ments of standing crop in oligotrophic lakes compared
with potential shifts in richer lakes (also, see Forsberg
1 ,0-
NUTRIENT LIMITATION
•LIGHT LIMITATION
MIKED DEPTH .METERS
• THEORETICAL VALUES
• 1968.STRATIFIEO
® I<»69,DESTRATIFIEO
D 1970.DESTRATIFIED
Figure 1. — Theoretical and observed peak biomass of algae in
Kezar Lake (Lorenzen and Mitchell, 1975).
and Shapiro in this volume on shifts in peak biomass
with changing total phosphorus levels).
Blue-green species often control their depth dis-
tribution via buoyancy regulation to take advantage of
specific optima in light, temperature, and nutrients
(Fogg and Walsby, 1971; Konopka, et al. 1978).
Artificial circulation disperses metalimnetic popula-
tions and causes overall decline of Oscillatoria
spp.(Bernhardt, 1967; Weiss and Breedlove, 1973).
Whatever the mechanism, Anabaena spp. are among
the most sensitive forms (Ridley, 1970; Knoppert, et al.
1970; Malueg, et al. 1971; Steichen, et al. 1974;
Barnett, 1975).
Chemical Mechanisms
Mixing techniques have been applied to reduce algal
blooms by curtailing recycling of nutrients from the
profundal zone (Toetz, et al. 1972; Dunst, et al. 1974;
Fast, 1979a). Although PO concentrations are
indeed lowered by destratification (e.g., Haynes, 1973;
Toetz, 1979), the flux of nutrients from profundal
sediments to the overlying water and subsequent
uptake by the plant community could actually increase.
Under aerobic conditions, the higher temperatures in
the sediments after destratification will stimulate
decomposition and release of phosphorus to overlying
waters (Margrave, 1969; Fast, 1971 a; Kamp-Nielsen,
1975). Simultaneously, nutrient exchange across the
mud-water interface is facilitated by increased flow of
water over the sediments and invasion of burrowing
macroorganisms which mix the sediments vertically,
e.g., chironomid larvae (Porcella, et al. 1970; Gallepp,
et al. 1978). Lastly, it is unlikely that circulation
techniques can reduce internal loading of nutrients
from other sources such as "leakage" from littoral
macrophytes (Demarte and Hartman, 1974; Lehman
and Sandgren, 1978).
Even if artificial circulation does reduce phosphorus
regeneration from the sediments, significant changes
in the biota will occur only if internal loading of
nutrients is large relative to input from the watershed
and algal growth is limited by phosphorus (Fast, 1975).
Although the latter appears true in many instances
(Likens, 1972), lakes with nuisance algal blooms are
usually eutrophic and, by definition, experience high
external loading. Also, Lane and Levins (1 977) caution
against overreliance on the concept of a single limiting
nutrient.
The effects of destratification on the concentrations
of dissolved inorganic nutrients in the upper waters of
a lake are unpredictable due to interactions with biota
and organic factions (Toetz, et al. 1972; Fast, 1975). For
example, nutrients may be released by lysed algae
(Robinson, et al. 1969; R. S. Kerr Res. Center, 1970;
McNall, 1971), by decomposition of resuspended
detritus (Hooper, et al. 1953; Fast, 1971 a; Haynes,
1973), or by an abundant zooplankton population
(Devol, 1979).
Biological Mechanisms
An effective destratification often causes a dramatic
shift in species composition of the phytoplankton
community, from dominance by one or a few species of
-------
AERATION/MIXING AND AQUATIC PLANT HARVESTING
127
blue-greens to a predominant assemblage of green
algae (Table 1). Zooplankton readily graze on green
algal species, whereas they reject the inedible and
sometimes toxic blue-greens or grow poorly on them
(Arnold, 1971; Porter, 1973; Webster and Peters,
1978). On the other hand, some gelatinous greens
actually profit from passage through the gut of a
Daphnia (Porter, 1975).
Shapiro (1973; Shapiro, et al. 1975) has induced the
blue-green to green shift in experimental enclosures by
adding CO2 or HCI, both of which lower the pH of the
water. Moreover, adding NO9 and PO« facilitates the
shift. Since the blue-greens decline precipitously
before the greens begin growing rapidly, Shapiro, et al.
(1975) suggest that the shift is mediated by the action
of cyanophages (Shilo, 1971; Lindmark in Shapiro,
1979), rather than by direct competitive replacement.
Indeed, the release of large quantities of PO<~and NH
to the water after the sudden decline of blue-greens in
the enclosures suggests that lysis is occurring.
Destratification essentially mimics Shapiro's ex-
perimental treatments by adding COaand nutrients to
the surface waters through: (1) Mixing of hypolimnetic
CO2and nutrients into the surface layer; (2) recarbona-
tion of waters by atmospheric exchange; and (3)
decreasing the ratio of primary production to respira-
tion through deepening of the mixed layer. In
experimental enclosures, a change in algal species
composition occurs only at pH values less than 8.5, and
the results are unpredictable between pH 7.5 and 8.5
(Shapiro, et al. 1975). Lakes where pH decreased
following circulation also showed an increase in the
ratio of green to blue-green algae; whereas experi-
ments that failed to lower the pH also failed to produce
the shift to greens (Table 2).
In Kezar Lake during 1969, mixing caused a
temporary rise in pH, but after 20 days of aeration, the
pH dropped from 9.0 to 7.1, and at least a small
increase in the ratio of greens to blue-greens ensued
(N.H. Water Supply Pollut. Control Comm. 1971).
Oestratification by pumping hypolimnetic water to the
surface maintained relatively low pH in the epilimnia of
four Ohio lakes and prevented the usual fall blooms of
blue-green algae (Irwin, et al. 1966).
r
CIRCULATION
UQHT LJ
\
B" \
"\
e I
LjJ-
9 \
CYANOPHAQE 1
ACTIVITY 1
1 rJ SS
? 1 ™
E-HHEEN TO 1
EH ALGAE SHIFT 1
_»
Q 1
PHEDATION
ON ZOOPLANKTON
* SSSMT0"
9 INCREASE IN RESPONSE PARAMETER
6 DECREASE IN RESPONSE PARAMETER
Figure 2. — Beneficial effects of artificial circulation on
phytoplankton (Shapiro, 1979).
Table 2. — Eplimnetic pH changes associated with artificial circulation.
Lake
Group la
Cline's Pond
University Lake
Kezar Lake
Stewart Hollow
Caldwell Lake
Pine Lake
Vesuvius Lake
Buchanan Lake
Group llb
Pan/in Lake
Test Res. 1 & II
Starodworski Lake
Lake Calhoun
Ham's Lake
Arbuckle
Lake Catharine
Hyrum Res.
El Capitan Res.
Reference
Malueg, et al. 1971
Weiss and Breedlove, 1973
N H WS PC C 1971
Haynes, 1973
N H WS.P.C.C. 1971
Haynes, 1973
Irwin, et al. 1966
Irwin, et al. 1966
Irwin, et al. 1966
Irwin, et al. 1966
Irwin, et al. 1966
Brown, et al. 1971
Lackey, 1972
Knoppert, et al. 1970
Lossow, et al. 1975
Shapiro and Pfannkuch, 1973
Steichen, et al. 1974
Toetz, 1977b
Toetz, 1977b
Toetz, 1979
Kothandaraman, et al. 1979
Drury, et al. 1975
Fast, 1968
Direction
of Change
1968-
1969 +
-
0
0
0
0?
-
0
1973- 1975
0
1975- 1977
0
0
±
0
PH
Before
6.2-9.6c
7.6"
9.4
6.6
6.8
6.8
7.3
6.9-7.2
6.8-7.3
7.1
6.6-7.2d
?
9.0-9.4"
8.0-8.5"
8.5
>8.0
7.71"
~7.5d
>8.0"
7.8-8.9"
7.5-8.6"
Values
After
6.4-7.2
7.3, 7.0
6.7
9.0
5.5
6.5
7.0-7.5
6.7-7.1
6.8-7.0
6.7
6.7-7.2
>9.0
7.3-8.6
8.0-8.5
7.5
>8.0
7.39
-7.5
>8.0
7.2-9.2
7.7-8.3
" Group I = Lakes in which the ratio of green algae to blue-green algae increased after treatment
b Group II = Lakes in which the ratio of green algae to blue-green algae decreased or stayed the same after treatment
c Control section
" Control year, summer values
-------
128
RESTORATION OF LAKES AND INLAND WATERS
Figure 3. — Some adverse impacts of artificial circulation and
their role in promoting blue-green algae blooms (Shapiro,
1979).
Although mixing caused a temporary decrease o,
epilimnetic pH in Ham's Lake (1973 experiment) and
Starodworski Lake (Poland), the pH remained above 7.3
in both cases, failing to produce a shift from blue-green
algae to green algae (Tables 1 and 2). In Hyrum
Reservoir, where aeration caused microstratification
and a reduction in mixed depth, pH of the surface
waters rose sharply to 9.2 during a bloom of
Aphanizomenon (Drury, et al. 1975).
Figures 2 and 3 summarizes some of the important
mechanisms underlying the effects of artificial circula-
tion on phytoplankton. The risk of adverse impacts can
be minimized by proper design and application of the
mixing system (see Pastorok, et al. In press).
EFFECTS OF ARTIFICIAL CIRCULATION
ON ZOOPLANKTON AND SPECIES
INTERACTIONS IN OPEN WATER
Artificial circulation generally leads to an increase in
the abundance of zooplankton and an expansion of
their vertical distribution (Table 3). Several studies
reported no effects of mixing on the zooplankton but
this result is probably due to inadequate sampling
design (Eufaula Reservoir), incomplete mixing (Hyrum
Reservoir, Arbuckle Lake), or lack of control data
(Ham's Lake, Arbuckle Lake).
Depth Distribution
Most investigators have observed profound changes
in distribution of cladocerans (e.g., Shapiro, et al. 1975;
Brynildson and Serns, 1977). Lackey (1973b) found
that the depth distributions of Cladocera and rotifers
were generally unaffected by artificial circulation, but
Diaptomus spp. tended to occur in deeper water during
the treatment year. Even in lakes where zooplankton
occupy the entire water column before treatment (e.g.,
Ham's Lake and Starodworski Lake), circulation usually
shifts the vertical profile of the population toward lower
depths.
Zooplankton Abundance
Brynildson and Serns (1977) documented a fourfold
increase in Daphnia spp. after mixing of Mirror Lake in
September 1974. Althougn the density of small
cladocerans including Bosmina longirostris and.D/a-
phanosoma leuchtenbergianum showed no significant
Table 3. — Responses of zooplankton to artificial circulation3
Lake
Buchanan Lake
Lake Roberts
Lake Calhoun"
Stewart Lake
Indian Brook Reservoir
Mirror Lake
Reference
Brown, et al. 1971
McNall, 1971
Shapiro and Pfannkuch, 1973
Barnes and Griswold, 1975
Riddick, 1957
Brynildson and Serns, 1977
Abundance"
+
+
+
1974 1975
+
1973 +
Depth Ratio
Distribution Copepods: Cladocerans
+
+
+
+ +
1974
Parvin Lake
El Capitan Reservoir
Starodworski Lakec
Eufaula Reservoir"'6
Ham's Lake0
Arbuckle Lake0'6
Hyrum Reservoir"1"
Lackey, 1973b
Fast, 1971 b
Lossow, et al. 1975
Bowles, 1972
McClintock, 1976
McClintock, 1976
Drury, et al. 1975
+
+
0
0
0
0
+
+
0
0
0
0
0
+
0
o
o
0
= decrease, 0 = no significant change
° Weighted mean density or standing stock
c Zooplankton distributed to bottom before mix
0 Inadequate sampling design or lost samples
" Incomplete mix
-------
AERATION/MIXING AND AQUATIC PLANT HARVESTING
129
change after circulation, calanoid and cyclopoid
copepods increased during both experiments.
During aeration of Starodworski Lake, the relatively
large Daphnia hyalina appeared for the first time and
became especially abundant in lower water (Lossow, et
al. 1975). Bosmina longirostris declined to particularly
low densities during summer of the treatment and was
replaced by the larger B. coregoni. Chaoborus larvae
which are significant predators on small zooplankton
(cf. Pastorok, in press), declined during aeration,
relieving the predation pressure on Bosmina.
Shapiro, et al. (1975) found that the abundance of
Daphnia spp. increased five to eight times during
artificial circulation of Lake Calhoun compared with the
previous control year. Moreover, the large-bodied D.
pulex invaded the lake and became reasonably
common after treatment. Other zooplankters, including
cyclopoids, Diaptomus, Bosmina, and Diaphanasoma
increased less. Although Lackey (1973b) reported a
significant decline in the population of D. schodleriand
Cladocera in general during treatment of Parvin Lake,
the control year may have been unusual due to
absence of the late summer bloom of Aphanizomenon
flos-aquae (cf. Lackey, 1973a).
The growth of zooplankton populations following
destratification could be caused by several factors: (1)
Resuspension of detritus, creating additional food
resources for filter-feeders (Saunders, 1972); (2) the
shift from blue-green algae to green algae (Table 1);
and (3) habitat expansion for both zooplankton and
planktivorous fishes (Table 3). The dimly lit bottom
waters serve as a refuge for zooplankton, protecting
them from visual predators (Zaret and Suffern, 1976).
The reduction in encounter rate between fish and their
prey lessons predation pressure on the zooplankton,
allowing population growth and invasion of large-
bodied forms, especially Daphnia (Shapiro, 1979; cf.
Hrbacek, et al. 1961; Andersson, et al. 1978;
DeBernardi and Guissani, 1978).
In turn, large herbivores such as Daphnia pulicaria
are more effective grazers of algae than are small
zooplankton (Haney, 1973; Hrbacek, et al. 1978). They
also release less phosphorus per unit body weight than
the smaller forms (Bartell and Kitchell, 1978).
Andersson, et al. (1978) found that dense populations
of fish in experimental enclosures resulted in low
numbers of planktonic cladocerans, high concentra-
tions of chlorophyll, and blooms of blue-green algae. In
enclosures without fish, large cladocerans prospered
and grazed the phytoplankton down to low levels.
EFFECTS OF ARTIFICIAL CIRCULATION
ON BENTHIC MACROINVERTEBRATES
The responses of benthic communities to lake
aeration/circulation have been relatively consistent;
i.e., increases in number of taxa, diversity, and
biomass, especially in profundal areas (Table 4). In two
lakes receiving low nutrient inputs, Parvin Lake and
Section Four Lake, population densities showed a
generalized decline or no change. Although the
hypolimnion of Parvin Lake was normally anoxic during
late summer while the deeper areas of Section Four
Lake remained high in oxygen, the mechanisms
producing declines in chironomid densities may have
been similar. Both lakes normally had dominant
chironomid assemblages in deep water prior to
aeration. The decline in overall densities may have
resulted from increased midge emergence due to the
warmer bottom temperatures during lake circulation. In
both lakes, the other insect larvae and invertebrates
such as Asellus and Hyalella, which were abundant in
littoral areas, did not invade the hypolimnion following
aeration. Therefore, overall profundal biomass de-
clined.
Four of the five lakes in which Chaoborus formed a
significant component of the profundal benthos
displayed a general decline in larval density following
aeration (Table 4). The exception was Parvin Lake in
which Chaoborus density did not change. In Cox Hollow
Lake there was pronounced decline in Chaoborus
associated with replacement of C. punctipennis by C.
albatus (Wirth, et al. 1970). Prior to aeration Chaoborus
was the only profundal macroinvertebrate in Stewart
Lake, but following treatment the larvae were almost
completely absent, having been replaced by oligo-
chaetes and chironomids (Barnes and Griswold, 1975).
The distributional characteristics of Chaoborus are
consistent with the observed declines in Chaoborus
densities during lake aeration. Aeration of bottom
strata removes the anoxic refugia of Chaoborus, thus
exposing the larvae to intense fish predation. Since
third and fourth instar Chaoborus are relatively large
organisms (6 to 15 mm), they are a preferred food item
for zooplanktivorous fish (Northcote, et al. 1978; von
Ende, 1979). Field studies have shown that the
migratory C. punctipennis occurs in lakes with fish
while the non-migratory C. americanus is excluded
from fish lakes (von Ende, 1979). Moreover, introduc-
tion offish predators into lakes has virtually eliminated
C. americanus and markedly reduced the densities of C.
trivittatus, a deeper dwelling species (Northcote, et al.
1978).
In lakes showing declines in Chaoborus densities
during aeration, the profundal areas were occupied by
increased densities of other fauna such as oligo-
chaetes, chironomids, and other insect larvae (e.g.,
Wilhm and McClintock, 1978. Sikorowa, 1978). These
detritivores responded to the generally rich deposits of
organic material by establishing relatively high stand-
ing crops. Thus, aeration may modify the trophic
structure of the community by reducing zooplankton
predators (i.e., Chaoborus) and increasing benthic
detritivores. Organisms such as chironomid larvae are
important food items for a variety of fish species. The
high utilization of benthic fauna and the influence of
fish predation on prey population densities are
indicated in field studies such as Andersson, et al.
(1978).
EFFECTS OF ARTIFICIAL CIRCULATION
ON FISHES
In stratified lakes, coldwater fishes such as sal-
monids may be compressed into a narrow layer of
available metalimnetic habitat by warm water above
and anoxic conditions below. In all cases where depth
distribution has been evaluated, fish have been
observed to expand their vertical distribution in
response to lake destratification.
-------
130
RESTORATION OF LUKES AND INLAND WATERS
Prior to destratification of Mirror Lake, trout and
yellow perch were confined to the epilimnion and
metalimnion (Brynildson and Serns, 1977). During the
spring and late summer the maximum depth occur-
rence of the two fish species was about 5 and 7 meters,
respectively, corresponding to a dissolved oxygen level
of about 3 to 4 mg/l. After destratification fish were
distributed throughout the water column to the
maximum depth of 13 meters. Trout were essentially
evenly distributed while yellow perch occurred from 4
to 13 meters.
After partial destratification of Lake Arbuckle in late
summer, gizzard shad, freshwater drum, white crappie,
and black bullhead all displayed increased depth
distributions when compared with pre-circulation
conditions (Gebhart and Summerfelt, 1976). In 1975
the total available fish habitat (as defined by the 2 mg/l
DO isopleth) increased from 53 percent of lake volume
in August to 99 percent of total volume in September
following treatment. Habitat expansion has also been
observed in El Capitan Reservoir for channel catfish,
threadfin shad, and walleye (Fast, 1968), and in Lake
Calhoun for yellow perch, bluegill, and crappie (Shapiro
and Pfannkuch, 1973). Aeration of Casitas Reservoir
has allowed the establishment of a year-round trout
fishery (Barnett, 1975).
It is generally assumed that an expanded habitat
benefits fish populations because of increased food
supply and alleviation of crowding into epilimnetic
strata during the summer. Comprehensive studies at
Lake Arbuckle did indicate increased growth of bottom-
feeding fishes, but the results varied with species and
year of study (Gebhart and Clady, 1977). Increased
growth at Stewart Lake (Barnes and Griswold, 1975)
was apparently caused by selective elimination of
stunted bluegills. Although the stimulation of fish
growth and production is a conceivable benefit of
aeration, it has not been evaluated in most projects.
Most studies were of limited duration and fish
populations may not have reached equilibrium with the
modified lake environment. Moreover, some of the
lakes contained already stressed populations (e.g.,
overcrowded and stunted centrarchids) slow to respond
to habitat improvements (e.g., Wirth, et al. 1970).
In lakes with severe winter-kill problems, aeration
during fall and winter reduces mortality rates (Halsey,
1968). In warmer areas, circulation of the lake in
summer will increase the heat budget and may result
in adverse water temperatures for salmonids, such as
occurred in Puddingstone Reservoir (Fast and St.
Amant, 1971). Localized aeration and partial destratifi-
cation could allow for some cooler areas with sufficient
DO for trout survival (e.g., Casitas Reservoir); however,
the potential for other benefits such as water quality
changes would be considerably less.
Adverse impacts follow destratification whenever
the oxygen demand associated with resuspended
particulates and reduced compounds lowers dissolved
oxygen in the entire lake to levels below 2 to 3 mg/l. A
dissolved oxygen concentration of at least 5 mg/l is
generally required for maintenance of good game fish
populations (U.S. EPA, 1976).
Table 4. — Responses of benthic macroinvertebrates to artificial circulation.
Lake
Reference
Organism
density
No. of Species
(or diversity)
Ham's Lake
Starodworskie Lake
Lake Catherine
Parvin Lake
El Capitan Reservoir
Cox Hollow Lake
University Lake
Stewart Lake
Section Four Lake
Wilhm and McClmtoch, 1978
Sikorowa, 1978
Kothandaraman, et al. 1979
Lackey, 1973c
Inland Fish. Branch, 1970
Wirth, et al. 1970
Weiss and Breedlove, 1973
Barnes and Griswold, 1975
Fast, 1971 a
varied"
a Chironomids only
" Chironomids -, others 0
REFERENCES
Andersson, G. et al. 1978. Effects of planktivorous and
benthivorous fish on organisms and water chemistry in
eutrophic lakes. Hydrobiologica 59:9.
Arnold, D. E. 1971. Ingestion, assimilation, survival, and
reproduction by Daphnia pulex fed seven species of blue-
green algae. Limnoi. Oceanogr. 16:906.
Barnes, M. D., and B. L. Griswold. 1975. Effects of artificial
nutrient circulation on lake productivity and fish growth.
Spec. Conf. Lake Reaeration Res. Am. Soc, Civil Eng.,
Gatlinburg, Tenn.
Barnett, R. H. 1975. Case study of reaeration of Casitas
Reservoir. Spec. Conf. Lake Reaeration REs. Am. Soc. Civil
Eng , Gatlinburg, Tenn.
Bartell, S. M., and J, F. Kitchell. 1978. Seasonal impact of
planktivory on phosphorus release by Lake Wingra
zooplankton. Verh. Int. Verein. Limnoi. 20:466.
Bernhardt, H. 1967. Aeration of Wahnbach Reservoir without
changing the temperature profile. Jour. Am. Water Works
Assoc. 9:943.
Biederman, W. J., and E. E. Fulton. 1971. Destratification
using air. Am. Water Works Assoc. 63:462.
Bowles, L. G. 1972. A description of the spatial and temporal
variations in species composition and distribution of pelagic
net zooplankton in the central pool of Eufaula Reservoir,
Oklahoma, with comment on forced aeration destratification
experimentation. Trans. Kans. Acad. Sci. 75:156
-------
AERATION/MIXING AND AQUATIC PLANT HARVESTING
131
Brown, D. J., et al. 1971. Progress report on the
destratification of Buchanan Lake. Ontario Water Res.
Comm., AID for Lakes Program (Artificially Induced
Destratification).
Brynildson, O. M., and S. L. Serns. 1977. Effects of
destratification and aeration of a lake on the distribution of
planktonic Crustacea, yellow perch and trout. Tech. Bull. 99.
Wis. Dep. Nat. Resour.
DeBernardi, R., and G. Giussani. 1978. Effect of mass fish
mortality on zooplankton structure and dynamics in a small
Italian lake (Lago di Annone). Verh. Int. Verein. Limnol.
20:1045.
DeMarte, J. A., and R. T. Hartman. 1974. Studies on absorption
of 32P, 59Fe, and 45Ca by water-milfoil (Myriophyllum
exalbescens FERNALD). Ecology 55:188.
Devol, A. H. 1979. Zooplankton respiration and its relation to
plankton dynamics in two lakes of contrasting trophic state.
Limnol. Oceanogr. 24:893.
Drury, D. D., D. B. Porcella, and R. A. Gearheart. 1975. The
effects of artificial destratification on the water quality and
microbial populations of Hyrum Reservoir. PRJEW 011-1.
Utah Water Res. Lab.
Dunst, R. C., et al. 1974. Survey of lake rehabilitation
techniques and experiences. Tech. Bull. 75. Wis. Dep. Nat.
Resour.
Fast, A. W. 1968. Artificial destratification of El Capitan
Reservoir by aeration. Part 1: Effects on chemical and
physical parameters. Fish. Bull. 141. Calif. State Dep. Fish
Game. "
1971 a. The effects of artificial aeration on lake
ecology. Water Pollut. Control Res. Ser. 16010 EXE 12/71.
U.S. Environ. Prot. Agency.
1971b. Effects of artificial destratification on
zooplankton depth distribution. Trans. Am. Fish. Soc.
100:355.
1973. Effects of artificial destratification on
primary production and zoobenthos of El Capitan reservoir,
California. Water Resour. Res. 9:607.
1975. Artificial aeration and oxygenation of lakes
as a restoration technique. Symp. Recovery of Damaged
Ecosystems, Virginia Polytechnic Institute and State
University, Blacksburg.
1979. Artificial aeration as a lake restoration
technique. Proc. Natl. Conf. Lake Restoration. U.S. Environ.
Prot. Agency.
Fast, A. W., and M. W. Lorenzen. 1976. Synoptic survey of
hypolimnetic aeration. Jour. Environ. Eng. Div. Am. Soc. Civil
Eng. 102:1161.
Fast, A. W., and J. A. St. Amant. 1971. Nighttime artificial
aeration of Puddihgstone Reservoir, Los Angeles County,
California. Calif. Fish Game 57:213.
Fast, A. W., B. Moss, and R. G. Wetzel. 1973. Effects of
artificial aeration on the chemistry and algae of two Michigan
lakes. Water Resour. Res. 9:624.
Fogg, G. E., and A. E. Walsby. 1971. Buoyancy regulation and
the growth of planktonic blue-greerfalgae. Mitt. Int. Verein.
Limnol. 19:182.
Gallepp, G. W., J. F. Kitchell, and S. M. Bartell. 1978.
Phosphorus release from lake sediments as affected by
chironomids. Verh. Int. Verein. Limnol. 20:458.
Garton, J. E., R. G. Strecker, and R. C. Summerfelt. 1978.
Performance of an axial flow pump for lake destratification.
Pages 336-346 in W. A. Rogers, ed. Proc. 13th Annu. Conf.
S.E. Assoc. Fish Wildl. Agencies.
Gebhart, G. E., and M. D. Clady. 1977. Effects of mechanical
mixing in reservoirs on seasonal and annual growth rates of
fishes. Tech. Comple. Rep. A-069-OKLA. Oklahoma Water
Resour. Res. Inst.
Gebhart, G. E., and R. C. Summerfelt. 1976. Effects of
destratification on depth distribution of fish. Jour. Environ
Eng. Div. Am. Soc. Civil Eng. 102:1215.
Halsey, T. G. 1968. Autumnal and over-winter limnology of
three small eutrophic lakes with particular reference to
experimental circulation and trout mortality. Jour. Fish. Res.
Board Can. 25:81.
Haney, J. F. 1973. An in-situ examination of the grazing
activities of natural zooplankton communities Arch.
Hydrobiol. 72:87.
Margrave, B. T. 1969. Epibenthic algae production and
community respiration in the sediments of Marion Lake.
Jour. Fish Res. Board Can. 26:2003.
Haynes, R. C. 1973. Some ecological effects of artificial
circulation on a small eutrophic lake with particular
emphasis on phytoplankton. I. Kezar Lake experiment.
Hydrobiologia 43:463.
Holdren, G. C., Jr., and D. E. Armstrong. 1980. Factors
affecting phosphorus release from intact lake sediment
cores. Environ. Sci. Technol. 14:79.
Hooper, F. F., R.C. Ball, and H. A.Tanner. 1953. An experiment
in the artificial circulation of a small Michigan Lake. Trans.
Am. Fish Soc. 82:222.
Hrbacek, J., B. Desortova, and J. Popovsky. 1978. Influence of
the fishstock on the phosphorus-chlorophyll ratio. Verh. Int.
Verein. Limnol. 20:1624.
Hrbacek, J.,et al. 1961. Demonstration of the effect of the fish
stock on the species composition of zooplankton and the
intensity of metabolism of the whole plankton association.
Verh. Int. Verein. Limnol. 14:192.
Inland Fisheries Branch. 1970. Effects of artificial
destratification on distribution of bottom organisms in El
Capitan Reservoir. Fish. Bull. 148. Calif. Dep. Fish Game.
Irvin, W. H., J. M. Symons, and G. G. Robeck. 1966.
Impoundment destratification by mechanical pumping. Jour.
San. Eng. Div. Am. Soc. Civ. Eng. 92(SA6):21.
Kamp-Nielsen, L. 1974. Mud-water exchange of phosphate
and other ions in undisturbed sediment cores and factors
affecting the exchange rates. Arch. Hydrobiol. 73:218.
1975. Seasonal variation in sediment-water
exchange of nutrient ions in Lake Esrom. Verh. Int. Verein.
Limnol. 19:1057.
Knoppert, P L., et al. 1970. Destratification experiments at
Rotterdam. Jour. Am. Water Works Assoc. 62:448.
Konopka, A., T. D. Brock, and A. E. Walsby. 1978. Buoyancy
regulation by planktonic blue-green algae in Lake Mendota,
Wisconsin. Arch. Hydrobiol. 83:524.
Kothandaraman, V., D. Roseboom, and R. L. Evans. 1979. Pilot
Lake restoration investigations: Aeration and destratification
in Lake Catharine. III. State Water Survey.
Lackey, R. T. 1972. Response of physical and chemical
parameters to eliminating thermal stratification in a
reservoir. Water Res. Bull. 8:589.
1973a. Artificial reservoir destratification effects
on phytoplankton. Jour. Water Pollut. Control. Fed. 45:668.
1973b. Effects of artificial destratification on
zooplankton in Parvin Lake, Colorado. Trans. Am. Fish. Soc.
102:450.
1973c. Bottom fauna changes during artificial
reservoir destratification. Water Res. 7:1349.
Lane, P., and R. Levins. 1977. The dynamics of aquatic
systems. 2. The effects of nutrient enrichment on model
plankton communities. Limnol. Oceanogr. 22:454.
Lehman, J. T., and C. D. Sandgren. 1978. Documenting a
seasonal change from phosphorus to nitrogen limitation in a
small temperate lake, and its impact on the population
dynamics of Asterionella. Verh. Int. Verh. Limnol. 20:375.
-------
132
RESTORATION OF LAKES AND INLAND WATERS
Likens, G. E., ed. 1972. Nutrients and eutrophication: The
limiting-nutrient controversy. Am. Soc. Limnol. Oceanogr.
Spec. Symp. 1.
Lorenzen, M. W., and A. W. Fast. 1977. A guide to
aeration/circulation techniques for lake management. Res.
Ser. EPA-600/3-77-004. U.S. Environ. Prot. Agency.
Lorenzen, M. W., and R. Mitchell. 1975. An evaluation of
artificial destratification for control of algal blooms. Jour.
Am. Water Works Assoc. 67:373.
Lossow, K., eta 1.1975. Results of research on the influence of
aeration on the physico-chemical systems and biological
complexes in the Starodworskie Lake obtained hitherto. Pol.
Arch. Hydrobiol. 22:195.
Malueg, K. W., et al. 1971. Effect of induced aeration on
stratification and eutrophication processes in an Oregon
farm pond. Geophys. Monogr. Ser. 17:578.
McChntock, N. 1976. Effects of artificial destratification on
zooplankton of two Oklahoma reservoirs. M.S. thesis.
Oklahoma State University.
McCullough, J. R. 1974. Aeration revitalizes reservoir. Water
Sewage Works. 121:84.
McNall, W. J. 1 971. Destratification of lakes. Federal AID Proj.
F-22-R-11, J of C-8, Job Prog. Rep.
Mortimer, C. H. 1941, 1942. The exchange of dissolved
substance between mud and water in lakes. Jour. Ecol.
29:280; 30-147.
1971. Chemical exchanges between sediments
and water in the Great Lakes — speculations on probable
regulatory mechanisms. Limnol. Oceanogr. 16:387.
New Hampshire Water Supply and Pollution Control
Commission. 1971. Algae control by mixing. Concord, N.H.
Northcote, T. G., H.W. Lorz, andJ. C. MacLeod. 1964. Studies
on diel vertical movement of fishes in a British Columbia lake.
Verh. Int. Verein. Limnol. 15:940.
Northcote, T. G., C.J.Walters, and J. M. B. Hume. 1 978. Intitial
impacts of experimental fish introduction on the
macrozooplankton of small oligotrophic lakes. Verh. Int.
Verein. Limnol. 20:2003.
Oskam, G. 1978. Light and zooplankton as algae regulating
factors in eutrophic Biesbosch reservoirs. Verh. Int. Verein.
Limnol. 20:1612.
Pastorok, R. A. In press. Selection of prey by Chaoborus larvae:
A review and new evidence for behavioral flexibility. Am.
Soc. Limnol. Oceanogr. Spec. Symp. 3.
Pastorok, R. A., T. C. Ginn, and M. W. Lorenzen. In press.
Evaluation of aeration/circulation as a lake restoration
technique. U.S. Environ. Prot. Agency.
Porcella, D. B., J. S. Kumagai, and E. J. Middlebrooks. 1970.
Biological effects on sediment-water nutrient interchange.
Jour. San. Eng. Div. Am. Soc. Civil Eng. 96:911.
Porter, K. G. 1973. Selective grazing and differentia I digest ion
of algae by zooplankton. Nature 244:179.
1975. Viable gut passage of gelatinous green algae
ingested by Daphnia. Verh. Int. Verein. Limnol. 19:2840.
Riddick, T. M. 1957. Forced circulation of reservoir waters
yields multiple benefits at Ossming, New York. Water
Sewage Works 104:231.
Ridley, J E. 1970. The biology and management of eutrophic
reservoirs. Water Treat. Exam. 19:374.
Ridley, J. E., P. Cooley, and J. A. P. Steel. 1966. Control of
thermal stratification in Thames Valley reservoirs. Proc. Soc.
Water Treat. Exam. 15:225.
Robinson, E. L., W. H. Irwin, and J. M. Symons. 1969.
Influence of artificial destratification on plankton populations
in impoundments Trans. Ky. Acad. Sci. 30:1.
R. S. Kerr Research Center. 1970. Induced aeration of small
mountain lakes. Water Pollut. Control Res. Ser. 16080-
11/70. U.S. Environ. Prot. Agency.
Saunders, G. W. 1972. The transformation of artificial
detritus in lake water. Mem. 1st. Ital. Idrobiol.
29(Suppl.):261.
Shapiro, J. 1973. Blue-green algae: Why they become
dominant. Science 197:382.
1979. The need for more biology in lake
restoration. Proc. Natl. Conf. Lake Restoration. U.S. Environ.
Prot. Agency.
Shapiro, J. and H. 0. Pfannkuch. 1973. The Minneapolis
chain of lakes. A study of urban drainage and its effects.
Interim Rep. 9. Limnol. Res. Center, University of
Minnesota.
Shapiro, J., V. Lamarra, and M. Lynch. 1975. Biomanipula-
tion: An ecosystem approach to lake restoration. Pages 85-
95 in P. L. Brezonik and J. L. Fox, eds. Proc. Symp. Water
Quality Manage. Through Biological Control. University of
Florida and U.S. Environ. Prot. Agency, Gainesville.
Shilo, M. 1971. Biological agents which cause lysis of blue-
green algae. Mitt. Int. Verein. Limnol. 19:206.
Sikorowa, A. 1978. Changes in the distribution and number
of the bottom fauna as an effect of artificial lake aeration.
Verh. Int. Verein. Limnol. 20:1000.
Smith, S. A., D. R. Knauer, andT. L. Wirth. 1975. Aeration as
a lake management technique. Tech. Bull. 87. Wis. Dep.
Nat. Resour.
Steichen, J. M., J. E. Carton, and C. E. Rice. 1974. The effect
of lake destratification on water quality parameters. Annu.
Meet. Am. Soc. Agric. Eng.
Symons, J. M., et al. 1967. Impoundment destratification for
raw water quality control using either mechanical or
diffused-air pumping. Jour. Am. Water Works Assoc.
59:1268.
Symons, J. M., J. K. Carswell, and G. G. Robeck. 1970.
Mixing of water supply reservoirs for quality control. Jour.
Am. Water Works Assoc. 62:322.
Thomas, E. A. 1966. Der Pfaffikersee vor, wahrend, und nach
kunstlicher Durchmisching [In German]. Verh. Int. Verein.
Limnol. 16:144.
Toetz, D. W. 1977a. Biological and water quality effects of
whole lake mixing. Final Tech. Rep. A-068-OKLA. Okla.
Water Resour. Res. Inst.
1977b. Effects of lake mixing with an axial flow
pump on water chemistry and phytoplankton. Hydrobiologia
55:129.
1979. Biological and water quality effects of
artificial mixing of Arbuckle Lake, Oklahoma, during 1977.
Hydrobiologica 63:255.
Toetz, D., J. Wilhm, and R. Summerfelt. 1972. Biological
effects of artificial destratification and aeration in lakes and
reservoirs — Analysis and bibliography. Unit Rep. REC-ERC-
72-33. Okla. Coop. Fish.
U.S. Environmental Protection Agency. 1976. Quality criteria
for water. U.S. Government Printing Office, Washington,
D.C.
von Ende, C. N. 1979. Fish predation, interspecific predation,
and the distribution of two Chaoborus species Ecology
60:119.
Webster, K. E., and R. H. Peters. 1978. Some size-dependent
inhibitions of large cladoceran filterers in filamentous
suspensions. Limnol. Oceanogr. 23:1238.
Weiss, C. M., and B. W. Breedlove. 1973. Water quality
changes in an impoundment as a consequence of artificial
destratification. Rep. 80. N.C. Water Resour. Res Inst
-------
AERATION/MIXING AND AQUATIC PLANT HARVESTING 133
Wilhm, J., and N. McClintock. 1978. Dissolved oxygen
concentration and diversity of benthic macro-invertebrates
in an artificially destratified lake. Hydrobiologia 57:163.
Wirth, T. L, and R. C. Dunst. 1967. Limnological changes
resulting from artificial destratification and aeration of an
impoundment. Fish. Res. Rep. 22. Wis. Conserv. Dep.
Wirth, T. L. et al. 1970. Manipulation of reservoir waters for
improved quality and fish population response. Rep. 62.
Wis. Dep. Nat. Resour., Madison.
Zaret, T. M., and J. S. Suffern. 1976. Vertical migration in
zooplankton as a predator avoidance mechanism. Limnol.
Oceanogr. 21:804.
-------
134
PREDICTING THE ALGAL RESPONSE
TO DESTRATIFICATION
B. R. FORSBERG
Graduate Student, Limnologieal Research Center
University of Minnesota
Minneapolis, Minnesota
J. SHAPIRO
Director, Limnologieal Research Center
University of Minnesota
Minneapolis, Minnesota
ABSTRACT
The response of phytoplankton communities to artificial destratification has been quite variable.
The mechanisms underlying this variability were investigated in eight field experiments on two
Minnesota lakes. Polyethylene enclosures were used in controlled experimental designs to
investigate specific response mechanisms. A mathematical model was developed to describe the
community response under different mixing regimes. The peak concentration and total amount of
chlorophyll a in the mixed layer were predicted to either increase, decrease, or remain the same
depending on changes in the mixed depth and the concentration of total phosphorus in the mixed
layer following destratification. Changes in species composition during artificial circulation
depended on the mixing rate achieved. Blue-green algae increased in relative abundance at the
slower mixing rates while green algae and diatoms were favored at the fastest mixing rates. The
shift to green algae occurred only during conditions of low pH and high nutrient availability
associated with rapid mixing and is therefore most likely to occur when relatively deep productive
lakes are rapidly mixed.
INTRODUCTION
Artificial circulation often has been proposed as a
method for controlling algal blooms in lakes and
reservoirs. However, in practice, the response of the
phytoplankton to this treatment has been quite
variable. Pastorok, et al. (1 980) have recently summar-
ized the results of a large number of destratification
experiments. In 40 experiments where destratifcation
was relatively complete, they found that only 65
percent led to a significant change in algal biomass; of
these, 30 percent increased and 70 percent decreased
algal biomass. Changes in algal species composition
following destratification have also been variable. A
shift in dominance from blue-green to green species
has been reported by several investigators (Irwin, et al.
1966; Malueg, et al. 1971; Weiss and Breedlove,
1973). However, in some cases diatoms (Bernhardt,
1967) and in others blue-greens(Knoppert, et al. 1970;
Drury, et al. 1975) have increased in relative
abundance. These results indicate that the effects of
artificial circulation on the phytoplankton are not
always beneficial. It is therefore important that we
understand the mechanisms underlying these effects
so that our ability to predict the algal response will
improve and circulation techniques can be used more
effectively.
A number of mechanisms have been proposed to
explain the response to the phytoplankton during
artificial circulation. Several authors have constructed
mathematical models of algal growth to predict the
response at the community level (Murphy, 1962; Bella,
1970; Lorenzen and Mitchell, 1973; Oskam, 1978).
While these models often ignore important aspects of
algal growth (e.g. algal losses due to sinking and
grazing), preliminary tests (Lorenzen and Mitchell,
1975; Oskam, 1978) indicate that this general
approach may eventually provide a theoretical frame-
work for predicting the community response. Fewer
mechanisms have been proposed to explain shifts in
species composition associated with artificial circula-
tion. However, Shapiro (1973) has suggested that
shifts from blue-green to green species reported in the
literature might be related to decreases in pH and
increases in nutrient availability which sometimes
occur following destratification. He found similar shifts
when he reduced pH and added nutrients to natural
assemblages of algae in controlled field experiments.
The fact that most of the blue-green to green shifts
found during artificial circulation also occurred during
conditions of low pH (Irwin, et al. 1966; Haynes, 1971;
Weiss and Breedlove, 1973) tends to support this
hypothesis.
We present here an overview of the results from
eight field experiments which were conducted over a
period of 3 years on two Minnesota Lakes. These
experiments were designed to investigate specific
mechanisms proposed to explain variability in the algal
response during artificial circulation. Particular em-
phasis was placed on developing and testing a
-------
AERATION/ MIXING AND AQUATIC PLANT HARVESTING
135
mathematical model capable of describing the com-
munity level response and on evaluating the pH-shift
mechanism proposed by Shapiro (1973). Our purpose
in this paper is to summarize those results which have
a direct bearing on our ability to predict the algal
response during artificial circulation.
STUDY SITES
The experiments were conducted on two small
eutrophic lakes near Minneapolis and St. Paul during
the ice-free months between 1976 and 1978. Little
Lake Johannah, which has a surface area of 7.3
hectares and a maximum depth of 13 meters, was the
site of experiments 1, 2, 4, 5 and 7. Experiments 3, 6
and 8 were carried out on Twin Lake which has a
surface area of 15 hectares and a maximum depth of
12 meters. Additional details on the limnology of Twin
Lake and Little Lake Johannah are described else-
where (Allott, 1979; Shapiro, et al. in prep.).
EXPERIMENTAL DESIGNS AND
METHODS
The experiments were designed to simulate the
effects of mixing without circulating a whole lake. This
was accomplished by enclosing vertical columns of
lake water in polyethylene bags and then circulating
with compressed air. These enclosures were made of 6
mil extruded polyethylene cylinders with a diameter of
1 meter and depth of 8 meters. They were open at the
top, reinforced with PVC tubing on the sides, and either
open or closed at the bottom. Open bottom enclosures
were used to simulate natural conditions. They were
held open at the bottom by weighted PVC hoops and
lowered slowly from the surface to entrain an
undisturbed column of water.
Closed bottom enclosures were used to study the
effects of selected hypolimnetic constituents. These
bags were first filled with surface water and then
allowed to stratify through thermal conduction with
their surroundings (the thermocline was generally at a
depth of 3 meters in Little Lake Johannah and 5 meters
in Twin Lake during the experiments). Water was then
withdrawn from the hypolimnetic portions and, after
specific additions, returned to the same depth. These
additions included various combinations of nitrogen,
phosphorus, alkalinity, and carbon dioxide designed to
simulate natural hypolimnetic levels.
In each experiment several enclosures were sus-
pended from outriggers attached to rafts as shown in
Figure 1. Various treatments were then applied to the
different enclosures in controlled experimental designs
to investigate specific response mechanisms. In
addition to the manipulations of hypolimnetic chem-
istry in the closed bottom bags, the mixing rate and
mixed depth were varied in the enclosures by adjusting
the air flow rate and depth of air release, respectively.
The flexibility of this general design made it possible to
simulate a wide range of mixing conditions. For a
detailed description of specific experimental designs
and analytical procedures refer to Shapiro, et al., in
prep.
dlfluser
Figure 1 ./ns/ft/apparatusfor suspending
experimental enclosures.
RESULTS AND DISCUSSION
By varying the mixing regime within the different
experimental designs it was possible to produce a
range of algal responses similar to that observed in
whole lake destratification experiments. This made it
possible to evaluate specific response mechanisms
proposed to explain this variability in the algal
response. Mechanisms of potential importance at both
the community and species level were evaluated.
The algal community response during artificial
circulation. A mathematical model of algal growth was
constructed to provide an appropriate theoretical
framework for evaluating the community response
during artificial circulation, (refer to Forsberg and
Shapiro, 1980, and Shapiro, et al. in prep, for a more
detailed development of the model) While it contains
elements of expressions presented by Tailing (1957),
Megard (1974), and Senft (1978) the development of
the model follows directly from the work of Lorenzen
and Mitchell (1973). They considered the effects of
nutrient and light limitation independently and devel-
oped separate expressions to predict nutrient and light
limited peak algal biomass following destratification.
We also evaluated the effects of both nutrient and light
limitation on algal growth, but, instead of treating them
separately, we considered their effects simultaneously
and derived a single expression for the peak
concentration of chlorophyll a
c* =
where,
DGZmBc + [Ln(zlo/lz,)Psatkq]/TP
c* = the peak concentration of chlorophyll a in
the mixed layer (mg Chi nrf2)
lo = the light intensity just below the surface
lz' = the light intensity at the depth r
T = the depth empirically defined as, z — (daily
integral rate of photosynthesis, mg C rrf 2d~1)/
-------
136
RESTORATION OF LAKES AND INLAND WATERS
(maximum daily volumetric rate of photo-
synthesis, mg C m~3d~1)
Psat= the maximum daily specific rate of
photosynthesis in a nutrient saturated mixed
layer (mg C mg Chl~1d~1)
D — the specific loss rate (day"1)
0 = the ratio of carbon to chlorophyll a in the algae
mg C mg Chi"1)
zm = the depth of the mixed layer (meters)
ec = the partial attenuation coefficient of chlorophyll
a (m2 mg Chi"1)
ew — the residual extinction coefficient of the water
(rrf1)
kq — the level which the ratio of total phosphorus to
'" chlorophyll a must exceed before photosyn-
thesis will occur (mg P mg CM"1)
TP = the concentration of total phosphorus in the
mixed layer (mg P m"3)
Data from 10 experimental enclosures in experiment
6 (Twin Lake) were used to evaluate the parameters in
equation 1 and provide a preliminary test of the model.
The enclosures were all closed at the bottom. Nitrogen,
phosphorus, and alkalinity were added to the hypo-
limnia of circulated enclosures before mixing in
amounts designed to simulate natural levels. Eight of
the enclosures were artificially circulated to a depth of
7 meters while two control bags remained stratified
with a mixed depth of 5 meters.
TP and Zm were the parameters in equation 1 which
changed significantly during artificial circulation. All
other factors were assigned constant values. The
model was then used to simulate the effects of changes
in TP and z™ on the peak concentration of chlorophyll a
in Twin Lake. The results of this simulation are shown
in Figure 2a. Each line in this figure indicates the effect
of changes in Zm on c* at a single level of TP The TP
levels chosen for the simulation were those observed
in the lake and in several of the enclosures at the point
of maximum yield (i.e., maximum observed ration,
chlorophyll/total phosphorus). The chlorophyll concen-
trations observed at the point of maximum yield
provided field estimates of c* and are indicated by black
circles on Figure 2a at the appropriate mixed depths.
The lines connecting these black circles to the
simulation lines represent the differences between the
predicted and observed c* values.
These differences are generally small indicating good
agreement between the simulation and field results.
The simulation results indicate that, at a given level of
TP, c* will decrease as the mixed depth is increased.
However, if the concentration of TP in the mixed layer
changes during mixing, c* may either increase,
decrease, or remain the same depending on the
direction and magnitude of the change. When the
mixed depth was increased from 5 to 7 meters in the
circulated enclosures the concentration of TP in-
creased in the mixed layer changes during mixing, c*
may either increase, decrease, or remain the same
depending on the direction and magnitude of the
change. When the mixed depth was increased from 5
to 7 meters in the circulated enclosures the concentra-
tion of TP increased dramatically, resulting in a large
increase in the concentration of chlorophyll a (c.f.
Figure 2a). Because of this increase in TP, the mixed
depth would have to be increased to a depth greater
than 20 meters before a reduction c* would occur.
Since the mean depth of Twin Lake is only 5.5 meters,
destratification would not be effective in reducing the
peak chlorophyll concentration.
The model was also used to simulate the effects of
changes in TP and Zm on peak algal biomass which was
defined as the total amount of chlorophyll a beneath a
square meter of lake surface or c*z™. The results of this
simulation are shown in Figure 2b. Again, the circles
represent field observation and the agreement be-
tween predicted and observed results is good. The
model predicts that, at a given level of TP, peak biomass
will reach a maximum level at a mixed depth of about
15 meters. Peak biomass will increase during artificial
circulation if the final mixed depth is less than this
value and may either increase, decrease or remain the
same at greater mixed depths. However, if the
concentration of TP increases during mixing, as it did in
the circulated enclosures, the level of c*zm achieved
will be higher than would otherwise be expected.
Peak biomass was much higher in the circulated
enclosures than in either the lake or the control bag
and this difference was primarily due to the increase in
TP which occurred during mixing. These results
indicate that, even if TP didn't change, Twin Lake would
have to be mixed to a depth greater than 30 meters
(about six times its mean depth) before a reduction in
peak biomass would occur.
Figure 2.-The effect of changes in the mixed depth, zm,
and the concentration of total phosphorus, TP (mg P
m-3), on (a) the peak concentration of chlorophyll a c*,
and (b) the peak algal biomass, c*zm in Twin Lake. The
black circles represent field observations from
experiment 6. The TP levels of 18 and 51 represent
values for the control enclosure and lake, respectively.
-------
AERATION/MIXING AND AQUATIC PLANT HARVESTING
137
Lorenzen and Mitchell (1975) presented a similar
analysis for Kezar Lake, N. H. Although they used a
different model which did not consider the effects of
changes in nutrient concentrations, the results were
qualitatively the same. They also predicted that peak
algal biomass would reach a maximum at a particular
mixed depth, which they defined earlier (1973) as the
optimum mixed depth, zopt. However the value of zopt
which they predicted for Kezar Lake (about 1.5 meters)
was much lower than the value found for Twin Lake (15
meter). The lower value for Zopt in Kezar Lake means
that a much smaller increase in zm would be required to
reduce peak biomass. This difference in Zopt between
lakes is apparantly due to differences in the response
characteristics of the two communities involved.
The results of the simulation for Twin Lake suggest
that increases in TP which often occur during artificial
circulation can significantly affect community response
and, in some cases, can increase the concentration of
chlorophyll a. Total phosphorus concentrations were
found to increase in the circulated enclosures in all
eight of the experiments conducted on Twin Lake and
Little Lake Johannah. These increases in TP generally
resulted in higher concentration of chlorophyll a. In six
of the eight experiments, a significant positive
relationship was found between the average concen-
trations of TP and chlorophyll a determined in the
enclosures. The regression lines from these relation-
ships are shown in Figure 3. The regression lines for
different experiments in the same lake were generally
similar (experiments 3, 6, and 8 were in Twin Lake;
experiments 1, 2 and 5 were in Johannah). However,
the response for each lake was quite different.
Increases in TP generally resulted in much larger
increases in chlorophyll a in Twin Lake than in Little
Lake Johannah.
This difference in community response is apparently
related to differences in the relative availability of
CJ
E
O>
75
50
O
25
3(46]
50
100
150
-3
TP mg-m
Figure 3.-The relationships found between
average concentrations of total phosphorus
and chlorophyll a in experimental enclosures
for circulation experiments in Little Lake
Johannah (1,2 and 5) and in Twin Lake (3, 6
and 8). Numbers in parentheses indicate the
ratios of IN/IP determined initially in .each
experiment.
nitrogen and phosphorus in the two lakes. The ratio of
inorganic nitrogen to inorganic phosphorus, IN/IP,
determined initially in each experiment is indicated on
Figure 3. The IN/IP ratios were always greater than 30
in Twin Lake and less than 10 in Johannah. Forsberg,
et al. (1978) surveyed a large number of lakes and
found that above an IN/IP ratio of 12 most
phytoplankton were P-limited, below a value of 5 they
were generally N-limited, and between 5 and 12 either
nutrient could limit algal growth. It is clear, then, that
the phytoplankton in Twin Lake were limited by
phosphorus while the low IN/IP and weaker com-
munity response found in Johannah suggest that the
phytoplankton there were probably limited by nitrogen.
These results indicate that, during circulation, the
phytoplankton will respond to changes in the concen-
tration of that nutrient which is in the shortest supply.
Changes in Species Composition during Artificial
Circulation. Changes in species composition during
artificial circulation were found to depend primarily on
the mixing rate. This effect was most apparent in the
open bottomed enclosures. When these enclosures
were mixed slowly surface levels of TP and pH
generally increased. These conditions often increased
the relative abundance of blue-green species such as
Anabaena circulinus and Microcystis aureginosis. At
the faster mixing rates, where complete chemical
destratification occurred, larger increases in TP and
nutrient availability were usually observed at the
surface. In addition, increases in the concentration of
CQz, which occurred as hypolimnetic water was
brought rapidly to the surface, generally resulted in
lowered pH levels. These conditions often increased
the relative abundance of green algae and diatoms. The
green algae: Sphaerocystis Schroederi, Ankistrodes-
mus falcatus, and Scenedesmus spp., grew particularly
well at these faster mixing rates as did the diatoms:
Nitszchia spp., Synedra spp. and Melosira spp.
There was some evidence that reduced sinking
losses may have given the diatoms an advantage at the
faster mixing rates. This was demonstrated in
experiments 1 and 2 where several enclosures were
mixed rapidly but without increasing the mixed depth
below the thermocline. This allowed us to separate the
direct effect of turbulence from other factors which
might change if the mixed depth were increased. The
growth rates of diatoms were found to increase to
much higher levels than those of green and blue-green
species as the level of turbulence increased. The
conditions of high nutrient availability and low pH
which prevailed at the faster mixing rates are similar to
those which Shapiro (1973) found to produce a shift
from blue-green to green species in his field
experiments. The fact that similar shifts to green
species also occurred in the rapidly mixed enclosures
suggests that a common mechanism might be involved.
Shapiro found that the growth of blue-green species
was suppressed at low pH. We also observed a decline
in the growth rates of blue-green algae as the pH
dropped in the rapidly mixed enclosures. While the
mechanism is not entirely clear, this disadvantage of
blue-greens at low pH may involve the activity of
cyanophage, Lindmark (1979) demonstrated dramatic
increases in the incidence of cyanophage infection as
-------
138
RESTORATION OF LAKES AND INLAND WATERS
the pH was lowered in laboratory cultures of blue-
green algae.
Alternatively, King (1970) has suggested that shifts
from blue-green to green species which he observed in
sewage lagoons at low pH might be due to differences
in carbon uptake kinetics between these two divisions.
Long (1979) provided considerable support for this
hypothesis. In a series of carbon growth and uptake
experiments for a large number of species he
demonsrated the general competitive superiority of
green algae over blue-greens at low pH. He also
showed how this competitive advantage shifts in favor
of the blue-green species at high pH levels. This latter
result may explain the dominance of blue-green algae
at high pH in the slow mixed enclosures.
The shift from blue-greens to greens was not always
observed in the rapidly mixed enclosures and appar-
ently depended on the magnitude of the pH drop during
mixing. It occurred most often in Johannah where
relatively low alkalinity and high levels of hypolimnetic
CCfe resulted in a large drop in pH during mixing. The
shift was seldom seen in Twin Lake where pH dropped
only slightly during mixing due to much higher levels of
alkalinity and lower hypolimnetic levels of 002.
SUMMARY
The results presented here suggest that the response
of a particular phytoplankton assemblage during
artificial circulation will depend on a number of factors.
The response at the community level is apparently a
complex function of many different lake and com-
munity characteristics which can only be described
within a theoretical framework which considers all of
these factors simultaneously. The model presented
here represents an improvement over earlier attempts
to provide such a framework.
Previous expressions did not consider the effects
which changes in total nutrient concentrations might
have on the phytoplankton. The results from the field
experiments (Fig. 3) indicate that these changes can
have a significant effect on the community response.
By considering the effects of nutrient and light
limitation simultaneously a single expression was
derived which could be used to predict the level of algal
biomass as a direct function of the mixed depth and
total phosphorus concentration. The simulation results
presented for Twin Lake demonstrated how both the
total amount and concentration of chlorophyll a in the
mixed layer could either increase, decrease, or remain
the same depending on the changes in TP and zm which
occur during artificial circulation. A different approach
may have to be taken in lakes such as Little Lake
Johannah where algal growth was apparently limited
by nitrogen instead of phosphorus.
However, it should be possible to develop a model,
similar to the one presented here, which would predict
the effects of changes in total nitrogen levels on the
community response. Shifts in species composition
during artificial circulation were found to depend
primarily on the mixing rate achieved. Blue-green
species were favored at the slowest mixing rates while
greens and diatoms were favored at the faster mixing
rates. While the mechanisms involved were not
entirely clear, the shift in competitive advantage from
blue-green species at the faster mixing rates was
apparently related to the low pH and high nutrient
levels achieved under these conditions. This shift
would therefore be most likely to occur in relatively
deep productive lakes with high hypolimnetic concen-
trations of CCh and nutrients where mixing will result
in a large drop in pH and increases in nutrient level at
the surface.
REFERENCES
Allott, N. 1978. Recent paleolimnology of Twin Lake near St.
Paul, Minn., based on a transect of cores. M.S. Thesis.
University of Minnesota.
Bella, D. A. 1970. Simulating the effect of sinking and vertical
mixing on algal dynamics. Jour. Water Pollut. Control Fed.
42:140.
Aernhardt, H. 1967. Moglichkeiten der verhinderung anae-
rober verhaltnisse in el ner trinkwassertalsperre wahrend-
dersommerstagnation. Arch. Hydrobiol. 63:4094.
Drury, D. D., D. B. Porcella, and R. A. Gearheart. 1975. The
effects of artificial destratification on the water quality and
microbial populations in Hyrum Reservoir, Utah. PRJEW
011-11. Water Res. Lab., Utah State University, Logan.
Forsberg, B. R., and J. Shapiro. 1980. The effects of artificial
circulation on algal populations. In Symp. Surface Water
Impoundments, June 2-5, Minneapolis, Minn.
Forsberg, C., et al. 1978. Water chemical analyses and/or
algal assay — sewage effluent and polluted lake water
studies. Mitt. Int. Verein. Limnol. 21:352.
Haynes, R. C. 1971. Artificial circulation in a small eutrophic
New Hampshire lake. Ph.D. Thesis. University of New
Hampshire.
Irwin, W. H., J. M. Symons, and G. G. Robeck. 1966.
Impoundment destratifcation by mechanical pumping. Jour.
San. Eng. Div. Proc. Am. Soc. Civil Eng. 92:21.
King, D. L. 1970. The role of carbon in eutrophication. Jour.
Water Pollut. Control Fed. 42:2035.
Knoppert, P. L. et al. 1970. Destratification experiments at
Rotterdam. Jour. Water Works Assoc. 62:448.
Lindmark, G. 1979. Interaction between Lpp-1 virus and
Plectonema boryanum. Ph.D. Thesis. University of Lund,
Sweden.
Long, E. B. 1979. The interaction of phytoplankton and the
bicarbonate system. Ph.D. Thesis. Kent State University.
Lorenzen, M. and R. Mitchell 1973. Theoretical effects of
artificial destratification on algal production in impound-
ments. Environ. Sci. Technol. 7:939.
1975. An evaluation of artificial destratification
for control of algal blooms. Jour. Am. Water Works Assoc.
67:372.
Malueg, K., et al. 1971. The effect of induced aeration upon
stratification and eutrophication processes in an Oregon
farm pond. Int. Symp. Manmade Lakes, Knoxville, Tenn.
Megard, R. 0., and P. D. Smith. 1974. Mechanisms that
regulate rates of growth of phyto-plankton in Shagawa
Lake, Minn. Limnol. Oceanogr. 19:279.
Murphy, G. I. 1962. Effect of mixing depth and turbidity on the
productivity of freshwater impoundments Trans Am Fish,
Soc. 91:69.
Oskam, G. 1978. Light and zooplankton as algae regulating
factors in eutrophic Biesbosch Reservoirs Verh Int Verein
Limnol. 20:1612.
-------
AERATION/MIXING AND AQUATIC PLANT HARVESTING 139
Pastorok, R. A., T. C. Ginn, and M. W. Lorenzen. 1980.
Evaluation of aeration/circulation as a lake restoration
technique. EPA Draft Final Rep. TC-3947. U.S. Environ.
Prot. Agency.
Senft, W. H. 1978. Dependence of light-saturated rates of
algal photosynthesis on intra-cellular concentrations of
phosphorus. Limnol. Oceanogr. 23:585.
Shapiro, J. 1973. Blue-green algae: why they become
dominant. Science 179:382.
Shapiro, J., et al. 1980. Report in preparation to the EPA on
Res. Contr. R 803870.
Tailing, J. F. 1957. The phytoplankton as a compound
photosynthetic system. New Phytol. 56:133.
Weiss, C. M., and B. W. Breedlove 1973. Water quality
changes in an impoundment as a consequence of artifical
destratification. Rep. 80. Water Resour. Res. Inst. University
of North Carolina.
-------
140
RESERVOIR MIXING TECHNIQUES:
RECENT EXPERIENCE IN THE UK
D. JOHNSON
J.M. DAVIS
Water Research Centre
Medmenham Laboratory
Marlow, Buckinghamshire
England
ABSTRACT
The United Kingdom has about 180 impounding and pumped storage reservoirs of sufficient size
and location which are likely to stratify thermally. Of these at least 30 have, or plan to install, a
destratification system. The two systems most commonly used are submerged jetted inlets for
pumped storage reservoirs and perforated-pipe air-mixing systems for impounding reservoirs.
When operating, these systems maintain temperature differences between surface and bottom
waters of 1 to 3 °C compared with differences of 8 to 9 °C under stratified conditions. The chemical
quality of the water is also maintained at a higher standard.
INTRODUCTION
The United Kingdom has about 400 raw-water
reservoirs having a capacity exceeding 20,000 m3.
Most of these are impounding reservoirs typical of
upland storage while the remainder are either bunded
pumped-storage systems, or pumped-storage im-
poundments, primarily situated in lowlands.
Approximately half the water consumed in the UK is
supplied from storage reservoirs with direct river
abstraction; ground water meets the remaining
demand. The quality of reservoir water depends to a
large extent on size, geographical location, the quality
of the water used for refill, and the extent of thermal
stratification.
Of the reservoirs in the UK, about 180 are likely to
stratify in the spring and summer. Frequently asso-
ciated with thermal stratification is hypolimnetic
deoxygenation, and the poor chemical quality of this
water severely restricts its use for water supply or river
regulation. Excessive amounts of iron, manganese, and
humic acids must be removed by water treatment and
this may not always be achieved easily. Ammonia may
interfere with sterilization by chlorine and possibly
increase treatment costs. Sulfides cause unpleasant
odors, demand much chlorine, and corrode iron and
concrete.
For river flow regulation purposes this water is also
undesirable. Most of the decomposition products
contained in the water will exert an oxygen demand on
the river when the water is used to augment the low
summer flows. At such times the oxygen is most
needed and the content is at its lowest. Sulfides and
ammonia in sufficient concentrations are toxic to fish,
especially in combination with low oxygen levels(Calif.,
State Water Qual. Control Board, 1 963; Herbert, 1961).
Precipitated iron oxide may coat macrophytes and
settle on stream beds, consequently disturbing the
stream ecosystem.
Two techniques have been used to mix reservoirs in
the UK — inlet jetting and air injection. Both these
methods achieve a high degree of mixing throughout
the reservoir depth and in doing so maintain
approximate isothermal conditions during the summer;
this prevents chemical deterioration of water quality.
Although artificial mixing is primarily intended as a
method of improving the chemical quality of the water
and making a greater volume of stored water available
for use, it has been suggested (Steel, 1972; Lorenzen
and Mitchell, 1975) that some control of algal
populations is also achieved.
MIXING SYSTEMS USED IN THE UK
Two main directions have been followed in the UK to
break down and prevent thermal stratification. The
technique used for pumped-storage reservoirs involves
pumping the incoming water through a nozzle situated
near the bed of the reservoir, Figure 1. Water is
entrained from the lower layers and this results in
vertical and horizontal mixing which not only maintains
the reservoir in an approximately isothermal state but
ensures continual re-aeration at the surface. A
diagrammatic representation of the induced circulation
pattern is shown in Figure 2.
A technique used in impounding reservoirs is air
injection, using either a confined or unconfined device.
Confined-air injection consists of releasing com-
pressed air through a series of vertical, free-standing
polyethylene tubes positioned on the bed of the
reservoir. The tubes are usually 2 to 3 meters long with
a diameter of 0.3 to 0.4 meters. Two confined systems
which have been employed over the last decade are the
Aero-Hydraulics gun (Bryan, 1964) and the Helixor
-------
AERATION/MIXING AND AQUATIC PLANT HARVESTING
141
Figure 1. — A normal and a jetted-inlet system in a pumped
storage reservoir.
Surface
Figure 2. — A diagramatic representation of the induced
circulation resulting from inlet jet mixing.
Figure 3. — A perforated-pipe system anchored by concrete
blocks at the foot of an impounding dam.
Figure 4. — Circulation of water induced by rising air bubbles
from a perforated pipe.
(made by Polcon Corp., Montreal). In the former system
a . 'eady flow of large single air bubbles, which act like
expendable pistons, force water through the tube,
inducing circulation. The Helixor system, originally
designed for wastewater treatment, is similar to the
Aero-Hydraulics gun, but rather than a large single air
bubble, the device relies on many smaller bubbles.
These follow a spiral path within the tube itself
entraining bottom water and inducing the necessary
circulation.
Unconfined air injection has been used more
commonly. In this system compressed air is pumped
through a perforated pipe or a diffuser dome anchored
near the bed of the reservoir. A perforated-pipe system
usually has 100 to 200 meters of pipe with 0.8 mm
diameter holes at 0.3 meter centers. An anchored pipe
section is shown in Figure 3. Clean, oil-free air is
supplied to the pipe by a compressor which may be
sited some distance away. A diagrammatic representa-
tion of the circulation pattern induced by the rising
bubbles is shown in Figure 4. Oxygen demands are met
primarily through surface re-aeration.
A list of the reservoirs in which these systems have
been installed, or are going to be installed, is shown in
Table 1.
SUGGESTED DESIGN PROCEDURE FOR
MIXING SYSTEMS
When considering the thermal behavior of water
bodies, the year in the southern part of the UK, can be
considered to be made up of four quarters. These are
defined in Table 2 by the thermal effect and range of
water temperatures associated with them. Clearly the
critical time is the constant heating quarter, and any
system designed to prevent stratification from becom-
ing established must be designed to overcome the
effects of this period.
Jetted-lnlet System
The objectives of a jetted inlet are:
1. To entrain water from the lower layers carrying it
to, or near to, the surface where re-aeration takes
place.
2. By their direction and momentum to circulate the
general body of the water, giving an overall mixing
effect.
The characteristics of a jet system may be identified
by the densimetric Froude number (F) defined as
U
F =
eq. 1
where,
U =mean jet velocity (ms~1)
g = acceleration due to gravity (ms~2)
D = diameter of jet nozzle (m)
Ap = absolute density difference between pumped and
ambient water (kgm~3)
p — density of pumped water (kgrrf3)
As F tends to infinity, plume buoyancy is negligible
and momentum dominates, while as F tends to zero
momentum is negligible and the plume rises almost
-------
142
RESTORATION OF LAKES AND INLAND WATERS
Reservoir
Kielder
Rutland Water
Datchet
Wraysbury
Carsington
Bewl Bridge
Broad Oak
King George VI
Wimbleball
Queen Elizabeth II
Hollowell
Blithfield
Loch Turret
Farmoor II
Bough Beech
Blagdon
Staunton Harold
Clatworthy
Pitsford
Farmoor I
Upper Glendeven
Lower Glendevon
Castlehill
Cropston
Ardleigh
Ravensthorpe
Wistland Pound
Hawkridge
Lower Lliw
Grimsbury
Court Farm
Table 1. — Reservoirs
Volume 106m3
204
124
38
35
35
31
24
20
20
20
18
18
18
9
9
8
7
5
5
5
5
4
3
3
2
2
2
1
1
0.3
0.3
in the UK having artificial mixing
Method of Filling
Impounding
Pumped & Impounding
Pumped
Pumped
Impounding
Pumped & Impounding
Pumped & Impounding
Pumped
Impounding
Pumped
Impounding
Impounding
Impounding
Pumped
Pumped & Impounding
Impounding
Pumped & Impounding
Impounding
Impounding
Pumped
Impounding
Impounding
Impounding
Impounding
Pumped & Impounding
Impounding
Impounding
Impounding
Impounding
Impounding
Pumped
systems.
Mixing System
Perforated pipe*
Helixor
Inlet & recirculation jets
Inlet & recirculation jets
Perforated pipe*
Perforated pipe
Perforated pipe*
Diffuser blocks
Perforated pipe
Inlet & recirculation jets
Perforated pipe
Perforated pipe
Air gun
Inlet & recirculation jets
Recirculation jet
Perforated pipe
Perforated pipe
Perforated pipe
Perforated pipe
Inlet & recirculation jets
Air gun
Air gun
Helixor
Perforated pipe
Helixor & perforated pipe
Perforated pipe
Perforated pipe
Perforated pipe
Perforated pipe
Perforated pipe
Perforated pipe
'Planned installation
immediately after leaving the nozzle.
Steel (1976) has reviewed the literature related to
inlet jets and their characteristics. For a given nozzle
diameter approximate relationships have been derived
between jet trajectory, orientation, and densimetric
Froude number.
spring and summer. For design purposes in temperate
climates the proportion of solar radiation which
appears as heat energy may be taken as approximately
5 J m d In addition, the efficiency of energy
transmission (17) associated with a jet system is about 2
to 5 percent which means that the rate of energy (E)
required at the jet may be estimated by
Z = X sec0 ) sin 0 +
0.048 X
-( sec0r J. ; 0<0<45U
eq. 2
Where: Z is the height of the jet trajectory above the
jet orifice (m)
X is the related distance from the jet orifice (m)
6 is the orientation of the jet to the horizontal
To meet the objectives stated earlier it is necessary to
maximize trajectory length, thereby maximizing the
total volume of water entrained, while ensuring that
the entrained water reaches the surface layers. The
choice of jet orientation and nozzle diameter will be a
compromise to suit the range of operating conditions.
This involves taking into account different inlet
pumping rates and density differences between the
inlet water and that in the reservoir. Indeed it may well
be prudent, costs permitting, to have a choice of jet
inlets of different diameter and orientation. One
approach to designing such a system would be to
determine the rate of energy required at the jet to
maintain approximate isothermal conditions during
5 x surface area
E = (J
86,400 x n
eq. 3
This energy rate may be related to mean nozzle
velocity (U) for a range of inlet pumping rates(Qi, mV1)
through,
= 0.5pQiU2
(J s"1)
eq. 4
where p is the density of the incoming water.
The mean nozzle velocity can then be determined
from equation 4 as,
,= / 2-l\
V'QI/
U= I — 1 (m
The diameter of the nozzle follows from
eq. 5
(m)
eq. 6
-------
AERATION/MIXING AND AQUATIC PLANT HARVESTING
143
and the Froude number is calculated from equation 1.
The orientation of the jet may now be chosen using
equation 2 so that for a range of operating conditions
the design criteria are met.
Air Injection: Perforated-Pipe System
There are two main requirements of a perforated
pipe destratification device.
1. After the onset of thermal stratification in spring or
early summer the device must be capable of mixing the
greater part of the reservoir volume in a reasonably
short time (say 5 to 10 days depending on reservoir
volume) so that approximate isothermal conditions
exist over the depth.
2. During operation of the device the oxygen demands
within the water column should be met initially
through mixing between the upper and lower water
layers and in the longer term through surface re-
aeration.
A design procedure for this device (Davis, 1980)
relies on assuming a design temperature profile
corresponding to conditions in spring or early summer.
Typically in the UK the design temperature profile will
correspond to a 4°C temperature difference between
the epilimnion and hypolimnion, with the temperature
of the hypolimnion being around 8°C. From this the
stability of the reservoir (the energy required to
completely mix a stratified reservoir) is calculated. An
estimate of the total energy required (E) to destratify
the reservoir is obtained by adding to the stability value
the solar heat energy input (approximately 5Jm"2d"1)
during the destratification period.
An estimate of the free air required at the
compressor can be obtained from the following
relationship.
Energy input by the perforated pipe
to destratify reservoir
Total theoretical energy required (E)
= 20
eq.7
The numerator in equation 7 is obtained by assuming
isothermal conditions and a bubble pressure just
sufficient to overcome the hydrostatic head. This is a
function of the free air supplied by the compressor.
After re-arranging terms, equation 7 becomes
Q =
0.196 E
Tin ( 1 +H/10.4)
(I s'1)
eq. 8
Where: Q is the free air delivered
T is the time for destratification (s)
H is the depth of water above the pipe (m)
To calculate the length of perforated pipe required (L)
the following relationship is used:
Volume of water entrained by
air bubbles to destratify reservoir
volume of reservoir (V)
= 2.5
eq. 9
The volume of water entrained for a given free air
flow has been investigated by Bulson (1961); his
empirical relationship is used.
Equation (9) then becomes
L = 3.73
V3[ 1 +H/10.4]
(m)
eq. 10
The air pressure required at the compressor can now
be calculated taking into account hydrostatic head and
friction losses in the pipe work.
CASE STUDIES
The results from three different reservoirs employing
inlet jetting, jetted recirculation, and perforated pipe
air-mixing are presented. The morphometry of these
reservoirs is given in Table 3.
Reservoir 1: Jetted-lnlet Systems
Water is pumped into the reservoir from the adjacent
river to balance treatment plant demand and maintain
the reservoir at top water level. River water is pumped
in through a low-velocity 0.76 m diameter pipe or a
high-velocity jetted inlet (0.3 m or 0.38 m diameter)
inclined at 22.5°. Cost of jetting operation is
approximately 2 percent over low velocity pumping
cost.
3
2 -
1 •
°C
S
4
3
2
1
°C
la) 1970
°C
March April
May Juna July Auguit Saptember Octobar
Figure 5. — Temperature differences between surface and
bottom water (10 m) when (a) unjetted; (b) jetted after
stratification was established; (c) jetted throughout the
summer.
-------
144
RESTORATION OF LAKES AND INLAND WATERS
When the reservoir was first constructed the inlet
pipes were not jetted and the water velocity at the inlet
was not sufficient to prevent thermal stratification.
Temperature differences of 8.5°C between surface and
bottom water were observed (Figure 5a). One inlet pipe
was then modified by fitting a reduced diameter nozzle
inclined, at 22.5° to the horizontal. In 1973 the
reservoir was allowed to stratify. The inclined jet was
then brought into use and the temperature differential
decreased from 7.5 to <3°C (Figure 5b).
In subsequent years the system has been operated to
prevent, as far as possible, any stratification taking
place. Although temperature differences of up to 3°C
(Figure 5c) have occurred, the bottom waters have not
become anoxic, a minimum value of 50 percent of
saturation being recorded.
Reservoir 2: Jetted Recirculation
Ninety percent of the stored volume of this reservoir
is pumped from the river source between September
and April; the only water entering the reservoir during
the remainder of the year flows in from a natural feeder
stream. Water is taken from the bottom of a draw-off
tower near the dam wall and returned to the reservoir
about 500 meters up the valley through a 0.46 m
diameter jet inclined at 8° to the horizontal. A pumping
rate of 0.53 m3 s~1 gives a mean nozzle velocity of
3.2 ms~1. Pumping costs are currently estimated at £11
per day.
Filling of the reservoir began in 1969 to a depth of 12
meters. Summer stratification resulted in low dissolved
oxygen levels with increases in dissolved iron,
manganese, and ammonia. The following year the
reservoir was at maximum depth by April. Summer
stratification again resulted in low dissolved oxygen
levels with increases in dissolved silica, phosphate,
ammonia, and manganese, but not iron. The recircula-
tion pumps were run during June and July for short
periods and this smoothed the thermal profile although
chemical stratification persisted. Natural overturn took
place in October.
In subsequent years the reservoir stratified and as a
result high levels of dissolved manganese occurred in
the hypolimnion. The recirculation pumps were used
intermittently to lessen the degree of stratification and
Table 2. — Thermal quarters of water bodies in the southern UK
(Steel, 1976).
Quarter Thermal effect
Months
Water temperature
1
2
3
4
Constant cold
Constant heating
Constant warm
Constant cooling
Jan. - March
Apr. - June
July - Sept
Oct. - Dec.
~4°C
— H5°C month"1
~ 20°C
- -5°C month"1
decrease the levels of dissolved manganese. A typical
pattern of events is shown in Figure 6, indicating the
rapid improvement in water quality following operation
of the recirculation system.
Reservoir 3: Perforated-Pipe System
The water supply to this reservoir is entirely from
natural feeder streams. This reservoir has had a history
of thermal stratification and hypolimnetic deoxygena-
Recirculation
Jet in use
mg.r
April
May
July August September October
Figure 6. — Dissolved manganese in the bottom water (21 m)
and temperature differences between the surface and bottom
for a jetted-recirculation system.
Total manganese
14 18 22 26 30 4 8 12 16 20 24
Figure 7. — Dissolved-oxygen and total-manganese levels in
the bottom water (21 m) during operation of a perforated-pipe
system.
Table 3. — Details of the reservoirs used in the case studies.
Reservoir
1
2
3
Filling
method
Pumped storage
Pumped and
Impounding
Impounding
Mixing
system
Inlet jet
Recirculation
Jet
Perforated pipe
Volume
106m3
4.5
8.9
1.6
Area
ha
50.6
115.0
16.6
Depth
max m.
11.0
22.9
23.1
Depth
mean m.
8.9
7.7
9.4
-------
AERATION/MIXING AND AQUATIC PLANT HARVESTING
145
tion giving rise to re-solution of iron and manganese
from the bottom sediments. Total manganese con-
centrations of 3 mg T1 had been common and had on
occasions reached 7 mg f1 causing treatment
problems at times of high demand. On some occasions
the problems were exacerbated by high algal popula-
tions in the epilimnion.
A polyethylene pipe, of which an 85 m length was
perforated by 0.8 mm dia. holes at 0.3 m centers, was
installed in the deepest part of the reservoir near the'
dam wall. The pipe was attached to a compressor
delivering57mg I"1 of free air compressed to 3 bar. The
compressor was turned on when the reservoir was
already stratified with a column temperature difference
of about 7°C. The result of this and the changes in
water quality prior to the operation is shown in Figure
6. The total manganese concentration of 2.3 mg I"1
was rapidly reduced to 0.2 mg f1 and at the same time
the dissolved oxygen content at the bottom of the
reservoir reached saturation with a column tempera-
ture difference of <1 °C.
When the compressor was turned off, however,
thermal stratification re-established and a fall in
dissolved oxygen resulted in re-solution of manganese.
Two periods of operation were sufficient to improve the
water quality. Following this the compressor was used
for about 8 hours each day until early autumn to
maintain the water quality at a treatable level. The
initial high concentration of algae in the surface water
was distributed by mixing over the entire depth and this
dilution also reduced water treatment problems. Fuel
charges for operation of the compressor were
approximately £28 per day.
CONCLUDING REMARKS
Inlet jetting and air injection, the latter by means of a
perforated pipe system, have been the main methods of
artificial mixing in the UK to date. Inlet jetting is ideally
suited to pumped-storage schemes but provision for
this must be made at the reservoir construction stage.
Perforated pipe devices have been used almost
exclusively in impounding reservoirs and although they
are preferably installed during construction of the dam
they may also be installed easily, as and when
required, following impoundment.
Experience with both destratification devices has
shown that thermal stratification can be prevented
from becoming established during the summer and the
water quality problems associated with stratification
can be controlled.
Lorenzen, M. W., and R. Mitchell. 1975. An evaluation of
artificial destratification for control of algal blooms. Jour.
Am. Water Works Assoc. 67:373.
Steel, J. A. 1972. The application of fundamental limnological
research in water supply system design and management.
Symp. Zool. Soc. Lond. 29:41.
1976. The management of Thames Valley
reservoirs. Page 371 -419 in Proc. Symp. Effects of Storage
on Water Quality. Reading, 1975. Water Res. Centre,
Medmenham.
REFERENCES
Bryan, J. G. 1964. Physical control of water quality. Br. Water
Works Assoc. Jour. 46:546.
Bulson, P. S. 1961. Currents produced by an air current in
deep water. Dock Harbour Authority Jour. 42:15.
California State Water Quality Control Board. 1963. Water
quality criteria. 2nd ed. Sacramento.
Davis, J. M. 1980. Destratification of reservoirs — a design
approach for perforated-pipe compressed air systems.
Water Serv. August.
Herbert, D. W. 1961. Freshwater fisheries and pollution
control. Proc. Soc. Water Treat. Exam. 10:135.
-------
146
CASE STUDIES OF AQUATIC PLANT MANAGEMENT FOR
LAKE PRESERVATION AND RESTORATION IN BRITISH
COLUMBIA, CANADA
PETER R. NEWROTH
Ministry of Environment
Victoria, British Columbia
ABSTRACT
A wide variety of aquatic plant management technologies have been tested and evaluated as part
of the British Columbia Aquatic Plant Management Program. These detailed studies are valuable
to guide planning and long range management of aquatic macrophytes in diverse aquatic
environments. The recent introduction of Eurasian water milfoil (Myriophyllum spicatum L.) to
British Columbia has caused concern about environmental quality and has interfered with multiple
water resource uses in a number of water bodies. Case studies are presented in this paper to
illustrate the application of aquatic plant management technologies in preservative or restorative
strategies. The technologies are described and in each case study the history and management
approaches are summarized. The limitations.costs, and benefits of the strategies have been highly
variable but this variability reflects the political, social, and technical complexities of aquatic plant
management.
INTRODUCTION
Nuisance aquatic vegetation can threaten or inter-
fere with multiple use of water resources. In British
Columbia, there were few reports of problems with
aquatic macrophytes until recent years, although some
species such as Elodea canadensis Rich, in Michx.,
Potamogeton crispus L, and Ceratophyllum demersum
L. have been recorded as nuisances for some time.
However, the introduction and spread of Eurasian
water milfoil (Myriophyllum spicatum L.) in British
Columbia waters during the past decade has created a
demonstrable environmental problem which has
warranted a substantial management effort. M.
spicatum has become a widespread management
problem in the eastern United States.
While control projects have been necessary in many
areas where this plant is now established, the efforts of
the B.C. Ministry of Environment are unique because of
the diversity, complexity, and comprehensiveness of
the management approaches and of the attempts to
document the results. A historical perspective of major
elements of this aquatic plant management program
was presented elsewhere (Newroth, 1979).
The timeliness and suitability of applying technology
to specific problems is particularly important in
successfully allocating resources for maximum man-
agement benefits. Case studies are presented here to
exemplify management strategies; technologies and
approaches described reflect the changing philosoph-
ies and policies developed since 1972.
DESCRIPTION OF TECHNOLOGIES
The B.C. Ministry of Environment has extensively
reviewed or field tested a wide range of technologies.
Also, the Province has encouraged private enterprise in
technological development as well as in original
research and development in the following categories:
1. Prevention Since several major regions of the
Province are not infested with M. spicatum (and
experience has demonstrated the difficulty of eradica-
ting established populations), several preventive ap-
proaches have been developed. Because boaters are
suspected of spreading aquatic weeds, quarantine
projects have been used to encourage voluntary public
cooperation; boats leaving infested areas are checked
(Scales and Bryan, 1979; Dove and Malcolm, 1980).
Also, surveillance for M. spicatum has been performed
at selected noninfested lakes, with emphasis on
marina and boat launch areas of those with high public
recreational value. Data gathered from the quarantine
projects have aided in selecting lakes known to be
frequented by boaters who have visited infested water
bodies.
2. Intensive control While eradication of Eurasian
water milfoil appears to be a remote possibility, the
timely application of a variety of technologies may
provide a cost-effective means to contain nuisance
aquatic plants. Depending on the situation, tech-
nologies may be used independently or in an integrated
fashion. Detailed descriptions of technologies suitable
for intensive control are provided in reports prepared by
the Ministry of Environment (Anon., 1978; Bryan, 1980;
Maxnuk, 1979; Goddard, 1980). Brief descriptions and
major limitations of these techniques are outlined here:
(a)2,4-D Butoxyethanol Ester. Laboratory and
extensive field testing in the Okanagan Valley lakes
since 1976 has indicated the utility and environmental
safety of selectively using this herbicide in a granular
formulation (Aqua-Kleen 20) (Goddard, 1980). 2,4-D
has proven ability to kill most roots and shoots of
Eurasian water milfoil in treated areas, and it can be
-------
AERATION/MIXING AND AQUATIC PLANT HARVESTING
147
applied rapidly to large areas without inducing plant
fragmentation. However, this herbicide cannot be used
in localities where the public may become exposed to
the herbicide at detectable concentrations following
treatment. Also, the most effective applications have
coincided with quiescent water conditions. Herbicides
can be integrated successfully with other technologies
including mechanical harvesting, diver dredging, and
rotovating. Although at present the Ministry of
Environment only uses 2,4-D for intensive control, in
some situations it may be practical for more cosmetic
management.
(b) Diver-operated dredges: The use of scuba divers
and hand suction dredges has been refined since early
trials in 1976. This method is slow, relatively costly, but
highly effective in reducing the density of M. spicatum
infestations (Maxnuk, 1979). The localized activity of
divers discourages dissemination of viable plant
fragments. The best application of this approach is in
small areas, newly populated by Eurasian water milfoil,
where visibility is good and non-target vegetation is
sparse (ideally in spring and early summer).
J. Semi-cosmetic or cosmetic control: In water
bodies where M. spicatum has become widely
distributed or where treated areas are subject to rapid
reinfestation from major sources of viable plant
fragments, the most practical approach is to minimize
the nuisance cosmetically and cheaply. Three types of
technology have been developed and used by the
Ministry of Environment:
(a)Shallow water tillage: Large, shallow, littoral
areas may be treated using amphibious vehicles
drawing agricultural implements (including plows and
discs), to stress and remove plant roots and to lessen
the density of nuisance growth. This is most effective
when the area is subject to winter drawdown, but
obstacles and water pipelines may prevent efficient
operations.
(b)Deep water tillage: Barge mounted rotovators
have been extensively tested in efforts to achieve root
removal in water up to 4 m deep (Bryan, 1980).
Although this method is relatively slow and costly (as
compared with harvesting) and is limited to areas
where obstacles or rocks do not interfere with
operation, the main nuisance growth may be reduced
for several seasons. In certain cases rotovating may be
successfully integrated with more long-term manage-
ment technologies and root removal may be achieved
in the winter months.
(c)Plant harvesting: Mechanical harvesting pro-
vides cosmetic management by removing the tops of
nuisance plants. Four large machines are now
employed to harvest M. spicatum in the Okanagan
Valley and a similar machine is used in southern
Vancouver Island for control of Ceratophyllum, Elodea,
and other macrophytes. The main limitations of this
technology include the relatively slow speed (as
compared with 2,4-D), the need to time harvesting to
coincide with nuisance growth, and the repetitive,
seasonal nature of the operation. Although future
research may demonstrate that frequent or repetitive
harvesting may contribute to declines in vigor of M.
spicatum populations, in some situations the use of
harvesters may further spread this plant. Where
massive infestations prevail, the harvesting approach
is a practical management tool although high capital
costs may limit the number of machines.
4. Passive control— fragment barriers: A variety of
containment devices have been developed for deploy-
ment around equipment or to reduce downstream
spread of buoyant, viable fragments of M. spicatum.
Portable floating booms have been used around
rotovating and harvesting areas, where warranted, and
appear to reduce the escapement of much floating
plant material. Stationary fragment barriers have been
used in river channels and interconnections between
lakes since 1976 and have been developed to the
degree that they are generally successful in reducing
fragment movement (Stephenson and Baillie, 1980).
Practical limitations often restrict the successful
application of barriers; they must be maintained
routinely, using suction pumps to clear the fragments
trapped on the mesh.
CASE STUDIES
As part of the Ministry of Environment aquatic plant
management program, over 900 water bodies in British
Columbia have been inspected and records made on
aquatic macrophytes observed. About 30 lakes with
high recreational value are under continuing study.
Aquatic plant management technologies have been
applied to about 10 additional lakes and case studies of
four of these lakes are outlined here to illustrate the
approaches, expectations, and results which have been
achieved. Figure 1 shows the location of British
Columbia lakes now infested with Eurasian water
milfoil: Wood, Kalamalka, and Okanagan Lakes are
directly interconnected and are located in the Okana-
gan Valley of south-central British Columbia; Cultus
Lake is situated south of the Fraser River in the
southwestern part of the Province.
Okanagan Lake
Okanagan Lake is the largest lake in the mainstem
lake chain (area 34,800 hectares) and is described as
oligotrophic (Stockner and Northcote, 1974). Nuisance
growths of Myriophyllum in the northern part of this
lake were the first reported in British Columbia. The
present management program has developed in
response to requests for assistance from local
agencies. As part of the pilot studies to document
aquatic vegetation in Wood, Kalamalka, Okanagan,
Skaha, Vaseux, and Osoyoos Lakes (see Figure 1),
voucher collections from all these lakes were made in
1972. Although Myriophyllum specimens were collect-
ed in all these lake's, subsequent analyses of these
early collections by chromatography (Ceska, 1977)
have verified Myriophyllum spicatum only among
specimens collected in Okanagan Lake. It is possible
that the initial infestation of this plant was in Okanagan
Lake and that subsequent downstream spread to other
lakes and transport by boaters has led to the
infestations recorded later in other areas. Accurate
definition of the extent of M. spicatum prior to 1972 is
impossible, although study of aerial photographs taken
prior to the development of the present conspicuous
colonies indicated that major expansion occurred in
northern Okanagan Lake in the early 1970's.
-------
148
RESTORATION OF LAKES AND INLAND WATERS
FIGURE I.
MAP OF BRITISH COLUMBIA SHOWING
EURASIAN WATER MILFOIL INFESTED
AREAS
\
*i~ \
Al ,.-i>
ALL LAKES AFFECTED
BY Mynophyllum spicotum
Figure 1. — Map of British Columbia showing Eurasian water milfoil-infested areas.
Although localized surveys of aquatic plant popula-
tions were performed in 1972-74, a more com-
prehensive survey was performed in 1975 (Nijman,
1976). This documentation has been expanded each
year since 1975; Table 1 indicates the changes in area
occupied by M. spicatum in the lakes discussed here.
In retrospect, it appears that early containment and
immediate intensive control efforts should have been
considered seriously for Okanagan Lake. However,
limited resources were available in 1972 and the major
nuisances caused by the rapid spread and growth of M.
spicatum were not publicly recognized until 1973 and
Table 1. — Recent records of M. spicatum in case study lakes.
Lake
Wood
Kalamalka
Okanagan
Cultus
Approximate
Littoral
Area (ha)
85
145
1,930
37
Area affected
1975 1976
<1 <1
< 3
233 288
? 7
by M.
1977
3
11
358
13
spicatum (ha)
1978
12
12
398
17
1979
17
4
403
16
1974; in many ways, the nuisance potential of M.
spicatum was underestimated. Despite the experiences
of agencies in other areas (especially Florida and the
Tennessee Valley) with this plant, funding necessary to
begin effective work on the growing problem was not
forthcoming for several years.
Recent applications of aquatic plant management
technologies in Okanagan Lake are summarized in
Table 2. Because the potential negative impact of
Eurasian water milfoil was clearly apparent in 1974,
resources were made available in 1975 and 1976 for
extensive testing and monitoring of:
1. Contact herbicides, diquat and paraquat in 1974 and
1975 (Bryan, et al. 1977) and the systemic herbicide
2,4-D (Lim and Lozoway, 1977).
2. Bottom barriers (Armour, et al. 1979).
3. Mud Cat hydraulic dredge (Bryan, 1978).
4. Three types of rotovating machines and a hydraulic
washing device (Maxnuk, 1979; Bryan, 1980).
Most of these treatments were performed in popular
recreational areas and somewhat relieved nuisance
conditions, but as shown in Table 1, this activity
coincided with a rapid expansion (1975-1977) of
Eurasian water milfoil in Okanagan Lake. Because of
uncertainty, lack of experience as to the most
appropriate technology, and lack of funding, no
-------
AERATION/MIXING AND AQUATIC PLANT HARVESTING
149
concerted effort was planned until later. Thorough
evaluations of the trials showed some benefits from the
treatments but reinfestation from outside areas and
regrowth of M. spicatum roots remaining in treated
areas indicated that the large scale use of 2,4-D was
the only practical technology that might substantially
reduce Eurasian water milfoil in Okanagan Lake. Also,
containment and overall control in Okanagan Lake
became a lower priority because of the rapid growth of
M. spicatum in Kalamalka and Wood Lakes during
1975-1977.
In 1978 nearly 400 hectares of shoreline in
Okanagan Lake were colonized by Eurasian water
milfoil and many public and private recreational areas
were affected to some degree. A major control program
was planned for 1978 and designed to achieve a major
reduction of the primary nuisance colonies in Okana-
gan Lake; however, the Herbicide Use Permits
necessary for large scale use of 2,4-D were not
received. As shown in Table 3, major efforts were made
using harvesters and other machines but only about 12
hectares of experimental 2,4-D treatments could be
performed. Reluctantly, it was recognized in 1978 that
no practical options for overall control of Eurasian
water milfoil in Okanagan Lake were available. The
1979 and 1980 programs have been designed to
relieve nuisances at relatively low costs using
harvesters only. Basic harvesting costs (excluding
administrative and capital costs) averaged between
$1,500 and $1,900 per hecatare in 1978 and about
$1,200 per hectare in 1979. Because efforts have
continued to improve these machines (reduce down-
time and increase overall harvesting efficiency) it is
hoped that the 1980 costs will be slightly lower than
those documented in 1979. Because of the ongoing
need for harvesting, local authorities have been
encouraged to share costs of cosmetic harvester
operation, with 75 percent of the operating costs borne
by the Province.
Kalamalka and Wood Lakes
The Kalamalka-Wood Lakes Basin in the Okanagan
region has been the subject of intensive management
study; both lakes (see Figure 1) are considered of high
value for recreation and water supply purposes (Anon.
1974). Both lakes support populations of Eurasian
water milfoil although they are completely different in
character (Wood Lake, area 930 hectares, eutrophic:
Kalamalka Lake, area 2,590 hectares, oligotrophic)
(Stockner and Northcote, 1974). Preserving the
aesthetic beauty of Kalamalka Lake is considered a very
high priority by government and residents of the area.
Because dense aquatic plant growth would impair the
attractive white marl littoral areas of Kalamalka Lake
and interfere with beach use and boating, aquatic
macrophyte control has been recognized as an
important objective (Anon. 1974). Wood Lake dis-
charges through a short canal to Kalamalka Lake
downstream so successful management requires
simultaneous attention to both lakes.
Independent of the aquatic plant management
program, water quality studies to preserve Kalamalka
Lake and improve Wood Lake have been sponsored by
the Province and Federal agencies. These studies have
Table 2. — Summary of aquatic plant management
technologies applied to Okanagan Lake.
Year of application and approximate
area treated (ha)
Technology
Harvesting
Rotovating
Dredges
a) Mud Cat
b) Diver-operated
Herbicides
1975
15
Nil
2.9
2.9
Nil
1976
Nil
551
Nil
Nil
Nil
1977
45
42
Nil
Nil
Nil
1978
55
4
Nil
Nil
1.0
1979
47
Nil
Nil
Nil
Nil
a) Paraquat/Diquat 1.6 Nil
Nil
Nil
Nil
b) 2,4-D (B.E.E.) Nil 1.2 7.0 12.0 Nil
'Area treated by tractor drawn rotovator (5 ha), amphibious rotovator
(24 ha) and barge mounted rotovator (26 ha).
Barge mounted rotovator only from 1977 on.
Table 3. — Summary of aquatic plant management
technologies applied to Kalamalka and Wood Lakes.
Year of application and
approximate area
treated (ha)
Lake
Wood
Kalamalka
Technology
Diver-operated
Dredge
2,4-D (B.E.E.)
Rotovating
Diver-operated
Dredge
2,4-D (B.E.E.)
1976
Nil
Nil
0.5
Nil
Nil
1977
Nil
Nil
10
5
Nil
1978
11
Nil
10
24
<1.0
1979
5
7
Nil
28
13
revealed a trend of increasing water transparency over
the past 3 or 4 years (Nordin, 1980). This phenomenon
may be linked to the dramatic increase of area affected
by M. spicatum in Wood Lake between 1976 and 1979
(see Table 1).
Although minor nuisances caused by growth of
Potamogeton crispus had been reported as early as
1972, surveys for aquatic plants did not locate M.
spicatum until 1974. The first collections (1974) of M.
spicatum in Kalamalka Lake were confirmed several
years later by chromatography (Ceska, 1977). Because
of the close morphological similarity of this species to
M. exalbescens Fern., positive identification was first
confirmed in 1975 when small populations of
characteristic vigorous growth were located at the
north end of Kalamalka Lake and the south end of
Wood Lake (Nijman, 1976). Both populations had
developed adjacent to boat launching and marina
facilities.
Prior to the observation of established colonies of
Eurasian water-milfoil in these lakes, posting of signs
to discourage the introduction of fragments and
immediate efforts to eradicate M. spicatum were
recommended (Newroth, 1975). These recommenda-
tions coincided with extensive technological testing
which began in 1975; pilot scale rotovating was
performed in the fall, 1976, in the most dense M.
spicatum colony in Kalamalka Lake. Table 3 sum-
marizes the applications of management technologies
in Wood and Kalamalka Lakes. As shown in Table 1,
major expansion of Eurasian water milfoil was
documented in Kalamalka Lake between 1975 and
1976, but no major change in M. spicatum growth was
seen in Wood Lake during this interval.
-------
150
RESTORATION OF LAKES AND INLAND WATERS
Details of the management efforts in Kalamalka Lake
are presented elsewhere (Bryan, 1976; Maxnuk, 1979;
Armour, et al. 1980): the objectives, efforts and results
achieved in this lake are illustrative and are reviewed
briefly here. Following the trials with a prototype
rotovator in Okanagan Lake in 1976, about 4 hectares
of moderately infested area of Kalamalka Lake were
treated in early 1977. A new, improved rotovator
treated about 6 hectares of Kalamalka Lake in the fall,
1977. Also, several versions of diver-operated dredges
were evaluated and used to treat about 5 hectares of
littoral area in this lake during the spring and fall, 1977.
These efforts and subsequent treatments of large areas
of Kalamalka Lake by dredging (24 hectares) and
rotovating (10 hectares) in 1978 were intended to
achieve the maximum possible degree of control of
Eurasian water milfoil. However, reinfestation of
previously rotovated areas was observed despite short
term effectiveness of up to 90 percent root removal.
The diver dredges concentrated in 1977 on removing
roots and shoots missed by rotovating, and achieved up
to 97 percent effectiveness (Maxnuk, 1979). In 1978,
larger diver dredges were made available and
continued to minimize regrowth in previously rotovated
areas of Kalamalka Lake. Also, these improved
machines with larger crews were deployed widely to
search for and remove small, new Eurasian water
milfoil infestations on the littoral zones of Kalamalka
and Wood Lakes.
In the case of Kalamalka Lake, the combination of
early and extensive integrated control efforts in 1977
and 1978, combined with good underwater visibility,
appeared to have stabilized Eurasian water milfoil
populations (see Table 1). However, it was recognized
that although rotovating and diver dredging could give
very good cosmetic control and prevent nuisance
conditions, fragmentation of M. spicatum and the slow
speed of mechanical treatments would lead to further
spread and were unlikely to achieve the needed
reduction in size of dense major colonies. In 1979, 13
hectares of Kalamalka Lake were treated with 2,4-D;
most of this target area was the same shoreline that
had been treated previously with machines. Diver
dredging in 1979 (28 hectares) was concentrated on
eliminating new colonies in Kalamalka Lake.
Operating costs for rotovating have varied widely,
depending on the stage of mechanical development
and suitability of the area treated. Excluding ad-
ministrative, overhead capital costs, and monitoring
expenses, the last major rotovating work (Kalamalka
Lake, 1978) averaged $2,200/ha. In 1979, the diver-
operated dredging in Wood and Kalamalka Lakes was
estimated to cost about $1,600/ha. Herbicide treat-
ments have been relatively expensive because of high
administrative costs associated with the research and
monitoring objectives of all treatments to date.
Excluding some of these costs, but including costs of
the monitoring and alternate water supply systems
required by regulation of the Herbicide Use Permit, the
2,4-D treatments averaged $4,300/ha for Kalamalka
and Wood Lakes in 1979.
Documentation of aquatic plant populations in Wood
Lake has been frustrated by poor visibility until recent
years. It is suspected that Eurasian water milfoil growth
and expansion were limited by turbidity until about 1977.
At this time, a major increase in Eurasian water milfoil
was documented and in 1978 diver-operated dredges
were deployed to attempt control of these populations.
Unfortunately, a fourfold increase in M, spicatum was
recorded in 1978 (see Table 1) and plans were made for
major herbicide applications in 1979. Approximately 7
hectares of affected area at the south end of Wood Lake
were treated with 2,4-D in 1979 and limited diver
dredging was performed in an additional 5-hectare area
to reduce further fragmentation and expansion. In 1979
and 1980, aquatic plant fragment barriers were installed
in the canal between Wood and Kalamalka Lakes and
maintained to minimize fragment movement with the
current from Wood to Kalamalka Lake.
At this time, the Ministry of Environment is
continuing intensive management of both Kalamalka
and Wood Lakes. Because the apparent success in
achieving a major reduction in 1979 of area affected by
M. spicatum in Kalamalka Lake has been maintained in
1980, spot treatments using 2,4-D and intensive diver
dredging are continuing. It is hoped that containment of
Eurasian water milfoil can be maintained and that
overall expenses of this preventive maintenance can be
reduced. Monitoring of the Eurasian water milfoil
colonies must be continued unabated to ensure that
adequate resources are allocated to preservation. The
dramatic expansion of Eurasian water milfoil recorded
in Wood Lake in 1 979 has continued in 1 980 and major
2,4-D treatments of substantial shoreline areas are
being contemplated. Because of water use constraints
and the large colonies of plants, it is feared that
extensive management of Wood Lake may be more
difficult than Kalamalka Lake.
Cultus Lake
Cultus Lake is an oligotrophic lake of considerable
recreational importance because of its proximity to the
City of Vancouver and other major urban areas in the
Lower Mainland region of southeastern British Colum-
bia. The littoral area is relatively small (about 37
hectares) compared to the surface area (630 hectares).
Aquatic plant surveys were first performed in this lake
by the Ministry of Environment in 1977 as part of
baseline studies throughout southern British Colum-
bia. Surveys in 1977 and 1978 showed that a number
of other water bodies in the Lower Mainland were
affected. Because there were no documentation
studies prior to 1977, there are no adequate records to
indicate the probable sites of initial infestations in the
Lower Mainland. Approximately 13 hectares of littoral
areas of Cultus Lake were found to be affected by M.
spicatum. The populations located in Cultus Lake in the
fall of 1977 appeared to have resulted from recent
introduction and most plants were distributed then in
the marina area and in a downstream direction toward
the lake outlet.
Since numerous sites in Cultus Lake and several
adjacent water bodies were already populated with M.
spicatum, major control efforts directed toward
containment or eradication did not appear practical at
the outset. Because of public concern from residents
and local government and recreational agencies
associated with Cultus Lake, various options for
-------
AERATION/MIXING AND AQUATIC PLANT HARVESTING
151
managing the existing populations in this lake were
reviewed. Since M. spicatum had become established
in the river outlet of Cultus Lake, fragment barriers
were considered of limited value. Control by har-
vesters, rotovators, or other tillage equipment was
considered. However, the relatively sparse infestations
were not suitable for harvesting, and further spread of
fragments was feared; also, M. spicatum colonies were
established in many areas where obstacles (such as
docks and submerged logs) or rocky bottom conditions
would preclude effective treatment with large ma-
chines. The use of 2,4-D herbicide and/or diver-
operated dredges appeared to be most practical.
Because of controversy about use of 2,4-D and concern
about the public use of water from the lake, diver
dredging was selected. In cooperation with a local
group, the Cultus Lake Milfoil Action Committee,
formed in 1978, the Ministry of Environment per-
formed feasibility studies in 1978 to assess the use of a
diver-operated dredge.
The objective of the 1978 trials and subsequent
treatments performed under a cost-sharing agreement
(75 percent Provincial: 25 percent local funds), has
been to reduce the density of M. spicatum; this may
have reduced the rate of spread. In 1979 about 1.3
hectares were treated at a cost of about $50,000 and in
1980 about 2 hectares are proposed for treatment at a
cost-of approximately $60,000. Because of uncertainty
about the future impacts of M. spicatum's continued
expansion, and to provide detailed information on the
treatment method and its effectiveness, the Cultus
Lake operational program has been monitored closely.
The Cultus Lake example is intermediate between
the Kalamalka-Wood Lakes case (intensive manage-
ment) and the Okanagan Lake example where only
cosmetic control appears practical. Local circum-
stances, including factors such as recreational de-
mand, local cooperative interest, and the present and
potential impacts of plant growth, have determined the
management approach. As long as the Eurasian water
milfoil colonies in Cultus Lake appear to be relatively
stable in extent and density, semi-cosmetic manage-
ment at minimum cost appears to be justified. Because
there remains concern that M. spicatum may be
transported to other, noninfested lakes by the heavy
traffic of boaters who utilize Cultus Lake, efforts to
clear plants adjacent to the boat launch ramps are a
high priority. Public information and a voluntary
aquatic plant quarantine check station will be
employed by the Ministry of Environment in 1980 to
reduce the chance of further spread by boaters.
DISCUSSION
The aquatic plant managment experiences and
objectives of the B.C. Ministry of Environment are
illustrated by the case studies presented here.
Objectives of the comprehensive program that has
developed in British Columbia include:
1. Identification and evaluation of the conflicts to
multiple use of water bodies caused by unwanted
growth of aquatic macrophytes (and especially exotic
species such as Eurasian water milfoil).
2. Research and evaluation of all practical aquatic
plant management technologies and application strat-
egies.
3. Response to public need and relief of nuisance
conditions in an environmentally acceptable and
effective manner, with documentation of the results.
4. Prevention of undue further spread of Eurasian
water milfoil to presently unaffected water bodies in
British Columbia.
Experience has shown the complexity and high
degree of difficulty of successfully achieving these
objectives. Some of the major considerations and
constraints are summarized in context with the case
studies.
No management can be effective unless the problem
is clearly identified and public and political sentiment
support both the need and the means. Because of
general ignorance about the identification, biological
capacity, and ecological impacts of introduced aquatic
plants, there has been uncertainty about the nature of
aquatic plant problems. The initial reports of Eurasian
water milfoil in Okanagan Lake were followed by a
period during which the plants were identified correctly
and local nuisance conditions were experienced and
documented (1972 to 1975). The next phase (1975 to
1977) included the testing and development of
technologies believed most suitable to control the
infestations. Another very important preventive phase,
a major effort to contain the relatively small initial
infestations, was not initiated in time in Okanagan
Lake. Consequently, it now appears that the initial
infestations in Okanagan Lake were rapidly spread
downstream by water movement and possibly to
Kalamalka, Wood, and Cultus Lakes (and others) by
boaters. Spread of potential nuisance organisms is
difficult to monitor, especially in the aquatic environ-
ment, and infestations of aquatic weeds are virtually
impossible to locate at an early stage.
Experiences with aquatic plant management tech-
nologies have demonstrated that complete elimination
of exotic aquatic plants is exceedingly difficult. Also,
the degree of difficulty increases in proportion to the
area affected and the number of sources of reinfesta-
tion. In addition to the costs of technological
development, surveys to locate new infestations and
documentation of the trials and continuing evaluations
of ongoing work (i.e., harvesting benefits in Okanagan
Lake; measurement of diver dredging effectiveness in
Cultus Lake) are also costly.
To respond to predictable and projected management
needs, considerable staff with diverse skills must be
assembled and trained. This organization must im-
plement consistent, long-term policies in consultation
with local authorities to secure public support and
cooperation and to share management responsibilities
and expenses. The Cultus Lake project involves local
agencies in selection and cost-share funding of
appropriate management techniques. Responsibilities
for cosmetic control also will be turned over to local
authorities in the Okanagan Valley. The Province will
provide technical guidance and secure major funding
for the necessary implementation work.
Needs for long-term aquatic plant management, and
a preventive approach wherever practical, are apparent
because of the importance of water-based recreation to
the general public and to the tourism industry.
-------
152
RESTORATION OF LAKES AND INLAND WATERS
However, data are not available on the dollar value of
certain beach or marina areas. These data will be
important to determine and assess priorities for future
management. Comparisons of cost and benefit of various
technologies is difficult and may not be particularly
worthwhile until routine procedures are established.
Each technology is a somewhat specific tool appropri-
at to certain locations, situations, and seasons;
combinations of technologies gained from the intensive
management of Kalamalka and Wood Lakes may be
applicable to some other presently infested lakes or to
noninfested water bodies. These lakes present the best
opportunity to achieve a high standard of control at a
minimal cost of annual maintenance.
Th"e case studies illustrate the interactions of
planning, field observation, and deployment of appro-
priate technologies. Effective management requires a
successful combination of comprehensive policies and
long-range planning with technical knowledge, ade-
quate funding, and public support. Operational pro-
grams must be developed within a framework of clear
and realistic objectives. Continual feedback on the
needs and benefits of the restorative or preservative
program is essential to guide policymakmg and
implementation. All of the cases described here were
reactive situations. Experience has shown that the
rapid spread and potential nuisance impacts of exotic
aquatic plants should not be underestimated. Also,
control and management, once an exotic plant is
established, may be very difficult and costly, or
unsatisfactorily cosmetic and require an ongoing
commitment. This reality may pose a heavy burden for
senior government and the local authority or taxpayer.
Goddard, J. M. 1980. Studies on aquatic macrophytes. XXX.
Control of Myriophyllum spicatum in Kalamalka and Wood
Lakes using 2,4-D butoxyethanol ester in 1979. I: Data Rep.
Water Invest. Branch Rep. 2824.
Lim, P. G., and K. R. Lozoway. 1977. Studies on aquatic
macrophytes. X. A field experiment with granular 2,4-D for
control of Eurasian water milfoil, 1976. Water Invest.
Branch Rep. 2613.
Maxnuk, M. 1979. Studies on aquatic macrophytes. XXII.
Evaluation of rotovating and dive dredging for aquatic weed
control in the Okanagan Valley. Water Invest. Branch Rep.
2823.
Newroth, P. R. 1975. Management of nuisance aquatic
plants. Water Invest. Branch Rep. 2337.
1979. British Columbia Aquatic Plant Manage-
ment Program. Jour. Aquat. Plant Manage. 17:12.
Nijman, R. A. 1976. Studies on aquatic macrophytes. VII.
Aquatic plant documentation, Okanagan Basin, 1975.
Water Invest. Branch Rep. 2424.
Nordin, R. N. 1980. Strategies for maintaining water quality
in two British Columbia lakes. Can. Water Res. Jour. (In
press.)
Scales, P., and A. Bryan. 1979. Studies on aquatic
macrophytes. XXVII. Transport of Myriophyllum spicatum
fragments by boaters and assessment of the 1978 Boat
Quarantine Program. Water Invest. Branch Rep. 2761.
Stephenson, W., and D. D. Baillie. 1980. Studies on aquatic
macrophytes. XIII. Aquatic plant fragment barriers in the
Okanagan Basin, 1976-1979. Water Invest. Branch Rep. (In
preparation.)
Stockner, J. G., and T. G. Northcote. 1974. Recent
limnological studies of Okanagan Basin lakes and their
contribution to comprehensive water resource planning.
Jour. Fish. Res. Board Can. 31:955.
REFERENCES
Aiken, S. G., P. R. Newroth, and I. Wile. 1979. The biology of
Canadian weeds. 34. Myriophyllum spicatum L. Can. Jour.
Plant Sci. 59:201.
Anonymous. 1974. Kalamalka-Wood Lake Basin water
resource management study. Water Invest. Branch Rep.
2246.
1978. A review of mechanical devices used in the
control of Eurasian water milfoil in British Columbia. Water
Invest. Branch, Inf. Bull. IV,
Armour, G., D. Brown, and K. Marsden. 1979. Studies on
aquatic macrophytes. XV. Bottom barriers for aquatic weed
control. Water Invest. Branch Rep. 2801.
Bryan, A., ed. 1977. Studies on aquatic macrophytes. IX.
Experimental herbicide treatment for aquatic weed control,
Kelowna Boat Basin, Okanagan Lake, 1975. Water Invest.
Branch Rep. 2627.
1978. Studies on aquatic macrophytes. VIII.
Experimental hydraulic dredging for aquatic weed control in
Vernon Arm, Okanagan Lake, 1975. Water Invest Branch
Rep. 2727.
Bryan, A. 1980. Studies on aquatic macrophytes. XI.
Rototilling and hydraulic washing for aquatic weed control
in Okanagan and Kalamalka Lakes, 1976. Inventory Eng.
Branch Rep. (In preparation.)
Ceska, 0. 1977. Studies on aquatic macrophytes. XVII.
Phytochemical differentiation on Myriophyllum taxa collect-
ed in British Columbia. Water Invest. Branch Rep. 2614.
Dove, R F., and D. R. B. Malcolm. 1980 Studies on aquatic
macrophytes. XXXII. The 1979 aquatic plant quarantine
project. Inventory Eng. Branch Rep. (In preparation.)
-------
153
GERMAN EXPERIENCE IN RESERVOIR
MANAGEMENT AND CONTROL
JURGEN CLASEN
Wahnbachtalsperrenverband
(Association of Wahnbach Reservoir)
Siegburg, Federal Republic of Germany
ABSTRACT
In Germany reservoir management and control begins prior to construction. Monitoring programs
are set up for chemical and biological investigations of the tributaries. From the results various
control measures are derived, such as diverting of local wastewater inputs, designing a treatment
plant, etc. A traditional control measure has been to build pre-reservoirs. Their original purpose
was silt sedimentation, but more recently it has become obvious that pre-reservoirs act as nutrient
sinks and therefore can protect reservoirs from eutrophication. Another control measure is
establishing protective zones with legal restrictions for various activities. In cooperation with
forest experts, rules for afforestation of the slopes around reservoirs have been worked out.
Different control measures are applied to protect reservoirs from eutrophication. In the case of
point sources of nutrients wastewater is pumped through a pipeline to a treatment plant which is
situated downstream from the reservoir. In the case of diffuse sources nutrients are removed from
the tributaries. In small tributaries, this is done by constructing seepage trenches or filtering
through aluminum oxide columns; chemical precipitation and subsequent filtration in a plant or
sedimentation in the reservoir itself are used in large tributaries. Only hypolimnetic aeration is
widely used as a control measure within the reservoir.
INTRODUCTION
There are two aspects of reservoir management.
Either management relates to water quantity or it
relates to water quality. Although management of
water quantity is a well-established practice, methods
for water quality management are still being devel-
oped. Water quality problems are of special importance
in the drinking water reservoirs. These are reservoirs
from which water is pumped directly to a treatment
plant which supplies the public with potable water.
This paper deals exclusively with management and
control of water quality in drinking water reservoirs,
above all, considering the problem of eutrophication.
THE IMPORTANCE OF WATER
QUALITY PREDICTIONS
Management and control of a reservoir should begin
prior to its construction. This means that as many
measures as possible should be applied to protect
water quality when the reservoir is still being planned.
In Germany experience has shown that control
measures are much easier and less costly to put
through during this early stage than later.
The most important tool for such management is the
prediction of water quality of the reservoir which is
being planned. Up to rather recent times such
estimations were based on a saprobiological survey of
the main tributary and an analysis of only a few
randomly taken water samples. It was assumed that
these few samples were representative for a long
period of time and that the water quality of the
reservoir should not differ considerably from the quality
of its tributaries. This concept failed when reservoirs
could no longer be created in remote woodland; for
example, Wahnbach Reservoir had to be built in
densely populated agricultural land (Clasen, 1979).
Today, predictions of reservoir water quality are
based on detailed monitoring programs. The control of
point sources of nutrients is based on the results of
such investigations. For concentrations of nutrients
which cannot be controlled in this way the trophic
status of the planned reservoir is estimated using
Vollenweider's approach (1976). This prediction is then
used to design the planned treatment plant. The
concept which is described here is at present applied to
several reservoirs which are just being planned or built
in FRG.
REMOVAL OF SOIL AND
TREE TRUNKS
Water quality of a reservoir not only depends on
water quality of the tributaries but also on the quality of
the basin. This refers mainly to the soil,which may
release nutrients and undesirable organic substances.
As a consequence, removing the soil and tree trunks
-------
154
RESTORATION OF LAKES AND INLAND WATERS
may be desirable, but financially prohibitive. Since the
impact of soil is strongest during the first year after
impounding in such cases water which had been
dammed up might have to be discarded in the course of
one year. Very recently this happened in a new
reservoir in the Federal Republic of Germany.
INTAKE DEVICES OF
TREATMENT PLANTS
Experience has shown that treatment plants should
be able to take raw water from several different levels.
It is therefore possible to avoid concentration peaks
caused by fast flowing floods, development of algae, or
release of unpleasant substances from the sediment.
Of course intake at different levels is advantageous
only if at least a minimum of monitoring of water
quality in the reservoir takes place. Recently the author
of this paper was consulted by a treatment plant which
had run into great trouble because extraordinary
numbers of diatoms penetrated the filters. The
reservoir had not been monitored before and there was
an intake device at only one level. An in situ turbidity
profile revealed that this coincided exactly with the
peak of the heavy diatom bloom. In this case
management had to consist of waiting until the
breakdown of the bloom.
Different intake levels give several advantages.
During stratification periods raw water can be taken
close to the bottom in order to renew the water
overlaying the sediment. As soon as too much
manganese, etc., is released from the sediment one
has to switch to the next higher intake level. In several
drinking water reservoirs in the FRG experience has
shown that the hypolimnion should not be too small as
compared to the amount of raw water which is
processed during the period of stagnation (Bernhardt,
Clasen, and Nusch, 1973). Otherwise, the zone of
decay (hypolimnion) will become too small toward the
end of the stratification period as compared with the
production zone (epilimnion), which means an over-
loading of the mineralization capacity. These observa-
tions show that sometimes quality problems are closely
related to water quantity. This aspect should be
considered in a very early stage of planning.
PRE-RESERVOIRS
Traditionally, German reservoirs are provided with
pre-reservoirs in which at least the major tributaries
are impounded. The original purpose of these pre-
reservoirs was to retain silt. Recently, however, it has
become evident that they also retain nutrients and thus
to some extent protect reservoirs from eutrophication.
Basic work on this has been done mainly by Bendorf,
Putz, and Henke (1975) in the German Democratic
Republic. As opposed to Vollenweider, Bendorf and co-
workers only consideredo-PCU. Whereas Vollenweider
mainly considered water bodies with retention times of
more than 1 year, Bendorf deals with water bodies with
retention times of less than 3 months.
The phosphorus elimination in pre-reservoirs is
based on an extensive amount of bioproductivity.
Phosphorus becomes fixed in the biomass in the pre-
reservoir. This biomass is retained to a great extent by
sedimentation. The phosphorus remains fixed on the
bottom if there is sufficient oxygen available for this
purpose.
The P-elimination rate can be predicted if the
following parameters are known:
a) o-PO4 ortho-phosphate concentration in the
tributary P(/ug/l P)
b) Water temperature T (°C) in the reactor (pre-
reservoir) V
c) Average light intensity in the uppermost 3 m layer
of the pre-reservoir 1 m(cal/cm2.d)l
d) Calculated water retention time t(d)
e) Actual water troughput q(mVd)
From the parameters listed under a d the critical
retention time t can be calculated, t is identical with
t , if the output loss of phytoplankton is not greater
than the production rate.
\0.5+P / \10 + l / \ 20 /
— tkrit
From the ratio of actual (t ) to critical retention time the
elimination rate of o-PO :
tc VR
tknt q-t
- = n
Since Bendorf's formula also includes light-intensity
and temperature, estimations can be made for every
season. In winter when there is little bioactivity the
elimination figure for O-PCU is much smaller than
during summer. Wilhelmus, Bernhardt, and Neumann
(1978), who carried out similar investigations at
various pre-reservoirs in the FRG, calculated for
Wahnbach pre-reservoir an average o-PCUelimination
capacity of 40 percent over a total examination period
of approximately 2 years. For short periods, when the
inflow was very low this even increased to 90 percent.
Nusch and Koppe (1975) state that the P-elimination
rate of the pre-reservoir of the Moehne-Reservoir is40
to 80 percent. The elimination rate of another pre-
reservoir which did not have such a heavy P-load was8
to 65 percent, depending on the time of year. These
examinations show that the retention time of the water
in the pre-reservoir must be at least 15 days during
normal water flow to achieve a 60 percent P-
elimination. For calculations only the upper 3 to 5
meters of water are taken into account. The fact that
elimination capacity is low during the winter is a
disadvantage of the biological phosphorus elimination
process.
SEEPAGE TRENCHES
Although the beneficial effect of pre-reservoirs has
been known for decades, the mechanisms involved
have been understood only recently. This has resulted
from another method of reservoir control, the seepage
trenches. These are a means of reducing the P-content
of small streams rich in nutrients which originate
chiefly from diffuse sources. Phosphorus is eliminated
when the water passes through the ground and this
process is even more effective in predominantly fine-
-------
AERATION/ MIXING AND AQUATIC PLANT HARVESTING
155
grained sandy clay. The phosphorus becomes fixed in
the upper soil layers.
This method has been successfully used for years by
the Wuppertal Municipal Works for protecting the
Kerspe Reservoir, a drinking water reservoir in the
Bergisches Land. This reservoir contains 15.5 million
m3 and has eight important tributaries draining a
catchment area with an average of 60 percent
woodland. The rest of the area is made up of fields and
pastures. The population numbers 623 who live in
scattered settlements. The eight tributaries flow via a
pre-basin into seepage trenches. The o-phosphate ion
content of these streams is between 3 and 60 /ug/l, 17
/ug/l on an average. The total phosphorus concentra-
tion amounts to an average of 40 to 60 jug/lp.
According to Grau's examinations (1978), the use of
seepage trenches and pre-basins decreases the
dissolved o-phosphate ions from an average 17 /ug/l to
an average 7 /ug/l which corresponds to 61 percent
elimination. The seepage capacity of the one stream is
55 mVhr and that of the other is 250 mVhr at a
maximum, thereby guaranteeing complete treatment of
the influent, at least during the summer. In another
seepage trench, the average o-phosphate ion con-
centration is reduced by 50 percent from 25 /ug/l P to
13/jg/l P. During the total period it was reduced by 37
percent from an average of 19 A
-------
156
RESTORATION OF LAKES AND INLAND WATERS
reservoir is carried in by the river Wahnbach, the
reservoir's main tributary.
To change the reservoir from eutrophic to an
oligotrophic or mesotrophic state, the phosphorus
entering the reservoir had to be reduced by approxi-
mately 90 percent. The average total phosphorus
concentration of the river Wahnbach is reduced from
90 //g/l to 5 yug/l P. The chemical process used entails
precipitation with iron-lll-salts, subsequent floccula-
tion, and multi-layer filtration. The plant has a
maximum capacity of 18,000 mVhr and a filter surface
of 1,100 m2 A new method of precipitation,
flocculation, and filtration was developed, the 'Wahn-
bach System' This system not only eliminated the
phosphorus from the water but an average of 77
percent of the COD, more than 50 percent of the
dissolved organic compounds as well as 99 percent of
the bacteria. This system therefore gives an additional
improvement to the quality of the water in the
Wahnbach Reservoir.
The method of adding iron- or aluminum salts
directly to a lake or its tributaries so that the
phosphorus compounds can precipitate is widely used
in Dutch reservoirs but in Germany is applied only to a
part of the reservoir 'Halterner Stausee' (Gelsenwasser
AG).
TERTIARY TREATMENT
In the case of point sources the phosphorus
elimination processes carried out usually are con-
nected with normal mechanical-biological sewage
treatment. In the FRG this method is applied in Moehne
Reservoir where the outflow of the sewage treatment
plant is diverted to the pre-reservoir.
DIVERSION CHANNELS
The use of diversion channels (circular channels) is a
widespread means of preventing wastewater from
entering lakes and causing eutrophication. In the FRG
this method is applied mainly to natural lakes but also
to at least two reservoirs: Soese Reservoir and Innerste
Reservoir, both situated in the Harz mountains in
Northern Germany (Harzwasserwerke des Landes
Niedersachsen).
AERATION
Aeration is a method applied to various lakes but
especially reservoirs (Wahnbach Reservoir and Ennepe
Reservoir in the Federal Republic of Germany). The
main aim of aeration is to artificially restore the
equilibrium between bioproduction and respiration
capacity. Special apparatus is used to pump oxygen
(atmospheric air) into the tropholytic zone and
particularly into the sediment-water zone to compen-
sate the oxygen depletion processes there and to
maintain aerobic conditions in the micro-layer during
summer stagnation.
Thus aeration represents a process which prevents
phosphorus from being transported from the sediment
into the free water zone and thus into the production
zone. By creating oxidative conditions in the sediment-
water using artificial aeration phosphorus can be fixed
in the sediment to which it has been carried along
various paths and the sediment actually becomes a
phosphorus trap. Artificial oxygen input into the
sediment also means that the organic algal substances
collecting there can mineralize. This prevents oxygen-
depleting substances from accumulating.
The extent of the reductive solution of manganese
and iron compounds, the conversion of nitrate to
nitrogen or to ammonium ions, and the conversion of
sufate ions to hydrogen sulfate all depend on the size of
the reduction potential. If these reduction processes
cannot take place then concentrations of dual
manganese and iron ions cannot increase and
ammonium ions cannot form in the water in the
tropholytic zone. There can be no formation of sulfide
ions and methane gas cannot exist. These intercon-
nected processes have a considerable influence on the
size of the phosphorus cycle in a lake and thus on
plankton production (Ohle, 1953).
From experience gathered during work on the
Wahnbach Reservoir aeration must be carried out to
such an extent that the oxygen content of the micro-
layer is not allowed to sink to below 3 mg/l oxygen
during stagnation. If this is achieved, then one can be
sure that the upper layer of the sediment which is in
contact with the water body has sufficient oxidation
potential. Reduced ions and phosphate ions will be
prevented from being released into the free water zone.
The process of hypolimnic aeration which was
developed by the Wahnbach Reservoir Association over
10 years ago has proved particularly effective
(Bernhardt, 1978).
OTHER CONTROL METHODS
WITHIN RESERVOIRS
Facilities for other control measures within the
reservoirs are rather limited. The application of
herbicides such as copper sulfate to control algae has
never been considered in Germany. Growth limitation
by artificial mixing which is widely used in the
Netherlands and the United Kingdom is not applied in
the FRG for several reasons. Since the non-biotic
extinction coefficient is usually much lower than in
Dutch or English reservoirs the growth-limitation by
artificial mixing would be also much lower. Further-
more, water temperature would become too high in
summer since the FRG has a rule that temperature of
drinking water should not rise above 15°C.
ADMINISTRATIVE CONTROL
MEASURES
The hitherto described control methods deal with the
management of water quality of tributaries or of the
reservoir itself. Generally speaking, this depends on
the structure of the catchment area. As a consequence,
within the catchment area of German reservoirs
protective zones can be legislated to restrict various
activities (Bernhardt, 1975). These restrictions refer to
housebuilding, fertilizing, transportation and storage of
fuel, and so forth. The restrictions are severest close to
the water edge. The public is not even allowed to
approach the shoreline of drinking water reservoirs
The use of these drinking water reservoirs for
-------
ecreation is completely forbidden. The only exception
n some reservoirs is sport-fishing from the shore by a
imited number of people who have to renew their
ishing permit every year. In cooperation with forest
jxperts the German water managers have worked out
ules for the afforrestation of the slopes around
eservoirs (Bernhardt, in press). Coniferous trees are
avored to protect the reservoirs from leaf litter.
IEFERENCES
'.endorf, J., K. Putz, and W. Henke. 1975. Die Funktion der
Vorsperren zum Schutz der Talsperren vor Eutrophierung.
Wasserwirtshaft — Wassertechnik 25:19.
ernhardt, H. 1975. Richtlinien fur Trinkwasser-Schutzge-
biete II, German Assoc. Gas Water Works. DVGW-
^rbeitsblatt W 102. ZfGW-Verlag, Frankfurt.
Aeration of Wahnbach Reservoir without chang-
ng the temperature profile. Jour. Am. Water Works Assoc.
59:943.
1978. Die hypolimnische Beluftung der Wahn-
sachtalsperre. Gas- und Wasserfach 119:177.
_. In press. Behandlung des Waldes in Schutzge-
Dieten DVGW-Merkblatt W 105. ZfGW-Verlag, Frankfurt.
•..if
arnhardt, H., J. Clasen, and E. A. Nusch. 1973. Vergleich-
ande Untersuchungen zur Ermittlung der Eutrophierungs-
i/organge an Riveris- und Wahnbachtalsperre. Vom Wasser
40:245.
ernhardt, H., and H. Schell. 1979. Die verfahrenstechnische
Konzeption der Phosphor-Eliminierung am Wahnbach mit
Hilfe der Flocken-filtration (System Wahnbach). Z.f. Wasser-
und Abwasserforschung 12:78.
lasen, J. 1979. Das Ziel der Phosphorelimimerung am
Zulauf der Wahnbachtalsperre im Hinblick auf die Oligo-
trophierung dieses Gewassers. Z.f. Wasser- und Abwasser-
forschung 12:65.
elsenwasser AG. Personal communication.
rau, A. 1978. Der Einsatz einer Hangversickerung zur
Vorreinigung verschmutzter Wasser im Einzugsgebiet einer
Trinkwassertalsperre. Vom Wasser 50:101.
larzwasserwerke des Landes Niedersachsen. Personal
communication.
Hotter, F. G. 1979. Die technische Losung der Konzeption der
Phosphor-Eliminierungsanlage an der Wahnbachtalsperre.
Z.f. Wasser- und Abwasserforschung 12:119.
Nusch, E. A., and P. Koppe. 1975. Die Veranderung der
Wasserqualitat durch Stauhaltung in Talsperren. Wasser-
wirtschaft 65:8.
Ohle, W. 1953. DerVorgang rasanter Seen-Eutrophierung in
Holstein. Naturwissenschaften 40:153.
Schwertmann, U., and H. Knittel. 1972. Phosphatsorption
einiger Boden in Bayern. Jour. Plant Nutr. Soil Sci. 134:1.
Vollenweider, R. A. 1976. Advances in defining critical
loading levels for phosphorus in lake eutrophication. Mem.
1st. Ital. Idrobiol. 33:53.
Wilhelmus, B., H. Bernhardt, and D. Neumann. 1978.
Vergleichende Untersuchungen uber die Phosphor-Elimi-
nierung von Vorsperren. DVGW-Schriftenreihe Wasser 16.
ZfGW-Verlag, Frankfurt.
-------
158
THE EFFICACY OF WEED
FOR LAKE RESTORATION
HARVESTING
DARRELL L. KING
THOMAS M. BURTON
Institute of Water Research
Michigan State University
East Lansing, Michigan
ABSTRACT
Harvesting macrophytes is a useful means of restoring eutrophied lakes to a less nutrient-rich
status only where nutrient loading is reduced to a low level. Submerged macrophyte biomass yield
of 50 to 400 g dry wt/mVyr in northern lakes and 1 50 to 650 g dry wt/mVyr in southern lakes
and phosphorus content of the plants of from 0.2 to 0.4 percent of dry weight limit net phosphorus
removal. Harvest of plants such as water hyacinths which gain their carbon dioxide from the air
may offer the opportunity for net phosphorus removal from some lakes. Macrophyte harvest is a
sound ecological means of managing macrophyte abundance for recreational and aesthetic
purposes and may limit internal phosphorus loading in some shallow lakes.
INTRODUCTION
An increase in the rate of phosphorus addition to
most lakes is accompanied by an increase in biomass of
one or more of the aquatic plant forms. The division of
this increased photosynthetic activity between phyto-
plankton, filamentous algal periphyton, and macro-
phytes is determined by a complex interaction between
a great many factors peculiar to the individual lake,
including lake morphology and hydrology; clarity,
alkalinity, and nitrogen content of the lake water; type
of benthic sediment and the remainder of the biotic
community. But the propensity phosphorus has to sorb
on lake bottom sediments allows an early increase in
phosphorus available to rooted macrophytes in lakes
challenged by increased nutrient loading. Continued
phosphorus loading of the benthic sediments permits
rooted macrophytes to expand throughout the littoral
zone. Eventual equilibrium saturation of the benthic
sediments increases phosphorus concentrations in the
lake water sufficient to produce planktonic algal
blooms throughout the lake.
The early burgeoning growth of macrophytes
responding to increased sediment phosphorus concen-
tration in the shallow waters of lakes is a harbinger of
problems often accompanied by public desire to do
something about them. Depending on the prior history
of the lake, this public demand for macrophyte control
may occur when as little as 1 percent of the total lake
area is infested with macrophytes.
Direct harvest and removal of the offending plant
mass is one solution to public demand for macrophyte
control which, in addition, removes phosphorus from
the lake. In fact, harvest and removal of macrophytes
has been suggested as a means of restoring lakes to
some former less nutrient-rich status. The purpose of
this discussion is to consider the potential for lake
restoration and management offered by harvest of
aquatic plants.
POTENTIAL LAKE RESTORATION
HARVEST OF AQUATIC PLANTS
BY
Restoration of a lake to a less phosphorus-rich state
by harvesting aquatic macrophytes obviously requires
that the amount of phosphorus removed from the lake
exceed the annual net input of phosphorus to the lake.
The amount of phosphorus removed by plant harvest
depends on the rate of production .of harvestable
aquatic plant mass, the phosphorus content of that
plant mass, and the efficiency of harvesting. All
parameters vary considerably from lake to lake and
from plant to plant.
Production of Aquatic Plants
Early estimates of the nutrient removal potential
offered by harvesting were based on observations of
macrophyte abundance in wastewater ponds. Devel-
opment of standing crops of Ceratophyllum demersum
of 700 g dry wt/,m2 in 60 to 70 days in a wastewater
pond led McNabb and Tierney (1972) to conclude that
as many as three crops of 700 g dry wt/m2 could be
harvested in a 180-day growing season. However, even
with the abundance of nitrogen, phosphorus, potassi-
um, and other required nutrients characteristic of
wastewater ponds, successive harvests of macro-
phytes would not be possible because of limits imposed
by carbon availability (King, 1972).
Impressive amounts of carbon dioxide available from
the carbonate-bicarbonate alkalinity and bacterial
respiration of organic matter accrued over the winter
are sufficient to produce one crop of macrophytes in
wastewater ponds. Harvest of this first macrophyte
crop would not leave sufficient carbon for any
-------
AERATION/MIXING AND AQUATIC PLANT HARVESTING
159
significant subsequent regrowth of the plants. McNabb
(1976) later revised his estimate of potential harvest of
macrophytes from wastewater ponds to about 400 g
dry wt/mVyr.
Most aquatic plants appear to fix carbon by the Cs
pathway and become carbon limited at aqueous carbon
dioxide concentrations about 1 /umole CO/A . Craig
(1978) noted that Ceratophyllum demersum had a
required aqueous carbon dioxide threshold for net
growth of 1.3/uCO2/A and Liehr (1978) found that
the specific net carbon fixation rate decreased and the
required threshold aqueous carbon dioxide concentra-
tion increased for C. demersum with decreasing light.
Thus, submerged macrophyte growth is limited by
carbon availability even in sewage lagoons where
significant carbon dioxide is produced by bacterial
respiration of waste organics (Craig, 1978).
Development of planktonic algal blooms is often
observed following harvest of macrophytes. These
algae are able to continue carbon fixation to lower
aqueous carbon dioxide concentrations than those
which limit submerged macrophytes (King, in press)
and the simultaneous reduction of the aqueous carbon
dioxide level and depth of light penetration by the
planktonic algae markedly limits the potential for
regrowth of the submerged macrophytes. Emergent
plants gain most of their carbon dioxide from the air
and are not affected much by aqueous carbon dioxide
concentrations unless they are harvested to below the
water surface.
McNabb's (1976) estimate of 400 g dry wt/mVyr for
wastewater ponds appears to be about the maximum
harvest of macrophytes that can be expected from lakes
in the northern United States. After an extensive
literature review, Burton, et al. (1978) concluded that
potential harvest of submerged macrophytes would
range from 50 to 400 g dry wt/m2/yr in lakes in the
northern United States while submerged macrophyte
harvest from southern United States lakes could be
expected to range from 150 to 650 dry wt/m2/yr.
Emergent and floating plants which gain their carbon
from the air are capable of producing much greater
biomass. However, these forms generally are more
abundant in very shallow water and typically do not
comprise a significant harvestable component in most
lakes. Therefore, in most lakes, particularly in the
north, macrophyte harvest will be limited largely to
submerged forms.
Nutrient Content of Plant Biomass
The content of nitrogen and phosphorus in aquatic
plant tissues varies with the nutrient content of the
water (Gerloff and Krumbholz, 1966; Adams, et al.,
1971; McNabb and Tierney, 1972). At nutrient
concentrations less than the critical value required by
the plants, nutrient increases increase plant pro-
duction. Nutrient additions to lakes with nutrient
concentrations above the critical value for aquatic
plants do not yield increased production but are
accompanied by increased nutrient content of the plant
tissue through "luxury" uptake and storage (Gerloff
and Krumbholz, 1966; Gerloff, 1975; Wetzel, 1975).
Nutrient content of macrophytes in wastewater
ponds can be as high as 1.6 percent phosphorus and
in natural waters range from 0.05 to 0.75 percent
phosphorus and from 1.5 to 4.3 percent nitrogen on a
dry weight basis. From the literature review by Burton,
et al. (1979), it appears that mean values of from 0.2 to
0.4 percent phosphorus and 2.7 to 3.0 percent nitrogen
on a dry weight basis would be expected for
macrophytes from most lakes.
Potential Nutrient Removal by Plant Harvest
The amount of plant nutrient which can be removed
from a lake by harvest of macrophytes will depend upon
the density of the macrophyte growth, the nutrient
content of the macrophytes, and the percent of the total
lake covered by the plants. All three of these factors
vary significantly from lake to lake and each must be
assessed before the potential for effective nutrient
removal by macrophyte harvest can be estimated for an
individual lake. Knowledge of these three parameters
and the annual phosphorus loading to a lake allow
calculation of the degree of net nutrient removal
offered by harvest of macrophytes according to the
following equation.
% Removal of Annual P Loading =
Where:
(AP)(B)(PB)(100)
(PN)(AT)
Ap =Area of lake covered by macrophytes (m2).
B = Average biomass of plants in areas covered by
plants (g dry wt/m2/yr).
PB = Phosphorus content of plants (g P/g dry wt.).
PN =Net annual phosphorus loading (g P/m2 of lake
surface/year).
AT — Total area of lake (m2).
This equation and the assumption of a phosphorus
tissue concentration of 0.3 percent dry weight for
macrophytes were used to construct Figure 1 which
illustrates the amount of plant biomass which must be
harvested to remove an amount of phosphorus equal to
net phosphorus loading as a function of the percent of
the total lake area occupied by macrophytes.
With submerged macrophytes yielding an annual
biomass of from 50 to 400 g dry wt/mVyr, it is
apparent from Figure 1 that macrophyte harvest would
allow a phosphorus removal equal to net loadings of
only about 0.1 to 1.0 g P/mVyr even if the entire lake
bottom was occupied by macrophytes. Since phos-
phorus loadings of 0.1 to 1.0 g P/mVyr are the lower
end of those usually considered excessive and since
few lakes are entirely occupied by macrophytes,
macrophyte harvest by itself does not offer much hope
of restoring most lakes to a less nutrient-rich status.
Obviously, macrophyte harvest would be of some
benefit in nutrient removal in those lakes where
phosphorus input could be simultaneously reduced.
Inspection of Figure 1 indicates that a plant harvest
of from 2,000 to 3,000 g dry wt/m2/yr would be
required from a significant percentage of the total lake
to offer much net nutrient removal from most lakes
subject to excessive phosphorus enrichment. This level
-------
160
RESTORATION OF LAKES AND INLAND WATERS
where plants such as water hyacinths gained their
required carbon dioxide from the air (Burton, et al.
1979).
Even in those lakes where circumstances appear to
favor net nutrient removal by harvesting, success will
be predicted on regrowth of the macrophytes in the
following years. The effect of harvest on regrowth of
many submerged species is not clear, but Myrio-
phyllum regrowth can be diminished by harvest
(Nichols, 1974; Neel, et al. 1973; Grinwald, 1968). In
addition, abundance of submerged macrophytes may
decline because of other factors. Carpenter (1979)
hypothesized that invasion of Myriophyllum spicatum
may follow a wave pattern through lakes showing
explosive growth followed by a decline because of
unknown variables. Such lack of predictability makes
difficult any reasonably accurate estimate of continued
nutrient removal by macrophyte harvest.
Thus, it appears that in most lakes macrophyte
harvest by itself offers no real potential of removing
sufficient nutrients to yield some less nutrient-rich
status. Macrophyte harvest would be important in lake
restoration only as an adjunct to other restoration
measures.
REQUIRED PLANT HARVEST
(gDRY WT./M2/ YR.)
&
to —
G
J>
o
Figure 1. — Plant harvest required to equal net phosphorus
loading as a function of lake area covered by harvestable
macrophytes assuming a plant phosphorus content of 0.3
percent P (dry weight).
MACROPHYTE HARVEST AS A
MANAGEMENT TOOL
Despite a general inefficiency in nutrient removal,
macrophyte harvest offers many very real advantages
in managing macrophyte abundance in lakes. It is
particularly useful in controlling plant masses to allow
recreational use of the lake while minimizing some of
the undesirable aspects associated with complete
removal of the macrophytes (Burton, et al. 1979).
Removal of intact macrophytes reduces oxygen
demand and the potential for winter kill of fish
associated with bacterial respiration of the macrophyte
biomass.
Since macrophytes can cause significant nutrient
transfer from the lake bottom to the open water (Lie,
1979; Welch, et al. 1979), their harvest can help
reduce internal phosphorus loading to lakes, particu-
larly to shallow lakes. While this may reduce the
recyling rate of the nutrient in the lake, the harvest
itself can increase phosphorus levels in the water. The
nutrients contained in macrophytes cut but not
removed can be recycled rapidly to the water while the
cut stems can pump at least as much as 0.43 mg P/m2
into the lake water (Carpenter and Gasith, 1978).
CONCLUSIONS
The general worth of macrophyte harvest to lake
management and restoration depends to a large extent
on a great many variables peculiar to a given lake.
Harvest offers a direct, ecologically sound control of
aquatic weed abundance without adding foreign
materials. In addition, it removes oxygen-demanding
materials and some nutrients. As a management tool,
macrophyte harvest can control macrophyte over-
abundance for recreational and aesthetic purposes but,
by itself, is not effective for restoring most lakes to a
less nutrient-rich state.
REFERENCES
Adams, F. S., et al. 1971. The influence of nutrient pollution
levels upon element constitution and morphology of Elodea
canadensis Rich. Mich. Environ. Pollut. 1:285
Burton, T. M., et al. 1979. Aquatic plant harvesting as a lake
restoration technique. In Lake restoration Proc. Natl. Conf.
EPA 440/5-79-001. U.S. Environ. Prot. Agency, Washing-
ton, D.C.
Carpenter, S. R. 1979. The invasion and decline of
Myriophyl/um spicatum in a eutrophic Wisconsin lake.
Pages 11-31 in Proc. Aquatic Plants, Lake Management,
and Ecosystem Consequences of Lake Harvesting Conf.
Center for Biotic Systems, Inst. Environ. Stud. University of
Wisconsin, Madison.
Carpenter, S. R., and A. Gasith. 1978. Mechanical cutting of
submerged macrophytes: Immediate effects on littoral
water chemistry and metabolism. Water Res. 12:55.
Craig, J. 1978. Carbon dioxide and growth limitation of a
submerged aquatic plant. M. S. Thesis. Michigan State
University, East Lansing.
Gerloff, G. C. 1975. Nutritional ecology of nuisance aquatic
plants. EPA-660/3-75-027. U.S. Environ. Prot. Agency,
Washington, D.C.
Gerloff, G. C., and P. H. Krombholz. 1966. Tissue analysis as a
measure of nutrient availability for the growth of
angiosperm aquatic plants. Limnol. Oceanogr. 11:529.
Grinwald, M. E. 1968. Harvesting aquatic vegetation.
Hyacinth Control Jour. 7:31.
King, D. L. 1972. Carbon limitation in sewage lagoon. Pages
98-110 in Nutrients and eutrophication. Spec. Symp. Vol. 1
Am. Soc. Limnol. Oceanogr.
King, D. L. In press. Some cautions in applying results from
aquatic microcosms. In J. Giesy, ed. Microcosms in
ecological research. Savannah River Eco. Lab. Aiken, S.C.
Lie, G. B. 1979. The influence of aquatic macrophytes on the
chemical cycles of the littoral. Pages 101-126 in Proc.
Aquatic Plants, Lake Management and Ecosystem Con-
sequences of Lake Harvest Conf. Center for Biotic Systems,
Inst. Environ. Stud. University of Wisconsin, Madison.
Liehr, S. K. 1978. Interacting carbon and light limits to
macrophyte growth. M.S. Thesis. Michigan State University,
East Lansing.
-------
161
McNabb, C. D., Jr. 1976. The potential of submerged vascular
plants for reclamation of wastewater in temperate zone
ponds. Pages 123-132 in J. Tourbier and R.W. Piersort, Jr.,
eds. Biological control of water pollution. University of
Pennsylvania Press, University Park.
McNabb, C. D., Jr. and D. P. Tierney. 1972. Growth and
mineral accumulation of submerged vascular hydrophytes
in pleioeutrophic environs. Tech. Rep. 26. Inst. Water Res.,
Michigan State University, East Lansing.
Neel, J. K., et al. 1973. Weed harvest and lake nutrient
dynamics. EPA-660/3-77-001. U.S. Environ. Prot. Agency,
Washington, D.C.
Nichols, S. A. 1974. Mechanical and habitat manipulation for
aquatic plant management. Tech. Bull. 77. Dep. Nat.
Resour., Madison, Wis.
Welch, E. B., et al. 1979. Internal phosphorus related to rooted
macrophytes in a shallow lake. Pages 81-99 in Proc.
Aquatic Plants, Lake Management and Ecosystem Con-
sequences of Lake Harvest Conf. Center for Biotic Systems,
Inst. Environ. Stud. University of Wisconsin, Madison.
Wetzel, R. G. 1975. Limnology. W. B. Saunders Co.,
Philadelphia.
-------
162
LAKE RESTORATION - A HISTORICAL PERSPECTIVE
KENNETH M. MACKENTHUN
Enwright Laboratories, Inc.
Greenville, South Carolina
ABSTRACT
The Federal Government became involved in lake restoration in the mid-1970's, but the fruits from
the effort will be sampled in the 1980's, and in the years beyond. Efforts to enhance lake water
quality began in some States decades before the 1970's. As we proceed with this environmental
program it is well to recall the stepping stones of our progress. These began with scientific efforts
by Birge and Juday to understand lakes in the early 1900's; it followed with State laws directed
toward aquatic plant control of the late 1930's and early 1940's, the lengthy battle to divert
sewage from the Madison, Wis., lakes in the late 1950's, the Lake Washington story, and the
dedicated efforts by then Senator Mondale to enact Federal legislation to restore lakes. The
present chapter in the lake saga deals with the initial Federal grant program. The program has
stimulated public involvement, State lake restoration legislation, and scientific investigation.
Substantial Federal, State, and other funds have been committed to lake restorative activities. The
program's scientific success cannot be judged with any degree of accuracy for another decade.
However, the probability for success of the lake restoration program is excellent.
We hear much in advertising these days about "a
product of the 1980's, and the years beyond." Lake
restoration with the involvement of the Federal
government began in the mid-1970's, but the fruits
from the effort will be sampled in the 1980's, and
hopefully in the years beyond. Efforts to enhance lake
water quality began in some States decades prior to the
1970's. As we proceed with a program to enhance this
segment of the environment, it is often well to recall
the stepping stones of our progress. These can be
divided into investigative, legislative, and administra-
tive actions.
INVESTIGATIVE ACTIONS
The investigation of lakes historically is anchored
firmly to the comprehensive works of Birge, Juday and
Smith (Birge and Juday, 1911, 1922; Juday, 1914;
Smith, 1920, 1924) in Wisconsin and Forbes (1925) in
Illinois. These investigators developed methods, pro-
voked scientific interest, stimulated students, and with
their prolific writings enriched the scientific literature.
Even after seven decades, students of lakes would do
well to examine the early writings of that era.
The stepping stones of our investigative progress
have been compiled, from time to time, in several
notable books or proceedings. These publications
contain scientific facts and other information that
remain germane today. Perhaps the first useful
reference that addressed lake conditions is "The
Microscopy of Drinking Water' (Whipple, Fair, and
Whipple, 1927). This was first copyrighted by George
Chandler Whipple in 1899. "Problems of Lake Biology"
was published in 1939 (Moulton). In the foreword
Moulton stated that perhaps no other biological subject
involves a greater variety of interrelated factors than
lake biology.
A symposium on hydrobiology was held at the
University of Wisconsin in 1941. In the proceedings,
James G. Needham wrote with pleasure of his personal
knowledge of Stephen A. Forbes, Charles A. Kofoid,
and R. E. Richardson in Illinois and of Edward A. Birge
and Chancey Juday of Wisconsin. Hydrobiology, he
said, is an offshoot from the old maternal rootstock of
natural history, with ecology as an intimate associate.
In the 1960's, many will recall the Cincinnati, Ohio,
seminar on Algae and Metropolitan Wastes (U. S. Dep.
Health, Edu. Welfare, 1961) and the Madison, Wis.,
Symposium on Eutrophication (Natl. Acad. Sci. 1969).
The latter developed as a result of a "...recognition of
growing concern over problems associated with
eutrophication of lakes," and was held on a very hot,
humid day in June 1967, coinciding with a failure in
the air conditioning system at the University of
Wisconsin. Six hundred people representing 11 foreign
countries and the United States attended.
Noteworthy publications in the 1970's include the
"Environmental Phosphorus Handbook" (Griffith, et al.
1973), a comprehensive review of lake rehabilitation
techniques and experiences (Dep. Nat. Resour. 1974),
and the Minneapolis, Minn, conference proceedings on
lake restoration (Uc S. EPA, 1979).
The convoluted story of improving the Madison, Wis.
lakes entails many years of investigation and con-
troversy prior to eventual diversion, in December 1958,
of treated sewage effluent around the two lower lakes.
There are written reports of algal nuisances occurring
in Madison's lakes as early as 1881. In 1884, this city
of 12,000 used privies, cesspools, and direct drains to
the lakes to dispose of sewage. In 1894, the City
Council was told that the lakes are not to be used as
receptacles for sewage in the crude state. In 1897, a
sewage treatment plant was constructed but it fell far
short of the claim that it would produce an effluent as
pure as the water of Lake Mendota, and it was
abandoned in January 1901. A new treatment plant
-------
PUBLIC BENEFIT AND INSTITUTIONAL PROBLEMS
163
was constructed in 1902, but periods followed when
the system became overloaded and new plants had to
be built.
Madison's Nine Springs treatment plant was put into
operation in 1928 and flowed through the Nine Springs
Creek into the Yahara River, and thence into Lakes
Waubesa and Kegonsa. A Clean Lakes Association was
formed in 1931 to prevent pollution of the lakes.
Extensive treatment with copper sulfate to control the
algae, a controversial program, was instituted and
continued for many years. Sawyer and Lackey were
commissioned to study the problem in the early 1940's
and the results of their efforts are classic in lake
investigations (Sawyer and Lackey, 1943, 1944).
Sawyer published his conclusions regarding nutrients
and algal growths in 1947.
Following abortive legislative efforts, protestations,
public hearings. State orders, and court appeals, the
State of Wisconsin Supreme Court upheld the State
Board of Health and the State Committee on Water
Pollution, and the Sewerage District in 1951 was
forced to prepare plans for effluent diversion. The
Madison lakes saga was recorded by Sarles in 1961,
based largely on Flannery's historical research (1949)
during his graduate studies at the University of
Wisconsin.
The Lake Washington, Wash., story also extends over
many years (Edmondson, 1977). Starting in 1941, Lake
Washington went through a period of eutrophication
because of secondary sewage effluent. The initial
effects of increased abundance of algae and decreased
transparency coupled with predictions of the con-
sequences of further enrichment produced consider-
able concern among area residents. A public vote in
1958 established the Municipality of Metropolitan
Seattle with the responsibility for improving sewerage
in the region, including diverting effluent from Lake
Washington. The amount of effluent entering the lake
progressively decreased from 1963 to 1968. With the
first diversion of about one-third of the effluent, the
lake stopped deteriorating, and with further diversion it
began to recover, as measured by increasing trans-
parency and decreasing amounts of phytoplankton. By
1972 the lake began to come into equilibrium with its
new circumstances.
These, and many other studies, were placed in
perspective by Vollenweider (1968) in his excellent
effort to determine loading rates that may be
associated with biotic problems.
The phosphorus-in-detergents issue arose in 1971.
On September 15 of that year, the Council on
Environmental Quality, the Department of Health,
Education, and Welfare, and the Environmental
Protection Agency issued a joint news release on the
subject. The release made four principal statements: (1)
Nitrilotriacetic acid should not be used in detergents;
(2) the health hazards of using highly caustic
substitutes for phosphates in laundry detergents is a
serious concern; (3) States and localities should
reconsider laws and policies which unduly restrict the
use of phosphates in detergents; and (4) EPA will begin
an intensive study to identify those water bodies with a
potential or actual eutrophication problem caused by
phosphates. EPA also pledged assistance to States and
local governments in reducing phosphates through the
treatment of municipal wastes.
Subsequently, on October 27,1971, Russell E. Train,
Chairman, Council on Environmental Quality, in
testimony before the Subcommittee on Conservation
and Natural Resources of the House Government
Operations Committee stated that the principal
strategy in controlling eutrophication would be through
adequate waste treatment. Two days later, William D.
Ruckelshaus, Administrator, Environmental Protection
Agency, in testimony before the same Committee,
reaffirmed the initiation of a comprehensive National
Eutrophication Survey to identify those lakes where
municipal waste treatment plants should install
phosphate control equipment, or where industrial
nutrient sources should be controlled through the
Refuse Act permit program.
On May 7, 1972, EPA announced plans to use
specialized Army aircraft to collect samples "in a
project beginning this month'' to study eutrophication
in lakes and impoundments. EPA said the survey would
provide appropriate knowledge about whether a lake
could be improved by reducing municipal phosphates.
LEGISLATIVE ACTIONS
Although State legislation to financially support
restoration of particular lakes is of recent origin, some
States have had legislation for many years to control
aquatic nuisances. Wisconsin was one of the first. In
1941, the Wisconsin legislature passed an act ( —
144.025 (2) (i)) calling upon the Committee on Water
Pollution to supervise chemical treatment of waters to
suppress algae, aquatic weeds, swimmer's itch, and
other nuisance-producing plants and organisms. The
Committee was authorized to purchase equipment and
to charge for its use and for any services performed in
such work. This cost is covered by representative
taxation in a town's sanitary district.
On October 18, 1972, P. L. 92-500 was enacted.
Section 314 on clean lakes (33 USC 1324) provides the
legislative framework for the lake restoration program.
On February 5, 1980, cooperative agreements for
protecting and restoring publicly owned freshwater
lakes were published as a final rule (40 CFR 35.1600).
Clean lakes section 314 had a lengthy gestation
period. In 1966, Senator Mondale from Minnesota with
the support of Senators Burdick from North Dakota,
Douglas from Illinois, and Nelson from Wisconsin
introduced Senate Bill 3769, the Clean Lakes Act of
1966 (Congress. Rec. 1966). This bill would have
authorized the Secretary of the Interior to award grants
and contracts to State or local agencies for compre-
hensive pilot programs to improve and revitalize the
Nation's lakes by controlling pollution. In introducing
this bill, Senator Mondale stated that minimal attention
had been given to lake pollution and that there was no
Federal assistance program to help States clean
polluted lakes. This initial thrust was designed
principally to finance pilot projects.
Virtually the same bill was introduced to the
Congress again on March 21, 1967 (Congress. Rec.
1967). This bill, S. 1341, passed the Senate but did not
get through the House Committee on Public Works
(Congress. Rec. 1968).
-------
164
RESTORATION OF LAKES AND INLAND WATERS
On October 8,1969, the Senate passed Senate Bill 7,
the basis for the Water Quality Improvement Act of
1970. It contained provisions for basic research into
the cause, cure, and prevention of lake pollution.
Senator Mondale introduced the Clean Lakes Act of
1970 on April 8, 1970 (Congress. Rec. 1970a). He
called this Act an extension of the clean lakes research
provisions that he had introduced in 1966. This new
clean lakes bill would establish a coordinated program
of increased waste treatment and lake cleaning using
the latest technology to rehabilitate those lakes in
particularly poor condition. Twenty-six senators co-
sponsored this legislation (Congress. Rec. 1970b). This
bill was reintroduced into the Senate on February 26,
1971, as the Clean Lakes Act of 1971 (Congress. Rec.
1971) The full Senate Public Works Committee voted
October 18 to include the bill as part of the 1971
Federal Water Pollution Control Act Amendments. On
October 18, 1972, the lake restoration section 314
became a reality in the Federal Water Pollution Control
Act Amendment of 1972.
ADMINISTRATIVE ACTIONS
Particularly in the Great Lakes States, Iowa,
Michigan, Minnesota, and Wisconsin, there was early
concern about the growth and control of algae and
vascular aquatic plant nuisances. In Wisconsin during
1949, 78 chemical treatment projects were supervised.
This total included 33 projects to eradicate aquatic
vegetation, eight to control swimmer's itch, five to
reduce bacteria in bathing beaches, and one to spray
DDT over water. This was 22 percent more than
projects completed in 1948 (Mackenthun, 1950).
State-owned equipment purchased in 1941 was
operated on a rental basis to chemically control aquatic
nuisances. The program's early growth was governed
by the ability of the operating crew to treat as many of
the proposed projects as possible. It was soon found
that this procedure could not keep abreast of the
demands. Therefore, in 1949, the use of State-owned
equipment was discontinued. Sponsoring organiza-
tions were given the opportunity to select one of two
options in conducting the work: they might enter into
private contract with a commercial operator or they
might apply the chemical by using their own equipment
(Mackenthun, 1958).
Federal interest in lake restoration began late in
1971. A relatively small discretionary grant fund was
established to be used for adding a nutrient removal
capability to wastewater treatment plants. A specific
requirement of using such funds was that the addition
of such capability was necessary to prevent eutrophi-
cation of a freshwater lake.
Controversy surrounded the National Eutrophication
Survey, which was initiated prior to the enactment of
P.L. 92-500. An initial requirement in selecting the
lakes for study in this program was that they be
receiving waters for effluents from a municipal sewage
treatment plant within 25 miles of the lake. Thus, a
primary purpose of the survey was to develop a need
for a point source phosphorus control program. There
was concern that such a program was not consistent
with the needs of the Clean Lakes Program under
section 314. Later, for those lakes surveyed west of the
Mississippi River, the lake selection criteria were
broadened to include nonpoint source and other lake
problems.
Initially, also, the identification and classification of
all publicly owned lakes was considered to be a
problem for States to solve without Federal financial
assistance. An amendment in P. L. 95-217 clarified this
issue and directed the Administrator of EPA to provide
financial assistance to States to prepare the identifica-
tion and classification surveys required in section 314.
Such assistance, subsequently, was provided in the
form of matching funds on July 10, 1978 (43 F.R.
29617).
The Environmental Protection Agency did not
request budgetary support for a Clean Lakes Program
until fiscal year 1979. EPA believed that other
programs, which required full use of available
personnel had a higher priority in the goal for
environmental improvement. Prior to this budgetary
request, the Congress appropriated as an add-on, $4
million in FY 1975, $15 million in FY 1976, $4 million
in the transition quarter, $15 million in FY 1977, and
$2.3 million in 1978. Largely through the persistent
encouragement of then Senator Mondale, it was
possible to award the first lake restoration grant in
January 1976.
THE FUTURE
There was concern within the Federal bureauracy
during the program's formative years that the Clean
Lakes Program might parrot the construction grants
program in eventual magnitude. I believe that fear now
is abated.
The Clean Lakes Program was founded on a sound
technical base. Federal funding has so far been
sustained as an identifiable, consistent entity. The
need for such a program is becoming increasingly
apparent to a larger sector of the population. Public
participation is increasing and is coming from a broader
base of constituents. The future, I believe, is bright for a
sustaining Clean Lakes Program.
REFERENCES
Birge, E. A., and C. Juday. 1911. The inland lakes of
Wisconsin: the dissolved gases of the water and their
biological significance. Bull. 22. Sci. Ser. 7. Wis. Geol. Nat.
Hist. Surv. Madison.
1922. The inland lakes of Wisconsin: The
plankton. I. Its quantity and chemical composition. Bull. 64.
Sci. Ser. 13. Wis. Geol. Nat. Hist. Surv. Madison.
Congressional Record. 1966. Senate, August 26. 20839,
20774.
1967. Senate, March 21. 7453.
1968. Senate, September 26. 28321.
1970a. Senate, April 8. 10818.
1970b. Senate, April 16. 12177.
1971. Senate, February 26. 4095.
Department of Natural Resources. 1974. Survey of lake
rehabilitation techniques and experiences. Tech. Bull. 75.
Madison, Wis.
-------
PUBLIC BENEFIT AND INSTITUTIONAL PROBLEMS 165
Edmondson, W. T. 1977. Trophic equilibrium of Lake
Washington. EPA 600/3-77-087. U.S. Environ. Prot.
Agency.
Flannery, J. J. 1949. The Madison lakes problem. M.A.
Thesis. University of Wisconsin, Madison.
Forbes, S. A. 1925. The lake as a microcosm. III. Nat. Hist.
Surv. Bull. 15:537.
Griffith, E. J., et al. 1973. Environmental phosphorus
handbook. John Wiley and Sons, New York.
Juday, C. 1914. The inland lakes of Wisconsin: The
hydrography and morphometry of the lakes. Bull. 27. Sci.
Ser. 9. Wis. Geol. Nat. Hist. Surv. Madison.
Mackenthun, K. M. 1950. Cleaner lakes for Wisconsin. Health
9:12.
_. 1958. The chemical control of aquatic nuisances.
Comm. Water Pollut. Madison, Wis.
Moulton, F. R., ed. 1939. Problems of lake biology. Am. Assoc.
Adv. Sci. The Science Press.
National Academy of Sciences. 1969. Eutrophication:
Causes, consequences, correctives. Proc. Symp., Wash-
ington, D.C.
Sarles, W. B. 1961. Madison lakes: Must urbanization destroy
their beauty and productivity? Pages 10-16 in Algae and
metropolitan wastes. SEC TR W61-3, Robert A. Taft San.
Eng. Center, Cincinnati, Ohio. U.S. Dep. Health Edu.
Welfare.
Sawyer, C. N. 1947. Fertilization of lakes by agricultural and
urban drainage. Jour. New England Water Works Assoc.
61:109.
Sawyer, C. N., and J. B. Lackey. 1943. Investigation of the
odor nuisance occurring in the Madison lakes, particularly
Monona, Waubesa, Kegonsa from July 1942 to July 1943.
Rep. Governor's Committee.
1944. Investigation of the odor nuisance
occurring in the Madison lakes, particularly Monona,
Waubesa, Kegonsa from July 1943 to July 1944. Rep.
Governor's Committee.
Smith, G. M. 1920. Phytoplankton of the inland lakes of
Wisconsin, Part I. Myxophyceae, Phaeophyceae, Hetero-
konteae, and Chlorophyceae exclusive of the Desmidiaceae.
Bull. 57. Sci. Ser. 12, Wis. Geol. Nat. Hist. Surv. Madison.
1924. Desmidiacea. Part II. Bull. 57. University of
Wisconsin, Madison.
Symposium on Hydrobiology. 1941. University of Wisconsin
Press, Madison.
U.S. Department of Health, Education, and Welfare. 1961.
Algae and metropolitan wastes. Trans. 1960 Seminar. SEC
TR W61-3. Robert A. Taft San. Eng. Center, Cincinnati,
Ohio.
U.S. Environmental Protection Agency. 1979. Lake restora-
tion: Proc. Natl. Conf., Minneapolis, Minn. Off. Water Plan.
Stand. Washington, D.C.
Vollenweider, R. A. 1968. Scientific fundamentals of the
eutrophication of lakes and flowing waters, with particular
reference to nitrogen and phosphorus as factors in
eutrophication. Tech. Rep. DAS/CSI/68.27. Organ. Econ.
Coop. Dev. Paris.
Whipple, G. C., G. W. Fair, and M. C. Whipple. 1927. The
microscopy of drinking water. 4th ed. John Wiley and Sons,
Inc., New York.
-------
166
BENEFITS AND PROBLEMS OF
EUTROPHICATION CONTROL
D. J. GREGOR
Inland Waters Directorate
Environment Canada
Burlington, Ontario, Canada
W. RAST
International Joint Commission
Washington, D.C.
ABSTRACT
The objectives of eutrophication control are commonly thought of in terms of limnological
considerations, such as turbidity, algal biomass, nutrient concentrations and so forth. However,
management of water resources varies greatly among different geographic areas as a direct result
of different uses of the water body, as well as the historical perspectives of the user of the water. In
one area, a water body considered to be enriched in nutrients could in another area be considered
to be underproductive with respect to fish yields, or alternatively to be relatively pristine. It has
been determined, for example, that in the Laurentian Great Lakes the trophic conditions of some of
these lakes will be improved, while other lakes will be maintained simply in their present
oligotrophic condition, as a result of recommended phosphorus control objectives. Some estimates
of the costs and the anticipated limnological benefits of such an approach will be discussed.
Socioeconomic factors, such as shoreline property values, municipal water supply costs and fish
production, will also be considered, even though these factors were largely ignored in determining
the recommended phosphorus loading objectives for these lakes. Comparisons will be made with
other water bodies where conditions differ greatly, to demonstrate the need to look beyond
limnological concerns alone when attempting to address eutrophication problems.
INTRODUCTION
Both the lake manager and the scientist concerned
with water quality should reflect on why they consider
lake management and restoration to be worthwhile.
The purpose of such efforts must be the preservation
and/or restoration of a natural resource for the
greatest benefits to the greatest number of people,
including scientists, managers, and the public. This
paper focuses on the need to integrate considerations
of the public's use of water and its perception of water
quality, as well as scientific/limnological considera-
tions, in developing effective lake management
programs to control eutrophication.
Despite increasing emphasis in many countries on
other environmental problems, such as acid ram and
toxic and hazardous substances, controlling eutrophi-
cation is still an important issue of world-wide concern.
The scientist is usually concerned primarily with the
chemical, physical, and biological aspects of these
problems. The lay person, while vaguely appreciating
the technical aspects of water pollution, probably
assesses water quality most often on its aesthetic
value.
Traditionally, impacts of eutrophication have been
assessed almost solely on the basis of limnological
considerations. Commonly-used limnological indica-
tors include in-lake phosphorus concentrations, Secchi
depth (a measure of water clarity), algal biomass (often
expressed as chlorophyll), and hypolimnetic oxygen
depletion. These and other variables have been
discussed by Sawyer (1947), Sakamoto (1966), Vol-
lenweider (1968), Lee (1971), Burns and Ross (1972),
Dillon and Rigler (1974), and Schindler and Fee (1974),
to mention just a few authors.
As a result of this initial scientific emphasis,
eutrophication control has been based primarily upon
the views of scientists and engineers, with little public
input or scrutiny. Consequently, water quality may be
better or worse, depending on the circumstances, than
the public might desire on the basis of aesthetic or
economic concerns.
Public perception of waiter quality and eutrophication
control is not, however, characterized by the rigorous
analysis and review typical of the scientific view.
Rather, public perception of desirable water quality can
vary considerably, and is dependent primarily upon the
intended use of the water and its historical perspective.
For example, most North Americans look upon lakes as
a focus of recreational activities, such as fishing,
swimming, and boating, and as objects of intrinsic
beauty. Other uses exist, of course, including com-
mericial fishing, shipping, water supply, and waste
assimilation. Obviously, not all these uses are
complementary, nor are they necessarily mutually
exclusive. By contrast, in less-developed countries, the
production of fish as a food supply may supersede all
other uses. Thus, a clear, oligotrophic lake, which
-------
PUBLIC BENEFIT AND INSTITUTIONAL PROBLEMS
167
produces a limited quantity of fish, would be
considered an unproductive expanse of water.
Because of its technical and often complex nature,
the limnology of a water body may be unknown — or
unintelligible — to the public. When public perceptions
of desirable water quality conflict with scientific
opinion, it is necessary to recognize these differences
and educate the general public as to their significance.
Therefore, although global relationships exist which
describe relative trophic conditions and nutrient load-
response relationships within water bodies (e.g., OECD
International Cooperative Programme on the Monitor-
ing of Inland Waters), effective lake management must
not be restricted to limnological considerations, but
should also include public desires. In the western
world, few people would quarrel with the need for
phosphorus control for eutrophic water bodies. There
are frequently differences in opinion, however, re-
garding the in-lake conditions necessitating phos-
phorus control and/or to what degree such control is
required. This illustrates the need for integrating both
scientific and public opinions as to desirable water
quality in setting goals for control of eutrophication.
PUBLIC BENEFITS AND PROBLEMS
OF EUTROPHICATION CONTROL
Studies relating public benefits and problems to the
technical control of eutrophication are scarce; they are
difficult to generalize. By contrast, there is considerable
literature concerning benefits and problems associated
with eutrophication in a limnological sense. Several of
the public benefits and problems associated with
eutrophication control, as reported in the literature, are
discussed here.
Beneficial Effects
Increased fertilization of a lake usually produces
increased algal biomass, which in turn may increase
the total fish production (Oglesby, 1977; Lee and
Jones, 1980). Nevertheless, although total fish pro-
duction is increased with increased fertilization,
serious eutrophication of water bodies produces
changes in fish species from the more desirable game
fish (e.g., lake trout) to less desirable coarse fish
species (e.g., perch). The point at which increased
production of desirable game species is surpassed by
the increased production of less desirable fish species,
however, is not clear.
Alternatively, in situations where food supply is of
primary concern, gross fish production is the most
important factor, both as a social and economical goal.
As one additional benefit occasionally cited, Lefevre
(1964) has reported that some algal species in
eutrophic lakes have been found to produce active
substances of therapeutic value for treating patients
with ulcers and patients with atomic wounds.
Adverse Effects
The adverse effects of eutrophication, by contrast,
have received much greater attention in the literature
than the beneficial effects. That is due in part, of
course, to the fact that from the perspective of man's
use of water, there are usually more adverse than
beneficial effects.
Although not a general concern, human health can
be affected by eutrophication (Landner, 1976), though
the effects tend to be chronic and are often not
apparent because of inexperience on the part of
physicians in recognizing the toxic effects of causative
algae. Toxic effects of algae on humans may be
classified as gastrointestinal, respiratory, and derma-
tological. As cited in Landner (1976), Dillenberg and
Dehnel (1960) and Senior (1960) have reported cases
of severe diarrhea, vomiting, and other discomfort
occurring after swimming and/or swallowing water
heavily infested with Microcystis and Anabaena in
hypereutrophic lakes in western Canada. Heise(1951),
also cited in Landner (1976), noted respiratory
impairments in bathers in waters infested with
Oscillatoriae; Cohen and Leif (1953) reported derma-
tological problems in bathers following their swimming
in a lake with Anabaena blooms. Toxic effects resulting
from algal populations (mainly Aphanizomenon) have
also been noted for fish, poultry, and horses.
Welch (1978), citing a 1967 survey of State sanitary
engineers, noted 56 percent of the total municipal
surface water supplies in the United States ex-
perienced water treatment problems related to eu-
trophication. Cleveland water treatment plants (south
shore of central Lake Erie) are frequently subjected to
excessive clogging of their sand filters as a result of
excessive quantities of algae in the intake waters.
Taste and odor problems have also been noted in
municipal water supplies (Am. Water Works Assoc.,
1966). Similar problems occur in highly eutrophic
embayments and nearshore areas of the Great Lakes.
Welch (1978) also notes impacts of eutrophication on
industrial water supplies, shorefront property values,
commercial fisheries, and recreational activities, citing
several studies on the Great Lakes. The relationships
between the eutrophication process in a water body
and its effects on the use of the water are often
ambiguous. While eutrophication control measures
have often been related to improving water quality, few
attempts have been made to relate these measures to
their effects on the use of the water. To illustrate
possible relationships between eutrophication control
measures and public benefits, three specific examples
are considered. These three cases, involving the
Canadian portion of the Great Lakes Basin, are
intended as examples of public versus limnological
benefits, rather than as comprehensive analyses of
eutrophication control.
PUBLIC BENEFITS OF EUTROPHICATION
CONTROL IN THE GREAT LAKES BASIN
The Governments of Canada and the United States
have agreed that eutrophication control through
reduced total phosphorus loads is necessary in the
Great Lakes Basin. Phosphorus loading objectives have
been proposed for each of the lakes or sub-basins (U.S.
Dep. State, 1978). The proposed target loads for Lakes
Erie and Ontario and Saginaw Bay are 11,000, 7,000
and 440 metric tons/yr, respectively, all being
substantial reductions from present loads. As noted in
the 1978 Great Lakes Water Quality Agreement, these
-------
168
RESTORATION OF LAKES AND INLAND WATERS
loads can be achieved through a combination of point
and nonpoint source control measures. By contrast, the
target loads for the other lakes require minimal
additional reductions in phosphorus loads. The goals of
the proposed phosphorus load reductions for Saginaw
Bay and for Lakes Erie and Ontario are outlined in Table
1. It is noted these goals are expressed entirely in
limnological terms, except for Saginaw Bay, with no
attempt to relate them to associated public benefits.
Table 1. — Goals of proposed phosphorus target loads for
Saginaw Bay, Lake Erie and Lake Ontario.
Water Body Phosphorus Control Goal
Saginaw Bay Reduce filter-clogging and taste and
(440 metric odor problems in drinking water by
tons/yr) maintaining an average annual total
phosphorus concentration of 15 ,ug/l
in the inner bay.
Lake Erie Reduce anoxia by approximately 90
(11,000 metric percent in the central basin.
tons/yr) Prevent any substantial release
of phosphorus from the sediments.
Lake Ontario Minimize degradation of the ecosystem
(7,000 metric by maintaining an annual average total
tons/yr) system concentration of approximately
10jug/l.
Source: Task Group III (1978)
The United States and Canada have already invested
substantial sums of money in attempting to reduce
point source phosphorus loads and other pollution
problems associated with municipal sewage treatment
plants in the Great Lakes Basin. Capital expenditures
for construction and expansion of municpal sewage
treatment plants and sewerage works from 1971 to
1978 within the Basin are summarized in Table 2. The
total commitment is large in absolute terms. It is,
however, not very substantial when viewed on a per
capita basis. Achieving the proposed target loads will
require even further public expenditures for point
source and nonpoint source controls. It is the
translation of these costs into public benefits which is
of interest here. Commonly-cited benefits of eutrophi-
cation control:
1. Enhanced shorefront property values;
2. Enhanced recreational values;
3. Improved commercial fisheries; and
4. Reduced costs for municipal and industrial water
supplies.
Three of these benefits (shorefront property values,
commercial fisheries, and municipal treatment) are
discussed further in an attempt to determine whether a
distinct example of associated public benefit resulting
from eutrophication control can be identified.
Shorefront Property Values
Omerod (1970) considered the impact of algae-
fouled beaches on property values along the Canadian
shorefront of Lake Erie. He compared their real estate
values (average value per foot of water frontage) for
three categories of algae-fouling: (1) no algal cover; (2)
light algal cover; and (3) heavy algal cover. Omerod
Table 2. — Funds committed for municipal wastewater
treatment plant construction in the Great Lakes Basin
(millions of dollars).
Year
1971
1972
1773
1974
1975
1976
1977
1978
Total
$ per capita (1975)
Canada
57
66
138
103
112
174
150
191
991
144
United States
370
313
419
509
950
429
716
618
4324
146
Source International Joint Commission (1979)
determined that statistically there was no significant
difference between shorefront property values with
either light or heavy algal cover. However, the
combined light and heavy algal-fouled frontage
exhibited property values 15 to 20 percent below those
of the shorefront areas with no algal cover.
A recent study by Sudar (1980) compared property
values for the Bay of Quinte with contiguous eastern
Lake Ontario shorelines. This comparison was based
on sales information for 1971-73, converted to 1973
dollars. The median sale price per metre of shorefront
for the Bay of Quinte and for eastern Lake Ontario were
$495 and $449, respectively. As noted in Figure 1,
however, Lake Ontario water quality, measured in
terms of total phosphorus, chlorophyll a, and Secchi
depth, is considerably better than that observed in the
Bay of Quinte. These data demonstrate that factors
other than water quality must account for these
differences in shorefront property values. It is likely, for
example, that high lake levels during the mid-1970's,
and accompanying shore erosion and flooding, had a
greater impact on shorefront property values than
water quality degradation alone. It would appear, at
least in this instance, that better water quality does not
necessarily translate into higher shorefront property
values.
Commercial Fisheries
Considerable literature exists concerning the im-
pacts of various cultural stresses on the commercial
fisheries of Lake Erie. One is cultural eutrophication,
with its associated hypolimnetic oxygen demand and
potential anoxic conditions in the central basin
hypolimnion. This concern was expressed in both the
1972 and 1978 Canada-United States Great Lakes
Water Quality Agreements (U.S. Dep. State, 1972,
1978). Dobson and Gilbertson (1971) suggested that
the critical hypolimnetic oxygen depletion rate produc-
ing anoxia in the central basin was reached about
1 960. It is also noted that major coldwater species such
as lake trout, lake sturgeon, lake whitefish, and lake
herring disappeared as a component of the commercial
catches between the years 1940 and 1960 (Regier and
Hartman, 1973; Christie, 1974).
-------
PUBLIC BENEFIT AND INSTITUTIONAL PROBLEMS
169
BAY OF QUINTE
EASTERN LAKE ONTARIO
TOTAL FRONTAGE SOLD (m) 8,128 3,009
NUMBER OF SALES 227 91
MEDIAN SALE PRICE ($) 495 449
PER METRE OF FRONTAGE
024
2.3/SECCHI DEPTH (nT1)
PROPERTY VALUES BASED ON 1971,1972 & 1973 SALES
CONVERTED TO 1973 DOLLARS-WATER QUALITY DATA
ARE SUMMER,SURFACE DATA COMBINED FOR 1972 & 1973
A.
Figure 1.
However, other stresses, including over-fishing,
parasitism and/or predation by lamprey eels and other
exotic species, and loss of habitat along the shoreline
and in spawning streams, have also been cited by
Regier and Hartman and Christie as factors affecting
the commercial fishery of Lake Erie. It is not clear,
therefore, that reducing the Lake Erie phosphorus load
to the target level will insure the return of a viable
coldwater fishery. A more recent evaluation of Lake
Erie hypolimnetic oxygen data by Charlton (1980)
suggests that historic increases in the apparent
hypolimnetic oxygen depletion rate were not as great
as formerly believed and, furthermore, that the
differences which did occur in the oxygen status of the
hypolimnion were more related to variations in the
thickness of the hypolimnion than to changes in the
Lake Erie phosphorus loads.
This example illustrates two factors which are very
important in managing phosphorus loads to Lake Erie.
First, the scientific information is subject to different
interpretations and may, in fact, suggest a public
benefit which will not necessarily occur. Second,
commercial fish production in Lake Erie, the most
eutrophic of the Great Lakes, is the highest in the Great
Lakes system (Figure 2). Thus, if the Lake Erie
phosphorus load is reduced, and results in decreased
productivity, the impact upon the Lake Erie commercial
fishing industry may not be positive from the point of
view of the commercial fisherman. This is an instance
in which public benefit and, in particular, user-specific
concerns (i.e., commercial fishery) may be of more
importance in the lake management decisionmaking
process concerning phosphorus control than limnolo-
gical concerns alone.
1978 1974 1970 1966 1962 1958 1954 1950
Figure 3.
Municipal Water Supplies
Data for the city of Belleville water treatment plant
on the eutrophic Bay of Quinte (eastern end of Lake
Ontario) for the period 1950-1978 are being evaluated
(Gregor, 1980) as part of an attempt to assess the
impacts of eutrophication on the plant's operation. The
average filter-clogging rate for the month of July for 29
years of data, based on a preliminary analysis, as well
as the chlorination rate, are presented in Figure 3.
-------
170
RESTORATION OF LAKES AND INLAND WATERS
Available data for total phosphorus loads to the Bay of
Quinte are also provided. Plant records indicate that
filter-clogging by algae had become such a problem by
1958 that a micro-screen to strain large algae from
water intakes had to be installed at the plant. The
impact of the micro-screen is apparent in Figure 3. A
slight downward trend is noted from about the mid-
1960's to 1978. Nevertheless, it is noted that the 1978
results were obtained without the use of the micro-
screen, which had been operated every summer prior
to 1978. The chlorine application rate curve tends to
parallel the filter-clogging rate curve, although
changes in operating policy likely account for much of
the upward trend in the early years. Interestingly, the
1978 -.chlori nation rate dropped considerably, even
though the micro-screen was not used that year.
At present, these data are inconclusive and warrant
further evaluation. They do indicate, however, that
eutrophication control efforts are likely to produce
improvements in water treatment plant operations.
CONCLUSION
The control of eutrophication is, without question, a
worthwhile goal, especially in the more developed
western nations, which use water for multiple
purposes and therefore generally require better water
quality. However, it is incumbent upon scientists and
lake managers alike to consider the goals of
eutrophication control in other than strictly technical
terms. There is little point from a societal viewpoint in
improving water quality if the views of the public as to
desirable water quality are not considered. Conversely,
if the public is not educated as to the scientific basis for
phosphorus control efforts, such initiatives may be
unnecessarily or irreversibly restricted. The effective
management of water quality, to achieve the maximum
beneficial uses consistent with limnologically desirable
water quality, requires that these various, sometimes
diverse viewpoints be integrated into overall eutrophi-
cation control efforts.
PUBLIC PERCEPTION OF GREAT LAKES
WATER QUALITY
It is not yet possible in the Great Lakes Basin to
demonstrate clearly and quantitatively the public
benefits to be expected from eutrophication control. It
is interesting, nevertheless, to note the public's general
perceptions of Great Lakes water quality. A study,
summarized by the International Joint Commission
(1978), indicated that 38 percent of the people in
southern Ontario, Canada, used the Great Lakes for
diverse leisure activities. Perceptions of water quality
trends by the user public for Lakes Ontario, Erie, and
Huron are summarized in Table 3. It is interesting to
note that Lake Erie was the only lake in the Great Lakes
system perceived by a majority of the respondents to be
improving, contrary to what one would expect based on
examining the pollutant loads to the lakes. Other
conclusions from this study were that:
1. The number of respondents who perceived that
water quality was getting worse was decreasing with
time;
2. More than 50 percent of the respondents were
unaware of direct governmental measures to improve
water quality; and
3. Most respondents were willing to have more of
their tax money directed toward maintaining good
water quality.
Table 3. — Public perceptions in Southern Ontario of Great
Lakes water quality (1977).
Perceptions
Lake Ontario Lake Erie
Lake Huron
(percent of respondents')
Better
Worse
No change
Do not know
32 49
56 38
6 5
6 9
32
45
10
13
" respondents do not include non-user public
Source. International Joint Commission (1978)
REFERENCES
American Water Works Association. 1966. Nutrient-asso-
ciated problems in water quality and treatment. Rep. Task
Group 2610-P. Jour. Am. Water Works Assoc. 58:1337.
Baldwin,N. S., et al. 1979. Commercial fish production in the
Great Lakes, 1867-1977. Tech. Rep. 3, Great Lakes Fish.
Comm. Ann Arbor, Mich.
Burns, N. M., and C. Ross. 1972. Oxygen-nutrient relation-
ships within the central basin of Lake Erie. Pages85-119/>j
N. M. Burns and C. Ross, eds. Project Hypo— an intensive
study of the Lake Erie central basin hypolimnion and related
surface water phenomena. Pap. 6. Canada Centre for Inland
Waters, Burlington, Ontario, Canada.
Charlton, M. N. 1980. Oxygen depletion in Lake Erie: Has
there been any change? Can. Jour. Fish. Aquat. Sci. 37:72.
Christie, W. J. 1974. Changes in the fish species composition
of the Great Lakes. Jour. Fish. Res. Board Can. 31:827.
Cohen, S. G., and C. B. Reif. 1953. Cutaneous sensitization to
blue-green algae. Jour. Allergy 24:452.
Dillenberg, H. 0. and M. K. Dehnel. 1960. Toxic water bloom
in Saskatchewan. Can. Jour. Comp. Med. 24:26.
Dillon, P. J. and F. H. Rigler. 1974. The phosphorus-
chlorophyll relationship in lakes. Limnol. Oceanogr. 19:767.
Gregor, D. J. 1980. Unpublished data. Environ. Can.,
Burlington, Ontario.
Heise, H. A. 1951. Symptoms of hay fever caused by algae. II.
Myerocystis. Ann. Allergy 9:100.
International Joint Commission. 1978. Great Lakes water
quality. Sixth Annu. Rep. Great Lakes Water Quality Board
to the Int. Joint Comm. Great Lakes Reg. Off., Windsor,
Ontario.
1979. Great Lakes water quality. Seventh Annu.
Rep. Great Lakes Water Quality Board to the Int. Joint
Comm. Great Lakes Reg. Off., Windsor, Ontario.
Landner, L. 1976. Eutrophication of lakes — its causes,
effects and means for control, with emphasis on lake
rehabilitation. ICP/CEP 210. Swedish Water Air Pollut. Res.
Lab. IVL, Stockholm, Sweden.
Lee, G. F. 1971. Eutrophication. Pages 315-338 in
Encyclopedia of chemical technology, 2nd ed. John Wiley
and Sons, Inc., New York.
-------
PUBLIC BENEFIT AND INSTITUTIONAL PROBLEMS 171
Lee, G. F. and R. A. Jones. 1980. Effects of eutrophication on
fisheries. Rep. prepared for the Am. Fish. Soc. Eng. Res.
Center, Colorado State University, Fort Collins.
Lefevre, M. 1964. Extracellular products of algae. Pages 337-
367 in D. F. Jackson, ed. Algae and man. Plenum Press,
New York.
Oglesby, R. T. 1977. Relationships of fish yield to lake
phytoplankton standing crop, production and morphoe-
daphic factors. Jour. Fish. Res. Board Can. 34:2271.
Omerod, G. K. 1970. The relationship between real estate
values, algae and water levels. Lake Erie Task Force, Dep.
Pub. Works, Ottawa, Ontario.
Regier, H. A., and W. L. Hartman. 1973. Lake Erie's fish
community: 150 years of cultural stresses. Science
180:1248.
Sakamoto, M. 1966. Primary production by phytoplankton
community in some Japanese lakes and its dependence on
mean depth. Arch. Hydrobiol. 62:1.
Sawyer, C. N. 1947. Fertilization of lakes by agricultural and
urban drainage. Jour. N.E. Water Works Assoc. 61:109.
Schindler, D. W., and E. J. Fee. 1974. Experimental lakes
area: Whole lake experiments in eutrophication. Jour. Fish.
Res. Board Can. 31:937.
Senior, V. E. 1960. Algal poisoning in Saskatchewan. Can.
Jour. Comp. Med. 83:1151.
Sudar, A. 1980. Personal communication. August 15.
Environ. Can., Burlington, Ontario.
Task Group III. 1978. Fifth-year review of Canada-United
States Great Lakes Water Quality Agreement. Rep. Task
Group III, a technical group to review phosphorus loadings.
U.S. Dep. State, Washington, D.C.
U.S. Department of State. 1972. Great Lakes Water Quality
Agreement, with annexes and texts and terms of reference,
between the United States and Canada, signed at Ottawa
(Ontario), April 15, 1972. Washington, D.C.
1978. Great Lakes Water Quality Agreement, with
annexes and terms of reference, between the United States
and Canada, signed at Ottawa (Ontario), November 22,
1978. Washington, D.C.
Vollenweider, R. A. 1968. Scientific fundamentals of the
eutrophication of lakes and flowing waters, with particular
reference to nitrogen and phosphorus as factors in
eutrophication. Tech. Rep. DAS/CSI/68./27. Organ. Econ.
Coop. Develop. Paris.
Welch, J. L. 1978. The impact of inorganic phosphates in the
environment. EPA-560/1-78-003. Off. Toxic Subst., U.S.
Environ. Prot. Agency, Washington, D.C.
-------
172
THE POLITICS OF BENEFIT ESTIMATION
DAVID J. ALLEE
College of Agriculture and Life Sciences
Cornell University
Ithaca, New York
ABSTRACT
Several studies of a newly created lake carried out in the late 1960'sare used to review the current
state of the art in economic benefit estimation. Techniques for direct user benefits and for indirect
economic effects in the community have become well established. The basic concepts are
willingness to pay and the household income partition of local business multipliers. These
techniques applied to public water project investments have stressed the effects of intermediate
products: For example, irrigation, power, property damage from floods. Final consumption products
have been slighted, such as recreation, trauma from floods, and amenity values. Likewise, when
indirect economic effects are considered, they stress employment and money income. Rarely is
there a consideration of option values and/or income redistribution. Such a mix in evaluation fits
the politics of traditional water resource development projects, i.e., distributive politics. But this
does not fit the politics of many environmental problems which are redistributive or regulatory in
character. Evaluation plays an important role in achieving political agreement. Different politics
call for different evaluation techniques. Those benefits slighted by traditional analysis must be
developed in evaluation if environmental restoration is to be achieved with less conflict.
INTRODUCTION
Water attracts people. A lake can be a joy to a
community. Its utilitarian values are an endless list. It
can perform them all while still serving as a magnet for
recreation and refreshment. Just going by it everyday
can be a reminder of things enjoyed, of a well ordered
life and environment.
At Cornell University we have Cayuga and the other
Finger Lakes at our doorstep. Cayuga must be one of
the most studied lakes in the world. Besides the many,
many scientific investigations reposing in the libraries
of the university, it has been studied by virtually every
kind of resource management and planning program
that Federal, State, and local government agencies can
devise. A few years ago Cornell's Environmental
Research Center completed an educational and
research effort on lake management that concentrated
on Cayuga and Owasco Lakes among others. This
paper is based in part upon that experience as well as
an earlier series of studies of a nearby reservoir in the
Susquehanna River Basin at Whitney Point, N.Y.
Lakes are parts of larger hydrologic systems and
must be managed as a part of those systems. They
enjoy a distinction from the rest of the system in that
they are much more easily noticed. Many people won't
know which way the water in a lake flows, but they will
know it's there and are more likely to be aware of its
condition than surrounding streams. Thus, a lake is
more likely to receive management by the public even
though it may not come under a separate management
entity.
Public management of such a resource includes a
number of actions that are more effective the more the
system to be managed is understood. But who must
understand what? Many public expenditures are
involved. Rules governing both public and private
activities have to be devised to protect and enhance the
values of the system. Monitoring, evaluation, and
administrative decisions will proceed. Through it all the
people involved must learn to understand and respond
to the opportunities for system management. Without
public management users will add their demands to a
natural system until carrying capacity has been
exceeded and diminishing returns set in. Investments
to increase capacity likewise are limited and finally
demand must be managed. Not the least of the process
is accommodating water related values to nonwater
interests. Water systems interact so extensively that
public management means multiple agency, inter-
governmental management. The result of this is that if
an agency is put in charge of a natural system it never
has enough control and must influence others by many
means. Also, some systems are managed well enough
without someone obviously in charge.
Public management decisions by definition involve
politics — the balancing of a variety of interests
through the structure and processes of government.
We academics are quite used to the idea that decisions
can be grouped by kinds of information required. But
we are less aware that different decisions call forth
different kinds of politics, even different sets of
decisions. Public management of a water system
requires a planner who in his official capacity must
serve as a broker between the major interests with a
direct stake in that system. The planner may be in the
U.S. Corps of Engineers examining water level control
options, or a consulting engineer designing a sewage
treatment plant, or a regional official trying to deal with
an aquatic weed problem. Whatever they try to do, they
must deal with a series of veto points in their decision
processes; some of these will turn out to be bargaining
-------
PUBLIC BENEFIT AND INSTITUTIONAL PROBLEMS
173
arenas where other interests, compatible or conflicting,
will have to be accommodated or be accommodating.
The intent of this paper is to explore the differences
in decisions, information, and politics and how they
may be changing. In particular, the focus is on the
problem of estimating benefits of public management
options. Not only is the problem one of estimating what
those benefits are, but it appears that it will be
increasingly important to examine whose benefits are
at stake.
BENEFIT ESTIMATION AND
DISTRIBUTIVE POLITICS
Comparing benefits and costs began in earnest after
the Flood Control Act of 1936 called for their display in
any proposal for a federally funded water project. The
dominant politics is that of the local public works
project largely funded by nonlocal government. Here
benefit to cost evaluations serve as a screen to deny or
delay the allocation of the public benefit to otherwise
deserving recipients. It makes manageable the com-
petition for limited public funds, and does it on the
basis of a measure of general welfare akin to the
economist's concept of efficiency and gross national
product. As carried out by the agencies, the concept is
sufficiently flexible that it also structures support at the
local level for the project. Traditionally, it has
minimized conflict at both the local and national level.
Distributive politics is a term coined by Lowi to
describe the classification into which most water
projects seem to fall. A large number of intensely
organized interests operate with their major common
interest being obtaining the government action at
hand. Each project is dealt with separately, and
"mutual noninterference," "log-rolling," and "pork
barrel" are used to describe the coalition building
processes involved. Leadership is executed by broker-
age and is more likely to be expressed in the legislature
or in an agency rather than by the executive. Policy is
arrived at more through cooptation rather than conflict
and compromise. Avoiding conflict at both local and
national levels is an essential ingredient to the success
of their model of politics and can lead to its change. The
focus is on gaining something rather than balancing of
costs and returns to different groups. Costs are so
diffused as not to be perceived nor well represented.
Most important, the decision on how to solve a problem
— the output of the policy— is made when those with
the problem first approach an agency for help.
But no real situation ever fits only one model
perfectly, and public policy in water resources is no
exception. A second, if not dominant, model applies,
labeled redistributive politics by Lowi. This is the
politics of the "rules of the game." It is expressed more
through an elite such as economists or environment-
alists who hold important positions in the policy
process. Class is more important than group. Ideology
shapes policy and choice more than the distribution of
benefits and costs. In redistributive politics there are
rarely more than two sides to an issue, e.g.,
environment vs. development; and one elite for each
side, rather than many separate groups. While still a
gross oversimplification, two significant elite values in
the redistributive political sense have been important
in water resources. The first is the pressure for a
rational-analytic structure and process with its two
branches — orthodoxy and objectivity in economic
analysis and holisitic system management. The first
branch stresses proper and comprehensive economic
analysis kept close to market based values and has
frequently been a vehicle for asserting executive
branch authority over the process. The second branch
stresses river basin studies and draws some support
from some State and Federal agency program
managers as a way for them to communicate.
More recently and to much more effect, the
environmentalist's vigorous opposition to water devel-
opment projects, and indifference, if not hostility, to
waste treatment works have reshaped the informal
rules for distributive politics. Ingram has detailed the
problem of restricting conflict over water projects when
the opponents have a local as well as a national base
and where there is little that the agencies can give
them. What they want is no less than a different
system of politics. It is well not to lose sight of the fact
that in a democracy elites and ideological values prevail
only when they are widely understood and accepted.
Then the elites are acting with acquiescence if not with
much organized support. Obviously, the body politic
will accept rules that limit largess and freedom when
the limits can be justified by appeal to some higher
value. The makers of the principles of redistribution are
indeed the holders of the command posts as Lowi
argues, but the rules-of-the-game command real
adherence only when everyone expects them to be
enforced. When enforcement is not expected and
sometimes, even when it is, requirements for such
things as benefit cost analysis or environmental impact
statements will be honored symbolically rather than as
a substantive part of decisionmaking.
The rational-analytic model in both its economist and
planner versions, with embellishments and additions
from the environmentalist including a preference for
demand management (conservation) and nonstructura!
alternatives takes second place to'distributive politics.
THE BENEFITS FROM WATER
RESOURCE MANAGEMENT
Benefits are estimated, in part, to be compared to
each other and to the costs involved in creating them.
Fair comparability can be an elusive goal, but
comprehensiveness is even more difficult to achieve.
Since, unlike toothpaste, there are no open markets for
water services, indirect measures of direct benefits
must be devised. To be comparable they must
approximate the values an open market would assign.
An example long used and widely accepted is the case
of flood control benefits. Repair costs likely to be
avoided by flood control offer a measure of what the
beneficiaries should be wiHing to pay for flood
protection. Reasonable people should be able to agree
on what beneficiaries should be willing to pay if data
are available to estimate likely damages.
Methodology is more easily developed and used to
estimate the value of water used to produce products
that are in turn sold in a market. These values are then
derived from observable prices. If the shipper didn't use
the waterway or the power company didn't use the
-------
174
RESTORATION OF LAKES AND INLAND WATERS
hydropower, or the farmer didn't irrigate, what would
they do as an alternative? The "with and without"
situations are estimated and the differences in net
returns are struck. Money returns are easily estimated,
debated, and decided. They represent a common
denomination in the analysis and help identify self
interest in a decision that will affect those returns.
Evaluations that stress flood control, irrigation,
power, and navigation fit the traditional expectations
for water development programs and some of the
public interest in them. As long as the conflicts were
the interests of upstream and downstream water
users, these benefits served the purposes of evalua-
tion. But other interests have been seriously involved at
least since "Earth Day" in 1970.
The addition of recreation to the recognized benefits
of water development has stimulated other questions.
Recreation is a product of water that is directly used.
Early efforts to simply use the associated expenditures
of recreationists as a proxy for the benefits created
were soon disparaged. Those were costs analogous to
the costs of harvesting an irrigated crop, and there was
no logical reason to expect them to approximate or
even correlate with the size of the net benefit to those
individuals after they had paid associated costs. What
should the users be willing to pay for access to the
water? Examining what a consumer should be willing
to pay as opposed to using the water to produce
something for further sale suggests a number of
questions.
Consumers, except for the very rich, can be expected
to require a larger compensation for giving up an
experience than they feel able to pay for using it. The
willingness to sell is higher than the willingness to pay
because of the constraint of income. Also something
used to produce a good for sale is apt to have closer
substitutes. Likewise, looking to the future, the direct
consumption of resources such as wilderness rec-
reation is apt to expand in demand relative to the more
commercial values of the resource, while technology is
less likely to benefit its supply.
And note that the directly enjoyed values can take on
some very interesting characteristics in terms of
distribution between people as individuals and as part
of the community. One person's enjoyment doesn't
necessarily reduce another's. Indeed, even nonusers
may take more pleasure in knowing that they may
become users and that others are using the resource.
Also, the cost of restricting use to only those who pay
for it may be very high.
The net result of these characteristics is that the
stake involved in the directly enjoyed use is apt to be
spread over many people and is more likely to be a
small part of each of their income or satisfaction with
life. Commercial uses are apt to mean a great deal to a
smaller number of people. Thus, the cost of getting
organized to either bid for the resource or to represent
their interest politically is much greater for the
diffused, direct enjoyment users than it is for the
commercial users. An important exception would be
where this disadvantage is widely recognized and
political leaders are supported in efforts to tilt the rules
of the game in favor of individuals.
Extending these concepts to water quality manage-
ment, fish and wildlife values, aesthetic, cultural, and
spiritual values has been suggested. Also, flood control
benefits based only on property damage can be seen to
be deficient when the value of trauma avoided is
considered. It would seem logical that the more violent,
harder to deal with floods, may be more trauma
producing. If true, developing valuations for flood
trauma may reinforce flood plain management al-
ternatives at the expense of structural measures for
flood control.
Always a concern in methodology development is
whether two or more approaches are actually
measuring the same thing, and whether the instru-
ment imparts a bias of its own. For example, when you
ask people what they will pay, will they in fact do so?
Probably not, but what to do instead, and how much is
the bias? Will respondents shade their answers to give
what they think the investigator wants, and what will
that be? Will they expect to have to pay if they say they
will? Or will they raise their values to induce more of
what they want to be provided, since they can't imagine
that they can be made to pay? The alternative to asking
questions is to use costs incurred by users such as
travel and time to relate differences in cost to quantity
enjoyed. Such a travel cost method is employed to
relate quantity to price in a manner similar to demand
studies in a conventional open market.
In a study carried out at a reservoir in central New
York in 1966, Romm compared 10 techniques. Table 1
summarized the results. While the range is large —
four and five to one — plausible arguments suggest a
part of the variation is due to differences in what is
being measured. For example, travel cost has many
elements that are not variable with the trip. The facility
in question was very heavily used by young people and
given our youth culture this may have added to the
larger value for willingness to have government spend.
Table 1. — Alternative estimates of the value of recreational use of a
reservoir, Whitney Point, N. Y., 1966.
Summary of method
Benefit Per User
U.S. dollars (1966)
Travel cost, without time value adjustment 0.29
Additional distance willing to travel — 0.35
hypothetical bids
Willingness to pay fee — hypothetical bids 0.39
Combined distance and fee — hypothetical bid 0.63
Willingness to pay in addition to taxes — 0.26
open end question
Willingness to pay asked after question on 0.45
government investment
How much should government spend per user $1.31
day?
Compared to next best priced alternative, $1.03
recreation preferred
Compared to priced alternative, recreation $1.35
preferred and rejected
Cost of requirements for the same activity 0.75
Source: Romm, 1969.
Bishop and Heberlein report on an experiment that
created a market where one had not been before. They
bought special goose hunting permits in Wisconsin (for
an average of $63) and then compared that result with
hypothetical offers of willingness to sell ($101),
willingness to pay ($21), and travel cost estimate with
no time value ($11) up to one-half the income level
-------
PUBLIC BENEFIT AND INSTITUTIONAL PROBLEMS
175
($45). The relative positions of the estimates were as
expected, but the magnitude of the differences — a
range of ten to one — was surprising.
Batie and Shabman point out a critical problem in the
development of this type of methodology. Estimates
almost always value the complete, directly enjoyed
service. But the planner and the policymaking process
must deal with policy measures that affect less than
the whole value. In other words, they have not
addressed the "with and without" problem. Focusing
on the control variables is surely an important
challenge for research. But that is not likely to explain
why planners and agency analysts have moved slowly
and selectively in estimating directly enjoyed benefits.
This requires an explanation that deals with the politics
or institutional setting for the use of such information.
The methodology of manageable costs has been
available, could have been adapted to the "with and
without" problem, and gives results that are no less
precise than the more traditional benefit estimates for
flood control, irrigation, navigation, and hydropower
and reservoir recreation.
SEEDS FOR A SHIFT IN THE POLITICS OF
WATER RESOURCES
Planners estimate benefits and costs (including the
loss of existing benefits) to match the demands for
information of the kind of politics in which they find
themselves. Distributive politics requires finding local
agreement and translating that into national agree-
ment. Increasingly, conflict cannot be contained.
Formalized public participation processes have been
introduced partly to deal with this and partly to respond
to a more recently emphasized change in the rules ol
the game by the holders of the command posts.
An example is the recent national water quality
planning exercise under section 208 of P.L. 92-500
where various advisory committees were required:
technical, public, and local government. Contracts
were signed with various nontraditional water quality
clientele (agricultural and forestry agencies, general
regional planners) for plan elements. Farm and forestry
agencies that are developing water quality programs
will probably continue as part of the water quality
network locally and nationally. In some localized
situations, local governments have used the oppor-
tunity to develop permanent management capacity. But
many communities lacked a publicly demonstrable,
immediate problem; the permit process covered
industrial discharges and federally funded municipal
plants were controlling municipal discharges. There
was little basis to devise new institutional arrange-
i ments. But what of the future? Will the very high costs
''of advanced treatment and the extreme system
burdens posed by toxics and nonpoint pollutants
provide the impetus for a regional approach? Will the
rational-analytical model at the basin level be given
new life because it can do the job more cheaply? If so,
much of the bargaining over pollution standards and
enforcement would shift to the regional level from the
State-Federal focus it has enjoyed in recent years.
Achieving agreement in the face of environmentalist
opposition may force something similar on the dam and
channel building agencies.
In water resource development, informal public
participation has never been lacking. The support
requirements for dam and channel projects have been
substantial from the time studies begin to the last
Federal dollar spent decades later. For public sewage
treatment plants first-come, first-served rules and a
complex process of reviews have tried the persistence
of none-too-enthusiastic local officials. As Holden
observed in 1966, water quality in both the permit and
treatment plant construction activities have always
involved a substantial amount of bargaining between
polluter and enforcement official. This is a type of
politics that fits neither the distributive nor the
redistributive models but still a third general theory of
politics which Lowi calls regulatory politics. It can apply
to much more than the public activities usually
designated as regulatory. Changes in both water
development and water quality management may
increase the significance of this type of politics and
with it change the kinds of information, including
benefit estimates.
The essential features of regulatory politics are
captured in the pluralist tradition of political science.
Policy is the result of group conflict and the groups are
large and well organized. It is not the result of log
rolling by many small groups who have nothing else in
common but groups whose interests collide. It is not a
question so much of colliding values which in our
system may go forever without being resolved. Rather
one group cannot continue to enjoy its values unless it
can achieve an accommodation by another group.
Rules for accommodation tend to be broad and give the
appearance of inflexibility. Examples include "zero
discharge" and "non-degradation." Subsidies are more
openly identified. Leadership and coalition members
may be too unstable to fit the term of an elite in the
political authority sense. Bargaining, mediation,
agreements, and acceptability characterize an em-
phasis on process. Information on the benefits and
costs enjoyed by different groups might become grist
for the mill rather than symbolic accommodation of a
general value or ideology.
But water resources management is a localized and
sectionalized phenomenon. The focus is on the lake
and the watershed and the associated communities.
Also for more effective future management, many of
the functions jealously performed by local governments
will need to be used — land use controls are a case in
point. The distributive politics involved will still
dominate at the national level. This suggests that the
scope for expanding regulatory politics is at the local
level to achieve consent and agreement that can be
transmitted to the national level.
Where distributive politics are hamstrung by local
conflict, the search for a broader coalition should look
attractive. The key step will be in avoiding early
commitment to particular means to solve problems. But
this will require moving away from the presumption
that Federal money will be available only for dams,
channel works, sewers, and treatment plants except
where mitigation and similar bargaining yield funds for
fish and wildlife and recreation facilities. Broader
-------
176
RESTORATION OF LAKES AND INLAND WATERS
access to alternative means will attract new support
groups and encourage accommodation to environ-
mental interests.
A planning process that emphasizes identifying more
of the benefits earlier — even before they can be
refined to fit the specifics of particular options —
suggests that conflicts may surface while they can still
be accommodated in the planning process. If no conflict
arises, distributive politics can proceed as usual. If it
does and no accommodation appears possible, the
unsatisfied interests — whether because of deeply felt
value conflicts or otherwise — will have received a
more obvious application of political due process.
Planners will have a better chance to display
accommodations which may still not be acceptable to
conflicting ideologies but which others find acceptable
in their behalf. Remember the test for willingness to
pay is not that the beneficiary is indeed willing, but that
reasonable people agree that the beneficiary should be
willing.
Ogelsby, R. T. and D. J. Allee, ed. 1969. Ecology of Cayuga
Lake and the proposed bell station. Publ. 27, Water Resour.
Mar. Sci. Center, Cornell University, Ithaca, N.Y.
Romm, J. 1969. The value of reservoir recreation. A.E. Res.
296. Dep. Agric. Econ. Cornell University, Ithaca, N.Y.
REFERENCES
Allee, D. J. 1980. Education techniques and planning for
water resources development, Phase III. Dep. Agric. Econ.,
Cornell University, Ithaca, N.Y.
Allee, D. J. and H. M. Ingram. 1 971. Interview for the Natl.
Water Comm. Denver, May.
Allee, D. J. and B. T. Osgood. 1980. Housing and private
human costs of floods in the Tug Fork river valley. Inst.
Water Resour. U.S. Army Corps Eng.
Batie, S. S., and L. Shabman. 1979. Valuing nonmarket goods
— conceptual and empirical issues: Discussion. Am. Jour.
Agric. Econ. December: 931.
Bishop, R. C. and T. A. Heberlein. 1979. Measuring values of
extramarket goods: Are indirect measures biased? Am.
Jour. Agric. Econ. December: 926.
Dobrowolski, F., and L. Grille. 1 977. Experience with the 303-
208-201 study relationships. Water Resour. Bull. 13:455.
Freeman. A. M., III. 1979. Approaches to measuring public
goods demands. Am. Jour. Agric. Econ. December: 915.
Hinman, R. C. 1969. The impact of reservoir recreation on the
Whitney Point microregion .of New York State. A.E. Res.
295. Dep. Agric. Econ. Cornell University, Ithaca, N.Y.
Ingram, H. M. 1977. The changing decision rules in the
politics of water development. Pap. 72106. Water Resour.
Bull.
Ingram, H. M. and D. J. Allee. 1975. Balancing national and
regional objectives: The shifting information requirements.
Staff Pap. 75-5. Dept. Agric. Econ. Cornell University,
Ithaca, N.Y.
Krutilla, J. V., and A. C. Fisher. 1976. The economics of
natural environments. The Johns Hopkins University Press,
Baltimore, Md.
Lord, W. B. 1979. Conflict in Federal water resource
planning. Water Resour. Bull. 15:1226.
Lowi, T. J. 1964. American business, public policy, case-
studies and political theory. World Politics XVI.
1966. Distribution, regulation, redistribution: The
functions of government In R. B. Ripley, ed. Public policies
and their politics. W. W. Norton and Co., Inc. New York.
Mann, D. E. 1 973. Political incentives in U.S. water policy: The
changing emphasis on distributive and regulatory politics.
Paper for Int. Political Sci. Assoc. August.
-------
177
CLEAN LAKES ESTIMATION SYSTEM
NEILS B. CHRISTIANSEN
U.S. Environmental Protection Agency
Corvallis Environmental Research Laboratory
Corvallis, Oregon
ABSTRACT
In administering the Clean Lakes Program the U.S. Environmental Protection Agency, Office of
Water Regulations and Standards has taken a number of steps to assure efficient management.
Those steps include establishing a regulation for administration, seeking additional funds to
support the Program and establishing a Program Strategy. This paper describes a part of that
strategy: the Clean Lakes Estimation System and its role in the Clean Lakes Evaluation System.
This combines Federal, State, and local decisionmakers with an information system which can be
used to estimate the various ecological and human consequences of a 314 project and human
value of all those consequences. The Estimating System being developed will provide information
for use in evaluating the results of the entire Clean Lakes Program and also in choosing projects
for funding. The components of the Estimation System are: (1) Procedures to estimate the
limnological and other ecological outcomes of various treatments; (2) procedures to estimate the
various human impacts of a 314 project; (3) procedures to identify the value of the various impacts
in terms of standard economic values; (4) procedures to identify the value of all consequences in
terms of stated local, State, and Federal goals.
In administering the Clean Lakes Program, the U.S.
Environmental Protection Agency, Office of Water
Regulations and Standards, has taken steps to help
assure efficient management: establishing a regulation
for administration (Code Fed. Reg., 1980), seeking
additional funds to support the program, and establish-
ing a Clean Lakes Program Strategy (U.S. EPA, 1980).
This paper describes an essential component of that
strategy. An improvement in the Clean Lakes Evalua-
tion System which we term the Estimation System.
First is a general description of the proposed Estimation
System and how it will fit into the total Evaluation
System.
EVALUATION SYSTEM
The Clean Lakes Evaluation System has a dual
purpose: (1) To select, in as wise a way as possible, the
best restoration or protection projections for funding
under section 314; and (2) to provide continuous
evluation of the entire Clean Lakes Program for use in
improving the Program and in assessing its benefits.
The Evaluation System is used to select projects for
funding by identifying project goals and forecasting the
degree to which those goals will be met. Projects which
promise success through high goal attainment are the
best candidates for funding.
The results of the Clean Lakes Program are evaluated
by ascertaining the degree of goal attainment in
completed projects. Thus, the System evaluates
proposals for funding by estimating the likely results. It
evaluates the Program by estimating actual results.
Figure 1 shows the elements of the Evaluation
System and their interrelationships. A 314 project
proposal arises because some decisionmakers* (Figure
GOALS
I
I
I
I
i_
I
ACTION
I
DECISION
MAKERS
I
REVIEW
I
IMPLEMENTATION
I
PROJECT
SELECTION
MULTIOBJECTIVE
ANALYSIS
1
ECONOMIC
MEASUREMENT
1
HUMAN
IMPACTS
,
ENVIRONMENTAL
IMPACTS
~l
1
1
1
J
-|
OTHER
INFORMATION
OTHER
SIGNIFICANT
IMPACTS
i
HU
CHARAC1
MAN
[ERISTICS
ENVIRONMENTAL
CONDITIONS
-ESTIMATING SYSTEM-
Figure 1. — Clean Lakes Evaluation System.
1: top, center) judge some current and potential
environmental conditions (bottom, right) to be change-
-------
178
RESTORATION OF LAKES AND INLAND WATERS
able and inconsistent with goals, i.e., undesirable. The
actions taken to carry out the project, combined with
existing environmental conditions, lead to various
environmental impacts. The environmental impacts
propagate through webs of cause and effect until they
lead to some human impacts, including desirable
changes (positive impacts) in at least some of the
conditions which led to the 314 project proposal. There
may also be impacts (either positive or negative) which
have not been anticipated. The nature and magnitude
of the human impacts are affected by the human
characteristics involved. For example, urban dwellers
often differ from their rural cousins in attitudes about
natural resources.
EnvironmentaJ and human impacts are often so
numerous that we are forced into some degree of
abstraction when identifying those to consider in the
evaluation process. In principle, we choose those
impacts which are of greater significance. Significance
is based on the number of people impacted and the
enormity of the impact on each individual. A guide to
the significance of the various impacts can be found in
the goals of a 314 project, for those goals led to the
proposed actions.
To date, identifying goals, determining human
impacts, and evaluating the positive and negative
aspects to arrive at a good project design and then
judging the desirability of the project has been a
qualitative process, and in many cases implicit rather
than explicit. Thus, in terms of Figure 1, the evaluation
of a proposed project has tended to have no economic
measurement or multiobjective analysis. Most impacts
are of the other significant impacts sort and outcomes
are weighed in qualitative terms. A project design is
evolved and that design is analyzed to decide whether
or not to select the project for funding, if so,
implementation occurs with various results which,
through review, provide feedback to the decision-
makers. This feedback is combined with other
information to either reaffirm or modify goals for
subsequent proposals.
ESTIMATION SYSTEM
The Clean Lakes Evaluation System is composed of
two major parts. One consists of the various
decisionmakers at the Federal, State, and local levels
involved in designing, selecting, and implementing 314
proposals, as well as those who administer, fund, and
otherwise influence the entire Clean Lakes Program.
The second part is made up of the information systems
used by the decisionmakers to obtain estimates of the
outcomes (either expected or realized) of individual
projects, or of the Program as a whole.
The Office of Water Regulations and Standards
intends to make certain portions of the information
systems explicit to improve both the quality of the
information and the ease of its communication to
various decisionmakers. This will improve the office's
evaluation of proposed projects, and thereby its
selection process, as well as provide a more objective
basis for funding and improving the entire Clean Lakes
Program. The improvements all relate to the bottom
four boxes, and their associated arrows, in the center
column of Figure 1 and their duplicates in review.
Accordingly, the improvements will include procedures
to estimate:
1. The limnological and other ecological impacts of
various treatment possibilities in a 314 project.
2. The various human impacts of a project.
3. The value of the various impacts, where possible,
in standard economic units, i.e., dollars.
4. The value of as many as possible of all the impacts
in terms of stated local, State, and Federal goals
through a multiobjective analysis.
Together these four sets of procedures are termed
the Clean Lakes Estimation System, as shown in Figure
1. The Estimation System is intended to be com-
prehensive in two dimensions. One, which we might
term the vertical dimension, includes the whole series
of impacts stemming from restorative (or protective)
action to ecological impact to human impact to the
values of those impacts. The second, or horizontal,
dimension includes all human values affected by a 314
project and the ecological and human impacts which
impinge on those values. Comprehensiveness in the
vertical dimension is necessary to insure that
evaluations are made in terms of human values, and
that only changes in these values which can be traced
back through the social and ecological systems to the
actions undertaken are credited to the project.
Comprehensiveness in the horizontal dimension is
necessary if myopia is to be avoided and the spirit of the
National Environmental Policy Act (42 USC 4321 ) and
the Principles and Standards of the Water Resources
Council (Fed. Reg., 1973) and other statements
regarding the management of-public resources are to
be followed. This is particularly necessary because
many 314 projects include a wide variety of treatment
activities both in a lake and throughout the watershed.
However, some impacts and values will always be
unknown and some which are known will not be
contained in formal estimating procedures. Therefore,
the comprehensiveness of the Estimation System, like
all other aspects of the Clean Lakes Program, will be
continually subject to review and modification.
ESTIMATING IMPACTS
The first step in developing the Estimation System
will be to develop a list of significant environmental and
socioeconomic impacts which are anticipated from
various types of 314 projects.
For each type of significant impact, a procedure will
be developed for estimating the magnitude of that
impact on the basis of certain, presumably causal,
factors. The procedure will, in essence, be a series of
equations in which the impacts of concern are the
dependent variables and the causal factors are the
independent variables. To illustrate:
suppose S = b0
P - b2A
* The elements named in Figure 1 are italicized when first used in the
text.
-------
PUBLIC BENEFIT AND INSTITUTIONAL PROBLEMS
179
where S = number of swimmers using a lake
P = population of the community
A = concentration of algae in the lake
bi =various coefficients developed through
analysis or otherwise
To employ such an equation requires values of the
independent variables (P and A). The variable A is the
link back to the 314 project, for presumably A is to be
reduced. The pre-implementation value Ao, leads to an
estimate So,, and the post-implementation value, A-i,
leads to an estimate Si. Of course, the value of P may
also change. The change in S(Si-So) is therefore one
of the human impacts of the restoration.
To develop precise and cost effective estimating
procedures, equations must be formulated on a
regional basis with certain independent variables
included which allow the prediction to be fitted to a
given 314 project. Using the illustration presented
earlier, let the number of swimmers be determined as
follows:
S = bo + biP - b2A + b3U
where S, P, A and the B are defined as before, and U is
a variable representing urbanization.* If city dwellers
have a significantly different attitude toward swimming
than rural dwellers, than U will be important; including
it makes it possible to determine one equation for a
whole region which can be applied with precision to a
given locality.
VALUES
As described so far, the System yields estimates of
the impacts of a 314 project. Some of these impacts
can be validly measured in economic terms, some
cannot. Some values measureable in economic terms
are costs of restoration, tax impacts, property value
changes, business activity, some damages caused by
floods, etc. These impacts will occur in ordinary
markets and so long as these markets are relatively
well organized they will reflect the values involved.
Other values not explicitly identified in a market can
still be measured in economic terms by virtue of their
close relationship to a market. Two notable examples
are benefits to recreationists and costs of modifying
farm management practices. The field of recreation
economics has devised a variety of tools for estimating
the value of recreation based on a proxy price, such as
travel cost. Modified farm management practices, if
instituted on a wide scale, will ultimately be reflected in
the market prices of farms. In the meantime, farm
management models can be used to estimate the
impacts on farm profits. Profits, like travel costs, are the
market values which can be used to estimate the value
of a closely related impact.
Finally, some impacts are so removed from ordinary
market processes that economic valuation is not
possible. Economic measures of the value of items
such as community cohesion, education, and research
often have little validity. However, one must be careful
* Urbanization is a qualitative concept. An acceptable quantification
will need to be determined. Simple formulations such as population
density or city size may be adequate, or more complex formulations
may be needed.
in making such judgments. In a specific context it may
be possible to derive a valid, useful economic
measurement of almost anything. Consider a human
life: Can its value be expressed in economic terms? In
general, I'd say no. Such approaches as discounted
earnings or life insurance carried, etc., reflect only
poorly and partially the contribution of a person to his
child or to society. But, in certain specific situations we
can measure the value society implicitly places on
human life. Consider the case of plane passengers: By
computing the costs of safety regulations and the
number of deaths per million passenger miles we can
derrive an implied value per life. This value can be
useful to decisionmakers concerned about possible
modifications in safety regulations, particularly when
compared to other modes of transportation and other
types of activity.
The boundary between values which can, and those
which cannot, be correctly measured in economic
terms is fuzzy since it depends upon the decision being
contemplated and how the measurement will be used.
This is another way of saying that a correct economic
valuation procedure depends on the goals involved. The
value of a human life appropriate for determining
public safety regulations is quite possiblly different
from one appropriate to a damage suit involving
negligence. And neither is likely to adequately measure
the value to one's child. For this reason, Figure 1 shows
goals to be a determinant of economic measurement in
addition to their role in determining the significance of
human impacts. It is also true that over time, certain
categories of value become measurable in economic
terms as the field advances: Fifty years ago economic
valuation of publicly provided recreation benefits would
have been impossible.
EPA intends to identify the value of those impacts
which cannot be correctly obtained in standard
economic terms through a process called decision
analysis by some and multiobjective analysis by others
(Keeney and Raiffa, 1976). The values would be
obtained through the goals and utility functions of
decisionmakers. The utility functions specify the
degree of goal attainment associated with various
levels of some objectively measurable variables. The
measurable variables are those impacts estimated by
the System as described previously.
Consider the following simplified example. Suppose
two goals exist for a 314 project:
1. Maximize net financial worth of recreation
benefits minus project costs.
2. Maximize wildlife diversity and abundance
through the use of wetlands as nutrient sinks to create
new and diverse habitats.
These two goals cannot be met simultaneously since
maximizing goal 2 implies a very large budget which
would lead to a less than maximum value for goal 1.
Assume the decisionmakers judge goal 1 to be three
times as important as goal 2. Suppose the utility
functions for each goal are as shown in Figure 2. If a
particular project design promises a net financial worth
of $1.5 million (utility .9) and a wetlands budget of
$250,000 (utility = .5) the value (V) of the project V =
3(,9) + 1(.5) = 3.2. This magnitude of V could be
compared with that of other designs to pick the best
-------
180
RESTORATION OF LAKES AND INLAND WATERS
design and with that of other projects to select the best
projects for funding.
I.2
I.O
0.8
0.6
0.4
0.2
0.5 I.O 1.5
NET WORTH
(million dollars)
2.0
1.2
I.O
0.8
0.6
0.4
0.2
0 IOO 200 300 400
WETLANDS BUDGET
(thousand dollars)
Figure 2. — Utility associated with each of two goals.
SELECTION AND REVIEW
The Estimation System can also be used any time
after implementation to see the degree to which goals
were actually met. The pre-implementation and post-
implementations results could differ if any of three
things occur: Errors in estimation, shifts in value
structure, or inadequate management.
Consider the example presented earlier involving two
goals: Financial worth and wildlife. Suppose the design
having a value of 3.2 was implemented. Suppose upon
review a value of only 2.0 resulted. Suppose further the
difference in value was due to two sources: A net worth
of only $1.0 million occurred instead of the $1.5 million
anticipated, and the wildlife utility function fell. The
lower net worth might be traced to higher than
expected costs or lower than expected benefits of some
specific type. The downward shift in the wildlife utility
function might reflect the fact the local populace
considered the wetland habitat to be less useful than
they had anticipated. This, in turn, might result from a
number of causes. Perhaps the area was insufficiently
accessible. Perhaps the community had been initially
oversold on prospective benefits and the shift in utility
was a reflection of reality. Or, perhaps a public
education program was needed for the community to
observe and appreciate the wildlife impact.
The pre- and post-implementation values of this
project could also be compared with those of a number
of other projects. Do they all show a decrease between
anticipated and actual value? If so, perhaps the
program needs to be reduced. Do some projects show
large increases and others large decreases? If so, this
may indicate the presence of some important variables
not presently accounted for in the Estimation System.
Some such variables can be analyzed for inclusion.
Others may be identified, but accounted for in only a
subjective way.
One such variable is management. For instance, a
successful implementation may require effective
management of a lake district or the coordination of
several local, State, and Federal agencies. If this
effective management or the necessary institutional
framework is absent, a well planned restoration may
not succeed. Therefore, it will be necessary to ascertain
what institutional or administrative factors are asso-
ciated with successful and unsuccessful projects.
Since no plan is constructed with perfect foresight, a
deviation does not necessarily indict management.
What counts is whether or not the final result is
considered to have been worth the effort. In either
case, there is something to be learned from the process
which can help future implementations succeed.
It is clear the review process is exceedingly
important. It can detect errors in the Estimation
System, ascertain shifts in value structures, and
identify important new variables, for specific attention
in the evaluation process.
INFORMATION
A point worth emphasizing is that the Estimation
System, as envisioned, is a set of procedures to
estimate what would occur and what did occur in the
ecosystem and social system, given certain conditions
and actions. Therefore, the Estimation System does not
make decisions, it generates data for use by
decisionmakers. The System is confined to answering
questions about what would (or did) happen if
This is obvious in the estimation of environmental and
human impacts. It is equally true in economic
measurement and multiobjective analysis. These last
two items estimate values which occur given certain
conditions and actions. Weighing the various alter-
natives to finally arrive at project selection or Program
modification requires additional inputs. In Figure 1
these inputs are symbolized by the box "other
significant impacts" which include not only human
impacts and values unaccounted for in the Estimation
System, but also judgments by the decisionmakers of
the validity and precsion of the various components of
the Estimation System.
From the illustrations given, the Estimation System
may appear to be totally quantitative. This need not,
and hopefully will not, be so. Estimates may just as well
be qualitative as quantitative. Indeed, at any one
moment in time, qualitative estimates may contain
more information than is possible with currently
available quantitative technology. The Estimation
-------
PUBLIC BENEFIT AND INSTITUTIONAL PROBLEMS 181
System is intended to be an information generating
system, not an information destroying system. There-
fore, if comprehensiveness is to be obtained in the
vertical and horizontal dimensions as stated earlier, the
System will inevitably contain qualitative elements.
However, it is also true that since quantitative
information is easier to transmit and transform than
qualitative information, there will be a tendency to
develop quantitative expressions wherever appropri-
ate. This leads to three recommendations for those
developing and using the Estimation System:
1. Those who develop quantitative information
should be humble enough to state what their numbers
do not contain. This includes error terms in the
statistical sense. It also includes statements about
omissions resulting from the way the problem was
formulated to make it amenable to quantitative
analysis.
2. Those who develop qualitative information should
be concise — not quantitative, but concise — if they
don't want their information treated like excess
3. Users of the Estimation System should view it
with a jaundiced eye: Its output is probably in error. The
need is to know where those errors lie, how significant
they are, and how to compensate for them.
REFERENCES
Code of Federal Regulations. 1980. Cooperative agreements
for protecting and restoring publicly owned freshwater
lakes. Part 35, Subpart H.
Federal Register. 1973. 38:24778. September.
JACA Corp. 1980. Economic benefits assessment of the
section 314 Clean Lakes Program. Fort Washington, Pa.
Keeney, R. L, and H. Raiffa. 1976. Decisions with multiple
objectives: Preference and value tradeoffs. John Wiley and
Sons, New York.
U.S. Code. 42 4321.
U.S. Environmental Protection Agency. 1979. Limnological
and socioeconomic evaluation of lake restoration projects:
Approaches and preliminary results. EPA-600/3-79-005.
Environ. Res. Lab. Corvallis, Ore.
1980. Clean Lakes Program strategy. Criteria
Stand. Div. Washington, D.C.
-------
182
IMPACTS OF LAKE PROTECTION ON
A SMALL URBAN COMMUNITY
NICOLAAS BOUWES, SR.
Economics, Statistics, and Cooperatives Service
U.S. Department of Agriculture
and
Department of Agricultural Economics
University of Wisconsin-Madison
LOWELL KLESSIG
Environmental Resources Unit
University of Wisconsin-Extension
Madison, Wisconsin
STEPHEN LOVEJOY
Department of Agricultural Economics
Purdue University
Purdue, Indiana
PETER CAULKINS
DOUGLAS YANGGEN
Department of Agricultural Economics
University of Wisconsin-Madison
ABSTRACT
The Waupaca City Council was the first municipal body in Wisconsin to form a lake management
district to take advantage of new legislation and funding for lake cleanup. Efforts to manage Mirror
and Shadow Lakes began in the 1960's when it became apparent that the two small lakes located
within the city limits were experiencing water quality problems. Research revealed that excessive
nutrients were entering the lakes through storm sewers. In 1975 the Waupaca Lake District
requested State and Federal financial aid for a stormwater diversion project, and became one of
the first awardees under the Clean Lakes Program. An evaluation grant accompanying the EPA
implementation grant examined hmnological, sociological, and economical impacts of such a
project. The evaluation of the sociological impacts examined the effects on individuals, groups, and
local government. The project generated only mild sociological benefits beyond those associated
with water recreation itself. In the economic evaluation it is the critical explanatory variable, water
quality, that must be accounted for in the relevant models, such as predicting the impacts on
property values and the recreation response. The methods used in estimating these impacts and
the interpretation of the results obtained from these models are presented in this paper.
INTRODUCTION
Under the Clean Lakes provision of Public Law 92-
500, the Environmental Protection Agency embarked
on a major program of cost-sharing grants to
implement lake restoration projects. Since requests for
financial assistance exceed available funds, an evalua-
tion of project impact is crucial to sound decisions on
future applications. Both potential grantees and EPA
need to know how past efforts have fared. Project
justification, optimal level of implementation, and
relative priority of individual projects depend on such
evaluations. Public investment decisions can best be
made if potential impacts can be predicted based on
systematic evaluations of the following procedures:
1. Limnological evaluation to determine whether
water quality has been improved (or maintained);
2. Economic analysis to determine monetary benefits
relative to investment; and
3. Sociological assessment of non-monetary impacts
on individuals and groups.
In this paper we are primarily concerned with the
economic and sociological impacts of the Mirror/
Shadow Lake Project in Waupaca, Wis.
Waupaca
The city of Waupaca has experienced a steady
population increase since 1960, exceeding the average
growth rate for rural communities which reflects a
nationwide shift away from urban areas (Waupaca
County Outdoor Recreation Planning Committee,
1978). The current population of the city is approxi-
mately 5,000.
-------
PUBLIC BENEFIT AND INSTITUTIONAL PROBLEMS
183
During the summer, the area's population more than
doubles as second-home owners and vacationers
immigrate to Waupaca and the surrounding chain-of-
lakes region. These summer residents participate in
pleasure boating, canoeing, swimming, hiking, picnic-
ing, and scenic driving. They contribute an estimated
$1,050,000 a year to the local economy (Cooper and
Powers, 1976). Clearly, the maintenance of environ-
mental quality, especially the water resources, is
critical to the continued economic well-being of the city
of Waupaca and surrounding areas.
Mirror/Shadow Lakes
These small lakes, of 5 and 17 hectares respectively,
are located within the city of Waupaca. South Park, on
the municipally-owned west shore of Shadow Lake,
provides the only swimming beach in the immediate
area. The park is heavily used for picnicking as well as
swimming. In 1978 the attendance at South Park was
estimated to be 83,809 local users, with nonresidents
accounting for an additional 10 to 15 percent of that
amount (Bouwes, et al. 1980). Both lakes began
experiencing algae problems and dissolved oxygen
depletion in the late 1960's. The quality of Shadow
Lake was still acceptable, but the drainage of low
quality Mirror Lake water into Shadow Lake concerned
lake users.
In response to that concern, Waupaca created a lake
district in 1974, the first year the Wisconsin Lake
Management Law went into effect providing for such
local management units (Chapter 33, Wisconsin
Statutes). Studies revealed that most of the phos-
phorus entering the lakes could be traced to storm
sewers emptying directly into them. With technical
assistance from the Wisconsin Department of Natural
Resources, the lake district proposed a three-phase
project to deal with the problem:
1. Eliminate most of the phosphorus loading by storm
sewer diversion;
2. Treat the lakes with alum to precipitate the
phosphorus in the water column and seal off the
phosphorus-rich sediment; and
3. Aerate Mirror Lake to promote turnover since
natural turnover by wind action is inhibited by its depth
(13 meters) and sheltered location.
The storm sewers were diverted in 1976, alum was
added in 1977, and aeration began in 1977 at a total
cost of approximately $430,000. This cost was shared
by EPA (50 percent), DNR (30 percent), and the local
lake district (20 percent).
LIMIMOLOGICAL IMPACTS
Limnological evaluations, which have been con-
ducted since 1977, have revealed that phosphorus
levels have dropped in both lakes and oxygen levels
have increased to again support fish life in Mirror Lake.
Water clarity has not improved. Oscillator/a rubescens,
a blue-green alga which lives deep in the water
column, has been replaced by green algae which are
characteristic of less eutrophic lakes and support a
better aquatic food chain, but they also grow closer to
the surface where they are more visible.
ECONOMIC IMPACTS
A thorough analysis of the economic impacts of a
project should include both allocative efficiency and
distributional equity considerations. The efficiency
issue examines whether the reallocation of resources
to the project, e.g., those used for water pollution
control, increases the net value of the output produced
by the resources. Ideally, one would wish to determine
not only if the resources had been optimally allocated
among alternative uses, but whether they are optimally
allocated for a given project. The equity issue examines
the welfare redistribution associated with a project;
that is, the distribution of the project benefits and costs.
Efficiency
One of the common tools employed to examine
project impacts is a benefit-cost analysis. Such an
analysis seeks to answer the question: "Are the
benefits, i.e., increases in welfare, generated by a
project greater than the costs necessary to realize the
project?" In the case of a project with water quality
impacts, it is necessary to employ a methodology which
will allow values to be imputed to water quality as a
nonmarket good, since the market fails to provide
prices (values) directly.
Economic theory and earlier research indicate that
the benefits associated with the Mirror/Shadow Lakes
improvement project will be capitalized in the
surrounding property values. Consequently, a property
value model was used to estimate these impacts
(Dornbusch, 1974).
The basic premise of this model is that water
resource projects have value to the general public
which, in the absence of a market for direct sale of this
output, are adequately reflected in the market prices of
those properties situated near the resource.
When a public project enhances productivity or
utility, benefits accrue to the affected firms and
households. These benefits increase the value of
certain locations, and as a result, the initial equilibrium
in land markets will be perturbed. Eventually, new
equilibrium land values are established. The total
benefits from such a program equal the sum, over all
firms and households, of the changes in productivity
and utility, and, therefore, are equal to the sum of the
changes in land values from the initial equilibrium to
the new equilibrium.
The empirical model postulates first that a change in
property values is due to perceived changes in water
quality by area residents, and secondly, that the impact
on property values decreases as distance from the lake
increases. These two aspects are incorporated into the
following equations:
a. PWQIExP= I akBi)k
b. PWQIfles = -24.778+ 0.463 (PWQIExP) + 15.5 (Public
Access)
c. bi = e6'398 (PWQIpes) 0.492 e1'180 (WBT Lake)
e0'991 (WBT Bay)
-------
184
RESTORATION OF LAKES AND INLAND WATERS
1
(DWmax)
Equation (a) determines the experts' perceived water
quality index (PWQIsxp) which represents a limnol-
ogist's ex ante estimates of how much each of seven
different water quality parameters would change both
with and without the project. These seven parameters
are (1) industrial wastes in the water, (2) debris in or on
the water, (3) clarity of the water, (4) algae in the water,
(5) odor from the water, (6) wildlife support capacity of
the water body, and (7) the recreational opportunities
affected by the water level. BNk reflects the change from
water quality condition i to water quality condition j for
the kth parameter. The relative importance of each of
the seven parameters is represented by the weighting
factor 3k.
Equation (b) expresses the perceived water quality
index rating by residents (PWQIRes) which is a linear
function of the expert's perception of water quality and
the degree of public access available at the lake.
Consequently, by being able to predict residents'
reactions to a given water quality change, this equation
provides a vital link which allows for an ex ante
evaluation.
Equation (c) is used to determine coefficient b which
is a function of the residents' perceived water quality
index and whether the water body type is a lake or bay.
Equation (d) determines the constant term b which
serves the function of making the change in property
values equal to zero at the outermost limit of the area
impacted by the project. Equation (e) represents the
mathematical expression of the model's relationship
where the percentage change in property values
(AP%d) is a function of both the perceived water quality
changes by residents as predicted by experts (embodied
in bi) and the average distance from the water zone d
(1/DWd).
The period of project analysis was determined to be
34 years, 1976-2010; water resource experts estimate
this to be the longest time period for a positive (with
project) or negative (without project) change in water
quality to occur either on Mirror or Shadow Lakes. With
a limnologist's predictions of the status of the seven
water quality parameters for each year, equations (a-e)
are calculated, and the incremental, annual percentage
change in property values is determined for different
distances from the lakes.
To simplify the calculations, the impact area around
the lakes was divided into separate distance-from-the-
lake zones. Since Shadow Lake has ample public
access, all residential property values throughout the
City of Waupaca were assumed to be impacted by the
project. Non-residential property was excluded (Lind,
1973). Ten distance-from-the-lake zones were con-
structed emanating out from Shadow Lake. For Mirror
Lake only one distance-from-the-lake zone was
constructed. As there is little public access, only
lakefront owners were assumed to benefit from
improvements in that lake's water quality.
The direct project benefits are calculated by
multiplying the incremental percentage change in
property values for each distance zone by the
corresponding sum of property values in that zone; this
is done for each year for both lakes. However, these
benefits are spread over the entire life of the project,
and to compare this stream of project benefits with the
time stream of project costs, each must be reduced to a
single number — their present value. The discount rate
is the crucial parameter in this calculation. There are
numerous, conflicting schools of thought regarding the
appropriate discount rate. We used two rates to bi acket
this range: 7 1 /8 percent and 15 percent, which reflect
the rate suggested by the Water Resources Council to
discount Federal projects and the opportunity cost of
capital in the private sector as approximated by the
prime lending rate, respectively.
The present values (1977 dollars) of project benefits
and costs were used to determine the benefit-cost ratio
for the project. If the ratio is greater than one, the
present value of discounted project benefit exceeds
that of discounted project costs and the project has met
at least a minimum standard of economic efficiency.
Discounted project costs were $439,872 and $469,650
for 7 1 /8 percent and 15 percent discount rates, respect-
ively. Discounted project benefits were $1,049,269
and $833, 958 for 7 1/8 percent and 15 percent
discount rates, respectively. The two benefit-cost ratios
generated in this study result from the sensitivity
analysis performed with respect to the discount rates
used. The corresponding benefit-cost ratios are 2.385
and 1.776 with 71/8 percent and 1 5 percent discount
rates, respectively. These results indicate that regard-
less of the discount rate the project is justifiable using
economic efficiency criteria.
Equity
Benefit-cost ratios address the issue of allocating
scarce resources in an efficient manner. However, the
preceding analysis ignores equity considerations
regarding the distribution of benefits and costs. There
are several equity considerations involved with the
Mirror/Shadow Lakes Project.
We have assumed that the Federal, State, and local
cost shares have been appropriately determined and
we will only examine distribution of the local cost
share. There are basically two approaches for
subsidizing the satisfaction of public wants — the
"ability to pay approach" and the "benefits received
approach." The former is typically implemented
through taxes on property and the latter by special
assessment taxation in proportion to benefits received.
Local revenues for this project were raised by levying
a 2-year 0.9 mill rate tax on all real property in the lake
district (city). Each property owner then paid for a
portion of the project according to his/her assessed
property (e.g., approximately $90 for a $50,000 home).
However, the well-being that each residential property
owner enjoys as a result of the project does not vary in
the same manner as the amount of taxes each had to
pay. One reason for this discrepancy is that the
increases in residential property values, because of a
perceived improvement in water quality, diminishes as
distance-from-the lake increases. An example of the
discrepancy between the distribution of project
benefits and costs can best be demonstrated by
examining the Mirror Lake properties. The analysis
-------
PUBLIC BENEFIT AND INSTITUTIONAL PROBLEMS
185
reveals that 31 percent of the benefits accrued to these
properties; however, only 5 percent of the costs were
paid by these property owners.
If local financing was meant to be distributed on
benefits received basis, then instead of a uniform mill
rate being levied on all property owners, a special
assessment based on a graduated rate could be
implemented to reflect diminishing property value
benefits for homes farther away from the lakes. Those
homes that have lake frontage on Mirror Lake could be
levied an even higher special assessment to reflect the
exclusive benefits they enjoy from the improvements in
that lake's water quality. And if financing was meant to
be distributed on an ability-to-pay basis special
consideration should still be given to the expected flow
of benefits as the higher valued properties in this
instance are also the ones to benefit the most since
they are the ones located in the zones closer to the
lakes.
SOCIOLOGICAL IMPACTS
Many public projects, especially large Federal
projects, have been criticized as being insensitive to
human needs. Decisions to undertake a project are
often based on narrow economic criteria. While
economic benefits are calculated to be greater than
economic costs, the social costs to the residents and
community are often greater than the social benefits
(Dixon, 1978). The controversy surrounding such
projects is the result of inadequate attention to social
impacts.
Such a controversy did not develop before, during, or
after the lake project in Waupaca; the social impacts
analyzed were neutral or positive.
Citizen Participation
Since a city council can create and operate a lake
district under Wisconsin law (Klessig, 1979), the
Waupaca City Council could act without a petition from
landowners and without extensive involvement by
citizens. The city could also use its administrative staff
to implement the project and supervise contractors.
The minimal citizen participation is shown by
comparing attendance figures at annual meetings in
Waupaca with those of a similar project in a small
population rural setting at White Clay Lake. Only 2
percent of Waupaca residents attended a lake district
annual meeting. In contrast, half of the White Clay Lake
residents attended such meetings. In the rural area of
White Clay Lake, without an incorporated local
government, 14 percent attended as many as eight
annual and special meetings (Klessig and Lovejoy,
1980).
Environmental Understanding
To determine the impact of the project on the
knowledge level of citizens, a series of questions on
lakes were asked of the Waupaca sample, the White
Clay Lake property owners, and a statewide control
group. Table 1 shows that Waupaca residents scored
very close to the statewide average. They scored
highest on a storm sewer question — one directly
related to their project. Beyond that specific issue,
there appeared to be little increase in knowledge about
lakes. On the other hand. White Clay Lake residents
generally scored substantially above the State average.
This difference may reflect the greater participation by
White Clay Lake citizens. Rural location and farm
occupations may have also contributed to greater
knowledge of the lake ecosystem.
Community Cohesion and Development
in many situations, a local community's involvement
with large projects yields valuable experience in terms
of personal familiarity with granting agencies, know-
ledge of technical and financial assistance, and
assertiveness in dealing with bureaucrats. In other
cases it yields frustration, bitterness, distrust of
government and unwillingness to participate in future
programs. Thirty-four percent of Waupaca residents
felt the project experience would be useful to Waupaca
in the future. Most of the remaining residents were not
aware of the project.
There was little evidence that residents of Waupaca
felt the projecit had damaged community development
or community cohesion. At no time did the project open
or reopen wounds between segments of the com-
munity. The lake project was not the type of
development that pitted old against young, newcomers
against traditional families, or developers against
environmentalists. When asked whether the lake
management project made their community a more or
less desirable place to live, a majority felt that their
community was a desirable place before, and the
project had not affected that. Thirty-six percent felt the
project had made Waupaca more desirable. No
respondent felt the project had had a negative impact.
Table 1. — Educational impacts of lake projects in percent correct
responses. Italicized words are correct answers.
Waupaca White Clay Statewide
1. City and village storm drains
can empty into nearby lakes
without hurting the quality
of lake water.
Agree or disagree! 73 80 69
2. The major cause of lake fish
dying — or fish kills — in the
winter months is that the
water gets too cold for the
fish to live.
Agree or disagree) 70 97 72
3. If farmers near lakes
fertilize their fields by
spreading manure only in the
winter, the amount of
pollutants running to the
lakes would be reduced.
Agree or disagree? 45 71 49
4. Marshes around lakes act as
a filter because they keep
out material which would
otherwise pollute lakes.
Agree or disagree? 58 86 60
5. The lakes would always
remain clear, clean and
fresh if there were no
people around to cause
pollution.
Agree or disagree? 41 31 39
(N = 140) (N = 35) (N = 1,342)
-------
186
RESTORATION OF LAKES AND INLAND WATERS
Table 2. — Excellent or good ratings on agency water quality
protection activities by respondents who were aware of the agency's
activity (N).
U.S. Environmental
Protection Agency
Wisconsin Department
of Natural Resources
University of Wisconsin-
Extension
U.S. Soil Conservation
Service
Regional Planning Commission
Statewide
49% (416)
56% (820)
74% (316)
60% (417)
42% (296)
Waupaca
68% (34)
72% (82)
86% (43)
84% (51)
78% (27)
White
Clay Lake
55% (31)
47% (32)
74% (27)
75% (28)
42% (24)
The project appeared to be perceived as one of a
number of activities that were important in keeping
Waupaca desirable.
Alienation/Agency Image
Another common result of large projects is im-
personal decisionmaking. Citizens often feel over-
whelmed by bureaucratic processes. They feel helpless
to cope with big government, big labor, or big business.
Decisions always seem to be made by an anonymous
person in a faraway city.
The Waupaca lake project did not increase aliena-
tion. Both the Federal and State programs are offered
to local communities rather than carried out directly by
the agency. Table 2 shows how Waupaca residents.
White Clay residents, and a statewide control group
rated five related agencies on water quality activities.
In comparison to the statewide sample, Waupaca
residents rated all agencies higher. Two-thirds or more
felt the agencies were doing a good or excellent job.
The University of Wisconsin Extension received the
highest rating, 86 percent, and the lowest was 68
percent for EPA. The anti-government feeling was
much stronger in the statewide sample which had not
experienced a lake project.
More Tourists
Tourism was the one project-related concern evident
in Waupaca. Out-of-town residents made up a
substantial portion of lake users prior to the project. A
majority of the Waupaca residents indicated that they
would not favor any increase in tourists. Over 80
percent were not in favor of increases over 25 percent.
Tourists present a special dilemma for lake manage-
ment programs. State and Federal assistance is
premised on use of local lakes; if the general public
can't use a lake, why should their tax dollars be
invested there? Local citizens, on the other hand, are
reluctant to invest their time and money to manage the
lake if they might be crowded out by "rowdy outsiders."
Economic benefits of projects are often calculated in
terms of increased use by tourists who stimulate the
local economy with their purchases. Local businesses
may promote a project for this reason. However, local
property owners usually would rather not share their
lake with any more users. Tourists increase density at
local facilities, recreational and commercial; this
increased density may negatively affect local citizens
and could promote community conflict.
SUMMARY
The Waupaca Lake District carried out a major
project of storm sewer diversion, alum treatment, and
aeration over a period of 5 years without any major
setbacks or negative impacts. In economic terms, the
project is generating more benefits than the $430,000
invested. While those near the lake might have been
expected to pay a higher share of the local costs, the
uniform property tax was modest and was simplest to
collect with a uniform mill rate.
Social benefits have been positive and modest with
the single exception that the project could become a
liability for many residents, if tourism significantly
increased crowding. Because the city council con-
ducted the affairs of the lake district, the project
provided little experience in self-governance for
citizens or education in aquatic ecosystems. The
residents liked Waupaca before the project and the lake
project maintained that image. The image that
residents held of government agencies was substan-
tially improved during the course of the project. Most
significantly, the project has not caused the community
to suffer social costs, especially those which cut lasting
divisions into the community structure. The Waupaca
lake project went very smoothly; it enhanced the lake
and maintained a functional social structure.
REFERENCES
Bouwes, N., et al. 1980. Socio-economic impacts of lake
improvement projects at Mirror/Shadow Lakes and White
Clay Lake. In preparation for U.S. Environ. Prot Agency.
University of Wisconsin-Extension.
Cooper, R., and J. Powers. 1976. Waupaca chain-of-lakes
second homes owners: Expenditures, perceptions, char-
acteristics, economic impact. Recreation resour. Center,
University of Wisconsin-Extension.
Dixon, M. 1978. What happened to Fairbanks? The effects of
the Trans-Alaskan oil pipeline on the community of
Fairbanks, Alaska. Westview Press, Boulder, Colo.
Dornbusch, D. 1974. The impact of water quality improve-
ments on residential property prices. Natl. Comm. Water
Quality, Washington, D.C.
Klessig, L. 1979. Lake districts—a unique organization with
a special purpose. Fisheries 4:10.
Klessig, L., and S. Lovejoy. 1980. Necessary conditions for
resource allocation and management. 45th N.A. Wildl. Nat.
Resour. Conf., Miami Beach.
Lind, R. 1973. Spatial equilibrium, the theory of rents and the
measurement of benefits from public programs. Q. Jour.
Econ. 87.
Waupaca County Outdoor Recreation Planning Committee.
1978. Waupaca County outdoor recreation plan. East
Central Wis. Regional Plan. Comm.
-------
187
LAKE MANAGEMENT AND COST-BENEFIT
ANALYSIS IN ONTARIO
PETER A. VICTOR
Victor & Burrell
Toronto, Canada
ABSTRACT
Cost-benefit analysis is a highly developed, formal, economic methodology for evaluating the
benefits and costs of a wide range of activities. It offers a systematic means for comparing lake
management options so that an optimal mix of uses (e.g., recreation, waste disposal, potable water
supply) can be identified. Such an approach would seem to be especially attractive during a time in
which economic problems appear paramount and protection of the environment should be secured
at the least possible cost. It is significant therefore, that the regulatoryauthorities in Ontario make
little use of cost-benefit analysis for lake management purposes. This paper examines the reasons
for this. Considerable emphasis is placed on the institutional framework for lake management in
Ontario, and a liberal use is made of real life examples. The paper concludes with some
recommendations on the role of cost-benefit analysis for lake management in Ontario, and by
extension, in other similar jurisdictions.
INTRODUCTION
Ontario is Canada's third largest Province with a total
area in excess of 1,036,001 square kilometers
(400,000 square miles). More than 10 percent of the
Province is covered by water. This is divided about
equally between Canada's share of the Great Lakes
and the rest of Ontario's lakes and rivers. In this age of
quantification, there has been no official enumeration
of Ontario's lakes. While they are not literally
countless, they remain uncounted.
The abundance of Ontario's inland lakes has two
important and contrary implications for water quality
management in the Province. Because they are so
numerous and frequently so large, the likelihood of
serious, widespread contamination is reduced. How-
ever, the extent to which resources must be stretched
to monitor and regulate activities which could damage
the lakes makes it difficult for the regulatoryauthorities
to perform their role effectively.
It should also be remembered that while the average
population density of the Province is little more than
3/sq. mile, over 80 percent of the population live in
urban centers. Most of the population is located in the
southern part of the Province, which is the industrial
heartland of Canada accounting for about 40 percent of
the gross national product. This combination of high
economic activity, with its related environmental
impacts, and localized concentrations of a population
accustomed to a wide variety of outdoor recreation,
places a considerable burden on many of the Province's
more accessible lakes.
Supplementing these domestic sources of actual and
potential adverse impacts on Ontario's lakes are those
for which people outside the Province are responsible.
Each year some 20 million Americans visit Ontario,
many for recreational purposes. Furthermore, sulfuric
and nitrous oxides transported through the air
internationally may be having irreversible impacts on
numerous lakes in the Province.
All of these circumstances taken together present a
formidable challenge to rational and effective lake
management in Ontario. It is a challenge which has
been met, though by no means with complete success,
primarily at the Provincial level of government rather
than Federal or municipal. Accordingly, this paper
focuses on the approach to lake management and
especially benefit assessment, taken by the Provincial
authorities. However, in the case of the Lake Simcoe-
Couchiching basin study which is discussed at some
length, the role of the municipal authorities is readily
acknowledged.
The next section outlines lake management policies
and implementation procedures in Ontario, and insofar
as the treatment is specific, it deals with water quality
management in lakes other than the Great Lakes.
There then follows an account of a recent attempt to
develop an environmental strategy for Lake Simcoe-
Couchiching, southern Ontario's largest body of water
after the Great Lakes. The limited consideration given
to the benefits from improved water quality in this
otherwise comprehensive study is especially note-
worthy.
Finally, the paper considers, from a somewhat
critical standpoint, the role that cost-benefit analysis
might play in Ontario's lake management and
examines the reasons why the Province has made
relatively little use of this evaluation technique.
ONTARIO'S APPROACH TO LAKE
MANAGEMENT
Though several Provincial Ministries have a role to
play in lake management in Ontario, the primary
responsibility for water quality management rests with
the Ontario Ministry of the Environment. The goal of
-------
188
RESTORATION OF LAKES AND INLAND WATERS
the Ministry with respect to surface water quality is:
"to ensure that the surface waters of the Province are
of a quality which is satisfactory for aquatic life and
recreation" (Ontario Minist. Environ. 1978).
It is believed by the Ministry of the Environment that
"water which meets the water quality criteria
(designated as Provincial Water Quality Objectives) for
aquatic life and recreation will be suitable for most
other beneficial uses, such as drinking water and
agriculture"(Ontario Minist. Environ. 1978).
A major policy implication of this general goal is that
the use of stream classification, where specific rivers
and lakes in the province are designated for various
and different uses, is not permitted since all surface
water must be suitable for all uses. In fact, this goal has
not been achieved and implicitly stream classification
is practiced in Ontario, at least in relation to setting
timetables for compliance with effluent discharge
objectives.
The Ministry of the Environment has set Water
Quality Objectives for lakes and rivers which, if
satisfied, will fulfill the Ministry's water quality
management goal. These Objectives are both quantita-
tive, e.g., expressed as concentrations, and qualitative
where conditions of the receiving waters are declared
unacceptable. The Objectives were not established to
balance costs and benefits. Benefits are assumed, and
cost considerations only enter in cases where it is
acknowledged that the Objectives cannot be met owing
to the accumulation of past discharges.
The stage in the regulatory process where benefits
(and costs) do play some role is in the Ministry's
compliance programs for point source discharges. (To
date, the Ministry has done little to regulate nonpoint
sources of wastewater contaminants.) For industrial
sources the Ministry works out discharge objectives on
a case by case basis. Compliance schedules are
negotiated and a company may receive a "program
approval." Providing the terms of the approval are not
contravened, the company cannot be prosecuted for
pollution until the time period of the approval has run
out. Failure to comply with the terms of a program
approval is not in itself an offense. Partly because
program approvals have not achieved abatement
objectives, Control Orders and Requirements and
Directives have been increasingly emphasized. These
are legally enforceable statements requiring com-
panies to undertake studies and to control their waste
discharges. Non-compliance with either of these
regulatory instruments is punishable with a fine.
Note that Ontario has no effluent discharge
standards for water pollution. In effect, a Control Order
establishes a source specific standard but the issue of a
Control Order is entirely at the discretion of the
Ministry of the Environment, as is an amendment to a
Control Order. Likewise, "Certificates of Approval"
which are licenses required by anyone wishing to
operate a potential source of pollution, are issued by
the Ministry and are tailored to the conditions of the
receiving waters.
The other major category of point source discharges
into Ontario's lakes and rivers is municipal sewage
treatment plants. Typically these have been built by the
Ministry of the Environment (and its predecessor, the
Ontario Water Resources Commission) and in most
cases, turned over to the municipalities for operation
once satisfactory performance has been achieved. This
close involvement of the Ministry in the construction
and operation of these plants has facilitated somewhat
more effective control than is the case for many
industrial sources. Nevertheless, problems can arise,
especially when the Ministry seeks improvements in
the performance of municipal sewage treatment plants
beyond the original design capability, since the
municipalities may be reluctant to incur the associated
increase in costs.
This brief description of Ontario's approach to
achieving and maintaining a satisfactory level of water
quality in lakes and rivers should not obscure the fact
that it is all part of a comprehensive regulatory
framework. In addition to surface quality the Ministry of
Environment has policies, objectives, and implementa-
tion procedures for surface water quantity manage-
ment, which limit water withdrawals, and for ground-
water quality and quantity management. This activity is
supported by an extensive and sophisticated research
capability. Moreover, the Ministry, working through its
regional and district offices as well as the head office,
liaises with staff in other Ministries and government
agencies whose responsibilities also impinge on lake
management in the Province. Important among these
are the Ministry of Natural Resources, which regulates
the exploitation of such resources as wildlife and
forests, and the municipally-based Conservation Au-
thorities, with responsibility for land-use and develop-
ment activities as they affect flood-plain areas and the
total watershed.
Notwithstanding all this planning and regulatory
activity, and the enormous production of studies and
data, little explicit consideration is given to the benefits
of lake management in Ontario. This is further
illustrated in the development of an environmental
strategy for the Lake Simcoe-Couchiching Basin
described in the next section.
THE LAKE SIMCOE—COUCHICHING
BASIN ENVIRONMENTAL STRATEGY:
THE ROLE OF BENEFIT ASSESSMENT
Lakes Simcoe and Couchiching have a combined
area of some 777 square kilometers (300 square miles)
draining a land area of 2,434 square kilometers (940
square miles) (see Figure 1). Proximity to large centers
of population and outstanding natural features have led
to increasing use of the lakes for swimming, fishing
(sports and commercial), boating, water supply (cot-
tages and small townships), and waste disposal
(municipalities and cottages). A 4-year study of the
lakes culminated in a report published in 1975 (Ontario
Minist. Environ. 1975) which concluded that local
problems aside, the general water quality of the lakes
was good, but problems were emerging. Chief
concerns centered around increasing algal growths
and changing fishing success. These were related to
excess nutrient material, particularly phosphorus,
being discharged into the lakes.
As a result of public meetings and the further
accumulation of data by Provincial Ministries, two
committees were formed: a Report Committee and a
Steering Committee. The Report Committee, consisting
-------
PUBLIC BENEFIT AND INSTITUTIONAL PROBLEMS
189
of staff from Provincial Ministries and municipalities,
was directed by the Provincial Cabinet Committee on
Resources Development (CCRD) to: (1) Assess the types
and magnitudes of environmental problems in the Lake
Simcoe-Couchiching area; (2) identify the causes of
these problems; and (3) propose a strategy for dealing
with the problems.
The proposals of the Report Committee were adopted
by the Steering Committee whose members were
drawn primarily from the local municipalities. After
receiving the report CCRD accepted all of the
recommendations except that referring to the critical
issue of phosphorus loadings to the lakes.
The Report Committee estimated the total phos-
phorus loading to the lakes in 1979 to be 103 tons,
broken down by sources as shown in Table 1.
The Report Committee also identified other environ-
mental problems in the watershed: Important marsh
and wildlife areas are being encroached upon, forested
areas are being diminished, poorly managed mining
activities are threatening ground waters, and sensitive
ecological areas are subjected to stress. All of these
factors affect the basin environment and the recom-
mended environmental strategy addresses them all.
However, for the purposes of this paper it is sufficient
to focus on the evaluation of benefits in relation to a
reduction in phosphorus discharged to the lakes. This is
also consistent with the emphasis given to this matter
in the Simcoe-Couchiching report and in CCRD's
response.
Three alternative environmental development strat-
egies were considered by the Report and Steering
Committees:
1. Maintain existing environmental quality (i.e.,
maintain present water quality, fishery, and general
Dasin environment);
2. Improve environments! quality;
3. Allow environmental deterioration.
This third option was rejected by both committees on
the grounds that it is contrary to the Provincial policy on
surface water quality management and to the desires
of the local community. No attempt was made to
identify and evaluate the benefits that would be
foregone if this alternative were adopted. It was
assumed implicitly that these would outweigh any
savings in costs that such a strategy would allow.
According to the report, maintaining existing en-
vironmental quality in the face of local population
growth and the necessary use of the lake for recreation
requires deliberate actions to limit total phosphorus
loadings to the current 103 tons/year. Any combina-
tion of schemes to control the individual sources of
phosphorus which have the effect of limiting the
phosphorus loading to this level was deemed a
potentially acceptable strategy.
The benefits expected from maintaining existing
water quality were believed to stem from "an
abundance of warm-water species in Lake Simcoe and
a precarious cold-water fishery." (Lake Simcoe-Couch.
Rep. Comm. 1979). This in turn would benefit the
tourism industry which "relies on the quality of the
recreational experience which relates to good water
quality and fisheries." (Lake Simcoe-Couch. Rep.
Comm. 1979).
Figure 1. — Map of Lake Simcoe-Couchiching drainage basin.
Table 1. — Estimated phosphorus loadings into Lakes
Simcoe-Couchiching, 1979.
Tons/Year*
Phosphorus
The field leakage — cottages
Precipitation
Rivers under natural watershed
conditions
Urban storm wastes
Agriculture and other land-use
disturbances
Sewage treatment plant effluents
.(with P removed to 1 mg/l)
3
21
26
9
22
22
103
* 1 ton (metric) = 2,000 Ibs.
Source: Lake Simcoe-Couchiching Basin Environmental
Strategy, 1979, p. 4.
An improvement in water quality, the last option to
be considered would result in "the elimination of
scums on the lakes, decreased weed growth, improved
water clarity in some areas, and a more stable fishery
with healthy, self-reproducing populations of whitefish
and lake trout." (Lake Simcoe-Couch. Rep. Comm.
1979). (Under a strategy of maintaining water quality
the whitefish population is expected to be extinct in 20
years and the lake trout population might be
maintained through stocking.) As for the phosphorus
loading required by this strategy of improvement it
would have to be reduced progressively to 75
-------
190
RESTORATION OF LAKES AND INLAND WATERS
tons/year to achieve a self-reproducing cold water
fishery. (Lake Simcoe-Couch. Rep. Comm. 1979).
Again, the study did not give any detailed consideration
to the value of the benefits that might be derived from
such a strategy. No more on benefits that has been
given in this brief summary of the study was produced
for comparison with the roughly estimated strategy of:
1. Maintaining existing water quality: $3,000,000/yr
(1979 Canadian dollars) with a basin population
increasing from 190,000 to 300,000.
2. Improving water quality — $4 to $5,000,0007year
with no increase in basin population.
The Report Committee recommended a total phos-
phorus loading objective of 103 tons/year. This was
reduced somewhat by the Steering Committee to 95
tons/year to allow for some uncertainty in the
performance of the proposed control measures.
Subsequently, the Cabinet Committee on Resources
Development reduced the phosphorus objective to 87
tons/year. What is especially significant here is that
this was based on a belief that "the value of the cold
water fishery in terms of tourism and the economy, as
well as an indicator of the ecological health of an
important water resource, warrants a major effort to
restore water quality to a level which will support such
a fishery. .Improved water quality, and the resulting
reduction in slime and aquatic weeds, will make the
basin a more attractive recreation and tourism
destination and could increase property values" (Cab.
Comm. Resour. Dev. 1979).
This objective of 87 tons/year substantially exceeds
the 75 tons/year stated by the Report Committee to be
necessary for a self-reproducing cold-water fishery. Yet
apparently, the CCRD had no new evidence on which to
base its belief that substantial benefits from a thriving
cold-water fishery would ensue if this revised objective
is achieved. To give such weight to anticipated benefits
when the supporting documentation is so obviously
lacking underlines the importance of improving benefit
estimation in developing lake management strategies.
It is the potential of cost-benefit analysis in this regard
that is discussed in the next section.
COST—BENEFIT ANALYSIS: BENEFIT
ESTIMATION AND LAKE MANAGEMENT
Cost-benefit analysis is a highly developed, formal,
economic methodology for evaluating the benefits and
costs of a wide range of activities. It offers a systematic
means for comparing lake management options so that
an optimal mix of uses (e.g. recreation, waste disposal,
water supply) can be identified.
Such an approach would seem to be especially
attractive during a time in which economic problems
appear paramount and protection of the environment
should be secured at the least possible cost. It is
significant therefore, that the regulatory authorities in
Ontario make little use of cost-benefit analysis for lake
management purposes. In particular, the estimation
and evaluation of benefits within the cost-benefit
framework is seldom practiced in Ontario." The
* Cost-effectiveness, a subcomponent of cost-benefit analysis is sometimes used to
determine the least costlv means of achieving some prescribed objective
Furthermore, the Ontario Ministry of the Environment has recently (July 1980)
funded studies of the damages due to actd precipitation
discussion which follows will examine possible
reasons for this. It will also consider whether the use of
such an approach might improve lake management in
Ontario and, by extension, in other sijTjHarjjjrisdictions.
The first possibility is that the benefits from lake
management are already adequately accounted for in
the regulatory process. This does not seem to be the
case either in the establishment of Ontario's Surface
Water Quality Goal or in the setting of the Provincial
Water Quality Objectives. When it comes to the actual
process of regulation, which involves a considerable
degree of informal negotiation between the regulators
and those responsible for waste discharges, the picture
is less clear. Nevertheless, the development of the Lake
Simcoe—Couchiching management strategy, which
stands out as a relatively comprehensive and system-
atic approach to lake management, underlines the
casual way in which benefits are addressed. It does not
appear, therefore, that adequate consideration is
already given to benefits in Ontario's approach to lake
management.
A second possible reason for not using cost-benefit
analysis in Ontario's lake management is that the
approach is not well understood by Ontario's regulatory
authorities. The Ontario Ministry of the Environment,
like many other environmental agencies, is staffed
principally by people with backgrounds in engineering
and the natural sciences, who characteristically
approach environmental management differently than
an economist. The notion of optimizing across a range
of environmental, social, and economic objectives is
alien to them and this tends to weaken the appeal of an
approach such as cost-benefit analysis.
The reasons given so far question the adequacy of
the existing approach, and the orientation of key
personnel within the regulatory agencies. The possi-
bility that deficiencies in cost-benefit analysis might
account for its lack of use must now be examined.
Cost-benefit analysis in any application is essentially
a process of market simulation (Mishan, 1976). The
analysis attempts to identify the potential gainers and
losers from a proposed project, program, or policy and
to estimate the maximum sum of money the gainers
would be willing to pay for the benefits and the
minimum that the loser would require as compensation
for incurring the losses. Only if the benefits exceed the
costs, according to these definitions, does the proposal
being analyzed pass the cost-benefit test.
The technical problems involved in conducting a
cost-benefit analysis are challenging even to the most
well-trained and experienced economist. The informa-
tion requirements, alone, may be too demanding,
especially on the benefits side, to allow the use of this
approach for lake management. But even if these
difficulties can be overcome there are other problems
with the approach.
First of all, the proper description of gainers and
losers may be ambiguous. In the case of lake
improvements requiring pollution abatement the
gainers might be those who would benefit from
improvements and the losers those who will have to
incur costs to abate pollution. Alternatively, the same
project could be looked at from the viewpoint of
maintaining the existing level of water quality. In that
case the gamers would be those not having to further
-------
PUBLIC BENEFIT AND INSTITUTIONAL PROBLEMS
191
control their pollution and the losers would be those
having to forego the benefits from enhanced water
quality. A cost-benefit analysis could be conducted
from either perspective and the results could well be
different.
Typically, economists presume that the status quo is
an acceptable point from which to start and so the
gainers would be those who would benefit from the
improved water quality. By implication, this confers a
right to pollute on existing waste dischargers and this
may be unacceptable to the regulators and to the public
at large.
Another way in which the status quo is often given
normative significance in cost-benefit analysis is with
respect to the distribution of incomes and wealth.
Estimates of willingness to pay for benefits and
compensation required for costs depend upon people's
economic situation. A change in this will change the
estimates. Again it may be unacceptable to the
regulators and to the public for decisions about the
proper use of the publicly shared environment to reflect
the distribution of private incomes and wealth. Yet this
is built into cost-benefit analysis and can only be
amended through arbitrary adjustments in the benefit
and. cost estimates.
Some economists argue that it is appropriate for
cost-benefit analysis to deal only with "economic
efficiency" (Mishan, 1976). Considerations of equity or
fairness, such as those alluded to here, should be
addressed in the political arena. Insofar as lake
management poses problems of equity, both among
contemporaries and across generations, this further
justifies limiting the role of cost-benefit ana lysis in lake
management. Moreover, it is questionable whether
efficiency (getting the most from the least) and equity
(sharing the benefits and costs fairly) can be separated
in this way when the measures of costs and benefits
depend upon the distribution of incomes and wealth.
Other issues which pose difficulties for cost-benefit
analysis in lake management relate to the scope of the
analysis both temporally and spatially. How should
benefits and costs expected in the distant future be
compared with those likely to be incurred in the near
term? Should the gainers and losers include people
from beyond the jurisdiction responsible for lake
management? (This is particularly important in Ontario
where many of those who benefit from improvements
in lake quality come from other Provinces and
countries.) But these are questions that any rational
approach to lake management must confront. Though
they may be inadequate, answers to them are provided
within the benefit-cost framework. Intertemporal
comparisons of benefits and costs are made using a
discount rate which has the effect of giving more
weight to benefits and costs the sooner they are
expected. The scope of a benefit-cost study in terms of
who is included typically reflects the extent of the
jurisdiction within which the analysis is being done.
Finally, cost-benefit analysis may be regarded with
some skepticism by those who believe that decisions
on lake management should reflect concerns that
override those of human beings alone. Responsibility
for the protection of the environment goes beyond
questions of its optimal use. Whether or not this
perspective is compatible with the cost-benefit frame-
work, the obvious anthropocentric bias of cost-benefit
analysis has no doubt deterred some people from
taking it more seriously.
CONCLUSION
This paper has shown that systematic benefit
estimation plays a minor role in lake management in
Ontario. It has also suggested several reasons to
explain why cost-benefit analysis has not been used
more extensively, especially for evaluating benefits.
The question remains whether, despite all its real and
perceived shortcomings, cost-benefit analysis could
usefully contribute to the formulation of lake manage-
ment strategies and programs. The greatest danger in
this regard lies in the possibility that poorly conducted
cost-benefit analyses, from which numerous important
considerations are omitted, should come to supplant
the sort of participative approach being developed in
studies such as that for the Lake Simcoe-Couchiching
Basin. While much may be inadequate about that
process, the opportunity for a wide range of affected
parties to interact, provide information, and learn is
impressive. Yet, a greater emphasis on benefits,
however approached, might improve the process
considerably. At the very least, the categories of
benefits (e.g., water supply, recreation, habitat,
commercial fishing) expected from lake management
could be specified, case by case. A modest effort might
provide quantitative estimates of the relationship
between various levels of lake water quality and some
of these benefits measured in natural units (e.g.,
gallons of potable water, recreation days, acres of
habitat, tons of catch). Ascribing dollar values to these
benefits, so that they may be aggregated and compared
with the costs of lake management, is the final step in
cost-benefit analysis and no doubt the most treach-
erous one. But if it is not taken some other possibly
inferior means must be found to evaluate lake
management options.
Cost-benefit analysis is one way to increase the
attention given to benefits and costs in lake manage-
ment; it makes explicit issues that have to be dealt with
one way or another in any case. Providing the
estimates of benefits and costs are properly presented
and understood, they could enhance the planning and
regulation of lakes in Ontario and elsewhere.
REFERENCES
Cabinet Committee for Resources Development. 1979.
Information for meeting with the Lake Simcoe— Couchi-
ching Steering Committee. November 1.
Mishan, E. J. 1976. Cost-benefit analysis. 2nd ed. Allan and
Unwin.
Lake Simcoe—Couchiching Report Committee. 1979. Lake
Simcoe—Couchiching Basin Environmental Strategy.
Ontario Ministry of the Environment. 1975. Lake Simcoe
Basin, a water quality and use study. June.
1978. Water management. November.
-------
192
THE LEMAN COMMISSION
GUY BARROIN
Station d'Hydrobiologie Lacustre
Institut National de la Recherche Agronomique
Thonon-les-Bains, France
ABSTRACT
The Commission Internationale pour la Protection des eaux du lac Leman et du Rhone centre la
pollution was officially established in 1960, originating from an informal Franco-Swiss institution
founded 10 years before. Study of the sanitary and trophic status of a lake is followed by a
technical subcommission involving several laboratories of both countries. The studies concern the
evolution over time and space of the different components of the whole ecosystem: The lake (582
km2) and its drainage basin (7,390 km2). The studies are planned on a 5-year basis and distributed
to the different laboratories according to their specialization (chemistry, biology, microbiology). The
results are published in annual reports and constitute the basis for the Commission's
recommendations. The main practical objective was to diminish eutrophication by domestic and
industrial sewage treatment, including phosphorus elimination. More recently the Commission
has been faced with problems of mercury and PCB pollution.
INTRODUCTION
With 8,582 km,3 the Leman —Lake Geneva— is the
greatest lake and the largest freshwater reserve in
western Europe. The lake and its drainage basin are
shared by France and Switzerland. An international
commission has been founded to solve pollution
problems through a joint effort of everybody concerned.
THE LAKE AND ITS DRAINAGE BASIN
The Leman includes two sub-basins: The Grand Lac
upstream, and the Petit Lac downstream (Figure 1).
Table 1 gives their main physical characteristics. The
Petit Lac looks more like an enlarged river than a lake; it
is more realistic to express its mean residence time
versus depth:
4 to 5 years from the surface to 50 meters deep
10 years from 50 meters to 200 meters
20 years from 200 meters to the bottom.
The area distribution between France and Switzer-
land is given in Table 2.
The Leman receives water from a surface area of
7,390 km2, 80 percent of which lies in Switzerland
(Figure 2). With 176 m3 S"1 , the Rhone River is the
main tributary and represents 75 percent of the water
input. Its drainage basin, totally under Swiss control,
culminates at an altitude of 4,638 meters and is 5,220
km2 wide; 16 percent of the surface area is covered
with glaciers. The Dranse River is the second largest
tributary, draining 535 km2 in France for a contribution
of 20 m3 S~1 Urban, industrial, and agricultural
activities are concentrated in the plains and dispersed
in the mountains. The principal activity seen in the
mountains is tourism.
COMMISSION ORIGINS
During a meeting in Lyon, France, in the spring of
1950, the Association of Rhodanians (Union Generale
des Rhodaniens) pointed out the disastrous conse-
Table 1. — Main physical characteristics of the Leman.
Surface area (km2)
Surface area (%)
Volume (km3)
Volume (%)
Maximum depth (m)
Mean depth (m)
Grand lac
503.5
86.5
85.69
96.4
309.7
172.4
Petit lac
78.8
13.5
3.23
3.6
76.5
37.8
Leman
582.3
100.0
88.92
100.0
309.7
152.7
Table 2. — Distribution between France and Switzerland.
Swiss Canton or
French Department
Vaud, Switzerland
Geneve, Switzerland
Valais, Switzerland
Surface area
293.98 km2
36.35 km2
10.50 km2
= 340.83 km2
Figure 1. — Bathymetric map of the Leman.
Haute-Savoie, France
241.47 km2
-------
RESTORATION OF LAKES AND INLAND WATERS
193
Figure 2. — Map of the drainage basin.
quences for man and his environment of dumping
sewage into the Rhone. Dr. Messerli from Lausanne,
Switzerland, was asked to establish an informal
commission to study sewage pollution and to try to find
a solution through coordinated action of the concerned
populations. This commission consisted of many
experts generally connected with French or Swiss
authorities.
After a few years devoted to standardizing analytical
methods and elaborating sampling programs, the first
systematic and coordinated survey began in 1957. But
because of its informal status, this commission was
limited to observation and research without any
practical application. Therefore, on November 9,1960,
Swiss and French authorities decided to recognize the
commission officially, as the Commission Internation-
ale pour la protection des eaux du lac Leman et du
Rhone centre la pollution. A Franco-Swiss agreement
became effective November 1, 1963.
THE FRANCO-SWISS CONVENTION
Established between Switzerland's Conseil Federal
and the French government to coordinate their efforts
to protect Lake Geneva against pollution, the Conven-
tion extends the Commission's jurisdiction to the Swiss
border at the lake outlet and includes all superficial and
deep waters. It establishes the Commission's four main
objectives:
1. To organize and monitor all research aimed at
determining the nature, extent, and origin of pollution
and to use the results.
2. To recommend governmental measures to cure
today's pollution and to prevent future pollution.
3. To prepare the basis for establishing international
regulations in the case of incompatibility between
legislation of the respective governments.
4. To investigate all questions concerning water
pollution.
The result is that only strictly-applied research is
developed and entrusted to a sub-commission (the
Sous-Commission Technique). Occasional working
groups may be constituted for solving specific
problems. Finally, the Convention limits the Commis-
sion's intervention to recommendations, all decisions
being made by the governments.
THE COMMISSION'S MEMBERSHIP
The Commission consists of two delegations; the
French delegation is led by Mr. J. Leclerc from the
Foreign Office in Paris and is composed of eminent
officials from concerned governmental authorities; the
Swiss delegation is led by Mr. Pedroli, Director of the
Federal Office for Environmental Protection in Berne.
The delegation leaders alternate in presiding at
commission meetings. The Commission is organized as
shown in Figure 3.
COMMISSION INTERNATIONALE POUR LA
PROTECTION DES EAUX DU LEMAN
CONTRE LA POLLUTION
SECRETARIAT
PERMANENT
VERIFICATEUR
DES
COMPTES
Figure 3. — Organization of the Leman Commission.
The S.C.T. (Technical Subcommission)
The Sous-Commission Technique is the Commis-
sion's executive body. It plans research on a 5-year
basis and elaborates recommendations. It includes
experts in several disciplines: Physicians, engineers,
biologists, microbiologists, chemists, etc., who are
selected by the delegation leaders. It is presided over
alternately by a French or a Swiss representative. The
Sous-Commission Technique is also divided in two
delegations, functions according to internal regulation,
and has its own administrative board.
The Secretariat Permanent (Permanent
Secretariat)
This technical and scientific secretariat assists the
Commission in everything concerning data treatment,
annual report preparation, public information, etc.
Franco-Swiss collaboration on oil pollution
The Collaboration Franco-Suisse en cas d'accident
par les hydrocarbures is totally independent of the
Sous-Commission Technique and is directly attached
to the Commission. It organizes and coordinates action
taken against oil pollution. Its existence and its
effectiveness result from a specific agreement, dated
November 18, 1977, which solves problems concern-
ing crossing the border on land, lake, and air.
-------
194
RESTORATION OF LAKES AND INLAND WATERS
Table 3. — Program of the purification plants construction.
Sewage treatment plants
Number
LEMAN
Valais
Vaud
Haute-Savoie
Ain
Geneve
Total
With phosphate
elimination
RHONE'S BASIN
Haute-Savoie
Ain
Geneve
Total
General total
1968
9
8
1
1
3
22
5
4
9
31
1973
18
35
4
3
3
63
11
2
14
27
90
1978
36
58
6
4
3
107
48
15
3
13
31
138
1979
38
57
8
4
3
110
45
19
3
13
35
145
1968
21,100
262,000
3,000
500
4,900
291,500
11,200
400,000
411,200
702,700
Treatment capacity
1973
316,190
507,000
109,000
29,500
5,050
966,740
39,000
18,500
451,000
508,500
1,475,240
1978
762,845
666,630
127,000
33,000
6,950
1,596,425
1,448,760
119,150
20,000
460,505
599,655
2,196,080
1979
770,645
668,200
129,400
33,000
6,950
1,608,195
1,449,730
133,950
20,000
460,505
614,455
2,222,650
THE COMMISSION'S AIMS
Lake Geneva's water quality must be appropriate for
drinking water, bathing, and salmonids. The definition
of water quality is based on European Economic
Community parameters. According to the Commis-
sion's recommendations, many dispositions have been
taken in connection with:
Domestic sewage: Table 3 gives the program of
treatment plant construction realized since 1968. On
January 1, 1979, the rate of collected populations was
73 percent in Lake Geneva's watershed and 56 percent
in the downstream watershed; 45 of 110 treatment
plants are currently eliminating phosphorus. Treatment
efficiency is generally inspected once a year, 50
percent of the plants being checked four times. This
frequency of inspection will be generalized; analysis
will concern a 24-hour sample. BOD5 elimination is
better (<20 mgf1) than COD elimination and much
better than phosphorus elimination. During 1978, only
five plants serving 400,000 persons each, were able to
reach the 1 mgPI'1 limit. The effluents of 20 plants
contained between 1 and 2 mg. The situation is now
improving, thanks to the Commission's encouraging
work towards better phosphorus elimination. Recently,
the Commission proposed establishing an international
fund to buy reagents necessary for phosphorus
elimination.
Phosphorus in detergents: The Commission recom-
mends reducing or eliminating phosphate detergents.
In the meantime, it recommends prohibiting or limiting
television publicity for these products.
Nutrient nonpoint sources: Investigations are con-
ducted to estimate the agricultural and the natural
contributions to the nutrient budget. The Commission
also recommends a better focus on pollution from
touristic and commercial navigation through technical
and legislative measures relating to boats and harbors.
Industrial pollution: During 1970-1972, investiga-
tions conducted on the Rhone demonstrated that 10 to
15 kg of mercury were introduced daily into the lake's
ecosystem. Public opinion reflected significant agita-
tion, exacerbated by mass media and by the Minamata
affair. Necessary measures were rapidly taken, and in
1978 new investigations indicated a reduction to 1 to3
kg of mercury per day. During the next 5 years heavy
metals (lead, chromium, cadmium) and PCB will be
intensively surveyed in sediments and fish populations.
CONCLUSIONS
Although the realizations appear to be slow, it must
be remembered that the Commission is not invested
with any supranational authority and that its direct
action is limited to simple recommendations. It is at
least a consolation to see they are not ineffective.
REFERENCES
Documents concerning the Leman Commission may be
obtained from:
Dr. R. Monod
Commission Internationale pour
la Protection des Eaux du lac
Leman centre les Pollutions
Case Postale 80
1000 LAUSANNE 12 Chailly
Switzerland
Tel. 021/33.14.14
-------
195
STRUCTURE, AIMS AND ACTIVITIES OF THE
INTERNATIONAL ALPINE COMMISSIONS IN EUROPE
0. RAVERA
Commission of the European Communities
Ispra, Italy
G. BARROIN
Lacustrine Hydrobiology Station
Thonon, France
D. IMBODEN
Federal Institute for Water Resources
and Water Pollution Control
Dubendorf, Switzerland
G. WAGNER
Institute for Lake Research and Fishery
Lengenargen, Germany
ABSTRACT
Structure, aims, and activities of the International Alpine Commissions in Europe are discussed.
Particular emphasis is given to the constitution of the commissions and the most important
problems concerning the common water of the five countries.
The boundaries between countries result from
historical events, and, therefore, they do not coincide
with the boundaries of the watersheds. As a
consequence, several lakes and rivers mark the
boundaries between countries or cross them. To
effectively manage these water resources and protect
them against pollution, the governments of the
countries concerned must agree upon common rules
and actions concerning this problem. On this basis,
over the last two decades in the Alpine region of
Europe several conventions between two or more
governments have been ratified. These conventions
entrust the protection of the surface and ground waters
to international commissions. This important problem
was exhaustively discussed at the OECD Seminaire sur
la pollution transfrontie're dans les bassins hydro-
graphiques internationaux (June 6-10, 1977).
The Alpine International Commissions for water
protection are:
1. Lake Constance, created in 1959 and signed the
following year by the governments of Austria, Baden-
Waerttemberg, Bavaria, and Switzerland.
2. Lake Geneva, officially established in 1960 and
ratified by the French and Swiss governments in 1963.
3. Lake Maggiore and Lake Lugano, created and
signed in 1960 by the governments of Italy and
Switzerland (and again in 1972 for subsequent water
courses and the ground waters of their watersheds).
The structures of these commissions are very
similar. Each has a president and is composed of
delegates from member governments and a limited
number of high officers of those governments. The
presidency changes after defined periods. As a rule, the
commissions meet at least once a year.
The most important duties of the commissions are
the following:
1. To examine problems concerning water pollution
and to judge proposals on studies of these waters as
well as on the actions to be taken to reduce the
pollution level;
2. To illustrate to the governments the water
problems and propose actions for protecting them
against pollution;
3. To establish financial plans to support the studies
sanctioned by the commission.
4. To prepare the basis for establishing international
regulations in tKe case of incompatibility between the
legislation of the respective governments (as in the
case of Lake Geneva).
As consultant agencies for the governments, the
commissions cannot decide on rules and actions
connected with environmental protection.
Technical and scientific sub-commissions serve as
official consultants to commissions; their number is
also limited. The sub-commissions have the following
duties:
1. To study the scientific and technical problems
proposed by the commission and examine the research
carried out by other organizations;
2. To elaborate on the research program to be
submitted to the commission;
-------
1 96 RESTORATION OF LAKES AND INLAND WATERS
3. To prepare the reports on the research sanctioned
by the commission.
(The sub-comissions have working groups for
studying special problems; external experts may be
invited to collaborate with these groups).
As an example of the sub-commission activity, some
of the most important problems investigated by the
sub-commission are described here.
For the Italian-Swiss sub-commission these are:
1. Pluriannual researches on the trophic evolution of
Lake Maggiore and Lake Lugano and the evaluation of
the nutrient load (nitrogen and phosphorus com-
pounds) from the watershed of these lakes;
2. Studies on the microbiology of Lake Maggiore
waters and, particularly, the coastal area;
3. Studies for unifying criteria and methods to control
the sanitary conditions of the swimming waters of Lake
Maggiore and Lake Lugano. These studies have been
carried out on the basis of a directive of the European
Communities Council (December 8, 1975).
4. Comparison between plans for protecting and
ameliorating the quality of the Swiss-Italian water
bodies;
5. Elaboration of a unified plan for the alarm and
intervention in the case of an accident produced by the
release of hydrocarbons or other noxious substances in
the Swiss-Italian water bodies.
In Lake Constance some important working groups
are: Lake water research; river control; sediment
investigations; oil, heavy metals, and organic alloch-
thonous compounds (and other scientific problems);
investment program for treatment plants; collaboration
on oil pollution prevention and other technical
problems. There are no activities on sanitary problems.
Some of the most important programs of the Lake
Geneva commission are: Franco-Swiss collaboration
on oil pollution protection; domestic sewage; phos-
phorus in detergents; nutrient nonpoint sources; and
industrial pollution.
The sub-commissions meet four to five times per
year. Their working methods are: (1) A problem is
identified and discussed by the sub-commission
experts; (2) a proposal for its solution is worked out and
is given to the commission to judge. The financial
support for major projects is guaranteed by the member
governments.
At the sessions of the different working groups a
regular reporting on the progress of the investigations
takes place. The results are summarized in final reports
and, as a rule, published. For the publications, special
redaction committees exist. The publications are
available from:
Lake Geneva Commission, Secretary at Lousanne
Chailly (Dr R. Monond); Lake Maggiore and Lake
Lugano Commissions, Secretary at Locarno-Casa
Rusca (Dr. A. Rima); Lake Constance Commission,
Ministries for Environmental Protection of the collabo-
rating countries (Bern, Munich, Stuttgart, Vienna.)
-------
197
INSTITUTIONAL ARRANGEMENTS FOR SHORELAND
PROTECTION AND LAKE MANAGEMENT IN WISCONSIN
DOUGLAS A. YANGGEN
University of Wisconsin Extension
Madison, Wisconsin
LOWELL L. KLESSIG
University of Wisconsin Extension
Stevens Point, Wisconsin
ABSTRACT
Increasing development brings problems to many lakes. Proper shoreland development can be a
key factor in avoiding lake problems. Where problems already exist, certain management practices
often can prevent further degradation or repair past damage. Institutions with appropriate legal
authority, financial resources, technical capability, and an interest in the resource are central to
lake and shoreland management. In Wisconsin, there are two complementary State-local
programs: The Shoreland Protection Program and the Lake Management Program. Their
legislative development, their regulatory powers, and the results of their implementation are
discussed.
INTRODUCTION
Lake Problems
Increasing development brings problems to many
lakes. The amenities of a natural shoreline are replaced
by ribbons of structures. Dwellings may be squeezed
onto undersized lots too close to each other and too
close to the water. With the removal of shoreland
vegetation, native plant communities are destroyed and
wildlife habitat disappears. Erosion problems also
intensify as vegetation is removed. Road building,
grading, and filling during development exposes raw
earth and causes additional erosion. Silt muddies the
water and impairs aquatic life. In some places,
municipal and industrial wastes and agricultural runoff
are the major polluters. Septic tank systems which
serve most recreation developments can add excess
nutrients to the lake, if improperly installed or
maintained. Other lakes are free from these sources of
pollution but need management to protect their quality.
Solutions to Lake Problems
Proper shoreland development can be a key factor in
avoiding lake problems. Maintaining a natural strip of
shoreline can provide a buffer between land and water.
Buffer strips are natural preservers of water quality as
they trap nutrients and retard erosion. Preservation of
the shoreland buffer also conserves the unique scenic
qualities of the lakeshore. Density standards to avoid
overcrowding and proper installation of on-site waste
disposal systems can also help preserve lakes.
Where problems already exist, certain management
practices can, in some cases, prevent further degrada-
tion or repair past damage. There are two general
approaches to lake protection and rehabilitation: (1)
Limiting fertility; and (2) treating the products of over-
fertilization. Limiting fertility can consist of measures
such as diversion, nutrient inactivation with chemicals,
and dilution through lake flushing. Treating the
products of overfertilization includes measures such as
mechanical aeration to increase oxygen levels and
mechanical harvesting of excess weeds (Born and
Yanggen, 1972).
Institutions
Institutions with appropriate legal authority, financial
resources, technical capability, and an interest in the
resource are central to lake and shoreland manage-
ment. In Wisconsin, there are two complementary,
State-local programs: The Shoreland Protection Pro-
gram and the Lake Management Program. The
Shoreland Protection Program establishes minimum
State standards for mandatory local zoning and
sanitary (septic tank) and subdivision ordinances to
protect the shoreline environment. Shoreland ordi-
nances which include lands within 1,000 feet of lakes
contain, among other things, provisions governing:
Minimum lot size and width; waterline setbacks;
removal of vegetation; on-site waste disposal, filling,
grading, and dredging; and wetland protection.
The Lake Management Program authorizes the
creation of special purpose districts at the local level
and provides these lake districts with technical and
financial assistance from the State. A district has
power to tax, levy special assessments, borrow, and
bond to raise money. It may make contracts, hold real
estate, and disburse money for lake protection and
rehabilitation projects, but does not possess the
regulatory power.
-------
198
RESTORATION OF LAKES AND INLAND WATERS
THE SHORELAND PROTECTION
PROGRAM
The Enabling Statute
In adopting sections 59.971 and 144.26 of the
Wisconsin statutes, the legislature (1) created special
shoreland protection corridors, i.e., unincorporated
lands within 1,000 feet from a lake, pond or flowage,
300 feet from a river or stream or to the landward side
of the flood plain if this is a greater distance; (2)
established special regulatory objectives to protect
water quality and shoreline amenity values; and (3)
provided that if a county failed to adopt shoreland
regulations meeting minimum State standards within
48 months after passage of the law, a State agency
was required to adopt the regulations.
To implement the broad mandate of this law, a
number of difficulties had to be overcome:
1. The regulations had to be capable of being applied
on a statewide basis involving many miles of lake and
stream shores;
2. Few counties had modern zoning, subdivision
control, and sanitary regulations, or experience in
adopting and administering them;
3.There was a lack of detailed resource data on
specific shoreland characteristics such as soil types,
slope, vegetative cover, land use development patterns,
direction of groundwater flow, and water quality
parameters, which may vary widely for individual water
bodies;
4. Limited scientific data made it difficult to general-
ize about the potential pollutional effects of various
uses;
5. Constitutional limitations on the use of the
regulatory powers, such as those that prohibit the
taking of private property without compensation had to
be considered;
6. It was necessary to strike a balance between the
concern with shoreland problems and the desire for
economic revenue from shoreland development to
maintain political support for the regulatory program;
7. The regulations had to be designed to be feasibly
administered and enforced.
The Department of Natural Resources, which was
charged with supervising county compliance, was
assisted by the University of Wisconsin and State and
Federal agencies in preparing a shoreland protection
manual and model shoreland protection ordinance. The
purpose of these publications was to provide those
counties lacking professional assistance with informa-
tion to help meet the 18-month deadline.
The Shoreland Protection Ordinance
The model shoreland protection ordinance is essen-
tially a natural resource oriented development code
(Yanggen and Kusler, 1968). The basic land use
controls available to local government, that is, zoning,
subdivision regulation, and sanitary codes, are com-
bined in an integrated package. Special provisions not
usually found in these regulatory devices are added to
meet the special objectives of the shoreland protection
law. Many of the regulatory standards are keyed to the
physical characteristics of the site. This information is
generated at the time of an application for development
permission. Certain special uses with potential prob-
lems require a case-by-case evaluation by an adminis-
trative agency according to standards set forth in the
ordinance. The resulting ordinance consists of broad
regulations applicable to all shorelands, together with a
basic three-district zoning use classification.
Certain controls apply to all shoreland areas
regardless of the zoning district in which they are
located. These regulations include minimum standards
for water supply and waste disposal, tree-cutting
controls, setbacks for structures from highways and
navigable waters, minimum lot sizes and widths, filling
and grading limits, lagooning and dredging controls,
and subdivision regulations. These provisions con-
stitute the central core of the recommended regula-
tions.
The manner in which common shoreland uses are
developed usually threatens the quality of shoreland
areas more than the encroachment of incompatible
uses. Typical lakeshore development consists of
cottages, residences, and resorts, with occasional
taverns, groceries, or other commercial buildings on
some lakes. Few recreational areas are threatened by
severe nuisance uses like factories or junkyards.
The main problems in shoreland areas are over-
crowding, deterioration of water quality, and de-
struction of shore cover and natural beauty, stemming
from: (1) Inadequate lot sizes, side yards, and setbacks
from the roads and water; (2) improperly functioning
sewage disposal facilities; (3) development practices
which lead to extensive erosion; and (4) indiscriminate
tree cutting and filling of wetlands. These problems
result not so much from the particular use placed on
the lot, but from the size of the lot, its suitability for on-
site waste disposal, and the manner and placing of
development. The basic development code is geared to
meet these problems.
Septic systems: The soil of the absorption field is an
integral part of a septic tank system and a source of
frequent failure. When a system fails, the bacteria-
laden effluent backs up into the house or runs out onto
the land surface, causing health hazards. This is
particularly serious when the effluent reaches a water
supply or open water.
Failing septic tank systems and even efficiently
operating systems may contribute to a more subtle type
of pollution. As septic tank effluent seeps through the
soil, filtration of nutrients is incomplete. Depending on
the direction of the groundwater flow, these nutrients
may enter a lake or river. With these additional
nutrients from sewage effluent, weeds and algae may
overproduce and cause nuisances. This problem is
particularly serious in lakes, which do not have the
assimilative capacity of streams.
Since domestic waste disposal is a serious problem
in shoreland areas, a sanitary permit for private
sewage disposal facilities is required prior to building
any structure intended for human occupancy. A permit
will not be issued for areas which cannot properly
absorb septic tank effluent (that is, steep slopes, high
bedrock, high ground water, and impermeable soils)
unless these limitations can be overcome. Sites are to
be checked for limiting conditions by on-site inspection,
including soil borings and percolation tests, and the
use of detailed soil surveys where available. Assuming
-------
PUBLIC BENEFIT AND INSTITUTIONAL PROBLEMS
199
that a site exists with suitable physical properties for
soil absorption of liquid wastes, there are additional
provisions in the "sanitary code" portion of the
ordinance. Detailed standards pertaining to the
construction, location, and maintenance of septic tank
systems are included.
Tree-cutting: These regulations apply to a strip
paralleling the shoreline and extending 35 feet inland
from the water. No more than 30 percent of the length
of this 35-foot-deep strip may be cleared to the depth of
the strip. The cutting of the 30 percent must not create
clearcut openings greater than 30 feet in width. In the
remaining 70 percent, cutting must leave sufficient
cover to control erosion and to screen cars and
structures (except boathouses) visible from the water
unless a special cutting plan is permitted by the board
of adjustment.
Tree-cutting regulations are designed primarily to
protect the scenic beauty of timbered shorelines while
still allowing a view of water from the lot. There are
important secondary benefits. Retaining shoreland
vegetation makes land less vulnerable to erosion.
Substantial shoreland cover can also reduce the
amount of nutrients and other pollutants reaching the
water. The shoreland vegetation uses nutrients
contained in effluent and fertilizers as food. The
vegetation can block other pollutants and debris from
entering the water.
Setbacks: In addition to setbacks from highways
(typical of conventional zoning) all structures except
piers, wharves, and boathouses must be set back from
the water. Setbacks help preserve shore cover, natural
beauty, and wildlife along the land-water fringe. A 75-
foot setback from the water is required of all structures
except boathouses. Increased setbacks are recom-
mended for bodies of water that possess outstanding
fish and aquatic life, shore cover, natural beauty, or
other ecological attributes.
Lot size: A minimum area of 20,000 square feet and
minimum width of 100 feet are required for all new
shoreland lots not served by public sewers. This is the
minimum size considered necessary to achieve other
dimensional requirements such as setbacks from the
water and roads, separating distances between private
sewage disposal facilities and wells or navigable
waters, side yards and parking areas; and the shore-
cover protection strip along the water.
Filling and grading: These provisions are aimed at
reducing erosion from raw soil and controlling filling of
wetlands. Land that has surface drainage toward the
water and is within 300 feet of a navigable waterway
can be filled or graded only by special permit if the
exposed area and the slope exceed a minimum figure.
The permit must be obtained from the board of
adjustment, which can attach a variety of conditions to
minimize erosion.
Excavating: Lagooning and dredging provisions are
designed to protect wetlands, prevent slumping of
sides of excavated areas, and protect fish from oxygen-
depleted conditions which may prevail in improperly
constructed lagoons. A special permit, contingent upon
overcoming these problems, is required for dredging or
constructing any waterway or lagoon, or pond within
300 feet of a navigable water.
Subdivision controls: Regulating the division of land
into lots for sale is an important part of the shoreland
protection ordinance. Percolation tests, soil borings,
detailed soil surveys, and other physical data are used
to determine that a specified percentage of each lot
within the subdivision is free from physical limitations
such as impermeable soils, high ground water, near-
surface bedrock, excessive slopes, and flooding. Lot
size is geared to the degree to which an area is free
from a combination of these limiting conditions. The
20,000 square-foot lot area and 100-foot width is the
minimum size permitted. Lots with less favorable
physical site factors must have a correspondingly
larger size. The presence of limiting factors beyond a
certain point prohibits subdivision.
Planned unit development: These provisions allow a
developer greater flexibility to arrange lots in clusters
rather than in long strips along the shore. The
minimum lot size for each dwelling can be reduced if an
equivalent portion of the subdivision is restricted to
permanent open space. Clustering lots on suitable
terrain reduces land improvement costs and makes
common sewerage and water systems economically
feasible. Wetlands, steep slopes, and other difficult-to-
develop areas can be preserved as scenic assets. One
subdivision in northern Wisconsin is laid out with all
residential development in offshore clusters. The entire
lake is buffered by undeveloped land extending back
200 feet from the shoreline. This shoreline strip is
owned in common by purchasers of residential lots.
The residential clusters, in turn, are linked with each
other and certain recreational facilities by the
shoreland buffer and other commonly-owned green-
ways. The profit the developer foregoes by not
subdividing the high value shoreline property is more
than compensated for by the increased value of the
more numerous offshore lots. Planned unit develop-
ment provisions can permit thoughtful design which
preserves environmental resources while enhancing
property values.
Wetlands: The model approach suggests that all
substantial wetlands in regulated shoreland areas be
placed in "conservancy districts." Wetlands are
defined as areas where ground water is at or near the
surface of the ground much of the year. These areas
are either delineated on U.S. Geological Survey maps
or detailed soil survey maps. Some wetlands along
water provide fish spawning grounds, whereas others
may be prime wildlife habitat. Wetlands are seldom
suitable for building because of septic tank failure,
unstable soil conditions, and seasonal flooding. For
these reasons, ^the conservancy district regulations
limit building development.
Permitted uses of land in conservancy districts
include harvesting of wild crops, forestry, wildlife
preserves, hunting, fishing, the display of certain signs,
and other uses that do not include residential
structures and that have relatively minimal effects on
the natural environment. Special exception uses
include dams, flowages, removal of topsoil or peat,
general farming, cranberry bogs, and other uses that
may substantially affect the environment. Filling and
drainage are also special exceptions and may be used
to overcome the natural development limitations of
-------
200
RESTORATION OF LAKES AND INLAND WATERS
some of the areas. If the board of adjustment grants a
special use permit for filling or drainage, and if the
wetland area is made suitable for building development
in conformance with the conditions imposed, the
county board can then amend the district boundaries to
place the area within another zoning district.
Implementation: To take constitutional constraints
and informational limitations into account, a number of
techniques were used in drafting the ordinance: (1 (The
objectives of the regulations are set forth in consider-
able detail in an introductory statement of purpose
spelling out the relationship between various object-
ives and the means used in the ordinance to
accomplish them. The rationale of various regulatory
sections is elaborated in portions of the ordinance, and
wherever possible, the relationship to public health and
water pollution control is indicated.
(2) Many of the regulatory standards are keyed to the
physical characteristics of the site. This information is
generated at the time of an application for develop-
ment. For example, proposals for all uses involving on-
site sewage disposal must be accompanied by detailed
information about soil permeability, slopes, depth to
bedrock, and height of ground water. This information
is based on percolation and soil boring tests conducted
by a licensed technician.
(3) Uses which are potential sources of pollution or
could have other adverse effects are evaluated on a
case-by-case basis. The applicant for a special
exception permit must supply detailed information
about the proposed use to the county board of
adjustment. The board investigates the likely effects of
the proposed use and decides whether to refuse, grant,
or conditionally grant the special exception permit.
Standards for the board's investigation and conditions
which may be attached to the permit to minimize
detrimental effects are set out in ordinance. If needed,
technical assistance is available from field representa-
tives of the Department of Natural Resources, Division
of Health, Soil Conservation Service, and other
agencies. This combination of detailed standards and
availability of technical assistance lessens the likeli-
hood of arbitrary decisionmaking and the attendant
danger that a court will find the regulations invalid.
WISCONSIN LAKE MANAGEMENT
PROGRAM
Legislative History
Wisconsin's lake management program originated in
the Inland Lake Demonstration Project, a joint venture
by the Wisconsin Department of Natural Resources and
University of Wisconsin-Extension (Born, 1974). The
project was designed to demonstrate the technical
feasibility of several lake protection and rehabilitation
techniques and to examine the institutional capability
for using those lake management tools that showed
potential for practical application.
After 6 years of testing physical methods for
improving water quality in Wisconsin lakes with
various characteristics (seepage, flow-through, reser-
voir) and a review of lake renewal work in other States
and countries (Dunst, et al, 1974), project personnel
concluded that applied lake management, though still
in its infancy, could be carried on in a general
management program. Concurrent examination of
various units of government (State, county, town,
sanitary district) and private groups (lake associations)
indicated that no existing institution had both the
interest and the legal structure to provide the
necessary authority, financing, and long-term com-
mitment for managing lake resources (Klessig, 1973).
The State had neither the staff nor the financial
resources to individually manage 9,800 lakes. Counties
and towns were more concerned with providing
services, such as roads, that permanent residents
(voters) demanded. Sanitary districts had bent toward
lake management, but their enabling authority to do so
was shaky. Lake associations often exhibited strong
interest, but as voluntary groups they had no legal
authority to manage a public resource (Klessig and
Yanggen, 1973, 1975).
The final task of the Inland Lake Demonstration
Project was to draft legislation setting up the necessary
institutional structure. This legislation became Chapter
33 of the Wisconsin Statutes.
The Enabling Legislation
Chapter 33 provides the legal framework for creation
of special-purpose units of government to manage
lakes and for State assistance to these lake districts.
The people who own property around a lake must
initiate information of the lake district under the
legislation, and the lake district must be established by
an official resolution of a general purpose unit of local
government (county, city, village, or town) (Klessig,
1976).
Although the lake district is legally independent, the
law provides that a member of the county board (soil
and water conservation district) and a member of the
local municipal governing body be represented on the
lake district's board of commissioners. This overlap of
governing bodies is designed to facilitate communica-
tion and cooperation among the general purpose units
of local government, which retain all police powers,
such as zoning; the soil conservation district, which
helps landowners retard erosion; and the new lake
district, which focuses on water quality management
but must also cope with the impact of land use
patterns.
Operating under the directives of the annual meeting
and through a board of commissioners, the lake district
is the functional agency for comprehensive manage-
ment of a given lake or chain of lakes. It not only
develops and adopts plans for managing the lake, but
also directs any protection or treatment work and takes
on long-term responsibility for the community re-
source. As a unit of government, the lake district has
the full range of powers to make contracts, hold real
estate, disburse money, and levy a property tax. Its
specific lake management powers include, but are not
limited to; (1) Study of the causes of existing or
potential lake problems; (2) prevention and control of
aquatic weeds; (3) prevention and control of algae; (4)
prevention and control of swimmer's itch; (5) aeration;
(6) nutrient diversion, removal, or inactivation; (7)
erosion control (voluntary cooperation and financial
-------
PUBLIC BENEFIT AND INSTITUTIONAL PROBLEMS
201
assistance for landowners); (8) dredging; (9) treatment
of bottom sediments; and (10) construction and
operation of water-level control structures.
The second part of Chapter 33 provides for technical
and financial aid to lake districts. This aid is based on
the premise that lakes are used by the general public,
and the general public should bear a portion of the
costs of lake management. (A lake district can only be
formed around lakes with public access.) The Wis-
consin Department of Natural Resources is the lead
agency in providing this aid through its Office of Inland
Lake Renewal (Klessig, et al. 1978).
In addition to its new program of assistance to offical
lake districts, the DNR retains its statutory responsi-
bility as the trustee of public water. It must approve
lake management plans submitted by districts and
issue permits for most in-lake activities carried out by
districts.
Information for lake property owners and public
officials is provided by lake resource management
specialists and county-based resource agents of the
University of Wisconsin-Extension.
Organizing a Lake District
The process of lake district development usually
begins when community leaders attend a regional
conference where lake management is discussed by
State lake management professionals. If the com-
munity leaders feel the program is applicable in their
case, they initiate the district formation process. The
County Extension Office plans an educational meeting
in the community. Often accompanied by a DNR
resource manager, a University of Wisconsin-Exten-
sion specialist makes a presentation and provides
attendees with educational materials and guidelines on
using Chapter 33 (Klessig, 1977).
At the conclusion of the meeting, an ad hoc
organizing committee of property owners, with the
assistance of Extension, Soil Conservation Service and
DNR officials, goes about the arduous task of defining a
proposed lake district boundary. Depending on the
character of the lake basin and political realities, the
district may include a narrow strip of cottage lands
around a lake or encompass an entire watershed.
The committee then collects signatures of land-
owners within that boundary. Once the petition has
been signed by a majority of landowners, the town
board or county board holds a hearing and uses the
following criteria from Chapter 33 to decide whether or
not to create the district;
1. Has the petition been signed by at least 51 percent
of the landowners or owners of at least 51 percent of
the land?
2. Is the district necessary?
S.Will the public health, comfort, convenience,
necessity, or public welfare be promoted?
4. Will the included property benefit?
5. Will the district cause or contribute to long-range
environmental pollution?
Operating a Lake District
Five commissioners govern a lake district. They
include three residents or property owners within the
district elected at the annual meeting of the lake district
each summer; one member of the town board, village
board, or city council with the highest assessed
valuation (equalized) in the district and appointed by
that governing body; and one supervisor of the county
soil and water conservation districts (who in Wisconsin
is also typically a county board member), appointed by
the county board.
The commission applies for technical assistance
from the Office of Inland Lake Renewal by compiling all
existing information on the lake and its watershed. A
feasibility study design is prepared by the centralized,
interdisciplinary staff of the Office of Inland Lake
Renewal. If the district decides to proceed, it contracts
with a private consulting firm to collect additional data
as prescribed in the design. A State grant pays for 60
percent of the study cost. The remaining 40 percent is
paid by the lake district through its taxing powers
and/or by local volunteer efforts that reduce the cash
cost of the feasibility study.
The results of the study are returned to the
interdisciplinary team in the Office of Inland Lake
Renewal for analysis and formulation of alternative
methods of protection or rehabilitation. The lake district
selects and modifies the alternatives to conform with
local values and financial resources. The lake district
plan is then submitted to the regional planning
commission and the soil and water conservation
district for comment and finally to DNR for approval.
Following a public hearing held by DNR in the local
area, DNR may approve the plan and provide up to 80
percent funding. Depending on the character and scope
of the project, the Office of Inland Lake Renewal may
prepare a grant application for the district and submit
the proposed project to the Environmental Protection
Agency. If Federal funding is also approved, EPA funds
50 percent of the project, DNR 30 percent, and the lake
district 20 percent.
The lake district decides at an annual meeting
whether to proceed. If a majority of the resident and
non-resident property owners present favor imple-
mentation of the plan, the district commissioners sign
the necessary contracts, and implement a manage-
ment plan. Simultaneously, Extension provides lake
district commissioners with newsletters, handbooks,
personal consultation, and workshops on operating the
new unit of government (Klessig, 1979).
SUMMARY AND CONCLUSIONS
The Shoreland Protection Program and the Lake
Management Program are complementary in philo-
sophy and objectives. Improved water quality is the
goal of both programs.
They both involve a strong State role—the shoreland
program requires that counties adopt regulations
meeting minimum State standards if the county wishes
to avoid State level regulations. The lake program
involves State participation through financial and
technical assistance to lake districts and by requiring
State approval of lake management plans.
Both programs also involve a strong local govern-
mental role. The county is the main actor in the
shoreland program. All Wisconsin counties have
adopted shoreland regulations without direct State
-------
202 RESTORATION OF LAKES AND INLAND WATERS
intervention. These regulations, some of which exceed
State minimum standards, are administered on the
county level. The lake program has resulted in the
formation of over 120 lake protection districts. These
local special purpose districts are undertaking a variety
of lake protection and rehabilitation activities.
These State-local cooperative programs reinforce
each other. While shoreland regulations are largely a
protective measure, the lake program includes both
rehabilitation and protection. Rehabilitative activities
undertaken include storm sewer diversion, chemical
inactivation of nutrients, aeration, dam construction,
and dredging. Protective activities include serving as a
forum to encourage proper administration of shoreland
regulations or upgrading of shoreland standards,
monitoring of septic systems through attainment of
sanitary powers, cost-sharing for manure storage, and
purchase of grass waterways and wetlands.
The Shoreland Protection Program provides a
countywide framework for regulating land use at the
fragile land and water interface. The Lake Management
Program provides the citizens who live near that
interface with a mechanism to manage the specific
resource that attracted them.
REFERENCES
Born, S. 1974. Inland lake demonstration project. University
of Wisconsin-Extension and Wis. Dep. Nat. Resour.
Madison.
Born, S. and D. Yanggen. 1972. Understanding lakes and lake
problems. Publ. G2411. University of Wisconsin-Extension,
Madison.
Dunst, R., et al. 1974. Survey of lake rehabilitation
techniques and experiences. Tech. Bull. 75. Wis. Dep. Nat.
Resour. Madison.
Klessig, L. 1973. Recreational property owners and their
institutional alternatives for resource protection: The case
of Wisconsin lakes. University of Wisconsin-Extension,
Madison.
1976. Institutional arrangements for lake man-
agement in Wisconsin. Jour. Soil Water Conserv. 31:152.
1977. Ten years of education on lakes. Pages 313-
322 in Trans. 42nd N.A. Wild. Nat. Resour. Conf.
_. 1979. Lake districts: A unique organization with a
special purpose. Fisheries 4:10.
Klessig, L., 0. Williams, and G. Gibson. 1978. A guide to
Wisconsin's lake management law. 4th ed. University of
Wisconsin-Extension and Wis. Dep. Nat. Resour. Madison.
Klessig, L. and D. Yanggen. 1973. Town sanitary districts in
Wisconsin: Their legal powers, characteristics, and activi-
ties. University of Wisconsin-Extension, Madison.
1975. The role of lake property owners and their
organizations in lake management. Publ. G2548. University
of Wisconsin-Extension, Madison.
Yanggen, D., and J. Kusler. 1968. Resource protection
through shoreland regulation. Land Econ. 44.
-------
203
SAMPLING STRATEGIES FOR ESTIMATING
CHLOROPHYLL STANDING CROPS IN STRATIFIED LAKES
ROBERT E. STAUFFER
Water Chemistry Laboratory
University of Wisconsin
Madison, Wisconsin
ABSTRACT
The spatial distributions of chlorophyll a in Lakes Mendota and Delavan, Wis. were studied during
the 1971-72 stratified seasons. Both lakes are calcareous and normally support large epilimnetic
chlorophyll standing crops (CSC) composed of planktonic blue-green algae. Profile sampling the
central lake station provides a nearly unbiased estimate (perhaps negative by 2 to 4 percent) of
lake-average CSC with a long-term coefficient of variation (c.v.)-~16 percent. Lake stations at the
margins are individually strongly biased for lake-average CSC and have large c.v.'s. The bias terms
result from the prevailing southwest afternoon breezes, and intermittent north/north west front
passages affecting Wisconsin. The large c.v.'s reflect the important wind shifts between sampling
dates. For3Mendota the mean square deviation (squared percent) in CSC increases approximately
as 100ds where ds is the station separation distance (km). Based on this relationship, the c.v.
(percent) for lake-average CSC is given approximately by: (1) 7 a2/3n~1'2 for a "simple random
sample;" (2) 7 a2/3n~5'6 for a "stratified random sample;" where a is the average length (km)of the
lake's major and minor axes, and n is the number of chlorophyll profiles sampled. For Lake
Mendota, with n = 9 and a stratified design, the estimated c.v. for lake-average CSC is~4 percent.
Stauffer (1980b) examined wind stress effects on the
position-dependent chlorophyll concentration profile
and time scales for lateral redistribution of epilimnetic
chlorophyll standing crops in eutrophic Lakes Delavan
and Mendota in southeast Wisconsin. In this sequel I
consider expected mean square sampling errors for
several total lake (epilimnion) chlorophyll estimators
and the error in estimating changes in lake chlorophyll
standing crop over time.
EXPERIMENTAL
The chlorophyll sampling and analytical procedures
follow Stauffer et al. (1979). Let C(x,y,z,t) be a point
estimate of chlorophyll a concentration (mg rrf3),,
CSQx,y,t) the position-dependent chlorophyll standing
crop (mg rrf2), and t, the lake-average CSC at
time t.
RESULTS
Coefficient of Variation for CSC: Effect of
station separation distance
I now consider the effect of station separation
distanced = I(x2 - x()2 + (y2 - yi)2)1on the root mean
square (RMS) sampling variation between CSC (XL yi, t)
and CSC(x,y,t ).
I divide this by
-------
204
RESTORATION OF LAKES AND INLAND WATERS
Table 1. — Regional coefficients of variation for CSC: Lake Mendota, 1972.
Date
June 24
June 30
July 8
August 6
August 10
August 14
August 22
August 27
September 4
September 21
September 27
October 7
October 13
Lake region* and
sampling stations
NW
ENE
SE
CB
SW
CB
ENE
ENE
ENE
ENE
NW
ENE
NW
ENE
SE
SW
CB
NE
CB
ENE
CB
ENE
9,15
2,16
20,203,7
10,22
4,5
1,1,1,1,*
2,16
2,23
2,23
2,23
9,15
2,23
9,15
2,23
20,21
4,30,55
1,6,10
12,2
1,6
2,23
6,6*
2,23
2.10
1.00
.075
1.25
1.00
1.60
1.00
1.40
1.40
1.40
2.10
1.40
2.10
1.40
0.60
0.80
1.40
1.00
1.25
1.40
1.60
1.40
Mean CSC
region
67
65
137
120
77
146
172
168
268
199
.91
130
110
156
180
148
162
159
59
51
110
151
(mg rrf2)
Lake
76
76
102
102
102
149
149
181
239
216
109
109
140
140
140
161
161
161
57
57
91
133
c.v.(CSC)0
6.7
0.1
39.0
11.6
12.9
8.0
8.9
1.8
2.8
12.4
0.9
26.1
17.6
6.4
4.1
4.6
6.4
6.2
1.8
16.8
10.2
1.6
d.f/
1
1
1
1
1
3
1
1
1
1
1
1
1
1
1
2
2
1
1
1
1
1
"CB = Central Basin. = mean separation distance between stations within region (km), ' estimates for 8 July and 7 October CB based on
the length of scale, u'At , for the relevant time period. "Within region 'd.f = degrees of freedom
Table 2 — Coefficient of variation for CSC: Between lake regions: Lakes Mendota and Delavan, 1972.
Lake and date
Number of
stations
lake
(mgm
CSCmax:CSCm,n
c.v.(CSC) Ratio mean
(%) squares1'
Significance
Mendota:
June 10
June 16
June 24
June 30
July 8
July 17 a.m.
July 17 p.m.
July 23
July 24
July 31
August 6
August 10
August 14
August 22
August 27
September 4
September 13
September 21
September 27
October 7
October 13
Delavan:
247
260
76
102
149
215
226
194
194
206
181
239
216
109
140
161
138
57
57
91
133
"Ratio sigmricani at ovo level.
"Ratio significant at 1% level.
Inter Regional Grand mean regional (Mendota)
1.23
1.28
1.76
3.70
1.57
1.33
3.67
1.58
1.30
1.42
1.18
1.33
1.49
1.79
1.92
1.43
1.58
1.45
1.67
1.83
1.50
10.7
13.0
22.5
47.0
18.4
15.4
54.6
24.4
12.3
14.5
6.9
10.2
12.3
26.2
22.9
13.9
17.3
13.4
17.4
20.9
16.0
1.21
1.78
5.36
23.30
7.86
2.50
31.50
6.29
1.61
2.22
0.50
1.10
1.61
7.25
5.54
2.05
3.17
1.89
3.20
4.62
2.71
June 6
June 20
July 5
July 18
July 30
3
3
3
4
4
105
226
208
268
312
1.24
2.69
1.42
3.42
1.63
10.6
44.3
20.2
54.5
21.6
1.18
20.80
4.32
31.40
4.93
_
**
*«
• •
•*
-------
SPECIAL PROJECTS AND TOPICS FOR ADDRESSING THE TROPHIC STATE
205
I also examined c.v. (CSC) for all station pairs when
Lakes Mendota (n 658) and Delavan (n 25) were
sampled during 1971 and 1972. Because the air
friction velocity, u , is a measure of the surface wind
drift current (Hicks, 1972), the alternative metric,
d* = (d2 + u2At2)1'2was also examined because lake
sampling could not be made perfectly synoptic (Figure
2). Estimates of u. were based on algorithms reported
by Stauffer (1980a).
Figure 2 reveals a triangular scatter pattern with
individual Mendota c.v. (CSC) estimates varying widely
at the longer separation distances. Less scatter is
evident in the Delavan points. The "no-intercept" (valid
becau§e the local c.v. (CSC) estimates-»0) polynomial
regressions for the Mendota and Delavan data are:
(note the very small differences between equations 1
and 1a and between 2 and 2a).
Mendota: np = 658
cV(CSC) = 8.0 ds- 0.79df
(0.6) (0.14)
cV(CSC) = 7.1 dl - 0.59 d*s2
(0.6) (0.13)
Delavan: np = 25
cV(CSC)=13.4ds
(1.9)
cV(CSC)=12.9ds*
(1.8)
R2 = 0.537
FT = 0.541
R2 = 0.685
R2 = 0.693
eq. 1
eq. 1a
eq. 2
eq. 2a
(The standard errors of the regression coefficients are
in parentheses.) Platt, et al. (1970) also found a rapid
increase in c.v. (CSC) for short station separation
distances and then little change for sampling quad-
rants exceeding 2.6 km2.
CVCSC VS. OS
LHKES MENDOTfl RND OELflVBN 1971. 1972
Figure 2. — Variation of c.v. (CSC) vs. ti, for all station pairs:
crosses = Mendota, open symbols = Oelavan: 1971-1972. A:
Best fit "no-intercept" quadratic polynomial (regression
Mendota data). B: Delavan quadratic polynomial regression. C:
Curve based on da relationship inferred from Mendota local
vs. regional vs. inter-regional analysis.
Persistent Regional CSC Variations
I now contrast the regional CSCt values with the
lakewide means tfor the ensemble of available
sampling dates. For the i'th (fixed) lake region and the
t'th sampling date, define the statistic z :
Zit = 100
CSC.(t)
t
"
1
J
eq. 3
Clearly, independent estimates of Zit are generated for
each of the dates when the i'th region was sampled as
part of a lakewide chlorophyll survey. The statistical
moments are:
N 1=1
,I
,R
eq. 4
eq. 5
If the numerator and denominator of Eq. 3 are kept
independent and we approximate c.v. (1/0
= c.v. t, then, following Goodman (1960),
the long term sampling standard deviation (in %) for the
i'th lake region is approximately:
C.V.i
eq. 6
where n is the number of lake station chlorophyll
profiles sampled on the t'th date. Finally, the
approximate long term mean square percentage error
(MSE) in estimating lake-average CSC incurred by
sampling one station profile in the i'th defined lake
region is as follows:
MSE, (CSC) = z + (cAv., (CSC)}2
Z is the "persistent" regional CSC bias of the i'th
lake region, expressed in percent.
The five identified subregions of Lake Mendota and
the three identified subregions of Lake Delavan have
distinct statistical patterns in chlorophyll standing crop
(Tables 3 and 4). The Mendota Central Basin stations
were very nearly unbiased( =-2%):or CSC (not
significantly different than zero even at the a = 0.2
level) over the two year span, and also exhibited the
smallest regional c.v. (CSC) among the regions tested.
Pooling the results for the two Mendota CB stations,
and both years, c.v. (CSC) 14.1 percent, based on 62
d.f. The central basin bias for Delavan is also negative
(-4.0 percent), but the effect lacks statistical signifi-
cance. Both the northeast and southeast regions of
Mendota are biased positively; the converse is true for
-------
206
RESTORATION OF LAKES AND INLAND WATERS
the southwest and northwest lake regions. The higher
mean CSC levels in the eastern region of Lake Mendota
are probably the result of prevailing westerly winds (cf.
Stauffer, 1980a, b). The southwest region of Lake
Delavan has a negative, and the northeast region a
positive, CSC bias (Table 4).
The CSC means of station pairs on opposite ends of
the fetch axes (NW—SE, SW—NE) are very nearly
unbiased and display a reduced variance. For the 14
dates in 1971 when both the Middleton Bay (SW) and
NE stations were visited, the difference between these
two stations averaged 43 percent of the lake-average
CSC, while the mean of the pair differed from the
lakewide mean by an average of only + 2.6 percent. The
comparable percentages were 31 and + 1.8 percent for
Table 3. — Statistical summary of regional CSC departures*
from lake-average: Mendota, 1971-1972.
* CHLOROPHYLL
LAKE MENDOTA
7 JULY 72
Year
1971
cVi(CSC)
VMSE
1972
av.i(CSC)
VMSE
Central
D.H.
-3.3
16.8
17.1
-2.1
13.4
13.6
Basin0
M.L.
-5.0
11.4
12.4
+2.8
14.6
14.9
Peripheral regions
NE NW SW SE
+15.1
26.4
30.4
+2.3
20.3
20.4
-9.8
16.7
19.4
-5.2
24.9
25.4
-11.0
23.2
25.7
-9.7
12.8
16.1
+4.3
23.5
23.9
+13.2
38.0
40.2
"In percent.
°D.H. = Deep hole station. M.L. - Midlake station (Fig. 1).
Table 4. — Statistical summary of regional CSC departures
from lake-average: Delavan 1972.
Percent difference by lake region
Date
June 6
June 20
July 5
July 18
July 30
A., (CSC)
VMSE
Central
(Station 5)
-0.4
+13.8
-13.4
-17.3
-2.9
-4.0
12.2
12.8
Northeast
(Station 7)
-10.4
+35.8
-9.8
+53.5
+25.8
+ 19.0
28.3
34.1
Southwest
(Station 2)
+10.8
-49.6
+23.2
-36.2
-22.9
-14.9
31.0
34.4
eves: vs. DS*
LRKES MEN00TR RND DELHVRN I971. 1972
Figure 3. — Lake Mendota chlorophyll spatial distribution:
July 17, 1972. p.m.
DEPTH BELOW SURFACE meters
Figure 4. — Chlorophyll profiles for Fish Lake, Dane County,
Wis. showing seasonal displacement of chlorophyll maximum
to greater depths accompanying stable stratification, and
progressive nutrient depletion of the eiplimmon. Fish Lake isa
small (A = 0.88 km2), mesotrophic, spring-fed, marl lake with
a late-spring-summer epilimnion boundary at 3-5 m, and a
summer Secchi transparency averaging 4.2 m. Dissolved
oxygen levels are typically highly supersaturated (+5 mg L~1)m
the upper metalimnion.
the 11 SE-NW pairs observed in 1971. The calculated
mean CSC values for the SE-NW and SW-NE station
pairs during the windy afternoon of July 17,1972 were
228 and 224 mg/m2, values which differ from the
morning lakewide mean (215 mg/m2) by a maximum 6
percent. Chlorophyll standing crops at the NW and SE
stations differ by 270 percent in the afternoon (Figure
4).
Anisotrophic Character in Inter-regional c.v.
(CSC)
A detailed analysis of the events of July 17, 1972
(Stauffer, 1980b) showed that lateral gradients in CSC
intensify along persistent wind-fetch axes, i.e., c.v.
(CSC) is anisotropically related to d Thus, station
separation distances corresponding to the inter-
regional level can feature either extreme differences in
CSC (SE vs NW, July 17 p.m.) or modest differences in
CSC (SE vs SW, July 17 p.m.). This explains the
triangular pattern of the Mendota data points in Figure
2. Less scatter was observed in the Delavan data
because of the lake's linearized morphology (cf.
Stauffer, 1980b), with sampling stations lying only
along the principal wind fetch axis.
Stauffer (1980b) showed that lateral gradients
(expressed as percent oftl) increase with
i, probably because large standing crops are
buoyance prone, hence susceptible to wind advection,
Curve C (Figure 2) shows this relation for hyper-
eutrophic Lake Delavan.
Expected error in t
I consider now the expected mean square error in
lake chlorophyll sampling. Assume initially, that
detailed information on antecedent weather conditions
is either lacking, or that shifts in wind magnitude and
-------
SPECIAL PROJECTS AND TOPICS FOR ADDRESSING THE TROPHIC STATE
207
direction during the 48 hours prior to sampling lead to
indeterminancy in the vector predictor of CSC gradients
(Stauffer, 1980b). Hence, we have no conditional
expectation of regional differences in CSC. Assume
also that the lake is box-shaped with side length, a= 6
km; hence, surface area (36 km2) is only slightly less
than that of Mendota (39.1 km2). Recalling that:
M.S.(CSC) = 95 d4'3
eq. 8
CSC for the
what are the expected errors in
following sampling strategies:
1. A simple random sample (srs) of chlorophyll
profiles of size n?
2. A stratified random sample (str) of n equal-area
squares which partition area a2? (Readers unfamiliar
with sampling theory should consult Cochran, 1963.)
For two points randomly placed within a square of
the side length a, the expected (denoted E) mean
square distance btween the points can be shown to be
exactly a2/3. From this relationship, we can
approximate E d!3as a*3/2.08, and
M.S.a(CSC) = 46 a4
eq. 9
where the subscript a denotes that the mean square
variation in CSC is for points randomly placed in a
square with side length a. The expected error in
t * 6.8 a2 3/n1 2
eq. 10
For a = 6 km, the expected error for srs decreases from
22.5 to 11.2 to 7.5 percent as n goes from 1 to 4 to 9.
If we partition the original lake area into component
squares and conduct a stratified random sample with
one profile per component square, the expected error is
c.v. s.r t » 6.8 a2 3/n5'6
eq. 11
In particular, for n = 4,9, the stratified random sample
has expected error of 7.1 and 3.6 percent, respectively.
As in the case of any commodity which is overdis-
persed in the sampling space, a stratified random
sample gives higher precision than a simple random
sample.
A systematic sample can yield improved precision
over a stratified sample and is easier to execute. Earlier
it was shown that a single chlorophyll profile near the
midpoint of the Mendota Central Basin is very nearly
unbiased and estimates lake CSC with an RMS error of
~ 16 percent. A systematic sample for larger n can be
constructed by sampling the lake's midpoint and
station pairs equidistant from the midpoint along the
principal and transverse axes. The relative precision of
the systematic sample (as compared to a stratified
sample) can be expected to increase with increasing
magnitude and predictability of the downwind CSC
gradient. Cochran (1963) notes that whenever the
correlogram for a sampling variate is concave upward,
a systematic sample is more precise than a stratified
sample; this condition is met here.
The expected error in lake-average CSC was directly
estimated for Lake Mendota by comparing the mean
CSC estimates for the morning (3 stations) and
afternoon (4 stations) of July 17, 1972 and the mean
CSC estimates of September 20 and 21, 1971 (3, 5
stations, respectively). Based on 2 d.f., CSC had an
uncertainty of 6.0 percent; this includes the effect of
time for the two short intervals. The estimated error is
gratifyingly close to the prediction of Eq. 11.
If we now assume that the wind has been relatively
steady and strong during the 48 hours prior to
sampling, CSC(x.y) will increase monotonically with
increasing distance downwind along the fetch axis
(Stauffer, 1980b). Hence, the conditional expectation of
CSC(x,y) is biased positive or negative, depending on
whether (x,y) falls in the downwind or upwind region. A
single station near the midpoint of the fetch axis will be
nearly unbiased for lake-average CSC. A systematic
sample involving the midpoint and two symmetrically
placed stations out along the fetch axis will provide a
strong, nearly unbiased estimate of lake CSC. However,
care must be exercised in chlorophyll transect
sampling. Significant elapsed time between stations
can lead to biases in 'because of confounding
between the die! windpower cycle and the station
visitation sequence (Stauffer, 1980a).
Optimal allocation of sampling effort
Expected lateral gradients in CSC influence the
optimal density of the sampling stations within
separate lake regions. In fact, "Neyman optimal
allocation1' dictates that the regional station density
(stationsKm"2) increase approximately as the square of
the expected regional CSC (Cochran, 1963).
Until now we have assumed that chlorophyll samples
were systematically obtained at integral meter depths
below the surface. Under what circumstances should
this design be modified? Again by Neyman allocation;
sample density (samplesm'1) within the profile should
increase with the expected gradient, dc/dz. How then
do we form our gradient expectations? Clearly, under
conditions of elevated turbulence, the vertical spacing
of the epilimnion chlorophyll samples can be increased
markedly, but only until the epilimnion-metalimnion
boundary, h, is approached (Figure 3).
A simple set of decision rules regarding allocation of
sampling effort within profiles can be formulated as
follows:
1. If oxygen supersaturation occurs with the
metalimnion and/or the Secchi transparency depth>0.5
h , then a sharp temperature-dependent chlorophyll
maximum can be expected below h (Denman, 1977;
Fee, 1976; Figure 4). In such cases, the density of
chlorophyll samples along the z axis should be
proportional either to the temperature or oxygen
gradients.
2. If h 3 X Secchi depth, ignore chlorophyll
concentrations below that boundary.
3. Above h, sample proportionally to the temperature
gradient.
-------
208
RESTORATION OF LAKES AND INLAND WATERS
Inferences concerning changes in lake chloro-
phyll over time
The variance of the difference of two random
variables (Hogg and Craig, 1970)
Var(X Y) = Var X + Var Y - 2 Cov(X.Y)
eq. 12
can be used to estimate the expected error in
chlorophyll standing crop changes over any time
interval(ti - to).Under the reasonable assumption that
sampling and analytical errors are independent of the
sampling dates, the Cov term disappears. If a stratified
random sample of size n profiles was obtained on each
date, then under the null hypothesis that standing crop
was in fact equal on the two dates, equations 11, 12,
show that
,) = 9.6 a2/3/n5
eq. 13
where f Ais expressed in percent of«CSC». For n =
1, a 64 percent change in lake CSC would be required
to reject the null hypothesis (time equivalence) at the
.05 level For a systematic sample of n= 1 (Midlake
station), approximately a 45 percent change in CSC is
required. However, for a stratified sample of size n= 4
on each date, only a 20 percent change in is
required for rejection of the null hypothesis at the
a = .05
In some instances, research focuses on documenting
chlorophyll response to a natural or cultural lake
perturbation. The specific effects of storm-activated
nutrient transport or sewage diversion projects are
examples. If an extended ' record is available,
bracketing the "event" date, a time series "interven-
tion analysis" employing a "dummy" variable yields
the strongest inferences on the specific intervention
effect.
Figure 5 illustrates the change in epilimnion
chlorophyll accompanying the powerful July 12-13,
1971 cold front on Lake Mendota (Stauffer and Lee,
1 973). ,.
, or 12 percent of the estimated change in
CHLOHOPH TLL
FISH LAKE , I972
O)
E
I2 0 13 0 14 0
DEPTH BELOW SURFACE meters
Figure 5. — Lake Mendota vertical chlorophyll distributions:
July 8, 14, 1971.
CHLOROPHYLL IN LflKE MENDOTA
WISCONSIN
Figure 6
Table 3. — Statistical summary of regional CSC departures*
from lake-average: Mendota, 1971-1972.
Year
1971
A
cv.,(CSC)
\/MSE
1972
ov.,(CSC)
VMSE
Central
D.H.
-3.3
16.8
17.1
-2.1
13.4
13.6
Basin0
M.L.
-5.0
11.4
12.4
+2.8
14.6
14.9
Peripheral regions
NE NW SW SE
+15.1
26.4
30.4
+2.3
20.3
20.4
-9.8
16.7
19.4
-5.2
24.9
25.4
-11.0
23.2
25.7
-9.7
12.8
16.1
+4.3
23.5
23.9
+13.2
38.0
40.2
"In percent.
°D.H. = Deep hole station. M.L. - Midlake station (Fig. 1).
REFERENCES
Cochran, W. G. 1963. Sampling techniques. 2nd ed. John
Wiley & Sons, New York.
Denman, K. L. 1977. Short term variability in vertical
chlorophyll structure. Limnol. Oceanogr. 22:434.
Fee, E. J. 1976. The vertical and seasonal distribution of
chlorophyll in lakes of the Experimental Lakes Area,
northwestern Ontario: Implications for primary production
estimates. Limnol. Oceanogr. 21:767.
Goodman, L. A. 1960. On the exact variance of products.
Jour. Am. Stat. Assoc. 55:708.
Hicks, B. G. 1972. Some evaluations of drag and bulk transfer
coefficients over water bodies of different sizes. Boundary-
Layer Meteorol. 3:201.
Hogg, R. V., and A. T. Craig. 1970. Introduction to
mathematical statistics. 3rd ed. Macmillan, New York.
Plait, T., L. M. Dickie, and R. W. Trites. 1970. Spatial
heterogeneity of phytoplankton in a near-shore environ-
ment. Jour. Fish. Res. Board Can. 27:1453.
Stauffer, R. E. 1980a. Windpower time series above a
temperate lake. Limnol. Oceanogr. 25:513.
1980b. Windstress effects on chlorophyll distri-
bution in stratified lakes. Limnol. Oceanogr. (In Press.)
Stauffer, R. E., and G. F. Lee. 1973. The role of thermocline
migration in regulating algal blooms. Pages 73-82 in E. J.
Middlebrooks, ed. Modeling the eutrophication process,
Ann Arbor Science, Ann Arbor, Mich.
Stauffer, R. E., G. F. Lee, and D. E. Armstrong. 1979.
Estimating chlorophyll extraction biases Jour Fish Res.
Board Can. 36:152.
-------
SPECIAL PROJECTS AND TOPICS FOR ADDRESSING THE TROPHIC STATE 209
ACKNOWLEDGEMENTS
I am especially grateful for Alan Kutchera's tireless
assistance in sampling lakes and performing analyses
during 1972. Partial financial support was provided under
U.S. EPA Research Grants No. Eutrophication 16010-EHR
and R805281010 and Training Grant No. 5P2-WP-184-04.
-------
210
THE INFLUENCE OF NUTRIENT ENRICHMENT ON
FRESHWATER ZOOPLANKTON
OSCAR RAVERA
Department of Physical and Natural Sciences
EURATOM, J.R.C.
Ispra (Varese) Italy
ABSTRACT
Any relevant change of the environment produces quantitative and/or qualitative effects on the
biota. As a consequence, in eutrophicated lakes, variations of the zooplankton biomass and
structure may be expected. To solve this problem two approaches have been adopted: (1) Relate
the change in zooplankton to the trophic evolution of the environment; and (2) compare the
zooplankton characteristics from lakes with different trophic level. Interrelation between phyto-
and zooplankton, selective predation by vertebrates and invertebrates and competition between
species populations belonging to the same zooplankton association may have a significant
influence on the zooplankton structure and biomass. In addition, introduction of toxic substances
and manmade change in lake hydrology may also modify the zooplankton association.
Consequently, we need more quantitative information on the relationships to estimate the actual
influence of eutrophication on zooplankton and to ascertain if zooplankton are directly or indirectly
influenced by nutrient enrichment. From comparing several lakes, it is evident that the relationship
between trophic level of a water body and structure of its zooplankton seems rather complex. As a
consequence, in spite of the great amount of information on this subject it seems practically
impossible to use zooplankton to classify water bodies on a trophic scale basis, although useful
information on their trophic evolution may be obtained.
INTRODUCTION
Significant modification of the physical and chemical
characteristics of the environment causes more or less
noticeable changes in the biota. Consequently, modifi-
cations in the zooplankton structure and biomass can
be expected in a water body during passage from the
oligotrophic to the eutrophic stage. For the same
reason significant differences should be evident
between zooplankton associations living in lakes with
different trophic stages. Two approaches are commonly
adopted for studying zooplankton modifications follow-
ing the trophic evolution of a water body. In the first
approach the present characteristics of the plankton
are compared with those of the past. The second
approach consists in comparing the characteristics of
zooplankton from lakes with different trophic levels.
Another paper (Ravera, 1980a) discusses causes of
methodological error and the uncertainty in interpret-
ing the results obtained by both methods. Much
information is available on the qualitative and
quantitative changes of the zooplankton in lakes during
their trophic evolution and the characteristics of the
zooplankton from lakes with different trophic levels
(e.g., Edmondson, 1969; Brooks, 1969; Patalas, 1972;
Bonacina, 1977; Ravera, 1977, 1978).
In spite of this abundant information, several aspects
of the problem are still a matter of discussion. In this
paper we briefly describe the most important causes
affecting variations in the zooplankton biomass and
structure in eutrophicated lakes (e.g., changes in
phytoplankton density and species composition). The
opinions of different authors on the most important
points of this problem are compared, and some
examples of zooplankton changes in relation to the
nutrient enrichment of -the lake are reported.
INFLUENCE OF PHYTOPLANKTON
It is generally accepted that nutrient enrichment
modifies phytoplankton structure and increases its
biomass and production. As a consequence of
quantitative and qualitative changes in the phyto-
plankton, variations can be expected in the structure
and biomass of the zooplankton.
This apparent discrepancy between the conclusions
of certain authors (e.g., Patalas, 1972), who noted that
the increase of zooplankton is related to that of
phytoplankton, and those of others (e.g., Nelson and
Edmondson), 1955) who observed a great increase of
the phyto- but not of the zooplankton can perhaps be
justified by the following considerations. A relationship
between phytoplankton and zooplankton production
can be found only where algal production is a limiting
factor. According to Poulet (1978) an increase of
zooplankton is limited by the food supply, if this is
scarce or if consumption is greater than 50 percent of
phyto-production. In several productive lakes, a
considerable quantity of the phytoplankton cannot be
used by zooplankton because the production of the first
is excessive, and zooplankton increase is limited by
other factors. The aliquot of phytoplankton not eaten by
herbivores passes into the detritivorous chain and
accumulates to some extent in the sediments.
-------
SPECIAL PROJECTS AND TOPICS FOR ADDRESSING THE TROPHIC STATE
211
Information is scarce on the importance of the
organic suspended particles in the diet of zooplankton
of lakes with different trophic level; but it is clear that
when algal production is low, zooplankton can use
organic detritus as a food (e.g., Nauwerk, 1963; Heinle
and Flemer, 1975). This change in the diet may mask
the relationship between producers and consumers.
An increase in predator pressure, reported for several
eutrophic lakes, may neutralize the production of
zooplankton previously enhanced by the increase of
primary production (Brooks, 1969).
In addition, zooplankton variations should be more
closely related to variations in the amount of algae
suitable as food than to the total phytoplankton
biomass. For example, the high correlation between
phyto- and zooplankton production found by Hakkari
(1978) resulted from the high percentage of algae
suitable for zooplankton.
With increasing trophy the percentage of small algae
decreases and they are partially replaced by large ones
(e.g., Pavoni, 1963; Gliwicz 1967). According to Nilssen
(1978) large filter feeders (Daphnids) preferentially use
nannoplankton and filtration is more difficult if large
algae are present. Large algae are easily ingested by
grazers (e.g., Cyclopoids) while the smaller zoo-
planktons (Rotifers and Crustacean larvae) select
nannoplankton from a suspension of algae of different
sizes. As a result, large filter-feeders dominate in
oligotrophic lakes, whereas small zooplankters and
Cyclopoids are abundant in nutrient-enriched lakes.
Hrbaceck, et al. (1961) observed that the large filter-
feeders (e.g. Daphnia) have a more efficient filtration
rate than the smaller ones (e.g., Rotifers). Therefore,
the large filter-feeders are more favored than other
zooplankters in water with low phytoplankton density
(oligotrophic lakes).
Interesting results have been obtained by Poulet
(1978) from experiments on feeding five marine
copepod species. They select particles in relation to
abundance and taste but not size. Because of this
behavior population growth may be limited by food, but
interspecific competition grows stronger. These results
seem to contrast with those of other authors. For
example, Hutchinson (1951) justifies the presence of
more than one Diaptomid in the same lake with the
different size of the algae selected by each species
according to its body size.
From experiments carried out on three species of
Daphnia, Korinek (1978) concludes that the duration of
the preadult stage decreases with increase in the algal
concentration. According to this relationship a balance
could be attained between primary production rate and
Cladoceran biomass. Consequently, the growth rate
should be greater in eutrophic than in oligotrophic
lakes under similar thermal conditions. Goulden, et al.
(1978) observed that larger Cladocerans have a higher
fecundity than the smaller and these, in turn, a shorter
growth rate than the former.
On this basis, and considering that fecundity
increases with food density, the author concluded that,
in the absence of predators, Daphnia should be
dominant in water bodies rich in phytoplankton, and
Bosmina in those with a low algal concentration. This
interesting hypothesis will have to be tested by further
investigation because some authors do not agree on
the faster growth rate in Bosmina. In Bosmina,
Ceriodaphnia, and Diaphanosoma a slower growth rate
has been measured than in Daphnia (Novakova, et al.
1978).
While the biomass and the quality of phytoplankton
influence zooplankton, they, in turn, influence phyto-
plankton by their grazing as well as by regenerating
nutrients by excretion (e.g., Harris, 1959). In some
lakes nutrient regeneration assumes great importance.
For example, the quantity of phosphorus excreted by
zooplankton in Lake George (East Africa) amounts to
862 tons/year and that of nitrogen to 3,212 tons(Ganf
and Blazka; 1974). The ecological importance of P-
release by zooplankton and the indirect control exerted
on this process by fish predators are discussed by
Bartell, et al. (1978).
INFLUENCE OF FISH
It is well known that an abundance of food produces
increases in the numbers and growth rate in fish which
feed on zooplankters. The effect of fish predation on
zooplankton has been the subject of several studies
(e.g. Hrbacek, et al. 1961; Brooks, 1968; Hutchinson,
1971; Stenson, 1972).
Because planktivorous fish capture their prey
visually, they select in relation to size but not to
numbers (e.g. Giussani, 1974). The same fish species
may well vary its diet with the season; for example,
Coregonus sp. from Lake Maggiore prefers Daphnia
from April to July and Bythotrephes from August to
November (Giussani, et al. 1977). Although Cyclopoids
are not generally considered to be prey for fish (Allan,
1976), in some lakes and seasons copepodits and
adults are preyed upon by salmonids (Klementsen,
1968; Jacobsen, 1974).
The influence of predation on the body size of the
zooplankters has been clearly demonstrated by Gal-
braith (1966) in a study on a small Jake in Michigan. No
fish lived in this lake for 4 years and Daphnia pulex was
abundant and reached 3 mm length; 4 years after the
introduction of fish, the size of Daphnia decreased to
1.5 mm. Because these small individuals cannot
reproduce, Daphnia pulex has been completely re-
placed by the smaller Daphnia galeata, D. retrocurva,
Bosmina sp., and small copepods. Another example
has been given by de Bernardi and Giussani (1978) for
the eutrophic Lake Annone (northern Italy). Mass fish
mortality occurred in one of the two basins composing
this lake but not in the other. After about 1 month,
small zooplankters dominated the one basin with fish;
in the basin without fish the most frequent filter-feeder
(Daphnia hyalina) increased its size from less than 1
mm to more than 1.6 mm in length. In 10 Norwegian
lakes Langeland (1978) observed that a high frequency
of arctic char (Salvelinus alpi'nus) seems to have a
noticable effect on the large zooplankton prey: for
example, Daphnia galeata and Holopedium gibberum
decreased their body size significantly.
According to some authors (Hrbacek, 1962; Brooks
and Dodson, 1965) intensive predation by fish may
completely eliminate larger zooplankton species, while
the smaller ones become more frequent. In water
bodies without fish, large zooplankters may dominate
-------
212
RESTORATION OF LAKES AND INLAND WATERS
because their filtration rate is higher than that of
smaller zooplankters.
Because fish generally prefer Daphnia to Calanoids
of the same size, other characteristics of the
zooplankton must influence food selection. Zaret, et al.
(1976) observed that in a tropical lake, fish preferred
smaller and less numerous Cladocerans (e.g. Bosmina,
Ceriodaphnia) to larger and more abundant Diaptomus
gatunensis, but in the laboratory the same fish ate
Diaptomus in almost the same amounts as Clado-
cerans. The authors conclude that in natural conditions
the fish feeds only in the surface waters and
Diaptomus migrates towards this layer during the
night, when the fish cannot see potential prey.
la a temperate lake the same authors observed a
similar migration behavior in Daphnia galeata, probably
aimed at reducing predation. Jacobsen (1974) noted a
similar strategy in Megacyclops gigas. It is clear that in
these examples the behavior of the prey protects it
more efficiently than its size.
From experiments on Daphnia pu/ex Jacob (1978)
concludes that fish predation rate depends upon the
combined effect of prey size and light intensity. Some
very interesting preliminary results have been obtained
by Northcote, et al. (1978) who introduced fish into two
small Canadian lakes (Eurice and (Catherine) in which
there were previously no fish but abundant Chaoborus.
The substitution of Chaoborus by fish seems to have
more evident effects on the body size of the prey than
on zooplankton density and composition. For example,
Bosmina, which is not eaten by fish, increased its mean
body size, because in the absence of Chaoborus larger
individuals became more frequent. Among the 10
zooplankters the fish preferred the larger specimens of
Diaptomus kenai and Daphnia rosea, leading to a
decrease in the mean body size of these species.
In spite of this, the authors do not believe that fish
predation is the most important cause of a reduction of
prey body size. Indeed, in late summer, when fish
predation was negligible, the body size of the two
species remained small and in other seasons, when
predation was very active, the number of fish was too
low for elimination of all the large specimens. In
addition, the body size of the crustaceans preyed upon
was very similar in both lakes, in spite of the greater
abundance of fish in Lake Eurice. In conclusion, the
increase in the frequency of juvenile stages of D. kenai
and D. rosea, an effect of the elimination of Chaoborus,
is the most probable cause of the body size reduction.
Heisey, et al. (1977) observed that fish predation on
Daphnia galeata and Daphnia magna should be more
active in the euphotic zone and, consequently, the
larger and/or more pigmented zooplankters are more
vulnerable than the smaller and transparent ones. In
the layers in which light is attenuated and oxygen
concentration is moderate, zooplankters, which syn-
thesize hemoglobin at lower concentrations of oxygen
(and are consequently transparent), will be less subject
to predation than others. These considerations indicate
that in an eutrophicated environment the partial or
complete substitution of one species by another may be
due to several causes and their combinations.
Since fish may have a considerable influence on
zooplankton, any variation in the behavior or frequency
of these predators is reflected in the zooplankton. In
eutrophic lakes, the population density of a coarse fish
increases at the expense of salmonids and coregonids
(Larkin and Northcote, 1969). For example, in Lake
Constance, during the last 50 years the total amount of
fish has tripled and in Baldeggersee (Switzerland)
around 1940, the white fish were replaced by perch,
carp, and pike. New species of fish may modify the
predation on zooplankton. In addition, some plank-
tivorous fish, because of an increase of plankton
biomass and/or other causes (i.e. reduction of
macrophytes) prey upon pelagic zooplankton (Quartier,
1965; Pignalberi, 1967). Any cause producing high
mortality in fish (i.e., industrial effluents, water
acidification) reduces the predation pressure on
zooplankton. A decreased predation may also be
caused by fish diseases that are more frequent in
eutrophic waters than in oligotrophic. For example, in
some eutrophic lakes of Northern Italy mass mortality
of bleak (Alburnus alborella) due to branchiomycosis,
has been frequently observed. This infection seems to
be favored by high concentrations of un-ionized
ammonia often occurring in eutrophic waters (Gi-
ussan et al. 1976).
PREDATION EFFECT BY
INVERTEBRATES
The relationships between invertebrates, predator
and fish, show a double aspect. In other words, an
invertebrate predator may be used as food by fish or
they may compete for food. Kajak, et al. (1970)
calculated that in two Polish lakes Chaoborus larvae
daily remove 7 and 13 percent of the total zooplankton.
In Leechmere, 94 percent of the Daphnia mortality was
caused by Chaoborus predation (Dodson, 1962).
Gliwicz, et al. (1978) estimated that in some water
bodies predation pressure by fish is far less important
than that of Chaoborus, Crustaceans, and Rotifers.
Size selection by fish is the opposite of that by
invertebrate predators, because the first prefer larger
zooplankters and the latter the smaller (Landry, 1978).
For example, Brandl, et al. (1978) observed that Cyclops
vicinus and Mesocyclops edax prey selectively upon
smaller zooplankters, such as Rotifers.
Consequently, where predation by fish is severe,
small zooplankton dominate, whereas larger forms
could be abundant in lakes dominated by invertebrate
predators.
If this is true, the abundance of large or small
zooplankters is controlled more by the nature of the
predation than by the trophic evolution of the water
body. For example, O'Brien (1975) observed that in
those lakes of the Noatak drainage basin (Alaska) in
which there were no fish but a Calanoid predator
(Heterocope septentrionalis), there were no small
zooplankton. Fish prey visually, whereas zooplankton
predators use mechanoreceptors and chemoreceptors.
Consequently, the prey may escape from plantivorous
fish by migrating into deep layers (Zaret, et al. 1976),
but by means of alternative behavioral responses (i.e.,
"dead-man" response) and morphological structure
(i.e., spines) zooplankton prey protect themselves from
zooplankton predators (Friedman, etal. 1975). Stenson
(1976) observed that the ratio of Bosmina coregoni to
Bosmina longirostris, calculated for eight small lakes,
-------
SPECIAL PROJECTS AND TOPICS FOR ADDRESSING THE TROPHIC STATE
213
is controlled by fish as well as by other predators (e.g.
Chaoborus, Cyclops, Bythotrephes), which, in their
turn, are preyed upon by fish. A study by Dodson (1975)
of the predation behavior of 12 zooplankton species
revealed that each predator was preyed upon by some
other species. These examples and others clearly
demonstrate that prey-predator relationships are
generally complex.
Zaret (1978) divided predators into three classes: (1)
gape-limited predators (GLP), planktivorous fish that
select visible prey; (2) size-dependent predators (SDP),
invertebrates that prey upon small zooplankters. The
third class arises from the coexistence of GLP and SDP.
The zooplankters may escape from predators of the
third class by means of their small size, and
morphological structures which make capture more
difficult. In water bodies, in which predation is the most
important factor controlling the zooplankton structure,
the dominant forms should be Ceriodaphnia, Bosmina.
Diaphanosoma. and small Diaptomus if GLP are
abundant. In the presence of abundant SDP Daphnia
pulex and large Diaptomus could attain high density
values. With predation by both GLP and SDP, Daphnia
galeata. Holopedium gibberum and Ceriodaphnia
lacustris could be the most frequent species.
CHANGES IN ZOOPLANKTON
STRUCTURE AND DENSITY
Lake Lugano. This deep lake, lying between
Switzerland and Italy, is now heavily eutrophicated
(Ravera, 1980b), but the first effects of nutrient
enrichment were observed more than 30 years ago
(Baldi, et al. 1949; Jaag, et al. 1970). Relevant
variations in zooplankton structure have occurred
during the last decades (Ravera, 1978, 1980a);
Eudiaptomus padanus, Mixodiaptomus laciniatus, and
the genus Sida have been eliminated and population
density of Daphnia obtusa has increased. Because
Cladocera and Rotifers have no Diaptomid competitor,
they have probably attained higher population density
than in the past, but due to the different sampling
method the reliability of a comparison between our
data and those of the preceeding authors is rather
small. In addition, the abundance of Daphnia hyalina
could indicate a moderate fish predation. Our finding
agrees with McNaught (1975), who explains the
succession of the Cladocerans to the Calanoids with
the better pre-adaptation of the former to the eutrophic
environment and particularly to the algal size and
frequency.
Lake Maggiore (Northern Italy). In 1950 this deep
lake was oligotrophic. The first Oscillatoria rubescens
blooms were observed in 1967 (Ravera, et al. 1968)
after a heavy bloom of Tabellaria fenestrata had
occurred some years before. From 1909 to 1973 the
considerable increase of Daphnia hyalina and Chy-
dorus sphaericus and the decrease of Mesocyclops
leuckarti seem to show the increasing eutrophication
of this lake. During the last 35 years Diaptomids
decreased from 40 to 37 percent and Cyclopids from 23
to 8 percent; consequently, the Cyclopoids/Diaptomids
ratio diminished (Bonacina, 1977). This is not in
agreement with Patalas (1972), who observed a
reduction of the population size of Calanoids and an
increase of Cyclopoids. The change in zooplankton
structure was more rapid in Lake Lugano than in Lake
Maggiore probably because the eutrophication rate is
slower for Lake Maggiore as compared with that of
Lake Lugano.
Lake Mergozzo (Northern Italy). Until 1967 this deep
lake was a water body moderately enriched by
nutrients. In 1969-1970 blooms of Oscillatoria ru-
bescens were observed (Zutshi, 1976) and some
modifications of the benthos structure testified to the
progressive eutrophication of this lake. This seemd to
be caused by an increase of the nutrient loading and
the temporary increase of the hydrological renewal
time. During 1975 the chemical characteristics of the
pelagic waters and those of the phytoplankton
demonstrated that this lake has returned to a
mesotrophic stage (Saraceni, et al. 1978). From a
comparison of the data on zooplankton collected from
1949 to 1975, a significant increase in the population
density of Copepods, Cladocerans, and Rotifers is
evident. In addition, Sida cristallina, Mixodiaptomus
laciniatus, and some Rotifers disappeared, whereas
Daphnia hyalina increased and Mesocylops hyalinus
has been almost completely replaced by M. leuckarti
(De Bernardi, et al. 1978). The increase of total
zooplankton density from 1949 to 1970 may be caused
by nutrient enrichment (Ferrari, et al. 1976).
The increase of zooplankton density from 1970 to
1975 (in spite of an improvement in the conditions of
the lake) is not clear and may be the effect of the
delayed response of the zooplankton to the decrease of
nutrient concentration in the pelagic waters. A strict
relationship between trophic level and zooplankton
density and biomass has been found by Godeanu
(1978) in several lakes of Northern Germany.
Lakes ofBrianza (Northern Italy). These shallow lakes
(Annone, Oggiono, Alserio, Pusiano, Segrino, and
Montorfano) have a high trophic level, except for the
latter which may be considered oligo-mesotrophic.
From the data reported by Bonomi, et al. (1967) and
Gerletti and Marchetti (1977) the following conclusions
can be drawn: (1) From 1954 to 1972 the population
density of Copepods increased in Lake Segrino and
decreased in Lake Motorfano; (2) in Lakes Alserio,
Pusiano, and Segrino the total zooplankton density was
higher in 1972 than in 1957 but lower in Lakes
Annone, Oggiono, and Montorfano; and (3) the
population density was roughly in agreement with the
nutrient enrichment.
It is rather difficult to explain the changes in
population density from 1954 to 1972 as being the
effect of nutrient load. Indeed, the changes in
population density did not follow a trend, except for
Lake Segrino and Lake Montorfano. On the other hand,
there is no evidence that in Lake Montorfano this
decrease of nutrient load occurred to justify the
decrease of zooplankton. For the same reason it is
rather difficult to explain the considerable reduction in
zooplankton density, particularly of Copepods, during
1967. It is probable that the changes in zooplankton
density from 1954 to 1972 are long-term fluctuations
controlled by normal meteorological conditions. This
hypothesis agrees with the conclusion drawn by
-------
214
RESTORATION OF LAKES AND INLAND WATERS
Edmondson (1972) for Lake Washington. In this lake no
significant change was evident in the zooplankton
structure, in spite of the considerable nutrient load
received over a long period of time. As far as the
increase of zooplankton biomass is concerned, Beeton
(1965) and Patalas and Salki (1973) observed that this
increase occurs in some eutrophicated lakes, but it
cannot be generalized. Relevant quantitative and
qualitative modifications in the zooplankton have been
observed during the trophic evolution of Lake Nemi
(Central Italy) by Stella, et al. (1 978). These changes are
partly due to the massive fungal epidemics affecting
zooplankton as well as phytoplankton and fishes.
A clear example of the influence of the nutrient
loading from domestic and agricultural effluents on the
zooplankton is given by Lake Valencia (Venezuela) (De
Infante, 1978). From 1968 to 1976 the population
density of zooplankton almost tripled, probably because
of an increase of phytoplankton of a factor of 103. This
increase varied with the taxa; a conspicuous increase
of Rotifers and significant decrease of Cladocerans
were observed. Predation by fish and Chaoborus most
likely reduced the Cladoceran density. In addition,
Notodiaptomus venezo/anus has been replaced by
Thermocyclops hyalinus.
INDICATORS OF TROPHIC CONDITIONS
Some zooplankton species have been identified as
indicative of very productive water bodies (e.g. Bosmina
longirostris (Brooks, 1969); Daphnia cucullata, Acan-
thocyclops bicuspidatus (Dussart, 1969); Brachionus
angularis, Brachionus calyciflorus (De Beauchamp,
1 965). Pejler (1 957) proposed a list of Rotifer indicators
of different trophic stages for Swedish lakes.
According to McNaught (1975) oligotrophic waters
are more suitable than eutrophic for Diaptomus
because of their high filtration and ingestion rate at low
density of nannoplankton. There are some exceptions
to this statement; for example, Eudiaptomus padanus
may attain a high population density in very eutrophic
lakes (e.g. Lake Varese and Comabbio) and Eudi-
aptomus gracilis is distributed in both eutrophic and
oligotrophic lakes. Mixodiaptomus laciniatus seems to
be more sensitive to eutrophic conditions. In fact, this
species has completely disappeared from Lake Lugano,
in which it was very abundant when this lake was
oligotrophic (1947), and its percentage was decreased
considerably in Lake Maggiore because of progressive
•nutrient enrichment (Tonolli, 1962). The same species
disappeared from Lake Mergozzo (Northern Italy) from
1970 to 1 975 for the same reasons (de Bernardi, et al.
1976). In oligotrophic waters, Bosmina generally have
low population density because of comparatively low
filtration-rate, but its small size, high birth rate, and
capacity to feed on algae of different size, favor its
diffusion in eutrophic waters (McNaught, 1975). The
same author observed that Daphnia occurs both in
oligotrophic and eutrophic lakes, because of its high
birth rate and filtering capacity and because it can
ingest algae of different sizes. On the other hand, its
large size facilitates fish predation.
In many cases it is rather difficult to classify the
degree of trophy of a lake only on the basis of a list of
the species living in it. For example, some Rotifers
considered indicators of eutrophicated water (Keratella
quadrata) have also been collected from oligotrophic
lakes. This discrepancy may be due to the different
ecogenotypes composing these species (Hutchinson,
1967). The well-known examples reported by Minder
(1938)for Lake Zurich and by Deevey (1942) for Linsley
Pond (Connecticut), demonstrate that progressive
enrichment of the water bodies causes replacement of
the larger Bosmina coregoni by the smaller B.
longirostris. According to Hutchinson (1967) the
presence of Bosmina longirostris in productive lakes
and of B. coregoni and B. longispina in the less
productive ones is not an absolute rule.
This has been demonstrated by Findenegg (1943),
who collected Bosmina longirostris from the epilimnion
and B. coregoni from the hypolimnion of the same lake.
In addition, a rich population of B, coregoni (highest
value attained is 38,000 individuals/liter) has been
studied by Dumont (1967) in the very eutrophic Lake
Donk (Belgium). Because the substitution of B.
coregoni with B. longirostris does not seem to be a
general rule, Mikulski (1978) proposed to apply the
ratio Chydorus/Bosmina for estimating the trophic
stage of the lake. The value of this index of
eutrophication increases with the trophic evolution of
the water body. For example, its value for mesotrophic
lakes is about 0.5 whereas it ranges between 1 to 7 for
eutrophic.
For some species we need sophisticated taxonomical
identification. For example, Cyclops scutifer scutifer
lives principally in oligotrophic waters, while C. scutifer
wigrensis is a typical form of meso- and eutrophic lakes
(Hutchinson, 1967). To evaluate the relationship
between the trophic evolution of a water body and the
zooplankters inhabiting it, all available information on
the ecology and the recent history of the lake should'be
taken into account.
DISCUSSION AND CONCLUSIONS
Algal growth nutrients cannot directly influence
zooplankton, because they are not able to use mineral
nutrients as phytoplankton do. On the other hand,
nutrient concentration is normally too low to be toxic
for most zooplankton species. Nutrient enrichment
indirectly influences zooplankton in different ways: for
example, by increasing the primary production, varying
the biomass and composition of phytoplankton, or
depleting hypolimnetic oxygen. This represents the
fundamental difficulty in establishing a clear relation-
ship between trophic evolution and zooplankton
density and structure.
For the same reason phytoplankton seems to be a
more useful indicator of the trophic conditions than
zooplankton. In a stable water body zooplankton
maintains constant structure, biomass, and production,
obviously with seasonal variations and annual fluctua-
tions. As a result of natural causes and anthropogenic
influences, these ideal conditions are not the rule, but
the exception. Qualitative and quantitative modifica-
tions of the zooplankton are often evident. The
abundance and structure of the zooplankton are
controlled by food, predators, and parasites, in addition
-------
SPECIAL PROJECTS AND TOPICS FOR ADDRESSING THE TROPHIC STATE
215
to the physical and chemical factors. Therefore, to
survive, zooplankton species must develop a complex
strategy, which consists essentially of adapting itself to
the available food, competing successfully with other
species, and protecting itself against predators. When
the physical environment is modified by eutrophication,
the zooplankton strategy must generally be changed
because of variations in the predation pressure and
food produced by the trophic evolution. Any species
which cannot adapt itself to the changed conditions —
or whose adaptation turns out to be inefficient — is
eliminated or reduced in number and biomass.
When a species is eliminated because its food supply
has become too scarce, its niche disappears; but if the
species is eliminated by predation, its niche may be
occupied by other species better protected against
predators. In addition, the trophic evolution favoring
some phytoplankton species (recently immigrated or
previously rare), may create new niches available for
new zooplankton species.
The relative importance of predation, feeding, and
competition varies in different communities. Conse-
quently, all the hypotheses proposed to explain the
modification of the zooplankton abundance and
composition are probably reliable, but their importance
varies with the lake under consideration. Indeed, in
spite of the various studies on this subject very few
results can be generalized, with the exception of an
increase in the zooplankton frequency and biomass
observed in eutrophicated lakes and the influence of
the predator on the body size of the prey.
There is general agreement on the latter relation-
ship, but more hypotheses have been proposed for
identifying the mechanisms responsible for it. Several
authors have observed that fish select large prey and
may eventually eliminate the prey. It would be
interesting to know the fate of the predator fish
following the extinction of the large prey. We may
imagine that the extinction of the fish follows that of
the prey, or the fish may switch its predation from the
large forms to the small. If the first hypothesis is true,
fish population density should decrease simultane-
ously with the decrease of the prey, but before the
latter are gone. Unfortunately, quantitative information
on this subject is almost absent. If the second
hypothesis is reliable it seems rather unlikely that the
fish eat large prey exclusively. It seems more likely that
the fish, with the decrease of its preferred prey,
introduces into its diet an ever-growing percentage of
small prey. If this is true the reduced predation
pressure could permit an increase in the density of
large prey in agreement with the Volterra-D'Ancona
model. This increase would have to be more rapid and
consequently more probable, in those large partheno-
genetic prey species having a high intrinsic rate of
natural increase such as Daphnia, than in the small
Cladocerans. When there is no previous information,
the absence of large zooplankters in water bodies rich
in planktivorous fish, does not necessarily demonstrate
that fish have eliminated the large zooplankters, but
that, under certain circumstances the fish diet may
consist only of small forms.
The problems concerning predation and competition
require more information, but our knowledge of the
interactions between these processes is even scarcer.
One of the few examples is given by Bossone, et al.
(1954) on the competition between two Diaptomids and
the predation on them by Heterocope sal/ens in a small
alpine lake (Northern Italy). In addition, studies in
natural and semi-natural conditions, e.g. "micro-
ecosystems",nfluence of zooplankton on phytoplankton
exerted by grazing and nutrient mineralization are
fundamental for understanding the interrelationships
between nutrients and zooplankton changes.
In several cases it is rather difficult to discriminate
between the effects on the zooplankton structure and
biomass caused by eutrophication and those produced
by other causes interfering with the trophic processes
(i.e., industrial and mining pollution, introduction of
new species, overfishing, yearly meteorological chang-
es). As a result, one may suppose that eutrophication is
the only cause of zooplankton modifications if these
show a well-defined trend over a series of years, and if
there are no other apparent influences acting upon the
zooplankton.
REFERENCES
Allan, J. D. 1976. Life history patterns in zooplankters. Am.
Nat, 110:165.
Baldi, E., L. Pirocchi, and V. Tonplli, 1949. Relazione
preliminare sulle ricerche idrobiologiche condotte sul Lago
di Lugano. Ispettorato Federals Svizzero per la Pesca,
Berna:1.
Bartell, S. M., and J. F. Kitchell. 1978. Seasonal impact of
planktivory on phosophorus release by Lake Wingra
zooplankton. Verh. Int. Verein. Limnol., 20:466.
Beeton, A. M. 1965. Eutrophication of the St. Lawrence Great
Lakes. Limnol. Oceanogr. 10:240.
Bonacina, C. 1977. Lo zopplancton del Lago Maggiore:
situazione attuale e modificazini a lungo termine delta
struttura comunitaria. Mem. 1st. Ital. Idrobiol. 34:79.
Bonomi, G., C. Bonacina, and I. Ferrari. 1967. Caratteristiche
chimiche plancton e benton nel quadro evolutive recente dei
laghi briantei. Mem. 1st. Ital. Idrobiol. 21:241.
Bossone, A., and V. Tonolli. 1954. II problema del la
convivenza di Arctodiaptomus bacillifer (Koelb), di A-
canthodiaptomus denticornis (Wierz.) e di Heterocope
saliens, Lill. Mem. 1st. Ital. Idrobiol. 8:81.
Brandl, Z., and C. H. Gernando. 1978. Prey selection by the
cyclopoid copepods Mesocyclops edax and cylops vicinus.
Verh. Int. Verein. Limnol., 20:2505.
Brooks, J. L. 1968. Eutrophication and changes in the
composition of the zooplankton. Proc. Symp. Eutrophication,
Natl. Acad. Sci.
Brooks, J. L, and S. I. Dodson. 1965. Predation, body size and
composition of plankton. Science 150:28.
De Beauchamp, P. 1965. In P. P. Grasse. Traite' de zoologie.
IV (III) Masson et C. .e: 1347.
De Bernard!, R., and G. Giussani. 1978. Effect of mass fish
mortality on zooplankton structure and dynamics in a small
Italian Lake (Lago di Annone). Verh. Int. Verein. Limnol.
20:1045.
De Bernardi, R. and E. Soldavini. 1976. Long-term
fluctuations of zooplankton in Lake Mergozzo, Northern
Italy. Mem. 1st. Ital. Idrobiol. 33:345.
Deevey, E. S. Jr. 1942. Studies on Connecticut Lake
sediments. Am. Jour. Sci. 240:233.
1969. Cladoceran populations of Rogers Lake,
Connecticut, during late and postglacial time. Mitt.
Int.Verein. Limnol. 17:56.
-------
216
RESTORATION OF LAKES AND INLAND WATERS
De Infante, A. 1978. Lakes. Central and South America.
Zooplankton of Lake Velencia (Venezuela) I. Species
composition and abundance. Verh. Int. Verein. Limnol.
20:1186.
Dodson, S. L. 1972. Mortality in a population of Daphnia
rosea. Ecology 53:1011.
1975. Predation rates of zooplankton in arctic
ponds. Limnol. Oceanogr, 20:426.
Dumont, H. J. 1967. A five day study of patchiness in
Bosm/na coregoni, Baird in a shallow eutrophic lake. Mem.
1st. Ital. Idrobiol. 22:81.
Dussart, B. 1969. Les copepodes des eaux continentales
d'Europe Occidentale. II: Cyclopoides et Biologie M. Boubee
et C. ie Ed., Paris.
Edmondson, W.T. 1969. Cultural eutrophication with special
reference to Lake Washington. Mitt. Int. Verein. Limnol.
17:19.
1972. The present conditions of Lake Washington.
Verh. Int. Verein. Limnol. 18:284.
Hrbacek, J., et al. 1961. Demonstration of the effect of the
fish stock on the species composition of zooplankton and the
intensity of metabolism of the whole plankton association.
Verh. Int. Verein. Limnol. 14:192.
Hutchinson, B. P. 1971. The effect of fish predation on the
zooplankton of ten Adriondack lakes, with particuar
reference to the alewife, Alosa pseudoharengus. Trans. Am.
Fish Soc. 2:325.
Ferrari, I., and 0. Rossi. 1976. Composition and seasonal
succession of the zooplankton in Lake Mergozzo (Northern
Italy). Mem. 1st. Ital. Idrobiol. 23:85.
Findenegg, L. 1943. Zur Kenntnis der planktischen Clado-
ceren Karntens. Carintia II: 47
Friedman, M. M., and J. R. Stnckler. 1975. Chemoreceptors
and feeding in calanoid copepods (Arthropoda: Crustacea).
Proc. Natl. Acad. Sci. 72:4185.
Galbraith, M. G. Jr. 1966. Size-selective predation on
Daphnia by rainbow trout and yellow perch. Trans. Am.
Fish. Soc. 96:1.
Ganf, G. G., and P. Blazka. 1974. Oxygen uptake ammonia
and phosphate excretion by zooplankton of a shallow
equatorial lake (Lake George, Ugande). Limnol. Oceanogr.
19:313.
Gerletti, M., and R. Marchetti 1977. Indagine sui laghi della
Bnanza. Quaderni dell 'Istituto di Ricerca sulle Acqu, 19:7.
Giussani, G. L., and R. de Bernardi. 1977. Food selectivity in
Coregonus sp. of Lago Maggiore: an energetical approach.
Mem. 1st Ital. Idrobiol. 34:121.
Giussani, G. L., I. Borroni, and E. Grimaldi. 1976. Role of un-
ionized ammonia in predisposing gill apparatus of Alburnus
alburnus alborella to fungal and bacterial diseases. Mem
1st. Ital. Idrobiol., 33:161.
Gliwicz, A. M. 1967. The contribution of nannoplankton in
pelagial primary production in some lakes with varying
trophy. Bull. Acad. Polonaise Sci. Ser. Sci. Biol. 15:343.
Gliwicz, Z. M., A. Hillbricht-llkowska, and T. Weglenska.
1978. Contribution of fish and invertebrate predation to the
elimination of zooplankton biomass in two Polish lakes.
Verh. Int. Verein. Limnol. 20:1007.
Goulden, C. E., and C. Wilson Horing. 1978. Why do large
zooplankton species dominate? Verh. Int. Verein. Limnol
20:2457.
Hakkari, L., 1978. Interrelationship of phytoplankton and
zooplankton production in some lakes of Central Finland.
Verh. Int. Verein. Limnol. 20:905.
Harris, E. 1959. The nitrogen cycle in Long Island Sound. Bull.
Bingham Oceanogr. Collect. 17:31.
Heinle, D. R. and D. A. Flemer. 1975. Carbon requirements of
a population of the estuarme copepod Eurytemora affins
Mar Biol. 31: 235.
Heisey, D., and K. G. Porter. 1977. The effects of ambient
oxygen concentration on filtering and respiration rates of
Daphnia galeata mendotae and Daphnia magna. Limnol.
Oceanogr. 22:839.
Hrbacek, J. 1962. Species composition and the amount of
zooplankton in relation to fish stocks. Rozpr. Cesk. Akad
Ved., 72:1.
1951. Copepodology for ornithologist.
Hutchinson, G. E.
Ecology 32:571.
1967. Introduction to lake biology and the
limnoplankton (II). John Wiley and Sons, Inc., New York.
Jaag, 0., and E. Marki. 1970. Recent developments and
present-day situation in relation to the lake of Lugano and
its tributaries. In Euthophication in large lakes and
impoundments. Uppsala Symp. OECD, Paris.
Jacobs, J. 1978. Influence of the prey-size, light intensity,
and alternative prey on the the selectivity of plankton
feeding fish, Verh. Int. Verein. Limnol. 20:2461.
Jacobsen, O. J. 1974. Feeding habits of the population of
whitefish (Coregonus lavaretus (L.) in Haugatjern — a
eutrophic Norwegian lake. Norw. Jour. Zool. 22:295.
Kajak, Z., and B. Ranke-Rybicka. 1970. Feeding in predation
efficiency of Chaoborus flavicans Meigen (Diptera, Culcidae)
larvae in eutrophie and dystrophie lake. Pol. Arch.
Hydrobiol. 17:225.
Klementsen, A. 1967/68. On the feeding habits of the
population of brown troup (Salmo trutta, L.) in Jrflstervann,
West Norway, with special reference to the utilization of
planktonic crustaceans. Norw. Jour. Zool. 15:50.
Korinek, V. 1978. Slow growth rates of Daphnia populations
from oligotrophic lakes. Verh. Int. Verein. Limnol. 20:2467.
Landry, M. R. 1978. Predatory feeding behavior of a marine
copepod, Labidocera trispinosa. Limnol. Oceanogr. 23:1103.
Langeland, A. 1978. Effect of fish (Salvelinus alpinus, arctic
char) predation on the zooplankton in ten Norwegian lakes.
Int. Verh. Verein. Limnol. 20:2065.
Larkin, P. A. and T. G. Northcote. 1969. Fish as indices of
eutrophication. Pages 256-273 in Eutrophication: Causes,
consequences, correctives. Natl. Acad. Sci., Washington
D.C.
McNaught, D. C. 1975. A hypothesis to explain the succession
from calanoides to cladocerans during eutrophication. Verh.
Int. Verein. Limnol. 19:724.
Mikulski, J. S. 1978. Value of some biological indices in case
histories of lakes. Verh. Int. Verein. Limnol. 20:992.
Minder, L. 1938. DerZurichseeals Eutrophierungsphanomen.
Geol. Meere Binnengewasser 2:284.
Nauwerck, A. 1968. Die Beziehungen zwischen Zooplankton
und Phytoplankon im See Erken. Symp. Bot. Upsal. 17:1.
Nelson, P. R., and W. T. Edmondson. 1955. Limnological
effects of fertilizing Bare Lake, Alaska. Bull. U.S. Fish Wildl.
Serv. 56:414.
Nilssen, J. P. 1978. Eutrophication, minute algae and
inefficient grazers. Mem. 1st. Ital. Idrobiol. 36:121.
Northcote, T. G., C.J.Walters, and J. M. B. Hume. 1978. Fish
and fisheries. Initial impacts of experimental fish introduc-
tions on the macrozooplankton of small oligotrophie lakes.
Verh. Int. Verein. Limnol. 20:2003.
Novakova, J., et al. 1978. Growth rate as a factor in the
competition among Cladocera. Verh. Int. Verein. Limnol.
20:2468.
O'Brien, J. 1975. Some aspects of the limnology of the pond
and lakes of the Noatak drainage basin, Alaska. Verh. Int.
Verein. Limnol. 19:472.
Patalas, K. 1972. Crustacean plankton and the eutrophication
of St. Lawrence Great Lakes. Jour. Fish. Res Board Can.
19:1451.
-------
SPECIAL PROJECTS AND TOPICS FOR ADDRESSING THE TROPHIC STATE 217
Patalas, K., and A. Salki. 1973. Crustacean plankton and the
eutrophication of lakes in the Okanagan Valley, British
Columbia. Jour. Fish. Res. Board Can. 30:519.
Pavoni, M. 1963. Die Bedeutung des Nannoplanktons im
Vergleich zum Netz-plankton. Qualitative und quantitative
Untersuchungen im Zurich-see, Pfaffikersee und anderen
Seen. Schweiz. Z. Hydro). 25:219.
Pejler, B. 1957. Taxonomical and ecological studies on
planktonic Rotatoria from Central Sweden. Kungl. Svenska
Vetenskapsakad. Handlingar 7:5.
Pignalberi, C. 1967. Importanza dei diversi gruppi di
organism! nell'alimentazione estiva delle specie ittiche
litorali del Lago maggiore. Mem. 1st. Ital. Idrobiol. 21:89.
Poulet, S. A., 1978. Comparison between five coexisting
species of marine copepods feeding on naturally occurring
paniculate matter. Limnol. Oceanogr. 23:1126.
Quartier, A. A. 1965. devolution de la peche dans le lac de
Neuchatel. Bibl. Mus. Ville de Neuchatel, Rapport 1965:1.
Ravera, 0.1977. Effects of eutrophication on the zooplankton
of a subalpine lake: Lake Lugano. Pages 97-104 in W. K.
Downey and G. Ni Did, eds. Lake pollution eutrophication
control. Natl. Sci. Counc. Dublin.
1978. Eutrophication of Lake Lugano "read" by
means of planktonic remains in the sediment. Schweiz. Zeit.
Hydrol. 40:40.
1980a. Effects of eutrophication on zooplankton.
IQWFR Post Conf. Seminar P5 Eutrophication of deep lakes.
Girfvik, Norway (June 19-20, 1978). (In press.)
1980b. Interventions and controls introduced
during and subsequent to monitoring program and the
effect achieved in OECD Regional Project Alpine lakes. (In
press.)
Ravera, O., and R. A. Vollenweider. 1968. Oscillatorie
rubscens, D.C. as an indicator of Lago Maggiore eutrophic-
tion. Schweiz. Z. Hydrol. 30:374.
Saraceni, C. D., Ruggeri, and M. Nakanishi. 1978. Phyto-
plankton dynamics chlorphyll-a and phaeophytin in Lago di
Mergozzo (Northern Italy). Mem. 1st. Ital. Idrobiol. 36:215.
Stella, E., L. Ferrero, and F.G. Margaritora. 1978. Alteration
of the plankton in a much polluted lake in central Italy
(Latium), the volcanic Lake Nemi. Verh. Int. Verein. Limnol.,
20:1049.
Stenson, J. A. E. 1972. Fish predation effects on the species
composition of the zooplankton community in eight small
forest lakes. Rep. Inst. Freshw. Res. Drottingholm 52:132.
1976. Significance of predator influence on
composition of Bosmina spp. populations. Limnol. Ocean-
ogr. 21:814.
Tonolli, V. 1962. L'attuale situazione del popolamento
plancetonico del Lago Maggiore. Mem. 1st. Ital. Idrobiol.
15:81.
Zaret, T. M. 1978. A predation model of zooplankton
community structure. Verh. Int. Verein. Limnol. 20:2496.
Zaret, T., and S. Suffern. 1976. Vertical migration in
zooplankton as a predator avoidance mechanism. Limnol.
Oceanogr. 21:804.
Zutshi, D.P. 1976. Phytoplankton productivity, algal dynamics
and trophic status of Lake Lergozzo (Northern Italy). Mem.
1st. Ital. Idrobiol. 33:221.
-------
218
USING TROPHIC STATE INDICES TO EXAMINE
THE DYNAMICS OF EUTROPHICATION
ROBERT E. CARLSON
Department of Biological Sciences
Kent State University
Kent, Ohio
ABSTRACT
To use trophic state indices solely for lake classification overlooks their greater potential as
diagnostic aids in examining relationships between various factors in lakes. Carlson's index (1977)
uses simple models and regression relationships to provide correlated index values for chlorophyll,
transparency, and total phosphorus. The values from these three commonly used variables are
transformed into a common scale. Frequently, deviations of these variables provide more
information about the lake than does their coincidence. The index is used to identify possible
nitrogen-limitation, non-algal turbidity, and the impact of zooplankton grazing.
INTRODUCTION
The concept of trophic state has proved to be a
durable and useful concept in limnology, although it is
plagued with lack of definition and of measurement.
Various indices have been proposed to aid in evaluating
trophic state. These indices, by one method or another,
attach a label or number to the lake, thus classifying it.
Although trophic classification is certainly an important
aspect of limnological investigations, the pinning of a
trophic label on a lake can actually obscure important
differences among lakes within that classification.
A unique classification index devised by Carlson
(1977) derives index values which calculate the index
individually for three variables: Chlorophyll, Secchi disk
transparency, and total phosphorus. Under "normal"
circumstances all three index values should be similar.
Although the values derived can certainly be used to
classify lakes, a far more important use of the index has
been to examine deviations of index values from those
predicted by the other values. These deviations can
reveal basic differences in the ecological functioning of
certain aquatic systems. This paper illustrates this use
of the index.
DATA FOR LAKE SURVEYS
Often agencies are confronted with data reduction
and interpretation of large amounts of data from
surveys of many lakes. Typically, these surveys have
limited numbers of samples, taken at different times of
the year. Interpretation of such data sets is difficult
because of the lack of "normal" references with which
to compare the data. The trophic state index can
provide such a reference line, as it is based on the
assumption that the index values should be the same in
the same samples. Obviously "normality" in this case
depends on the original data set used in formulating
the index's regression equations, but extensive use of
the index on other lakes suggests coincidence of the
three variables in common, and that reasons do exist
for consistent deviations.
To illustrate the utility of the index in checking large
data sets, I used data collected in the National
Eutrophication Survey (U.S.EPA, 1978a, b, c). The sets
consisted of data from 105 natural lakes and 386
reservoirs located in the southern, central, and western
United States. Most of the lakes had been sampled
three times during 1 year and the data I used was the
median or average value obtained from these three
samples. Trophic state index values were calculated
from the data for mean Secchi disk transparency,
median total phosphorus, and mean chlorophyll a. The
calculated index values were compared two at a time to
a 1:1 index correspondence line based on the
assumption of perfect correlation of the indices.
The first comparisons indicated that impoundments
showed different relationships among the indices than
did natural lakes. For this reason, natural lakes were
compared separately from impoundments.
The correspondence between the chlorophyll index
and the transparency index in natural lakes is quite
good (Figure 1), although there are some outlying lakes.
The index comparison quickly identified deviant lakes,
which could then be examined more intensively. When
the total phosphorus and transparency indices are
compared (Figure 2), however, there is a systematic
deviation of a large number of lakes, especially at high
total phosphorus values. The asterisks indicate that all
of these lakes have a total nitrogen to total phosphorus
ratio of less than 15:1 by weight. This same trend is
seen in the total phosphorus-chlorophyll comparison
(Figure 3). Lakes with low nitrogen to phosphorus
ratios seem to show significant deviations of the algal
indicators from the total phosphorus.
The effect of TN/TP ratios is best seen in Figure 4,
where the difference between the chlorophyll and
phosphorus index is plotted against the nitrogen-
phosphorus ratios. Lakes with a TN/TP ratio of less
-------
SPECIAL PROJECTS AND TOPICS FOR ADDRESSING THE TROPHIC STATE
219
than 26:1 by weight have significant deviations in the
algal-nutrient relationship. It appears to be a con-
tinuous function, with the difference increasing
exponentially as the TN/TP ratio decreases. A deviation
of the indices at low TN/TP ratios should be expected,
since correspondence of the indices is predicated on
the algae being phosphorus-limited. If the algae are
nitrogen-limited, no correspondence should be ex-
pected.
CHLOROPHYLL TSI
Figure 1. — The relationship between chlorophyll and
transparency indices in natural lakes. (+ = TN/TP > 15; * =
TN/TP <15).
TOTAL PHOSPHORUS TSI
Figure 2. — The relationship between total phosphorus and
transparency indices in natural lakes. Symbols as in figure 1.
Impoundments also show this deviation based on the
TN/TP ratio, but other overriding factors determine the
TSI relationships in these artificial bodies. The
chlorophyll-transparency relationship (Figure 5) shows
a large number of errant lakes, unlike the relationships
found in natural lakes. There is also a poor
correspondence between the total phosphorus and
chlorophyll indices (Figure 6), again with the most
deviation being from those lakes with low TN/TP ratios.
However, the total phosphorus-transparency relation-
ship is relatively good, with many of the deviants being
lakes with low TN/TP ratios (Figure 7).
The simplest interpretation of the fact that trans-
parency is better related to total phosphorus than
53 75
TOTAL PHOSPHORUS TSI
Figure 3. — The relationship between total phosphorus and
chlorophyll indices in natural lakes. Symbols as in Figure 1.
2 10
-10.
a.
-20 .
-33 .
-40 .
-50 .
-60
Figure 4. — The deviation of the chlorophyll index from the
total phosphorus index as a function of the total nitrogen to
total phosphorus (by weight) ratio.
CHLOROPHYLL TSI
Figure 5. — The relationship between chlorophyll and
transparency indices in impoundments. Symbols as in Figure
1.
-------
220
RESTORATION OF LAKES AND INLAND WATERS
chlorophyll is that the major attenuator of light in many
of the impoundments is a non-algal material, which,
however, does contain phosphorus. This non-algal
material may be eroded soils or clay. Interpretation
requires further information, but again comparing the
indices produced evidence that impoundments vary
considerably from natural lakes in their nutrient-algal
relationships. Perhaps it should be expected that
impoundments would be muddier than their natural
counterparts, but if this is the case, it would be
foolhardy to use predictive models using coefficients
derived from natural lakes to predict water quality in
impoundments.
TOTAL PHOSPHORUS TSI
Figure 6. — The relationship between total phosphorus and
chlorophyll indices in impoundments. Symbols as in Figure 1.
TOTAL PHOSPHORUS TSI
Figure 7. — The relationship between total phosphorus and
transparency indices in impoundments. Symbols as in Figure
1.
SEASONAL CHANGES WITHIN LAKES
When adequate seasonal data exist, the three TSI
variables within a single lake can be compared.
Sometimes, striking deviations in the indices are
uncovered. Figure 8 illustrates this in Halsted Bay, Lake
Minnetonka, Minn. In each of the 3 years I studied this
lake there was a marked decline of algae in late May, at
the time the thermocline became established. This type
of spring decline in algae has been reported by others,
and has been attributed to die-offs of spring species,
sinking of the heavier diatoms (Knoechell and Kalff,
1975), and to zooplankton grazing (Fogg, 1975).
The seasonal plot of the indices indicated a marked
deviation of the chlorophyll and transparency indices
from the phosphorus index in the spring. The increased
transparency is certainly the result of decreases in
algal chlorophyll, but the amount of phosphorus in the
water remains unchanged. Actually, particulate phos-
phorus decreased but ortho-phosphorus increased. If
the algal cells had simply fallen to the bottom, some
decrease in total phosphorus might have been
expected. The cells must have either lysed while in the
epilimnion or have been eaten and the phosphorus
excreted (Peters and Rigler, 1973). The coincidence of
the year's maximum zooplankton abundance at the
time of the decline strongly supports the latter
explanation. Other lakes I have examined have also
shown this chlorophyll-transparency deviation at the
time of high zooplankton densities. Mogadore Reser-
voir, Ohio was studied by myself and G. D. Cooke in
1976. Two major deviations of the chlorophyll-
transparency indices from the total phosphorus index
Figure 8. — Upper Graph: The season fluctuations of total
phosphorus (0--0), chlorophyll (0—0) and trans-
rency (•—•) indices in Halsted Bay, Lake Minnetonka,
Minnesota.
Lower Graph: The dry weight of zooplankton (mg/l) over the
same season.
Svcdl. Oni T *lmu,r •
I «0
Figure 9. — The seasonal fluctuations of total phosphorus
chlorophyll, and transparency indices in Mogadore Reservoir,
Ohio.
-------
SPECIAL PROJECTS AND TOPICS FOR ADDRESSING THE TROPHIC STATE
221
are seen: one in late April, the other in September
(Figure 9). Both periods are characterized by the
highest densities of the herbivore Daphnia galeata.
Other aspects of Mogadore's index fluctuations are
also enlightening. If an index value of 50 is taken to
represent the lower limits of eutrophy (Carlson, 1979)
then the lake did not exhibit eutrophic levels of algae
until mid-July. This lake would have been classified as
oligotrophic based on the chlorophyll levels found in
May. In Mogadore, these changes are internally driven
(Carlson and Cooke, unpubl.), with the rapid rise in
phosphorus in late August associated with a major die-
off of macrophytes. The wide seasonal variation in
trophic state in the reservoir makes using a single
trophic designation for a lake questionable. This lake
exhibited eutrophic algal characteristics for only 2 to 3
months of its open-water season. The purpose of lake
management is better served by a trophic concept and
corresponding index that assumes that trophic state is
not a constant for a lake but a seasonally changing,
dynamic assessment of the lake's condition. Proper
management requires measuring the duration of a
problem, as well as its extent and cause. Illustrating the
seasonal changes in trophic state could have an impact
on the assessment of the lake's condition and on the
methods used in its management.
Knoechell, R., and J. Kalff 1975. Algal sedimentation: the
cause of a diatom-blue-green succession. Verh. Int. Verein.
Limnol. 19:745.
Peters, R. H., and F. H. Rigler. 1973. Phosphorus release by
Daphnia. Limnol. Oceanogr. 18:821.
U.S. Environmental Protection Agency. 1978. A compendium
of lake and reservoir data collected by the National
Eutrophication Survey in eastern, north-central, and
southwestern United States. Work. Pap.475. Corvallis, Ore.
1978. A compendium of lake and reservoir data
collected by the National Eutrophication Survey in the
central United States. Work. Pap. 476. Corvallis, Ore.
1978. A compendium of lake and reservoir data
collected by the National Eutrophication Survey in the
central United States. Work. Pap. 477. Corvallis, Ore.
CONCLUSIONS
In this paper a trophic state index has been used to
identify situations of nitrogen-limitation, non-algal
turbidity, and zooplankton-induced algal declines. The
index can do this because it provides a set of expected
relationships against which data from other lakes can
be compared. This method surpasses the simple
comparison of raw data because often smaller or
regional data sets have internal correlations which may
imply relationships that cease to exist when compared
against a more global data set. Certainly the idea of
what is a normal lake is dictated by the original data set
used in the index, but data from some of the world's
clearest and worst lakes were included in the original
data.
Certainly these relationships could be examined
using the original regression relationships rather than
the transformed index values. The importance of the
transformation is that the comparisons are made in the
context of the trophic state concept. The major
importance of this concept is that it implies that many
aspects of a lake will change as a lake assumes
eutrophic characteristics. Even when the index does
not measure hypolimnetic oxygen depletion or changes
in fish species, the interconnectedness of the lake's
biological components is implied. Thus the TSI values
take on meaning far greater than do the raw data.
REFERENCES
Carlson, R. E. 1977. A trophic state index for lakes. Limnol.
Oceanogr. 22:361.
1979. A review of the philosophy and construction
of trophic state indices. Pages 1 -52 in T. Maloney, ed. Lake
and reservoir classification systems. Ecol. Res. Ser. EPA-
600/3-79-074. U.S. Environ. Prot. Agency.
Fogg, G. E. 1975. Algal cultures and phytoplankton ecology.
The University of Wisconsin Press.
-------
222
REGRESSION ANALYSIS OF RESERVOIR WATER
QUALITY PARAMETERS WITH DIGITAL SATELLITE
REFLECTANCE DATA
HERBERT J. GRIMSHAW
SUSAN MEYER TORRANS
Oklahoma Water Resources Board
Oklahoma City, Oklahoma
THOMAS LERA
U.S. Environmental Protection Agency
ABSTRACT
Nine Oklahoma reservoirs were sampled monthly from June through October 1979. Water
samples were collected concurrent with satellite "passes" and were analyzed for nitrate, nitrite,
ammonia and kjeldahl nitrogen, dissolved and total orthophosphate, color, filtrable residue,
chlorophylls a, b and c, pheophytm, nephelometric turbidity, and total alkalinity Additionally, in
situ data were also gathered concerning temperature, pH, dissolved oxygen, conductivity, wind
speed, and Secchi disk extinction depth. Water quality and satellite monitored reflectance data
were analyzed utilizing multiple regression techniques. Equations were generated which permit
the prediction of chlorophyll a, pH, turbidity, color, total alkalinity, and total orthophosphate
concentration in Oklahoma reservoirs Generalization of these relationships to areas outside
Oklahoma requires further testing.
INTRODUCTION
Since the launch of the first LANDSAT in July 1972,
several studies have evaluated applying satellite based
multispectral scanner data to lake or water quality
monitoring programs (Bukata, et al. 1 974; Rogers, et al.
1975; Bohland, 1976; McKeon, et al. 1977; Scarpace,
et al. 1978; Bohland, et al. 1979; Grimshaw and
Torrans, 1980).Two of these studies have incorporated
concurrently obtained water quality and satellite
monitored reflectance data collected from a number of
different water bodies on several dates (Scarpace, et al.
1978; Grimshaw and Torrans, 1980). Scarpace, et al.
developed regression relationships between trophic
class and reflectance values averaged over an entire
lake. This study develops regression relationships
between LANDSAT monitored reflectance values and
specific water quality parameters based upon samples
obtained from discrete sampling stations. These
relationships will permit the estimation of the
concentration of several water quality parameters in
reservoirs which were not surface sampled.
SITE SPECIFICITY
To evaluate the use of satellite based multispectral
scanner data in water quality monitoring programs it
was necessary to compare discrete, site specific water
quality data to concurrently obtained site specific
reflectance values. Mean reflectance values, obtained
by averaging reflectance over the entire water body,
were considered inappropriate for use. Consequently,
triangulation procedures were used to determine the
latitude and longitude of each sampling station. This
procedure permitted an accurate comparison of water
quality data and satellite monitored reflectance values.
RESULTS
As would be expected, generally more favorable
relationships were obtained when the data set was
restricted to concurrent satellite and water quality data.
Inspection of Table 1 illustrates this point by
demonstrating the improvement which can be achiev-
ed in the correlation between Band 4 (500 to 600
nm)/Band 5 (600 to 700 nm) ratio and log transformed
chlorophyll a concentration when only concurrently
obtained data are used in the statistical analysis. The
extent of this improvement would probably have been
even more pronounced had our total data set not been
collected very near to the actual satellite coverage
dates. Water quality data were collected 1 day before
satellite coverage on five occasions; 11 data elements
were collected ±11 days or less, and only two entries
were off by 18 days. The remaining 22 data elements
were collected concurrent with satellite coverage.
Table 1 also illustrates the improvement in the Band
4/Band 5 ratio, log chlorophyll a correlation coefficient
which can be obtained by subjecting the site specific
water quality data to a turbidity based cluster analysis
prior to correlation with satellite reflectance data.
Figure 2 illustrates this relationship between
chlorophyll a concentration and LANDSAT multispec-
tral scanner Band 4/Band 5 ratio data. Triangularly
shaped symbols represent turbid water samples,
defined here to refer to water samples with nephelo-
-------
SPECIAL PROJECTS AND TOPICS FOR ADDRESSING THE TROPHIC STATE
223
Figure 1. — Reservoirs included in this study.
Table 1. — Correlation coefficients between Band 4/Band 5
ratio and log chlorophyll a.
Criterion
Total data set
Concurrent data
Low turbidity
concurrent data
Correlation
coefficient
0.50142
0.57421
0.65726
Significance
level
.01
.01
.01
n
40
22
16
metric turbidity values of 20 N.T.U.'s or greater, while
the circular symbols represent clear water samples.
Solid symbols indicate that reflectance values were
obtained from LANDS AT 3 while open symbols indicate
that they were obtained from LANDS AT 2.
A semilog plot of the turbid water Band 4/Band 5
reflectance ratio against chlorophyll a concentration
approximates a horizontal line, exhibits no statistically
significant correlation, and consequently demonstrates
the lack of any relationship between these variables. A
similar plot (see Figure 2) of the clear water samples,
however, indicates that a clear and reasonably strong
relationship does exist between these variables, as
evidenced by their highly significant correlation (see
Table 1).
01 02 03 01 05 06 07 OB 09 10
Figure 2. — Semilog plot of chlorophyll a concentration and
Band 4/Band 5 reflectance ratio.
Further inspection reveals that there is considerable
scatter within the data. For example, a reflectance
value of 1.05 relates to a chlorophyll a concentration of
16.8 mgrrf3, while a reflectance value of 1.08 relates to
a chlorophyll a concentration of 4.4 mgm"3 . This
relationship is exactly opposite to what one would
expect. Because of chlorophyll's absorption and
reflective characteristics, a higher Band 4/Band 5 ratio
should be related to a higher, rather than a lower,
chlorophyll concentration.
These observations suggested that information from
more than one band width or band ratio is required to
predict chlorophyll a concentration with any reason-
able degree of accuracy, in spite of the highly
significant correlation which had previously been
demonstrated between the Band 4/Band 5 ratio and
Table 2. — Regression equations for concentration estimation of Oklahoma reservoir water quality parameters.
Regression equation
Log chlorophyll a = 1.094 + 0.092 (Band 4) - 0.107 (Band 5)
pH = 9.526 - 0.049 (Band 4) - 0.040 (Band 5) + 0.147 (Band 7)
Turbidity = 15.725 - 4.365 (Band 4) + 4.911 (Band 5) - 0.443
(Band 7)
Color = 32.826 - 4.570 (Band 4) + 4.356 (Band 5)
Log total alkalinity = 3.877 + 0.033 (Band 4) - 0.031 (Band 5)
-0.053 (Band 6) - 1.181 (Band4/Band5)
Secchi = 0.811 + 0.048 (Band 4) - 0.053 (Band 5)
Log total ortho-phosphate = -1.022 - 0.064 (Band 4) + 0.058
(Band 5)
Coefficient
determination
(%)
88.8
89.2
88.7
80.9
72.3
53.6
41.8
Significance
level
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
0.0001
Standard error
of estimate
0.42
0.26
7.92
7.95
0.37
0.21
0.25
-------
224
RESTORATION OF LAKES AND INLAND WATERS
chlorophyll a. Consequently, multiple regression anal-
ysis was undertaken, utilizing log transformed chloro-
phyll a data as the dependent variable and Bands 4, 5, 6
(700 to 800 nm), 7(800 to 1,100 nm)and band ratios of
4/5, 4/6 and 5/6 as the independent variable's (see
Table 2). The significance of each independent
variables contribution to the variance explained was
evaluated using a t-test. Variables which did not
contribute in a highly significant (.01 level) manner
were not used in the equations which were ultimately
developed. Results of this analysis are presented in
Table 2. Inspection reveals that the multiple regression
procedure has successfully explained approximately 89
percent of the variance in the chlorophyll a data set.
Regression equations were also subsequently devel-
oped to predict pH, turbidity, color, total alkalinity,
Secchi disk extinction depth, and total orthophosphate
concentration. These equations were developed using
the entire data set, while the chlorophyll a equation
was developed using exclusively concurrent, clear
water data. In spite of this fact, several dependent
variables, notably pH, turbidity, and color, exhibited
very high coefficients of determination.
DISCUSSION
These results suggest the possibility of monitoring
several water quality parameters using LANDSAT's
multispectral scanner. A more extensive evaluation of
the utility of these regression equations, for purposes
of trophic classification, will be published in the near
future. Initial efforts in this regard, however, are
presented in Table 3, where trophic state indices (T.S.I.)
(Carlson, 1977) calculated from both observed and
predicted chlorophyll a concentrations, are tabulated.
Graphic analysis of these data (see Figure 3) indicate
that the chlorophyll a regression equation predicts
T.S.I, values with an accuracy of ±7 T.S.I, units.
Predictions were most accurate when the chlorophyll a
concentrations ranged from 13 to about 21 mgrrf3
Index estimates obtained when chlorophyll a concen-
trations were from 4 to 11 mgrrf3 overestimated the
T.S.I, by approximately 6 units. Progressively increas-
ing underestimates were obtained when the chloro-
phyll a concentration was in the 32 to 49 mgrrf 3ranae.
Table 3. — Observed and predicted chlorophyll a
concentrations and trophic state indices.
OBSERVED T S I
OBSERVED
Chlorophyll a
mgm
48.8
10.6
14.0
15.8
20.8
7.8
4.3
10.2
5.8
10.2
15.5
18.8
31.8
14.4
13.7
16.8
T.S.I.
68.7
53.8
56.5
57.7
60.4
50.8
44.9
53.4
47.8
53.4
57.5
59.4
64.5
56.8
56.3
58.3
PREDICTED
Chlorophyll a
mgm
22.1
20.6
10.6
18.2
18.1
14.3
8.0
18.5
10.6
18.8
18.1
17.5
21.8
15.2
13.3
7.6
T.S.I.
61.0
60.3
53.8
59.1
59.0
56.7
51.0
59.2
53.8
59.4
59.0
58.7
60.8
57.3
56.0
50.5
Figure 3. — Comparison of observed and predicted trophic
state indices.
CONCLUSIONS
1. Highly significant multiple regression relation-
ships have been demonstrated to exist between several
water quality parameters and LANDSAT multispectral
scanner data.
2. These equations appear to permit the prediction of
chlorophyll based trophic state indices with an
accuracy of ±7 T.S.I, units.
REFERENCES
Barb, C. E., and J. A. Harrington. 1980. Low cost satellite
digital image analysis. Symp. on Surface-Water Impound-
ments, June 2-5, Minneapolis, Minn.
Boland, D. H. 1976. Trophic classification of lakes using
LANDSAT-1 (ERTS-1) multispectral scanner data. Ecol. Res.
Ser. EPA-600/3-79-123. U.S. Environ. Prot. Agency.
Boland, D. H., et al. 1979. Trophic classification of selected
Illinois water bodies: Lake classification through
amalgamation of LANDSAT multispectral scanner and
contact-sensed data. EPA-600/3-76-037. U.S. Environ.
Prot. Agency.
Bukata, R. P., G. P. Harris, and J. E. Bruton. 1974. the
detection of suspended solids and chlorophyll a utilizing
digital multispectral ERTS-1 data. Can. Symp. on Remote
Sensing. Guelph, Ontario. 2:552.
Carlson, R. E. 1977. A trophic state index for lakes. Limnol.
Oceanogr. 22:361.
Grimshaw, H. J., and S. M. Torrans. 1980. Correlation
analysis of reservoir water quality parameters with digital
satellite reflectance data. Symp. on Surface-Water Im-
poundments. June 2-5. Minneapolis, Minn.
McKeon, J. B., R. H. Rogers, and V. E. Smith. 1977. Production
of a water quality map of Saginaw Bay by computer
processing of LANDSAT-2 data. 11th Int. Symp on Remote
Sensing of Environment. Ann Arbor, Mich. April 25-29.
(Bendix Aerospace Sys. Div. No. BSR4277)
-------
SPECIAL PROJECTS AND TOPICS FOR ADDRESSING THE TROPHIC STATE 225
Rogers, R. H., et al. 1975. Application of LANDSAT to the
surveillance and control of eutrophication in Saginaw Bay.
Proc. 10th Int. Symp. on Remote Sensing of Environment.
Scarpace, f. L, K. Holmquist, and L T. Fisher. 1978.
LANDSAT analysis of lake quality for a statewide lake
classification program. Proc. Am. Soc. Photogrammetry.
44th Annu. Meet. February 26-March 4. Washington, D.C.
-------
226
LAKE ASSESSMENT IN PREPARATION FOR A
MULTIPHASE RESTORATION TREATMENT
WILLIAM H. FUNK
HARRY L. GIBBONS
GARY C. BAILEY
Department of Civil and Environmental Engineering
Washington State University
Pullman, Washington
ABSTRACT
Liberty Lake is a 288 hectare body of water located in eastern Washington. It is a heavily utilized
recreational lake when its waters and swimming beaches are not plagued by massive blooms of
blue-green algae. In 1974 alum treatment of the lake aimed at late summer and fall release of
phosphorus successfully demonstrated the need to control internal cycling of nutrients (especially
phosphorus) as well as surface and subsurface input. Macrophytes growing in rich sediments
acted as nutrient pumps releasing phosphorus above the floe layer. This event as well as flushing
of the bird refuge and marshland to the south of the lake and continued input of septic tanks
overcame the alum treatment within 3 years. The 3-year respite was the first in 10 years from
blue-green algae problems. Restorative efforts began in 1978-79 with sewering of the lake
periphery. Marsh runoff diversion was completed in 1979-80. Suction dredging followed by alum
treatment is scheduled for fall 1980. Extensive monitoring of water quality parameters began in
late 1977 and has continued to assess each phase of the restoration. The initial results give reason
for cautious optimism.
INTRODUCTION
While many of the intricacies of accelerated lake
eutrophic processes remain poorly understood, the role
and implication of excessive nutrients, especially
phosphorus, has been well elucidated (Sawyer, 1947,
1952; Ohle, 1953; Thomas, 1969; Vallentyne, 1974;
Edmondson, 1972; Wetzel, 1975).
To deaccelerate, reverse, or at least stabilize the
deterioration of a lake's water by overenrichment, the
sources of the nutrients must be defined. In addition,
the contribution of each source must be determined as
accurately as possible and a mechanism set in place to
divert, reduce, or mitigate' that source of nutrient
inflow. Thirdly, unless a concomitant educational
program is established, the mitigating efforts may
come to naught because a social, economic, or political
decision may countermand restorative efforts. Such an
action may introduce a new or overload a formerly
insignificant nutrient source.
Finally, restorative efforts must be evaluated to
predict the future of the lake in question and add to the
scientific body of knowledge for lakes of similar
background. This paper deals with assessment of
multiphase restorative efforts at Liberty Lake, Wash.
STUDY AREA
Liberty Lake (Figure 1) is a softwater lake (288 ha) of
glacial origin enclosed on three sides by a small
mountain range 300 to 500 meters above the lake
surface. Most of the watershed (3,445 ha) lies in this
horseshoe-shaped basin, forested by Ponderosa pine,
SPOKANE RIVER
DRAINAGE BASIN
OUTLINE
ROUND MTN.
SCALE h 62,500
Figure 1. — Liberty Lake and drainage basin.
-------
SPECIAL PROJECTS AND TOPICS FOR ADDRESSING THE TROPHIC STATE
227
grand fir, Douglas fir, larch, white pine, and aspen. The
major tributary, Liberty Creek, originates in the higher
southeastern slopes and passes through Moscow and
Springdale soils before reaching the Spokane and
Semihoo muck series adjacent to and in a marsh. The
stream flows along the eastern margin of the marsh
(and until recently, overflowed into it) before entering
the lake. Most of the tributary area is underlain by
quartz - feldspar - biotite paragneiss. Residential areas
occupy 87 percent of the shoreline and overlie
relatively shallow soils (Spokane series); gneiss (west
side and north shore) and Columbia River basalt (west
shore) form the bedrock. A small creek enters the lake
from the northwest side. Until 1978-79, waste disposal
had been by septic system and an old sewer system
built in 1910 which served approximately 40 percent of
the residents. In late 1978 and 1979, a collection
system was built. It serves about 2,000 permanent
residents and diverts 95 percent of the domestic
sewage from the lake basin.
The mean residence time of lake waters is 3 years.
Approximately 2.76 x 106 mVyr is lost by seepage,
presumably through the bottom at the northern end of
the lake. The lake may become weakly stratified for
short periods of time during the mid and late summer
period.
The lake is heavily used, 80,000 to 100,000 visits per
season when the swim areas and beaches are not
plagued by massive blue-green algal blooms (Gloeo-
trichia, Anabaena, and Aphanizomenon).
A 3-year respite from algal nuisances was made
possible by an alum treatment in 1974. Details are
described in Funk, et al. (1975) and Gibbons (1979).
ASSESSMENT METHODS
Earlier estimates of stream inflow, lake level, and
outflow had been made from Gurley current meters
and staff gages in the lake. Precipitation was measured
by standard rain and snow collector devices. From
these data, Orsborn (1973) developed a water balance.
Those estimates were later refined by Copp, et al.
(1976) and a nutrient budget developed for the lake by
volume weighting flows with phosphorus and nitrogen
data collected biweekly during the summer period and
monthly during the winter period. During the low flow
period of 1977, Parshall flumes were installed on the
main stems of Liberty. The flumes were equipped with
Manning F 3000 series flowmeters and model S-4040
discrete samplers for continuous flow measurements
and water sample collection. Provision was also made
with Spokane County Parks personnel for daily
inspection and reading of gages in event of equipment
failure. Parks personnel were also contracted to read
rain gages and evaporation pans. Gurley meters were
used to estimate the flow of several small intermittent
streams. Storm events and runoff were similarly
measured.
Ground Water Inflow
Attempts to measure ground water were made by
circumscribing the lake with 17 banks of piezometers
as described by McComas (1977).
Sample frequency has been biweekly in the summer
and fall and monthly in the winter. Vandalism problems
arose when sites were obvious.
Seepage meters (Lee, 1977) were installed in the
near shore areas to complement the piezometer banks.
Similar vandalism occurred more frequently. Fifteen
barrels were set out and only two remained un-
disturbed; no data were obtained with this method.
Macrophyte Evaluation
Macrophyte growth was estimated by scuba methods
along six transects demarked by 100 to 500 m nylon
lines laid out on the lake bottom. Plants were collected
at 1 m increments of depth using a round metal
sampler (.2m2 area) to delineate the sample area. Three
samples were collected at each depth to give a total
area sampled of .15m2 The macrophytes, including
roots, were placed in plastic bags, tied, inflated by
exhaust air from the diver's tank, and allowed to float to
the surface. The boat crew collected and labeled each
bag.
In the laboratory, the samples were rinsed to remove
sediments, drained, and weighed for wet weight.
Subsamples were dried for 24 hours at 100°C. The
dried sample was then weighed and ashed at 550°C to
obtain ash free dry weight. Additional subsamples
were digested to determine nitrogen and phosphorus
content.
The data were plotted on a contour map and
planimetered to estimate the area of macrophyte
growth, weight, and nutrient content.
Macrophyte measurements began each spring at ice
off (March or April) and continued until late September
or October when at least half of the macrophytes
senesce and deteriorate. First estimates were made in
fall 1974. Measurements for restoration evaluation
began in March 1978 and have continued to date.
Sediment Assessment and Nutrient
Release Studies
To assess the contribution of nutrient release from
sediments, 10 cores were driven in 1974 by Ewing
piston corer. An additional 28 were taken in 1978 by a
modified hand-driven piston corer. This latter device
used a 12 x 155 cm clear PVC tube as both coring tube
and liner. The previously used Ewing corer appeared to
force the flocculent sediments away and compacted
the upper layers, probably resulting in lower phos-
phorus values when analyzed. Fourteen cores taken in
1978 were analyzed for nutrient content (phosphorus
and nitrogen) and selected metals following methods
outlined in the U.S. EPA Laboratory Manual for Bottom
Sediments (1969) and Am. Pub. Health Assoc. (1975).
Four were subjected to phosphorus release tests.
These core samples were augmented by 70 Ekman
grab samples taken randomly across the lake to
observe the sediment appearance and texture, and to
construct a bottom-sediment map.
Additional smaller (12 x 33 Cm) cores were taken by
scuba methods using a small stainless steel piston
corer. The small corer could deposit a relatively
undisturbed intact core into a laboratory test column.
-------
228
RESTORATION OF LAKES AND INLAND WATERS
Both long- and short-term anaerobic, facultative, and
aerobic conditions were run on these latter cores to
determine release rates of total phosphorus (TP), total
soluble phosphorus (TSP), and soluble reactive phos-
phorus (SRP).
Benthic Invertebrates
Extensive collection of benthic invertebrates began
in March 1978. Prior to that time, collections were
made on the basis of monitoring the intial alum
treatment of the lake. Sampling until August 1978 was
carried out at 17 locations and consisted of random
Ekman grab samples in each of the areas delineated by
the bottom sediment map. For the past 2 years, a
modified stratified random sampling method has been
adopted which takes into account the three general
substrate types (Nelson, 1980). The first can be
described as a silt composed of finely divided organic
matter, relatively decomposed, and is characteristic of
the middle portion and southern end of the lake under 6
to 9 m of water. The second is adjacent to and partially
in the littoral zone. This area experiences heavy
macrophyte growth during the spring and summer with
Elodea, Potamogeton, and Ceratophyllum being the
dominant plants. The sediment is muck-like in
appearance and contains partially decomposed plants
with some parts readily distinguishable. The third
major sediment type is composed of a wide variety of
inorganic particle sizes ranging from sand to gravel and
rock. Overlying waters are 0 to 6 m in depth.
A random site is selected by superimposing a grid
system on a map of the lake. Each grid represents a 60
x 60 m square. The grids within each substrate are
numbered and a site selected by random number
tables. The number of samples taken from each site
depends upon the variation encountered during the
previous sample period. The usual number of samples
taken is 12. Sample frequency is once every 3 weeks
during spring, summer, and fall and once per month in
the winter.
Ekman grab samples are used in the open waters and
a box type sampler with a bladed screen as described
by Minto (1977) is used in areas of heavy macrophyte
growth. Material from the samples is initially screened
in the field, using three screens in tandem with a high
wall on the first. The screen set is composed of
numbers 6, 20, and 28 U.S. standard sieve sizes. The
screened materials are placed in plastic bottles
preserved with 40 percent formalin solution mixed with
rose bengal dye (Mason and Yevich, 1967) to facilitate
laboratory sorting. Additional screening occurs in the
laboratory with U.S. standard sieve sizes 6, 10, 16, 25,
30, and 40. After microscope scanning, the material is
placed in a subsampling tray divided into 70, 3 cm2
units and mixed for random distribution. Subsamples
for extensive identification are collected from five of the
3 cm2 units as determined by a random number table.
The data obtained from each subsample are checked
for randomness by use of the chi squared test (Elliot,
1971).
Lake Water Quality and Productivity
Measurements
Standard field physicochemical measurements were
made biweekly during the growing season and monthly
in the winter at 1.0 m intervals top to bottom at two
lake stations. Measurements included temperature,
light transmission, dissolved oxygen, alkalinity, pH,
conductivity, and Secchi disk. Both wet chemistry and
calibrated probes were used (light - Kahlsico Gemware;
conductivity, dissolved oxygen, and pH Hydrolab II).
Phyto and zooplankton samples as well as samples
for laboratory chemistries and chlorophyll a were
collected by rapid pump methods at the same time,
location, and interval as the field measurements. Water
for in situ carbon-14 productivity measurements were
collected in the same manner. Pump manifold and
intake funnel were clear PVC; pump line was clear
Tygon 1.3 gm (id). Pumping rate was approximately 14 I
per minute depending upon depth and was calculated
for each sampling date. Zooplankton samples were
collected by pumping water through a 60 prr\ mesh
nylon plankton net and cup. Killing and preservation
was by formalin, ethanol, and glycerin (Schwoerbel,
1970). Identification and counts were made according
to methods outlined in Edmondson (1959), Edmondson
and Winberg (1971), Brooks (1957), and Pennak
(1978). Successive 1 and 5 I subsamples were
examined in the laboratory for abundant and scarce
individuals until 50 of the most common individuals
were obtained or 20 ml of sample had been analyzed.
Generally, phytoplankton counts were made upon
subsamples from 1.0 /^unpreserved samples within 24
hours of collection. Preliminary statistical analysis
involving the chi squared test (variance to mean ratio)
was employed to check for subsample homogeneity.
The remaining portions of the phytoplankton sample
were preserved by modified Lugols solution (Schwoer-
bel, 1970) for additional identification and measure-
ment. If concentration was necessary, centrifugation
was employed. Strip count methods were as outlined
by Edmondson (1974). Volume measurements of
phytoplankton were made as described by Vollen-
weider, et al. (1974) and Wetzel (1975). Zooplankton
biomass (//g/m3) was determined using the values of
Hall, et al. (1970), Peterka and Knutson (1970), and
Bottrell, et al. (1976).
Chlorophyll a samples were fixed in the field with
MgCCb Upon return to the laboratory they were
immediately filtered through a .45 m Millipore filter
and frozen. Chlorophyll a was extracted in 90 percent
aqueous acetone solution by Bonification procedures
and measured by reading absorbances on a Beckman
model DU 2 spectrophotometer before and after
acidification with 1N HCI, Chlorophyll a and pheophy-
tin concentration were calculated using formulas
contained in Vollenweider (1974) and Am. Pub. Health
Assoc. (1975). Calibration of the spectrophotometer
was checked periodically by using purified chlorophyll
extract (Sigma Chemical Co.). Carboh-14 procedures
were carried out in situ at both lake stations at 2.0 m
intervals. Fifty ml aliquots were filtered through .45 £im
Millipore filters and counting was done with a Nuclear
Chicago — Mark II Scintillation System.
-------
SPECIAL PROJECTS AND TOPICS FOR ADDRESSING THE TROPHIC STATE
229
RESULTS AND DISCUSSION
Water and Nutrient Budget
Restoration of Liberty Lake has been a multiphase
program. Preliminary efforts at reducing nutrient flow
to the lake began with a temporary diversion structure
built in 1978 in cooperation with Spokane County
Parks personnel at the bifurcation of Liberty Creek. The
diversion allowed a large portion of spring runoff to
pass down the West Fork of Liberty Creek without
flushing through the marsh to the lake (Figure 2). After
the first year of operation it was noted that with
capacity stream flow in the East Fork branch nearly 30
percent of the water flow was still being lost to the
marsh through breaks in the streambanks and overflow
caused by stream bed obstructions. Even with some
flushing action occurring, nutrient loading by stream
flow was reduced from approximately 148 kg P and
1,137 kg N before diversion to132.5 P and 831.6 kg N.
A permanent diversion structure was completed in
1979 by M. Kennedy Engineers; the stream channel
was also repaired and cleaned. Phosphorus loading
was further reduced in 1979 to 117.5 kg but nitrogen
was higher (1,191 kg), possibly due to greater
interaction between flowing waters and the newly
cleaned stream bed. Flow data for 1980 should reflect
still lower phosphorus and nitrogen values because of
less overflow and flushing. Stream hydrographs and
nutrient loading for the East and West Fork inflows are
shown in Figures 3 and 4.
ISO
I20
in
2 90
x
in «0
2
so
EAST FORK.
LIBERTY CREEK
JFMAMJJASONOJFMAMJJASOND
JFMAMJJASONDJFMAMJJASOND
Figure 3. — Hydrograph and nutrient inflow East Fork, Liberty
Creek.
WEST FORK. LIBERTY CREEK
Figure 2. — East and West inlets to Liberty Lake and flooded
area of marsh.
JFM4MJJASONOJFMAMJJ4SOND
-I49 ,
Figure 4. — Hydrograph and nutrient inflow West Fork, Liberty
Creek.
-------
230
RESTORATION OF LAKES AND INLAND WATERS
Ground Water Measurements
Preliminary estimates of groundwater input made
prior to completion of the sewer collection system
indicated that the immediate area (784 ha) around the
lake contributed approximately 6.3 * 105m3 of water
which carried approximately 1 50 kg P and 1,717 kg N
toward the lake (Rector, 1979). The remaining portion
of the watershed is relatively undisturbed (2,660 ha)
and its nutrient addition is believed to be minor in
comparison, contributing only about 12 kg as P and 57
kg as N to the lake. Table 1 shows mean values of
piezometer samples for 1978-79.
It is interesting to note that at Dreamwood Bay (site
8),^ number of homes have been built on fill across the
natural drainage. After the collection system was
completed in 1979, the piezometer samples show
sharply reduced nutrient values. It is expected that
most other areas will not show such immediate
improvement because of prolonged drainage bed
leaching and perched water table contribution. These
later phenomena were graphically demonstrated when
trenches were opened for the collection system in
1979. Percolation had occurred in the sandy loam soils
until a plugging of soil pores occurred or a clay lens
prevented downward movement. The effluent then
ponded and moved to the surface or laterally until it
reached the soil surface on the downgrade.
Considerable groundwater outflow from the lake
occurs in the northern portion of the Lake; Orsborn
(1973) estimated the flow to be 2.76 x 106 mVyr. He
used existing well logs to show a sand and gravel
aquifer 6 m in thickness and approximately 1.5 km
wide. One well (25/145-16 Cl) is centrally located in
the aquifer and its elevation closely follows that of the
lake. We estimate a loss of 40 to 50 kg of P/yr from the
lake bottom to the sand gravel aquifer based upon
analysis of the well water and flow.
Table 1. — Mean groundwater concentration of phosphorus
and nitrogen at Liberty Lake.
Drainage Basin
Northwest Side
North End
East Side
MacKenzie Bay
County Park
Dreamwood Bay
Main Watershed
Piezometer Total Total
Sample Site Phosphorus Nitrogen
5,12, 13, 14
22, 23,24,25 +
Northwest side
values
3
6, 7
9, 10
8
1, 2, 28, 29, 30
.26
.153
.26
.065
.036
.25
.039
2.54
1.41
1.78
.54
.44
5.72
.008
Sediment Contribution
Earlier investigations (Funk, et al. 1976, 1979) had
shown that much of the deposited material in Liberty
Lake such as the gravel, clay, and sand release
relatively few nutrients. Recent intensive laboratory
work (Mawson, 1980) has verified that release of
nutrients from sediments located in the southern
portion of the lake could in conjunction with
macrophyte decline accountfor huge algal populations.
Two sediment types were tested, an organic refractory
silt (ROS) representing about 70 ha and a heavy
organic muck (HOM) making up 68 ha. In one test
series oxygen was added to waters overlying the
sediment column (aerobic). In another series of tests
(faculative) the columns were open at the top and
atmospheric oxygen was allowed to equilibrate with
column water. Finally, in an anaerobic series oxygen
was removed by adding sulfide. Conditions during tests
are shown in Table 2. Summarized results for the ROS
sediments are shown in Table 3 and results for the
HOM sediment type are shown in Table 4.
Table 2. — Average dissolved oxygen concentration, average
pH and standard deviations for HOM and ROS (Mawson,
1980).
DO Standard Standard
(mg/l) Deviation pH Deviation
Anaerobic
Facultative
Aerobic
Anaerobic
Facultative
Aerobic
0.0
2.71
8.36
0.0
3.1
6.84
HOM
—
0.41
0.95
ROS
—
0.53
0.96
6.65
6.63
6.97
6.55
6.70
7.02
0.31
0.27
0.46
0.16
0.15
0.38
Table 3. — Summary of number of observations (n), slope of
concentration over time (k), correlation coefficients (r), release rates
(k'), and confidence levels for average observed concentrations for
ROS sediment (Mawson, 1980)
Anaerobic
Facultative
Aerobic
Anaerobic
Facultative
Aerobic
Anaerobic
Facultative
Aerobic
N
28
30
16
34
26
17
34
26
17
K(mg/l-day)
0.007
0.001
0.000
0.001
0.000
0.000
0.001
0.000
5.283x10~B
r
0.909
0.330
0.092
0.772
0.455
0.319
0.807
0277
0.000
k'Cug/m'-hr)
12.7
1.22
0.186
7.10
1.40
03057
2.75
0.393
9.79X10"5
Confidence
levels
99%
<99%
<95%
99%
95%
<95%
99%
<95%
<95%
Analysis
T-P
T-P
T-P
TSP
TSP
TSP
SRP
SRP
SRP
Table 4. — Summary of number of observations (n), slope of
concentration over time (k), correlation coefficients (r), release rates
(k*), and confidence levels for average observed concentrations for
HOM sediment (Mawson, 1980).
Anaerobic
Facultative
Aerobic
Anaerobic
Facultative
Aerobic
Anaerobic
Facultative
Aerobic
N
34
82
26
34
82
26
34
82
34
K(mg/l-day)
0.003
0.001
0.000
0.002
0.001
6.9x10~s
0.002
0.001
1.23X10"5
r
0.693
0.839
0.411
0.707
0.779
0.026
0.832
0.750
0.290
k*Gug/m2-hr)
5.42
2.48
0.399
2.693
1.41
0.011
3.071
1.10
1.87X10"2
Confidence
levels
99%
99%
95%
99%
99%
<95%
99%
99%
<95%
Analysis
TP
TP
TP
TSP
TSP
TSP
SRP
SRP
SRP
As expected, maximum release rates occurred under
anaerobic conditions for the phosphorus component
measurements. Release rates from ROS sediments
were also much greater, almost by a factor of 2, than
the HOM sediments (with the exception of SRP).
-------
SPECIAL PROJECTS AND TOPICS FOR ADDRESSING THE TROPHIC STATE
231
Mawson (1980) believes that this may be due to more
interstitial phosphorus present in the HOM as well as
greater biological activity. The release of phosphorus
from the ROS appears to follow a diffusion pattern
while that of the HOM is much more complex. Bottom
sediment types are shown in Figure 5.
WEST FORK INLET
SAND
ROCK. GRAVEU OR SHINGLE
RELATIVELY LIGHT ORGANIC DEPOSITS WITH CLAY AW PACKED SAND
MODERATE ORGANIC DEPOSITS
MODERATELY HEAVY ORGANIC DEPOSITS,UNCONSOLIDATED
HEAVY MJCK AND ORGANIC DEBRIS, HIGHLY UNCONSOLI OAT ED
CONSCL1OATED FIBROUS PEAT
30 H
5.7 M
112 H
70 H
31 H
36 M
S H
CZI
Figure 5. — Bottom sediment characteristics.
Macrophyte
Macrophytes were collected along a series of
transects as shown in Figure 6. First measurements
were made in 1974 and at that time approximately
58,000 kg (dry weight) of plants covered 80 ha of
bottom area. Approximately half of the aquatic plants
die in the fall. In 1977 at the time of fall dieback a
moderately heavy algal bloom of Anabaena flos-aquae
immediately developed. This increased shading of
viable plants and stimulated production of blue-green
aglal cellular products, causing additional macrophyte
deterioration. These events coupled with drought
conditions and unseasonably high water temperatures
increased blue-green algal growth, leading to addition-
al shading and byproducts. Ultimately, over 75 percent
of the macrophyte beds deteriorated and floated to the
surface in large rafts. By the onset of the present
investigation in 1978, maximum standing crop was
about 60 percent of earlier years and the previously
dominant macrophyte Ceratophyllum demersum had
been replaced by Elodea canadensis, especially in the
southern portion of the lake. As noted by Figure 7 and a
mean of 3 years of data, the most prolific growth is in
the richer, more unconsolidated sediments of the
southern end. Since this area has been proposed for
dredging (to remove the top 0.5 meter of rich
sediments) a considerable amount of prime macro-
phyte bed area will also be removed.
"853?
LEGEND
WTER QUALITY SAMPLING STATION
BENTKIC INSECT SAMPLING STATION
MACROPHYTE SAMPLING TRANSECT
Figure 6. — Sampling stations and macrophyte collection
transects.
UNNAMED OUTLET
EMERGENT jAjghAfl SP BEDS
MOSTLY PQTAMQGE TON AMPgFQLIUS
MOSTLY EjJOEA CANAOENSI^ (WITH NITELL4
CERAIQPHJLIUM OE;MERSuM W'TH STELLA Sf AND E_ QANAD£NSJS_
#HATQPK'fLJ-vJM DOMINANCE DISPLACED BY £_ CftNAPENSIS *J 1978
PQTAMOGE TON PANQRMlTANUg AND P PECTINATgS INCREASES at GREATER DEPTHS
PATCH
g AND P
Y DISTRIBUTION OF ElOQEA
* &NP PQTftMOGETW SPP
Figure7. — Distribution of aquatic macrophyte species.
-------
232
RESTORATION OF LAKES AND INLAND WATERS
Table 5. — Macrophyte biomass during maximum standing crop, August 2, 1978.
Location
Southern end and along
east side of lake
including MacKenzie Bay
Depth
(m)
0-1
1-2
2-3
3-4
4-5
5-6
Area
(m2)
14.6 x 103
18.8 x 103
38.6 x 103
132.0 x 103
176.0X 103
86.6 x 103
Sub-total southern
Eastern portion of lake
2-4
36.5 x 103
Ash free dry
weight per area
(9/m2)
31.7
126.0
171.0
98.5
41.9
13.0
portion of lake
38.9
Total
org. mass
(kg)
463
2,370
6,600
13,000
7,390
1,130
31,000
1,420
% P/
contour
.20
.21
.21
.16
.13
.13
.19
Total
P/contour
(kg)
.93
4.98
13.86
20.80
9.61
1.47
51.65
2.70
from MacKenzie Bay north to
public launch area
Dreamwood Bay 3-4 5.22 x103 20.5
Off southern end of
Wicomico Beach 2-3.5 13.6 x103 171.0
107
2,320
.18
.19
.19
4.41
Total lake
35,000
58.95
Table 6. — Macrophyte biomass during maximum standing crop, July 17, 1979.
Location
Southern end and along
east side of lake
Including MacKenzie Bay
Eastern portion of lake
Depth
(m)
0-1
1-2
2-3
3-4
4-5
5-6
6-7
Sub-total
2-4
Ash free dry
Area weight per area
(m2) (g/m2)
14.6 x 103
18.8 x 103
38.6 x 103
132.0 x 103
176.0 x 103
86.6 x 103
8.7 x 103
southern portion of lake
36.5 x 103
104
89
101
121
169
114
116
120
Total
org. mass
(kg)
760
2,670
5,021
17,491
14,465
2,338
433
43,178
5,291
% P/
contour
.15
.20
.39
.23
.38
.49
.36
.12
Total
P/contour
(kg)
1.14
5.34
19.58
40.23
54.97
11.46
1.56
134.28
6.35
from MacKenzie Bay north to
public launch area
Launch area to
Wicomico Beach
13.6 x 103
184
4,243
.23
9.76
Total lake
57,712
150.39
Table 7. — Dominant macrobenthic fauna of Liberty Lake. (Preliminary list by class, order or family where possible.)
Family Chironomidae
Ablasbesmyria
Chironomous
Cryptocladopelma
Cryptochironomous
Endochironomous
Glyptodendipes
Polypodium
Procladius
Pseudochironomous
Family Chaoboridae
Chaoborus
Family Ceraptopogonidije
Palpomyia
Alloaodomynia
(In addition, individuals of the classes
Pelecypoda, Gastropoda and Oligochaeta are
present in moderate to high numbers.)
Order-Ephem eroptera
Order-Odonata
Benthic Invertebrate Assessment
To assess effects of restoration efforts upon higher
aquatic food chains, extensive collections of benthic
invertebrates began in March 1978. Attempts are
being made to classify the organisms to genera and to
species where possible. Preliminary results indicate
that higher numbers of organisms are found in the mid-
lake and southern end of the lake. Large chironomids
were found in abundance in the mid-lake sediments
which intergrade between the heavy organic deposits
characteristic of the southern end and the lighter
organic material found in the northern end of the lake.
Dominant benthic forms found in the soft silty
sediments were dipteran larvae of the genera
Chironomus, Procladius, and Chaoborus. Tanytarsus
was also present in fewer numbers. Chironomus sp
were also found in numbers of 1,700 to 2,500
-------
SPECIAL PROJECTS AND TOPICS FOR ADDRESSING THE TROPHIC STATE
233
organisms/m2 in the northern and mid-lake sections as
well as 430 to 1,460 organisms/m2 in the heavier silt
and organic debris at the southern end. Figure 8
indicates the distribution found to date. Table 7 lists the
dominant forms.
Nelson (1980) will be publishing pre- and post-
treatment results. Our earlier studies indicate that with
an alum treatment bent hie invertebrates appeared to
increase moderately for a short period of time before
returning to previous population levels. Narf (1978)
also indicates that this may be the case, as well, in
alum-treated Wisconsin lakes. As only a portion of
Liberty Lake will be suction dredged in the fall of 1980,
it will be possible to study the effects of dredged and
undredged areas in conjunction with alum treatment.
In other field studies concerning benthic organisms,
over 100 fish stomach samples were collected during
the spring and summer of 1978 with the aid of
fishermen and the homeowners association. Much of
the data appears to be invalid because of the intensive
put and take fisheries. Most of the stomachs contained
corn, artificial eggs, hamburger bits, and pull tabs
rather than benthic forms. Two permit requests made
by our fisheries biologist to take fish by shocking or
netting were denied by the Washington State
Department of Game.
In laboratory studies being completed now (Lamb,
1980) a chironomid, Tanytarsus dissimilus, is being
subjected to acute and long-term alum toxicity tests. At
this time 240, 320, 560, and 750 mg/l of alum have
not shown acute toxicity to the organisms. In tests to
date all individuals appeared to remain fairly active and
have had no difficulty moving through the floe layer
(1.1 mm to 3.5 mm in depth) and actively injesting algal
cells, Selenastrum capricornutum, supplied for food. It
is expected that these data will be reviewed and
submitted for publication shortly.
\
LEGEND WEST FORK INLET
RELATIVELY LOW BENTHIC INSECT MAflERS fO.OOO ORGANISMS/M*)
MODERATE 6ENTHIC INSECTS NIW8CRS {1.000 10 2,000 ORGANISMS/M*!
MODERATELY HIGH BENTHIC INSECT NUMBERS ( 2.000 lo 4.000 ORGANISMS/M*) \ .
HIGH BENTHIC INSECT NUMBERS (4,000 lo 6,000 ORGANISMS/M1) ^, ' J
Lake Water Quality and Productivity
Measurements
Physicochemical Conditions
Generally the lake exhibits weak thermal stratifica-
tion during the summer with loss of oxygen in the
hypolimnion and occasional anoxic conditions (Gib-
bons, 1976). During this investigation the lake did not
stratify but oxygen levels were reduced to below 2.0
mg/l and eventually to zero near the sediments by
August. For purposes of brevity Table 8 lists only
maximum, minimum, and mean measurements taken
to date. Extensive measurements taken at 1.0 m
intervals are on computer file at WSU and in LEI data.
The concentration of phosphorus versus time has
revealed no discernible seasonal pattern. The usually
low concentration of phosphorus is probably due to its
almost instantaneous uptake by algae and macro-
phytes. Nitrogen to phosphorus ratio was 17:1 and
Schindler (1978) suggests that when N:P is more than
10:1 phosphorus is the most limiting of the two. As
Wetzel (1975) has noted, the concentrations of SRP
and TSP are not as significant as the rate of
interchange between SRP and TSP and paniculate
phosphorus in the water. The ability of the blue-greens
to accumulate phosphorus far in excess of their
immediate needs (Fogg, et al. 1973; Whitton, 1973),
helps to account for our repeated observation that
masses of Anabaena flos-aquae, A. spiroides, and
Gloeotrichia echinulata first appear in the vicinity of
decaying macrophytes and near the bottom sediments
before rising in the water column.
Table 8. — Summary of 1978-79 water quality conditions at
Liberty Lake (/ug/l except where noted).
Southeast
Parameter
TP
TSP
SRP
N-Ammonia
N-Nitrite-Nitrate
Total N
Alkalinity (mg/l)
HCO3
CO3
CO2
D.O. (mg/l)
pH (-log H+)
Secchi Disk (M)
Temperature
Chlorophyll a
Mean
30.0
7.5
1.5
10.0
15.0
400.0
21
<1.0
2.0
10.0
3.3
18.2
8.0
Min.
10.0
2.5
1.0
5.0
10.0
275.0
13.0
0.0
0.0
8.0
6.5
1.8
0.0
1.0
Max.
75.0
17.5
5.2
60.0
78.0
550.0
36.0
3.0
7.0
16.0
8.5
6.0
25.0
40.0
Northwest
Mean
3.5
7.5
4.0
10.0
20.0
360.0
20.0
<1.0
2.0
8.0
3.6
18.2
10.0
Min.
8.0
2.6
1.0
6.0
10.0
320.0
13.0
0.0
0.0
0.0
6.0
1.8
0.0
1.0
Max
78.0
12.5
5.6
70.0
58.0
545.0
36.0
3.0
17.0
16.0
8.6
6.0
25.0
25.0
Figure b. — Benthic invertebrate distribution.
Phytoplankton Productivity
The phytoplankton of Liberty Lake produced ap-
proximately 8.6 x 10s kg of organic carbon in 1978 and
5.9 x 10B kg of organic carbon in 1979. The annual rate
of productivity was estimated at 300 g C mVyr in 1978
and 205 g C m2 in 1979. We would like to attribute the
reduced productivity to the first restoration measures,
the completion of the sewage collection system and the
diversion of spring runoff waters from flooding the
marsh. However, part of the reduced phytoplankton
productivity is most likely due to competition from
increased macrophyte growth as can be noted by
-------
234
RESTORATION OF LAKES AND INLAND WATERS
comparing Tables 5 and 6; cooler weather and changed
patterns of precipitation are also other factors. Gibbons
(1980) has observed that diatoms made up over 50
percent of the standing crop during the same period. He
also noted that according to Hutchinson (1967) three of
the four prominent species (Fragilaria crotonensis,
Melosira granulata, and Tabellaria fenestrata) are
indicative of eutrophic waters. Figure 9 shows primary
productivity for 1978-79. Figures 10, 11, 12, and 13
show mean cell volumes of dominant species of blue-
greens and diatoms measured over the same period.
'A'M'J'J'A'S'O'N'D'J'F'M'A'M'J J A S 0
I978 I979
Figure 9. — A monthly summary of carbon 14 productivity
(integrated) for Liberty Lake Stations.
80
40-
0
750-
500-
250-
0
25H
- 0-
x
•g 1500-
i
„" 12001
I 900-
O)
O
- 600-
300-
0
750-
500-
250-
o-
Microcystis aeruginoso
Gloeotrichia echinulata
Coelosphaerium Naegelionum
Anabaena spiroides
Anaboeno flos-oquae
A'M' J' J'A'S'O'N'D'J'F'M'A'M'J 'J 'A'S'O
I978
1979
Figure 10. — Contribution to biomass of blue green algae at the
Southeast Liberty Lake Station.
Ceratium hirundinella, a relative newcomer and
prominent contributor to lake phytoplankton popula-
tions, is also shown.
20-
10-
0
600-
300-
0
E 20-
&
o
X2IOO-
|_I800-
oT 1500-
| 1200-
=5 900-
o
600-
300-
0'
400-
200-
0
Microcystis oeruginosa
Gloeotrichia echinulato
JIA
Coelosphoerium Noegelianum
Anobaena spiroides
Anaboena flos-aquoe
M'A'M'J'J'A'S'O'N'D'J'F'M'A'M'J'J'A'S'O
1978 1979
Figure 11. — Contribution of biomass by blue-green algae at
the Northwest Liberty Lake Station.
Chlorophyll a
Chlorophyll a has shown a sharp reduction over the
past 2 years; in 1977 lake values exceeded 30 /ug/l for
a 2-week period and ranged as high as 240//g/l. This
occurred after the early decline of macrophytes and
subsequent rise of Anabaena flos-quae, A. spiroides,
Gloeotrichia echinulata blooms (Figure 14). Aphani-
zomenon flos-aquae appeared for a short period under
the ice the first week of January 1979 after a late die-
off of some of the remaining macrophytes. Chlorophyll
a levels generally ranged from 2 to 20 /ug/l at the
southeast station (Figure 15). Chlorophyll a did reach
40 /jg/l at one level at the southeast station in late
October 1979.
Zooplankton
Thirty-four species and 28 genera of zooplankton
have been recognized by Gibbons (1980) at Liberty
Lake during the 1978-1979 study years. Eleven species
-------
SPECIAL PROJECTS AND TOPICS FOR ADDRESSING THE TROPHIC STATE
235
were Rotifera, 17 were Cladocera, five were Eucope-
poda, and one was a Diptera. Rotifera dominated the
community by numbers (87 percent) but made up only
4.3 percent of the biomass. The grazing Cladocera
made up 5 percent of the numbers but accounted for 58
to 67 percent of the biomass. The Calanoida
contributed 9 percent of the abundance but 25 percent
of the biomass over the study period. The contributions
made by the major species are shown in Table 9. The
careful documentation of the species composition and
numbers should aid in assessing the effects of
restoration measures upon secondary production.
200-
Ceratium hirundinella
I^c]: Tabellaria fenestrata
I75J
150-
100-
50-
I 600
1 J920
E
&
IO
O
X
K>
400-
200-
0-
700-
450J
300-
150-
o
900-I
Melosira /\ granuloto
Fragilaria crotonensis
1420
— -
" I
400-
200-
100-
1 I1830
Asterionella
formosa
M'A'M'J'J'A'S'O'NVJ'F'M'A'M'J'J'A'S'O
I978 I979
Figure 12. — Contribution to biomass by phytoplankton other
than blue-green algae at the Southeast Liberty Lake Station.
400
200^
I I
750-
200-
IOO-
S.
O)
I
0
1200-
750^
500-
250-
0
300-
200-
Ceratium hirundinella
/\
Tabellaria fenestrata
265
Melosiro gronulata
3482
Fragilaria crotonensis
3143
A
QfV") -
' Asterionella formosa
I \ I
600H
300-
0 'A'M'J'J'A'S'O'N'D'J^'M'A'M'jTrA'S^O^
1978 1979
Figure 13. — Contribution to biomass by phytoplankto other
than blue-green algae at the Northwest Liberty Lake Station.
Figure 14. — Southeast Liberty Lake chlorophyll a
measurements (ug/l) July— October, 1977.
SUMMARY
The low buffering capacity, shallow depth, and
relatively long-term detention time of Liberty Lake
waters has precluded a large one-time monophase
restoration effort. In addition, the high usage rate by
residents, park visitors, fishermen, and water sports
enthusiasts has made a stepwise treatment the most
judicious route to follow. This procedure also insures
the protection of commerical interests whose livelihood
depends upon seasonal use of the lake.
The most distinctive asset in terms of restoration
protection has been the undisturbed upper watershed
of Liberty Lake. Phosphorus and nitrogen content of
inflowing waters is very low unless overflow flushes
additional nutrients from the adjoining marsh.
A 15 to 20 percent reduction in phosphorus input has
been achieved by diverting flood waters away from the
marsh and to the lake through repair of the West Fork
of Liberty Creek.
Sewering 95 percent of the residential area around
the lake has diverted another 150 to 170 kg of nitrogen
-------
236
RESTORATION OF LAKES AND INLAND WATERS
AMJJASONDJFMAMJJASO
Figure 15. — Chlorophyll a measurements at Liberty Lake
Southeast Station (ug/l).
Table 9. — Summary of principal taxa comprising the zoo-
plankton community of Liberty Lake, Wash. (9/27/78-
8/22/79), including percent composition in terms of density
(#'s/m3) and dry weight standing crop {/jg/rn3) over all dates at
the northwest and southeast stations (Gibbons, 1980).
Taxon
Percent Composition
(#'s/m3)
Cladocera
Bosmina longirostris
Chydorus sphaericus
Ceridaphnia lacustris
Diaphanosoma brachyurum
Daphnia pu/ex
Daphnia schodleri
Daphnia galeata mendotae
Daphnia thorata
Daphnia immatures
Leptodora kindtii
Rotifera
Keratella cochlear/s
Kellicottia longispina
Trichocera cylindrica
Gastropus sp.
Conochilus sp.
Polyarthra spp.
Calanoid Eucopepoda
Nauplii
Copepodids
Diaptomus reighardi
Cyclopoid Eucopepoda
Copepodids
Mesocyclops Leuckarti
Macrocyclops albidus
Diptera
Chaoborus sp.
NW
4.1
*
0.5
*
0.1
0.5
0.3
1.3
0.8
0.5
*
86.3
24.8
38.3
2.9
0.6
0.8
18.8
9.3
7.2
1.6
0.5
0.3
0.3
*
#
*
#
SE
2.8
*
0.6
*
0.3
0.3
*
0.9
0.4
0.3
*
87.2
41.0
28.2
1.8
1.0
0.2
14.8
9.8
8.3
1.1
0.4
0.2
0.2
*
*
*
*
+g/m3
NW
66.5
0.1
1.9
0.1
0.6
18.2
9.4
19.7
12.0
4.2
0.1
4.3
1.7
0.7
*
*
*
1.8
25.5
2.0
8.6
14.9
2.9
1.4
0.2
1.3
0.8
0.8
SE
57.7
0.2
3.2
0.1
3.0
13.9
2.6
21.4
8.9
3.9
0.2
8.0
4.6
0.9
*
*
#
2.2
30.1
3.8
10.0
16.3
3.5
1.1
0.4
1.8
0.7
0.7
* '60.1%
from its yearly movement to the lake. It is estimated
that septic tank drainage beds will continue to leach for
another 4 to 7 years depending upon the hydraulic
pressure placed on them.
The importance of reducing the macrophyte beds and
their rich substrata cannot be overemphasized.
Mawson (1980) has shown that when the sediments
become anaerobic a potential 150 to 280 kg
phosphorus could be released. Based upon Macken-
thun and Ingram's (1967) estimates of phosphorus
contained in algae, the potential release of phosphorus
from macrophytes and sediments theoretically pro-
duces 77 to 100 metric tons of algae in the water. After
dieoff and decay of macrophyte beds and subsequent
algal blooms in 1971 and 1973, we estimated
approximately 60 to 70 metric tons of debris on the
beaches alone (Funk, et al. 1975).
Suction dredging is planned for fall 1980 to remove
up to 33 percent of the rich top sediment and about 50
percent of the heavier macrophyte growth area.
Another alum treatment of 10 mg/l is planned to
coincide with the dredging to reduce phosphorus
released by suspended sediments. This latter treatment
will also help seal freshly exposed sediments and aid in
breaking the nutrient cycle without overwhelming the
buffering capacity of the lake.
A program is being planned with M. Kennedy
Engineers to reduce urban runoff both mechanically
and by educating the residents on the importance of
minimum lawn fertilization, litter control, and clean
streets.
The ultimate goal is to reduce controllable nutrient
input (especially phosphorus) by about 40 percent and
to assess as accurately as possible the value of each
restoration measure put into practice.
Barring undesirable land use practices beyond
controlled areas and massive influx of human
populations we are optimistic about the future of
Liberty Lake.
REFERENCES
American Public Health Association. 1975. Standard methods
for the examination of water and wastewater. 14th ed.
Bottrell, H. H., et al. 1976. A review of some problems in
zooplankton production studies. Norw. Jour. Zool. 24:419.
Brooks, J. L. 1957. The systematics of North American
Daphnia. Mem. Conn. Acad. Arts Sci. 13.
Copp, H., et al. 1976. Investigation to determine extent and
nature of nonpoint source enrichment and hydrology of
several recreational lakes of eastern Washington. Wash.
Water Res. Center Rep. 26. Washington State University,
Pullman.
Edmondson, W. T., ea. 1959. Fresh-water biology. 2nd ed.
John Wiley & Sons, New York.
Edmondson, W. T. 1972. Nutrients and phytoplankton in Lake
Washington. In G. E. Likens, ed. Nutrients and eutrophica-
tion: The limiting-nutrient controversy: special symposium.
Limnol. Oceanogr. 1:172.
1974. A simplified method for counting
phytoplankton. Pages 14-16 in R. Vollenweider, ed. A
manual on methods for measuring primary production in
aquatic environments. IBP handbook 12.
Edmondson, W. T, and G. G. Winberg, eds. 1971. A manual
on methods for the assessment of secondary productivity in
freshwaters. IBP handbook 17. Blackwell.
Elliot, J. M. 1971. Some methods for the statistical analysis
of benthic invertebrates. Sci. Publ. Freshw Biol. Assoc.
25:1.
Fogg, G. E., et al. 1973. The blue-green algae. Academic
Press, New York.
Funk, W. H., H. L. Gibbons, and G. C. Bailey. 1979. Effect of
restoration procedures upon Liberty Lake. First status
report. In Limnological and socioeconomic evaluation of
lake restoration projects. EPA-600 3/79/005. U.S. Environ.
Prot. Agency.
-------
SPECIAL PROJECTS AND TOPICS FOR ADDRESSING THE TROPHIC STATE
237
Funk, W. H., et al. 1975. Determination, extent, and nature of
nonpoint source enrichment of Liberty Lake and possible
treatment. Wash. Water Res. Center Rep. 23, Washington
State University, Pullman.
1976. Investigation to determine extent and
nature of nonpoint source enrichment and hydrology of
several recreational lakes of Eastern Washington. Part II.
Wash. Water Res. Center Rep. 26. Washington State
University, Pullman.
Gibbons, H. L. 1976. The primary productivity and related
factors of Liberty, Newman, and Williams Lakes in Eastern
Washington. M.S. thesis. Washington State University,
Pullman.
1980. Phytoplankton production and regulating
factors in Liberty Lake, Washington with possible implica-
tions of restoration activities on the primary productivity.
Ph.D. thesis. Washington State Unversity, Pullman.
Gibbons, M. V. 1980. An investigation of the zooplankton
community of Liberty Lake, Washington with special regard
to seasonal succession. M.S. thesis. Washington State
University, Pullman.
Hall, D. J., W. E. Cooper, and E. E. Werner. 1970. An
experimental approach to the production dynamics and
structure of freshwater animal communities. Limnol.
Oceanogr. 15:839.
Hutchinson, G. E. 1967. A treatise on limnology. Vol. II.
Introduction to lake biology and the limnoplankton. John
Wiley & Sons, New York.
Lamb, D. S. 1980. Acute and chronic effects of alum to midge
larvae (Chironomidae). M.S. thesis. Washington State
University, Pullman.
Lee' D. R. 1977. A device for measuring seepage flux in lakes
and estuaries. Limnol. Oceanogr. 22:140.
Mackenthun, K. M., and W. M. Ingram. 1967. Biological
associated problems in freshwater environments. Fed.
Water Pollut. Control Admin. U.S. Dep. Inter. Washington,
D.C.
Mason, W. T. Jr., and P. P. Yevich. 1967. The use of phloxine B
and rose bengal stains to facilitate sorting benthic samples.
Trans. Am. Microsc. Soc. 86:221.
Mawson, S. 1980. The impact of the sediments on the
phosphorus loading of Liberty Lake as a result of diffusion.
M.S. thesis. Washington State University, Pullman.
McComas, M. R. 1977. State of the art of measuring
groundwater flow to lakes. Presented at Mechanics of Lake
Restoration Conf. Madison, Wis. (Preprint).
Minto, M. L. 1977. A sampling device for the invertebrate
fauna of aquatic vegetation. Freshw. Bio. 7:425.
Narf, R. P. 1978. An evaluation of past aluminum sulfate lake
treatments: present sediment aluminum concentrations
and benthic insect communities. Wis. Dep. Nat. Resour.
Rep. (Manuscript).
Nelson, R. O. 1980. An evaluation of the relative efficiencies
of benthic macrophyte samplers used at Liberty Lake. M.S.
thesis data. Washington State University, Pullman.
Ohle, W. 1953. Phosphor als Initialfactor de Gewassereu-
trophierung. Vom. Wasser 20:11.
Osborn, J. F. 1973. Water balance of Liberty Lake,
Washington. Rep. to Futrell, Redford, and Saxton.
Pennak, R. W. 1978. Fresh-water invertebrates of the United
States. 2nd ed. John Wiley & Sons, New York.
Peterka, J. J., and K. M. Knutson. 1970. Productivity of
Phytoplankton and quantities of zooplankton and bottom
fauna in relation to water quality of Lake Ashtabula
Reservoir, North Dakota. OWRR Project No. A-011 -NDAK.
North Dakota State University, Fargo.
Rector, Thomas. 1979. Groundwater nitrogen and phos-
phorus, yearly contribution to Liberty Lake (Unpublished
data).
Sawyer, C.N. 1947. Fertilization of lakes by agriculture and
urban drainage. Jour. N.E. Waste Works Assoc. 51:109.
1952. Some new aspects of phosphates in
relation to lake fertilization. Sewage Ind. Wastes 24:768.
Schindler, D. W. 1978. Factors regulating phytoplankton
production and standing crop in world's freshwaters.
Limnol. Oceanogr. 23:478.
Schwoerbel, J. 1970. Methods of hydrobiology (freshwater
biology). Pergamon Press, Toronto.
Thomas, E. 1969. The process of eutrophication in central
European lakes In Eutrophication: Causes, consequences,
correctives. Natl. Acad Sci. Washington, D.C.
U.S. Environmental Protection Agency. 1969. Chemistry
laboratory manual. Bottom sediments. Great Lakes Regional
Committee on Analytical Methods. Fed. Water Qual. Admin.
Vallentyne, J. R. 1974. The algal bowl lakes and man. Misc.
Publ. 22. Dep. Environ. Ottawa, Canada.
Vollenweider, R. A. 1974. A manual on methods for
measuring primary production in aquatic environments. 2nd
ed. IBP handbook 12.
Wetzel, R. G. 1975. Limnology. Saunders Co.
Whitton, B. A. 1973. Freshwater plankton. In N. G. Carr and B.
A. Whitton, eds. The biology of blue-green alage. University
of California Press.
ACKNOWLEDGEMENTS
The authors acknowledge the continuing contribution of
Maribeth Gibbons and Forrest Woodwick in zooplankton
identification; Ralph Nelson for identification of microin-
vertebrates; Barry Moore, field and laboratory chemical
analyses; Simon Mawson and Stephen Breithaupt, sedi-
ment nutrient studies and chemical analyses; Paul J.
Bennett, laboratory chemical analyses; David Lamb and
Sydney Harper, aluminum toxicity studies. All have aided in
field work in addition to biologists Tom Rector and Phillip
Kaufmann of M. Kennedy Engineers and Don Secor, County
Parks ranger. We are also indebted to W. T. Edmondson and
Ami Lift for aid with some difficult zooplankton identifica-
tions.
-------
238
THE CONTINUING DILUTION OF MOSES
LAKE, WASHINGTON
E. B. WELCH
M.D. TOMASEK
Department of Civil Engineering
University of Washington
Seattle, Washington
ABSTRACT
The quality of Moses Lake during 1977-79 was markedly improved over that in 1969-70 by the
addition of low-nutrient Columbia River dilution water. Total N, total P and chlorophyll a improved
by 50 percent or greater while Secchi transparency and the fractional composition of blue-green
algae in the phytoplankton improved about 40 percent. This resulted from average spring-summer
water exchange rates of 10 percent per day in Parker Horn (8 percent of volume) and 0.8 percent
per day throughout the whole lake Although the response of phytoplankton is rather complex,
dilution water primarily reduces biomass and chlorophyll a by diluting total N in the inflow and
causing some instability in the water column through increased circulation, which discourages the
accumulation of large blue-green blooms. Other factors such as iron limitation of N fixation in
rates and reduction of allelopathic substances may also contribute, but are less discernable in the
data Future control efforts will be oriented toward better distribution of about the same amount of
dilution water added throughout the lake in the spring-summer of 1977-79.
INTRODUCTION
Large quantitites of low nutrient Columbia River
water have been diverted through existing facilities
into Moses Lake, a large (2,853 ha) rather shallow (z =
5.6 m) eutrophic lake in eastern Washington. The
dilution water has been added to Parker Horn from the
Eastlow Canal via Rocky Coulee Wasteway and Crab
Creek (Figure 1) at varying rates on seven occasions
during the spring-summer of 1977-1979 (Table 1).
These inputs resulted in hypothetical lake water
exchange or renewal rates of about 10 percent per day
for Parker Horn and about 0.8 percent per day for the
whole lake, which is about 10 times greater than
normal.
Substantial improvement in Moses Lake quality was
expected. It was hoped that the total phosphorus
content would decrease about 50 fjg f1 and as a result
control chlorophyll a to an average of about 20 fjg I
Although it was known that soluble nitrogen was
normally depleted and was therefore the apparent
growth rate limiting nutrient during summer, total P
was expected to better determine biomass because of
the prominance of N-fixing blue-green algae. Algal cell
loss through simple washout was not expected to
appreciably affect biomass.
The results from 1977-1979 that are briefly
described here show that total N rather than P probably
accounted for much of the 53 percent decrease in
chlorophyll a. However, the addition of dilution water
may also physically deter blue-green algal blooms by
decreasing water column stability.
Although dilution water was more completely
distributed into the Main Arm of the lake than
expected, two additional phases of the restoration
Figure 1. — Moses Lake, Washington: Showing sampling
transect locations and treated sewage affluent (SE) and pipe
connecting Parker and Pelican Horns proposed in Phase II.
-------
SPECIAL PROJECTS AND TOPICS FOR ADDRESSING THE TROPHIC STATE
239
project are planned by Brown and Caldwell Engineer-
ing and involve delivering water to Pelican Horn and
the Upper Main Arm (Figure 1). Essentially the long-
term program for the lake includes a continuous total
input of 20 m3 sec"1 from April 1 to June 15, and from
20 to a minimum of 7 m3 sec"1 from June 15 to August
15. The principal change for the long term is to add
about the same amount of water during the spring-
summer period, but to distribute it more evenly. Total
dilution water input has been 203,111, and 250 x 106
m3 during 1977, 1978, and 1979, respectively. The
continuous flows described here amount to an April 1
to August 15 total of either 264 or 178 * 106 m3. The
continuous addition, but at a lower rate, is expected to
•.provide at least as good a quality as did the 1977-79
patterns of dilution. The diversion facility from Parker to
Pelican Horn should be completed by summer 1981.
Table 1. — Dilution water inflow rates to Parker Horn, Moses Lake via
Crab Creek showing hypothetical water exchange rates for Parker
Horn (8 percent volume) and the whole lake (100 percent volume)
during April through September of the 3 years.
Mean percent of surface light available (I) to
phytoplankton was calculated by
Year
1977
1978
1979
Dilution
period
3/20-5/07
5/22-6/04
8/14-9/18
(96 days)
4/20-6/18
(60 days)
4/03-6/04
7/11-8/28
9/20-10/18
(138 days)
Mean flows in m3 sec"1
dil. water
33.6
10.5
17.3
21.7
25.1
16.3
23.2
Crab Creek
0.4
1.3
2.5
1.7
1.5
April-September
exchange rate in days"1
Parker Horn whole lake
0.11 0.009
0.07 0.006
0.13 0.010
PROCEDURES
Water was continuously sampled at least weekly
along seven horizontal transects in 1977 and eight in
1978 and 1979 at depths of 0.4 meters (Figure 1). The
composites were analyzed for total P, soluble reactive P
(SRP), nitrate N(NO3-N), total N, chlorophyll a, and
specific conductance. Station 12 was added in 1978. At
the midpoint of each transect vertical profiles of these
variables as well as temperature and Secchi disk depth
were determined. The surface values for constituents
in the profile were used only occasionally in this
analysis. The horizontal transect composites were used
as primary data to compare with values from 1969-70,
which were determined by the same collection
techniques. Procedures for nutrient and chlorophyll a
analysis followed those of Strickland and Parsons
(1972).
Phytoplankton analysis involved counting cells or in
the case of blue-greens, unit colonies, in the horizontal
transect composites. All counts were converted to
volumes based on appropriate cell measurements and
geometric configurations. Blue-green count-to-volume
conversions were based on procedures by Strathmann
(1977). Data from 1970 are based on daily counts made
on surface samples taken from a point midway
between stations 5 and 7 during four 2-week periods in
the summer. Means for those periods were converted
to cell volume.
- e"KZ)
KZ
where Z is the depth of mixing estimated from
temperature profiles and K is the extinction coefficient
estimated from Secchi disk measurements, assuming
that the depth of maximum visibility was 10 percent of
lo, the surface intensity. The depth of mixing was either
2 or 4.5 m at station 7 and 1,2,6, or 9.2 m at station 9
depending on whether or not the temperature change
exceeded 1 .0°C.
Percent lake water (% LW) was estimated by
assuming that 100 percent LW would be represented
by the conductance of Crab Creek water and 0 percent
LW by the conductance of Columbia River water. The
percent residual LW was thus calculated on each
sample date by
%LW = -
LWSC - CRSC
CCSC - CRSC
where LWSC, CRSC, and CCSC are specific con-
ductances for lake water, Columbia River water, and
Crab Creek water.
Water column stability was estimated by the change
in temperature between surface and bottom at each
station and indicated by A°C.
RESULTS
Quality Improvement
Marked improvement occurred in nutrient content,
phytoplankton biomass composition, and transparency.
Total P decreased by 45 percent from a volume
weighted mean of 156//gf1 in 1969-70 to 86 fjg I"1
in 1977-79, chlorophyll a by 53 percent from45,-21jugr1
and Secchi disk transparency increased from 0.9 to 1.5
m. These changes are based on values from all stations
except 12 and represent 58 percent of the lake volume
for the May to September period. The average rate of
water replacement was calculated as 1.4 percent per
day for that volume. Degree of improvement was even
about 10 percent greater at station 7 in lower Parker
Horn (replacement 10 percent per day) in terms of
nutrient content and more so in terms of chlorophyll a.
Of course, chlorophyll a in the lower lake was less than
in Parker Horn even in the predilution years in spite of
little change in nutrient content between the two
areas.
Although average nutrient content and chlorophyll a
were reduced by about 50 percent, large blooms of
blue-greens nevertheless occurred in 1977 and 1979
and were characterized by chlorophyll a content
exceeding 100/ug I"1 The blooms were considerably
delayed by dilution, however, compared to the non-
dilution years of 1969-70. Once the input of dilution
water stopped, phytoplankton biomass began to
increase after 3 to 4 weeks. Blooms were not
pronounced in 1978 and were delayed much longer —
about 2 months.
-------
240
RESTORATION OF LAKES AND INLAND WATERS
II-DILUTIOH 1969-70 VERSUS DILUTION 1977-79
PARKER HORN (7) LOWER LAKE (9)
L
150
100
50
NON-DILUTION VERSUS DILUTION WITHIN 1977-79
PARKER HORN (7) LOWER LAKE (9)
I
f, ^-"
Figure 2. — Mean values for Total P, Total N ( ), chlorohyll a,
and % lake water (LW) in Parker Horn and the lower lake
during dilution (stippled bar) and non-dilution (solid bar)
periods between years (1977-79) versus 1969-70) and within
years (1977-79).
formation (caused by high concentrations of blue-
greens).
Controlling Nutrients
Dilution water appears to cause nitrogen rather than
phosphorus to be the principal macronutrient control-
ling phytoplankton biomass. Figure 3 suggests that
decreasing the total N to less than 600 /jg f1 by adding
dilution water reduces the phytoplankton chlorophyll a.
The N values during non-dilution periods (solid
squares) tend to be higher and associated with higher
chlorophyll a than those during dilution periods.
Phosphorus, on the other hand, appears to be less
important as a biomass control. Because N values tend
to lie to the left and P to the right on the scale of 10N:1P
by weight, a reasonable requirement ratio for algal
growth, N would appear to be most limiting to biomass
formation.
sor
S 20
O •
•D
o
o
0
04>-
200
100 600 800
TOTAL N IN uG L'1 AND TOTAL P IN uG
The rather temporary effect of dilution water input in
batches is illustrated in Figure 2 by comparing dilution
and non-dilution period means for nutrients, chloro-
phyll a, and % LW. Lag times of 10 and 20 days were
allowed in calculating means for Parker Horn and the
lower lake. Note that chlorophyll a for the non-dilution
period is much greater than that for the dilution period
in Parker Horn compared to those in the lower lake.
That is, recovery of phytoplankton was greater in
Parker Horn. Note also that chlorophyll a at the lower
lake station was quite different between dilution and
non-dilution periods in spite of little differences in %
LW.
The phytoplankton is still dominated by blue-green
algae, especially Aphanizomenon, during June through
September as it was in the pre-dilution years. However,
the fraction of the volume contributed by blue-greens
decreased to 55 percent during 1977-79 compared to
96 percent in 1970. Although the observations in 1970
were at a point midway between stations 7 and 5 they
are probably comparable. Little difference in composi-
tion was noted between stations 7 and 9 in 1977 and
1978. Thus, less biomass coupled with a greatly
reduced fraction of blue-greens resulted not only in
clearer water, but also in a lower frequency of scum
Figure 3. — Chlorophyll a related to Total N (closed symbols)
and Total P (open symbols) as measured during dilution
(circles) and non-dilution (squares) periods at stations 7 and
Parker Horn and Lower Lake, with 10- and 20-day lag times,
respectively, during 1977-79.
Chlorophyll a:cell volume ratios suggest that growth
rate was indeed nutrient limited and the soluble N:P
ratios indicate also that N was in shortest supply during
both pre-dilution and dilution years (Table 2). A few
measurements of NhU-N during maximum growth
indicated that the ratios of total soluble N:P were
probably as much- as four times greater than those
listed in Table 2. Chlorophyll a:cell volume ratios were
near or within the zone of "moderate nutrient
deficiency" suggested by Healy (1978). Thus, cell
growth was N limited during dilution just as it was prior
to dilution, largely as a result of excess P.
Dilution did not necessarily alter the pattern of
nutrient limitation but it further restricted the
availability of N, the most limiting nutrient, with
differences being apparent in the dilution years. For
example, total N was substantially less in 1978 and
1979 than in 1977, 535 and 550 versus 600 yug l~1 in
the lower lake. The ratio of organic N:P (total minus
soluble) was also higher in 1977 (12) than in 1978 or
-------
SPECIAL PROJECTS AND TOPICS FOR ADDRESSING THE TROPHIC STATE
241
1979 (8 and 9). Phosphorus was initially thought to be
the key nutrient to control by adding dilution water to
limit biomass. Nitrogen fixation by the most abundant
blue-green, Aphanizomenon. was considered capable
of providing the N needs to match the available P.
However, N fixation apparently supplied insufficient N
to use the available P; a reduction in total N through
dilution therefore effectively lowered average algal
abundance.
Table 2. — Mean chlorophyll a cell volume ratio, SRP and
NOs - N during 1977-1979 in Tower Moses Lake (station 9).
Nutrient means are from May-September data and
phytoplankton variables are for June-September.
1969-70
1977
1978
1979
chl a:
cell volume
2.8*
4.1 .
2.5
1.0
NOs-N
^gr1
23
48
19
15
PCU-P
^gr1
48
56
24
26
N
P
0.5
0.9
0.8
0.6
* cell volume data from Buckley (1971) at point midway
between stations 5 and 7 and chl a station 7.
Effect of Washout, Light, and
Stratification
Although dilution water input decreased the amount
of the most limiting nutrient (N), physical changes are
thought to have influenced the persistence of blue-
green blooms and in that way also controlled biomass
and possibly species composition. One such physical
effect is washout, or elimination of biomass from the
system at a greater rate than the growth rate can
supply new cells. From Figures 4 and 5 it is apparent
that on some occasions chlorophyll a declined in
proportion to % LW after the beginning of a dilution
period followed by a recovery and mid-summer bloom.
The response of chlorophyll a to % LW is most
pronounced in Parker Horn where the two variables
appear closely related following dilution water inputs in
the spring of 1977 and 1979. In the lower lake,
however, the response in chlorophyll a to a decrease in
% LW was much greater than would have been
expected from simple dilution or cell washout. In 1977
chlorophyll a decreased from 137 to 10//g I"1 in a
period of 1 week while % LW decreased only 25
percent from 52 to 39. In 1979 chlorophyll a continued
to increase in August while % LW decreased from 46 to
28. The hypothetical water exchange rates in the lower
lake during the late summers of 1977 and 1979 were
about 5 percent per day. If indeed equally mixed, such
rates of exchange are too low to substantially exceed
growth rate and account for that magnitude of
chlorophyll a decrease. However, the surface layer
containing the greatest concentration of chlorophyll a
would tend to be preferentially lost through the lake
outlet.
An algal bloom did not begin to develop in 1978 until
late August (Figure 6) even though there was only one
dilution period following which percent LW recovered
as in 1977 and 1979. Furthermore, stratification was
as great in June and July of 1978 as in 1977 and 1979.
80 100
60 75
40 50
20 20 -
60 75
10 50
20 25
Figure 4. — Chlorophyll a (open circles), % LW (solid circles)
and A°C (squares) in Parker Horn (upper) and Lower Lake
(lower) druing 1977. Dilution periods are indicated by
brackets.
SJ 1UJ .
60 75
40 50 -
20 25
60 75
40 50
20 25
Figure 5. — Chlorophyll a (open circles), % LW (solid circles)
and A°C (squares) in Parker Horn (upper) and Lower Lake
(lower) during 1979. Dilution periods are indicated by
brackets.
-------
242
RESTORATION OF LAKES AND INLAND WATERS
80 100
60 75
10 50 -
20 25 •
60 75 •
tfl 50 -
20 25 -
Figure 6. — Chlorophyll a (open circles), % LW (solid circles)
and A°C (squares) in Parker Horn (upper) and Lower Lake
(lower) during 1978. Dilution periods are indicated by
brackets.
Physical washout no doubt slowed biomass buildup
in Parker Horn during the first dilution inputs in April
when exchange rates were 20 to 25 percent per day.
However, blue-green algae were shown to grow at a
maximum rate of 50 percent per day in 0.5 m deep
plastic bags in situ (Buckley, 1971). Deliberate
exchange of water in the bags at 10 percent per day
demonstrated that blue-green increase was minimally
affected by physical dilution (washout). This is shown
in Figure 6 where the rate of increase in chlorophyll a
was similar (about 40 percent per day in 100% LW)
whether or not exchanged at 10 percent per day.
Light intensity apparently did not greatly influence
bloom formation or dieoff during 1977 or 1979. Table 3
shows that I, the average %l (surface intensity)
available in the mixed layer, was not different 1 month
prior to compared with 1 month after the maximum
bloom biomass. Further, the June and July means in
Parker Horn and the lower lake were 22, 45, and 37
percent for 1977, 1978, and 1979, respectively. Thus,
the failure of a large bloom to develop in 1978 was
probably not due to low light levels resulting either
from exceptionally well mixed conditions or high
extinction coefficients.
The factor that appears most related to bloom
development and its subsequent crash is water column
stability. Decreases in stability following dilution water
input may have been caused largely by the greater
mixing produced by the increased rate of water
exchange. However, wind influences vertical mixing in
Moses Lake and it is rather difficult to separate effects
of wind and water exchange rate. Periods of increased
stability are normal in Moses Lake, which allows the
surface temperature to increase causing even firmer
stratification. The buoyant blue-greens tend to be
favored by a stable water col.umn and gradually
accumulate in the surface layer at concentrations in
excess of 100/L/gf1 chlorophylls.
Figure 4 indicates that the blooms in both Parker
Horn and the lower lake developed under rather stable
conditions, but stability tended to break up following
the third dilution input and the bloom crash nearly
coincided with the decreased stability possibly causing
the surface accumulation of chlorophyll a to disperse.
The picture was not quite so clear in 1979 (Figure 5).
The bloom crash in the lower lake did not coincide with
decreased stratification as in 1977, but the 1979 bloom
was also not as large as that in 1977. Chlorophyll a in
Parker Horn in 1979 remained rather high during the
dilution period and the water column remained
moderately stable as well. Nevertheless, average water
column stability, indicated by A°C, was substantially
different preceding and following bloom maxima during
both 1977 and 1979 (Table 3).
Stability and % LW remained apparently favorable for
bloom development during June and July in both
Parker Horn and the lower lake in 1978, but for other
reasons, including lower total N content, a substantial
bloom did not occur.
60
20
60
-------
SPECIAL PROJECTS AND TOPICS FOR ADDRESSING THE TROPHIC STATE
243
Table 3. — Average (± SE) water column stability, percent of
surface light (I) and surface (0.5 m) chlorophyll a 1 month
preceding compared to 1 month following blue-green algal
blooms at stations 7 and 9 during 1977 and 1979.
Chi a: fjg f1 Stability A°C
pre maximum
post maximum
58 ± 8
28 ± 4
3.0 ± 0.45
1.6± 0.45
24 ± 3.3
21 ± 2.8
DISCUSSION
The addition of low nutrient Columbia River water to
Moses Lake has markedly improved its quality.
Improvements have either equaled or exceeded 50
percent for most variables. This has occurred with rates
of water exchange averaging only 1.4 percent per day
for nearly two-thirds of the lake. Improvements were
slightly greater in the lake section where dilution water
entered (Parker Horn) and exchange rates averaged 10
percent per day with maximumsof 20 to 25 percent per
day. Dilution had a greater relative effect in Parker
Horn probably because of the large fraction of soluble
nutrients that normally enter with irrigation return
flows in Crab Creek.
The exact cause(s) for the overall improvement is not
entirely clear. Although N and P totals decreased,
soluble fractions throughout the lake were not greatly
less and in some cases averaged even more than in
predilution years. As in pre-dilution years, N remained
the most limiting nutrient for growth rate. The data
suggest that total N set the limit on average chlorophyll
a and probably biomass as well. N fixation by the
abundant N fixing blue-green Aphanizomenon was
apparently insufficient to utilize the available P.
At the rates of water exchange employed, phyto-
plankton cell loss through washout probably reduced
biomass significantly only in Parker Horn where
exchange rates during peak spring inflow were at times
20 to 25 percent per day. The decreased biomass in
other parts of the lake, where exchange rates during
peak inflows were less than 10 percent per day and on
the average 1 to 2 percent per day, was probably not
caused by washout. On the other hand, increased
instability in the water column indicated by small
temperature differences with depth (A C), may have
been brought about by increased rates of water
exchange. This was suggested by significantly lower
A°C,for the 1 -month periods following blooms
compared to values during the month preceding
blooms. However, weekly observations may not
realistically reflect the nature and cause of stability,
which contributes to or detracts from bloom formation.
Although the degree of stability appears related to
bloom formation and demise and may have been
markedly influenced by dilution water input, much of
the instability may also have been caused by wind.
Ahlgren (1979) has suggested that the dominance of
blue-green algae in eutrophic lakes may be caused to a
large extent by their tolerance of low light and stable
conditions. Given sufficient nutrient availability, diatom
and green algae biomass could increase until light
extinction limited their growth. Blue-greens would
then be favored because of their lower light require-
ments in addition to their buoyancy allowing them to
rise in the water column. Under stable conditions, their
advantage would be even greater because diatoms and
green algae would sink. If rates of water exchange on
the order of 5 to 10 percent per day prevent stability in
the water column and in so doing discourage blue-
green algae, possibly that is an additional benefit of
dilution water, even if the added water is not low in
nutrients. Eutrophic, productive lakes, in which blue-
green algae are not dominant, are not uncommon.
Palmont (1980) suggested yet another cause of
biomass control in Moses Lake. Soluble iron was found
at concentrations on the order of 1 /ugl~1 at station 9 in
August 1978. Such low iron values are of particular
interest when compared with results of bioassays in
Clear Lake, Calif., a similar hypereutrophic lake in an
arid region. Wurtsbaugh, et al. (1978) found adding
ferrous iron to lake water containing 2 to 30 AigT1
dissolved iron greatly increased N-fixation rates and
chlorophyll a. Dilution water in Moses Lake may have
diluted iron, reducing N-fixation. This would have
allowed total N flowing into the lake to act more as a
control on biomass and chlorophyll a The lower
concentration of total N in Columbia River water in
1978 than 3977 (190 versus 360 /ugT1) may have
accounted in part for the lower chlorophyll a in 1978
than 1977 (16 versus 26 /ugl'1). Aside from iron
limiting N fixation, the relatively slow rate of N fixation
(5 percent per day: Home and Goldman, 1972) in a
system with rather high water exchange rates may
have also contributed to the inability of fixation to
supply enough N to utilize the available P
Another potential cause for the dilution water effect
in Moses Lake is that of reducing allelopathic
substances excreted by blue-green algae. The dilution
of such substances, which have been shown to
suppress the growth of non-blue-greens (Keating,
1978), may have contributed to blue-greens becoming
less abundant in favor of more diatoms and green algae
(Welch and Patmont, in press). The difficulty with this
hypothesis is that blue-greens represented as sizable a
fraction of the biomass in Parker Horn (station 7),
where % LW averaged less, as in the lower lake (station
9) where % LW was greater. Although there may be
other explanations to the similar fractional contribution
of blue-greens in the two lake areas, there is
nevertheless no discernible direct relationship be-
tween % LW and blue-green fractional composition.
REFERENCES
Ahlgren, I. 1979. Lake metabolism studies and results at the
Institute of Limnology in Uppsala. Arch. Hydrobiol. Beih.
Ergebn. Limnol. 13:10.
Buckley, J. A. 1971. Effects of low nutrient dilution water and
mixing on the growth of nuisance algae. M.S. thesis.
University of Washington, Seattle.
Healy, F. P. 1978. Physiological indicators of nutrient
deficiency in algae. Mitt. Int. Verein. Limnol. 21:34.
Home, A. J., and C. R. Goldman. 1972. Nitrogen fixation in
Clear Lake, California. I. Seasonal variation and the role of
heterocysts. Limnol. Oceanogr. 17:678.
Keating, K. I. 1978. Blue-green algal inhibition of diatom
growth: transition from mesotrophic to eutrophic com-
munity structure. Science 199:971.
-------
244 RESTORATION OF LAKES AND INLAND WATERS
Patmont, C. R. 1980. Phytoplankton and nutrient responses
to dilution in Moses Lake. M.S. thesis. University of
Washington, Seattle.
Strathmann, R. R. 1977. Estimating the organic content of
phytoplankton from cell volume or plasma volume. Limnol.
Oceanogr. 12:411.
Strickland, J. D., and T. R. Parsons. 1972. A practical
handbook of seawater analysis. Bull. Fish. Res. Board Can.
167.
Welch, E. B., and C. B. Patmont. 1980. Lake restoration by
dilution: Moses Lake, Wash. Water Res. 14:1377.
Wurtsbaugh, W. A., A. J. Home, and S. R. Vasak. 1977. Iron in
a eutrophic lake: its importance for algal growth and
nitrogen fixation. Presented at the 58th Annu. Meet. Pac.
Div. AAAS, San Francisco.
-------
245
MANAGING AQUATIC PLANTS WITH
FIBERGLAS SCREENS
MICHAEL A. PERKINS
Department of Civil Engineering
Water and Air Resources Division
University of Washington
Seattle, Washington
ABSTRACT
Vinyl coated Fiborglas mesh screening materials have been used to control nuisance growth of
Eurasian watermilfoil (Myriophyllum spicatum L) in selected areas of Lake Washington over the
past two growing seasons (1978, 1979). The screens were immediately effective in providing a
plant-free water column and greatly retarded regrowth after removal. Two months' coverage in
early spring reduced plant biomass by approximately 80 percent throughout the summer. Areas of
potential use, methods of application, ecosystem impacts, and economics are discussed.
INTRODUCTION
Lake restoration is often viewed in the context of its
relationship to the process of lake eutrophication.
Many measures which have been used to mitigate
adverse impacts associated with nutrient enrichment
of lakes have attempted to alter nutrient loading
characteristics. The premise is that longer term
benefits should derive by addressing the causes of an
observed impact rather than the consequences of that
impact. While the rationale is clear, it may not, in all
cases, constitute the most practical approach. This
would seem particularly true in regard to higher
aquatic plants. Perceived nuisance conditions associ-
ated with the growth of higher aquatic plants are often
related to the introduction and proliferation of exotic
plant species and may be localized within discrete
areas of a lake. The identification and treatment of
distinct causes of such growth may be difficult and
perhaps impossible once the plant has become
established. Whole lake manipulation to control a
particularly troublesome plant may not be necessarily
warranted and could have ramifications for the whole
system, beyond the initial intent of the manipulator.
A case in point would be the relatively recent
infestation of Eurasian watermilfoil (Myriophyllum
spicatum L.) within selected bays and nearshore areas
of Lake Washington. This lake, located in metropolitan
Seattle, Wash., represents a classic example of
successfully using sewage diversion as a restoration
methodology (Edmondson, 1978). Apparently, the
presence of milfoil may not be related to nutrient
conditions within Lake Washington. While further
nutrient input reduction schemes have been suggest-
ed, they are not considered to be of practical
significance in preventing or controlling the growth and
spread of milfoil. Current restoration efforts within
Lake Washington are now concentrated on the
cosmetic approach of removing the localized nuisance.
A variety of management techniques has been
applied to control excessive aquatic plant growth and
restore beneficial use within impacted waters. Chemi-
cal control techniques using herbicides such as
endothall, diquat, and various formulations of 2,4
dichlorophenoxyacetic acid have largely dominated the
aquatic plant management field. As more attention has
been directed toward the potential detrimental impacts
associated with using chemical control techniques,
interest has increased in developing nonchemical
alternatives. Mechanical techniques such as harvest-
ing have provided an attractive alternative in those
areas where the use of chemicals is limited by either
label restrictions or the philosophical attitudes of user
groups. Harvesting, while offering potential benefits
beyond the simple removal of nuisance plant growth
(Carpenter and Adams, 1977,1978) is limited by water
depth, site accessibility, and the requirement for
multiple treatments during a single growing season. In
some instances, harvesting may also aggravate a
situation in that the process may generate an increased
number of viable plant fragments which may spread
the particular target plant (Kimble, 1980). Harvesting
can also lead to the accumulation of a considerable
mass of aquatic plant tissue, causing a disposal or
reuse problem.
Using bottom barriers as a mechanical means of
aquatic plant control has derived largely from work
with black polyethylene sheeting materials (Born, et al.
1973; Nichols, 1974). Additional studies have been
conducted with a variety of bottom-covering materials
and sand/gravel blankets (Nichols, 1974; Armour, et
al. 1979; Cooke and Gorman, 1980; Engle, pers.
comm.). The results of these applications have varied.
In general, where successfully applied, bottom cover-
ing would appear to be an effective technique for
reducing nuisance conditions associated with the
presence of dense aquatic plant growth, at least in the
short term.
-------
246
RESTORATION OF LAKES AND INLAND WATERS
Problems related to the use of bottom coverings
derive from the nature of the sheeting materials used
and the modes of application. Most materials described
in the literature have relatively low specific gravities
which make them bouyant. This bouyancy hinders the
application process and renders the material suscep-
tible to lifting by wave action once in place. Even when
securely anchored, sheeting materials such as poly-
ethylene tend to be easily torn and dislodged (Armour,
et al. 1979). Additionally, unless perforated, most
sheeting materials tend to trap gases produced as a
result of benthic decomposition which also leads to
lifting of the materials from the bottom.
Sheeting materials, whether used by themselves or
in conjunction with sand/gravel blankets are usually
installed permanently. The accumulation of sediments
and detritus on the surface of bottom coverings can be
rapid. As this accumulated material constitutes a
substrate for continued aquatic plant growth, the
effective period of treatment can be greatly reduced.
The impacts of these coverings upon benthic inverte-
brate communities are largely undefined.
The use of Fiberglas screens as a bottom covering
was first reported by Mayer (1978). Mayer's work in
Chautauqua Lake, N.Y., indicated that many of the
problems associated with bottom covering materials
were circumvented with the screens. The screen, a
negatively bouyant permeable barrier, was highly
effective and could be temporarily placed. Larger scale
applications and more detailed studies in regard to
efficacy, timing and duration of placement, and
ecological impacts were conducted in Lake Washington
(Perkins, et al. 1979, 1980; Perkins, 1980; Boston,
1980).
LAKE WASHINGTON STUDIES
The screening material used in Lake Washington
was the same as that described by Mayer (1978): A
polyvinylchloride-coated Fiberglas mesh having 64
apertures per cm2, each aperture measuring 1 mm2,
and a specific gravity of 2.54 (known commercially as
Aquascreen). The screens were built to 9 * 24 meters
and equipped with grommets at approximately 2 meter
intervals along the edges. Concrete reinforcing bar
stakes were placed through the grommet holes to
secure the panels to the lake bottom.
Treatment plots were delineated within a plant
infested embayment at the outlet of Lake Washington
(Union Bay). Eight panels were placed in July 1978, the
treatment variables being duration of placement (over
winter, 1, 2, and 3 months of coverage) and water
depth (0 to 2 and 2 to 3 meters). The 1979 work
involved the application of three panels in April and
three panels in June. Duration of coverage for both the
April and June applications were 1, 2, and 3 months for
the individual panels. One further application in 1979
involved recovering one-half of two of the 1978
treatment plots. Scuba divers placed all installations.
The effectiveness of Aquascreen was evaluated by
following variations in mean dry weight biomass in
both treatment and control plots. Estimates of mean dry
weight biomass were obtained by randomly selecting
five samples from each treatment and control plot at
monthly intervals following screen removal. Samples
were taken by scuba diver using a cylindrical sampler
having a cross-sectional area of 0.25 m2. Plant samples
were washed free of debris, sorted by species, and
oven dried at 60°C for 48 hours prior to weighing.
The results of both the 1978 and 1979 samplings
indicated that the screens were highly effective for
removing nuisance conditions associated with aquatic
plant growth, maintaining a plant-free water column
for the duration of placement, and significantly
reducing regrowth after panel removal (Perkins, 1980;
Boston, 1980). Two to three months of coverage
resulted in biomass reductions ranging from 78
percent to virtual elimination of all aquatic plants. One
month of coverage provided plant control only for the
period of time during which the screen was in place.
The results from both years of screening are
summarized in Figure 1. Placement of panels in the
early spring was more advantageous in that the
installation was facilitated by the less dense plant mass
occurring at that time and a longer term and more
effective biomass reduction was obtained. The results
would indicate that screens placed for 2 or 3 months
during the period April to June could be removed and
a. 0
e
I976
AQUASCREEN
SHALLOW PLOTS
E 200
o
m
A M J
1979
APRIL APPLICATION
Z
<
_l
Q_
_J
^ 200 ,— 1979
O JUNE APPLICATION
Figure 1. — Variation in dry weight plant biomass within
control and screened plots over the 1978 and 1979 growing
seasons. Vertical lines represent a 2 SE deviation about the
mean for n = 5 replicates.
-------
SPECIAL PROJECTS AND TOPICS FOR ADDRESSING THE TROPHIC STATE
247
transferred to another plant infested area, hence
doubling the effective area of treatment with a single
piece of material.
The release of inorganic nutrients and oxygen-
demanding organics as a result of plant decomposition
beneath the screens was a potentially significant
impact associated with their use. A variety of
techniques including field observation and monitoring,
field enclosures, and laboratory growth tanks was used
to evaluate these factors.
A monthly monitoring of water quality characteristics
(dissolved oxygen, pH, alkalinity, dissolved organic
carbon, total phosphorus, molybdate reactive phos-
phorus, total nitrogen, ammonia, nitrate plus nitrite,
various cations, and chlorophyll a) both within and
outside of the screen treatment area failed to indicate
any impact of screening upon the water column. The
method of assessment, however, was undoubtedly
insensitive to such changes if they occurred, as the
area was open and subject to fairly rapid water
exchange (Perkins, et al. 1980).
To circumvent dilution effects associated with water
exchange within the treatment area, we employed 1
square meter polyethylene field enclosures during the
1979 growing season. The enclosures were establish-
ed in August and monitored at weekly intervals through
October.
One and two weeks after the experiment began,
dissolved oxygen levels in the screened enclosures
were significantly reduced relative to control enclo-
sures. The mean values (±1 S.E. for n = 3 replicates) for
the screened enclosures were 3.1 ±0.65 and 2.1 ±
1.70 mg 02 • liter "1 for the first and second weeks,
respectively. In two of the enclosures, screen place-
ment was such that plant materials were completely
compressed into the sediment; complete compression
was not achieved in the third enclosure. In those two
enclosures where plant compression was complete,
dissolved oxygen rapidly decreased to anoxic condi-
tions (0.0 and 0.9 mg Oz liter~1after 2 weeks) while the
third enclosure did not differ significantly from the
controls.
Concomitant with plant decomposition, one might
also anticipate increases in the concentration of
dissolved organic carbon (DOC) and soluble inorganic
nutrients. With the exception of molybdate reactive
phosphorus (MRP), such increases were not observed.
Concentrations of DOC and soluble inorganic nitrogen
in the screened enclosures were not significantly
different than the controls. The average concentration
of MRP increased by approximately 60 percent relative
to the control within 2 weeks. The increase was caused
by high concentrations in the two enclosures which
went anoxic; the third again did not differ significantly
from the controls.
The results of the enclosure studies over the first 2
weeks suggest that when sediment contact is achieved
in the installation, plant decomposition is rapid and
localized at the sediment-water interface. Presumably,
both organic and inorganic materials become entrained
within the sediments, creating an increased sediment
oxygen demand. Release of inorganic phosphorus to
the water column would appear to depend upon
whether or not anoxic conditions develop but may also
involve the phosphorus retention capacity of the
sediments (Boston, 1980).
After the second week, conditions within the
screened enclosures rapidly improved. The results of
the third week's sampling indicated no significant
differences in the various parameters measured. Algal
biomass (chlorophyll a) increased significantly by the
fifth week of sampling and remained at levels higher
than the control for the duration of the experiment.
These increases occurred in late September and may
have been related to increased light availability
because the plant canopy was removed.
An assessment of impacts upon the benthic
invertebrate community was limited by resources and
time. As a result, our efforts were limited to an
evaluation of mean densities and composition within
treatment plots covered for a period of 3 months only.
Samples for identification and enumeration were taken
by coring techniques and standard Ekman grab. The
results indicated that 3 months of cover with screen
had no significant influence upon either mean density
(numbers per square meter) or composition of the
invertebrate community (Perkins, 1980).
CONCLUSION
Overly dense growth of aquatic plants in lakes
creates problems not only for recreational users and
private waterfront owners but may also have ramifica-
tions for whole ecosystem functioning. This may range
from increased rates of sediment accumulation to
adverse effects upon the food chain. That aquatic plant
management is within the realm of lake restoration
seems evident and it is also clear that, whatever
technique is applied, it should maintain plant com-
munities intact. This would imply using techniques
which could be limited to well-defined areas of need
rather than indiscriminant broad scale treatment.
Fiberglas screens would appear to have considerable
merit in terms of reducing the nuisance characteristics
of aquatic plant growth in a manner commensurate
with maintaining the ecological integrity of the system
being treated. The screen, if properly applied, can be
among the most effective of aquatic plant management
techniques. Control can be placed precisely where it is
desired, there are no plant disposal problems to
contend with, and there need be no adverse impacts
associated with plant decomposition. As a nonselective
method of control, screen should be used only where
complete elimination of plants is the desired end. From
this standpoint alone, a limited use strategy would
seem most appropriate.
Further, the initial high costs (in excess of
$12,000/acre installed) would also argue in favor of
the judicious use of screen. By limited use strategy, we
are referring to localized nearshore areas where
complete control is the desired end and the area of
treatment need not be acres. For individual property
owners requiring plant control around docks for boat
moorage, swimming and diving, or for use on beach
areas, the screen seems ideally suited. Season-long
maintenance of a 1,000 square foot area with screen
could be more effective and accomplished at com-
parable expense to other management techniques. We
-------
248 RESTORATION OF LAKES AND INLAND WATERS
have estimated the annual costs for 1,000 square feet
of treatment, when allocated over a 5-year life
expectancy for the material, to range from $110 to
$140.
REFERENCES
Armour, G. D., D. Brown, and K. Marsden. 1979. Studies on
aquatic macrophytes. Part XV. An evaluation of bottom
barriers for control of Eurasian watermilfoil in British
Columbia. Water Invest. Branch, Ministry Environ., Province
of British Columbia, Victoria.
Born, S. M., et al. 1973. Restoring the recreational potential
of small impoundments. The Marion Millpond experience.
Tech. Bull. 71. Dep. Nat. Resour., Madison, Wis.
Boston, H. L. 1980. Factors related to the effectiveness and
environmental impacts of Fiberglas screens used for the
control of aquatic plants. Master's Thesis. Dep. Civil Eng.
University of Washington, Seattle.
Carpenter, S. R., and M. S. Adams. 1977. The macrophyte
tissue nutrient pool of a hardwater eutrophic lake:
Implications for macrophyte harvesting. Aquat. Bot. 3:239.
1978. Macrophyte control by harvesting and
herbicides: Implications for phosphorus cycling in Lake
Wingra. Jour. Aquat. Plant Manage. 16:20.
Cooke, G. D., and M. E. Gorman. 1980. Effectiveness of Du
Pont Typar sheeting in controlling macrophyte regrowth
after overwinter drawdown. Water Res. Bull. (In press).
Edmondson W. T. 1978. Trophic equilibrium of Lake
Washington. Ecol. Res. Ser. EPA-600/3-77-087. U.S.
Environ. Prot. Agency.
Engle, S. Personal communication. Wis. Dep. Nat. Resour.
Kimbel, J. C. 1980. Factors influencing potential intralake
colonization by Myriophyllum spicatum L. and the implica-
tion for mechanical harvesting. Master's Thesis. University
of Wisconsin, Madison.
Mayer, R. J. 1978. Aquatic weed management by benthic
semi-barriers. Jour. Aquat. Plant Manage. 16:31.
Nichols, S. A. 1974. Mechanical and habitat manipulation for
aquatic plant management. A review of techniques. Tech.
Bull. 77. Dep. Nat. Resour. Madison, Wis.
Perkins, M. A. 1980. Evaluation of selected non-chemical
alternatives for aquatic plant management. Municipality of
Metropolitan Seattle, Wash.
Perkins, M. A., H. L. Boston, and E. F. Curren. 1979.
Aquascreen, a bottom covering option for aquatic plant
management. In Breck, Prentki, and Loucks, eds. Aquatic
plants, lake management, and ecosystem consequences of
lake harvesting. Inst. Environ. Stud. University of Wiscon-
sin, Madison.
1980 Use of Fiberglas screens for control of
Eurasian watermilfoil. Jour. Aquat. Plant Manage. 18:13.
-------
249
RELATIONSHIPS BETWEEN AGRICULTURAL PRACTICES
AND RECEIVING WATER QUALITY
FRANK J. HUMENIK
Biological and Agricultural Engineering
North Carolina State University
Raleigh, North Carolina
ABSTRACT
Results from several studies will be summarized to more clearly define water quality in typical
agricultural areas in relationship to forested or background areas and relationships between
agricultural practices and water quality. Studies to be reviewed include a 3-year EPA project
analyzing samples from the forested and agricultural piedmont plus well and poorly-drained
coastal plain in the Chowan River basin to determine the nature of rural runoff on an areawide
basis. Subsequently, a more detailed evaluation of the relationships between agricultural practices
and water quality was conducted at four sites in the Chowan River basin, two of which had been in
the previous EPA study. The major goal of this continuing study is to determine water quality
changes resulting from the best management practices to control agricultural nonpoint sources.
As part of the statewide 208 agricultural planning process, three small agricultural watersheds in
the coastal plain and piedmont were evaluated to determine relationships between management
practices and water quality. Water quality differences between regions and individual watersheds
within regions have been analyzed in light of agricultural practices determined from detailed
producer surveys within the. study watersheds. The study showed that the only significant
difference in agricultural activities for watersheds with increased concentrations of nitrogen and
phosphorus was greater animal production. Major goals of continuing studies are to document
relationships between agricultural practices and receiving water quality and subsequent quality
changes.
INTRODUCTION
Assessment and regulation of sources impacting
water quality over an entire river basin are complex
problems. There are so many rural nonpoint source
inputs throughout a drainage basin that complete
spatial and temporal evaluation of associated water
quality impacts is impractical. Additionally, a basic
principle commonly overlooked in areawide water
quality assessment is that stream flow from un-
disturbed lands provides nutrients essential for
productive aquatic ecosystems. Therefore, it seems
necessary to establish areawide natural or background
water quality levels as a basis for assessing the impact
of increased inputs from human activities. The
tremendous complexity of direct cause and effect
relationships between agricultural practices and water
quality on an areawide basis make identification of the
nature and extent of nonpoint sources very difficult.
These dilemmas in no way lessen the importance of
water quality planning and evaluation activities to date,
but merely emphasize the need to build upon these
initial activities to obtain a sound data base for
evaluating and directing nonpoint source or rural
watershed pollution control programs.
SAMPLING
Historically, judgmental sampling has dominated
water quality investigations and evaluations of re-
lationships between agricultural practices and receiv-
ing water quality. With judgmental sampling, sites and
times are professionally selected as typical of the
activities and conditions being studied. This process is
characterized by subjective judgment and in this fact
lies both the strength and the weakness of this method.
The advantages of judgment sampling are that the
investigator can select sites and times according to his
experience as those best for overall program needs,
specific technical requirements, and best use of
sampling resources. Disadvantages include the in-
troduction of personal bias in the selection process, the
difficulty of determining a sampling error, the fact that
the selection process is not reproducible by another
investigator because of the personal element, and the
lack of a defined universe from which to extrapolate
results.
Probability or random sampling has not been used
much in studying water quality. Here, after rigorously
defining the project scope and specific sampling
details, the study sites and sampling times are selected
from the total universe being considered. Often the
total project may be subdivided into smaller, nonover-
lapping but all-inclusive parts which can be sampled
independently.
Probability sampling has- been widely applied to
many scientific and social problems. Its advantages are
that unbiased estimates may be derived by repro-
ducible methods; inference may be made from the
sampling results to the defined universe; a statistical
sampling error can be estimated; and with certain
assumptions about statistical distribution, confidence
limits may be set about the estimates. The ability to
-------
250
RESTORATION OF LAKES AND INLAND WATERS
estimate sampling error provides a rational basis for
considering the optimal allocation of sampling effort.
Disadvantages of probability sampling are that
without effective stratification, too much of the field
effort may be spent at the sites and times, which do not
result in maximum efficiency for achieving study goals.
Some randomly selected sites may seem inappropriate
or nonrepresentative but yet every site is unique in
some way. Additionally, some randomly selected sites
cannot be sampled but because of this should not have
been included in the sampling frame.
Either automated or grab methods may be used to
obtain water samples and measure flow. Judgmental
or probability sampling may be used to direct these
sa.mpling strategies. An automated installation is costly
but it can provide a continuous record of stream flow
and a programmed series of water samples. Grab
sampling costs less per site and is therefore a more
flexible method when covering many sites, but it
provides information about flow and water quality only
for the time of the sampling visit.
STUDY ON SAMPLING TECHNIQUES
AND AREAWIDE WATER QUALITY
A 3-year EPA project entitled, "Probability Sampling
to Measure Pollution from Rural Land Runoff,"
(Humenik, et. al. 1980) investigated the feasibility of
using probability sampling in describing rural water
quality not affected by point sources on an areawide
basis. The study also examined the substantive results
of this sampling effort for greater insight into
relationships between land activities and receiving
water quality.
The Chowan River Basin which was selected as the
study area is about 209 kilometers long and drains an
area of 12,802 km2 in southeastern Virginia and
northeastern North Carolina (Figure 1). The upper
5,180 km2 of the basin lie m the gently rolling hills of
Virginia's piedmont plateau. The remaining 7,700 km2
Geographic location of
Chowan River Basin
A-Agricultural Piedmont
F-ForMted PMmont
W-Well-Drained Coaital Plain
P-Poorly-Drained Coastal Plain
0 10 20 30 MILES
SCALE
Figure 1. — Study basin location and delineation of sampling
areas.
lie in the flat coastal plain of Virginia and North
Carolina where the soils can be broadly classified as
either poorly drained or well-drained sands. Within the
piedmont there was no logical reason for stratification
by soil type because nearly all the soils are loamy.
However, because water quality differences are likely
Table 1 — Land use in Chowan river study subbasm.
Subbasin
Drainage area
Forest
Crop
Pasture
Developed
sq
sq km
Logged
Ponds
Poorly-Drained Coastal Plain
P-8
P-10
P-11
P-13
4.51
3.74
490
38.04
11.68
9.69
12.69
98.52
77 1
72.1
69.0
66.2
17.4
25.9
23.0
26.4
Well-Drained
W-3
W-4
W-8
W-10
F-1
F-2
F-3
F-7
6.27
0.20
3.31
6.37
5.21
6 14
14.06
6.04
16.24
0.52
8.57
16.50
13.49
15.90
36.42
15.64
44.7
52.6
56.2
48.5
72.1
91.9
90.3
82.8
53.5
46.2
41.9
43.3
Forested
15.0
3.7
7.0
10.6
4.2
2.0
3.0
4.7
Coastal Plain
1.3
1.2
0.4
7.0
Piedmont
8.1
2.8
1.0
3.0
1.0
0.0
4.0
1.5
0.3
0.0
1.5
0.9
0.0
0.2
0.1
2.9
0.3
0.0
1.0
1.1
0.2
0.0
0.0
0.3
4.8
1.4
1.6
0.7
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Agricultural Piedmont
A-1
A-4
A- 8
5.57
4.28
1.75
14.43
11.09
4.53
63.6
55.6
38.1
27.0
29.9
20.6
7.3
14.3
32.5
0.2
0.2
1.0
1.7
0.0
7.6
0.2
0.0
0.2
-------
RURAL WATERSHED POLLUTION CONTROL
251
to result from different land uses, a forested land was
selected to represent background conditions and an
agricultural area was selected to measure land-use
impact. Thus the four study areas were forested
piedmont, agricultural piedmont, poorly-drained coast-
al plain, and well-drained coastal plain. The areas
designated for sample site selection were from a
restricted part of the watershed to allow more
convenient sample retrieval; they represented about 25
percent of the total watershed area.
The sampling site was defined as a point located at
the first railroad or road crossing below the confluence
of two first-order streams on a U.S. Geological Survey
,1:250,000 map above which there were no sources
that would require discharge permits. All such sites
(above 90) that met the definition within the study
areas were identified and comprised the sampling
universe. Then by random selection four sites were
chosen from three of the areas and three sites from the
fourth area for a total of 15 sampling sites. The
drainage area of these sites ranged from 0.5 to 100 km2
and the general land use ranged from 40 to 90 percent
forested (Table 1). Such stream channels were from 1.5
to 15 meters wide and at base flow from 0.15 to 0.60
meters deep.
Two years (November 1974-November 1976) of data
were collected at the forested piedmont, poorly-drained
coastal plain, and well-drained coastal plain sites, and
18 months (June 1975-November 1976) of data were
collected at the agricultural piedmont site. The grab
sampling plan involved time stratification to ensure
that measurements were obtained at a uniform rate
throughout the study. Basically, the stratification was
such that each stream was monitored 26 times per
year at the rate of two visits (chosen by a restricted
random sampling) per 28-day period. During each grab
site visit the flow rate was measured by standard USGS
procedures and a depth-integrated water sample was
obtained manually at the midpoint of the stream.
Automated sampling systems were established at 5
of the 15 statistically selected grab sampling sites.
Stream stage was recorded continuously by analog
recorders and the automated samplers had the
capability of collecting 28 discrete 500-ml water
samples. The sampler was activated by stage change
with a subsample to be taken at each 76 mm rise or fall
in stream stage. The time each subsample was
obtained was recorded on a stage strip chart by a relay
activated pen so that each sample bottle was assigned
the mid-point time between samplings for mass
transport analyses.
COMPARISON OF GRAB AND
INSTRUMENTED SAMPLING DATA
Many water quality assessment agencies have
limited monetary resources so that the grab sample
approach is often used to determine regional water
quality. The sample frequency varies, but typically the
sample design is such that the stream is periodically
instead of randomly sampled. The 22-month data base
from one poorly drained coastal plain site that had both
grab and instrumented sampling was employed to
compare the results of such sampling on flow and
concentration estimates. For illustrative purposes the
value obtained from the instrumented sampler for the
22-month period was also plotted at the 24-sample per
year frequency .(Figures 2 through 6). As anticipated,
the range of grab sampling values generally increased
with decreasing sample frequency. This trend is not
surprising and can be predicted by sampling theory.
Additionally, in some cases the high frequency grab
sampling value is very similar to the instrumented
value, and in other cases markedly different. Realizing
such parameter variation with space, time, and
sampling costs as a function of the number of sites and
visits to each site is most important in developing a
technically sound monitoring scheme within given
budget requirements.
I I
FHCOUf MCV tAWLt I/YEAR
Figure 2. — Range of Mean Flow Velocity as a function of Grab
Sample Frequency.
ci
mo/l
A-Autom«Kl EitlmlU
l-Qrab Eitimitt Ringt
A
X
23 4 6 8 8 12
FREQUENCY SAMPLES/YEAR
Figure 3. — Range of Mean Chloride Concentration as a
function of Grab Sample Frequency.
Data from the five sites which were both in-
strumented and grab sampled were compared for the
total 22-month sampling period. Constituent mean
values -~t all sites indicated that differences existed (p<
0.10) between chemical oxygen demand (COD), total
organic carbon (TOC), total phosphorus (TP), total
Kjeldahl nitrogen (TKN), chloride(cr)concentrations
measured by grab and automated samples while there
was no evidence of ^differencejp< 0.10) for nitrate
(NOs) concentrations. The grab mean concentrations
were less than the automated mean concentrations by
-------
252
RESTORATION OF LAKES AND INLAND WATERS
TP04
mg/l
0.7 -
0.1
0.0
A-Automited Estimate
l~Qrab EitlmitB Range
I I
23456 9 12
FREQUENCY SAMPLES/YEAR
Figure 4. — Range of Volume Average Mean Total Phosphorus
Concentration as a function of Grab Sample Frequency.
TKN
mg/l
A-Automated Estimate
l-Grab Estimate Range
234 5 6
9 12
FREQUENCY SAMPLES/YEAR
Figure 5. — Range of Mean Volume Average Total Kjeldahl
Nitrogen Concentration as a function of Grab Sample
Frequency.
0.6
0.4
A-AutomatadEttimate
l-Grab Estimate Range
9 12
FREQUENCY SAMPLES/YEAR
Figure 6. — Range of Mean Volume Average Nitrate
Concentration as a function of Grab Sample Frequency.
36, 39, 28, 19, and 8 percent for COD, TOC, TP, TKN,
and Cl, respectively.
Relatively large differences were observed between
grab and instrumented estimates of annual volume
average concentration for a given site. Because the
grab to automated annual volume average concentra-
tion ratio varied among constituents at a site, no
consistent factor related the two estimates. Further, for
each water quality constituent the grab to automated
concentration ratio varied among the sites; thus a
factor relating the two estimates which was in-
dependent of site did not exist. Statistical testing
indicated (P>0.10) that annual volume average grab
concentration estimates of COD, TOC, and TP which
were about 50 percent less than automated sampling
values were statistically significant. For TKN, NO-iand
Cfgrab sampling was less by an average 13 percent,
not significant statistically.
The annual grab and automated water yield
estimates obtained at different sampling sites resulted
in relatively large differences between paired samples.
However, statistical testing for paired samples provided
no evidence of a statistically significant difference
between estimates by the two methods.
Conclusions
Statistical sampling methods can be used to measure
the mean areawide contribution of chemical species
from rural nonpoint sources as an alternative to the
more difficult and often impractical complete monitor-
ing approach. Grab and instrumented sampling are two
common methods of assessing stream water quality
which can be employed in a statistically designed
sampling program. Although differences were found
for grab and instrumented sampling estimates of
concentration and a flow resultant data, analyses from
both sampling methods supported the same general
conclusions.
Annual volume average concentration and water
yield estimates were obtained at five sites by routine
operation of both stage activated, instrumented
samplers with stage recorders and simple time
stratified grab sampling. Results indicated that about
50 percent lower COD, TOC, and TP estimates were
obtained by the grab sampling method. Although rather
large differences were observed for the water yield
estimates, the data did not indicate any statistically
significant difference between the two methods. Due to
the confounding factors associated with the two
sampling methodologies, it was not possible to
completely define- reasons for the differences.
Comparing annual water yield estimates from the 15
statistical survey sites to historic values for the study
region verified that simple time-stratified grab sam-
pling provided reasonable estimates of areawide
annual water yield; however, the precision of the
individual site estimates was low.
RELATIONSHIPS BETWEEN LAND USE
AND WATER QUALITY
Prediction of present and future rural water quality is
often based upon models which relate water quality to
macro land-use factors. These models are probably
-------
RURAL WATERSHED POLLUTION CONTROL
253
best supported by a national water quality land-use
study of 928 sites reported by Omernik (1977). That
study found nitrogen and phosphorus concentrations
increased with increased agricultural intensity. Water
quality land-use relationships for the 15 statistical
survey sites were investigated to determine if similar
relationships were evident in this watershed. These
analyses considered both in-stream conditions and
exported water quality as a function of percent forested
area. The land-use summary presented in Table 1 for
these subbasins showed that the forested area, plus
agriculture area, was approximately equal to the total
subbasin area at each site. Thus in these subbasins
decreased forested area indicated increased human
activity which was primarily agriculture. Therefore,
major trends which were observed for water quality
constituents with respect to percent forested area
would also exist (only inversely) with respect to percent
agriculture.
Site water quality versus land-use relationships for
COD, TP, TKN, and NOa-N concentrations during June
1975-November 1976 are presented in Figure 7. For
each of the four parameters, the grab sampling mean, a
mean plus or minus one standard deviation, maximum,
and minimum values for each site are plotted as a
function of percent forested area. From a water quality
perspective, the COD, TP and TKN vs land-use graphs
show no meaningful mean concentration increase as
the percent forested area decreased. Indeed, even the
range of values for these parameters was relatively
constant for all sites. The mean NOa-N concentration
also does not display a concentration increase with
decreased forested area, but the values vary con-
siderbly from site to site, indicating that the sites were
probably differentially impacted by the general land-
use activities. The major conclusion obtained from
examining these graphs was that the mean time
average concentration was relatively uniform throug-
hout the watershed and did not increase as agricultural
activity increased.
I he effects of land use on receiver system water
quality was assessed by analyzing flow weighted
concentrations of COD, TP, TKN, and NOa-N versus
percent forested area for the 15 statistical survey sites
during June 1975-November 1976. These data are
liki
I.
"
11
[
J.»
i
I
A
J ,.i .II. .
roue i TED ARIA «
displayed in Figure 8. The flow weighted concentra-
tions do not present any clear relationship between
water quality and percent forested area so regression
models were not attempted.
Nutrient levels in streams usua'ly increase as
agricultural intensity increases, but of ton a wide range
of confidence limits exists for regression models
developed for large geographic areas. For example, one
regression model (Omernik, 1977) relating mean total
phosphorus concentration to percent agriculture plus
urban area for the eastern United States had broad
confidence limits as measured by a ratio for the plus or
minus 1 sigma range to predicted mean value of 130
percent. The Chowan concentration versus land-use
graphs point out the need for caution when employing
model predictions to specific cases. Regression models
employing macro land-use factors do not account for
varying agricultural cropping patterns and manage-
ment practices, annual weather conditions, stream
border buffer systems, or other factors which can
impact agricultural effects on water quality.
Figure 7. — Site arithmetic data summary versus land use for
grab sampling (June 1975 to November 1976).
Figure 8. — Site flow weighted concentrations versus land use
for grab sampling (June 1975 to November 1976).
Conclusions
While direct water quality land-use relationships for
the Chowan data were weak, differences related to
geographic area, season, and size were observed.
Evaluation of the water quality data obtained during
this study led to the following conclusions concerning
nature of rural runoff on an areawide basis as
developed and summarized in the project report
(Humenik, et al. 1980).
1. Neither in-stream (arithmetic average) nor net
export (flow weighted) concentration data presented
any clear relationships between water quality and
macro land-use factors as measured by percent of
forested land. This result points out the need for
caution when applying model predictions to specific
cases. Macro land-use factors do not account for
varying agricultural cropping and management prac-
tices, annual weather conditions, stream border buffer
systems, or other factors which can minimize the
impact of agricultural activities on water quality.
2. The comparison of geoclimatic areas demon-
strated that the dominant variation was between the
piedmont and coastal plain with only minor variations
occurring within these two physiographic regions. The
-------
254
RESTORATION OF LAKES AND INLAND WATERS
differences between the piedmont and coastal plain
were judged to result from naturally occurring
physiographic variations in (a) basin characteristics
such as vegetation, soil type, and stream hydraulics;
and (b) ocean proximity.
3. Average stream water quality for the four
sampling areas was relatively uniform. Quality was
generally good compared to proposed standards, but
elevated TP concentrations even in the forested
piedmont area demonstrated the need for basing water
quality assessments on measured local conditions,
especially background or natural levels. Stream sample
concentrations usually displayed large variations with
respect to mean values indicating that rural nonpoint
sources are highly variable in both space and time.
4. Analyses for seasonal trends indicated that water
yield and the associated nutrient yields were greater
during the winter and spring seasons than during the
summer and fall, reflecting rainfall and evapotranspi-
ration cycles. In analyzing seasonal flow weighted
average concentration data, the models generally
demonstrated significant relationships but had rather
low r2 values.
5. Measured concentrations did not display any
consistent functional relationship to flow (water yield)
levels. However, data showed that NO-j-N concentra-
tions were elevated during flow conditions at a small
(0.5 km2 or 0.2 mi2)site but not at a larger (20 km2 or 8
mi2)site but not at larger (20 km2 or 8 mi2) site in the
well-drained coastal plain with similar land use. The
NOs-N attenuation was judged to be the result of in-
stream dynamics.
6. The impact of channelizing coastal plain streams
was most pronounced with respect to high NOs-N
concentrations in the channelized streams as com-
pared to the unchannelized streams which have
natural swampy flood plains and channels that
increase NOa-N attenuation by denitrification and
biological uptake.
7. Assessment of point and nonpoint source impacts
in one small basin verified classic point source
concentration spikes with subsequent decline to
intermediate levels for all investigated constituents
except chloride and nitrate. Therefore, for the studied
stream reach, nitrogen and phosphorus inputs which
appeared to come from treatment plant effluents are
reduced to headwater background levels as long as
stream assimilatory capacity is not overwhelmed or
natural inputs change background levels.
8. Point-in-time comparisons between headwater
and downstream constituent concentrations showed
small differences on a water quality basis.
208 CRITICAL AREA STUDIES
Two areas of intensive agricultural production in
major North Carolina river basins were selected for
critical area studies under the statewide 208 water
quality planning and implementation program for
agriculture. Within both the coastal plain and piedmont
study area three predominantly agricultural water-
sheds between 1 3 to 26 sq. km were randomly selected
for intensive monitoring (Homey, et al. 1978).
The coastal plain study subbasins in the Neuse River
have intensive cropping (predominantly corn, tobacco
and soybeans) and considerable swine production. The
selected watersheds have slopes ranging from 1 to 6
percent with sandy loam and its associates.
The piedmont study area subbasins have steeper
topography and slate soils that increase the signifi-
cance of erosion and sedimentation. Slopes of the
study watersheds generally range from 4 to 10 percent
although some as steep as 18 to 20 percent were
found. Major crops are corn, sorghum, and soybeans.
Some land is in permanent pasture for beef production,
but swine and poultry production predominate.
Instrumented sampling stations were installed at the
lower boundary of each watershed. Rainfall was
measured by recording rain gage and a digital stage
recorder measured stream stage from which flows
were determined. Water samples were taken during
storm events by an automated sampler adapted to
sample across the runoff hydrograph (Koehler, et al.
1978). In the coastal plain study area a background
station had been established by USGS as representing
undisturbed forest lands with an area of 2 sq. km of
which 95 percent is forested, 5 percent is in
agriculture, and less than 1 percent is paved roads.
This site was grab sampled during storm events over
the study period.
Data Analysis
Concentration data from the three monitored
watersheds in each study area are combined in Table 2.
Data from the coastal plain background station are also
presented to allow comparison with storm grab
sampling data for such a relatively undisturbed area.
Data are included from State monitoring stations on
the river stems closest to the priority areas in the two
study watersheds. These main river stations were grab
sampled quarterly and therefore do not represent
runoff conditions.
Mean concentrations in the study streams draining
agricultural areas were higher than the background
and river stations. As expected, the background station
had the lowest mean constituent concentrations. Since
the watershed sampling program was designed to
measure water quality during runoff conditions, these
concentrations can be attributed mainly to rainfall
runoff transport. Study area runoff concentrations
were higher than the major receiver stream average
flow concentrations by a factor of 1.4 to 4 times in the
coastal plain and 5 to 10 in the piedmont.
Average data for all sites in a region showed mean
concentrations for all constituents were higher in the
piedmont than the coastal plain study areas (Table 2).
An analysis of variance was performed for nitrogen and
phosphorus data to determine whether the regional
differences in mean concentrations were significant.
While statistically high for all N and P forms, only 9 to
23 percent of the variance within the data was due to
the regional separation; therefore, there must have
been considerable variation between and within
individual watersheds over time.
Water quality data for each site collected from April
1978 to June 1979 in the coastal plain study areas and
July 1978 to June 1979 in the piedmont study areas
are summarized in Table 3. Median, maximum, and
minimum concentrations are included as well as the
-------
RURAL WATERSHED POLLUTION CONTROL
255
Table 2. — Mean constituent concentrations (mg/l) in study area, background, and river stations.
Coastal Plain -
Background
(n' = 23)
Parameter
Ammonia Nitrogen
(NH3-N)
Total Kjeldahl Nitrogen
(TKN)
Nitrite + Nitrate Nitrogen
(NO2~-NO3~-N)
Total Nitrogen
(TN)
Total Phosphorus
(TP)
Total Residue
(TR)
Total Nonfilterable
Residue (TNR)
Chemical Oxygen Demand
(COD)
Mean
mg/l
<0.05
0.18
0.13
0.30
<0.05
52.0
7.0
12.0
S.D.'
0
0.14
0.26
0.28
0.01
24.0
7.0
9.0
Coastal Plain -
Watersheds
(n = 314)
Mean
mg/l
0.21
0.90
0.93
1.83
0.32
232.0
39.0
46.0
S.D.
0.62
1.44
1.34
1.97
0.52
706.0
69.0
34.0
Coastal Plain -
River
(n = 7)
Mean
mg/l
0.05
0.56
0.70
NA*
0.22
NA
NA
26.0
S.D.
0.03
0.08
0.31
0.06
6.0
Piedmont -
Watersheds
(n = 261)
Mean
mg/l
0.64
2.36
2.27
4.63
0.51
901.0
617.0
63.0
S.D.
1.61
3.02
2.48
4.46
0.64
2005.0
1397.0
63.0
Piedmont -
River
(n = 12)
Mean
mg/l
0.07
0.31
0.62
NA
0.09
119.0
NA
9.0
S.D.
0.06
0.12
0.29
0.10
23.0
5.0
'S.D. - Standard Deviation
NA - Not Available
n - Number of Observations
Table 3. — General statistics for water quality parameter concentrations (mg/l) in each monitored watershed.
CP-1 (74 Observations)
CP-2 (100 Observations)
CP-3 (99 Observations)
Parameter
NH3-N
TKN
NOz-3-N
TN
TP
TR
TNR
COD
Median
•COS
0.40
0.65
1.07
0.06
70.0
12.0
38.0
Mean
0.05
0.46
0.71
1.18
0.07
449.0
13.0
40.0
S.D.
0.07
0.21
0.38
0.50
0.06
1179.0
11.0
13.0
Min
<.05
0.20
<.05
0.33
<.05
40.0
0.0
22.0
Max
0.46
1.70
1.50
2.60
0.39
4650.0
54.0
94.0
Median
0.16
0.60
1.50
2.50
0.26
93.0
22.0
43.0
Mean
0.48
1.51
1.68
3.19
0.58
121.0
47.0
58.0
S.D.
1.04
2.36
2.10
2.97
0.78
73.0
61.0
47.0
Min
.05
0.10
0.13
0.54
0.05
41.0
1.0
12.0
Max
7.70
19.0
22.0
22.3
4.80
429.0
369.0
280.0
Median
0.08
0.60
0.49
1.09
0.17
98.0
24.0
31.0
Mean
0.09
0.64
0.61
1.24
0.24
243.0
50.0
35.0
S.D.
0.07
0.24
0.57
0.52
0.22
736.0
92.0
20.0
Min
<.05
0.30
<.05
0.43
0.06
38.0
4.0
17.0
Max
0.37
1.40
1.90
2.40
1.40
5150.0
670.0
180.0
P-1 (80 Observations)
P-2 (81 Observations)
P-3 (66 Observations)
Parameter
NH3-N
TKN
NO2-3-N
TN
TP
TR
TNR
COD
Median
0.23
1.50
1.80
4.10
0.39
283.0
128.0
47.0
Mean S.D.
0.33 0.38
2.65 2.63
3.15 3.11
5.79 4.61
0.85 0.82
1669.0 2346.0
1326.0 2091.0
93.0 84.0
Mm
<.05
0.30
0.27
0.77
0.06
88.0
6.0
14.0
Max
1.80
15.0
18.00
21.2
3.00
9620.0
9330.0
290.0
Median
0.56
1.60
1.30
3.52
0.34
217.0
83.0
41.0
Mean
1.13
3.23
2.51
5.78
0.53
889.0
536.0
59.0
S.D. Min
2.05 <.05
3.84 0.40
2.40 0.20
5.13 0.98
0.62 0.05
2514.0 81.0
1172.0 5.0
58.0. 17.0
Max
15.0
18.0
11.0
28.0
4.0
21700.0
6810.0
380.0
Median
0.11
0.90
0.66
1.66
0.18
173.0
86.0
38.0
Mean
0.16
1.23
0.91
2.04
0.26
416.0
185.0
42.0
S.D.
0.17
0.79
0.78
1.42
0.20
707.0
275.0
20.0
Mm Max
<.05 0.85
0.20 4.60
<.05 3.80
0.34 8.40
•COS 0.97
60.0 4040.0
4.0 1730.0
1.0 110.0
standard deviation to indicate data distribution. Similar
mean and median values indicate a fairly even
distribution about the mean while a mean value
considerably higher than the median indicates a few
extreme values.
An analysis of variance was performed to determine
whether differences among watersheds within each
study area were significant. Differences were found to
be statistically high for the nitrogen and phosphorus
forms with a variance of 35 to 48 percent. This
indicates that water quality differences are a function
of the individual watershed as well as regional location.
Over 50 percent of the variance is not explained by this
classification indicating the highly variable nature of
small streams as reported from a previous study
(Humenik, et al. 1980).
Conclusions
General conclusions resulting from this study as
developed in the total project report (Homey, et al.
1979) are:
1. Streams draining predominantly agricultural wa-
tersheds have higher nitrogen and phosphorus runoff
concentrations than in undisturbed forested water-
sheds in the same area and long-term averages in
receiver river streams.
2. Concentration levels are generally higher in the
piedmont than in the coastal plain; however, variation
between individual watersheds within regions are as
important as regional differences.
3.The watershed in both the piedmont and coastal
plain study areas with the highest nitrogen and
phosphorus concentrations also has the highest level
-------
256 RESTORATION OF L^KES AND INLAND WATERS
of livestock production. Since no direct animal waste
point sources were found, many such waste inputs
would have been from nonpoint sources.
4. Excessive use of chemical fertilizers was found in
some fields in all areas, but crop management
practices were found to significantly differ from one
watershed to another and thus relationships to water
quality differences could not be determined on the
basis of cropping patterns.
Analysis of nitrogen forms indicated possible
differences in nitrogen transport related to observed
livestock production systems and waste management
techniques. Better understanding of these sources and
transport mechanisms should result from a more
complete analysis of production-waste management
practices and stream flow data.
SUMMARY
The relatively similar average concentrations from
sampling sites receiving rural nonpoint source inputs
from different land use and geoclimatic regions and in
main rivers draining these sites indicate that more data
on background conditions and relative impact of
nonpoint sources are needed before wide implementa-
tion of best management practices is required,
particularly in areas with heterogenous land use. It also
seems most important that agencies developing
regulatory criteria be responsive to ambient conditions
and not entertain standards requiring better than
background water quality particularly for relatively
undisturbed or pristine areas such as the forested
regions evaluated in these studies.
Tools must be developed to assess areawide
relationships between agricultural practices and re-
ceiving water quality. Efficient use of currently
available sampling and modeling techniques can
provide more cost-effective assessment of both
management practices to control agricultural nonpoint
sources and areawide water quality over time and
space for existing conditions and different planning
strategies. The effects of agricultural practices on the
water quality of both a stream reach and an area must
be defined to answer the important and difficult
questions concerning cost effectiveness and technical
feasibility of management practices to control agri-
cultural nonpoint sources and achieve areawide water
quality goals.
REFERENCES
Homey, L. F., et al. 1978. North Carolina 208 case study.
ASAE Paper 78-2584, Chicago.
Humenik, F. J., et al. 1980. Probability sampling to measure
pollution from rural land runoff. EPA 600/3-80-035,
Athens, Ga.
Koehler, F. A., F. J. Humenik, and E. P. Harris. 1978. Simple
sampler activation and recording system. Eng. Div. ASCE
Vol. 104, No. EE5.
Omernik, J. M. 1977. Nonpoint source-stream nutrient level
relationships, a nationwide study. EPA 600/3-77-105.
-------
257
SOURCE CONTROL OF ANIMAL WASTES FOR
LAKE WATERSHEDS
LYNN R. SHUYLER
Animal Production Section
Robert S. Kerr Environmental Research Laboratory
U.S. Environmental Protection Agency
Ada, Oklahoma
ABSTRACT
Controlling wastes from animal production facilities is conceptually well defined. There are many
components of an animal waste management system that have been proven with use. The trick to
designing a successful system for a specific location is to know what component will work in that
climate and under the management regime of the farm. Containing these wastes is basically an
engineering problem; the disposal or utilization becomes an agronomic problem, and the resultant
runoff or pollution from disposal is an environmental problem. All of these must be addressed in
the planning stage and not after the fact. Controlling animal wastes produced from a nonpoint
source or pasture production system becomes a more difficult problem. This type of control relates
almost totally to pasture management and livestock management with very little engineering
involved. Therefore, the solutions are termed best management practices (BMP) and are selected
from a set of management practices proposed for a region of the United States.
NATURE OF ENVIRONMENTAL
PROBLEM
Livestock and poultry industries are a significant part
of the U.S. economy. We currently have approximately
44 million beef cows, 11 million dairy cows, 10 million
beef feeders, 50 million swine, 16 million sheep, 400
million layers, 460 million broilers, and 46 million
turkeys.
Livestock and poultry industries annually produce
about 112 million tons of dry residue to be applied to
crop or pasture land. Contained in this residue are
about 4.1 million tons of nitrogen (N), 1.1 million tons
of phosphorus (P), and 2.4 million tons of potassium (K).
U.S. agriculture uses about 9.2 million tons of chemical
fertilizer N annually; therefore, it is easy to see the
commercial impact of animal wastes if it could all be
collected and used to replace so'me of the chemical
fertilizers. Further information can be obtained from a
USDA publication entitled "Estimating U.S. Livestock
and Poultry Manure and Nutrient Production."
VARIETY IN TYPES OF ANIMAL
PRODUCTION UNITS
In the beef cattle industry only one-fourth of the 54
million animals are in confined feedlots. The remainder
are on pasture or in some partial confinement system,
such as winter feeding of cows. About 95 percent of
those confined are in open dirt lots; of the 137,700
feedlots in operation, less than 20 percent are under
the restrictions of the effluent guidelines. Most of the
feedlots store manure in open piles for application to
land during the spring and/or fall. Only about 1 to 2
percent of the nearly 400,000 dairy farm units are
subject to discharge permits. About 25 percent of the
farms use milking centers and cow yards or pastures;
the other 75 percent use some type of confinement.
Most of the confinement systems use daily cleaning,
and the manure is quite often spread daily regardless of
weather conditions, i.e., frozen ground and/or snow
cover.
Hog production in the U.S. is about equally divided
into open dirt lot production and covered, paved units,
with most in units of less than 200 heads. The wastes
from many of the new housed production units are
liquid in nature and are stored in lagoons or pits for
later land application.
Poultry are nearly all produced in housed units,
turkeys being the exception with nearly 75 percent of
them on range or in open confinement pens. Most of
the collectable waste from housed units is in a solid
form, with only a few units using liquid manure
systems.
IMPACT ON WATER QUALITY
The effect of animal wastes on water quality has
been known for many years but has been dramatically
demonstrated during the past 20 years, when fish kills
were related to the runoff from feedlots entering the
streams. However, the relationship between land
application of manure and the water quality of a stream
receiving runoff from the application site has not been
fully established. The effect on chemical water quality
can be estimated by using dilution factors or by actual
stream quality measurements. Such stream water
quality measurements are expensive; few have been
made in the past, and only recently have large scale
studies been initiated. To actually study the nonpoint
source component of the stream quality in a large basin
would be very expensive. Moreover, effects on the
-------
258
RESTORATION OF LAKES AND INLAND WATERS
biological community within receiving waters are
unknown for the most part; this subject needs
considerable effort in the near future.
Annual pollution loads resulting from land applica-
tion of animal wastes change dramatically because of
the management practice selected. The magnitude of
such a change can be illustrated by using the entire
State of New York as an example. Assuming that New
York has approximately 900,000 dairy cows producing
enough manure to supply the nitrogen needs of
450,000 acres of cropland, the following two manage-
ment practices demonstrate the difference in annual
nitrogen loading to the waters of the State. If this waste
is applied daily without incorporation into the soil, the
total load of nitrogen to the waters of the State would
be approximately 3.6 million pounds. When the
assumed best management practice of proper applica-
tion timing and total incorporation of the waste is
evaluated, the nitrogen load to the waters would be
only 0.8 million pounds annually. This would be a
reduction of 2.6 million pounds of nitrogen annually,
which could have a very beneficial effect on the water
quality of the receiving streams and lakes. Estimates
and calculations of this nature can be made by using a
joint EPA-USDA document entitled "Animal Waste
Utilization on Cropland and Pastureland."
SOLUTION TO THE PROBLEM
The best way to solve any water pollution problem is
not to allow the pollutant to reach the water in the first
place. This has been the goal of the effluent guidelines
developed several years ago under P.L. 92-500 which
require zero discharge from confined animal pro-
duction facilities unless a certain storm magnitude and
frequency is exceeded. This regulation does not allow
the producer in the U.S. to discharge effluent from any
type of treatment system for animal wastes. The only
options open to the animal industry then are to use the
wastes in some manner or totally destroy them.
There are many ways to use animal wastes, in the
production of fuel or feed, or as a fertilizer. Fuel and
feed production require a technology not commonly
found on the average farm, and, therefore, are not
widely used today. Also, some of the processes used
will produce some type of effluent that must be
managed under the same zero discharge guidelines.
Putting animal waste back on the land as fertilizer
seems to be the onlv real answer to the problem.
Man has long recognized the beneficial effects of
animal manures on crop growth and soil condition.
Applying manure to the land completes the natural
cycle of growth, death, and decay on land where crops
are produced. The land contains legions of organisms
capable of decomposing organic wastes of plants and
animals into useful humus and the various elements
essential for continued crop production. The applica-
tion of animal wastes to the land sounds very simple,
easy, and environmentally sound, but as with many
other things, there are right and wrong ways to
accomplish this task. Improper use or application of
animal wastes to the land provides a potential pollution
hazard to both surface and ground water. Incorrectly
designed or managed manure collection and storage
systems may lead to direct water pollution and can
reduce the value of the animal wastes prior to delivery
to the field.
The term best management practice (BMP) is used in
the nonpoint source planning and implementation
programs. Many take this to mean that for animal
production there is a short list of practices that can be
used to solve the problem. This is not quite correct. In
fact, there are very long lists of components from which
to choose to build a complete animal waste manage-
ment system for a given set of conditions. However, the
conditions will vary from one region to another and
from one farm to another, even though the farms may
be located side by side. What we have for BMP's then is
a list of management system concepts which can be
selected for a given area, and then we have a long list
of ideas, components, and existing facilities which the
farm planner may select and modify to finally design a
true BMP for a given farm.
Many factors must be considered in selecting a BMP;
only a few will be mentioned here to give some insight
into the complexity of the system. Starting with the
production unit, the number of head on the farm must
be known to establish the daily volume of wastes to be
considered. The form of the waste must be known,
either liquid, slurry, or solid or some combination of
each. The structural needs for the system may include
runoff control ponds, manure storage areas, or manure
pits below production buildings.
If storage is for slurry or liquid and odors become a
problem, mechanical equipment may be added to
control the odors. The size of the storage will depend
upon number and size of animal and the length of
storage time. Storage time is dictated by how often the
land will be suitable for application due to weather and
cropping patterns.
The cost of each proposed management system must
also be evaluated, for it does no good to design a
system that cannot be implemented because of
economics. Many systems and their components are
evaluated in a document developed by Ohio State
University for EPA entitled "A Manual on Evaluation
and Economic Analysis of Livestock Waste Manage-
ment Systems." Computer models for sizing runoff
control systems and applying the runoff to cropland
have been developed by Kansas State University and
Oregon State University. Also, the Cooperative Ex-
tension Service in each State has many fact sheets and
other information that can be very useful to the
developer of BMP's for animal waste problems.
The problem posed by pasturing of animals is
different in that it is truly a nonpoint source problem,
and the approach must be different from that used for
confined animals. The pastured animals spread their
own waste back on the grassland which produced the
feed originally. They do not, however, spread it evenly.
The wastes accumulate near watering locations and in
shady areas. These areas are usually located very near
stream banks and therein lies the potential problem.
Some States have laws which require the fencing of
streams which feed directly into public water supplies.
This is a very positive way of approaching the problem.
There are other solutions: Shade can be provided
away from streams; water tanks can be located near
the new shade; and the trees could be removed from
along the streams. There seems to be some correlation
-------
RURAL WATERSHED POLLUTION CONTROL 259
between the amount of forage remaining on the
pasture and the quality of the runoff from the pasture.
The more forage remaining, the better the quality of the
runoff from that area. This possible correlation
indicates that good forage and cattle management may
provide the necessary BMP's for most of the nonpoint
source (NPS) pasture problems.
The problem of bacterial pollution is one that will
always be of concern in the NPS programs. Coliform
counts from pastures have been reported to exceed
stream standards at most locations. Several studies on
water quality from grassland with and without animals
indicate that bacterial pollution counts vary greatly and
can be very high from the ungrazed grassland. This
indicates that domestic animals are not the only
problem; removing them from the areas may only allow
wildlife numbers to increase. A great deal remains to
be learned about the effectiveness of BMP's for
unconfined animal production.
DECIDING ON BEST MANAGEMENT
PRACTICES
Agencies or individuals charged with developing
programs to control pollution resulting from animal
production will need to use a systematic procedure to
be able to properly identify NPS problems and to
recommend practical BMP's to solve the problems. The
most effective way of developing these local NPS
programs will be to bring together a group of specialists
to identify the problems and develop specific guidelines
for the localized area. The responsible agency should
seek assistance from all Federal, State, and university
agencies active in the area. The group should include
agronomists, soil scientists, hydrologists, economists,
engineers, biologists, and most importantly, farmers
who know the local area. These representatives
working as a group and using all of the available
information on animal waste management should be
able to pinpoint problem areas, develop realtistic BMP
concepts, and design evaluation programs that will be
suitable for implementation and will solve the problem.
REFERENCES
Gilbertson, C. B., et al. 1979. Animal waste utilization on
cropland and pasture — a manual for evaluating agronomic
and environmental effects. Sci. Edu. Admin. U.S. Dep. Agri.
Utilization Res. Rep. 6. EPA-600/2-79-059. Off. Res.
Develop. U.S. Environ. Prot. Agency, Washington, D.C.
Miner, J. R., R. B. Wensink, and R. M. McDowell. 1979.
Design and cost of feedlot runoff control facilities. EPA-
600/2-79-070. Off. Res. Develop. R. S. Kerr Environ. Res.
Lab. U.S. Environ. Prot. Agency, Ada, Okla.
Van Dyne, D. L, and C. B. Gilbertson. 1978. Estimated U.S.
livestock and poultry manure and nutrient production. Econ.
Stat. Coop. Serv. Rep. ESCS-12. U.S. Dep. Agric.
Washington, D.C.
White, R. K., and D. L Forster. 1978. A manual on : Evaluation
and economic analysis of livestock waste management
system. EPA-600/2-78-102. Off. Res. Develop. R. S. Kerr
Environ. Res. Lab. U.S. Environ. Prot. Agency, Ada, Okla.
Zovne, J. J., and J. K. Koelliker. 1979. Application of
continuous watershed modelling to feedlot runoff manage-
ment and control. EPA-600/2-79-065. Off. Res. Develop. R.
S. Kerr Environ. Res. Lab. U.S. Environ. Prot. Agency, Ada,
Okla.
-------
260
USDA SOIL CONSERVATION SERVICE STANDARDS FOR
LIVESTOCK MANURE MANAGEMENT PRACTICES
CHARLES E. FOGG
Soil Conservation Service
U.S. Department of Agriculture
Washington, D.C.
ABSTRACT
Studies of nonpoint sources of pollution throughout the United States indicate that livestock and
poultry operations often significantly reduce water quality in surface waters to which they drain.
This can be in the form of excess nutrients, reduced dissolved oxygen, or increased coliform
bacteria. The Soil Conservation Service (SCS) has developed standards for planning, designing,
constructing, and operating several practices for managing manure to minimize pollution of
surface and ground waters. The first step is developing an overall waste management system plan.
Known as Practice 312, Waste Management System, it sets forth all the system components
needed on a farm or ranch to properly manage manure from the time it is excreted to its ultimate
use as a beneficial resource — usually for improvement of soil tilth and fertility. The overall waste
management system may logically include other soil and water conservation practices to prevent
degradation of water, air, soil, or plants. Minimum standards for these practices are set forth in the
SCS National Handbook of Conservation Practices. These standards are supplemented by States
as necessary to meet local requirements. A waste management system requires not only careful
planning initially, but careful, consistent operation and maintenance by the owner.
INTRODUCTION
The Soil Conservation Service is an agency within
the U.S. Department of Agriculture. We provide
technical assistance to landowners through locally
organized soil and water conservation districts to
protect soil and water for long-term production. This
technical assistance includes planning, designing, and
supervising installation of systems for managing
livestock and poultry manure.
Waste management systems serve two main
purposes — efficient management of manure for
beneficial use, and control of pollution. The need for
manure management systems became apparent as
livestock production under confined conditions in-
creased in the late 1960's and early 70's. SCS modified
existing soil and water conservation practices and
developed new ones to meet this need.
Our first systems were simple in concept. They were
aimed at controlling pollution from confined livestock
areas. Typically, they included clean water diversions
to prevent such water from reaching livestock areas,
polluted runoff diversions or collection ditches to
intercept flow before it reaches surface waters holding
ponds to retain polluted runoff, and irrigation or other
equipment to apply collected liquids on available land.
Systems now include waste storage structures, filter
strips, and designation of waste utilization areas to
provide more efficient management of the manure and
runoff water and make beneficial use of contained
nutrients.
SCOPE OF PROBLEM
While SCS provides technical assistance in installing
about 3,500 waste management systems a year, this
does not meet the need for such systems across the
country. There are some 1,800,000 farms in the United
States with some type of livestock or poultry. Possibly
800,000 of these can be considered confined opera-
tions.
Estimates of the number and size of confined feeding
operations having a potential for polluting surface
waters were made in 1976 by USDA and State
research and extension personnel in major livestock-
producing States. The estimates were made for 18
States representing 95 percent of beef production, 15
States with 90 percent of swine production, and 24
States with 85 percent of dairy production. The animal
waste subcommittee of the USDA Environmental
Quality Committee summarized the estimates and
provided them to the U.S. Environmental Protection
Agency to help in determining the impact of proposed
feedlot regulations.
About 94,500 operations in the major livestock
producing States pose a potential pollution threat
because of discharge in manmade waste conveyances
to surface waters, in a watercourse traversing the
operation, or in operations with 1,000 animal units or
more with a discharge reaching surface waters from a
storm of less magnitude than a 24-year-frequency, 24-
hour storm event. About 14,000 beef, 32,000 dairy,
and 48,500 swine operations make up this total and '
represent 20, 19, and 11 percent of total production,
respectively.
-------
RURAL WATERSHED POLLUTION CONTROL
261
In addtion, approximately 105,000 operations with
less than 1,000 animal units have discharges reaching
surface waters from storms of less than a 25-year-
frequency, 24-hour magnitude. This includes about
23,000 beef, 29,500 dairy, and 54,500 swine
operations and 8, 16, and 11 percent of the total
production, respectively.
The estimated 200,000 confined livestock operations
with potential pollution problems represented 28
percent of the 719,000 total operations in the major
production States.
POINT AND NONPOINT SOURCES
1OF POLLUTION
SCS is not a regulatory agency. It is our job to provide
technical assistance to land owners and operators to
help them comply with local State and Federal
regulations. When a livestock or poultry operation is
relatively large or is found to be a significant source of
pollution, it is designated as a "point source" by the
U.S. Environmental Protection Agency or the State
regulatory agency. As a point source, the owner or
operator must comply with the National Pollutant
Discharge Elimination System. An NPDES permit is
needed, and compliance requires that pollutants not be
discharged to surface waters. This requirement means
that all runoff from storms up to a 25-year, 24-hour
event must be retained on the farm. While EPA
regulations are uniform across the country, State
regulatory agency designations of concentrated animal
point sources vary considerably.
Smaller livestock and poultry operations are gen-
erally considered "nonpoint sources" of pollution.
Control of such nonpoint sources involves the use of
best management practices (BMP's). Depending on
specific site conditions, BMP's can range from the no-
discharge systems used with point sources to simply
directing any discharge through grass filter areas. Even
for the smaller operations, however, SCS generally
recommends installing a livestock waste management
system that prevents discharge of polluted water to
surface waters.
LIVESTOCK WASTE MANAGEMENT
SYSTEMS
SCS standards require that a complete waste
management system be planned before individual
practices are installed. This is to prevent the owner
from investing in a component that may not be a logical
part of a total system needed for that particular
enterprise. Our concept of a complete waste man-
agement system is to provide facilities for management
of manure from its production to utilization — usually
on the land. While national standards must be met,
practice standards may vary from State to State to
recognize differences in climate. State and local
regulations, and types of livestock and poultry
operations. National practice standards are revised
periodically on the basis of improved technology and
experience. SCS national practice standards for a
waste management system and its various com-
ponents are summarized as follows.
PRACTICE 312 - WASTE
MANAGEMENT SYSTEM
This practice comes before all others. It evaluates all
liquid and solid waste sources on a farm and develops a
complete system including all necessary components
to manage them without degrading air, soil, or water
resources. The practice considers the waste from the
time of production to its ultimate use on the land. This
practice determines if there is sufficient land to utilize
contained nutrients in the waste and if the land is
available at times compatible with crop management
and labor requirements. While we emphasize man-
agement of wastes in a manner that conserves
nutrients, there are situations where sufficient land is
not available to utilize the nutrients. In such cases it
may be best to plan practices where nitrogen loss is
maximized.
A waste management system may consist of a single
practice such as a clean water diversion, or a
combination of several practices. It is important that
individual practices be installed in a sequence that
insures that each will function as intended without
being hazardous to others. For example, a lagoon or
holding pond should not be installed until planned
diversion of outside sources of runoff has been
accomplished.
Components of complete waste management sys-
tems may include, but are not limited to, the following
practices: Debris basins, dikes, diversions, fencing,
filter strips, grassed waterways, irrigation systems,
pond sealing or lining, subsurface drains, waste
storage ponds, waste storage structures, waste
treatment lagoons, and waste utilization.
Another important element of a complete waste
management plan is guidance for operation and
maintenance. An operation plan is prepared for the
owner, providing specific details for operation of each
component of the system. Typically, such a plan should
include:
1. Timing, rates, volumes, and locations for applica-
tion of waste and, if appropriate, approximate number
of trips for hauling equipment and an estimate of the
time required.
2. Minimum and maximum operation levels for
storage and treatment practices and other operations
specific to the practice, such as estimated frequency of
solids removal.
3. Safety warnings, particularly where there is
danger of drowning or exposure to poisonous or
explosive gases.
4. Maintenance requirements for each of the
practices.
PRACTICE 425 - WASTE STORAGE
POND
Waste storage ponds are used to temporarily store
liquid and solid wastes, wastewater, and polluted
runoff until it can be safely applied to land or otherwise
used without polluting surface or ground water. They
are constructed of earth and may have paved entrance
ramps and bottoms to facilitate emptying.
-------
262
RESTORATION OF LAKES AND INLAND WATERS
A common use of waste storage ponds is to store
polluted runoff from concentrated livestock areas such
as feedlots and barnyards. Another common use is for
storage of manure and milkho.use or milking center
wastes.
Diversions or dikes are usually used in conjunction
with waste storage ponds. Some divert clean water
away from concentrated livestock areas, and others are
designed to collect polluted runoff and direct it to the
storage pond. They are designed to handle the same
storm event which governs the storage pond design,
usually the runoff from a 25-year, 24-hour rainfall.
Design of the waste storage pond must consider the
maximum period between emptying. This varies
according to climate, crops, and labor. The design
volume must equal or exceed the total of the following:
With drainage area — Without drainage area —
1. Manure, wastewater, 1. Manure and waste
and normal runoff1 water'
2. Normal precipitation 2. Normal precipitation
less evaporation on less evaporation on
pond surface1 pond surface1
3. 25-year, 24-hour 3. 25-year, 24-hour
runoff precipitation on
pond surface
4. Solids accumulation2 4. Solids accumulation2
'Accumulated during the storage period.
2For the period between solids removal. This applies
mainly to ponds used to store wastewater and polluted
runoff and refers to the residual solids after the liquids
have been removed.
Additional storage is often provided to meet
management goals or State regulations. The most
common comment heard from owners is that, if they
were to construct a pond again, they would provide
more storage to allow more flexibility in applying the
manure or wastewater ic^the land.
When storing polluted runoff or liquid wastes, it is
advisable to direct the waste through some type of
solids removal device. This cuts down on the frequency
of removing accumulated solids from such ponds and
reduces problems when spray irrigation type equip-
ment is used to apply liquids to the land. Common
solids removal facilities include debris (settling) basins,
low-gradient channels, and vegetative filter strips.
Various mechanical devices are also available for this
purpose.
There is some concern that waste storage ponds will
pollute ground water. Certain gravelly soils and
shallow soils over fractured or cavernous rocks should
be avoided unless special precautions are taken.
Research indicates and our experience shows that
waste storage ponds rapidly seal in all except very
coarse solids if the waste is organic and solids content
exceeds about 0.5 percent. This seal is a result of
settling of fine solids and biological action at the
interface of the soil with the waste. It is good practice to
retain some liquids in the ponds to maintain the seal.
Extended drying breaks down the biological seal. The
seal is reestablished when wastes are again intro-
duced.
PRACTICE 359 - WASTE
TREATMENT LAGOON
When animal or other agricultural wastes must be
treated, waste treatment lagoons may be a logical
component of a waste management system. Treatment
may be needed for odor control or, where land for
application is limited, to reduce nitrogen content of the
waste.
Waste treatment lagoons are designed as anaerobic,
aerobic, or aerated lagoons. For livestock wastes, they
are often used to biologically treat milkhouse or milking
center wastes, liquid manure from flush systems, or
other types of liquid wastes.
Anaerobic lagoons are the most common type for
livestock wastes. They require much less area and
volume than aerobic lagoons and do not require the
energy input of aerated lagoons. Volatilization of
ammonia causes substantial loss of nitrogen from
anaerobic lagoons, often allowing use of smaller land
areas for the waste. When the lagoons are properly
designed and managed, odors are not usually a
problem in agricultural areas. However, when the
lagoons are overloaded and often when they are being
emptied, odors may be objectionable in residential
areas. The effluent from anaerobic lagoons is not of
sufficiently high quality for discharge to surface
waters.
Aerobic lagoons are sometimes used for milkhouse
or milking center wastes and for other relatively weak
agricultural wastes. Because of large surface areas
involved, they are not often used for treating liquid
manure. Rarely is the effluent from aerobic lagoons of
sufficient quality for discharge to surface waters.
Aerated lagoons are used primarily for odor control.
The cost of the energy required generally prohibits their
use for complete mixing and treatment of strong
agricultural wastes. When used for odor control, they
are designed to aerate the surface of the lagoon — the
remainder is anaerobic. Once again, the effluent
should not be discharged to surface waters.
Anaerobic lagoons are designed on the basis of
volatile solids (VS) loadings ranging from about 3
pounds VS in the north to 7 pounds VS in the south per
1,000 cubic feet per day. Aerobic lagoons are designed
on the basis of 5-day biochemical oxygen demand
(BOD5) ranging from about 20 pounds BOD5 in the
north to 60 pounds BOD5 in the south per acre of
surface area per day. Aerobic lagoons treating animal
wastes with a high chemical oxygen demand to BOD5
ratio often are aerobic only near the surface. Virtually
all aerobic lagoons treating organic agricultural wastes
have an anaerobic zone at the liquid-soil interface.
Waste treatment lagoons are not designed to treat
polluted runoff. The irregularity of runoff events and
variability of pollution loading are not amenable to
rational design. Uncontrolled outside runoff is excluded
from lagoons for these reasons.
PRACTICE 313 - WASTE STORAGE
STRUCTURE
Waste storage structures include storage tanks and
manure stacking facilities. In contrast to waste storage
ponds, they are made of materials such as reinfciced
-------
RURAL WATERSHED POLLUTION CONTROL
263
concrete, coated steel, wood, and masonry. As
components of waste management systems, waste
storage structures are used to temporarily store liquid
or solid wastes until they can be safely applied on the
land or otherwise used. Storage tanks are used for
liquid and slurry wastes. Stacking facilities are used for
wastes that behave as solids.
Waste storage structures are sized to store accumu-
lated wastes, bedding, wastewater, and any needed
dilution water for the maximum period that such
wastes cannot be beneficially used. The period of cold
and rainy weather often dictates how long wastes must
be stored; however, stage of crop growth or availability
of labor may influence length of storage. Structures are
designed to insure that they are sound and of durable
materials commensurate with the required service life,
cost, and maintenance. Anticipated service life is
broken into three categories — 10 to 20 years, 20 to 50
years, and over 50 years — depending on the specific
enterprise and the owner's desires.
Provisions, such as entrance ramps and pumping
and agitating ports, are provided for emptying waste
storage structures. The owner or operator must have
equipment available for his use in filling and emptying
the structures, and he must provide warning signs,
ladders, ropes, rails, and other devices as necessary for
the safety of human beings and livestock. Proper
ventilation of enclosed structures is a critical concern
to prevent accumulation of explosive and toxic gases.
While waste storage structures are expensive
components of waste management systems, they offer
many advantages over waste storage ponds. Advan-
tages include preservation of nutrient content of stored
wastes, minimization of odors, management flexibility,
and improved aesthetics. Occasionally, State regula-
tory agencies prefer or require waste storage struc-
tures rather than waste storage ponds.
PRACTICE 633 — WASTE
UTILIZATION
The purpose of the waste utilization practice is to
safely use wastes to provide fertility for crop, forage, or
fiber production; to improve or maintain soil structure;
to prevent erosion; to produce energy; and to safeguard
water resources. It completes a waste management
system.
With most animal waste management systems,
waste utilization refers to where and when manure
should be applied to land. This practice is developed to
match conditions in each State. Available nutrients are
determined based on type and number of livestock and
nitrogen losses related to method of management. For
example, manure stored on an open lot and spread
annually may lose up to 70 percent of its nitrogen
whereas manure stored in a deep tank and incorpo-
rated into the soil before drying may lose only 20
percent.
Land areas available for application of manure and
crops to be grown are determined. Amounts of manure
to be applied are based on nutrient content of the
manure and nutrient needs of the crop. Timing of
applications is based on stage of crop growth and
availability of labor and equipment.
Many other factors are considered in planning a
waste utilization practice. They include rates of release
of nutrients from manure, soil types, climate, and
moisture need. If a lot of land is available, it may be best
to apply only enough manure to meet phosphorus need
and apply supplemental commercial nitrogen. If land is
limited, it may be best to supply all the crop's nitrogen
needs with manure. The most important factor is the
owner's preference.
CONCLUSION
PRACTICE 393 — FILTER STRIP
Filter strips have a definite place as components of
waste management systems. While they have been
used for years to filter sediment from water flowing
from cropland, their use as a formal practice in waste
management is relatively new. Their purpose is to
remove sediment and other pollutants from runoff by
filtration, infiltration, absorption, adsorption, decom-
position, and volatilization.
Filter strips can be considered a useful and relatively
inexpensive practice for reducing sediment and other
nonpoint pollutants. To date, national design criteria do
not spell out the limits of the effectiveness of filter
strips. They are currently being used between feeding
areas and streams for livestock on pasture, between
areas where wastes are stored and surface waters,
below feedlot areas to filter solids from polluted runoff
before runoff is directed to holding ponds, and to a
limited extent as facilities for reducing pollutants in
runoff from concentrated livestock areas. It is hoped
that additional research and experience will provide
improved guidelines relative to length of filter area and
reduction of various pollutants.
The U.S. Department of Agriculture is emphasizing
greater use of organic wastes to improve soil tilth and
fertility. Working with farmers and ranchers across the
country, SCS is planning, designing, and supervising
installation of waste management systems to abate
pollution and to use organic wastes as resources.
We have discussed the general content of waste
management practices as developed at the national
level. Minimum standards are set forth in the SCS
National Handbook of Conservation Practices. The SCS
staff in each State supplements these standards as
necessary to meet local conditions.
The key to a successful waste management system is
the owner or operator. If he considers management of
waste a nuisance and an unpleasant chore, the best
conceived system just will not work. However, if he has
a real interest in making beneficial use of the wastes,
he will make his system work regardless of any
shortcomings. A waste management system and its
components require careful operation and mainten-
ance. If a lagoon or waste storage pond or tank is not
agitated and wastes removed on a regular basis, the job
becomes much more difficult. Once weeds and brush
with their massive root systems begin to form on the
floating mat of a dairy manure storage pond, for
-------
264 RESTORATION OF UXKES AND INLAND WATERS
example, it becomes virtually impossible to agitate and
empty the facility.
SCS is emphasizing that complete waste manage-
ment systems be planned before individual com-
ponents are installed, that the plan include designation
of areas for beneficial use of the waste, and that a
written plan for operation of the system be developed
with the land owner or operator.
REFERENCES
Animal Waste Subcommittee. 1976. Implications of EPA
proposed regulations of November 20, 1975, for the animal
feeding industries. Environ. Qual. Comm. U.S. Dep. Agric.
Soil Conservation Service. National handbook conservation
practices. U.S. Dep. Agric., Washington, D.C.
-------
265
AGRICULTURAL NONPOINT SOURCE CONTROL OF
PHOSPHORUS AS A REMEDY TO EUTROPHICATION
OF A DRINKING WATER SUPPLY
MARK P. BROWN
MICHAEL RAFFERTY
Bureau of Water Research
New York State Department of Environmental Conservation
Albany, New York
ABSTRACT
The water quality goal of the West Branch Delaware River Model Implementation Program, a U.S.
EPA-USDA sponsored agricultural and silvicultural nonpoint source remedial program, is the
deceleration of eutrophication in the Cannonsville Reservoir, a New York City drinking water
supply. The river, which provides 75 percent of the reservoir's water, drains a 85,000 hectare
forest (70 percent) and dairy agriculture (22 percent) watershed. Data collected from the reservoir
by various investigators during 1971 -1975 document the reservoir as a eutrophic system largely
controlled by phosphorus. Total phosphorus (TP) loading estimates made from 1972-1975 data
range from 39 to 117 tons/year. Based upon 1976-1978 data (n = 70),TP loading to the reservoir
has been reduced to approximately 22 tons/year, largely because of improvements in point source
discharges. TP loading remains near the level considered dangerous with respect to
eutrophication. The program has directed approximately 75 percent of its Federal cost-sharing
funds toward animal waste management practices, primarily barnyard runoff controls (275 dairy
farms generate about 200 tons of phosphorus annually). A research and monitoring program will
evaluate the effectiveness of barnyard runoff controls and the impact of the program on
phosphorus export from the watershed and eutrophication in the reservoir.
INTRODUCTION
Increased attention is being given to the section 208
(P.L. 92-500) requirement calling for the development
of processes to control agricultural and silvicultural
nonpoint sources of water pollution. Responding to this
requirement, the U.S. Department of Agriculture and
the U.S. Environmental Protection Agency initiated the
USDA/USEPA Model Implementation Program. The
program employs existing agency funds to provide for
land management in a select number of model
watersheds to improve water quality.
In late 1977 the New York State USDA section 208
advisory committee nominated the West Branch of the
Delaware River as a candidate for a Model Implemen-
tation Program, citing use impairment of the Cannons-
ville Reservoir as a New York City drinking water supply
as the ultimate target. EPA had estimated that the
reservoir was receiving over three times the phos-
phorus loading considered dangerous with respect to
eutrophication. Nonpoint source phosphorus derived
from cropland and animal wastes was suspected to
account for a significant portion of the total load.
A technical conference convened in early 1978 after
approval of the WBDR-MIP application provided the
following conclusions to guide the MIP (N.Y. Dep.
Environ. Conserv. 1978):
1. Nutrient (phosphorus) enrichment of the Can-
nonsville Reservoir is the critical water quality problem
to be addressed.
2. Nonpoint source control measures with the
greatest potential for reducing phosphorus loadings
must receive the highest priority.
3. Measures relating to dissolved phosphorus must
receive higher priority than those relating primarily to
total phosphorus.
4. Measures not having a substantial relationship to
the phosphorus loading, as defined in the guidance
document, must receive a low priority.
The conference defined milkhouse wastes, barn-
yards, manure storage, and manure spreading as high
priority sources of phosphorus, particularly dissolved
phosphorus. Cropland, initially targeted for a large
share of attention, was considered a low priority source
because of its relatively unknown potential for
phosphorus loading. The conference concluded that
owing to the complex nature of the reservoir and the
contribution of point sources, measurable improve-
ment in the water quality of the Cannonsville Reservoir
would not be expected during the program's 3-year
period or shortly thereafter. The guidance document
expressed the need for a baseline analysis of existing
information on the river and the Cannonsville
Reservoir.
This paper summarizes the available data for total
phosphorus from previous investigations and attempts
to provide a clearer perspective of the water quality
goals of the program. While dissolved phosphorous is
probably more relevant to eutrophication, the paucity of
reliable dissolved phosphorus data for the WBDR
-------
266
RESTORATION OF LAKES AND INLAND WATERS
prohibits its use to analyze trends in dissolved
phosphorus loading.
STUDY AREA
The West Branch of the Delaware River watershed,
located in the southeastern portion of New York State,
is approximately 90 kilometers in length and drains in a
southwesterly direction into the Cannonsville Reser-
voir (Figure 1). The reservoir is owned and operated as
a public water supply by the City of New York. Land
uses in the 116,000 hectare watershed (Table 1) are
predominantly woodland (70 percent) and dairy
agriculture (22 percent). Urban areas occupy only 1
percent of the study area. The population of the
watershed is approximately 17,000. Watershed topog-
raphy is rolling to mountainous. The average water-
shed slope is moderately steep (approximately 20
percent).
Climate is humid continental with temperatures
averaging 8°C annually and precipitation averaging
over 100 centimeters annually. Stream runoff from this
precipitation averages 64 centimeters annually. The
ground generally is frozen in Delaware County from
December to March.
The Cannonsville Reservoir, formed by damming the
West Branch of the Delaware River near Stilesville,
N.Y., began filling in 1963. The surface area of the
reservoir at a useful storage capacity (0.362 km2) is
19.43 square kilometers and its mean depth is 18.6
meters. Seventy-eight percent of the reservoir water-
shed is drained by the river. The remaining 22 percent
is accounted for by a number of small tributaries and
direct input. Flow in the river below the dam is
maintained by a hypolimnetic discharge during periods
when crest capacity is not exceeded. The annual
average turnover time is 0.45 year. Normalized flow
data for the reservoir (U.S. EPA, 1974) indicate 55
percent water replacement during the period February-
May during a normal year based upon the assumption
of complete mixing. In light of the morphometry of the
long and narrow West Branch arm of the Cannonsville
Reservoir, complete mixing is probably a weak
LEGENL
|5I7 ICN WATERSHEL NUMBER
SCALE l" = 4 Ml
WEST BRANCH DELAWARE RIVER
WATERSHED BOUNDARY MAP
assumption and indeed much of the reservoir's water
could be replaced during high flow by nondispersive
advection.
Table 1 — Land use in the Cannonsville Reservoir
watershed (Soil Conserv. Serv., 1977)
Cannonsville
Reservoir
ha Percent*
Woodland
Cropland
Pastureland
Former cropland
Urban
Other
81,160
16,318
8,891
3,342
1,013
5,515
116,239
70
14
8
3
1
5
"Figures Total > 100 due to rounding.
Figure 1. — West Branch Delaware River watershed boundary
map.
RESERVOIR EUTROPHICATION AND
PHOSPHORUS LOADING
The available water quality data for the reservoir are
from four sources: a phytoplankton survey of the
Delaware River Basin by Schumacher and Wager
(1973), the National Eutrophication Survey (U.S. EPA,
1974), a limnological study of the reservoir by Wood,
(1979), and the New York City Department of Water
Resources routine water quality monitoring data. The
U.S. Geological Survey maintained water quality
stations at Walton and at Beerston where the river
enters the reservoir and at Deposit (Figure 1 (below the
reservoir from May 1973 through April 1975. All
concluded the reservoir was eutrophic.
Using an analysis developed by Vollenweider (1975),
U.S. EPA (1974) estimated that the reservoir would
remain eutrophic if annual surface loading exceeded
1.05 g P/m2/year. If loading was reduced below this
limit, the reservoir could eventually become meso-
trophic. EPA (1974), employing 1972-1973 data,
estimated surface loading to be 4.26 g P/m2 year. In
this estimate the river accounted for 97 percent of the
load. Wood (1979), using 1974-1975 data estimated
surface loading to be 1.37 to 1.66 g P/m2 year. The six
estimates of TP loading to the Cannonsville Reservoir
made prior to 1980 range from 27 to 117 tons/year. In
these estimates the river accounted for a minimum of
77 percent of the load. Known point sources accounted
for 27 to 41 tons/year. An estimate of 23 tons/year
was generally employed for a cheese processing plant
in Walton.
Much of the variability in the estimates can be
accounted for by the assumptions concerning stream
discharge and sample collection. While some investi-
gators employed long-term average or normal flow,
others considered flow for a specific year. The EPA
(1974) and Geological Survey (1974, 1975, 1976) data
were collected from November 1972 to April 1975. All
of the pre-1980 loading estimates with the exception of
Wood's (1979) have used the combined EPA and
Survey data (n = 36). Wood's (1979) estimate is the only
one incorporating a substantial amount of data from
1975. His substantially lower estimate probably
reflects the reduction in TP loading resulting from the
phaseout in the use of phosphorus-based cleaners at
the cheese plant in Walton during 1975. Brown and
-------
RURAL WATERSHED POLLUTION CONTROL
267
Rafferty (1980) examined estimates of phosphorus
loading to the reservoir in greater detail and addressed
the limitations of the 1972-1975 data. They concluded
that the sampling program for the river below Walton
may have biased results due to diurnal and weekly
variations in the cheese plant's effluent quality. Such a
bias could have resulted in overestimating the annual
load to the reservoir.
A statistical summary of the entire TP data set for the
river above Walton and at Beerston, where the river
enters the reservoir, from 1973 to 1979 is presented in
Figure 2. It is readily evident that the range of TP in the
pre-1975 data is substantially greater for Beerston
than for the river above Walton. Based upon 95 percent
confidence limits as presented in Figure 2, the meanTP
concentration at Beerston was significantly higher
than above Walton during 1973-1974. Similarly, the TP
concentration at Beerston during 1973-1974 was
significantly higher than for the period 1974-1979. The
dramatic reduction in TP concentration that occurred in
the Beerston record in 1975 corresponds in time to a
phaseout in the use of phosphorus-based cleaners at
the cheese plant in Walton. Based upon incremental
drainage area, stream flow at Beerston is generally 10
percent greater than above Walton. Incremental flow
enters diffusely from a predominantly forested water-
shed.
Observations by Schumacher (pers. commun.) cor-
roborate the reduction in nutrient loading that occurred
in the Walton area during 1975. Schumacher and
Wager (1973) had noted that unlike the river above
Walton high concentrations of fungi were found in
samples approximately 9 kilometers downstream from
Walton. The fungi there were dense enough to prohibit
phytoplankton counting. Schumacher observed that
the fungi were no longer visible in the same area
during and after 1975, and were replaced by
filamentous green forms.
Log TP concentration on log flow linear and second
order polynomial regressions performed on the 1976-
1979 data yielded no significant relationships on which
to base a loading estimate. Based upon the 1976-1978
mean TP concentration at Beerston, .023 mg/l, mean
stream discharge for the same period, 25 mVs and 95
percent confidence limits on the mean TP concen-
tration results in an annual TP loading estimate of
19,300 ± 4,800 kg/year. Employing the EPA(1974)TP
loading estimates from tributaries, immediate drainage,
and direct precipitation to the reservoir, 2,700 kg/year,
the total loading to the reservoir would be 22,000
kg/year or 1.14 g/mVyear. Using Vollenweider's
(1975) model, EPA estimated that the eutrophic rate of
phosphorus loading was 1.05 g/m2.
Mean Annual Total Phosphorus (TP)
Concentration in the West Branch
Delaware River
Mean TP Concfntralion i 95% CF,
a.
1
i Mean TP1
I B«i
1973 1974 1975 1976 1977
Y.or
1170 1979
Figure 2. — Mean annual total phosphorus (TP)
concentration in the West Branch Delaware River.
Table 2. — Summary of TP loading estimates to the Cannonsville Reservoir
Investigator
Hydroscience (1974)
EPA (1974)
Bricke (1975)
Bricke (1975)
Goodale (1975)
Goodale (1975)
Wood (1979)
Brown and Rafferty
(1980)
Data
USGS (1974, 1975)
EPA (1974)
Nov. 72 — Oct. '74
EPA (1974)
Nov. '72 — Oct. 73
EPA (1974)
USGS (1974, 1974)
EPA (1975)
USGS (1974, 1975)
EPA (1974)
USGS (1974, 1975)
EPA (1974)
USGS (1974, 1975)
Wood (1979)
NYC Dep.
Water Resour.
unpubl. data,
1976-1978
Estimate
metric tons/year
76 — 117
83
39
49
67
40
27-32
22
Methods
Product of mean TP concentration and
mean flow for different periods
of interest
Sum of products of normalized monthly
flow and empirically adjusted
TP concentration
Log [TP] on log flow linear regression to
determine concentration for flow duration
analysis.
Log [TP] on log flow 2° polynomial
regression to determine concentration
for flow duration analysis.
Product of mean annual flow and mean
TP concentration.
Log [TP] on log flow linear regression to
determine concentration for flow
duration analysis.
Not reported.
Product of mean TP concentration and
mean flow for the same period (1 — 2
order polynomial regressions of log (TP)
on log flow were not significant.)
-------
268
RESTORATION OF LAKES AND INLAND WATERS
INVENTORY OF PHOSPHORUS
SOURCES IN THE CANNONSVILLE
RESERVOIR WATERSHED
There are five municipal sewage treatment plants, a
meat processing plant, and a hubcap plating plant that
discharge directly to the river. The City of New York
Department of Water Resources has been monitoring
the quality of these effluents since 1977. Their data is
summarized in Table 3. Known point sources con-
tribute about 9,900 kg/year of TP Approximately 75
percent of the point source load comes from the Walton
sewage treatment plant, which has a mean TP effluent
concentration of 8.0 mg/l.
In addition to point sources, EPA(1 974) collected and
analyzed 14 samples during 1972-1973 from each of
three streams tributary to the reservoir draining
forested land. The mean TP concentration in Dry Brook
and Maxwell Brook was 0.015 mg/l. The third stream,
Dryden Brook, had a mean TP concentration of 0.018
mg/l. Assuming unit area runoff from forested land
equals that for other land uses, and employing the
1976-1978 mean stream discharge at Beerston (25.2
mVs), aTP concentration of 0.015 mg/l corresponds to
a unit area TP load of 0.14 kg/ha/year. This load is well
within the range reported for forested land in other
areas (Uttormark, et al. 1974). Approximately 11,000
kg/y would be delivered to the Cannonsville Reservoir
from forested land.
Remaining avenues of TP loading to the river and the
reservoir include export from agricultural and silvi-
cultural activities, urban runoff, septic systems,
landfills, and direct precipitation. Estimates of these
nonpoint sources are summarized in Table 3.
Logging operations appear to be minor contributions
of phosphorus. Results of an aerial survey conducted
by New York foresters during March 1979, indicated
over 70 sites in the watershed had been logged since
1974. Forty of these sites were surveyed. The average
size of the surveyed logging operations is 19.3 hectares
of which logging roads on the sites averaged 0.7
hectares. Logging roads are generally the major sites of
erosion on logging sites in the watershed; cleared
areas typically have adequate soil cover (Trotta, 1980).
Slavicek (1980) estimated that only eight of the 40 sites
contributed sediment to streams. Unit area sediment
loading rates developed for 12 transects, using the
Universal Soil Loss Equation (Wischmeier and Smith,
1965), indicated sediment loading ranged from 4.29 to
247.0 tons/ha/year.
Agricultural sources of TP are related primarily to
dairy farming although some beef and poultry farms
are located in the basin. These sources include loading
from barnyards, cropland and pastureland, and from
handling of animal and milkhouse wastes. Because
little data are available, estimates of TP loading from
cropland, pastureland,- and grassland in addition to
urban runoff (Table 3) should be viewed in an order of
magnitude sense.
Loading from barnyards, animal and milkhouse
waste handling are more difficult to estimate, though
their potential can be addressed. The 15,000 dairy
cows and 5,000 replacements on a total of approxi-
mately 275 farms in the watershed (Soil Conserv.
Sen/.,1978) would be expected to produce approxi-
mately 300,000 tons of manure per year containing
approximately 200 tons of phosphorus.
Table 3. — Inventory of phosphorus sources to Cannonsville Reservoir (Brown and Rafferty, 1980)
River distance
from reservoir
km
Point Sources
Sewage treatment plants:
"Stamford 72
'Hobart 69
'South Kortright 61
•Delhi 35
•Walton 8
Industrial effluents:
•Parnett 53
'Delchrome 8
Nonpoint Sources
Land runoff:
"Forest
Cropland
Pastureland
Grassland, other
Urban
Precipitation
Other Diffuse Sources
Barnyards
Milkhouse wastes
Septic systems
Total
Mean flow
m3/d
2,498
568
57
1,287
2,498
57
114
Area
ha
81,160
16,318
8,891
8,857
1,013
1,942
TP waste
load
kg/y
12,000
3,000-12,000
1,750-3,500
[TP]
mg/l
1.25 (n =50)
1.42 (n =47)
1.95 (n =41)
2.00 (n = 51)
8.00 (n = 14)
3.56 (n = 43)
2.20 (n =41)
Unit area
load
kg/ha-y
0.14
0.3
0.1
0.1
1.1
1.02
Delivery
ratio
7
?
?
Annual
load
kg/y
1,140
294
41
940
7,294
74
92
11,362
4,895
889
886
1,114
1,981
31,002
'Data available from the watershed.
-------
RURAL WATERSHED POLLUTION CONTROL
269
Based upon time spent in the barnyard by dairy cows
(Mattern, pers. commun.), 12,000 kg/year of TP would
be deposited on barnyards. Approximately 25 percent
of these barnyards are adjacent to streams and 70
percent are within 67 meters of streams. Draper, et al.
(1980) found that generally less than 10 percent of
manure phosphorus is exported from open lots and
employed a "best estimate" of 5 percent in analyzing
phosphorus loading to the Great Lakes from livestock in
the Ontario Basin.
While soils generally have a strong affinity for
inorganic phosphorus, spreading manure on frozen
ground has been shown to result in the export of up to
13 percent of manure phosphorus (Minshall, et al.
1970). In the West Branch watershed few farmers have
manure storage facilities. Manure is spread daily
during winter at locations often determined by
accessibility. It should be noted that terrestrial systems
conserve phosphorus, and only a small portion of that
stored in undisturbed watersheds is exported (Hobbie
and Likens, 1973; EI-Baroudi, 1975). Klausner, et al.
(1974) demonstrated that even combining high
phosphorus fertilization rates (49 kg/ha/year) and poor
soil management, resulted in export of less than 1
percent annually of inorganic phosphorus.
The major source of TP in milkhouse wastes is
phosphorus-based cleaners used to clean bulk tanks
and pipelines. These cleaners contain up to 8.7 percent
phosphorus by weight. Milk, which contains approxi-
mately 1 g/l phosphorus, accounts for an insignificant
portion of TP in milkhouse wastes due to the relatively
small quantity wasted. Assuming 0.34 to 1.4 kg/day of
8.7 percent phosphorus cleaners are used on each of
the 275 farms in the watershed, a general range of
3,000 to 12,000 kg/year of phosphorus is handled as
milkhouse wastes. Part of these wastes enters septic
tank-leach field systems; the remainder is discharged
to dry wells, tile fields, or surface ditches.
DISCUSSION
Phosphorus loading to the Cannonsville Reservoir
has been significantly reduced through managing point
sources. However,.preliminary monitoring data collect-
ed during summer 1980, indicate the reservoir is still
eutrophic. The remaining point sources and nonpoint
sources of phosphorus are summarized in Table 3. The
most credible estimates are those for point source
discharges and forest runoff, because of availability of
data from the watershed. Other estimates should be
viewed in an order of magnitude sense. The sum of
individual source loading estimates for the river at
Beerston is 24,500 kg/year (Brown and Rafferty, 1980)
as compared to 19,300 ± 4,800 kg/year based upon
New York City data for the river at Beerston. The
conservative approach used in estimating nonpoint
source loads where local data were unavailable
probably contributes to the general agreement of the
two estimates, considering the lack of high flow data in
the New York City data. For example, while TP loading
from cropland is probably underestimated, the rather
short hydrologically active periods when the greatest
percentage of phosphorus from cropland and other
nonpoint sources would be exported were not well
represented by the New York City monitoring.
Although the estimates of phosphorus loading from
point sources to the river are generally good due to the
availability of data, calculating a nonpoint source load
by difference between the TP estimate made from data
collected at Beerston is an especially tenuous
procedure. It is likely that the TP estimate made from
New York City data for the river at Beerston
underestimates the true annual load. In addition, the
assumption of conservative transport of sewage
phosphorus inherent to the calculation may not be
operationally valid for the river. Carlson, et al. (1978)
demonstrated the attenuation of soluble phosphorus
flux in streams below sewage treatment plants. The
mechanism of attenuation was incorporation with bed
sediments. If sewage phosphorus incorporation with
bed material is a significant mechanism in the river, the
delivery of this point source phosphorus to the
reservoir would be synonomous with the major
nonpoint source loads during those hydrologically
active periods that received no special consideration in
the New York City monitoring program.
With respect to eutrophication, that portion of the
total phosphorus load which is available for algal
growth is relevant. Lee, et al. (1980) suggest that
available phosphorus from urban and rural runoff
generally includes the soluble molybdate reactive
phosphorus fraction and approximately 20 percent of
the paniculate phosphorus. Orthophosphorus ac-
counted for approximately 60 percent of TP in samples
collected from the river at Beerston by EPA (1974).
The larger portion of the phosphorus from point
sources is probably available phosphorus. In streams
tributary to the Cannonsville Reservoir draining
forested land, approximately 50 percent of TP was
present as orthophosphorus (EPA, 1974). Much of the
phosphorus exported from cropland is likely to be
sorbed to or precipitated on soil particles (Burwell, etal.
1977; Alberts, et al. 1978) and thus not immediately
available. Much of the phosphorus loaded directly to
the lake by precipitation could be unavailable (Lee, et
al. 1980). Clearly, a better definition of phosphorus
sources, particularly nonpoint sources and their
bioavailability is needed.
The Walton sewage treatment plant appears to be a
major contributor of phosphorus to the river. Con-
sidering the morphometry of the long and narrow river
arm of the reservoir, the increase in phosphorus
concentration caused by the Walton sewage treatment
plant (.047 mg/l for normal July and August flow)
probably insures the eutrophication of at least the
upper reaches of the reservoir during summer months.
It is more difficult to speculate on the plant's impact on
the reservoir as a drinking water supply, since the
drinking water tunnel inlets are located approximately
20 kilometers down the 27 kilometer-long reservoir.
While costs and effectiveness associated with point
source improvements are well defined, the effective-
ness of nonpoint source controls is poorly understood.
For agriculture, the New York Model Implementation
Program is directed primarily toward controlling
barnyard runoff. Of the 154 barnyards given priority
based upon proximity to streams «100 m), 90 farms
participated in the program during 1978-1979; 67
signed up for installation of barnyard runoff controls.
Of the $449,000 in Federal cost sharing funds
-------
270
RESTORATION OF LAKES AND INLAND WATERS
Table 4. — Practices used in the West Branch, Delaware River Model Implementation Program
Agricultural (Based upon data from
Barnyard Runoff Controls
Practice
Roof gutters
Tile
Drop inlet
Concrete pad & gravel
Fencing
Diversions
Waterway
Open ditch
Land grading
Costs
Other
Temporary field storage of animal waste
Animal storage facility
Milkhouse waste filter strip
Critical area planting
Activity (as of 9/79)
Recon for Management Advice
Inspections
Management Plans
Prepared
Revised
Timber Stand Improvement Marking
Sawtimber Marked
Pulpwood Marked
Fuelwood Harvested
Information & Education
Log Road Erosion Control (6,450 m)
farms where practices were
Numbers of farms
where applied
27
45
11
46
36
26
9
7
3
50
1
2
1
16
Silvicultural
Number of
participants
34
30
8
2
19
4
1
1
54
installed during 1978-1979)
Average quantity
per farm
36m
68 m
1 Unit
1 Unit
150 m
93 m
0.3 m
48 m
0.4 ha
$3,137
04 ha
Total
hectares
114
128
256
45
66
36
24
5
committed during the first 2 years, approximately 75
percent is directed toward animal waste management
practices: in particular, barnyard runoff controls to limit
the size of barnyards, restrict direct access of cows to
streams, divert surface flow from entering the
barnyard, increase drying of barnyards, and facilitate
collection of deposited manure. The controls consist of
any combination of roof gutters and leaders, drop
inlets, land grading, concrete pads and gravel, buffer
strips, fencing, tile drainage, diversions, waterways,
and open ditches (Table 4).
While the primary emphasis has been placed on
barnyard runoff controls, a number of animal waste
storage facilities, access roads to cropland, and a
grassed filter strip for processing milkhouse wastes
have been applied on farms in the watershed. The
implementation of agricultural practices is being
directed locally by the Delaware County Soil and Water
Conservation District and the Soil Conservation Service
in Walton.
The Silvicultural activities are being administered by
the U.S. Forest Service and the New York State
Department of Environmental Conservation's Division
of Lands and Forests. Their work has included an aerial
reconnaissance of logging operations in the watershed
and recruitment of landowners' and loggers' coopera-
tion through a public information campaign. After the
participating landowner's or logger's site is inspected,
he can be advised of proper logging site management,
have logging trails and timber marked by foresters, and
possibly have a complete management plan prepared
for the site (Table 4). In light of the increasing demand
being placed upon wood as a n energy source, growth in
logging operations should coincide with erosion control
practices to maintain logging's status as a minor
contributor of phosphorus.
The evaluation of the water quality impact of the
program resulting particularly from agricultural man-
agement practices, is the goal of a research and
monitoring effort sponsored by the U.S. Environmental
Protection Agency and conducted by investigators from
the New York State Department of Environmental
Conservation, the State College of Agriculture and Life
Science, Cornell University, and the Soil Conservation
Service. It could be reasonably assumed that the
WBDR-MIP is targeting a less than 10 percent
reduction in TP loading to the reservoir through the
management of barnyard runoff. The evaluation will
address three major processes:
1. Phosphorus loading to tributaries from barnyard
runoff, manure spreading, and milkhouse wastes.
2. Phosphorus delivery to the Cannonsville Reservoir
from the WBDR.
3. Eutrophication of the Cannonsville Reservoir.
These processes differ in the analytical sensitivity
required to detect changes in them. The cumulative
effect of the program on phosphorus loading to the
Cannonsville Reservoir and its eutrophication may not
be discernible in measurements taken at the reservoir
because of the complexity of phosphorus sources and
-------
RURAL WATERSHED POLLUTION CONTROL
271
phosphorus transport and the relatively small target
loading reduction. However, the ongoing research and
monitoring of individual barnyard sources will permit at
least an a priori assessment of the program's impact on
the reservoir. In addition, the effectiveness of specific
management practices in controlling nutrient export
from barnyards will be evaluated.
REFERENCES
Bricke, K. 1975. An updated computation of the phosphorus
loading to the Cannonsville Reservoir. Unpubl. report. EPA
Region II, New York.
Brown, M. P., and M. Rafferty. 1980. A historical perspective
of phosphorus loading to the Cannonsville Reservoir as it
relates to the West Branch Delaware River Model
Implementation Program. Tech. Rep. 62. Bur. Water Res.
N.Y. State Dep. Environ. Conserv., Albany.
Carlson, G. A., L. J. Hetling, and W. W. Shuster. 1978.
Transport and loss of sewage phosphorus in streams. Am.
Soc. Civil Eng., New York.
Draper, D. W., J. B. Robinson, and D. R. Coote. 1980.
Estimation and management of the contribution by manure
by livestock in the Ontario Great Lakes basin to the
phosphorus loading of the Great Lakes. In R. Leohr, et al.,
eds. Best management practices for agriculture and
silviculture. Ann Arbor Science, Ann Arbor, Mich.
EI-Baroudi, H. 1975. Inventory of forms of nutrients stored in
a watershed. Rensselaer Polytechnic Inst. Troy, N.Y.
JGoodale, B. 1975. An analysis of phosphorus input to
Cannonsville Reservoir. File Rep. N.Y. State Dep. Environ.
Conserv., Albany.
Hobble, J. E., and G. E. Likens. 1973. Output of phosphorus,
dissolved organic carbon, and fine particulate carbon from
Hubbard Brook watersheds. Limnol. Oceanogr. 18:734.
Hydroscience. 1974. Discussion of water quality analysis of
the West Branch Delaware. Westwood, N.J.
Klausner, S. D., P. J. Zwerman, and D. F. Ellis. 1974. Surface
runoff losses of soluble nitrogen and phosphorus under two
systems of soil management. Jour. Environ. Qual. 3:42.
Lee, G. F., R. A. Jones, and W. Rast. 1980. Availability of
phosphorus to phytoplankton and its implication for
phosphorus management strategies. In Phosphorus man-
agement strategies for lakes. Ann Arbor Science, Ann
Arbor, Mich.
Mattern, P. Personal communication. Cooperative Extension
Agent, Delaware County Coop. Extens., Hamden, N.Y.
Minshall, N. E., M. S. Nichols, and S. A. Nitzel. 1970. Stream
enrichment from farm operations. Jour. San. Eng. Div. Am.
Soc. Civil Eng. 96:513.
New York State Department of Environmental Conservation.
1978. Water quality guidance, West Branch Delaware River
Model Implementation Program. Albany, N.Y.
Schumacher, G. J. Personal communication. Prof. Biolog.
Sci., State University of New York, Binghamton.
Shumacher, G. J., and D. B. Wager. 1973. A study of the
phytoplankton in the Delaware River basin streams in New
York State. Delaware River Basin Comm., Trenton, N. J.
Slavicek, R. L. 1980. The West Branch of the Delaware River
Model Implementation Program — survey of logging road
erosion and sediment production. Forest Serv. U.S. Dep.
Agric.
Soil Conservation Service. 1977. USDA/Model Implementa-
tion Program application.
1978. West Branch Delaware River watershed
nonpoint sources water pollution study. U.S. Dep. Agric.,
Syracuse, N.Y.
Trotta, P. Personal communication. Forester, N.Y. State Dep.
Environ. Conserv., Stamford, N.Y.
U.S. Environmental Protection Agency. 1974. Report on
Cannonsville Reservoir, Delaware County, N.Y. In National
Eutrophication Survey. Working Pap. 150.
U.S. Geological Survey. 1974. Water resources data for New
York, water year 1973. Water-Data Rep. NY-73-1.
1975. Water resources data for New York, water
year 1974. Water-Data Rep. NY-74-1.
_. 1976. Water resources data for New York, water
year 1975. Water-Data Rep. NY-75-1.
Uttormark, P. D., J. D. Chapin, and K. M. Green. 1974.
Estimating nutrient loadings of lakes from nonpoint
sources. Ecol. Res. Ser. EPA-660/3-74-02. U.S. Environ.
Prot. Agency.
Vollenweider, R. A. 1975. Input-output models. Can. Centre
Inland Waters, Burlington, Ontario, Canada.
Wischmeier, W. H., and D. D. Smith. 1965. Predicting rainfall
losses from cropland east of the Rocky Mountains.
Handbook 282. Agric. Res. Serv. U.S. Dep. Agric.
Washington, D.C.
Wood, L. W. 1979. The limnology of Cannonsville Reservoir,
Delaware County, N.Y. Environ. Health Rep. 6. N.Y. State
Dep. Health.
-------
272
RESERVOIR PROTECTION BY IN-RIVER
NUTRIENT REDUCTION
HEINZ BERNHARDT
Wahnbachtalsperrenverband
(Association of Wahnbach Reservoir)
Siegburg,Federal Republic of Germany
ABSTRACT
If in the catchment area of a reservoir the portion of phosphorus compounds from diffuse sources
prevails, phosphorus input can be reduced by chemical treatment of the main tributary. This
scheme has been applied by Wahnbachtalsperrenverband for the oligotrophication of Wahnbach
Reservoir (volume 40,000,000 m3). At the point where River Wahnbach flows into the reservoir the
incoming water taken from the pre-reservoir which serves as a reserve basin is treated by
precipitation, flocculatton with iron-lll-salts at pH 6.4, and filtration. With this the total P-
concentration is reduced by 99 percent to 4 /jg/l P as an average. Turbidity also is reduced to a
residual of 0.05 FTU and dissolved organic carbon is reduced by 60 percent. This is achieved by
energy-input controlled direct filtration (Wahnbach system) developed by Wahnbachtalsperren-
verband. The treatment process includes precipitation of dissolved phosphates, destabilization of
colloids and suspensoids, agglomeration of formed microfloc to large, well filtrable floes and three
layers' filtration with maximum 15 m/h filtration velocity. The maximum throughput of the plant
amounts to 18,000 mVh. The 3 years' run of the plant shows, that by drastically reducing the
annual average P-concentration from 100/jg/l to 4 /jg/\ the eutrophic Wahnbach Reservoir is
transformed from the eutrophic to the oligotrophic-mesotrophic status. The annual average
concentration of all tributaries including precipitation was reduced in 1 979 for the first time to 1 6
/ug/l, distinctly lower than the tolerable annual average concentration of 20 fjg/\. At present, the
dominating phosphorus load comes in via small marginal tributaries of the reservoir. This input
will be reduced by further special measures.
INTRODUCTION
The origins of phosphorus in the catchment area of a
reservoir are varied and detailed examinations have to
be carried out to determine the most important
phosphorus sources. One differentiates here between
'point' and 'diffuse' phosphorus sources. If diffuse
phosphorus sources dominate in a catchment area,
then there are only a few methods of reducing loading
(Bernhardt, 1978).
One of these methods entails treating the whole of
the inflowing main tributary using chemical processes
to control the phosphorus input. This chemical
treatment has been used on the Wahnbach Reservoir
in the Federal German Republic for 3 years (Bernhardt,
Clasen, and Schell, 1971; Bernhardt and Schell, 1979).
It can be applied to those reservoirs which have only
one or two tributaries or a gathering channel from a
neighboring catchment area and in which phosphorus
is a minimum factor.
EUTROPHICATION OF THE WAHNBACH
RESERVOIR
Increased phosphorus input has gradually eu-
trophied the Wahnbach Reservoir (40,000,000 m3
content) since impounding began in 1957. This
eutrophication process made it more and more difficult
to treat the raw water taken from the reservoir for
drinking water. At the end of the 1960's and beginning
of the 1970's, masses of blue-green algae Oscil/atoria
rubescens appeared. This not only colored the water
but the Oscillatoria broke through the filter. Using a
special flocculation process with a double dose of
flocculant combined with a dose of polyelectrolyte, we
generally mastered this calamity (Bernhardt and
Clasen, 1973).
Sometimes the mass development of large diatoms
such as Melosira ita/ica or Melosira islandica caused
shorter filter-run times because the rapid sand filter
became blocked after 4 or 5 hours. Later, the small
blue-green algae, e.g., Coelosphaerium naegelianum
grew in increasing quantities and despite a 90 to 99
percent reduction could not be totally eliminated from
the water because they were present in too large a
concentration (20,000 to 200,000 cells/ml in the raw
water). This was unsatisfactory for obtaining drinking
water from the Wahnbach Reservoir.
Difficulties particularly arose every autumn during
drinking water treatment as a result of algal organic
compounds (Bernhardt and Wilhelms, 1978). They also
disturbed the flocculation and disinfection and were
partly precursors for the development of trihalo-
-------
RURAL WATERSHED POLLUTION CONTROL
273
methane (Bernhardt and Hoyer, 1979). All these factors
compelled the Wahnbach Reservoir Association to take
steps to reduce the high input of phosphorus
compounds into the reservoir from its catchment area
so that the amount of bioproductivity in the reservoir
would become tolerable again.
THE CONCEPTION OF THE
OLIGOTROPHICATION OF THE
WAHNBACH RESERVOIR
Experiments carried out over a period of several
years showed that phosphorus compounds in the
reservoir are the limiting factor (Figure 1). Every year
orthophosphate was depleted in the reservoir down to
a concentration of <10/L/g/l owing to bioproductivity.
However, concentrations of nitrogen and carbon hardly
decreased at all. This meant that it was sufficient to
reduce the phosphorus in the lake only so that algal
development would be controlled.
Detailed studies on the origin of the phosphorus
compounds from the catchment area of the Wahnbach
Reservoir showed that more than 50 percent originated
from diffuse sources. Only a small part of them came
from locatable point sources (Bernhardt, et al. 1978).
For this reason the Wahnbach Reservoir Association
decided to erect a plant to decrease the phosphorus
content in the main tributary flowing into the reservoir;
this tributary transports 90 percent of the annual
reservoir influent and approximately 90 percent of the
phosphorus entering the reservoir. Figure 2 shows the
site of this plant. After flowing into the pre-reservoir
which serves as a floodwater retention basin, the water
is pumped into the phosphorus elimination plant (PEP)
and treated by precipitation, flocculation, and filtration
which aims at reducing the total phosphorus content to
<10A
-------
274
RESTORATION OF LAKES AND INLAND WATERS
Before the plant began operating, most of the
phosphorus was transported into the reservoir by the
River Wahnbach. The quantities of phosphorus which
reached the reservoir via lateral inflowing tributaries
and precipitation were insignificant. After the plant
was put into operation, it happened that more
phosphorus was transported into the reservoir via
these lateral tributaries (white columns) than had been
transported by the main tributary. Thus the phosphorus
loading of the reservoir caused by the lateral streams
became decisively important.
I
PI
D
fl
n
P- load of Wahnbach
before commencement of PEP
Eliminated
P-load of Wahnbach since
commencement of PEP
P-load due to
overflow of pre-reservoir
P-load from PEP
outlet; input to reservoir
after commencement of PE P
P-load from lateral tributaries
P- load due to precipitation on
reservoir surface
spring overturn P-concentration was below 10 to 15
/jg/l P. To ensure that this concentration was attained
in the Wahnbach Reservoir, it was decided to reduce
the P-concentration in the main tributary to 10/ug/l P.
Today we know that this concentration in the main
reservoir tributary is still too high if the reservoir is to
become oligotrophic to mesotrophic. If one applies
Vollenweider's formula (1976) for estimating the
tolerable Ptot-concentration in all the reservoir inflows
["FT] .considering the average Ptot-concentration in a
reservoir [PT] A = 10///I calculated over a period of 1
year,
[PT]i,c =10(1+ VTW )
one then obtains for ["FT] i,c=20 A//lif the retention time
of the water in the reservoir rw = 1 year.
Three years of operating the plant have shown that
the Ptot-concentration of an average of 100 /ug/l in the
Wahnbach (60-180 //g/l Ptot) has been reduced to an
average of 4 fjg/\ Ptot in the plant's outflow. This
means that the Ptot-concentration of all the inflows into
the reservoir including precipitation was reduced to 16-
20 /ug/l Ptot. This figure corresponds to the calculated
Ptot-concentration of all the inflows.
tot
1,4 —|
1,3 -
',2 -
1,1 -
1,0 -
0,9 -
0,8 -
0,7 -
0,6 -
05-
0.4 -
03 -
0,2 -
0,1 -
0 -
J FMAMJJASONDUFMAMJJASONDUFMAMJJASONDl
1977
1978
1979
Figure 3. — Phosphorus load of Wahnbach Reservoir from
different sources before and after commencement of
phosphorus elimination plant.
THE AIM OF PHOSPHORUS REMOVAL
AT THE MAIN TRIBUTARY
When the pilot plant for removing phosphorus was
installed on the main tributary of the Wahnbach
Reservoir, it was not precisely known how far the P-
concentration in the main tributary had to be lowered
to achieve a tolerable water quality in the reservoir. At
that time one could only rely on Sawyer (1966), who
had found out from practical experience that water
quality caused no problems in those lakes in which the
PRINCIPLE OF THE PHOSPHORUS
ELIMINATION PLANT
The phosphorus elimination plant (Figure 4) is
designed for a maximal flow of 5 mVsec. Thus the
fivefold amount of the long-time average flow of the
River Wahnbach (1 mVsec) can be treated. Together
with the storage capacity of the pre-reservoir which
serves as a water retaining basin with a capacity of
500,000 m3, up to 8 m3/sec can be treated, at least for
a limited period.
The phosphorus elimination plant should meet the
following requirements:
1. It should run for several weeks on full capacity.
2. Rapid variation of flow capacity between 3,000
and 18,000 mVh.
3. Operation for a few hours with intervals of several
days, frequent switching on and off without decreasing
quality of filtrate.
4. No drop of efficiency at water temperatures of 0 °C
(winter running).
5. Treatment of water with high turbidity (up to 100
mg/l content of solids (105°C) without shortening the
duration of filter runs to less than 10 hours at the
maximal filtration rate of 16 m/h.
6. Decrease of the total phosphorus content to values
S5 HQ/\.
1. Treatment should be arranged in such a way that
^ 99 percent of the plankton occurring during the
summer months (max. 400,000 cells/ml) can be
eliminated from the water. Algal cells which break
through the filter cause high concentrations of
undissolved and dissolved organically bound phos-
phorus compounds in the filtrate. This means an
undesired phosphorus load in the reservoir.
8. Removing 99 percent of inorganic turbidity flushed
into the reservoir after the erosion of arable land which
is rich in phosphates.
-------
RURAL WATERSHED POLLUTION CONTROL
275
9. Flocculation is carried out with 4-10 mg/l Fe+3
The total iron concentration in the effluent should not
exceed 50 /ug/l Fe.
10. The plant should be constructed as compactly as
possible to be economically and technically worth-
while. The filter area is 1,100 m2. With a throughput of
4-5 mVsec. one has a filter velocity of 16 m/h.
To achieve all this we developed the energy-input
controlled direct filtration called 'Wahnbach System'
with the following steps (Figure 4):
1. Precipitation of the o-phosphate ions present in
the water by adding iron-lll-ions in the acid pH zone (pH
6.0-7.0, average pH 6.4).
2. Destabilization of the colloids and suspensoids in
the raw water to which the precipitated iron phosphate
products also belong, This is done by adding iron-lll-
ions as a flocculation agent.
3. Agglomeration by means of transport. The
microflocs unite in a subsequent agglomeration step
and form larger, partly visible floes. By adding a cationic
polyelectrolyte they are made suitable for filtering. The
type of cationic polyelectrolyte used changes according
to the time of the year.
Required for both destabilization and agglomeration,
the amount of special energy-input is adapted to the
quality of the raw water and the throughput. The filter
of the phosphorus elimination plant consists of three
layers of various granulations and densities:
1. Upper layer — 30 cm active carbon, granulation 3-
5 mm, effective grain size 3.9 mm.
2. Middle layer — 125 cm hydro anthracite,
granulation 1.5-2.5 mm, effective grain size 1.7 mm.
3. Bottom layer — 50 cm quartz sand, granulation 0.7
-1.2 mm, effective grain size 0.9 mm.
EFFICIENCY OF THE PHOSPHORUS
ELIMINATION PLANT
The quality of the filtrate is illustrated by Table 1. The
average and the minimum and maximum values and
the average elimination are shown. About 95 to 99
percent of the phosphorus compounds are eliminated.
The total phosphorus concentrations in the filtrate
amount to 4 (jg/\ 1 on a 2-year average. They are 60
percent less than the P-concentration we aimed at
achieving.
Agglomeration
o-PO4"*"- Elimination
andDestabilisation
Filtration
Pumpmgstation Fe'
30cm activ carbon
125cm hydro anthracite
50cm quart? sand
6 Pumps of 3000 m3/h
orthokinetic
effect
G-values-50sec~
G-t 20000-50000
Figure 4. — Principle of the direct-filtration with controlled
energy input, 'Wahnbach System'.
The decrease of soluble organic compounds varies
depending on the sum parameter chosen for character-
ization. We have 58 percent elimination of dissolved
organic carbon because there is only 25 percent
elimination of the organic compounds with a molecular
weight of <1000 and 70 to 80 percent elimination of
the organic compounds with a molecular weight
>1000. More than 99 percent of the algae, expressed
as chlorophyll, are eliminated during development
periods. The decrease in turbidity is always higher than
99 percent. The very low remaining turbidity of <0.1
FTU, the small remaining concentrations of bacteria,
algae, and iron (<30 fjg/\) show that the quality of the
filtrate is practically that of drinking water.
The decrease in P-input has caused a considerable
decrease in the P-concentration in the Wahnbach
Reservoir. As a consequence, there is less algal growth
resulting in an increase in the Secchi depth as can be
seen from Figure 5. The years 1969 and 1970 were
chosen for comparison with the conditions after the
commencement of phosphorus elimination as their
hydrological conditions were similar. Not only Secchi
depth increased, but there was also a considerable
change in the composition of algal flora (Figure 6).
Various species of tiny blue-green algae, e.g.,
Coelosphaerium and Aphanothece. completely dis-
appeared after the plant went into operation. Chlorella
which used to grow in large quantities is now present
only in small amounts. In the spring diatoms such as
Asterionella formosa and Melosira dominate.
ESTIMATION OF THE INFLUENCE OF
P-ELIMINATION ON THE TROPHIC
GRADE OF WAHNBACH RESERVOIR
If one tries to classify the Wahnbach Reservoir using
the data from the OECD-Cooperative Program for
Monitoring of Inland Waters (Vollenweider and
Kerekes, in prep.),this reservoir would be with a
probability of more than 50 percent mesotrophic in
1979 (Tables 2 and 3). During years of extensive
Oscillatoria growth (e.g., 1969) with annual average
chlorophyll concentrations of 25 /ug/l, the reservoir
was clearly eutrophic.
If one uses the total phosphorus concentration
(annual average figures) instead of the average
chlorophyll concentration, then the Wahnbach Reser-
voir would be with a probability of more than 50
percent oligotrophic.
If one applies the registered Secchi depths for
classifying the reservoir, then the reservoir would be
classed as oligotrophic or mesotrophic. The reservoir
should be mesotrophic to eutrophic during the years
1969 and 1970 (Figure 5).
One should not forget that the data in Table 2 is
based on the statistical evaluation of a large amount of
data (Vollenweider and Kerekes, in prep.; Vollenweider,
1979). It is worth noting that the OECD cooperative
program showed, for example, that the chlorophyll
concentrations that actually occurred fluctuate to a far
greater extent than the annual average values. This
means that far higher concentrations of chlorophyll
can occur for short periods of time in a mesotrophic
lake. In 1979 peak concentrations of chlorophyll in the
Wahnbach Reservoir were, however, only 10/yg/l.
-------
276
RESTORATION OF UVKES AND INLAND WATERS
Table 1 _ Elimination of several substances by the phosphorus elimination plant (1.10.1977-31.5.1980).
Parameter
PEP-lnflow
min.—max.
x ± s
PEP-Outflow
min.—max.
x + s
Coliforms
bacteria/100 ml
Colony-count (22°C)
colonies/ml
Chlorophyll ijg/\
Turbidity FTU
COD mg/l
Spectral absorp-
tion coefficient
254 nm m~1
DOC mg/l
Total P mg/m3
560
560
433
515
107
561
563
569
0 - 68,000
5,979 ± 8,818
285 - 290,000
12,504 + 20,530
1.0 - 204.3
25.15 + 27.69
0.6 - 48.7
10.4 ± 5.25
3.7 - 22.3
11.13± 4.52
3.4 - 20.8
8.14 ± 2.86
0.9 - 7.3
2.37 ± 0.83
27 - 480
116.5 ± 49.2
479
479
360
515
97
482
484
485
0 - 171
8 ± 15
0 - 17,100
263 ± 1,338
0.1 - 17.3
1.28 ± 1.81
0.01 - 0.8
0.06 ± 0.09
0.1 - 6.3
2.56 + 1.12
0.3 - 4.7
2.40 + 0.68
0.4 - 2.2
1.00 ± 0.30
1 - 13
4.3 ± 1.7
99.87
97.90
94.9
99.3
77.0
70.5
57.8
96.3
Elimination
%
Table 2. — Preliminary classification of trophic state (OECD-Cooperative Program). The geometric mean (based on log 10
transformation) was calculated after removing values x 2 SD.
Oligotrophic
Total Phosphorus
mg/m3
Chlorophyll a
(annual mean values)
mg/m3
Transparency
Secchi depth
m
x
x ± 1 SD
x + 2 SD
Range
x
x + 1 SD
x + 2 SD
Range
x
x ± 1 SD
x ± 2 SD
Range
8.0
4.85-
2.9
3.0
1.7
0.8
0.4
0.3
9.9
5.9
36
5.4
13.3
22.1
17.7
3.4
7.1
4.5
16.5
27.5
28.3
Mesotrophic
26.7
14.5-
7.9-
10.9-
4.7
3.0-
1.9-
3.0-
4 2
2.4 -
1.4 -
1 5-
49.0
90.8
95.6
7.4
11.6
11.0
Eutrophic
48
16.
16
6.
3.
2.
84
-
8-
2-
14
7 -
1 -
7 -
.4
189
424
386
3
31
66
78
2.45
7 4
13.0
8.1
1.
0.
0.
5 -
9-
8-
4.
0
6.7
7.
0
x geometric mean SD standard deviation (shortened from (10, 11})
SUMMARY
The phosphorus elimination plant developed by the
Wahnbach Reservoir Association has been in operation
at the point where the River Wahnbach flows into the
reservoir since the end of 1977. Its operation has
produced very clear water in the main reservoir and an
average total phosphorus concentration of below 10
//g/l This value was only exceeded for short periods
of time, particularly when the inflow was higher than
the retention capacity of the pre-reservoir and the
efficiency of the plant.
The decrease in total phosphorus resulted in a shift
in the algal species from blue-green algae to green
algae and then to diatoms. This shift in species was
typical of the change in the trophic state from eutrophic
to oligotrophic-mesotrophic. Whereas blue-green algae
have disappeared almost completely from the reservoir,
the population of green algae has been reduced to such
an extent that they have no dominating influence
compared with diatoms. The clear decrease in the
phosphorus input into the reservoir has caused a
change in the trophic state of the lake which was
eutrophic.
-------
RURAL WATERSHED POLLUTION CONTROL
277
Table 3. — Classification of trophic state of the Wahnbach Reservoir
(annual mean values (x)).
Oligotrophic Mesotrophic Eutrophic
Total Phosphorus mg/m3
1969
1970
1977
without PEP
25
26
16
1 978 PEP in operation
1979
Chlorophyll a mg/m3
1969
1970
9
6
25
11
1977
1978
1979
Transparency m
Secchi depth
1969
1970
1977
1978
1979
7
8
5
3
3
5
6
6
0 Oscillatoria
M Melosira
S Synura
A Astenonella
^.lMF>y;.^AM.I.it.^lMl.m^f..?*.J|Mlm^lLmlM.1,il^'.^lP.^.<.9l
flood
REFERENCES
Bernhardt, H. 1978. Water control in lakes and reservoirs.
Pages 313-225 in Prog. Water Technol. Vol. 10. Pergamon
Press, Great Britain.
Bernhardt, H., and J. Clasen. 1973. Die Aufbereitung
planktonreicher Talsperrenwasser zu Trinkwasser. Fort-
schritte der Wasserchemie und ihre Grenzgebiete 15:137.
Bernhardt, H., and 0. Hoyer. 1979. Characterization of
organic water constituents by the kinetics of chlorine
consumption. Pages 110-137 in Oxidation technique in
drinking water treatment. Drinking Water Pilot Proi. Report
IIA, Karlsruhe, FRG. EPA-570/9-79-020.
Bernhardt, H., and H. Schell. 1979. The technical concept of
phosphorus-elimination at the Wahnbach estuary using
floe-filtration. Zeitschr. fur Wasser- und Abwasserfor-
schung 12:78.
Bernhardt, H., and A. Wilhelms. 1978. Der Einfluss
algenburtiger organischer Verbindungen auf den Floc-
kungsprozess bei der Trinkwasseraufbereitung. Pages 112-
146 in Organische verunreinigungen inder umwelt. Erich
Schmidt Verlag, Berlin.
Bernhardt, H., J. Clasen, and H. Schell. 1971. Phosphate and
turbidity control by flocculation and filtration. Jour. Am.
Water Works Assoc. 63:355.
Bernhardt, H., et al. 1978. Phosphor — wege und verbleib in
der Bundesrepublik Deutschland. Verlag Chemie Wein-
heim, New York.
Sawyer, C. N. 1966. Basic concepts of eutrophication. Jour.
Water Pollut. Control Fed. 38:737.
Vollenweider, R. A. 1976. Advances in defining critical
loading levels for phosphorus in lake eutrophication. Mem.
1st. Ital. Idrobiol. 33:53.
1979. Das nahrstoffbelastungskonzept als grun-
dlage fur den externen eingriff in den eutrophierung-
sprozess. Zeitshcr. f. Wasser- und Abwasserforschung
12:10.
Vollenweider, R. A., and J. Kerekes. In preparation.
Cooperative programme for monitoring of inland waters
(eutrophication control). Synthesis Rep.
flood
Figure 5. — Secchi-depths in the Wahnbach Reservoir before
(1969/70) and after (1978/79) the begin of operation of the
plant.
1975 1976 1977 1978 1979
cells /ml 1,6
1,2
0,8.
0,4-
••*. /
j
I
"•~\J/"''
Coelosphaerium
V _
i
cells/ml i2o^I
80
40
:
~' ^ •
^''•W/'^
'/
' ?•/. / •:
V
\vt. !
Aphanothece
f':+A commencement ot PEP
f^,.
n v._
cells/ml 20] V 1 :,.; I
10 1? x.^f~jf "'-L_
values X1000 o il I, ,
Chlorella
I
Figure 6. — Change in the occurrence of species of green and
blue-green algae after the plant began to operate (values x
1000).
-------
278
AGRICULTURAL POLLUTION CONTROL IN THE
NETHERLANDS
H. L GOLTERMAN
Biology Station
La Tour du Valat le Sambuc
Aries, France
ABSTRACT
The work of a Dutch Royal Commission to prepare "an inventory of agricultural pollutants disposed
of - purposefully or inadvertently - into aquatic ecosystems" has taken some 8 years. I chaired a
working group to quantify the disposal of chemical pollutants with manure and artificial fertilizers.
The major efforts were directed to nitrogen- and phosphate pollution. In the study the peculiar
structure of Dutch agricultural land had to be taken into account. The largest part lies below sea
level with groundwater tables often 10 to 30 cm below soil surface; for agricultural use the
groundwater table must be maintained at a fixed level. This means an export of rain water by
pumping during winter and inlet of riverwater - mainly Rhine water - during summer. Thus the
phosphate of the Rhine accounts for 50 percent of the input in the phosphate balance of Dutch
waters. Other sources of phosphate come from layers of peat. Therefore, no reliable estimate could
be made of the agricultural contribution because there are no areas where the (semi)natural input
can be measured or quantified. However, partly due to the phosphate holding capacity of the soils
the impression was obtained that neither manure nor artificial fertilizers contribute significantly to
the phosphate input. This situation is completely different in the higher sandy soils, where
intensive husbandry of cattle is performed. Considerable quantities of phosphates enter the waters
in these regions. The situation is again different for the nitrogen balance. Considerable quantities
of nitrogen reach the canals, lakes, and rivers, both in the form of ammonia and nitrate. Quantitave
assessment of these data was not possible. The same difficulties as for the phosphate studies
were met, while possible denitrification in the soil appeared to be an unknown factor of some
importance. No proposals have been formulated for a control of these inputs. A manure balance for
the whole country was established; no excess of manure seems to exist.
For the complete paper, please contact Dr. Golterman at the following address.
Dr. H. L. Golterman
Station Biologique D
La Tour du Valat le Sambuc
F-13200 Aries, France
Phone: (90) 98. 90. 13
-------
279
URBAN STORMWATER/COMBINED SEWAGE
MANAGEMENT AND POLLUTION ABATEMENT
ALTERNATIVES
RICHARD P. TRAVER
Municipal Environmental Research Laboratory
U.S. Environmental Protection Agency
Edison, New Jersey
ABSTRACT
Overflow points are the built-in inefficiencies of combined sewers. Untreated overflows from
combined sewers are a serious water pollution source during both wet and dry weather periods. In
urban areas, the principal nonpoint source concern is stormwater management. A nationwide
survey of public works officials in 1976 identified urban flooding and its associated pollutants
caused by inadequate storm sewers as the number one urban problem. As suburban land
continues to be developed, the problems with separate storm and combined sewer systems will
increase. Although stormwater runoff and combined sewer overflows typically occur only during
brief periods, the quantities of sediment, nutrients, chemicals, and toxic metals dumped into
streams during the storm periods dwarf the quantities of such materials released by the municipal
treatment plants throughout the entire year. This problem has serious implications for
communities using the streams for water supply as well as for other downstream users. Urban
runoff management is a continuous process. Essential to its success is a constant process of
innovation, demonstration, assessment, implementation guidance, and active program feedback.
This paper reviews the innovative technology available today for implementation in our Nation's
fight to protect and preserve our recreational receiving waters.
THE PROBLEM
Overflow points are the built-in inefficiencies of
combined sewers. Untreated overflows from combined
sewers are a serious substantial water pollution source
during both wet and dry weather periods. Nationwide,
there are roughly 15,000 to 18,000 combined sewer
overflow points.
In urban areas, the principal nonpoint source
concern is stormwater management. A nationwide
survey of public works officials conducted in 1976
identified urban flooding and its associated pollutants
caused by inadequate storm sewers as the number one
urban problem. As suburban land continues to be
developed, the problem will increase. In response, a
few urban areas have initiated programs to improve
stormwater management.
Although combined sewer overflows and stormwater
runoff typically occur only during brief periods, the
quantities of sediment, nutrients, chemicals, and toxic
metals dumped into streams during the storm periods
dwarf the quantities of such materials released by the
municipal treatment plants throughout the entire year.
This problem has serious implications for communities
using the streams for water supply as well as for other
downstream users.
Urban stormwater management is a continuous
process. Essential to its success is a constant process
of innovation, demonstration, assessment, implemen-
tation guidance, and active program feedback. Based
upon January 1978 dollars, the total national needs to
control pollution from combined sewer overflows were
approximately S21.16 billion (U.S. EPA, 1978). Such a
control program must be founded on proven capabili-
ties, comparable methodologies and assessment
criteria, an expanding data base, and a continuous,
effective technology transfer.
Because of the unique nature of urban runoff
abatement technology, control and/or treatment of
storm sewer discharges and combined sewer over-
flows is a major problem in water quality management.
Over the past 14 years much research has generated a
large amount of data, primarily through the actions and
support of the EPA's Storm and Combined Sewer
Section.
Every metropolitan area of the United States has a
stormwater problem, whether served by a combined
sewer system (approximately 29 percent of the total
sewered population) or a separate sewer system (Lager
and Smith, 1974).
The problem is best quantified when discharges are
compared on the basis of mass loadings released over
discrete periods of time encompassing one or several
consecutive storm events. In many cases, however,
aesthetics or beneficial uses (such as maintaining
receiving water quality above body contact use
standards) are of primary concern.
Each metropolitan area should, therefore, be directly
involved in setting its goals for a stormwater
management program.
-------
280
RESTORATION OF LAKES AND INLAND WATERS
URBAN RUNOFF CHARACTERIZATION
Figure 1 illustrates representative strengths of
wastewaters. The average 5-day biochemical oxygen
demand (BOD5.)concentration in combined (domestic
and storm) sewer overflow is approximately one-half
the raw sanitary sewage BODs However, storm
discharges must be considered in terms of their
shockloading effect. A common rainfall can produce
flow rates up to 100 times dry-weather flow. Even
separate stormwater is a significant source of
pollution, having solids concentrations equal to or
greater than untreated sanitary wastewater, and
BODV.s approximately equal to secondary effluent.
Bacterial contamination of separate stormwater is two
to four orders greater than concentrations considered
safe for water contact (Field, Tafuri, and Masters,
1977).
Because flow quantities are high, control — whether
through flow balancing, multiple uses of facilities,
runoff retardation, or combinations thereof — is the
focus of cost-effective planning.
^| RAW
t22 COMBINED
I I STORM
RAW
COMBINED
I I SIORM
TOTAL COLIFORM TOTAL TOTAL
MPN/100 ml NITROGEN PHOSPHORUS
Figure 1. — Representative strengths of wastewaters (flow
weighted means in mg/l).
THE ATTACK
The existing tools for reducing urban runoff pollution
provide many-faceted approach techniques to individu-
al situations. These tools are constantly being
increased in number and improved upon as part of a
continuing research and development program guided
by the EPA Storm and Combined Sewer Section.
Continuing progress is being made in the variations
and refinements of storage concepts. From the
sophisticated computer controlled systems utilizing in-
line storage capacities as found in Seattle and Detroit;
monumental undertakings as the Chicago Tunnel and
Reservoir Plan; Chippewa Falls off-line storage basin;
Akron's underground void space storage; Sandusky,
Ohio's underwater storage bag; and most recently, the
evaluation of using static flow energy dissipators
coupled with small off-line storage tanks and bulk-
headed interceptors.
Simplified mathematical models based upon the
general storage equation and operated off real
(continuous) rainfall data provide an excellent tool for
equating the effectiveness of alternate storage vol-
umes and treatment rates.
Control and treatment of stormwater introduce many
unique operation and maintenance requirements.
These include automated control, startup and shut-
down procedures, maintenance and surveillance
between storms, and solids handling and disposal.
Much emphasis is currently being placed on
controlling stormwater pollution by attacking the
problem at its source, as opposed to potentially more
costly downstream treatment facilities. These source
controls, termed Best Management Practices (BMP),
can either be directed toward planning control for
further development or redevelopment efforts. The
program has instituted research in using natural
drainage features, erosion controls, operation and
maintenance practices, highway deicing, street sweep-
ing, collection system and catchbasin maintenance,
and most recently, sewer flushing during dry weather
to reduce receiving water impacts from first flush loads
during storms.
Management alternatives for stormwater pollution
abatement are generally categorized into four areas:
Source control, collection system control, storage and
treatment, and integrated (complex) systems.
SOURCE CONTROL
Source controls are defined as those measures for
reducing stormwater pollution that involve actions
within the urban drainage basin before runoff enters
the sewer system. Examples include planning surface
flow attenuation, using porous pavements, controlling
erosion, restricting chemical use, and improving
sanitation practices (street cleaning, more frequent
refuse pickup, etc.).
Planning
Preventing and reducing the source of stormwater
pollution best applies to developing urban areas, where
man's encroachment is yet minimal, or at least
controllable, and drainage essentially conforms to
natural patterns. Such lands offer the greatest
flexibility in preventing pollution. They must be
developed in such a way that runoff remains close to
natural levels. In these new areas proper management
can prevent long-term problems.
-------
URBAN AND POINT SOURCE POLLUTION CONTROL TECHNOLOGY
281
On-site Storage of Runoff
The objective of on-site storage of runoff is either to
prevent storm flow from reaching the drainage system
or to change the timing of the runoff by controlling the
release rate. Retention is the term for total contain-
ment, and detention is the common term for controlling
the release rate to smooth out the peak flows.
The precipitation/infiltration process is the most
important method of replenishing the groundwater
reservoirs that serve as potable water supplies for
many areas of the country. The decreased infiltration
and increased water demand caused by urbanization
will stress groundwater supplies unless recharge areas
are set aside as basins develop. Although large-scale
urban stormwater recharge programs have not been
implemented because of potential groundwater pollu-
tion, on-site retention and recharge have been
developed for small watersheds. Retention basins are
usually variable-depth ponds designed with no outlet or
only a bypass for exceptionally high flow conditions.
Retention is also practiced as controlled on-site
storage where groundwater recharge is not important.
In a typical example, the California Division of
Highways has built retention basins to dispose of
highway runoff in the San Joaquin Valley. These
basins were developed from 0.4 to 2.4 hectare
depressions that had originally been excavated for
embankment material. Infiltration capacity is some-
times improved by excavating 1.8 to 3.1 meter deep
trenches or vertical drains and backfilling with porous
material. Maintenance is minimized by providing low-
velocity channels ahead of the basins to help settle
suspended particles. The areas are scarified once a
year to decrease the surface clogging effects of organic
solids.
A demonstration in Cleveland, Ohio (funded by
Region V Great Lakes National Program Office with
technical guidance from SCSS) is trying to obtain a
quanitity and quality control on a portion of a combined
sewer system by reducing overflows to receiving
waters during rainfall events, and reducing residential
basement flooding caused by combined sewer sur-
charging {U.S. EPA, 1970). Stormwater runoff is
prevented from entering the already overburdened
combined sewer as shown in Figure 2. Based upon
applying the Dorsch HVM computer model to simulate
runoff and backwater effects for each sub-catchment
area, four upstream off-line storage tanks have been
strategically placed to detain stormwater from entering
the already surcharged combined sewer. The four units
are constructed of corrugated structural steel pipe and
measure: 2 tanks 156' x 87" x 63"; 163' x 48"; 170' x
93" x 67". Catchbasins are directing surface and street
runoff into the tanks from which the stormwater will be
discharged at a controlled rate into the combined
sewer. The flow rate of the discharge will be regulated
by a Hyrdrobrake internal energy dissipator. This small
device located at the downstream end of the tank will
deliver virtually a constant discharge rate regardless of
head variations. This is accomplished without moving
parts or external energy sources.
Approximately 22 3 and 4 " diameter Hydrobrakes
will be placed in existing catchbasins upstream from
the larger detention tanks to maximize storage
capacities and surface ponding.
Porous Pavement
An interesting technological answer to the problem
of preserving pervious area is paving with an open
graded asphaltic concrete. Experiments have shown
that it will serve as a porous pavement, allowing as
much as 64 cm/hr of stormwater to infiltrate through
the pavement (see Figure 3).
Preliminary investigations have shown that this
material can withstand stability, durability, and freeze-
thaw tests, and that it compares in cost with
conventional paving with drainage. Long-term tests
will have to be made of its resistance to clogging and
the effects on the quality of water that filters through
the pavement. If the soil under the pavement and base
is free draining, the rainwater will infiltrate quickly into
the ground; however, porous pavement can also serve
as a ponding device if storm quantities exceed soil
capacity. The porus nature of the pavement permits
water to be stored in the pavement. A pavement with a
10 cm surface course and 15 cm base course could
store 6.1 cm of runoff in its voids (Thelan, et al. 1972).
The proven use of porous pavement can be an
important tool in preserving natural drainage.
EXCEEDED THE
MINIMUM MAR-
SHALL STABILITY
CRITERION FOR
MEDIUM TRAFFIC
USES
AEROBIC ACTIVITY
UNDER PAVEMENT
NOT IMPAIRED
DURABILITY TEST
INDICATED THAT
HEIGHTENED EX-
POSURE TO AIR OR
WATER DID NOT PRO-
DUCE ASPHALT
HARDENING
AGGREGATE GRADED TO ALLOW
A WATER FLOW OF 76"/HOUR
5.5% BYWT.OF
85-100 PENETRATION
ASPHALT CEMENT
BINDER
SUBJECTED TO 265
FREEZE-THAW CY-
CLES WITH NO
CHANGES IN PHYS-
ICAL DIMENSIONS
MARSHALL STABILITY
VALUES OR FLOW
RATES
Figure 2. — Stormwater detention tank with hydrobrake.
Figure 3. — Porous asphaltic-concrete features.
-------
282
RESTORATION OF LAKES AND INLAND WATERS
Surface ponding is the most common form of
detention being used by developers. In most cases, the
facilities are carefully planned so that the ponding area
is a dual-use facility that enhances the value of the site.
Variable level ponds have a permanent water level
during dry weather and increased holding capacity
during storm conditions. The permanent lakes have
aesthetic and recreational appeal, increasing lot
values. Basins that are dry between storms are often
designed to be used as baseball fields, tennis courts,
and general open space. Parking lots can serve as low-
depth storage ponds by sloping the sides and
constructing drain outlets. Side slopes are restricted to
about 4 percent for traction in the winter, and the pond
depth is limited by the need for people to reach their
vehicles during storm events. Obviously, a truck
terminal lot can be allowed to pond to a greater depth
than a supermarket lot. Table 1 indicates various
surface ponding locations.
EROSION CONTROLS
Controlling erosion from construction and developing
sites will have a major impact on the total pollution
loads imposed on receiving waters. Current estimates
indicate that approximately 3,900 km2 (1,500 mi2) of
the United States is urbanized annually. All of this land
is exposed to accelerated erosion (White and Franks,
1978).
From a knowledge of erosion and the guidelines that
have been written concerning erosion control, several
basic principles for control of erosion are apparent:
1. Reduce the area and duration of soil exposure.
2. Protect the soil with mulch and vegetative cover.
3. Reduce the rate and volume of runoff by
increasing infiltration rates and surface storage and by
diverting excess runoff.
4. Dimmish runoff velocity with planned engineering
works.
5. Protect and modify drainage ways to withstand
concentrated runoff resulting from paved areas.
6. Trap as much sediment as possible in temporary or
permanent sedimentation basins.
7. Maintain completed works and assure frequent
inspection for maintenance needs.
These principles can be implemented by a variety of
simply constructed facilities. Detailed descriptions and
design criteria are available in the literature. Costs for
some of the basic erosion control alternatives are
presented in Table 2.
Chemical Use Control
One of the most often overlooked measures for
reducing pollution from stormwater runoff is reducing
the indiscriminate use and disposal of toxic substances
such as fertilizers, pesticides, oil, gasoline, and
detergents.
Table 2 — Erosion control costs per developed acre.
First year
Initial place- maintenance
ment cost, cost,
Vegetative measures $/acre $/acre
Seeding: seedbed preparation, seed and
application, mulching at 2 tons/acre
Temporary seeding by machine 240-330 50-120
Temporary seeding by hand 335-415 50-120
Permanent seeding by machine 790-1,220 50-120
Sodding, including seedbed preparation 2,400-3,600 240-2,900
Mulch, 2 tons/acre
By hand
By machine
120-140
90-120
_.
Mechanical measures
Earth diversion berms
Straw bale barriers
0.15-0.30
0.75-1.10
1 20-3.60
1 20-3.60
Silt basins with earth dam, watershed
area
2 acres to 5 acres
25 acres to 100 acres
100 acres to 200 acres
600-1,200 500-750
1,200-3,500 750-1,200
3,500-5,000 1,200-1,800
$/acre x 2 469 = $/ha
acre x 0,405 - ha
tons/acre x 2240 - kg/ha
Table 1. — Surface ponding.
Site
Description
Cost estimate, $
With surface Without surface
ponding ponding
Earth City, Missouri
Consolidated Freightways,
St. Louis, Missouri
Ft Campbell, Kentucky
Indian Lakes Estates,
Bloomington, Illinois
A planned community including
permanent recreational lakes
with additional capacity for
Storm flow
A trucking terminal using its
parking lot to detain storm
flows
A military installation using
ponds to decrease the required
drainage pipe sizes
A residential development
using ponds and an existing
small diameter drain
2,000,000
115,000
2,000,000
200,000
5,000,000
150,000
3,370,000
600,000
-------
URBAN AND POINT SOURCE POLLUTION CONTROL TECHNOLOGY
283
Operations such as tree spraying, weed control, and
fertilization of parks and parkways by municipal
agencies, and the use of pesticides and fertilizers by
individual homeowners can be controlled by increasing
public awareness of the potential hazards to receiving
waters, and providing instruction as to proper use and
application. In many cases over-application is the major
problem, where moderate use would achieve equal
results. The use of less toxic formulations is another
alternative to minimize potential pollution. Direct
dumping of chemicals, crankcase oil, and debris into
catchbasins, inlets, and sewers is a significant problem
that may only be addressed through educational
programs, ordinances, and enforcement (Am. Publ.
Works Assoc. 1969).
Street Sweeping
Street sweeping is used by most cities to remove
accumulated dust, dirt, and litter from street surfaces,
but cleaning is usually done for aesthetic reasons. In
many neighborhoods the amount of paper tolerated by
the public governs cleaning frequencies. Street
cleaning practices have been shown to be an effective
way to attack the source of stormwater-related
pollution problems.
Removal rates as reported in the literature vary
considerably. In one study, the range was from 11 to 62
percent of the initial solids loading (McGuen, 1975). In
another study, overall removal has been estimated at
33 percent of all pollutants on the street surface
(McPherson, 1976).
Litter Control
Discarded containers from food and drink, cigarettes,
newspapers, sidewalk sweepings, lawn trimmings, and
a multitude of other materials become street litter.
Unless this material is prevented from reaching the
street or is removed by street cleaning equipment, it
often is found in stormwater discharges. Enforcement
of antilitter laws, convenient location of sidewalk waste
disposal containers, and public education programs are
just some of the source control measures.
COLLECTION SYSTEM CONTROL
Collection system control includes all alternatives
pertaining to collection system management. Examples
include inflow/infiltration control, the use of improved
regulator devices, temporarily increased line-carrying
capacities using polymer (friction reducing) flow
additives, catchbasin maintenance, sewer separation,
the use of remote monitoring/control systems, and the
flushing of combined sewers during dry-weather
periods.
Detailed knowledge of how collection systems
respond to wet-weather flow is almost universally
lacking in municipalities today. As a result, demonstra-
tion projects frequently reveal previously unknown
relief points and crossovers critical to proper function-
ing. Such conditions emphasize the need for early and
intensive monitoring .and modeling for predictive
responses.
Inflow and Infiltration
Extraneous flows entering a sewer can be generally
categorized as either inflow or infiltration. Inflow
usually occurs from surface runoff via roof connec-
tions, cross connections between sanitary and storm
sewers, yard drains, or flooding of manhole covers.
Infiltration usually occurs by water seeping into the
pipe or manholes from leaky joints, crushed or
collapsed pipe segments, leaky lateral connections, or
other pipe failures. By reducing effective collection
system and treatment plant capacities, extraneous flow
may cause unnecessary pollution (Sullivan, et al.
1977). Table 3 presents rehabilitation cost estimates.
Table 3. — Rehabilitation cost estimates for inflow elimination.
Inflow source
Leakage around
manhole covers
Holes in man-
hole covers
Foundation drains
Roof leaders
Cross connection
Catchbasin
Flowrate,
gal/min
10-20
50-100
10
10
250-450
300
Rehabilitation
cost (ENR 2000), $
50-75
100-125
300:1200
50-75
100-500
3000-5000
Ditch or storm sewer-
infiltration sanitary
sewer (per manhole reach) 60-80
Area drains
50-200
500-2500
50-350
gal/min x 0.0631 = L/s
Stormwater Regulations
The swirl regulator/concentrator is of simple
angular-shaped construction and requires no moving
parts. An isometric view of the final form of the device
is shown in Figure 4. Again, the swirl provides a dual
function: regulating flow by a central circular weir
spillway while simultaneously treating combined
wastewater by swirl action, separating solids from
liquid. Dry-weather flows are diverted through a
cunette-like channel in the floor of the chamber into
the bottom orifice or foul underflow (located near the
water downshaft) to the intercepting sewer for
subsequent treatment at the municipal plant. During
higher flow storm conditions, the low-volume con-
centrate (3 to 10 percent total flow) is diverted via the
same bottom orifice leading to the interceptor, and the
excess, relatively clear, high-volume supernatant
overflows the central circular weir into a downshaft for
storage, treatment, or discharge to the stream. This
device is capable of functioning efficiently over a wide
range of combined sewer overflow rates and can
separate settleable lightweight matter and floatable
solids at a small fraction of the detention time normally
-------
284
RESTORATION OF UVKES AND INLAND WATERS
required for sedimentation. Figure 5 shows estimated
costs for the swirl and helical bend units (Masters and
Field, 1977).
The helical bend flow regulator is based on the
concept of using the secondary helical motion imparted
to fluids at bends, employing a total angle of
approximately 60 degrees and a radius of curvature
equal to 16 times the inlet pipe diameter (Sullivan, et
al. 1975).
Figure 6 illustrates the device. The basic structural
features of the helical bend are: The transition section
from the inlet to the expanded straight section before
the bend; the overflow side weir and scum baffle (not
shown); and the foul outlet for concentrated solids
removal and controlling the amount of underflow going
to the treatment works.
Dry-weather flow goes through the lower portion of
the device and outlets to the intercepting sewer and
onto treatment. As the liquid level increases from
storm conditions, secondary helical motion begins and
the polluted solids are drawn to the inner wall and drop
to the lower level of the channel leading to the
treatment plant. As with the swirl, the proportion of
concentrated discharge will depend on the particular
design. The relatively clean combined sewer overflow
passes over a side weir and discharges to the receiving
waters, storage and/or subsequent treatment. Float-
ables are prevented from overflowing by a scum baffle
along the side weir; they collect at the end of the
chamber and are conveyed to the treatment plant when
the storm flow and liquid level subside.
The hydraulic model studies of the helical bend
regulator indicate that this flash method of solids
removal can efficiently remove settleable solids with
reasonably sized units and without using mechanical
appurtances. Although its costs are greater than the
swirl, structural and hydraulic head requirements may
render it more appropriate.
LEGEND
j Inlet ramp
h flow deflector
c Scum ring
d Overflow weir and
e Spoiler*
I Floalables trap
g Foul tewer outlet
h floor fullers
i Dowmhall
Z
Helical Separator
100% Grit Removal
Swirl Concentrator
100% Grit Removal
Swirl Concentrator
90% Grit Removal
Dijcharqe - CFS
Figure 5. — Estimated construction costs — helical bend and
swirl concentrator regulator.
CHANNEL FOP
OVERFLOW
TRANSITION SECTION
\ 15D
STRAIGHT
'\ SECTION
50
' HELICAL
BEND 60"
R - 160
NOTES
2. Dry-weather flow shoi
OUTLET TO PLANT
Figure 4. — Isometric view of the swirl.
Figure 6. — Isometric view of helical bend regulator.
Catchbasin Maintenance
A catchbasin is defined as a chamber or well, usually
built at the curbline of a street, for admitting surface
water to a sewer or subdrain; at its base is a sediment
sump designed to retain grit and detritus below the
point of overflow. The distinction is made between
catchbasins as devices which intentionally trap
sediment and storm inlets which do not have sumps
and as a result should not retain sediment.
Historically, the role of catchbasins was to minimize
sewer clogging by trapping coarse debris and to reduce
odor emanations from low-velocity sewers by providing
a water seal. With improvements in steet surfacing and
design for self-cleaning velocity in sewers, their
benefits were considered marginal as far back as 1900.
Despite the purported reduced need, catchbasins are
-------
URBAN AND POINT SOURCE POLLUTION CONTROL TECHNOLOGY
285
still widely used. Catchbasins receive pollutants
through the washoff of street surfaces and deliberate
dumpings of crankcase drainings, leaves, grass
clippings, pet feces, etc. (Lager, et al. 1977a)
Cleaning methods fall into four main categories:
Hand cleaning, bucket cleaing, educator cleaning, and
vacuum cleaning. Comparison of American Public
Works Association survey data from 1959 to 1973
show that, on a national basis, the median cleaning
frequency has decreased from twice per year to once
per year. This trend is obviously detrimental from a
water quality aspect; many problems associated with
catchbasins may be traced to inadequate maintenance.
In general, catchbasins should be used only where
there is a solids transporting deficiency in the
downstream collection drains or at specific sites where
surface solids are unusually abundant (such as beach
areas, construction sites, unstable embankments, etc.).
The advantages to be considered in converting existing
catchbasins to inlets are (1) a direct reduction in the
"first flush" pollutant load, (2) a reduction in required
maintenance, and (3) the opportunity to relocate the
conserved labor. Where catchbasins are required, they
should be cleaned more often than once a year to limit
the sediment buildup to 40 to 50 percent of the sump
capacity. Figure 7 shows average solids removal
efficiencies for catchbasins and the recommended
design configuration.
Figure Recommended design.
0.19 - 2.0 ••
o.io - 2.0 ««
234587
BASIN INFLOW. It3/!
2 1 < 5 I 7
BASIH INFL01. M3 'I
Figure 7. — Solids removal efficiencies.
SEWER FLUSHING
Regular flushing of sewers can ensure the continu-
ing capability of sewer laterals and interceptors to carry
their design capacity as well as alleviate the solids
buildup that pushes solids into overflow. Sewer
flushing can be particularly beneficial on sewers with
very flat slopes (i.e., too flat for average flows to
maintain sand and grit particles, with their associated
contaminants in suspension at all times). If a small
quantity of water is discharged through these flat
sewers periodically, small accumulations of solids can
be washed from the system. This cleaning technique is
generally effective only on freshly deposited solids
(Pisano, et al. 1979).
Internal automatic flushing devices have been
developed for sewer systems. An inflatable bag is used
to stop flow in upstream reaches until a volume
capable of generating a flushing wave is accumulated.
When the correct volume is reached, the bag is deflated
by a vaccum pump releasing impounded water.
STORAGE
Storage facilities possess many attributes desired in
stormwater treatment: (1) They may equalize flow and,
in the case of tunnels, provide flow transmission; (2)
they respond without difficulty to intermittent and
random storm behavior; (3) they are relatively
unaffected by flow and quality changes; and (4)
frequently, they can be operated with regional dry-
weather flow treatment plants for benefits during dry-
and wet-weather conditions.
Storage facility variations include concrete holding
tanks, open basins, tunnels, underground and under-
water containers, underground "silos," granular
packed beds (void space storage), abandoned facilities,
and existing sewer lines.
System controls using in-line storage represent
promising alternatives in areas where conduits are
large, deep, and flat (i.e., backwater impoundments
become feasible) and interceptor capacity is high.
Reported costs for storage capacity gained in this
manner range from 10 to 50 percent of the cost of
similar off-line facilities. Because system controls are
directed toward maximum utilization *of existing
facilities, they rank among the first alternatives to be
considered.
Constructing new separate sanitary sewers to
replace existing combined sewers largely has been
abandoned because of the enormous cost, limited
effectiveness, inconvenience to the public, and extend-
ed time required for implementation.
Costs associated with in-line storage systems are
summarized in Table 4. Costs include regulator
stations, central monitoring and control systems, and
miscellaneous hardware (Lager, et al. 1977b).
Off-line storage is used to attenuate storm flow
peaks, reduce storm overflows, and capture the first
flush, or provide treatment in the form of sedimentation
when storage capacity is exceeded. Off-line storage
facilities may be located at overflow points or near dry-
weather treatment facilities, depending on the type and
function of the storage facility to be used. Off-line
storage may also be used for on-site storage of runoff.
Table 5 presents costs of off-line storage facilities
(Lager, et al. 1977b).
Disadvantages of storage facilities include their large
size, high cost, and dependency on other treatment
facilities for processing the retained water and settled
solids.
-------
286
RESTORATION OF LAKES AND INLAND WATERS
Table 4. — Summary of in-line storage costs".
Location
Storage
capacity
Mgal
Drainage
area,
acres
Capital
cost, $
Storage
cost,
$/gal
Cost per
acre,
$acre
Annual operation
and maintenance
$/yr
Seattle, Washington
Control and
monitoring system
Automated
regulator stations
Minneapolis-St. Paul
17.£
13,120
3,500,000
3,900,000
7,400,000
0.42
564
73,000
219,200
292,000
NA = not available
a. ENR 2000
$/acre x 2.47 = $/ha
$/gal x 0.264 = $/L
Mgal x 3785 = m3
NA 64 000 3 000 000 47
140 89600 2810000 002 31
PHYSICAL TREATMENT ALTERNATIVES
Physical treatment alternatives are primarily applied
to remove suspended solids from wastestreams, and
are of particular importance to storm and combined
sewer overflow treatment to remove settleable and
suspended solids and floatable material. Physical
treatment systems have demonstrated they can handle
high and variable influent concentrations and flow
rates and operate independently of other treatment
facilities, with the exception of treating and disposing
of the sludge/solids generated from these faciliites.
The principal disadvantage is when equipment sits idle
during dry weather. When implemented on a dual use
basis as either pretreatment or effluent polishing of
conventional sanitary sewage treatment plant flows,
capital investment may be reduced by continuously
using the physical treatment system.
Physical treatment processes that have been
demonstrated on either a pilot or prototype scale
include: Sedimentation and chemical clarification;
solids concentration and flow regulation (swirl con-
centrator/flow regulator); screening; dissolved air
flotation; high rate filtration; and a relatively new
process, magnetic separation (Allen and Sargent,
1978). Many prototypes employ combinations of these
processes to form integrated treatment systems, or use
physical treatment processes in conjunction with
biological and disinfection to produce desired water
quality goals. Table 6 shows various removal efficien-
cies for physical treatment.
Biological Treatment
Biological treatment of wastewater, used primarily
for domestic and industrial flows, produces an effluent
of high quality at comparatively low cost. For treatment
of storm flow, however, the following are serious
drawbacks: (1) The biomass used to assimilate the
waste constituents must either be kept alive during
times of dry weather or allowed to develop for each
storm event; and (2) once developed, the biomass is
highly susceptible to washout by hydraulic surges and
organic overload.
Examples of biological treatment applications to
stormwater include (1) the contact stabilization
modification of activated sludge, (2) high-rate trickling
filtration, (3) bioadsorption using rotating biological
contactors, and (4) oxidation lagoons of various types.
The first three are operated conjunctively with dry-
weather flow plants to supply the biomass, and the
fourth approaches total storage of the flows (detention
times of 1 to 10 days). Table 7 summarizes various
biological treatment installations.
Integrated (Complex) Systems
The most promising approaches to urban storm flow
management involve the integrated use of control and
treatment systems with an areawide, multi-disciplinary
(water use, land use, wet- and dry-period discharges,
etc.) perspective.
Storm flow treatment processes can be most
effectively used following some form of storage (flow
equalization). This yields not only longer running
periods, reduced shock effects, and buffer flexibility for
startup and shutdown, but also, frequently, lower
overall costs.
SUMMARY
Nonstructural and low structurally intensive alter-
natives offer considerable promise as the first line of
action to control urban runoff pollution. By treating the
problem at its source, or through appropriate legisla-
tion curtailing its opportunity to develop, multiple
benefits can be derived. These include lower cost,
earlier results, and an improved and cleaner neighbor-
hood environment.
The greatest difficulty faced by BMP's is that the
action-impact relationships are almost totally un-
quantified. It is clear that on-site storage, for example,
can be closely related to reduced downstream conduit
requirements but the net water quality benefits are far
-------
URBAN AND POINT SOURCE POLLUTION CONTROL TECHNOLOGY
287
Location
Akron, Ohio (21)
Milwaukee,
Wisconsin (13)
Humboldt Avenue
Boston,
Massachusetts
Cottage Farm
Detention and
Chlorination
Station (17)"
Charles River
Marginal Conduit
Project (19)
New York City
New York (22, 23, 25)
Spring Creek
Auxiliary Water
Pollution
Control Plant
Storage
Sewer
Chippewa Falls,
Wisconsin (18)
Storage
Treatment
Chicago, Illinois
(2, 11,26)
Tunnels and pumping
Reservoirs
Total storage
Treatment
Sandusky, Ohio (16)
Washington, D.C.
(2, 15)
Columbus, Ohio
(2, 3, 12)
Whittier Street
Cambridge
Maryland (14)
Table
Storage
capacity
Mgal
1.1
3.9
1.3
1.2
12.39
13.00
25.39
2.82
2.82
2 998
41,315
44313
44,313
0.36
0.20
3.75
0.25
5. — Summary
Drainage
area,
acres
188.5
570
15,600
3,000
3,260
3,260
90
90
240 000
240,000
240,000
14.86
30.0
29,250C
20
of off-line storage
Capital
cost, $
455,700
1,744,000
6,495,000
9,488,000
11,936,000
11,936,000
744,000
189,000
933,000
870000,000
682,000,000
1,552,000,000
1,001,000,000
2,553,000,000
520,000
883,000
6,144,000
320,000
costs8.
Storage
cost,
$/gal
0.41
0.45
5.00
7.91
0.96
0.47
0.26
0.26
0.29
0.02
0.04
0.04
1.44
4.41
1.64
1.28
Cost per
acre,
$acre
2,420
3,110
416
3,160
3,660
3,660
8,270
2,100
10,370
3,630
2,840
6,470
4,170
10,640
35,000
29,430
210
16,000
Annual operation
and maintenance
$/yr
2,900
51,100
80,000
97,600
100,200
100,200
2,700
8,000
10,700
8,700,000
6,200
3,340
14,400
a. ENR 2000.
b. Estimated values; facilities under design and construction.
c. Estimated area.
$/acre x 2.47 = $/ha
$/gal x 0.264 = $/L
Mgal x 3785 = m3
less defined. Similarly, cleaner streets and neighbor-
hoods and enforced legislation will eradicate gross
pollution sources but to what limit should these be
applied and who will bear the cost? The final answers
will not be found short of implementation.
However, one thing we can be assured of is that in
view of the various documents which outline correct
evaluation procedures and the continually developing
state-of-the-art technologies, many local authorities
will be able to significantly reduce urban runoff
pollution in a cost-effective manner.
The technologies and procedures for combating
stormwater pollution and combined sewer overflows
are available today and are expanding rapidly. The EPA
-------
288
RESTORATION OF L^KES AND INLAND WATERS
Table 6. — Comparison of typical physical treatment removal efficiencies for selected pollutant parameters.
Percent reduction
Suspended Settleable
Physical unit process solids BODs COD solids
Sedimentation
Without chemicals 20-60 30 34 30-90
Chemically assisted 68 68 45
Swirl concentrator/flow
regulator 40-60 25-60 .. 50-90
Screening
Microscreens 50-95 10-50 35
Drum screen 30-55 10-40 25 60
Rotary screens 20-35 1-30 15 70-95
Disc strainers 10-45 5-20 15 ....
Static screens 5-25 0-20 13 10-60
Dissolved air flotation8 45-85 30-80 55 93b
High rate filtration0 50-80 20-55 40 55-95
High gradient magnetic
separation" 92-98 90-98 75 99
a. Process efficiencies include both prescreening and dissolved air flotation with chemical addition.
b. From pilot plant analysis.
c. Includes chemical addition.
d. From bench scale and small scale pilot plant operation, 1 to 4 L/mm (0.26 to 1.06 gal/mm).
Table 7. — Summary of typical biological stormwater treatment
Type of Tributary Design No.
biological area, capacity, Major of
Total
phosphorus
20
20
10
12
10
55
50
installations.
Project location treatment acres Mgal/d process components units Total size
Kenosha Contact 1,200 20 Contact tank 2
Wisconsin stabilization Stabilization tank 2
Milwaukee Rotating 35 0.05a 3 ft diameter 24
Wisconsin biological RBC units
contactors
Mt. Clemens,
Michigan
Demonstration Treatment lagoons 212 1.0b Storage/aerated lagoon 1
system in series with Oxidation lagoon 1
recirculation Aerated lagoon 1
between storms
Citywide full- Storage/treatment 1,471 4.0b Aerated storage basin 1
scale system lagoons in series Aerated lagoon 1
with recircula- Oxidation lagoon 1
tion between
storms Aerated/oxidation lagoon 1
New Providence, Trickling filters 6.0 High-rate plastic media 1
New Jersey High-rate rock media 1
Shelbyville, Treatment lagoons:
Illinois Southeast site 44 28C Oxidation lagoon 1
Southwest site 450 110 Detention lagoon plus 1
2-cell facultative lagoon
Springfield Treatment lagoon 2,208 67 Storage/oxidation lagoon 1
Illinois
a. Design based on average dry-weather flow; average wet-weather flow — 1 Mgal/d.
32,700 ft3
97,900ft3
28,300 ft2
750,000 ft3
1,100,000 ft3
930,000 ft3
4,440,000 ft3
508,000 ft3
1,100,000ft3
922,000 ft3
36 ft diameter
65 ft diameter
255,600 ft3
2,782,700ft3
5,330,000 ft3
b. Design flowrate through lagoon systems. Total flowrate to facilities is 64 Mgal/d for the demonstration project and
system.
c. Estimated using a 50% runoff coefficient at a rainfall rate of 1.95 m/h.
acres x 0.405 = ha
Mgal/d x 0.0438 = m3/s
ft3 x 0.0283 = m3
ft2 x 0.0929 = m2
ft x 0.305 = m
in/h x 2.54 = cm/h
Total Kjeldahl
nitrogen
38
30
17
10
8
35
21
Period
of operation
1972 to 1975
1969 to 1970
1972 to 1975
Under construction
1970 to present
1969 to present
1969 to present
1969 to present
260 Mgal/d forcitywide
-------
URBAN AND POINT SOURCE POLLUTION CONTROL TECHNOLOGY 289
recognizes the magnitude of the problem which is
facing local communities and is ready to help in their
fight to protect the quality of their receiving waters.
REFERENCES
Allen, D. M., and R. Sargent. 1978. Treatment of combined
sewer overflows by high gradient magnetic separation. EPA
600/2-78-209. U.S. Environ. Prot. Agency.
American Public Works Association. 1969. Water pollution
aspects of urban runoff. EPA 11030 DNS 01/69. U.S.
Environ. Prot. Agency.
Field, R., A. N. Tafuri, and H. E. Masters. 1977. Urban runoff
pollution control technology overview. EPA 600/2-77-047.
U.S. Environ. Prot. Agency.
Lager, J. A., and W. G. Smith. 1974. Urban stormwater
management and technology: an assessment. EPA 670/2-
74-040. U.S. Environ. Prot. Agency.
Lager, J., et al. 1977a. Catchbasin technology overview and
assessment. EPA 600/2-77-051. U.S. Environ. Prot.
Agency.
1977b. Urban stormwater management and
technology: Update and users guide. EPA 600/8-77-014.
U.S. Environ. Prot. Agency.
Masters, H. E., and R. Field. 1977. Swirl device for regulating
and treating combined sewer overflows. EPA 625/2-77-
012. U.S. Environ. Prot. Agency.
McGuen, R. H. 1975. Flood runoff from urban areas. Off. Res.
Technol. Tech. Rep. 33.
McPherson, M. B. 1976. Utility of urban runoff modeling. In
Proc. Spec. Session, Spring Annu. Meet., Am. Geophys.
Union, Washington, D.C., April 14, 1976; Am. Soc. Civil
Eng. Urban Water Resour. Res. Progr. Tech. Memo. 31.
Pisano, W. C., et al. 1979. Dry-weather deposition and
flushing for combined sewer overflow pollution control. EPA
600/2-79-133. U.S. Environ. Prot. Agency.
Sullivan, R. H., et al. 1975. The helical bend combined sewer
overflow regulator. EPA 600/2-75-062. U.S. Environ. Prot.
Agency.
1977. Sewer system evaluation rehabilitation and
new construction — a manual of practice. EPA 600/2-77-
017d. U.S. Environ. Prot. Agency.
Thelan, E., et al. 1972. Investigation of porous pavements for
urban runoff control. EPA 11034 DUY 03/72. U.S. Environ.
Prot. Agency.
U.S. Environmental Protection Agency. 1970. Stormwater
detention tank/hydrobrake demonstration for flood control
and combined sewer overflow pollution abatement. EPA
Proj. No. S-005370.
1978. Report to Congress on control of combined
sewer overflows in the United States. EPA 430/9-78-006.
White, C. A., and A. L. Franks. 1978. Demonstration of erosion
and sediment control technology. EPA 600/2-78-208. U.S.
Environ. Prot. Agency.
-------
290
THE GREAT LAKES: AN EXPERIMENT IN
TECHNOLOGICAL INNOVATION AND
INSTITUTIONAL COOPERATION
MADONNA F McGRATH
Director, Great Lakes National Program Office
U.S. Environmental Protection Agency
Chicago, Illinois
ABSTRACT
Restoration and preservation of good water quality in the Great Lakes ecosystem is of great
importance because of the magnitude of the resource and its many present and potential uses
serving the basin's 40,000,000 population in the United States and Canada. Within U.S. EPA the
Great Lakes National Program Office in Chicago serves as the coordinator and catalyst in dealing
with Great Lakes problems, including the complex inter-media and inter-governmental aspects of
this international resource. Congress has funded for staff and research under section 104(f) of the
Clean Water Act and for demonstrations of pollution abatement technology under section 108(a).
A variety of demonstrations is underway addressing agricultural nonpoint source control through
improved techniques and management practices, and urban nonpoint sources through control of
construction runoff, combined sewer overflows, and stormwater flows.
The Great Lakes National Program Office head-
quartered in the U.S. Environmental Protection
Agency's Region V, Chicago, is responsible for
managing and coordinating U.S. EPA's abatement and
control programs as they affect the water quality of the
Great Lakes. The Office serves as the Agency's catalyst
to identify and recommend solutions to lakewide and
transboundary pollution problems which cross cut
traditional lines of authority. Externally, the Office is
the principal U.S. focal point for communication,
coordination, and cooperation for Great Lakes pollution
issues with Canadian environmental agencies, the
States, and the public.
The Office concentrates most of its scientific and
technical resources on three key areas:
1. Revision and implementation of a Great Lakes
monitoring program with particular emphasis on toxic
organics, nutrients, and toxic metals.
2. Special investigations of serious "hot spot"
problem areas, with emphasis on developing control
measures for the full range of pollutant sources such
as land, water, and air.
3. Increased State and public involvement in Great
Lakes decisionmaking through the State/EPA agree-
ment process.
Since the principal goal of U.S. EPA's Great Lakes
effort is to restore and enhance water quality in the
Great Lakes Basin ecosystem so that public health,
welfare, and the environment are protected, the Great
Lakes National Program Office relies heavily on the
expertise of regional program offices, State pollution
control agencies, and EPA research laboratories in
finding solutions to these complex Great Lakes
pollution problems.
It may be asked by those unfamiliar with either the
breadth or the majesty of the Great Lakes ecosystem,
why the special concern for these bodies of water? Size
and use alone hold some of the answers. By volume,
the Lakes contain 6 quadrillon gallons of fresh water —
20 percent of the world's fresh surface water and over
95 percent of the United States' supply. More than 40
million people — nearly 20 percent of the U.S.
population and 50 percent of Canada's, live in the
Great Lakes Basin. More than 23 million of those
people depend on the Great Lakes for their drinking
water. While those statistics in and of themselves
indicate the vastness of the Lakes, they only begin to
convey the problems which have resulted from such
varied and intensive use.
The Great Lakes have been among the most abused
waters in our country, and that abuse has had far-
reaching effects. Since the area was first settled, the
Great Lakes have been a convenient disposal site for
every form of human waste and refuse. Industries,
municipalities, and communities found it all too easy to
discharge toxic substances, solid refuse and garbage,
and biological wastes into the Great Lakes and the
rivers feeding them. Runoffs from heavy rains and
spring thaws of winter snows flowed into the streams,
rivers, and the Great Lakes, carrying large amounts of
fertilizers and pesticides with them. By the late 1960's,
worldwide attention had focused on the severe
contamination and pollution problems in the Great
Lakes, which required direct and immediate action.
On the international scene, an institutional mech-
anism was already in place to guide those actions. Both
Canada and the United States had long recognized the
importance of the Great Lakes as a shared resource. In
1909, Canada and the United States signed the
Boundary Waters Treaty, which concerns all the waters
which form or cross the border between the two
countries. The Treaty created the International Joint
-------
URBAN AND POINT SOURCE POLLUTION CONTROL TECHNOLOGY
291
Commission to deal with boundary water problems,
including those of the Great Lakes. Many studies were
conducted on the Lakes over the years and finally in
1972 the two Governments developed the first Great
Lakes Water Quality Agreement. The IJC was asked to
determine the pollution in Lakes Superior and Huron
and also to determine the extent of pollution from land
drainage. The studies were completed and reports
submitted to the two Governments. In 1977-78 a
review of the 1972 Water Quality Agreement was
made and public hearings held to improve upon it.
In 1978 a new Water Quality Agreement was signed
by the United States and Canada. The new agreement
is more comprehensive than the first one in several
ways. It includes the entire Great Lakes System — the
land surrounding the Lakes, the streams flowing into
them, and the Lakes themselves. It involves more than
water quality. The ecosystem approach which recog-
nizes the complex interrelationships among water,
land, air and living things (plants, animals, and man) is
found throughout the agreement.
The new agreement also emphasizes the need to
understand and manage toxic substances. It reinforces
the importance of controlling phosphorus pollution. It
renews Government's commitment to control pollution
from shipping and dredging, and to collect the data
necessary to monitor water quality effectively. The
Agreement of 1978 requires programs to determine
the impacts and sources of airborne pollutants, and
new measures to control pollution from various land
uses. The agreement's general and specific objectives
are designed to achieve and preserve a certain level of
quality in the Great Lakes ecosystem.
But what have been some of the problems of the
Great Lakes? In the 1960's enforcement conferences
were held for each of the Great Lakes to determine
their pollution status. Federal and State investigations
involved water sampling and chemical/biological
analysis to diagnose the Lakes' problems. It was
determined then that nutrients were a major problem,
especially in Lake Erie. Oil and grease, suspended
solids, and organic contaminates were unsightly,
damaging to wildlife, and caused problems with many
water users. Untreated and/or inadequately treated
industrial and municipal wastes were being discharged
directly to rivers and lakes. Combined sewer overflows
were causing bacterial pollution of beaches along with
debris. Stormwater overflows were in some cases
discharging toxic materials directly to surface waters.
Schedules were set to remedy many of the problems
but the law did not provide the teeth to enforce a
cleanup effort.
In 1972 the Clean Water Act, Public Law 92-500,
gave the U.S. EPA regulatory authority to enforce water
pollution cleanup. Also during the 1970's other
environmental laws were passed to further strengthen
EPA's position. These laws included the Amendment to
the Clean Water Act-1977, the Safe Drinking Water
Act of 1974, the Resource Conservation and Recovery
Act, the Toxic Substances Control Act of 1976, and the
Clean Air Act Amendments of 1970 and 1977.
Back when Lake Erie was headlined as a "dead lake"
and Rachel Carson's book entitled "Silent Spring' was
stimulating environmental interest, State and Federal
Governments concluded that phosphorus was the
element that could best be controlled through waste
treatment practices to reduce giant algal blooms in the
lakes and the rapid aging taking place. Waste treatment
processes were discussed and researched to see what
could be done. Wastewater treatment requirements for
municipal plants were set to provide secondary
treatment with phosphorus removal. Industry was
required to correct its discharge problems. In 1972 the
Clean Water Act provided billions of dollars to upgrade
municipal wastewater treatment plants to meet the
Nation's pollution abatement needs.
Detergent phosphate bans were imposed in all of the
Great Lake States but Ohio and Pennsylvania. Studies
have indicated these bans significantly reduced
phosphorus. At present a 1 mg/l effluent phosphorus
limit is the target goal for wastewater treatment plants
of 1 million gallons per day size or larger on the Great
Lakes. To achieve the Agreement's target loadings may
require not only greater point source control activity but
also some nonpoint source controls.
Five billion dollars has been spent by EPA in the last
decade to help clean up the Great Lakes. Additional
billions of dollars have been spent by State and local
governments and industries. While this expenditure of
public and private funds has enabled us to abate the
most visible Great Lakes pollution, it is what we do not
see, taste, or smell that may cause severe problems in
the years ahead. Clearly, the future challenges of lake
restoration are in the area of toxic substance control.
The most serious threat is the existence of persistent
toxic chemicals in Great Lakes' water, fish, wildlife,
and sediments. These substances affect all portions of
the Great Lakes in varying degrees. Many have the
capacity to bioaccumulate; they have been found in the
Lakes' fish and wildlife in alarming concentrations.
Fish from Lake Ontario are heavily contaminated by
Mirex. Lake Michigan fish cannot be sold commercially
because of high levels of PCB's. Fish from Lake St. Clair
had high levels of mercury that restricted their use for
several years.
These substances reach the aquatic environment
through direct discharges from industries, in runoff
from agricultural and urban activities, and from the
atmosphere after evaporation or insufficient incinera-
tion. While the effect of toxic substances on aquatic
organisms is not well understood, severe adverse
health effects on mammals and birds are well
documented.
The National Program Office is checking the Lakes
for toxic chemical "hot spots." One way we find these
areas is through an extensive fish tissue and analysis
program, which concentrates on fish found both in the
open waters and in the nearshore tributary streams.
Scientists combine findings from these surveys with
results of intensive sediment studies to identify toxic
chemical problem areas in selected harbors and
tributary basins. We then use this information to
identify specific sources and remedial measures.
Regulation assessments are underway or planned in
the following areas: the Ashtabula River in Ohio,
Buffalo River in New York, Raisin River in Michigan,
Indiana Harbor Canal in the vicinity of Gary, and
Milwaukee, Wis.
-------
292
RESTORATION OF LAKES AND INLAND WATERS
But we are really only on the threshold of toxic
substance control. We have some analytical and
enforcement tools, but require much more. Our record
is much better in the development of techniques to
control conventional pollutants. We are striving to
apply the knowledge gained in this area to the
perplexing toxic questions. For example, the National
Program Office administers the section 108(a) demon-
stration grant program which provides that the EPA can
enter into agreements with any State, political
subdivision, interstate agency, or other public agency
to carry out projects to demonstrate new methods and
techniques and to develop preliminary plans for the
elimination or control of pollution, within all or any part
of the watersheds of the Great Lakes drainage basins.
The Great Lakes National Program Office has
entered into a number of demonstration projects with
State and local entities of Government to develop and
implement new methods and techniques for sediment
and related pollutants from rural runoff and for
reducing pollutants from urban runoff. Institutional and
educational methods have been developed to help
implement rural and urban nonpoint source pollution
controls. Technical seminars have been held and
project reports have been published for national
distribution.
The section 108 program is closely coordinated with
the section 208 water quality management planning
and trie Office of Research and Development. Data and
technical information derived from Section 108
projects have impacted national nonpoint source
guidance as well as States and local legislation and
ordinances.
The section 108 program has been used to provide a
systems approach to solving Great Lakes water
pollution problems. We have tried to bridge the gap in
EPA water pollution programs to tie planning and
implementation together in one continuous effort. We
use the planner, the institutional structure of State or
local government, and citizen involvement. The
program tries not to duplicate but rather enhance other
EPA programs. An example would be Washington
County, Wis., where the State Board of Soil and Water
Conservation Districts was the grantee. They worked
through the County Soil and Water Conservation
District and the University of Wisconsin. Project Staff
spent much time and effort with local public officials
and land owners. Through these efforts and the use of
grant funds to provide incentives for best managment
practices demonstrations and to monitor results,
individual involvement in nonpoint control efforts has
been stimulated and some local governments have
adopted construction runoff ordinances in the project
area giving the Soil and Water Conservation District a
role in reviewing subdivision plats.
Major water pollution problems in the Great Lakes
that are high priority considerations for funding are as
follows:
1. Toxic or hazardous substance control.
2. Combined sewer overflow pollution control.
3. Storm sewer overflow pollution control.
4. Rural nonpoint source pollution control.
To date this program has provided Congress and EPA
Headquarters with data on nonpoint source pollution
that have helped to develop the 1977 amendments to
Public Law 92-500. Three major section 108 projects
are the Black Creek project, the Washington County
project, and the Red Clay project. Within these projects
we have developed educational films and curricula for
informing the public about nonpoint source pollution
and the solutions to it.
We have developed and/or evaluated erosion control
and a series of best management practices on the Black
Creek Project that will improve water quality. We have
also developed a watershed management model
(ANSWERS) that accurately predicts sediment runoff
during storm events; the model relates to the land
management practices used, soil type characteristics,
and slope of land.
The Black Creek Project began by evaluating 33
practices found in the Soil Conservation Service
technical manual. A small number of the practices was
found to be of major importance in the study area.
Some sources originally thought to be important such
as stream bank erosion turned out to be far less
significant than others. Tillage practices, increasing
crop residue and surface roughness, grassed water-
ways, livestock exclusion from streams, pasture
planting, sediment control basins and terraces all
proved to be of considerable use. A further general
discovery was the importance of targeting critical areas
rather than the original attempt of treating all areas.
The Washington County project investigators have
provided much of the basic material and support that
helped the State of Wisconsin pass its recent Sediment
and Erosion Control legislation (Wisconsin Fund). All
projects have achieved pollution reductions. Data and
information from these section 108 projects have also
been used in preparing the IJC Pollution from Land Use
Activities Reference Group report and its remedial
program recommendations.
Numerous technical reports have been published
and distributed on section 108(a) activities. We have
also encouraged some changes in Soil Conservation
Service's procedures in dealing with land management
practices as they affect water quality.
The National Program Office has worked closely with
the Office of Research and Development at Edison, N.
J., to demonstrate new techniques to reduce and
remove pollutants from combined sewer overflows. We
have three active projects dealing with urban combined
sewer treatment and control.
At Rochester, N. Y. the Rochester Pure Waters
District studied its drainage systems and developed a
combined sewer overflow abatement program that
recommended implementation of a best management
practices system coupled with construction grant
programs.
The demonstration of best management practices is
underway and is scheduled to be completed about
December 1980. This project involves the implementa-
tion and evaluation of minimal structural and non-
structural techniques to control urban storm and
combined sewer overflow discharges. This represents
the first phase of the master plan developed under a
previous section 108(a) grant for the Rochester Pure
Waters District. The best management practice
program in conjunction with facilities provided under
concurrent abatement programs is projected to result
in an 80 to90 percent reduction in the combined sewer
-------
URBAN AND POINT SOURCE POLLUTION CONTROL TECHNOLOGY •
293
overflow pollutant load to the Genesee River and the
Rochester embayment of Lake Ontario. This is expected
to significantly improve the water quality of several
Lake Ontario-Rochester beaches.
At Saginaw, Mich, we are working with the Saginaw
Department of Public Utilities to demonstrate that a
swirl concentrator and degritter are acceptable, cost-
effective methods for controlling and treating storm-
water overflow from a combined sewer system. A study
of the alternative treatment and control techniques to
solve the City of Saginaw combined sewer overflow
problem using swirl concentrators over retention
basins can save the city $19,000,000 in capital costs
and .$206,0007year in operation and maintenance
costs. This project is in its preconstruction stage. It is
due.to be completed in March of 1982.
At Cleveland, Ohio we are working with the City
Department of Public Utilities and the Northeast Ohio
Regional Sewer District to demonstrate control of
sewer overflows during rain events. The process to be
demonstrated is a controlled discharge of stormwater
runoff according to the designed capacity of the
individual sewer line. The catch basins are discon-
nected from the existing sewer and hooked into a
detention tank from which the storm water will be
discharged at a controlled rate by gravity through an
internal energy dissipator(Hydro-brake) which requires
no sources of energy, and has no moving parts. Capital
savings in the order of 50 percent, compared to any
conventional alternative for rehabilitation of combined
sewer systems, are indicated. Evaluation of this
process will start soon.
We are trying to get innovative technology demon-
strated at the size and level such that consultant
engineering firms will begin to factor these methods
into their alternative treatment and control costing
requirements under the municipal facilities planning
exercise. We still have many gaps to bridge to get new
technology into the system. We hope our projects can
assist in filling this need.
But what of the future? If pollution contaminates
more groundwater sources, even more millions of
people will look to the Great Lakes as a source of
drinking water. The energy situation may require that
we use the Great Lakes even more intensively for
navigation, power production, and possibly natural gas,
for which Canada already drills in the western end of
Lake Erie. Recreation close to home will continue;
popular resort areas already face overbuilding and
resulting strains on water treatment systems. Other
emerging problems, such as increased levels of sodium
and chlorides, also may affect the ecological balance
within the Lakes and their interconnected systems.
Finding solutions to these problems requires both
interstate and international partnership, a highly
dedicated scientific community, and heightened public
awareness. The key role of that public cannot be
underestimated — for without their support, both
financially and philosophically — the efforts to
understand and help Lake processes may well be for
naught.
REFERENCES
Andrews, S. C., D. S. Houtman, and W.J. Lontz. 1979. Impact
of nonpoint pollution on western Lake Superior. Red Clay
Proj.-summary. Final Rep. EPA-905/9-79-002. U.S. En-
viron. Prot. Agency.
Lake, J. E., and J. Morrison. 1977. Environmental impact of
land use on water quality. Black Creek Final Rep. Summary.
EPA 905/9-77-007-A. U.S. Environ. Prot. Agency.
Madison, F. W., etal. 1980. Development and implementation
of a sediment control ordinance or other regulatory
mechanism: Institutional arrangements necessary for
implementation of control methodology on urban and rural
lands. Washington County Proj. Final Rep.
McGrath, M. F. 1979. Great Lakes National Program Strategy
Document. Region V, U.S. Environ. Prot. Agency, Chicago,
1980. GLNPO: Great Lakes is their concern.
Environ. Midwest March 1980.
1980. The Great Lakes. EPA Jour. 6:18.
McGuire, J. 1980. How the acts unfolded. Great Lakes
Communicator 10:5.
U.S. Department of State. 1978. United States Treaties and
other international agreements. U.S. — Canada Great Lakes
Water Quality Agreement of 1978. November 22.
-------
294
DESIGN OF STORAGE/SEDIMENTATION FACILITIES
TO CONTROL URBAN RUNOFF AND COMBINED
SEWER OVERFLOWS
W. MICHAEL STALLARD
WILLIAM G. SMITH
RONALD W. CRITES
Metcalf & Eddy, Inc.
Sacramento, California
GEORGE TCHOBANOGLOUS
University of California
Davis, California
ABSTRACT
Urban stormwater runoff and combined sewer overflows are potentially significant sources of
water pollution. Storage/sedimentation facilities have been recognized, both in the United States
and Europe, as cost-effective measures for stormwater treatment and control. This paper
summarizes a manual currently being prepared for the U.S. Environmental Protection Agency,
detailing procedures for planning and design of various storage/sedimentation techniques. Such
techniques as upland attentuation, inline storage, and end-of-pipe storage and treatment are
detailed. Pollutants and watershed characteristics of stormwater mariagement are discussed,
including the range of water quality expected in urban stormwater runoff and combined sewer
overflow. Data for the specific study area must be used. Models to evaluate the runoff problem and
select effective solutions are listed. European practice in stormwater storage and sedimentation is
described. Current practice in the United States in storage/sedimentation is discussed based on
the American Public Works Association survey of 1980 and several case histories by Metcalf &
Eddy. Recommended design practice is specified. Water quality benefits of both urban stormwater
and combined sewer overflow storage/sedimentation are discussed.
INTRODUCTION
As municipal wastewater treatment is upgraded in
accordance with the Federal Clean Water Act, urban
stormwater runoff and combined sewer overflows are
emerging as significant sources of surface water
pollution in the United States. A 1975 survey of 56
public agencies located throughout the United States
revealed that " control (of) stormwater pollution
from sources other than erosion " ranked second
only to flood control as a stormwater management goal
(Poertner, Draft). The 1978 Needs Survey prepared by
the U.S. Environmental Protection Agency estimated
that $87.4 billion is needed by the year 2000 to bring
combined sewer overflows and urban stormwater
runoff into compliance with the requirements and
goals of the Federal Water Pollution Control Amend-
ments of 1972 (U.S. EPA, 1979).
Urban runoff is not a new problem. Traditionally, the
goal of stormwater control has been to reduce or
eliminate flooding. Temporary storage of runoff, a
widely used method of flood control, is gaining wider
application in the United States as a means of reducing
the pollutant load of stormwater runoff. The U.S.
Environmental Protection Agency is preparing a Design
Manual for Storage/Sedimentation and Combined
Sewer Overflows. This paper summarizes the contents
of that manual, which will be available early in 1981.
THE MANUAL'S PURPOSE
In recent years, EPA has been committed to identify
pollution sources other than municipal wastewater
discharges and to develop viable methods for their
control. A large amount of information has been
developed over the past decade on stormwater runoff,
and particularly, combined sewer overflow characteris-
tics, receiving water impacts, and treatment. The
Design Manual is to summarize the existing informa-
tion and detail step-by-step procedures for stormwater
storage/sedimentation treatment facilities. The Manu-
al's audience is not only the hydrologist and
stormwater control engineer, but the local decision-
maker and land development engineer, as well.
The Design Manual is organized into six chapters.
Chapter 1 is an introduction and guide to its contents.
The second chapter, written for the nontechnical
decisionmaker, overviews urban runoff and combined
sewer overflow as pollution sources, and describes
how storage/sedimentation facilities can be used to
reduce the pollutant load. Chapters 3 and 4 outline the
basic operating principles of storage/sedimentation
facilities, and detail the data needs and design
procedures for five types of facilities. Chapter 5
describes, through examples, the application of
storage/sedimentation facilities in an overall storm-
water management system. An important part of
-------
URBAN AND POINT SOURCE POLLUTION CONTROL TECHNOLOGY
295
Chapter 5 is devoted to explaining how existing flood
control storage facilities can be retrofitted to provide
better pollution control. The final chapter of the Manual
draws heavily on European practice to suggest ways in
which regional design guidelines can be developed and
applied to controlling stormwater pollution.
A STORMWATER OVERVIEW
It is important when selecting or designing a
stormwater control system to understand the types of
pollutants contained in urban runoff and combined
sewer overflows, the characteristics of the watershed
that may influence the quantity and quality of
stormwater, and the possible impacts of the storm-
water on the receiving water.
Table 1 compares the concentrations of pollutants
most commonly found in stormwater runoff with
concentrations of the same pollutants in receiving
water and sanitary wastewater. However, a wide
variety of other pollutants, particularly toxic sub-
stances, may also be present in stormwater.
The pollutant concentrations shown in Table 1
should be used to identify relative magnitudes only.
Runoff and combined sewer overflows are highly
variable in the concentrations of pollutants present, as
well as in quantities and rates of flow. Among the
characteristics that may influence runoff quantity and
quality from a watershed are hydrology, land use,
physical characteristics of the surface such as soil type
and percentage of area covered by impervious
structures, and the type and configuration of the
stormwater drainage system. The importance of
collecting data on the stormwater runoff characteris-
tics and treatability specific to the area cannot be
overemphasized.
The severity of surface discharge of wastewater
depends on the natural self-purification mechanisms of
the receiving water. The goal of wastewater control is
to reduce the pollutant load so that it can be
assimilated without impairing the receiving water. In
the United States, the impact of an urban stormwater
discharge on the assimilative capacity of the receiving
water must also be evaluated in light of other point and
nonpoint discharges.
STORAGE/SEDIMENTATION OPTIONS
Generally, urban runoff and combined sewer
overflow pollution occurs during periods of peak
rainfall and runoff when the infiltration capacity of the
ground surface and the transport and/or treatment
capacities of the drainage system are exceeded.
Temporary storage of stormwater runoff can reduce the
peak rates of flow so that the transport and treatment
capacities are exceeded less often. When the storage
capacity is exceeded, storage basins may be designed
to provide sedimentation treatment for the excess flow.
Storage/sedimentation facilities can be categorized by
disposal method. Detention storage facilities are those
in which the stored runoff excess is released to the
sewers at a reduced rate when capacity is available.
The captured flows are usually treated before
discharge when the storage aim is pollution reduction.
Retention storage facilities capture flows which then
are allowed to evaporate or percolate to the ground
water without release from the facility.
DETENTION STORAGE FACILITIES
Three types of detention storage facilities are
covered in the Design Manual: (1) Upland attenuation
facilities, such as rooftop, parking lot, and plaza
storage; (2) inline storage facilities; and (3) detention
storage/sedimentation basins. The first two types are
usually designed principally for storage. In most cases,
maximum use of available upland and inline storage is
made in combination with some downstream control
facility. Detention storage/sedimentation basins are
generally placed downstream of the storm or combined
sewer system to provide both storage and sedimenta-
tion control.
Detention storage/sedimentation basins may be
operated in a variety of modes. Excess runoff or
combined sewer overflows are routed to the basins
until the basins are full. At this point, all flows may
continue to be routed through the basins, subjecting
them all to sedimentation treatment. If a significant
first flush is exhibited, as in small catchments with
combined sewers, flows greater than the basin storage
capacity may be bypassed to the receiving water. In this
Table 1. — Comparison of stormwater discharges to other pollutant sources, (mg/l unless otherwise noted.)
Background
levels
Stormwater
runoff
Combined
sewer overflow
Sanitary
wastewater
TSS
5-100
415
370
200
VSS
-
90
140
150
BOD
0.5-3
20
115
200
COD
20
115
367
500
Kjeldahl
nitrogen
—
1.4
3.8
40
Total
nitrogen
0.05-0.5"
3.10
9.10
40
Total
PO4-P
0.01-0.2°
0.6
1.9
10
OPO.-P
—
0.4
1.0
7
Fecal
Lead conforms"
<0.1
0.35 13,500
0.37 670,000
--
a. ORGANISMS 100/ ML.
B. NO; as N.
c. Total phosphorus as P.
-------
296
RESTORATION OF LAKES AND INLAND WATERS
way, the first flush is captured without risking
resuspension by later flows. The sediment captured in
the basins is usually returned to the sewers when
capacity becomes available for later treatment before
discharge. The contents also may be released directly
to the receiving water slowly so as not to exceed
assimilative capacity.
When designing detention storage/sedimentation
facilities, many factors are taken into consideration.
Important data needs include watershed characteris-
tics and hydrology, runoff pollutant concentrations and
treatability, sewer and treatment plant capacities, and
identification of available sites. A design procedure
might follow these steps:
1. Quantify expected stormwater flows and pollutant
loadings. Very often, computer simulation based on
collected data is necessary. It is important that the
distribution of runoff and pollutants within storm
events be assessed.
2. Identify waste load reductions required. To ensure
that receiving waters are protected, water quality
impacts must often be assessed. Once the problem
pollutants and required removal efficiencies have been
identified, a decision can be made to design either for
complete capture or for overflow sedimentation.
3. Identify feasible basin sites.
4. Capture basin design. Capture basins generally are
located on very small catchments where a first flush of
pollutants is most pronounced. The most important
considerations are the degree of first flush exhibited,
sewer and treatment capacities, and the removal of
captured runoff and, particularly, solids after the runoff
rate subsides.
5. Sedimentation basin design. Sedimentation basins
can be very effective in removing paniculate and
floatable materials from urban runoff. The removal
efficiency is a function of particle size and density,
surface overflow rate, and horizontal velocity. Other
important considerations include elimination of short
circuiting, weir and basin depth design to cut down
scouring of settled solids, and captured material
removed.
RETENTION FACILITY DESIGN
The design discusses design principles and pro-
cedures for two types of retention storage facilities.
Percolation/retention ponds, also called dry ponds, are
earthen basins in which runoff is stored and allowed to
percolate, usually within a few days. In wet ponds,
excess runoff is stored in a permanent pond by varying
the water level.
Retention storage facilities are generally very
effective in reducing the pollutant loads, both sus-
pended solids and BODb, for the runoff captured. They
also have the added advantage of providing ground-
water recharge. Because retention facilities depend on
percolation and evaporation for emptying, these
facilities are usually very large and shallow ponds.
During overflow conditions, the deposited solids may
be resuspended and carried over the overflow weir.
Retention facilities are therefore most effective when
operated as capture basins for first flush containment,
with a total bypass of excess flows.
The data needs for designing retention storage
facilities include watershed characteristics and hy-
drology, identification of available sites, and site soil
characteristics. General design procedures include:
1. Quantify expected stormwater flows and pollutant
loadings. The occurrence interval of runoff events is an
important consideration, as well as runoff volumes and
pollutant content. Retention facilities must be sized to
allow sufficient emptying between events.
2. Identify the waste load reduction required. An
assessment of receiving water impacts may be
necessary or the required waste load reduction may be
determined by a regulatory agency.
3. Identify feasible sites. The large area requirement
and need for suitable soils are often the factors limiting
the use of retention ponds.
4. Investigate the most promising sites for suitability
of soils. It is important to keep in mind that silt from the
runoff will tend to seal the soil surface and that the
ponds must be sized according to the frequency with
which the pond bottom will be scarified or dredged.
5. Quantify expected evaporative losses. For wet
ponds, evaporation may be a major factor in the
hydraulic balance.
6. Size the basins based on a water balance of all the
hydrologic factors.
REGIONAL STORMWATER CONTROL
GUIDELINES
For many generations, Europeans have used stor-
age/sedimentation to control pollution from combined
sewer overflows. In many cases, the Europeans have
developed simple and easy to follow guidelines for
designing these facilities. This approach is made
possible because the guidelines are applied to very
limited areas, in which storm patterns, land use,
pollutant washoff functions, and water quality impacts
are sufficiently similar to allow generalization. This
same approach is being developed in some areas of the
United States, such as Montgomery County, Md.,
Fairfax County, Va., and Denver, Colo.
The final chapter of the Manual looks at this regional
guideline approach to stormwater management. It
covers European practice in Scotland, Switzerland
(Kanton), and Germany (Bavaria). Each is presented on
a case study basis, including an evaluation of its
effectiveness by regulations and agencies.
REFERENCES
Field, R. Trip report on 1978 tour of European stormwater
control facilities. U.S. Environ. Prot. Agency, Edison, N.J.
(Unpublished).
Lager, J. A.,et at. 1977. Urban stormwater management and
technology: Update and users' guide. EPA 600/8-77-014.
U.S. Environ. Prot. Agency.
Metcalf & Eddy, Inc. 1980. Urban stormwater management
and technology: Case histories. EPA 68-03-2117. U.S.
Environ. Prot. Agency.
Nussbaum, G. Remarks on the treatment of rain water in the
sewer system. Source unknown.
Poertner, H. 1974. Practices in detention of urban
stormwater runoff. Spec. Rep. 43. Am. Pub. Works Assoc.
Chicago, III.
-------
URBAN AND POINT SOURCE POLLUTION CONTROL TECHNOLOGY 297
Draft. Manual on stormwater management. Am.
Pub. Works Assoc. Chicago, III.
U.S. Environmental Protection Agency. 1976. Areawide
assessment procedures manual. EPA 600/9-76-014.
1979. 1978 needs survey cost methodology for
control of combined sewer overflow and stormwater
discharge. EPA 430/9-79-003.
Wanielista, M. P. 1978. Stormwater management: Quantity
and quality. Ann Arbor Science, Ann Arbor, Mich.
-------
298
SWEDISH EXPERIENCE OF NUTRIENT REMOVAL
FROM WASTEWATER
CURT FORSBERG
SVEN—OLOF RYDING
Institute of Limnology
University of Uppsala
Uppsala, Sweden
ABSTRACT
Water quality preservation steps in Sweden have been focused on chemically treating wastewater
for phosphorus removal. In early 1979 more than 750 chemical or biological-chemical wastewater
treatment plants were operating, treating about 75 percent of the total amount of wastewater from
urban areas. This paper describes removal efficiencies for different process combinations, process
improvements, sludge disposal, and treatment costs. The decreasing pollution load has improved
many Swedish waters. Examples are given from brackish and fresh waters. Correlations between
phosphorus and transparency indicate that reducing phosphorus will not markedly decrease the
chlorophyll a content and thereby increase transparency until the phosphorus concentration in
lake water is depressed below 0.1 to 0.2 g/m3.
INTRODUCTION
During the 1950's and 1960's several Swedish lakes
and coastal waters became markedly eutrophicated.
Increasing population, increasing numbers of water
closets and the use of phosphorus-containing synthetic
detergents rapidly increased the P-load during a short
space of time. The effects of household detergents
were intensively discussed during the 1960's. Some
modifications of these products reduced the P
originating from them in sewage to approximately 30
percent (Natl. Swed. Environ. Prot. Board, 1972).
Water quality preservation steps in Sweden have
focused on total nutrient removal, on expanding the
chemical treatment of wastewater for phosphorus
removal. This paper summarizes the development of
sewage treatment, describes process combinations
and efficiencies, efforts to improve treatment methods,
sludge handling, and also gives examples of lakes
where recovery has been observed following nutrient
removal.
EXPANSION OF WASTEWATER
TREATMENT PLANTS
The expansion of wastewater treatment plants in
Sweden from the mid-1950's to the mid-1970's is
described by Ulmgren (1975). The large extension of
chemical sewage treatment began in 1968, with the
purpose of reducing the phosphorus content in
wastewaters. In early 1979 more than 750 municipal
wastewater treatment plants were operating with
chemical or combined biological and chemical treat-
ment (Table 1), corresponding to about 75 percent of
the total amount of wastewater from urban areas.
Table 1 — The number of sewage treatment plants and
processes used in densely populated areas in Sweden,
January 1, 1979 (Natl. Swed. Environ. Prot. Board, 1979).
Type of sewage
treatment
No treatment
Sedimentation
Biological
Chemical
Biological +
chemical
Complementary
Number of
plants
156
380
141
625
18
1,320
Number of persons
served
7,000
181,000
1,398,000
324,000
4,833,000
107,000
6,850,000
QUALITY OF INCOMING WASTEWATER
The amount of wastewater entering a Swedish
treatment plant is about 400 liters per person per day,
including water from smaller industries, etc. Water
consumption in households is about 200 liters per
person per day.
Ulmgren (1975) analyzed incoming wastewater at 50
wastewater treatment plants (Table 2) and found the
main change during the first half of the 1970's was a
more than 20 percent decrease in the phosphorus
content.
Table 2, — Quality of incoming wastewater to 50 Swedish treatment
plants, g/m3 (Ulmgren, 1975).
Parameter
Organic matter, BOD7
Organic matter, COD
Suspended solids
Total Phosphorus
Total Nitrogen
Average
value
123
226
122
5.7
26
Median
value
116
259
103
5.5
24
Standard
deviation
± 60
±115
± 66
± 2.5
± 9
Number of
analyses
122
78
122
124
89
-------
URBAN AND POINT SOURCE POLLUTION CONTROL TECHNOLOGY
299
MONITORING SEWAGE EFFLUENT
QUALITY
An effluent control program for studying the
efficiency of the Swedish wastewater treatment plants
has been directed by the Environment Protection
Board. In particular, chemical oxygen demand (COD)
and total phosphorus are being analyzed. At treatment
plants serving more than 2,000 people, the samples for
COD and total phosphorus are preserved in weekly flow
proportional samples. The frequency of sampling
increases with plant size. At plants >20,000 people,
flow proportional, continuous sampling is conducted.
At several plants samples are being taken continuously
and analyzed by the minitest method (Elf ring, Forsberg,
and Forsberg, 1975).
The municipalities forward the results to the local
county administration. The National Swedish Environ-
ment Protection Board then summarizes and evaluates
the results annually (Natl. Swed. Environ. Prot. Board,
1979).
COMBINATIONS OF PROCESSES
AND EFFICIENCIES
As illustrated in Table 1, the main process consists of
biological and chemical treatment. Where poor receiv-
ing conditions prevail in relation to the discharge,
complementary treatment, mainly in the form of
postfiltration, is prescribed.
In early 1979 chemical sewage treatment was
employed according to processes and sizes of
treatment plants listed in Table 3. Post-precipitation
dominated the treatment. At that time complementary
treatment was used at 18 sewage works, serving about
100,000 people. Most of the sewage from the
Stockholm area was treated by pre- or simultaneous
precipitation.
Earlier, aluminum sulfate was the dominant precipi-
tant. Today iron salts are also frequently used. At about
50 smaller plants, lime is the precipitating agent.
Ryding
Table 3. — Flocculation processes and size distribution of
wastewater treatment plants, January 1, 1979 (Natl. Swed.
Environ. Prot. Board, 1979).
Number of treatment plants
Flocculation designed for pe
process <500
Direct preci-
pitation 28
Pre-precipi-
tation
Simultaneous
precipitation 3
Post-precipi-
tation 58
501-
2000
71
1
13
194
2001-
5000
28
1
4
137
5001-
20,000
10
2
7
119
>20,000 Total
4
10
6
69
141
14
33
576
Phosphorus removal efficiencies for different pro-
cess combinations have been discussed recently
(Gronquist, et al. 1978; Hultman, 1978,1979). In spite
of similar processes, precipitants, size of load, etc., the
results from different plants vary widely. This illus-
trates that factors not normally monitored have a great
influence on the treatment efficiencies. In cases where
there are no significant process disturbances, the
phosphorus concentrations listed in Table 4 refer to
permanent full scale operation. Hultman (1978)
pointed out that the data in this table are comparatively
old. New evaluations, at present being compiled, will
probably change the ranges given in Table 4. Normally
loaded plants operating with lime precipitation, for
instance, seem to be more efficient than indicated.
Table 4 — Phosphorus removal efficiences for different processes.
Modified after Hultman (1978).
Effluent
P-concentrations
g/m3
0.5-1.2
0.5-0.8
0.2-0.4
0.15-0.3
Processes
Post-precipitation, Al-sulphate, pH 6.5-7.2
Post-precipitation, lime
Pre-precipitation
Simultaneous precipitation
Post-precipitation, Al-sulphate, pH 5.5-6.4
Post-precipitation, lime (low loaded)
Post-precipitation, Fe3* + sludge recircu-
lation
Pre-precipitation, + filtration
Post-precipitation, Al-sulphate, pH 5.5-6.4
+ filtration
Simultaneous precipitation -(-contact
filtration
Pre-precipitation + contact filtration
DISPOSAL OF MUNICIPAL SLUDGE
The growing demand for more advanced wastewater
treatment has considerably increased the volume of
municipal sludge. Since 1960 the amount has
increased about threefold (Tullander, 1975).
Sludge can be disposed of either at sludge disposal
sites as landfill, or for agricultural use in enriching soil.
Using sludge for agricultural food production poses a
number of hygienic and environmental hazards. The
Swedish National Board of Health and Welfare has
investigated this problem and published instructions in
1973. Standards for evaluating the quality of sludge
are given by Tullander (1975).
Special attention has been devoted to the content of
heavy metals. At present it is not possible to make a
definite evaluation of the biological effects of these
metals. Asa general rule, frequent and long-term use
of sludge on any one field should be avoided. Similarly,
sludge having excessive levels of heavy metals should
be avoided in agriculture. From an environmental
viewpoint, the maximum amount of sludge spread on
individual fields should not exceed 5 tons of dry
solids/ha during a 5-year period. Declaration of
contents is recommended for sludge. This will simplify
adherence to the standards and give the treatment
plants people valuable information on the composition
of the wastewater and also on the need for further
improvements of the treatment processes.
COSTS OF SEWAGE TREATMENT
The costs of sewage treatment have been examined
by Hultman(1978). His values for post-precipitation are
reproduced in Table 5, showing that capital costs are
somewhat higher than operating costs. Centralizing
-------
300
RESTORATION OF LAKES AND INLAND WATERS
the wastewater treatment in bigger plants will
decrease treatment costs.
The chemicals necessary for nutrient removal
require about 25 percent of the operating costs for
post-precipitation. The cost of chemical precipitation is
15 percent and of sludge conditioning 10 percent.
Table 5. —Approximate costs for sewage treatment, 1978 (Hultman,
1978).
Number of person Costs for post-precipita- Additional costs
Equivalents (p.e.) tion plants (including for deep-bed
sludge treatment) filtration
Capital Operating Capital Operating
costs* costs* costs* costs'
2,000
5,000
20,000
50,000
130
100
60
45
100
70
50
40
25
10
7
8
4
3
* in Swedish Crowns per capita per year.
Notes: 1 Swedish Crown = 0.23 $
In calculation of capital costs the annuity used is 10% and
13% for post-precipitation plants and deep-bed filters,
respectively.
PROCESS IMPROVEMENTS
After the rapid expansion of advanced wastewater
treatment, efforts are now concentrated on reducing
the operating costs and promoting efficiency. Important
work is being done within the Nordic Cooperative
Organization for Applied Research (NORDFORSK). A
project concerning management of municipal waste-
water treatment plants has resulted in five reports
dealing with flow equalization in sewer systems
(Stahre, 1978), wastewater filtration (NORDFORSK,
1 978), evaluation of continuously operating measuring
instruments (Holmstrom, 1979a), guidelines for moni-
toring programs (Balmer, et al. 1979), and simulta-
neous precipitation (Gronqvist and Arvin, 1979).
Very promising results have been obtained with
methods where phosphorus is chemically reduced
before the final precipitation step. Simultaneous
precipitation followed by contact filtration gave an
average effluent concentration of 0.24 g phosphor-
us/m3. The operational cost at this small plant (about
2,000 people) was reduced by about $8,000 per year
(Holmstrom, 1979b). Recirculation of post-precipitated
sludge to the activated sludge process has improved
effluent quality at reduced cost. Examples from
Uppsala and Eskilstuna have been reported (Hultman,
1979; Forsberg, 1977), where the effluent phosphorus
was decreased to about 0.3 g/m3- The recirculation
makes it possible to decrease the precipitant dose. In
1977 this reduced the cost for the precipitant in
Uppsala (200,000 people) by about $60,000 (Forsberg,
1977). Sludge with improved settling and dewatering
properties is also often obtained as a result of this
recirculation of chemical sludge.
Two-stage precipitation, i.e., simultaneous precipita-
tion followed by post-precipitation, also gave con-
centrations of effluent phosphorus corresponding to
0.3 g/m3.
Other methods tested are regulating the alkalinity of
the wastewater to reduce the requirement of alumi-
num sulfate or lime in post-precipitation, and using
automatic control to save chemicals and energy
(Hultman, 1979).
DECREASING POLLUTION LOAD
The comprehensive development of municipal
wastewater treatment has markedly reduced the
pollution load on Sweden's water courses and coastal
waters. The biological oxygen demand (BOD ?) load was
about 80,000 tons/year around 1960; the phosphorus
load above 7,000 tons/year at the end of 1960. At the
end of 1970 these figures had been lowered to about
20,000 and 2,500 tons/year, respectively(Falkenmark,
1977). For the city of Uppsala the phosphorus load has
been reduced to that observed about 50 years ago
(Forsberg, 1979), a situation probably prevailing in
many cities served by advanced wastewater treatment
It must also be mentionea that intensified anti-
pollution efforts within the industry have markedly
contributed toward reducing the total pollution load on
Swedish water bodies (Falkenmark, 1977). Both
municipal and industrial anti-pollution measures have
been supported by State grants.
RECOVERY OF POLLUTED WATERS
The decreasing pollution load has improved many
Swedish waters. Table 6 shows decreasing phosphor-
us concentrations and increasing transparency in
brackish and fresh waters in the Stockholm areas and
in two of the largest lakes in Sweden. Transparency
here mainly reflects algal turbidity. Because of the
improved conditions in Lake Malaren, open air bathing
has once again become possible in the most central
parts of Stockholm.
To study more in detail the effects of nutrient
removal, a comprehensive program was started by the
National Swedish Environment Protection Board in
1972 for analyzing the loadings on and the conditions
in a number of different recipient lakes (Forsberg,
Ryding, and Claesson, 1975). Results from some lakes
showing both improvements and delayed recovery
have been presented (Ryding and Forsberg, 1976;
Forsberg, et al. 1978). Results from 22 lakes have been
evaluated and briefly summarized in Table 7. The
majority of these lakes have responded positively with
lowered concentrations of total phosphorus and
organic matter. Half showed lowered chlorophyll a
values, but in only six lakes did transparency increase
significantly. Nitrogen increased in 10 lakes. Increased
nitrogen values have also been observed in other
Swedish waters analyzed during the 1970's. The four
lakes showing decreasing nitrogen content are lakes
where sewage has been totally diverted.
Correlations between phosphorus and chlorophyll a
and between chlorophyll a and transparency have been
presented for these lakes (Forsberg and Ryding, 1979).
In waters where transparency is principally influenced
by algal turbidity a correlation can be expected between
phosphorus and transparency, at least within the
concentration range where phosphorus is the primary
algal growth-limiting nutrient. Similar correlations
have also been demonstrated (Lee, Rast, and Jones,
1978). Table 6 indicates close correlations between
these parameters.
-------
URBAN AND POINT SOURCE POLLUTION CONTROL TECHNOLOGY
301
Table 6. — Total phosphorus and transparency in the
Stockholm area (Riddarfjarden, Blockhusdden, Tralhavet,
Cronholm and Bennerstedt, 1978, central part of Lake
Malaren (S. Bjorkfjarden) and Lake Vattern, Ahl, pers. comm.
LAKE GLANINOGN
Water body
Riddarfjarden
Blockhusudden
Tralhavet
S. Bjorkfjarden
L. Vattern
Period
1968-70
1971-73
1974-76
1968-70
1971-73
1974-76
1968-70
1971-73
1974-76
1965-69
1974-78
1968-70
1978-80
Total-P, g/m3 Transparency, m
0.072
0.040
0.032
0.173
0.091
0.049
0.060
0.048
0.027
0.035
0.020
0.010-0.015
0.007-0.008
2.1
3.1
4.5
1.9
2.0
2.2
2.4
2.6
2.9
3.0
4.3
7-8
10-12
Table 7. — Change in water quality observed in 22 lakes after nutrient
removal.
Number of lakes
Changes in Total Total Organic Chloro- Transpa-
concentrating nitrogen phosphorus matter phyll a rency
Decreasing
No signifi-
cant change
Increasing
4
8
10
14
7
1
15
7
0
11
9
2
3
13
6
In heavily polluted (hypertrophic) lakes a reduction of
phosphorus will not markedly decrease the chlorophyll
a concentration and thereby increase transparency
until the phosphorus concentration is depressed below
0.1 to 0.2 g/m3. This is illustrated in Figure 1, where
seasonal averages of phosphorus are plotted against
the corresponding values of transparency for 12 of the
22 lakes evaluated and listed in Table 7. Analyses
showed that a comparatively large change in annual P-
load must occur, a reduction by about 70 percent of the
pre-diversion data, to achieve any significant im-
TOTAL PHOSPHORUS
CU Of 0.8
Figure 1. — Total phosphorus versus transparency in 12
wastewater receiving lakes. Surface water (0-2 m). Average
values based on one sample/week, June-September. Bo =
Lake Boren, Dj = L. Djulosjon, Ek = L. Ekoln, Fi = L. Finjasjon,
Ha = L. Hacklsjon, Ka =L. Kalven, Ky = L. Kyrkviken, Ma = L.
Malmsjon, Ry = L. Ryssbysjon, SB = L. Sodra Bergundasjon, Sa
= L. Sabysjon, Tr = L. Trehorningen. For geographical positions
see Forsberg and Ryding, 1979.
Figure 2. — Phosphorus content in lake water, discharge and
external in- and output of phosphorus in Lake Glaningen and
Lake Malmsjon after diversion of sewage in early 1974.
Monthly average values, 1972-1977.
provement in clarity in Lake Boren and Lake Ekoln
(Forsberg, et al. 1978)..A reduction by 30 to 40 percent
revealed less marked .improvements.
A great amount of phosphorus is known to be
released — and recycled — from the sediments in
shallow, polluted lakes. The recovery after decreased
external nutrient load in these lakes will be delayed, as
seen in several lakes treated in Table 7. A high water
flow through a lake is favorable for washing out
phosphorus, illustrated (Figure 2) by data from two
lakes where sewage was totally diverted in 1974. Lake
Glaningen, having a higher flushing rate than Lake
Malmsjon, attained stable conditions more rapidly. The
phosphorus decrease in Lake Malmsjon was still
significant during the fourth summer period after the
diversion. A similar trend has also been reported for
Lake Norrviken (Ahlgren, 1977).
DISCUSSION
When the rapid development of advanced waste-
water treatment for phosphorus removal started in
Sweden about 10 years ago, the knowledge of
biological-chemical treatment was comparatively limit-
ed. The rapid development was positive and valuable,
greatly reducing pollution loads and improving the
water quality in rivers, lakes, and coastal waters.
Experiences obtained during this decade indicate,
however, that several treatment plants do not operate
as efficiently as expected. Depending on lack of
-------
302
RESTORATION OF LAKES AND INLAND WATERS
operating experience and guidelines and unsuitable
process technology and equipment, effluent values of
total P were not below the required limit, 0.5 g/m3 for
about 40 percent of the plants (Hultman, 1978).
Therefore, it seems necessary to improve or change
processes to obtain better effluent quality at reduced
operating cost.
As most of the Swedish population now is served by
biological-chemical treatment plants, it seems un-
realistic to expect that new advanced technologies will
be introduced if they are not adaptable to already
operating plants. Therefore, further efforts to improve
tt\e effluent quality and to reduce the operating costs
will be concentrated on recirculation of chemical
sludge, stepwise precipitation, and additional steps,
e.g., by filtration. To maximize the efficiency it will also
be important to have an effective emergency service in
case of technical mishaps. It will be also necessary to
have well prepared and competent employees handling
the plants, especially when the treatment processes
become more complex.
The recovery of polluted waters will be influenced by
many different factors, such as hydrological and
morphometrical conditions, the size and rate of
nutrient reduction, the chemical and biological charac-
ter of the recipient water, etc., making it a complicated
process. For a more general evaluation of the effects of
improved wastewater treatment, results and experi-
ences over a long period and from a great number of
different waters are needed.
The Swedish experiences show that natural waters
can recover rapidly after nutrient removal, e.g., by
advanced wastewater treatment (Forsberg, et al. 1 978).
To obtain visible results a comparatively large change
in the annual P-load must occur. The improvement is
easier to achieve in deep stratified waters {Table 6)
than in shallow lakes influenced by internal loading
from the sediments (Forsberg, 1979; Ahlgren, 1977;
Bengtsson, et al. 1975; Ryding, 1978). In several
shallow, polluted lakes the internal loading is a serious
problem during the vegetation period, i.e., that period
when public concern for good water quality is at a
maximum (Ryding and Forsberg, 1977, 1980a; Fors-
berg and Ryding, 1980).
Strong winds induce an increased vertical mixing
and thereby an increased internal loading, implying
that climatic fluctuations have to be monitored when
studying the response to nutrient removal measures in
shallow lakes (Ryding and Forsberg, 1977; Forsberg,
1978). Multiplying the duration of critical wind
directions by the force of the wind produces a close
correlation between "stirring capacity" and the
chlorophyll concentration (Ryding and Forsberg,
1980b).
During high phosphorus content in these shallow,
internally loaded lakes (July-September) the water flow
through the lake is often very low. To avoid recycling, a
high flushing rate is therefore necessary for wash out of
substantial amounts of phosphorus. To improve
conditions in the shallow, hypertrophic Lake Finjasjon in
Skane (south Sweden), where advanced wastewater
treatment does not help, a temporary damming (within
natural fluctuation levels) during July-August, followed
by a rapid lowering of the water level has been
suggested as a way to wash out comparatively large
amounts of phosphorus (Ryding and Forsberg, 1980a).
Even in deeper and larger lakes the hydraulic
residence time has been found to accurately describe
the trophic state (Vollenweider, 1975; Sonzogni,
Uttormark, and Lee, 1976). One way to further refine the
nutrient load-lake response concept is to apply an
estimate of the load that is more related to the actual
hydrological conditions for each separate lake for the
growing season compared to annual loading figures.
Calculations of a so-called hydraulic relevant phos-
phorus load (i.e., the amount of imported phosphorus
during the growth period and one "filling time'' prior to
it) adequately described the summer phosphorus
content in Lake Boren (Ryding and Forsberg, 1980b). As
is evident, Sweden has gained much experience in
nutrient removal; results are both positiveand negative.
The big "cleaning up" occurred during a period of good
economy. Today, costs and energy problems make it
necessary to find new and more inexpensive methods.
The Lake Finjasjon model (Ryding and Forsberg, 1 980a)
for lake restoration may be one approach.
REFERENCES
Ahl, T. Personal communication.
Ahlgren, I. 1977. Role of sediments in the process of recovery
of an eutrophicated lake. Pages 372-377 in H. L. Golterman,
ed. Interactions between sediments and fresh water. Junk,
The Hague.
Balmer, P., et al. 1979. Guidelines for monitoring prog-
rammes. In Management of municipal wastewater treat-
ment plants. Rep. 4 (in Swedish). Nordic Coop. Organ. Appl.
Res.
Bengtsson, L., et. al. 1975. The Lake Trummen restoration
project. I. Water and sediment chemistry. Verh. Int. Verein.
Limnol. 19:1080.
Cronholm, M., and K. Bennerstedt. 1 978. Water conditions in
the Stockholm archipelago after the introduction at biological
and chemical purification of wastewater. Prog. Water
Technol. 10:273.
Elfving, E., A. Forsberg, and C. Forsberg. 1975. Minitest
method for monitoring effluent quality. Jour. Water Pollut.
Control Fed. 47:720.
Falkenmark, M. 1977. Water in Sweden. Natl. Rep. U.N. Water
Conf., Ministry Agric.
Forsberg, B. 1978. Phytoplankton in Lake Uttran before and
after sewage diversion. PM 1029. Natl. Swed. Environ. Prot.
Board.
Forsberg, C. 1977. Advances in eutrophication control in
Sweden. Proc. Seminar on lake pollution and eutrophication
control, Killarney, Ireland, May.
1979. Responses to advanced wastewater
treatment and sewage diversion. Arch. Hydrobiol. Beih.
13:278.
Forsberg, C., and S. 0. Ryding. 1979. Eutrophication
parameters and trophic state indices in 30 Swedish waste-
receiving lakes. Arch. Hydrobiol. 89:189.
1980. Water quality in Lake Ringsjohn 1975-1979.
Mellanskanes Planeringskommitte, Rep. 1980:1 (in
Swedish).
Forsberg, C., S.-O. Ryding, and A. Claesson. 1975. Research
on recovery of polluted lakes. A Swedish research program
on the effects of advanced wastewater treatment and
sewage diversion. Water Res. 9:51.
Forsberg, C., et al. 1978. Research on recovery of polluted
lakes. I. Improved water quality in Lake Boren and Lake
Ekoln after nutrient reduction. Verh. Int Verein Limnol.
20:825.
-------
URBAN AND POINT SOURCE POLLUTION CONTROL TECHNOLOGY 303
Gronqvist, S., and E. Arvin. 1979. Evaluation of simultaneous
precipitation. In Management of municipal wastewater
treatment plants. Rep. 5 (in Swedish). Nordic Coop. Organ.
Appl. Res.
Gronqvist, S., et al. 1978. Experiences and process
development in biological-chemical treatment of municipal
wastewaters in Sweden. IAWPR 9th Int. Conf. Stockholm,
Sweden, June 12-16.
Holmstrom, H. 1979a. Evaluation of continuously operating
measuring instruments. In Management of municipal
wastewater treatment plants. Rep. 3 (in Swedish). Nordic
Coop. Organ. Appl. Res.
1979b. Simultaneous precipitation in combination
with contact filtration. In Management of municipal
wastewater treatment plants. Rep. 2. Nordic Coop. Organ.
Appl. Res.
Hultman, B. 1978. Chemical precipitation in Sweden —
present situation and trends in process improvement and
cost reduction. Paper presented at 2nd Int. Congr. Environ.,
Paris, December 4-8.
: 1979. Reduction of phosphorus at municipal
wastewater treatment plants. S.W. Water Waste Water
Works Assoc. (In Swedish).
Lee, G. F., W. Rast, and R.A. Jones. 1978. Eutrophication of
waterbodies: Insights for an age-old problem. Environ. Sci.
Technol. 12:900.
The National Swedish Environment Protection Board. 1972.
Household detergents and water protection. (Typewritten
rep.)
1979. Sewage treatment in densely populated
areas in Sweden, January 1, 1978.
NORDFORSK. 1978. Seminar on wastewater filtration. In
Management of municipal wastewater treatment plants.
Rep. 2 (in Swedish). Nordic Coop. Organ. Appl. Res.
Ryding, S.-O. 1978. Research on recovery of polluted lakes.
Loading, water quality and responses to nutrient reduction.
Acta Univ. Upsal. Abstr. Uppsala dissertations. Faculty of
Science. No 459.
Ryding, S.-O, and C. Forsberg. 1976. Six polluted lakes: A
preliminary evaluation of the treatment and recovery
processes. Ambio 5:151.
1977. Sediments as a nutrient source in shallow,
polluted lakes. Pages 227-234 in H. L. Golterman, ed.
Interactions between sediments and fresh water. Junk,The
Hague.
1980a. Lake Finjasjon 1976-1978. Hydrology,
loading and water quality. Rep. from the Natl. Swed.
Environ. Prot. Board. Research on recovery of polluted lakes.
Inst. Physiologi. Bot., Uppsala (in Swedish.)
1980b. Short-term load-response relationships in
shallow polluted lakes. SIL Workshop on Hypertrophic
Ecosystems. Hydrobiology (in press).
Sonzogni, W. C., P. D. Uttormark, and G. F. Lee. 1976. The
phosphorus residence time model. Water Res. 10:429.
Stahre, P. 1978. Flow equalization in sewer systems. In
Management of municipal wastewater treatment plants.
Rep. 1 (in Swedish). Nordic Coop. Organ. Appl. Res.
Tullander, W. 1975. Final disposal of municipal sludge in
Sweden. Jour. Water Pollut. Control Fed. 47:688.
Ulmgren, L. 1975. Swedish experiences in chemical
treatment of wastewater. Jour. Water Pollut. Control Fed.
47:696.
Vollenweider, R. A. 1975. Input-output models with special
reference to the phosphorus loading concept in limnology.
Schweiz. Z. Hydrobiol. 33:53.
-------
304
STORMWATER POLLUTION CONTROLS
FOR LAKE MANAGEMENT
WILLIAM C. PISANO
GERALD L ARONSON
Environmental Design & Planning, Inc.
Cambridge, Massachusetts
ABSTRACT
This paper presents an overview of an on-going stormwater management project for Lake
Quinsigamond, Mass. The overall project is funded by the National Urban Runoff Program (NURP)
and Section 314 Lakes Restoration Program. Lake Quinsigamond is located in central
Massachusetts and is the deepest manmade lake in the State. Presently, pollution point sources
have been eliminated. However, urban/commercial growth is high in the watershed and
eutrophication has been notably quickened by the accelerated stormwater solids, organic and
nutrient loadings. It is envisioned that solids separation devices may be an important low-level
structural control for eliminating high concentration urban runoff solids (and nutrient) loadings to
the lake. Current settleability tests of collected stormwater samples are also described in this
paper. These tests are meant to determine the relative fractions of easily removable floatable,
settleable, and suspended solids versus the more difficult light, colloidal material.
BACKGROUND INFORMATION
Lake Quinsigamond is located in the middle of
Massachusetts between the City of Worcester and the
Town of Shrewsbury. The extreme southernmost
portion of the lake lies within the Grafton boundary.
Figure 1 shows Lake Quinsigamond and its tributary
system. The lake lies in a north-south direction and is
crossed by three major highways: Interstate 290, Route
9, and U.S. Route 20.
Lake Quinsigamond is separated into two distinct
sections: The deep, narrow, northern basin and the
shallow southern basin known as Flint Pond. The total
area of the lake is 312 hectares comprised of 192
hectares in the northern basin and 120 in Flint Pond.
The lake has a maximum depth of 26 meters and an
average depth of 6 meters. The lake is approximately 8
kilometers long with the width varying from 76 meters
to 1.6 kilometers. The lake volume is estimated at
19.43 million cubic meters.
Being situated in a highly urban area, the lake
supports multiple recreational uses including fishing,
boating, water skiing, and bathing. The entire periphery
of the lake is densely settled with many private homes
and some commercial establishments. Two State
parks, several private beaches, and marinas are located
along the shorefront.
Lake Quinsigamond presently meets the water
quality standards required for water contact recreation.
The main body of Lake Quinsigamond has passed
through the mesotrophic stage and is in an early
eutrophic stage. Although the water quality of the lake
is satisfactory, intensive development of the drainage
basin has accelerated the lake's natural aging process,
and may limit the lake's recreational value in the
future.
Stormwater runoff from the drainage basin is
believed to be the major factor causing the accelerated
rate of eutrophication. Stormwater contributes signifi-
cant loadings of phosphorus and inorganic nitrogen to
the lake. Stormwater carries large amounts of solids
into the lake, increasing the turbidity of the lake water
and creating sandbars that make boating hazardous
and provide areas for rooted aquatic plant growth.
Stormwater degrades the bacteriological quality of the
lake.
Concern about the deteriorating water quality
combined with the tremendous desire to use the
recreational assets of the lake has produced wide-
spread concern for the future of Lake Quinsigamond.
Consequently, over the last several years investiga-
tions of the water quality to the lake and its feeder
streams have been undertaken by State and local
agencies, conservation groups, university departments,
and private citizens.
Recently, U.S. EPA awarded to the State of
Massachusetts (Division of Water Pollution Control
(DWPC) and Division of Environmental Quality Engi-
neering (DEQE)) Section 314 Lake Restoration and
National Urban Runoff Program (NURP) funds to
develop a pollution-related lake management program
for Lake Quinsigamond. Both projects are currently
underway. The DWPC is conducting the Section 314
diagnostic study. In January 1980 the DEQE solicited
engineering services to prepare the NURP stormwater
management plan for Lake Quinsigamond. Environ-
mental Design & Planning, Inc., Cambridge, Mass, was
awarded the overall engineering study. Meta Systems,
Inc., Cambridge, Mass., a subcontractor, will perform
the water quality impact modeling analysis.
-------
URBAN AND POINT SOURCE POLLUTION CONTROL TECHNOLOGY
305
LAKE QUINSIGAMOND NURP
PROGRAM
The objectives of the NURP project for Lake
Quinsigamond are as follows:
1. Develop an overall framework control strategy and
plan for mitigating the impact of nonpoint source
pollution on Lake Quinsigamond to achieve/maintain
class B water quality;
2. Focus on and develop reasonably detailed
engineering information at several catchment sites
supportive of implementation of stormwater treat-
ment/control demonstration facilities as part of the
continuing 314 program effort;
3. Develop a calibrated methodology for estimating
causal relationships between pollutant emissions and
water quality impacts for continuing planning, control,
monitoring efforts; and
4. Develop a sound data base of quantified land
use/emission pollutant loadings, rainfall/impervious-
ness/runoff characteristics, and effective control/
treatment alternatives that can be input into the NURP
data files as well as provide information for similar
studies in the New England region.
The work program for the Lake Quinsigamond NURP
project is shown in Figure 2. The 314 program is in
concurrent operation but is not described for the sake
of brevity. A short description will be presented of only
the stormwater measurement and control program
formulation tasks with emphasis on settleability
experiments and potential solids separation devices.
Task 1: Stormwater Measurement Program
The overall stormwater measurement program
consists of three parts. The first program uses
automotive equipment to measure flow/water quality
at five locations within the Lake Quinsigamond
watershed for 20 storm events. This information is
meant to better define the land-use emission factors
used in the runoff models. The second program
consists of manual grab sampling at a number of
secondary sites concurrent with the primary program.
The aim of the secondary program is to obtain auxiliary
data at other locations in the watershed. The third and
final program entails obtaining, during storm events,
large samples (151.4) liters of runoff at several key
measurement locations and performing settling col-
umn tests. These tests will be used to define types of
realistic controls in the watershed. The settling column
tests will help to define the relative fraction of
grit/easily settleable versus light colloidal material.
Nutrient analyses will also be performed as part of the
settling analysis so that relative fractions of nutrients
attached to particles of differing sizes (settling
velocities) can be ascertained.
Stormwater Solids Settleability Characteristics
Efficient and rational designs of solids separator
devices center on the knowledge of the settling velocity
characteristics of the solids particles and fractions to be
removed from them. These devices may play an
important role in the 314 implementation program for
Lake Quinsigamond.
Since eutrophication of the lake is a major issue, the
effectiveness of solids separating devices will also
depend upon the partitioning of nutrients (especially
phosphorus) between the dissolved and suspended
fractions. Samples will be collected to permit mea-
surement of each fraction. Because of the distribution
of particle sizes and the general tendency for smaller
(less easily removed) particles to contain/absorb
greater quantities of phosphorous per unit mass,
settling tests and evaluations of solids' concentrating
devices should include direct phosphorus measure-
ment. The bioavailability of the sediment phosphorous
phase will be assessed on some representative fraction
of the samples using algal growth potential tests
and/or extraction procedures.
Settling characteristics of urban runoff are difficult to
determine accurately using conventional procedures
because of the presence of both large (quick to settle)
particles such as sand and grit, and small, light
fractions (long settling times). Conventional procedures
such as hand-operated rotation and stirring or using
compressed air for pre-mixing create undesirable
characteristics including solids degradation, incom-
plete mixing, and generation of eddies and currents. A
U.S. EPA study, "Characterization of Urban Runoff
Settleability Characteristics," describes a new method
developed by Environmental Planning & Design for
obtaining representative and accurate characteristics.
The state-of-the-art column is shown in Figure 3.
The concept encompassed steady-state, bi-direc-
tional rotation coupled with flow stators inside the
cylinder to facilitate mixing. Two electric motors were
wired through variable speed controllers to offer a high
degree of uniformity and flexibility to the mixing
process. The flow stators can be conceptualized as
minimum disturbance deflectors assisting mixing by
developing uniform, low velocity currents opposing the
centrifugal forces generated by axial rotation. This
balance of forces was considered an attempt to gain
uniform distribution of solids across a cross-section of
the column. The longitudinal rotation was the
mechanism by which the solids would be dispersed
throughout the horizontal axis of the column. The
device has been used with artificial media of known
particle size and seems to closely replicate theoretical
settling rates. Comparative investigations using the
new device and conventional methods such as air
diffusion and plunger mixing showed significant
differences in settling velocity curves for the same
sample of combined sewer overflow.
The importance of settleability information for
rational design of solids separating devices is depicted
in Figure 4. Combined sewer overflow samples (151.4
liters) were obtained from three locations in Dorchester
and were analyzed using the new approach. Swirl
regulators can be designed to remove particles with
settling velocities exceeding 0.15 cm/sec. Complete
removal of grit (5 cm/sec.) can be expected using a
properly designed swirl. The three figures show a
range of partial solids removal and the two areas of no
removal (A) and complete removal (B).
-------
306
RESTORATION OF LO.KES AND INLAND WATERS
Task 2: Develop Control Program
Pollutant removal levels required to allow the lake
and each of its net urban drainage tributaries achieve
the water quality criteria established in the water
quality goals for the lake defined in Task 2, will be
defined by interactive catchment area emission/water
quality input analysis. Emission modeling approaches
will be used to roughly ascertain hydrologic design
criteria such as design storm capture volume/rate/
frequency coupled with expected pollutant loadings.
The settleability results will be used to further
fractionate controllable pollutant loads by class of
treatment/control devices, i.e., what can be expected
by 'solids separators such as Swirl/Helical Bend
regulators versus removable loads by street sweeping,
and microstrainers. In-situ lake treatment control
techniques will also be investigated if the required
removal of solids/organics/nutrients cannot be fea-
sibly attained by emission control of pollutants entering
the lake.
PHYSICAL TREATMENT SOLIDS
SEPARATION DEVICES
Physical treatment alternatives are primarily applied
for removing floatable, settleable, suspended solids
and their associated pollutants from wastestreams,
and are particularly important to stormwater and
combined sewer overflow treatment. Physical treat-
ment systems have demonstrated a capability to handle
high and variable influent concentrations and flow
rates and operate independently of other treatment
facilities, with the exception of treatment and disposal
of the sludge/solids residuals.
Swirl Regulator/Concentrator
The Swirl concentrator has demonstrated the
capability to handle high and variable influent
concentrations and flow rates with relatively high
removal efficiencies. (See Figure 5-A.) This device can
be designed to completely remove sand and grit,
partially remove (40 to 60 percent) lightweight
settleable particles, and substantially remove (45 to 80
percent) floatable solids at a fraction of the detention
time (1 to 2 minutes) normally required for convention-
al sedimentation. Swirls are designed for hydraulic
loading rates ranging from 37,850 to 151,400
liters/mVday, depending on the application. The
device is perfectly suited for treating intermittent
discharges (wet weather runoff) containing both
settleable and floatable pollutants.
Helical Bend Regulator/Concentrators
Helical bend regulator/concentrators have been
modeled, and design criteria as well as comparative
cost evaluations have been developed and are
presented in handbook form. Helical bends appear
practical as in-line regulator devices commensurate
with swirl. (See Figure 5-B.)
West Roxbury Swirl/ Helical Bend R&D Facility
A major U.S. EPA effort is underway involving the
design, fabrication, installation, and operation of two
full-scale state-of-the-art stormwater pollution abate-
ment devices. A treatment complex consisting of a
swirl and helical bend regulator solids separator are
being tested, side by side, for their ability to remove
stormwater and simulated combined sewer pollution
loads from a 65 hectare catchment area situated in
West Roxbury, Mass., tributary to the Charles River.
Environmental Planning & Design has designed, shop-
fabricated, field-installed, and is monitoring wet
weather pollutant removal effectiveness of the two
devices over a 1-year period.
A plot plan of the facility is shown in Figure 8. The
design flow for each unit (based on 3-week recurrence
interval) is 6 cfs. Both units can be driven up to peak
discharge of 18 cfs each. Discharge into each unit is
evenly split using motor-activated bottom-opening
sluice gates. Foul sewer underflows from both units
containing the removed pollutants are flow controlled
at 3 percent of the unit design flow by Hydrobrakes.
These devices provide nearly uniform discharge under
fluctuating lead conditions. No clogging problems have
been experienced to date. Pertinent dimensions of the
swirl are as follows: diameter— 3 meters, height— 1.4
meters, influent/effluent (clear) diameter — 0.6, and
foul sewer effluent — 0.3 meters. The helical bend is
20 meters long and is 1.3 meters high; piping sizes are
similar. The evaluation program commenced late in
1979 and will continue to the end of 1980. Pollutant
removals to date approximate primary treatment.
It is believed that the swirl and helical bend flow
regulator/solids separators will be very useful to
communites as inexpensive, maintenance-free tools
for combating stormwater pollution problems. For
storm drain systems these devices can be installed on
separate storm drains before discharge and the
resultant foul underflow could be stored in relatively
small tanks since concentrate flow is only a few
percent of total flow.
In another approach, devices such as the swirl
degritter or sedimentation basins may be used to
provide final dewatering of the concentrate underflow
suitable for disposal. Stored underflow could be later
directed to the sanitary sewer for subsequent
treatment during low-flow or dry-weather periods, or if
capacity is available in the sanitary system, the foul
underflow may be diverted without storage. This
method of stormwater control would be cheaper in
many instances than building huge holding reservoirs
and it offers a feasible approach to treating separately
sewered urban stormwater.
-------
307
AN EXAMPLE OF URBAN WATERSHED MANAGEMENT
FOR IMPROVING LAKE WATER QUALITY
MARTIN P WANIELISTA
YOUSEF A. YOUSEF
Department of Civil Engineering and Environmental Sciences
University of Central Florida
Orlando, Florida
ABSTRACT
Many investigators have identified the urban environments as those producing high levels of
water pollutants relative to other land uses. In a 55 hectare (136-acre) urban watershed in
Orlando, a stormwater system discharges to an 11 -hectare (27-acre) lake. The lake water quality is
.characterized by frequent algal blooms, odor, and in general, reduced human activities. The lake is
one of the focal social areas of the city. Previously reported work on algal assays, bottom mud
inactivation, and trophic analysis indicated that a mass of phosphorus should be removed to
reduce algal blooms and improve the general aesthetic appearance of the lake. Lake water quality
and stormwater impacts not previously published are presented in this paper. Stormwater runoff
pollution mass and concentrations were estimated from a hydrograph related and composite
sampling program. The average loadings and concentrations were compared to national data. A
wide range of values was noted among storm events. Stormwater management procedures were
established based on the runoff sampling program and a target reduction of phosphorus and
metals. Estimates for the cost and benefits of the abatement program were completed.
Management of stormwater for the removal of phosphorus was accomplished by diversion for
retention of the first flush of pollutants. The efficiency cost curves were estimated from field
performance data. For average yearly removals over 80 to 85 percent per year, these curves reflect
rapidly increasing cost. Below 80 to 85 percent linear curves were typical.
INTRODUCTION
Stormwater may be a significant source of surface
water pollution in urban areas (Weibel, 1969;
Wanielista, 1977; Yousef, 1980). Lake impacts have
been and continue to be studied on an international
level. There exists in the United States a National
Eutrophication Research Program (Gakstatter, 1975)
and an international program with U.S. participation
(Rast, 1978).
This paper documents stormwater impacts on an
urban lake. The impact was first defined by visual
observation. In a U.S. Government funded 208 program
(East Central Florida, 1978), stormwater was reported
to be the major pollution source. There were no point
sources of industrial or domestic wastewaters. Thus,
an investigation of the stormwater impacts was
initiated and the results are reported here and
elsewhere (Yousef, 1980). It was necessary to estimate
stormwater composition, mass loadings, and impacts
to determine a combination of management practices.
Evaluations of stormwater management practices
have been completed prior to this work, such as those
for urban areas (Field, 1977), and others (Wanielista;
1978). However, the critical relationship between a
management practice and receiving water quality has
not been well documented, except for some dissolved
oxygen responses in rivers. This work aids in
evaluating stormwater management practices to
reduce impacts on lakes.
WATERSHED AND LAKE
CHARACTERISTICS
The drainage area studied here is the Lake Eola
watershed located in central Florida within the city of
Orlando. The stormwater system is separate from the
sanitary sewage system. The stormwater system
drains a watershed of approximately 55 hectares (136
acres), composed of 31.7 hectares (78.2 acres) of
commercial and 23.5 hectares (57.8 acres) of residen-
tial areas discharging to an 11 -hectare (27-acre) lake.
In addition, 4.5 hectares (11.2 acres) of parkland
surrounds the lake. The watershed area was deter-
mined from storm sewer drawings and visual observa-
tion during rainfall events. Streets and parking lots
comprise approximately 16.6 hectares (41 acres) of the
watershed within a total of about 29.5 hectares (73
acres) of impervious lands. The pervious area is only 9
hectares (22 acres), most of which is in the residential
areas.
Thirty-five parking areas discharge onto the street
surfaces. Their total area is about 10 hectares (24
acres). These parking areas were identified as possible
areas for management of stormwater. Since runoff
waters discharge to the land-locked lake, one of the
parking lots was designated a sampling location for
runoff waters.
The 11 hectare (27 acre) lake is known for its
picturesque setting. Its picture is the logo for the city of
Orlando. It was once a natural lake, and historical
-------
308
RESTORATION OF LAKES AND INLAND WATERS
records indicate that surface waters did not discharge
from the lake. The level of the lake is usually
maintained between 26.5 meters (87.0 feet) and 27
meters (88.5 feet) above sea level. Physical character-
istics of the lake are shown in Table 1. It is a shallow
lake with a mean depth of approximately 3 meters;
about 73 percent of its volume is located within the 0 to
3 meter frustrum layers. Most of the 5,000 urban lakes
in central Florida have similar physical characteristics,
BENEFIT
The benefits of the lake and its surroundings are
evident, but difficult to quantify. The lake is a city focal
point for residents and tourists, with frequent music
concerts, arts/crafts shows, a children's park, and
relaxation areas. The land values of property surround-
ing the lake bring top value because of its location. Lake
Eola is one of the main reasons for the economic health
of the downtown area. Estimates of the dollar benefits
from lake activity are presented in Table 2.
Table 1. — Physical characteristics for Lake Eola, Florida.
Lake Eola water was found to be somewhat alkaline
with pH ranging from 8.4 to 9.5. Measurements of pH
in Lake Eola indicate the rate of algal production. The
average annual value and the measured range of
values for pH, chlorophyll a, inorganic and organic
carbon, and Secchi disk transparency are shown by
Table 3.
The average values shown in Table 3 can be
compared to values reported in the literature for
eutrophic lakes, as indicated in Table 4. Lake Eola has
some of the characteristics of an eutrophic lake.
Table 3. — Values for selected parameters measured in Lake Eola, Fla.
between July 1978 and August 1979.
Number of Average Standard Range of
Parameter samples value Units deviation values
Chlorophyll a
Organic carbon
Inorganic carbon
PH
Secchi disk
64
67
68
57
32
25.4
10.9
18.8
8.9
106
mg/m±
mg/l
mg/l
—
cm
8.8
6.7
6.4
—
13.0
9.0- 36.4
3.0- 29.1
13.8- 40.6
8.4- 9.5
90 -120
Parameter
Quantity
Approximate surface area
Approximate volume
Mean depth
Maximum depth
Length of shoreline
Shoreline development
Volume development
109,270 mV11.0 hectares/(27.0 acres)
3.30 X 105m3/(8.73 X 10e gallons)
3.20 m/(9.92 ft)
6.8 m/22.3 ft)
1417 m/(4650ft)
1.21
1.72
Average height above sea level 26.8 m/(68 ft)
Table 2. — Estimated Lake Eola benefits.
Activity
Approximate Approximate
frequency/year people-visits/year
$/year
Music concerts1
Arts/crafts2
Tourist visits3
Fish-a-thons1
Food concessions2
Paddle boats2
Children's park1
Relaxation/aesthetics1
Jogging1
Land value"
35
3
Constant
3
Constant
Constant
Constant
Constant
Constant
Constant
TOTALS
87,500
60,000
180,000
3,000
—
5,000
125,000
200,000
50,000
—
710,500
262,500
300,000
90,000
9,000
100,000
20,000
187,500
600,000
150,000
600,000
2,319,000
' Based on estimated attendance and an expenditure of $3 per person
visit.
2 Based on concession money received by the City of Orlando and an
estimated attendance.
3 Greyline of Orlando estimated visits as a portion of a larger tour.
4 Based on lakefront vs. non-lakefront property taxes.
LAKE IMPACTS
This section summarizes the lake impact work
completed to date. A more complete report is published
elsewhere (Harper, 1980; Wanielista, 1980). Visual
observation and analytical data reveal that Lake Eola
has persistent algal blooms virtually year round.,
Populations of the macroscopic algae, Chara, and the
filamentous green algae, Spirogyra, covered up to 30
percent of the lake surface area during the summer
rainy season.
Table 4. — The range of values for selected parameters in eutrophic
lakes, as reported by Wetzel (1975).
Source
Chlorophyll a Total P Secchi disk Organic carbon
ma/m ua/l cm ma/I
Wetzel (1975)
EPA-NES (1974)
10-500
12
10-30
20
200
Concentrations of dissolved oxygen in Lake Eola,
although usually at or above saturation near the
surface, drop periodically during the spring and
summer months to less than 1 mg/l in deep areas of 4
meters or more water column. Phosphorus from the
bottom sediments was released up to a level of 250
mg/m2 after 2 months of anoxic conditions (Marshall,
1980). This anaerobically released phosphorus has the
potential for increasing water column phosphorus by
11.6//g POV3 - P/l,or about 50 percent of the average
orthophosphorus concentration in the lake (23/jg/l).
When the concentration of orthophosphorus in Lake
Eola was less than 0.10 mg/l, algal production was
regulated by adding orthophosphorus alone. Above this
concentration it appears that an excess of phosphorus
was available, and algal growth was regulated by the
N:P ratio. However, in most cases the concentration of
orthophosphorus in Lake Eola water was below 0.04
mg/l, and algal production was most likely limited by
the concentrations of added phosphorus alone.
In contrast to the enhanced algal growth conditions
experienced during the summer rainy months, runoff
entering the lake after prolonged periods of drought
also produces several toxic effects on aquatic life in
Lake Eola. Contaminants are allowed to accumulate
within the watershed, and when a storm event occurs,
the mass loading to the lake is many times larger than
that experienced during frequent rainfall periods. This
influx of toxic and oxygen-demanding wastes can kill
many forms of aquatic life. Evidence of such a
phenomenon was recorded in March 1979 when it
rained after a dry period of 6 weeks. Concentrations of
organic carbon as high as 400 mg/l were measured in
-------
URBAN AND POINT SOURCE POLLUTION CONTROL TECHNOLOGY
309
stormwater runoff entering the lake during this rain.
Two days later, dissolved oxygen concentrations had
been reduced from saturation near the surface to 4
mg/l at a depth of 1 meter and to near zero below 2
meters. Numerous large-mouth bass averaging 2 to 3
pounds were found floating in the water, and large
masses of dead filamentous algae had accumulated in
thick mats over much of the lake's surface. During
1979 a total of six fish kills were reported, one dead
bass fish (about 2 kilograms) floating at the surface was
brought to the laboratory, processed and analyzed for
metal concentrations in selected organs (heart, gall
bladder, liver, stomach) and flesh. From the limited data
available it appears that nickel and lead concentrated in
the gills, iron concentrated in the heart, and zinc and
copper concentrated in the liver. However, at this time,
it is not known that these metals were directly
responsible for the fish kill.
A pathogen isolation study was conducted over 1
year. One hundred twenty-nine water and sediment
samples were collected. Fourteen were composites of
runoff, 32 were bottom samples, and 83 were lake
water. Clostridium was isolated from the bottom
sediments of the lake and Salmonella was isolated
from the lake water samples.
There are domestic ducks in the Lake Eola waters
and park areas. On the average, they number
approximately 20, with decreasing populations noted
over the past 5 years. The population decrease is
believed to be caused by Clostridium botulism.
Microbiologists at the Orange County Pollution Control
(Adams, 1977) have speculated that during site visits,
gas production from the anaerobic sediments is
increased in the summer months. This anaerobiosis
promotes growth of the botulism organism which
produces a toxin which, in turn, concentrates in the
small insect larvae of the sediments. When ducks eat
the larvae they can die.
Two dead ducks were sent to the State Veterinary
Laboratory for autopsies. They were selected from
among 35 dead ducks by the Humane Society. After
autopsies, the Humane Society of Orlando reported
that botulism caused the duck deaths in the lake
(Orlando Sentinel Star, 1977).
STORMWATER
Stormwater pollutants and flow rate were first
estimated by sampling stormwater relative to the
hydrograph. Eight rainfall/runoff events were quanti-
fied in this manner. Next, a composite sampling
program was completed with seven rainfall/runoff
events. One major question was the percentage of
dissolved pollution materials present in the runoff. The
sampling program indicated that the dissolved nutri-
ents and organics were approximately 50 percent or
more of the total, while the dissolved fraction of lead
was 20 percent. From the 15 runoff samples, estimates
were made for average mass loading (as discharged)
and average concentrations. These averages are
shown in Table 5. Estimates of loading rates from both
commercial and residential areas were calculated from
the runoff studies.
The Lake Eola study loading site data are compared
with the loadings of SWMM/level I analysis (Heaney,
1976) and other national data (Wanielista, 1979). The
suspended solids and BOD data (Table 6) appear to
reasonably agree. However, total nitrogen data are
higher in the Lake Eola watershed. Possible reasons
are that the residential areas should be classified as
commercial areas when considering loading rate data,
the landscaping maintenance places an additional
nitrogen load, and the heavy rainfall (130 cm) is greater
than the national average. Most likely, a combination of
these reasons caused the increase.
Table 5. — Concentration and loading rate runoff summary (hydrograph related and composite sampling programs).
Parameter
Suspended solids
Volatile suspended
NVSS
BODs
COD
TOC
TKN
Ammonia-N
Total phosphorus
Zinc
Cadmium
Arsenic
Nickel
Copper
Magnesium
Iron
Lead
Chromium
Calcium
Sample
size (storms)
14
7
7
8
6
13
10
12
14
9
g
8
9
9
8
9
9
9
9
Mass loading
range (kg/ha-yr)
470 — 2,368
234 — 610
76 — 587
40 — 315
130 — 1,776
53-2,572
10 — 87
0.2 — 10.4
1.8-16.4
1.2 — 5.5
0.09—1.0
0.17 — 1.76
0.06 — 0.54
0.12—1.39
2.58-31.25
2.9—16.46
1.1—9.5
0.07 — 0.51
99.7 — 487
Loadings
kg/ha-yr
991.0
538.0
453.0
98.0
711.0
946.0
32.0
4.1
4.8
3.7
0.28
1.02
0.28
0.68
9.86
9.52
4.26
0.25
308.0
Averages*
Concentration
mg/l
131.0
71.0
60.0
13.0
74.0
99.0
3.3
0.43
0.48
0.38
0.03
0.11
0.03
0.07
1.03
0.99
0.44
0.03
32.10
"Both commercial and residential
-------
310
RESTORATION OF LAKES AND INLAND WATERS
Table 6. — Loading rate comparisons.
Lake Eola
Commercial
Residential
+SWMM/Level I
Commercial
Residential
++National averages
Commercial
Residential
+ Heaney, 1976
++ Wanielista, 1979
SS
1,076
827
1,255
922
941
470
BOD5
196
87
181
45
97
39
TOC TN
1,167 32.0
757 40.5
- 16.7
7.4
14.5
6.6
TKN PO4-P TP
27.8 1.7 3.5
36 1 31 6.2
4.3
1.9
3.0
2.0
The commercial and residential land use pollution
contribution to the total was estimated to be 98 percent
for suspended solids, 96 percent for BODs, 95 percent
for total organic carbon, 94 percent for TKN, and 91
percent for TP. The total contribution was defined as
the sum of stormwater, atmosphere, and ducks living
on the lake.
The sampling program and the lake impact work led
to the following conclusions: (1) stormwater is the
major source of lake related pollution; (2) phosphorus
and,other stormwater pollutants had to be removed; (3)
sedimentation was possibly not the choice method for
stormwater management because of the large per-
centage of dissolved pollutants.
TARGET PHOSPHORUS REDUCTION
The major question is to what degree should the
bottom sediment and stormwater be treated to
economically reduce the nutrient enrichment, fish and
duck kills, and algal activity to an acceptable level?
Using the trophic state models, a target reduction level
of phosphorus loadings in the oligotrophic/meso-
trophic level may reduce algal blooms. In addition, a
chlorophyll a mean concentration of 7 A9/I may
indicate a mesotrophic state. Table 7 illustrates the
target level and the need for an approximate 90 percent
reduction in phosphorus load and phosphorus concen-
tration.
Table 7. — Target reductions.
Models
Before
Target Reduction Levels
Vollenweider 2.33 g-P/sq m/year 0.2 g-P/sq m/year
Dillon 0.49 g-P/sq m 0.05 g-P/sq m
Larsen-Mercier 0.48 mg/l 0.05 mg/l
OECD/chlorophyll1 269 mg-P/m3 70 mg-P/m3
1 Reduction corresponding to a chlorophyll a of 7 /jg/l (Gakstatter.
1975)
In the National Eutrophication Study total phos-
phorus concentration of less than 10 /jg/\ in the water
column was noted as a target reduction to classify
lakes as oligotrophic. A combination of stormwater
treatment and bottom sediment inactivation may
produce a water column concentration of less than 10
//g/l. The bottom sediments were estimated to
contribute 11.6/t/g/l of the average water column
concentration of 23 /ug/l
STORMWATER MANAGEMENT
SELECTION
Each stormwater management practice that could be
defined in terms of cost and efficiency and was
practical for the watershed was evaluated for storm-
water control. The selection of the best combination of
practices was based on those which meet cost and
efficiency constraints. With many practices and control
locations within the Lake Eola watershed, the selection
of the best combination (least cost) could be aided by a
computer analysis.
Cost-efficiency curves (present value dollars versus
removal quantities) were developed for each subwater-
shed of the Lake Eola watershed. Removal efficiencies
from the literature (Field, 1977; Lager, 1977) and local
208 programs (Calabrese, 1977; Wanielista, 1979)
were used. These 208 efforts in the central Florida area
had defined the efficiencies and costs for diversion/
percolation basins, swales, underdrains, and vacuum
sweeping nonpoint source management methods. In
the highly impervious urban areas, the cost of land is
expensive, and land intensive activities (detention and
retention basins) are sometimes not aesthetically
pleasing. Thus, street sweeping, diversion with
retention underground, and catchbasin cleaning ap-
peared probable for urban areas. Dutch drains, rooftop
storage, coagulation, filtration, and concentrators were
other management methods under investigation.
These formed the basis for determining optimal
combinations of practices. This was accomplished
using a computer program written for this work. A
linear programing network routing model was in-
corporated. The cost-efficiency curves were estimated
by "piecewise" linear approximation (Calabrese, 1979).
One limitation on stormwater control was the use of
private property. Thus, it was decided to do all
management within the city right-of-way. The alter-
natives considered for management of the stormwater
1. Diversion of stormwater to the sanitary sewer
system for treatment.
2. Street cleaning by both broom and vacuum
sweepers.
3. Diversion of stormwater into percolation basins.
4. Conversion of inlets to catchbasins.
5. Coagulant addition with sedimentation.
6. Silt removal from lake, and drawdown every 5
years.
7. Natural "living filter" treatment.
8. Fabric bag filters.
9. "Best" combination of any or all of the above
alternatives.
10. Diversion of stormwater into infiltration trenches.
11. Others, such as sand filtration, swirl concen-
trators, and diatomite filters.
The first alternative was eliminated because it was
not considered as a general solution for other areas
and it required replacing over 7,000 meters of sanitary
sewer lines, thus the capital cost of pipe and pumping
stations was over $600,000. The number 11 alter-
-------
URBAN AND POINT SOURCE POLLUTION CONTROL TECHNOLOGY
311
natives were not developed because of lack of technical
data on performance (cost vs. efficiency data) in a storm
sewer system.
Before the percolation alternatives could be con-
sidered the infiltrative capacity of the soils was
estimated. This was done by defining the type of soils
and the location of the water table. The water table is at
least 2 meters (6 feet) below the ground surface for a
ground elevation of 29 meters (95 feet) or higher.
Borings close to the lake indicate the water table is
near elevation 27 meters (88 feet). In addition, sandy
soil is available to about 6.5 meters (20 feet) below
ground level. Percolation of stormwater is possible for
parking lot and street drainage for those areas whose
ground elevation is above 29 meters (95 feet).
All alternatives were evaluated in terms of estimated
cost and yearly pollutant removal efficiencies. The cost
for the natural living filter areas were estimated from
local contractors and the city of Orlando records. The
vegetation selected is native vegetation and has been
used in other lakes. All other cost data were obtained
from recently bid sewer projects.
The solution selected was based on minimum
present value cost and maximum removal efficiencies.
The fabric bag alternative had a lower capital cost, but
poor removals relative to other alternatives. The
locations of the best management practices were near
parking lots and immediately before lake discharge.
Parking lot and street diversion were designed to
percolate the first 0.6 to 1.25 cm (1/4 to 1/2 inch) of
every storm chosen. This results in a removal efficiency
of 90 percent on a yearly basis. The resulting capital
cost for stormwater management and lake restoration
was approximately $6,250/hectare (S2,500/acre) of
watershed.
CONCLUSIONS
Based on citizen concern and historical water quality
data on fish and duck kills, oxygen depletion, and algal
blooms, it was evident that the factors causing the
water quality impact had to be identified. Trophic state
analysis indicated that the lake was estimated as
eutrophic. In laboratory tests, algal productivity was
related to stormwater. Also, the bottom sediments
were shown to contribute to the phosphorus con-
centration in the water column.
Based on the runoff quality and quantity data with
lake limnological data, an implementation plan for
stormwater management was developed. Since phos-
phorus is most likely the limiting nutrient, it will be
controlled. The two major sources of phosphorus are
stormwater and lake bottom mud recycle. By reducing
stormwater phosphorus mass, re-stocking, littoral zone
planting, and coagulant coverage of bottom muds, it is
predicted that the effects of stratified conditions
(anaerobic) will be minimized and algal blooms will be
reduced.
The stormwater management will be done by
diversion/percolation of parking lot runoff and limited
street runoff (approximately 17 hectares, (40 acres)). In
addition, those areas not managed with this method
will be diverted into trench storage for infiltration into
the lake (approximately 28 hectares, (65 acres)).
REFERENCES
Adams, J. 1977. Lake Eola fish kill. Letter to Walt Lawson,
City of Orlando, Fla.
Calabrese, M. M. 1979. Optimization of stormwater
management practices and processes. M.S. Thesis. Univer-
sity of Central Florida, Orlando.
Calabrese, M. M., and M. P. Wanielista. 1977. Stormwater
management practices manual. East Central Fla. Regional
Plan. Counc. Orlando.
East Central Florida Regional Planning Council. 1978.
Orlando area 208. Winter Park.
Field, R., et al. 1977. Urban runoff pollution control
technology overview. EPA-600/2-77-047. Munic. Environ.
Res. Lab. U.S. Environ. Prot. Agency, Cincinnati, Ohio.
Gakstatter, J. H. M. 0. Allum, and J. M. Omernik. 1975. Lake
eutrophication results from the national Eutrophication
Survey. Corvallis Environ. Res. lab. U.S. Environ. Prot.
Agency.
Heaney, J. P., W. C. Huber, and S. J. Nix. 1976. Stormwater
management model: Level I — preliminary screening
procedures. EPA-600/2-76-257. U.S. Environ. Prot. Agen-
cy, Cincinnati, Ohio.
Lager, J. A. 1977. Catchbasin technology overview and
assessment. EPA-600/2/77-051. Munic. Environ. Res.
Lab. Cincinnati, U.S. Environ. Prot. Agency, Ohio.
Marshall, F. 1980. Phosphorus interactions with Lake Eola
bottom sediments. M.S. Thesis. University of Central
Florida, Orlando.
Orlando Sentinel. 1977. Botulism killed ducks. August 3.
Rast, W.,and G. F. Lee. 1978. Summary analysis of the North
American (U.S. portion) OECD eutrophication project-
Nutrient loading — lake response relationships and trophic
state indices. EPA-600/3-78-008. Corvallis Environ. Res.
Lab. U.S. Environ. Prot. Agency.
Wanielista, M. P. 1979. Stormwater management: quantity
and quality. Ann Arbor Science, Ann Arbor, Mich.
1980. Stormwater management to improve lake
water quality. EPA final Rep. to be published; available from
the University of Central Florida, Orlando.
Wanielista, M. P., and E. E. Shannon. 1978. Stormwater
management practices evaluation. East Central Florida
Regional Plan. Counc. 208 study. Winter Park.
Wanielista, M. P., Y.A. Yousef, and W. M. McLellon. 1977.
Nonpoint source effects on water quality. Jour. Water
Pollut. Control Fed. 12:441.
Weibel, S. R. 1969. Urban drainage as a factor in
eutrophication. Proc. Symp. Eutrophication. Natl. Acad. Sci.
Washington, D.C.
Yousef, Y. A. 1980. Proc. urban stormwater and combined
sewer overflow impact on receiving water bodies conf. To be
published by U.S. Environ. Prot. Agency. Cincinnati, Ohio.
ACKNOWLEDGEMENTS
The work reported in this paper was sponsored by the
Environmental Protection Agency, the Storm and Combined
Sewer Section, the State of Florida Department of Environ-
mental Regulation, the City of Orlando, and the University of
Central Florida.
-------
312
LAKE RESTORATION BY EFFLUENTS
DIVERSION IN FRANCE
GUY BARROIN
Station d' Hydrobiologie Lacustre
Institut National de la Recherche Agronomique
Thonon les Bains, France
ABSTRACT
Sewage diversion has been applied by French authorities to protect two main lakes against
eutrophication. The first one, Lake Annecy, is 27 square kilometers wide and showed signs of
deteriorating after World War II because of domestic sewage. A pipe was thus constructed to
collect these effluents all around the lake for treatment in a downstream purification plant. The
construction started in 1962 and was completed in 1976. The response to reduced nutrient influx
has been an obvious lake oligotrophication. The second lake. Lake Le Bourget, is 44 square
kilometers in area. Eutrophication had been deteriorating its water quality for almost 30 years. To
reduce the nutrient influx, essentially urban generated, most of the treated sewage was collected
and diverted outside the drainage basin into the Rhone by a 12.3 kilometer-long tunnel excavated
through a mountain. The diversion operation started in 1979 and no improvement has been
observed until now.
INTRODUCTION
Diversion of sewage from a lake that is being
enriched by the algal nutrients present in wastewater
is the radical solution French authorities have found for
the eutrophication problem in two main water bodies:
Lake Annecy and Lake Le Bourget. These lakes are
located in the same alpine area, less than 100
kilometers southwest of Geneva (Figure 1). Their main
physical characteristics are given in Table 1.
Table 1. — Lake Annecy and Lake Le Bourget: main physical
characteristics.
Lake Annecy Lake Le Bourget
Altitude above sea level (m)
Surface area (km2)
Maximum depth (m)
Mean depth (m)
Volume (km3)
Mean residential time (months)
Drainage area (lake excepted)(km2)
446.5 231.5
27.04 44.62
64.7 145.4
41.5 81 14
1.1235 3.6203
44.0 36-48
251.0 560.0
LAKE ANNECY AND ITS PERIPHERAL
COLLECTOR
Lake Annecy (Figures 2,3) began showing real signs
of eutrophication during the decade after World War II,
an evolution considered likely since 1943. Professional
and amateur fishermen were perturbed at the marked
decline of salmonid populations such as omble and
trout. They rightly associated this with the rapid
efflorescence of phytoplankton blooms resulting from a
higher concentration of nutrients. Furthermore, the
authorities responsible for monitoring the drinking
water supply drawn from the lake were more worried
by laboratory reports of rising bacteria counts,
especially during the summer. The people really
SWITZERLAND
LAKE 'I ^J A!X LES BAms
LE BOURGET
CHAMURV
Figure 1. — Location of Lake Annecy and Lake Le Bourget.
concerned by this deterioration were, of course, those
living in nearby towns and lakeside communities, a
population totaling 75,000. This group called in
authorities to assess the state of the lake. It embarked
simultaneously on a public education campaign, aimed
first at the leaders of the lakeside communities.
-------
URBAN AND POINT SOURCE POLLUTION CONTROL TECHNOLOGY
313
PURIFICATION PLANT
PUMPING STATIONS
COLLECTORS
Figure 2.
system.
— Lake Annecy: bathymetric map and sewerage
The situation appeared so serious that two technical
solutions were proposed. One was a chain of
purification plants for all wastewater then being
discharged into the lake. The other involved collecting
this water in main collector pipes running all around
the lake and leading to a single large treatment plant.
This plant would in turn discharge into the Fier, a
tributary of the Rhone, below Annecy. The second
alternative was chosen.
Initially, a syndicate of eight communities was
formed in July I957 to cleanse the lake. Today, the
group includes 21. The main difficulty was persuading
these communities, set in their traditional way of life,
to commit themselves to expenditures likely to burden
taxpayers for the foreseeable future. The construction
began in 1962 and was completed in 1976.
A schematic diagram of the sewerage system is
shown in Figure 4. Table 2 gives the repartition of these
280 kilometer pipes collecting wastewater from
103,000 inhabitants (Figure 2). The treatment plant
capacity is today 135,000 gallons per inhabitant; it
purifies 40,000 mVday-2 by a two-step process which
is both mechanical and biological (activated sludges.) In
addition, there is a 120 ton/day-2 composting plant and
a 50 ton/day12 incineration plant for dealing with
domestic and industrial waste. A total investment of
110 million French francs(1975) has been realized but
this represents only half the total amount projected for
satisfying the demand predicted for the year 2000.
SEPARATE SEWERAGE
(LEFT LAKESMORE)
Figure 4. — Lake Annecy: schematic diagram of the sewerage
system.
Table 2. — Lake Annecy. technical characteristics of the sewerage
system.
Principal
collector
Secondary
collector
Figure 3. — Lake Annecy: map of the drainage basin.
Pumping Pumping
Length Diameter Station Length Station
(km) (mm) (Number) (km) (Number)
Separate sewerage 19 800-200 4 64 10
(Right lakeshore)
Separate sewerage 28 700-200 7 132 9
(Left lakeshore)
Combined sewerage 37
(Annecy)
-------
314
RESTORATION OF LAKES AND INLAND WATERS
A few years of recuperation has improved the water
quality, first in the epilimnion of the northern basin, the
first to be protected. Nutrient concentrations are now at
oligotrophic levels, transparency is restored, phyto-
plankton is dominated by diatoms, and the water has
again become drinkable after only a limited filtration
and ozonation process. However, a new problem
recently appeared: The quantitative decrease in the fish
catch may be related to restoring the lake's oligotrophic
state.
LAKE LE BOURGET AND ITS
DIVERSION TUNNEL
Lake Le Bourget (Figures 5,6) is the largest lake in
France. It was described as oligotrophic in 1947. But
around 1952-1955 it became clear that the lake was
deteriorating because profuse algal blooms appeared.
First localized along the most populated shores, they
finally invaded the whole lake. Transparency decreased
simultaneously with hypolimnetic oxygen and sal-
monid populations.
Preliminary investigations showed the nutrient input
must be reduced by 95 percent to restore the lake. The
construction of orthodox treatment plants appeared
insufficient and tertiary treatment by chemical or
biological nutrient elimination seemed to be too
unreliable. Therefore, it was decided first to enlarge the
collector network to a 95 percent capacity; secondly, to
treat all collected effluents (primary and secondary
treatment); and thirdly, to divert all treated sewages
outside the drainage basin.
Two syndicates were created. Between 1960 and
1973, the two first steps were realized, especially in
the districts of Chambery and Aix-les-Bains, whose
effluents represented 85 percent of the point source
input. This program has continued until now; a 400
kilometer sewer network collects wastes from more
than 95 percent of the population and nonagricultural
industries.
The third step began in 1973. Treated sewage from
Chambery, Aix-les-Bains, and Le Bourget du Lac were
collected and drained directly to the Rhone by a tunnel
excavated through a mountain. Figure 6 shows the
location of the different operations, Figure 7, a
schematic diagram, and Table 3, their main technical
characteristics. This sanitation action was reinforced
Cans/ o,9
AIX LES BAINS
0 1 2 Km
CHAMBERY
Figure 5. — Lake Le Bourget: bathymetric map.
Table 3. — Lake Le Bourget: technical characteristics of the
diversion operations.
Chambery — Le Bourget du
Lac discharge pipe
Aix-les-Bains — Le Bourget
du Lac discharge pipe
Mont du Chat diversion
tunnel
Chindrieux — Rhone
diversion pipe
Mains
Diameter Cross Sectional
area
(mm) (m2)
1,200
600
4.35
400
Flows
Length
(km)
8.2
7.6
12.3
5.2
Type
by gravity
by pumpage
by gravity
by gravity
Volume
(1.3-1)
1,630
580
7.500.103
35
-------
URBAN AND POINT SOURCE POLLUTION CONTROL TECHNOLOGY
315
by an incineration plant, the biggest in France, and a
mobile device for collecting floating detritus. A total of
FF 280 million has been spent since 1965, 170 for the
diversion. As this diversion functioning started only at
the beginning of 1980, it is too early to observe any
improvement.
1+144 DIVERSION PIPE
_ DISCHARGE PIPE
• PURIFICATION PLANT
DIVERSION TUNNEL
•
CHI RIEUX ; | > *••.
••••••••*
REFERENCES
Documents concerning the Lake Annecy purification maybe
obtained from; Syndicat Intercommunal des Communes,
Riveraines du lac d' Annecy (SIRCLA), B.P. 739 — CRAN
GEVRIER, 74015 ANNECY CEDEX, France, Tel. 507
57.15.28.
Documents concerning the Lake Le Bourget purification may
be obtained from: Syndicat Intercommunal du lac du
Bourget (SILB), 73000 AIX LES BAINS, France, Tel.
79/35.00.51.
Syndicat Intercommunal d'Assainissement de la Region de
Chambery (SIARC).Rue Aristide Berges, 73000 Chambery,
France, Tel. 79/69.58.69.
8 Km
Figure 6. — Lake Le Bourget: map of the drainage basin and
location of the diversion operations.
Mont du Chal
Figure 7. — Lake Le Bourget: schematic diagram of the
diversion operations.
-------
316
PHOSPHORUS BALANCE AND PREDICTIONS:
LAKE CONSTANCE, OBERSEE
G. WAGNER
Environmental Protection Board
Institute for Lake Research and Fishery
Langenargen/Lake Constance
Baden-Wurttemberg, Germany
ABSTRACT
Results obtained from a dynamic model for the phosphorus balance of Lake Constance are
discussed. Accordingly, the eutrophication process is based upon an increase of the phosphorus
loading from waste waters, especially from detergent phosphorus. When the planned sanitation
measures are finished, the lake will not return to its original state. However, it is able to react
quickly upon a decrease of the loading because of large sedimentation rates. It is pointed out that
the model requires development: Consideration of the river waters and loadings entering deeper
layers of the lake, a further division of the water body during stagnation, and the calculation of
fluctuations of the lake's volume.
INTRODUCTION
During the last 5 to 10 years numerous models to
judge the trophic states of lakes have been proposed
(Schindler, 1978; Vollenweider, 1976; Schroeder and
Schroeder, 1978). Morphological, hydrological, chemi-
cal, and biological parameters were combined, con-
sidering loads and concentrations of important ele-
ments or compounds, residence time of water, density
of biomass, and production. Emphasis has been on
comparing lakes differing in degree of pollution and
describing their conditions. The reaction of lakes after
change in a parameter can be derived, but there are
uncertainties in the quantitative prognosis of the
development of the lakes and of the chances of
practical measures succeeding in a single case. This is
more possible with a dynamic model fitted to the
special lake. With such a model behavior of other lakes
also can be understood.
This paper presents experiences gained with a
simple dynamic model for the phosphorus budget of
Lake Constance (Wagner, 1976a). With its aid the
following questions should be answered:
1. How did the yearly phosphorus loading of the lake
develop and from what sources?
2. What phosphorus concentrations are to be
expected in the lake without sanitation measures?
3. How do these measures affect the phosphorus
balance?
4. What has to be taken into account towards future
development of the model?
5.Which phenomena need more research?
In the mid-1930's about 5 mg P/m3 were measured
and o-phosphate could not be found. Since the 1950's
the phosphorus concentration has increased: slowly at
first, then more rapidly. The consequences were
increasing the density of biomass and the algal blooms,
and decreasing oxygen in the hypolimnion of the
summer stratified lake. O-phosphate always was
present in the hypolimnion. Today, a research program
is investigating loading from the tributaries, the waste-
water treatment plants, the atmospheric precipitation,
and the concentrations in the lake at several stations.
Lake Constance is used as a drinking water reservoir,
for recreation, and by fisheries. In 1960 the Interna-
tional Commission for Water Protection was founded. It
decided to treat plants with phosphorus precipitation
and decided against a ring channel. The main reason
for this decision was that a large catchment area would
have had to be connected to the ring channel. The
measures were started in the early 1960's, with
building since about 1970 at great financial expense.
Most of the treatment plants have begun to work after
1975. The sanitation program will end in the 1980's.
Estimations of phosphorus balance data were
available on the mentioned phosphorus sources, on
water discharge, concentrations in the lake, lake
stratification, and statistics on turnover of poly-
phosphate, fertilizers, and on the development of the
human population.
ROLE OF THE PHOSPHORUS SOURCES
Less phosphorus always leaves the lake than has
been added (even after subtraction of the phosphorus
within the suspended matter of rivers). This means that
both paniculate phosphorus and a large amount of the
originally dissolved compounds remain in the lake. We
can assume that the paniculate fraction enters the
sediment within a few months. There are between
2,000 and 3,000 tons during high water years and less
than 1,000 tons P/yr. during low water. Suspended
matter in the less polluted rivers Rhine and Bregenzer
Ach adsorb o-phosphate from the lake water, while
those of the other more polluted rivers add phosphate
-------
MODELING AND ASSESSMENT OF THE TROPHIC STATE
317
to the water (Wagner, 1976b). But altogether the
effects of these processes seem to be compensated. In
the case of Lake Constance it was more advisable to
hold this paniculate phosphorus not responsible for the
eutrophication process. Therefore, the balance is
restricted to the sum of all phosphorus compounds
with the exception of particulate ones from the rivers.
These data were used to estimate the yearly
phosphorus loadings from polyphosphate, sewage,
precipitation of the catchment area and the surface of
the lake, fertilizers and the geological formations for
the years from 1930 to 1975 (Figure 1).
tP
r2000
poly
•1500
•1000
- 500
L 0
1930 40
50
60
70
80
Figure 1. — Estimation of the yearly phosphorus loading of Lake
Constance (without particulate compounds from rivers) and its
separation into sources: poly = polyphosphate; sew = sewage
without polyphosphate; atm = atmospheric precipitation; fert =
fertilizers; geol = geologic formation.
PHOSPHORUS BALANCE AND
PREDICTIONS
Before modeling, the following assumptions had to
be made (Wagner, 1976a): (a.(The annual phosphorus
loading for the present enters the epilimnion; (b.) the
suspended material from the rivers deposits immedi-
ately; (c.) morphological particulars of the lake are not
considered; (d.) by precipitation, sedimentation, and
circulation phosphorus enters deeper layers of the
water body; (e.) the different phosphorus compounds in
the lake are interchangeable. The model works with
monthly intervals; the waterbody is divided by the
thermocline only. Inverse temperature stratification
does not occur. Only one circulation period exists.
Input data include the monthly phosphorus loading
and water discharge. Output data and fitting para-
meters have been mean phosphorus concentrations in
the epilimnion, the hypolimnion, and at the end of the
circulation period (March/April) of the total lake as well
as the yearly load in the outlet of the lake. Using the
coefficients in the terms for sedimentation rates from
epi- and hypolimnion, the model has been adapted.
These terms take over all downward phosphorus
transports.
After adapting the data, the large amount of the
yearly (calculated!) sedimentation rates of originally
dissolved phosphorus compounds were noticed (Figure
2). Phosphorus may reach the sediment transported by
skeletons of organisms, in the course of calcite
precipitation (Rossknecht, 1980), adsorbed by sinking
particles, or chemically precipitated.
- 500
1930
Figure 2. — Whereabouts of the yearly phosphorus loading of
Lake Constance (without particulate compounds from rivers):
out = outlet of the lake; ace = accumulated in the water body; sed
= deposited in the sediment. In 1944/45 impoverishment in the
water body (blank).
In the period of increasing phosphorus loading the
budget was never in a steady state. During years with
low water (1971-72) a large amount of phosphorus
accumulated in the water body contrasted with a
relative small amount during high water. Also changes
in the yearly water discharge caused a non-steady
state which equalized over a long period. This means
that high water periods equal decreasing phosphorus
concentrations in the lake; low water periods equal
increasing concentrations.
After the rapid increase in loading, the phosphorus
budget now seems to be in a steady state (temporarily?)
and a concentration plateau has been reached. Though
in the last years the densities of phytoplankton biomass
did not develop proportionally to the phosphorus (self-
shading? Buergi and Lehn, 1978), still phosphorus
limitation seems to exist, at least during August/
September. It can be said that a significant diminution
of the phosphorus loading also decreases biomass and
a recovery of the oxygen budget.
The simulation of steady states resulted in the
following relations between yearly loading (without
particulate P from rivers) and concentration at the end
of overturn (total P March/April) in the lake:
2,000 tons P 135 mg P/m3
1,500 tons P 90 mg P/m3
1,000 tons P 45 mg P/m3
500 tons P 17 mg P/m3
The future turnover of phosphorus in the lake will lie
within the observed ranges only as considered in the
adaptation of the model. Therefore, prognosis based on
a time extrapolation may be allowed. About 2,000 tons
P/year were calculated for the mid-1970's. Without
sanitation measures, more than 100 mg P/m3 would
have been expected (Figure 3) in the lake. Lake
Constance has rather a long residence time (4.4 years).
Nevertheless, the lake would quickly react to a
-------
318
RESTORATION OF LAKES AND INLAND WATERS
decreased pollution: Within 10 to 15 years after total
cessation of phosphorus input, phosphorus would
disappear from the lake waters (simulated example)
because of the great phosphorus uptake (Edmondson,
1979; Imboden and Gachter, 1978) and sedimen-
tation. The rest of the loading — after the planned
measures have been effected — amounts to about
1,200 tons P/year (without the suspended matter of
the rivers). The resulting steady-state phosphorus
concentration in the lake will then go down to 60 mg
P/m3. But a large increase in the phosphorus turnover
in the catchment area will affect the lake again.
Today, the observed concentration of 80 to 85 mg
P/m3 (close to steady-state) corresponds to a simulated
loading of less than 1,500 tons P/year. Apparently the
measures have already decreased phosphorus. If these
sanitation measures are not completed and develop-
ment continues with more homes discontinuing septic
disposal and connecting to the sewage plants,
phosphorus will again increase. Information about this
will be available after a current investigation has been
completed. However, the effectiveness of treatment
plants can be improved and the phosphorus content of
detergents can be reduced.
/ugP/l
150-,
2000 t P
100-
50-
1930
40
50
60
70
1200 t p
Figure 3. — Overturn concentrations of total phosphorus in
Lake Constance; calculated steady state concentrations during
a period of permanent loading of 2,000 tons P/Yr (without
paniculate compounds from rivers), decrease of this loading to
1,200 tons P/Yr within 5 years as an effect of measures or to
zero after an Utopian total stop of phosphorus input.
DEVELOPMENT OF THE MODEL
The difficulties of getting data for a balance do not
occur in the investigation of the lake itself but in the
record of the seasonal variation of the loading to the
lake. Experiences show that the load of suspended
matter in the rivers, for instance, can be determined by
special programs only. Also, particulate compounds
should be separated from dissolved ones because of
their different behavior. With respect to judging the
success of sanitation, the known difficulties exist with
determining the origin of the loads. The use of statistics
is risky and the expense of chemical investigations all
over the catchment area is rather high. So an attempt
has been made to calculate and separate point source
and diffuse loading of dissolved compounds with data
only from the mouths of rivers (OECD, 1979). However,
further separation of the suspended matter cor-
responding to its origin, also in the future, might be
uncertain (interim sedimentation!).
The adaptation of the model with the aid of the
chosen terms for sedimentation rates was satisfactory.
However, a number of processes occurring in the lake
had to be temporarily disregarded.
Meanwhile, it has been confirmed that phosphorus
release rates from profundal sediments of Lake
Constance (Obersee) are very small, because more
than 1 mg oxygen always exists (Frevert, 1980). It will
not be relevant to the phosphorus balance.
Also, the influence of the large concentrations of
suspended matter during high water on the behavior of
river water within the lake has been neglected too
much. Investigation of the relationship between
density of river water and seasonal course of
temperature, suspended matter concentrations de-
pending upon rain falls, and concentrations of
dissolved salts have shown (Wagner and Wagner,
1978) that the densities of lake and river water differ
significantly during a year. Large concentrations of
suspended matter during high water (up to about 5 g/l,
grain size median 10 / um! Wagner, 1976b) exceed
the effects of temperature.
The water flow through the lake decisively influences
the phosphorus budget. After summer stratification
begins, the snow in the Alps starts to melt, carrying
waters low in dissolved phosphorus. The melt water is
cold and rich in suspended matter it pushes forward
into deeper layers of the lake. The outflowing waters
from the lake mainly originate from the warmer
epilimnion, where the production of biomass has
already started. Phosphorus uptake, sedimentation,
and displacement of epilimnic waters result in a quick
decrease of phosphorus at the surface of the lake
during May (Figure 4). The reasons for the short-term
maximum of total phosphorus in April/May are still
unsettled, because the differentiation is difficult
between phosphorus in plankton and in fine grained
matter from rivers.
The metalimnion is an efficient barrier for the
incoming river water. From May to October — with the
exception of high waters rich in particles— river water
remains near the thermocline above 50 meters. But
from November to March it mainly flows into a depth of
more than 50 meters. So phosphorus from the rivers is
not available for phytoplankton until it is circulated to
the upper layers. The epilimnic impoverishment in
phosphorus during August/September certainly re-
sults from throttled supplies. Therefore, considering
the fact that the largest portion of the dissolved and the
particulate phosphorus compounds of the river waters
mixes into the hypolimnion, a change for the better in
modeling results can be expected. Instead of dividing
the water body by the thermocline only, an additional
sectioning (at least epilimnion, metalimnion, and
hypolimnion) will be necessary. Besides, the large
-------
MODELING AND ASSESSMENT OF THE TROPHIC STATE
319
fluctuation of the water input also causes fluctuations
of lake volume (and water level) which have to be
considered by an additional hydrological model unit.
Finally, monthly calculations will be made of the
relations between phosphorus and biomass, produc-
tion and oxygen. Some of the numerous empirical
model functions will be used, because modeling real
processes is not yet possible. However, improvements
can be made by using additional diverse connecting
parameters and functions (e-functions, geometric
functions, iterative determination of coefficients). In
any case the model should be as simple as possible.
The model functions mentioned at the beginning
have the incontestable advantage of simple handling.
On the other hand, dynamic models need a long time
for development and adaptation to given facts; this
usually makes long-term operators necessary. Some-
times experiences and ideas disappear if such a
specialist changes his employment. At universities
such models are evolved, too. But there the long-term
data sets and the experiences of the practical men are
not readily available. In contrast on-the-spot-modeling
frequently is not possible for different reasons. Only
large institutions with a sufficient number of col-
laborators are able to connect both. At present in
Germany a discussion of these problems is taking
place.
/ugP/l
lOO-i
50-
IV V VI VII VIII IX X XI XII I II III
Om
% p*,
+2 — 2 — 5—10 — 12— 17—10— 30- •
SO
250
Om
V.R,
lh.-2— 2 — S — 10—12—17—20—30-
part
50
250
Figure. 4 — Seasonal phosphorus concentrations 1977-79 at
the surface of Lake Constance (upper figure) and calculated
phosphorus input from the rivers into different layers of the
lake expressed as percentage (sum of all circle areas = 100
percent yearly loading; Ppart = phosphorus within the suspended
matter; Pdiss = dissolved phosphorus compounds; th = depth of
thermocline).
SUMMARY
A phosphorus balance of Lake Constance is
calculated by a dynamic model. Data since 1935 are
available. Loading after World War II was largely based
on sewage (polyphosphate and feces). The main portion
of the yearly loading enters deeper layers of the lake.
Paniculate phosphorus from the rivers seems not to
influence the balance decisively. Today, the turnover
concentration of total phosphorus is 80 to 85 mg P/m3.
Nevertheless, in August/September o-phosphates still
seems to limit the algal production.
The model simulates steady states for different levels
of loading. The main results are: Without sanitation
measures turnover concentrations of total phosphorus
of more than 100 mg P/m3 can be expected. After the
planned sanitation is finished, the lake will not return
to its original state, for pollution from other sources
remains too high. The concentration then will amount
to about 60 mg P/ m3. The lake is able to respond rather
quickly to a decrease in the loading because the
sedimentation rates are high; they include the
suspended fraction from the rivers plus more than 50
percent of the rest of the phosphorus loading.
To improve the phosphorus balance model, the
supply of river water and phosphorus to the deeper
layers of the lake and a further division of the water
body during stagnation and seasonal volume variations
ought to be considered.
REFERENCES
Buergi, H. R.,and H. Lehn. 1978. Die langjahrige Entwicklung
des Phytoplanktons im Bodensee (1965-1975) Teil 2:
Obersee. Int. Gewasserschutzkomm. Bodensee 22. (In
press.)
Edmondson, W. T. 1979. Lake Washington and the
predictability of limnological events. Arch. Hydrobiol. Beih.
Ergebn. Limnol. 13:234.
Frevert, T. 1980. Dissolved oxygen dependent phosphorus
release from profundal sediments of Lake Constance
(Obersee). Hydrobiologia 70. In press.
Imboden, D. M., and R. Gachter. 1978. A dynamic lake model
for trophic state prediction. Ecol. Model. 4:77.
Organization for Economic Cooperation and Development.
1979. Cooperative programme for monitoring of inland
waters (eutrophication control). Regional Proj. Alpine Lakes.
Draft Rep.
Rossknecht, H. 1980. Phosphatelimination durch autoch-
thons Calcitfallung im Bodensee-Obersee. Arch. Hydrobiol.
88:328.
Schindler, D. W. 1978. Factors regulating phytoplankton
production and standing crop in the world's freshwaters.
Limnol. Oceanogr. 23:478.
Shroeder, R., and H. Schroeder. 1978. Ein Versuch zur
Quantifizierung des Trophiegrades von Seen. Arch. Hy-
drobiol. 82:240.
Vollenweider, R. A. 1976. Advances in defining critical
loading levels for phosphorus in lake eutrophication. Mem.
1st. Ital. Idrobiol. 33:35.
Wagner, G. 1976a. Simulationsmodelle der Seeneutroph-
ierung, dargestellet am Beispiel des Bodensee-Obersees.
Teil II: Simulation des Phosphorhaushalts des Bodensee-
Obersees. Arch. Hydrobiol. 78:1.
_. 1976b. Die Untersuchung von Sinkstoffen aus
Bodenseezuflussen. Schweiz. Z. Hydrol. 38: 191.
Wagner, G., and B. Wagner. 1978. Zur Einschichtung von
Flubwasser in den Bodensee-Obersee. Schweiz. Z. Hydrol.
40:231.
-------
320
PREDICTION OF TOTAL NITROGEN IN LAKES
AND RESERVOIRS
ROGER W. BACHMANN
Department of Animal Biology
Iowa State University
Ames, Iowa
ABSTRACT
The basic Vollenweider input-output model was adapted to predict total nitrogen concentrations in
standing waters. Data from randomly selected group of lakes in the U.S. Environmental Protection
Agency National Eutrophication Survey were used to develop the coefficients for the model, and
data from a different group of lakes from the same survey were used for verification The 95
percent confidence interval for predicting total nitrogen in a lake is from 41 to 255 percent of the
calculated value for the best models. The same equations could be used equally well for natural
lakes or artificial reservoirs.
Phosphorus and nitrogen have long been recognized
as the two elements most likely to limit biological
production in inland waters: thus, their cycles have
been the subject of intensive research. An important
advance was made by recognizing the importance of
continuing nutrient inputs in the determination of
trophic state (Vollenweider, 1968) and the develop-
ment of input-output models for predicting nutrient
concentrations on the basis of nutrient loading, lake
morphometry, and hydraulic flushing rate (Vollen-
weider, 1969). Since that time, a number of empirical
models have been developed to predict total phos-
phorus concentrations (Vollenweider, 1975; Kirchner
and Dillon, 1975; Chapra, 1975; Jonesand Bachmann,
1976; Larsen and Mercier, 1976; Reckhow, 1977,
1979; Canfield, 1979). Yet little effort has been
expended on developing similar models for the other
important element, nitrogen. The purpose of this study
is to develop and test an input-output model for total
nitrogen in natural and artificial lakes.
Unlike phosphorus with only one valence state in
natural waters, nitrogen is found in four different
states of oxidation. One of these, nitrogen gas, is
relatively inert and is not included in the total nitrogen
measurement; however, it can be incorporated into the
cycle through biological fixation by blue-green algae or
can be lost from the biological cycle through the action
of denitrifying microorganisms on nitrates, thus
reducing the total nitrogen concentration. By analogy
with the general development of the phosphorus
models (Vollenweider, 1969), the change in total
nitrogen concentration per unit time equals the rate of
loading of nitrogen from external sources per unit area
divided by the sum of mean depth, plus internal loading
from the sediments plus the rate of nitrogen fixation
minus losses through the outlet minus losses to the
sediments minus denitrification losses. Some of the
parameters in this equation are easily measured or
estimated (total nitrogen concentration, areal nitrogen
loading, lake mean depth, and hydraulic flushing rate),
but the rest are very difficult if not impossible to
measure. These factors (internal loading, sedimenta-
tion losses, nitrogen fixation, and denitrification) are
grouped together as attenuation losses and are
expressed as:
attentuation losses = a TN
were
a = attenuation coefficient, yr~1
TN = the concentration of total nitrogen in the lake,
mg. m23
The differential equation is given as:
dTN/dt L/z TN TN (1)
where
t = time
L = annual nitrogen loading per unit of lake surface
area,
mg-rrr2yr1
z mean depth of lake, m
p = hydraulic flushing rate, yr'1
The steady-state solution is:
TN = L/(Z(a + p))(2)
This is the same as the solution for the total
phosphorus model (Vollenweider, 1975) with the
exception that the sedimentation coefficient has been
replaced with an attenuation coefficient.
DATA BASE
The basic data were obtained from the results of the
U.S. Environmental Protection Agency National Eu-
trophication Survey. Data were tabulated for all lakes
on annual areal total nitrogen loading rates, median
total nitrogen concentrations, lake mean depths,
hydraulic flushing rates, chlorophyll a concentrations,
total phosphorus concentrations, and total phosphorus
areal loading rates. The median total nitrogen
concentration was taken to represent the steady-state
total nitrogen concentration, agreeing with Reckhow
(1977) that the median would be less affected by
-------
MODELING AND ASSESSMENT OF THE TROPHIC STATE
321
extreme measurements. Nitrogen attenuation coeffi-
cients for each lake were estimated from the data by
assuming steady state and rearranging the terms in
Equation 1: ,,-,-K,-^
a = L/(TNZ) p
All the errors in estimating the total nitrogen
concentration, areal loading, lake mean depth, and
hydraulic flushing rate are incorporated into the
attenuation coefficient. Negative values for this
coefficient might indicate a lake that has a net
production of nitrogen through nitrogen fixation, is not
in steady state, or where the errors of estimation may
result in a negative value.
The sample includes all the EPA-surveyed lakes with
a complete set of data. In the first year of that survey,
total nitrogen concentrations were not measured, thus
reducing the size of the sample. The remaining 95
natural and 384 artificial lakes include a wide range of
lake types with mean depths from 0.5 to 307 meters,
total nitrogen concentrations from 125 to 7,185
mg rrf3, nitrogen loading from 1,500 to 14,900,000
mg rrf2yr~1 , and attenuation coefficients from -5 to
392 yr'1 (Table 1).
The lakes were randomly sorted into two data sets.
One data set (model development) with 49 natural and
199 artificial lakes was used to develop the predictive
models, and the other data set (model verification) with
46 natural and 185 artificial lakes was used to test the
predictive abilities of the empirical models and
establish confidence limits. Because the values of most
parameters spanned several orders of magnitude and it
was reasonable to assume that variances were
proportional to means, all data were transformed to
their natural logarithms before statistical analyses
(unless stated otherwise.).
NITROGEN ATTENUATION
COEFFICIENTS
Because the nitrogen attenuation coefficient cannot
be directly measured, I investigated the possibility that
it could be related to some other measurable variable.
Correlations between the coefficient and several
limnological variables are shown in Table 2. In general,
stronger correlations were found for artificial than for
natural lakes. The best correlations were obtained with
various measures of water or nitrogen loading, with
greater rates of input being associated with greater
fractional losses of nitrogen from the lake water (Figure
1).
In addition, a nitrogen retention coefficent was
calculated following the procedures that Dillon and
Rigler (1974) used for phosphorus. The retention
coefficient and its logarithms also were used in the
same correlation matrix, but the resulting correlations
were less strong than those found by using the
attenuation coefficient. I also attempted to fit a nitrogen
settling velocity following Chapra's (1975) work with
phosphorus, but it also was less satisfactory.
The strongest correlations were found with the
volumetric nitrogen loading, the areal nitrogen loading,
and the hydraulic flushing rate: however, this does not
prove a cause-and-effect relationship for any one
variable. Indeed, these three variables are all inter-
correlated (Table 3); any one of them could influence
nitrogen attenuation, or there could be an important
unmeasured variable that also is correlated with either
nitrogen or water inputs.
10,000
100 1000
CALCULATED TOTAL NITROGEN MG/M3
10,000
Figure 1. — Relationship between nitrogen attenuation
coefficients and volumetric nitrogen loading for both natural
and artificial lakes combined.
Table 1. — Mean values and related statistics for annual areal total nitrogen loading rates (mg-m'V'1), total nitrogen
concentrations (mg-rrT3), mean depths (m), hydraulic flushing rates (yr"'), and calculated attenuation coefficients (yr"1), for 479
natural and artificial lakes included in this study.
Variable
Areal nitrogen
loading (L)
Total nitrogen
(TN)
Mean depth (z)
Hydraulic flushing
rate (p)
Attentuation co-
efficient (a)
Lake Type
natural
artificial
natural
artificial
natural
artificial
natural
artificial
natural
artificial
No. in
sample
95
384
95
384
95
384
95
384
95
384
Mean
60894.0
139092.0
1441.0
1027.0
11.0
9.2
4.9
14.4
4.8
8.7
Standard
deviation
172457.0
635435.0
1223.0
974.0
32.4
8.4
8.5
40.4
14.3
31.3
Range
Minimum
1500.0
1700.0
125.0
220.0
0.5
0.6
0.002
0.019
-8.3
-50.0
Maximum
14900000
11155000
6040
7185
307
59
45
365
130
392
-------
322
RESTORATION OF LAKES AND INLAND WATERS
Table 2. — Correlation coefficients (r) between various
limnological parameters and attenuation coefficients.
Logarithmic transformations were used. All coefficients
significant at the 5% level except those marked NS not
significant).
Parameter
Volumetric nitrogen loading
Areal nitrogen loading
Hydraulic flushing rate
Areal water loading
Mean depth
Ratio total nitrogen to
total phosphorus
Chlorophyll a
Total nitrogen concentration
Natural
Lakes
0.60
0.67
0.63
0.61
-0.12
-0.25
0.00 NS
-0.14 NS
Aritifical
Lakes
0.78
0.75
0.74
0.66
-0.30
-0.18
-0.06 NS
-0.04 NS
Both
Combined
0.74
0.74
0.72
0.65
-0.25
-0.20
-0.05 NS
-0.01 NS
Table 3. — Correlations between the logarithms of
chlorophyll a (CHLA), areal phosphorus loading rate (LP),
areal nitrogen loading rate (L), volumetric nitrogen loading
rate (L/Z), total nitrogen (TN), total phosphorus (TP), ratio of
total nitrogen to total phosphorus (TN/TP), hydraulic flush ing
rate (p), and the ratio of the areal nitrogen loading rate to the
areal phosphorus loading rate (L/LP).
CHLA
LP
L
L/Z
TN
TP
TN/TP
P
L/LP
CHLA LP L
1.00 0.07 0.05
1.00 0.63
1.00
L/Z
0.33
0.56
0.86
1.00
TN
0.67
0.22
0.32
0.56
1.00
TP
0.69
0.34
0.22
0.47
0.66
1.00
TN/TP
-0.22
-0.22
0.04
-0.05
0.15
-0.64
1.00
P
0.12
0.53
0.83
0.89
0.27
0.26
-0.08
1.00
L/LP
-0.25
-0.20
0.06
-0.02
-0.08
-0.52
0.61
-0.02
1.00
Regression equations were developed with the
model-development data set for the relationships
between the nitrogen attenuation coefficients and the
volumetric nitrogen loading, areal nitrogen loading,
and hydraulic flushing rates. These were determined
for natural and artificial lakes both separately and
combined (Table 4). The attenuation coefficients were
then substituted back into Equation 2 to yield the
various predictive models for total nitrogen.
MODEL VERIFICATION
I tested the abilities of these models to predict the
measured total nitrogen concentrations of the lakes in
the model-verification data set. Correlation coefficients
were calculated between measured and calculated
total nitrogen concentrations, and empirical 95 percent
confidence limits were determined for the calculated
total nitrogen concentrations of each model by
calculating the standard deviation of the mean
difference between the logarithms of the measured
and calculated total nitrogen concentrations. Average
errors and average percentage errors also were
calculated from the untransformed calculated and
measured total nitrogen values. These four measures
of precision were used to evaluate the respective
models.
For the models based on volumetric loading, areal
loading, and flushing rate, similar results (Table 4)
were obtained whether separate equations were used
Table 4. — Comparison of calculated and measured total
nitrogen concentrations for the model-verification data set
with use of models based on volumetric nitrogen loading
(L/Z), areal nitrogen loading (L), and hydraulic flushing rate
(p). Error estimates include the average error (AE),
percentage error (PE), and 95% confidence limits as
percentages of the calculated total nitrogen value (CL).
Model
Correlation
coefficient
r
Error estimates
AE PE CL
Based on L/z
natural lakes with
In a = -0.345 + 0.505ln (L/z)
and artificial lakes with
In a = -0.434 + 0.618ln (L/z) 0.80 410 38 41-253
both with
In a = -4.144 + 0.594ln (L/z) 0.80 419 46 47-286
Based on L
natural lakes with
In a = -6.506 + 0.724ln L and
artificial lakes with
In a = -6.430 + 0.709ln L 0.82 382 37 41-255
both with
In cr = -6.426 + 0.710ln L 0.82 3823741-255
Based on p
natural lakes with
In a = -0.485 + 0.5861ln p and
In or = -0.291 + 0.5821 In p 0.76 5135936-325
both with
In o = -0.367 + 0.5541 In p 0.77 4985636-315
for natural and artificial lakes or a single equation was
used for both. This indicates that the same coefficients
can be used in nitrogen models for both natural and
artificial lakes. This contrasts with phosphorus models
in which different coefficients for the respective lake
types lead to a greater degree of precision (Canfield,
1979).
The best results were obtained for the models based
on volumetric nitrogen loading (Figure 2) or areal
nitrogen loading although the model based on flushing
rate also gave acceptable results. It was originally
thought that a simple nitrogen model would give poorer
results than a phosphorus model because of the
greater complexity of the nitrogen cycle. This was not
found, for by comparison, the best available model for
predicting total phosphorus (Canfield, 1979) had an
average percentage error of 44 percent and 95 percent
confidence limits of 31 to 288 percent when applied to
a similar set of lakes. My best nitrogen models have a
37 percent percentage error and 95 percent confidence
limits of 41 to 255 percent.
NITROGEN, PHOSPHORUS, AND
CHLOROPHYLL a
Stronger correlations (Table 5) were found between
total nitrogen and chlorophyll a in natural lakes than in
artificial lakes. This agrees with similar findings by
Canfield (1979) for the total phosphorus-chlorophylls
relationship and presumably results from a greater
chance of light limitation in artificial lakes because of
greater concentrations of inorganic paniculate mate-
rials.
-------
MODELING AND ASSESSMENT OF THE TROPHIC STATE
323
10° I01
10* I05 I06 I07
VOLUMETRIC NITROGEN LOADING MG/M
Figure 2. — Relationship between measured total nitrogen and
total nitrogen calculated with separate regressions for natural
and artificial lakes on the basis of volumetric nitrogen loading
(Table 4). The best-fit linear regression line is shown.
Tables. —Correlations (r) between logarithms of chlorophyll
a and total nitrogen and total phosphorus for natural and
artificial lakes.
Total phosphorus
Total nitrogen
Natural lakes
0.84
0.81
Artificial lakes
0.59
0.59
The relatively high correlation (r = 0.81) between
total nitrogen and chlorophyll a was unexpected,
because most of the lakes were thought to be
phosphorus-limited on the basis of the ratios of total
nitrogen to total phosphorus (Table 6). Vallentyne
(1974) reported that aquatic plants characteristically
have ratios of nitrogen to phosphorus of about 7,
considerably smaller than the average ratio of 23.7 in
the sample lakes. Most likely, the high correlation is
because of the fact that total nitrogen and total
phosphorus concentrations in lakes are highly corre-
lated with each other (Table 3), so that they both would
be correlated with chlorophyll even though phosphorus
may have been the limiting nutrient in most instances.
Other similarities were noted between the behavior
of nitrogen and phosphorus in the lakes in this sample.
In general, the lakes were sinks for both elements with
similar loss rates for both as indicated by the finding
that the average ratios of nitrogen to phosphorus
within the lakes were not significantly different from
the ratios in the inputs (Table 6). There were no large
shifts in the ratio indicating differential losses or
substantial effects of nitrogen fixation by blue-green
algae.
The only major difference was found in those lakes
(27 of 479) where negative nitrogen attenuation
coefficients indicated that more nitrogen was being
produced in the lake than was being lost. It may be
significant that the average N:P ratio in the inputs to
those 27 lakes (17.4) is significantly different from the
ratio (24.5) in the other 452 lakes with positive
Table 6. — Frequency distributions of the ratios of total
nitrogen (TN) to total phosphorus (TP) within the lakes in the
sample and the ratios of the annual surface loading of total
nitrogen (L) to the annual surface loading of total phosphorus
(LP). The differences between the averages of the two ratios
are not statistically significant.
Ratio TN:
2
4
6
8
10
12
14
16
18
20
30
40
60
80
Average
Std. Dev
TP % of lakes with
a smaller ratio
0.6
1.9
7.0
12.1
19.8
27.4
34.9
42.3
49.6
56.8
79.1
89.1
96.6
99.4
= 23.7
. = 20.9
Ratio L:LP
2
4
6
8
10
12
14
16
18
20
30
40
60
80
% of lakes
a smaller
0.2
2.9
9.2
15.8
23.2
30.5
39.4
47.3
54.3
58.2
77.5
86.2
95.2
98.1
with
ratio
Average = 24.1
Std. Dev. = 27.5
attenuation coefficients, but the N:P ratios within the
two groups (23.9 and 23.7, respectively) are not
different. This could illustrate the proposal by Schindler
(1977) that lakes with small ratios of N:P in their inputs
will have enhanced rates of nitrogen fixation, with a
subsequent elevation of the N:P ratio within the lakes
themselves.
Lastly, the nitrogen attenuation coefficient and the
analogous phosphorus sedimentation coefficient are
both strongly correlated with the loading rates of the
respective elements as well as with the water loading
rates (Canfield, 1979), leading to similar forms for their
respective prediction equations. The reasons for this
are poorly understood. The strong affinity of phos-
phorus for particulate materials has been used as an
explanation for its behavior (Canfield, 1979), but this
does not seem likely for nitrogen. Clearly more work is
needed to understand the factors controlling nitrogen
concentrations in lakes.
REFERENCES
Canfield, D. F. 1979. Prediction of total phosphorus
concentrations and trophic states in natural and artificial
lakes: The importance of phosphorus sedimentation. Ph.D
dissertation. Iowa State University, Ames.
Chapra, S. C. 1975. Comment on "An empirical method of
estimating the retention of phosphorus in lakes" by W. B.
Kirchner and P. J. Dillon. Water Resour. Res. 11:1033.
Dillon, P. J., and F. H. Rigler. 1974. The phosphorus-
chlorophyll relationship in lakes. Limnol. Oceanogr. 19:767.
Jones, J. R., and R. W. Bachmann. 1976. Prediction of
phosphorus and chlorophyll levels in lakes. Jour. Water
Pollut. Control Fed. 48:2176.
Kirchner, W. B., and P. J. Dillon. 1975. An empirical method
of estimating the retention of phosphorus in lakes. Water
Resour. Res. 11:182.
Larsen, D. P., and H. T. Mercier. 1976. Phosphorus retention
capacity of lakes. Jour. Fish. Res. Board Can. 33:1742.
Reckhow, K. H. 1977. Phosphorus models for lake
management. Ph.D. dissertation, Harvard University.
-------
324 RESTORATION OF LAKES AND INLAND WATERS
1979. Uncertainty applied to Vollenweider's
phosphorus criterion. Jour. Water Pollut. Control Fed.
51:2123.
Schindler, D. W. 1977. The evolution of phosphorus
limitation in lakes. Science 195:260.
Vallentyne, J. R. 1974. The algal bowl: lakes and man. Fish.
Res. Board Can. Misc. Publ. 22.
Vollenweider, R. A. 1968. Scientific fundamentals of the
eutrophication of lakes and flowing waters, with particular
reference to nitrogen and phosphorus as factors in
eutrophication. Organ. Econ. Coop. Dev. Tech. Rep.
DAS/CS1768.27.
1969. Possibilities and limits of elementary
models concerning the budget of substances in lakes (in
German). Arch. Hydrobiol. 66:1.
1 975. Input-output models with special reference
to the phosphorus loading concept in limnology. Schweizer-
ische Zeitschrift fur Hydrologie. 37:53.
ACKNOWLEDGMENTS
I would like to thank Dr. Jack Gakstatter of the U.S. EPA
Corvallis, Oregon, Laboratory for providing data from the
National Eutrophication Survey and Ms. Debra Hoffmaster
who assisted with data reduction, computer programing, and
statistical analyses.
-------
325
AN INCREMENTAL PHOSPHORUS LOADING CHANGE
APPROACH FOR PREDICTION ERROR REDUCTION
KENNETH H. RECKHOW
School of Forestry and Environmental Studies
Duke University
Durham, North Carolina
ABSTRACT
Lake quality management planning necessitates projecting the impact of proposed watershed
activity and land use changes on lake quality. In most cases, the change is relatively small in
comparison to the watershed characteristics that are expected to remain constant over the
planning period. Prediction using a lake loading model is probably unnecessary for these
unchanging land uses, since existing lake data represent the resultant water quality impact. In
those situations, lake loading model prediction may be required only for the proposed watershed
land use changers). With sufficient representative lake quality data, the future projection reliability
is improved when the model prediction is calculated for the change only. This is manifested in a
reduction in the total projection error, which is a function of lake data variability (for unchanging
land use), and model and loading error (for changing land use).
INTRODUCTION
Effective lake quality management planning neces-
sitates the use of quantitative methods or models to
relate relevant human activities and natural character-
istics to lake water quality. Models, in turn, are of
particular value to the planning process when
reliability can be directly assessed. Reliability, or its
converse, uncertainty, serves three vital functions in
planning studies:
1. Reliability represents an estimate of the value of
information. If the reliability associated with a
prediction is low, the prediction is uncertain and
imprecise, and the predictive information is not
particularly valuable. Alternatively, if the reliability of a
prediction is high, the prediction is precise, and the
predictive information can be valuable.
2. Important factors that are poorly characterized
(i.e., have high uncertainty) may be identified when
reliability is assessed. The analysis of uncertainty, or
error, helps model developers and model users in a
sensitivity analysis exercise. Specifically, estimation of
errors allows the analyst to identify those character-
istics that have significant error and have a significant
effect on the prediction. The analyst then realizes that
in order to obtain a precise prediction, these uncertain,
prediction-sensitive terms must be better defined.
3. Discrimination among control strategies may be
explicitly evaluated with an assessment of reliability.
Without uncertainty analysis, one is given the
impression that prediction differences of one micro-
gram per liter or less are significant and indicate a well-
defined ordering of quality states. With uncertainty
analysis, the prediction interval, or confidence interval
defined by the prediction error, identifies a region in
which land use strategies may be predictively
indistinguishable. The error analysis allows the
planner to determine when land use strategy impacts
can be predictively distinguished, given the error
associated with model applications.
Several phosphorus lake models have been proposed
recently that incorporate a procedure for estimating
prediction uncertainty (Chapra and Reckhow, 1979;
Reckhow, 1979a, b; Reckhow and Simpson, 1980;
Reckhow and Chapra, 1980; Reckhow, et al. 1980).
These approaches represent an improvement over the
purely deterministic analyses presented in older
literature, since the error estimate is a measure of
prediction information value.
However, there are two problems or shortcomings
with the existing error analysis methods. Errors arise in
model applications because of error in the model, the
model parameters, and the model variables. In a more
fundamental sense, one may also say that the errors
are caused by natural variability, inadequate sampling
design, measurement error and bias, and model
specification error. When the data set variables used to
construct a model contain error, then this error is
transmitted to the model error term for the fitted model.
Since virtually all limnological statistics contain error
as a result of the aforementioned causes, then lake
models developed from these data contain error
associated with the data error. This means that the
model standard error term includes an error compo-
nent associated with errors in the model variables.
This is an unwanted component, yet it is unavoidably
there given present knowledge and data.
Models can be employed in a descriptive or a
predictive mode. When used descriptively, models may
be used to relate observed inputs (the independent
variables) to observed outputs (the dependent variable).
In a descriptive application, when all variables are
directly measured in the same manner as the variables
in the model development data set were measured,
-------
326
RESTORATION OF L^KES AND INLAND WATERS
then there is no need to add additional application lake
variable error. This is because the appropriate variable
error is already contained in the model error term.
However, when the model is used in a predictive mode,
the dependent variables generally cannot be measured
(because the predictive nature implies conditions not
yet physically realized). Predictive applications of a
model require that the analyst extrapolate variable
values from other points in time and/or space. This
extrapolation process introduces error beyond that
already contained in the model error term. Thus,
predictive use of a model should be accompanied by an
error analysis that includes variable error. This errors-
in-variables analysis must be undertaken thoughtfully,
however, to avoid "double counting" errors (due to the
errors-in-variables term already contained in the model
standard error term). The first problem of existing error
analysis methods, therefore, is that their application
may lead to error double counting.
The second shortcoming associated with existing
methods is of greater importance, given the fact that
with care, double counting can probably be kept at an
acceptably low level. The second problem relates to the
magnitude of the error term. The input-output
empirical phosphorus lake models of concern here are
developed from cross-sectional analyses. The models
are simple, our knowledge of limnology is limited, and
all ptiosphorus-settling processes are aggregated into
one empirically-determined model parameter. As a
result, the model error term is large, since it represents
cross-sectional variability, measurement and sampling
error, and model specification error. In addition, errors
in the model variables, particularly in phosphorus
loading, can be substantial for certain applications. The
combined effect of these error terms is a large
prediction error using existing error analysis methods.
The magnitude of ths error term, and the associated
prediction intervals, is such that the analyst is often
unable to find "statistically significant differences"
among competing lake management options.
A PROPOSED ERROR ANALYSIS
METHODOLOGY
An alternative error analysis methodology will
substantially reduce prediction error over existing error
analysis techniques for most applications. This pro-
cedure exploits two features that are common to many
lake quality management planning situations:
1. For most projected planning scenarios, the land
use area expected to change is small in comparison to
the land use area that is expected to remain constant
during the planning period. Stated another way, the
impact of the change is generally small in comparison
to the impact of the existing land uses.
2. Existing phosphorus lake data reflect the impact of
present land use conditions. Furthermore, the variabil-
ity in these data represent the variability in impact
response. These data could already be in existence or
they could be acquired upon initiation of this modeling
program.
To see how these features can lead to prediction
error reduction, consider a planning scenario in which
no change is projected so that existing land use and
future land use are equivalent. In that case, two
methods may be used to predict future lake phosphorus
concentration (ignoring temporal variability for the
moment):
1. A phosphorus lake model may be applied to relate
land use to phosphorus concentration through litera-
ture export coefficients. This standard procedure is
accompanied by a high prediction error.
2. Existing phosphorus lake data may be used to
describe future lake quality under unchanging water-
shed conditions. Here the error term is a function of the
standard error of the estimate for the data and of the
representativeness of the data.
In virtually all cases, with even a modest amount of
phosphorus lake data, the error for the second method
will be considerably smaller than the error for the first
method, given the size of the phosphorus loading and
model error terms.
If this scenario is modified slightly to a situation
common in lake quality management planning, the
new error analysis methodology may be outlined.
Consider a planning scenario in which a relatively
small land use change is projected. The new modeling
and error analysis methodology stipulates that the
analyst use:
1. Existing lake data to evaluate the impact of
unchanging land uses, and
2. The model to evaluate the impact of the land use
change and the impact of hydrologic variability.
Existing error analysis methods do not permit the
analyst to distinguish between land uses that are
projected to change and land uses that are to remain
constant. This means that the impacts of all watershed
land uses on lake phosphorus concentration are
evaluated through the model. Since the impact of
unchanging land uses is manifested in recent lake
phosphorus concentration data, information (the lake
phosphorus concentration data) is wasted and high
prediction errors result.
To indicate the magnitude of the error reduction
associated with the procedure outlined herein, con-
sider the following model (Reckhow, 1979b):
P=-
11.6 + 1.2qs
Eq. (1)
where:
P =lake phosphorus concentration (mg/l)
L = annual areal phosphorus loading (g/m2-yr)
qs = annual areal water loading (m/yr)
The model standard error is .128 in logarithmically-
transformed concentration units. This translates to
about a ± 30 percent prediction error when the antilog
is determined for a particular concentration. The
difference between the existing and proposed error
analysis methodologies may best be stated through
hypothetical comparisons.
1. With the existing error analysis procedures, the
model is used to predict the impact from all land uses.
The model error alone (to which errors in variables
must eventually be added) is approximately ±30
percent. For oligotrophic lakes, this model error term is
relatively small. However, planning frequently occurs
on lakes with phosphorus concentrations ranging from
-------
MODELING AND ASSESSMENT OF THE TROPHIC STATE
327
.020 mg/l to .060 mg/l. Plus or minus 30 percent error
would amount to ± .006 mg/l to ± .018 mg/l I for these
concentrations. This is a substantial error term, and it
may both discourage the planner from using error
analysis and obscure the differences among manage-
ment strategy impacts.
2. With the error analysis procedure proposed
herein, the model is used to predict the impact for the
changing land uses only. The analyst must use the
model to evaluate the impact for both the old and the
new land used. Most projected changes in land use
have a relatively minor impact on lake phosphorus
concentration in comparison to the impact from all
watershed land uses. For comparison purposes
assume that a land use change from forest to
agriculture is to occur in a watershed. Assume that
model predictions indicate that this forested land
contributed 2 percent of the total phosphorus loading to
the lake and that the new agricultural use is expected
to contribute about 10 percent. Since the model must
be used to evaluate the impact of both old and new
changing land uses, the result is a 12 percent (10 + 2)
loading change to be evaluated using the model. For
the range of lake phosphorus concentrations of .020
mg/l to .060 mg/l, and a model error of ± 30 percent
the model prediction error term is .00072 mg/l to
.00216 mg/l. If a reasonable amount of lake sampling
for phosphorus concentration has occurred (under a
good sampling design), then the impact of unchanging
land uses may be evaluated objectively. Even for
modest amounts of data, the standard error will usually
be small.
For example, Reckhow (1979c) evaluated phos-
phorus data variability in a cross-sectional study and
found that the interquartile range is equivalent to about
half the median phosphorus concentration. If it is
assumed that the interquartile range is approximately
twice the standard deviation,and the mean and median
are equivalent, then the standard deviation is about
one-fourth of the mean. For the phosphorus con-
centration range of .020 mg/l to .060 mg/l, the
estimated standard deviation is about .005 mg/l to
.015 mg/l. With a relatively small data set of perhaps
20 to 30 phosphorus concentration measurements, the
standard error of the estimate is (1A/n times the
standard deviation) .00091 mg/l to .0034 mg/l.
Combining this error term with the model prediction
error term for changing land uses (square the error
terms, add, and calculate the square root), the error
ranges from .0012 mg/l to .0040 mg/l.
This comparison does not include all error terms for
either methodology (see Reckhow, 1980, for an
example with more detail), but the major error terms
are calculated. Note that the proposed methodology
reduces prediction error by 75 to 80 percent in the
hypothetical example. The analyst must realize,
however, that this error reduction associated with the
new methodology is contingent on the magnitude of
the land use change that must be evaluated using the
model. Obviously, as the magnitude of the projected
impact increases, the advantage of the proposed
procedure diminishes.
The hypothetical example comparing error analysis
procedures includes three of the four basic terms for
the proposed methodology. The four error terms, and
their interpretations in modeling applications, are:
1. Uncertainty in the assessment of current lake
phosphorus concentration. If adequate data exist, this
error term may be represented by the mean square
error of the data. Data "adequacy'' should be
determined by whether existing data are representative
on a spatial and temporal (within and across years)
basis. In situations with inadequate data, this error
term may be estimated through regression analysis
with more comprehensive data sets on correlated
variables (e.g., Secchi disk transparency) or through
subjective determination.
2. Uncertainty in the hydrology variable,qs. Cross-
sectional error in qa already exists in the model
standard error for the reason identified earlier. This qs-
component of model error has unknown magnitude
and may be sufficient for lakes with low inflow-outflow
variability. Further, when several years of in-lake
phosphorus concentration data exist (described in error
component number 1), these data already exhibit the
effect of qs-variability, making qs-error analysis
unnecessary. Therefore, this additional error term may
be considered optional. In cases with substantial
variability in year-to-year values of the hydrology
variable (qs>, and limited in-lake phosphorus data, a q
error term should be included, propagated through the
model (using first order analysis: see Benjamin and
Cornell, 1970). This error term should represent the
year-to-year variability and the estimation or mea-
surement error associated with the determination of qs.
3. Uncertainty in the prediction of the impact of
the projected new land use on lake water quality.
This term is estimated using the phosphorus lake
model and a procedure like that presented in Reckhow,
et al. (1980). This error component includes model
error and error in the estimate of phosphorus loading
for the projected land use change. Note that for minor
land use changes (relative to the entire watershed) the
impact of this error term is small (despite the inclusion
of model error) because the fractional phosphorus
loading addition is small.
4. Uncertainty in the prediction of the impact of
the existing land use in the area to undergo change.
To properly assess the anticipated change, the analyst
must determine the impact of both the old and the new
land uses using the modeling/error analysis pro-
cedure. These calculations are undertaken in the same
manner as are the calculations for error term described
in Number 3. Note that here, too, the error is small
when the fractional phosphorus loading subtraction is
small.
In summary, the two error analysis methodologies
may be compared with the aid of Figure 1. At the top of
the figure, the traditional method is undertaken by
estimating the phosphorus loading, and the loading
estimation error, for all land uses in the lake
watershed. The phosphorus loading and the loading
error are propagated through the model for the
calculation of the predicted lake phosphorus concen-
tration. Prediction error for this procedure is deter-
mined by the loading error and the model error. Since
the model error term is proportional to the phosphorus
loading magnitude propagated through the model, and
-------
328
RESTORATION OF LAKES AND INLAND WATERS
since all phosphorus loading is propagated through the
model under the traditional procedure, the model error
term is large. As a result, the total prediction error for
the traditional procedure is often ± 30 to ± 40 percent
The new procedure often leads to a prediction error
reduction because it is not required that the model (and
model error) be used to predict all land use impacts.
The watershed may be divided into land uses that are
expected to remain constant over a planning period and
land uses that are expected to change. Similarly the
average phosphorus concentration in a lake may be
divided into a fraction contributed by unchanging land
use and a fraction contributed by land use that is
expected to undergo change. For the new error analysis
procedure, existing phosphorus lake data represent the
impact of all existing land uses. The variability in these
data reflect estimation uncertainty. Since no predictive
model was required to assess this impact, the
uncertainty term is often small. Added to this
uncertainty is the prediction error associated with the
determination of the impact of all changing land uses
calculated using the model. However, since a fraction
(often a sizable fraction) of the land use impacts is
assessed without the model, total prediction error for
the new procedure is generally much lower than it is
for the old procedure.
ISSUES FOR CONSIDERATION
An effort has been made in this brief paper to stress a
conceptual discussion of error analysis, forsaking at
present applications and the mechanics of calculations.
Continuing along this line of approach, some issues
that were alluded to warrant yet further consideration.
These issues are identified here in the hope they will
stimulate additional analysis of this topic.
1. What is, or should be, the meaning of the lake
phosphorus measurements error term? It is intended to
represent the impact of unchanging land use on lake
water quality.
a. Can we determine whether the lake is in
steady state relative to watershed land uses?
b. It has been indicated that the lake phosphorus
measurements should represent spatial and temporal
variability. How does the need for temporal variability
representation conflict with the need for "steady state"
and the likelihood that most lakes undergo continuous
small land changes?
c. Can "adequate" sampling design be defined
objectively?
2. To what extent is time series variability
represented in the lake phosphorus measurements and
to what extent must it be included in the q -error term?
3. Time series data for qs-variability could be
extrapolated from other similar watersheds or perhaps
from precipitation data. In those situations, an
additional error term should be included, representing
possible bias associated with the use of extrapolated
data.
4. The analyst should be aware of the difference
between the standard deviation and the standard error
of the estimate. The standard deviation is a measure of
the variability in a set of data. The standard error of the
estimate, which may often be calculated from the
standard deviation by dividing by vn, reflects the
error in a statistic. The error analysis yields a standard
error of the estimate that represents the error in the
prediction; it does not directly reflect the variability to
be expected for (in this case) lake phosphorus
concentration.
5. The error propagation equation (Benjamin and
Cornell, 1970) is to be used to calculate the impact of
errors in the variables and errors in the parameters on
the total prediction uncertainty. One term in the error
propagation equation represents the error contribution
associated with variable (or parameter) correlation.
Should this term be computed for the correlation
between the old, changing land use phosphorus
loading and: (a) the new land use phosphorus loading,
and/or (b)qs? This question is largely of a conceptual
nature, since the impact on total prediction uncertainty
in either case is undoubtedly quite small. Nevertheless,
it illustrates the type of conceptual problem that must
be considered as error analysis methodologies are
proposed and refined.
REFERENCES
Benjamin, J. R., and C. A. Cornell. 1970. Probability,
statistics, and decision for civil engineers. McGraw-Hill,
New York.
Chapra, S. C., and K. H. Reckhow. 1979. Expressing the
phosphorus loading concept in probabilistic terms. Jour.
Fish. Res. Board Can. 36:225
Reckhow, K. H. 1979a. Empirical lake models for phosphorus:
Development, applications, limitations, and uncertainty. In
Pages 193-221. D. Scavia and A. Robertson, eds.
Perspectives on lake ecosystem modeling. Ann Arbor
Science Publishers, Ann Arbor, Mich.
1979b. Quantitative techniques for the assess-
ment of lake quality. EPA-440/5-79-015. U.S. Environ.
Prot. Agency, Washington, D.C.
1979c. Lake data analysis and phosphorus
variability. Paper presented at the North Am. Lake Manage.
Conf., Michigan State University, East Lansing.
1980. Lake modeling error analysis for in-
cremental land use changes. Unpubl. mss.
Reckhow, K. H., and S. C. Chapra. 1980. Engineering
approaches for lake management: Data analysis and
modeling. Ann Arbor Science, Ann Arbor, Mich. (In Press.)
Reckhow, K. H., and J. T. Simpson. 1980. A procedure using
modeling and error analysis for the prediction of lake
phosphorus concentration from land use information. Can.
Jour. Fish. Aquat. Sci. (In Press.)
Reckhow, K. H., M. N. Beaulac, and J. T. Simpson. 1980,
Modeling phosphorus loading and lake response under
uncertainty: A manual and compilation of export coef-
ficients. U.S. Environ. Prot. Agency, Washington, D.C. (In
Press.)
-------
329
APPLICATION OF PHOSPHORUS LOADING MODELS TO
RIVER-RUN LAKES AND OTHER INCOMPLETELY
MIXED SYSTEMS
STEVEN C. CHAPRA
Great Lakes Environmental Research Laboratory
National Oceanic and Atmospheric Administration
Ann Arbor, Michigan
ABSTRACT
Theoretical calculations are used to illustrate how river-run reservoirs tend to retain a larger
fraction of their phosphorus loading than completely mixed lakes because of the effect of
incomplete mixing or the sedimentation process. Empirical models are used to demonstrate the
correlation between flushing characteristics and sedimentation. Enhanced settling is also ascribed
to the higher proportion of solid-associated phosphorus in the loadings of incompletely mixed
systems. The importance of solids to lake phosphorus budgets is demonstrated with a
nutrient/phytoplankton model for a river-run lake.
INTRODUCTION
The phosphorus loading concept provides a variety of
mathematical and graphical models to predict trophic
state as a function of simple expressions of a lake's
morphometry, hydrology, and loading. Because they
can be used to make inexpensive, order-of-magnitude
estimates of water quality, these models have been
widely applied for lake management. However, as with
any mathematical idealization, there is a residual
variability which these models do not explain.
While the variability of phosphorus loading models
results from a variety of factors, it can be divided
generally into two components (Chapra, 1980). The
first, called "perceptual error," relates to our ability to
perceive the actual state of an individual lake. Thus,
perceptual error is caused by factors such as
measurement errors and year-to-year meteorological
variations that cause a lake to vary from its most likely
condition. Hypothetically, if enough of the proper
measurements are taken, the perceptual error (or the
mean) would approach zero and we would obtain an
accurate estimate of the "true" state of the lake.
The second component of the variability, called "lake
uniqueness," relates to the fact that, even if the
perceptual error is reduced to zero, an individual lake
will still differ from model predictions because of
biological, chemical, and physical factors not ac-
counted for by simple phosphorus loading relation-
ships. A case in point is Lake Washington where, even
though phosphorus levels remained constant, its water
clarity has recently increased because of changes in its
zooplankton assemblage (Edmondson, 1978). In fact,
Shapiro (1979) has suggested that biological factors
not accounted for by simple models could represent
viable control options for lake rehabilitation.
Whereas a strong case has been made for biological
factors, less has been done to elucidate physical
mechanisms that bear on phosphorus loading predic-
tions (Chapra, 1979). The present paper is devoted to
one of the more important physical aspects of lake
uniqueness — incomplete horizontal mixing.
The theoretical basis of most phosphorus loading
relationships developed to date is the completely mixed
model (Figure 1). In this idealization, it is assumed that
phosphorus inputs are instantaneously dispersed
throughout the lake's volume so that the concentration.
in the water is homogeneous. While this is an excellent
model for many lakes, there are a variety of systems
where it does not apply (Figure 2). For example, river-
run lakes and reservoirs typically exhibit strong
horizontal gradients near river mouths and sewage
outfalls that would not be accounted for by a well-
mixed approach. Additionally, incomplete mixing is
Loading
Flushing
Sedimentation
Figure 1. — A completely mixed model showing the majoi
mechanisms governing the level of total phosphorus in a
lake.
-------
330
RESTORATION OF LAKES AND INLAND WATERS
(a) River-run lake
Figure 2. — Overhead views of some incompletely mixed
systems. The darkshaded areas represent heightened
phosphorus levels near river mouths and the arrows
designate the direction of flow.
relevant to the modeling of lake sub-areas such as
embayments and the littoral zone where human
influence, use, and perception of the water body are
intense. Thus, the application of completely mixed
models to such systems, neglects, or averages out, the
inhomogeneities that represent critical aspects of their
water quality
Of equal importance is the fact that neglecting these
gradients and their underlying mechanisms may lead
to faulty predictions when applied to incompletely
mixed systems. In the present paper, this is illustrated
by comparing the prediction of a well-mixed model with
the prototype incompletely mixed system — the river-
run lake or reservoir. These elongated lakes are ideal
for such a contrast since they exhibit most of the
characteristics of other incompletely mixed systems yet
have a simple one-dimensional transport regime that
allows a clear perception of the processes underlying
their dynamics. The basic conclusion of the comparison
is that incompletely mixed systems are more efficient
sedimentation basins than well-mixed lakes. The
interrelationship of settling and lake hydraulics is also
demonstrated by a theoretical analysis of some
empirical phosphorus loading models. Finally, the
importance of solids to the dynamics of incompletely
mixed systems is demonstrated by a nutrient-food
chain-suspended solids model for a river-run lake.
THEORETICAL COMPARISON OF
COMPLETELY MIXED AND RIVER-RUN
BUDGET MODELS
Input-output or budget models predict a lake's
contaminant level by determining fluxes of the
substance across the system's boundaries. The input of
phosphorus consists of loadings such as sewage
effluents and tributary discharges that enter the lake at
its periphery or atmospheric loadings that enter
through its surface. In general, two major processes
characterize phosphorus losses. The first, sedimenta-
tion,represents the net amount of phosphorus in-
corporated into the lake's bottom along with settling
particulate matter. The second, flushing, represents
the loss of phosphorus carried by water flowing
through the lake's outlet.
As depicted in Figure 1, the assumption of complete
mixing allows these processes to be modeled in a very
simple fashion. For example, since the point of entry of
the inputs is irrelevant, a single term can be used to
represent the total loading. In a similar fashion,
sedimentation and flushing can be represented by
simple formulations. For systems where mixing is not
complete, however, adequate characterization of in-
lake water motion or transport is required.
As will be shown, the more complex transport
regime in turn has an impact on the magnitude and
structure of the flushing and sedimentation processes
and requires that the location of inputs be specified.
Before demonstrating the importance of these factors
for a river-run lake, however, the completely mixed
model will be reviewed briefly as a point of reference
for the subsequent discussion.
The Completely Mixed Model
A phosphorus budget model for a well-mixed lake
can be expressed mathematically as (Vollenweider,
1969; Chapra, 1975)
dp
V—= W —Qp — vAsp
dt
(D
(accumulation) = (inputs) - (flushing) - (sedimentation)
where V is lake volume (106m3), p is its phosphorus
concentration (mg m~3), t is time (yr), W is the rate of
mass input of phosphorus (kg yr"1), Q is the rate of
water flow through the lake's outlet (106 m3 yr~1),isthe
apparent settling velocity of total phosphorus (m yr"1)
and As is the lake's surface area (106 m2).
At steady state (i.e., dp/dt = 0), Eq. 1 can be solved for
P — Pout =Pln(-
v/qs
(2)
where Pout is the total phosphorus concentration of the
outlet (ring m"3), Pm is the concentration of the inputs
(mg m~3)= W/Q, and qs is the areal water loading (m
yr"1) = Q/A9. Thus, Eq. 2 is a relationship that can be
used to calculate in-lake phosphorus concentration as
a function of loading and parameters related to the
lake's flushing and sedimentation characteristics. Note
that since the lake is well-mixed, the concentration of
outflowing water is equivalent to that at mid-lake. In
the following section, this equivalence is contrasted
with systems when in-lake concentrations are hetero-
geneous.
-------
MODELING AND ASSESSMENT OF THE TROPHIC STATE
331
The River-Run Model
In contrast to the disorganized or turbulent flow
regime of the well-mixed lake, a river has a well-
organized, unidirectional flow as depicted in Figure 3a.
For the ideal case where no longitudinal mixing is
present, the river flow would not change the identity of
the substance being transported. Thus, as in Figure 3b,
a conservative dye (i.e., one which does not react or
settle), would merely move downstream along with the
water flow. In the engineering lexicon, such systems
are called plug-flow reactors. For the case where a
substance settles at a first order rate as it flows, it can
b'e shown (Reckhow and Chapra, in press) that the'
steady state concentration downstream from a con-
stant point source can be calculated as
p = pm exp[-(vw/Q)x]
(3)
where w is the width of the river (km) and x is the
distance downstream from the waste source (km). As in
Figure 3c, note that while the substance maintains its
longitudinal identity, it gradually diminishes in con-
centration because of settling losses. However, in
comparison to the well-mixed model, the concentration
along the longitudinal axis of the river is not
homogeneous. Thus, the outlet concentration differs
from the mid-lake value. If the "outlet" for the river is
defined as being at distance L downstream from the
waste source, Eq. 3 can be used to calculate the
concentration at that point as
Pout = Pin exp(-v/qs)
(4)
Between the idealizations of complete mixing and
plug flow are those lakes where both advection and
turbulent mixing are important. Such river-run lakes,
as depicted in Figure 4a, are typically long and narrow
with a major tributary at one end and an outlet at the
other. A key feature of such sytems is that advective
water movement due to inflow and outflow is large
enough to have a comparable effect on material
transport as that caused by turbulent mixing due to,
winds and density difference. For such systems a plug
of conservative dye introduced at the head end of the
lake would move downstream along with the net water
flow but would also spread out due to turbulent mixing
as in Figure 4b. A steady state solution for such
systems comparable to Eqs. 2 and 3 can be obtained
(Reckhow and Chapra, in press) and is depicted
graphically in Figure 4c. Note that for high levels of
turbulent mixing, the solution becomes equivalent to
the completely mixed model and for zero turbulence
converges on the plug-flow model.
This exercise leads to the general conclusion that, all
other things equal, a river-run reservoir is a more
efficient settling basin than a completely mixed lake.
This can be seen by observing that the outlet
concentration (i.e., at x = L) for the well-mixed system is
higher than for the river-run lakes. Thus, the amount of
phosphorus retained by the latter would be higher. This
is a necessary consequence of the direct, linear
proportionality with concentration that is used to
characterize sedimentation for both systems. In the
well-mixed lake, sedimentation is uniform throughout
the reactor since concentrations are homogeneous. In
contrast, for the river-run lake settling is greater near
the inlet where concentrations are high. These losses
are proportionately more efficient than the reduced
sediment losses near the outlet and the effect is that
the net removal is higher than for the well-mixed case.
(a)
W
(b)
(c)
P
Pin
x=0 x=L
Distance Downstream
Figure 3. — A river with waste source at x -, (a)
overhead view, (b) movement of a plug of conservative
dye downstream, and (c) steady state profile of
concentration normalized to concentration® x = 0fora
substance that settles at a first order rate.
(b)
(c)
P
Pin
Complete Mixing
x=0
Distance Downstream
x=L
Figure 4. — A river-run lake with waste source at x=0,
(a) overhead view, (b) movement of a plug of
conservative dye through the lake, and (c) steady state
profiles of concentration normalized to inflow
concentration for a substance that settles at a first
order rate. The different profiles are for varying
degrees of turbulent mixing.
-------
332
RESTORATION OF LXKES AND INLAND WATERS
This exercise provides a theoretical basis for the
importance of sedimentation in incompletely mixed
systems. While it has been limited to river-run lakes,
similar processes would be evident in embayments and
near-shore areas where loadings enter at the system's
periphery and concentration gradients are pronounced.
Before pursuing this subject with a more realistic
model, the following section presents some empirical
evidence along the same lines.
EMPIRICAL EVIDENCE LINKING
FLUSHING AND SEDIMENTATION IN
LAKES
A number of phosphorus loading models have been
developed by fitting equations to budget data from sets
of lakes. While some of these models have a semi-
theoretical basis, many are strictly empirical and it is
often difficult to determine what they imply regarding
the cause and effect relationships underlying lake
dynamics. For example, several models have been
developed to predict the fraction of a lake's loading that
does not exit via the outlet. The first of these retention
models was that of Kirchner and Dillon (1975)
RP = 0.426 exp(-0.271 qs) + 0.574 exp(-0.00949 qs) (5)
where Rp is the retention coefficient. Eq. 5 seems to
suggest that retention is solely dependent on the lake's
hydraulic characteristics. However, as discussed
previously, Rp is also a function of its sedimentation
rate. The use of a single coefficient to define the
combined magnitude of these processes can, therefore,
obscure their individual effects. In contrast, theoretical
models provide a means for keeping the mechanisms
separate. For example, a theoretical retention coeffi-
cient for a completely mixed lake can be derived
(Chapra, 1975) by rearranging Eq. 1 at steady state to
yield
RP
v+qs
(6)
Note that in contrast to Eq. 5, the theoretically derived
coefficient has separate terms for the flushing (qs) and
sedimentation (v) effects. Algebraically, Eq. 6 can be
rearranged to yield
RPqs
v =
— RP
(7)
Eq. 5 can then be substituted into Eq. 7 to give
(8)
VKD =
[0.426 exp(-0271 qs)+ 0.574 exp(-0.00949 qs)]qs
— 0.426exp(-0.271qs)— 0.574exp(-0.00949qs)
where VKD is the apparent settling velocity of the
Kirchner-Dillon model.
In essence, the above operation has separated the
flushing and sedimentation effects that were con-
founded in the original model. The validity of the
derivation depends on the assumption that the flushing
mechanism for the lakes used to fit the Kirchner-Dillon
model obeys the simple theoretical relationship in Eq.
1, i.e., that the outflow of mass equals the product of
flow and concentration. If this is true, the manipulation
has essentially removed the flushing effect from the
retention relationship so that the residual (Eq. 8) is
solely representative of the sedimentation process.
This operation can also be performed on other
phosphorus loading models (Reckhow and Chapra, in
press) with the results displayed in Figure 5. The
surprising result is that even after the correction for
flushing is made, it appears that sedimentation is still
related to qs.
Reasons for the positive correlation of v with qs are
presently a matter of speculation. A possible explana-
tion is that the assumption of complete mixing and
ideal flushing is being systematically violated. Chapra
(1975) speculated that lakes with high values of qs
could be governed by different mechanisms than lakes
with low qs. Reckhow (1977) has suggested that lakes
with qs>50 m/yr typically receive 90 percent or more
of their inflow from one tributary. In other words, they
may actually represent a distinctive class in that they
are frequently just widened sections of rivers (i.e.,
river-run lakes). As shown previously, such lakes have
different sedimentation characteristics than completely
mixed water bodies; this might account in part for the
effect. Further, fakes whose inputs come primarily from
a major tributary (rather than from treatment plants)
may have more of their phosphorus loading associated
with eroded particulate matter that would settle quickly
upon entering the lake. Thus, such lakes would have a
higher apparent settling velocity.
While the foregoing is somewhat speculative it
suggests the importance of sedimentation mechanisms
for incompletely mixed systems. For this reason, the
following section presents some theoretical computa-
tions to assess the impact of solids dynamics on a
river-run lake.
100 I—
75
1 50
25
Kirchner and Dillon (1975)
Reckhow and Simpson (1980)
Jones and Bachmann (1976)
Vollenweider (1976), Larsen
and Mercier(1976)
50
100 150
qs(myr-i)
200
Figure 5. — Plot of apparent settling velocity (m yr 1)
versus a real water load (m ~1) for some commonly used
phosphorus loading models. (Note: the Vollenweider and
Jones and Bachmann mode Is a re also dependent on depth
with shallower lakes having smaller settling velocities
than deeper lakes.)
-------
MODELING AND ASSESSMENT OF THE TROPHIC STATE
333
THE EFFECT OF SOLIDS ON
PHOSPHORUS DYNAMICS OF A
RIVER- RUN LAKE
One objection to the use of total phosphorus loading
as a determinant of lake eutrophication is that a portion
of such input is associated with particulate matter that
settles rapidly upon entering a lake and, thus, never
influences mid-lake quality (Schaffner and Oglesby,
1978). As suggested by the previous analysis,
theoretical and empirical evidence suggests that such a
process could be especially important for incompletely
mixed systems. However,the models developed in the
previous sections of this paper are unsuitable for a
more detailed analysis of this phenomenon since they
use a single variable, total P, to define the nutrient.
Therefore, a more detailed model that differentiates
between various forms of phosphorus has been
developed. As illustrated in Figure 6, the model
consists of two particulate and two dissolved fractions.
Inorganic particulate P (i.e., associated with inorganic
particles such as fine-grained suspended sediments)
adsorbs and desorbs dissolved inorganic P via
equilibrium relationships. Phytoplankton P, on the
other hand, is modeled kinetically and takes up
dissolved inorganic P via a Michaelis-Menten relation-
ship and releases phosphorus to the dissolved organic
pool via a first order reaction. Dissolved organic P is, in
turn, recycled to the dissolved inorganic pool by a first
order reaction. In addition, the particulate fractions are
lost via sedimentation with the inorganic matter
settling at a somewhat higher rate. Details of the
model's structure are described elsewhere (Reckhow
and Chapra, in press).
Loading
1
Loading
1
Particulate
Inorganic
Phosphorus
(PIP)
Adsorption
Desorption
Dissolved
Inorganic
Phosphorus
(DIP)
Figure 6. — Schematic of multi-species phosporus model.
One-way arrows designate mass transfer mechanisms that
are modeled kinetically. The two-way arrow specifies that
sorption is treated as an equilibrium reaction (i.e., it is
modeled using a partition coefficient.
The model was applied to a river-run lake with
loadings of solids and phosphorus entering at the head
end. The results of the simulation are shown in Figure
7. Note that the inorganic particles are at a high level at
the beginning of the lake but eventually are removed
from the water column (along with considerable
quantities of adsorbed phosphorus) via sedimentation.
In addition, the solids affect productivity by light
attenuation with the result that phytoplankton growth
is suppressed for most of the lake. Thus, while
inorganic particles transport phosphorus into the
system, their tendency to diminish water clarity and to
remove P from the water via sedimentation tends to
inhibit productivity.
(a)
0.2 0.4 0.6 0.8
Distance from Inlet (km)
1.0
Figure 7. — Plots of (a) phosphorus conce' 'ation (mg rrT3)
and (b) phosphorus sedimentation flux (g F in / yr~1) versus
distance downstream from the inlet of a /iver-run lake.
The importance of these factors to remedial control
measures is demonstrated in Figure 8 where the
effects of two alternative phosphorus abatement
strategies are simulated. In the first, phosphorus
loading is controlled by lowering the dissolved
inorganic fraction with no effect on the incoming solids
as might be the case for point source treatment. The
result (Figure 8a) is that the phytoplankton levels are
decreased in proportion to the load reduction. Figure
8b, on the other hand, shows the results if the solids
loading is removed along with the phosphorus as might
be the case if land runoff control were implemented. In
this simulation, the peak phytoplankton level is higher
than in Figure 8a because less P is removed from the
water by sedimentation of inorganic particles. Addi-
tionally, the extent of phytoplankton growth increases
to encompass most of the lake because of the absence
of light attenuation by the inorganic solids. Thus, from
the standpoint of productivity the latter control
measure results in a more highly degraded lake than
before treatment. Although this computation is a
somewhat simple representation of a complex system,
it serves to illustrate the importance of solids to the
dynamics of incompletely mixed lakes.
-------
334
RESTORATION OF LAKES AND INLAND WATERS
200 400 600 800
Distance from Inlet (m)
1000
Figure 8. — Plots of phosphorus concentration versus
distance downstream from the inlet of a river-run lake where
(a) the phosphorus laoding is reduced by 50 percent with no
solids control and (b) the phosphorus loading is reduced by 50
percent and all participate inorganic solids are removed.
DISCUSSION
A specific objective of the foregoing analyses has
been to demonstrate how the modeling of the dynamics
of incompletely mixed systems is inextricably tied to the
fate of solids. It should be noted, however, that this
conclusion is also relevant to well-mixed lakes. As
stated previously, Schaffner and Oglesby (1978) have
suggested that dividing phosphorus loadings into
available and non-available (i.e., rapidly settling)
fractions could improve predictive models for well-
mixed lakes. In addition, while the present paper dwells
on horizontal features of lake physics, solids can also
have an effect on vertical aspects.
Aside from thermal stratification, the primary vertical
process influencing phosphorus dynamics is the
accumulation and release of phosphorus from the
bottom sediments. In a physical sense, solids may
influence sediment-water exchange via burial. Addi-
tionally, the chemical composition of allochthonous
particulate matter can have a decided effect on
sediment feedback (Armstrong, 1979). Finally, the
transport and fate of pollutants other than phosphorus
are inextricably tied to solids. For example, many
organic toxicants are extremely hydrophobic and when
introduced into a lake tend to associate with particulate
matter. The accurate modeling of these substances
therefore requires that adequate information on the
system's solids' budget be obtained.
In a more general sense, the foregoing analyses have
been intended to caution against applying phosphorus
loading models to systems where they are inappropri-
ate. To date there have been numerous cases where
empirical models developed from well-mixed lakes
have been applied to systems as diverse as embay-
ments, the coastal zone, and brackish estuaries. It is
hoped that the present paper will prevent such
misapplications in the future by showing how these
systems are fundamentally different from completely
mixed water bodies. Additionally, it is hoped that by
demonstrating the importance of solids to modeling
phosphorus dynamics, this paper represents a step
toward improving these models in the future.
REFERENCES
Armstrong, D. E. 1979. Phosphorus transport across the
sediment-water interface. Pages 169-176 in Lake restora-
tion. EPA 440/5-79-001. U.S. Environ. Prot. Agency,
Washington, D.C.
Chapra, S. C. 1975. Comment on "An empirical method of
estimating the retention of phosphorus in lakes" by W. B.
Kirchner and P. J. Dillon. Water Resour. Res. 11:1033.
1979. Applying phosphorus loading models to
embayments. Limnol. Oceanogr. 24:163.
1980. Application of the phosphorus loading
concept to the Great Lakes. Pages 135-152 in R. C. Loehr, et
al., ed. Phosphorus management strategies for lakes, Ann
Arbor Science, Ann Arbor, Mich.
Edmondson, W. T. 1978. A revolution in the zooplankton of
Lake Washington. Presented at the 41st Conf. Am. Soc.
Limnol. Oceanogr., Victoria, B.C.
Jones, J. R., and R. W. Bachmann. 1976. Prediction of
phosphorus and chlorophyll in lakes. Jour. Water Pollut.
Control Fed. 48:2177.
Kirchner, W. B., and P. J. Dillon. 1975. An empirical method
of estimating the retention of phosphorus in lakes. Water
Resour. Res. 11:182.
Larsen, D. P., and H. T. Mercier. 1976. Phosphorus retention
capacity for lakes. Jour. Fish. Res. Board Can. 33:1742.
Reckhow, K. H. 1977. Phosphorus models for lake
management. Ph. D. dissertation. Harvard University.
Reckhow, K. H., and S. C. Chapra. (In press). Engineering
approaches for lake management. Ann Arbor Science, Ann
Arbor, Mich.
Reckhow, K. H., and J. T. Simpson. (In press). A method for
the prediction of phosphorus loading and lake trophic
quality from land use projections. Can. Jour. Fish. Aquat.
Sci.
Schaffner, W. R., and R. T. Oglesby. 1978. Phosphorus
loadings to lakes and some of their responses. Part 1. A new
calculation of phosphorus loading and its application to 13
New York lakes. Limnol. Oceanogr. 23:120.
Shapiro, J. The need for more biology in lake restoration.
Pages 1 61-I67 in Lake restoration. EPA440/5-79-001. U.S.
Environ. Prot. Agency, Washington, D.C.
Vollenweider, R. A. 1969. Moglichkeiten und Grenzen
elementarer Modelle der Stoffbilanz von Seen. Arch.
Hydrobiol. 66:1.
1976. Advances in defining critical loading levels
for phosphorus in lake eutrophication. Mem. 1st Ital.
Idrobio. 33:53
-------
335
THE APPLICATION OFTHE LAKE EUTROPHICATION GAME
SSWIMS TO THE MANAGEMENT OF
LAKE GEORGE, NEW YORK
JAY A. BLOOMFIELD
WILLIAM B. MORTON
J. DOUGLAS SHEPPARD
New York State Department of Environmental Conservation
Albany, New York
ABSTRACT
During the past decade, limnologists have refined several concepts and techniques for managing
freshwater lakes. In this paper, we will present a computer-based game based on such concepts as
annual phosphorus and hydrologic budgets and empirical relationships among such indicator
variables as winter total phosphorus, summer chlorophyll a, summer Secchi disk depth, extent of
macrophyte growth, and hypolimnetic oxygen depletion. Although the game was originally
designed as a vehicle for instructing personnel of government agencies in New York State in the
subject of lake management, we shall discuss the use of the game in projecting the future quality
of Lake George in Warren County, N.Y. under varied assumptions concerning changes in land use
and waste water treatment.
INTRODUCTION
The dynamics of lake ecosystems are generally too
complex to understand through simple explanation. A
number of North American researchers are using
simulation modeling to more clearly understand
biological, physical, and chemical relationships in
lakes. This paper demonstrates the use of a simulation
gaming model developed to help personnel of New York
State agencies understand the consequences of
various decisions which affect lake ecosystems. Lake
George, in northern New York State, was selected as
the study area to demonstrate the use of the model
(Figure 1).
The simulation model consists of difference equa-
tions for lake winter total phosphorus, bottom anoxia,
and extent of rooted vegetation. Regression equations
developed for New York State lakes are used to
calculate average summer Secchi disk depth and
chlorophyll a from winter total phosphorus. Once the
geomorphometric parameters of a specific lake are
given to the program, the user may vary human
population growth, land use patterns, and degree of
treatment for sanitary wastes or stormwater. The
SSWIMS model then predicts phosphorus, Secchi disk,
chlorophyll a, bottom water oxygen, and the growth of
rooted aquatic plants over any specified time period.
Results may be displayed in either tables or plots.
Variables in the game or the actual lake may be easily
changed. The model program is written in FORTRAN, is
inexpensive to use, and is designed for interactive use.
The conceptual model used for deriving the
equations is shown in Figure 2. Its compartments
represent either algebraic or difference equations. The
Figure 1. —
-------
336
RESTORATION OF LAKES AND INLAND WATERS
Figure 2
PHOSPHORUS MODEL FOR A LAKE WATERSHED
Figure 2. — Phosphorus model for a lake watershed.
paths represent influence rather than transfer of an
entity such as energy or matter. The direction of the
arrow on each path indicates the active and passive
compartments.
The major component of the SSWIMS model is an
annual phosphorus budget (see Figure 3). Loadings
from various sources in the watershed are added to the
lake which is assumed to be well mixed. A certain
fraction of the phosphorus is retained in the lake yearly,
and under anoxic conditions, some is released from
bottom sediments. The portion of the model related to
fish production and quality of fishing is discussed only
briefly in this paper.
SSWIMS LAKE MODEL
Figure 3
THE MODEL
The model SSWIMS described here is a deterministic
version lacking fishery equations. Our experience with
driving the model with stochastic climatic inputs
(amount of precipitation and runoff) has not improved
the validity of the predictions, as continuous time
series for variables such as winter total phosphorus or
chlorophyll a do not exist for Lake George (Ferris, et al.
in press).
A. Phosphorus
The phosphorus equation predicts the annual change
in the amount of lake phosphorus (Xi). Its form is:
dX,
dt
= (RLOAD + RELS + HSUM + XNPSUM)
— ( 1 — CUP*VOL
TH
+ FCHAIN)*X,
eq. 1
where:
CUP = Phosphorus concentration in inflow from upstream
lake(s) (mg/m3)
RLOAD = Direct atmospheric loading to the lake (mg/yr)
and
eq. 2
Figure 3. — SSWIMS lake model.
RLOAD = CPR* PR'AREAL
where:
CPR = Phosphorus concentration in precipitation (mg/m3)
PR = Annual precipitation (m3/m2)
AREAL = Lake surface area (m2)
RELS = Anoxic release from bottom sediments (mg/yr.)
and
RELS = ALPHA* VOLAN* CHK/VOL eq. 3
where:
ALPHA = Anoxic release rate parameter (mg/yr.)
VOLAN = Summer volume of anoxic water (m3) (see
Section D)
VOL = Volume of lake (m3)
and
CHK = 1 — BETA* Xi/VOL eq. 4
where:
BETA = Concentration dependent release parameter
(m3mg)
HSUM=Phosphorus contribution of septage and sewage
(mg/yr.)
where:
HSUM =1 GAMMAi *(HPi(T) + HSi (T)/3))
i = 1
and
GAMMAi = Unit load per capita for treatment level;
(mg/cap-yr)
HPi(T) = Permanent human population served by treatment
level, at time T
-------
MODELING AND ASSESSMENT OF THE TROPHIC STATE
337
HSi(T) = Seasonal human population (3 months) served by
treatment level I at time T.
HPi(T) and HSi(T) are functions of time. For HPi(T), the
function is :
HPi(T) = HR(0) * (1 - PPI:)T
where:
eq. 5
HPi(T) = Population at any time T.
HPi(O) = Initial population at time TFIRST.
PPh = Fractional annual population increase
TFIRST = Initial year
XNPSUM = Phosphorus contribution from diffuse sources
(mg/yr)
where:
m
XNPSUM =AWS I XL *PLU/100
1 = 1
and
eq. 6
AWS = Watershed area (ma)
XU = Unit load of phosphorus from ith land use (mg/m2-yr)
PLUi = Percent of watershed in the ith land use (unitless)
% Urban land use is increased using the same function as
human population.
As urban use increases, forest and agriculture are decreased,
but the ratio between % forest and % agriculture remains
constant (an assumption).
FCHAIN = Exponential loss rate of phosphorus to food
chain and bottom sediment (yr~')
TH = Hydraulic retential time of lake (yr)
The annual amount of lake phosphorus is converted to
spring total phosphorus concentration by the equation
(Chapra and Tarapchak, 1976):
TPs, =
0.9
*VOL
eq. 7
B. Chlorophyll a:
Summer chlorophyll a (CHLOR) can be interpreted as
a simple estimate of lake food chain production. The
algebraic equation in SSWIMS is:
CHLOR = EXP (—0.51 + 0.86 *LN (TP.P))
eq. 8
C. Secchi disk depth:
Secchi disk depth (SECCHI) is a rough measure of
water clarity. The algebraic equation is:
•jprrui - I10-09 - 2-93 *LN (CHLOR) I
SECCHI - I 01Q I
L CHLOR ^ 30 mg/m3 J
CHLOR > 30 mg/m3
eq. 9
Equations 8 and 9 a re derived from data for New York
State lakes (Oglesby and Schaffner, 1975, 1978;
Bloomfield, 1978 a,b, 1980) concerning summer
chlorophyll a, Secchi disk depth, and winter total
phosphorus concentration. The data of Wood and Fuhs
(1979) concerning Lake George tend to follow these
relationships.
D. Extent of Anoxic Conditions:
A deficiency of oxygen in the bottom waters of a lake
during summer thermal stratification generally indi-
cates intense oxidation of organic materials in the
bottom sediments and adjacent waters. A lack of
oxygen in bottom waters during the summer is
significant for two reasons.
First, reducing conditions tend to increase the
solubility of phosphorus compounds, e.g., phosphorus
in lake bottom sediments may dissolve and thus
become available to stimulate algal and other plant
growth. Second, anoxic conditions often lock valuable
game fish into the cold bottom waters during the
summer.
A simple difference equation based on the work of
Welch and Perkins (1979) is used to simulate summer
hypolimnetic oxygen depletion. The simulated variable
(X2) represents the area! hypolimnetic oxygen depletion
rate. The equation is:
dX2
~dT
- = THETA * (ODR — X2)
eq. 10
where:
THETA = Decomposition rate constant (unitless)
ODR = Equilibrium depletion rate (mgO2/m2-day)
The equilibrium rate (ODR) is defined from Welch
and Perkins (1979), in our notation:
ODR = 38.02
X,
-)*
AREAL
I +FC*TH)
0.37
eq. 11
where all constants and variables have been previously
defined.
Hypolimnion dissolved oxygen at the end of summer
thermal stratification is then defined as:
DOHPO = DOSAT
eq. 12
DOHYPO = Average hypolimnetic dissolved oxygen at
the (mg O2/l) end of summer thermal
stratification
ANMAX = Maximum duration of thermal stratification
(days)
VHYPO = Volume of hypolimnion (m )
AHYPO = Area of hypolimnion (m )
DOSAT = Oxygen saturation value for hypolimnion (mgOz/l)
VHYPO and AHYPO are calculated from the
cumulative volume and area functions and the
following equation which was developed assuming a
simple relationship between minimum depth of anoxia
(ZANX) and hypolimnetic dissolved oxygen:
ZANX =ZST
where:
(DOHYPO I I DOHYPO I
1 — I +ZBOT' I I
DOSAT / \ DOSAT /
eq. 13
ZANX = Minimum depth of anoxia (m)
ZST = Depth of seasonal thermocline (m)
ZBOT = Depth of bottom (m)
-------
338
RESTORATION OF LAKES AND INLAND WATERS
and VOLAN, the anoxic volume (m3) is then
calculated from the cumulative volume relationship
and ZANX.
E. Macrophyte Growth
The dynamics of macrophytes (aquatic weeds) in
lakes have not been studied in enough detail to permit
quantitative simulation. The term "weeds' in this
paper will be limited to emergent and submergent
vascular plants and macroalgae such as Nitella. The
difference equation describing the area of the lake
covered by a discernible weed growth (Xa) is relatively
straightforward and has several assumptions implicit
in its formulation. They are:
1. For a specific lake, weed beds can only increase to
cover an ultimate area (AWMP, m2). This potential area
is defined by the morphometry of the lake and the
fertility of the bottom sediments.
2. Light is a major limiting factor to aquatic weed
growth.
3. Weed beds cannot extend into anoxic zones or into
areas of poor growing conditions (extreme hydrostatic
pressure, poor bottom conditions, high current activity,
etc.) The equation for weed growth is:
dX3
dt =WGROWX * (1 — Xs/AWMP)
where:
eq. 14
WGROW = Intrinsic rate of increase of weed beds (yrs'1)
AWMP = Area of potential weed penetration (m2)
and:
AWMP = AREAL * (1 — F (ZWMP)) *WPER eq. 15
where:
ZWMP
(0.83 + 1.22 'SECCHI \
ZANX 1
ZWMP < ZANX
ZWMP >ZANX
in press)
ZWMP = Depth of potential weed penetration (m2)
WSECC = Light dependent coefficient (unitless)
WPER = Percent of total lake area where weeds will
grow potentially (unitless)
F(ZWMP) = Depth, vs bottom area relationship for a
specific lake
(Dunst,
eq. 16
APPLICATION OF
LAKE GEORGE
SSWIMS TO
Lake George is located in the eastern Adirondack
Mountains of New York State and the southeastern
portion of the Adirondack State Park (Figure 1). It is one
of the most heavily used recreational waters in the
eastern United States. Most homes around the lake use
its water directly, often without disinfection.
Recognizing that the lake is a unique resource in the
northeastern United States, there has been a strong,
long-term State and local commitment to protect and
enhance water quality in Lake George. However,
efforts to protect water quality in Lake George have not
been entirely successful. Since the early 1920's when
Secchi disk measurements were first published, water
transparency has decreased from 10 to slightly over 6
meters (Needham, et al. 1922).
Recent efforts to stem and reverse the trend toward
eutrophy and increasing bacterial pollution in Lake
George have culminated in a plan to collect and divert
sewage out of the south basin of the lake. This
proposal, which enjoys strong support among many
lake residents and government officials, is not without
its critics who argue that sewering the lake will
facilitate its development. Critics also contend the
increased storm runoff from an expanding urban area
will more than offset the benefits of sewering the south
lake basin.
The interaction of these factors— population growth,
sewering vs. non-sewering, and controlling phos-
phorus in urban storm drainage — were compared
using the SSWIMS model. Fourteen scenarios with a
time frame to the year 2030, which corresponds to the
project design capacity of the proposed Warren County
sewer system, were simulated.
Examining three different population projections
revealed that the annual rate of population increase in
the southern basin of Lake George is about 1.4 percent.
An annual rate of increase of 2.0 percent served as the
upper limit of population growth on the assumption
that sewering the lake basin would speed up the rate of
development.
It was also assumed that urban expansion in the lake
basin would increase in proportion to population
growth, or that doubling the population would double
the urbanized area. Management policies for control-
ling 25, 50, and 90 percent of the phosphorus in urban
storm runoff were compared with each other and with
a policy of non-control.
The value of each parameter and constant used in
the simulation is shown in Tables 1 and 2. Their
sources are:
1. Physical Constants (TH, VOL, AREAL, AWS,
ZBOT). These values were obtained from Wood and
Fuhs (1979) and agree closely with other published
information on Lake George.
2. Land use parameters (XLU, PLU) were developed
from Fuhs (1972) and Hetling (1974). Since agriculture
and urban area is presently quite limited in the Lake
George watershed, unpublished information provided
by Nicholas L. Clesceri (pers. comm.) on stormwater
quality was used to confirm that the phosphorus
loading estimates made by Hetling (1974) were
reasonable.
3. Per capita phosphorus contribution parameters
(GAMMA) were assumed to be zero for the sewered
population. The sewered areas are served by two small
municipal plants at Lake George Village and Bolton
Landing. Each plant discharges to natural sand beds
and extensive field work has indicated that these plants
probably contribute less than 5 percent of the total
annual input of phosphorus to Lake George (Aulen-
bach, et al. 1976). The value of GAMMA for the
population served by septic tanks was determined from
the estimates of septic tank phosphorus contributions
made by Gibble (1974), Hetling (1974) and Ferris, et al.
(in press). None of these estimates has been checked
by field measurements.
-------
MODELING AND ASSESSMENT OF THE TROPHIC STATE
339
Table 1. — Parameters and constants representing conditions used in simulations of South and North Lake George.
SYMBOL
TH
VOL
AREAL
AWS
ZBOT
FCHAIN
THETA
DOSAT
ALPHA
BETA
ZST
WGROW
WPER
ANMAX
PPI
CPR
PR
PARAMETER OR CONSTANT
Hydraulic retention time of lake
Volume of lake
Surface area of lake
Watershed area
Maximum depth
Phosphorus retention rate
Decomposition rate parameter
Hypolimnetic saturation dissolved oxygen
Anoxic release rate parameter
Concentration dependent release parameter
Depth of seasonal thermocline
Aquatic vegetation growth rate
Potential percent of lake where
vegetation will grow
Maximum duration of thermal stratification
Growth rate parameters for human
population groups and developed area
Phosphorus concentration in precipitation
Annual precipitation
y
m3
m2
m2
m
yr"
unitless
mg-liter~1
mgP-yr"1
m3mgP~1
m
yr"
unitless
days
yr"
mgP-m
m
SOUTH
6.90
1.02x10"
5.8x10?
3.1x10°
58.0
0.28
1.0
13.6
1.0x10"
0.005
12.0
0.5
50.0
150.0
0.014
10.0
1.0
NORTH
4.49
1.08x10"
5.6X107
1.8x10"
53.0
0.25
1.0
13.6
1.0x10"
0.005
12.0
0.5
50.0
150.0
0.014
10.0
1.0
Table 2. — Parameters and constants related to human population and
land use.
LAND USE
PLU (percent)
XLU
South
North mgP-m~!Vr~1)
Forest
Cropland
Developed
(Non-contributing)
82
5
3
(10)
95
0
2
(3)
4.0
30.0
100.0
(0.0)
TREATMENT TYPE Population Served (1975)
GAMMA
South Basin North Basin (mgP-cap~1-yr~
HP HS HP HS
Sewered, no discharge 2.200 21,500 0 0
Septic tanks 2.90026,4001,1003,300
0.0
6.0 x 104
PLU — Percent of watershed in ith land use
XLU — Unit load of phosphorus from ith land use
HP — Permanent human population served by treatment level i
HS — Seasonal human population served by treatment level i
4. Phosphorus retention (FCHAIN) was determined
by dividing the difference between annual phosphorus
inputs and losses (outflow) by the annual average
phosphorus concentration of the lake water. Various
estimates of phosphorus retention for Lake George
made by Aulenbach (1973), Ferris, et al. (in press) and
Wood and Fuhs (1979) yield values for FCHAIN ranging
from 0.2 yr to over 1.5 yr
5. Anoxic zone and aquatic vegetation parameters
(THETA, ALPHA, BETA, ZST, WGROW, WPER, ANMAX)
were estimated using SSWIMS data from a variety of
New York State lakes and from Ferris, et al. (in press)
data concerning vegetation, thermal stratification, and
dissolved oxygen. Lake George's hypolimnion is at
present well oxygenated and aquatic vegetation often
is 9 to 10 meters deep.
6. Precipitation parameters (PR, CPR). The annual
average precipitation was estimated from U.S. Depart-
ment of Commerce data (1979) for the Glens Falls, N.Y.
station. The phosphorus content of precipitation was
derived from Wood and Fuhs (1979) and from 10 years
of data from the New York State precipitation chemistry
network (U.S. Dep. Inter. 1979).
7. Human population estimates (HP, HS) and
projections were derived from Lawler, Matusky, and
Skelly, Inc. (1975), New York Department of Environ-
mental Conservation (1976), and Hazen and Sawyer
(1977). Annual human population growth is estimated
at an average of 1.4 percent.
RESULTS
Figure 4 shows that of the 14 water quality
management alternatives compared with the model,
only six maintain existing water quality or increase
water transparency in Lake George. Of these, four are
not feasible because they would require more than 25
percent removal of phosphorus from urban storm
runoff. The lack of suitable terrain for constructing
control devices makes these alternatives technically
difficult to accomplish (Figure 4). Therefore, only 2, 6,
and 13 appear technically feasible.
42.7
§366
305
3244
ft ,,,
Summer Sec
a N a
3 — f» ci
14
12
10
a
4
2
1 Predicted Secchi Disk Readings far
1
Water Quality Management Alternatives in
South Bosin-Lalie George
Q° Year 2OOO
E3= Year 203O
1
|
|
' 2 ' 3 '
' '/
nn
n0n
I ',
|
I •
4 ' 5 ' 6 ' 7 ' 8 ' 9 ' 10 ' II ' 12' 1
Alternatives
ra •
hi
1
3' 14
Figure 4. — Predicted Secchi disk readings for water quality
management alternatives in south basin-Lake George.
Under alternative 13 (Table 3), population growth
would have to be limited and urbanization held to
present levels. Also, the proposed sewer system would
not be constructed and phosphorus in urban runoff
-------
340
RESTORATION OF LAKES AND INLAND WATERS
Year
1975
1980
1985
1990
1995
2000
2005
2010
2015
2020
2025
2030
Table 4.
Year
Table 3. —
Win Tot P
(Aig/0
9.
8.
8.
8.
8.
8.
8.
8.
8.
8.
8.
8.
— Construct
Win Tot P
Limit population growth and halt urban development
Chlor a
(pg/l)
4.
4.
3.
3.
3.
3.
3.
3.
3.
3.
3.
3.
, no sewerage system (south basin).
Secchi Z Anoxic Bloom Sev. % Area Weeds
6.1
6.4
6.4
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
6.5
sewer, reduce P in urban
Chlor a
Secchi Z
meters
57.4
57.0
57.0
57.0
57.0
57.0
57.0
57.0
57.0
57.0
57.0
57.0
storm runoff
None
None
None
None
None
None
None
None
None
None
None
None
by 25 percent.
13.6
14.4
14.7
14.7
14.7
14.7
14.7
14.7
14.7
14.7
14.7
14.7
1.4 percent
Anoxic Bloom Sev. % Area Weeds
Population
53,000
53,000
53,000
53,000
53,000
53,000
53,000
53,000
53,000
53,000
53,000
53,000
annual growth
Population
% Basin Urban
3.
3.
3.
3.
3.
3.
3.
3.
3.
3.
3.
3.
rate (south basin).
% Basin Urban
(yyg/l) (//g/l) meters
1975
1980
1985
1990
1995
2000
2005
2010
2015
2020
2025
2030
9.
7.
7.
7.
8.
8.
8.
8.
8.
8.
9.
9.
4.
3.
3.
3.
3.
3.
4.
4.
4.
4.
4.
4.
6.1
6.6
6.6
6.5
6.5
6.4
6.4
6.3
6.3
6.2
6.2
6.1
57.4
57.0
57.0
57.0
57.0
57.0
57.0
57.0
57.0
56.9
56.9
56.9
None
None
None
None
None
None
None
None
None
None
None
None
13.6
14.7
14.9
14.9
14.8
14.7
14.6
14.5
144
14.3
14.1
14.0
53,000
56,815
60,905
65,290
69,990
75,028
80,429
86,219
92,426
99,080
106,212
113,858
3.
3.
3.
4.
4.
4.
5.
5.
5.
6.
6.
6.
would not be controlled. Such indicators as predicted
nutrient enrichment, chlorophyll a, Secchi disk depth,
depth to anoxic conditions, bloom severity, and percent
area of the lake supporting weed growth, as revealed in
Table 3, show that as long as this policy remains in
effect the lake will enter into a steady state equilibrium
in which there would be no further impairment to
water quality as measured by these indicators.
Although water quality in Lake George could be
maintained at present levels for the indefinite future
under this alternative, there would be no net
improvement in water quality.
As Figure 4 and Table 4 reveal, water quality can be
maintained to the year 2030 by sewering the south
lake basin and reducing phosphorus in urban storm
runoff by 25 percent, provided that population growth
and urbanization do not increase above the current
projected annual rate of 1.4 percent (alternative 6).
An examination of Figure 4 reveals, however, that if
the annual growth in the south basin of Lake George is
allowed to increase to 2.0 percent, constructing the
sewer and reducing P in urban storm runoff by 25
percent as in alternative 9, is not sufficient to offset the
impact of accelerated growth and development.
DISCUSSION
An examination of the simulated scenarios reveals
relatively few options either for enhancing water
quality in the south basin of Lake George or
maintaining it at existing trophic levels. As the
predicted Secchi disk readings for the south basin in
Figure 4 show, only two water quality management
alternatives, 6 and 13, achieve this goal while
appearing to be technically feasible. The remaining
alternatives either fall short of maintaining present
levels of water quality in the south basin, or would be
technically difficult to accomplish.
Although technically sound, alternative 13 is
probably neither economically or politically feasible.
This alternative would require an immediate cessation
of growth and development in the south basin.
Presumably, strategies to limit growth would require
combining strict land use controls with regional
growth-inhibiting economic policies.
The findings do indicate that water quality can be
maintained at present levels to the year 2030 through
sewering the south lake basin and reducing P in urban
storm runoff by 25 percent, provided that population
growth and urbanization do not increase above the
current annual rate of 1.4 percent (alternative 4). With
a modest investment in control structures, backed by
land use planning, management, and controls, it
appears technically feasible to reduce P in urban storm
runoff by 25 percent.
There is, however, considerable uncertainty as to the
influence the sewer system will exert on rates of
growth and development in the south basin. Some
have suggested that the sewer system will accelerate
the rate of population growth and urban development.
If, for any reason, the annual rate of growth and
development exceeds 1.4 percent, the policy of
sewering the south lake basin and of reducing P in
urban storm runoff by 25 percent will fall short of
meeting the objective of maintaining the present level
-------
MODELING AND ASSESSMENT OF THE TROPHIC STATE
341
of water quality for the life of the sewer project, i.e., to
the year 2030 (see Figure 4).
A further implication is that, if it is concluded that
reducing P by more than 25 percent is not feasible and
if the sewer project accelerates development, then
controls which curb growth and development would
have to be instituted before 2030.
Faced with considerable uncertainty about future
rates of growth (if the goal is to maintain or enhance
water quality),prudent decisionmaking would dictate
that a water quality management strategy should be
based on a 2 percent annual rate of growth. Growth
would have to be limited to achieve the goal unless a
much greater control of P could be insured.
Other than simulating in-basin tertiary treatment
with 90 percent P removal (alternative 14), no
simulations were made for other in-basin wastewater
treatment alternatives in this paper. However, what is
clear from the SSWIMS simulations is that no matter
what the alternative for in-basin treatment may be, it
must provide almost 100 percent P removal, combined
with 25 percent P reduction in urban runoff if water
quality is to be maintained at present trophic levels.
Furthermore, in terms of a comprehensive approach to
water quality planning, in-basin strategies to effect
wastewater treatment may conflict with strategies for
reducing P in urban storm runoff. For example,
diverting wastewater outside the basin, as is currently
proposed, would make the sand filter beds at the Bolton
and Lake George Village sewage treatment plants
available for treatment of storm runoff. Presumably,
most approaches to in-basin wastewater treatment
would use the sand filter beds, thereby preventing their
potential use for treating urban runoff.
Figure 4 also contains some additional water clarity
information for comparison. On the far left are the
present average summer Secchi disk depths for the
north and south basins of Lake George as predicted by
SSWIMS. These values agree with information
presented by Clesceri, et al. (in press) and Wood and
Fuhs (1979). On the right of the diagram are the
average summer Secchi disk depths for three New York
State lakes with a morphometry similar to Lake George:
Canandaigua Lake in western New York, Cayuga Lake,
a Finger Lake impacted by point sources or municipal
waste, and Otsego Lake in east-central New York. All
presently exhibit water clarity inferior to almost all of
the scenarios predicted by SSWIMS for South Lake
George. This is presented only to show how unique the
present condition of Lake George is when compared to
other major New York State lakes.
THE QUESTION OF FISHERIES
MANAGEMENT
The original SSWIMS model included State variables
for management of lake fish community and associated
sport fisheries. Although on a global basis, Ryder's
Morphoedaphic Index (MEI, Ryder, et al. 1974) and the
more recent work of Oglesby (1977) have been used to
predict fish production from morphometry and indica-
tors of lake trophic status, it was clear to us at that time
that the information available for New York State lakes
was not sufficient to test the validity of either
technique. This is still true because of the extreme
effort required to accurately estimate fish numbers in a
large lake. The work of David Green on the fishery of
Canadarago Lake (Harr, et al. 1980) is one of the few
efforts documented in which changes in the fish
community were quantified concurrently with limno-
logical studies and improved wastewater treatment
measures. Although phosphorus inputs to the lake
were significantly reduced in 1973, only minor
fluctuations in water clarity, chemistry, and biota
occurred. However, major changes in the numbers of
various fish species were observed, probably related to
climatic conditions. It should be noted that reducing
phosphorus input to the lake did not affect the MEI
parameters.
Thus, the inclusion of fishery variables in eutrophi-
cation models remains limited by the future collection
of adequate fisheries data. However, recent compari-
sons of data in New York State Angler Surveys(Brown,
1973; NYS Dep. Environ. Conserv. 1978) with trophic
status of various lakes reveal that the total catch per
unit effort (Figure 5) and the relative proportion of
gamefish to forage species taken (Figure 6) are
somewhat related to such variables as summer
chlorophyll a. Ideally, the model should provide outputs
on the resulting fisheries and their associated
socioeconomic values. The challenge of meshing
fishery and limnological concepts lies in relating catch
per unit effort of various fish species to productivity
indices. In turn, these catches per unit of effort could be
translated into socioeconomic values which would help
decisionmakers in determining the best alternative.
The information from this segment of the model would
also be useful in assessing the effectiveness of the
various management techniques, such as seasons,
creel limits, size limits, habitat improvement, and
12
>
LJ
O
Relationship between total catch
rate and summer chlorophyll a
for eighteen New York State lakes.
05 10 15 20 25
SUMMER CHLOROPHYLL a
(mg/m3)
Note.- Line is for comparison purposes only.
Figure 5. — Relationship between total catch rate and
summer chlorophyll a for 18 New York State lakes.
-------
342
RESTORATION OF LAKES AND INLAND WATERS
stocking (see Figure 7). Perhaps the best we can hope
for at this time is to be able to project whether the
fishing (fish/angler-day) for a given species will be
good, fair, or non-existent, and subsequently weigh the
socioeconomic impacts. Nevertheless, the challenge to
do better should not be ignored.
50
o>
1C.
o
40
.C
-------
MODELING AND ASSESSMENT OF THE TROPHIC STATE 343
New York State Department of Environmental Conservation.
1976. N.Y. State Section 208 Interim Population Projec-
tions.
_. 1978. N.Y. Angler Survey. Preliminary Results
1976-1977. Albany.
1979. Significance and control of urban runoff in
Lake George. Proposal to the U.S. Environ. Prot. Agency,
Albany, N.Y.
Oglesby, R. T. 1977. Relationships of fish yield to lake
phytoplankton standing crop, production and morphoe-
daphic factors. Jour. Fish Res. Board Can. 34:2271.
Oglesby, R. T., and W. R. Schaffner. 1975. The response of
lakes to phosphorus. In K. S. Porter, ed. Nitrogen and
phosphorus, food production, waste, the environment. Ann
Arbor Science, Inc., Ann Arbor, Mich.
1978. Phosphorus loadings to lakes and some of
their responses. Part 2. Regression models of summer
phytoplankton standing crops, winter total P and transpar-
ency of New York lakes with known phosphorus loadings.
Limnol. Oceanogr. 23:135.
Ryder, R. A. , et al. 1974. The morphoedaphic index, a fish
yields estimator — review and evaluation. Jour. Fish Res.
Board Can. 31:663.
U.S. Department of Commerce. 1979. Climatological data for
New York State. Natl. Oceanic Atmos. Admin., Asheville,
N.C.
U.S. Department of Interior. 1979. Water resources data for
New York, 1978. U.S. Geol. Surv., Albany, N.Y.
Welch, E. B., and M. A. Perkins. 1979. Oxygen deficit —
phosphorus loading relation in lakes. Jour. Water Pollut.
Control Fed. 51:2823.
Wood, L. W., and G. W. Funs. 1979. An evaluation of the
eutrophication process in Lake George based on historical
and 1978 limnological data. N.Y. State Dep. Health, Albany.
-------
344
VARIABILITY OF TROPHIC STATE INDICATORS
IN RESERVOIRS
WILLIAM W. WALKER, JR.
Environmental Engineer
Concord, Massachusetts
ABSTRACT
As part of the Environmental Water Quality Operation Studies being conducted by the Army Corps
of Engineers, a data base has been compiled that describes the morphometry, hydrology, and
water quality of over 300 reservoirs throughout the United States. The data base will be used to
test and evaluate existing empirical models for assessing eutrophication problems and to develop
new methods, where appropriate. This work has been motivated by concerns over the application
of existing models to reservoirs, despite the fact that most have been developed using data bases
consisting entirely of northern, natural lakes. Existing methods may not be adequate for reservoirs
because of differences in morphometry, hydraulics, sedimentation, and region, that may influence
responses to nutrient loading. To provide preliminary insights into the effects of using different
data-reduction procedures and into the adequacy of the data for model testing purposes, EPA
National Eutrophication Survey data from 76 phosphorus-limited Corps impoundments are
analyzed and used in testing Carlson's (1977) Trophic State Indices. Seasonal effects and
variance/covariance components are identified at different averaging levels. Results indicate that
chlorophyll a levels in Corps reservoirs are generally less sensitive to phosphorus or transparency
than in the natural lakes used by Carlson in developing the index system. The use of error analysis
for assessing the adequacy of the data set for model testing purposes is demonstrated
INTRODUCTION
The development of phosphorus loading/trophic
state response models over the past decade has greatly
increased the feasibility of lake water quality planning.
Most of these models have been based upon empirical
studies of data from natural lakes in glaciated regions.
Their applicability to manmade impoundments is in
question because of lake/reservoir differences in age,
morphometry, hydrodynamics, sedimentation, and
region (Thornton, et al. 1980). To provide a basis for
testing available models, data describing the morpho-
metry, hydrology, water quality, and sedimentation
rates of over 300 active U.S. Army Engineer reservoirs
have been compiled (Walker, 1980a). During the next
year, this data base will be used in a systematic
assessment of phosphorus loading models and rela-
tionships among trophic state indicators in reservoirs.
The data base currently contains over two million
water quality observations. Testing empirical eutrophi-
cation models in reservoirs requires averaging water
quality measurements over spatial and temporal
scales. If within-pool water quality variations are not
random with respect to date, station, or depth, then
summary statistics for a given reservoir will depend to
some extent upon the particular data reduction method
employed. The choice of reduction method may, in turn,
influence conclusions regarding the adequacy of
existing models as well as the parameter estimates of
any new models which may be developed.
There is no standard data reduction procedure which
can be used prior to model development, testing, or
application. Methods have included, for example, (1)
taking the median or mean of all within-pool
observations (U.S. EPA, 1975); (2) sequential averaging
over depths, stations, and dates (Lambou, et al. 1976);
(3) sequential averaging within specific depth ranges
(Carlson, 1977); and (4) various weighted averaging
schemes which reflect morphometric characteristics
(Boyce, 1973). As compared with natural lakes, many
reservoirs pose special data reduction problems
because of extreme spatial and/or temporal variations
in conditions.
This paper describes investigations of the variability
of trophic state indicators among and within a group of
Corps reservoirs. The analysis covers seasonal re-
lationships, variance/covariance components, regres-
sion analyses, and error analyses. This work has been
undertaken to assess the implications of using different
data reduction procedures and to assess the adequacy
of the data for model testing purposes.
DATA BASE
National Eutrophication Survey (U.S. EPA, 1973)
data have been used as a basis for this analysis. The
relatively uniform sampling program designs used by
the survey provide data that are suitable for statistical
treatment. One drawback, however, is that under this
program reservoirs were typically sampled only three
times during one growing season. In future work, we
plan to examine data from other agencies, which, in
many cases, are more intensive and/or cover longer
periods. The Survey data have been screened to
eliminate data from 19 reservoirs which were
predominately nitrogen-limited (based upon bioassays)
and to eliminate all stations with fewer than three
sampling dates for total phosphorus, chlorophyll a, and
-------
MODELING AND ASSESSMENT OF THE TROPHIC STATE
345
transparency. The resulting file contains 963 observa-
tions from 306 stations in 76 reservoirs.
Surface total phosphorus, Secchi depth, and chloro-
phyll a values have been expressed in terms of
Carlson's Trophic State Indices (Carlson, 1977):
IP =4.2+ 33.2 logioP eq. 1
IT =60-33.2 log ,0Zs eq. 2
IB =30.6 + 22.6 logioB eq. 3
where,
P = total phosphorus concentration (mg/m3)
Zs = Secchi depth (m)
B = chlorophyll a concentration (mg/m3)
T = transparency
The indices are calibrated so that the three versions are
equivalent, on the average, when applied to mid-
summer, epilimnetic data from northern, natural lakes.
Expression of measurements on these scales tends to
reduce the skewness in the distributions of the
variables and provides benchmarks for assessing
reservoir trophic state relationships in comparison to
those typical of natural lakes.
The latitudes of 309 natural lakes sampled by the
Survey are compared with the latitudes of 106 Corps
reservoirs sampled by the Survey in Figure 1. The
distribution of natural lakes is bimodal, with a northern
peak (glacial lakes) and a southern peak (subtropical
lakes in Florida). Most of the Corps reservoirs may be
influenced by regional factors as well as the effects of
impoundment type.
0 10 20 30 40 50 0 10 20
NATURAL LAKES
!N=309)
CE RESERVOIRS
(N=106)
Figure 1. — Latitudes of natural lakes and Corps reservoirs
sampled by the EPA National Eutrophication Survey.
SEASONAL RELATIONSHIPS
Average seasonal variations in the index com-
ponents are depicted in Figure 2. Station means have
been computed and their effects removed from the data
prior to calculating the mean and standard error for
each month (March to November) and index com-
ponent. Analyses of variance indicate that fixed
monthly effects are significant (p<.0001) but explain
only 11 percent of the total within-station variance of
each index. The seasonal variations depicted in Figure
2 are characteristic of this collection of reservoirs but
not necessarily of each individual reservoir.
Average seasonal effects on phosphorus and
transparency are similar: Both tend to be lowest during
March and midsummer and highest during April and
November, possibly reflecting seasonal flow and
turbidity variations and the influences of turnover
periods. Monthly effects on chlorophyll a suggest a
spring maximum (April-May), followed by a June
depression, a midsummer maximum, and lower values
in November. Temperature and light effects may be
responsible for the relatively low chlorophyll a levels
during March and November. The June depression may
be caused by seasonal succession of algal species. A
more detailed examination of the data indicates that
lower June chlorophyll a levels are characteristic of
about half of the stations sampled in June, while the
rest have June levels more typical of May or July
values. In testing seasonal aspects of TSI behavior,
Carlson (1977) also noted a June depression in
chlorophyll a index relative to the phosphorus index in
three natural lakes.
Differences among various versions of the index
provide a measure of "lake-like" behavior, since the
index system is calibrated so that IP, IT, and IB,values are
equivalent, on the average, when applied to mid-
summer epilimnetic data from northern, natural lakes.
Figure 2 indicates that the range of index means is
generally lowest during midsummer and highest
during March, June, and November (approaching 15).
Minor recalibration of the phosphorus and/or trans-
parency index would bring I and I into agreement for
all seasons, since the monthly effect curves in Figure 2
are roughly parallel. Since seasonal chlorophyll a
behavior is fundamentally different, however, re-
calibration alone would not eliminate biases (i.e.,
significant differences between IB and IP or IT) for all
seasons.
62
58
X
g
50
46
42
mean ±1 std. error
5 7 9
MONTH
11
Figure 2. — Monthly variations in trophic state indices.
-------
346
RESTORATION OF LAKES AND INLAND WATERS
VARIANCE COMPONENTS
Trophic index observations can be classified in a
hierarchy defined by region, reservoir, station, and
sampling date. Variations at each level could account
for some portion of the total variance of each index. A
nested analysis of variance procedure (Statist. Anal.
Inst. 1979) has been applied to derive pooled estimates
of variance and covariance components according to
the following model:
Var(l) = al + a,2«j> + a!(d,n + 01
eq. 4
where,
aa = variance among regions, defined by Corps
districts
cr,2(d) = variance among reservoirs, within districts
crlid.fi = variance among stations, within reservoirs and
districts
crl = variance within stations
This model has been used to describe variations in
the data. It is of limited use for significance testing,
which would require randomness and serial inde-
pendence in the within-station variations, that can be
attributed to variations in time, sampling error, and
measurement error. As demonstrated in the previous
section,"some of the within-station variations can be
attributed to seasonal factors and are therefore
nonrandom. Given three observations per station
spaced at roughly bimonthly intervals, serial de-
pendence in the observations is not likely to be strong,
since conditions are known to vary in many reservoirs
at a much higher frequency, as influenced, for
example, by storm events and algal bloom occurrences.
Among-station, within-reservoir variations also show
some serial dependence, since spatial trends in the
indices are often apparent when station means are
displayed in a downstream order (Walker, 1980b).
The relative magnitude of the last term is of special
significance to modelling efforts. With relatively large
within-station variance, it would be difficult to obtain
much accuracy in station summary statistics (e.g.,
station mean) with limited data. This would reduce the
explainable variance of any model or index system
calibrated to the reduced data set, make it more difficult
to distinguish among alternative model formulations,
and increase the error associated with model para-
meter estimates.
Variance components estimated for each index are
displayed on the left side of Figure 3. Variations in the
phosphorus and transparency indices are similar at all
levels. Variance components of the chlorophyll a index
at the district, reservoir, and station levels are
considerably lower than would be predicted based
upon corresponding phosphorus and transparency
variance components. The within-station components
account for a major portion (—60 percent) of the total
chlorophyll a variability. Thus, on the average for this
data set, temporal variations in the chlorophyll a index
at a given station appear to be stronger than variations
among stations, reservoirs, and/or districts. The
within-station variance components correspond to
standard deviations of 6.5, 6.5 and 7.9 for IB, IT, and Ip
respectively.
The covariance components on the right side of
Figure 3 provide insights into relationships among the
indices at different averaging levels. Spatial co-
variances are positive in all cases. Thus, the various
versions of the index correlate positively among
districts, reservoirs within districts, and stations within
reservoirs. Appreciable temporal covariance is ob-
served only for phosphorus and transparency. This
covariance might be attributed, for example, to turbidity
variations following seasonal or short-term (storm
event) flow variations. Despite its positive covariance
spatially, the chlorophyll a index does covary tem-
porally with the other indices.
VARIANCES
COVARIANCES
w
20
0
c/3
z 60
01
z
£ 40
o
f)
20
LU
O
Z Q
a:
$ 20
0
u
~- 0
UJ U
u
-y
5 60
t£
^ 40
20
n
AMONG DISTRICTS
: H H pq
AMONG
\
\
\
\
\
\
RESERVOIRS
\
\
\
\
\
F\l
H
.AMONG STATIONS
n ^ _
.WITHIN STATIONS
N
K
\~
\
\
\
^\
\
\
\
s,
\
\
\
\
R R ^
R
\
\ n
\ N N '
> N N '
Rl rq R
R „ '
IP/IT IP/IB
Figure 3. — Variance/covariance components of trophic state
indices.
REGRESSION ANALYSES
The covariance components indicate that the indices
can be averaged by station with some loss of
information about the phosphorus/transparency re-
lationship, but without losing chlorophyll a predict-
ability. Figure 4 depicts relationships among station-
mean values of the indices. Results of standard and
geometric-mean regression analyses relating the
indices are given in Table 1. Geometric mean
regressions summarize functional relationships and
are appropriate to use when both the independent and
dependent variables are subject to natural variability
and measurement error (Ricker, 1973). Standard
regressions are appropriate for predictive purposes.
The phosphorus and transparency indices explain 37
and 29 percent of the variance in the chlorophyll a
index, respectively. Recalibration of the index system to
these reservoirs requires significant reductions in the
corresponding slopes. In contrast, the phosphorus
index explains 78 percent of the transparency index
-------
MODELING AND ASSESSMENT OF THE TROPHIC STATE
347
100
x
LU
:c
D-
O
CC
O
80
60
40
20
20 40 6D 80
TRANSPARENCY INDEX
100
100
O.
o
ae.
o
80-
60
40
20
20 40 60 80
PHOSPHORUS INDEX
100
TOO
80
60
CL
CO
cc
40
20
20 40 60 80
PHOSPHORUS INDEX
100
Figure4.— Relationships among station-mean index values(a
= line of equality, b = geometric-mean regression, c = standard
regression).
variance and requires an adjustment in the intercept
only. Thus, compared with chlorophyll a, the phos-
phorus/transparency relationship appears to be more
typical of the natural lakes used by Carlson in deriving
his index system.
Tne effects of using alternative data reduction
procedures on the regression analyses have been also
studied. Using only summer mean values reduces the
regression slopes and R2 values and increases mean
squared residual errors by 58, 46, and 94 percent for
the IB/IP, IB/IT, and IT/P regressions, respectively.
These increases in error result partially from loss of
within-station replication when spring and fall values
are eliminated. In future work, data from other
monitoring programs with more intensive summer
sampling will be investigated. Use of reservoir means
has little influence on the results, but increases the
standard errors of parameter estimates.
ERROR ANALYSES
Residual errors from the regressions can be
attributed to three types of error: parameter, data, and
model. The first reflects uncertainty in the model
coefficients; the second, errors in the predicted and/or
predictor variables; and the third, influences of factors
which are not considered in the model structure. The
results of the variance component and regression
analyses can be used to derive approximate estimates
of the data errors according to the following equation:
Var(R)D
Var(lY)E + b2Var(lx)E - 2b Cov(lY, IX)E
N
eq. 5
where,
Var(R)o = data-error component of mean-squared
residual
Var (!Y)E = within-station variance of predicted index
Var(lx>E = within-station variance of predictor index
b — slope of regression equation
COV(|Y,!X)E = within-station covariance of predicted
and predictor indices
N = number of observations per station (averaging 3.1)
This formula is approximate because it assumes serial
independence in the within-station variations, which
would tend to be more important at sampling intervals
less than the 2-month intervals characteristic of this
data set.
The results of applying this equation to the
regression models in Table 1 are given in Table 2. They
indicate that roughly half (50 to 59 percent) of the
residual errors from the regressions can be attributed
to data errors. These components could be reduced
with a more intensive sampling program (i.e, more
replications per station). The influences of parameter
uncertai ity on the total residual error are expected to
be relatively insignificant, since the parameter error
component is inversely proportional to the number of
stations used in the regression analyses and the
parameters are relatively well-determined (Walker,
1977). Thus, most of the remaining error can be
attributed to the effects of factors which are not
considered in the index system.
-------
348
RESTORATION OF LAKES AND INLAND WATERS
Since data errors do not explain all of the residual
variance, it may be possible to improve the index
system by modifying it to take other important factors
into account. One modification is suggested by these
results and by the turbid nature of many reservoirs.
Chlorophyll a/phosphorus and chlorophyll a/trans-
parency relationships may not be constant across
reservoirs because of variations in non-algal panicu-
late materials (turbidity), which would influence
measurements of total phosphorus and transparency
but not of chlorophyll a. The relative stability of the
phosphorus/tranparency relationship across lakes and
reservoi/s may be attributed to the fact that both types
of measurements are sensitive to algal and non-algal
particulate materials. Other factors which might
contribute to model error include kinetic effects in
reservoirs with short hydraulic residence times. It
might also be possible to modify the system to account
for nitrogen limitation, by including N-limited as well as
P-limited reservoirs in the data set. Expansion of the
index system to include hypolimnetic oxygen deficits is
another possibility (Walker, 1979). These approaches
will be investigated in future studies of Carlson's index
system and other schemes using more extensive and
intensive data sets derived from the Corps reservoir
data base.
CONCLUSIONS
This paper has demonstrated an analytical approach
which provides insights into the adequacy of data for
modeling purposes. Potential applications of the
approach to monitoring program design are discussed
elsewhere (Walker, 1980b). Results suggest that
chlorophyll a is considerably less sensitive to phos-
phorus or transparency in these reservoirs, compared
with the natural lakes used by Carlson in developing
the index system. The phosphorus and transparency
indices are not relative indicators of biomass in these
Table 1. — Results of regression analyses relating station-
mean index values.
IB
IB
IT
IB
IB
IT
Slope
standard
Equation error
R2
standard regressions
= 24.7 + .443 IP .033 .374
= 25.2 + .403 IT .036 .291
= 11. 6 +.854 IP .026 .774
geometric mean regressions
= 9.9 + .724 IP .033 .224
= 5.7 + .747 IT .036 .079
= 5.4 + .971 IP .026 .760
Mean
squared
error
37.2
42.1
24.1
46.1
54.7
25.6
reservoirs, possibly because they are influenced by
non-algal materials. In future work the approaches
demonstrated here will be applied in evaluating
alternative schemes for summarizing relationships
among measures of trophic state in reservoirs, using
an expanded data set.
REFERENCES
Boyce, F. M. 1973. A computer routine for calculating total
lake volume contents of a dissolved substance from an
arbitrary distribution of concentration profiles. Tech. Bull.
83, Inland Waters Directorate, Canada Centre for Inland
Waters, Burlington, Ontario.
Carlson, R. E. 1977. A trophic state index for lakes. Limnol.
Oceanogr. 22:361.
Lambou, V. W., et al. 1976. Prediction of phytoplankton
productivity in lakes. In W. R. Ott, ed. Environmental
modeling and simulation. Off. Res. Dev. Off. Plann. Manage.
U.S. Environ. Prot. Agency.
Ricker, W. E. 1973. Linear regressions in fishery research.
Jour. Fish. Res. Board Can. 30:409.
Statistical Analysis Institute. 1979. SAS User's Guide.
Thornton, K. W., et al. 1980. Reservoir sedimentation and
water quality — a heuristic model. Surface water
impoundments. Am. Soc. Civil. Eng. and University of
Minnesota, Minneapolis.
U.S. Environmental Protection Agency. 1973-1976. National
Eutrophication Survey Working Pap. Ser., Pacific N.W.
Environ. Res. Lab., Corvallis, Ore.
1975. National Eutrophication Survey Methods
1973-76. Working Pap. No. 175. Pacific N.W. Environ. Res.
Lab., Corvallis, Ore.
Walker, W. W. 1977. Some analytical methods applied to lake
water quality problems. Ph.D. thesis. Harvard University.
1979. Use of hypolimnetic oxygen depletion rate
as a trophic state index for lakes. Water Resour. Res.
15:1463.
1980a. Empirical methods for predicting eu-
trophication problems in impoundments — phase I data
base development. U.S. Army Corps Eng. Waterways Exp.
Sta. Vicksburg, Miss. Draft final rep.
1980b. Analysis of water quality variations in
reservoirs: Implications for monitoring and modelling
efforts. Symp. Surface Water Impoundments. Am. Soc. Civil
Eng. and University of Minnestoa, Minneapolis.
Table 2. — Results of error analyses.
Relationship*
IB/IP
IB/IT
IT/IP
Mean
squared
error
37.2
42.1
24.1
Data
error
21.8
22.0
12.1
Percent
data
error
59%
52%
50%
Standard regressions in Table 1
-------
349
RESERVOIR WATER QUALITY SAMPLING DESIGN
KENT W. THORNTON
ROBERT H. KENNEDY
A. DALE MAGOUN
GARY E. SAUL
Waterways Experiment Station
U.S. Army Corps of Engineers
Vicksburg, Mississippi
ABSTRACT
The design of monitoring programs often serves as a major source of error or uncertainty in water
quality data bases. Properly designed programs should minimize uncertainty or at least provide a
means by which variability can be partitioned into recognizable components. While the design of
sampling programs has received recent attention, commonly employed strategies for limnological
sampling of lakes may not be completely appropriate for many reservoirs. Reservoirs differ from
natural lakes in that they are generally larger, deeper, and morphologically more complex.
Reservoirs also receive a majority of the inflow from a single tributary located at considerable
distance from the point of outflow. The result is the establishment of marked physical, biological,
and chemical gradients from headwater to dam. The existence of horizontal as well as vertical
gradients, and their importance in water quality sampling design were the subject of intensive
transect sampling efforts at DeGray Lake, a U.S. Army Corps of Engineers reservoir in southern
Arkansas. Data collected were used to partition variance, identify areas of similarity, and
demonstrate how an equitable sampling program might be designed.
INTRODUCTION
Recent legislation, including the Federal Water
Pollution Control Act (P.L 92-500) and the Amend-
ments of 1972 and 1977, requires water quality
monitoring programs to identify problems and assess
management procedures. As a result, Federal, State,
and local agencies spend millions of dollars annually
monitoring water quality in rivers, lakes, and reser-
voirs. Often overlooked in the final analysis, however,
is the error or uncertainty associated with these
estimates. This uncertainty may result from experi-
mental design, sampling variability, analytical error,
intrinsic variability, or all of these. In many instances,
the sampling program's design is the major source of
bias or error in the data. Designs are often
inappropriate because of ambiguous objectives, lack of
knowledge about the system, or manpower and
funding constraints. Monitoring programs must, there-
fore, receive careful review and consideration prior to
their implementation if meaningful information is to be
obtained.
Although sampling design and the problem of
uncertainty have recently received attention (Kwiat-
kowski, 1978; Liebetrau, 1979; Reckhow, 1979, 1980;
Reckhow and Chapra, 1979; Ward, et al. 1979), water
quality sampling design for reservoirs has not been
adequately addressed. This is due, in part, to the tacit
assumption that lakes and reservoirs are similar. The
purpose of this paper is to: (1) Generally describe
several differences between reservoirs and lakes that
influence sampling design; (2) discuss an intensive
water quality sampling program conducted at an U.S.
Army Corps of Engineers reservoir; and (3) describe
one approach for designing reservoir water quality
sampling programs.
LAKES AND RESERVOIRS
Although reservoirs are incorporated in the formal
definition of lakes (Hutchinson, 1957), several signifi-
cant differences between lakes and reservoirs suggest
that reservoirs are unique lentic systems (Ryder, 1978;
Thornton, et al. 1980). A comparison of 309 natural
lakes and 107 USAE reservoirs included in the 1972-
75 U.S. Environmental Protection Agency National
Eutrophication Survey indicated reservoirs had greater
drainage and surface areas, drainage/surface area
ratios, mean and maximum depths, shoreline develop-
ment ratios, and areal water loads than did natural
lakes (Table 1). Reservoirs also had shorter hydraulic
residence times and lower total phosphorus and
chlorophyll concentrations despite higher total phos-
phorus and nitrogen loadings.
In addition to these differences in scale, reservoirs
also exhibited pronounced longitudinal gradients, a
phenomenon not unexpected considering the impor-
tance of advective and unidirectional transport in
reservoirs (Baxter, 1977). Impoundment of meandering
rivers and their floodplains often creates long, narrow,
highly dendritic reservoirs that receive most of their
inflow from a single tributary located a considerable
distance from the outflow or dam. This promotes the
development of physical, chemical, and biological
gradients in space and time (Gloss, et al. 1980;
Hamblin and Carmack, 1978; Hebbert, et al. 1979;
Hyne, 1978; Johnson and Merritt, 1979; Kennedy, et
al. 1980; Kimmel and Lind, 1972; McCullough, 1978;
-------
350
RESTORATION OF LAKES AND INLAND WATERS
Table 1. —A comparison of geometric means on selected variables for natural lakes and USAE reservoirs (Thornton, etal. 1980).)
Variable
Drainage Area (Km2)
Surface Area (Km2)
Drainage/Surface Area
Mean Depth (m)
Maximum Depth (m)
Shoreline Development Ratio
Areal Water Load (m/yr)
Hydraulic Residence Time (yr)
Total Phosphorus (fjg/\)
Chlorophyll a (/ug/l)
P Loading (g/m2-yr)
N Loading (g/m2-yr)
*Hutchinson, 1957
"Leidy and Jenkins, 1979
Natural Lakes
(N = 309)
222.0
5.6
33.0
4.5
10.7
2.9 (N = 34)+
6.5
0.74
54.0
14.0
0.87
18.0
USAE Reservoirs
(N = 107)
3228.0
34.5
93.0
6.9
19.8
9.0 (N = 179)++
19.0
0.37
39.0
8.9
1.7
28.0
Probability
Means are
Equal
<0.0001
<0.0001
<0.0001
<0.0001
<0.0001
<0.001
<0.0001
<0.0001
0.02
<0.0001
<0.0001
<0.0001
Thornton, et al. 1980). These gradients should be
considered in designing reservoir water quality
sampling programs.
DEGRAY LAKE
Description and characterization of lateral, longi-
tudinal, and vertical water quality gradients were the
objectives of intensive water quality transect samplings
conducted on DeGray Lake, a USAE reservoir located in
southern Arkansas. DeGray Lake has a 53.4 km2
surface area, a mean and maximum depth of 9 and 60
m, respectively, and is located in a large (1,162 km2)
predominately forested watershed. It is highly dendritic
(shoreline development ratio of 1 3) and exhibits strong
thermal stratification. DeGray Lake has an average
hydraulic residence time of 1.2 years and is operated
for hydropower production. The outlet structure has the
capability for selective withdrawal and can discharge
epilimnetic, metalimnetic, or hypolimnetic water to
meet downstream requirements.
METHODS
Sampling transects were established from the dam
to the headwaters. Stations were located on each
transect to sample over the old river channel, in the
littoral area on each shore, and at intermediate
distances between these locations. Fifteen transects,
averaging five stations per transect, were established
in the main body of DeGray Lake (Figure 1). Three to
four transects were also established on the two major
embayments. Water samples, pumped to the surface
from depths of 0, 2, 4, 6, and 10 m and at 5-m intervals
thereafter to within 0.5 m of the bottom, were stored in
acid-washed polyethylene bottles. Sampling occurred
during July 1978, and January and October 1979. All
samples were collected on the same day during each
sampling trip. The lake was thermally stratified during
the summer and isothermal during the winter. During
stratified periods, the lake was divided into epilimnion,
metalimnion, and hypolimnion by defining the meta-
limnion as those depths at which temperature changed
by 1°C or more per meter of depth. The lake was
completely mixed during the January sampling.
Laboratory determinations included total phosphor-
us, chlorophyll a, and turbidity analyses since these
parameters generally characterize the distribution of a
representative nutrient, phytoplankton biomass, and
physical factors affecting light regime, respectively.
Total phosphorus samples were analyzed by persulfate
digestion/ascorbic acid reduction according to stan-
dard methods (Am. Pub. Health Assoc. 1 976). Turbidity
analyses were conducted on a Hach turbidimeter.
Chlorophyll samples, taken only at 0 and 4 m, were
stored in the dark at 4°C in polyethylene bottles, and
filtered within 8 hours. The trichromatic method, with
pheophytin correction, was used for chlorophyll
analyses (Am. Pub. Health Assoc. 1976). Chlorophyll
samples that could not be analyzed immediately were
frozen after filtering and analyzed within 2 weeks.
Data analyses were performed using the Statistical
Analysis System (SAS 79.4, 1979). A variance com-
ponent analysis with random group effects was used to
test for differences with depth, sampling 'station, and
transect. Since samples were not replicated, sampling
depth was nested in the analysis of variance and
included sampling and analytical error as well as the
variability among" sampling depths.
Figure 1. — Map of DeGray Lake, Arkansas, and locations of
sampling transects (bold lines).
RESULTS
DeGray Lake exhibited marked longitudinal and
vertical variation during all sample trips as evidenced
by the percentage of the total variance contributed by
-------
MODELING AND ASSESSMENT OF THE TROPHIC STATE
351
Table 2. — Percent of total variance.
Epilimnion
Variable
Turbidity
Total P
Chi a
Turbidity
Total P
Chi a
Turbidity
Total P
Chi a
Tran
76
55
88
73
18
83
97
60
15
Sta
0
0
0
Depth
Metalimnion
Tran Sta
24 69 0
45 56 0
12
OCTOBER
3 24
8 74
0 17
JANUARY
0
2
26
3
38
60
40
61
(no
Hypolimnion
Depth Tran
31
44
28 32
4 35
stratification)
10
33
48
98
Sta
9
13
2
0
Depth
81
54
51
2
. 1 I • • B g . I g 8 i .
I I I I I I I I I I I I I
0.06|—
0,02
0,01
lllll
II I I I I
I I I I
I '
£ 0.02
000 - I I I I I I I I I I I
IE 14 13 12
S 4 3 2 1
II'. ,
• I
1 1 1 1 1 1 1 1
16 14 13 12 11
II
i.
15 14 13 12 11 10 9 B 7
TRANSECT
I |
54321
i j ! I ' I
I I I
64321
Figure 2. — Changes in mean (solid circle) and median (square)
chlorophyll a concentration (ug/l)b with transect in January
1978 (top), July 1979 (middle), and October 1979 (bottom).
Vertical lines indicate 1 S.D.
0.06
0.04
0.03
0.02
0.01
0.00
1
-'
-
-
1
-
-
I
1 I ll
* » I * * , I •
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
6 14 12 10 B 6 4 2 0
TRANSECT
Figure3. — Changes in mean(solidcircle)andmedian(square)
total phosphorous concentration (ug/l)b with transect in
January 1978 (top), July 1979 (middle), and October 1979
(bottom). Vertical lines indicate 1 S.D.
transect and depth, respectively (Table 2). Since
variance associated with stations within transects was
minimal, subsequent comparative analyses were
performed using volume-weighted transect means for
epilimnion, metalimnion, and hypolimnion.
A comparison of these transect means indicated that
longitudinal variation could be attributed to the
existence of gradients from headwater to dam. In
general, turbidity and total phosphorus concentrations
in the epilimnion, metalimnion, and hypolimnion
-------
352
RESTORATION OF LAKES AND INLAND WATERS
I I I I
J I L
l I I I I
rll
• i
0.0 i
-I
I I I I
I I I I I
Figure 4.— Changes in mean (solid circle) and median (square)
turbidity (NTU's) with transect in January 1978 top), July 1979
(middle), and October 1979 (bottom) Vertical lines indicate 1
S.D.
decreased with distance downstream during all
sampling trips. Epilimnetic chlorophyll a concentra-
tions, while decreasing in a downstream direction in
July and October, were lower and similar for all
transects during the January sampling (Figure 2). Of
limnological significance is the fact that differences
among transects were most pronounced in the upper
region of the reservoir nearest the tributary inflow.
Differences among transects from mid-reservoir to
dam, when they existed, were minimal. For example.
mean epilimnetic total phosphorus concentrations for
headwater transects ranged from 0.03 to 0.04 mgP/l
during all sampling months, decreased to approxi-
mately 0.01 mgP/l by transect 6,11, and 9 in January,
July, and October, respectively, and then remained
unchanged between these mid-reservoir locations and
the dam (Figure 3). Mean epilimnetic turbidity values,
which were lowest during the July sampling, exhibited
longitudinal changes similar to those for total
phosphorus (Figure 4). The variance about transect
means was, in general, greatest at upstream transects
on all sampling dates and lowest at transects near the
dam.
These longitudinal gradients present a unique
problem in designing a water quality sampling
program. Since it is not possible to sample at a single
station and adequately characterize reservoir water
quality, the number and location of required sampling
stations must be determined. One approach for
locating multiple stations will be illustrated using
epilimnetic total phosphorus, turbidity, and chlorophyll
a data from DeGray Lake.
Analyses of variance and Duncan's multiple range
test indicated no significant differences among certain
transects as well as significant differences among
others. Various linear models were used to describe
the change in epilimnetic mean total phosphorus and
chlorophyll a concentrations, and turbidity longitu-
dinally down the reservoir. These linear models
provided an initial estimate of the minimum number of
stations required to characterize these areas. If, for
example, ANOVA procedures are to be used to
characterize areas, and the slope (b) of a model among
transect means is not significantly different from zero,
then a minimum of one station would be required to
characterize this area. If a linear function (b ^ 0)
accounts for a significant portion of the variance
among other transect means, a minimum of two
stations might characterize this area; a quadratic
function would require three stations; a cubic function,
four stations, etc. It should be noted that the suggested
numbers of stations are minimums, since increasing
the number of stations would provide greater reliability
and statistical confidence in all estimates, particularly
those described by statistical models. These linear
models and appropriate transects can then be
compared for all water quality variables of interest and ^
all sampling dates to identify areas of similarity and
overlap.
For DeGray Lake, transects 1 through 5 exhibited
similar means for all variables over all dates, so a single
station could be selected to characterize this area. The
relation among transects 6 through 13 (i.e., mid-
reservoir) was generally linear, thus requiring a
minimum of two sampling stations to characterize
water quality gradients in this area. Since transects 14
and 15 were either distinct or linear, a separate station
could be established for each transect. A minimum of
five sampling stations, then, would be required to
characterize longitudinal water quality gradients in
DeGray Lake. These might logically be located on
transects 3, 10, 12, 14, and 15. Since there was no
significant lateral variability, the stations could be
located over the deepest point on each transect.
-------
MODELING AND ASSESSMENT OF THE TROPHIC STATE
353
The number of samples to be collected at each
station depends on the variability of the water quality
constituent and the desired precision of the estimate.
Estimates of variability can be obtained by reviewing
existing data or, as in the case of DeGray Lake, by
preliminary surveys of the system to be sampled.
Precision will be dictated by analytical capabilities
and/or the purposes for which the data will be used. A
general formula for random sampling can be used to
obtain initial estimates for sample size (Cochran,
1963). A Student's t value for 30 degrees of freedom
can be used to initiate the procedure. The formula is
applied iteratively until n converges on the sample size.
n =
where n = number of samples
t = appropriate value from Student's
t distribution
s2 = sample variance
d = desired precision about the mean
Assuming a fixed cost for a sampling trip, the cost of
sampling several variables can be estimated by
C(n)=Co
C,a
1=1
where C0 = fixed cost of sampling
C, =unit sample cost for variable i
n, = number of samples for variable!
Frequently, the total computed sampling cost will
exceed the funds available and the sampling effort
must be reduced. Since precision is incorporated in the
sampling formula, a matrix can be developed to
indicate concomitant reductions in cost and precision
of various water quality constituents (Table 3). For
example, we may wish to estimate means for total
phosphorus at the five suggested sampling stations in
DeGray Lake within 5 /ugP/l with 95 percent confi-
dence. Since means for these stations differ, percent
precision will vary from 50 percent at transect 3 (mean
of 9 /ugP/l) to 15 percent at transect 14 and 15 (mean
of 34 jugP/l). Finding the appropriate entry in Table 3
indicates that 12 samples will be required at transect
(or station) 3, 18 samples each at transects 10 and 12,
and 23 samples each at transects 14 and 15. The total
number of samples for all stations would thus be 94.
Assuming a unit analytical cost of $13, the total
analytical cost per sample trip would be $1,222. If we
would also like to estimate means for chlorophyll
within 25, 25, and 15 percent, and turbidity within 50,
20, and 20 percent at downstream, mid-reservoir, and
headwater stations, respectively, with 95 percent
confidence, analytical cost could then be obtained by
summing the cost for all three variables. In this
example, the total analytical cost per sampling trip
would be $3,085. Assuming a fixed sample collection
cost of $412 per sample trip, the total cost of this three-
variable sample program would be $3,497 per sample
trip.
If it is determined that funds would be insufficient to
support this sampling effort, then some decision
concerning the quality of the data would have to be
made. For instance, if the objectives of the study
required precise information for total phosphorus (e.g.,
± 5/ugP/l but less precise information for chlorophyll
and turbidity (i.e.,>25 , 25, and 15 percent and>50,
20, and 20 percent, respectively), then the number of
samples for chlorophyll and turbidity could be reduced.
Table 3. — Decision matrix for DeGray Lake epilimnetic sampling.
Transect 3 Transect 10 or 12 Transect 14 or 15
Total
Phosphorus
Turbidity
Chlorophyll
Unit Cost
Mean
Precision
Probability
Sample No.
Total Cost
Unit Cost
Mean
Precision
Probability
Sample No.
Total Cost
Unit Cost
Mean
Precision
Probability
Sample No.
Total Cost
$13 $13 $13
SfjgP/\ 20/jgP/l 34 /jgP/l
±50% ±100% ±25% ±50% ±15% ±30%
95% 90% 80% 95% 90% 80% 95% 90% 80% 95% 90% 80% 95% 90% 80% 95% 90% 80%
12 9544 3 18 13 865 3 32 23 14 10 7 5
156 117 65 52 52 39 234 169 104 78 65 39 416 299 182 130 91 65
$3 $3 $3
1.3NTU's 5.3NTU's 5.3 NTU's
±50% ±100% ±20% ±40% ±20% ±40%
95% 90% 80% 95% 90% 80% 95% 90% 80% 95% 90% 80% 95% 90% 80% 95% 90% 80%
13 10 6543864421 864421
39 39 18 15 12 9 24 18 12 12 6 3 24 18 12 12 6 3
$20 $20 $20
2 //g/l 7 Aig/l 1 1 /ug/l
±25% ±50% ±25% ±50% ±15% ±30%
95% 90% 80% 95% 90% 80% 95% 90% 80% 95% 90% 80% 95% 90% 80% 95% 90% 80%
19 13 9 6 5 3 13 10 6 5 4 3 21 15 9 7 5 4
380 260 180 120 100 60 260 200 120 100 80 60 420 300 180 140 100 80
NOTE: 1. Mean values are those expected based on three sampling dates.
2. Percent precision based on approximate levels of analytical precision for each test or requirements of the study.
3. Total cost calculated as product of unit cost and sample number.
-------
354
RESTORATION OF LAKES AND INLAND WATERS
This, in turn, would reduce analytical costs. For
example, decreasing the percent precision for chloro-
phyll and turbidity by a factor of two while continuing to
estimate total phosphorus within ±5/ugP/l would cost
$1,885. Assuming 1 2 sample trips per year, this would
save $14,000 per year.
Construction of a similar matrix for all water quality
variables, including information for the metalimnion
and hypolimnion, would allow design of a total
sampling program. This matrix permits balancing
precision, confidence, and cost within monetary
constraints. While loss of precision may be scientifi-
cally undesirable, at least the uncertainty associated
with the data could be accounted for in comparing and
discussing the data.
Finally, sampling dates, which will depend in part on
the objectives of the sampling program and site-
specific characteristics, must be identified. Since
reservoirs are strongly influenced by advective trans-
port of nutrients and particulate material, sampling
during elevated flows is necessary. Sampling should
also occur during periods of stratification and during
low flow periods. One potential sampling strategy for
DeGray Lake might be: Once during January; three
times during elevated flows from March to mid-April;
once during May; three times during stratification from
mid-June through July; once in August; three times
during low flow from mid-September to mid-October;
and once in mid-November. The total number of
samples approximates a monthly sampling effort but
the sampling program now incorporates hydrologic and
limnological factors responsible for many of the
observed water quality gradients. Obviously, the exact
dates must reflect the objectives of the sampling
program, potential problems, and specific hydrologic
and site-specific characteristics. In temperate regions,
spring runoff is a critical period that needs to be
incorporated in the reservoir sampling program. Winter
kill in some reservoirs may require more frequent
winter sampling.
DISCUSSION
Minimizing uncertainty or partitioning variability into
recognizable components is the purpose of experi-
mental design. To be effective, sample design should
allow the researcher to adequately discuss the
characteristics of the system as well as permit
comparative evaluations through time and/or across
systems. The presence of non-partitioned variability in
the data reduces its informational content and thus its
value to the researcher or manager.
Proper design of a sampling program will depend in
part on the objectives of the study. An objective
common to many limnological surveys is characteriza-
tion of the water quality of an entire body of water.
However, it is frequently assumed that horizontal
heterogeneties are insignificant relative to those
occurring vertically, and thus a single, deep station is
often established. In lakes such as DeGray, which
exhibit marked longitudinal gradients, such an ap-
proach to sample design would be inappropriate. For
instance, using Vollenweider's (1968) criteria to
classify lakes with respect to phosphorus and
chlorophyll concentrations, DeGray Lake could be
classified as either oligotrophic, mesotrophic, or
eutrophic depending on the location of the single
sampling station. On all sampling dates, the headwater
area would be classified as eutrophic, mid-reservoir as
mesotrophic, and the lower reservoir as oligotrophic.
DeGray Lake is not unique in this regard, as similar
classification problems have arisen in other reservoirs
(Hannan, et al. 1980).
The approach to sample design outlined here for
DeGray Lake provides a means for obtaining informa-
tion which can improve characterization of the entire
lake as well as the existence of gradients which may be
of limnological or management significance. Charac-
terization of the entire lake, if it must be attempted,
could be more realistically accomplished by volume-
weighting observations from representative portions of
the lake, while station-by-station evaluation of the
same information would assess possible problem
areas.
An additional consideration in the design of a
sampling program is to balance desired precision
against the reality of monetary limitations. A review of
historical data, educated guesses, or, as in the case of
DeGray Lake, data from a pilot study will be required to
estimate initial means and associated variability, and
thus, initial estimates of sample size (Cochran, 1963).
For long-term monitoring programs, the expenses of a
pilot study may be more than offset by eliminating
samples which do not significantly increase the
informational content of the data base. Such studies
would also identify the existence of gradients,
information which may be important in meeting study
objectives. Identification of the general location of
these gradients prior to beginning a monitoring
program would permit logical positioning of sampling
stations.
CONCLUSIONS
Reservoirs are advectively dominated systems that
often exhibit pronounced gradients in water quality.
Evaluation of water quality conditions in reservoirs will,
therefore, require sampling programs designed to
adequately assess changes in both space and time.
Presented here is one approach for designing such a
program, using DeGray Lake as an example. Similar
studies currently being conducted at other reservoirs
will allow further evaluation of this approach and
hopefully provide a basis for generalizing sampling
design methodologies for use in many reservoirs.
REFERENCES
American Public Health Association. 1976. Standard meth-
ods for the examination of water and wastewater. 14th ed.
Washington, D.C.
Baxter, R. M. 1977. Environmental effects of dams and
impoundments. Ann. Rev. Ecol. Systemat. 8:255.
Cochran, W. G. 1963. Sampling techniques. 2nd ed. John
Wiley and Sons, Inc., New York.
Gloss, S. P., L. M. Mayer, and D. E. Kidd. 1980. Advective
control of nutrient dynamics in the epilimnion of a large
reservoir. Limnol. Oceanogr. 25:219.
Hamblm, P F., and E. C. Carmack. 1978. River-induced
currents in a fjord lake. Jour. Geophys. Res. 83:885.
-------
MODELING AND ASSESSMENT OF THE TROPHIC STATE
355
Hannan, H. H,, D. Barrows, and D. C. Whitenburg. 1980. The
trophic status of a deep-storage reservoir. Proc. Symp.
Surface-Water Impoundments. Am. Soc. Civil Eng. Minne-
apolis, Minn. June. (In press.)
Hebbert, B., et al. 1979. Collie River underflow into the
Wellington Reservoir. Am. Soc. Civil Eng. Jour. Hydraul
Eng. Div. 105:533.
Hutchinson, G. E. 1957. A treatise on limnology. 1 st ed. John
Wiley and Sons, Inc., New York.
Hyne, N. J. 1978. The distribution and source of organic
matter in reservoir sediments. Environ. Geol. 2:279.
Johnson, N. M,, and D. H. Merritt. 1979. Convective and
advective circulation of Lake Powell, Utah-Arizona, during
1972-1975. Water Resour. Res. 15:873.
Kennedy, R. H., K. W. Thornton, and J. H. Carroll. 1980.
Suspended-sediment gradients in Lake Red Rock. Proc.
Symp. Surface-water Impoundments. Am. Soc. Civil Eng.
Minneapolis, Minn. June. (In press.)
Kimmel, B. L., and 0. T. Lind. 1972. Factors affecting
phytoplankton production in a eutrophic reservoir. Arch
Hydrobiol. 71:124.
Kwiatkowski, R. E. 1978. Scenario for an ongoing chlorophyll
a surveillance plan on Lake Ontario for non-intensive
sampling years. Jour. Great Lakes Res. 4:19.
Liebetrau. A. M. 1979. Water quality sampling: some
statistical considerations. Water Resour. Res. 15:1717.
McCullough, J. D. 1978. A study of phytoplankton primary
productivity and nutrient concentrations in Livingston
Reservoir. Texas Jour. Sci. 30:377.
Reckhow, K. H. 1979. The use of a simple model and
uncertainty analysis in lake analysis. Water Resour. Bull.
15:601.
1980. Techniques for exploring and presenting
data applied to lake phosphorus concentration. Can. Jour.
Fish. Aquat. Sci. 37:290.
Reckhow, K. H., and S. C. Chapra. 1979. A note on error
analysis for a phosphorus retention model. Water Resour.
Res. 15:1643.
Ryder, R. A. 1978. Ecological heterogeneity between north-
temperate reservoirs and glacial lake systems due to
differing succession rates and cultural uses. Ver. Int. Ver.
Limnol. 20:1568.
Thornton, K. W., et al. 1980. Reservoir sedimentation and
water quality — a heuristic model. Proc. Symp. Surface-
water Impoundments. Am. Soc. Civil Eng., Minneapolis,
Minn. June. (In press.)
Vollenweider, R. A. 1968. Scientific fundamentals of the
eutrophication of lakes and flowing waters, with particular
reference to phosphorus and nitrogen as factors in
eutrophication. Tech. Rep. DAS/CSI/68.27. Organ. Econ.
Coop. Dev.
Ward, R. C., et al. 1979. Statistical evaluation of sampling
frequencies in monitoring networks. Jour. Water Pollut.
Control Fed. 51.2292.
-------
356
HEALTH ASPECTS OF EUTROPHICATION
MICHAEL J. SUESS
ROBERT B. DEAN
Regional Office for Europe
World Health Organization
Copenhagen, Denmark
ABSTRACT
Increasing eutrophication makes water more difficult and expensive to treat. Soluble organic
matter derived from algae contributes to taste and color and produces chloroform and other
trihalomethanes when the water is chlorinated. These compounds are suspected of being weakly
carcinogenic for humans and are indicators of the formation of other chlorinated compounds that
have been less well studied. The adverse effects directly related to eutrophication are, in general,
diseases transmitted by vectors, such as mosquitos and snails which spend at least part of their
life in water. Different mosquito species have different water requirements. Yet, many water-
related health problems exist independently of eutrophication. Waterborne bacteria and viruses,
as well as protozoa and other parasites that do not have an intermediate host, are simply carried by
water from an infected person to susceptible persons and eutrophication has no direct influence
on the passage. It is not possible to state that increased eutrophication will affect the incidence of
mosquito-borne diseases. Snails are quite adaptable and reducing eutrophication is not likely to be
an effective method for controlling schistosomiasis. Direct health effects of algae produced in
eutrophied lakes include objectionable taste and odor, dermatitis, asthma, other allergic
responses, and low grade chronic toxicity. Few cases of human poisoning by algae have been
reported, probably because the associated taste and odor causes people to seek other sources of
water whenever possible.
INTRODUCTION
The voluminous literature on eutrophication contains
remarkably few references to health effects. To find
such information one must search under "artificial or
manmade lakes," "sewage lagoons," or "oxidation
ponds.' One of the most comprehensive treatises on
the subject is the book "Man-made Lakes and Human
Health," a collection of 31 papers (Stanley and Alpers,
1975).
Most manmade lakes pass through an initial first-fill
eutrophic stage produced by nutrients leached from the
soil or decaying vegetation. If there are no major
nutrient inputs to the lake, this stage may be transitory
as nutrients are flushed out or buried in sediment.
Sewage lagoons are heavily loaded with nutrients
which produce a high state of eutrophication at the
"clean" end of a series.
Many water-related health problems exist indepen-
dently of the state of eutrophication. Waterborne
bacteria and viruses, as well as protozoa and other
parasites that do not have an intermediate host, are
simply carried by water from an infected person to
susceptible persons and eutrophication has no direct
influence on the passage.
While a high degree of eutrophication is universally
bad, a moderate degree may have some beneficial
effects. Increasing eutrophication makes water treat-
ment more difficult and more expensive. An example of
this problem, which is classical in its development, is
the Sou Regreg water supply reservoir in the Atlas
mountains above Rabat in Morocco. The treatment
plant has been found inadequate because of the
serious taste and odor problem caused by algae.
Treatment with large quantities of activated carbon
now appears to be necessary to produce a water that
will have an acceptable taste (WHO). If the water is not
acceptable, there will be a serious health risk because
consumers will turn to other, possibly unsafe drinking
water sources (Eng. Sci. Med. 1979).
The adverse effects directly related to eutrophication
are, however, diseases transmitted by vectors, such as
mosquitos and snails which spend at least part of their
life cycle in water. Of the very large number of species,
only a few can carry disease organisms that infect man,
while others are a general nuisance, their bites causing
allergic reactions and sometimes secondary infection.
MOSQUITOS
In the case of malaria, early on it was discovered that
it was important to concentrate on the species that
carried the disease in the specific area in question
(Waddy, 1975; Surtees, 1975). Different mosquito
species have different water requirements: Some
hatch from clear water in sunlight, others hatch from
-------
HEALTH-RELATED PROBLEMS
357
shaded pools of water held by plant leaves, and still
others flourish in mats of algae on eutrophied lakes.
Therefore, it is not possible to state that increased
eutrophication will increase or decrease the incidence
of mosquito-borne diseases. In many cases, the
mosquitos hatch from small pools of water adjacent to
the main body of the lake and only few, if any, can
survive in open water where they are subject to
predation by fish and disturbance by waves.
Underwater rooted vegetation and heavy mats of
algae or floating water plants increase the number of
sites where mosquitos can hatch and, therefore, may
increase the incidence of arthropod-carried disease,
including malaria, dengue fever, disease due to
arboviruses and filariasis (Waddy, 1975).
One objection to sewage treatment lagoons, which
certainly are highly eutrophied, is that they provide
breeding places for Culex mosquitos, the primary
vectors of several types of encephalitis virus. In the
midwestern and southwestern United States, improp-
erly maintained lagoons were found to be a major
source of these mosquitos (Hopkins, 1960). Control
was achieved by removing rooted vegetation and algae
mats. Fish that eat mosquito larvae are frequently
cultivated in lakes, ponds, and rice paddies (Jackson,
1975). These mosquito-fish are too small to contribute
very much to human nutrition, so their only effect is to
control the mosquito population. High levels of
eutrophication produce shelter for mosquito larvae
where fish cannot reach them. Many mosquitos do not
require a high level of nutrients and the Arctic, which is
famous for numerous clear oligotrophic lakes, still
swarms with myriads of mosquitos and small flies
during the summer.
Nutrients from sewage discharge in the Baltic Sea
have been blamed for increased growth of the
Phragmites reed in shallow water near coastal cities.
These dense growths probably increase the number of
mosquitos which, in this climate, are not serious
vectors of disease but certainly interfere with man's
well-being.
SNAILS
In the tropics, schistosomiasis (bilharziasis), a
debilitating but rarely fatal disease caused by a
trematode worm, seems invariably to accompany the
construction of lakes and irrigation schemes (Jordan,
1975 ; Burch, 1975). The schistosome spends half of its
life cycle in a freshwater snail. Infected humans
excrete worm eggs, larva hatch from these eggs and
can develop in certain species of snails if excreta are
discharged into standing or slow moving water. The
developed infectious larvae leave the snail after a few
days and enter the skin of humans who may be wading
in the water. Theoretically, this chain of infection can
be broken by good sanitation, keeping excreta
(including urine) out of surface water, controlling
snails, chemotherapy of infected persons, and keeping
humans, especially children, out of the water. The
snails that harbor the schistosomes require a tropical
climate and this is, unfortunately, the very climate that
encourages children to play, bare-legged, in the water.
The host snails thrive in shallow waters that contain
dissolved organic matter and are mildly eutrophic. They
grow best when the water has moderate light
penetration and is not turbid. An abundance of
submerged aquatic plants provides shelter for egg
laying. However, the host snails are quite adaptable
and reducing eutrophication is not likely to effectively
control schistosomiasis.
ALGAE
Direct effects of algae produced in eutrophied lakes
include objectionable taste and odor. Dermatitis,
asthma, and other allergic responses, and low grade
chronic toxicity have also been reported in a few cases
(MacGregor and Keeney, 1975). Diarrhea has been
demonstrated in experimental animals, and livestock
have been killed by drinking water that was heavily
infected with blue-green algae (cyanophytes). In the
latter case, the livestock had no choice but to drink the
algae with the water. However, it is uncertain whether
the toxic material was in solution or in the algae cells.
On the other hand, only a few cases of human
poisoning by algae have been reported, probably
because the associated taste and odor have caused
people to seek other sources of water whenever
possible (Kay, Sykora, and Burgess, 1980). Eutrophica-
tion in salt water may be manifested as a red tide of
dinoflagellates. These small organisms are filtered out
by shellfish, rendering them toxic. The prevalence of
dinoflagellates in the warmer summer months is
probably the origin of the old stricture against eating
shellfish in any month without an R in it (May-August).
Soluble organic matter derived from algae con-
tributes taste, odor, and color to water but also
produces chloroform and other trihalomethanes when
the water is chlorinated. These compounds are
suspected of being weak carcinogens for humans and
are indicators of the formation of other chlorinated
compounds that have been less well studied (WHO,
1978).
The secondary effects of high levels of eutrophica-
tion, though not directly affecting health, nevertheless
detract from man's well-being. A high concentration of
algae in bloom causes extreme shifts in dissolved
oxygen and pH, and may kill fish and other, less
tolerant, species of algae. Even the algal species
responsible for the bloom will eventually die of
overcrowding. Bacterial decay of the dead plant and
animal matter removes dissolved oxygen and then
generates hydrogen sulfide which further contributes
to objectionable odors and foul black deposits of ferrous
sulfide. Aquatic weeds can completely clog lakes and
waterways that receive excess nutrients, rendering
them unfit for navigation, fishing, or even as a water
supply (Ferguson, 1968).
OTHER EFFECTS
Eutrophication may have an indirect and possibly
beneficial effect on the transmission of waterborne
diseases. There is evidence that algae and plants are
antagonistic to enteric bacteria so that the effluent of a
well-run set of oxidation ponds has a lower count of
indicator organisms than the effluent from a trickling
filter or activated sludge plant (Hopkins, 1960). A less
direct effect of algae in a water supply is to increase the
-------
358
RESTORATION OF LAKES AND INLAND WATERS
probability that the water will be treated prior to
consumption. Treatment that reduces turbidity also
removes many infective agents. Man is all too willing to
drink polluted water if it looks and tastes fairly good and
eutrophication may serve as a warning that treatment
should be applied.
The construction of large artificial lakes displaces the
local population, causing interruptions in food supplies,
dependence on government hand-outs, and psycho-
logical disturbances. If the lake supports a suitable
population of fish, they can make a very real
contribution to the nutrition and well being of the
displaced population (Hopkins, 1960). First-fill eutroph-
ication usually provides excellent fishing in the first
few years of a reservoir's life. If no nutrients are added,
the fish catch will eventually decline as the original
nutrients are depleted. Good fishing requires a
controlled state of eutrophication. Even clean-water
fish, such as salmon, may be more productive if
controlled amounts of fertilizer are added to the water
(Le Brasseur, McAllister, and Parsons, 1979).
WHO ACTIVITIES
A number of WHO activities have been concerned,
directly or indirectly, with the health aspects of
eutrophication (Deom, 1976; Landner, 1976). Malaria
and other parasitic diseases have a high priority and
are handled through specific programs by both WHO
headquarters and the regional offices. Moreover, the
World Health Organization cooperates in the inter-
national effort to provide safe drinking water and
sanitation for all by the end of this decade. To achieve
this goal, pollution from sewage and agriculture must
be reduced; one of the results will be a decrease in the
eutrophication of lakes and slow moving rivers.
CONCLUSION
The effects of eutrophication on man's well-being
are not very extensive and they may be only partially
negative at low levels of eutrophication. A high level of
eutrophication is generally bad, but even here the
effects on disease are only indirect, and there seems to
be no justification at present for special programs
dealing with the effects of eutrophication on disease.
Present programs to control excessive eutrophication
will, to the extent that they are successful, have a
favorable effect on human disease as well as make a
positive contribution to the health of the total
environment.
Deom, J. 1976. Water resources development and health, a
selected bibliography. Document MPD/76.6, and Adden-
dum I, 1977 Document MPD/77.7. World Health Organi.,
Geneva.
Ferguson, F. F. 1968. Aquatic weeds and man's well being.
Hyacinth Control Jour. 7:7.
Hopkins, G. 1960. Waste stabilization lagoons. U.S. Publ.
Health Serv. Region VI, Kansas City. (Review by: D. M.
Pierce. Water Sewage Works 408-411, October 1960).
Jackson, P. B. N. 1975. Fish. Pages 259-275 in N. F. Stanley
and M. P. Alpers, ed. Man-made lakes and human health.
Academic Press, London.
Jordan, P. 1975. Schistosomiasis-epidemiology, clinical
manifestations and control. Pages 35-50 in N. F. Stanley
and M. P. Alpers, ed. Man-made lakes and human health.
Academic Press, London.
Kay, G. P., J. L. Sykora, and R. A. Burgess. 1980. Algal
concentration as a quality parameter of finished drinking
waters in and around Pittsburgh, Pa. Jour. Am. Water
Works Assoc. 72:170.
Landner, L. 1976. Eutrophication of lakes. Document
ICP/CEP 210. World Health Organ. Regional Off. Europe,
Copenhagen.
Le Brasseur, R. J., C. D. McAllister, andT. R. Parsons. 1979.
Addition of nutrients to a lake leads to greatly increased
catch of salmon. Environ. Conserv. 6:187.
MacGregor, A. N. and D. R. Keeney. 1 975. Nutrient reactions.
Pages 237-257 in N. F. Stanley and M. P. Alpers, ed. Man-
made lakes and human health. Academic Press, London.
Obeng, L. E., ed., 1969. Man-made lakes: The Accra
Symposium. Ghana Universities Press, Accra. (Sole distrib-
utor outside Ghana: Oxford University Press, London.)
Stanley, N. F. and M. P. Alpers, ed. 1975. Man-made lakes
and human health. Academic Press, London.
Surtees, G. 1975. Mosquitoes, arboviruses and vertibratis.
Pages 21-34 in N. F. Stanley and M. P. Alpers, ed. Man-
made lakes and human health. Academic Press, London.
Waddy, B. B. 1975. Mosquitoes, malaria and man. Pages 7-20
in N. F. Stanley and M. P. Alpers, ed. Man-made lakes and
human health. Academic Press, London.
WHO Eastern Mediterranean Regional Office. 1978. The
prevention and control of vector-borne diseases in water
resources development projects. Document VBC/EM/78.1.
WHO Regional Office for Europe. 1978. Treatment agents and
processes for drinking-water and their effects on health.
Document ICP/CEP 101(6).
Various consultant reports. Moroccan Project
MOR/RCE 001.
REFERENCES
Ackermann, W. C., G. F. White, and E. B. Worthington. 1973.
Man-made lakes, their problems and environmental effects.
Am. Geophys. Union, Washington, D.C.
Anonymous. 1979. Engineering science and medicine in the
prevention of tropical water related disease. Prog. Water
Technol. 11:1.
Burch, J.B.I 975. Freshwater molluscs. Pages 311 -321 in N.
F. Stanley and M. P. Alpers, ed. Man-made lakes and human
health. Academic Press, London.
Chadwick, W. L., ed. 1978. Environmental effects of large
dams. Am. Soc. Civil Eng., New York.
-------
359
GENERAL IMPACTS OF EUTROPHICATION ON
POTABLE WATER PREPARATION
HEINZ BERNHARDT
Wahnbachtalsperrenverband
(Association of Wahnbach Reservoir)
Siegburg, Federal Republic of Germany
ABSTRACT
With increasing eutrophication, the preparation of potable water from lakes and reservoirs is so
disturbed by algal blooms and the products of algal metabolism and decay, that the security of
potable water supply at times becomes questionable. Particularly the following disturbing factors
can be enumerated: Penetration of algae through the treatment plant into potable water, inducing
aftergrowth of bacteria, caused by the decay of algal substances in distribution systems,
reservoirs, and end points; dissolved organic substances originating from algae consuming
chlorine impair disinfection of potable water; if chlorine is used for oxidation and disinfection these
compounds are precursors for the formation of trihalomethanes; considerable algal counts in raw
water clog filters; certain algae release taste and odor substances that cause taste and odor
impacts on water. Eutrophication of water depletes the oxygen concentration and leads to the
reductory release and increasing concentration of iron- and mangan-ions in the hypolimnion.
Mangan-ll-ions are often difficult to remove from water. Frequently eutrophic water contains
increased concentrations of ammonia which disturb the disinfection process. Drinking water
reservoirs should be kept in an oligotrophic, eventually mesotrophic, but by no means in an
eutrophic status.
INTRODUCTION
Experience in Germany with treating water from
reservoirs for drinking water during the last decades
has shown that the main disturbance in the treatment
process was caused chiefly by eutrophication of the
impounded water. The following influential factors
have negative effects on the quality of the impounded
water:
1. Domestic, sometimes also industrial sewage.
2. Effluents from farms and cattle-breeding.
3. Effluents from cultivated land (erosion, flushing).
4. Natural phenomena, e.g., moor water, decaying
leaves, humates, mine drainage (e.g., lead, cadmium,
zinc).
5. Biocides (used in agriculture and forestry).
6. Dangerous organic substances, especially those
that are persistent.
This paper is concerned only with the effects of
eutrophication on stagnant waters used as drinking
water.
LOADING WITH PHOSPHORUS AND
NITROGEN COMPOUNDS
From the point of view of usage, the eutrophication of
a stagnant water body represents an undesirable
change in the quality of the water. It is the constant
inflow of too many phosphorus and nitrogen com-
pounds which is particularly responsible for this
negative quality of the water. Phosphorus is of chief
importance because it acts as a limiting factor in most
stagnant water bodies and determines the extent of
phytoplankton production.
If the average annual P-concentration in the
tributaries exceeds the specific concentration limit
tolerable for a given stagnant water then plankton
production is likely to increase to such an extent that
more biomass is produced in the lake than can be
decomposed under aerobic conditions on the bottom.
An input of nitrogen compounds does not influence
primary production if phosphorus is the limiting factor.
Despite this, one should attach a certain amount of
importance to an inflow of nitrate ions, nitrite ions and
ammonium ions because they can either disturb the
water treatment process or impair the quality of the
drinking water.
DETRIMENTAL EFFECTS OF
EUTROPHICATION ON WATER QUALITY
IN LAKES AND RESERVOIRS
The manmade eutrophication of lakes and especially
of drinking water reservoirs causes profound changes,
primarily detrimental, in the quality of the water. These
detrimental effects impair the process of obtaining
drinking water from lakes and reservoirs.
Primary detrimental effects of algal development:
1. Change in algal population as green and blue-
green algae increase. The growth of blue-green algae
is especially detrimental.
2. Extensive occurrence of particulate organic sub-
stances (phytoplankton, zooplankton, bacteria, fungi,
detritus).
3. Occurrence of dissolved organic compounds
which impart odors and tastes. They are released by
the plankton and other microorganisms as products of
metabolism or cellular decomposition.
-------
360
RESTORATION OF LAKES AND INLAND WATERS
4. Formation of organic compounds with chelating or
complexing properties.
5. Formation of humic substances during the
decomposition of organisms.
6. Occurrence of water colored by plant pigments.
Secondary detrimental effects of algal mass de-
velopment:
I.The oxygen budget of the water body becomes
highly overstrained by the decomposition of biogenic
organic substances. This creates water zones free of
oxygen expecially in the sediment-water contact area
and above it.
2. Incomplete mineralization of organic substances
and release of methane. The sediment and the water
near the sediment become enriched with non-
mineralized organic substances.
3. Reductive release of iron—and manganese ions
from the sediment and subsequent increase in their
concentrations in the water.
4. Reduction of nitrate to nitrogen and sulfate to
hydrogen sulfide. Increasing concentrations of am-
monium ions.
IMPAIRMENT OF PRODUCTION AND
DISTRIBUTION OF DRINKING WATER
FROM EUTROPHIC RESERVOIRS
SHOWN BY EXAMPLES
Following is an overview of the effects which a
decrease in the quality of the water of eutrophic
reservoirs has on the treatment process of this water
and on the distribution of this drinking water to the
consumer. This report is based on the Wahnbach
Reservoir Association's experience with treating water
taken from their reservoir which was eutrophic.
Experience with other eutrophic reservoirs was similar.
Owing to the limited space of this publication an
overview of literature on this subject cannot be given.
Excessive Algal Development
Excessive algal development can impair the treat-
ment process to a considerable extent. During the
summer, algae are frequently limited to the upper
layers of the lake which means that the raw water
taken from the hypolimnion is normally low in algae.
However, during the periods of partial and full
circulation, algae reach all the depths of the lake and
thus also the raw water intake zone.
Insufficient elimination of algae using floccula-
tion and filtration
Over a period of some 10 years the blue-green algae'
Oscillatoria rubescens grew in very large quantities in
the Wahnbach Reservoir. Figure 1 shows the depths to
which it spread during 1969. Some years during spring
and autumn the reservoir water turned red. Large algal
accumulations on the surface of the reservoir were not
exactly aesthetically beautiful and not a particularly
good advertisement for a drinking water reservoir.
There were always severe difficulties during the
treatment process in the winter and autumn when the
algal filaments in the raw water (up to 300
filaments/ml) broke through the sand filter and
reached the water supply system.
Algal elimination by means of flocculation using
alum would only have been possible if 200 mg/l
aluminum sulfate and 100 mg/l natrium carbonate had
been added; this is impossible in practice. The
Wahnbach Reservoir Association developed a treat-
ment process which was reasonably satisfactory. It
entails adding a double dose of alum with a maximum
of 20 mg/1 aluminum sulfate and a single dose of 1 to
2 mg/1 of anionic flocculant aid. Despite this new
process, it was impossible to prevent the Oscillatoria
filaments from breaking through the filter from time to
time and thus reaching the drinking water.
Special treatment is required to eliminate the algae
from the water. Some species are more difficult to
eliminate than others. For example, only 90 percent of
the relatively large species Oscillatoria rubescens can
be removed by using several flocculation processes in
succession. The highest elimination rate using floc-
culation and filtration, especially for diatoms, is 99
percent, or in the case of treatment plants which
operate exceedingly efficiently it is 99.9 percent. If
there are mass algal developments in a water body,
these elimination rates are insufficient and the
particulate organic substances cannot be reduced to
small amounts. They form deposits in the distribution
system and storage tanks and act as a basis for the
growth of bacteria, macro-zoo benthos, water-lice,
mussels, etc.
0 50 100 TOO 300 5OO 0 200 0 100 200 Ind./m
Figure 1. — Distribution of the blue-green algae Oscillatoria
rubescens in the Wahnbach Reservoir in 1969.
Filler clogging caused by algae
Mass populations of algae, particularly the large
diatoms, rapidly clog a filter and can thus bring
operations of a water treatment plant nearly to a
standstill. This applies to rapid and slow sand filters.
Because these mass algal blooms occur at relatively
short notice, little can be done to combat them. This
means considerable investment and maintenance
expenditure and there is a limit to even these treatment
processes. Filamentous algae cannot be retained in a
microstrainer satisfactorily.
For example, in the Wahnbach Reservoir the diatom
Melosira islandica developed in large quantities
(approximate chlorophyll concentration of 25 mg/m3)
form Novembver 1972 to January 1973. Asa result the
filter run time was reduced to 8 hours. At the same
time, the blue-green algae, Coelosphaerium naegeli-
anum, which form colonies, appeared in increasing
quantities. Many of the colonies usually disintegrated
-------
HEALTH-RELATED PROBLEMS
361
into individual cells. The small cells (up to 1,000/ml)
could only be eliminated in an unsatisfactory way
despite a double dose of flocculant and an additional
anionic flocculant aid.
However, it was just this additional anionic floc-
culant which encouraged the rapid clogging of the filter
owing to the presence of diatoms so that eventually
filter run times were reduced to 4 hours. Technically,
such short filter runs are no longer worthwhile. The
water throughput in the plant had to be reduced by 30
percent for a period of 5 weeks until the population of
diatoms suddenly broke down in the reservoir at the
end of January.
This was the first time that the mass development of
various nuisance species of algae disturbed treatment
operations to such an extent that the plant throughput
was limited. This occurrence must be considered very
seriously to demonstrate the fact that development of
algae in a stagnant water body can impair the
production of drinking water and can even stop it
altogether.
Disturbances Caused by Taste and
Odor Substances
The occurrence of taste and odor substances in raw
water often has a detrimental effect on the drinking
water (taste of cucumber, fish, cod-liver oil, earth, etc.).
Actinomycetes which often grow following a mass
occurrence of blue-green algae are particularly
disagreeable and the substances released by these
microorganisms impart taste and odor. For example,
the organic substance 'geosmine' was isolated from
actinomycete populations and blue-green algae. Even
when it is highly diluted geosmine still imparts an
intense earthy smell.
The mass development of the diatom Melosira italica
which appeared in the Wahnbach Reservoir during
winter circulation in concentrations of up to 100,000
cells/ml imparted a fishy, oily smell and taste to the
water; this originated from products of metabolism and
decomposition. Specific examinations showed that
trimethylamine was present in the water. Even adding
activated carbon to the water failed to eliminate the
impaired taste and smell. What was particularly
unpleasant was the fact that these organoleptically
effective substances remained in the filters for weeks
after a mass diatom bloom in the reservoir, and they
reached the filtrate which meant that the odor and
taste of the water was impaired for over 4 weeks. We
received numerous complaints from the population
during this period.
Disturbances Caused by Dissolved Organic
Compounds
One consequence of extensive algal growth in a
reservoir is the occurrence of dissolved organic
compounds as the products of algal metabolism and
decompositon. Even low concentrations of these
dissolved organic substances (i.e., between 0.5 to 2
mg/1 DOC) can disturb the water treatment process.
Flocculation Disturbance
Flocculation using iron- and aluminum salts is
occasionally disturbed by organic .substances of this
type. These products are of different molecular weights
(Figure 2). Apart from the products of algal metabolism,
other products of cellular decomposition are probably
also of significance. They are the products of
decomposition after decay, and the products of the
metabolism of the zooplankton which arise in
connection with their feeding on phytoplankton. All
these substances undergo further decomposition
owing to the bacteria in the water. Some of them are
completely mineralized; others are converted into high
molecular compounds, e.g, humates which decompose
with difficulty. One assumes that some of them are
acid polysaccharides. Several of these substances
disturb the flocculation process because they (a) are
compounds which have a complexing effect; (b) are
compounds with strongly acid groups which enrich the
negative surface charge of the colloids and suspen-
soids which are already negatively charged; (c) form
insoluble compounds with the iron-lll-and aluminum
ions added for flocculation.
All these processes hinder or prevent the hydrolytic
formation of the polynuclear hydroxcomplexes of the
iron-Ill or aluminum ions, respectively, which are
required for destabilizing the negative colloids and
suspensoids. These processes can also impair colloid
discharging.
C% 100—i
75-
50 -
25 -
0 -1
Molecular - Weight
1 000 Boundaries
'
55
10,000
24
30,000
12
Figure 2. — Percentual proportions of the four fractions of
molecular weights of the DOC in the water of the pre-reservoir
(November 1979).
Figure 3 shows an example of flocculation disturbed
by algal organic substances in the filtrate of an
Oscillatoria rubescens suspension taken from the
Wahnbach Reservoir during a mass algal development.
The filtrate contained 14 mg/1 dissolved organic
carbon. 10 mg/1 bentonite was added as turbid
material for flocculation tests. Experiments showed
that it was only after a dose of some 300 mg/1
aluminum sulfate that the negatively charged particles
discharged; a dose of 500 mg/1 was required before
the particles re-charged. If one compares the course of
electrophoretical motion of the bentonite suspension
using reservoir water free of algae, it is easy to see that
the bentonite particles were already discharged after a
-------
362
RESTORATION OF LAKES AND INLAND WATERS
dose of 20 mg/1 aluminum sulfate and they become
positively charged after 40 mg/1 aluminum sulfate. It
was possible to decrease the remaining turbidity only
after adding about 300 mg/1 aluminum sulfate.
Normally, less than 10 mg/1 aluminum sulfate are
sufficient. Flocculation using iron or aluminum salts is
disturbed when algal-borne organic substance is
present in concentrations of 3 mg/1 dissolved organic
carbon. However, during mass blooms of Oscillator/a
rubescens. we registered algal organic substances in
the Wahnbach Reservoir in concentrations of up to 50
mg/1 dissolved organic carbon.
2.0 -
Reservoire Water
Filtrate of Suspension
0.4 -i
I I
AI3+ln The
Filtrate
0 0.8-
p
O 0.4-
T' ET o-
o- Ej u
O "I
H si
0 1 1
UJ
f
Reservoire Water
MF Filtrate
Osc. Rub Susp.
20 40 100 300
mg/l AI2 (SO4)3-18H2O
500
Figure 3. — Disturbance of the flocculation process caused by
mass development of Oscillatoria rubescens in the reservoir.
Results of Jar-tests with the filtrate of the algal suspension.
Disturbance of the Disinfection Process
Increased concentrations of dissolved organic com-
pounds in the water of eutrophic lakes and reservoirs
have an unfavorable effect on disinfection using usual
methods. Specific examinations showed that the
process of germ extinction using chlorine in water
containing organic compounds is much slower and
sometimes incomplete. The presence of these sub-
stances not only causes chlorine depletion or con-
sumption of chlorine dioxide if chlorine or chlorine
dioxide are used for disinfecting, but decreases the
speed of the disinfection process. Although chlorine
dioxide dose not act like chlorine with these organic
substances, it is reduced by organic compounds just
like chlorine and chlorite ion forms which are harmful
to humans. Not more than 0.5 mg/1 chlorine dioxide
should be added to the water to prevent the chlorite ion
concentration in the water from exceeding this figure.
If organic substances are present in a colloid form, they
can form a protective coating around the micro-
organisms, and the disinfectant has difficulty penetra-
ting this coating. Under these conditions far more
disinfectant has to be used than is required under
normal water treatment conditions.
On the whole, disinfection is impaired in the
presence of a dissolved or colloidal organic algal
substance. Therefore, these substances decrease the
total safety of the drinking water supply. On the other
hand, a higher dose of, for example, chlorine as a
disinfectant, impairs the taste and odor of the water
and creates undesirable toxic substances such as
trihalomethane, or chlorite ions when chlorine dioxide
is the disinfectant.
Alga! Organic Substances as Precursors for the
Formation of Trihalomethanes
In treatment plants which take water from eutrophic
water bodies, it may be necessary to use pre-
chlorination to exclude problems during coagulation
and filtration and to guarantee terminal disinfection
with chlorine. However, this measure leads to the
formation of trihalomethanes. It is a well-known fact
that humic acids react with chlorine forming trihalo-
methanes. We were also able to confirm the fact that
algal organic substances act as precursors (Figure 4).
This figure shows the trihalomethane concentration
plotted against the reaction time of the chlorine added
to the water (chlorine depletion time) of unfiltered algal
cultures (thick line) and to the water of the pre-
reservoir during algal bloom (300,000 ind./ml) (dotted
line). In these tests the da: Cwas kept at 3. These
graphs show the similarity of the amount of halo-
methanes produced at the same reaction time of
chlorine with either the organic algal substances in
these cultural solutions (unfiltered) or with pre-
reservoir water.
Figure 5 shows the linear connection between the
concentration of TOCI and THM compounds after 20
hours of chlorine depletion and the content of dissolved
organic substance (precursor) expressed as the
spectral absorption coefficient at 280 nanometers
(water samples were taken from the eutrophic pre-
reservoir and from the mesotrophic main reservoir).
This figure shows that under constant chlorine
depletion conditions the concentrations of formed
TOCI-compounds and the THM-compounds are pro-
portional to the content of organic substances in these
water samples. Some of these organic substances act
as precursors for the formation of THM-compounds.
In the Federal Republic of Germany trihalomethane
concentrations of 25/ug/l as an annual average are
considered permissible but measures should be taken
to reduce the trihalomethane concentrations in
drinking water to far below this figure. The best way of
doing this is to keep the presence of precursors in the
raw water as low as possible. This is achieved mainly
by controlling the extent of algal development.
-------
HEALTH-RELATED PROBLEMS
363
DC
III
O
O
O
I
5
z
300
200
100
, _— Fragilaria
Pre - Reservoire
f I
I
0 5 10 15 20 25 (h)
REACTION - TIME
Figure 4. — Formation of THM as a result of the reaction of
chlorine with algal organic substances compared with pre-
reservoir water rich in algae.
200 r-
150
2
c
o
I 100
CM
O
50
246 8 10 (rrT1)
Absorption Coeffic.
Figure 5. — Formation of theTOCI and THM depending on the
DOC in the water (Absorption coefficient at 280 nm, C\s-
reaction time 20 h).
The Overstraining of the Oxygen Budget
of a Stagnant Water-Body
If the mass of organic substance formed in a
stagnant water body during the productivity period
exceeds the lake's capacity to decompose this
substance under aerobic conditions in the tropholytic
zone, then oxygen is depleted and finally disappears in
the hypolimnic water, especially on the bottom of the
reservoir. Anaerobic conditions lead to the reductive
mobilization of various compounds.
Iron and Manganese
During their oxidation, the algal organic substances
consume most of the oxygen dissolved in the water and
then reduce the manganese-IV-oxide hydroxide hy-
drates. This increases concentrations of manganese-ll-
ions in the water, a strong indication of oxygen
depletion and of the development of a reduction zone
on the bottom of the lake (see Figure 6). Later, the iron
oxide hydroxides are reduced and iron-ll-ions are
released. For example, we have had concentrations of
manganese-ions up to 10 mg/1 during summer
stagnation in our reservoir after the oxygen had been
used up.
Drinking water should contain only extremely small
concentrations of iron and manganese compounds.
This means extensive iron and manganese elimination
in those drinking water plants in which large quantities
of manganese and iron are present in the raw water
can be a difficult process from a technical point of view,
FUTURE PROSPECTS
These examples demonstrate to what degree the
eutrophication of a stagnant water body mainly used
for drinking water supply can impair the production and
distribution of drinking water. In Germany we claim
that drinking water reservoirs should be in an
oligotrophic to mesotrophic state to guarantee a
definite supply of drinking water. Despite extensive
water treatment there is no certainty of a safe supply of
drinking water from an eutrophic reservoir. For this
reason, the Wahnbach Reservoir Association con-
structed a phosphorus elimination plant at the point
where the River Wahnbach flows into the reservoir to
turn the eutrophic impoundment into a mesotrophic to
oligotrophic state by drastically decreasing the phos-
phorus input.
MAY JUNE
SEPT. OCT.
mg/l
6
4
2
1
0.25
0
MAY JUNE
Figure 6. — Connection between the occurrence of Mn2*- ions
and the increased 02-depletion in the water on the bottom of
the Wahnbach Reservoir (Investigations, 1969).
-------
364
ORGANIC CONTAMINANTS IN THE GREAT LAKES
DAVID WEININGER
Environmental Research Laboratory
U.S. Environmental Protection Agency
Duluth, Minnesota
DAVID E. ARMSTRONG
Water Chemistry Laboratory
University of Wisconsin
Madison, Wisconsin
ABSTRACT
Anthropogenic organic chemicals have been detected in large numbers in the Great Lakes. The
major problem has been the accumulation of stable, lipophihc organochlorine chemicals (PCB's,
DDT, DDE, mirex) to high concentrations in fish. The presence of these chemicals raises concern
over their effects and the recovery time required following implementation of restoration methods.
The compounds detected in the Great Lakes and chemical properties related to behavior in the
environment are reviewed briefly. A mass balance for PCB's in Lake Michigan is presented. Data
are evaluated on the rate of decline of the DDT-group pesticides in coho salmon in Lake Michigan.
This data indicate the residence time for DDT in the water column is about 1.74 years. Transport to
coho via pelagic and benthic food chains corresponds to approximately 80 to 20 percent,
respectively, of the coho body burden of total DDT prior to the ban on DDT use. Transport of PCB's
and DDT are expected to be similar. Consequently, reducing or eliminating the input of PCB's and
similar chemicals into the Great Lakes should result in a fairly rapid and substantial decrease in
the concentrations present in fish.
INTRODUCTION
Contamination by organic chemicals has become a
major problem in the Great Lakes (Delfino, 1979). The
presence of a wide range of anthropogenic organic
chemicals has raised concern over the potential for
harmful effects on human health and on the flora and
fauna of the Great Lakes ecosystems. The potential
adverse effects include reduced reproductive potential,
reduced resistance to disease, behavorial abnormali-
ties, cancer, genetic mutations, and physical deformi-
ties. Accumulation of some compounds by aquatic
organisms to levels considered unsafe for human
consumption has been the major health-related
problem identified. However, information is sparse on
the effects of organic contaminants on aquatic
organisms in the Great Lakes.
The problem of contamination by organic chemicals
is not restricted to the Great Lakes. Almost all aquatic
and terrestrial ecosystems are contaminated to some
degree. However, the Great Lakes provide an important
example for evaluating the problem. This information
can be used to assess the possible behavior of organic
chemicals in other aquatic ecosystems.
To illustrate the magnitude of the problem, a partial
listing of anthropogenic organic compounds detected in
the Great Lakes or their tributaries is given in Table 1.
Although this is not an all-inclusive listing, it does
suggest the scope of the problem of identifying the
compounds, locating their sources, and determining
their fate and effects in the Great Lakes.
The most widely documented examples in the Great
Lakes are the DDT-group pesticides, polychlorinated
biphenyls (PCB's), and mirex. The DDT-group pesticides
and PCB's are distributed throughout the Great Lakes
at relatively high levels, and Lake Ontario is con-
taminated with mirex. All three groups of compounds
resist degradation, and are lipophilic, leading to high
concentrations in pisciverous fishes. These com-
pounds, in particular, have raised concern over the
potential for similar behavior by other compounds,
especially lipophilic organochlorine compounds. Con-
sequently, a major need exists for understanding the
environmental processes and chemical properties
controlling organic chemicals and for developing a
predictive capability for their transport, distribution,
and fate in the Great Lakes.
In this paper some of the environmental processes
and properties of organic chemicals controlling their
transport and fate in the Great Lakes are discussed and
transport processes are evaluated through analyzing
data on DDT and PCB's in Lake Michigan. Implications
for the rate of recovery from contamination are
discussed.
ENVIRONMENTAL PROCESSES AND
CHEMICAL PROPERTIES
The amount of an organic contaminant in an aquatic
ecosystem can be viewed in terms of the balance
between input and loss rates. In general, the rate of
change of the concentration of a contaminant in the
-------
HEALTH-RELATED PROBLEMS
365
Table 1. — A partial listing of organic compounds reported in the Great Lakes and their tributaries, including fish samples.
Compound or Group
Location
Reference
Polychlorinated biphenyls
DDT-group pesticides
Dieldrin
Chlordane
Mi rex
Hexachlorobenzene
Chlorobenzenes
Chlorophenols
Pentachloroan isole
Chlorostyrenes
Chlorobutadienes
Nonachlor
Polychlorinated dibenzodioxins
Polychlorinated dibenzofurans
Polychlorinated naphthalenes
Dehydroabetic acid
Polyaromatic hydrocarbons
Hydrocarbons (petroleum)
All Great Lakes
All Great Lakes
All Great Lakes
Green Bay (L. Michigan)
Ashtabula River (L. Erie)
L. Ontario
Saginaw River (L. Ontario)
Grand River (L. Michigan)
Tittabawasee River (L. Huron)
Maumee River (L. Erie)
Saginaw River (L. Ontario)
L. Superior
Fox River (L. Michigan)
Detroit River (L. Erie)
Fox River (L. Michigan)
Ashtabula River (L. Erie)
Saginaw River (L. Huron)
Ashtabula River (L. Erie)
Grand River (L. Michigan)
Tittabawasse River
(L. Huron)
L. Michigan
L. Michigan
Fox River (L. Michigan)
Nipigon Bay (L. Superior)
Fox River (L. Michigan)
L. Michigan
Several, e.g., Veith, 1975; Veith, et al. 1977
Glooshenko, et al., 1976; Swain, 1978
Eisenreich, et al. 1979;
Several, e.g., Reinert, 1970; Veith, 1975
Veith, et al. 1977; Glooshenko et al., 1976;
Swain, 1978
Ibid.
Veith, et al. 1979
Kaiser, 1974, 1978; Holdrinet, et al. 1978
Veith, et al., 1979
Ibid.
Ibid.
Swain, 1978
Peterman, et al. 1980
Veith, et al., 1979
Peterman, et al. 1980
Veith, et al. 1979c
Ibid.
Ibid.
Ibid.
Dougherty, et al. 1979
Dougherty, et al. 1979
Ibid.
Peterman, et al. 1980
Kaiser, 1977
Peterman, et al. 1980
Haile, 1977
waters of the Great Lakes is controlled by the following
factors:
1. Inputs
(a) Discharges from industrial areas.
(b) Municipal waste waters.
(c) Land drainage by rivers and streams.
(d) Atmospheric wet and dry deposition.
2. Losses
(a) Surface water discharge.
(b) Volatilization.
(c) Sedimentation.
(d) Degradation.
(e) Harvesting.
3. Recycling
(a) Resuspension and/or release from bottom
sediment.
(b) Biological recycling.
The variety and complexity of the sources and
transport routes have been major problems in
controlling contamination of the Great Lakes by organic
chemicals. Pesticides applied to land areas are
transported by air and water. Other compounds
produced or used in industrial processes near the Great
Lakes or their tributaries, may be transported directly
by air or water or stored in various forms, such as
chemical stocks, industrial products (e.g., electrical
transformers and capacitors containing PCB's), waste
disposal sites (e.g., landfills), or in bottom sediments.
Storage may be temporary or permanent. Leakage of
transformers discarded in landfills may lead to PCB
transport by air and water. Contaminants in harbor
sediments may be transported gradually by sediment
resuspension. However, storage may act as an output
for the system. For example, sedimentation in the
Great Lakes can bury organic chemicals in the bottom
sediments and inhibit their return to the biochemical
cycle.
-------
366
RESTORATION OF LAKES AND INLAND WATERS
The relative importance of these factors depends on
the nature of the organic compound and the lake's
characteristics. For specific compounds, the rates of
losses by volatilization, sedimentation, and degradation
may vary to some extent among lakes due to variations
in mass sedimentation rates, concentrations of
suspended material, and temperatures of waters and
sediments. Furthermore, recycling by resuspension
and biological transport may reflect differences in lake
morphometry and aquatic organism populations.
However, differences in chemical properties are
probably more important than lake differences in
controlling the rate of change in organic contaminant
levels.
Concern over the environmental fate and behavior of
organic chemicals has focused attention on the use of
physical and chemical properties for predicting en-
vironmental behavior. The association of lipophilic or
non-polar character with a tendency for persistence
and bioaccumulation is widely recognized. Important
examples are PCB's and organochlorine pesticides in
the Great Lakes. The octanol-water partition coefficient
(Kow)is often used as a measure of polarity. For many
compounds, Kow values are tabulated (Leo, et al. 1971),
and Kow values can be obtained fairly readily by direct
measurement (Kanckhoff, et al. 1979), prediction
based on water solubility (Chiou, et al. 1977) or high
pressure liquid and chromatography (HPLC) measure-
ments or calculations based on molecular struc-
ture (Hansch and Leo, 1979) (Veith, Austin, and
Morris, 1979). In turn, K0w values have been correlated
with bioconcentration factors (BCF) measured in the
laboratory (Neely, et al. 1 974; Chiou, et al. 1 977; Veith,
DeFoe, and Bergstedt, 1979) and with sediment-water
partition coefficients (K ) for organic compounds
(Karickhoff, et al. 1979; Chiou, et al. 1979; Hassett, et
al. 1 980). Acute toxicity to fish is also highly correlated
with Kow Consequently, a physical property (Kow)
shows considerable promise in predicting important
aspects of environmental behavior. In the case of
adsorption of some compounds by sediments, the Kp of
a given compound depends mainly on the organic
carbon (OC) content of the sediment. This allows use of
an OC-based partition coefficient (K0c) obtained by
dividing K by the fractional OC content of the sediment
and predicting adsorption from the sediment OC
content and the Kow for the organic compound.
However, as K0w and OC values decrease (more polar
compounds) adsorption estimated from these two
parameters may become less accurate.
While obviously important, adsorption and biocon-
centration are only two of several processes and
factors controlling the behavior of organic chemicals in
the environment. Measurements or predictions of
hydrolysis, photolysis, volatilization, and biodegrada-
tion rates and bioaccumulation through food chain
transport are also required. Bioaccumulation from
consumption of contaminated food may not be
predicted by laboratory measurements of bioconcentra-
tion directly from water. Measurements and evalua-
tions of these processes are being actively researched.
Information rates of the important processes
controlling environmental distribution and fate can
provide the basis for modeling the behavior of an
organic chemical discharged into an aquatic ecosystem
(e.g., Smith, et al. 1977). While holding promise for
assessing the wide range of compounds of environ-
mental concern, such models are at a fairly early stage
of development and testing. An alternative approach is
the analysis of data on contaminants widely distributed
in the environment such as PCB's and DDT.
TRANSPORT OF ORGANIC
CONTAMINANTS IN LAKE MICHIGAN
Two basically different approaches can be used to
obtain information on the transport of organic
contaminants. The first approach is extrinsic in nature,
involving the mass balance of PCB's in Lake Michigan.
The second, intrinsic, approach involves extrapolating
the behavior of PCB's in Lake Michigan from
measurements made on certain components of the
lake. Both approaches are presented and compared to
gain insight into the factors controlling the response of
the system to external changes.
A Tentative Mass Balance of PCB's
in Lake Michigan
Estimates of PCB loading (input) and losses for a lake
can be combined with an estimate of the contaminant
"standing crop" to provide some understanding of
transport and distribution within the lake (Eisenreich,
et al. 1979; Pavlou and Dexter, 1979). Important
sources and sinks may be distinguished and con-
taminant residence time estimated.
The amounts of PCB's stored in Lake Michigan can
be estimated based on measurements of the PCB
concentrations in the major reservoirs, the lake water,
and the bottom surficial sediments. Estimates are
summarized in Table 2. Relatively little data are
available on PCB concentrations in Lake Michigan
Table 2. — Tentative mass balance of PCB's in Lake Michigan.
PCBs in Lake Water (kg)
Dissolved
Paniculate (<20% of total)
Total (2 ng/l)
PCBs in Bottom Sediments (kg)
0-2 cm layer (0.1,ug/g)
2-5 cm layer (0.025 /jg/g)
Total
PCB Inputs (kg/year)
Atmospheric
Particulate
Vapor
Wet
Total Atmospheric Estimate'
Tributaries (0.05 fjgA)
Industrial Discharges
Total
8200
1600
9800 kg
30,000
30,000
60,000 kg
1200
0 to 2700
1100 to 4800
5000
1650
6650 kg/year
PCB Losses (kg/year)
Volatilization
Surface water discharge
Sedimentation
Total Loss Estimate^
Oto 3100
100
2600
2700 kg/year
Assumes wel deposition is 1100 kg/year,
^Assumes volatilization loss is 0
-------
HEALTH-RELATED PROBLEMS
367
waters. Murphy and Rzeszutko (1977) found concentra-
tions in the range of 30 to 40 ng/l for a few samples,
similar to the concentrations reported by Haile (1977).
However, recent unpublished data indicate concentra-
tions may be as low as 1 to 2 ng/l. Although the
concentrations are low in either case, the lake water
represents a major reservoir of PCB's, and accurate
information on the concentrations present is highly
important in evaluating fluxes, residence times, and
the rate of response of the system to changes in
external loadings. Based on a concentration of 2 ng/l
and a lake volume of 4.9 x 103 km3 (Klein, 1975), the
estimated mass in the lake water is about 9,800
kilograms. The sediment-water partition coefficient for
PCB's is estimated to be about 1 * 105(Karickhoff, etal.
1979; Pavlou and Dexter, 1979). Based on this partition
coefficient and suspended particulate matter con-
centrations in the range of 0.5 to 2 ng/L (our recent
measurements), less than 20 percent (1,600 kg) is
expected to be present as "particulate" PCB's.
The bottom sediments also represent a major
reservoir of PCB's. Our recent data indicate PCB
concentrations in surficial sediments in southern Lake
Michigan range from 0.005 to 0.2 /ug/g. These a re the
same range as concentrations reported in Lake
Superior sediments (Eisenreich, et al. 1979). The
variations in concentration with location in southern
Lake Michigan are apparently related in part to
variations in sedimentation rate and depth of mixing of
the surficial sediments. Based on estimated average
PCB concentrations (dry-weight basis) of 0.1 //g/g and
0.025 fjg/g and sediment porosities of 75 and 60
percent for the 0 to 2 and 2 to 5 cm layers, respectively,
the estimated sediment PCB reservoirs are 30,000 kg
in the 0 to 2 cm layer and 30,000 kg in the 2 to 5 cm
layer.
The major sources of PCB's to the lake are
atmospheric deposition, tributaries, and industrial
discharges. Recent information indicates atmospheric
input is important and may be the major source of
PCB's to Lake Michigan. Based on field measurements.
Murphy and Rzeszutko (1977) estimated the input in
precipitation was 4,800 kg/year. Atmospheric input by
both wet and dry deposition was estimated by Doskey
(1978). Wet deposition was estimated to be about
1,100 kg/year based on measurements of PCB
concentrations in air over Lake Michigan and calcu-
lated PCB washout. This lower value is used to
calculate the PCB input in Table 2. The estimated
atmospheric input of particulate PCB's was 1,200 pg/g
based on measured concentrations and an estimated
deposition velocity. Uncertainty exists over the vapor
input because of uncertainty in the Henry's Law
constant (air-water partition coefficient) for PCB's and,
thus, whether air-water transfer is gas-phase or liquid-
phase controlled (Doskey, 1978). However, laboratory
measurements indicate transfer is probably gas-phase
controlled. This means net vapor transfer would be
from air to water; the estimated input is 2,700 kg/year.
Consequently, the combined atmospheric input is
about 6,650 kg/yr. The wide distribution of PCB's in
the environment is consistent with the importance of
atmospheric transport. Examples are present of PCB's
in Lake Superior (Glooshenko, et al. 1976; Veith, et al.
1977; Swain, 1978; Eisenreich, et al. 1979) and the
north Atlantic (Harvey and Steinhauer, 1976).
Tributary inputs of PCB's to Lake Michigan are
probably less than the amounts received from the
atmosphere. However, comprehensive data on tribu-
tary inputs are lacking, partly because of the large
number of tributaries and expected variations with both
location and time. In 1970 to 1971, PCB concentrations
ranging up to 0.45 fjg/l were observed in tributaries
entering Green Bay (Veith, 1972). Municipal and
industrial wastes are known to be sources of PCB's in
tributaries. For example, PCB's were detected in
effluents from wastewater treatment plants in the
Milwaukee River watershed (Veith and Lee, 1971),
southeastern Wisconsin (Dube, et al. 1974), and
Michigan (Hesse, 1976). PCB's were also found in
effluents from pulp and paper mills (Kleinert, 1976;
Peterman, et al. 1980). Inputs to tributaries from these
sources may be declining with decreasing use of PCB's.
However, leaching from landfills and other discharges
associated with disposal may represent continuing
sources. Based on the available data on concentrations
in streams and wastewater effluents, Murphy and
Rzeszutko (1977) estimated the input to Lake Michigan
from these sources was approximately 1,650 kg/year.
For an average tributary flow of 33 kmVyear, this
would correspond to an average PCB concentration of
about 0.05 /ug/l .
Losses of PCB's from the lake system occur through
surface water discharge and permanent sedimenta-
tion. Biodegradation, volatilization, and harvesting
losses are considered negligible. Although PCB's can
be partially degraded by microorganisms (Furakawa
and Matsumura, 1976), our laboratory experiments
(Flotard, 1978) involving incubation of Aroclors in
sediments and measuring changes with time in major
peaks indicated negligible degradation in sediments.
More recent experiments in our laboratory showed
some degradation of low-chlorine PCB's in sediments
but indicated degradation was retarded by PCB
adsorption on sediments. In Lake Michigan sediments,
adsorption and low temperatures may completely
inhibit degradation. Volatilization is also uncertain. If
air-water transfer were liquid-phase controlled, vola-
tilization could amount to 3,100 kg/year in Lake
Michigan. However, the evidence for gas-phase control
indicates volatilization losses may be negligible
(Doskey, 1978). Losses through harvesting are slight
because of the small proportion of PCB's contained in
the fish population. The loss through surface water
discharge (water residence time 100 years) is about
98 kg/year based on the water concentration of 2 ng/l
and an outflow rate of 49 kmVyear.
The major mechanism for PCB removal from the
system is sedimentation and burial in the bottom
sediments. Transport of PCB's to the bottom sediments
can be estimated from the mass sedimentation rate
and the concentration of PCB's in the depositing
sediment. The average mass sedimentation rate for the
southern basin of Lake Michigan has been estimated to
be 7 mg/cmVyear based on 210Pb measurements
(Edgington and Robbins, 1976). Allowing for de-
composition of organic matter after deposition
(assumed to be 50 percent), a mass deposition rate of
-------
368
RESTORATION OF LAKES AND INLAND WATERS
about 15 mg/cm2/year is obtained. Our recent data
indicated the PCB concentration in surficial fine-
textured sediment in Lake Michigan is about 0.2
/jg/g. Allowing for some dilution by mixing with non-
contaminated sediment, an estimated PCB concentra-
tion in depositing sediment of 0.3/ug/g is obtained.
This is probably high. For example, calculations based
on a sediment-water partition coefficient of 1 * 10,5, a
suspended sediment concentration of 1 mg/l, and a
lake water total PCB concentration of 2 ng/l indicates
the concentration in the suspended (depositing)
sediment should be about 0.18 A<9/9 . However, if the
mass sedimentation rate for the southern basin is
assumed to represent the average for the entire lake,
the calculated PCB deposition rate based on a mass
deposition rate of 15 mg/cmVyear and a PCB
concentration of 0.3 fJQ/g is about 2,600 kg/year.
The mass balance calculations indicate either the
system is not in steady-state with respect to PCB's or
the mass balance is in error. The estimated input to the
water column (6,650 kg/year) exceeds the loss (2,700
kg/year). Furthermore, if PCB input is assumed to have
occurred at this rate over a 20-year period, the
estimated inputs (—133,000 kg) exceed the estimated
amounts in the water and sediments (69,800 kg) plus
the amounts lost by surface water discharge (100 kg) or
about 70,000 kg. Several possible explanations exist
for these discrepancies:
1. The inputs may be high.
2. The lake water and/or sediment concentrations of
PCB's may be low.
3.The sedimentation rate for PCB's may be low.
4. Biodegradation and/or volatilization may con-
tribute to PCB losses.
As discussed previously, some uncertainty exists
over the importance of biodegradation and volatiliza-
tion losses. However, available evidence supports the
assumption of negligible losses by these pathways.
While the calculated sedimentation rate for PCB's
could be low, evidence suggests the value is probably
high. Estimates for lake water and sediment PCB
concentrations are conservative, therefore, the esti-
mates of PCB's stored in these in-lake reservoirs may
be low. Consequently, the most likely explanations for
the discrepancies seem to be an underestimate of lake
water and sediment values and/or an overestimate of
input rates. The loading rates do not include any
estimates for contributions from point sources such as
the previous industrial discharge at Waukegan (Mur-
phy and Rzeszutko, 1977). Consequently, inputs from
other sources may be overestimated.
If the systems were in approximate steady-state, the
apparent residence time for PCB's in the lake water
could be calculated from the amount of PCB's (9,800
kg) in the lake water and the PCB loss rate (2,700
kg/year), or the input rate (6,650 kg/year). The
residence time would be 3.6 years based on the loss
rate and 1.5 years based on the input rate. More
accurate estimates of PCB input, losses, and storage
are needed to estimate residence times using the mass
balance approach.
Analysis of Data on t-DDT and PCB Concen-
trations in Fish
Fish accumulate microcontaminants directly from
water via their gills (direct uptake) and from their food
(consumptive uptake). Direct uptake can be responsible
for efficient bioconcentration of compounds which are
not eliminated by fish. Long-term laboratory studies
have shown brook trout can accumulate PCB's to levels
8,000 to 25,000 times ambient water concentrations
(Snarski and Puglisi, 1976). Such findings agree with
the 2,4,5,2',5'-PCB bioconcentration factor (BCF) of
14,500 expected for rainbow trout based on aqueous
solubility-BCF correlations presented by Chiou, et al.
(1977).
Although direct bioconcentrations of these magni-
tudes are dra matic, they do not account for the fish PCB
levels found in the environment. If PCB's were
irreversibly accumulated by Lake Michigan lake trout
solely from a concentration of about 2 ng/l "dissolved"
in water, the expected concentrations of PCB's in these
fish would be between 0.016 and 0.05 ppm, based on
these BCF's. A bioenergetics-based model relating
direct PCB exposure to oxygen uptake demonstrated
that the amount of PCB reaching the gills of an adult
Lake Michigan lake trout at a lake water PCB
concentration of 1 ng/l could account for a maximum
whole fish concentration of 0.09 ppm (Weininger,
1978). Lake trout in Lake Michigan contain PCB levels
of 15 to 35 ppm (Willford, 1977; Veith, 1975;
Weininger, 1978).
This implies that lake trout receive PCB's primarily
from their foods. A simple approach to evaluate the
importance of food chain biomagnification of miqro-
contaminants involves comparing the amount of
contaminant ingested to the achieved growth of an
organism. Dividing the contaminant concentration in
the diet by the gross conversion efficiency (GCE) of
growth for the fish provides a measure of the maximum
expected contaminant level in the fish attributable to
dietary accumulation. Adult lake trout in Lake Michigan
feed primarily on adult alewives containing 4 to 7 ppm
of PCB (Eck, 1977; Veith, 1975; Weininger, 1978). The
mean gross conversion efficiency of these trout has
been estimated to range between 23 percent (2 to 3
years old) and 14 percent (7 years old) (Weininger,
1978; Stewart, et al. 1980). Assuming a diet containing
5.5 ppm of PCB, lake trout are expected to accumulate
as much as 24 to 40 ppm of PCB via dietary exposure.
The data showing PCB concentrations in lake trout are
in this range support the conclusion that dietary
accumulation is predominantly important in this
system.
Recognizing PCB and PCB-like contaminants are
principally transported to predatory fish via the food
chain, two primary pathways can be described (see
Figure 1). The first is a pelagic pathway: water —
phytoplankton and suspended particulates — zooplank-
ton — macroinvertebrates -~ forage fish — piscivorous
fish. The portion of contaminants currently reaching
fish via the pelagic pathway has never been removed
from the water to the bottom sediment; this portion is
expected to have a fairly short residence time in the
water column. A second, benthic pathway may also
-------
HEALTH-RELATED PROBLEMS
369
exist: water paniculate matter — sediment —benthic
invertebrate — forage fish — pisciverous fish. The
portion of a contaminant transported in this manner
comes from a sediment "reservoir". In lakes where the
sedimentation rate is low, the benthic pathway is
expected to continue to make persistent contaminants
available to the lake biota for a long time.
[PISCIVEROUS FISHESl
•s^ rci_MWiw
3"- MACROINVERTEBRATES
[PELAGIC FORAGE FISHES| [BENTHIC
£T
FISHES)
PELAGIC
t r
BENTHIC
INVERTEBRATES
_
- [>|ZOOPLANKTONl
~
=!>|PHYTOPLANKTON|
(EBZ
UJi
1
Figure 1. — Pelagic (solid) and benthic (cross-hatched)
pathways of contaminant transport to pisciverous fishes.
We evaluated data on the t-ODT levels in Lake
Michigan coho salmon to obtain insight into the
relative importance of the pelagic and benthic
pathways. The conclusions from this analysis have
been summarized briefly elsewhere (Francis, et al.
1979). A more detailed analysis follows. The wide-
spread use of DDT resulted in a high degree of
contamination of Lake Michigan fishes during the
1960's. In 1970, the use of DDT was banned. DDT
degrades to a limited number of products; by examining
the total concentration of DDT and its degradation
products (t-DDT) during subsequent years, information
on the transport of such chemicals in the environment
can be obtained. Coho salmon are short-lived, fast
growing fish and feed almost exclusively on alewives
during their adult years. Contaminant concentrations
in coho salmon are therefore expected to respond
rapidly to changes in the levels of environmental
contamination.
Following the DDT ban in 1970, t-DDT concentra-
tions in coho decreased rapidly. However, the t-DDT
concentrations in coho seem to approach a new level
distinctly higher than zero (Figure 2). It is hypothesized
that these concentrations result from a two-part
phenomenon. The rapid decrease reflects the removal
of t-DDT from the lake water column and corresponds
to the direct and pelagic food chain transport of t-DDT
to coho. Under these assumptions, "pelagic" portion of
t-DDT transport can be modeled simply as an
exponential decrease with time. The second part of the
model reflects the benthic transport route. Since
sedimentation in Lake Michigan is low and the age of
the sediment mixed zone is high (Robbins and
Edgington, 1975), a very slow decrease in transport via
this route is expected. Although transport from the
sediments will decline slowly due to gradual burial of
PCB's in the sediments, the benthic pathway contribu-
tion is modeled here as remaining constant. The
simplified model is thus written:
where y = t-DDT concentration in coho salmon
t = time, in years since DDT input
elimination
a, b, c = constants.
A weighted non-linear regression provides the
following equation of best fit:
y 11.8 exp (-.5751) + 2.59
Figure 2 shows this model as well as the
individual components. The portion of the 1970 t-
DDT levels in coho salmon attributed to the benthic
pathway (c) is about 18 percent. The residence time
of t-DDT in the water column (1/b) appears to be
short (1.74). This is attributed to t-DDT removal by
sedimentation, i.e., adsorption of t-DDT by the
depositing particulate matter.
-
-
68
\ ^^*_
hftnthic sourrn \ ^ -_^
^_____
1970 1972 1974 1976 1978 1980
YEAR
Figure 2. — Concentrations of t-DDT ( + ) and PCBs( *) in Lake
Michigan coho salmon and 95 percent errors (vertical lines).
Data from Willford (1977). Solid line shows result of weighed
nonlinear regression model; dashed lines show estimated
pelagic and benthic components.
Comparing Results
The short pelagic residence time obtained for t-
DDT in Lake Michigan (1.74 years) agrees approxi-
mately with the residence times (1.5 to 3.6 years) for
PCB's estimated by the mass balance approach.
PCB's and t-DDT are expected to behave similarly
with respect to sedimentary removal. This supports
the validity of assumptions made in calculating the
mass balance. The 1.75-year residence time
calculated from observed coho levels is thought to
be accurate; this calculation procedure would tend
to result in over- not under-estimation of residence
time. Reports of similar residence times for other
insoluble substances (Koide and Goldberg, 1961)
further support this contention. Unfortunately, time
series data on the t-DDT level in Lake Michigan
water, which would verify this result, are not
available.
Differences in the behavior of t-DDT and PCB's
could lead to some differences in their residence
times in Lake Michigan water. Differences in
-------
370
RESTORATION OF LAKES AND INLAND WATERS
suspended particulate matter-water partition coeffi-
cients might lead to differences in sedimentation
rates and residence times. However, the partition
coefficients are in the same range (Chiou, et al.
1979). Differences in volatilization rates and/or
degradation rates are also possible. In aerobic
systems, DDT degrades to DDE (see Guenzi and
Beard, 1976a) but DDE is stable. In anaerobic
sediments, DDT degrades through ODD to other
products not measured as t-DDT (Guenzi and Beard,
1976b). Considering the aerobic nature and low
temperature of Lake Michigan bottom waters, the
degradation of DDT to ODD in Lake Michigan
sediments is probably slow. If t-DDT degradation in
sediments between 1970 and 1976 was significant,
part of the decline in coho t-DDT levels could have
resulted from decreasing sediment t-DDT levels and
decreasing transport via the benthic pathway. This
would result in an underestimate of the t-DDT
residence time in the lake water, i.e., some of the
decline in coho t-DDT levels would be caused by
declining t-DDT levels in the surficial sediments.
However, this effect would be small because the
decline in lake water levels is relatively rapid and the
proportion of the 1970 levels accumulated from the
water column is relatively large.
The PCB mass balance is tentative because of
uncertainties in input and output magnitudes. If the
system is approximately in steady-state and the
residence time for PCB's in the Lake Michigan water
column is 1.74 years, the PCB loading or loss
required to maintain the observed amount of PCB's
in the water column can be calculated:
annual
loading = standing crop = 9,800 kg -5532 kg/year
or loss residence time 1.74 years
This value falls between the estimated loading (6,650
kg/year) and loss(2,700 kg/year)calculated inthe mass
balance (Table 2). This loading is based on the lower
range of reported estimates of atmospheric input by wet
deposition (1,100 kg/year). The general agreement
between the input-output rates based on the t-DDT
residence time and mass balance approaches supports
the validity of the calculated mass balance while
illustrating the lack of precision involved in its
calculation.
The PCB levels in Lake Michigan coho salmon have
not shown the decline observed for the t-DDT levels
(Figure 2). Similar observations have been reported for
Cayuga Lake, N.Y. (Wszolek, et al. 1979). Assuming
similar transport for the two groups of compounds, this
indicates the input of PCB's to the Lake Michigan system
has not been dramatically reduced, although the use of
PCB's and DDT was limited at nearly the same time.
(DDT was banned from use in 1970; use of PCB began to
decline in 1971-1972). This indicates available
reservoirs of PCB's exist in the harbors, rivers, drainage
ditches, and landfills in the Great Lakes Basin.
Apparently, transport of PCB's from these reservoirs to
Lake Michigan is continuing.
MANAGING THE USE OF NEW
CHEMICALS
Approximately 1,500 new chemicals per year are
created and marketed for a wide variety of industrial,
agricultural, and domestic uses. Measures are needed
to ensure that they do not become future con-
taminants. An adequate system of screening chemicals
must be developed by Federal agencies in the near
future. To be effective, the screening process should be
mated to internationally consistent certification pro-
cedures. The rewards for efforts in this regard can be
expected to be indirect, but substantial; it is ultimately
cheaper to refrain from polluting than to restore an
ecosystem.
A variety of methods might be used to evaluate the
hazards of new chemicals (Dickson, 1979). OECD in
Europe and the Office of Toxic Substances in the
United States are developing a tier-structured screen-
ing program similar to that proposed by Cairns (1980).
Depending on the proposed use, a chemical would be
required to pass a set of tests. The first-tier (screening)
tests are rapid and relatively inexpensive. Proposed
screening tests include the BOD test (a chemical
should show at least 60 percent of its theoretical
biochemical oxygen demand), acute toxicity tests (LC
for Daphnia and fathead minnows), the Ames test, and
a test for photosynthetic inhibition activity. If a
chemical does not pass a screening test, or if its
proposed usage warrants, further more sensitive (and
more expensive) tests may be required. These might
include a long-term rodent test for carcinogenicity,
chronic toxicity tests with fish (embryo-larval growth),
and a test for bioaccumulation and persistence
potential. The most extensive tests are reserved for
chemistry which will be introduced into the environ-
ment such as high usage industrial chemicals. These
tests are expected to be primarily field studies and will
be quite^expensive.
The responsibility for conducting certification tests for
new chemicals falls to industry. In most cases, chemical
industries have triedtodevelopand use safe substances
or to recommend adequate disposal techniques to their
industrial customers. Their mandatory participation in
certification screening is a logical extension of this
effort.
Control of substances currently in use (and for which
disposal permits are already issued) is the responsibility
of Federal, State, and Provincial governments. The
success of such a program will depend upon the
establishment of consistent usage-related certification
and availability of adequate disposal sites for hazardous
wastes. Chemicals in current use can be screened by
the same methods proposed for new chemicals;
consideration should be given both to uses and to
disposal. Disposal sites must be made available and
enforcement procedures established to ensure their
use. Disposal site adequacy can be insured only by
appropriate monitoring.
-------
HEALTH-RELATED PROBLEMS
371
MANAGEMENT TECHNIQUES FOR
CONTROLLING PERSISTENT
POLLUTING SUBSTANCES IN THE
GREAT LAKES
The approaches available for reducing the levels of
contaminants in the Great Lakes are limited mainly to
various forms of loading reductions. Reducing the
contaminant loading to a lake will be followed by a rapid
reduction in the contaminant levels in the lake's fish
populations (e.g., DDT in Lake Michigan coho salmon).
The long-term residual contamination of a system
appears to result chiefly from benthic recycling. Even in
the case of t-DDT, the benthic contribution is small ( <20
percent of the 1970 levels).
The short residence time of insoluble organic
compounds demonstrates the potential for large lakes to
rapidly clear their pelagic zones of these compounds and
justifies a comprehensive program to remove
contaminant sources. The first part of this program must
consist of a systematic effort to identify the point
sources of pollutants on a harbor-by-harbor, river-by-
river basis. Removing these sources by both discharge
elimination and contaminant-reservoir containment
(dredging) or destruction (incineration) can provide
substantial rewards in a short time.
Diffuse sources, such as landfills, are difficult to
control, and are expected to continue to release volatile
contaminants, such as PCB's, for many years. The
potential for reducing atmospheric sources is high
because their residence time in the atmosphere is
relatively short (20 to 60 days; Bidelman and Olney,
1974). However, lateral atmospheric transport is rapid,
and truly effective control requires worldwide
compliance. Current stores of organic contaminants
must be identified and practical methods of disposal
developed.
REFERENCES
Bidelman, T. F., and C. E. Olney. 1974. Chlorinated
hydrocarbons in the Sargassso Sea atmosphere and surface
water. Science 182:516.
Cairns, J. Jr. 1980. Estimating hazard. BioScience 30:101.
Chiou, C. T., L J. Peters, and V. H. Freed. 1979. A physical
concept of soil-water equilibria for nonionic organic
compounds. Science 206:831.
Chiou, C. T., et al. 1977. Partition coefficient and
bioaccumulation of selected organic chemicals. Environ. Sci.
Technol. 11:475.
Delfino, J. J. 1979. Toxic substances in the Great Lakes.
Environ. Sci. Technol. 13:1462.
Dickson, K. L, A. W. Maki, and J. Cairns, Jr. 1979. Analyzing
the Hazard Evaluation Process. Am. Fish. Soc.
Doskey, P. V. 1978. Transport of airborne PCB's to Lake
Michigan. M.S. thesis. University of Wisconsin, Madison.
Dougherty, R. C., et al. 1979. Negative chemical ionization
studies for polychlorodibenzodioxins and dibenzofurans in
environmental samples. Abstracts, Am. Chem. Soc Div.
Environ. Chem., 178th Natl. Meet. Washington, D.C.
Dube, D. J., G. D. Veith, and G. F. Lee. 1974. Polychlorinated
biphenyls in treatment plant effluents. Jour. Water Pollut.
Control Assoc. 46:969.
Eck, G. 1977. Fish in the diet of Lake Michigan trout. Presented
at Sea Grant-Great Lakes Fish. Conf. December, Madison,
Wis.
Edgington, D. N., and J. A. Bobbins. 1976. Records of lead
deposition in Lake Michigan sediments since 1800. Environ.
Sci. Technol. 3:266.
Eisenreich, S. J., G. J. Holland, and T. C. Johnson. 1979.
Accumulation of polychlorinated biphenyls (PCB's) in
surficial Lake Superior sediments. Atmospheric deposition.
Environ. Sci. Technol. 13:569.
Flotard, R. D. 1978. The degradability of PCB's in Lake
Mendota sediment. Ph.D. thesis. Water Chemistry Program,
University of Wisconsin, Madison.
Francis, G. R. et al. eds. 1979. Rehabilitating Great Lakes
ecosystems. Rep. Great Lakes Fish. Comms.
Furukawa, K., and F. Matsumura. 1976. Microbial metabolism
or polychlorinated biphenyls. Studies on the relative
degradability of polychlorinated biphenyl compnents by
Alkaligenes sp. Agric. Food Chem. 24:251.
Glooschenka, W. A., W. M. J. Strachan, and R. C. J. Simpson.
1976. Distribution of pesticides and polychlorinated
biphenyls in water, sediments, and seston of the Upper Great
Lakes-1974. Pestic. Monit. Jour. 10:61.
Guenzi, W. D., and W. E. Beard. 1976a. The effects of
temperature and soil on conversion of DDT to DDE in soil.
Jour. Environ. Qual. 5:243.
1976B. DDT degradation in flooded soils as related
to temperature. Jour. Environ. Qual. 5:391.
Haile, C. L. 1977. Chlorinated hydrocarbons in the Lake
Ontario and Lake Michigan ecosystems. Ph. D. thesis. Water
chemistry Program, University of Wisconsin, Madison.
Hansch, C., and A. J. Leo. 1979. In Substituent constants for
correlation analysis in chemistry and biology. Chapter 4. The
FRAGMENT method of calculating partition coefficients.
John Wiley and Sons, New York.
Harvey, G. R.andW. G.Steinhauer. 1976. Transport pathways
of polychlorinated biphenyls in Atlantic water. Jour. Mar.
Res. 34:561.
Hassett, J. J. etal. 1970. Sorption of dibenzthiophene by soils
and sediments. Jour. Environ. Qual. 9:184.
Hesse, J. L. 1976. Polychlorinated biphenyl usageand sources
of loss to the environment in Michigan. In Natl. Conf.
Polychlorinated Biphenyls. EPA-560/6-75-004.
Holdrinet, M. V. H., etal. 1978. Mirex in the sediments of Lake
Ontario. Jour. Great Lakes Res. 4:69.
Kaiser, K. L. E. 1974. Mirex: An unrecognized contaminant of
fishes from Lake Ontario. Science 185:523.
1977. Organic contaminant residues in fishes from
Nipigon Bay, Lake Superior. Jour. Fish. Res. Board Can.
34:850.
1978. The rise and fall of mirex. Environ. Sci.
Technol. 12:520.
Karickhoff, S. W., D. S. Brown andT. A. Scott. 1979. Sorption
of hydrophobic pollutants on natural sediments. Water Reds.
13:241.
Klein, D. H. 1975. Fluxes, residence times, and sources of
some elements to Lake Michigan. Water Air Soil Pollut. 4:3.
Kleinert, S. J. 1976. Sources of polychlorinated biphenyls in
Wisconsin. Natl. Conf. PCB's. Conf. Proc. EPA-560/6-75-
004. U.S. Environ. Prot. Agency, Washington, D.C.
Koide, M., and E. D. Goldberg. 1961. Lead 210 in natural
waters. Science 134:98.
Konemann, W. H. 1979. Quantitative structure-activity
relationship for kinetics and toxicity of aquatic pollutants
and their mixtures in fish. Ph.D. thesis. Dep. Vet.Pharmacol.
Toxicol. Rijksuniversiteit te Uterecht, Netherlands.
Leo, A., C. Hansch, and D. Elkins. 1971. Partition coefficients
and their uses. Chem. Rev. 71:525.
-------
372 RESTORATION OF LAKES AND INLAND WATERS
Murphy, T.J., andC. P. Rzeszutko. 1977. Precipitation inputsof
PCB's to Lake Michigan. Jour. Great Lakes Res. 3:305.
Neely, W. B., D. R. Branson, and G. E. Blau. 1974. Partition
coefficient to measure bioconcentration potential of organic
chemicals in fish. Environ. Sci. Technol. 8:1113.
Pavlou, S. P., and R. N. Dexter. 1979. Distribution of
polychlorinated biphenyls (PCB) in estuarine ecosystems.
Testing the concept of equilibrium partitioning in the marine
environment. Environ. Sci. Technol. 13:65.
Peterman, P. H.,etal. 1980. Chloro-organic compounds in the
lower Fox River, Wisconsin. In B. K. Afghan and D. Macay,
eds. Hydrocarbons and halogenated hydrocarbons in the
aquatic environment. Plenum Publishing Corp.
Reinert, R. E. 1970. Pesticide concentrations in Great Lakes
fish. Pestic. Monit. Jour. 3:233.
Robbins, J. A., and D. N. Edgington. 1975. Determination of
recent sedimentation rates in Lake Michigan using Pb-210
and Cs-137. Geochim. Cosmochim. Acta 39:285.
Smith, J. H., et al. 1977. Environmental pathways of selected
chemicals in freshwater systems. Part I. Background and
experimental procedures. EPA-600/7-77-113. U.S. Environ.
Prot. Agency.
Snarski, V. M., and F. A. Puglisi. 1976. Effects of Aroclor
1254 on brook trout, Sa/ve/inus font/na/is. EPA-600/3-76-
112. Environ. Res. Lab., U.S. Environ. Prot. Agency, Duluth,
Minn.
Stewart, D. J., et al. 1980. An energetics-based population
model for lake trout, Sa/ve/inus namaycush, in Lake
Michigan.
Swain, W. R. 1978. Chlorinated organic residues in fish,
water, and precipitation from the vicinity of Isle Royale, Lake
Superior. Jour. Great Lakes Res. 4:398.
Veith, G. D. 1972. Recent fluctuations of chlorobiphenyls
(PCB's) in the Green Bay, Wisconsin, Region. Environ. Health
Perspect. (April) 51.
1975. Baseline concentrations of polychlorinated
biphenyls in Lake Michigan fish, 1971. Pesic. Monitor Jour.
9:21.
Veith, G. D., and G. F. Lee. 1971. Chlorobiphenyls (PCB's) in
the Milwaukee River. Water Res. 5:1107.
Veith, G. D., N. M. Austin, and R. T. Morris. 1979. A rapid
method for estimating log P for organic chemicals. Water
Res. 13:43.
Veith, G. D., D. L. DeFoe, and B. V. Bergstedt. 1979. Measuring
and estimating the bioconcentration factor of chemicals in
fish. Jour. Fish. Res. Board Can. 36:1040.
Veith, G. D., et al. 1977. Residues of PCB's and DDT in the
Western Lake Superior ecosystem. Arch. Environ, Contam.
Toxicol. 5:487.
Veith, G. D., et al. 1979. Polychlorinated biphenyls and other
organic chemical residues in fish from major watersheds of
the United States, 1976. Pestic. Monitor Jour. 13:1.
Weminger, D. 1978. Accumulation of PCB's by lake trout in
Lake Michigan. Ph.D. thesis. University of Wisconsin,
Madison.
Willford, W. 1977. PCB data on Lake Michigan lake trout.
Available from Bur. Sport Fish. Wildl., Ann Arbor, Mich.
Wszolek, P. C., et al. 1979. Persistence of polychlorinated
biphenyls and 1,1-dichloro-2,2'-bis (p-chlorophenyl)
ethylene (p,p'-DDE) with age in lake trout after 8 years.
Environ. Sci. Technol. 13:1269.
-------
373
ORGANOCHLORINATED COMPOUNDS IN DRINKING
WATER AS A RESULT OF EUTROPHICATION
GERARD DORIN
Environment Directorate
Organisation for Economic Cooperation and Development
Paris, France
ABSTRACT
The impact of eutrophication on drinking water supplies may be serious. It makes its preparation
more difficult and costly. Moreover, it generally reduces the final quality of drinking water
distributed, by giving rise to unpleasant and persistent taste, as well as leading to the formation of
hazardous organochlorinated compounds. Eutrophied waters are, in general, very rich in organic
substances, arising from the metabolism and decay of algae and other aquatic plants. When any
chlorination treatment is applied, organic substances react readily with chlorine, forming soluble
organochlorinated compounds which are very persistent and cannot be efficiently removed. The
high content of organic and nutrient substances contributes to the "dirtiness" of the distribution
network and may increase risk of bacterial growth which also encourages higher final chlorine
application. The formation of organochlorinated compounds will continue in the distribution
system as long as both chlorine and organic substances persist. Although organochlorinated
compounds may already be present as pollutants in raw waters,the chlorination treatment itself is
usually by far the main source of these substances in drinking water. Trihalomethanes are the
volatile compounds, and can easily be identified, but on average they represent only a modest
proportion (20 percent) of all organochlorines in drinking waters. The non-volatile compounds (up
to 80 percent of organochlorines) are still very poorly identified but may well contain more
hazardous compounds. The health risk (essentially cancer) from organochlorines, cannot yet be
fully evaluated but toxicity tests and epidemiological studies suggest that extensive measures
need to be urgently considered to prevent their presence in drinking waters.
INTRODUCTION
Eutrophied waters contain a substantial quantity and
variety of organic substances arising mainly from the
metabolism and decay of algae and other aquatic
plants. These substances may cause unpleasant taste
but in general are not directly toxic to man.
Nevertheless, the utilization of eutrophied raw waters
generally substantially increases the chlorine dosage
and use in drinking water treatment. Instead of a
moderate application for disinfection, at the end of
treatment, chlorine may be used extensively through-
out the whole system: (a) during raw water transporta-
tion to prevent the growth of fixed organisms in the
pipes; (b) during the treatment itself to control
organisms, breakdown of ammonia and other sub-
stances, etc.; (c) as a final disinfection; and (d) to
maintain an increased chlorine residual in the
distribution system (because of the increased con-
sumption of chlorine by the organic substances still
present). Eutrophied waters, because of their high
organic content and increased chlorine treatment,
clearly encourage increased formation of organo-
chlorinated compounds.
EXTENT OF THE PROBLEM
The primary concern in traditional drinking water
treatment has been to control micro-organisms which
cause waterborne diseases (such as typhoid and
cholera) and to provide an aesthetically acceptable
water (taste, odor, color). This goal has been achieved
largely by using chlorine and other oxidants in
conjunction with other treatment processes. Recently,
however, the presence of chemical pollutants in
drinking water and their possible health hazards have
caused increasing concern. With new analytical
techniques and instrumentation, such as gas chro-
matography and mass spectrometry, several hundred
specific organic pollutants have been identified in low
concentrations* in various drinking water supplies.
These compounds generally originate to a minor extent
from the polluted raw waters but to a larger extent from
the drinking water treatment itself. Concentrations of
these pollutants vary from virtually nil in drinking water
drawn from protected ground water to substantial
amounts in drinking water derived from contaminated
surface and ground waters which are chlorinated.
Potable water treatment may considerably increase
the content of synthetic chemicals in drinking water;
recent studies in many countries indicate that the large
number of halogenated products formed by chlorina-
tion are often a major portion of the indentifiable
synthetic chemicals in drinking water. These by-
products are found especially in drinking water derived
from water containing precursors (such as eutrophied
waters) when treated with chlorine; they can be
present at concentrations of up to several hundred
* Typically from 0.01 to 100 microgram/litre (ug/l).
-------
374
RESTORATION OF LAKES AND INLAND WATERS
micrograms per litre. The formation of organochlorines,
in waters rich in organic precursors, is roughly
proportional to the extent and intensity of the chlorine
treatment. For instance, compared to the same treated
waters receiving no chlorine application, a chlorine
disinfection alone may increase the organochlorinated
compounds by 10 times or more, and breakpoint
chlorinations 100 times or more (see Table 1).
Table 1. — Effect of chlorination on the occurrence of some
halogenated compounds in tapwater; concentration range in pg/\
(data from the Netherlands)
Parameters
Type of treatment with chlorine
None
Number of supplies
studied
Chloroform
Bromodichloromethane
Dibromochloromethane
Dichloroiodomethane
Bromochloroiodomethane
Brornoform
1, 1 Dichloroacetone
Trichloromtromethane
13
0.01 - 2.0
001 - 0.9
0.01 - 0 1
0.01
0.01
0.01
0.005
0.01
Final
disinfection
only
4
01-10
0.1-10
0.01 - 5
0.01 - 0.3
0.01 - 0.03
001 - 1.0
0005
0.01 - 3.0
Breakpoint
chlorination
3
25- 60
15-55
3- 10
0.01 - 10
0.01 - 0 3
30-10
01 10
0.01 - 3 0
The organohalogens (see Appendix) formed are
mainly organochlorinated compounds but brominated
and iodinated compounds can also be present. Only a
portion (about 20 percent) of the organohalogens
present in drinking water can currently be identified
(Figure A). These are mainly the volatiles such as the
trihalomethanes (THM's) which include chloroform.
The other identified compounds which may originate
from the raw water account for about 2 percent but
represent a large number of compounds (chloro-
phenols, PCB's, pesticides, etc.). The non-volatile
compounds (up to about 80 percent) are difficult to
identify with current analytical techniques (gas
chromatography, mass spectroscopy). They represent a
large number of compounds and some may be of
greater toxicological significance than the identified
portion (THM's). Their overall level in water can be
measured by the TOCI test (Total Organic Chlorine).
The total amount of organohalogens reaching the
consumer may be higher than the amount measured in
the water leaving the treatment plant, because these
chemicals continue to form in the distribution system
as long as precursors and chlorine are present.
Byproducts may also be formed when using alternative
oxidants such as ozone or chlorine dioxide but probably
to a lesser extent; very limited knowledge exists on
these.
Knowledge of the relationship between the trophic
state of waters, the production of organic precursors,
and the potential formation of organohalogens still
seems to be relatively modest. Basic factors determin-
ing the "yield" of organohalogens during drinking
water treatment are not only the quantities of chlorine
and organic precursors, but also the pH and tempera-
ture.
The following figure is intended to illustrate a typical
distribution of organohalogenated compounds that may be
found in raw waters drawn from rivers in industrialized
countries and in treated waters (after chlorine treatment).
of organohalogenated compound! that may be foum
iters drawn from riven in industrialized countries am
RAW WATERS'
n treated water, (after chloi
RANGE TOCI «10 to 100(ig/l
RANGE THM =2 to 20 /jg/l
tified
10*
Figure A. —Typical distribution of organohalogensin waters in
industrialized countries.
1. Chlorine. As already stated, an increase in the
trophic level of the raw waters used generally
increases chlorination application: in transportation of
raw water, and in treatment, disinfection and distribu-
tion of drinking water.
2. Precursors. Eutrophied water is, of course, much
richer in a variety of organic precursors. Humic
substances (decay of cellulose and lignin) are generally
the main ones; chlorophyll and its derivatives are also
precursors.
3. pH: Under eutrophic conditions, algal activity tends
to consume the CO2 present in water which alters the
carbonic equilibrium, with a corresponding rise in the
pH (which can reach 9). Moreover, a higher pH
increases the yield of volatile organohalogens (tri-
halomethanes). Under certain conditions, with the rise
of 1 unit of pH, this yield may double. Little is known so
far about the corresponding variation of the non-
volatile organohalogens. It seems that the proportion of
volatiles in drinking water (approx. 25 percent) vis-a-vis
the non-volatiles (75 percent) tends to increase with a
pH rise. However, the question is: do the non-volatiles
really decrease in absolute value, remain stable or even
slightly increase with a pH rise? More knowledge on
this point would be desirable.
4. Temperature: A rise in temperature increases the
yield of organohalogens (for instance at 20°C the yield
of organohalogens is about 50 percent higher than at
4°C).
-------
HEALTH-RELATED PROBLEMS
375
The temperature factor shows that in summer the
conditions for organohalogen formation are at their
maximum and strategies to control organochlorine
compounds should pay special attention to this fact. In
practical terms, the storage of water before treatment
has proved to be useful in decreasing the content of
various pollutants (including organic precursors).
However, the organic precursors may rise again rapidly
(after 15 days, for instance) as a result of eutrophica-
tion, especially in summer.
The parameters currently used in most countries for
controlling drinking water treatment plants do not
generally include measuring contamination by or-
ganics and organohalogen byproducts. There is a clear
need for suitable test procedures and also for
operational control of the processes to minimize the
formation of these byproducts. So far, four analytical
techniques are available to measure actual or potential
organic contamination: (1) Total Organic Chlorine, (2)
trihalomethane analysis, (3) Total Organic Carbon, (4)
oxidant demand measurement.
Techniques 3 and 4 give no indication of the
byproducts that may be formed during treatment with
oxidants but measure the precursors present in the
water. Trihalomethane analysis is becoming widely
known and is within the analytical capability of at least
major water treatment works' laboratories. However, it
measures only a small part of the total halogenated
byproducts formed, and could be considered as a
"marker." TOCI is the most comprehensive and
relevant test and should be developed as a standard
test (it does not indicate, of course, which individual
chlorinated compounds are present).
At the low individual concentrations at which some
organic compounds may occur in drinking water, the
primary concern is for their potential contribution to
chronic health risks, e.g., cancer. Although the specific
causes of cancer are not yet fully understood, there is
growing agreement among scientists that exposure to
carcinogenic contaminants in man's total environment
which include food, water, and air, may contribute to
the incidence of cancer which accounts for up to one-
third of the annual mortality in OECD countries. Many
organohalogenated compounds may be found in
drinking water at low concentrations. Even at the
concentration of some micrograms per litre, the
aggregate exposure to such chemicals from a lifetime
of water consumption contributes a potential risk to
human health. In addition, not only is the exposure to
each of these compounds separately of concern, but
also the possibility of synergistic effects. Furthermore,
certain sections of the population are at greater risk
because of age, physical state, environmental stresses,
and possibly genetic disposition.
The assessment of the effects of synthetic organic
chemicals on man is mainly based on animal tests and
on epidemiological studies using statistical data on
human diseases and mortality. In 1976, the U.S.
National Cancer Institute published a study which
showed that under laboratory conditions, cancer was
caused in rats and mice by daily exposure to high doses
of chloroform. Long-term toxicity tests carried out in
France on mice and rats with organic micropollutant
extracts from chlorinated drinking water, showed a
significant increase in the incidence of various types of
malignant tumors.
Various epidemiological studies have explored the
association between organohalogens, or some sur-
rogate parameter found in drinking water, and various
types of cancer. Epidemiological investigations in the
United States have indicated correlations between
increased cancer rates and areas where poor quality,
chlorinated surface waters supply the drinking water.
An epidemiological study in the Netherlands of 4.6
million inhabitants has suggested that where drinking
water is prepared from surface waters of poor quality,
which are chlorinated, a higher cancer mortality rate
was found (especially esophagus and stomach) than in
areas where it is prepared from ground waters of good
quality and generally not chlorinated. Although it is not
yet possible to fully evaluate and quantify the health
hazard resulting from drinking water chlorination, it is
thought that there may be no "safe" or "no-effect"
levels for organohalogens. Other than estimates on
health risks from chloroform, knowledge is still lacking
on the potential hazards from the large number of other
unidentified organohalogenated compounds in water.
Thus prudence is required and it is justifiable to
maintain organochlorine concentration as low as
feasible in drinking water supplies.
TREATMENT AND DISINFECTION
PROCESSES
Evaluation of Possible Approaches
Over the past few decades, potable water supply has
generally been characterized by (1) a net decrease in
the quality of many raw waters used (pollution,
eutrophication), and (2) the consequent intensification
of the treatment applied. The parallel increase of
organic pollutants in waters and chlorine levels used in
treatment (such as breakpoint chlorination) has led to
high organohalogen concentrations in a number of
drinking water supplies.
Unfortunately, the current practice in many drinking
water treatment plants is still to use chlorine
extensively throughout the system. Although its use
corresponds to specific functions, organohalogen
formation will take place all along the system. Any
realistic control policy should carefully consider these
stages:
1. Raw water transportation: Chlorine is used here
for its biocidal effect, i.e., to prevent growth of fixed
organisms in the mains. Other techniques can be used
such as preliminary filtration and clarification of raw
waters before transportation, mechanical cleaning, etc.
2. Purification treatment: Oxidants are used here for
several purposes:
a. The oxidant effect is aimed mostly at removing
various organic and inorganic contaminants such as
ammonia and substances causing taste, odor, or color.
Breakpoint chlorination is frequently carried out to
remove ammonia but various alternatives can be
applied such as biological removal, storage in
reservoirs, or ion exchange. Color can often be
effectively removed by coagulation, and powdered
activated carbon dosing usually controls taste and odor.
-------
376
RESTORATION OF LAKES AND INLAND WATERS
b. The biocidal effect is used in different parts of
the treatment (filters, settling tanks, etc.) to prevent
growth of algae and other organisms. High chlorine
doses may be used, especially in summer, for this
purpose. Various alternatives (physical, mechanical, or
chemical) exist for this application.
c. Other miscellaneous effects, such as action on
colloids and sludge, can be replaced by other
approaches.
3. Disinfectant effect: This is the primary purpose for
which chlorine and other oxidants are used:
a. For waters drawn from polluted sources,
disinfection is necessary. Various approaches exist:
The use of an oxidant such as chlorine, ozone, chloride
dioxide, etc., or ultra-violet treatment. Various filtration
techniques such as slow sand filtration, bankside
filtration, surface infiltration (on soil or dunes) are very
effective and substantially reduce the need for
disinfection;
b. For waters drawn from unpolluted and well-
protected sources (this is the case for ground waters
especially). Different viewpoints exist, however, in
various countries: (1) in a number of countries it is
judged that systematic chemical disinfection of waters
of good biological quality is unnecessary; thus often it
is not applied; (2) in a few countries, however, regular
chlorine disinfection is applied systematically, even to
high quality waters.
4. Residual "bacteriostatic" effect in the distribution
network: This is also a controversial point. In some
countries, it is not common practice to maintain a
chlorine residual in the distribution network under
normal conditions, although this may be done on some
occasions (when a network is not in a good state of
maintenance for instance). Other persistent disin-
fectants can also be used such as chlorine dioxide or
chloramines. A clean and well-maintained network is a
desirable policy to minimize application of a final
disinfectant.
The different oxidants used as treatment reagents
and disinfectants have advantages and disadvantages,
both in terms of their effectiveness and the byproducts
they may generate. The main alternatives to chlorine,
which have already been used in full-scale operation
over a certain period and for which experience exists,
are ozone and chlorine dioxide. Ozone has been used
for potable water disinfection since the beginning of
this century. It is an efficient oxidant and a powerful
disinfectant but does not leave a residual in the
distribution, system; therefore, where necessary, a
bacteriostatic agent (such as chlorine dioxide, chlora-
mine or chlorine) may be added. Chlorine dioxide is also
an efficient oxidant and a very good disinfectant; it
leaves, like chlorine, a residual in the distribution
system. It does not remove ammonia. Little is known
about the possible byproducts of using ozone and
chlorine dioxide and there is concern about the chlorite
and chlorate generated when using chlorine dioxide.
Ozone and chlorine dioxide are more satisfactory for
taste and odor problems than chlorine and have been
used for this reason.
The cost of water treatment is generally a small
fraction of the consumer's cost for drinking water. In
many cases minor modifications to existing treatment
aimed at minimizing the precursors before applying
oxidants and optimizing oxidant application without
endangering the biological quality of the water, will be
effective for little or no cost in substantially reducing
byproducts. As the cost involved is usually moderate, it
is prudent to carry out these modifications where
feasible. Using certain treatments such as granular
activated carbon or resins to remove organochlorine
byproducts after formation would be by far the most
costly option, and probably only needs to be considered
in those cases where water quality is so poor that other
conventional technologies cannot sufficiently reduce
oxidant demand and precursors. Control options
available to small water systems differ considerably
from those available to large systems because small
systems have higher per capita costs, less access to
trained operating personnel, and less capacity to
monitor sufficiently. Using high quality raw waters is
thus particularly important in this case as it makes the
whole treatment and distribution far easier and safer.
Alternative Approaches
Controlling organohalogens in drinking water in-
volves either preventing their formation, or removing
them after they have been formed. The latter approach
is not at present practicable since organochlorines,
once formed, are generally very persistent and pass
through conventional treatment. The preventive ap-
proach is the safer and better method and in general
may be achieved in the following ways:
1. By encouraging the selective use of raw waters of
better quality (non-polluted and non-eutrophic). Where
this is feasible, the use of chlorine (or other oxidants)
can be avoided or at least minimized.
2. By using alternative purification processes (filtra-
tion, precipitation, etc.) which minimize the use of
chlorine or other oxidants at any stage. This approach
is particularly advisable in the case of raw waters
which are moderately or not polluted.
3. By minimizing the dose of chlorine applied and
limiting its use to final disinfection only. This approach
may be practical in many situations (small water supply
installations, for instance) and will be easier if
combined with the alternative processes considered in
2. Other oxidants can also be used.
4. By minimizing the organic precursors before any
chlorine application is made (at the very end of
treatment). This is a basic approach for raw waters of
mediocre quality where both the precursor content and
chlorine application may be substantial.
5. By carefully controlling the conditions of raw
water transportation and potable water distribution, as
these may be major sources of organochlorines in
drinking water: (a) chlorination of raw waters (a
neglected but frequently important organochlorine
source) should be avoided and replaced by alternative
approaches (clarification of water before transporta-
tion, mechanical cleaning, etc.). Using oligotrophic
waters would favorably resolve the problem; (b) when
and where a chlorine residual is judged necessary in
the distribution network, it should be kept as low as
possible; good bacteriostatic agents such as chlor-
amines or chlorine dioxide may also be used. Good
-------
HEALTH-RELATED PROBLEMS
377
maintenance and cleanliness of the networks are of
great importance; they contribute to biological safety
and help minimize the formation of organochlorines
(through lower dosing of chlorine residual and lower
organic content in the pipes).
Careful and moderate use of chlorine is not
condemned, but more prudence and selectivity are
required in its use as there is concern about operating
practices that are not cognizant of the problems of
byproducts and do not attempt to minimize them.
Chemically and biolgically safe water remains the
central goal of drinking water supplies.
CONCLUSIONS
1. Although organochlorinated substances may al-
ready be present as pollutants in raw waters,
chlorination of water containing natural or synthetic
organic precursors is generally by far the main source
of halogenated organic chemicals in drinking water, as
most surface waters contain substantial amounts of
precursors. This is especially true of eutrophied waters
which are generally very rich in organic substances.
2.Trihalomethanes (including chloroform) are cur-
rently the more easily identified organohalogen
byproducts, but normally they represent only a modest
proportion (about 20 percent) of the total organo-
halogens present in drinking water, and not necessarily
the most hazardous substances. Although a large
proportion of organohalogens present in drinking water
are still unidentified, useful overall or partial para-
meters have been developed to permit a closer
assessment of the presence or potential formation of
organohalogens in drinking water. Total Organic
Chlorine is the most relevant test as it is com-
prehensive and applies to the whole range of
organochlorines present (however, it does not in-
dividually identify compounds). Total Organic Carbon is
a useful complementary test for assessing the potential
amounts of precursors. The Trihalomethane analysis is
a relatively easy test but only gives a partial view of the
total mix of chemicals present.
3. Oxidants such as chlorine, ozone, chlorine dioxide
and to a lesser extent chloramines, are effective
reagents in drinking water treatment, especially for
disinfection, their essential function. However, being
chemically very active, they may produce a variety of
byproducts by reacting with the organic precursors
present in waters. Up to now, organochlorinated
byproducts have received most of the attention for a
number of reasons: They are frequently encountered in
significant levels in drinking water; a number of
organochlorines are known or suspected to present
health hazards, and they can currently be detected with
present techniques. Although knowledge is very
limited, it would be prudent to consider the possible
effects of the byproducts which may arise from the use
of other oxidants.
4. In many drinking water treatment installations,
chlorine is used extensively throughout the system
from the initial raw transportation to final drinking
water distribution. It is clear that chlorine applications,
particularly from the early stages when water may still
contain substantial levels of organic precursors, will
lead to significant organochlorine formation. A better
control of organohalogens in drinking water requires
more selectivity in the use of chlorine, which should, as
far as possible, be kept to its essential role of final
disinfection. For the bacteriostatic effect in the
network, a chlorine dioxide or chloramine residual is an
effective alternative.
5. In principle, processes which remove or reduce
contaminants (physical and biological treatment)
should be preferred to processes such as chemical
treatment which transform them into other chemicals
with undesirable or unknown effects. Authorities
should also specify and control the quality of additive
chemicals used in potable water treatment.
6. The gradual decrease frequently noted in the
quality of raw waters used over the past few decades,
has intensified treatment. The parallel increase of both
organic pollutants in waters and chlorine applications
all along the treatment system has lead to the
organohalogen levels currently encountered in drink-
ing waters. Using good quality raw waters is thus
fundamental to controlling organohalogens and other
trace pollutants in potable water.
7. Breakpoint chlorination, commonly practiced for
ammonia removal, may lead to high levels of
organochlorines in drinking water. Thus it seems
advisable to use other ammonia removal methods such
as biological removal, storage, resins, better protection
of the source, or combinations of these processes..
Under exceptional circumstances (e.g., during cold
periods) when breakpoint chlorination is used, it should
be carried out as a final disinfection treatment stage,
after removal of organic precursors.
8. Under certain geographical and geological condi-
tions, raw waters of good quality may, however, have a
high content of humic and fulvic acids. Although these
substances may not in themselves present a real
hazard to human health, they react readily with
chlorine to form organochlorinated compounds. Pre-
cautions should be taken with these waters so that the
processes used throughout the water transportation,
treatment and distribution system, minimize the
formation of organohalogens. Similar caution is
required with sources subject to sea water intrusion
and bromide contamination, as chlorination will lead to
the formation of significant amounts of both organo-
bromides and organochlorides.
9. Where chlorine is used for purposes other than
disinfection (e.g., keeping the treatment plant clean
and free of biological growth) alternative approaches
should be adopted (such as shock-dosing and rinsing of
installations).
'10. Chlorination of raw waters during transportation,
carried out for secondary purposes only (control of fixed
organisms) may be a very important source of
organohalogens; however, it is generally underesti-
mated or neglected because it does not take place in
the plant. A number of alternative processes such as
clarification of water before transportation, mechanical
cleaning, shock-dosing and rinsing, etc., can be used
successfully.
11. When a distribution system is in poor condition,
high chlorine dosing is often used to maintain
substantial disinfectant residual, and in the presence of
-------
378 RESTORATION OF LAKES AND INLAND WATERS
precursors the formation of organohalogens will
continue as long as chlorine persists in the system.
Good maintenance of the distribution network contri-
butes to biological and chemical safety of water as it
minimizes the use of a chlorine r.esidual.
12. The microbiological quality of drinking water is,
like its chemical quality, of prime importance, and
biological safety should not be compromised when
improving the chemical quality. Sufficient technologies
are available to optimize both biological and chemical
purity of water at a cost which, especially for larger
water systems, is generally a modest fraction of the
consumer's cost for drinking water. Thus, it is false
economy to sacrifice drinking water quality by not
applying optimal treatment.
13. The increased risk of cancer due to organo-
chlorinated compounds in drinking water cannot yet be
fully evaluated. Besides the estimates on health risks
from chloroform, knowledge is still lacking on the
potential hazards from the large number of unidentified
organohalogenated compounds encountered in drink-
ing water; they may be much higher. As there may be
no "safe" level for these substances, it is justifiable to
maintain their concentration at the lowest practical
level.
14. It is desirable that guidelines be fixed, preferably
in terms of total organic chlorine at the consumer's tap.
Sufficient flexibility should be left in application,
especially in different cases. However, unless the goals
expressed in the guidelines or standards are stringent
enough they may have a negative effect on a large
number of water works which are already within the
limits fixed, and act as a disincentive for any further
improvement. In other words, they must not be the
lowest common denominator but should be focused on
the best levels realistically attainable.
APPENDIX
Precursors are natural or synthetic organic compounds
capable of reacting with chlorine (and other halogens)
and producing organochlorinated compounds (or more
generally organohalogenated compounds).
Organohalogens (or organohalogenated compounds):
Organic compounds whose molecule contains 1 or
more halogens (such as chlorine, bromine and iodine).
Organochlorines (or organochlorinated compounds):
Organic compounds whose molecule contains 1 or
more chlorine atoms.
Volatile organohalogens: the molecule contains less
than 4 atoms of carbon.
Non-volatile organohalogens: the molecule contains 4
or more carbon atoms.
Trihalomethanes (THMs) are volatile organohalogens
whose molecule contains 1 atom of carbon, 1 atom of
hydrogen and 3 atoms of halogen. When there are 3
atoms of chlorine it is Chloroform. When there are 3
atoms of bromine it is Bromoform. For instance, when
there is 1 atom of bromine and 2 atoms of chlorine, it is
monobromodichloromethane, etc.
-------
379
THE IMPACT OF TOXIC TRACE ELEMENTS ON INLAND
WATERS WITH EMPHASIS ON LEAD IN LAKE MICHIGAN
ALAN W. ELZERMAN
Environmental Systems Engineering
Clemson University
Clemson, South Carolina
ABSTRACT
Because of their widespread distribution and potentially significant detrimental effects, trace
metals as pollutants have been extensively researched. Attention has focused on point sources
and gross pollution, selected heavy metals, metal-organic associations and complexes, extremely
toxic metals, transport mechanisms, metal species which undergo biological transformations,
nonpoint sources, and sinks of trace metals. Non-metal, non-nutrient trace elements like As and
Se have also become of concern. Fundamental understanding of the impact of trace elements on
inland waters has not yet, however, been accomplished. Analytical and sampling limitations
coupled with low concentrations encountered and possible multiplicity of species present have
hampered progress. Also, the criteria for measuring impacts are not well developed. In particular,
little is known of the biological effects of low level chronic exposures and potential synergistic
effects. Some progress has been made. Considerable information is now available on common
sources and sinks of trace metals. Modeling efforts and analytical developments are advancing
toward fundamental predictive capabilities, and toxicological research has progressed beyond
simple acute response measurements. Nonpoint sources, especially atmospheric inputs,
paniculate phase transport, sediment processes, and biochemical transformations are recognized
as being critical to management strategies. Trace elements most likely to have adverse impacts
have been identified, and strategies for impact assessment have been developed. Lead in the
sediments of Lake Michigan offers a useful study example to review current knowledge and
capabilities.
INTRODUCTION
Earlier considerations of trace metal pollution have
focused mainly on metals in drinking water. Standards
for lead, copper, and zinc, for example, were first set by
the U.S. Public Health Service in 1925 (Pojasek, 1977).
Progressively more complex problems have been
approached as analytical capabilities have improved
and environmental concerns have broadened. Increas-
ing inputs of trace metals to natural waters from man's
activities have been documented. Non-metal, non-
nutrient trace elements, like As and Se, are also
potentially significant pollutants.
Fundamental and comprehensive understanding of
the impact of trace metals on inland waters has not,
however, been accomplished. Analytical and sampling
limitations, low concentrations, and numbers of
species present have hampered progress (Stumm and
Morgan, 1970; Stumm and Bilinski, 1973; Brewer and
Spencer, 1975; Am. Chem. Soc. 1978). Sources and
sinks of trace metals are relatively easily identified, but
important species, cycling processes, and controls on
concentrations are more difficult to determine. Also,
the criteria for measuring impacts are not well
developed. In particular, knowledge of the biological
effects of low level chronic exposures and potential
synergistic effects is incomplete.
Some progress has been made. Currently available
knowledge can be used to improve investigative,
management, and restorative practices. This paper
reviews some current knowledge and shows its
application to Lake Michigan.
ASSESSMENT OF TRACE METAL
POLLUTION IN NATURAL WATERS
Complete understanding of the inputs and outflows,
the physical, chemical, and biological forms and
interactions of trace metals within the system, and the
significance of changes in these characteristics, is the
unattained goal of trace metal aquatic pollution
research. Extensive and valuable information is
available and has been reviewed in detail (Stumm and
Morgan, 1970; Martell, 1971; Schnitzer and Kahn,
1972; Singer, 1973; Stumm and Baccini, 1978)
Reviews of analytical techniques for trace metals,
which point out the limitations of currently available
data and the possibilities for advances based on new
analytical approaches are also available (e.g. Mancy,
1971; Burrell, 1974; Quinby-Hunt, 1978). Pojasek
(1977) and Ketchum (1972) present systems ap-
proaches to impact assessment, and James (1978) and
Jenne (1979) have edited reviews of modeling
approaches that are relevant to trace metal pollution.
Many problems, like metal speciation and its relation
to toxicity, result directly from underdeveloped study
techniques (Andrew, et al. 1976; Cantillo and Segar,
1975), but some shortcomings relate to broader
concepts. Three generalizations summarize the prob-
lem: (1) Investigations are often too limited in scope to
-------
380
RESTORATION OF LAKES AND INLAND WATERS
contribute to general advances — they focus on areas
prescribed by current "fads"; (2) attempts to synthesize
information have often become merely mathematical
techniques; and (3) the importance of the relationship
between data, regulations, and environmental impacts
is often subjugated to other considerations. For
example, a list of fads in aqueous trace metal research
might include gross pollution from point sources, waste
treatment processes, metal-organic associations and
complexes, adsorption, biomethylations, and nonpoint
sources. Evidence supporting the second generaliza-
tion might include box models complete with arrows
and labels for concentrations and transfer coefficients
but with little indication of interest in obtaining that
information. Finally, an example of the third generaliza-
tion is the current emphasis on expedient effluent
guidelines rather than receiving-water effects (An-
drew, et al. 1976).
The point of these generalizations is not to be critical.
Progress in investigations usually requires limiting
their scope. Simplifications are necessary. The art of
science is making the best simplifications. The point is
to recognize each generalization as a step only and not
as an end in itself, and to ask many and varied
questions, including the difficult ones. Despite years of
investigation and vast amounts of information, the
integrative and predictive capabilities required for
impact assessment and effective management of
inland waters are not adequate.
What are current capabilities and how should impact
assessment and management be approached? Nu-
merous answers are possible and many different ones
would have validity. The approach taken here does not
claim to be original, nor to overcome the limitations
mentioned. It is, rather, intended to represent some
current approaches and knowledge.
Definition of the System
Impact assessment first requires delineating a scope
of interest that attempts to include all relevant factors
but remains manageable in size. For trace metal
pollution, minimum consideration includes: (1) Organ-
isms or ecological subsystem affected or to be
protected; (2) elements and forms of elements added
and present; and (3) physical locality. Note that the
three areas of decision are not independent of each
other. Essentially, definition of the system results from
a combination of assumptions, previously available
information, and perceived interests. Flexibility in
changing the system must be maintained. The validity
of the system chosen is critical to both the attainment
and the usefulness of the results. It is impossible to
answer all questions, so the goal must be to answer the
right questions.
Organisms or ecological subsystems: Understand-
ably, primary consideration of health effects normally
centers on humans. However, the importance of
broader concerns of environmental impact has been
established and protection of many aquatic organisms
is desired. In the case of Lake Michigan, no direct
health effects on humans resulting from trace metal
pollution of the water or fish are evident (Torrey, 1976;
Andrew, et al. 1976; Int. Joint Comm. 1978, 1980). In
fact, the offshore water of Lake Michigan meets all
International Joint Commission target criteria for trace
metals which also consider effects on aquatic
organisms (Torrey, 1976; Int. Joint Comm., 1978; and
Table 3). Hg in fish has been of concern in some of the
lower lakes (Int. Joint Comm. 1978). Hg has received
considerable attention in other systems, of course,
including adverse effects on human health.
If all criteria are being met, should further
consideration be abandoned? The answer is no. The
fact that all criteria are being met can mean either we
know all the answers, or we didn't ask the right
questions. In the words of Brown (1976), "we might be
in danger of outsmarting ourselves." The complicated
nature of toxic reactions has become evident and
sublethal effects (see Table 1) as well as acute effects
must be considered. The common approach of setting
standards for individual metals in aquatic systems
overlooks possible additive, synergistic, and antago-
nistic effects as well as variations in external
environmental factors and previous history of the
organism (Zitko, 1976; Anderson and Weber, 1976;
Cairns, et al. 1976). For example, International Joint
Commission standards for individual metals were set
below levels thought to have measurable toxicity to
algae, but a mixture of all of the metals at
concentrations just 10 percent of the standards proved
toxic to test algae (Int. Joint Comm. 1978; Wong, et al.
1978).
Present abilities to predict or even measure toxicity,
especially under environmental conditions, are im-
proving but are still limited (Zitko, 1976; Dagani, 1980;
Water Pollut. Control Fed. 1980; Am. Chem Soc. 1978).
Research is being conducted on new measures of
toxicity like ATP activity (Riedel and Christensen,
1979), low level in situ techniques (e.g. Marshall and
Mellinger, 1978), multiple factor and synergistic
toxicity (e.g. Anderson and Weber, 1976; Vernberg,
1978), and continuous sublethal monitoring (Dagani,
1980; Bruber, et al. 1979). However, in most cases
management decisions must still be based on assumed
or potential impacts, and impact assessments often
rely on arbitrary safety factors rather than detailed
knowledge.
Although younger life stages and some species are
more susceptible than others, universally acceptable
indicator species are not available (Dagani, 1980;
McKim, 1977; Brown, 1976). Few impact assessments
or management decisions will be afforded the luxury of
limiting concern to one or two organisms, but as a
practical matter, choices will have to be made.
Table 1 — Examples of sublethal effects (U.S. EPA, 1979).
Disruption of Normal Behavior (feeding, breeding,
locomotion)
Interference with Thermoregulation in Birds and Mammals
Abnormal Biological Processes
Decrease in Reproductive Success
Change in Growth Rates
Effects on Competitive Balance and Predator-Prey
Relationships
Shifts in Population Age Structure
Mutagenicity, Teratogenicity, and Carcinogenicity
-------
HEALTH-RELATED PROBLEMS
381
Table 2. — (Brown, 1976; Wood, 1975; Ketchum, 1972).
Widespread Toxic
Elements
Co, Bi, Ni,
Cu, Zn, Sb,
Cd
Widespread Toxic
Elements Also Po-
tentially Present
as Metal-alkyls
Sn, Se, Te,
Pd, Ag, Pt,
Au, Hg, Tl,
Pb
Metals with
Element
Sn
Sb
Pb
Fe
Cu
Zn
Mo
Hg
Ag
Ni
High Anthropogenic Mobilization
Rates
Man-induced rate
natural rate
110.
31.
13.
13.
12.
11.
4.4
2.3
1.4
1.1
Elements and forms of elements: The impact of any
trace metal input is a complex function of (1) the
amount and timing of the input, (2) the form of the input
and the forms present after any transformation in the
water system, (3) the exposure of organisms to the
input, and (4) the innate toxicity of forms present.
Exposure is governed substantially by the physical
characteristics of the water system relative to
ecological habitats, and the nature of the input material
is governed by the source. The forms present are
controlled by the characteristics of the water system
and the elements.
Of the 92 elements from hydrogen to uranium, all but
22 are metals. Several other relatively toxic elements,
especially As and Se, have some metallic and some
non-metallic characteristics which give them complex
aqueous chemistries (Holm, et al. 1979). The term
"trace" metal is a relative term that has exact meaning
only in specific cases but generally infers a concentra-
tion below 1 mg/l (Brown, 1976).
Adverse impacts could result from adding sufficient
amounts of any metal. A combination of high degree of
toxicity, tendency to accumulate in organisms, and
widespread distribution has been used to indicate
potential hazards. Since trace metals are naturally
ubiquitous and many are essential in biochemical
processes, a further refinement in targeting potential
impacts has been to use the ratio of anthropogenically
mobilized metal relative to natural flux rates. Table 2
summarizes elements which appear as problems in
analyses based on these criteria. Table 2 overlooks
many factors, but eight elements, Sn, Sb, Pb, Cu, Zn,
Hg, Ag, and Ni, appear on both lists and are,
presumably, especially worthy of consideration. Stud-
ies conducted for the International Joint Commission
have also developed a list of elements of concern: Pb,
Cu, Zn, Hg, As, Se, Cd, Cr, and V (Int. Joint Comm.
1978,1980). Commission objectives and example Lake
Michigan concentrations for these elements are
presented in Table 3.
The International Joint Commission objectives, like
most standards set for trace metals, are based on total
metal concentrations, with the exception of Hg, which
pertains to filtered samples. As discussed in the trace
metal aquatic chemistry reviews summarized in Table
4, numerous forms of any one element can exist in a
natural water system. Toxicity is known to be a function
of the specific forms present (Lee and Hoadley, 1967;
Table 3. — Offshore water concentrations and objectives for
designated elements (Int. Joint Comm., 1978; Torrey, 1976;
Elzerman and Armstrong, 1979).
Element IJC Objective Typical Offshore Lake Michigan
(fjg/\) Total Concentration (/Kg/I)
Hg
Pb
Cr
Cd
Cu
Zn
Se
As
0.2
25.
50.
0.2
5.0
30.
10.
50.
0.02-0.20
0.8
3.0
0.03
1.2
1.2
0.1
1.1
Table 4. — Summary of major forms of trace metals in aquatic
addition to variations in oxidation state).
Dissolved
Non-Living Particulate
Living
1. "Free" Ion 1. Adsorbed, as any of 1. Adsorbed
(hydrated only) dissolved forms
2. Inorganic com- 2. Precipitated; Amorphous 2. Absorbed
plexes or crystalline, including
substitutions and co-pre-
cipitates
3. Organic complexes
4. Alkylated or other
organic
Am. Chem. Soc. 1978; Andrew, et al. 1976). Some
progress is being made in relating forms of metals in
aquatic systems to toxicity and organism accumulation
(Andrew, et al. 1976, 1977; Whitfield and Turner,
1979; Magnuson, et al. 1979; Vernberg, 1978). In
many cases, the most toxic form of the metal seems to
be the "free" cation (hydrated only), except that
alkylated species, when present, are generally even
more toxic than inorganic forms. The form of metal
present probably affects potential accumulation in
organisms as well as direct toxicity (e.g., Dodge and
Theis, 1979).
On the basis of input rates, tendencies to be
biomethylated, and potential accumulation in sedi-
ments and biota, the International Joint Commission
considers Pb and Hg to be of greatest concern in the
Great Lakes (Int. Joint Comm. 1978). Concentration
objectives, however, are still given in relation to total
concentrations (except for Hg). Current knowledge does
not allow setting criteria for specific forms since
-------
382
RESTORATION OF LAKES AND INLAND WATERS
Table 5. — Lead distribution in southern Lake Michigan (Elzerman, 1976; Int. Joint Comm. 1978; Edgington and Bobbins, 1976).
Reservoir
Water: Dissolved
particulate
Biota: Plankton
Fish
Sediments: top cm
1-4 cm
Typical Cone.
0.7/yg/l
0.1 fjgfl
?
0.5/ug/l
120/jg/g
80 /ng/g
Reservoir Vol.
1.5x 1015I
1.5x 1015I
Negligible
Negligible
1.8 x 101" cm3
5.4 x 101"cm3
Total Mass
1.1 x 1015g
0.2 x 1015g
Negligible
Negligible
4.3 x 109
1.7x 1010
% of Total
85
15
Negligible
Negligible
Negligible
Negligible
(porosity = 0.9; solids density = 2 g/cm3)
TOTAL 1.3x 10lag
interconversion of forms is an insufficiently understood
possibility. Distinction between dissolved and partic-
ulate forms and careful control of ambient parameters
are minimum requirements for future investigations. A
general consideration of trace metal impacts must, at
this stage, still consider total concentrations.
Physical limits of the system: The physical limits of
the system chosen for consideration cannot be
arbitrary. The boundaries delineate the sources, sinks,
and internal processes of the system. Judicious choice
of boundaries can greatly simplify a study. For example,
if As from a point source in a harbor is to be considered
(e.g. see Holm, et al. 1979), a logical choice of
boundaries may be the sides of the harbor and the
outlet flow to the point where the As concentration is
diluted to background lake water levels. In some cases,
it might make more sense to follow the system to
where sediment concentrations decrease to normal
regional values. Partial physical barriers often make
convenient boundaries; Lake Michigan is often divided
into a southern more industrialized and a northern less
industrialized basin by the submerged ridges running
between Milwaukee and Grand Haven.
Collection and Analysis of Information
After the scope of an impact assessment has been
defined, the next task is accumulating needed
information. Obtaining as much relevant information
as possible and abstracting the most useful is
frequently beneficial. Already available information
may need to be supplemented by investigation to obtain
new information. All information must be analyzed for
quality and significance to determine its importance to
the assumed goals of the assessment.
Distribution in the system: Knowing the distribution
of a trace metal in different reservoirs can be useful.
Representative values for southern Lake Michigan are
given in Table 5. Note that values given are estimates
of variable reliability and the original references should
be consulted for further information. Assumptions
made ignore the higher lead concentrations in near-
shore waters and the uneven distribution of lead in the
sediments (Edgington and Robbins, 1976; Int. Joint
Comm. 1978). Although crude, the estimates indicate
the water column is the most significant reservoir in
terms of mass of lead. High concentrations of lead are
found in the sediments and significant amounts are
buried below the top active zone (here taken as the top
4 centimeters), but the overlying water actually
contains more lead. Little information on Pb levels in
Lake Michigan biota, especially plankton and bacteria,
is available. Pb bioaccumulation factors are generally
in the range of 102 to 103(Ketchum, 1972, Callahan, et
al. 1979). The relatively insignificant mass of biota
makes it a negligible reservoir. Similarly, large
concentrations of lead can be found in the surface
microlayer (Elzerman and Armstrong, 1979), but its
relatively small volume makes the mass of lead in this
reservoir insignificant.
Of course, the fact that a reservoir contains a small
fraction of the total lead does not mean lead
interactions are insignificant. For example, alkylation of
lead has been observed (Wong, et al. 1975; Chau, et al.
1979), but whether it occurs primarily in the
sediments, water column, or organisms is unknown.
Metal alkylation in the environment is known to be a
complex process (Ridley, et al. 1977).
Forms present: Information on thermoaynamically
expected forms for most trace metals in aqueous
systems is now readily available and , although subject
to the limitations of thermodynamic predictions, very
useful. Numerous computerized models, like REDEQL,
MINEQL, and GEOCHEM, are widely used for species
prediction (Jenne, 1979), and summaries of metal
speciation are available (e.g. Callahan, et al. 1979).
Kinetic controls on concentrations and the nature of
particulate phases present are not as easily predicted.
Soluble forms of Pb are influenced greatly by pH and
the anions present, especially carbonate (Callahan, et
al. 1979; Davis, 1976). The free ion (Pb+2) dominates
only at low pH. At intermediate pH's, species such as
PbCOS, PbOH+, and Pb(OH)2 are important. Sufficient
Cf of SO42 can lead to the presence of PbCT and
PbSO° Pb appears to be strongly complexed by
organic materials and readily adsorbed by particles in
natural waters. Particulate lead in the water column is
probably mostly adsorbed or in organisms. Elzerman, et
al. (1979) found evidence of significant fluxes of high
Pb concentration (>5,000 /ug/g) atmospheric particles
to the lake and almost all Pb in the surface microlayer
to be in particulate form. More recent evidence
(Elzerman, et al. 1980) suggests that some of the Pb
quickly dissolves from the aerosol in the lake water and
-------
HEALTH-RELATED PROBLEMS
383
may then be readsorbed by particles in the surface
micro-layer. In sediments, PbSO4 (PbS if S"2 is present),
PbCOa, and complexed or adsorbed Pb are likely solid
components.
Boundary fluxes: An important consideration, es-
pecially to contemplated management measures, is the
sources and sinks of a trace element to the system. The
major source of Pb to Lake Michigan seems to be the
atmosphere (Edgington and Bobbins, 1976; Cogley,
1974; Eisenreich, 1980). Atmospheric inputs to Lake
Michigan have been extensively studied (e.g. Win-
chester and Nifong, 1971; Klein, 1975; Gatz, 1975;
Eisenreich, 1980, summarized in Table 6). Estimates of
the fraction of the total Pb input attributable to the
atmosphere range from 60 to almost 100 percent.
International Joint Commission estimates (1978) of Pb
inputs to the whole lake are 190 metric tons per year
from point sources and 1,670 metric tons per year from
nonpoint sources (including the atmosphere). The only
substantial sink of Pb from the system is the sediments,
where it seems to be essentially immobile (Edgington
and Robbins, 1976; Cogley, 1974).
Mass balances and residence times can be estimated
from boundary fluxes and reservoir loadings (Bowen,
1975). For example, Edgington and Robbins (1976)
estimated the flux of Pb to southern Lake Michigan in
1972 to be 270 metric tons per year (240 from the
atmosphere). Approximately the same flux to the
sediments was found. A simple calculation for the
water column based on this flux and the mass of Pb in
the water column reservoir (Table 5) indicates a
residence time of many thousands of years, but
processes like transport to the sediments in particles
make the actual residence time much less (Brewer and
Spencer, 1975). As a result, increases in Pb inputs to
the lake are exhibited as increased concentrations of
Pb in recent fine-grained sediments of active deposi-
tional regions (Int. Joint Comm. 1978; Edgington and
Robbins, 1976; Leland, et al. 1973).
Table 6. —Trace metal inputs to Lake Michigan (Eisenreich, 1980.)
Element
Pb
Zn
Ca
Cu
Mn
Cd
Fe
Mg
Al
Co
Tributaries
180
500
18,400
230
850
12
36,000
8,800
17,500
15
Shore Erosion Atmosphere
103 kg/yr
240 640
1,800 1,100
280,000 79,800
540 120
4,100 640
75 11
2,300 2,770
250,000 15,500
75,000 4,990
700 25
% Atmsopheric
60.
32.
21.
13.
11.
11.
6.7
5.7
5.1
3.3
Internal cycling and transport: Biological, chemical,
and physical transformations within an aquatic system
are numerous, complex, and not well understood,
especially in relation to the rates at which they occur
(see Stumm and Baccini, 1978, for review). Physical
transport mechanisms, including advection, dispersion,
and diffusion, have been more successfully described
and probably account for the major horizontal
movements of Pb within the lake. Vertical dispersion
coefficients tend to be much smaller than horizontal
dispersion coefficients (Thibodeaux, 1979); therefore
the major vertical movement of Pb probably results
from sorption by particles or organisms followed by
sinking (Brewer and Hoa, 1979; Ferranti and Parker,
1977; Brewer and Spencer, 1975). Internal cycling and
transformation affect the forms of Pb present, the
exposure of organisms to Pb, and the removal of Pb
from the system. Leland, et al. (1973) have reviewed
factors controlling the high concentrations of Pb in
sediments.
Integration of Information and Objectives
Regardless of the quality and quantity of information
accumulated, final integration of information and
objectives is necessary, often difficult, and likely to
require non-scientific decisions. Frequently, substan-
tial modeling efforts have been undertaken to improve
interpretation and implementation of results. Modeling
has not always been successful, especially for
comprehensive problems, but useful approaches like
EXAMS (Lassiter, et al. 1979) and transfer models (e.g.
Wiersma, 1979) have developed. Models are probably
most useful for sensitivity analyses (estimating
responses to various perturbations) and as part of a
general systems approach (Pojasek, 1977; Ketchum,
1972). Throughout the investigation, and particularly
when arriving at conclusions, the applicability of the
data to intended uses must be reviewed. For example,
information presented on Pb in Lake Michigan does not
define the impact of alkylated Pb compounds on Lake
Michigan fish, but it can be used in predictive models to
evaluate different remedial measures to control Pb
inputs to the lake. Control of atmospheric inputs would
be most significant, but difficult to achieve. Heidtke, et
al. (1980) have shown that control of rural and urban
runoff sources would be expensive and only of limited
usefulness. Consequently, lead inputs to the lake are
likely to continue at significant levels and the need for
further assessment of the fate and effects of Pb in the
lake remains.
REFERENCES
American Chemical Society. 1978. Cleaning our environ-
ment: a chemical perspective. Washington, D.C.
Anderson, P. D., and L. J. Weber. 1976. The multiple toxicity
of certain heavy metals: additive actions and interactions.
Pages 263-282 in R. W. Andrew, et al. eds. Toxicity to biota
of metal forms in natural water. Int. Joint Comm., Windsor,
Canada.
Andrew, R. W., K. E. Biesinger, and G. E. Glass. 1977. Effects
of inorganic complexing on the toxicity of copper to Daphnia
magna. Water Res. 11:309.
Andrew, R. W., P. B. Hodson, and D. E. Konasewich, eds.
1976. Toxicity to biota of metal forms in natural water. Int.
Joint Comm., Windsor, Canada.
Bowen, H. J. 1975. Residence time of heavy metals in the
environment. Toronto.
Brewer, P. and D. W. Spencer. 1975. Minor element models
in coastal waters. Pages 80-96 in T. M. Church, ed. Marine
chemistry in the coastal environment. Am. Chem. Soc. Ser.
18. Washington, D.C.
-------
384
RESTORATION OF LAKES AND INLAND WATERS
Brewer, P. G. and W. M. Hoa. 1979. Oceanic elemental
scavenging. Pages 261-274 in E. A. Jenne, ed. Chemical
modeling in aqueous systems. Am. Chem. Soc. Ser. 93.
Washington, D.C.
Brown, V. 1976. Aspects of heavy metal toxicity in fresh
water. Pages 59-76 in R. W. Andrew, et al. eds. Toxicity to
biota of metal forms in natural waters. Int. Joint Comm.
Windsor, Canada.
Burrell, D. C. 1974. Atomic spectrometric analysis of heavy-
metal pollutants in water. Ann Arbor Science, Ann Arbor,
Mich.
Cairns, J., D. Messenger, and W. F. Calhoun. 1976.
Invertebrate response to thermal shock following exposure
to acute sublethal concentrations of chemicals. Arch. Fuer
Hydrobio. 77:164.
Callahan, M. A. et al. 1979. Water related fate of 129 priority
pollutants. 1. EPA 440/4-79-029a. U.S. Environ. Prot.
Agency.
Cantillo, A. Y., and D. A. Segar. 1975. Metal species
identification in the environment: a major challenge for the
analyst. Pages 27-31 inf. C. Hutchinson, ed. Proc. Int. Conf.
Heavy Metals in the Environment. Toronto, 1975.
Chau, Y. K., et al. 1979. Determination of tetraalkylead
compounds in water, sediment, and fish samples. Anal.
Chem. 51:187.
Cogley, A. C. 1974. Large-scale mass balance for lead in
southern Lake Michigan. Water Resour. Center Res. Rep.
85. University of Illinois at Urbana-Champaign.
Dagani, R. 1980. Aquatic toxicology matures, gains im-
portance. Chem. Eng. News June 30:18.
Davies, P. H. 1976. The need to establish heavy metal
standards on the basis of dissolved metals. Pages 93-126 in
R. W. Andrew, et al. eds. Toxicity to biota of metal forms in
natural water. Int. Joint Comm., Windsor, Canada.
Dodge, E., and T. L. Theis. 1979. Effect of chemical speciation
on the uptake of copper by Chironomus tentans. Environ.
Sci. Technol. 13:1287.
Edgington, D. N., and J. A. Robbins. 1976. Records of lead
deposition in Lake Michigan sediments since 1800. Environ.
Sci. Technol. 10:266.
Eisenreich, S. J. 1980. Atmospheric input of trace metals to
Lake Michigan. Water Air Soil Pollut. 13:287.
Elzerman, A. W. 1976. Surface microlayer-microcontaminant
interactions in freshwater lakes. Ph.D. Thesis. University of
Wisconsin, Madison.
Elzerman, A. W., and D. E. Armstrong. 1979. Enrichment of
Zn, Cd, Pb, and Cu in the surface microlayer of Lakes
Michigan, Ontario, and Mendota. Limnol. Oceanogr.
24:133.
Elzerman, A. W., D. E. Armstrong, and A. W. Andren. 1979.
Particulate zinc, cadmium, lead, and copper in the surface
microlayer of southern Lake Michigan. Environ. Sci.
Technol. 13:720.
Elzerman, A. W., T. J. Overcamp, and L. F. Owens. 1980. Rate
of solution of ambient aerosols: initial results for lead.
Presented at Am. Chem. Soc. fall conf., San Francisco, Aug.
25-29.
Ferrante, J. G., and J. I. Parker. 1977. Transport of diatom
frustrules by copecod fecal pellets to the sediments of Lake
Michigan. Limnol. Oceanogr. 22:92.
Gatz, D. F. 1975. Pollutant aerosol deposition into southern
Lake Michigan. Water Air Soil Pollut. 5:239.
Gruber, D., et al. 1979. Initial testing of a recent biological
monitoring concept. Jour. Water Pollut. Control Fed.
51:2744.
Heidtke, T. M., D. J. Schef low, and W. C. Sonzogni. 1980. U.S.
heavy metal loadings to the Great Lakes: Estimations of
point and nonpoint contributions. Great Lakes Environ.
Plan. Study Contrib. 12. Great Lakes Basin Comm., Ann
Arbor, Mich.
Holm, T. H., et al. 1979. Heterogeneous interactions of
arsenic in aquatic systems. Pages 712-736 in E. A. Jenne,
ed. Chemical modeling in aqueous systems. Am. Chem.
Soc. Symp. Ser. 93. Washington, D.C.
International Joint Commission. 1978. Environmental
management strategy for the Great Lakes system. Windsor,
Canada.
1980. Pollution in the Great Lakes basin from land
use activities. Rep. to U.S. and Canadian governments.
Windsor, Canada.
James, A. ed. 1978. Mathematical models in water pollution
control. John Wiley and Sons, New York.
Jenne, E. A., ed. 1979. Chemical modeling in aqueous
systems. Am. Chem Soc. Symp. Ser. 93. Washington, D.C.
Ketchum, B. H., ed. 1972. The water's edge: Critical problems
in the coastal zone. Mass. Inst. Technol. Press, Cambridge.
Klein, D. H. 1975. Fluxes, residence times, and sources of
some elements to Lake Michigan. Water Air Soil Pollut. 4:3.
Lassiter, R. E., et al. 1979. Research and development of an
exposure analysis system (EXAMS). Environ. Res. Lab., U.S.
Environ. Prot. Agency, Athens, Ga.
Lee, G. F., and A. W. Hoadley. 1967. Biological activity in
relation to the chemical equilibrium composition of natural
waters. Am. Chem. Soc. Adv. Chem. Ser. 67:319.
Leland, H. V., S. S. Shukla, and N. F. Shimp. 1973. Factors
affecting distribution of lead and other trace elements in
sediments of southern Lake Michigan. Pages 89-129 in P.
C. Singer, ed. Trace metals and trace metal-organic
interactions in natural waters. Ann Arbor Science, Ann
Arbor, Mich.
Magnuson, V. R., et al. 1979. Relationships of activities of
metal-liquid species to aquatic toxicity. Pages 635-653 in E.
A. Jeanne, ed. Chemical modeling in aqueous systems. Am.
Chem. Soc. Ser. 93. Washington, D.C.
Mancy, K. 1971. Instrumental analysis for water pollution
control. Ann Arbor Science, Ann Arbor, Mich.
Marshall, J. S. and D. L. Mellinger. 1978. An in situ study of
cadmium stress in a natural zooplankton community. Pages
316-330 in J. H. Thorp and J. W. Gibbons, eds. Energy and
environmental stress in aquatic systems. Natl. Tech. Inf.
Serv. CONF.-771114.
Martell, A. E. 1971. Principles of complex formation. Pages
239-264 in S. P. Faust and J. C. Hunter, eds. Organic
compounds in aquatic environments. Marcel Dekker, New
York.
McKim, J. M. 1977. Evaluation of tests with early life stages of
fish for predicting long-term toxicity. Jour. Fish Res. Board
Can. 34:1148.
Pojasek, R. B. 1977. Mobilization of metal contaminants in
surface water supplies. Pages 329-344 in R. B. Pojasek, ed.
Drinking water quality enhancement through source
protection. Ann Arbor Science, Ann Arbor, Mich.
Quinby-Hunt, M. S. 1978. Monitoring metals in water. Am.
Lab. Dec. 1978:17.
Ridley, W. P., L. J. Dizikes, and J. M. Wood. 1977.
Biomethylation of toxic elements in the environment.
Science 197:329.
Riedel, B., and G. Christensen. 1979. Effect of selected water
toxicants and other chemicals upon adenosine triphosphate
activity in vitro. Bull. Environ. Contam. Toxicol. 23:365.
Schnitzer, M., and S. U. Kahn. 1972. Humic substances in the
environment. Marcel Dekker, New York.
-------
HEALTH-RELATED PROBLEMS 385
Singer, P. C., ed. 1973. Trace metals and trace metal-organic
interactions in natural waters. Ann Arbor Science, Ann
Arbor, Mich.
Stumm, W., and P. Baccini. 1978. Manmade chemical
perturbation of lakes. Pages 91-126 in A. Lerman, ed.
Lakes: chemistry, geology, physics. Springer-Verlag, New
York.
Stumm, W., and H. Bilinski. 1973. Trace metals in natural
waters: difficulties of interpretation arising from our
ignorance of their speciation. Adv. Water Pollut. Res.
1973:39.
Stumm, W., and J. J. Morgan. 1970. Aquatic chemistry.
Wiley-lnterscience, New York.
Thibodeaux, L. J. 1979. Chemodynamics. John Wiley and
Sons, New York.
Torrey, M. 1976. Environmental status of the Lake Michigan
region. 3. Chemistry of Lake Michigan. Argonne Natl. Lab.
ANL/ER-40(3), Argonne, III.
U.S. Environmental Protection Agency. 1979. A small oil spill
off west Falmouth. EPA 600/9-79-007.
Vernberg, F. J. 1978. Multiple factor and synergistic stresses
in aquatic systems. Pages 726-747 in J. H. Thorp and J. W.
Gibbons, eds. Energy and environmental stress in aquatic
systems. Natl. Tech. Inf. Serv. CONF-771114.
Water Pollution Control Federation. 1980. Literature reviews:
effects of pollution on freshwater invertebrates, macro-
invertebrates, and freshwater fish. Jour. Water Pollut.
Control Fed. 52:1670.
Whitfield, M., and D. R. Turner. 1979. Critical assessment of
the relationship between biological, thermodynamic, and
electrochemical availability. Pages 657-682 in E. A. Jenne,
ed. Chemical modeling in aqueous systems. Am. Chem Soc.
Symp. Ser. 93. Washington, D.C.
Wiersma, G. B. 1979. Kinetic and exposure commitment
analyses of lead behavior in a biosphere preserve. Rep. 15.
Monitor. Assess. Res. Centre, Chelsea College, University of
London.
Winchester, J. W., and G. P. Nifong. 1971. Water pollution in
Lake Michigan by trace elements from pollution aerosol
fallout. Water Air Soil Pollut. 1:50.
Wong, P. T. S., Y. K. Chau, and P. L. Luxon. 1975. Methylation
of lead in the environment. Nature 253:263.
1978. Toxicity of a mixture of metals on
freshwater algae. Jour. Fish. Res. Board. Can. 35:479.
Wood, J. M., et al. 1975. Metabolic cycles for toxic elements in
the environment. Pages49-68 inT. C. Hutchinson, ed. Proc.
Int. Conf. Heavy Metals in the Environment. Toronto.
Zitko, V. 1976. Structure-activity relations and the toxicity of
trace elements to aquatic biota. Pages 9-32 in R. Andrew, et
al., eds. Toxicity to biota of metal forms in natural water. Int.
Joint Comm. Windsor, Canada.
ACKNOWLEDGMENTS
The support of Clemson University during the preparation
of this manuscript is gratefully acknowledged.
-------
386
WATERBORNE GIARDIASIS
EDWIN C. LIPPY
Health Effects Research Laboratory
U.S. Environmental Protection Agency
Cincinnati, Ohio
ABSTRACT
Waterborne disease outbreaks occur in the United States at an average rate of 35 per year and
have been increasing since 1950. Giardiasis is a waterborne illness of major concern and
accounted for 36 outbreaks affecting 16,000 people in the past 15 years. The number of cases is
thought to be considerably underestimated. The most serious outbreaks have occurred in
communities using reservoirs as a source of water supply. Water systems using reservoirs as a
source of supply depend on long storage time and environmental factors to reduce turbidity and
permit microbiological die-off to occur. Therefore, minimum treatment is provided and in many
cases chlormation is the only protection. Giardia cysts, unlike other pathogenic agents, can survive
for long periods of time in the water environment and overcome the natural barriers provided by
reservoir storage. They are also more resistant to chlorination. Chlorine dosage must be increased
to inactivate cysts, creating the undesirable side effect of producing additional trihalomethanes.
Increased chlorination provides a solution for an acute disease outbreak but may contribute to a
chronic health problem.
BACKGROUND
Waterborne disease outbreaks in the United States
have been recorded in the literature since 1920. For an
historical perspective, the annual number of outbreaks
based on averaging data for 5-year periods, is shown in
Figure 1. A peak occurred during 1936-45 and it
appears that we are approaching that peak in the
current 5-year period. The trend has been increasing
since the 1950's and has caused some concern. The
largest outbreak recorded occurred in 1926 in Detroit,
Mich, and affected 45,000 to 50,000 people with acute
gastrointestinal illness. The most recent large outbreak
occurred this year in Texas and affected over 8,000
people.
Waterborne giardiasis outbreaks are relatively new
in this country with the first reported in 1965. A total of
36 have been reported through 1979, as shown in
Table 1. Two large outbreaks not included in the table
are now thought to be waterborne giardiasis. One
occurred in Portland, Ore. in 1954-55 affecting an
estimated 50,000 persons and the other in Boulder,
Colo, in 1972 with 300 cases. The Portland outbreak
occurred at a time when there was considerable doubt
about the pathogenicity of Giardia even though it was
detected in stools of those experiencing symptoms that
are now known to be typical of the illness. The Boulder
outbreak was termed inconclusive because the
organism was n.t identified in the water system and
young adults were the group mainly affected, con-
tradicting a belief that the high risk group was children.
In approximately 55 percent of the reported waterborne
outbreaks, the causative agent is not determined so the
true incidence of waterborne giardiasis outbreaks
could be considerably more than the number shown in
Table 1. There is also general agreement that many
outbreaks occur that are not investigated and con-
sequently not reported.
Table 1. —Waterborne giardiasis outbreaks U.S. (1965-1979).
Period
1965-69
1970-74
1975-79
TOTAL
Outbreaks
2
12
22
36
Cases
142
361
15,407*
15,910
"4,000 cases estimated for preliminary 1979 data
Figure 1. — Waterborne disease outbreaks. United States
1920-79.
-------
HEALTH-RELATED PROBLEMS
387
The location of giardiasis outbreaks is shown in Table
2. They occur predominantly in the mountainous areas
of the country, particularly in the Rocky Mountains,
New England, and the Pacific Northwest. Many of
these areas depend on sources of water supply that are
not influenced by wastewater discharges and, con-
sequently, minimal water treatment measures are
employed. In many cases, chlorination is the only
treatment used. Chlorination, as presently practiced by
most water utilities, is not effective in inactivating
Giardia cysts.
Table 2. — Location of waterborne outbreaks of giardiasis,
U.S. (1965-1979).
State
Colorado
Utah
New Hampshire, New York,
Oregon
California, Montana, Vermont,
Washington
Arizona, Idaho, Pennsylvania,
Tennessee
Outbreaks/
State
11
4
3
2
1
Total
11
4
9
8
4
36
THE ORGANISM AND THE DISEASE
Giardia is a single-celled protozoan organism with
two distinct stages in its life cycle. In a human or
animal host it exists in an active or reproductive stage,
termed a trophozoite. Outside the host it exists in an
inactive or cyst stage. The cycle of infection for a
human begins when the cyst is ingested either through
contaminated food or water. When the cyst enters the
upper small intestine, it excysts to the trophozoite stage
and attaches to the epithelial lining where reproduction
by binary fission, or splitting, occurs. Attachment to the
lining of the small intestine apparently interferes with
the digestive process, causing watery diarrhea,
bloating, abdominal pain, and cramps. Eventually,
trophozoites detach from the lining and begin to encyst
in the small intestine and are excreted in the feces in
the cyst stage. This is the form in which they are
usually found in the feces; however, in some cases of
severe watery diarrhea they are identified in the
trophozoite stage. In the cyst stage they can survive for
long periods in the water environment and have been
reported as surviving for more than 3 months.
Feeding studies in the early 1950's determined the
number of cysts required to produce an infection in
humans. Prisoners who volunteered for the study were
given cysts in their drinking water and it was found that
10 cysts were sufficient to cause infection. It is of
interest to note that a person who is infected will shed
an average of 15 x 106 cysts per gram of feces. A
normal human stool weighs about 150 grams so the
potential for one carrier in transmitting the disease is
tremendous. (This translates to a capability of one
person being able to contaminate a 50 mg reservoir to
an infectious dose of 10 cysts/I.)
In this country, the illness is treated with three drugs.
Quinacrine is normally the drug of choice and has a
cure rate of about 95 percent.
PROBLEMS RELATED TO WATER
SUPPLY
Twenty-nine of the 36 waterborne outbreaks of
giardiasis were related to using inadequately treated
surface water. As previously noted, the outbreaks
occur where sources of water supply are not influenced
by wastewater discharges and this explains the
minimal treatment. In most cases, the communities
relied on reservoirs for raw storage to permit natural
forces to reduce turbidity and microbial populations.
Treatment consisted of chlorination to destroy bacteria
so the systems complied with drinking water standards
for coliform bacteria.
Until 1975, not much attention was paid to
waterborne outbreaks of giardiasis. During that year a
large outbreak affecting nearly 5,000 people occurred
in Rome, N.Y. Prior to that time 14 outbreaks in small
water systems had affected about 500 people. The
Rome outbreak was notable not only because it
affected so many people, but because it lasted for 6
months. There were no wastewater discharges in the
watershed; however, a few malfunctioning septic tanks
were discovered in the 200 square mile drainage area.
Only four of 257 samples collected from the
distribution system during the outbreak showed
evidence of coliform contamination.
One year later, an outbreak at Camas, Wash, affected
600 people. This outbreak was especially notable in
that Giardia cysts were, for the first time, easily
identified in raw and finished water, and the organism
was found in beaver living near the water intake. In the
past 2 years, beaver have been implicated in 8 of 12
outbreaks by identifying cysts in beaver feces, or by
necropsy of animals trapped from the watershed. It has
also been determined through feeding experiments
that cysts isolated from beaver feces can infect humans
and, conversely, cysts from human feces can infect
beaver.
The management concerns that have developed
because of waterborne outbreaks of giardiasis included
control of beaver in reservoirs and watersheds
especially where a water supply may be affected,
sampling and laboratory methodology to identify the
cyst and determine whether it is still viable, and
treatment technology to remove and/or inactivate the
cyst.
Control of Beaver
Controlling beaver in reservoirs and watersheds may
be difficult, depending on how it is done. It also cannot
or need not be applied in every situation. It obviously
cannot be applied in large watersheds because of the
cost and logistical requirements, and beavers need not
be controlled in locations where water treatment
facilities are adequate. Control is considered to be a
viable alternative in relatively small watersheds and
reservoirs, especially where water treatment is
marginal.
Control does not imply destroying the animal. Where
this is the method used, trouble can be expected from
an aroused public. A more acceptable method is to live-
trap the animals and relocate them where the impact of
water quality is minimized. If suitable holding facilities
-------
388
RESTORATION OF LAKES AND INLAND WATERS
are available, it would be desirable to hold them in
captivity until the infection can be cured. Other
measures include replacing deciduous trees around
reservoirs with coniferous species, thereby removing
the beavers' food supply and discouraging habitation.
It is surprising how upset people become at the
thought of dead-trapping or sacrificing a few beaver.
They somehow have perceived the beaver as a
harmless animal rather than a member of the rodent
family which can be quite destructive in a reservoir and
around waterways. They are indiscriminate in selecting
trees for felling and can create turbidity and trash
problems in reservoirs. It is not difficult to determine
whether beaver are present in reservoirs as stripped
cuttings are usually piled up against spillways, dam
faces" and around outlet structures.
Sampling and Laboratory Methodology
Regulatory agencies and water utilities have for
years relied on coliform bacteria as indicators of
drinking water quality and safety. Absence of the
coliform group from drinking water normally indicates
the water is free of pathogenic or disease-causing
bacteria and virus. Laboratory tests for coliforms are
relatively simple, inexpensive, and universally ac-
cepted. To produce water to meet the coliform
standard, simple chlorination is used to effectively
reduce bacterial counts to acceptable levels. In
practice, 30 minutes' contact time is normally needed
for chlorine to react with and destroy coliform bacteria;
by current standards the water then is safe to drink. For
the concentration of chlorine and contact time usually
employed to achieve compliance with the coliform
standards, disease-causing bacteria (Shigella, Sal-
monella, enterotoxigenic E. coli) and virus (Polio,
Coxsackie, ECHO) are destroyed or inactivated. Therein
lies the paradox.
Giardia cysts are much more resistant to chlorine
than the coliform group of bacteria. They are not
inactivated at concentrations and contact times used by
many water systems. Traditional monitoring of drinking
water for coliforms as required by law will indicate that
the water is safe when it may be contaminated with
Giardia cysts.
Current sampling methodology for cysts requires
filtering large volumes of water (2,000 liters) through
an orlon fiber filter tube to trap the cysts. Cysts are then
removed from the tube through a laborious laboratory
procedure which requires cutting the fibers from the
tube and stirring the mass in a blender. The fluid
expressed from this procedure is taken through a
flocculation process to separate the organic and
inorganic particulate matter also trapped on the filter,
to obtain a suspension hopefully containing the cysts.
The suspension is centrifuged and a few drops of the
centrifugate are examined microscopically for the
presence of cysts. Besides being time-consuming, the
method is only about 6 percent efficient. Because of its
inefficiency, a negative finding does not indicate
absence of cysts, so its use in monitoring a water
system is somewhat limited. The methodology is quite
difficult and requires personnel with specialized
training who are not normally available to laboratories
that conduct routine analyses for water utilities. It also
has the inherent disadvantage of not being able to
determine viability of cysts viewed under the micro-
scope. This has important implications related to water
treatment.
Cyst viability or ability to produce infection is
determined by feeding the flocculated suspension
obtained in the laboratory to specific pathogen-free
beagle puppies. A positive test is development of an
infection in the pups which requires about 7 to 10 days.
Only one location in the United States has the
capability to conduct the feeding experiments. Need-
less to say, EPA is involved in extensive research to
address monitoring and laboratory methodology prob-
lems.
Water Treatment and Control Technology
The multiple barrier concept which requires placing
protective systems between the water consumer and
actual as well as potential sources of contamination is
of primary importance to insure the delivery of safe
drinking water. Reservoirs play an integral part in the
multiple barrier concept as they aid in reducing
turbidity and microbial populations to improve water
quality and assure a dependable supply of water during
low flow periods. These assets are perhaps used to an
unfair advantage by communities that rely on
chlorination as the only means of treatment, especially
so where Giardia contamination is a potential threat.
Conventional water treatment including chemical
coagulation and filtration is effective in removing
Giardia cysts from water and should be employed as an
additional barrier where the surface water and
reservoirs are used as a source of supply. Where
surface water without an impounding reservoir is used
as a source of supply, communities have had to install
filtration facilities to reduce turbidity and produce
drinking water of acceptable quality. As previously
mentioned, however, outbreaks of giardiasis have
occurred in areas where water supply sources are not
influenced by wastewater discharges and raw water
quality is better than average. In these areas,
dependence has been placed solely upon reservoirs
and treatment by chlorination which are not adequate
barriers during Giardia cyst challenge.
Where outbreaks have occurred under these con-
ditions, two emergency measures are implemented. A
boil water order is issued and chlorination is increased
to inactivate the cyst. Boiling water for 1 minute
destroys the cyst but the extent to which the
community complies with the order is not generally
known. With energy costs now an important considera-
tion in the family budget, people resent the added cost
and burden of boiling their water. Increasing chlorina-
tion to destroy the cysts produces objectionable side
effects of creating an unaccustomed taste and odor
problem and may contribute to the building of
trihalomethane concentrations. EPA has promulgated
regulations to control trihalomethanes because of their
cancer-causing potential, so by increasing chlorination
to control an acute health problem, a potential chronic
health situation may be created.
The long term solution is to fully implement the
multiple barrier concept and provide adequate treat-
ment with a desirable adjunct of controlling the beaver
population in the watershed where it is practical.
-------
HEALTH-RELATED PROBLEMS 389
Adequate treatment includes conventional unit opera-
tions capable of removing cysts, and disinfection at an
acceptable concentration over a sufficient period of
time for inactivation to occur, as an added measure of
protection. The State of Colorado adopted regulations
in 1977 requiring communities using surface water as
a source of supply to provide treatment to remove
Giardia cysts. Other States have similar regulations
under consideration.
SUMMARY
Outbreaks of giardiasis are increasing in frequency
and severity and occur predominantly in communities
using surface water as a source of water supply. The
most serious outbreaks have occurred in communities
depending on reservoirs, with minimal treatment
facilities as barriers of protection. While these barriers
are adequate to produce water complying with the
coliform standard, they are inadequate under Giardia
cyst challenge.
Beaver have been increasingly associated in out-
breaks as carriers of Giardia cysts. Feeding studies
have shown that the cyst infecting beaver also infects
humans and the converse is true. Control of beaver is
appropriate and should be incorporated into the
multiple barrier concept in certain situations.
There are problems in monitoring and laboratory
methodology which require additional research. Im-
proved sampling techniques, laboratory processing of
samples, and a methodology for determining cyst
viability are required. EPA is addressing these needs
through in-house studies and research grants and
contracts.
Water treatment technology is available to reduce
and inactivate Giardia cysts and this technology should
be applied to prevent outbreaks of giardiasis.
REFERENCES
Craun, G. F. Waterborne outbreaks in the United States,
1971-78. (Submitted for publ.)
Eliassen, Ft., and R. H. Cummings. 1948. Analysis of
waterborne outbreaks, 1938-45. Jour. Am. Water Works
Assoc. (May).
Gorman, A. E., and A. Wolman. 1939. Waterborne outbreaks
in the United States and Canada and their significance.
Jour. Am. Water Works Assoc. 31:225.
Taylor, A. Jr., etal. 1972. Outbreaks of waterborne disease in
the United States, 1961-70. Jour. Infect. Dis. 125:3.
U.S. Environmental Protection Agency. 1979. Waterborne
transmission of giardiasis. Proc. Symp. EPA-600/9-79-001.
Natl. Tech. Inf. Serv., Springfield, Va.
Weibel, S. R., et al. 1964. Waterborne disease outbreaks,
1946-60. Jour. Am. Water Works Assoc. 56:8.
-------
390
RESIDENTIAL WELL WATER QUALITY IN
WISCONSIN INLAND LAKE COMMUNITIES
GEORGE R. GIBSON, JR.
Environmental Resources Unit
University of Wisconsin-Extension
Madison, Wisconsin
ABSTRACT
Older inland lake communities in Wisconsin are more likely than many areas to have degraded
water supplies. Many homes sit on sandy soils with high water tables, have shallow wells close to
their own or a neighbor's septic system, and may not comply with State sanitation or well codes.
Concern over this condition led to an investigation of groundwater quality at two lakefront
residences suspected of having failing septic tank systems. A statewide Extension education
program for lake communities was also created which includes screening tests of home drinking
water for at least coliform bacteria, nitrate-nitrite-N, and chlorides. Test results suggest that about
half of the lake or river community wells tested appear to be contaminated to some degree and that
better residential well water management is needed.
BACKGROUND
Inland lake communities (particularly older ones) in
the Great Lakes Region comprise a relatively high risk
area for degraded drinking water supplies. Soils in
these areas are often sandy with a rapid rate of
groundwater movement and a high water table. Wells
frequently shallow, driven sand points; and the on-site
waste disposal system aging and inadequate. In
addition, the popularity of lake property leads to small
lots and crowded conditions around the lake. These
circumstances, combined with possible non-compli-
ance with State well and sanitation codes, could
contaminate the shallow groundwater and create a
health risk when it is tapped by residential wells.
The data presented in this paper are derived from two
sources: a 1979 lake research project, and a University
of Wisconsin-Extension information program initiated
in the summer of 1978. The program involves a
screening test of residential well water samples
collected by inland lake homeowners. Tests conducted
include total coliform, chlorides, and nitrate-nitrite-N.
The samples are analyzed at local university facilities
and test results returned to the participants at a
community meeting. At the meeting the parameters
tested are explained, local hydrologic relationships
discussed, and advice provided for the protection and
use of the groundwater resource. Residents whose
samples indicate unusually high chlorides or nitrate-N
or have coliform or general bacterial colonies are
advised to have their water further analyzed by a
certified laboratory and to determine whether their
water systems meet the minimum standards of the
State well code.
PREVIOUS STUDIES
Ellis (1971) and Ellis and Childsf 1973) demonstrated
the groundwater intrusion and lateral movement of
septic system effluent at Gull Lake and at Houghton
Lake, Mich. This point was also made by Brandes
(1975). However, these studies were primarily based
on the measurement of nutrients and other chemical
constituents of effluent. Brandes did observe fecal
coliform movement up to 17 meters from drainfields
and Mack (1972) reported the transport of both polio
viruses and coliform bacteria from a restaurant
drainfield to its well water supply 300 feet away. (This
transport distance may have been facilitated by the
fractured limestone underlying the study area.) In
1979, the Office of Inland Lake Renewal, Wisconsin,
Department of Natural Resources, applied the Ellis and
Childs study design to a series of residences on a
central Wisconsin lake(Knauer, 1980). At two of these
sites, bacterial samples were taken from a series of
shallow monitoring wells placed in a line between the
septic system disposal field and the lakeshore. At both
sites, the number of colonies per 100 ml in the
groundwater increased and peaked between the septic
tanks and the lakeshore. In both instances, and for all
parameters measured, these peaks were noticably
greater than the background level measured in a
control well located upgradient from the septic tank on
each property. Parameters measured were: total
coliform, fecal coliform, fecal streptococcus, and
Pseudomonas aeruginosa.
While this cursory investigation and the references
cited certainly do not indict on-site waste disposal
techniques, they do raise the issue of possible
wellwater contamination from this source, particularly
in lake communities or areas of shallow groundwater,
poorly constructed wells, and sandy soils.
-------
HEALTH-RELATED PROBLEMS
391
INITIAL STUDY OF BACTERIAL
MOVEMENT IN GROUNDWATER FROM
SEPTIC TANK SYSTEMS
The Ellis and Childs technique was adapted to a lake
community in south central Wisconsin. A series of
three variable depth monitoring wells was installed at
roughly 1/3 distance intervals between the septic tank
system and the lakeshore at two homesites suspected
of having defective septic tank systems. The soils at
both sites are of mixed glacial origin, but are mostly
sandy clay. Each sampling site consisted of three
adjacent wells set 15 centimeters, 60 centimeters, and
120 centimeters below the normal groundwater table.
Figure 1 illustrates the placement of the wells. The
primary intent of sampling these wells for bacteria was
to determine if such contamination could be demon-
strated for a Wisconsin lake community.
Sampling was accomplished by first pumping each
well to waste, waiting 3 hours, and then pumping out
the sample, using sterile tubing. A sterile collection
bottle was inserted in the line, and suction provided by
hand pump. The tube was lowered to well bottom and
then withdrawn a few inches to reduce sediment
intake. Fresh sterile equipment was used at each well
site. These precautions were, however, compromised
to some extent since other investigators had used the
same wells to collect chemical water quality samples.
Figure 1. — Schematic representation of the test well
system installed at two lake front residences on a lake in
south central Wisconsin. Each cluster of three test wells (A,
B, C) consists of a well set at the normal water table level
(shallow), one set 2 feet deeper (middle), and one 4 feet
deeper (deep). The control well (ch'k) sampled was
equivalent to the middle depth.
Immediately after collection, the samples were
returned to the laboratory for analysis. All analyses
were by the most probable number (MPN) technique, in
accordance with Standard Methods for the Analysis of
Water and Wastewater(Am. Pub. Health Assoc., 1975).
The results of the investigation are presented in Fig-
ure 2.
At both sites, the bacterial parameters in all cases
increased between the drainfield and downgradient
lakeshore. These increases consistently exceeded
background levels as indicated by counts taken from a
control well at each site located upgradient from the
septic tank system and test wells. Later that summer
attenuated Type I polio virus was introduced to the
septic tank at site 2 as a tracer. The virus was later
recovered from all of the test wells at the site and from
the adjacent lake water and sediments (Stramer,
1980). Household water supplies apparently were not
threatened by either of these septic systems since their
wells were located elsewhere on the property, but the
groundwater was being contaminated locally and the
coliform, streptococcus, and Pseudomonas organisms
remained alive and culturable at distances up to and
exceeding 30 meters downgradient from the drain-
fields.
DATA GATHERED FROM THE
EXTENSION INFORMATION DRINKING
WATER PROGRAMS
While this study was admittedly rudimentary in
nature, it was concluded that this evidence in con-
junction with the body of existing literature was
sufficient to justify further development and expansion
of a pilot drinking water information program for
Wisconsin lake communities. An assessment of data
gathered from testing the residential drinking water in
these communities would itself support or contradict
the presumption of contamination risks peculiar to lake
settings. Subsequently, 351 well water screening tests
of residential well water systems have been conducted
in 15 lake and river communities in Wisconsin (Fig-
ure 3).
The samples are analyzed for total coliform bacteria,
nitrate-nitrite-N, chloride, and (variably) pH, specific
conductance, hardness, and iron, in accordance with
Standard Methods for the Examination of Water and
Wastewater (Am. Pub. Health Assoc., 1975). In field
settings, a portable kit augments standard laboratory
techniques. All coliform analyses are conducted on
100 milliliter samples using the membrane filtration
technique employing Millipore Corp. apparatus and
disposable 5 centimeter diameter petri dishes and
filters. Sample bottles are sterilized by autoclaving;
field equipment is either packaged and sterilized by
autoclave, or sterilized for reuse at the site by
ultraviolet irradiation. The culture medium used is MF
Endo Medium (8BL). All coliform test samples are
filtered and incubated within 8 hours from time of
collection. Confirmation is based upon the presence of
metallic sheen colonies observed when the incubated
cultures are read at 24 and 36 hours. Chemical tests
are completed wihtin a maximum of 48 hours. Blanks
and known reference samples are run with each set of
chemical tests, and sterile water blanks and a positive
control sample of contaminated water with each set of
bacterial tests cultured. Usually no dilutions or
replicates are made.
Water quality data compiled for the 15 programs are
presented in Table 1. Total coliform colonies in the
tests ranged from negative to "too numerous to count."
Cultures which obviously generated colonial growth,
but were lacking the classical metallic sheen were
reported as atypical and are shown in column five. As a
further indication of possible health risk, nitrate-nitrite
nitrogen concentrations greater than 10 mg/l (EPA
criteria re: risk of methemoglobinemia in infants) are'
shown in the next column. Chloride concentrations
greater than background levels in the sample com-
munity are a possible tracer indicating effluent in the
groundwater; they are also shown. Common sources of
high chlorides are sewage, especially septic tank
effluents containing water softener back flushes, road
salt runoff and infiltration, animal waste leaching,
fertilizer leaching, and natural salt deposits. The final
column of total suspect samples is a summation of
those water samples which exceeded the levels for any
or more of the constitutents as shown in the table.
-------
392
RESTORATION OF LAKES AND INLAND WATERS
Values and percentages are not cumulative across the
table because a given sample may reveal more than
one suspect characteristics.
Fecal Streptococcus
- 100.000
' 10,000
<100
Cti'k s.T. ABC Lake
PseudQmonas aeruqinosa
-1,000
Ch'k S.T. A
Site 1
Site 2 / /.
Ch'k S.I. ABC Lake
Fecal Conform
Ch'k S.T. A B C Lake
Ch'k S.T. A „ C Lake Ch'k S.T. ABC Lake
Ch'k S.T. ABC Lake Ch'k S.T. ABC Lake
Figure 2. — Graphic presentation of the data collected from
a single samplingof groundwater test wells at two lake front
residences in south central Wisconsin. Bacteria
concentrations are shown on the vertical axes, scales are
not consistent. The horizontal axes repeat the linear
arrangement of the test wells from the up-gradient control
(ch'k) well, past the septic tank system (S.T.) to three equally
spaced monitoring well clusters (A, B, C) each of which
consists of three separate depth wells — shallow at the
normal water table depth; medium 2 feet deeper; and deep 2
more feet deep. "Lake" indicates a sample taken at the lake
edge in line with the wells. X=1 data from shallow wells,• =
data from medium wells and also the control and lake
observations, A - 1 data from deep wells. Dashed lines
indicate missing samples because the shallow wells were
often dry. Dashes were also used between groundwater
data and lake data. The trend is for bacteria concentrations
to increase down gradient from the septic tank systems.
Figure 3. — Wisconsin counties in which Extension drinking
water quality information programs have been held, 1978-
1980.
-------
HEALTH-RELATED PROBLEMS
393
Table 1: Residential well water quality data collected in fifteen Wisconsin lake or river communities from summer, 1978 to and
including summer, 1980. The number of suspect water samples is less than the sum of values for each row because a given sample
may be suspicious for more than one parameter.
Lake
Commu-
nity
A
B
C
D
E
F
G
H
I
J
K
L
M
N
O
N
24
14
26
34
32
29
14
32
27
6
24
26
27
16
20
351
No. of samples
(100 ml) with
"total coliform"
colonies
2( 8%)
3(21%)
4(15%)
5(15%)
3( 9%)
9(31%)
2(14%)
7 (22%)
1 ( 4%)
1 (17%)
4(17%)
1 1 (42%)
3(11%)
8 (50%)
0
63(18%)
No. of samples
with "TC"col-
onies greater
than one
(EPA Drinking
Water Criteria)
1 ( 4%)
2(14%)
2( 8%)
3( 9%)
1 ( 3%)
6(20%)
1 ( 7%)
5(16%)
1 ( 4%)
0
2( 8%)
7 (27%)
3(11%)
6(38%)
0
40(11%)
No. of samples
with bact. growth
but not identified
as "TC"
5(21%)
2(14%)
14(54%)
12(35%)
14(44%)
Not reported
1 ( 7%)
12(37%)
3(11%)
1 (17%)
14(58%)
22 (85%)
16(59%)
13(81%)
_2(10%)
131 (37%)
No. of samples
with NOz-NOa-N
concentration
greater than
10 mg/l
1 ( 4%)
0
0
1 ( 3%)
0
0
0
1 ( 3%)
0
2 (33%)
2( 8%)
2( 8%)
0
0
0
9 (2%)
No. of samples
with Cl concen-
tration notice-
ably greater than
local background
levels (usually
10 mg/l
4(17%)
4 (29%)
6 (23%)
3( 9%)
4(12%)
17(59%)
1 ( 7%)
9 (28%)
2( 7%)
2 (33%)
3(13%)
6 (23%)
3(11%)
1 ( 6%)
3(15%)
68(19%)
Total
Suspect
Samples
8 (33%)
8 (57%)
17(65%)
16(47%)
19(59%)
22 (76%)
3(21%)
19(59%)
6 (22%)
4 (66%)
18(75%)
23 (88%)
17(63%)
15(94%)
_5(25%)
200 (57%)
DISCUSSION
Some of the variations involved in this approach
include: samples are collected by the householders
themselves; they are not instructed to flame the faucet
before collection, but do purge the line; considerably
less time elapses between collection of the sample and
culturing than when samples are mailed to a
laboratory; and incubation time is a minimum of 24
hours with a second inspection of the culture plates
again at 36 hours. This additional 12 hours of
incubation was elected because experience has shown
that small colonies may be missed when the plates are
incubated for only 24 hours. Colonial development
after the initial 24 hour incubation may be stressed
coliform bacteria or some other form, such as
Pseudomonas or Aeromonas. It has been practical so
far to verify such subsequent results.
The likely effect of these variations is to produce a
greater frequency of coliform and/or atypical colonial
growth than might be reported by conventional
sampling and laboratory methods. When the multiple
tube dilution technique is employed by a lab, further
deviation might occur because most laboratories do not
confirm non-gas forming, but nonetheless cloudy
fermentation tubes which might be masking coliform
occurrences.
This entire question of atypical colonial growth
remains to be addressed. The EPA Microbiological
Methods for Monitoring the Environment (Bordner, et
al, 1978) states". . . groundwaters frequently contain
high total counts of bacteria with no coliforms. Such
waters pass Interim Drinking Water Regulations but
technical judgment must conclude these are not
acceptable as potable waters." Wisconsin health
standards similarly related only to the presence of the
coliform indicator. While non-coliform colonial growth
may or may not reflect pathogenicity, it does suggest
that the well water has been recently exposed to the
surface soil or atmospheric environments. The cause of
such growth may be as innocuous as incidental organic
or construction contamination; but it could also reflect
a broken seal or too shallow a well of only marginal
safety, which may be a pathway for contamination from
surface runoff. The admission of ubiquitous soil
organisms, not coliform in nature, to a drinking water
supply may not violate present health codes, but it
certainly should be reason for concern since some of
these organisms have been shown to be opportunistic
pathogens. Such results, while relatively frequent, are
certainly not the norm for good quality drinking water.
Similarly, concentrations of nitrites and nitrates
and/or chlorides in a residential well considerably
higher than those of one's neighbors may not violate
present water quality standards, but should induce at
least further investigation.
The number of samples tested so far does not
demonstrate a correlation between the presence of
coliform colonies, nitrates, and/or increased con-
centrations of chlorides. This is not surprising since
these components have different mobility character-
istics in soils and need not necessarily be derived from
the same source. For example, one lake community
studied revealed extremely variable chloride data with
well samples varying from a chloride concentration of
less than 1 to more than 100 mg/l. There was no
spatial pattern to these results that could equate the
data to groundwater movement. There also was no
correlation with nitrates which might have implied
septic system sources of the chlorides. However,
nitrates would be low if the drainfield was in the water
table and nitrification inhibited by low oxygen. Road
salting is not reported to be a local practice and
apparently few homeowners have water softeners.
In the communities studied, many people had no idea
how deep their wells were; when they were installed;
-------
394
RESTORATION OF LAKES AND INLAND WATERS
whether they met current code specifications; or when
they had been tested for safety. Of those households
where this information was available, there was no
evident relationship between the depths of wells and
instances of suspect samples.
Motivation of the voluntary participants in the
program may influence the nature of the test results
obtained. There is no specific evidence available to
suggest just what prompts an individual householder to
participate in or to avoid the program (but it is evident
that very few of the homeowners encountered ever
have their water tested after their well is installed).
Some people signed up because they perceived the
chance to take advantage of a bargain in the free
screening test. Others may have done so from a sense
of responsibility for theirs and their friends' health. But
others have stated that they avoided the program
because they suspect a water quality problem and don't
want it identified if remedial expenses may be involved.
CONCLUSIONS
Data analysis from this small, non-random sample
indicates a surprising number of well water samples
containing coliforms, indeterminate bacterial popula-
tions, and/or unusually high nitrate-nitrite or chloride
concentrations. Of 351 screening tests in these
communities from 21 to 94 percent of the samples
tested were of suspicious water quality. The overall
average was 57 percent. Even if the abnormal and
unexplained chloride data for community "F" is deleted
from the data, 52 percent of the 351 samples remain
suspicious. While further investigation is indicated,
particularly with respect to extended incubation times
and the identification of atypical bacteria observed, it is
evident that better residential water supply manage-
ment is needed in lake communities. This improved
management should begin with efforts to encourage
homeowners to have their well water tested on an
annual or semiannual basis, and to avail themselves of
professional remedial services where indicated.
Hendricks, C. W., ed. 1978. Evaluation of the microbiology
standards for drinking water. EPA PB-297 119. Natl. Tech.
Inf. Serv., Springfield, Va.
Knauer, D. 1980. Unpublished manuscript. Off. Inland Lake
Renewal, Wisconsin Dep. Nat. Res., Madison.
Mack, W. N. 1977. Total coliform bacteria. Spec. Tech. Publ.
635. Am. Soc. Test. Mater.
Mack, W. N., Y. Lu, and D.B. Coohon. 1972. Isolation of
poliomyelitis virus from a contaminated well. Health Serv
Rep. 87:271.
McCoy, E., and W. A. Ziebell. 1975. The effects of effluents on
groundwater: bacteriological aspects. Paper presented at
the Nat. San. Foundation Conf. on Onsite Wastewater
Systems, Nov. 5-7.
Stramer, S. 1980. Personal communication. Based on Ph.D.
dissertation data on virus survival in septic tank systems.
Tentative publ. date, 1981. Dep. Bacteriol. Food Res. Inst.
University of Wisconsin, Madison.
U.S. Congress. 1974. Public Law 93-523, Safe Drinking
Water Act. Washington, D.C.
Wisconsin Department of Natural Resources. 1978. Safe
drinking water, Administrative Code, Chapter NR 109.
Madison.
REFERENCES
Allen, M. J., and E. E. Geldreich. 1975. Bacteriological criteria
for ground water quality. Ground Water 13.
American Public Health Association. 1975. Standard meth-
ods for the examination of water and wastewater. 14th ed.
Washington, D.C.
Bordner, R. J. Winter and P. Scarpino, eds. 1978.
Microbiological methods for monitoring the environment.
EPA-600/8-78-017. Natl. Tech. Inf. Serv., Springfield, Va.
Brandes, M. 1974. Studies on subsurface movement of
effluent from private sewage disposal systems using
radioactive and dye tracers (Interim Rep. I). Ontario Ministry
Environ., Toronto.
1975. Studies on subsurface movement of
effluent from private sewage disposal systems using
radioactive and dye tracers. (Part II). Ontario Ministry
Environ., Toronto.
Ellis, B. G. 1971. Gull Lake investigations: nutrient input
studies. Res. rep. Kalamazoo Nature Center. Dept. Crop Soil
Sci. Michigan State University, E. Lansing.
Ellis, B. G., and E. Childs. 1973. Nutrient movement from
septic tanks and lawn fertilization. Tech. Bull. 73-5. Mich.
Dep. Nat. Res., Lansing.
-------
395
PHOSPHORUS INACTIVATION: A SUMMARY OF
KNOWLEDGE AND RESEARCH NEEDS
G. DENNIS COOKE
Department of Biological Sciences
Kent State University
Kent, Ohio
ROBERT H. KENNEDY
U.S. Army Corps of Engineers
Waterways Experiment Station
Vicksburg, Mississippi
ABSTRACT
Phosphorus (P) precipitation and inactivation are lake improvement techniques which can lower
the lake P content sufficiently to retard algal growth. P precipitation has been shown to have at
least short-term effectiveness in improving lake trophic state. Longer term lake improvement is
more likely to occur through control of P release from lake sediments (inactivation); in one case,
significant improvement over a 5-year period occurred. A basis for determining a safe maximum
dose to lake sediments has been developed. There is a need to improve application procedures to
lower treatment costs. .Toxicity problems seem to be absent, but additional research is needed.
Long-term monitoring of P inactivation treatments to establish cost-effectiveness should be
encouraged.
INTRODUCTION
Phosphorus (P) precipitation/inactivation is a lake
improvement technique to lower the P concentration in
the water to a level sufficient to reduce standing crop
and/or productivity of planktonic algae. This is
accomplished by either removing P from the water
column (precipitation) or by controlling P release from
nutrient-rich sediments (inactivation). This treatment is
used to accelerate lake improvement after nutrient
diversion, particularly in those cases where internal P
release represents a significant contribution to the P
budget (Cooke, et al. 1977; Larsen, et al. 1979). While
this procedure, as it is now understood, may be
effectively used to remove material (e.g., phosphorus,
silt) from the water column, its principal objective is
long-term control of P release from lake sediments
through the sorptive action of a layer of colloidal
aluminum hydroxide on the sediments.
Our knowledge of this lake improvement technique
has been summarized by Cooke and Kennedy (1980a,
b), and the reader is referred to these works for details
of effectiveness, dose determination, application pro-
cedures, problems with toxicity, case histories, and
costs. Funk and Gibbons (1979) have also reviewed the
technique, including a useful discussion of costs.
The purposes of this paper are to briefly describe
what we know of this technique and what we need to
know.
PHOSPHORUS PRECIPITATION-
INACTIVATION
Phosphorus Precipitation
It is now well-known that the addition of aluminum
salts to lake waters, principally aluminum sulfate
(Al2(SO4>3), will bring about a prompt lowering of
phosphorus concentration. Removal of P can occur as
AIPO4 precipitate, by sorption to the surface of
AI(OH>3 polymers or floe (which is formed when
AI.2(SO4)3 is added to water with carbonate alkalinity),
or by entrapment of paniculate P in the AI(OH)3 floe.
Removal of particulate and inorganic P is dependent
upon the quantity of floe and upon pH (Eisenreich, et al.
1977; Cooke and Kennedy, 1980a). Removal of
dissolved organic molecules which contain P is
considerably less effective (Browman, et al. 1973,
1977; Eisenreich, et al. 1977), a factor which could be
of major significance in the prompt return of blue-
green algal blooms since some nuisance species of this
phylum synthesize alkaline phosphatases at low
inorganic P levels and thereby remove P from dissolved
organic molecules (Heath and Cooke, 1977).
At this writing, we are aware of 28 lake and pond
treatments to remove (precipitate) or inactivate P, 19 of
which have had the objective of P removal (see Table 1
of Cooke and Kennedy, 1980a, which lists all of these
treatments and summarizes their results). Of these 19,
only four appear to have any amount of published
information (Jernelov, 1970; Peterson, et al. 1973;
May, 1974; Funk and Gibbons, 1979). In each of these
cases it was clearly demonstrated that addition of
-------
396
RESTORATION OF LAKES AND INLAND WATERS
aluminum sulfate (ferric alum and aluminum sulfate in
the case of May, 1974) can effectively remove a large
percentage of P in the water column and bring about at
least short-term improvement in lake trophic state.
Documentation of any long-term lake improvement,
with the exception of Horseshoe Lake, Wis. (Peterson,
et al. 1973) is not now available for any of these P
precipitation or removal projects. At Horseshoe Lake, a
dose of 2.1 g Al/m3, as slurried aluminum sulfate, was
applied in May 1970. Figure 1 illustrates the reduction
in hypolimnetic P concentration which was achieved,
and the control of P through the summer of 1972.
According to Born (1979), P has never reached the
levels found before treatment, although it has
increased slightly each year since application.
JAN FEB MAR APR NUT JUNE JULY AIX> SEPT OCT NOV DEC
Figure 1. — (Cooke & Kennedy)
Control of Phosphorus Release from lake
Sediments
More recent treatments of lakes with aluminum salts
have been based on the recognition that lake
sediments can be an important source of P to the water
column (May, 1974; Kennedy, 1978; Cooke, et al.
1978; Gasperino and Soltero, 1 978; Knauer (this book);
Dominie (this book)), and that long-term lake improve-
ment may occur if this important P source is also given
long-term control. That is, aluminum salts are added
primarily to cover lake sediments with a P-sorbing floe
of A!(OH>3, and not for P removal from the water
column. A stated but as yet untested hypothesis of this
approach is that longer, more complete control of P
release will occur in proportion to the amount of
aluminum added.
Only two lakes, Dollar and West Twin in Ohio (Cooke,
et al. 1978; Kennedy, 1978; Cooke, 1979), have
received sufficient monitoring to substantiate the
conclusion that a large dose of aluminum sulfate to the
lake sediments, well in excess of that needed to remove
P from the water column, will bring about a long-term
improvement in lake trophic state. In July 1974, the
hypolimnion of Dollar Lake (A = 2.2 ha., Z = 3.9m was
treated with 9 metric tons of liquid aluminum sulfate.
One ton was added to the surface. In July 1975, West
Twin's (A - 34 ha., Z = 4.4m) hypolimnion was treated
with 100 metric tons of liquid aluminum sulfate (26 g
Al/m3). In both cases a procedure for adding a
maximum safe dose, described in Cooke, et al. (1978),
Kennedy (1978), Cooke and Kennedy (1980a), and
Kennedy and Cooke (this book) was followed. Figures 2
and 3 illustrate the results.
P content was sharply lowered and has remained so
for West Twin through 1980. Dollar Lake, a seepage
lake, had a slight increase in P content in 1978,
probably reflecting accumulation of P from cultural
sources. Table 1 lists the changes in trophic state
(using the Carlson, 1977, Trophic State Index), and
shows that the lakes are now in the mesotrophic range.
Significant lake improvement thus occurred, as
evidenced by higher transparency, lower total P
concentration, and decreased planktonic algae. Dollar
and West Twin have remained in this improved state
for 6 and 5 years (through summer, 1980), respectively.
East Twin also improved since it obtains most of its
water from West Twin. All lakes have more macro-
phytes than before, perhaps due to the higher
transparency.
At this writing, our knowledge about the long-term
effectiveness of both P precipitation and P inactivation
in controlling nuisance planktonic algae is not
complete enough to warrant extended conclusions
about longevity of effect, cost-effectiveness, or any
long-term detrimental changes. Phosphorus removal
seems to be effective for at least 2 years (Horseshoe
Lake), and P inactivation seems to bring about
significant lake improvement for at least 5 years.
Table 1. Mean (May-September) Carlson Trophic State
Index (from surface measures; adapted from Cooke, 1979,
based on Total Phosphorus
Year
1971
1972
1973
1974
1975
1976
1978
1980"
West Twin
57.58
62.75
61.36
59.84
55.85
52.36
44.25
46.81
East Twin*
53.68
58.91
56.48
58.89
57.14
56.62
47.27
46.81
Dollar
no data
no data
64.31
no data
50.22
50.65
47.79
46.81
Table 1 (continued). Mean (May-September) Carlson (1977)
Trophic State Index (Cooke, 1979), based on Secchi Disk
transparency.
Year
1968
1969
1971
1972
1973
1974
1975
1976
1978
1980"
West Twin
ND
50.0
61.0
48.3
43.2
49.9
51.4
46.7
46.4
43.5
East Twin"
ND
51.6
50.4
52.8
49.0
50.5
51.9
51.4
45.5
45.0
Dollar
66.3
ND
ND
ND
63.8
ND
50.7
47.9
47.8
48.2
"Untreated downstream reference lake
"Based on average of 2 measure-
ments, July 1980
-------
NUTRIENT PREVENTION AND INACTIVATION
397
At the Dollar-West Twin (Ohio) hypolimnetic applica-
tions, the Wisconsin system was modified to include
on-shore aluminum sulfate storage, and a distribution
pipeline from shore to a mid-lake platform where
application barges would return to re-fill. Another
adaptation of the Wisconsin system was that of
Dominie (this book), who used a three-compartment
tank truck mounted on a barge, to add a mixture of
aluminum sulfate and sodium aluminate to a soft-
water lake. May (1974; this book) added ferric alum to
ponds by suspending blocks of chemical in the water
and allowing them to dissolve.
Actual application to the lake is usually accomplished
by dividing the lake into small, well-marked sections of
known area and volume. Pre-application calculations
then permit the barge operator to know the volume of
chemical to be added to each section.
Thus, adequate application procedures have been
developed for the P removal-P inactivation lake
improvement method. However, these procedures
(Cooke and Kennedy, 1980a) are expensive (labor costs
range from 1 to 4 man-days per hectare for the six
treatments for which such data are available) and
tedious, and represent an obstacle to the general use of
this method. As well, equipment design and construc-
tion are often difficult. Some new and effective
procedures need to be developed.
Determination of Effective Dose of Aluminum
Many lake treatments with aluminum sulfate
apparently had no basis at all for the dose added.
Regrettably, much of this work is thus of little value in
attempting to understand the results and in developing
a systematic procedure.
Most of the projects in which P removal was the
primary objective had dose based upon AI/P ratios,
following the model from water treatment plants. The
amount of aluminum needed to remove the desired
amount of P was determined in jar tests and the total
aluminum dose was then obtained by multiplying
amount of aluminum needed by the P content of the
lake. This procedure produced low doses of aluminum
(0.5 to about 10 g Al/m3), and usually adequate P
removal.
The first stated attempt to control P release from
sediments was the Cline's Pond Project in April 1971
(Sanville, et al. 1976), in which 10 g Al/m3 as sodium
aluminate, plus HCI to prevent high pH, were added to
the pond surface. The treatment was successful for at
least 1 year in controlling P concentration, but algal
blooms then returned. While the authors suggest that
their short-lived success may have been caused by
continued external loading, sediment disturbance, and
breakup of the floe, an equally plausible explanation is
simply insufficient dose.
What is a sufficient dose to control P release from
lake sediments for a prolonged period? The answer is
not known but it is assumed that it is necessary to put
as much aluminum hydroxide over the sediments as
possible, short of causing adverse environmental
impact. Kennedy and Cooke (1974) and later Kennedy
(1978) were the first to suggest a basis for determining
a maximum dose for P inactivation. A maximum dose of
aluminum sulfate was defined as that amount which
150
100
SO
ISO
WO
50
0
EAST TWIN LAKE
' "*•*." •' ". •-.•."/.
"... •••"•• -v\ " " '. •'..'..*• —"• \'-' '"•'— ' ' ".. :\. .'.'.\- .
- . •" . * .«!/ *" " * *•* *"• '"*
1873 1974 1975 1976
"• - * *. j ALUM APPLICATION
' • ,v * .
"* . " •*'* "•"
" ."" * .". • " "• . ""
V«* '
WES1 TWIN LAKE •"." ,*^***" . ... "w*". •'*'*' '"
1973 1974 1»'B " **197C
<
O
0
z
£
*-
<
-"••—
' 1978
J.FJAAJAJ.J.A.S.O.Nifl
»- 1
o
o
z
"
.. ^.^
""1978
ISO
100
50 5
g
0 J
i
150 -0
100
SO
0
Figure 2. — (Cooke & Kennedy)
25
a.
220
<
K
O
015
DOLLAR LAKE
ALUM APPLICATION
-V"'—
••'.*•
MJJA'S'O'N'D'J'F'M'A'M'J j VSONDJ f MAMJ j ASONOJ FM»MJ j »SOND JFMAMJ j» SON D
1973 «74 «« «™ ""
Figure 3. — (Cooke & Kennedy)
-------
398
RESTORATION OF LAKES AND INLAND WATERS
Or, the maximum dose is that amount which can be
added until pH 6.0 is reached, a pH at which little
dissolved aluminum appears. Details of determining
dose are completely described for an actual lake
treatment in Cooke, et al. (1978), for any lake in Cooke
and Kennedy (1980a), and in additional detail by
Kennedy and Cooke (in this book). For softwater lakes,
the dose would be very small using this definition. In
such cases, as exemplified by the work of Dominie (in
this book), a mixture of sodium aluminate and
aluminum sulfate is added and pH and dissolved
aluminum remain well within acceptable limits.
Thus dose procedures for P precipitation or removal
and P inactivation are known, and are sufficiently well
understood to allow any user of the technique to apply
the proper amount of chemical.
Application Methods
The basic application system, which has been used
throughout the 12-year history of this method, was
designed for the 1970 Horseshoe Lake treatment
(Peterson, 1973). Aluminum sulfate was stored on
board a barge as a solid, mixed with water and then
pumped to a manifold trailing behind the barge, and
applied to either surface or deep waters (later projects)
as a slurry. The usual procedure has been to treat
surface waters for P removal, and to add aluminum to
hypolimnetic waters where P inactivation is the object.
One reason for a hypolimnetic application is that there
will be no exposure of epilimnetic and littoral biota to
the chemical. Also, less volume equals less cost.
Toxicity
Unlike herbicides, adding aluminum salts to lake
waters does not involve a substance whose toxicity to
plants is the factor in controlling algae. The procedure
works by lowering P concentration to a level which
controls productivity. However, adding aluminum to
lakes can pose a hazardous condition to the biota if lake
pH is sufficiently lowered (addition of aluminum
sulfate) or raised (addition of sodium aluminate) to
bring about solubilization of aluminum hydroxide and
an increase in dissolved aluminum. The toxicity of
aluminum has been reviewed by Burrows (1977) who,
along with Everhart and Freeman (1973), pointed out
that few investigations of aluminum toxicity have
considered the complex chemistry of aluminum in
water. The amount of dissolved aluminum which will
appear in the water after a treatement is pH dependent
and will vary from lake to lake as a function of lake
alkalinity and amount of the aluminum salt which has
been added. It is in this dissolved form that aluminum
could be hazardous. Cooke and Kennedy (1980a, b)and
Kennedy and Cooke (1974; this book) have thus
suggested that a maximum dose of aluminum sulfate
be one in which pH does not fall below 6.0 nor
dissolved aluminum increase above 50 ug Al/m3.
No toxicity to fish has been observed at any of the
full-scale lake treatments. However, no systematic
investigation has been made of this possibility, or of the
possibility of aluminum in fish muscle tissue. The
hazards posed here are small because of the very low
toxicity of aluminum to humans (Berry, et al. 1974).
Narf (1978) has investigated benthic invertebrate
populations following the several lake treatments in
Wisconsin and reports no adverse changes. Moffett
(1979) found a significant persistent (at least 3 years
after a hypolimnetic application) reduction in the H'
species diversity of planktonic microcrustacea after the
West Twin treatment. Causes and importance of this
observation are unknown.
Despite indications of little or no hazards after adding
aluminum salts to lakewater, few systematic investiga-
tions of aluminum toxicity to aquatic populations or
communities are available. It may be of particular
importance to note that acid precipitation may bring
about sufficient lowering of a treated lake's pH so that
previously insoluble aluminum hydroxide yields signifi-
cant quantities of dissolved aluminum. This could be of
real significance in soft water lakes.
RESEARCH NEEDS
The evidence strongly supports the belief that an
aluminum application for control of P release can be an
effective, and apparently long-lasting, method of
controlling algae. Procedures for determining dose and
applying the chemical are available. The toxicity of
aluminum to aquatic communities has not been
evaluated with the exception of Narf's work on benthos
and Moffett's study of planktonic microcrustacea. This
is a critical need since the technique is effective in
controlling algae, and it may be used with increasing
frequency and with maximum doses. It is important to
note here that our research need is not with toxicity to
laboratory organisms, but with the short- and long-
term impacts on the actual level of biological
organization to which aluminum salts are applied,
namely the community level. This means we need
studies of changes in community metabolism, mineral
cycling, species diversity, and other attributes of lake
state. The use of the LD or the Maximum Acceptable
Toxic Concentration (MATC) will not be very useful for
evaluating a lake improvement agent since it is the lake
which is treated, not animal species.
Rapid application techniques might make the P
inactivation method less costly. The use of suspended
blocks of alum, as described by May (1974; this book) is
one approach which could prove to be economical and
effective. As well, for somewhat larger systems, use of
shore-based high velocity hoses could reduce man-
power costs. For larger lakes, barge applications now
seem to be most cost-effective. There is a great need to
develop new, innovative methods of aluminum applica-
tion.
Is P inactivation cost-effective? We do not know the
answer to this question partly because we lack
published information about it. Also, support for long-
term monitoring of this and other lake restoration
techniques is very small. Accurate cost-benefit cannot
be accurately assessed without such monitoring.
SUMMARY
The effectiveness and longevity of P inactivation
following nutrient diversion has been demonstrated on
a few lakes, but the majority of demonstrations of
control of P release are new and most data are as yet
-------
NUTRIENT PREVENTION AND INACTIVATION
399
unreported. This symposium will add greatly to our
knowledge. It is clear at this point that if sufficient
aluminum is added, control of P release will occur and
remain effective at least 5 years, and that this
approach, rather than P precipitation, is the method of
choice for most situations.
A basis for determining dose for the P precipitation
and the P inactivation techniques has been developed
which gives adequate assurance that at least immedi-
ately toxic levels of aluminum will not be reached.
Aluminum application may pose a long-term hazard to
lake biota, but to date there has been little systematic
research about this possibility. At present, based on a
long-term monitoring of benthic invertebrates and
short-term monitoring of planktonic microcrustacea,
the data are equivocal about change in lake communi-
ties.
Costs and benefits of the technique need further
assessment. Funk and Gibbons (1979) and Cooke and
Kennedy (1980a) indicate a wide range of costs, but
without long-term monitoring, costs vs. effectiveness
cannot be stated.
It is recommended that future research include these
projects:
1. Monitoring of change in lake trophic state after P
inactivation and P precipitation treatments;
2. Studies of changes in field and experimental
(enclosures, microcosms) communities after an alumi-
num application;
3.Development of rapid aluminum application tech-
niques; and
4.Evaluation of the possible effects of acid rainfall
upon the pH and subsequent release of dissolved
aluminum in treated lakes.
REFERENCES
Berry, J. W., D. W. Osgood, and P. A. St. John. 1974.
Chemical villains: A biology of pollution. C. V. Mosby Co., St.
Louis, Mo.
Born, S. M. 1979. Lake rehabilitation: A status report.
Environ. Manage. 3:145.
Browman, M. G., R. F. Harris, and D. E. Armstrong. 1973.
Lake renewal by treatment with aluminum hydroxide. Draft
rep. to Wis. Dep. Nat. Resour. Madison.
1977. Interaction of soluble phosphate with
aluminum hydroxide in lakes. Tech. Rep. 77-05. Water
Resour. Center, University of Wisconsin, Madison.
Burrows, W. D. 1977. Aquatic aluminum: Chemistry,
toxicology, and environmental prevalence. CRC Crit. Rev.
Environ. Control 7:167.
Carlson, R. E. 1977. A trophic state index for lakes. Limnol.
Oceanogr. 22:361.
Cooke, G. D. 1979. Evaluation of aluminum sulfate for
phosphorus control in eutrophic lakes. OWRT Proj. No. A
053-OHIO. Final Rep. Ohio Water Resour. Center, Colum-
bus.
1980a Precipitation and inactivation of phos-
phorus as a lake restoration technique. Ecol. Res. Ser. U.S.
Environ. Prot. Agency (in press).
Cooke, G. D., and R. H. Kennedy. 1980b. State of the art
summary of phosphorus inactivation as a lake restoration
technique. Proc. Algae Manage. Control Workshop. U.S.
Army Corps of Engineers and U.S. Environ. Prot. Agency (in
press).
Cooke, G. D., et al. 1977. The occurrence of internal
phosphorus loading in two small, eutrophic, glacial lakes in
Northeastern Ohio. Hydrobiology 56:129.
1978. Effects of diversion and alum application on
two eutrophic lakes. EPA-600/3-78-033. U.S. Environ.
Prot. Agency.
Eisenreich, S. J., D. E. Armstrong, and R. F. Harris. 1977. A
chemical investigation of phosphorus removal in lakes by
aluminum hydroxide. Tech. Rep. Wis. Water Resourc.
Center 77-02. University of Wisconsin, Madison.
Everhart, W. H., andR.A. Freeman. 1973. Effects of chemical
variations in aquatic environments. Vol. II. Toxic effects of
aqueous aluminum to rainbow trout. EPA-R3-73-011 b. U.S.
Environ. Prot. Agency.
Funk, W. H., and H. L. Gibbons. 1979. Lake restoration by
nutrient inactivation. Pages 141-151. in Lake restoration,
Proc. Natl. Conf., Minneapolis, Minn. EPA-440/5-79-001.
U.S. Environ. Prot. Agency.
Gasperino, A. F., and R. A. Soltero. 1978. Restoration of
Medical Lake: Engineering design and preliminary findings.
BN-SA-807. Battelle Northwest, Richland, Wash.
Heath, R. T., and G. D. Cooke. 1977. The significance of
alkaline phosphatase in a eutrophic lake. Verh. Int. Ver.
Limnol. 19:959.
Jernelov, A. 1970. Aquatic ecosystems for the laboratory.
Vatten 26:262.
Kennedy, R. H. 1978. Nutrient inactivation with aluminum
sulfate as a lake restoration technique. Ph.D. Dissertation.
Kent State University, Kent, Ohio.
Kennedy, R. H., and G. D. Cooke. 1974. Phosphorus
inactivation in a eutrophic lake by aluminum sulfate
application: a preliminary report of laboratory and field
experiments. Conf. Lake Protect. Manage., Madison, Wis.
Larsen, D. P. et al. 1979. The effect of wastewater
phosphorus removal on Shagawa Lake, Minnesota: Phos-
phorus supplies, lake phosphorus and chlorophyll a. Water
Res. 13:1259
May, V. 1974. Suppression of blue-green algal blooms in
Bra id wood Lagoons with alum. Jour. Aust. Inst. Agric. Sci.
40:54.
Moffett, M. 1979. Changes in the microcrustacean communi-
ties of East and West Twin Lakes, Ohio, following lake
restoration. M.S. Thesis. Kent State University, Kent, Ohio.
Narf, R.P. 1978. An evaluation of past aluminum sulfate lake
treatments: Present sediment aluminum concentrations
and benthic insect renewal. Wis. Dep. Nat. Res., Madison.
Peterson, J. 0. et al. 1973. Eutrpphication control: Nutrient
inactivation by chemical precipitation at Horseshoe Lake,
Wisconsin. Tech. Bull. 62. Wis. Dep. Nat. Res., Madison.
Sanville, W. D., et al. 1976. Studies on lake restoration by
phosphorus inactivation. EPA-600/3-76-041. U.S. Environ.
Prot. Agency.
ACKNOWLEDGEMENTS
The development of portions of this manuscript was supported
by an Inter-Governmental Personnel Agreement between
Kent State University and the Corvallis Environmental
Research Laboratory of the U.S. Environmental Protection
Agency.
-------
400
CONTROL OFTOXIC BLUE-GREEN ALGAE IN FARM DAMS
VALERIE MAY
National Herbarium of New South Wales
Royal Botanic Gardens
Sydney, Australia
ABSTRACT
Light, warmth, and polluted water in a lake or inland water reservoir are likely to lead to
Cyanophyte algal blooms, which often include toxic strains. Preventing or reducing the entry of
soluble plant nutrients and carbohydrates is a necessary preliminary to the restoration of the water
quality. This includes controlling the quality of the drainage from both point and nonpoint sources.
Inactivation of the nutrients already present in the system is a second essential step. In small farm
dams this inactivation has been achieved by pre-summer dosing of the water with ferric alum
blocks suspended in the water. This treatment has proved to be a functional and ecologically
satisfactory method of reducing or controlling toxic algae in these cases. Similarly, for larger water
storages, liquid alum has proved effective, perhaps in conjunction with sodium aluminate, so that
aluminium hydroxide floe is produced near the sediment. Destratification or hypolimnion aeration
is another treatment producing good results.
CONDITIONS PROMOTING GROWTH OF
TOXIC BLUE-GREEN ALGAE
Slow moving or stagnant water frequently contains
soluble plant nutrients and carbohydrates. If it also is
subject to adequate light and warmth, then heavy
growths of both bacteria and algae are likely to develop.
The bacteria use the carbohydrates and oxygen,
increasing CO2 levels and decreasing 02 content of the
water. Anaerobic conditions (Sylvester and Anderson,
1964) free solutes and nutrients from the previously
enriched sediments. This includes a rise in the
concentration of phosphorus (Patrick and Khalid,
1974).
Algal growth becomes stimulated by an increase in
available soluble nutrients, including phosphorus,
obtained either from this bacterial action or from any
local direct drainage. The algae then proceed to
increase photosynthetic carbon uptake, resulting in an
increased pH value (King, 1970.)
These conditions usually occur in deep waters of
stratified dams, but may occur also during drought in
small dams, which then suffer severe evaporation and
consequent increase in mineral content, together with
lowered aeration.
Conditions for algal growth then may include low
light intensity (because of the depth of the water or its
turbidity), low 62 levels, and high CC>2 levels, which,
because of high pH values, may be of low availability.
It has been shown that each of these conditions
(Holm-Hansen, 1967; Stewart and Pearson, 1970;
King, 1970, respectively) favors the growth of
Cyanophytes as contrasted with green algae. Hence
massive growth of Cyanophytes might well be
expected.
Further, many species of blue-green algae develop
specialized gas vacuoles, which seem to allow vertical
mobility, so that the plant can benefit from both deeper
water (i.e., higher nutrient levels) and shallower water
(i.e., more light). Hence, these species are likely to be
among those which develop in excessive numbers.
A further property of certain blue-green algae is their
ability to fix atmospheric nitrogen, and this again aids
their growth under certain conditions.
Thus it seems that, starting with adequate light,
warmth, and enriched water, together with the
naturally good distribution mechanism of freshwater
organisms, it is almost inevitable that dams will be
subject to blooms of blue-green algae.
These blooms are perhaps unsightly, but our main
concern is that toxic strains occur among the most
prevalent bloom-forming species. In Australia these
species are Anacystis cyanea (Kuetz.) Dr. & Dail.
(Microcystis aeruginosa Kuetz.), Anabaina circinalis
Rabenh. (including A. flos-aquae as far as concerns
the literature of toxic Cyanophyte algae, see May
1980), and Nostoc spumigena (Mert.) Drouet (Nodularia
spumigena Mert.). These are known to kill horses,
cows, sheep, fowls, turkeys, laboratory guinea pigs and
mice, and probably various wild animals including birds
and fish. The lethal agent is an endotoxin which affects
the liver and can cause death, in some cases within a
few minutes.
Each of these bloom species may occur in the field
either in almost pure culture, or, in the case of
Anacystis and Anabaina, at times as codominants.
The particular species which develops in a dam may
depend on the preceding crop of algae growing there.
Thus Lam and Silvester (1979) record thai Microcystis
(Anacystis) inhibits the growth of Anabaina sp; they
suggest this might be caused by the production of
inhibitory extracellular products by Microcystis. Fitz-
gerald (1964) stresses the transient effect of products
such as these, since there is a very rapid development
of succeeding species of algae.
-------
NUTRIENT PREVENTION AND INACTIVATION
401
Figure 1 shows the periods of occurrence of
Anacystis and of Anabaina in Carcoar Dam, N.S.W.
Each year Anacystis was present before, and continued
after, the Anabaina. From these figures it seems that in
this case the conditions favoring the growth of
Anabaina are perhaps more limited than are those
allowing the growth of Anacystis. This variation could
link with nutritional requirements; Fitzgerald (1969)
showed that in co-dominant bloom of Anabaina sp. and
Microcystis sp. (Anacystis sp.) phosphorus was a
limiting factor for the growth of one genus while
nitrogen was for the other.
OBSERVATIONS
TO DATE
OCT. NOV. DEC. JAN. FEB. MAR. APR. MAY JUNE JULY AUG. SEPT.
ANABAINA
ANACYSTIS
Figure 1. — Total occurrences of Anacystis cyanea and
Anabaina circinalis in all collections from Carcoar Dam, New
South Wales, from October 1977 to May 1980.
CONTROL OF TOXIC BLUE-GREEN
ALGAE
Biological control would be the most satisfactory of
all treatments. Some work has indicated blooms could
be controlled by particular viruses (Safferman and
Morris, 1964). but this treatment has not so far been
practical.
Many different algicides have been used in the past
to rid water of unwanted blue-green blooms, 1 ppm of
copper sulfate possibly being the best and cheapest.
However, a bloom treated by an algicide can recur in as
short a time as a week, and an algicide may cause
much ecological harm either where used or down-
stream. Hence it would be preferable to use a method
of decreasing or preventing the occurrence of such
blooms, especially if this method were ecologically
acceptable.
To break sequence of development of a bloom
stimulated by the described conditions, it follows that
one would need to limit one or more of the following:
(1) light, (2) temperature, (3) carbohydrates, (4) high pH,
(5) low oxygen levels, or (6) high levels of nutrients.
1. Light could be excluded by covering the water with
lightproof material, but this is often impractical,
particularly for large areas. Plastic covers (Anon. 1979)
or even numerous floating black ping-pong balls have
been used to cover small areas.
2. It is quite impractical to reduce temperature in the
open, especially in large storage reservoirs.
3. Carbohydrates are so prevalent, provided by either
land or aquatic plants or animals, macrophytes, or
plankton, that it is only for excessive quantities such as
from sewerage or urban drainage that control is at all
practical. Here, of course, it is highly desirable in order
to restrict later intense de-oxygenation caused by
bacterial action.
4. High pH. Since any heavy photosynthetic growth
will increase pH, and since most blue-green algae
control seems directed to increasing the alternative
growth of green algae, control of this factor for long
periods seems likely to be impractical.
This leaves two areas — "low oxygen levels'' and
"high levels of soluble nutrients" — as possible
avenues to decrease blooms of Cyanophytes. Both
avenues have been tried.
5. Overcoming low oxygen levels. To overcome low
oxygen levels, a compressor is often used to release air
(or oxygen ) into the hypolimnion water, fairly close to
the reservoir bottom. The resultant artificial mixing or
destratification of a water body has been very widely
employed. This treatment not only reduces the amount
of nutrient solution coming from the sediment, but also
breaks the thermal stratification of the storage water, if
present, and makes the all-over water temperature
more uniform. Improvement in the quality of the bottom
water often overcomes additional problems, such as
high concentrations of iron and manganese, and
perhaps of hydrogen sulfide, which affect later water
use. Further, low oxygen concentrations in the water
released from low outlets in a dam wall could also
damage biological communities downstream. Destrati-
fication in Australia is relatively new (Bowles, et al.
1979).
An early report on the advantages of destratification
by aeration is given by Howard (1972). He reported the
widespread effect within a reservoir of aeration applied
at only one site. Tolland (1977) advises that, while
destratification is useful in overcoming existing water
quality problems, it is better used in a preventive rather
than curative role.
A submerged hypolimnion aerator which preserved
thermal stratification was described by Fast, et al.
(1975). This was designed to aerate the deep water but
did not introduce the rich inorganic nutrients of the
hypolimnion into the photic zone. It also had the added
advantage of maintaining a suitable habitat for cold-
water fish. Certain problems for fish, because of
nitrogen gas supersaturation, need to be guarded
against with this treatment, particularly with deeper
lakes.
The high costs of installing the apparatus make these
treatments unsuitable for small dams.
6. Prevention of high levels of soluble nutrients. The
first and obvious way to achieve nutrient prevention
has been to stop adding nutrients by way of drainage,
sewerage discharge, aerial spraying or verge-pollution,
i.e., to control all sources of added nutrients, of both
point and nonpoint origin.
When sewerage and/or polluted drainage is diverted
from a water storage, eutrophication in the latter
decreases markedly. This treatment, unfortunately, is
expensive and the benefit may be delayed if the
sediment is already highly polluted, so that it continues
to release nutrients to the water during periods of
anaeroby for some time after the diversion.
-------
402
RESTORATION OF LAKES AND INLAND WATERS
Purifying the discharge itself (usually sewage) is also
expensive, but usually not nearly as difficult as
diverting it. This also leads to obvious benefits.
Catchment areas should be freed as much as
possible from nearby manurial material, and distant
siting arranged for contaminating activities such as
abattoirs and piggeries. Leaving the verge of a water
storage clear of animal use (such as by using water
troughs rather than dam edges for stock drinking) is a
simple expedient, but worthwhile.
It is perhaps interesting to diverge here to an
observation about a river, the Peel River, N.S.W., which
I am at present studying with serial collections. I have a
number of observation stations along the length of the
river, on which a dam has recently been constructed.
During the time that the dam wall has been under
construction, and before the flow of the river was
affected at all, there was massive disturbance of soil at
the construction site. It was just below this that
Anabaina circinalis was recorded, not frequently, but
more often than at higher or lower stations. Doubtless
this was caused by increased nutrients from the soil,
even though conditions in the flowing river were not
conducive to heavy growth.
Controlling blooms by a constant natural removal of
nutrients from the water system by biota is obviously
an extremely satisfactory method. Constant removal of
fish containing the nutrient minerals, might seem
feasible. Seasonal blooms, however, do not provide a
year-round basis for a food chain. Perhaps alternative
food could be made available at other times of year.
An interesting study by Weir (1976) investigated the
possibility of developing macrophyte beds of the
angiosperms Typha and Eleocharis by planting zones of
them around a bloom-susceptible northern N.S.W. lake
and its connected swamp. These plants would remove
mineral nutrients, stabilize the mud banks, and reduce
erosion of soil. It was further suggested that they might
also be grown in artificial floating rhizome beds, and
could be managed by cropping routinely to maximize
the uptake and removal of nutrient. Hopefully, this
would also lead to some profit, as. the cropped weed
could be used for stock feed, since it was shown that
periodic cropping reduced fibre content and maintained
unusually high levels of phosphorus and sulfur in the
plant material. Thus, digestibility of the crop was
increased while removal of nitrogen and phosphorus
from the water continued at a high level.
A similar scheme is in use in Holland where bulrush
reed ponds allow treatment of sewage from holiday
camps with a periodic input of sewage at weekends;
the reeds are cropped annually and later burned to
return nutrients to the agricultural system (de Jong
1975, cited Weir, 1976). This sort of scheme is likely to
be particularly effective in controlling the entry of
nonpoint sources of nutrients from the edges of a
reservoir. It does not suffer from the disadvantage of
using a floating plant such as water hyacinth, where
pest quantities of growth can affect navigation.
Since extremely nutrient-enriched sediment may
prolong enrichment to the water, sometimes it has
been considered necessary either to dredge (Hudson
and Marson, 1970), de-silt (as is usual in farm dam
management), or cover it, as suggested by Theis, et al.
1978, who placed fly ash on the eutrophic sediment.
Another reported cover for the sediment was a nylon-
covered fabric supported on a polyurethane grid (Anon.
1972).
Next, attention has focused on which of the soluble
nutrients should be particularly reduced or inactivated
to control the growth of unwanted Cyanophytes.
Trace elements could be considered, but with any
normal catchment area it would be extremely difficult
to control contamination by such small quantities as
needed. Nicholas (1980), however, has suggested that
sodium tungstate may inhibit the growth of Anabaina
since this material is antagonistic to molybdenum, a
trace element necessary for growth of the alga.
Most work has suggested that phosphorus or
nitrogen limitation is likely to be the most functional
approach to controlling unwanted blooms. Of these
elements, controlling nitrogen, even where it may be a
limiting factor, seems impractical since some of the
toxic blue-green algae (e.g. Anabaina) are nitrogen-
fixing. It is useless to spend one's efforts reducing the
nitrogen level if the algae are going to replace this
element from atmospheric nitrogen.
Most studies seem to concur that controlling the
level of phosphorus in the water is a practical method
of reducing Cyanophyte blooms, and that reducing
phosphorus at wastewater treatment plants is a
necessary, practical, and economic plan. This is despite
it being known that surplus ("luxury") phosphorus can
be held by the plant and that sometimes heavy growths
are recorded when the phosphorus concentration in
the water is low.
My work was directed toward controlling phosphorus
in small farm dams, particularly during summer, since
this is where and when we in Australia suffer most
stock losses.
My first work on this project was on a small inland
dam at Braidwood, N.S.W. Here it appeared that
blooms (of Anacystis and /or Anabaina) occurred
whenever the level of phosphorus rose to 0.5 ppm or
higher (May, 1972). As a control measure I hoped to
use a chemical which, applied before the summer rise
in phosphorus levels, would combine with this
phosphorus before the unwanted algae could absorb it.
I applied alum and block ferric alum (Alumina ferric R.),
at a combined concentration of 200 mg/1. This
treatment is effective through the absorption of
phosphorus by the aluminum hydroxide floe, which is
formed when these treatment chemicals are added to
alkaline water (May and Baker, 1978). The blocks were
much easier to handle than is loose alum. They were
suspended in the water from floats, to prevent them
sinking into the underlying mud, and were replaced at
intervals. Following this treatment, for the first time in
5 years no blooms developed in the treated dam (May,
1974) although apparently some did occur in untreated
dams in the same district and in the same year. Later,
other dams with similar histories were treated similarly
and again no blooms developed (May, 1974).
Prior to the next field experiment, preliminary in vitro
investigations were carried out by Harvey Baker, my co-
author in some of this work (May and Baker, 1978).
This study indicated that, at the alkalinities usually
present in dams, treatment with aluminum sulfate or
ferric alum blocks reduced a range of initial phosphorus
-------
NUTRIENT PREVENTION AND INACTIVATION
403
levels to below 0.5 mg/1, a concentration deemed
critical.
For the field study a series of farm dams, all of which
had recently suffered from toxic algae, was studied
concurrently, Some dams were untreated and were
considered as controls, while others had a single pre-
summer treatment with ferric alum blocks, and the
third group of dams was treated this way at recurrent
intervals (May and Baker, 1978).
All of these dams showed a summer rise in the
concentration of total phosphorus, but the presence of
the alum reduced the phosphorus levels and also
decreased the incidence of algal blooms. The dosage
\jsed (50 mg/1) was evidently sufficient to reduce but
not eliminate bloom occurrence; a somewhat higher
dose is therefore now recommended (100 mg/1 — this
equals approximately 9 i/g/l aluminum). It appears that
the earlier (Braidwood) dosage (200 mg/1) would be
unnecessarily high.
The alum treatment not only reduced excessive
growth of these bloom algae, it also increased the total
number of algal species. In addition, the average
number of species per collection occurring in the
treated dams increased, i.e., there was greater algal
diversity. Evidently this alum treatment leads to
conditions more like those prevailing in pre-eutrophic
times.
This treatment is proving satisfactory in reducing
bloom formation in our numerous small dams, many of
which previously presented a repeated threat to our
stock. It is most satisfying to feel that this treatment, in
contrast to the use of algicides, is ecologically pleasing.
Indeed, this method of nutrient inactivation appears to
offer a promising routine for farm dam protection and
restoration.
Other speakers at our Symposium are to tell of the
dosing of larger volumes of water with alum, with or
without sodium aluminate, and the generally success-
ful effect this has had in reducing the level of
phosphorus and the incidence of Cyanophyte blooms. It
should be noted that besides reducing the concentra-
tion of phosphorus already in the water, or later freed
from the sediment, aluminum sulfate can also reduce
the pH of the water. If this drops below — and stays
below — 6, this condition alone makes the growth of
toxic algae less likely.
Any treatment, by alum or aeration, can have its
benefits masked by enriched incoming drainage. It
follows that these treatments are more effective where
such external sources of enrichment are absent or
minimal; their use in other cases is probably limited to
making the end result less damaging than it would
have been otherwise.
Hence, for good control one needs both lake
treatment and also drainage management, both
localized (point) and general (nonpoint).
In conclusion, lake restoration depends on a series of
processes:
1. a. Diverting all massive polluted water drainage
(such as sewage) from the storage area, or at least
cleaning this water of its pollutants, both mineral and
carbohydrate.
b. Reducing nonpoint (runoff) drainage enrich-
ment as far as practical. This source of pollution is of
varying relative importance in different waterways.
Any residual drainage of this sort is probably best
cleansed by using some sort of littoral harvesting.
2. The best method so far available to control the
annual enrichment of the water internally from an
already polluted sediment depends on the size of the
impoundment.
a. If the water impoundment is small, treatment
with suspended block alum at a dosage of 100 mg/1
seems best. This also is likely to cope with some
general runoff pollution.
b. If the water impoundment is large, either alum
treatment or destratification or hypolimnion aeration
should be suitable.
3. If these treatments prove inadequate, i.e., in dams
with excessively polluted water and sediment, then
direct treatment of sediment, such as by dredging or
covering the sediment, may also be necessary.
REFERENCES
Anonymous. 1972. Nylon-coated fabric used to rehabilitate
reservoir. Water Sewer. Works. Jan. 49.
Anonymous. 1979. Floating covers protect New England
reservoirs. Water Wastes Eng. Mar. 58.
Bowles, B. A., I. J. Rowling, and F. L. Burns. 1979. Effects on
water quality of artificial aeration and destratification of
Tarago Reservoir. Aust. Water Resour. Counc. Tech. Pap.
46. Aust. Govt. Publ. Serv. Canberra.
Fast, A., V. Dorr, and R. J. Rosen. 1975. A submerged
hypolimnion aerator. Water Resour. Res. 11:287.
Fitzgerald, G. P. 1964. The biotic relationships within water
blooms. Pages 300-306 in D. F. Jackson, ed. Algae and man.
Plenum Press, New York.
1969. Field and laboratory evaluations of
bioassays for nitrogen and phosphorus with algae and
aquatic weeds. Limnol. Oceanogr. 14:206.
Holm-Hansen, O. 1967. Recent advances in the physiology of
blue-green algae. Pages 87-96 in Environmental require-
ments of blue-green algae. Proc. Symp. Water Pollut.
Control Fed. Admin. University of Washington, U.S. Dep.
Inter.
Howard, R. G. 1972. Reservoir destratification improves
water quality. Reclamation Era 58:6.
Hudson, E. J., and H. W. Marson. 1970. Eutrophication: With
particular reference to the role of phosphates. Chem. Ind.
1449.
King, D. L. 1970. The role of carbon in eutrophication. Jour.
Water Pollut. Control Fed. 42:2035.
Lam, W. Y., and W. B. Silvester. 1979. Growth interactions
among blue-green (Anabaena oscillarioides. Microcystis
aeruginosa) and green (Ch/orella sp.) algae. Hydrobiologia
63:135.
May, V. 1972. Blue-green algal blooms at Braidwood, New
South Wales (Australia). New South Wales Dep. Agric. Sci.
Bull. 82.
1974. Suppression of blue-green algal blooms in
Braidwood Lagoon with alum. Jour. Aust. Inst. Agric. Sci.
40:54.
1980. The occurrence of toxic Cyanophyte blooms
in Australia. (In press.)
May, V., and H. Baker. 1978. Reduction of toxic algae in farm
dams by ferric alum. New South Wales Dep. Agric. Tech.
Bull. 19.
Nicholas, D. I. D. 1980. Mineral nutrient requirements and
utilization by algal flora of freshwater lakes. Aust. Water
Resour. Counc. Tech. Pap. 50. Austr. Govt. Publ. Canberra.
-------
404 RESTORATION OF LAKES AND INLAND WATERS
Patrick, W. H. Jr., and R. A. Khalid. 1974. Phosphate release
and sorption by soils and sediments. Effect of aerobic and
anaerobic conditions. Science 186:53.
Safferman, R. S., and M. E. Morris. 1964. Control of algae
with viruses. Jour. Am. Water Works Assoc. Sept. 1217.
Stewart, W. D. P., andH. W. Pearson. 1970. Effects of aerobic
and anaerobic condtions on growth and metabolism of blue-
green algae. Proc. R. Soc. Lond. B. 175:293.
Sylvester, R. O., and G. E. Anderson. 1964. A lake's response
to its environment. Jour. Am. Soc. Civil Eng. San. Eng. Div.
90:1.
Theis, T. L, and P. J. McCabe. 1978. Retardation of sediment
phosphorus release by fly ash application. Jour. Water
Pollut. Control Fed. 50:2666.
Tolland, H. G. 1977. Destratification aeration in reservoirs.
Water Res. Centre Tech. Rep. TR50. August.
Weir, J. 1 976. Natural and agricultural control of eutrophica-
tion. Proc. Aust. Water Resour. Counc. Symp. Eutrophica-
tion. Dep. Nat. Resour. Austral. Govt. Publ. Serv. Canberra.
-------
405
ALUMINUM SULFATE DOSE DETERMINATION
AND APPLICATION TECHNIQUES
ROBERT H. KENNEDY
U.S. Army Corps of Engineer Waterways Experiment Station
Vicksburg, Mississippi
G. DENNIS COOKE
Department of Biological Sciences
Kent State University
Kent, Ohio
ABSTRACT
Nutrient diversion alone does not always adequately reduce in-lake phosphorus concentration
because of nutrient-rich sediments. Certain lakes and reservoirs may continue to experience
nuisance algal blooms and will require additional restorative steps. The phosphorus
precipitation/inactivation technique is a procedure to remove phosphorus from the water column
and to control its release from sediments. The salts of aluminum have long been used in advanced
wastewater treatment to remove phosphorus and this technology was logically extended to lake
rehabilitation. However, specific guidelines for dose calculation and application to lakes and
reservoirs are lacking. An objective of all aluminum treatments, although often unstated, is to
control phosphorus release from bottom sediments. The suggested approach to dose
determination allows maximum application of aluminum to bottom sediments and thus
emphasizes long-term control of phosphorus recycling. Such a dose can be calculated directly from
the alkalinity of the water to be treated. Titration of several lake-water samples of varying alkalinity
will allow the establishment of the relationship between residual dissolved aluminum, alkalinity
and dose which can then be employed for lake-scale applications of alum to lakes and reservoirs.
Application equipment and procedures will depend on site characteristics and treatment
objectives. General equipment requirements include lakeside storage, a distribution pipe, and an
application barge and manifold. In addition to phosphorus removal and control, alum may be used
to meet other restoration objectives including the treatment of problem inflows and the reduction
of paniculate concentrations.
INTRODUCTION
Aluminum sulfate application for postdiversion
phosphorus control in eutrophic lakes is an increas-
ingly popular management tool. A logical adaption of
water and waste treatment technology, aluminum
addition provides a direct ameliorative methodology for
high phosphorus concentrations in lakes and small
reservoirs. However, the current popularity of the
method, attributable to its simplicity and the fact that it
produces immediate reductions in lake phosphorus
concentrations, has misled many lake managers to
view alum as a panacea. Despite 12 years and 25
reported uses (Cooke and Kennedy, 1980), aluminum
treatments remain more of an art than a well
understood technological alternative in lake restora-
tion.
Problems arise from the failure of many lake
managers to establish limnologically appropriate
objectives or to fully understand the aqueous chemistry
of aluminum. Although the importance of phosphorus
recycling from anaerobic sediments in delaying the
response of lakes to reduced external phosphorus
inputs (e.g. Cooke, et al. 1978, Larsen, et al. 1976) has
prompted the use of aluminum sulfate, only recently
has the control of sediment phosphorus been specifi-
cally identified as the primary treatment objective. The
use of aluminum to precipitate phosphorus from the
water column, still often identified as the primary
objective of many lake treatments, provides little more
than short-term relief. Sound decisions concerning the
relative importance of recycling from anaerobic
sediments must be made prior to treatment, and
treatment methodologies must concentrate on sed-
iment phosphorus control if long-term effectiveness is
to be realized.
Confusion concerning dose determination methods
for lake treatments is a related problem. Three
approaches, each dictated by treatment objective, have
been followed to date. The first involves incremental
additions of aluminum to aliquots of lake water until a
predetermined phosphorus removal efficiency is at-
tained (e.g. Peterson, et al. 1973). This dose, which is
then volumetrically scaled for lake application, clearly
optimizes treatment for phosphorus removal from the
water column. A second similar method, also optimiz-
ing dose for phosphorus removal, is modeled after dose
determination procedures employed in waste treat-
ment facilities. With pH controlled, aluminum additions
are made to constantly mixed lake water samples until
optimum phosphorus removal is achieved. The A1/P
molar ratio at maxium phosphorus removal and the
phosphorus concentration of the lake to be treated are
then used to determine lake-scale doses (e.g., Peterson
et al. 1974). These controlled laboratory conditions are
quite different than those which occur during lake
-------
406
RESTORATION OF LAKES AND INLAND WATERS
treatment, however, and can often underestimate
effective lake doses (Kennedy, unpublished). The third
method, initially employed by Kennedy (1978) and later
by Cooke, et al. (1978), maximizes aluminum input to
sediments, as dictated by the buffering capacity of the
overlying water, and thus emphasizes long-term
phosphorus control as a primary treatment objective.
If it can be demonstrated that nutrient-rich anaerobic
sediments will be a significant source of phosphorus
long after external sources are reduced, then alumi-
num treatments must be targeted against phosphorus
exchanges at the sediment-water interface. The
purpose of this paper is to provide dose determination
guidelines for such treatments and to suggest
appropriate application procedures and possible man-
agement strategies employing aluminum salts.
ALUMINUM CHEMISTRY
A considerable body of information concerning the
chemistry of aluminum is available. Since it is not the
purpose here to provide a comprehensive review of this
information, the reader should consult such reviews as
that of Hayden and Rubin (1974) for more detailed
discussions. However, a basic understanding of the
aqueous chemistry of aluminum, which is essentially
the chemistry of aluminum hydroxide, is necessary for-
making dose determination decisions.
The addition of aluminum salts (e.g. aluminum
sulfate) to water, which initially results in the hydration
of aluminum ions, is followed by a series of hydrolysis
reactions resulting in a decreased pH, and ultimately,
in the formation of low solubility aluminum hydroxide
precipitate. In natural waters a secondary consequence
is a decrase in carbonate alkalinity. Aluminum
hydroxide is amphoteric and thus is converted to the
soluble aluminate ion in basic solution.
Significant for dose determination is the fact that the
distribution of aluminum species is pH dependent
(Figure 1). While insoluble aluminum hydroxide
predominates between pH 6 and 8, soluble species
occur at higher (AI(OH«) and lower (AI(OH)* then AN-3)
pH. As aluminum is added to alkaline lake water,
hydrogen ion concentration increases, alkalinity is
titrated and pH decreases (e.g. Figure 2). Initially, at low
aluminum dose, pH changes are small. If solution pH
remains alkaline, the dissolved aluminum concentra-
tion (i.e. AI(OH4) will be predictably high. Further
aluminum additions decrease pH, favoring the forma-
tion of insoluble aluminum hydroxide precipitate and
dissolved aluminum concentrations decrease. As
aluminum additions continue, dissolved aluminum
concentrations again increase in acidic solution, with
AI+3 predominating below pH 4.
The importance of pH change is thus of direct
concern in dose determination since, in addition to
obvious consequences for exposed biota, the pH of
treated lake waters will dictate the concentration of
potentially hazardous soluble aluminum species, and
the quality and quantity of aluminum hydroxide
polymer. Although the toxicity of aluminum for aquatic
biota is poorly defined (Burrows, 1977), some
conservative estimates are available. Concentrations of
dissolved aluminum below 52 /ugAI/l had no obvious
effect on rainbow trout (Freeman and Everhart, 1971)
or salmon (Peterson, etal. 1974, 1976). These findings
prompted Kennedy (1978) and Cooke, et al. (1978) to
adopt 50 A/gAI/l as a safe upper limit for posttreatment
dissolved aluminum concentrations. Dose was defined
as the maximum amount of aluminum which would
still ensure low( /ugAI/l) concentrations. Since, based
on solubility, dissolved aluminum concentrations,
regardless of dose, would remain below 50 AigAI/l in
the range pH 5.5 to 9.0, a dose producing a
posttreatment pH in this range could also be
considered environmentally safe.
The formation of large aluminum hydroxide polymers
or floe, essential for the deposition of added aluminum,
would also be promoted in this pH range. Rapid
removal of floe from the water column is of concern
since prolonged suspension of fine aluminum hy-
droxide particulates further complicates the question of
toxicity and treatments targeted against specific areas
(e.g. anaerobic sediments) would be adversely im-
pacted by the mixing and dispersion of floe.
Figure 1. — pH — dependent distribution of aluminum
species (Eisenrich, et al. 1977).
10 15
ALUMINUM DOSE lmi)AI//l
Figure 2. — Changes in pH (closed circles) and post treatment
dissolved aluminum concentration (open circles) following
additions of aluminum sulfate to lake water (initial total
alkalinity of 98 mg CaCOs/M; pH 7.3).
-------
NUTRIENT PREVENTION AND INACTIVATION
407
PHOSPHORUS REMOVAL
The long-term effectiveness of alum treatments will
depend on the ability of deposited aluminum hydroxide
to retain phosphorus at the sediment/water interface
and thus curtail its internal recycling. Secondary,
short-term benefits can be realized if water column
phosphorus concentrations can be reduced during
treatment. Phosphorus removal can occur by coagula-
tion/entrapment of phosphorus-containing particu-
lates, precipitation of AIPCU (Recht and Ghassemi,
1970) or by sorption of phosphorus on the surfaces of
aluminum hydroxide polymers (Eisenreich, et al. 1977).
Successful removal of particulates will depend on the
quality of floe produced, which in turn is related to pH
and aluminum dose. Precipitation and sorption, both
influenced by pH and phosphorus concentration
(Stumm and Morgan, 1970), are apparently related
processes since sorption appears to occur by the
formation of aluminum-ion-phosphate bonds at the
surface of aluminum hydroxide polymers (Hsu, 1965).
At high phosphorus concentrations, such as those
encountered in wastewater facilities and low pH,
AIPO*4 is the predominant reaction product. However,
eutrophic lakes are characteristically alkaline and,
despite having biologically high phosphorus concentra-
tions, are relatively low in phosphorus. At low
phosphorus concentrations and higher pH, OH- reacts
more readily with aluminum than does phosphate and
thus aluminum hydroxide is the expected product.
Therefore, phosphorus removal from the water column
will be primarily by entrapment and sorption. Failure to
obtain maximal phosphorus removal at the stoichio-
metric AI/P molar ratio of 1.0 supports this suggestion.
This is particularly true in the case of lake treatments.
For example, maximum phosphorus removal from
Cline's Pond water occurred at AI/P molar ratios
ranging from 5.7 to 7.2 (Peterson, et al. 1976). AI/P
molar ratios in excess of 525 were required to achieve
90 percent P removal from unfiltered Lake Mendota
epilimnetic water (Eisenreich, et al. 1977).
Physical factors influencing phosphate sorption
include floe size and settling rate. As floe size
increases, specific surface area decreases. Increased
floe size also increases settling rates and thus
decreases contact time between the floe and the
surrounding lake water. Therefore, phosphorus re-
moval will be highest on the immediate area of
aluminum addition since pH decreases would be
greatest here and floe size would be small. As floe size
increases during mixing and settling, phosphorus
removal efficiency decreases.
Aluminum hydroxide gels deposited on anaerobic
lake sediments would be exposed to high interstital
phosphorus concentrations and relatively low pH.
Phosphorus removal would continue by further
sorption/precipitation and the AI/P molar ratio of
deposited gels would decrease with prolonged ex-
posure to high phosphorus concentrations. Kennedy
(1978) eluted laboratory-produced gel with phosphorus
solutions at pH 6, 7, and 8 (Figure 3). Phosphorus
removal was pH dependent, and although initially high,
was minimal at AI/P molar ratios ranging from 2 to 4.
Therefore, effectiveness and longevity will depend on
the amount of aluminum deposited relative to sediment
phosphorus concentration and the rate at which
phosphorus is made available.
Few direct field evaluations of the effectiveness of
deposited aluminum hydroxide gels in retaining
phosphorus have been reported. An exception is Dollar
Lake, Ohio, which was treated hypolimnetically in
1974 with 10 tons (197 mg Al/l) of aluminum sulfate
(Kennedy, 1978). Phosphorus concentrations of water
samples collected immediately above capped treated
and untreated anaerobic sediments were compared
during summers 1974, 1975, and 1976, with dif-
ferences expressed as percent reduction (Figure 4).
Percent reduction averaged about 90 percent following
treatment but decreased to about 75 percent and 65
percent in 1975 and 1976, respectively.
2
2 »
* t
5 2
HO 1100
ELUTN3N VOIUHE. ml
Figure 3. — Changes in the AI/P molar ratio of aluminum
hydroxide gels eluted with phosphate buffer at pH 6, 7, and 8
(Kennedy, 1978).
DOSE DETERMINATION
Lake managers have employed a number of different
dose determination methodologies based on two major
objectives: Phosphorus removal from the water column
and control of phosphorus release from sediments.
Immediate reductions in phosphorus concentration,
while often desirable, are generally secondary to long-
term treatment objectives. However, adoption of proper
dose determination methodology will allow calculation
of a dose accomplishing both objectives.
Phosphorus precipitation/sorption, aluminum hy-
droxide formation, and dissolved aluminum concentra-
tion are pH dependent. In the range pH 6-8, dissolved
aluminum concentrations will be minimal while
phosphorus removal and floe formation will be
maximal. Although the control of sediment release
requires maximal aluminum deposition, maximum
additions to the lake will be dictated by conditions in
the water column and by changes resulting from
treatment. Since excessive additions of aluminum will
produce undesirable side effects (e.g. low pH and
alkalinity, and high dissolved aluminum concentra-
tions), an optimum dose would be that dose which
reduces pH to about 6.0. This optimum dose would
maximize the amount of aluminum deposited over
sediments as dictated by lake conditions.
Data collected from two Ohio lakes treated hypo-
limnetically with doses of aluminum sulfate which
-------
408
RESTORATION OF LAKES AND INLAND WATERS
Table 1
Lake
Dose
Percent
Phosphorus Removal
mg Al/l
TP
SRP
SUP
PP
Reference
Dollar Lake
West Twin Lake
(Hypo. Treatment)
"Maximum" Dose
7.5(T) 83(T) 96(T) 62(T)
19.7(H) 91 (H) 97 (H) 65(H)
22.6 92(H) 99(H)
Other
Dose Methodologies
65(T)
63(H)
Kennedy, 1978
Cooke, et al., 1978
Mytajarvi
Langsjon
Cline's Pond
Lake of 4 seasons
Horseshoe Lake
Pickerel Lake
Liberty Lake
State Rearing Pond
Medical Lake
Lake Mendota"
Lower Nashotah"
Lake Wingra**
Little John**
12.2
4.5
10.0
1.8*
1.8
7.0
0.4
11.0*
—
16.2
16.2
16.2
16.2
40
57
-0
—
25(E)
14(H)
-25-30
—
90
55(E)
79(E)
81(H)
43(E)
40(H)
42(E)
27(E)
92
92
-0
90
-
—
—
96
—
80(E)
90(E)
72(H)
100(E)
100(H)
8(E)
82(E)
..
-
—
—
-
—
—
—
—
—
17(E)
30(H)
36(E)
29(H)
1(E)
1(E)
Dunst, et al. 1974
Jernelove, 1970
Sanville, et al. 1976
Dunst, et al. 1974
Petersen, et al. 1973
—
D.R. Knauer, per. comm.
Funk, 1977
Dunst, et al. 1974
A. Gasperino, per. comm.
Eisenreich, et al. 1977
—
Eisenreich, et al. 1977
Eisenreich, et al. 1977
Eisenreich, et al. 1977
Eisenreich, et al. 1977
Note: T = whole lake, E = epilimnion, H = hypolimnion.
* mg Al/m2.
** Results of laboratory experiments using lake water.
were calculated based on similar considerations,
indicated that optimum doses also substantially
reduced water column phosphorus concentrations
(Table 1). While the removal of soluble reactive
phosphorus (SRP) was similar to that obtained using
other dose determination methods (e.g. those employ-
ing AI/P molar ratios and additions for maximum
phosphorus removal), the use of a maximum (Kennedy,
1978) dose markedly increased the removal of total (TP)
and soluble unreactive (SUP) phosphorus. Therefore,
optimum doses will effectively accomplish both
treatment objectives and thus allow for the establish-
ment of a single dose determination method.
The following simple method for determining
optimum doses for aluminum sulfate applications
requires limited laboratory equipment and expertise,
and can be easily implemented by local lake managers.
Procedure:
1. Obtain representative water samples from the lake
to be treated. Care should be exercised in selecting
sampling stations and depths since significant
heterogeneities, both vertical and horizontal, com-
monly occur in lakes. Samples should be collected
as close to the anticipated treatment date as
possible.
2. Determine the total alkalinity and pH of each
sample. Total alkalinity, an approximate measure of
the buffering capacity of lake water, will dictate the
amount of aluminum sulfate (or aluminum) required
to achieve pH 6 and thus optimum dose. Additional
chemical analyses can be performed, depending on
the specific needs of the investigator. For example,
phosphorus analyses before and after laboratory
treatment would allow estimation of anticipated
phosphorus removal effectiveness.
3. Determine the optimum dose for each sample. Initial
estimates of this dose, based on pH and alkalinity,
can be obtained from Figure 5. More accurate
estimates should be made by titrating samples with
fresh stock solutions of aluminum sulfate of known
aluminum concentration using a standard burette or
graduated pipette. The concentration of stock
aluminum solutions should be such that pH 6 can be
reached with additions of 5 to 10 milliliters per liter
of sample. Samples must be mixed (about 2
minutes) using an overhead stirring motor and pH
changes monitored continuously using a pH meter.
Optimum dose for each sample will be the amount
of aluminum, which when added, produces a stable
pH of 6.0.
4. The relationship between total alkalinity and
optimum dose can be determined using information
from each of the above titrations by plotting
optimum dose as a function of alkalinity. This
relationship will allow determination of dose at any
alkalinity within the range tested.
APPLICATION TECHNIQUES
Estimation of the relationship between optimum
dose and alkalinity provides a means by which
laboratory determined doses may be scaled for lake
-------
NUTRIENT PREVENTION AND INACTIVATION
409
100
80 -
*
1
S-
;40
20
•_
I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I
1974 1975 1976
SAMPLING DATE
Figure 4. — Percent reduction in total phosphorus
concentration of sediment seepage water collected above
aluminum treated Dollar Lake sediments (Kennedy, 1978).
Dollar Lake was treated hypolimnetically with aluminum
sulfate July 1974.
ALVMIMM D0(t li«t*U/) TO OCTAIN pH CO
Figure 5 — Estimated aluminum sulfate dose (mg AIM)
required to obtain pH 6 in treated water of varying initial
alkalinity and pH. Based on equations in Ferguson and King
(1977) and assuming insignificant phosphorus concentrations.
treatment. Total treatment dose can be determined by
calculating a volume-weighted mean total alkalinity for
segments of the lake (e.g. depth strata) having similar
pH. Careful consideration should be given to the extent
to which mixing can be accomplished since this will
determine the volume of water to be treated. For
example, surface treatments of deep lakes would,
depending on application method and equipment,
involve only waters in upper depth strata.
Once the total application dose is determined,
provisions must be made for proper distribution with
respect to depth and volume. This is most easily done
by establishing a treatment grid system and calculating
dose allocations for each treatment quadrant (Kennedy,
1978; Cooke, etal. 1978; Funk, 1977; Peterson, 1973).
The grid system will also facilitate field procedures if
Figure 6. — Basic components of a lake application system
(Cooke and Kennedy, 1980).
the intersections of grid lines are marked with coded or
numbered buoys.
Aluminum sulfate may be purchased in granular
form or as a liquid and lake managers must determine
which is more convenient. The use of liquid alum
simplifies field procedures since it does not require
mixing and may be pumped directly from tank trucks to
application equipment. It does, however, involve
further dose calculations to account for liquor density
and temperature (see Cooke and Kennedy, 1980, for a
discussion of these calculations). Liquid alum has the
disadvantage of not being readily available in many
areas of the country, as well as presenting storage
problems. Granular alum, while more readily available
and easier to store, must be dissolved prior to use, thus
complicating field procedures.
Application equipment systems employed to date
have consisted of a shorebased storage/mixing facility,
a distribution pipe, an application barge, and an
application manifold (Figure 6). The exact design of an
application system, while generally requiring these
components, will reflect the lake manager's specific
objectives and site characteristics. In general, treat-
ment will involve pumping alum from a storage or
mixing tank through a floating pipe, tube or hose to a
smaller storage tank on an application barge and then
to a distribution or application manifold. Provisions for
pumping both alum and lake water to the manifold will
allow for flash mixing before discharge and provide a
means for adjusting alum discharge rate. Mixing
should be accomplished by movement of the barge
and/or by turbulence behind the manifold. Additional
mixing during surface treatments could be attained by
positioning the manifold behind the propeller of the
barge motor. In any case, mixing should be maximized.
Applications may be made to surface water or at
predetermined depth(s) depending on treatment ap-
proach. Surface treatments, while less complicated
and time consuming, will be less effective in reducing
hypolimnetic phosphorus concentrations, a concern in
cases requiring phosphorus removal. Surface treat-
ment will also expose near-surface organisms to
reduced pH. Applications at depth (e.g. immediately
-------
410
RESTORATION OF LAKES AND INLAND WATERS
above the hypolimnion) may be made by suspending
the manifold below the application barge on a rigid
frame.
Treatment costs are highly variable and will depend
on availability of chemicals and equipment, chemical
costs, available labor, lake size, and dose. Reported
costs for several lake treatments are reviewed in Cooke
and Kennedy (1980).
LAKE MANAGEMENT STRATEGIES
USING ALUMINUM SULFATE
High phosphorus and low dissolved oxygen concen-
trations do not necessarily indicate that aluminum
sulfate application will have a beneficial impact
following nutrient diversion. Lakes which have not had
a long history of excessive nutrient and organic loads
may respond immediately to curtailed loadings (Schin-
dler and Lee, 1974), while lakes receiving substantial
inputs of clay in addition to nutrients may contain
sediments with high sorptive capacities for phos-
phorus. Extensive growths of macrophytes may,
through senescence and decay, continue to be the
dominant phosphorus source long after nutrient
diversion (Barko and Smart, 1979). Therefore, the
relative importance of sediments as a phosphorus
source must be assessed prior to attempting costly
aluminum sulfate applications.
Nutrient budget calculations provide a simple means
for assessing the importance of internal loading.
Cooke, et al. (1977) employed such calculations to data
from two Ohio lakes, one of which was subsequently
treated with aluminum sulfate, and determined that
internal sources accounted for 65 to 100 percent of the
summer increase in phosphorus content. While such
calculations do not specifically identify sources within
the lake, they can, when viewed in light of other
observations, indicate the possible need for in-lake
phosphorus control measures. For example, the
presence of thick organic, nutrient-rich sediments and
a low macrophyle density in lakes experiencing
significant internal phosphorus recycling would at least
suggest that aluminum sulfate application would
hasten lake recovery following reductions in external
nutrient loads.
How effective and long lasting are aluminum
treatments for sediment-phosphorus control? This will
depend on sediment characteristics and the quantity of
aluminum added and, at best, can only crudely be
estimated. Since phosphorus sorption, which de-
creases as AI/P molar ratio decreases, would
effectively cease at an AI/P molar ratio of 2 to 4, 1 mg
of aluminum could potentially remove 0.6 to 0.3 mg of
phosphorus. However, a number of other factors,
including disruption of the floe layer (Browman, et al.
1977), deposition of sediments subsequent to treat-
ment, and phosphorus uptake by floe during settling
would affect this estimate.
If there is reason to believe that optimum dose
applications would supply insufficient amounts of
aluminum to sediments (e.g. low alkalinity lakes),
combined aluminum sulfate/sodium aluminate treat-
ments could be considered (Wirth, et al. unpubl.).
Doses for such treatments could be considered (Wirth,
et al. unpubl.). Doses for such treatments can be
estimated by modifying the laboratory procedure.
Sodium aluminate increases pH and additions must be
balanced by additions of aluminum sulfate. Once
optimum pH is reached using aluminum sulfate only,
further additions would be possible by adding 3 moles
of aluminum as sodium aluminate for every mole of
aluminum added as aluminum sulfate. Doses calcu-
lated in this manner would also allow larger additions
to sediments with the greatest potential for phosphorus
release, such as those deposited near nutrient-rich
inflows.
Aluminum sulfate treatments may also be employed
for purposes other than sediment-phosphorus control.
Ree (1963) treated three California water supply
reservoirs and tributaries to reduce turbidity caused by
storm runoff from construction sites and thus reduce
particulate loads to water treatment filter beds. Alum
may also be used to coagulate and sediment organic
particulates, including algae (Lin, et al. 1971), as a
means of reducing oxygen demand following intense
algal blooms or macrophyte die-off. Applications to
littoral areas, while having little inhibitory effect on the
growth of macrophytes (Dunst, et al. unpubl.), could
retain phosphorus released during decay (Funk, et al.
1977) and thus reduce phosphorus inputs to pelagic
areas following herbicide treatments. Periodic inputs of
phosphorus and organic particulates could be reduced
by temporarily retaining and treating storm runoff in
urban areas (Shapiro and Pfannkuch, 1973).
CONCLUSION
Phosphorus control by in-lake chemical treatment is
only one of many lake restoration techniques available
to lake managers; the decision to use aluminum sulfate
must be based on careful evaluation of lake conditions
and management objectives. Although such treat-
ments provide a simple method for removing particu-
lates and phosphorus from lake water, they are more
appropriate in management situations requiring con-
trol of phosphorus release from eutrophic sediments.
The dose determination method described here allows
calculation of doses which maximize aluminum input
to sediments. The degree to which such treatments
provide long-term control over internal phosphorus
recycling is not adequately documented; future
restoration efforts should include provisions for long-
term evaluation.
REFERENCES
Barko, J. W., and R. M. Smart. 1979. The role of
Myriophyllum spicatum in the mobilization of sediment
phosphorus. In J. E. Breck, R. T. Prentki, and 0. L. Loucks,
ed. Aquatic plants, lake management, and ecosystem
consequences of lake harvesting. Proc. Conf., Madison, Wis.
Browman, M. G., R. F. Harris, and D. E. Armstrong. 1977.
Interaction of soluble phosphate with aluminum hydroxide
in lakes. Tech. Rep. 77-05. Water Resour. Center, University
of Wisconsin, Madison.
Burrows, W. D. 1977. Aquatic aluminum: Chemistry,
toxicology, and environmental prevalence. CRC Crit. Rev.
Environ. Control 7:167.
Cooke, G. D., and R. H. Kennedy. 1980. Precipitation and
inactivation of phosphorus with aluminum and zirconium
salts. 1977. Eco. Res. Ser. U.S. Environ. Prot. Agency, (in
press).
-------
NUTRIENT PREVENTION AND INACTIVATION
411
The occurrence of internal phosphorus loading in
two small, eutrophic, glacial lakes in Northeastern Ohio.
Hydrobiology 56:129.
Cooke, G. D., et al. 1978. Effects of diversion and alum
application on two eutrophic lakes. EPA-600/3-78-033.
U.S. Environ. Prot. Agency.
Dunst, R. C., et al. 1974. Survey of lake rehabilitation
techniques and experiences. Tech. Bull. 75. Dep. Nat.
Resour. Madison, Wis.
Dunst, R., D. Knauer, and R. Wedepohl. Undated. Macrophyte
growth-effect of mixing aluminum sulfate in lake sediment.
Dep. Nat. Resour. Madison, Wis. Unpublished.
Eisenrich, S. J., D. E. Armstrong, and R. F. Harris. 1977. A
chemical investigation of phosphorus removal in lakes by
aluminum hydroxide. Tech. Rep. 77-02. Water Resour.
Center. University of Wisconsin, Madison.
Ferguson, J. F., and T. King. 1977. A model of aluminum
phosphate precipitation. Jour. Water Pollut. Control Fed.
49:646.
Freeman. R. A., and W. H. Everhart. 1971. Toxicity of
aluminum hydroxide complexes in neutral and basic media
to rainbow trout. Trans. Am. Fish. Soc. 100:644.
Funk, W. H., H. R. Gibbons, and S. K. Bhagat. 1977. Nutrient
inactivation by large scale aluminum sulfate treatment.
Conf. Mechanics of Lake Restoration, Madison, Wis., April.
Gasperino, A. F. Personal communication. Battelle. Pacific
Northwest, Richland, Wash.
Hayden, P. L, and A. J. Rubin. 1974. Systematic investigation
of the hydrolysis and precipitation of aluminum (III). Pages
317-381 In A. Rubin, ed. Aqueous-environmental chemistry
of metals. Ann Arbor Science, Ann Arbor, Mich.
Hsu, P. H. 1965. Fixation of phosphate by aluminum and iron
in acidic soils. Soil Sci. 99:398.
Jernelov, A. 1970. Aquatic ecosystems for the laboratory.
Vatten 26:262.
Kennedy, R. H. 1978. Nutrient inactivation with aluminum
sulfate as a lake restoration technique. Ph.D. Dissertation.
Kent State University, Kent, Ohio.
Knauer, D. Personal communication. Dep. Nat. Resour.
Madison, Wis.
Larsen, D. P. et al. 1975. Response of Shagawa Lake,
Minnesota, USA to point-source phosphorus reduction.
Verh. Int. Ver. Limnol. 19:884.
Lin, S. D., R. L. Evans, and D. B. Beuscher. 1971. Algal
removal by alum coagulation. Rep. Invest. 68. Illinois State
Water Survey. Urbana, III.
Peterson, J. 0., et al. 1973. Eutrpphication control: Nutrient
inactivation by chemical precipitation at Horseshoe Lake,
Wisconsin. Tech. Bull. 62, Dep. Nat. Resour. Madison, Wis.
1974. Nutrient inactivation as a lake restoration
procedure— laboratory investigations. EPA-660/3-74-032.
U.S. Environ. Prot. Agency.
1976. Laboratory evaluation of nutrient inactiva-
tion compounds for lake restoration. Jour. Water Pollut.
Control Fed. 48:817.
Recht, H. L, and M. Ghassemi. 1970. Kinetics and
mechanism of precipitation and nature of the precipitate
obtained in phosphate removal from waste water using
aluminum (III) and iron (III) salts. Water Pollut. Control Res.
Serv. 17010 EKI.
Ree, W. R. 1963. Emergency alum treatment of open
reservoirs. Jour. Am. Water Work. Assoc. 55:275.
Sanville, W. D., et al. 1976. Studies on lake restoration by
phosphorus inactivation. EPA — 600/3-76-041. U.S.
Environ. Prot. Agency.
Schindler, D. W., and E. J. Lee. 1974. Experimental lakes
area: Whole-lake experiments in eutrophication. Jour. Fish.
Res. Board Can. 31:937.
Shapiro, J., and H. Pfannkuch. 1973. The Minneapolis chain
of lakes: A study of urban drainage and its effects. Interim
Rep. 9 Limnol. Res. Center. University of Minnesota.
Stumm, J. W., and J. J. Morgan. 1970. Aquatic chemistry. An
introduction emphasizing chemical equilibria in natural
waters. Wiley-lnterscience, New York.
-------
412
A COMPARISON OF TWO ALUM TREATED LAKES
IN WISCONSIN
D. R. KNAUER
P J. GARRISON
Office of Inland Lake Renewal
Wisconsin Department of Natural Resources
Madison, Wisconsin
ABSTRACT
Two alum treated lakes. Pickerel and Mirror, are compared. The polymictlc nature of Pickerel Lake
rather than the alum appeared to be responsible for observed changes in phytoplankton changes in
the dimictic Mirror Lake Pickerel Lake demonstrated only minor fluctuations in total-P
concentrations following treatment. Pre- and post-treatment comparison of phytoplankton
biomass indicated a reduction following the alum application. However, after holomixis occurred in
July, phytoplankton biomass was similar to or greater than the pre-treatment year. Mirror Lake
total-P concentrations were reduced from 90 /jg/l to 20 ng/\ . The phytoplankton biomass during
the spring and fall periods has decreased and the principal nuisance alga Oscillatoria agardhiibas
been eliminated
INTRODUCTION
The positive results of aluminum applications for lake
restoration in Sweden (Jernelov, 1970) encouraged
similar treatments in the United States. The use of
aluminum for removing phosphorus from eutrophic
lake waters is an extension of water and waste
treatment processes. Aluminum hydroxide has a high
capacity for removing dissolved and suspended
phosphorus materials under conditions that are
common to lakes. In addition, the use of aluminum
salts is relatively inexpensive and has a low toxicity to
most forms of aquatic life.
The objective of the alum treatment at Pickerel and
Mirror Lakes was to rapidly remove available phos-
phorus from the lake and at the same time prevent
release of phosphorus from the lake sediments,
thereby limiting the growth of planktonic plants. The
decision to treat Pickerel Lake was based on a history of
previous algal problems and associated fish winter
kills. Mirror Lake was treated to enhance the recovery
rate following a nutrient diversion project.
SITE DESCRIPTION
Both Mirror and Pickerel Lakes are glacial seepage
lakes approximately 19 kilometers apart situated in
outwash plains formed during the recession of the Gary
ice sheet of the Pleistocene Glaciation about 10,000 to
14,000 years ago. The physical characteristics of each
lake are presented in Table 1. The volume of both lakes
is similar although Pickerel Lake has four times the
surface area of Mirror Lake. The mean depth of Mirror
Lake, however, is three times that of Pickerel Lake.
An important morphological difference between
Mirror and Pickerel Lakes is their respective relative
depths. The relative depth (Z) is defined as the
maximum depth as a percentage of the mean diameter
(Hutchinson, 1957). The larger the Zr value, the more
stable the lake. Pickerel Lake has a Z of 0.5 percent and
is considered polymictic while Mirror Lake has a Zr of
2.3 percent and prior to artificial mixing in the fall of
1977, was considered meromictic. Since the fall of
1977, Mirror Lake has been artificially circulated for
several weeks each spring and late fall.
In 1972, the influx of total-P to Pickerel Lake from
surface runoff, ground water and direct precipitation
was calculated to be 24 kg which is equivalent to a
phosphorus loading rate (P) of 0.13 gm/m2.yr
(Hennings, 1978). The allochthonous sources of
phosphorus to the lake are diffuse and the direct
drainage basin is forested with only one permanent
dwelling.
Mirror Lake is located in the city of Waupaca and
received an annual allochthonous total-P influx of 15 to
20 kg in 1972 and 1973 (Knauer, 1975; Peterson,
1974). In 1976 a storm sewer diversion project reduced
the allochthonous sources of phosphorus by 50 to 60
percent. As a result of the diversion project, the P has
been reduced to 0.12 gm/m2 yr, very similar to Pickerel
Lake.
RESULTS
Phosphorus — Pickerel Lake
Liquid aluminum sulfate was applied in Pickerel Lake
on April 17, 1973. The application rate was 170 kg
Al/ha to yield a concentration of 7.3 mg Al/l in the
lake. The liquid alum was released near the surface
and at mid-depth to ensure a reasonable distribution in
the water column.
As a result of the aluminum sulfate additon, the pH
declined from 8.2 to 7.1 and total alkalinity was
-------
NUTRIENT PREVENTION AND INACTIVATION
413
reduced by approximately 25 mg/l (110 to 85 mg/l as
CaCOa). There was no immediate effect on the
phosphorus concentrations in the lake (Figure 1a).
Total-P (weighted mean) increased from 28/ug/l in
early June 1973 to 73 fjg/\ by late July 1973.
Following holomixis the total-P concentration in
Pickerel Lake was reduced to 10 /yg/l The reduction of
total-P during fall overturn has been noted in other
alum treated lakes, e.g., Horseshoe Lake, Wis. and
Medical Lake, Wash. (Peterson, et al. 1973; Gasperino,
et al. 1980). The phosphorus dynamics appear to be
different in Pickerel Lake, however, when compared to
the other alum treated lakes. In Horseshoe, Medical,
and Mirror Lakes, the alum floe in the sediments'
prevented sediment phosphorus from being released
into the overlying waters. In Pickerel Lake, the
weighted mean total-P/2 concentrations increased
during the fall months from 10 g/l in September 1973
to 30 fjg/\ by November 1973 (Figure 1a) and 50 yug/l
by March 28, 1974. The increasing total-P concentra-
tions during the winter months suggest the alum floe
had not completely prevented sediment phosphorus
release. An analysis of aluminum concentrations in
sediment cores from Pickerel Lake before and after
treatment indicated the floe had been redistributed
towards the center of the lake basin following a series
of'holomictic occurrences throughout the summer and
fall of 1973. A large area of lake sediments was
subsequently interacting with the overlying waters
without the benefit of an aluminum floe covering.
Phosphorus — Mirror Lake
In the year following the storm sewer diversion, the
annual average phosphorus concentraion, 90 /L/g/l ,
showed little change from previous years, 88 and 93
09/I in 1972 and 1973, respectively (Smith, et al.
1975). During the summer, total-P concentrations in
the epilimnion were typically 20/yg/l; however, in the
hypolimnion total concentrations increased to 550
j"g/I.The hypolimnetic total-P concentrations appear to
be the result of sediment-phosphorus release when the
bottom waters were anaerobic (Mortimer, 1941; Kamp-
Nielsen, 1974). Data from nutrient regeneration
chambers placed on the lake sediments indicated a
total-P release rate of 1.8 mg P/m2day during 1977
(Knauer and Garrison, 1979). However, that rate
becomes limited as a concentration of 550 fjg/\ is
approached and a sediment-water phosphorus equili-
brium is achieved.
Alum was applied to Mirror Lake on May 17, 1978,
since it appeared that sediments could potentially
provide a significant source of phosphorus. Previous
studies by Ryding and Forsberg (1977) and Welch
(1977) suggest that as a result of internal phosphorus
loading, lakes from which large allochthonous nutrient
sources were eliminated recovered very slowly. The
alum was applied at a rate of 337 kg Al/ha, and at a
depth of 3 meters to achieve an aluminum concentra-
tion of 6.6 mg/l.
As a result of the alum treatment, weighted mean
total-P concentrations were reduced from 90 /ug/l to 20
/"9/I, a 78 percent reduction (Figure 1b). Owing to an
algal bloom during the treatment date, much of the
total-P was in the particulate fraction and the dissolved
H
Wi)401
Pickerel Coke
JFMAMJJASONCIJFMAMJ JASON
1972 1973
Figure 1 a. — Weighted average total phosphorous for Pickerel
Lake during 1972 and 1973. The lake was treated with
aluminum sulfate on April 17, 1973.
P 80-
(ug/l)
12
1977
Figure 1 b. — Weighted average total and dissolved reactive
phosphorus for Mirror Lake during 1977,1978, and 1979. The
hypolimnion of the lake was treated with aluminum sulfate on
May 17, 1978.
reactive-P (DRP) concentration was less than 4/ug/l
The phosphorus reduction following the alum treat-
ment was the result of physical entrapment of
particulate-P (algae). Carbon, nitrogen, and phosphorus
data from sedimentation traps at the 12 m depth
confirmed the fact that algae were carried to the
sediments with the alum floe.
As indicated in Figure 1 b, the alum treatment has
been successful in preventing release from the
sediments. Weighted mean total-P concentrations
have remained at 20 /jg/\ and DRP has been
undetectable for at least 2 years following the alum
treatment.
Phytoplankton — Pickerel Lake
In 1972, the algal biomass was relatively constant, 6
to 7 mm3/! (Figure 2a), throughout the summer and
fall. During the summer months, the major com-
ponents of the algal biomass were from the phyla
Chlorophyta and Pyrrhophyta. After holomixis in mid-
September, a short pulse of Microcystis aeruginosa
-------
414
RESTORATION OF LAKES AND INLAND WATERS
was observed, followed by a diatom pulse that
dominated the algal biomass through October. During
the month of November, the dominance shifted from
diatoms (Stephanodiscus) to the green alga (Ankis-
trodesmus fractus (Figure 3).
Following the alum treatment in mid-April 1973, the
algal biovolume remained approximately 3 mm3/!
through July (Figure 2b). This was half the 1972
biovolume concentrations; the alum treatment ap-
parently was effective. However, in late July the lake
completely mixed and by the end of August the biomass
reached 43 mm3/!. The algal assemblage was
dominated by Microcystis aeruginosa from August
.through mid-October (Figure 3).
In 1972 and 1973, M. aeruginosa was not observed
in the surface waters until the lake mixed. It is possible
that M. aeruginosa may have been present at or near
the sediment surface. Light measurements taken in
Pickerel Lake during the summer of 1973 with a
submarine photocell indicated 1 percent of surface
HChlorophylo
LJ CryplophycM«
LJ Pyrrhophyto
E23 Bacillariophyceae
LJ Cyanophyta
A
v
" -. _\. __ ^"^.- ' ' ^
"\\
- -,_.^
ay June July Aug. Sept. Oct. Nov
1972
Figure 2a. — Algal biomass in Pickerel Lake during 1972 atO.5
meters.
E&lChlorophr
DChry,ophy
D Cryplophy
D Pyrrhophylo
£3 Boelllorlophy
LJ Cyunophyla
Apr. May
1973
Aug. Sept. Oct. Nov.
Figure 2b. — Algal biomass in Pickeral Lake during 1973 atO.5
meters. The lake was treated with aluminum sulfate on April 17,
1973.
M J JASON
1972
Figure 3. — The seasonal succession of major phytoplankton
genera, 1972 and 1973 for Pickerel Lake at 0.5 meters.
light was present at the sediment surface through July
Positive primary productivity measurements were also
recorded at the 41/2 m depth during 1973. Our data also
showed that immediately after holomixis in 1972 and
1973, the biovolume of M. aeruginosa was similar, 4.5
mm3/1. Our data suggest that as a result of holomixis,
M. aeruginosa was distributed throughout the water
column in both years. Following holomixis in mid-
September, 1972, the environmental conditions (tem-
perature, light, etc.) were suboptimal for M. aeruginosa
and the population never expanded. Fallon and Brock
(1980) have also reported a rapid decline of Microcystis
during late September in Lake Mendota.
In 1973, holmixis occurred in late July and M.
aeruginosa was distributed throughout the water
column when environmental conditions were more
favorable for growth. The biovolume increased from 4.5
mm3/! following holomixis to 43 mm3/! by late August.
As in the previous year, the population rapidly declined
in late September.
Phytoplankton — Mirror Lake
Mirror Lake did not experience the summer algal
problems that are typical in many eutrophic lakes. The
problem alga in Mirror Lake was Oscillator/a agardhii
(Figure 4). This species dominated the phytoplankton
-------
NUTRIENT PREVENTION AND INACTIVATION
415
assemblage during the late fall and early winter
months and at spring overturn (Figures 5a and 5b).
Although this alga was present during the summer, it
remained only in the lower metalimnion, albeit in large
concentrations. The occurrence of O. agardhiithrough-
out the lake during the fall overturn was owing to the
redistribution of the metalimnetic population and not
an increase in the growth rate.
The biovolume of O. agardhii was similar in the
spring of 1977 and 1978,6 mmVI (Figures 5a and 5b).
Following the hypolimnetic alum treatment on May 17,
1978, a decline in the 0. agardhii biovolume was
observed during the fall of 1978 and 1979 and
subsequent spring of 1979 (Figure 5c). 0. agardhii was
not present during the spring of 1980 nor was it
present in metalimnion during the summer of 1980.
At times, O. agardhii has dominated the metabolism
of Mirror Lake. Other studies (Smith, et al. 1975) have
shown that during the summer, BOD's in the
metalimnion of Mirror Lake were five times higher than
elsewhere in the water column as a result of the
O.agardhii population. The reduction of phytoplankton
biomass and the elimination of 0. agardhii as a result
of the alum application have substantially improved the
water quality of of Mirror Lake.
In summary, the introduction of liquid aluminum
sulfate into the water column of Pickerel Lake for the
purpose of lowering the phosphorus concentration and
reducing algal biomass was not successful. It is our
opinion that alternative techniques for alum addition to
polymictic lakes, e.g. plowing into the sediments,
should be researched. The application of alum to the
water column of dimictic lakes appears to be a
successful technique to improve water quality, provid-
ing the phosphorus loading to the lake has been
reduced to an acceptable level.
AMJJ A 5 O N
AMJJAS o N
120,000
2000
edit/ml
Sphoerocyttis Schroeleri
AM J JASON
1979
Figure 4. — The seasonal succession of major phytoplankton species, 1977, 1978 and 1979 for Mirror Lake at 2.5 meters.
-------
416
RESTORATION OF LAKES AND INLAND WATERS
Table 1. — Morphometric data for Mirror and Pickerel Lakes.
Surface area
Maximum depth
Mean depth
Relative depth
Volume
Hydraulic residence time
Mirror Lake
5.1 ha
13.1 m
7.8 m
2.28%
4x1 05m3
4 years
Pickerel Lake
21.0 ha
4.7 m
2.4 m
0.46%
5x1 05m3
0.63 years
[gj] Chlorophyla
Q Chrysopliylo
Q Pyrrbopbyto
Q Cryptophyceae
[X] Soci/'onppftyceoe
£3 Cytmophyceoe
APR MAY JUN JUL AUG SEPT OCT NOV
Figure 5a. — Algal biomass in Mirror Lake during 1977 at 2.5
meters.
Figure 5b — Algal biomass m Mirror Lake during 1978 at 2.5
meters. They hypolimnion was treated with aluminum sulfate
on May 17, 1978
g| CMarppbyla D CryplopHycsae
cillanophycea
JUL AUG SEPT
1979
NOV
Figure 5c. — Algal biomass in Mirror Lake during 1979 at 2.5
meters.
REFERENCES
Fallon, R. D., and T. D. Brock. 1980. Planktonic blue-green
algae: Production, sedimentation, and decomposition in
Lake Mendota, Wis. Limnol. Oceanogr. 25:72.
Gasperino, A. F., et al. 1980. Medical Lake improvement
project: a success story. Proc. Int. Symp. Inland Waters Lake
Restoration, Portland, Maine, Sept. 8-12.
Hennings, R. G. 1978. The hydrogeology of a sand plain
seepage lake Portage County, Wis. M.S. Thesis. University
Wisconsin, Madison.
Hutchinson, G. E. 1957. A treatise on limnology. I.
Geography, physics, and chemistry. John Wiley and Sons,
Inc., New York.
Jernelov, A. 1970. Phosphate reduction in lakes by
precipitation with aluminum sulfate. Water Pollut. Res.
Conf. Stockholm, Sweden.
Kamp-Nielsen, L. 1974. Mud-water exchange of phosphate
and other ions in undisturbed sediment cores and factors
affecting the exchange rates. Arch. Hydriol. 73:218.
Knauer, D. R. 1975. The effect of urban runoff on
phytoplankton ecology. Verh. Int. Verein. Limnol. 19:893.
Knauer, D. R., and P. J. Garrison. 1979. A staius report on the
Mirror/Shadow Lakes evaluation project. Pages 8-54 in
Limnological and socioeconomic evaluation of lake restora-
tion projects. EPA-600/3-79-005. U.S. Environ. Prot.
Agency.
Mortimer, C. H. 1941. The exchange of dissolved substances
between mud and water in lakes. Jour. Ecol. 29:280.
Peterson, J. 0. 1974. Mirror and Shadow Lakes urban runoff
project. Unpubl data. University of Wisconsin Extension,
Madison.
Peterson, J. 0., et al. 1973. Eutrophication control: Nutrient
inactivation by chemical precipitation at Horseshoe Lake,
Wis. Wis. Dep. Nat. Resour. Tech. Bull. 62.
Ryding, S. 0. and C. Forsberg. 1977. Sediments as a nutrient
source in shallow polluted lakes. Pages 227-234 in
Interactions between sediments and fresh water. H. L.
Golterman, ed. W. Junk, The Hague.
Smith, S. A., D. R. Knauer, andT. L. Wirth. 1975. Aeration as
a lake management technique. Wis. Dep. Nat. Resour. Tech.
Bull. 87.
Welch, E. B. 1977. Nutrient diversion: Resulting lake trophic
state and phosphorus dynamics. Ecol. Res. Ser. 600/3-77-
003 Corvallis, Ore.
ACKNOWLEDGEMENTS
Financial support for the Pickerel Lake study was provided
by the Upper Great Lakes Regional Commission. The Mirror
Lake Study is funded by the U.S. EPA.
-------
417
HYPOLIMNETIC ALUMINUM TREATMENT OF
SOFTWATER ANNABESSACOOK LAKE
DAVID R. DOMINIE II
Maine Department of Environmental Protection
Augusta, Maine
ABSTRACT
Since the 1940's Annabessacook Lake in central Maine has experienced algal blooms resulting
from industrial and municipal wastewater inputs. A comprehensive water quality restoration effort
combining nutrient diversion, agricultural waste management, and in-lake nutrient inactivation
was completed in 1978. The nutrient inactivation technique used aluminum sulfate and sodium
aluminate in combination to precipitate phosphorus in the hypolimnion. The use of the two
chemicals was necessary to provide sufficient buffering capacity to mitigate potential pH shifts and
aluminum toxicity in the low alkalinity (< 20 mg/l) water of Annabessacook Lake. A segregated
dual injection system was designed capable of delivering the chemicals simultaneously to any
depth between 0 and 7 meters. The hypolimnetic treatment, covering approximately 121 hectares,
was completed in a 3-week period. Monitoring data showed a 50 percent decrease in hypolimnetic
P approximately 1 month after the application. One year after the aluminum treatment, water
quality as measured by total phosphorus, chlorophyll a, and Secchi disk visibility, had significantly
improved, thus giving rise to optimism about the future of the lake.
INTRODUCTION
Nutrient inactivation by chemical precipitation is a
phosphorus removal technique which has recently
been applied to lake restoration. The precipitation
agent which has received the most attention is
aluminum, although iron, calcium, rare earth metals,
and fly ash have also been investigated (Higgins, et al.
1976; Peterson, et al. 1976). Aluminum has often been
used in cases where nutrient recycling from bottom
sediments would otherwise prolong the recovery of
eutrophic lakes long after external sources have been
reduced (Knauer and Garrison, 1979; Cooke, et al.
1977).
Aluminum reacts in water at various pH's to
stoichiometrically combine with phosphorus to form
AIPO*, or undergo hydrolysis to form an amorphous
floe which physically sorps soluble phosphorus.
Aluminum offers the advantages of low toxicity,
effectiveness within the pH range of most natural
waters, and is inert to changing redox potentials.
BACKGROUND
Annabessacook Lake is a large (574 hectares) lake
located in central Maine (Figure 1). The lake has
experienced blue-green algal blooms since the 1940's
because of combined discharge of municipal and
industrial wastewater into the system. In 1972, an
estimated 80 percent of the external phosphorus load
to Annabessacook Lake (Scott, 1977), was diverted
from the watershed with the construction of a regional
wastewater collection system. In 1976, the remaining
point sources to the lake were similarly diverted.
Despite the elimination of these nutrient sources,
however, the lake continued to experience blooms.
Surface area 576 ha
Mix. depth It.9 a
Mean Depth 5.4 m
Volume 30.4 x 106
Uaterahed area 22,011 ha
T/?£A TM£tST
Figure 1. — Annabessacook Lake.
-------
418
RESTORATION OF LAKES AND INLAND WATERS
A 208 Water Quality Management Study (Sage and
Moran, 1977) identified agricultural runoff from area
farms and internal loading from the lake sediments as
the primary reasons for the continued blooms. Based
on that study, the Cobbossee Watershed District
applied for and received a 314 lakes restoration grant
from the U.S. Environmental Protection Agency. The
agricultural phase of the restoration project involved
developing and implementing agricultural waste man-
agement plans for three farms representing 90 percent
of the animal units in the watershed. The waste
management plans centered around the construction
of manure storage facilities, to eliminate the need for
winter manure spreading. This is expected to signifi-
cantly reduce the spring runoff, and thereby, the
phosphorus loading to the lake. For further information
on the agricultural phase of this project see the paper
by Gordon (1980) in this volume.
FEASIBILITY OF ALUMINUM
Preliminary studies were conducted to determine the
feasibility of an aluminum treatment in Annabessacook
Lake. A seasonal phosphorus budget was developed to
verify the significance of internal loading, and a series
of laboratory tests were performed to determine if
aluminum compounds could effectively tie up phos-
phorus in Annabessacook water without causing
adverse ecological impacts.
Phosphorus Budget
Data were obtained for a phosphorus budget by
collecting daily total phosphorus samples, and flow
measurements on each of the five tributary streams to
the lake, as well as the outlet. Daily lake level
measiirerhents were made, ancl lake" phosphorus
profiles (1 - to 2-meter intervals) were obtained at least
biweekly. Precipitation was measured at a site within
1,000 meters of the lake, and its phosphorus
concentration determined. Estimates of phosphorus
inputs from overland flow and ground water were
taken from previous studies (Sage and Moran, 1977;
Prescott and Attig, 1977).
The phosphorus budget (Table 1 ) showed that
between June 10 and September 15 approximately
225 kilograms of phosphorus entered the lake from
external sources, while 480 kilograms left via the
outlet. During the same period, the in-lake phosphorus
mass increased by 1,500 kilograms. Inserting these
figures into a mass balance formula:
P,nt = P,n-lake — (PrPo)
Pint = internal phosphorus loading
Pin-iake =change in phosphorus
mass in the lake
Pi —sum of phosphorus inputs
from external sources
Po = phosphorus loss from the
lake via the outlet
yielded a value of 1,800 kilograms of phosphorus
attributable to internal sources. Phosphorus concen-
trations during the open water season ranged from 17
A
-------
NUTRIENT PREVENTION AND INACTIVATION
419
testing. Ratios yielding solutions with pH's in the 6 to 7
range had "dissolved" (-45/u filtered) aluminum levels
below the detection limit of .08 mg AI/1. At pH's
outside this range, dissolved Al concentrations rose to
potentially toxic levels. As a result of these tests and
others, a volumetric alum/aluminate application ratio
of 1:1.6 was chosen.
Sample
A1
A2
A3
B1
B2
B3
C1
C2
na
Table
Initial
Aluminum
(mg/l)
50
50
50
20
20
20
10
10
m
2. Alum-aluminate ratio testing.
Alum: Aluminate
volumetric ratio
1:1
1:1.8
1:3
1:1
1:1.8
1:3
1:1
1:1.8
1-3
PH
4.5
5.1 — 5.2
8.0 — 8.2
4.8
6.2 — 6.4
7.8 — 8.0
5.2 — 5.3
6.4
Residual
Aluminum (mg/l)
.45/u filtered
15
BDL'
.24 — .47
1.5 — 1.6
BDL
.25 — .40
.09 — .17
BDL
71—7 9HDI
•BDL — below detection limit
Another set of tests was used to determine the
efficacy of P removal by Al3* over a range of AI3+ and P
concentrations. Table 3 shows inorganic phosphorus
removal exceeded 98 percent for all aluminum
dosages.
Table 3. — Phosphorus removal alum aluminate ratio 1:1.6.
Aluminum
(mg/l)
50
50
20
20
10
10
5
5
Initial P
(/jg/i)
500
1000
500
1000
500
1000
500
1000
Final P
(A<9/l)
.45/j filtered
2
2
2
2
2
2
2
5-6
PH
6.4 — 6.6
6.5
6.5 — 6.6
6.5 — 6.8
6.5 — 6.7
6.6 — 6.7
6.5 — 6.8
.67 — 6.8
In addition to chemical tests, bioassays using fish
and macroinvertebrates as test organisms were run in
an attempt to identify potential toxicity resulting from
aluminum application. Chironomus plumosus, one of
only two benthic species found in the Annabessacook
hypolimnion during summer months, was placed in
500 milliliter flasks containing anoxic lake water and
2.5 centimeters of bottom sediment. Aluminum
compounds were added on a sediment area! basis in
proportion to the maximum dosage anticipated for the
project. This amount of aluminum (70,000 mg Al/m2)
formed a thick layer of white floe in the test containers.
The flasks were viewed daily, and the test organisms
were observed to be lying on the floe, and moving
through it as they might through any natural flocculant
substrate. Mortality increased with test duration,
exceeding 50 percent by day 30. In all cases, however,
the controls displayed higher mortality than the test
flasks (Table 4).
Table 4. Macroinvertebrate bioassay (Chironomus
plumosus).
Time
0
4 day:
test
control
15 day:
test
control
30 day:
test
control
DO
0.6
0.2-0.8
0.2-0.3
0.4-0.5
0.2-0.4
0.2
0.2-0.4
PH
6.7
5.0
6.7
5.8
7.0
5.9
6.5
Alka-
linity
62
8
65
16
79
14
78
Temp
9.0
9.0
9.0
9.0
9.0
9.0
9.0
No. No. Survival
Alive Dead %
24
15
13
7
10
2
1
2
11
7
13
7
96
88
54
43
22
In addition to the macroinvertebrate tests, 96-hour
static bioassays using golden shiners (Notemigonus
crysoleucus, a lake inhabitant), were conducted. The
tests were carried out in 1-gallon glass jars, each
containing two fish in 3 liters of lake water. Test results
(Table 5) show no mortality over the entire range of
aluminum concentrations. The fish did not appear to be
stressed by the test conditions, nor did they avoid the
aluminum floe even in the 100 mg Al/l jars.
These tests indicated that by careful manipulation of
alum/aluminate ratios and dosages, phosphorus can
be effectively removed from lake water. At the same
time pH levels could be held within an acceptable range
to minimize potential toxicity problems.
Table 5. —
Aluminum
(mg/l)
0
1
10
100
96-hour static bioassay
crysoleucas).
pH
(0-hour)
6.8
6.7- 6.9
6.9 - 7.0
7.1
PH
(96-hour)
6.9-7.0
6.9 - 7.0
7.0- 7.1
7.0- 7.1
(Notemigonous
% Survival
at 96 hours
100
100
100
100
APPLICATION
From the phosphorus budget and lab test results, it
appeared that an aluminum treatment was a feasible
restoration technique for Annabessacook Lake. The
goal of the in-lake restoration phase was to mitigate
internal nutrient loading so that natural recovery of the
lake resulting from reduced external loading might be
accelerated.
Internal loading occurs from both oxygenated and
anoxic bottom sediment. (Lee, et al. 1977). Phosphorus
release rates are generally far greater in anoxic
sediment, but such conditions are generally confined to
hypolimnetic sediment. Low vertical diffusion rates in
the metalimnion can severely restrict phosphorus
movement from the hyponmnion into the epilimnion,
thereby limiting its availability for algal assimilation
(Schindler, Hesslein, and Kipphut, 1976; Sweers,
1970). However, significant nutrient transfer can occur
in lakes which experience a metalimnetic phosphorus
buildup, and are also subject to such phenomena as
thermocline migration and/or internal seiches (Stauf-
fer and Lee, 1974; Mortimer, 1971).
-------
420
RESTORATION OF LAKES AND INLAND WATERS
Phosphorus from oxygenated epilemnetic sediment,
though less rapidly released, is more immediately
available for primary production than hypolimnetic
released phosphorus, and can be a significant internal
source. In smaller lakes, where the littoral zones
comprise a relatively large portion of the lake area,
littoral inputs may be particularly important, especially
where inputs are enhanced by groundwater inflow and
macrophyte pumping (Twilley, et al. 1 977; McRoy, et al.
1972). Conversely, in larger eutrophic lakes, seasonal
nutrient inputs from the hypolimnion may dominate. In
most lakes it is likely that a number of internal sources
contribute nutrients, at least on a seasonal basis.
Annabessacook Lake has both a large anaerobic
hypolimnion and a macrophyte-covered littoral zone.
The lake annually shows a large phosphorus buildup
concurrent with the loss of oxygen in the hypolimnion.
This buildup extends into the thermocline by mid-
summer at which time it is likely that considerable
transfer to the epilimnion occurs. Also the lake has
historically experienced fall blooms that correspond
with fall overturn when the nutrient-enriched bottom
waters become incorporated in the rest of the lake. It
was felt that if hypolimnion P was made unavailable,
fall blooms might be less severe. The reduction of any
phosphorus transferred out of the hypolimnion to the
littoral zone during overturn might mean that this
phosphorus would not be available through some form
of littoral release at a future date. Although littoral zone
phosphorus release might be significant in Anna-
bessacook Lake, the vast size of the littoral zone (400
hectares) presented enormous logistical problems.
It was decided that a hypolimnetic aluminum
application, covering the entire area of anaerobic
sediment, would be most appropriate for Annabes-
sacook. Such an application would accomplish two
objectives. First, the application would strip the
hypolimnion of phosphorus by precipitation and
entrapment. This could be expected to be most effective
if done in mid to late summer when hypolimnetic
phosphorus concentrations in Annabessacook Lake are
greatest, and 95 percent of the hypolimnetic P was an
orthophosphate. Second, aluminum floe would chemi-
cally seal the sediment, preventing future phosphorus
release. Through a hypolimnion application, maximum
aluminum concentration could be achieved in the area
of greatest release. Aluminum application dosages in
the top meter of treated water were 25 and 34 mg AI/1
for areas 7 to 10 meters and over 10 meters deep,
respectively.
The simultaneous placement of two chemicals over
an area of 150 hectares at a depth of 7 meters
presented a sizable logistics problem. Because of .the
large amounts of time and travel required to resupply, a
single large capacity vessel rather than a number of
smaller vessels was chosen for the application. Both
the commercial aluminum sulfate and sodium alumi-
nate were obtained in liquid form for ease of handling.
The chemicals were delivered to the lakeside base of
operations at staggered intervals in 3,500 to 4,700-
gallon tank trucks. At the base, the chemicals were
temporarily stored in two 1.2 x 5.49 meter diameter
polyvinyl-lined swimming pools, erected especially for
this project. The pools, with a capacity of 7,600 gallons
each, proved to be very adequate holding facilities. The
chemicals were pumped from the pools approximately
25 meters to a three-compartment, 23 m3 mild-steel
tank truck, mounted on a 12 m x7.6 m barge (Figure 2).
The barge, a series of iron pontoons, was transported
from Portland, Maine and placed in Annabessacook
Lake by the Maine National Guard.
Figure 2. — Aluminum application barge.
The tank truck was valved so that each compartment
could deliver its product via pumps to a completely
segregated dual diffuser system. It was necessary to
maintain complete separation of the chemicals,
because contact led to the instantaneous formation of a
precipitate which could clog the diffuser. The diffuser,
constructed of 5 centimeter diameter black iron pipe,
formed a 8.8 m x7.6 m rectangle with 1.5 m extensions
at each end of one of the longer sides. The other 8.8 m
side rested in a cradle and straddled the barge
approximately amidship, with the elongated side
extending out over the bow of the barge. With this
arrangement, the diffuser was able to pivot from a
horizontal, above-water traveling position, to a vertical
applying position, reaching a maximum of 7.5 meters
below the water surface. The diffusing pipes were
drilled with 6 mm holes every 15 cm over their 11.9 m
lengths. The holes of the two pipes were positioned to
coincide with one another and were angled to provide
good chemical mixing when the system is as operation.
The overall diffuser system weighed over 360
kilograms and was raised or lowered by a winch in the
center of the barge, and a block-and-tackle on either
side. The diffuser was positioned at 7 meters below the
water surface during application.
The chemicals were delivered from the tank truck to
the diffuser by two 3-horsepower gasoline driven
pumps through sections of 5 cm diameter hose.
Valving, both on the tank truck and in-line, and flow
meters accurately regulated the chemical^ supply.
Dosage rates were coordinated with barge speed and
depth of area being treated. Surface trials of the system
showed excellent floe formation and even dispersal
along the length of the diffuser. The use of quick-
connect hoses permitted easy and efficient flushing of
the pumps and diffuser at the end of each day's
application. Flushing was necessary because of the
corrosive nature of the chemicals.
-------
NUTRIENT PREVENTION AND INACTIVATION
421
first, these buoys were kept to the outside of the barge,
and the previously dropped buoys were picked up by
trailing boats. This insured that the entire area would
be covered, using a minimum number of buoys.
The aluminum application took approximately 18
days, averaging 10 hours per day for a standard crew of
five persons. In addition, four to five local volunteers
manning two to three boats were necessary to
coordinate buoy placement and pickup.
There was a considerable amount of down-time
during the application phase due to engine failures,
tank leaks, and damage to diffusers. Despite these
problems, approximately 95 percent of the area
originally targeted for treatment received treatment.
The barge was propelled by two powerboats (75
horsepower and greater), one on each side toward the
stern of the barge. Steering was accomplished by
varying the engines' speeds, and/or direction, of
thrust. This combination provided excellent maneuver-
ability. The barge was able to maintain a speed of 1 to
2.5 miles per hour under moderate wind conditions.
Aluminum treatment was carried out where depths
exceeded 8 meters. The treatment route generally
followed a pattern of decreasing concentric circles. To
distinguish treated areas from untreated areas, buoys
were dropped off the inside of the barge every 100
meters. On the next pass, a smaller circle inside the
POST APPLICATION RESULTS AND
CONCLUSIONS
The hypolimnetic aluminum treatment, combined
with the agricultural waste management controls,
dramatically reduced phosphorus content in Annabes-
sacook Lake in 1979 (Figures 3 and 4). The maximum
phosphorus mass in the lake in 1979 was 1,030
kilograms, compared to over 2,200 kilograms in 1977,
a reduction of greater than 50 percent. Internal
recycling in 1979 contributed 625 kilograms phos-
phorus from spring overturn to mid-August when the
phosphorus mass in the lake reached its highest level.
This represents a 65 percent reduction from the 1,800
kjlograms attributed to internal loading in 1977.
^Phosphorus decreased in both the epilimnion and the
hypolimnion in the post-application year. The disparity
between the 2 years' concentrations became increas-
ingly great, especially in the hypolimnion, as the
summer proceeded. Over the summer, the seasonal
increase in hypolimnetic phosphorus mass was only a
third of the 1977 increase, 320 kilograms compared to
1,100 kilograms, despite similar temperature and
dissolved oxygen conditions for the 2 years. The
implication is that the aluminum floe effectively sorbed
sediment-released phosphorus. The increase in hypo-
limnetic phosphorus that did occur might have been
caused either by some fugitive sediment-released
material, or mineralization of sedimented algal cells
raining down on the floe.
Epilimnetic phosphorus levels were also reduced in
1979. This was especially true in the early summer,
when post-treatment concentrations showed very little
increase. Only later in the summer, when weather
conditions facilitated phosphorus transfer from the
hypolimnion did epilimnetic phosphorus concentra-
tions rise.
Figure 3. — Total phosphorus isopleths in Annabessacook
Lake, 1977 and 1979.
seoo-
8OO-
~<& 400-
•5.
L
7X/1 f>HOSf>
\ C
g
n
L
£f>
ffl
1=
/2.//1/A
ffi ^377-
1/37.9
'/OA/
-1
HYfOi /M/V/OM
8OO-
f800-
MAY
I i/i/A
J
-i
-
' I c/^/£ 1 Xt/ff 1 Jf^1
Figure 4. — Total phosphorus mass in Annabessacook Lake,
1977 and 1979.
-------
422
RESTORATION OF LAKES AND INLAND WATERS
The reasons for the reduced phosphorus levels in
1979 are not completely understood. It seems likely
that they center around the aluminum treatment,
although climatological conditions and reduced extern-
al loading from agricultural lands may also have
contributed. It appears that over the summer the
aluminum floe was effective in sorbing internally
regenerated phosphorus, thereby suppressing hypo-
limnetic levels, and eventual transfer to the epilimnion.
In addition, littoral loading may also have been reduced
following the movement of phosphorus-binding alumi-
num to the littoral zone during fall and spring
overturns.
Whatever the reasons, the dramatic results were
also manifested in the Secchi disk and chlorophyll a
data (Tables 6 and 7).
Table 6. Annabessacook Lake Secchi disk visibility (meters)
monthly means.
Table 8. — Secchi disk visibility days*
May
June
July
August
September
October
Mean
1975
4.1
4.2
2.6
2.1
2.1
3.0
1976
3.1
3.7
2.2
2.9
1.5
2.4
2.6
1977
3.2
2.0
1.2
1.1
2.3
1.9
2.0
1978
3.0
3.1
1.1
1.8*
1.8
2.5
2.2
1979
3.5
3.9
3.4
2.6
3.8
3.9
3.5
' Aluminum Application
Table 7. — Annabessacook Lake chlorophyll a (fjgfl) monthly
means.
1976
1977
1978
1979
May
June
July
August
September
October
Mean
5.4
4.5
6.2
6.8
14.9
15.6
8.9
11.5
11.5
18.7
8.7
24.4
17.8
15.4
6.2
29.3
23.6*
17.2
19.0
19.1
4.7
4.7
6.2
12.7
8.7
7.4
7.4
' Aluminum Application
Especially encouraging was the absence of a fall
bloom which in the past has been stimulated by the
incorporation of nutrient-rich hypolimnetic water into
the epilimnion.
The summer of 1979 was the first since records have
been kept in which Secchi disk visibility was always
better than the Maine-designated bloom level of 2
meters. A review of the restoration progress made on
Annabessacook Lake resulting from watewater diver-
sions, agricultural waste management, and the
aluminum treatment, and reflected by changes in
Secchi disk visibility is presented in Table 8.
To date the results of the aluminum application are
very encouraging, but the degree of success cannot be
judged until sufficient data are available.
It can be stated though, that the Annabessacook Lake
Restoration project has shown that a large scale
hypolimnetic aluminum application on a soft-water
lake is a feasible lake restoration technique.
Secchi disk
Visibility (Meters)
0 — 0.9
1 — 1.9
2 — 2.9
3 — 3.9
4 — 4.9
5+
1972
(Before
sewage
diversion
43 days
63 days
0 days
0 days
0 days
0 days
1977
(Before
lake rest-
oration
project)
10 days
67 days
28 days
1 day
0 days
0 days
1979
(After
aluminum
treatment
and agriculture
controls)
0 days
0 days
30 days
40 days
35 days
1 day
* based on a summer season from June 1 — Sept. 15.
REFERENCES
Browman, M. G., R. F. Harris, and D. E. Armstrong. 1973.
Lake renewal by treatment with aluminum hydroxide. Draft
rep. to Wis. Dep. Nat. Resour., Madison.
Cooke, G. D., et al. 1977. The occurrence of internal
phosphorus loading in two small, eutrophic, glacial lakes in
Northeastern Ohio. Hydrobiology 56:129.
Cooke, G. D., et al. 1978. Effects of diversion and alum
application on two eutrophic lakes. EPA/3-78-033. Environ.
Res. Lab. U.S. Environ. Prot. Agency, Corvallis, Ore.
Everhart, W. H.,andR.A. Freeman. 1973. Effects of chemical
variations in aquatic environments, Vol. II. Toxic effects of
aqueous aluminum to rainbow trout. Ecol. Res. Ser. EPA-
R3-73-0116. U.S. Environ. Prot. Agency.
Gordon, T. U. 1980. Local commitment to lake restoration:
The Cobbossee Watershed example. In Proc. In Symp.
Inland Waters Lake Restor. U.S. Environ. Prot. Agency,
Washington, D.C.
Higgins, B., S. C. Mohleji, and R. L. Irvine. 1976. Lake
treatment with fly ash, lime, and gypsum. Jour. Water
Pollut. Control Fed. 48:2153.
Kennedy, R. H. 1977. Personal communication. U.S. Army
Corps, Eng., Vicksburg, Mass.
Knaur, D. R., and P. J. Garrison. 1979. A status report on the
Mirror/Shadow Lakes evaluation project. Unpubl. rep. to
Off. Inland Lake Renewal Wis. Dep. Nat. Resour. Madison.
Lee, G. F., et al. 1977. Significance of oxic vs. anoxic
conditions for lake Mendota sediment phosphorus release.
In H. L. Golterman, ed. Symp. Interaction Between
Sediments and Freshwater. Amsterdam, 1976. W. Junk,
The Hague.
McRoy, C. P., R. J. Barsdale, andM. Nebert. 1972. Phosphorus
cycling in an eel grass (Zostera marina) ecosystem. Limnol.
Oceanogr. 17:51.
Mortimer, C. H. 1971. Chemical exchanges between
sediments and water in the Great Lakes - speculations on
probable regulatory mechanisms. Limnol. Oceanogr.
16:387.
Peterson, J. 0. 1977. Personal communication.
Peterson, S. A., et al. 1976. Laboratory evaluation of nutrient
inactivation compounds for lake restoration. Jour. Water
Pollut. Control Fed. 48:817.
Prescott, G., and J. W. Attig. 1977. Geohydrology of part of
the Androscoggin River basin, Maine. U.S. Geol. Survey
open file rep. 78-297.
Sage, K., and E. Moran. 1977. Annabessacook Lake study.
Cobbossee Watershed District, Winthrop, Maine.
-------
NUTRIENT PREVENTION AND IMAGINATION 423
Sawyer, C. N., and P. L. McCarty. 1967. Chemistry for
sanitary engineers. McGraw Hill Book Co., New York.
Schindler, D. W., R. Hesslein, and G. Kipphut. 1976.
Interactions between the sediments and overlying waters in
an experimentally eutrophied Precambrian shield lake. In H.
.L Golterman, ed Symp. Interaction Between Sediments and
Freshwater. Amsterdam, 1976. W. Junk, The Hague.
Scott, M. 1971. The estimated nutrient budget of Annabes-
sacook Lake. Presented at Workshop on the Reclamation of
Maine's Dying Lakes, March 24-25, Orono.
Stauffer, R. E., and G. F. Lee. 1973. The role of thermocline
migration in regulating algal blooms./nE.J. Middlebrooks.et
al, eds. Modeling the eutrophication process. Utah Res. Lab.,
Utah State University, Logan.
Sweers, H. E. 1970. Vertical diffusivity coefficient in a
thermocline. Limnol. Oceanogr. 15:273.
Twilley, R. R., M. M. Brenson, and G. T. Davis. 1977.
Phosphorus absorption, translocation and secretion in
Nuphar luteum. Limnol. Oceanogr. 22:1022.
-------
424
MEDICAL LAKE IMPROVEMENT PROJECT:
SUCCESS STORY
A.F. GASPERINO
M.A. BECKWITH
G.R. KEIZUR
Battelle, Pacific Northwest Laboratories
Richland, Washington
R.A. SOLTERO
D.G. NICHOLS
J.M. MIRES
Eastern Washington University
Cheney, Washington
ABSTRACT
Medical Lake is an alkaline lake in Eastern Washington State that has historically exhibited
nuisance algal blooms, extensive summer anoxia, and high nutrient concentrations. The lake is
located within the corporate limits of the town of Medical Lake and its eutrophic condition resulted
primarily from internal phosphorus cycling. The lake lies in a closed basin with a drainage area of
3.5 km2. Land use is predominantly residential. Muncipal sewage collection and treatment began
in 1964. The lake was treated with alum during August and September 1977, with dramatic
results. Total phosphorus and orthgphosphorus concentrations have declined from over 400 and
300 /ugP to less than 60 and 3 /jgl ', respectively. Chlorophyll a has remained below 5 fjgl~\ and
often is undetectable. Secchi disk depths have averaged 5 m and ranged from 2.5 to 11 m. Before
treatment Secchi depths averaged 1.2m. The extent of summer and winter anoxia has declined
and increased oxygen levels have improved the fishing habitat. Fifteen thousand 6.3 cm rainbow
trout fingerlings were planted in the lake during June 1978. These fish currently exceed 1 kg in
weight and 50 cm in length. Recreational use of the lake and shore front park has increased
accordingly. Activities include swimming, boating, and picnicking. Furthermore, public fishing is
expected to be permitted during 1981.
INTRODUCTION
Medical Lake is a freshwater lake located near the
town of Medical Lake, approximately 24 kilometers
southwest of Spokane, Wash. For several decades, a
high phosphorus concentration contributed to the
recurrence of algal blooms and floating mats of algae in
the lake. Along with the thick algal surface scum,
offensive odors were associated with decaying algae
and hydrogen-sulfide-laden bottom waters. These
conditions allowed only limited use of the lake.
To improve recreation, the town of Medical Lake
sponsored a project to restore the water quality of the
lake. The project was directed by Battelle, Pacific
Northwest Laboratories, with major support from
Eastern Washington University. Financial support
came from the U.S. Environmental Protection Agency,
the State of Washington Department of Ecology on a
matching basis, the town of Medical Lake, and Spokane
County. The objectives of the project were to reduce
phosphorus and algae levels, increase oxygen levels,
and improve water clarity to permit recreational use of
the lake and possibly establish a fishery.
Data collection and laboratory analyses showed that
the high phosphorus concentration came from an
internal phosphorus cycle. Consequently, itwasdecided
that the best method for improving water quality would
be to disrupt the cycle. Of the procedures available,
phosphorus inactivation by chemical precipitation
appeared to be the most effective and economical
method. The technique chosen for phosphorus
inactivation consisted of multiple applications of
aluminum sulfate (alum).
The project began in June 1977. The alum was
applied over a 5-week period beginning in August
1977 Water quality monitoring was conducted prior to,
during, and after the application. Preliminary results
reported by Gasperino, et al. (1978) indicated that the
treatment had reduced phosphorus and algae concen-
trations and increased water clarity. Water quality
monitoring continued until June 1980.
This paper presents results of the restoration project
through December 1979. Detailed results are available
in the project's final report (Gasperino, et al. 1980).
Additional data on methodology development can be
found in the preliminary report (Gasperino, etal. 1978).
A fisheries report which includes water quality data
through 1980 will be prepared during 1981.
-------
NUTRIENT PREVENTION AND INACTIVATION
425
LAKE HISTORY AND CHEMISTRY
Medical Lake lies in a closed basin within the
corporate limits of the town of Medical Lake. The basic
physical characteristics of the lake are as follows:
Area 158 acres 64 ha
Volume 5,026 acre-ft 6.2X106m3
Maximum Depth 60 feet 18m
Mean Depth 33 feet 10 m
Maximum Length 5,600 feet 1.7 km
Maximum Width 1,300 feet 0.4 km
The basin was scoured from basalt of the Columbia
River group by recurring glacial floods. The largest
flood occurred between 18,000 and 20,000 years ago.
Land use in the drainage area (3.5 km2) is predomin-
ately residential. Approximately 36 percent of the
shoreline length (5.0 km) is developed. Municipal
sewage collection and treatment have been operational
since 1964. Prior to 1964, septic tanks and cesspools
were employed.
A familiar vertebrate inhabitant of the lake is the
painted turtle (Chrysemys picta). The only known fish
populations to inhabit the lake prior to the alum
treatment were small stocks of tench (Tinea tinea) and
carp (Cyprinis carpio). Tench still inhabit the lake. Since
the completion of the alum treatment approximately
30,000 rainbow trout (Salmo gairdneri) have been
introduced to the lake over 3 successive years.
Bauman and Soltero (1978) described the limnology
of Medical Lake in detail in 1974. Their study showed a
high concentration of phosphorus. Furthermore, they
concluded that most of the phosphorus was being
recycled within the lake. Pretreatment surveys during
the restoration project confirmed these earlier results.
Prior to treatment, the major sources of phosphorus
within the lake were decomposing algae and bottom
sediment, which released the nutrient throughout the
summer. Phosphorus was then mixed throughout the
lake during the fall. Thus, the algal production of one
growing season stimulated algal growth during the
following growing season. Very little phosphorus
probably enters the lake from the surrounding basin
because the lake receives no known sewage effluent or
agricultural runoff and has no surface inlets.
LABORATORY ANALYSES
After analyzing the lake's characteristics, studying
the literature, and reviewing other lake restoration
techniques, alum treatment was selected as the most
appropriate method for inactivating the phosphorus in
Medical Lake. Previous literature on eutrophication
indicated that an 87 percent orthophosphorus reduc-
tion was probably necessary to reduce algal blooms in
Medical Lake.
Before the alum was applied, laboratory analyses
were made to determine the quantity rate of
application, and type of mixing required. The following
requirements were necessary to achieve an 87 percent
reduction in orthophosphorus*:
1.A whole-lake alum (Ab (S04); IShbO) concentra-
tion of 150 mg 1-1;
2.Vigorous mixing of the alum as a liquid slurry
rather than as dry crystals;
3. Multiple doses rather than a single dose; and
4. Combined surface and subsurface applications
rather than surface applications alone.
'.Orthophosphorus is soluble reactive phosphorus.
For the laboratory analyses, orthophosphorus was
measured as an indication of overall phosphorus
reduction.
Detailed results of these tests are presented in the
preliminary and final reports (Gasperino, et al. 1978,
1980).
ALUM DISPENSING SYSTEM
Once the parameters required for treatment were
determined, a dispensing system was needed to
provide a fast, safe, and efficient means of placing the
alum into the water. Two pontoon barges were used to
distribute alum in a well-mixed, uniform concentration
at prescribed depths: a 12-meter barge for deep areas,
and a 8.5-meter barge for shallow areas. Each barge
was equipped with tanks for carrying the alum, a
distribution pump, and an injection manifold. The
injection manifold allowed alum distribution at the
surface or 4.5 meters.
The time for dispensing a load of alum varied
depending on the type of application. Subsurface
injection took about 45 minutes with the 12-meter
barge and 25 minutes with the 4.5-meter barge.
Surface applications were faster. The barges could
increase speed because less manifold drag occurred
and, consequently, the pumping rate was increased to
keep the volume of alum dispensed per dispensing
zone constant. Figure 1 illustrates the application
barge.
MANIFOLD-SURFACE
APPLICATION
ALUM TANK
Figure 1. — Alum dispensing system.
For the alum application, the lake was divided into six
equal zones with marker buoys to facilitate a
systematic distribution. The barge pilots treated each
zone in a series of back and forth passes orienting
themselves by the marker buoys and landmarks on
shore. The sequence of passes was: two subsurface
applications, two surface applications, two more
subsurface applications, then one surface and one
subsurface application. The second application in a
zone was not made until all other zones had been
treated.
RESULTS
The results of the water quality monitoring through
June 1980 show that the alum treatment was highly
successful in decreasing phosphorus levels, eliminat-
-------
426
RESTORATION OF
INLAND WATERS
ing algal blooms, and increasing water clarity. Thirty-
two sampling cruises were completed between
January 17, 1978 and December 10, 1979. Biweekly
samplings were made June through September and at
monthly intervals for the balance of the study. One
station, at the deepest point, was sampled throughout
the project. Water samples were taken at 2-meter
intervals from the surface to the bottom with a 1 -liter
Kemmerer sampler. Also, a euphotic zone composite
was collected by combining samples(usually taken at 1
meter intervals) of equal volume from the surface to the
lower limit of the euphotic zone.
Complete profiles of chemical and biological para-
meters are presented in the final report (Gasperino, et
al. 1980). Figures 2 through 6 illustrate average lake
concentrations of important water quality data. Most of
these data indicate that a substantial improvement has
occurred as a result of the treatment.
Figure 2 shows the mean monthly total and
orthophosphorus concentrations from December 1976
through December 1979. Prior to the alum treatment,
the mean monthly total and orthophosphorus con-
centrations were approximately 0.47 and 0.31 mg I-1,
respectively. Concentrations for both fractions declined
immediately following treatment, but it was not until
fall turnover that the impact of the alum application
was fully realized. A comparison of overall mean total
and orthophosphorus concentrations, before October
and after November 1977, showed that each fraction
decreased approximately 87 and 97 percent, respect-
ively.
ALUM TREATMENT
FALL TURNOVER
DJMAMJJ ASONDJ FMAMJJ ASONDJ FMAMJJ ASOND
1976 1977 1978 1979
Figure 2. — Mean Monthly Total and Orthophosphorus
Concentrations (mg 1 1 P) Before, During and Following
Treatment.
The cause for the substantial reduction of phos-
phorus at fall turnover, particularly during 1977, is not
clearly understood. A possible explanation of the
decline could be that the sedimented floe still
possessed phosphorus sorptive properties. Therefore,
additional phosphorus removal was effected by
circulation of the entire water mass. This mechanism
might also explain why phosphorus concentrations
continued to decline in 1978 and 1979.
Figure 3 presents the mean monthly chlorophyll a
values for all months of study. The overall mean
chlorophyll a concentration prior to and during the
alum treatment (December 1976 to September 1977)
was approximately 25.2 mg m-3. The overall mean
concentration for 1978 and 1979 was 3.20 mg m-3, a
decrease of 87 percent. Before treatment the maximum
mean monthly chlorophyll a concentration was 59.8
mg m-3 in May of 1977, while the maximum post
treatment value was 17.5 m-3 in February 1978. Mean
growing season (May-September) values for chloro-
phyll a concentrations during 1977, 1978, and 1979
were 16.7, 2.19, and 2.50 mg m-3. respectively.
ALUM TREATMENT
JMAMJj ASONDJ F MAMJ J AS ON D J F MAMJ J ASON
Figure 3. — Mean Monthly Chlorophyll a Concentrations
(mg rrf3) Before, During and Following Alum Treatment.
Figure 4 shows the mean monthly dissolved oxygen
concentrations for the water column before (December
1976 to July 1977), during (July 1977 to September
1 977) and following (October 1 977 to December 1 979)
the alum treatment. Rather large fluctuations in
concentration were evident before and during the
treatment. Following treatment, variation in mean
monthly oxygen concentrations decreased with levels
tending to be near 5 mg 1~1 Decomposition of the
sedimented organic matter, resulting from the treat-
ment, has probably negated significant improvement in
the overall dissolved oxygen regime of the lake. In the
future, overall oxygen concentrations should signi-
ficantly increase with the stabilization of the sedi-
mented materials.
7.0
6.0
5.0
2 4.0
^ 3.0
2.0
1.0
0
ALUM TREATMENT
D
976
r-
L
-
J-
-
>.6
AMMONIA mj TOTAL
~
-
MAMJJASONO
1977
-1
-
- -
--
-
-
JFMAMJ JASONI
1978
r
I-.---
J FMAMJ JASOND
H79
Figure 4. — Mean Monthly Dissolved Oxygen Concentrations
(mg T1) Before, During and Following Alum Treatment.
-------
NUTRIENT PREVENTION AND INACTIVATION
427
Figure 5 presents the mean monthly total and
ammonia nitrogen concentrations before, during, and
following the alum application. Prior to the alum
treatment the total nitrogen concentrations for the
water column approached 3.2 mg I-1; however, post-
treatment to to 0.9 mg I-1 following treatment (about a
40 percent decrease).
B
a
u
H
H
"i"
D
I
i
4
ALUM TREATMENT
M
JMAMJ JASONDJ FM AMJJ AS 0 N D|J FMAMJ JA50ND
N7t 1977 1978 1979
Figure 5. — Mean Monthly Total and Ammonia Nitrogen
Concentrations (mg 1"1 N) Before, During and Following
Alum Treatment.
Water clarity has significantly improved since the
alum treatment (Figure 6). The overall mean Secchi
disk visibility before the treatment was 2.4 meters;
after the treatment it was 4.9 meters. The improved
water clarity is a tangible indication of reduced algal
growth.
|-|
ALUMT
t
REA
D J MAMJ J ASOND
»76 1977
TMEU
IT
-
-II-
F MAMJ J A S OND J
1978
-
-
pi
-,
F MAMJ j A SON D
1979
Figure 6. — Mean Monthly Secchi Disk Visibilities (meters)
Before, During and Following Alum Treatment.
In 1977, the phytoplankton community was domin-
ated by the Cryptophyceae and Cyanophyceae (Soltero,
et al. 1978). The Cryptophyceae reached a maximum
standing crop of 20.52 mm3 I-' in May 1977 while the
Cyanophyceae, primarily Microcystis aeruginosa,
reached bloom proportions August 3 just prior to the
alum treatment. Since the treatment, the Chlorophy-
ceae and Cryptophyceae have dominated the phyto-
plankton community. In 1978, the chlorophyceaen
standing crop steadily increased from a low of 0.09
mm3!"1 in February to a seasonal high of 3.17. The
maximum cryptophyceaen standing crop occurred in
March during both study years, 1.57 mm3!"1 and 2.01
mm3!"1 in 1978 and 1979, respectively. Except for a
small blue-green pulse in the summer of 1978,
primarily consisting of Synechocystis sp. and Oscil-
latoria tenuis, the mean monthly cyanophyceaen
standing crop during both study years was less than
0.07 mm3!"1. Contributions to the total cell volume by
the remaining phytoplankton classes were minimal
during both years.
In 1977, a shift in the primary growth limiting
nutrient (\oSelanastrum capricornutum as determined
by algal assay) from nitrogen to phosphorus occurred
following the alum treatment (Soltero, et al. 1978).
Algal assay results for 1978 show that phosphorus
continued to limit the growth of Selenastrum through-
out the study year. The extent of phosphorus limitation
during 1978 was also evident by the overall high total
inorganic nitrogen to orthophosphate ratio (approxi-
mately 50:1) determined in the euphotic zone.
Keizur (1978) suggested that the decline and
replacement of blue-green algae with more palatable
greens and cryptomonads could result in decreased
rotifer numbers with a corresponding increase in
daphnid and diaptomid density. He proposed that the
macroconsumers, such as Daphnia and Diaptomus,
would make up a greater majority of the zooplankton
standing crop in response to the phytoplankton shift.
He further suggested that if fish stocking was
implemented, the larger macroconsumers would
provide an excellent food source for the fish.
Since the alum application, a substantial reduction of
blue-green algae has occurred with a corresponding
replacement by greens and cryptomonads. In response
to this change, the rotifer population declined with a
corresponding increase in daphnid and diaptomid
populations. Thus, a greater proportion of the zoo-
plankton community now consists of rnacroconsumers.
Medical Lake was stocked with 14,000 rainbow trout
fingerlings (Salmo gairdneri) in June 1978, 12,000 in
June 1979, and an additional 4,000 fingerlings a year
later. Preliminary results of an ongoing fisheries study
indicate that fish growth and condition are excellent
(Knapp, pers. commun.). Some of these 6.3 cm
fingerlings have grown greater than 50 cm and weigh
more than a kilogram. Stomach analyses of the trout
revealed that the larger and abundant macroconsumer,
Daphnia pulex, has been an excellent food source.
Continuing studies suggest, however, that the fish
predation has caused a size reduction in D. pulex and
affected the appearance and increased the numbers of
smaller zooplankters (Mires, 1980). Since modification
of the zooplankton community as a result of selective
grazing can greatly influence the composition of the
phytoplankton standing crop (Brooks and Dodson,
1965), it is important that a large complement of
macroconsumers (i.e., D. pulex) be maintained in
Medical Lake to help minimize algal standing crop, a
primary objective of the restoration project. Elimination
of large grazers because of excessive fish predation
will only shorten the life expectancy of the restoration.
Present understanding of the food chain balance would
indicate that further fish stocking would be inadvisable
at this time. Further study of the trout, zooplankton, and
-------
428
RESTORATION OF LAKES AND INLAND WATERS
phytoplankton relationships will give more exact
information as to the number of fish Medical Lake will
support without compromising its improved water
quality.
PROJECT COSTS
Total project costs (Table 1) included the cost of alum;
labor costs for monitoring alum application, data
analysis, and project management; and equipment
rental and outfitting. Water quality monitoring and data
analysis accounted for a large part of the expenditures.
The price of the alum was also significant.
Table 1. — Project costs for Medical Lake restoration.
Alum
Labor
Monitoring (1/77-6/80)
(Biological Physical, Chemical)
Chemical Application (8/77-9/77)
Project Management, Planning,
Coordination and Data
Ana lysis (1/77-6/80)
$ 90,000
55,000
5,000
72,000
Equipment
Bond
Barge Rental
Vehicle Rental
Pumps, Supplies
TOTAL
8,000
1,150
7,250
1,500
$239,900
CONCLUSIONS
The alum treatment of Medical Lake has significantly
improved the lake's water quality. Mean concentra-
tions of total and orthophosphorus steadily declined
from January through December during both study
years. Mean orthophosphorus concentrations declined
5.5 and threefold in 1978 and 1979, respectively. Total
phosphorus concentrations decreased approximately
twofold during each study year.
Mean monthly total and ammonia nitrogen levels
were lower in 1979 than 1978. Since the alum
treatment, mean total and ammonia nitrogen concen-
trations have declined 22 and 40 percent respectively.
Algal assay results for 1978 showed that phosphorus
was the primary growth limiting nutrient to Selena-
strum. Phosphorus limitation was also evident by the
overall high 1978 euphotic zone total inorganic
nitrogen to orthophosphate ratio (approximately 50:1).
Chlorophyceae and Cryptophyceae dominated the
phytoplankton community during 1978 and 1979 with
the Chlorophyceae being the major contributor to the
total cell volume. Cyanophyceae, the dominant phyto-
plankton class in 1977, was a minor contributor to the
total cell volume in both study years. The overall mean
phytoplankton standing crop (mm3!'1) before and
during the application was reduced by 91 percent when
compared to the mean standing crop following
treatment. This decline in algal standing crop also
related well with the overall 87 percent reduction in
mean chlorophyll a concentration (mg m~3) following
treatment.
Water clarity has significantly improved since the
alum treatment. The overall mean monthly Secchi disk
visibility before and after treatment was 2.4 and 4.9
meters, respectively.
Hypolimnetic anoxia was evident during 1978 and
1979 with the periods of anoxia lasting 9 and 5
months, respectively. A sediment oxygen demand,
because of the stabilization of organic matter deposited
on the lake bottom as a result of the alum treatment,
was the probable cause of anoxia during both study
years. Any improvement in the dissolved oxygen
regime of Medical Lake as a result of the treatment
possibly has been negated by this sediment demand.
However, the volume of water that became anoxic in
1978 and 1979 was much less than that of 1977.
Heavy selective grazing, caused by recent fish
stockings, appears to have effected a reduction in mean
body length of D. pulex and may have promoted a
species composition shift as evidenced by the
appearance of other smaller zooplankters.
The improvement in water quality and clarity has
increased recreational use of the lake. Activities
include swimming, boating, picnicking, and fishing.
Property values around the lake have risen, and users
of the park have increased on summer weekends from
fewer than 100 to an estimated 1,000. As a result, the
town is seeking additional funds to improve the park
and docking facility. And for the first time, the town has
hired a lifeguard.
REFERENCES
Bauman, L.R., and R.A. Soltero. 1978. Limnological invest-
igation of eutrophic Medical Lake, Wash. Northwest Sci.
52:127.
Brooks, J.L., and S.I. Dodson, 1965. Predation, body size and
composition of plankton. Science 150:28.
Gasperino, A.F., et al. 1978. Restoration of eutrophic Medical
Lake, Washington, by treatment with aluminum sulfate:
Preliminary findings. Prepared for the town of Medical lake
by Battelle, Pacific Northwest Lab., Richland, Wash.
Gasperino, A.F., et al. 1980. Restoration of Medical Lake.
Final report. Prepared for the town of Medical Lake by
Battelle, Pacific Northwest Lab., Richland, Wash.
Keizur, G.R., 1978. An investigation of the zooplankton
community of Medical Lake, Washington, before, during
and after a whole-lake application of aluminum sulfate. U.S.
Thesis. Eastern Washington University, Cheney.
Soltero, R.A. et al. 1978. Limnological investigation of
Medical Lake, Washington, before, during and after a
whole-lake application of alum. Battelle, Northwest Spec.
Agreement: B-49803-B-H, Proj. Completion Rep. Eastern
Washington University, Cheney.
Mires, J.M. 1980. Zooplankton dynamics of Medical Lake,
Washington, after a whole-lake alum application and a
subsequent establishment of a trout fishery. M.S. Thesis.
Eastern Washington University, Cheney.
-------
429
DETERGENT MODIFICATION:
SCANDINAVIAN EXPERIENCES
CURT FORSBERG
Institute of Limnology
University of Uppsala
Uppsala, Sweden
ABSTRACT
In Scandinavia the concern over the role of detergents in water deterioration has been focused on
over-fertilization by tripolyphosphates. Agreements between authorities and industry have
lowered the phosphate content to about 7 percent as P. Detergents with lower P-content have
been based on NTA or soap. Detergents totally free of phosphate have also been manufactured,
based on a mixture of adipate-acetate or citrate. A special campaign for phosphate-free detergents
is running within the Lake Mjosa catchment area in Norway. The measures taken have reduced
phosphorus load on the water bodies, but are difficult to evaluate, as other P-sources also were
reduced parallel to the decrease of detergent P. In any case, no harmful effect of NTA has been
reported.
INTRODUCTION
In Scandinavia scientific concern about the role of
the synthetic detergents in water deterioration started
in Sweden at the end of the 1950's. The main interest
was focused on the role of the detergent polyphos-
phates in the over-fertilization of natural waters, the
eutrophication process.
During the 1960's the discussions became very
public and culminated when advanced wastewater
treatment for phosphorus removal became standard
during the early 1970's. When the discussions began,
however, there was no method for efficient phosphorus
removal at the sewage treatment plants. At this time,
replacement of phosphorus in the detergents by some
other builder — especially different types of chelating
agents — seemed to be the most rapid way to start
decreasing the P-load on our waters. It was also stated
that this was a partial solution and that a more total P-
removal was necessary to achieve real improvements,
especially in urban waters.
Different measures and modifications resulted,
among other things, in lower consumption of detergent
phosphorus. In Norway a special campaign for
increasing the sale of P-free detergents was found
necessary in the Lake Mjosa catchment area. In the
other Scandinavian countries little interest has been
devoted to the environmental effects of detergents.
DETERGENT MODIFICATIONS
IN SWEDEN
The most common commercial synthetic detergents
were tested and found to be excellent sources for algal
growth, although different inhibitory effects of the
other detergent components were noted (Forsberg,
Jinneraot, and Davidsson, 1967). Nitrilotriacetic acid
(NTA) was considered a good substitute for poly-
phosphate, stimulating numerous investigations. The
first reports on biological degradation of NTA appeared
in 1967 (Swisher, Crutchfield, and Caldwell, 1967;
Forsberg and Lindqvist, 1967a,b). Since that time many
papers dealing with NTA have been published.
Comprehensive lists of references have been present-
ed (Monsanto, 1977). NTA has been used in Sweden as
a substitute for polyphosphate since 1968. In addition,
other modifications and measures have also been
made, examples of which are given in Table 1.
Table 1. — Examples of detergent modifications in Sweden.
Compounds
Percent
Surface active agent
Sodium tripolyphosphate
Soap
NTA
EDTA
Sodium-adipate
Sodium-acetate
Sodiu m-carbonate
Sodium-silicate
CMC
Perborate
Mg-silicate
Optical whiteners
Sodiu m-su If ate
Perfume
H2O
A
13.0
30.0
0.5
18.0
8.0
1.0
I8.0
1.0
0.3
1.5
0.3
9.0
B
14.0
9.0
12.0
31.0
5.0
1.6
20.0
1.2
0.3
1.5
0.2
4.0
C
3.0
7.0
38.0
16.0
8.0
1.0
15.0
0.4
0.3
11.0
D
22.0
4.0
)39.0
2.0
4.0
0.5
27.0
0.1
1.4
The surface active agents earlier causing well-
known water pollution problems were changed to more
biologically degradable ones. This was finalized in
1969 for anionic tensides and in January 1973 for non-
ionic ones.
An agreement between the National Environment
Protection Board and industry, limited the phosphate
-------
430
RESTORATION OF LAKES AND INLAND WATERS
content to a maximum of 7.5 percent For 30 percent as
sodiumtripolyphosphate, and to 10 percent P in
machine dishwashing agents. Detergents with lower
P-content have been based on NTA or soap (Table 1, B
and C). Totally phosphate-free detergents have also
been manufactured, based on a mixture of adipate-
acetate (Table 1, D) or citrate.
Through these measures the phosphate amount in
household detergents was reduced from 4,100 tons in
1968 (calculated as P) to about 3,000 in 1970,
corresponding to 28 percent. After that the detergent P
in sewage constituted approximately 30 percent (Natl.
Swed. Environ. Prot. Board, 1972). The detergent-P per
capita was reduced from 1.8 to 2 g/p/day to 0.9-1.1
g/p/day.
The experiences obtained by using NTA as a
detergent chemical have not given the National
Environment Protection Board any reason either to
oppose this use or to work for a general changeover to
NTA. The Protection Board considers advanced waste-
water treatment for P-removal a safer and more
effective method to reduce the phosphate load on our
waters. The desire remains, however, for further
limitation of the detergent phosphates.
The NTA-based or totally P-free detergents have
never dominated the Swedish market. Therefore, it is
difficult to evaluate possible environmental effects of
these products. In any case, no harmful effects of NTA
have been reported.
In Sweden as well as in most European countries,
washing processes often are programmed at 80 to
90°C. Bleaching is then obtained by perborate, active
above 60°C. Environmental aspects of boron have been
summarized (R. Swed. Acad. Sci. 1970). Boron and its
compounds (including perborates) have not been found
to be acutely hazardous to the environment. The
discussions concerning possible harm to the environ-
ment also included fluorescent whitening agents and
enzymes. No special measures against these com-
pounds have been taken. Fluorescent whitening agents
were evaluated at the Stockholm Symposium in 1973,
arranged by the Center for Environmental Sciences,
Royal Institute of Technology, Stockholm (reported in
MVC-Report 2, 1973).
To be able to follow changes in detergent formula-
tions which might affect the environment, the
detergent producers submit annual reports to the
Environment Protection Board on detergent quantities
and ingredients. They also provide information on the
levels of phosphorus and organic chelators in each
individual product.
CAMPAIGN FOR PHOSPHATE—FREE
DETERGENTS IN NORWAY
Eutrophication problems exist mainly in southeast
Norway. Recently, the largest lake in Norway, Lake
Mjosa, with a surface of 365 km2 showed very bad
water quality due to a bloom of the blue-green alga
Oscillatoria bornetii fa. tenuis. The alga discolored the
water and gave the drinking water for 200,000 people
very unpleasant taste and odor (Holtan, 1979; Holtan,
et al. 1980). As phosphorus has been demonstrated to
be the algal growth-limiting nutrient (see e.g., Ryding,
1980), a campaign aimed at reducing phosphorus
pollution and saving this lake was initiated (Minist.
Environ. 1979). Only a total removal of P was
considered sufficient in the Mjosa area. This was
thought possible because of the normally very low
water hardness in the water supplies.
In the spring 1977, the environmental protection
authorities gave new guidelines for sale of detergents
in Mjosa's catchment area. The sale of phosphate-free
detergents was to be emphasized. After 6 months
phosphate-free detergents' share of the market rose
from 2 to 3 percent to 57 percent. To increase the use
of P-free detergents further, in February 1978, the
Ministry of Environment issued regulations in pur-
suance of the Act concerning Product Control. These
regulations "prohibit the exhibition of household
detergents containing phosphates near the front of
shop premises and also advertising of such deter-
gents." The lack of a total prohibition derives from
consideration for consumers using hard water. In early
summer 1978 the turnover of P-free detergents had
risen to about 70 percent. Decisions on whether taxes
on detergents or other measures are required will be
made.
The Mjosa campaign includes measures against
pollution from municipalities, rural areas, agriculture,
and industry at a cost of approximately 1 billion
Norwegian crowns.
From 1973 through 1976 the annual load of total-P
on Lake Mjosa averaged 320 tons/year. For 1977
through 1979 the corresponding values were between
230 to 252 (Holtan, et al. 1980). When the campaign is
completed, discharges of P will be about 200 tons/year
(Minist. Environ. 1979).
Since 1976 there has been no blue-green algal
bloom. An evaluation of the measures performed will
take some time as the period with reduced P-load also
had cold and rainy summers.
Generally, an agreement between government and
industry limits the P-content of fabric-washing pro-
ducts to a maximum 5.5 percent P For machine
dishwashing products there is no limitation. For the
Mjosa catchment area consumers are advised to use as
little as possible.
The government is considering whether NTA can be
used in detergents in Norway. The problems discussed
are the residual concentration of NTA which can
appear in natural waters and in drinking water, and
possibilities of heavy metal mobilization.
NO SPECIAL DETERGENT
MODIFICATIONS IN DENMARK
As far as is known, no comprehensive detergent
modifications resulting from environmental problems
have been performed in Denmark. Detergent P has not
been considered to be a problem in Denmark so far
since sewage is normally discharged into the sea. In
Finland an agreement within industry limits the P
content of fabric-washing products to 7 percent, as P.
NTA has been used as a substitute in Sweden.
DISCUSSION
The modified detergents have had no dominating
position on the Swedish market, which means that it is
-------
NUTRIENT PREVENTION AND INACTIVATION
431
difficult to evaluate their environmental influences.
The rapid development of advanced wastewater
treatment for P-removal eliminated the need for a more
total replacement of P in the detergents. In rural areas,
however, the detergent P, in addition to other P-
sources, may still contribute to the over-fertilization.
In Norway, the measures taken to reduce the
'detergent P-consumption are difficult to evaluate, as
other P-sources were reduced parallel to the decreas-
ing sale of P-phosphoric detergents (Minist. Environ.
1979).
The NTA-containing detergent is still used in Sweden
and Finland. As no large-scale conversion over to NTA
has been attempted, this type of detergent has been
used to a limited extent. In any case, no harmful effect
of NTA has been reported. According to detergent
experts, it is normally not possible to replace all sodium
tripolyphosphate by NTA. The reason is that sodium
tripolyphosphate has two different properties: (a)
softens the water; (b) prevents precipitation —
especially of calcium carbonate — during a normal
washing and rinsing cycle. The effect of (a) can be
achieved with, for example, NTA, citric acid, or other
carboxylic acids.
The effect described in (b) is difficult to achieve with
other substances, but requires only 5 to 10 percent
sodium tripolyphosphate even in hard water.
If a phosphate-free detergent is required, all the
components must be carefully chosen to prevent
precipitation of Ca salts in hard water. If precipitation
occurs, the visible washing result will be bad. This
means severe restrictions on using different com-
ponents and phosphate-free detergents have had no
success.
Because of the very high chelating power of NTA the
residual concentration of Ca ions is very low in the
washing liquor when a NTA-based detergent is used.
This gives a better washing power, especially on
pigment dirt, than a normal phosphate-based deter-
gent.
Totally P-free detergents have also been used. Some
washing problems in hard waters were reported when
these products were based on citrate (15 percent).
Unfavorable costs for the substitutes compared with
polyphosphates, increasing consumption of standard
"low price" products, and the development of effective
methods for P-removal in sewage, at present provide
no base for successful marketing of modified deter-
gents.
Ministry of Environment. 1979. The Mjosa campaign. Oslo.
(With English introduction and summary.)
Monsanto and Procter & Gamble Companies. 1977. Ecolo-
gical effects of non-phosphate detergent builders. Cincin-
nati!, Ohio.
National Swedish Environment Protection Board. 1972.
Household detergents and water protection. Typewritten
report.
Royal Swedish Academy of Sciences. 1970. Environmental
aspects on boron. Rep. 33. (In Swedish.)
Ryding, S. p. 1980. Monitoring of inland waters. OECD
eutrophication programme. The Nordic project. Nordic co-
operative organisation for applied research. Secretar.
Environ. Sci. Publ. 1980:1. Helsinki.
Swisher, R. D., M. R. Crutchfield, and D. W. Caldwell. 1967.
Biodegradation of NTA in activated sludge. Environ. Sci.
Technol. 1:820.
REFERENCES
Forsberg, C. and G. Lindqvist. 1967a. On biological
degradation of nitrilotriacetate (NTA). Life Sci. 6:1961.
1967b. Experimental studies on bacterial degra-
dation of nitrilotriacetate, NTA. Vatten 23:264.
Forsberg, C., D. Jinnerot, and L. Davidsson. 1967. The
influence of synthetic detergents on the growth of algae.
Vatten 23:2.
Holtan, H. 1979. The Lake Mjosa story. Arch. Hydrobiol. Beih.
Ergebn. Limnol. 13:242.
Holtan, H., et al. 1980. Evaluation of the pollution situation
and effects of possible water level regulations in Jotun-
heimen. Rep. from NIVA. 0-7909. (In Norwegian.)
-------
432
THE LONG RANGE TRANSPORT OF
AIR POLLUTION AND ACID RAIN FORMATION
BRYNJULF OTTAR
Norwegian Institute for Air Research
Lillestrom, Norway
ABSTRACT
The increasing acidification of the precipitation in Europe was first pointed out in 1968 by Oden,
who related this to the acidification observed m rivers and lakes in Scandinavia and the increasing
use of fossil fuels with a high content of sulfur. In the OECD project "Long range transport of air
pollutants" (1972/77), the acidification of the precipitation was quantitatively related to the
emission and transformation of sulfur dioxide to sulfuric acid in the atmosphere. It was shown that
extensive exchange of air pollutants took place between the European countries, and in orographic
precipitation areas frequently exposed to polluted air masses, excessive amounts of acid
precipitation were observed. Later studies have shown that the air pollutants from Europe also find
their way into the Arctic region, particularly in the winter. The main acid component of the
precipitation is sulfuric acid with an addition of 20 to 50 percent of nitrate and ammonium ions on
an equivalent basis. The sulfate content is largely explained by the sulfate in the aerosol phase.
The content of nitrate and ammonium ions is explained by the uptake of gaseous nitric acid and
ammonia from the atmosphere. Atmospheric dispersion is discussed in relation to the methods
used to describe the chemical transformations and the dry and wet deposition processes.
INTRODUCTION
The increasing acidification of the precipitation in
Europe was first pointed out in 1968 by Oden. Data
from the European Precipitation Chemistry Network
coordinated by the Institute of Meteorology at the
University of Stockholm, showed that a central area in
Europe with highly acid precipitation (pH 3 to 4) had
expanded to include also the southern part of
Scandinavia.This observation was associated with
observed acidification of the water in rivers and lakes in
Scandinavia, where in many places the fish population
had disappeared.
In addition, incidents of greyish snow were observed
in areas remote from pollution sources. Chemical
analyses of the polluted snow showed a high content of
sulfuric acid, soot, fly ash, and other pollutants. These
observations caused much alarm in Scandinavia, and
in 1969 the matter was brought to the attention of the
Organization of Economic Cooperation and Develop-
ment. In the subsequent OECD project "Long range
transport of air pollutants" (1972-1977) the acidifica-
tion of the precipitation was quantitatively related to
the emissions of sulfur dioxide in Europe (OECD, 1978;
Ottar, 1978a). It was shown that extensive exchange of
air pollutants took place among the European countries
so that national control programs may achieve only
limited improvements with respect to the total
deposition of sulfur within the national borders.
Subsequent studies have shown that the air
pollutants from Europe also find their way into the
Arctic regions, particularly in the winter.
As a result of the OECD project, a European
monitoring and evaluation program for the long-range
transfer of air pollutants (EMEP) was established under
the auspices of the U.N. Economic Commission for
Europe and in cooperation with the U.N. Environmental
Program and the World Meteorological Organization.
Its main objective is to "provide governments with
information on the deposition of air pollutants, as well
as on the quantity and significance of long range
transmission of pollutants and transboundary fluxes."
At present, about 20 countries from both eastern and
western Europe participate in the program; its design
broadly follows that of the OECD program.
In North America the long-range transport of air
pollutants is examined in several regional programs.
While the European and Canadian studies centered on
ecological problems resulting from the acidification of
the precipitation, the U.S. emphasis was initially on air
pollutant concentrations, health effects, and visibility.
In later years this has changed, and the U.S. studies
today also deal with acid precipitation problems. Most
of the North American programs are described in the
proceedings of the Dubrovnik symposium (Ottar,
1978b).
ACIDIFICATION OF THE PRECIPITATION
The general plan of the OECD project was simple.
Single layer atmospheric dispersion models, wind
trajectories, and an emission survey for sulfur dioxide
were used to calculate the concentration fields of
sulfur dioxide and sulfate on particles. The dry
deposition was assumed to be proportional to the air
concentration, and the annual deposition of sulfate by
precipitation was empirically found to be proportional
to the product sum of amount of precipitation and
sulfate aerosol concentration. Parameters for the
chemical transformation of sulfur dioxide to sulfate and
-------
THE ACID RAIN PROBLEM: MECHANISM AND EFFECTS
433
deposition rates were adjusted by fitting the model to
daily measurements from more than 70 ground
stations in the region. Aircraft sampling was used to
obtain information on the vertical distribution of sulfur.
The calculation was carried out in a grid system (a
side length of 127 km) covering the northwestern part
of Europe, and complete mixing was assumed up to a
height of 1,000 m. Trajectories were calculated each 6
hours from wind fields obtained from the WMO
Weather Service. With some improvements the same
general approach is used in EMEP.
In Europe the spatial distribution of sulfur dioxide
emissions follows the population density and location
of major industries (see Figure 1) (Semb, 1979). The
maximum concentration of sulfur dioxide is found near
the major emissions. In the central part the annual
mean concentration of sulfur dioxide is about 20
wg/m3. Because Europe is situated in the westerlies,
the maximum values are found slightly northeast of the
emissions (see Figure 2) (Eliassen, 1978). The annual
concentration pattern of sulfate particles is similar, but
because of the time required for sulfur dioxide to be
transformed into sulfate particles, the maximum
concentration level is lower, about 10 /ug/m3. The dry
deposition of sulfur dioxide is a significant factor in the
central part of the area and responsible for removing
about 50 percent of the total emission. Compared to
this, the dry deposition of sulfate is of less significance.
As shown in Figure 3, the annual deposition of sulfate
by precipitation is strongly influenced by the amount of
precipitation. Maximum deposition is found in oro-
graphic precipitation areas frequently exposed to
polluted air masses. Examples are the Scandinavian
mountains, the Alps, and mountains in Scotland. About
30 percent of the total emission is removed by
precipitation.
o a
i \i
'i
\
\
tl
p
i
to
a
1,
4
1-
4
t
rr
,•
*
171
A
n
to
»,
0-
•0
~^
1
—
/
A-
10
*"K
M
10
1
^
1
1
t
1
~
1
#
110
no
1H
•
10
'it
I
I
1
1
I
t
1
.9,
MB
HO
»0
100
i.
1
*
n
i
1
(
J|
»
T
10
.*
*
1
II*
lit
80,
»
-T
T
•r
H.
n
'
-t
\
i
-?
,i>
\,
•[
i1
i
n
m
•M
,,'
10
'f-,
7*
4
CM
UO
m
»
wo
K
'».
10
•
.0
)'
n
»
i
T
t.
i
1
"
•0
»
IS
>i*
IV
n
to
na
t»
1W>
*
10
it
,*'
T
«0'
J
*
»
A
1
1
»
»
1
^
«
10
•
1
(0
0
V
ii.
• 0
1-
no
10
10
te
V
f1
1
1
10
!0
10
l»
11
110
'1
111
.»
10
40
10
.11
70
*»
ItO,
no
too
100
1)
1°
100
DO
X"
1
10
1
1
1
1
»1
ID
10
10
.0
'*!
>>
104
tbo
IT*
to
7)
^
t
It
to
1
11
*,
f '
*
10
1*
It
1*
7*
'10
»
TO
tl
70'
»
to
4
1**
11
Y
'*)
i
,'•
i
i
i
'
i
0
1
»
ro
10
i*
to
'»
»
«0
IV
H
1»
•
10
10.
10
^
1
I
)
1
1
'
0
0
1
11
14
«
40
)0
11
II
II
«
T4
It
11.
1
10
00
yf
10
u
M
10
10
40
10
11
10
tl
DO-
MI
*
10
->o
30
»
1
10
lk
>
1
'
'
0
0
10 1
11 100
100 11
100 11
10 11
M 1
40 lo
% *¥•
10 H
loo to
£. I0
-JO 1W
40 10-
1 .10
10 .
10 "•-
p '
iL
.' *!
11 J
' V
1 1s
1 1
1 >
0 0
0
Figure 2. — Estimated mean concentration field for sulfur
dioxide for 1974. Observed mean concentrations given by
italic numbers Unit fjg SO2/m3.
Figure 1. — Estimated annual emission of SO2 (10J tonnes S)
in grid elements with length 127 km at 60O N.
Figure 3. — Estimated sulfur wet deposition pattern for 1974.
Unit: g S/mft
The day to day situation is very different from this
average picture. With southerly winds, concentrations
of 20 to 30 A/g/m3 of sulfur dioxide and sulfate particles
are frequently observed in the Scandinavian area in
places where such concentrations cannot be explained
-------
434
RESTORATION OF LAKES AND INLAND WATERS
by local sources. About 50 percent of the total annual
deposition of sulfate may result from about 10 episodes
with highly acid precipitation. Aircraft measurements
have shown that these polluted air masses lose little of
their pollution content by passing over the North Sea, a
distance of about 800 km. A similar situation is
observed in other remote areas exposed to orographic
precipitation. In 1978 an exceptional case of 10 mm
precipitation with a pH of 2.5 was observed in Iceland.
Precipitation with pH down to 2.4 is known from both
Scotland and the west coast of Norway.
Recent studies have shown that in winter consider-
able amounts of air pollutants find their way from
Europe and the Soviet Union into the Arctic (Larssen
and Hanssen, 1979; Rahn and McCaffrey, 1980).
Concentrations as high as 6 and 4 wg/m3 of SO2 and
sulfate have been measured at Bear Island and Ny
Alesund on Spitsbergen. These pollutants have been
traced all the way across the Polar Basin to Barrow in
Alaska. There is very little precipitation in this region
during the winter, and evidently the chemical
transformation rate of sulfur dioxide is much reduced.
The main acid component of the precipitation is
sulf uric acid with an addition of 20 to 50 percent nitrate
and ammonium ions on an equivalent basis. The
sulfate content of the precipitation is largely explained
by nucleation on ammonium sulfate and ammonium
hydrogen sulfate from the aerosol phase. The content
of nitrate is probably explained by the absorbtion of
nitrogen dioxide and gaseous nitric acid from the
atmosphere.
In Scandinavia the concentration of sulfate in
precipitation is generally highest during the spring,
while the emissions of sulfur dioxide in Europe reach a
maximum in January (about twice the emissions in
July-August). This delayed maximum sulfate con-
centration in precipitation can be attributed to a
precipitation minimum in western Europe during the
early spring, and more rapid conversion of sulfur
dioxide to sulfate with increased solar radiation
(Joranger, Schaug, and Semb, 1980). The seasonal
variation of the concentration of nitrate in precipitation
is similar but with a longer maximum period. For
further elucidation of these differences, comparison
should be made between air and precipitation
concentrations of nitrogen compounds as well as for
the sulfur compounds.
MODELING OF THE LONG RANGE
TRANSPORT
Our knowledge of the details of long-range transport
of air pollutants and the acidification of the precipita-
tion is limited by the methods used. The following
discusses the significance of some of these limitations.
Emissions
The sulfur dioxide emissions in Europe are due
mainly to the burning of sulfur-containing coal and oil.
The increased demands for energy after 1950 were
met by widespread introduction of petroleum products,
and as a result sulfur dioxide emissions in Europe
doubled from 1950 to 1970 (see Figure 4) (Semb,
1978).
1500-
1000-
500-
20-
15-
10-
Figure 4. — Fossil fuel consumption and estimated sulfur
dioxide emissions in Europe.
When this emissions increase is considered in
relation to the transport of the air pollutants to remote
areas, it may well be that polluted precipitation has
occurred for a long time without being noticed. Thus,
fish kills in rivers in southern Norway reported at the
beginning of this century may well have resulted from
long-range transport of sulfur pollutants. The decline in
fish populations has been much more dramatic in the
last 30 years, however.
The emission survey for the OECD study was
established in coooperation with the participating
countries. For other countries this survey was based on
national fuel consumption data collected from OECD
and ECE, emission factors, and population density. For
some countries the accuracy is at least within 10 to 15
percent. This survey is being further elaborated in
connection with EMEP.
The size of the grid element limits the geographic
resolution, and the atmospheric dispersion models can
give only a smoothed picture of the concentration
fields. Clearly, measurements used to verify the model
calculations should represent comparatively large
areas. Furthermore, the acidity of the precipitation also
depends on other chemical components present,
particularly the nitrate and ammonium ions. It is
therefore also necessary to know the emissions of
nitrogen oxides and ammonia. Detailed emission
-------
THE ACID RAIN PROBLEM: MECHANISM AND EFFECTS
435
surveys for nitrogen oxides have been constructed for
some European countries (Semb, 1979). Recently,
Bonis, Meszaros, and Pusey (1980) estimated the
nitrogen oxide and ammonia emissions for all Europe.
It may be asked how relatively inaccurate emission
surveys can yield useful information. The answer is
that, however uncertain, emission surveys are an
indispensable tool in understanding the occurrence
and dispersion of air pollutants. The accuracy in
general should be ±20 percent or better, and the
positions and relative emission strength of major
emission areas are reasonably well defined. Although
better data would be welcome, the accuracy of the
present survey is sufficient for dispersion model
calculations.
Transport
Aircraft measurements show that the air pollutants
usually remain below a mixing height of 1 to 2 km, and
100 to 200 km downwind of a source area; there is no
further rapid dilution of the pollutants. Beyond this
distance, which depends on the weather conditions,
the pollutants are slowly removed by dry deposition.
The only process which can rapidly clean the air is
precipitation.
When air pollutants have reached this state of
dilution, the transport of the polluted air masses is
conveniently described by wind trajectories in a grid
system. In the grid models calculations are based on
average values for each grid element with respect to
emissions, wind, rainfall, etc. The geographic resolu-
tion of these models is limited by the size of the grid
element. There is also a relation between the
geographical and the time resolution which can be
obtained. For instance, the effect of nocturnal ground
inversions cannot be described in a simple one-layer
dispersion model. Therefore, 24 hourly measurements
will fit the model better than 6 hourly measurements.
To include such variations one has to use a smaller grid
element, a two-level model, or a perturbation of the
vertical concentration profile within the grid element.
The main problem is the effort required to provide
measurements to verify the results of more detailed
calculations.
Two different types of models were used in the OECD
project (Eliassen, 1978). In the back trajectory model,
the uptake and deposition of air pollution is calculated
for an air parcel following the trajectory up to the point
of interest. In the OECD program the concentration for
each grid element was calculated from 48-hour back
trajectories, and compared with measured daily mean
concentrations. In the EMEP, 96-hour back trajectories
are used to reduce the amount of pollution of unknown
origin.
In this model the contributions to one grid element
from all other elements are easily separated, and the
model is regularly used to calculate the exchange of
pollution between the European countries. In the
Lagrangian model of the OECD project, forward
trajectories were used to calculate the concentration
field with regular time intervals. This model has an
unlimited memory and can be used to predict episodes
of air pollution using weather forecast data.
In both models the air parcel is assumed to follow a
calculated trajectory. However, this trajectory does not
represent a physical reality, as the lateral and vertical
dispersions are neglected. The small scale turbulence
is not significant, but the meso-scale wind variations
cannot be neglected. These are simulated in an indirect
way in the two models mentioned by the fact that the
concentration values represent averages for large grid
elements. This introduces a so-called psuedo-diffusion,
the magnitude of which is determined by the size of the
grid element, the time step used in the calculation, and
the numerical advection procedure.
Husar and Patterson (1979) have recently developed
a different model based on individual handling of a
stream of air pollution parcels from each emission
source. The source strength is given by the number of
parcels and not by the concentration of each parcel. To
account for the meso-scale dispersion, they have
introduced a random displacement of the air parcels
when they have passed along the calculated trajectory
for a specified time interval. Probability distribution
functions are used to account for chemical transforma-
tions and deposition probability.
A main advantage of this model is that the lateral
(and if necessary the vertical) dispersion is separated
from the choice of grid size. For models on a global
scale, this may be an essential feature. A serious
limitation of this model in its present state is the
requirement of linear chemical interactions. For sulfur
dioxide and the formation of sulfates this causes no
problem, but in the case of nitrogen oxides and nitrates
chemical reactions are far from linear; this raises the
important question of using simplified procedures.
Modeling the long range transport of air pollutants
involves a number of approximations, some of which
have been mentioned. Because of this, the day to day
agreement between observed and calculated concen-
trations is reduced. For mean values over extended
periods of time better agreement is usually obtained.
Principally, the same applies to mean values for larger
areas, but most of the measurements represent point
values, normally at ground level. In principle, mean
values for larger areas, perhaps observed from
satellites or aircraft, should give better agreement.
Chemical transformation and deposition rates
In calculating the long-range transport of air
pollutants constant transformation and deposition
rates are generally used, and the wet deposition is
often estimated from annual precipitation data. In the
OECD project wind fields at different levels and a
number of advection schemes were tried, but a
sensitivity analysis showed that it would be more
important to improve the modeling of the chemical
transformation and deposition. As a first step, daily
precipitation fields are estimated and used to calculate
the wet deposition in EMEP. In this case the necessary
data are available from the WMO Weather Service.
Available information on the dry deposition rate of
gases and aerosols (Garland, 1978) is generally limited
to results of special laboratory and field investigations.
Although there is considerable evidence of variations
in the dry deposition velocities for different surfaces.
-------
436
RESTORATION OF LAKES AND INLAND WATERS
seasons, and weather conditions, constant deposition
rates are generally used for all seasons and surface
areas in the long-range transport models.
This is not a satisfactory approach, particularly when
transport over very long distances is considered. In the
summer season the sea is colder than the air, and a
shallow, stable layer of air often forms over the sea
surface, reducing vertical mixing of the air. In winter
the North Sea and the Atlantic are generally warmer
than the air, leading to increased vertical mixing and
precipitation, while the continental land masses and
frozen water bodies are colder than the air, forming
stable stratification near the surface.
As a first approximation one might correct for these
effects by introducing different deposition rates for
summer and winter and for land and sea areas.
However, to justify this additional information is
required. A simple calibration of the model is highly
unsatisfactory.
Similar conditions apply to the chemical transforma-
tion rates. Studies in the Arctic region and statistical
analyses of data from the OECD project (Prahm, et al.
1979) strongly indicate that the transformation rate of
sulfur dioxide to sulfate decreases with concentration
and depends on temperature, sunlight, and the
presence of other pollutants. Again, more measure-
ments are needed to specify these conditions in the
dispersion models.
The oxidation of sulfur dioxide to sulfate follows two
main pathways. When sulfur dioxide is absorbed and
oxidized catalytically in cloud droplets, the absorbtion
stops if the droplets become too acid. Ammonia will
neutralize the acidity and thus make further absorbtion
possible. Over the sea, little or no ammonia is available,
and the reactions stop. Photochemical oxidation in the
gaseous phase then becomes relatively more impor-
tant, and this leads to the direct formation of sulfuric
acid droplets. Under these circumstances pH-values
down to 2.5 have been observed in coastal precipita-
tion.
The catalytic oxidation of sulfur dioxide is much more
rapid in plumes from coal combustion than from oil,
because of the manganese content in submicron fly
ash particles from coal. Thus, the transformation of
sulfur dioxide to sulfate may go faster when polluted air
from the European continent passes over the Scandi-
navian area. The gas phase oxidation of sulfur dioxide
is intimately related to the photochemical reactions of
the nitrogen oxides and the production of hydroxyl ions.
A normal rain shower in Scandinavia precipitates
approximately 1 ml of water from each m3 of air at the
level where the precipitation is formed. Comparisons of
the sulfate content in precipitation with the aerosol
sulfate concentration at ground level show that the
amount of sulfate in precipitation corresponds to
complete scavenging of the sulfate particles at the level
of rain formation with only a minor addition of sulfate
from the absorption of sulfur dioxide.
On the other hand, the experience that 20 to 50
percent of the acidity in precipitation may be from nitric
acid, while simultaneous measurements of particles
show little or no nitrate ions, is evidence that most of
the nitric acid in precipitation does not come from the
particles, but probably from gaseous nitric acid. More
measurements of gaseous nitric acid, ammonia, and
the composition of cloud droplets and aerosols are
needed to clarify the significance of these processes.
The sulfate particles responsible for acidifying the
precipitation are found in the accumulation phase of
the bimodal aerosol size distribution (0.1 to 2.5 /urn)
The sea salt particles are mainly found in the larger
fraction (above 2.5 //m). Samples collected at coastal
stations therefore are corrected for their content of sea
salt sulfate by analyzing for sodium, chloride, or
magnesium.
The small particles in the accumulation mode are
important in the long-range transport of air pollutants,
and their chemical composition is markedly different
from the larger particles. Size-segregated sampling
would prevent chemical reactions between the small
and the large particles on the filter, which for instance
mav result in a loss of hydrochloric or nitric acid, and
thus assist in interpreting the results.
CONCLUSIONS
The studies of the long-range transport of the
atmospheric sulfur pollutants that began in the 1970's,
have shown that the air pollutants are more widely
distributed than previously believed. The components
that are transported over long distances as gases and
as particles in the accumulation mode, include most of
the pollutants and their secondary products.
The acidity of the precipitation is governed mainly by
its content of sulfate, nitrate, and ammonium ions, and
may to a large extent depend on the pathway of the
polluted air masses. The lowest pH values are obtained
in air masses which have remained over the sea for a
longer period of time.
The modeling of the long-range transport and the
formation of acid precipitation includes many simplifi-
cations. For larger areas and longer periods of time the
agreement between observed and calculated values is
reasonably good. To improve the day to day agreement,
the chemical transformations taking place in the
atmosphere and the deposition processes have to be
described in more detail.
The photochemical oxidation of sulfur dioxide is
intimately connected with the photochemical reactions
of the nitrogen oxides; however, introduction of non-
linear chemical reactions in the atmospheric chemistry
of the dispersion models will seriously complicate the
models. Adequate simplified procedures must be
developed.
REFERENCES
Boms, K., E. Meszaros, and M. Pusay. 1980. On the
atmospheric budget of nitrogen compounds over Europe.
Period. Hungarian Meteorol. Serv. 84:57.
Eliassen, A. 1978. The OECD study of long range transport of
air pollutants: Long-range transport modeling. Atmos.
Environ. 12:479.
Garland, J. A. 1978. Dry and wet removal of sulphur from the
atmosphere. Atoms. Environ. 12:349.
Husar, R. B., and D. E. Patterson. 1979. Synoptic-scale
distribution of manmade aerosols. Proc. WMO Symp. on
Long-Range Transport of Pollutants. WMO No. 538,
Supplement, Geneva.
-------
THE ACID RAIN PROBLEM: MECHANISM AND EFFECTS 437
Joranger, E., J. Schaug, and A. Semb. 1980. Deposition of air
pollutants in Norway. In Proc. Int. Conf. Ecologcial Impact of
Acid Precipitation, Sandefjord, March 11-14, 1980. SNSF-
prosjektet, Oslo-As, Norway.
Larssen, S., and J. E. Hanssen. 1979. Annual variation and
origin of aerosol components in the Norwegian Arctic-
Subarctic region. WMO Conf., Boulder, Colo., 1979.
Oden, S. 1968. Nederbordens och luftens fororening den
orsaker, forlopp och verkan i olika miljor. Statens
Naturvetenskapliga Forskningsrad, Ekologikomiteen, Stock-
holm, Bull. No. 1.
Organization for Economic Co-operation and Development.
1978. The OECD Programme on Long-Range Transport of
Air Pollutants. Measurement and findings. Paris.
Ottar, B. 1978a. An assessment of the OECD study on long
range transport of air pollutants. Atmos. Environ. 12:445.
1978b. Sulphur in the atmosphere. In R. B. Husar,
J. P. Lodge Jr., and D. J. Moore, eds. Proc. Int. Symp.
Dubrovnik, Yugoslavia 7-14 Sept. 1977. Pergamon Press.
Prahm, L. P., et al. 1979. Regional source quantification
model for sulphur oxides in Europe. In Proc. WMO Symp. on
the Long-Range Transport of Pollutants. Sofia, 1979. (WMO
- No. 538).
Rahn, K. A., and R. J. McCaffrey. 1980. On the origin and
transport of the winter Arctic aerosol. Ann. N.Y. Acad. Sci.
338:486.
Semb, A. 1978. Sulphur emissions in Europe. Atmos.
Environ. 12:455.
1979. Emission of gaseous and paniculate matter
in relation to long-range transport of air pollutants. WMO
Symp. on the long-range transport of pollutants, Sofia, 1 -5
October 1979. WMO-No. 538, Geneva.
-------
438
EFFECTS OF ACID PRECIPITATION ON AQUATIC
AND TERRESTRIAL ECOSYSTEMS
ARNE TOLLAN
SNSF Project (Acid precipitation: Effects on forest and fish)
Agricultural University of Norway
As-NLH, Norway
ABSTRACT
Acid precipitation, characterized by high concentrations of H+, SO< and NOa, occurs over large
regions, notably in Europe and eastern North America. In areas susceptible to acidification, i.e.,
areas with sparse soil cover and bedrock geology poor in neutralizing minerals, acid precipitation
acidifies waters. The normal bicarbonate buffering system breaks down, and sulfate becomes the
dominant anion in acidified water sources. Increased inflow of aluminum from the soil to the lakes
is particularly important to aquatic life. Aquatic ecosystems in acidified areas often show a
simplified structure, where a few tolerant species dominate. Changes are seen on all trophic
levels. Populations of valuable fish species, especially salmonids, are reduced or wiped out in
many acidified districts. There are signs that nutrients like calcium and magnesium have been
reduced in some soils exposed to acid precipitation. Continued leaching may eventually have
negative effects on forest growth. At present, field evidence of reduced growth is inconclusive, and
experimental research has in some cases shown growth increase under acid conditions. This is
interpreted as a fertilizing effect of the nitrogen content in acid precipitation.
EXTENT OF FRESHWATER
ACIDIFICATION
Acidification of lakes and rivers during recent
decades is a regional problem in Scandinavia and
eastern North America. The acidified areas are
underlain mainly by siliceous (quartz-rich) bedrock with
sparse or thin soil cover. These same areas now
receive decidedly acidic precipitation (weighted aver-
age below pH 4.6), and the time trends in acidification
of precipitation and inland waters are parallel. Recent
acidification of freshwaters is normally not found in
geologically similar, sensitive areas which lie outside
the regions of acid precipitation (e.g., Likens, et al.
1979; Wright, et al. 1980).
This regional coincidence both in space and time
strongly suggests that aquatic ecosystems are being
acidified by atmospheric deposits. The extent of
acidification is known from a few existing observations
of water pH and other chemical characteristics over the
years, and indirectly through mapping of lakes and
rivers where fish populations have been reduced or lost
in recent years. Nothing but acidification with its
associated altered chemical conditions can explain the
present regional fish loss. Surveys of land use changes
associated with agriculture and forestry practices in
acidified parts of Norway show no systematic relations
with acidified lakes and fish population loss(Drablos, et
al. 1980).
Observations from 1920 to 1970 of pH in 128 lakes
in southern Norway have been compared to pH data
from the same lakes during the 1970's. Of these lakes,
63 percent had become at least 0.25 pH units more
acid and 12 percent had become at least 0.25 pH units
less acid. Only 4 percent of the lakes had pH below 5.0
prior to 1950, compared to 25 percent in 1977. Before
1950 none of 130 lakes in southern Sweden was
below pH 5.5. In 1977 28 percent were below pH 5.5
and 15 percent below 5.0. All the lakes that had
become more acidic are situated in areas of southern
Scandinavia that today receive acid precipitation with a
pH below 4.6 (Wright, 1977).
Similar observations have been made in North
America. Of 320 high elevation lakes in the Adirondack
Mountains of New York, about 70 percent had a pH
above 6.5 and only 4 percent below 5.0 in the 1929-
1937 period. In 1975, 51 percent of a group of 217 high
elevation lakes had pH below 5.0 and 90 percent of
these lakes were devoid of fish (Schofield, 1976).
SOIL PROCESSES AND WATER
ACIDIFICATION
Several processes are known to acidify soil:
1. Root uptake of cations during plant growth.
2. Carbonic acid formation from COa derived from
respiration of soil fauna and flora.
3. Oxidation of nitrogen and sulfur compounds to
nitric acid and sulfuric acid.
4. Organic acids produced during decomposition of
plant matter.
5. Atmospheric input of acidifying substances, nota-
bly sulfuric acid.
Close to emission sources the acidity produced as a
consequence of dry deposition of 0^2 may dominate
over acidity produced by precipitation.
When soils acidify, the most important effects
probably are increased mobility and leaching losses of
basic metal cations such as Ca+2, MgT', (C, and Al^ .
-------
THE ACID RAIN PROBLEM: MECHANISM AND EFFECTS
439
The ion exchange processes in soil are very important
for soil and water acidification. Soil particle surfaces
are normally negatively charged, and therefore sur-
rounded by cations.
The cations, including hydrogen ions, can be
exchanged between the soil particles and the soil water
solution percolating in the soil pores. What cations are
exchanged depends on their charge and other
properties, and on the relative amounts in solution and
adsorbed to the soil. At high concentrations, hydrogen
ions in the percolating water will tend to exchange with
calcium, magnesium, and aluminum ions. This results
in higher concentrations of Ca, Mg, and Al in the soil
water, and lower H* concentration, which means a
neutralizing effect on the soil water which eventually
enters lakes and streams. Net adsorption of H also
leads to a more acid soil, unless the H* ions are
consumed in weathering processes.
Acidification of soil is a slow process in nature, and
field detection of effects of additional inputs of
acidifying components is likely to be difficult. There is
agreement that sandy, well-drained soils of inter-
mediate pH are particularly susceptible to pH changes.
There are, however, very few indications from field
studies of soil acidification caused by atmospheric
deposition (Troedsson, 1980; Linzon and Temple,
1980).
As the amount of basic cations in a particular soil
profile is reduced, a smaller proportion of H* will be
adsorbed, and a greater proportion of inflowing
hydrogen made available for transport to the water-
courses.
The transport of cations from the soil to the water
systems depends on available anions to maintain
electrical charge neutrality. Sulfate ions are important
as vehicles for cation transport as in many soils they
are very mobile, and will be adsorbed only temporarily
(Cronan, et al. 1978; Johnson, 1980).
Sulfate therefore plays a decisive role in the
acidification of freshwater, as a mobile carrier of the
hydrogen ions whether the hydrogen ions stem from
atmospheric deposits or are produced in the catch-
ment. The input and output of sulfate to catchments are
in many cases close to balance over periods of several
years. There is, however, often a retention in the winter
snowpack and during dry summers, and releases
during spring snowmelt and autumn rains (Likens, et
al. 1977, 1980; Seip, 1980; Figure 1). These processes
may produce episodes of very acid stream water, as the
sulfate is washed out with equivalent amounts of
cations, which in the acidified regions will tend to be
hydrogen ions. To explain water acidification, it is
therefore probably more important to consider the
possibility for leaching of H+and other cations provided
by sulfate, than the total amount of H+ in the catchment
(Seip, 1980).
The relationship between hydrogen and sulfate ions
is demonstrated by data from regional surveys of
Norwegian headwater lakes in 1975-1978. Lakewater
chemistry was significantly correlated to precipitation
chemistry. Sixty to 80 percent of the variance in
lakewater content of hTand SO4 could be explained by
precipitation amount and content of (-T and excess
sulfate (Mohn, et al. 1980).
15,000
•„ 10,000-
5.000 •
1964 - 1974 SO/
JJ
Figure 1. — Monthly flux of sulfate for undisturbed
ecosystems of the Hubbard Brook Experimental Forest,
N.H., showing input (solid line) dominance during summer,
and output (dashed line) dominance during autumn and
spring. (Likens, et al. 1977.)
LAKE ACIDIFICATION
The acidification process of lakes exposed to acid
water inflow can be described as a large-scale titration
(Henriksen, 1979, 1980). Weathering of rock material
in the catchment provides bicarbonate,HCOa, which
normally is the major anion in soft-water lakes, with
calcium, Ca, and magnesium, Mg, as the major cations.
Lakes with high bicarbonate levels are well buffered
(i.e., they resist changes in pH levels) and have pH
above 5.5. Fish populations are usually normal. High
influx of strong acids, notably sulfuric, from the
atmosphere may deplete the bicarbonate buffer and
cause severe pH fluctuations resulting in physiological
stress, reproductive failure, and episodic kills of fish.
If the influx of acids is high enough to completely
exhaust the bicarbonate buffer, the lake will enter the
acidified stage characterized by pH well below 5.0,
sulfate instead of bicarbonate as the dominant anion,
and high concentrations of aluminum, Al. Fish stocks
are severely reduced or lost.
Calcium, which normally accompanies bicarbonate,
is a useful indicator of the geological influence from the
catchment upon water chemistry (Figure 2). Calcareous
rocks and soils are easily soluble and will produce lake
waters with high concentrations both of calcium
bicarbonate and other compounds which provide
buffering capacity.
Thus waters in areas with calcareous bedrock and
soils, such as much of central Europe, will not be
acidified in spite of the fact that the acidity of
precipitation is very high. However, when a major
emission source happens to be located close to
geologically susceptible areas, such as the metal
smelters at Sudbury (Ontario), Canada, which have
annual emissions near 1.35 million tons SOa, the
chemical and ecological effects on the environment
can be devastating.
-------
THE ACID RAIN PROBLEM: MECHANISM AND EFFECTS
440
Particularly high concentrations of hydrogen ions
and other substances are commonly observed during
early spring snowmelt. Both laboratory experiments
and field observations have shown that concentrations
can be 3 to 10 times higher in the first meltwater than
in the bulk snowpack (Johannessen and Henriksen,
1979). Although there is considerable contact between
meltwater and soil (Seip, 1980) modifying the chemical
properties, the snowmelt period often produces major
impacts on aquatic chemistry and biota.
Figure 2. — pH and calcium concentrations in Norwegian
lakes 1974-1977. Lakes in southeastern Norway (•)
receive highly acid precipitation, pH 4.2 — 4.5.
When calcium concentrations are assumed to be well
correlated with pre-acidification bicarbonate alkalinity, the
empirically drawn curve will distinguish between acidified
and nonacidified waters. (Henriksen, 1979.)
EFFECTS OF ACID WATER ON
AQUATIC LIFE
The recent acidification of freshwater in parts of
Europe and eastern North America has had profound
impacts on aquatic life. All trophic levels have been
affected. The most immediate concern to the people
living in the acidified regions is the major decline in fish
populations, but primary producers, decomposers, and
invertebrate animals also are affected. (Aimer, et al.
1978).
Reduced numbers of several algal species have been
observed in acid lakes, especially among green algae.
On the other hand, there is often a conspicuous heavy
growth of filamentous algae and mosses in many acid
lakes and streams. The algal accumulation is probably
caused by reduced activity of invertebrates feeding on
the vegetation, and reduced decomposition. The
dominance of a few plant plankton species in acid
water probably results from specific tolerance or
changed biological interactions. Many of the algae are,
however, photosynthetically inactive, and thus the
productivity per unit of biomass may be lower in acid
waters. A possible factor reducing productivity in lakes
of pH 5 to 6 is precipitation of phosphorus by aluminum
released to the lakes from the surrounding catchment
(Aimer, et al. 1978).
Expansion of sphagnum moss on bottoms of acidified
lakes is known from Sweden (Grahn, et al. 1974);
sphagnum mats are also reported from south Norway
(Hendrey, et al. 1976) and the acidic Lake Golden in the
Adirondack Mountains of New York (Hendrey and
Vertucci, 1980).
The silica-containing algae, known as diatoms, show
changes in community composition, shifting to more
acid-tolerant species in rivers and lakes under
acidification.
Diatom remains in sediments in south Norwegian
lakes indicate that lake water pH has declined 0.5 pH
units or more since about 1930 to 1945 (Davis and
Berge, 1980).
Among decomposing organisms in acidified lakes
there is a shift from bacteria to slow-acting fungi,
leading to increased accumulation of organic matter
and reduced availability of nutrients. This is observed
both in North America and Scandinavia.
The invertebrate fauna is an important link between
primary producers and fish in the aquatic food chain.
Both zooplankton, aquatic insects, non-planktonic
crustaceans, snails, and mussels are reduced in
abundance and diversity during water acidification. A
few examples from Norway may illustrate:
Norwegian studies of mayflies indicate that the mean
number of species is about three to four times higher in
water with pH6.5 to 7.0 than at4.0 to4.5(Leivestad, et
al. 1976).
The mayfly Baetis rhodani is usually a key organism
in the food chain in oligotrophic rivers, transferring
energy from plants to the higher stages. This species
comprises 60 to 80 percent of the mayflies or even
more in parts of Norway. The species occurs in less
acid rivers, ph > 6.0, all over the country, and produces
one to two generations per year. In water of 4.5 to 4.7
and low salinity (too = 30-35 /uS/cm), B. rhodani
cannot survive more than 2 days. At the same pH, but
higher salinity (too = 125- 130//S/cm), 10 percent of
the animals were still alive after 5 days. Field
observations indicate that B. rhodani fails to reproduce
and dies from physiological stress in water of ph < 5.0
(Raddum, 1979).
The freshwater shrimp, Gammarus lacustris, is one
of the most important food organisms for trout in
Norway. In one oligotrophic lake studied it constitutes
24 percent of the energy intake of trout (Lien, 1978). In
lowland lakes it has not been recorded below pH 6.6.
Experimental tolerance tests have shown that adult G.
lacustris can be eliminated during short-term acidifica-
tion below pH 5.5 (Hendrey, et al. 1976).
Freshwater snails, bearing calcareous shells, are
generally not found below pH 6.0. Only 5 of the 27
Norwegian species occur in lakes between pH 6.0 and
pH 5.2. Also the abundance of snails is reduced in acid
-------
THE ACID RAIN PROBLEM: MECHANISM AND EFFECTS
441
water. Small mussels also disappear around pH 6.0.
Only 3 of the 20 Norwegian species have been found in
lakes below pH 5.0 (Okland and Okland, 1980).
EFFECTS OF ACID WATER ON FISH
A regional decline in inland fisheries during the last
decades has been reported from acidified districts of
south Scandinavia, Canada, and the United States
(Muniz and Leivestad, 1980; Harvey, 1980; Schofield,
1976).
Fish decline in Canada was first reported in the
1960's from the La Cloche Mountains near Sudbury. In
this region 33 of 150 lakes were classified as "critically
acidic" with pH below 4.5 and 37 more lakes as
"endangered," pH 4.5 to 5.5.
Today, perhaps 200 lakes in Ontario are known to be
devoid of fish, because of acidification. In Nova Scotia a
dozen salmon rivers now show pH's in the 4.5 to 5.0
range, and the salmon catch is declining (Harvey,
1980).
Intensive studies of acid precipitation effects on fish
populations in 217 lakes in the Adirondack Mountains
showed that in 1975, more than half of the lakes had
pH below 5.0 and 90 percent of these lakes were
devoid of fish. Comparable data from 1929-1937
indicated that only 4 percent of these lakes were below
5 pH and devoid of fish. Entire fish communities (brook
trout, lake trout, white sucker, and others) were
eliminated over a period of 40 years, resulting from
decreased pH (Schofield, 1976).
In south Scandinavia the first effect of acidification
on fish became known early in this century when
salmon began to disappear from several southern
rivers, all of which are now acidic. Some of these rivers
have now lost their salmon completely. In Sweden, the
roach disappeared from some west coast lakes as early
as the 1920's and 1930's. it is estimated that in the
Swedish west coast region, which is most sensitive, 50
percent of the lakes now have pH below 6.0. For the
whole of Sweden, the number of lakes with pH below
6.0 is now about 10,000 (Dickson, 1975). Also fish
populations of char, perch, and pike have been
seriously affected. In south Norway, fish population
surveys of more than 5,000 lakes have shown that
within an area of 13,000 km2 fish life is now virtually
extinct. In an additional area of 20,000 km2 the lakes
are losing their fish.
Some lakes are already barren, many have sparse
and declining populations, and some still give
reasonable fish yields. The population status since
about 1940 is known for almost 3,000 lakes in the
affected districts in southernmost Norway. There is
evidence that the fish decline was moderate before
1940 and most pronounced since 1960. The number of
remaining trout populations is quickly being reduced.
At the present rate, the four southernmost counties of
Norway will have lost 80 percent of their trout
populations by 1990 (Sevaldrud, et al. 1980; Figure 3).
Regional data show a close correlation between
increased water acidity and loss of fish. In lakes with
low salt content the fish loss is greater than in lakes of
higher salt content and the same pH level. It is also
typical that small lakes at high altitudes lost their fish
populations first. Today about 80 percent of lakes above
Number of populations
3000 r
2500
2000
1500
1000
500
100
BROWN TROUT
Population changes
Present Recent
status changes
Sparse
Good
No data
on changes
Unaffected
— Decrease
No data
on changes
1940
1950 1960 1970 75
Figure 3. — Time trend for population losses of brown trout
from the four southernmost counties of Norway. (Muniz and
Leivestad, 1980.)
1,000 m above sea level are empty. The fish loss has
since gradually spread downstream.
High egg and fry mortality in acid water that reduces
younger age classes, is regarded as a main reason for
fish decline (Schofield, 1976), but other population
responses such as post-spawning extinction are also
known (Muniz and Leivestad, 1980). Massive fish kills
of adult fish during acid episodes, especially during
snowmelt, are well documented.
Seemingly contradictory results from field observa-
tions and laboratory tests for fish survival in artificially
acidified tap water have indicated that some toxic agent
other than acidity increases mortality under field
conditions. Aluminum which is present in high
concentrations in lakes in acidified districts, is now
held to be a critical element for fish mortality.
Exposure tests, field experiments, and physiological
research have in recent years led to the following
hypothesis for fish loss in acidified districts:
1. Aluminum is dissolved and leached from the soils
of catchments receiving acid precipitation, and Al ions
occur in high concentrations in acidified lakes and
streams.
-------
442
RESTORATION OF LAKES AND INLAND WATERS
2. In acid, clearwater lakes low in organic content,
the aluminum will mainly be present as inorganic
compounds, some of which are highly toxic to fish, and
probably to other aquatic animals as well.
3. The toxicity of Al in water varies with pH, having a
maximum around pH 5. Aluminum toxicity thus acts in
combination with the "pure" pH stress on fish
physiology.
4. Aluminum toxicity attacks the gills. Al content in
gills of acid-stressed fish may be six to seven times
higher than in reference fish. This leads to mucus
clogging. Aluminum disturbs the exchange of ions
across the gill membranes. High concentrations of
dissolved salts tend to ameliorate aluminum stress and
ion depletion.
5. The main physiological effects of pH/AI stress are:
(a) depletion of body salt content; (b) hyperventilation;
and (c) lowered blood oxygen tension. Other effects are
also observed.
6. In some species, like brown trout, metabolic
activity increases, possibly reducing energy available
for growth.
Salmonid fishes are generally more vulnerable to
acid than other important species.
EFFECTS OF ACID PRECIPITATION
ON VEGETATION
Anthropogenic sulfur may affect soil and vegetation
mainly by two pathways: Sulfur dioxide is a primary air
pollutant, acting directly on soil and vegetation growing
close to the emission sources. Acid precipitation
contains high concentrations of sulfate which is
derived from sulfur dioxide. Acid precipitation has a
much wider distribution than SO2 and may indirectly
affect vegetation through chemical or biological
changes in the soil (Dochinger and Seliga, 1977;
Abrahamsen, et al. 1976; Hutchinson and Havas,
1980).
The increase in anthropogenic sulfur emissions,
coupled with the increased height of emissions, have
led to the transport of sulfur pollutants over long
distances. Acid precipitation from the polluted air
masses may affect vegetation directly or indirectly by
interfering with important soil processes.
The direct contact between acid precipitation and
vegetation increases the leaching of some elements
from the foliage. Precipitation also washes off
substances dry-deposited on the vegetation. The total
effect is an increase in concentrations of most of the
compounds in throughfall compared to incident
precipitation.
Leaching from foliage is high in cations such as
calcium and potassium. There are indications of a pH-
dependent loss, possibly as a result of exchange with
r-T ions. Leaching can lead to the appearance of
deficiency symptoms in leaves. On the other hand,
vegetation acts as an efficient filter of the chemical
components in air and precipitation, and the cycling
from litter-fall to the soil to root uptake can be intense,
especially for plant nutrients such as nitrogen that are
in high demand. The levels of lead found in organic
matter in forest soils in remote areas in New England
were comparable with those in many heavily traveled
roadsides, and levels were rising (Reiners, et al. 1975).
Analyses of more than 500 moss and soil samples in
Norway (Hanssen, et al. 1980; Allen and Steinnes,
1980) show that long-range transport of trace
elements determines the distribution of lead, zinc, and
cadmium, and to some degree arsenic, antimony, and
selenium.
Direct effects of acid precipitation on forest trees
have been shown experimentally. The wax coating of
the outer layer (cuticula) of oak was eroded at
precipitation pH 3.2, possibly affecting water loss and
attacks by fungi and bacteria (Schriner, 1976).
Direct effects of acid precipitation on agricultural
crops depend on the particular cultivar, on precipitation
characteristics, and the growing conditions. Precipita-
tion at pH above 4 seems to present a low risk of
measurable reductions in growth or yield. At pH levels
between 4 and 3 many effects on crops have been
demonstrated, both positive and negative, and pH
below 3 seems to substantially increase the chances of
harmful effects on growth or yield (Jacobson, 1980).
PLANT NUTRIENTS AND FOREST
GROWTH
Loss of nutrient minerals, a natural process caused
by weathering, appears to be widespread and
enhanced from soils in areas with high deposits of
acidifying components. This has been observed in
input-output balances for the Hubbard Brook catch-
ment in New Hampshire (Likens, et al. 1977), from
studies of nine Norwegian catchments (Wright, et al.
1978), in the Soiling forest, the Federal Republic of
Germany (FRG) (Ulrich, 1980), and in lysimeter
experiments (Abrahamsen, 1979). The catchments
generally act as temporary sinks forhT, NOa, and NhU,
and as sources for Ca, Mg, Mn, and Al. Sulfate is
generally close to balance, but some studies (Uhrich,
1980; Andersson-Calles and Eriksson, 1979) indicate
an accumulation over several years in catchments. This
buildup is probably a recent process which started with
large scale emissions of SOz from fossil fuel burning. If
the deposition reaches a new and stable level, input
and output are expected to balance once more after
some time. Little is known of possible reemission of
sulfur in gaseous form after deposition.
Recently published data from the National Forest
Survey in Sweden illustrate the situation in that
country (Troedsson, 1980). Chemical data from 2,500
humus layer sites in the forested area between 59° and
61 °N show significant decreases in exchangeable
calcium, magnesium, and potassium between 1961-
1963 and 1971-1973. Exchangeable H+ and aluminum
have increased, but not significantly. The loss of Ca,
Mg, and K from the soil is interpreted partly as an effect
of atmospheric acid deposition. There is a strong
correlation between increasing age of the coniferous
forest and decreasing pH in the humus layer, and this
effect is stronger than the acidifying effect resulting
from atmospheric deposition (Troedsson, 1980).
When plant nutrients leach from soils exposed to
acid deposition faster than minerals weather (which
provides new dissolved compounds while consuming
H+ ions), the net loss may be important for plant
productivity. Loss of magnesium due to soil acidifica-
tion is already believed to restrict forest growth in parts
-------
THE ACID RAIN PROBLEM: MECHANISM AND EFFECTS
443
of central Europe (Ulrich, 1980). Concentrations of
aluminum in the soil solution are so high in some acid-
impacted soils that Al may possibly be toxic to tree
growth (Ulrich, Mayer, and Khanna, 1979; Voigt,
1980).
In soils with nitrogen and sulfur deficiency, acid
precipitation could have a positive growth effect.
Douglas fir stands in the Pacific Northwest region of
the United States are only one example. At the same
time it is suspected that acid precipitation may have
depleted potassium in some soils (Johnson, cited by
Roberts, 1980).
Effects of acid precipitation on forest growth are
therefore now considered a nutritional problem (apart
from possible direct effects). The increased deposition
of nitrogen and sulfur can be regarded as fertilization,
and the increased leaching of nutrient cations caused
by increased atmospheric deposition of sulfur com-
pounds will tend to cause nutrient deficiencies. Plant
requirements for different nutrients and soil properties
will determine whether the growth effects will be
negative or positive (Abrahamsen, 1980).
Experiments on the effect of artificial acidification on
forest growth under field conditions have been carried
out in Sweden and Norway. The Swedish experiments
have shown that increasing application of dilute H2SC>4
has significantly increased the basal area growth
(Tamm, et al. 1980). The Norwegian studies consist of
field plot experiments where artificial rain has been
produced by mixing ground water with HaSO-i to pH
values from 6 to 2. In one experiment with Scots pine,
increased height and diameter growth were observed
in 1976 and 1977 at the plots supplied with 250 mm of
water per year of pH 3, 2.5, and 2. In 1979, however,
the most acidified plots showed significantly less
growth than the other experiments (Tveite, 1980).
Although acidification seems to temporarily increase
the nitrogen availability in the soil, the increased
deposition of inorganic nitrogen from the atmosphere
is probably more important for growth increase (Wood
and Bormann, 1975; Abrahamsen, 1980). The nitrogen
deposition is currently 5 to 10 kg N/ha/year in
southern Scandinavia. As nitrogen is the main growth
limiting element in forests, increased deposition of
nitrogen will most likely increase forest growth.
Increased growth combined with increased leaching of
magnesium, calcium, and potassium may produce
future deficiencies in these elements {Abrahamsen,
1980).
Field investigations on possible growth effects in
boreal coniferous forests receiving acid-jarecipitation
have been inconclusive. Jonsson and Sundberg (1972)
classified areas in southern Sweden as relatively
resistant to acid rain and relatively susceptible to acid
rain, and compared growth trends in both areas by
measuring annual rings from groups of trees which
were otherwise nearly identical. They found a
statistically significant difference and "found no reason
for attributing the reduction in growth to any cause
other than acidification." These results, however, have
not been confirmed by Norwegian researchers (Abra-
hamsen, et al. 1976; Strand, 1980.)
A number of possible effects of acid precipitation on
the biological and biochemical processes in forest soil
have been identified, and are reviewed by Tamm (1976)
and Alexander (1980). Among these are:
1. Changes in soil microbiological populations, such
as decreases in bacteria and subsequent increase in
soil fungi. Effects on humus decomposition have been
noted.
2. Nitrogen turnover, which is connected to organic
matter decomposition. Effects, mostly reductions, have
been observed in N-mineralization, nitrification, and N-
fixation.
CONCLUSIONS
• Atmospheric transport of sulfur compounds and
other acidifying components has caused extensive
regional acidification of water courses in sensitive
areas, both in Europe and North America.
• The regions affected by acidification are presently
increasing in area. Lakes in these areas are now
characterized by low pH, high contents of sulfate, and
high concentrations of several metals, notably alumi-
num, which is leached from the catchments under
impact of acid precipitation.
• Acidification of inland waters has had major
effects on life in rivers and lakes. Investigations have
shown that all types of organisms in the freshwater
ecosystem are affected by acidification, ecosystem
structures are simplified, and the lakes probably have
become poorer in nutrients.
• A prominent feature of regional water acidification
is the extensive loss of fish populations, caused
primarily by reproductive failure. Physiological stress
and fish kills are caused by toxic combinations of water
acidity and high aluminum content.
• Acid precipitation and dry deposition of acidifying
components interact with vegetation surfaces, leading
both to adsorption and leaching from the foliage. Soils
which are impacted by acid deposition, lose basic
elements during the neutralization process. These
include, in particular, calcium and magnesium which
are important nutrients for plant growth. Aluminum
also is leached from soils under acidification, with toxic
consequences for aquatic life. The mobile sulfate ion
provided by acid deposition plays an essential role for
transport of cations in the soil solution.
• The possible negative effects on boreal forest
growth of nutrient deficiency caused by cation leaching
seem to be offset at least in the short term by the
fertilization effect by nitrogen compounds in acid
precipitation. Little is known of the time required for
possible long-term effects, for instance, of magnesium
deficiency to become extensive. Several important
biological and biochemical processes in soils are
affected by acid deposition.
REFERENCES
Abrahamsen, G. 1979. Impact of atmospheric sulphur
deposit.on on forest ecosystems. Proc. Symp. Oak Ridge
Natl. Lab. Life sci. ser. Gatlinburg, Tenn.
Abrahamsen, G., et al. 1976. Effects of acid precipitation on
coniferous forest. Pages 37-63 in F. H. Braekke, ed. SNSF-
project FR 6/76.
1980. Acid precipitation, plant nutrients and
forest growth. Proc. Int. Conf. Ecol. Impact Acid Precip.,
Sandefjord, Norway (in press).
-------
444
RESTORATION OF LAKES AND INLAND WATERS
Alexander, M. 1980. Effects of acid precipitation on
biochemical activities in soil. Proc. Int. Conf. Ecol. Impact
Acid Precip., Sandefjord, Norway (in press).
Allen, R. 0., and E. Steinnes. 1980. Contribution from long-
range atmospheric transport to the heavy metal pollution of
surface soil. Proc. Int. Conf. Ecol. Impact Acid Precip.,
Sandefjord, Norway (in press).
Aimer, B., et al. 1978. Sulphur pollution and the aquatic
ecosystem. In J. 0. Nriagu, ed. Sulphur in the environment,
part II. J. Wiley and Sons, New York.
Andersson — Calles, U. M., and E. Eriksson. 1979. Mass
balance in representative basins. Nordic Hydrol. 2-3:99.
Cronan, C. S., et al. 1978: Forest floor leaching: Contributions
from mineral, organic and carbonic acids in New Hampshire
subalpine forests. Science 200:309.
Davis, R., and F. Berge. 1980. Atmospheric deposition in
Norway during the last 300 years as recorded in SNSF lake
sediments. II Diatom stratigraphy and inferred pH. Proc. Int.
Conf. Ecol. Impact Acid Precip., Sandefjord, Norway (in
press).
Dickson, W., 1975. The acidification of Swedish lakes. Inst.
Freshw. Res., Sweden, Rep. 54:8.
Dochinger, L. S., and T. A. Seliga, eds. 1976. Impact of acid
precipitation on coniferous forest ecosystems. Forest
Service, U.S. Dep. Agric.
Drablos, D., I. Sevaldrud, and J. A. Timberlid. 1980. Historical
land-use changes — lake acidification — fish population.
Proc. Int. Conf. Ecol. Impact Acid Precip., Sandefjord,
Norway"(in press).
Grahn, O., H. Hultberg, and L. Landner, 1974. Oligotrophi-
cation — a self-accelerating process in lakes subjected to
excessive supply of acid substances. Ambio 3:93.
Hanssen, J. C., et al. 1980. Atmospheric deposition of trace
elements in Norway. Proc. Int. Conf. Ecol. Impact Acid
Precip., Sandefjord, (in press).
Harvey, H. H. 1980. Widespread and diverse changes in the
biota of North American lakes and rivers coincident with
acidification. Proc. Int. Conf. Ecol. Impact Acid Precip.,
Sandefjord, Norway (in press).
Hendrey, G. R., and F. Vertucci. 1980. Benthic plant
communities in acidic Lake Golden, New York: Spagnum
and the algal mat. Proc. Int. Conf. Ecol. Impact Acid Precip.,
Sandefjord, Norway (in press).
Hendrey, G. R., et al. 1976. Acid precipitation: Some
hydrobiological changes. Ambio 5:224.
Hendriksen, A. 1979. A simple approach for identifying and
measuring acidification of freshwater. Nature 5704:542.
1980. Acidification of freshwaters — a large scale
titration. Proc. Int. Conf. Ecol. Impact Acid Precip.,
Sandefjord, Norway (in press).
Hutchinson, T. C. and M. Havas, eds. 1980. Effects of acid
precipitation on terrestrial ecosystems. Plenum Press.
Jacobson, J. S. 1980. The influence of rainfall composition
on the yield and quality of agricultural crops. Proc. Int. Conf.
Ecol. Impact Acid Precip., Sandefjord, Norway (in press).
Johannessen, M., and A. Hennksen. 1978.Chemistry of
snowmelt water: Changes in concentrations during melting.
Water Resour. Res. 14(4):615.
Johnson, D. W. 1980. Site susceptibility to leaching byH SO
in acid rainfall. Pages 525-535. In T. C. Hutchinson, and M.
Havas, eds. Effects of acid precipitation on terrestrial
ecosystems. Plenum Press.
Jonsson, B. and R. Sundberg. 1972. Has the acidification by
atmospheric pollution caused a growth reduction in
Swedish forests? Skogshogskolan (Sweden) Inst. Skogs-
prod. Uppsatser 20.
Leivestad, H., et al. 1976. Effects of acid precipitation on
freshwater organisms. Pages 87-111. In F. H. Braekke, ed.
SNSF-project FR 6/76.
Lien, L. 1978. The energy budget of the brown trout
population of Ovre Heimdalsvatn. Holarctic Ecol. 1:279.
Likens, G. E., F. H. Bormann, and J. S. Eaton. 1980.
Variations in precipitation and stream water chemistry at
the Hubbard Brook experimental forest during 1964 to
1977. Pages 443-464 in Effects of acid precipitation on
terrestrial ecosystems. T. C. Hutchinson and M. Havas, eds.
Plenum Press.
Likens, G. E., et al. 1977. Biogeochemistry of a forested
ecosystem. Springer-Verlag.
Likens, G. E., et al. 1979. Acid rain. Sci. Am. 241:43.
Linzon, S. N. and P. J. Temple, 1980. Soil resampling and pH
measurements after an 18 year period in Ontario. Proc. Int.
Conf. Ecol. Impact Acid Precip., Sandefjord Norway (in
press).
Mohn, E. et al. 1980. Regional surveys of the chemistry of
small Norwegian lakes: a statistical analysis of the data
from 1974-1978. Int. Conf. Ecol. Impact Acid Precip.,
Sandefjord, Norway (in press).
Muniz, I. P., and H. Leivestad. 1980. Acidification — effects on
freshwater fish. Proc. Int. Conf. Ecol. Impact Acid Precip.,
Sandefjord, Norway (in press).
Okland, J. and K. A. Okland. 1980. pH level and food
organisms for fish: Studies of 1,000 lakes in Norway. Proc.
Int. Conf. Ecol. Impact Acid Precip., Sandefjord, Norway (in
press).
Raddum, G. 1979. Effects of low pH on insect larvae. SNSF-
project IR 45/79. (In Norwegian.)
Reiners, N. A., R. H. Marks, and P. M. Vitousek. 1975. Heavy
metals in subalpine and alpine soils of New Hampshire.
Oikos 26:264.
Roberts, T. M. 1980. Effects of acidity on nitrogen cycling
(Rapporteur's summary) Pages 599-600 in T. C. Hutchinson,
and M. Havas, eds. Effects of acid precipitation on terrestrial
ecosystems. Plenum Press.
Schofield, C. L. 1976. Acid precipitation: Effects on fish.
Ambio 5:228.
Seip, H. M. 1980. Acidification of freshwater— sources and
mechanisms. Proc. Int. Conf. Ecol. Impact Acid Precip.,
Sandefjord, Norway (in press).
Sevaldrud, I., I. P. Muniz, and S. Kalvenes 1980. Loss of fish
populations in southern Norway. Dynamics and magnitude
of the problem. Proc. Int. Conf. Ecol. Impact Acid Precip.,
Sandefjord, Norway (in press).
Shriner, D. S. 1976. Effects of simulated rain acidified with
sulfuric acid on host-parasite interactions. Pages 873-879.
in L. S. Dochinger, and T. A. Seliga, eds. Effects of acid
precipitation on terrestrial ecosystems. Plenum Press.
Strand, L. 1980. The effect of acid precipitation on tree
growth. Proc. Int. Conf. Ecol. Impact Acid Precip.,
Sandefjord, Norway (in press).
Tamm, C. 0. 1976. Acid precipitation: Biological effects in
soil and on forest vegetation. Ambio. 5-6:235.
Tamm, C. O. et al. 1980. Effects of artificial acidification with
sulphuric acid on tree growth and soil chemistry in Scots
pine forests. Proc. Int. Conf. Ecol. Impact Acid Precip.,
Sandefjord, Norway (in press).
Troedsson, T. 1980. Ten years acidification of Swedish forest
soils. Proc. Int. Conf. Ecol. Impact Acid Precip., Sandefjord,
Norway (in press).
Tveite, B. 1980. Effects of acid precipitation on soil and
forest. Tree growth in field experiments. Proc. Int. Conf.
Ecol. Impact Acid Precip., Sandefjord, Norway, (in press).
Ullrich, B. 1980. Production and consumption of hydrogen
ions in the ecosphere. Pages 255-282. in T. C. Hutchinson
and M. Havas, eds. Effects of acid precipitation on terrestrial
ecosystems. Plenum Press.
-------
THE ACID RAIN PROBLEM: MECHANISM AND EFFECTS
Ulrich, B., R. Mayer, and P. K. Khanna. 1979. Deposition von
Luftwerunreinigungen und ihre Auswirkungen in Waldo-
kosystemen in Soiling. Schr. Fortsl. Fak. Univ. Gottingen
und Niedersachs. Fortsl. Vers. Anst. 58.
Vigt, G. K. 1980. Acid precipitation and soil buffering
capacity. Proc. Int. Conf. Ecol. Impact Acid Precip.,
Sandefjord. Norway (in press).
Wood, T. and F. H. Bormann, 1975. Increases in foliar leaching
caused by acidification of an artificial mist. Ambio 4:169.
Wright, R. F. 1977. Historical changes in the pH of 128 lakes
in southern Norway and 130 lakes in southern Sweden over
the period 1923-1976. SNSF-project TN 37/77.
Wright, R. F., et al. 1978. Inputs and outputs of water and
major ions at 9 catchments in southern Norway, July 1974-
june 1975. SNSF-project TN 39/78.
1980. Acidified lake districts of the world: a
comparison of water chemistry of lakes in southern Norway,
southern Sweden, southwestern Scotland, the Adirondack
Mountains of New York, and southeastern Ontario. Proc.
Int. conf. Ecol. Impact Acid Precip., Sandefjord, Norway (in
press).
-------
446
CHANGING pH AND METAL LEVELS IN STREAMS AND
LAKES IN THE EASTERN UNITED STATES CAUSED BY
ACIDIC PRECIPITATION
JAMES N. GALLOWAY
Department of Environmental Sciences
University of Virginia
Charlottesville, Virginia
STEPHEN A. NORTON
DENIS W. HANSON
Department of Geological Sciences
University of Maine at Orono
Orono, Maine
JOHN S. WILLIAMS
School of Oceanography
University of Rhode Island
Kingston, Rhode Island
ABSTRACT
The average pH of precipitation falling east of the Mississippi River is less than 5.0, locally less
than 4.0. The pH of rain and snow has decreased locally up to 0.75 units in the last 25 years.
Aquatic ecosystems in many large areas are vulnerable to this acidic precipitation because of
geologic and soil conditions. Time studies of pH and alkalinity for sensitive surface waters exist
for North Carolina, Pennsylvania, the Adirondack Mountains area of New York, New Hampshire,
and Maine. The duration of observations ranges from 2 to50 years. All studies indicate generally
decreasing pH and alkalinity. Precipitation event and snow melt studies of pH and alkalinity in
Virginia, the Adirondack Mountains of New York, and at Hubbard Brook, N.H. indicate that only
mildly acidic (5 to6) or circum-neutral (6 to7+) streams may undergo severe pH depression (1 to
3 pH units). Heavy metal data for lakes and streams are sporadic and widely distributed.
Precipitation heavy metal data are even rarer. Paleolimnologic data from New England and the
Adirondack Mountains of New York indicate increasing atmospheric fluxes of many metals
(especially Pb and Zn). Increases in Pb are apparently related to atmospheric particulates. Zn is
chemically more mobile and in strongly acidified (pH < 5.0) aquatic ecosystems there is a net
loss from the system. Increases in Al in surface waters in the Adirondack Mountain area
correlate strongly with pH decreasing below6.0. Leaching of Mn, Zn, and Cafrom acidified soils
and lake sediments suggest that concentrations of these metals have increased in surface waters
over the last 50 years and may now be decreasing because of impoverished soils.
INTRODUCTION
Considerable literature evaluates the impact of
anthropogenic activities within drainage basins on
surface water quality (e.g., Likens, et al. 1970) and on
sediment chemistry (e.g., Shapiro, Edmondson, and
Allison, 1971; Bradbury and Megard, 1972) in the
United States. Most of these studies focused on gross
pollution or large disturbances of a steady state. Only
recently (Schofield, 1976; Davis, et al. 1978; Norton,
Hess, and Davis, 1980) has attention been focused on
aquatic ecosystems with no drainage basin disturb-
ances; there it is possible to isolate the effects of acidic
precipitation and associated metal loading on surface
water quality and sediment chemistry.
Polluted air and thus polluted precipitation are not
inventions of 20th century industrialized society
(TeBrake, 1975; Smith, 1872). However, only recently
has the regional (even hemispheric) scope of atmos-
pheric and precipitation pollution been recognized (in
the U.S., Cogbill and Likens, 1974); in Scandinavia,
Oden, 1976; in Greenland, Cragin, et al. 1975). One of
the few positive effects of thermonuclear bomb testing
has been the documentation of global dispersal of
reaction products (e.g., Cs137 and Sr90) (Toonkel, 1980)
and obviously other pollutants.
Historical data on the pH of precipitation and surface
waters in the United States prior to significant air
pollution are non-existent. Before the mid-1950's, pH
measurements were generally made with colorimetry,
making comparison with modern (electrode) measure-
ments difficult and somewhat ambiguous (Spikkeland,
1977; Boyd, 1980). Even measuring pH with electrodes
is difficult because of the low ionic strength of
precipitation and some surface waters (Galloway, etal.
1979). Early measurements of pH of surface waters
were performed downstream from waters which would
respond to changes in precipitation chemistry and in
lakes subject to direct human influence. Precipitation
-------
THE ACID RAIN PROBLEM: MECHANISM AND EFFECTS
447
pH measurements, until the establishment of the
National Atmospheric Deposition Program network
(NADP. 1980), have been non-regional, short-lived, and
difficult to compare for numerous reasons. Conse-
quently, strict correlation of the changing pH of
precipitation with surface water pH changes is not
generally possible.
Similarly, Lazrus, et al. (1970) were the first to
produce data for heavy metals in precipitation. More
recent studies are short-lived, non-regional, and
generally not comparable because of differing collec-
tion or analytical techniques (Galloway, et al. 1979,
1980). However, there is no doubt that concentrations
have increased on a local (Bertine and Goldberg, 1971)
and hemispheric scale (Herron, et al. 1976). Chemical
profiles in recent deposits in ombrotrophic peat bogs
(Livett, et al. 1979) suggest an increased atmospheric
concentration of certain metals, notably Pb and Zn.
Empirical studies along modern environmental
gradients have been undertaken because of the lack of
historical data for pH of, and metals in, precipitation
and surface waters. The basis of these studies is that if
one is able to control variables in certain environmental
parameters, it is possible to assess time dependent
processes related to increased atmospheric deposition
of acids and metals on (a) acidification of streams, (b)
acidification of lakes, (c) acidification of soils, and (d)
mobilization and/or accumulation of metals. Simple
unambiguous conclusions can be reached by these
types of studies. For example, small oligotrophic non-
dystrophic lakes, in the absence of watershed
disturbances, become acidified only where the precipi-
tation is acidic.
VULNERABILITY/SENSITIVITY
For significant changes in pH or metal concentra-
tions to occur in surface waters in response to changes
in the chemistry of precipitation, the aquatic ecosystem
must be vulnerable and sensitive.
Virtually all of the eastern half of the United States is
receiving precipitation with an annual average weight-
ed pH less than 5.0. The northeastern States (New
England, New York, New Jersey, and Pennsylvania) are
receiving precipitation with an annual pH less than 4.5.
Precipitation with a pH less than 4.0 is common even
as far northeast as central Maine; pH's less than 3.0
have been recorded (NADP, 1980; Likens, 1976). Acidic
precipitation also occurs in Washington (Gillion and
Horner, 1977) and California (McColl, 1980; Morgan
and Liljestrand, 1980) but the geographic distribution
on the west coast is relatively restricted at present.
Thus, half of the surface waters of the United States
are potentially vulnerable to low pH of the precipitation.
The response to additional acidic precipitation is a
measure of sensitivity. Chemically, sensitivity mea-
sures the proton assimilative capacity and assimilation
kinetics of the ecosystem. In the absence of anthropo-
genic disturbances, sensitivity is controlled by the soils
and bedrock geology. Assimilation may occur by:
A. Solution of rocks/minerals such as
AI(OH)3x1 + 3Haq = Alaq*+ 3H2Oaq and
CaCOa + Haq = CaVq + HCOaq or
B. Loss of alkalinity by such reactions as
HCOa + Haq = HaCOs and
aq ^ aq
H2PO"4 + Haq = H3PO4° and
aq aq
C. Cation exchange reactions such as
AI(OH);aq + 2Haq = AI(OH)aaq + 2H2Oaq or
CA++ — OrganiCsohd + 2H+ = Hz — Organicsoiw + Ca+
Neutralization of acidic precipitation in soils primarily
uses mechanisms A and C. Neutralization of surface
waters is dominated by mechanism B.
McFee (1980) and Norton (1980) have developed
maps based on soils and geologic criteria, respectively,
showing the distribution of sensitive areas in the
eastern United States. These maps enable prediction of
impact caused by acidic precipitation. A portion of one
of these maps is shown in Fig. 1 . It demonstrates the
scale of variability of sensitivity to be expected in a
geologically complex area. Similar results prevail for
the soils analysis. Complete coverage depicting
geologically sensitive areas for all of the eastern United
States is given in Hendrey, et al. (1980).
Figure 1. — Sensitivity of part of coastal Maine to acidic
precipitation. Sensitivity is indicated as follows: 1
surface waters will show measurable decline in pH and
alkalinity; 2 = surface waters will locally show measurable
decline in pH and alkalinity, particularly during
precipitation episodes; 3 surface waters will only
undergo pH and alkalinity decline during periods of
overland flow. Dashed lines are county boundaries.
pH OF FRESH SURFACE WATERS
Although pH data are abundant for lakes and
streams, they are generally not suitable for temporal
studies of changing pH for the following reasons:
I.Most of the studies on streams and lakes have
focused on populated areas where local anthropogenic
activity may dominate the chemistry and where
sensitivity has been lost by virtue of upstream
heterogeneous soils and geology (Fig. 1). For example,
some of the longest series of data are U.S. Geological
Survey gauging stations which are not located on first,
second, or third order streams.
-------
448
RESTORATION OF LAKES AND INLAND WATERS
KLARALVEN
. 1.0-
M
Figure 2. — Short-term pH and Al variations in an
Adirondack Mountain stream. New York. Generalized
from Schofield (1977). Note reciprocal relationship.
2. Episodic excursions of pH in streams are common.
Burns and Galloway (in Hendrey, et al. 1980) in
Virginia, Schofield (1977, 1979) in the Adirondack
Mountains, N.Y. (Fig. 2), Hornbeck, et al. (1977) at
Hubbard Brook, N.H., and Haines and Norton in Maine
(unpubl. data) have demonstrated that the pH of
unbuffered streams can oscillate as much as 2 pH units
over a few days, depending on the relative proportions
of overland and groundwater flow involved in stream
discharge.
3. Changing land use has been responsible locally for
short to long-term changes in stream and lake pH
(Likens, et al. 1970; Rosenqvist, et al. 1980).
Consequently, studies based on temporally paired
stream pH are suspect. Arnold, et al. (1980) obtained
paired data for 314 streams with pH and alkalinity
measured twice (more than 1 year apart). Of the 314
streams 107 (34 percent) showed a decrease in pH,
alkalinity, or both. Therefore, 66 percent were constant
or increased in both pH and alkalinity. Although Arnold,
et al. (1980) claim that 34 percent have been acidified
by acidic precipitation, it is far more likely they are
randomly more acidic on one individual measurement.
Only with a large sample of randomly distributed pairs'
could one hope to detect a statistically meaningful
temporal drift in pH. High pH surface waters are
particularly susceptible to large random variations
because of factors other than changing precipitation
pH. Ideally, low alkalinity streams should be sampled at
closely spaced intervals over a long period of time such
as has been done by workers in Sweden (Oden, 1976)
(Fig. 3).
Because of the volume and long-term flushing
characteristics of lakes, they may integrate, smoothing
out the pH variations caused by the changing pH of
precipitation. Nonetheless, variations in lake water pH
-I T
1966
T
r
1968 1970
1 T
1972
-74
- PH
1974
-5.8
Figure 3. — Long-term pH variation in an southern Sweden
stream. Generalized from Oden. (1976).
may occur which are unrelated to long-term changes in
the pH of precipitation. These variations may be caused
by:
1. Seasonal changes in surface runoff/groundwater
flow.
2. Photosynthesis/respiration in the water column.
3. Sediment/water interaction.
Consequently, comparison of historic data for pH
trend analysis for lakes is plagued with the same
general problems as for streams, perhaps to a lesser
degree. Again, numerous paired data, randomly
distributed in time (and separated by as much time as
possible) and randomly distributed with respect to all
variables, should reveal pH trends, particularly for low
alkalinity waters.
Fig. 4 shows a systematic shift in 27 paired North
Carolina stream pH's, separated by at least 15 years.
Twenty-three of 27 streams were more acidic in 1979
than in 1960-64. Random variations should distribute
the points equally about the diagonal line (Fig. 4).
Similar relationships exist for 35 streams showing
decreased alkalinity.
6.0-
6.5 70
pH, 1960-64
75
Figure 4. — pH (1979) versus pH (1960-1964) for North
Carolina streams. Diagonal line is the locus of no change
(from Hendrey, et al. 1980).
In the Adirondack Mountains of New York, Schofield
(1976) selected a group of high altitude lakes for
comparative studies. Most of these were located in
-------
THE ACID RAIN PROBLEM: MECHANISM AND EFFECTS
449
sensitive terrain and analysis revealed a marked
decline in pH of lakes. In 1929-37, 5 percent of 217
lakes had a pH below 5.0; in 1976,51 percent had a pH
below 5.0. Hendrey, et al. (1980) analyzed paired data
from lakes and streams in New Hampshire and found
relationships there to be similar to those in North
Carolina. Davis, et al. (1978) studied 1,368 low
elevation lakes in Maine and reported a similar trend
(Fig. 5). Of 37 low elevation oligotrophic lakes in Maine
(Davis, et al. 1979) 31 had decreased 0.2 to 0.7 pH
units between 1935-1945 and 1978.
72
6.8
PH
6.4
5.6
J31
1940 1950 1960
DATE
1970
Figure 5. — Annual mean pH'sfor a data set from 1368
lakes in Maine. Each mean is labeled with the number
of pH readings on which it isjjased (Davis, et al. 1978).
METALS IN FRESH SURFACE WATERS
Recent literature contains abundant data for rivers
and lakes on trace metals (other than the major
elements Na, K, Ca, Mg, and Si) such as Fe, Mn, AI,Zn,
Pb, Cd, etc. However, most of the studies are not useful
in determining temporal changes in trace metal
content caused by atmospheric deposition. Most of the
studies were initiated because of suspected pollution
by anthropogenic activities located within the drainage
basin. Commonly, the concentrations for these pollu-
tants far exceed the concentrations one would expect
to find related to changes (either pH or metal
concentration) in the chemistry of precipitation.
Additionally, most of these studies are on higher order
(third, fourth, or more) streams where effects of
changing precipitation pH are less pronounced. Also,
the techniques for chemical analysis have evolved
rapidly and are not strictly comparable with older
results. Only recently has it become possible to analyze
directly for some metals in the ug/l range (Zn, Pb, Hg,
Cd, Cu, Al) without pretreatment such as extraction or
evaporation (Kleinkopf, 1960).
Just as for pH, metal concentrations in surface
waters are subject to short-term variations. Con-
sequently, long-term studies of rivers and lakes are
necessary to assess long-term changes caused by
changing precipitation pH (and consequent changed
metal mobility) and atmospheric deposition of metals.
To our knowledge, no useful long-term data exist for
trace metals for lakes or streams in the United States
which enable assessment of precipitation-related
changes. Therefore, to anticipate such changes, one
must turn to either transect studies or paleolimnologic
evidence.
Geographic transects for metals (precipitation-
derived or leached) in surface waters of chemically
comparable water bodies have not been done in the
United States as they have in Norway (Henriksen and
Wright, 1978) where pH of precipitation relates to trace
metal content of low pH lakes. An alternative approach
is to evaluate the relationship between pH and the
concentration of some metal in a variety of surface
waters in a small area receiving relatively uniform
composition precipitation. Fig. 6 shows the relationship
between Al and pH for lakes at high altitude in the
Adirondack Mountains, N.Y. Similar relationships have
been observed in southern Norway. From this we might
anticipate that Al should increase in surface waters as
pH decreased with time (Norton, 1976). Similar
relationships should exist for Zn, Mn, and other metals
with pH sensitive solubilities (Norton, Henson, and
Campana, 1980) and have been noted by Schofield
(1976) in an area receiving relatively uniform precipi-
tation but with widely varying surface water pH. Data
for this type of analysis must be carefully evaluated
because Al (or any other metal) may vary drastically
with pH (Fig. 2) because of varying proportions of
overland and groundwater flow to lakes and streams.
2000
1000-
100
50
20
10
o o fa Oo
o°° °
"V °o8 °o° ooo
oo ° 8
QO o
_L
PH
Figure 6. — Aluminum versus pH for 217 high altitude
lakes in the Adirondack Mountains, N.Y. Generalized from
Schofield (1976).
-------
450
RESTORATION OF LAKES AND INLAND WATERS
Paleolimnologic chemical analyses of sediment cores
from unpolluted lakes have been used to evaluate
changes in metal concentrations in the water or fluxes
of the metals through the ecosystem. Although most
studies have focused on lakes with drainage-basin
sources of metals, several studies have deliberately
focused on lakes with undisturbed watersheds except
for natural successional changes and natural catas-
trophes such as fires, pests, floods, etc.
Lake sediments may behave as sinks for certain
metals (e.g., Pb). Consequently, an increased flux from
the atmosphere will be reflected in concentration
profiles in the sediment (Fig. 7). Other metals (e.g., Zn,
Lazrus, et al. 1970) although proportionately more
abundant in lower pH precipitation (NADP, 1980) may
accumulate in non-acidified ecosystems, reach steady
state in moderately acidified ecosystems, and decrease
in strongly acidified systems (Fig. 7). This corresponds
to increasing, constant, and decreasing concentrations,
respectively, in current sediments. The ubiquitous and
concurrent rise in heavy metals in sediments in
relatively pristine lakes (Fig. 8) suggests that atmo-
spheric deposition causes the observed changes.
GRRNITE
SPECK
UNNflMED
Zn
CONC. VS. DEPTH ICM)
Figure 7. — Pb and Zn profiles for sediment from four
New England Lakes: circum-neutral (Maranacook,
Maine); slightly acidic (pH. 5-6) (Granite, N.H.); and
acidic (pH < 5) kettle pond ("unamed" Pond, Maine).
(Norton, et al. 1980b).
Chemical profiles from strongly acidified lakes
(Williams, 1980) (see e.g., Dream Lake, pH =4.5, Fig. 9)
suggest that detritus reaching the lake has been
depleted of Ca, implying a temporary elevation of Ca in
surface waters until readily leached Ca is removed.
(Mn, Cu, Zn, and Mg also decrease, as expected during
acidification.) Malmer (1976) noticed a decrease with
time in Ca in southern Sweden surface waters as did
Thompson in Nova Scotia rivers (1980, mss.). Other
workers (e.g., Watt, Scott and Ray, 1979, also in Nova
Scotia) found no change in Ca with a decrease in pH in
certain lakes. Schofield (1976) found in 219 lakes a
decrease in Ca with decrease in pH. Presumably, these
variations represent different stages in the release of
Ca from soils during acidification. Hanson (1980) found
U
0
1O
20
30
0
1O
a
QJ
°20
0
4
8
12
WOODS LAKE,NY
background
MOUNTAIN LAKE
VA
i i i i i
i i i i i
1978
1918
1858
1978
1948 w
1848 %
1748 £
1648 LU
1548 <
1978 x
O
1930 JX
<
1880
1830
0 20 40 60 80 100 12O
ug Pb g'1
Figure 8. — Pb concentrations in sedimentfrom Mountain
Lake, Va. (Galloway, et al. 1980), Woods Lake, N.Y.
(Galloway and Likens, 1979), and Speck Pond, Maine
(Davis, et al. 1979).
H20
TI02
K20
Nfl20 HL203
CflQ
MGO
ORG
FEO
MNO
ZN
PB
SI02
1 5 75 100
PCI IGH PCI ION
CU
IN mi, OHT, on icNiitn SEDIMENT vs DEPTH (cm
Figure 9. — Chemical profiles of sediment from Dream Lake,
N.H. Williams, 1980.
-------
THE ACID RAIN PROBLEM: MECHANISM AND EFFECTS
451
decreasing Ca, Mn, Mg, and K in soil litter subjected to
increasingly acidic precipitation on a transect from
southern Vermont (site 1) to the Gaspe' Peninsula,
Quebec (site 14) (Table 1). Steady-state release (pre-
air-pollution) would be followed by increased Ca in
surface waters with lowered pH, followed by a
decrease in Ca to a new equilibrium steady state
release, the level depending on the new pH.
Table 1. — Chemistry of forest litter from high altitude fir
forests. Sample sites range from southern Vermont (site 1) to
the Gaspe' Peninsula, Quebec (site 14). The pH of
precipitation ranges from about 4.0 to 4.6 Details of collection
and analysis are in Hanson (1980).
Site
1
2
3
4
5
*6
7
8
9
10
11
*12
13
14
Dry
wt %
Ca
0.370
0.216
0.373
0.499
0.400
0.653
0.301
0.494
0.654
0.628
0.814
0.749
1.006
0.962
Ca/AI
1.03
0.27
0.63
1.00
0.68
1.52
0.37
1.27
1.60
1.40
2.81
0.95
2.05
2.53
ppm
Mn
110
49
122
182
278
373
270
364
259
297
424
255
752
552
Mn/AI
0.31
0.06
0.21
0.36
0.47
0.87
0.33
0.93
0.63
0.66
1.46
0.32
1.54
1.45
Dry
wt%
Mg
0.050
0.042
0.041
0.065
0.050
0.070
0.068
0.058
0.059
0.078
0.064
0.050
0.070
0.063
Mg/AI
1.39
0.53
0.69
1.30
0.85
1.63
0.84
1.49
1.44
1.73
2.21
0.63
1.43
1.66
"anomalous sites, probably contaminated with mineral soil.
SUMMARY
Acidic precipitation and associated metal deposition
in the eastern United States have caused the following
changes:
*
1. Decreasing pH in lakes and streams rendered
sensitive by soils and bedrock chemistry.
2. Decreasing alkalinity in the same lakes and
streams.
3. Increasing dissolved Al and Ca and probably other
cations in acidifying aquatic ecosystems.
4. Increased flux to the aquatic ecosystems of some
trace metals (e.g., Pb) and accumulation/steady
state/release/net loss from the system of other
metals, depending on the pH.
Changes that can be reasonably anticipated with
increasing acidification include.
1.Increasing levels of dissolved Al, Fe, and Mn and
other major elements (solution of soil minerals).
2. Increasing Ca, Mg, and K (desorption).
3. Increasing levels of metals (Cd, Cu, Zn) whose
mobility is increased by lower pH.
REFERENCES
Arnold, D. E., R. W. Light, and V. J. Dymond. 1980. Probable
effects of acid precipitation on Pennsylvania waters. EPA
600/3-80-012. U.S. Environ. Prot. Agency.
Bertine, K. K. and E. D. Goldberg. Fossil fuel combustion and
the major sedimentary cycle. Science 173: 233.
Boyd, C. E. 1980. Reliability of water analysis kits. Trans. Am.
Fish. Soc. 109: 239.
Bradbury, J. P. and R. O. Megard. 1972. Stratigraphic record
of pollution in Shagawa Lake, Northeastern Minnesota.
Geol. Soc. Am. Bull. 83:2639.
Braekke, F. H., ed. 1976. Impact of acid precipitation on forest
and freshwater ecosystems in Norway. Res. Rep. 6/76,
S.N.S.F., Oslo, Norway.
Cogbill, C. V. and G. E. Likens. 1974. Acid precipitation in
northwestern United States. Water Resour. Res. 10:1133.
Cragin, J. H., M. M. Herron, and C. C. Langway, Jr. 1975.The
chemistry of 700 years of precipitation at Dye 3, Greenland.
Cold Regions Res. Eng. Lab. Rep. 341.
Davis, R. B.,S. A. Norton, and D. F. Brakke. 1979. Heavy metal
deposition in bottom sediments and acidification of New
England lakes from atmospheric inputs, and effects on lake
biota (abs.). Conf. Great Lakes Res.
Davis, R. B., et al. 1978. Acidification of Maine (U.S.A.) lakes
by acidic precipitation. Verh. Int. Verein. Limnol. 20:532.
Galloway, J. N. and G. E. Likens. 1978. The collection of
precipitation for chemical analysis. Tellus 30:71
1979. Atmospheric enhancement of metal
deposition in Adirondack lake sediments. Limnol. Oceanogr.
24427
Galloway, J. N., B. J. Cosby, Jr., and G. E. Likens. 1979. Acid
precipitation: Measurement of pH and acidity. Limnol.
Oceanogr. 24:1161.
Galloway, J. N., S. J. Eisenreich, and B. C. Scott. 1980. Toxic
substances in atmospheric deposition: A review and
assessment. Unpubl. Rep. U.S. Environ. Prot. Agency.
Gillion, R., and R. Horner. 1977. Bear Lake: Current status
and the consequences of residential development. Unpubl.
mss. University of Washington, Seattle.
Hanson, D. W. 1980. Acidic precipitation-induced chemical
changes in subalpine fir forest organic soil layers. M.S.
Thesis. University of Maine at Orono.
Hendrey, G. R., et al. 1980. Geological and hydrochemical
sensitivity of the eastern United States to acid precipitation.
EPA 600/3-80-024. U.S. Environ. Prot. Agency.
Henrikson, A., and R. F. Wright. 1978. Concentrations of
heavy metals in small Norwegian lakes. Water Res. 12:101.
Herron, M. M., et al. 1976. Vanadium and other elements in
Greenland ice cores. Cold Regions Res. Eng. Lab. Res. Rep.
76-24. Natl. Sci. Found.
Hornbeck, J.W., G. E. Likens, and J. S. Eaton. 1977. Seasonal
patterns in acidity of precipitation and their implications for
forest stream ecosystems. Water Air Soil Pollut. 7:355.
Kleinkopf, M. D. 1960. Spectrographic determination of trace
elements in lake waters of northern Maine. Geol. Soc. Am.
Bull. 71:1231.
Lazrus, A. L., E. Lorange, and J. P. Lodge, Jr. 1970. Lead and
other metal ions in United States precipitation. Environ. Sci.
Technol. 4:55.
Likens, G. E. 1976. Acid rain. Chem. Eng. News 54:29.
Likens, G. E. et al. 1970. Effects of forest cutting and
herbicide treatment on nutrient budgets in the Hubbard
Brook watershed-ecosystem. Ecol. Mon. 40:24.
Livett, E. A., J. A. Lee, and J. H. Tallis. 1979. Lead, zinc, and
copper analyses of British blanket peats. Jour. Ecol. 67:865.
Malmer, N. 1976. Acid precipitation: chemical changes in the
soil. Ambio 5:231.
McColl, J. G. 1980. A survey of acid precipitation in northern
California. Calif. Air Resour. Board. Final Rep.
-------
452
RESTORATION OF LAKES AND INLAND WATERS
McFee, W. W. 1980. Sensitivity of soil regions to long term
acid precipitation. EPA 600/3-80-013. U.S. Environ. Prot.
Agency.
Morgan, J. J. and H. M. Liljestrand. 1980. Measurement and
interpretation of acid rainfall in the Los Angeles Basin. Calif.
Inst. Technol. Unpubl. mss.
National Atmospheric Deposition Program. 1980. Data Rep:
1, 2, 3, and 4: Nat. Resour. Ecol. Lab. Fort Collins, Colo.
Norton, S. A. 1976. Changes in chemical processes in soils
caused by acid precipitation. Pages 711-724 in L. S.
Dochinger and T. A. Seliga, eds. Proc. First Int. Symp. on
Acid Precip. and the Forest Ecosystem. U.S. Dep. Agric.
Gen. Tech. Rep. NE-23.
Geologic factors controlling the sensitivity of
aquatic ecosystems to acidic precipitation. In Atmospheric
sulfur deposition: Environmental impact and health effects.
Ann Arbor Science Publishers, Ann Arbor, Mich.
Norton, S. A., D. W Hanson, and R. J. Campana. 1980. The
impact of acidic precipitation and heavy metals on soils in
relation to forest ecosystems: Proc. Effects of Air Pollutants
on Mediterranean and Temperate Forest Ecosystems.
Riverside, Calif.
Norton, S. A., C. T. Hess, and R. B. Davis, 1980. Rates of
accumulation of heavy metals in pre- and post-European
sediments in New England lakes. In S.J. Eisenreich, ed.
Inputs of atmospheric pollutants to natural waters. Ann
Arbor Science Publishers, Ann Arbor, Mich.
Oden, S. 1976. The acidity problem — an outline of concepts.
Pages 1-36 in L.S. Dochinger, and T. A. Seliga, eds. Proc.
First Int. Symp. on Acid Precip. and the Forest Ecosystem.
U.S. Dep. Agric. Gen. Tech. Rep. NE-23.
Rosenqvist, I. T., P. Jorgensen, and H. Rueslatten. 1980. The
importance of natural H+ production for acidity in soil and
water (abs.). Vol. II. Abs. of Volun. Contrib. to Int. Conf. on
the Ecol. Impact of Acid Precip. Sandefjord, Norway.
ACKNOWLEDGEMENTS
Research results reported herein were supported by the
following grants:
Norton and Galloway, U.S. Environmental Protection
Agency Contract EPA 79-D-X-0672 to Brookhaven National
Laboratory; Norton, U.S. National Science Foundation Grant
DEB-78-10641 to the University of Maine, and U.S.
Department of the Interior, Office of Water Research and
Technology Grant A-048-ME (Co-P.I. with R.B. Davis);
Galloway, U.S. Department of the Interior, Office of Water
Research and Technology Grant A-067-NY.
Schofield, C. L. 1973. The ecological significance of air-
pollution-induced changes in water quality of dilute-lake
districts in the northeast. Trans. N.E. Fish Wildl. Conf.,
May 14-17, 1972.
_. 1974. Acid precipitation: Effects on fish. Ambio
5:228.
1976. Dynamics and management of Adiron-
dack fish populations: Final Rep. Proj. Number F-28-R,
State of New York.
1977. Acid snow-melt effects on water quality
and fish survival in the Adirondack Mountains of New
York State. In U.S. Dep. Inter. Off. Water Res. Technol.
Complete Rep. Proj. No. A-072-NY.
Shapiro, J., W. T. Edmondson, and D. E. Allison. 1971.
Changes in the chemical composition of sediments of
Lake Washington, 1958-70. Limnol. Oceanogr. 16:437.
Smith, R. A. 1872. Air and rain: The beginnings of a
chemical climatology. Longmans, Green, and Co.,
London.
Spikkeland, I. 1977. Acidtrofe vann og dammer i bygland,
Aust-Agder. En undersetkelse av hydrografi og limnetiske
og litorale Crustace'-Samfum. Hovedfagsoppgave i
spesiell zoologi til matematisknaturvitenskapligem-
betseksamen ved Universitet i Oslo, IWstemesteret
1977.
TeBrake, W. H. 1975. Air pollution and fuel crises in pre-
industrial London, 1250-1650. Technol. Cult. 337.
Toonkel, L. E. 1980. Environ. Measurements Lab. Environ.
Q. Appendix (EML-374). U.S. Dep. Energy.
Watt, D. W., D. Scott, and S. Ray. 1979. Acidification and
other chemical changes in Halifax County lakes after 21
years. Limnol. Oceanogr. 24:1154.
Williams, J. S. 1980. The relative contribution of local and
regional atmospheric pollutants to lake sediments in
northern New England M.S. Thesis. University of Maine
at Orono.
-------
453
VARIATIONS IN THE DEGREE OF ACIDIFICATION
OF RIVER WATERS OBSERVED IN ATLANTIC CANADA
MARY E. THOMPSON
EDWARD B. BENNETT
National Water Research Institute
Burlington, Ontario, Canada
ABSTRACT
Freshwater bodies in large portions of eastern Canada are adversely affected by acidic
precipitation, with resultant damage to fish and other components of the aquatic ecosystem.
Concern exists regarding the future degree of acidification in these regions, in the face of
increasing emissions of sulfurous and nitrous compounds. This report deals with several
questions which arise from such concern: (1) What changes in river water chemistry can be
attributed to acid loading? (2) What year-to-year variations in acid loading/response have been
observed in river systems? and (3) What fraction of acidification is associated with sulfate
deposition?
INTRODUCTION
Freshwater bodies in large portions of eastern
Canada are adversely affected by acidic precipitation,
with resultant damage to fish and other components of
aquatic ecosystems and elevated concentrations of
heavy metals. This situation is caused by a combination
of circumstances, namely, relatively high rates of acid
loading from the atmosphere, and relatively low
buffering capacity of the receiving watersheds (Figure
1, from Thompson, et al. 1980). Concern exists
regarding the future degree of acidification in these
regions, in the face of increasing emissions of
sulfurous and nitrous compounds. This report deals
with several questions which arise from such concern:
1 . What changes in river water chemistry have been
observed in eastern Canada that can be attributed to
acid loading?,
2. What year-to-year variations in acid loading/
response have been observed to occur in river
systems?, and
3. What fraction of acidification is associated with
sulfate deposition?
"NORMAL" CHEMICAL WEATHERING
Chemical weathering in watersheds that receive
normal precipitation is predominantly due to the action
of carbonic acid. Carbonic acid is formed by solution of
atmospheric carbon dioxide in rain water or surface
water:
CO2 + H2O ~ HzCOa
Reactions of carbonic acid with carbonates and
silicates can be represented as
stf
HaCO3 + CaCOa - Ca++
2H2CO3 + Ca-silicate — Ca++
2HCOa
2-silicate
eq. 1
Figure 1. — Atmospheric deposition of hydrogen ion in 1977
(mg / m2 a) and soft water regions of eastern Canada.
These show that carbonic acid supplies protons
which are exchanged for cations (Ca++, Mg+t Na+, K+) in
the crystal lattice of the minerals comprising the soils
or rocks, causing the release of cations and bicarbonate
jons_ into solution. The rate at which cations are
removed from the watershed (cation denudation rate or
CDR) is a measure of the reactivity of the basin or of the
rate at which chemical weathering proceeds in the
watershed. The rate of production of bicarbonate is
correlated with the CDR; indeed, in basins where no
other anion (sulfate, for example) is a weathering
product, the two rates are equal. Where the rocks are
resistant, the bicarbonate concentration will be low, as
will concentrations of all other ions derived from
weathering. Many of the watersheds of eastern
Canada are in this class, for they are composed of
-------
454
RESTORATION OF LAKES AND INLAND WATERS
resistant granitic and siliceous bedrock from which
extensive glaciation has stripped away any younger,
calcareous deposits that may have existed.
Another consequence of the carbonic acid system is
a positive correlation between pH and alkalinity
(bicarbonate concentration) in runoff water. This is a
manifestation of the fact that for water in equilibrium
with the atmosphere, the product of the concentrations
(activities) of the hydrogen and bicarbonate ions tends
to be a constant:
[H+]
= constant
eq.2
It follows that pH, alkalinity, and CDR are positively
correlated; high values of each are characteristic of
hard water, with low values for soft water.
Chemical weathering by carbonic acid always
increases the alkalinity of runoff water above that of
the precipitation which falls on the basin (notwith-
standing the effects due to concentration by evapora-
tion in the watershed). Consistent with equation (2),
normal or "pure" rain has a pH of about 5.6.
Accordingly, the water discharge from a basin that
receives such rain must have a pH higher than 5.6.
WEATHERING BY ACID PRECIPITATION
The fact that normal rain has a pH of about 5.6 is a
consequence of the carbonic acid system. The pH of
acid rain is lower than 5.6 because of the presence of
strong acids such as sulfuric acid. In a watershed the
rain-borne strong acids serve as a ready source of
protons for exchange with cations in weathering
processes:
H2SO4 + CaCO3
H2SO4 + Ca-silicate
• Ca++ + SO4 + CO2 + \-\zO
•Ca++ + SO4 + Hs-silicate
eq.3
These proton-cation exchanges are the same as when
carbonic acid is involved (equation (1)), but bicarbonate
alkalinity is not a byproduct. Accordingly, weathering
by acidified precipitation tends to produce discharge
water of pH relatively low compared to that resulting
from the action of normal rain.
In general, the weathering action of acid rain should
be considered to be of both the carbonic acid and strong
acid types, especially since in some basins the free
protons of the strong acid will be supplied at a rate
insufficient to match the CDR. Therefore, in any
watershed, the degree to which pH is depressed in the
discharge water depends on the strength of strong acid
in the precipitation falling on the basin, and on the
resistance of the rocks. The aquatic regimes in basins
composed of easily-weathered minerals may show
little adverse effects from acid precipitation because of
continued dominance of bicarbonate alkalinity. How-
ever, hard-rock watersheds with low potential buffer-
ing capacities will show relatively large pH changes
corresponding to a given change in rain acidity.
The rate of arrival of strong acids in precipitation
must be considered with respect to the rate of chemical
weathering; thus, even a relatively reactive watershed
might be temporarily adversely impacted because the
rate of acid input momentarily overwhelms the rate of
weathering, e.g., melting of snowpacks.
WATER CHEMISTRY CHANGES
It follows that time histories of pH in soft water
drainage basins with little local anthropogenic in-
fluence can be interpreted in terms of trends in the
acidic strength of the rain. Moreover, a change in pH of
the discharge water should be accompanied by a
corresponding difference in the concentration of
sulfate ion, when the rain acidity is due primarily to
sulfuric acid.
The question of specific contribution of sulfuric acid
to acidification can be addressed by noting that the sum
of the concentration of bicarbonate ion and twice that
of the sulfate ion tends to be constant at any sampling
location in an aquatic system:
[HCOa] + [SOS] = constant (CDR)
eq. 4
This means that if sulfate increases in water because
of a rise in the sulfurous acidity of rain, then alkalinity
must decrease; and if the sulfate concentration falls,
alkalinity rises. This relationship is consistent with
earlier remarks regarding weathering by acidic precipi-
tation, is derived from equations (1) and (3), and is true
as long as the CDR is constant. In addition, equation (4)
is not valid if the sulfate concentration exceeds an
upper limiting value that is related to the CDR.
However, in general, the relationship is a powerful tool
for analyzing or predicting changes in water chemistry
that can be attributed to changes in sulfate loading. The
strong/weak acid relationship (4) can be combined
with the acidity/alkalinity relationship (2) to yield
[H+
[H]0 [HCO3]o
[HCOa] + 2[SO4]o-2[SOi]
eq. 5
Here the acidity is related to prior or initial values of the
acidity and alkalinity (subscripts), and to the change in
sulfate concentration.
Equation (5) is useful for calculating expected values
of the acidity corresponding to observed or potential
changes in sulfate concentration, as long as the sulfate
concentration does not exceed the limiting value given
by
[SO<]L = 1/2 [HCOalo + [SO4]o
eq. 6
If the rain-borne sulfate loading is sufficiently high,
then the sulfate concentration in the runoff water may
exceed [SO^L, which means that the rate of sulfate
discharge exceeds the CDT of the basin. In this case the
chemical weathering processes can be considered to
be entirely of the strong acid type, and the acidity of the
runoff water to depend primarily on the difference
between the sulfate loading rate and the CDR. It is
equally correct to consider that the acid precipitation
neutralizes the alkalinity that would have been
produced by only carbonic acid weathering, and that
the pH of the runoff water is a function of the excess
acidity. In any case, equations (4) and (5) would no! be
-------
THE ACID RAIN PROBLEM: MECHANISM AND EFFECTS
455
valid for such instances of strong acidity. A further
consequence of equations (2), (4), and (6) is that a
"pristine" pH can be calculated for those watersheds
where the present sulfate content is caused entirely by
atmospheric deposition, that is, where there is
essentially no sulfate deriving from the minerals. The
original bicarbonate concentration would be
[HC03]o,o = 2 [SO;],. = [HC03]0 + 2 [SO<]o
while the original pH would be
pK + log[HCO3]o.o
eq. 7
where K is the constant in equation (2). Again, such
calculations may be made with the provisos that the
CDR is constant over the years, and that the
precipitation acidity is essentially due only to sulfuric
acid.
It is not yet possible to quantify directly the influence
of the atmospheric deposition of nitrogen compounds
on acidification of water systems because ammonia
and nitrate are value nutrients in terrestrial and aquatic
regimes, and their chemistry is uncertain. As it
happens, however, most of the acidification observed
in eastern Canadian aquatic regimes can be attributed
to sulfate deposition.
EXAMPLES OF ACIDITY CHANGES
River water chemistry data are compared here to
demonstrate the influence of acidic precipitation in a
few watersheds in Nova Scotia and Newfoundland. In
each case, the parameters used for comparison are
discharge-weighted annual mean values derived from
individual samplings, usually at monthly intervals,
within each year. Most of the information was collected
by Canada's National Water Quality Monitoring
Program that began in 1961; a valuable contribution
was also made in 1954/1956 in the Atlantic Region by
a monitoring program of Water Survey of Canada.
Generally decreasing values of pH have been
observed in the rivers and lakes in the Atlantic
Provinces and reported by Thompson (1980), Thomp-
son, et al. (1980), and Watt, et al. (1978). Typical
histories exist for the Tusket and Medway Rivers that
are located in southern Nova Scotia (Figure 2); there
the mean pH decreased by 0.7 units in the time interval
1954/1955 through 1973 (Figure 3). Because the CDR
of each basin is essentially unchanged, the observed
increase in acidity of the discharge water must be due
to an increase in the rate of acid loading. Accordingly,
the pH decreases should be accompanied by increases
in the sulfate concentration in the runoff water from
each basin. Listed in Table 1 is information necessary
to assess pH-sulfate interdependency for the two Nova
Scotia rivers. Of immediate interest are the increases
in concentration of non-marine (or sea salt corrected)
sulfate between 1954/1955 and 1972, and the close
agreement between the pH values observed in 1972
and those calculated from equation (5). These figures
show that the observed acidification is primarily caused
by an increase in sulfate deposition. Closer inspection
Figure 2. — Location of the Tusket River (1) and Medway River
(2) in Nova Scotia.
Table 1. — Some significant values of acidity and sulfate
concentration in two Nova Scotia rivers.
1954/5
1972
pH
HCO3- mg/l
SO< mg/l
SO< mg/l
pH calculated*
pH observed
AH+ calculated*
At-T observed
[SO4]L mg/l
pHo
Tusket R.
5.14
1.84
2.32
3.36
4.60
4.48
0.76
3.77
6.01
Medway R.
5.67
2.65
1.71
2.88
5.31
5.12
0.65
3.79
6.12
* calculated from equation (5)
of the ratios of the calculated to observed changes in
hydrogen ion concentration shows that sulfurous
acidity accounts for at least two-thirds of the observed
effect. It follows that if sulfate deposition to a
watershed is increased, then the acidity of the runoff
water is increased.
Estimates of the limiting sulfate concentration
beyond which pH is a function primarily of excess acid
loading were calculated according to equation (6) and
included in Table 1. The value of 3.8 mg/l sulfate was
obtained for the two basins, implying essentially
identical geology. This limiting value was closely
approached in 1972 in the Tusket River, and was
equaled there in 1973. In both years, sulfate
concentration was about 0.5 mg/l higher there than in
the Medway River, suggesting that the rate of acidic
loading was lower in the Medway watershed than in
the Tusket. Such a difference in loading could be
expected because of the locations of the watersheds
with respect to the continental source regions (Figure
2).
-------
456
RESTORATION OF LAKES AND INLAND WATERS
pH
15
MEDWAY R.
10
11
1T291
12
1955
1965
YEAR
1975
Figures. — SummaryofpH observations in (a) Tusket River and
(b) Medway River. The range and mean values, and the number
of observations are given for each year.
PH
TUSKET R.
81T2
12
12
12
12
12
1955
1965
YEAR
1975
These temporal and spatial variations in river
chemistry are due to variations in acid deposition on
the watersheds, independent of the rate of emissions
at the sources. It is to be expected that natural
variability in climatic factors could cause significant
variability in atmospheric deposition rates in any basin,
especially in fringe areas. This point is illustrated by the
history of mean pH and sulfate in the Rocky River,
located in southeastern-most Newfoundland (Figure 4).
The observations show not only the general upward
trend of both acidity and sulfate concentration in the
river water, but also simultaneous occurrences of
relative maxima and minima of the two properties. Of
particular interest are first, the 1973 data, which show
record high sulfate and record low pH values, and
second, in subsequent years, a return to higher pH
levels in association with a decline in sulfate
concentration. Some of the year-to-year changes are
too large to be due to possible variations in emissions
at the sources, and must therefore indicate significant
differences in deposition caused by variations in
climatic factors.
The fact that pH in the Rocky River rises when the
sulfate loading decreases is an important point, for it
illustrates that the geochemical response to acid
loading in a basin is a dynamic one. Thus relationships
such as that given by equation (5) can be used to
measure or predict not only the course of acidification
due to increasing sulfate loading, but also that of
natural recovery which would be associated with
reduced sulfur dioxide emissions.
REFERENCES
Thompson, M. E. 1980. Acidic atmospheric precipitation:
evaluation of its impact on some Canadian surface waters.
Natl. Water Res. Inst., Burlington, Ontario. Unpubl. mss.
Thompson, M. E., et al. 1980. Evidence of acidification of
rivers of eastern Canada. Proc. Int. Conf. Ecol. Impact of
Acid Precipitation. Sandefjord, Norway.
Watt, W. D., D. Scott, and S. Ray. 1978. Acidification and other
chemical changes in Halifax County lakes after 21 years.
Fish. Mar. Sen/., Halifax, Nova Scotia, mss.
Figure 4. — Annual values of pH and sulfate in the Rocky River,
Newfoundland.
Included in Table 1 are the calculated values of the
pristine pH for the two basins, estimated from equation
(7). How accurate these figures are is not known,
because of the unknown influence of nitrogen
acidification. However, pH values of 6.0 and 6.1 serve
to underscore the fact that, in general, watersheds that
today are markedly impacted by acid loading are those
which had very little buffering capacity initially, and
therefore had a relatively low initial pH.
-------
457
RESPONSES OF FRESHWATER PLANTS AND
INVERTEBRATES TO ACIDIFICATION
GEORGE R. HENDREY
Department of Energy and Environment
Brookhaven National Laboratory
NORMAN D. VAN
Water Resources Branch
Ontario Ministry of the Environment
KAREN J. BAUMGARTNER
Department of Biological Sciences
Dartmouth College
Hanover, New Hampshire
ABSTRACT
The biota of acidic, oligotrophic, clear waters often are similar. The phytoplankton Dinophyceae,
and to a lesser extent Chrysophyceae tend to dominate. Production of 25 Shield lakes (pH 6.1 to
7.1) ranged from 25 to 240 mg Cm'2 d"'. Published values for acidic lake production are bracketed
by this range. Both biomass and production appear to be controlled by the availability of
phosphorus rather than pH per se. We found little evidence of possible C limitation in lakes
susceptible to acidification [H ] and biomass density in lakes do not appear to be directly related, as
illustrated by the whole-lake manipulations of [H ] and total phosphorus (TP). These studies,
however, do not examine effects of acidification on the whole lake-watershed system. It is
suggested that watershed acidification processes such as leaching of Al may reduce TP loading to
lakes, even below pH 5. Zooplankton community biomass appears to be reduced at low pH and
small-bodied forms may dominate. Among the zoobenthos, biomass does appear to be reduced in
some lakes but not others. Various studies found shredders, collectors, and scrapers to be reduced
more than raptorial species. We hypothesize that removal of fish predation on benthos allows a
relative increase in the invertebrate predators, reduction of herbivores (chironomids are relatively
abundant), and the subsequent increase in benthic algae observed in many waters.
INTRODUCTION
Over the past decade, numerous studies of the
biota of waters acidified by the deposition of strong
acids have described certain changes in communities
of fish, zooplankton, bottom fauna, phytoplankton,
benthic algae, rooted aquatic plants and mosses, and
benthic decomposers. We will point out where
diverse responses to acidification have been observed
and suggest ecological ramifications of certain of the
more common observations which may be useful in
interpreting observed changes. Other recent reviews
on the effects of acidification on biota are provided by
Aimer, et al. (1974, 1978), Dochinger and Seliga
(1976), Hendrey, et al. (1976), Leivestad, et al. (1976),
and Harvey (1980).
PHYTOPLANKTON COMMUNITY
COMPOSITION
Regional surveys of Scandinavian lakes of pH —4.0
to 7.0 have demonstrated that the numbers of
species in phytoplankton communities is reduced in
acid lakes, especially over a pH range of 6 to 5 (Aimer,
etal. 1978). Species from a II classes are lost a (though
data derived from regional surveys suggest that
proportionally more species of Chlorophyta disappear
(Aimer, et al. 1974). More intensive collections from
Adirondack lakes confirm this observation (Figure 1).
Biomass of phytoplankton communities of non-acidic
oligotrophic lakes are typically dominated by Chryso-
phyceae (Schindler and Holmgren, 1971) or Bacillari-
ophyceae (Duthie and Ostrofsky, 1974). This pattern
changes as lakes acidify (Figure 2). Community
structures of four acidic lakes near Sudbury, Ontario
were compared to those of 10 non-acidic lakes.
Dinophyceae, notably Peridinium inconspicuum, re-
placed chrysophytes and diatoms as community
dominants (Van, 1979). Late summer plankton of 60
Swedish lakes in the pH range 4.60 to 5.45 were
dominated by the Dinophyceae, especially Peridinium
inconspicuum and Gymnodinium cf. uberrium, while
Chrysophyceae dominated in spring. Some acidic
lakes, however, were dominated by Oocysf/s(Chloro-
phyceae) (Aimer, et al. 1978).
In some acidic lakes Chrysophyceae remain as
community dominants. Three Adirondack Mountain
lakes (Woods, pH 4.7 to 5.2; Sagamore, pH 5.0 to 6.4;
Panther, pH 5.3 to 2.8) (Hendrey, 1980) show the
pattern of reductions in species richness in acidified
lakes that is typically observed with losses of species
of Chrysophyceae (Figure 1). Chrysophyceae domi-
-------
458
RESTORATION OF LAKES AND INLAND WATERS
80
o
£
l/>
u_
o
40
5
^
Z
20
EJ DINOPHYCEAE
B CHLOROPHYTA
Q CHRYSOPHYCEAE
EH CRYPTOPHYCEAE
• BACILLARIOPHYCEAE
m CYANOPHYTA
Figure 1. — Total number of phytoplankton species
observed in each of three Adirondack Mountain Lakes
arranged by classes. (Samples collected biweekly during
the ice-free season, monthly during winter from three to
five in each lake).
100
50
Chrysophyceoe and Bacillariophyceae
E
o
m
•
«••
•g 4.0 5.0 6.0 7.0
o
I PH
>t
£ 100 f Dinophyceae
S.
O
1 50
0 <-t
4.0 5.0 6.0 7.0
pH
Figure 2. — Distribution of phytoplankton
biomass among Dinophyceae, Chryso-
phyceae and Bacillariophyceae in Sudbury
area lakes (before manipulation) and
Haliburton area lakes. (These are all
softwater oligotrophic lakes. Each point is
monthly-weighted, ice-free period mean of
biweekly collected, morphometrically
weighed euphotic zone composits.)
JIFIMIAIMIJIJIAISIOINIDIJ
J I F I M I A I M I J IJ IAISIOINIDIJ
JIFIMIAIMIJIJIAISIOINIDIJ
EH Dinophyceoe
IBS Chlorophyta
• Bacillariophyceae
S Cyonophyta
DID Cryptophyceoe
E3 Chrysophyceoe
Figure 3. — Seasonal variation in the proportion of total
phytoplankton biomass contributed by each taxonomic group
in three Adirondack Mountain Lakes (biomass derived by
microscopic measurements of cell volume).
nate the biomass of acidic Woods Lake, although
Dinophyceae (particularly P. inconspicuum) comprise
a significant fraction of the biomass in the ice-free
season (Figure 3). Perhaps less commonly, Cyano-
phyta may also be important in acidic lakes
(Watanabe, et al. 1973 Chlorophyta outweighing
losses of); Conroy, et al. 1976; Kwiatkowski and Roff,
1976). Thus, while it has been observed that
Dinophyceae are commonly the dominant phyto-
plankters in acidic lakes (Aimer, et al. 1978; Van,
1979; Hornstrom, 1979) many exceptions to this
pattern exist.
PHYTOPLANKTON PRODUCTIVITY
AND BIOMASS
Although many studies show freshwater acidifica-
tion to be associated with major changes in
community structure, only a few have examined the
effects on phytoplankton productivity. In 26 Canadian
Shield lakes in the pH range 6.1 to 7.1, maximum
volumetric productivity ranged from 1.7 to 6.9 mg C
m"3 hr"1and areal production from 25 to 240 mg C
m~2 d~? These data (Harvey, 1980) bracket the
published productivity values for acidified lakes,
although such data are quite limited (Dillon, et al.
1979; Hornstrom, 1979; Hendrey, 1980).
-------
THE ACID RAIN PROBLEM: MECHANISM AND EFFECTS
459
Carbon Limitation
At equilibrium with atmospheric CO2(Pco2=1°35
atm.) inorganic carbon concentrations in lakes of pH
<5.0 will be under 15 /urn . Phytoplankton production
therefore may be reduced, i.e., limited by available
carbon unless carbon concentrations are maintained
near saturation by atmospheric invasion of COaor by
the respiratory activities of the biota. If, on the other
hand, these processes provide dissolved inorganic
carbon to the phytoplankton of acidic lakes suf-
ficiently rapidly, then the greater clarity of the lakes
may facilitate aerial production rates that are as high
or higher than in non-acidic lakes of comparable
nutrient status. No evidence has been presented, so
far, that lack of carbon might limit phytoplankton
production in oligotrophic acidic lakes. In Woods Lake,
with DIG values in the range 0.3 to 0.6 mg C I"1 (pH
4.7 5.1), for example, the maximum volumetric
hourly productivity rates (measured biweekly at five
depths through the ice free season in 1979) never
exceeded 2 percent of the available carbon (DIG
measured by gas chromatography, Stainton, 1973),
and the maximum daily production per square meter
never exceeded 6.4 percent of the DIG available at 10
a.m. as shown in Figure 4 (Hendrey, unpubl. data).
The productivity of Woods Lake is at the upper end
of the range of productivity observed in many
oligotrophic Canadian Shield lakes (Figure 5). Experi-
mental acidification of ELA Lake 223 reduced DIG
from around 1.2 mg. C I"1 to the range 0.06 to 0.6 mg
C f1 yet phytoplankton photosynthesis was not
carbon limited. Short-term bioassays with carbon
enrichment did not increase phytoplankton produc-
tivity (Schindler, et al. 1980).
With an average daily production rate (which
assumes no carbon limitation) of 13.8 m moles C m12
(166 mg C m"2), and a mean depth of 8.4 m,
phytoplankton production in Clearwater Lake (Dillon,
et al. 1979) would consume 1.64 //moles DIG I"1
day"1. This represents about 14 percent of the DIG
available if carbon concentrations in the lake are at
saturation at a lake pH of 4.2. Schindler and Fee
(1973) found that in northwestern Ontario lakes in
which production was not carbon limited (fertilized
lakes), daily production could reduce DIG by 1 to 7
percent. In lakes that were carbon limited, over 25
percent of the carbon available at dawn was
consumed during the day. Based on this analysis,
Clearwater Lake occupies an intermediate position
suggesting that on cloudless days phytoplankton
production may be carbon limited.
Phosphorus Limitation
Several studies show that the biomass of phyto-
plankton is correlated with the supply of phosphorus
in both acidic and non-acidic lakes (Schindler, 1971;
Schindler and Fee, 1974; Nicholls and Dillon, 1978;
Van, 1979; Hornstrom, 1979). Studies in which the
chemistry of whole lakes has been manipulated
especially demonstrate the control exerted by phos-
phorus on phytoplankton biomass. In the fall of 1973
adding base raised the pH of acidic Middle Lake (pH
4.4) to pH ca. 7.0. Total P (TP) levels did not increase;
in consequence there was no increase in phyto-
plankton biomass (Table 1), although species com-
position shifted from Dinophyceae to Chrysophyceae.
TP was experimentally increased in Middle Lake from
1975 to 1977 and a large increase in biomass
occurred. TP was also increased in acidic Mountain-
top Lake without elevating lake pH and phytoplankton
biomass increased. Hydrogen ion concentration also
increased by 20 percent after 1 year and by 75
percent after 2 years because of bicarbonate
generation from SOa reduction in the hypolimnion
(Table 1). Comparing biomass in Mountaintop Lake to
non-acidified Labelle Lake to which phosphorus was
also added, shows the strong dependence of biomass
in Labelle on TP and not on pH (Figure 6). Schindler, et
al. (1980) slowly acidified Lake 223 in the Experi-
mental Lakes Area (ELA) from pH 6.7 to 7.0 in 1976 to
pH 5.7 to 5.9 in 1978 without any apparent change in
either productivity or biomass of phytoplankton, or in
TP or dissolved P.
These studies indicate that pH change alone does
not alter phytoplankton production (to pH>5.7) or
biomass (pH>4.4) and that biomass is regulated by
the supply of P. However, acid precipitation does alter
not only the lake water pH but also watershed
processes, particularly the weathering of aluminum
from rock and soil, and consequently, possibly the
watershed chemistry of nutrients such as phos-
phorus. Whole-lake manipulations which treat only
the lake water have proved very useful in elucidating
portions of the lake acidification story. But an
ecosystem approach must still be used to interpret the
complex phenomena associated with lake-watershed
acidification and consequent biological effects.
Table 1. — Total phosphorus (TP), pH and biomass of
phytoplankton observed in two Sudbury Experimental Study
lakes. Data are ice-free period, monthly weighted means from
Dillon, et al. (1979) and Van (unpubl. data).
Lake
Middle
Mountaintop
PH
1973 4.4
1974 7.0
1975-77 6.5
1976 4.4
1977 4.5
1978 5.0
TP phyto-biomass chlora
Ougf) (mgf1) (^gf1)
7.3
7.1
11.6
43.0
58.0
75.0
0.46
0.16
0.68
0.72
2.05
6.35
.0.91
0.92
2.70
5.7
20.1
64.8
Aluminum and Phosphorus
In studies of 58 oligotrophic lakes in the Swedish
west coast region the lowest phytoplankton biomass
was found in eight lakes in the pH range 5.1 to 5.6
while biomass was higher on either side of this pH
range. This evidence has been used to support the
view that nutrient availability is lowest in this
intermediate pH range because of Al complexing of P
(Aimer, et al. 1978).
The solubility of apatite minerals increases in acidic
environments, but as soil studies have demonstrated
(Brady, 1974), this need not result in increased
bioavailability of phosphorus. Watershed acidification
elevates concentrations of aluminum in runoff waters
-------
460
RESTORATION OF LAKES AND INLAND WATERS
75
60
- WOODS
45
o
0>
\
o
D>
E
30
I 5
5/2 5/29 6/26 7/24 8/21 10/2
5/16 6/12 7/10 8/9 9/4 10/12
DATE 1979
Figure 4. — Ratio of phytoplankton production (mg I rrfJ
d"1) to available carbon (g DIG rrf2) in the euphotic zone of
Woods Lake (pH 4.7 tp 5.1). Production determed by 14C
tracer uptake, at five depths in situ.)
300
200
o
o>
E
I 00
W
6
PH
Figure 5. — Maximum observed daily production (mg c m
d~') of phytoplankton in 14 lakes of the Canadian Shield (•)
and three Adirondack Mountain lakes (W = Woods, S =
Sagamore, P = Panther). Observations were made at least
eight times in each lake.
(Cronan and Schofield, 1979). In 37 clearwater
Swedish lakes Al concentrations of 0.01 to 0.6 mg I'1
occurred with lake water pH 5.5 and were usually
less than 0.1 mg T1 at pH 5.5 and above. Dickson
(1978) has shown that the removal of phosphorus
from lake waters containing additions of 0.5 or 1.0 mg
Al T1 is highly dependent on pH, with the maximum
removal of P occurring at pH 5 to 6. This does not
necessarily mean that waters at pH<5 will contain
more soluble P than less acidic water.
The chemistry of Al and its interaction with
phosphorus in natural waters is rather complicated
and a full discussion is beyond the scope of this
review, but several relevant points are worth noting.
1. Driscoll (1980) has found solubility of free Al to
be controlled byAI(OH)3, but Al-organic complexes
were the dominant form of monomeric Al in surface
waters.
20 40 60 80
°
I
Figure 6. — Phytoplankton biomass in
Mountaintop (o) and Labelled*) lakes versus
total phosphorus concentration (TP) and pH.
(Average for the ice-free seasons of 1976-
1978 in Mountaintop and 1977-78 in
Labelle).
2. Dickson (1978) has shown that Al bound to
humic substances is unavailable to complex with
inorganic P.
3. The minimum solubility of AlPOo is about 10 fjg
f1 at pH 6 (see Figure 10-1 in Stumm and Morgan,
1970).
4. Total P concentrations in acidified watersheds
have more Al to interact in the softwater, oligotrophic
lakes of Scandinavia, the Canadian Shield, and the
Adirondack Mountains are typically less than 10//g
I"1 (Conroy, et al. 1974) and the fraction of the TP
which may be available to phytoplankton is likely to be
much lower.
For these reasons it is unlikely that the formation of
AlPO-t is an important mechanism in regulating P
availability in acidified lakes. On the other hand, it
seems likely that aluminum hydroxides (or ferric
hydroxide) which have much lower minimum solu-
bilities could remove inorganic P by flocculation, but
this mechanism, with respect to acid lakes, has
received little attention (Driscoll, pers. comm.).
While Dickson's data (obtained at rather high P
concentrations of 50 and 100 //g I"1) indicate less P is
removed at pH<5 than at pH 5 to 6 for a lake water
with 0.5 mg Al I"', it is also shown that Al
concentrations increase with watershed and lake
acidification. Lakes with lower pH in acidified water-
sheds have more Al to interact in the removal of P, not
only in the lake itself but also in the watershed soils.
Biomass
Some uncertainty still exists concerning the effects
of lake acidification on phytoplankton biomass.
Biomass is clearly controlled by the supply of TP but
-------
THE ACID RAIN PROBLEM: MECHANISM AND EFFECTS
461
whether acidification affects the rate of supply is
unknown. We have compared biomass of five lakes of
pH<5.1 with 21 lakes of pH>5.6 (Harvey, 1980) and
found them not significantly different. Similarly,
chlorophyll concentration was not found to be related
to pH in 37 oligotrophic Norwegian lakes with
different pH values. No phosphorus data were
presented for these lakes (Raddum, et al. 1980).
The variability in physical and chemical features
among lakes located in the same general area can be
rather large and confounds attempts to compare lakes
statistically. For example, the Canadian Shield lakes
listed by Harvey (1980) show a difference in TP
concentrations among non-acidic lakes of from 2 to
16/aeq I'1
If phytoplankton biomass were reduced by acidifi-
cation then one would anticipate increased water
transparency. Increased lake clarity has occurred in a
few lakes concomitant with increasing acidity
(Schofield, 1973; Aimer, etal. 1978) and many acidic
clearwater lakes are very transparent. Aimer, et al.
(1978) note that humic substances are readily
precipitated in the presence of Al in the pH range 4.0
to 5.0. It is their view that humic substances are
bound either in the forest soils, thus reducing their
input to the lake, or they are precipitated to the lake
sediments. Kwiatkowsky and Roff (1976), in contrast
to Raddum, et al. (1980), found water transparency
and chlorophyll concentrations to inversely correlate
with pH in six lakes (pH range 4.05 to 7.15) 51
kilometers south of Sudbury. The lower chlorophyll
levels in the acidic lakes most probably correlate with
lower concentrations of TP. The lower TP levels in the
acidic lakes could indicate the inherently low rates of
export of phosphorus from watershed soils of low
buffering capacity; potentially, acidification of water-
shed soils may reduce these export rates.
BENTHIC ALGAE
In southern Norway, where precipitation is strongly
acidic and many lakes and streams have been
acidified, dense growths of filamentous algae and
mosses have been observed in streams and, in some
cases, lakes (Hendrey, et al. 1976; Hendrey and
Vertucci, 1980; Lazarek, 1980).
Acidification of artificial stream channels using
water diverted from Ramse Brook in Tovdal, southern
Norway (Hendrey, 1976), resulted in heavy accumula-
tion of periphytic algae, especially Mougeotia spp. and
Binuclearia tatrana (chlorophyceae), Tabellaria floccu
osa and Eunotia lunaris (Bacillariophyceae). In
another study, Norris Brook in the Hubbard Brook
watershed (New Hampshire) was acidified experi-
mentally. This also significantly increased periphytic
algae (Hall, et al. 1980). In both of these stream
experiments the productivity per unit of biomass
(chlorophyll a) decreased with acidification.
In an experimental acidification of Lake 223 waters
in 10-meter diameter polyethelene tubes, Muller
(1980) used HaSCU to obtain treatment levels of
approximately pH 6, 5, and 4, in addition to a control
tube (no addition of acid) at pH 6.5. No trend was seen
in biomass or productivity relative to pH but
community changes did occur. The Chlorophyta,
particularly Mougeotia spp., dominated at low pH.
Diatoms, (Achnanthes minutissima) and Mougeotia
spp. dominated at higher pH. Both community
diversity and similarity decreased with pH. Both
Hendrey (1976) and Muller (1980) observed carbon
uptake by periphyton which had been removed from
their substrates and incubated in vitro. In the artificial
stream channels the rate of photosynthesis, repli-
cated in three separate experiments, increased with
decreasing pH due to the larger biomass at lower pH,
but the photosynthesis per unit biomass (P/B)
decreased with pH (Hendrey, 1976). While there were
obvious differences in the tolerance of the algal
species to low pH, and this must certainly enter into
community composition (Hendrey, 1976), it cannot
explain the increased biomass at low pH. For
example, Tabellaria flocculosa which dominated
acidified stream communities in three of five
replications of the Tovdal experiment, has been found
to have a pH optimum between 5.0 and 5.3 (Cholonky,
1968) or higher (Kallqvist, et al. 1975). The niche
breadth with respect to pH for this species may be
wider than that of others but its optimum is not at pH
4. Therefore, some explanation other than tolerance
(Muller, 1980) must be found for observed increases in
algal abundance in streams of low pH. We will discuss
this later in the paper.
MACROPHYTES
Evidence of changes in macrophytic- community
structure following acidification has been obtained for
a few lakes. Lake Golden, N.Y., which has been
acidified by ca. 97 micro equivalents per liter and now
has a pH near 4.9, was surveyed in 1932 and again in
1979 (Hendrey and Vertucci, 1980). Very dense
stands of benthic macrophytes are frequently ob-
served in acidic clearwater lakes. Sphagnum abun-
dance was found to increase with decreasing pH in
five Swedish lakes, and to have greatly expanded in
acidic Lake Orvattnet (pH ca. 4.6) (Grahn, 1976).
Sphagnum replaced more common species such as
Lobelia, Litorella, and Isoetes. Similar density of
Sphagnum pylaesii was observed in Lake Golden with
about 300 g dry weight m~2. Although Sphagnum is a
normal component of the submerged flora of many
oligotrophic softwater lakes, the extent of these
stands in the Swedish lakes and Lake Golden appears
to be exceptional even for acidic lakes of this type.
Grahn (1976) notes that it has a high ion exchange
capacity when both alive and dead (peat), and that the
dense mat inhibits cycling of materials between the
mineralized sediment and overlying water. Thus,
Sphagnum may contribute to reduced plant nutrient
availability in lakes where it is abundant. Utricularia
also forms very dense stands covering large areas in
Lake Golden and in Woods Lake (pH ca. 4.9), and may
contribute to reducing exchange with sediments.
Macrophyte biomass increased in acidic Clearwater
Lake, Ontario (Table 2). The total biomass of primary
producers thus may be increased by acidification.
Because of the increased clarity of acidified lakes,
sediments at great depths may be available for
colonization. Data on the productivity of the benthic
zone in acidified lakes are not available.
-------
462
RESTORATION OF LUKES AND INLAND WATERS
Table 2 — Macrophyte biomass (dryweight), lake pH and macrophyte. phytoplankton biomass ratios for three Canadian Shield
Lakes in 1978 (unpubl. data provided by I. Wilde and G. Miller, Ontario Minist. Environ.)
Biomass Ratios .
Harp
Red Chalk
Clearwater
Lake pH
6.8
6.6
4.4
Macrophyte
Biomass (g m 2
of vegetated zone)
66
57
240
Macrophyte
Phytoplankton
6.1
4.9
73.0
Macrophyte Shoots
Phytoplankton
2.2
2.1
19.7
.
Tables. — Comparison of standing stocks of bent hie invertebrates from ohgotrophic lakes of different pH's in Northern Ontario.
Lake
Ao(ha)
PH
Abundance of Benthos
numbers m"2
Source
Middle
Hannah
Lohi
Clearwater
Lumsden
George
Nelson
Lake Type D.
28
27
41
77
22
148
309
12-44
15
9
20
22
23
30
50
13
4.4
4.3
4.4
4.3
4.6
5.2
5.7
>5.7
650
1,200
1,100
1,000-
6,000-
3,600 -
1,100-
500-
4000
17,000
18,800
2400
1600
Scheider,
11
"
Beamish,
Scheider,
Hamilton,
et al.
1974
et al.
1971
1975, 1976a
1976b
ZOOPLANKTON
From regional surveys conducted in Canada and in
Scandinavia it may be concluded that acidification of
lakes is accompanied by a two to threefold reduction
in richness of species of crustacean zooplankton
(Aimer, et al. 1974; Hendrey and Wright, 1976;
Sprules, 1975). Although data are less abundant it
appears that rotifer community diversites are also
reduced (Aimer, et al. 1974; Roff and Kwiatkowski,
1977); some rotifer species, however, are excep-
tionally tolerant of acidic environments (Smith and
Frey, 1971).
Changes in community structure are most notice-
able at pH<5.0 In Ontario, two of the species most
commonly observed and numerically dominant in
non-acidic lakes, Diaptomus minutus and Bosmina
longirostris, become even more important in acidic
lakes as other less tolerant species such as Daphnia
decline (Sprules, 1975). As fish are usually absent at
these levels of pH, the dominance of the zooplankton
community by small-bodied herbivores contradicts
the observation that in the absence of fish predation,
the dominant zooplankton will be large-bodied
(Walters and Vincent, 1973; Dodson, 1974; Brooks
and Dodson, 1965). This may result simply from
differences in tolerance among species to depressed
pH. Physiological bases for such differences have
been demonstrated (Potts and Fryer, 1979). Domi-
nance by small-bodied herbivores also may be
attributed to low levels of invertebrate predation in
the absence of fish predation (Lynch, 1979). Van and
Strus (1980) and De Costa and Janicki (1978)
showed, for example, that dominance by Bosmina
longirostris in acidic lakes was most evident only after
the dominant cyclopoid predator had declined in
density. The domination by small-bodied herbivores
may also indicate a reduced availability of food
(Goulden, et al. 1978). In some acidic lakes
concentrations of dissolved organic matter and hence
of bacteria may be low. A large fraction of the
phytoplankton is comprised of dinoflagellates, which
are not preferred prey for filter feeding herbivores
(Porter, 1973). These three hypotheses for changes in
structure of zooplankton communities warrant in-
vestigation.
Few studies of zooplankton are sufficiently in-
tensive to assess whether acidification reduces
zooplankton standing stocks. Apparently, there are
very large year to year variations. Fo example, Van
and Strus (1980) showed that the biomass averaged
over the ice-free season in Clearwater Lake, an acid
metal-contaminated lake near Sudbury, Ontario,
could vary up to 300 percent from year to year. What
data do exist, however, suggest that acidification
reduces zooplankton community biomass because
both size and numbers of community dominants
decline (Harvey, 1980). A consequence of this
reduced biomass is reduced efficiency of energy
transfer from primary to secondary trophic levelsfYan
and Strus, 1980). Such a phenomenon has previously
been suggested to occur in lakes acidified by mine
drainage (Smith and Frey, 1971) and in acidified
streams (Hendrey, 1976; Hall, et al. 1980).
BENTHIC INVERTEBRATES
Synoptic and intensive studies of lakes and streams
have demonstrated that numbers of species of
benthic invertebrates are reduced along a gradient of
decreasing pH (Sutcliffe and Carrick, 1973; Conroy, et
al. 1976; Hendrey and Wright, 1976; Borgstrom, et al.
1976; Aimer, et al. 1978). The more commonly
observed invertebrates in acidified waters belong to
the Notonectidae (backswimmers), Corixidae (water-
boatmen), Gerridae (waterstriders), Chironomidae
(midges), and Megaloptera (alderfly). These may be
abundant, especially in waters where fish predation
-------
THE ACID RAIN PROBLEM: MECHANISM AND EFFECTS
463
has been eliminated. Trichoptera, Ephemeroptera,
and Plecoptera have many species which are
intolerant of low pH.
In laboratory studies, Bell (1971), Bell and Nebeker
(1969), and Raddum (1978) have measured the
tolerance of some invertebrates to depressed pH.
Tolerance seems to be in the order caddisflies >
stoneflies > mayflies. Raddum concluded that while
many stonefly species did not seem to be affected by
low pH, Amphinemura sulcicollis, Brachyptera risi,
and Leuctra hippopus exhibited increased death rates
and decreased caloric content when exposed to acidic
water. The mayfly Baetis rhodani, the dominant
mayfly in non-acidic Norwegian streams, did not
survjve experimental acidification. Roff and Kwiat-
kowski (1977) concluded that the diversity of benthic
fauna from the La Cloche Mountain lakes near
Sudbury, Ontario was reduced in two lakes of pH>5.0
While data on species occurrence are scanty,
quantitative data on biomass or abundance of benthic
invertebrates in acidic lakes are even less available.
Raddum (1978) studied six acidic and eight less acidic
lakes in Norway, all with similar substrate. He found
that densities and biomasses of benthic invertebrates
were reduced in the acidic lakes, but the most
common animal group was the chironomids. A
summary of available but scanty information from
acidic and non-acidic shield lakes in Ontario
suggests, in contrast, that abundance may not be
reduced by depressed pH (Table 3).
Okland (1969) surveyed snail population in 832
lakes in Norway and found no snails in lakes of pH<
5.2. No comparable North American data are
available. Following the observations of K. A. Okland
(1969), Borgstrom and Hendrey (1976) found that the
amphipod Gammarus lacustris adults and the tadpole
shrimp Lepidurus arcticies could not tolerate pH<
6.0. Okland (1980) indicates that the isopod Asellus
aquaticus may be restricted to lakes of pH>5 in
Norway. —
While these invertebrates are restricted to some
extent by acidification, it appears that air-breathing
aquatic insects, especially predators, are very tolerant
of acidic environments. Population densities of
Coleoptera, Corixidae, and Megaloptera increased in
acidic lakes, and in the most acid lakes, Odonata
species were more abundant (Raddum, 1978, 1980).
No studies on changes in populations of these larger
invertebrates are yet available from North America.
Following experimental acidification of Norris Brook
(Hall, et al. 1980) to pH 4 in March 1977 the
downstream drift of insect larvae increased 13-fold.
Organisms in the collector and scraper functional
groups were affected more than predators. There was
also a 37 percent reduction in insect emergence with
members of the collector group most affected.
Invertebrates taken out of the bottom samples were
reduced by 75 percent compared to the central zone,
while the reduction of invertebrates in debris
accumulations was 84 percent.
Low pH also appeared to prevent permanent
colonization by a number of invertebrate species,
primarily herbivores, of the acidified reaches of River
Dudden, England (Sutcliffe and Carrick, 1973).
Ephemeroptera, Trichoptera, Ancylus (Gastropoda),
and Gammarus were absent. Observations that
herbivorous invertebrates are especially reduced in
acidified streams, as reported in Norris Brook and
River Dudden, support our earlier discussion (Hen-
drey, 1976; Hall, etal. 1980) that this may contribute
to increased algal accumulations seen in Norwegian
streams, the artificial acidification at Tovdal and
Norris Brook, and accumulations of benthic algae in
acidic lakes.
Petersen (1980) investigated the processing of
coarse particulate organic material in leaf packs in
streams at different acidities. The "shredder" func-
tional group is apparently reduced in the acidic
stream. Traaen (pers. comm.; Leivestad, et al. 1976)
conducted similar experiments with litter bags in
Norwegian waters with differing pH. Invertebrates
appeared to make a greater contribution to accelerate
decomposition in less acidic waters. These studies
and those in Norris Brook and River Dudden indicate
that shredders, collectors, and scrapers are reduced
to a greater extent than are the predatory inverte-
brates.
Water Hardness
The great importance of water hardness in
regulating distributions of invertebrate species and
their ranges of tolerance to acidity has been
demonstrated, most recently by studies in Norway. K.
A. Okland and Kuiper (1980) found the number of
species of Sphaeriidae (small mussels) increases
rapidly with increasing Ca concentration, up to 2 mg
f1 hardness (as CaO) (this includes over half of the
1,320 Norwegian freshwaters included in the report);
the number of species also increases with pH over the
pH range 4.6 to 6.9. Half of the 20 Sphaeriidae
species were classified with respect to both pH and
hardness requirements. Gastropods may be present
at water hardness values as low as1.5 mg f1 but only
if pH is 6. At greater hardness values (>6 mg f1)
gastropods were found in Norwegian lakes at pH as
low as 5.2 (J. Okland, 1980).
The fact that some species are more abundant in
acidic softwater lakes does not necessarily imply they
prefer such conditions, but may be due to the
elimination of competitors who are intolerant of such
conditions. This may be the case, for example, with
Asellus aquaticus, as discussed by K. A. Okland
(1980). This species is found in Norwegian waters
with hardness ranging from 2.3 to 208 mg I"1 (as CaO)
and pH 4.8 to 8.8; it decreases in frequency in lakes
with hardness >20 mg I"1, and is prominent in acidic
Swedish lakes (Aimer, et al. 1974). Gammarus
lacustris, which is intolerant of low pH (Okland, 1969;
Borgstrom and Hendrey, 1976), has a distribution
limited to pH 6.0 and hardness > 4.5 mg I" (K. A.
Okland, 1980). K. A. Okland notes these two species
are ecologically very similar and have a nearly
allopatric distribution and, under favorable condi-
tions, Gammarus apparently replaces Asellus by
competitive exclusion, thus leaving the low pH,
softwaters to Asellus.
Borgstrom and Hendrey (1976) found low pH
inhibited moulting progression of Lepidurus arcticus
and suggested this might be due to interference of
-------
464
RESTORATION OF LAKES AND INLAND WATERS
calcium uptake (see also Sutcliffe and Carrick, 1973).
Malley (1980) studied the crayfish Orconectes virilis
(Hagen) taken from lake 223 and placed in aquaria
maintained at pH 3.0, 4.0, 5.0, and 6.0. Mortality (ca.
14 percent) at pH 3 and 4 occurred during molting.
Progression through molt stages was slowed by pH
5.0 and uptake of Ca++ was greatly retarded. Some
species of crayfish do occur in rather acid waters.
Cambarus bartoni, for example, is common in
Clearwater Lake (pH 4.2).
DISCUSSION
Each species has a unique set of environmental
factors at which its growth is optimized. For algae,
nutrient (P, N, C, S, Fe, etc.) availability, light intensity,
and temperature are the primary factors that must be
optimized. Secondary factors, however, including
grazing and microbial activity (which lead to nutrient
regeneration and increased light penetration), and
low concentrations of toxic substances such as heavy
metals, are also important in determining species
optima. Lake and watershed acidification alters all of
these factors and increases the H+concentration by
orders of magnitude.
Given these complexities it is not surprising that the
differences in the structure of some biotic communities
(e.g., phytoplankton) among acidic lakes are not yet
explicable. They may in fact never be explicable.
Schindler (1975) commented, for example, that despite
the very precise knowledge concerning nutrient input
rates, lake physics, and chemistry available for ELA
lakes in northwestern Ontario, the ecosystems of the
lakes were too complex to model responses to nutrient
additions incorporating taxonomic detail.
Complex interactions among various trophic levels
might contribute to some of the phenomena we have
discussed for acid lakes. One of the most obvious
features of lake acidification is that all fish have been
eliminated from many such lakes. The effects of fish
removal from non-acidic, oligotrophic Emmaline Lake,
Colo., were studied by Walters and Vincent (1973).
The zooplankton community was markedly altered,
with a shift to dominance by large species, especially
Daphnia middendorffiana', midge larvae populations
increased and benthic invertebrates became domi-
nated by large forms. In contrast, evidence from the
Canadian Shield lakes indicates that smaller forms
predominate in acidic lakes.
Periphyton standing crops increased greatly in the
2 years following fish removal from Emmaline Lake;
this may have contributed to a "bloom" of small
herbivorous midges. Periphyton was greatly reduced
in the following 2 years by grazing. Henrikson, et al.
(1980) note that the disappearance of fish from
acidified lakes leads to a rapid increase in abundance
of several species susceptible to fish predation.
Vertebrate predators are replaced by invertebrate
predators, notably Corixidae, Chaoborus, Dyticisdae,
and Odonata.
Studies which link various observations, e.g., species
reduction, to acidification on the basis of correlation
analyses, have been criticized by Nilssen (1980) as
ignoring recent analytical advances, particularly in
evolutionary biology. Nilssen states that "In research
on acidification the various investigated parameters
are frequently plotted against ambient pH, often based
on one sampling date only." The implication seems to
be that some investigators have tried to correlate a
parameter against one pH measurement. This, of
course, is not so. What has been frequently done in
regional studies is to make correlations between a
parameter and pH values collected once from each of
many lakes. This technique has been validly used for
both biological and chemical analyses in the Swedish
and Norwegian synoptic surveys. Nilssen is correct in
pointing out that correlation, e.g., decreased species
number with decreased pH, does not prove direct
cause. Decreased species number may prove to relate
more directly to increased Al concentrations and/or
removal of predators than to the concentration of FT
The driving variable in the changes we have discussed
in this review is increased H+ loading to lakes,
nonetheless.
Changes in functional guilds of organisms will
undoubtedly have affects at other trophic levels. We
present the following hypothetical linking of biological
observations in acidic lakes. The removal of fish
predation increases predatory invertebrate abundance.
This results in increased invertebrate predation on
collectors, shredders, and scrapers and reduces the
activities of these guilds. Accumulation of attached
algae and benthic litter would thus be enhanced by
these changes. Benthic plants, especially algae and
litter, are in fact abundant in many acidic lakes and
streams. This hypothetical sequence is composed of
elements in a chain that have not yet been conclusively
linked. Quantitative studies of transfers of mass or
energy between trophic levels in acidified lakes are
lacking. This is the type of research which could now be
most fruitful to provide an integrated understanding of
acidification impacts on aquatic flora and fauna.
REFERENCES
Aimer, B., et al. 1974. Effects of acidification on Swedish
lakes. Ambio 3:30.
1978. Sulfur pollution and the aquatic ecosystem.
Pages 271-311 in J. 0. Nriagu, ed., Sulfur in the
environment, Part II: Ecological impacts. John Wiley, New
York.
Bell, H. L. 1971. Effect of low pH on the survival and
emergence of aquatic insects. Water Res. 5:313.
Bell, H. L., and A. V. Nebeker. 1969. Preliminary studies on
the tolerance of aquatic insects to low pH. Jour. Kans.
Entomol. Soc. 42:230.
Borgstrom, Ft., J. Brittain, and A. Lillehammer. 1976.
Evertebrater og surt vann: oversikt over innsamlingsloka-
liteter. (Invertebrates and acidic waters: survey of sampling
sites.) SNSF IR 21/76. SNSF Proj. 1432 Aas-NLH, Norway
(in Norwegian).
Borgstrom, R., and G. R. Hendrey. 1976. pH tolerance of the
first larval stages of Lepidurus arcticus (Pallas) and mature
Gammarus lacustris (G. 0. Sars). IR 22/76. SNSF Proj.
1432 Aas-NHL, Norway.
Brady, N. C. 1974. The nature and properties of soils. 8th ed.
MacMillan Publishing Co. Ltd. New York.
Brooks, J. L., and S. I. Dodson. 1965. Predation, bodysizeand
composition of plankton. Science 150:28.
-------
THE ACID RAIN PROBLEM: MECHANISM AND EFFECTS
465
Cholonky, B. J. 1968. Die Okologie der diatomeen in
Binnengewasser. Cramer, Weinheim. (In German.)
Conroy, N., D. S. Jeffries, and J. R. Kramer. 1974. Acid shield
lakes in the Sudbury Ontario region. Proc. 9th Can Svmo
Water Pollut. Res. Canada.
Conroy, N., et al. 1976. Influences of the atmosphere on lakes
in the Sudbury area. Jour. Great Lakes Res. 2(suppl. 1 ):146.
Cronan, C., and C. L. Schpfield. 1979. Aluminum leaching
response to acid precipitation: effects on high-elevation
watersheds in the Northeast. Science 204:304.
DeCosta, J., and A. Janicki. 1978. Population dynamics and
age structure of Bosmina longirostris in an acid water
impoundment. Verh. Int. Verein. Limnol. 20:2479.
Dickson, W. 1978. Some effects of acidification of Swedish
lakes. Verh. Int. Verein. Limnol. 20:851.
Dillon, P. S., et al. 1979. Acidic lakes in Ontario, Canada:
characterization, extent, and response to base and nutrient
additions. Arch. Hydrobiol. Beih. Ergebn. Limnol. 13:317.
Dodson, S. I. 1974. Zooplankton competition and predation:
an experimental test of the size-efficiency hypothesis
Ecology 55:605.
Driscoll, C. T. 1980. Aqueous speciation of aluminum in the
Adirondack region of New York State, USA (abstract). Int.
Conf. Ecol. Impacts of Acid Precipitation. Sandefjord,
Norway.
Duthie, H. C., and M. L. Ostrofsky. 1974. Plankton, chemistry
and physics of lakes in the Churchill Falls region of
Labrador. Jour. Fish Res. Board Can. 31:1105.
Goulden, C. E., L. Hornig, and C. Wilson. 1978. Why dp large
zooplankton species dominate? Verh. Int. Verein. Limnol.
20:2457.
Grahn, 0. 1976. Macrophyte succession in Swedish Lakes
caused by deposition of airborne acid substances. Pages
519-530 in Proc. First Int. Symp. Acid Precipitation and the
Forest Ecosystem, Ohio State University, May 12-15,1975.
U.S. Dep. Agric. Forest Serv. Gen. Tech. Rep. NE-23.
Hall, R. J., et al. 1980. Experimental acidification of a stream
in the Hubbard Brook Experimental Forest, New Hampshire.
Ecology 6(5) (in press).
Harvey, H. H,, ed. 1980. Increasing acidity in the Canadian
aquatic environment. Natl. Res. Counc. Can. Rep. (In
preparation.)
Hendrey, G. R. 1976. Effects of low pH on the growth of
periphytic algae in artificial stream channels. SNSF Proj.,
1432 Aas-NLH, Norway.
1980. Effects of acidity on primary productivity in
lakes: phytoplankton. In D. Shriner, G. R. Richmond, and S.
E. Lindberg, eds. Atmospheric sulfur deposition: Environ-
mental impacts and health effects. Ann Arbor Science
Press, Ann Arbor, Mich.
Hendrey, G. R., and F. Vertucci. 1980. Benthic plant
communites in acidic Lake Golden, New York: Sphagnum
and the algal mat. Proc. Int. Conf. Ecol. Impact of Acid
Precipitation. Sandefjord, Norway.
Hendrey, G. R., and R. F. Wright. 1976. Acid precipitation in
Norway: effects on aquatic fauna. Jour. Great Lakes Res.
2(suppl.):192.
Hendrey, G. R., et al. 1976. Acid precipitation: some
hydrobiological changes. Ambio. 5:224.
Henrikson, L., et al. 1980. Changes in predator-prey relations
important for the development in acidified lakes (abstract).
Int. Conf. Ecol. Impact of Acid Precipitation. Sandefjord,
Norway.
Hornstrom, E. 1979. Kalknings forsurnings-effekter pa
vakstplankton i tre vastkustsjoar. Statensnaturvardsverde
Rapport SNV-PM 1220. Box 1312 171 25 Solna Sweden.
Kallqvist, T., R. Romstad, and J. Kotai. 1975. Orienterende
undersokelse av pH-effekter pa algesamfunn i forsoksren-
ner. (Preliminary study of pH-effect on algal communities in
experimental channels). SNSF-Proj. IR 7/75 (in Norwegian).
Kwiatkowski, R. E., and J. C. Roff. 1976. Effects of acidity on
the phytoplankton and primary productivity of selected
northern Ontario lakes. Can. Jour. Bot. 54:2546.
Leivestad, H., et al. 1976. Effects of acid precipitation on
freshwater organisms. Pages 87-111 in F. J. Braekke, ed.
Impact of acid precipitation on forest and freshwater
ecosystems in Norway. FR 6/76. SNSF Proj. 1432 Aas-NLH,
Norway.
Lynch, H. 1979. Predation, competition and zooplankton
community structure: an experimental study. Limnol.
Oceanogr. 24:253.
Malley, D. F. 1980. Decreased survival and calcium uptake by
the crayfish Orconectis virilis in low pH. Can. Jour. Fish.
Aquat. Sci. 37:364.
Muller, P. 1980. Effects of artificial acidification on the growth
of periphyton. Can. Jour. Fish. Aquat. Sci. 37:355.
Nicholls, K. H., and P. J. Dillon. 1978. An evaluatin of
phosphorus-chlorophyll-phytoplankton relationships for
lakes. Int. Rev. Ges. Hydrobiol. 63:141.
Nilssen, J. P. 1980. Acidification of freshwater and limnetic
organisms: complex biotic interactions or statistics? (ab-
stract). Int. Conf. Ecol. Impact of Acid Precipitation.
Sandefjord, Norway.
Okland, J. 1969. Distribution and ecology of the freshwater
snails (Gastropoda) of Norway. Malacologia 9:143.
1980, A group of fish-food organisms (snails,
Gastropoda) and environmental parameters in 1,000 lakes
in Norway, with emphasis on low pH habitats (abstract). Int.
Conf. Ecol. Impact of Acid Precipitation. Sandefjord,
Norway.
Okland, K. A. 1969. On the distriubtion and ecology of
Gammarus lacustris, G. O. Sars in Norway, with notes on its
morphology and biology. Nytt Mag. Zool. 17:111.
1980. Ecology and distribution of Ase/lus
aquaticus (L.) in Norway, including relation to acidification
in lakes. IR 52/80. SNSG Proj. 1432 Aas, NLH, Norway.
Okland, K. A., and J. G. J. Kuiper. 1980. Smamuslinger
(Sphaeriidae) i Ferskvann i Norge-utbredelse, okologi og
relasjon til forsurning. IR 61/80. SNSF Proj. 1432 Aas-NLH,
Norway.
Petersen, R. C. 1980. Dynamics of coarse paniculate organic
material processing in diversified pH regimes (abstract). Int.
Conf. Ecol. Impact of Acid Precipitation. Sandefjord,
Norway.
Potts, D. T. W., and G. Fryer. 1979. The effect of pH and salt
content on sodium balance in Daphnia magna and
Acantholeberius curvirostris (Crustacea; Cladocera). Jour.
Comp. Physiol. Ser. B 129:289.
Raddum, G. 1978. Invertebrates: quality and quantity as fish
food. In G. R. Hendrey, ed. Limnological aspects of acid
precipitation. Rep. Int. Workshop. Brookhaven Natl. Lab.
Rep. No. 51074,
1980. Benthic invertebrate fauna in lakes with
varying H* concentrations (abstract). Proc. Int. Conf. Ecol.
Impact of Acid Precipitation. Sandefjord, Norway.
Raddum, G. G., etal. 1980. Phytoplankton and zooplankton in
waters with varying H* concentration (abstract). Proc. Int.
Conf. Ecol. Impact of Acid Precipitation. Sandefjord,
Norway.
Roff, J. R., and R. E. Kwiatkowski. 1977. Zooplankton and
zoobenthos communities of selected northern Ontario lakes
of different acidities. Can. Jour. Zool. 55:899.
-------
466 RESTORATION OF LAKES AND INLAND WATERS
Scheider, W. A., J. Adamski, and M. Paylor. 1975.
Reclamation of acidified lakes near Sudbury, Ontario.
Ontario Minist. Environ.
Scheider, W. A., J. Jones, and B. Cave. 1976a. A preliminary
report on neutralization of Nelson Lake near Sudbury,
Ontario. Ontario Minist. Environ.
1976b. Reclamation of acidified lakes near
Sudbury, Ontario by neutralization and fertilization. Ontario
Minist. Envrion.
Schindler, D. W. 1971. Carbon, nitrogen, and phosphorus and
the eutrophication of freshwater lakes. Jour. Phycol. 7:321.
1975. Whole lake eutrophication experiments
with phosphorus, nitrogen and carbon. Verh. Int. Verein.
Limnol. 19:577.
Schindler, D. W., and E. J. Fee. 1974. Experimental Lakes
Area: whole-lake experiments in eutrophication. Jour. Fish.
Res. Board Can. 31:937.
1973. Diurnal variation of dissolved inorganic
carbon and its use in estimating primary production and CO2
invasion in Lake 227. Jour. Fish. Res. Board Can. 30:1501.
Schindler, D. W., and S. K. Holmgren. 1971. Primary
production and phytoplankton in the Experimental Lakes
Area, Northwestern Ontario and other low-carbonate
waters, and a liquid scintillation method for determining1^
activity in photosynthesis. Jour. Fish. Res. Board Can.
28:189.
Schindler, D. W., et al. 1980. Experimental acidification of
Lake 223, Experimental Lakes Area: background data and
the first three years of acidification. Can. Jour. Fish. Aquat.
Sci. 37:342.
Schofield, C. L. 1973. The ecological significance of air-
pollution-induced changes in water quality of dilute-lake
districts in the Northeast. Trans. N.E. Fish Wildl. Conf.
1972:98.
Smith, R. W., and D. G. Frey. 1971. Acid mine pollution
effects on lake biology. Water Pollut. Control Res. Ser.
18050 EEC 12/71. U.S. Environ. Prot. Agency.
Sprules, G. W. 1975. Midsummer crustacean zooplankton
communities in acid-stressed lakes. Jour. Fish. Res. Board
Can. 32:389.
Stainton, M. P. 1973. A syringe gas-stripping procedure for
gas-chromatographic determination of dissolved inorganic
and organic carbon in fresh water and carbonates in
sediments. Jour. Fish. Res. Board Can. 30:1415.
Stumm, W., and J. J. Morgan. 1970. Aquatic chemistry.
Wiley-lnterscience, New York.
Sutcliffe, D. W., andT. R. Carrick. 1973. Studies on mountain
streams in the English lake district I. pH, calcium and the
distribution of invertebrates in the River Duddon. Freshw.
Biol. 3:437.
Walters, C. J., and R. E. Vincent. 1973. Potential productivity
of an alpine lake as indicated by removal and reintroduction
of fish. Trans. Am. Fish. Soc. 102:675.
Watanabe, T., et al. 1973. Studies on the eutrophication of
lake Osoresan-ko, a remarkable acidotrophic lake in Japan.
Ann. Rep. Noto Mar. Lab. 13:39. (In Japanese, English
summary.)
Van, N. D. 1979. Phytoplankton community of an acidified,
heavy-metal contaminated lake near Sudbury, Ontario:
1973-1977. Water Air Soil Pollut. 11:43.
Van, N. D., and R. Strus. 1980. Crustacean zooplankton
communities of acidic metal-contaminated lakes near
Sudbury Ontario. Can. Jour. Fish. Aquat. Sci.
-------
467
RESPONSES OF FISHES TO ACIDIFICATION OF STREAMS
AND LAKES IN EASTERN NORTH AMERICA
TERRY A. HAINES
U.S. Fish and Wildlife Service
University of Maine
Orono, Maine
CARL L SCHOFIELD
Department of Natural Resources
Cornell University
Ithaca, New York
ABSTRACT
Precipitation in eastern North America is acidic and contains elevated levels of heavy metals. As a
result of this precipitation the pH of lakes and streams has declined and metal concentrations have
increased in several areas. Episodic decreases in pH and increases in metal concentrations are
associated with spring snowmelt and heavy rains. These changes in water quality adversely affect
resident fish populations, reducing growth rate, increasing frequency of skeletal deformities, and
eliminating sensitive species through mortality or reproductive failure. Small-scale remedial
action has been taken in some areas, including lake neutralization, hatchery stocking, and
selective breeding for acid tolerance. In spite of all of the studies and observations, the causes of
declining fish populations in acidified waters have not yet been identified. Classical field
observations and laboratory bioassays dp not provide enough information to demonstrate the
effects of acid precipitation on fish populations. Innovative experiments will be required to provide
definitive answers to these questions.
INTRODUCTION
In eastern North America precipitation is now more
acidic than in Scandinavia, where acid rain problems
have been documented for a number of years. The
median annual pH for 1978-79 ranged from 4.0 to 4.4
(Gibson and Baker, 1979; Atmos. Environ. Serv., 1979).
Following an examination of historical data, Cogbill
(1976) concluded that precipitation during the 1920's
was low in acidity (probably pH 5.5). Precipitation had
become more acidic than normal by 1955, and has
increased steadily since then. This decline in pH has
been attributed to increased NOx emissions, and to a
variety of factors that have increased long-range
transport of SO2 emissions, such as taller smokestacks
and more use of particle precipitators in smokestacks
(Likens and Bormann, 1974). Studies of the magnitude
and distribution of SO2 and NO. in eastern North
America have shown that in both cases the highest
emissions are located in the industrial Midwest and
Great Lakes areas of the United States and Canada
(Altshuller and McBean, 1979). Prevailing wind
directions and storm tracks then carry these emissions
to the northeast (Altshuller and McBean, 1979;
Schlesinger, Reiners, and Knopman, 1974).
The metal content of acidic precipitation is higher
than normal precipitation (Delisle, Kloppenburg, and
Sylvain, 1979; Elgmork, Hagen, and Langeland, 1973;
Lazrus, Lorange, and Lodge, 1970; Ruhling and Tyler,
1973; Schlesinger, Reiners, and Knopman, 1974).
Metals which have been found at elevated levels
include lead, zinc, copper, iron, manganese, nickel,
mercury, and cadmium. They probably come from fossil
fuel combustion and metal smelting (Likens and
Bormann, 1974).
The effect of acidic precipitation on an aquatic
system is determined by the geochemistry, geomorph-
ology, and hydrodynamics of the watershed. These
factors determine the capacity of soil and water to
neutralize acids and resist pH change in surface
waters. In watersheds where the resulting buffering
capacity is low, the pH of surface waters has
decreased. Such changes have been recorded in Nova
Scotia, Ontario, New York, New Jersey, New Hamp-
shire, and Maine (Watt, Scott, and Ray, 1979; Beamish
and Harvey, 1972; Schofield, 1976; Johnson, 1979;
Hendrey, et al. 1980; Davis, et al. 1978). Trace metal
levels in acidified lakes are higher than in comparable
lakes located in areas receiving normal precipitation
(Wright and Gjessing, 1976; Beamish and Van Loon,
1977). These elevated metal levels may result from
direct input from precipitation or leaching from soils or
sediments by hydrogen ions in the precipitation. The
latter mechanism is especially important for aluminum
and manganese (Cronan and Schofield, 1979; Henrik-
sen and Wright, 1977).
Contaminants brought to aquatic systems by precipi-
tation vary seasonally and with individual storm events.
In eastern North America the greatest input occurs
with snowmelt, when pollutants stored in the
snowpack are released over a short period of time
(Schofield, 1977; Jeffries, Cox, and Dillon, 1977).
-------
468
RESTORATION OF LAKES AND INLAND WATERS
OBSERVED IMPACTS ON FISH
Long-term and episodic changes in pH and metal
content of surface waters have affected the biota
inhabiting these waters. Effects on fish have been
widely reported, probably because fish are highly
sensitive to acid and heavy metals and are one of the
most visible components of aquatic ecosystems. The
observed effects include direct mortality, reproductive
failure, reduced growth, and skeletal deformities.
The disappearance of various fish species from
acidified lakes has been recorded. In a study of 68
Ontario lakes the number of species of fish present was
found to decrease as measured lake pH decreased
(Harvey, 1975). In North America, there are many
recorded instances of fish disappearing from lakes in
Table 1. — Field studies on effect of lake acidification on natural fish
pond.
Family and Species
pH at which population ceases
reproduction, declines or disppears
Salmomdae
Lake trout
(Salvelinus namaycush)
Brook trout
(Salvelinus fontmalis)
Aurora trout
(S. fontmalis timagamiensis)
Arctic char
(Salvelinus alpinus)
Rainbow trout
(Sa/mo gairdner)
Brown trout
(Sa/mo trutta)
Atlantic salmon
(Sa/mo salar)
Lake herring
(Coregonus artedii)
Esocidae
Northern pike
(Esox lucius)
Cyprinidae
Lake chub
(Coueslus plumbeus)
Roach
(Leuciscus rutilus)
Catostomidae
White sucker
(Catostomus commersoni)
Ictalundae
Brown bullhead
(Ictalurus nebulosus)
Percopsidae
Troutperch
(Percopsis omiscomaycus)
Gadidae
Burbot
(Lota lota)
Centrarchidae
Smallmouth bass
(Micropterus dolomieull
Rock bass
(Ambloplites ruperstns)
Pumpkinseed sunfish
(Lepomis gibbosus)
Percidae
Walleye
(Stizostedion v vitreum)
Yellow perch
(Perca flavescens)
European perch
(Perca fluviatilis)
5.2-5.5 (1) 5.2-5.8 (2)
4.5-4.8 (3) . 5 (7)
5.0-5.5(4)
-5 (5)
5.5-6.0 (3)
5.0 (3) 5.0-5.5 (6) 4.5-5.5 (8)
5.0-55 (3)
4.5-4.7 (1) < 4.7 (2)
4.7-5.2 (2)
4.5-4.7 (1)
5.3-5.7 (5)
5.3-5.7 (5)
4.7-5.2 (1) (2)
4.5-5.2 (1) (2)
5.2-5.5 (1)
5.5-6.0 (1) 52-5 8 (2)
5.5-6.0 (1) > 5.5 (2) - 5.8 (9)
4.7-5.2 (1) (2)
4.7-5.2 (1) (2)
5.5-6.0 (1) 5.2-5.8 (2)
4.5-4.8 (1) < 4.7 (2)
5.0-5.5 (10)
References:
(1) Beamish, 1976; (2) Beamish, et al. 1975; (3) Grande, Muniz, and
Anderson, 1978, (4) Anonymous, 1978, (5) Aimer, et al. 1974; (6)
Jensen and Snekvik, 1972, (7) Schofield, 1976, (8) Wright and Snekvik,
1978; (9) N.Y. State Dep. Envron. Conserv 1978; (10) Runn, Johansson
and Milbrmk, 1977
Ontario and New York (Beamish, et al. 1975; Schofield,
1976). The apparent pH at which they disappeared is
listed in Table 1. Either these fish died or they failed to
reproduce.
Acute mortalities of fish which may result rrom
acidification and/or metal toxicity are rarely observed
under field conditions, and mortalities during episodic
inputs of hydrogen ion may be more common than has
been commonly observed. Mortality of Atlantic salmon
(Sa/mo salar) and brown trout (Salmo trutta) have been
recorded in Norway, usually following a sudden spring
melt or heavy autumn rain (Jensen and Snekvik, 1972;
Leivestad and Muniz, 1976; Leivestad, et al 1976). The
pH measured during these mortalities ranged from 3.9
to 4.6. Mortality of brook trout (Salvelinus fontinalis)
occurred during spring snowmelt in a New York
laboratory which was rearing fish in water piped from a
nearby stream (Schofield,1977). The lowest pH was
about 5.2 and aluminum concentration reached 1
mg/l. Mortality of Atlantic salmon fry has been
observed in hatchery pools fed water with a pH of 5.0
from the Mersey River, Nova Scotia (Farmer, et al.
1980). Embryos and fry are generally more sensitive to
acid and metals than adult fish (McKim, 1977), and
mortality of these stages would be difficult to observe
in nature.
Reproductive failure has been reported in fish
inhabiting acidified lakes, resulting in their gradual
disappearance over a period of several years. Missing
age classes was first observed, then all young age
classes were absent, and finally the populations
consisted only of a few large, old fish (Beamish, et al
1975; Beamish, 1976; Ryan and Harvey, 1980). This
apparent reproductive failure could have resulted from
either failure of fish to reproduce and deposit the
normal number of viable eggs, or post-spawning
mortality of embryos, larvae, or fry. Failure to deposit
eggs were observed in populations of white sucker
(Catostomus commersoni) from acidified lakes in
Ontario (Beamish, 1976; Lockhart and Lutz, 1976).
Post-spawning mortalities of early life stages of
Atlantic salmon have been produced in laboratory
studies under conditions similar to those found in
acidified lakes and streams (Daye and Garside, 1977,
1979).
Reduced growth rates have been reported for white
suckers and yellow perch (Perca flavescens) from
acidified lakes (Beamish, 1974b; Ryan and Harvey,
1980). Conversely, increased growth under similar
conditions was observed for older yellow perch and
rock bass (Ambloplites rupestris) (Ryan and Harvey,
1977, 1980). Skeletal deformities have been reported
for white sucker (Beamish, 1974a).
CAUSES OF IMPACTS ON FISH
Laboratory studies of the effects of reduced pH on
fish are summarized in Table 2. These studies show
that, depending on species, adult and embryo life
stages are least sensitive to pH, while production of
viable eggs, egg hatchability, fry mortality or growth are
the most sensitive biological parameters. Laboratory
mortality data sho^ ! be interpreted with caution.
Acidification of high pH water without adequate
aeration will produce high free CO2 concentrations,
-------
THE ACID RAIN PROBLEM: MECHANISM AND EFFECTS
469
Table 2. - Reduced pH levels found in laboratory experiments to cause various adverse effects on several fish species. Duration
of exposure varied from 4 days to life cycle.
Reduced Reduced Increased Increased Increased
co -iwanrfo • w Viable Egg Embryo Fry Adult Reduced Ceased Tissue Bone
Family and Species Eggs Hatchability ''
Salmonidae
Brook trout 5-0(1)
(Salvelinus fontinalis)
Arctic char
(Salvelinus alpinus)
Rainbow trout
(Salmo gairdneri)
Brown trout
(Salmo trutta)
Atlantic salmon
(Salmo salar)
Esocidae
Northern pike
(Esox lucius)
Cyprinidae
Roach
(Leuciscus rutilus)
Fathead minnow fifing
(Pimephales promelas) ( '
Catostomidae
White sucker
(Catostomus commersoni)
Percidae
European perch
(Perca fluviatilis)
6.5(1)
5.6(12)
4.0(2)
4.0(2)
4.0-5.5(15)
5.6(7)
5.9(14)
4.5(13)
5.6(7)
4.5(9)
5.5(10)
4.3(18)
< 5.0(9)
3.6(5)
3.9(6)
5.0(9)
3.4-4.4(18)
5.0(8)
6.1(1) 4.1(3) 6.5(1)
3.5(17) 4.6(19) 5.2(16)
4.8(20)
4.1(10)
4.8(20)
4.0(5)
4.3(6)
4.3(18)
5.9(14) 2.1(14) 4.5(14)
5.3(13) 4.5(4) 4.5(4) 4.2(4)
4.0(4) 5.0(12)
References:
(1) Mendenez, 1976; (2) Carrick, 1979; (3) Robinson, et al. 1976; (4) Beamish, 1972; (5) Daye and Garside, 1977; (6) Daye and
Garside, 1979; (7) Johansson and Milbrink, 1976; (8) Johansson and Kihlstrom, 1975; (9) Johansson, Runn, and Milbrink, 1977;
(10) Kwain, 1975; (11) Runn, Johansson, and Milbrink, 1977; (12) Trojnar, 1977b; (13) Trojnar, 1977a; (14) Mount, 1973; (15)
Peterson, Daye, and Metcalfe, 1980; (16) Daye and Garside, 1976; (17) Daye and Garside, 1975; (18) Daye, 1980; (19) Lievestad, et
al. 1976; (20) Edwards and Hjeldnes, 1977.
which interfere with the mortality effects of hydrogen
ion. These data may therefore report mortality at higher
pH than would occur in the absence of free C02.
Chronic exposure to reduced pH is unlikely to kill
adult fish in acidified lakes. Lakes rarely have pH below
4.5 even in the most extreme cases; this probably will
not be toxic to adult fish. However, a sudden reduction
in pH can cause mortality at a much higher pH than
chronic exposure to gradually reduced pH. Episodic pH
changes in early spring can cause observed fish
mortalities. Death will also occur at higher pH in water
with very low ionic content than in water with
moderate or high ionic content, because of increased
osmotic stress (Fromm, 1980).
Mortality of fish at low pH has been attributed to
failure of ion regulation or to asphyxiation. Fish
collected from the Tovadal River, Norway, during an
acid-caused fish-kill had lower plasma sodium and
chloride levels than fish from unaffected sections of the
river. The reduced levels were comparable to those
found in fish killed by low pH in a tank experiment
(Leivestad and Muniz, 1976). Transfer from pH 7 to pH
4 caused a threefold increase in sodium loss from
brown trout, sufficient to cause death in 24 to48 hours
(Potts, 1979; McWilliams and Potts, 1978). The loss of
ions appeared to result from an increased efflux across
the gills, rather than a reduced influx (Packer and
Dunson, 1970; McWilliams and Potts, 1978; Fromm,
1980). Exposure to increases in hydrogen ion con-
centrations increases gill membrane permeability.
Hydrogen ions from the environment diffuse in and
sodium and other ions from the blood diffuse out. Gill
membrane permeability is mediated by calcium ion,
and possibly other divalent cations, with the presence
of calcium reducing permeability. Thus, with increasing
calcium levels, the loss of ions is reduced and the lethal
pH decreases (Fromm, 1980).
Low pH may interfere with respiration through
several different mechanisms. Exposure to elevated
hydrogen ion may cause excessive secretion of mucus
from the gills, thereby reducing the rate of oxygen
diffusion across the gill surface (Daye and Garside,
1976; Dively, et al. 1977). The increased influx of
hydrogen ion reduces blood pH, which in turn reduces
the oxygen-carrying capacity of hemoglobin. Fish
respond to chronic acid exposure by increasing the
hematocrit 'index, hemoglobin content of blood, and
hemopoietic activity to maintain oxygen carrying
capacity (Fromm, 1980).
It is not possible currently to determine whether
osmotic or respiratory effects, or both are responsible
for death of fish at low pH, or whether some other
factor may be involved. Fromm (1980) speculated that
reduced pH may destroy enzyme activity; however, he
also noted that trout are remarkably tolerant of
changes in blood pH per se, which would indicate that
enzyme activity was not affected. The action of reduced
pH to increase the toxic effects of metals, such as
-------
470
RESTORATION OF LAKES AND INLAND WATERS
aluminum, may be more important than the direct
effects of pH.
The failure of fish exposed to low pH to produce
viable eggs has been explained by changes in serum
calcium levels in females. Beamish (1976) and
Lockhart and Lutz (1976) observed that the failure of
female white suckers exposed to low pH to produce
viable ova was coincidental with lower than normal
serum calcium levels. Serum calcium, in the form of
complex calcium phospho-proteinate, normally in-
creases in females during the period of ova develop-
ment. Ruby, et al. (1977) showed that oogenesis in
flagfish (Jordanella floridae) was reduced by exposure
to reduced pH because protein production was
disturbed, leading to improper yolk formation. Calcium
is important in the transfer of protein to the yolk.
Reduced egg hatchability at low pH in some fish
species apparently results from failure of the chorio-
lytic enzyme to properly degrade the chorion. This was
observed in European perch (Perca fluviati/is) by Runn,
Johansson, and Millbrink(1977) and in Atlantic salmon
by Peterson, Daye, and Metcalfe (1980). This supports
Fromm's (1980) speculation concerning the effect of
reduced pH on enzyme activity.
The effect of reduced pH on growth is ambiguous.
Growth of some fish may either increase or decrease
following acidification. Increased growth of fish which
survive acidification may result from decreased
competition for food following the disappearance of
acid sensitive competitors (Ryan and Harvey, 1977,
1980). Conversely, the growth rate offish may decline
in the face of abundant food, which is explained as a
sublethal stress response to the increased acid
(Beamish, 1974b). One laboratory study produced
reduced growth in brook trout exposed to pH 4.6
(Leivestad, et al. 1976), but only during the first 3
months of exposure in another (Menendez, 1976), On
the other hand, reduced pH did not affect growth in
brown trout (Jacobson, 1977).
Skeletal deformities observed in natural populations
of white sucker (Beamish, et al. 1975) were also
produced by reduced pH in laboratory bioassays
(Beamish, 1972; Trojnar, 1977a). This may be caused
by loss of serum calcium under acid stress. However,
the effect has not been observed in other species.
In addition to reduced pH, metal concentration, either
alone or in conjunction with reduced pH, was
responsible for fish losses in many acidified lakes.
Cronan and Schofield (1979) and Schofield and Trojnar
(1980) showed that mortality of brook trout in New York
was caused by aluminum and pH in combination,
rather than by either factor singly. Similar results were
reported by Grahn (1980) for brook trout and ciscoe
(Coregonus albula), by Hermann and Baron (1980) for
brook trout, and by Muniz (1980) for brown trout.
The toxicity of aluminum varies with pH and the
presence of complexing agents. The most toxic forms of
aluminum (AI(OH)++, AI(OH)2+, Al (OHM are present at
ph 5, and the toxicity declines at both higher and lower
pH. Aluminum complexed with organic matter is not
toxic to organisms (Driscoll, et al. 1980). At pH 5,
aluminum concentrations of 0.2 mg/l or greater cause
gill hyperplasia and mucus secretion (Schofield and
Trojnar, 1980). These pH and aluminum levels are
common in acidified waters (Wright and Henriksen,
1978; Dickson, 1975; Schofield, 1977).
Van, Girard, and Lafrance (1979) reported mortality
of rainbow trout (Salmo gairdneri) stocked in an
acidified lake which had been chemically neutralized.
Fish had been eliminated from the lake by acidification.
The mortality was attributed to copper or copper and
zinc in combination. Copper concentrations were 42 to
67 ug/l and zinc concentrations were 23 to 33 ug/l.
Copper concentrations as low as 9.5 mg/l were
reported to be toxic to brook trout embryos and
juveniles in laboratory exposures (McKim and Benoit,
1971). Copper concentrations up to 450 mg/l have
been measured in lakes near Sudbury, Ontario
(Adamski and Michalski, 1975). The toxicity of many
metals increases as pH and calcium decline (Chrost
and Pinko, 1980; Franzin and McFarlane, 1980.)
REMEDIAL ACTION
Action has been taken to counter the effects of acid
precipitation on selected fish populations. A variety of
approaches is now under study, including lake
neutralization, hatchery stocking, and selective breed-
ing for acid tolerance. Lake neutralization is the most
widely used approach. Lakes are treated with CaCOs,
Ca(OH)z, or both. The combination treatment appears to
have given the most satisfactory results to date
(Scheider and Dillon, 1976). In New York, lakes
totalling 819 acres are now being treated with lime;
viable brook trout fisheries survive (Pfeiffer, pers.
comm.). In Ontario several lakes have been neutralized,
or neutralized and fertilized, and changes in biota
monitored (Scheider and Dillon, 1976; Dillon, et al.
1977). However, rainbow trout stocked in one such
treated lake survived only a few days, apparently as the
result of copper toxicity (Yan, Girard, and Lafrance,
1979). Metal concentrations in this lake had been
elevated by acidification and atmospheric input from
metal smelters. Apparently neutralization alone was
not sufficient to reduce metal toxicity.
Stocking hatchery fish at the fingerling stage avoids
exposure of sensitive early life history stages to
environmental stress and maintains fish populations in
lakes in New Hampshire and Maine, that would
otherwise not support fish (Haines, unpubl. data). Gunn
(1980) reports that in situ incubation of eyed eggs of
fish (species not stated) in spawning boxes filled with
limestone is a promising technique for maintaining fish
populations in acidified lakes. The limestone protects
embryos and larvae from toxic acids or metals in lake
water.
Various populations of brook and brown trout have
been shown to differ in tolerance to acid, and this
tolerance is heritable (Robinson, et al. 1976; Edwards
and Gjedrem, 1979). Thus a selective breeding
program to improve fish survival in acidified waters is
possible. This technique has been successfully em-
ployed in New York (Pfeiffer, pers. commun.).
-------
THE ACID RAIN PROBLEM: MECHANISM AND EFFECTS
471
DETERMINATION OF THE
MECHANISMS OF ACID PRECIPITATION
IMPACT ON FISH POPULATIONS
In spite of all of these studies, we remain ignorant of
the true relationship between acidification from
precipitation and loss of fish populations. Laboratory
findings of pH tolerance cannot be directly related to
field observations to determine the reasons for the loss
of fish. Possible explanations for these discrepancies
include effects of metals, singly or synergistically, and
episodic changes in water quality that occur during the
acidification process. It is possible that episodic
changes in pH and metal content stress critical life
history stages (e.g., eggs, fry, maturing adults).
Descriptive field studies and classical laboratory
bioassays do not provide sufficient information to
demonstrate how acid precipitation affects fish
populations. We believe that innovative approaches
will be required to provide these answers. We advocate
the following approaches:
1. Intensive case studies: application of population
dynamics research to fish populations in carefully
selected waters undergoing acidification. These studies
would demonstrate the relative importance of adult
mortality, reproductive failure, and reduced growth in
determining effects of acidification on the populations.
2.Experimental manipulations of pH in field situa-
tions: induce both chronic and episodic acidification in
sensitive waters not subject to acidic precipitation and
determine resulting fish population responses.
3. Detailed studies of pH-metal-ligand interactions:
the relationship between pH, metal form, and ligands in
determining toxicity must be detailed in laboratory
studies, and field sampling must include more detailed
chemical analyses than in the past.
These approaches will provide insight into the
acidification process which has not been obtained from
previous studies.
REFERENCES
Adamski, J., andM. Michalski. 1975. Reclamation of acidified
lakes — Middle and Lohi, Sudbury, Ontario. Verh. Int.
Verein. Limnol. 19:1971.
Aimer, B., et at. 1974. Effects of acidification on Swedish
lakes. Ambio 3:30.
Altshuller, A., and G. McBean. 1979. The LRTAP problem in
North America: a preliminary overview. Report prepared by
the US-Canada Research Consultation Group on the Long-
Range Transport of Air Pollutants.
Anonymous. 1978. Limnological observations on the Aurora
trout lakes. Water Resour. Asse. Ontario Ministry Environ.
Rep.
Atmospheric Environment Service. 1979. CANSAP data
summary. Downsview, Ontario.
Beamish, R. 1972. Lethal pH for the white sucker,
Catostomus commersoni (Lacepede). Trans. Am. Fish. Soc.
101:355.
1974a. Growth and survival of white suckers
(Catostomus commersoni) in an acidified lake. Jour. Fish.
Res. Board Can. 31:49.
1974b. Loss of fish populations from unexploited
remote lakes in Ontario, Canada as a consequence of
atmospheric fallout of acid. Water Res. 8:85.
1976. Acidification of lakes in Canada by acid
precipitation and the resulting effects on fishes. Water Air
Soil Pollut. 6:501.
Beamish, R., and H. Harvey. 1972. Acidification of the
LaCloche Mountain lakes, Ontario, and resulting fish
mortalities. Jour. Fish. Res. Board Can. 29:1131.
Beamish, R., and J. Van Loon. 1977. Precipitation loading of
acid and heavy metals to a small acid lake near Sudbury,
Ontario. Jour. Fish. Res. Board Can. 34:649.
Beamish, R., et al. 1975. Long-term acidifcation of a lake and
resulting effects on fishes. Ambio. 4:98.
Carrick T. 1979. The effect of acid water on the hatching of
salmonid eggs. Jour. Fish. Biol. 14:165.
Chrost, L., and L. Pinko. 1980. The effect of calcium on the
toxicity of zinc and lead compounds in the water medium.
Abstr. Voluntary Contrib. Int. Conf. Ecol. Impact of Acid
Precipitation, Sandefjord, Norway. March 11-14.
Cogbill, C. 1976. The history and character of acid
precipitation in eastern North America. Pages 363-370. in
L. Dochinger and T. Seliga, eds. Proc. First Int. Symp. Acid
Precipitation and the Forest Ecosystem. Foresst Serv. Gen.
Tech. Rep. NE-23. U.S. Dep. Agric.
Cronan, C., and C. Schofield. 1979. Aluminum leaching
response to acid precipitation: Effects on high-elevation
watersheds in the northeast. Science 204:304.
Davis, R. et al. 1978. Acidification of Maine (USA) lakes by
acidic precipitation. Verh. Int. Verein. Limnol. 20:532.
Daye, P. 1980. Attempts to acclimate embryos and alevins of
Atlantic salmon, Salmo salar, and rainbow trout, S.
gairdneri, to low pH. Can. Jour. Fish. Aquat. Sc. 37:1035.
Daye, P., and E. Garside. 1975. Lethal levels of pH for brook
trout, Salvelinus fontinalis (Mitchill). Can. Jour. Zool.
53:639.
1976. Histopathologic changes in surficial tissues
of brook trout, Salvelinus fontinalis Mitchill), exposed to
acute and chronic levels of pH. Can. Jour. Zool. 54:2140.
1977. Lower lethal levels of pH for embryos and
alevins of Atlantic salmon, Salmo salar L. CAn. Jour. Zool.
55:1504.
1979. Development and survival of embryos and
alevins of the Atlantic salmon, Salmo salar L., continuously
exposed to acidic levels of pH, from fertilization. Can. Jour.
Zool. 57:1713.
Delisle, C. J. Kloppenburg, and J. Sylvain. 1979. Mercury and
acidity in the precipitation of north-west Quebec. Eau Du
Quebec 12:76.
Dickson, W. 1975. The acidification of Swedish lakes. Inst.
Freshw. Res. Drottingholm 54:8.
Dillon, P. et al. 1977. Acidic lakes in Ontario: characterization,
extent, and responses to base and nutrient additions.
Ontario Ministry Environ. Rep.
Dively, J., eta 1.1977. Blood P ,P and pH changes in brook
trout (Salvelinus fontinalis) exposed to sub-lethal levels of
acidity. Comp. Biochem. Physiol. 57:347.
Driscoll, C., et al. 1980. Effects of aluminum speciation on
fish in dilute, acidified waters. Nature 284:161.
Edwards, D., and T. Gjedrem. 1979. Genetic variation in
survival of brown trout eggs, fry and fingerlings in acidic
water. Acid Precipitation-Effects on Forest and Fish (SNSF)
Res. Rep. 16.
Edwards, D., and S. Hjeldnes. T977. Growth and survival of
salmonids in water of different pH. Res. Rep. 10, SNSF, Aas,
Norway.
Elgmork, K., A. Hagen, and A. Langeland. 1973. Polluted
snow in southern Norway during the winters 1968-71.
Environ. Pollut. 4:41.
Farmer, G., et al. 1980. Some effects of the acidification of
Atlantic salmon rivers in Nova Scotia. Unpubl. mss.
-------
472
RESTORATION OF LAKES AND INLAND WATERS
Franzin, W., and G. McFarlane. 1980. Fallout, distribution
and some effects of Zn, Cd, Cu, Pb and As in aquatic
ecosystems near a base metal smelter on Canada's
Precambrian Shield. Abstr. Voluntary Contrib. Int. Conf.
Ecol. Impact of Acid Precipitation, Sandefjord, Norway.
March 11-14.
Franzin, W., G. McFarlane, and A. Lutz. 1979. Atmospheric
fallout in the vicinity of a base metal smelter at Flin Flan,
Manitoba, Canada. Environ. Sci. Technol. 13:1513.
Fromm, P. 1980. A review of some physiological and
toxicoiogical responses of freshwater fish to acid stress.
Environ. Biol. Fish. 5:79.
Gibson, J. and C. Baker. 1979. NADP first data report July
1978 through February 1979. Nat. Resour. Ecol. Lab.
Colorado State University, Fort Collins.
Grahn, 0. t980. Fish kills due to high aluminum concentra-
tions in lake water. Abstr. Voluntary Contrib. Int. Conf. on
the Ecol. Impact of acid Precipitation, Sandefjord, Norway.
March 11-14.
Grande, M., I. Muniz, and S. Andersen. 1978. Relative
tolerance of some salmonids to acid waters. Verh. Int.
Verein. Limnol. 20:2076.
Gunn, J. 1980. Effects of acid precipitation on fish stocks and
amelioration efforts in lakes near Sudbury, Ont. Abstr.
Voluntary Contrib. Int. Conf. Ecol. Impact of Acid Precipita-
tion, Sandefjord, Norway. March 11-14.
Harvey, H. 1975. Fish populations in a large group of acid-
stressed lakes. Verh. Int. Verein. Limnol. 19:2406.
Hendrey, G. et al. 1980. Geological and hydrochemical
sensitivity of the eastern United States to acid precipitation.
Rep. prepared for Ecol. Effects Div. Corvallis Environ. Res.
Lab. U.S. Environ. Prot. Agency.
Henriksen, A., and R. Wright. 1977. Effects of acid
precipitation on a small acid lake in southern Norway. Nord.
Hydrol. 8:1.
Herrmann, R., and J. Baron. 1980. Aluminum mobilization in
acid environments. Great Smoky Mountains National Park.
Abstr. Voluntary Contrib. Int. Conf. Ecol. Impact of Acid
Precipitation, Sandefjord, Norway. March 11-14.
Jacobsen, O. 1977. Brown trout (Salmo trutta L.) growth at
reduced pH. Aquaculture 11:81.
Jeffries, D., C. Cox and P. Dillon. 1979. Depression of pH in
lakes and streams in central Ontario during snowmelt. Jour.
Fish. Res. Board Can. 36:640.
Jensen K., and E. Snekvik. 1972. Low pH levels wipe out
salmon and trout populations in southernmost Norway.
Ambio 1:223.
Johansson, N., and J. Kihlstrom. 1975. Pikes (Esox lucius L.)
shown to be affected by low pH values during first weeks
after hatching. Environ Res. 9:12.
Johansson, N., and G. Milbrink. 1976. Some effects of
acidified water on the early development of roach (Rutilus
rutilus L.) and perch (Perca fluviati/is L.) Water Res. Bull.
12:39.
Johansson, N., P. Runn, and G. Milbrink. 1977. Early
development of three salmonid species in acidified water.
Zoon 5:127.
Johnson, A. 1979. Evidence of acidification of headwater
streams in the New Jersey pinelands. Science 206:834.
Kwain, W. 1975. Effects of temperature on development and
survival of rainbow trout, Salmo gairdneri, in acid waters.
Jour. Fish. Res. Board. Can. 32:493.
Lazrus, A., E. Lorange, and J. Lodge, Jr. 1970. Lead and other
metal ions in United States precipitation. Environ. Sci.
Technol. 4:55.
Leivestad, H., and I. Muniz. 1976. Fish kill at low pH in a
Norwegian river. Nature 259:391.
Leivestad, H., et al. 1976. Effects of acid precipitation on
freshwater organisms. Pages 86-111 in F. Braekke, ed.
Impact of acid precipitation on forest and freshwater
ecosystems in Norway. Acid Precipitation — Effects on
Forest and Fish (SNSF) Res. Rep. 6.
Likens, G., and F. Bormann. 1974. Acid rain: a serious
regional environmental problem. Science 184:1176.
Lockhart, W., and A. Lutz. 1976. Preliminary biochemical
observations of fishes inhabiting an acidified lake in
Ontario, Candada. Pages 545-569. in L. Dochinger and T.
Seliga, eds. Proc. First Int. Conf. Acid Precipitation and the
Forest Ecosystem. Forest Serv. Gen. Tech. Rep. NE-23. U.S.
Dep. Agric.
McKim, J. 1977. Evaluation of tests with early life history
stages of fish for predicting long-term toxicity. Jour. Fish
Res. Board Can. 34:1148.
McKim, J., and D. Benoit. 1971. Effects of long-term
exposures to copper on survival, growth and reproduction of
brook trout (Salvelinus fontinalis). Jour. Fish. Res. Board
Can. 28:655.
McWilliams, P., and W. Potts. 1978. Physiological effects of
low pH on brown trout, Salmo trutta. Jour. Comp. Physiol. B
126:277.
Menendez, R. 1976. Chronic effects of reduced pH on brook
trout (Salvelinus fontinalis). Jour. Fish. Res. Board Can.
33:118.
Mount, D. 1973. Chronic effect of low pH on fathead minnow
survival, growth and reproduction. Water Res. 7:987.
Muniz, I. 1980. Toxic effects of aluminum on the brown trout,
Salmo trutta. Abstr. Voluntary Contrib. Int. Conf. Ecol.
Impact of Acid Precipitation, Sandefjord, Norway. March 11 -
14.
NYSDEC. 1978. Acid waters and declining fisheries — the
next step. New York State Dep. Environ. Conserv. Bur. Fish.,
Albany. Mimeo Rep.
Packer, R., and W. Dunson. 1970. Effects of low environ-
mental pH on blood pH and sodium balance of brook trout.
Exp. Zool. 174:65.
Peterson, R., P. Daye, and J. Metcalfe. 1980. Inhibition of
Atlantic salmon (Salmo salar) hatching at low pH. Can. Jour.
Fish. Aquat. Sci. 37:770.
Potts, W. 1979. Physiological effects of low pH. In M. Wood,
ed. Ecological effects of acid precipitation. EPRI SOA77-
403. Electric Power Res. Inst. Palo Alto, Calif.
Robinson, G., et al. 1976. Differences in low pH tolerance
among strains of brook trout (Salvelinus fontina/is). Jour.
Fish. Biol. 8:5.
Ruby, S., J. Aczel, and G. Craig. 1977. The effects of
depressed pH on oogenesis in flagfish Jordanella floridae.
Water Res. 11:757.
Ruhling, A., and G. Tyler. 1973. Heavy metal deposition in
Scandinavia. Water Air Soil Pollut. 2:445.
Runn, P., N. Johansson, and G. Milbrink. 1977. Some effects
of low pH on the hatchability of eggs of perch, Perca
fluviatilis L. Zoon 5:115.
Ryan, P., and H. Harvey. 1977. Growth of rock bass,
Ambloplites ruperstris, in relation to the morphoedaphic
index as an indicator of an environmental stress. Jour. Fish.
Res. Board Can. 34:2079.
1980. Growth responses of yellow perch, Perca
flavescens (Mitchill), to lake acidification in the LaCloche
mountain lakes of Ontario. Environ. Biol. Fish. 5:97.
Scheider, W., and P. Dillon. 1976. Neutralization and
fertilization of acidified lakes near Sudbury, Ontario. Proc.
11th Can. Symp. Water Pollut. Res. Can.
Schlesinger, W., W. Reiners, and D. Knopman. 1974. Heavy
metal concentrations and deposition in montane eco-
systems of New Hampshire, USA. Environ. Pollut. 6:39.
-------
THE ACID RAIN PROBLEM: MECHANISM AND EFFECTS 473
Schofield, C. 1976. Acid precipitation: effects on fish. Ambio
5:228.
1977. Acid snow-melt effects on water quality
and fish survival in the Adirondack mountains of New York
state. Res. Proj. Tech. Completion Rep. Proj. No. A-072-NY
submitted to Off. Water Res. Technol., Washington, D.C.'
Schofield, C., and J. Trojnar. 1980. Aluminum toxicity to fish
in acidified waters. Pages 347-366 in T. Toribora, M. Miller
and P. Morrow, eds. Polluted rain. Plenum Press, New York!
Trojnar, J. 1977a. Egg and larval survival of white suckers
(Catostomus commersoni) at low pH. Jour. Fish. Res Board
Can. 34:262.
1977b. Egg hatchability and tolerance of brook
trout (Salvetinus fontinalis) fry at low pH. Jour. Fish. Res.
Board Can. 34:574.
Watt, W., D. Scott, and S. Ray. 1979. Acidification and other
chemical changes in Halifax County lakes after 21 years.
Limnol. Oceanogr. 24:1154.
Wright, R., and E. Gjessing. 1976. Acid precipitation: changes
in the chemical composition of lakes. Ambio 5:219.
Wright, R., and E. Snekvik. 1978. Acid precipitation: chemistry
and fish populations in 700 lakes in southernmost Norway.
Verh. Int. Verein. Limnol. 20:765.
Yan, N., R. Girard and C. Laf ranee. 1979. Survival of rainbow
trout, Salmo gairdneri, in submerged enclosures in lakes
treated with neutralizing agents near Sudbury, Ontario.
Ontario Ministry Environ. Tech. Rep. LTS 79-2.
-------
474
FUTURE TRENDS IN ACID PRECIPITATION
AND POSSIBLE PROGRAMS
JAMES R. KRAMER
Department of Geology,
McMaster University
Hamilton, Ontario
ABSTRACT
Remedial action programs focus on sources, terrestrial and aquatic effects, and socio-economic
aspects of acidic precipitation. One research thrust is to assess the critical height of a source
emission with respect to the production of acidic aerosols; these soruces then should be given
priority in an abatement program. Modeling seems to be the best approach to this problem, and
field data exist for various point source emissions for various heights to calibrate the model.
Abatement of aquatic effects can be effected through liming, nutrient enrichment, and
modification of watershed hydrology. Mass budget calculations and field studies are required to
evaluate these alternatives. Finally, abatement costs should be compared with the costs of tax
reduction and direct subsidies to the user to decrease energy consumption and thus emissions.
The recent U.S. National Academy of Science energy report suggests that a 50 percent per capita
reduction in energy consumption is achievable. Assuming a proportional correlation between
emissions and deposition, this energy saving is comparable to the billm dollars required to abate
existing emission sources with NO , SO , and paniculate controls.
INTRODUCTION
Any attempt to propose research and abatement
programs for acid precipitation must consider a number
of related phenomena:
1. Emissions resulting in acid precipitation are
complex and tied to the energy demands, technological
and economic conditions, and lifestyles of one or more
countries.
2. Acid aerosols and acid precipitation cover a wide
area, containing many millions of square kilometers
(Hoffman and Rosen, 1980).
3. The impacts of aerosols and acid precipitation are
multiple, negatively affecting manmade structures,
health, aquatic ecosystems, and possibly forestry and
agricultural productivity. The economic effects of SC>2—
SO4 have been estimated in the tens of billions of
dollars (U.S. EPA, 1979).
4. There is a strong possibility that acid aerosol and
precipitation impacts will become more severe in the
next 20 to 30 years.
5. There will be a strong correlation between energy
use and the form of energy and acid precipitation in the
next 20 to 30 years.
The rationale used here to arrive at suggested
research studies and programs to abate acid precipita-
tion is (1) consideration of emissions, transport of
pollutant, and impacted areas; and (2) consideration of
immediate (and generally temporal) abatement, and
also long-term (decades) and generally permanent
abatement. Controlling emissions is considered both
from technological abatement ("scrubbing") and from
the reduction of energy losses ("waste"). Transport of
pollutants focuses on the sensitivity of stack height to
long-range transport and on chemical conversion
processes. Impacted area abatement focuses directly
upon treatment of aquatic watersheds ("liming").
The purpose here is to project emissions into the
21st century using energy models as a base and to
suggest general approaches and specific research
programs for remedial action.
FUTURE ENERGY DEMANDS AND
PROGRAMS
The kind of fuel and amount of waste energy is
almost directly related to acid precipitation aerosols.
Therefore it is pertinent to examine various energy
projections focusing on emissions. Perhaps the most
profound statement regarding future energy demands
and programs is that there are many uncertainties for
decisionmakers. Unfortunately, the large scale of the
problem demands decisions and investments now to
produce payoffs 20 to 30 years hence.
Some of the most significant factors that will affect
energy demand and policy are the evolving relation-
ships between economic factors and energy consump-
tion, conservation, and technological change. Equally
significant will be the reactions of society's next
generation to altered lifestyles. All of these are related
in a most complex way, and can be very much affected
by consumer attitude, government policies, pricing, and
technological change. For example, there has been
much discussion between economists relying on
supply/pricing strategy and resource scientists seeing
finite limits to availability of energy material (Hayes,
1979). Furthermore, there is no agreement that a
stable economy requires a nearly constant "GNP/
energy consumption" ratio; some suggest that this
ratio may double over time with a stable GNP growth of
-------
THE ACID RAIN PROBLEM: MECHANISM AND EFFECTS
475
2 percent and "business as usual" (Natl. Res Counc
1980).
The large degree of flexibility and uncertainty is
perhaps best summarized by the recent compilation
(Marshall. 1980) showing changes made in forecasts
from 1972 to 1978 by "low-growth" and "high-
growth" proponents for energy requirements in the
year 2000 to 2010. The 1972 "low-growth" projection
of 125 quads (10'5 BTU) equals the 1978 prediction of
the "high-growth" advocates; and the overall range of
forecasts in this 6-year period is sixfold (33 to 190
quads) for total energy demand. These figures can be
compared to total energy input of about 78 quads in
1977.
Amid these uncertainties, there are, however, some
strong future probabilities that directly relate to energy
consumption and pollution emission of acidic aerosols.
The following use the final report of the Committee on
Nuclear and Alternate Energy Systems (CONAES) (Natl.
Res. Counc. 1980) as a guide.
(1) Coal will be a significant fuel in the future,
especially for thermal generating plants. Energy
production using coal will about double in the next 20
to 30 years. (2) Electrical generation from thermal
generating plants will become a more important form
of energy, decreasing the ratio of net energy/gross
energy. As an approximate and limiting assumption,
the difference between gross and net energy may be
assumed to be proportional to the emission loadofSO2.
Figure 1 shows the forecasts from CONAES models
comparing total energy demand to various price
assumptions. Figure 2 shows the coal energy demand
for the same price assumptions. The important
difference between these two diagrams is the
comparison relative to present (1975) values. For
example, total energy demand projections for 2010 are
a little less than 1975 demand for case A, maximum
price with conservation. But coal energy demand
(assumed proportional to unabated SO2 emissions)
projection for case A is 1.6 times 1975. In short, there
is not a quantitative comparison between energy
demand and emissions. This is due to the increased
use of coal in projection, but more important, the
decreased efficiency due to the increased use of
electricity.
This model suggests then that emissions of SO2
without scrubbing will increase at least 60 percent in
the 21 st century. Other models need to be studied from
an environmental emissions perspective in order to test
this conclusion.
The distribution of coal energy demand by the
consuming sector focuses on areas requiring most
attention. The three sectors defined are transportation,
buildings, and industry; transportation has almost no
coal energy demand, and industry consumes about 62
percent of coal energy demand for all scenarios except
the lowest price where it decreases.
Another CONAES scenario, "CLOP", assumes im-
plementation of advanced technology, a strong en-
vironmental conscience, low material consumption,
etc. (Natl. Res. Counc. 1980). The CLOP projections
result in the same total energy consumption as does
the highest price plus conservation model (A*);
however, presumably (not stated by CONAES) the
energy loss and especially the pollutant emissions
would be lower due to increased use of solar energy.
In addition, the following summary questions
regarding energy appear to be important with respect
to emissions and the acid rain problem:
1. Assuming a major increase in coal use, can we
learn to use coal cleanly?
2. Given that 50 percent of the fuel consumption by
automobiles is made for trips of less than 5 km (Cook,
1973), can cities and transportation be restructured in
the next 20 to 30 years?
3. Can cogeneration facilities become a significant
function in the next 20 to 30 years?
4. Can we learn to use wastes, many of which are
organic pollutants now, as an important source of fuel?
5. Can industrial energy consumption per unit of
production be reduced considerably (20 to 50 percent)
by waste recycling, product substitution, and tech-
nological innovation?
SOURCE EMISSION ABATEMENT
As previously mentioned, source emission abate-
ment focuses upon coal burning plants. Abatement in
this context can be achieved by technological scrub-
bing, by reduced energy consumption, and by use of
alternate energy sources. In the following discussions,
it is assumed that abatement will take place. Therefore,
scrubbing costs might be replaced, for example, by
costs of new technology or costs of developing
alternate energy. Estimated efficiencies for removing
SOa and associated pollutants from plants are 80 to 90
percent for new installations and 50 percent for
retrofits (U.S. EPA, 1979a, b) with a cost of billions of
dollars per year for treatment and disposal in the
United States. These efficiencies and costs serve as a
reference for alternate approaches to coal emission
solutions.
Present sulfur abatement procedures for coal consist
of physical cleaning, flue gas desulfurization (FGD),
fluidized bed combustion (FBC), liquefaction, and
gasification. In addition, chemical cleaning may
become important (Ruether, 1979). Physical cleaning
will remove inorganic sulfur, which accounts for about
50 percent of the total sulfur, but only 10 to 20 percent
of U.S. coals can be reduced to acceptable levels by this
technique. FGD can attain 90 percent removal
efficiencies, given the availability of trained personnel
(U.S. EPA, 1979b). FGD removal should be considered
for cycled water in some fly ash lagoons, because some
fly ash slurries attain a pH of 11 to 12; this pH should
increase the SO2 oxidation rate. Other fly ash slurries
are acid. The characteristics of fly ash in different coals
should be investigated, since using fly ash would cut
chemical costs and solid waste disposal. Ponding of fly
ash and water would remove the possibility that trace
metals would be mobilized in surface and ground
water; furthermore, the trace metals might be exploited
as a resource in the waste.
Some recent studies have characterized the chemi-
cal nature of fly ash (Talbot, Anderson, and Andren,
1978; Edzwald and DePinto, 1978). During the past
decade, studies and proposals have been made for
using fly ash in acid mine neutralization (Tenney and
-------
476
RESTORATION OF LAKES AND INLAND WATERS
Echelberger, 1970). This requires a detailed mineral-
ogical and chemical study of the solid along with a
solution study of the aqueous slurry. Clearly much
more research must focus on coal and its wastes, their
abatement and resource potentials.
Many have emphasized the use of other "clean"
sources of energy. Achieving abatement by using
alternate sources and/or emissions involves many
uncertainties as well as many possibilities. Selecting
alternate sources of energy hinges on economic
feasibility, amount of energy required to obtain the
alternate sources, potential duration of supply, and the
likelihood of developing other environmental problems.
Nuclear and solar sources are often discussed;
geothermal sources may be significant to specific
areas. Nuclear generation is a contentious issue (e.g.,
Natl. Res. Counc., 1980; Holden, Smith, and Morris,
1979) and is not considered as an alternate here.
Solar energy devices are normally dismissed as
requiring large investments to achieve long-term
returns (Hayes, 1979; Natl. Res. Counc. 1980); only
passive collectors for direct heating are thought to be
feasible for immediate use. Furthermore, it is possible
that, through the year 2000, the energy required to
manufacture solar devices may approximate the energy
produced from such devices (Whipple, 1980); pre-
sumably this energy would be from coal or nuclear
fuels. High initial energy consumption would be
required to develop solar energy, but over its lifetime
the solar device would engender fewer emissions than
would other sources; the energy payoff period would be
approximately 10 years.
Other novel sources of energy such as wind, solar
electric, ocean thermal, and tides are presently ruled
out because their costs are estimated to be approxi-
mately 10 times that of available sources of energy
(Natl. Res. Counc. 1980); however, environmental
costs are not considered in these estimates.
Obviously, there is a balance between investment in
industrial innovation to minimize emissions and the
scrubbing costs of these emissions. It is not clear that
the costs of reducing emissions either by scrubbing or
by minimizing energy used have been carefully
considered in developing most energy productions. It
would be desirable to focus specifically on the overall
abatement aspects of these projections in context to
the whole, to test the effect of various mechanisms
such as price, emission regulations, and alternate
energy sources on the kinds and amounts of emissions.
For example, according to projections, increasing
energy prices fourfold will not modify the kind of energy
or the amount used in the industrial sector as much as
one might imagine. Moving into the real world, one
wonders what investments in innovative technology in
fuel supplies (i.e., solar) and in industrial technology
might be brought on by such a large price change.
Energy price and conservation seem to be important
factors in energy projections for buildings to the year
2000 to 2010. The CONAES models suggest energy
consumption in buildings will decrease by 60 percent
with a doubling of price adjusted for inflation. Presently
enacted conservation programs which include stan-
dards for appliances, thermal performance for new
construction, and retrofitting have been projected to
decrease energy by 20 percent (Hirst and Mammon,
1979). Other important considerations re energy
consumption in buildings will be architectural design
and the development of district heating systems.
Architectural design must minimize space per function
and concentrate upon closed space rather than open
space design; the latter permits the efficient use of
computer controlled zone heating and cooling systems.
Many projections suggest 50 percent of building
energy use will be from purchased electricity. If this is
so, a marked improvement in energy efficiency (about
50 percent from Swedish experiences) can be attained
from implementing district heating systems in con-
junction with electrical generating plants. Research
and incentives toward these ends are needed.
Decreases in energy used for transportation appear
to be equally sensitive to price increases and
conservation efforts. A doubling of price with conserva-
tion effort would decrease energy consumption by
about 50 percent. Oil is projected to be the only
significant fuel in 2000 for transportation.
In summary, a doubling of price with probable
conservation will markedly reduce energy consumption
in buildings and transportation. It is not clear what
effect other energy reducing activities, especially
design, will have on energy consumption. Transporta-
tion will depend entirely upon oil, and buildings will be
50 percent or more dependent upon coal. Projections of
industrial energy consumption, however, appear to be
less sensitive to price and conservation, and a switch to
coal would probably at least double emissions. Again it
is not clear how factors such as waste cycling of energy
intensive material, increased product quality, tax
structures and rebates, technological innovation,
trends in consumer demand, and the development of a
conservor ethic will affect industrial energy consump-
tion and coal consumption. It is conceivable that
industrial energy consumption could pose an increased
emission threat when total energy demand decreases.
Suggested research topics and demonstrations fall
into two groups: Modeling with emphasis on environ-
mental impact, and technical studies. A total energy
modeling effort is needed for various scenarios,
focusing on environmental emissions rather than
overall energy demand. An important area to examine
in detail is the industrial sector. Specific activities
should be broken down as to probable emissions, and
scenarios developed based upon alternate energy
sources, alternate product demand, and possible
technological innovation, particularly in material sub-
stitution. The assumption that coal will replace oil and
gas should be examined critically as this assumption
appears to be a major determinant of projected
emissions. In addition, special incentives such as
taxation (on coal) and tax rebate (for pollution emission)
should be considered as to their bearing on emission
reductions. Finally, a scenario focusing on environ-
mental emissions needs to be developed for a
"conservor' society with changed lifestyles; presum-
ably the results of this study would project minimum
emissions in the future.
Other research programs involve research and
demonstration studies. It is suggested that demonstra-
tion studies be considered as "contests" in which
individuals, industry, and communities could compete.
The rewards could be tax rebates or direct funding.
-------
THE ACID RAIN PROBLEM: MECHANISM AND EFFECTS
477
Some areas of effort include (1) dwelling and building
design emphasizing space factors; (2) district heating
development; (3) designing and using various tax
rebates to promote energy efficiency for individuals,
industry, and communites (Hirst, 1979); (4) community
and industrial design development to minimize trans-
portation requirements; and (5) replacement of energy
intensive and cold-dependent processes, products, and
uses in home and industry with alternates.
There appears to be a need for more research
emphasis on coal and coal wastes with the focus on
environmental abatement. The feasibility of using fly
ash and other wastes for sulfur gas scrubbing needs to
be studied.
These are but a few suggestions for research and
development projects. The main thrust here is to focus
on atmospheric emission reduction rather than energy
conservation as a goal. Pricing, alternate approaches,
and utilization of waste (Spilhous, 1970) need to be
emphasized in context to existing studies. Most of the
implementation of the above proposals is long term;
however, the market place can respond amazingly fast,
given the right incentives.
ATMOSPHERIC POLLUTANT
TRANSPORT
Long range transport is the particular phenomenon
associated with acid aerosols and acid precipitation.
Particular reference has been given to tall stacks in this
regard. It is important to ascertain the sensitivity of
stack height in a particular setting to the development
of acid aerosols and transport. By defining the key
sources, this sensitivity could be used to develop a
priority for abatement by scrubbing.
Modeling appears to be the best approach to
developing a stack height sensitivity and acid aerosol
formation and transport (Fischer, 1979). Lofting factors
which include vertical stratification, thermal rise
diffusion, and transport relative to the kinetics of SOz.
oxidation need to be emphasized. The master variable
given a specific setting would be stack height.
This research is needed now. There is sufficient
technical information and models available to carry out
the effort. Calibration of the model using aircraft should
be carried out.
CHEMICAL TREATMENT
Chemical modification of acid lakes from treatment is
a short-term abatement effort. It is generally suitable
for research purposes, but it may be feasible for certain
other lakes also. Studies have focused on a specific
lake except in Norway where parts of rivers have also
been treated. Hydrologic modifications using ground
water may be feasible in certain areas where a more
permanent acidification abatement may result.
Basic materials (Ca(OH)2, CaCO3, CaMg(CO3)2,Ca,
Mg-silicates have been added to lakes and the
surrounding land to increase the pH and the acid
neutralizing capacity. Various neutralizing materials
have been studied (Grahn and Hultberg, 1975); these
materials have generally beenCa(OH)2 orCaCO3 and
they are usually added directly to the lake (Scheider
and Dillon, 1976; Wiklander, et al, 1972). These studies
have used one lake or a small portion of a drainage
system.
The chemical treatment of acid lakes can be
considered as a titration of water and surrounding
sediment in contact with the water. Treatment of water
only is temporal due to the runoff of the water
containing the acid neutralizing capacity. A more
efficient and permanent treatment would be to treat
the surrounding soil to increase the base saturation. In
this case the acid neutralizing capacity stays in the soil
to react with the acid precipitation.
Many drainage basins with soils of low exchange
capacity have the upper 10 m or so depleted in base
saturation, but deeper layers may contain base
saturation or acid neutralizing capacity (above pH 5.5
critical to lakes). This phenomenon results from the
depletion of base saturation by the acid weathering of
soils over the past 10,000 years or so. In this situation,
there may be a large acid neutralizing capacity of soils
at depth, but there is little contact of surface waters
with these deeper soils. Therefore, hydrologic engi-
neering can cause permeation of deeper soil zones and
the neutralization of acid precipitation. This kind of
treatment would consider an entire drainage basin in
most cases. It is worthwhile noting that little
information is known about the chemical properties of
soils below the surface layer.
OL
Q_
50 100
QUADS
150
Figure 1. — Total energy demand from CONAES scenarios
for 2010. + — A* scenario (4 times price plus conservation).
O — B' scenario, 3 percent GNP growth. Compare to 1975
energy demand.
a:
a. .
un '
r^
en
05 '
UJ
+v
I 1975
I . 1
\
10
20
QUADS
Figure 2. — Coal energy demand. Symbols same as in figure
1, except primes refer to industry. Note increase in A*
estimate (+) over present demand.
-------
478
RESTORATION OF LAKES AND INLAND WATERS
SUMMARY
Energy conservation may not be directly coupled to
emission decreases. Research using energy models
focusing on emissions and sources of emissions is
needed. Variations in the amount and kind of energy
used and the resulting emissions can probably equal
the 90 percent reduction attainable by scrubbing of
emissions. Contests are suggested to implement
innovation and change.
Research on emission scrubbing should focus on the
use of solid wastes from coal as well as the exploitation
of metals in these wastes.
Research on short-term abatement can identify
important emission stacks using modeling to predict
sensitive stack height.
In chemically treating acid waters, aquatic water
systems should focus on the entire drainage basin and
especially soils. Deep soils and sediments have
sufficient neutralizing capacity in many cases to
neutralize acid precipitation.
Tenney, M. S., and W. F. Echelberger, Jr. 1970. Fly ash
utilization in the treatment of polluted waters. W.S. Bur.
Mines circ. 8488:237
U.S. Environmental Protection agency. 1979. Sulfur emis-
sion: control technology and waste management. EPA
600/9-79-019. Off. Energy Minerals Ind. Washington.
1979. Sulfur oxides control in Japan. EPA 600/9-
79-943. Off. Environ. Engi. Technol. Washington.
Whipple, C. 1980. The energy impacts of solar heating.
Science. 208:262.
Wiklander, A., andT. Ahl. 1972. The effects of lime treatment
to a small lake in Bergslagen Sweden. Vatten 5:431.
REFERENCES
Cook, C. S. 1973. Energy: planning for the future. Am. Sci.
61:61.
Ezwald, J. K., and J. V. DePinto. 1978. Recovery of
Adirondack acid lakes with fly ash treatment. Dep. Civil
Environ. Eng. Clarkson College of Technology, Potsdam,
N.Y.
Fisher, B. E. A. 1978. Long-range transport and deposition of
sulfur oxides. Pages 243-296 in J. 0. Nriagu, ed. Sulfur in
the environment. Part 1. The atmospheric cycle. Wiley
Interscience. New York.
Grahn, 0., and H. Hultberg. 1975. The neutralizing capacity of
12 different lime products used for pH adjustment of acid
water. Vatten 8:120.
Hayes, E. T. 1979. Energy resources available to the United
States, 1985 to 2000. Science 203:233.
Hirst, E. 1979. Understanding energy conservation. Science
206:513.
Hirst, E., and B. Hammon. 1979. Effects of energy
conservation in residential and commercial buildings.
Science 205:656.
Hoffman, D. J., and J. M. Rosen. 1980. Stratospheric sulfuric
acid layer: evidence for an anthropogenic component.
Science 208:1368.
Holdren, J. P., K. R. Smith, and G. Morris. 1979. Energy:
calculating the risks (II). Science 204:564.
Marshall, E. 1980. Energy forecasts: sinking to new lows.
Science 208:1353.
National Reserarch Council. 1980. Energy in transition 1985-
2010. Comm Nucl. Alternate Energy Sys. Final rep. W. H.
Freeman and Co., San Francisco, Calif.
Ruether, J. A. 1979. Chemical coal cleaning. Science
205:540.
Scheider, W., and P. J. Dillon. 1976. Neutralization and
fertilization of acidified lakes near Sudbury, Ontario. Pages
93-100 inProc. 11 th Can. Symp. Water Pollut. Res. Can. 11.
Spilhaus, A. 1970. The next industrial revolution. Science
167:1673.
Talbot, R. W., M. A. Anderson, and A. W. Andren. 1978.
Qualitative model of heterogeneous equilibria in a fly ash
pond. Environ. Sci. Technol. 12:1056
-------
479
MUTUAL RELATIONSHIP pH/EUTROPHICATION
ACID RAIN
H. L GOLTERMAN
Biology Station
La Tour du Valat le Sambuc
Aries, France
ABSTRACT
In older literature the pH of lake waters was used in attempts to quantify the eutrophication
process. Because most of the data were taken from Swiss lakes the results were really only valid
for hard waters. In addition, the influence of the temperature on the differences between summer
and winter values of pH was not sufficiently taken into account. In eutrophic lakes, two quite
different processes may take place after an increase of the pH value. In hard waters in which
calcium concentration may control the phosphate solubility the formation of apatite (calcium
phosphate) will counteract the eutrophication by withdrawal of phosphate from solution. If, on the
other hand, the phosphate concentration is controlled by ferric hydroxide - as suspended clay
component or as free hydrated ferric hydroxide - an increase of the pH may solubilize phosphate
from the sediments, stimulating eutrophication. Two interesting processes make a theoretical
approach of the calcium carbonate system extremely difficult: (a) the occurrence ofCaCOs
supersaturation has been known for a long time. Recently, however, it was found, that the degree
of supersaturation is related to the pH. (b) Diffusion of CO:into the lake seems to be a much more
complicated process than simple models predicted, as wind stress and microstratificatipn in the
lake depend more on physical-climatic factors than can be theoretically quantified. The
combination of the pH-eutrophication relationship with the acid rain problem causes interesting
thoughts for speculation. In poorly buffered, soft waters, the increase in acidity will decrease the
availability of carbon. The waters may easily become carbon limited, especially if heavily fertilized,
unjustifiably reviving the old carbon-phosphate controversy. In hard waters, the acid rain may
decrease the pH value relatively little: but such a pH decrease may render the apatite more soluble
and thus more available. Increased acidity in rain may in certain rock areas or types increase
phosphate erosion.
For the complete paper, please contact Dr. Golterman at the following address:
Dr. H. L. Golterman
Station Biologique D
La Tour du Valat le Sambuc
F-13200 Aries, France
Phone: (90)98. 90. 13
-------
480
AN EVALUATION OF METHODS FOR MEASURING THE
GROUNDWATER CONTRIBUTION TO PERCH LAKE
DAVID ROBERT LEE
PETER J. BARRY
Atomic Energy of Canada Limited
Chalk River, Ontario, Canada
ABSTRACT
Efforts to determine elemental fluxes within lakes have often been limited by incomplete
knowledge of inputs and outputs of water and solutes. The major problem in determining lake
budgets is the ground water component. In this paper we summarize the results of five methods
used to estimate the volumetric flow of ground water to a small (45 hectare) lake in Eastern
Ontario. The methods are : (1) a water-budget method, which uses estimates of evaporation,
change in storage, and surface inflows and outflows to calculate the net groundwater inflow; (2) a
Darcy approach involving estimates of hydraulic gradient, hydraulic conductivity, and area of
aquifers communicating with the lake; (3) a chemical method, which is based on the isotopic
dissimilarity between ground and-surface water; (4) a point-dilution method, which is similar to the
Darcy approach but uses groundwater velocity rather than hydraulic conductivity and gradient; and
(5) a seepage-meter method, in which fluxes across the sediment-water interface are measured
directly. None of the methods is completely satisfactory, and in most cases, investigators may have
to use more than one approach. None of these methods adequately solves the problem of nutrient
exchanges between ground water and lakes because of the reactivity of most nutrients with the
sediments through which seepage occurs.
INTRODUCTION
The "invisible" contribution of ground water is the
most difficult to measure, and sometimes the most
mysterious, of all the components of a lake water
budget. When attention is given to surface water
quality, water resource managers often require
information on the nature and volume of all inflows and
outflows. Few managers can shake the gnawing
feeling that the groundwater component is inade-
quately understood. Indeed, some workers (Uttormark,
1974; Lee et al. 1980) have suggested that ground
water may flush nutrients from sediments into the
overlying water. And Cartwright, et al. (1979) have
found upward groundwater flow potentials in the
littoral and pelagic sediments of Lake Michigan. Where
contaminants are present in ground water (as Love
Canal publicity reminds us) concerns are directed
toward finding and monitoring groundwater discharge
locations.
Our purpose here is to present an evaluation of work
done on the groundwater component of a small lake
(0.45 km2) in eastern Ontario. This evaluation is base_d
largely on work presented in detail elsewhere (Barry,
1975).
The methods compared in this report are:
1. Water budget from surface hydrology measure-
ments.
2. Classical hydrogeologic method based on the
Darcy equation.
3. Stable isotope ratios.
4. Point dilution.
5. Seepage-meters/mini-piezometers.
Table 1 summarizes the equipment needed and the
variables to be measured for each of these methods.
Surface hydrology
may be written:
AS = I O + P-E + G
The water balance equation
(eq. 1}
where AS is the change in amount of water stored in
the lake
I the surface-water inflow
0 the surface-water outflow
P the precipitation
E the evaporation and
G the net groundwater flow.
Because all terms except E and G are directly
measurable, independent estimates of either E or G
make it possible to estimate the other. For Perch Lake,
with values of E available from detailed energy-budget
measurements (Barry, et al. 1979), the water budget
was used to calculate G. Unfortunately, the available
values of E were obtained when the lake was free of ice
(i.e., May-October). An evaporation value (10.6 cm-yr"1)
for the period of ice cover was obtained from Bruce and
Weisman (1967) and from the assumption (also from
Bruce and Weisman, 1967) that the long-term average
annual evaporation equals 1.18 times the open-season
evaporation.
Average water budget figures for Perch Lake (Table
2) indicate that the annual groundwater inflow(2.89 x
105m3) is of the same order of magnitude as the direct
precipitation (3.63 x 105m3) and the evaporation (3.14 x
105m3). Ground water contributes 14 percent of the
total annual inflow.
-------
SPECIAL TOPICS
481
Table 1. — Equipment and variable requirements for the five methods.
Streamflow weirs, net radiometers,
hygrothermographs, rain and snow
gages, water temperature probes,
water-level
Observation wells and piezometers,
drilling rig, water pumps, water-
level recorders
Streamflow weirs, evaporation pan,
rain gage, piezometers, mass
spectrometer
Drilling rig, borehole wells,
tracers, tracer-dectector, down hole
mixing pump
Seepage-meter", drive casing
hammer, drive points, plastic
tubing, plastic bags
water budget
classical hydrogeologic
stable isotope
point dilution
seepage-meter/mini-
piezometer
continuous records of Streamflow, precipitation,
evaporation, lake level, lake morphometry, vertical
profiles of water temperature, relative humidity,
air temperature, net radiation
weekly groundwater levels, estimates of hydraulic
conductivity and aquifer thickness around the lake
18O/16O ratio Streamflow, ground water
precipitation, evaporation pan, and lake
distribution of the volumetric flux of ground water,
aquifer thickness around the lake
groundwater discharge distribution across lakebed,
hydraulic conductivity
'Described in text
The net groundwater inflows on a monthly basis are
shown in Figure 1. With spring rains and snowmelt,
groundwater flow increases rapidly and reaches a peak
in April. During May, evaporation increases as the
vegetation leafs out and groundwater flow declines
through the summer. In September with the onset of
autumn rains, the first killing frost, and declining
vegetation, groundwater rates begin to increase. They
reach a maximum in November when recharge ceases
as rain gives way to snow. Surface flows display similar
seasonal changes, probably for the same reasons.
However, the contribution of ground water to the total
inflow to the lake varies from a low of less than 10
percent in May to a high of 30 percent in August and
September.
Surface hydrology provides the most obvious and
most readily accepted method of estimating the
groundwater component. However, this method suffers
from several problems:
1. Only the net groundwater flow is determined. It
tells nothing about the source areas or significant
leakages through the lakebed.
2. The groundwater flow is a residual term which, for
many lakes, is represented as a small difference
between large numbers.
3. Energy budgets, required for the evaporation term,
and surface flow monitoring are expensive.
Hydrogeologic or Darcy approach — Darcy's
equation may be written:
Ah
Q = KA—
Ax
where Q is the volumetric groundwater flow, A, the
cross-sectional area of the flow path, h/ x, the
hydraulic gradient along the flow path, and K, hydraulic
conductivity of the geologic material.
Determining the pattern and rates of groundwater
flow into a lake requires knowledge Of (1) the vertical
cross-sectional area of aquifer materials that transmit
water to the lake, and (2) the distribution of hydraulic
head in that vertical cross section. Hydraulic head is
measured with piezometers and observation wells. In
the example shown in Figure 2, the piezometer water
levels are higher with greater depth. This indicates
there is an upward component in the groundwater
velocity at the site. Decreasing water levels with depth
would indicate downward flow.
I-
: V \
SEPT OCT NDV DEC JAN FEB MAR APR MAY JUN JUL AUG
MONTH
Figure 1. — Monthly groundwater flow to Perch Lake from
surface water budgets 1970-1977.
In the Perch Lake study it was possible to employ the
classical hydrogeologic or Darcy method on the sub-
basin aquifer contributing a substantial share of
ground water to the lake. Hydraulic conductivities
obtained by several standard methods at over 65
locations in a land area of 0.5 km2, were judged to have
upper and lower estimates of 1 x 1CT3 to 5 x 10~3 cm-s~1
for the sands and
basal silt and clay.
-------
482
RESTORATION OF LUKES AND INLAND WATERS
. _WATER ^"^
: TABLE
LAKE
PIEZOMETER
SCREEN —_
Figure 2. — Vertical section in the line of groundwater flow at
the lakeshore. Dark horizontal bars in piezometers show water
levels (hydraulic head) at the piezometer screens. A situation of
increasing head with depth indicates groundwater flow
potential into the lake.
Average annual hydraulic gradients were deter-
mined for various parts of the section. The representa-
tive gradients ranged from 0 to 0.034. The total
calculated flow had an upper value of 364 m3-d~' (or
0.91 mm-d~1 over lake surface) and a lower value of 73
m3-d~1 (0.18 mm-d"1). Even with the high sampling
density used, data were insufficient to assign different
gradients to different seasons or months.
The major difficulty with the classical hydrogeologic
approach (Darcy method) is determining the magnitude
and spatial variation of hydraulic conductivity. Al-
though the work of Cherry, et al. (1975) at Perch Lake
was one of the most thorough groundwater/lake
investigations in an area of its size, it resulted in
estimates of groundwater discharge that varied by a
factor of 5.
Isotope method — Where the ground water is
chemically different from the surface waters into which
it flows, groundwater flow can often be estimated. The
technique used at Perch Lake is based on the idea that
the heavy isotope of oxygen, O, which is naturally
present in water as H218O can vary relative to the
H O because of different rates of evaporation and
condensation. These processes cause isotopic frac-
tionation that results in "fingerprinting" different
water masses according to their 18O/16O ratios. At
Perch Lake the ground water is isotopically lighter
(higher H216O to H218O, ratio) than the lake water.
Different sources of ground water also differ in their
O enrichment. However the major uncertainty in the
Perch Lake study (Welhan and Fritz, 1977) lay in the
18O/16O ratio of water evaporating from trie lake.
If the groundwater flow is small (as in Perch Lake),
the isotope method probably has an error of up to 50
percent. Other obvious problems with the isotope
method are:
1. The necessity for having a uniform isotopic ratio in
ground water or a way to assign proportional amounts
of groundwater inflow from isotopically different zones.
2. The necessity for a fairly large number of sampling
points and information on geologic units that transmit
water to the lake.
3. A sufficiently large isotopic difference between the
surface and ground waters.
4. Of all the parameters that must be measured the
isotopic composition of evaporating moisture is the
most difficult and can result in an uncertainty of more
than 50 percent in the calculated lake evaporation rate
(Zimmerman and Ehhalt, 1970).
Point methods - Two techniques, which complement
both the surface hydrology and the classical hydro-
geologic approaches, are point-dilution measurements
of groundwater velocity and seepage-meter/mini-
piezometer methods. They complement these methods
because they directly measure groundwater flux at
specific points. These techniques are of particular
interest where there are known or suspected sources
of onshore groundwater contamination or where a
lakebed-aquifer system is fairly homogeneous. Neither
of these methods has been employed over an area wide
enough in Perch Lake that groundwater inflow
calculations can be given.
Seepage-meter and mini-piezometer methods
This approach relies on the fact that significant
groundwater inflows tend to occur through sediment
(peats, sands, gravels) in shallow nearshore areas. A
seepage meter is a cylindrical enclosure on the lakebed
to which a deflated submerged plastic bag is attached
(Lee, 1977). Where groundwater inflow is upward, the
flow is determined by measuring an increase in the
water volume of the bag over a period of time, generally
several hours. In fairly homogeneous systems the
seepage rate declines exponentially with distance
offshore (Lee, 1977). If a smooth pattern of seepage
flux is found, measurement points can be used to
estimate groundwater inflow through an area of
lakebed (Lee, et al. 1980). Mini-piezometers and small
bundle-type samplers are an inexpensive, manual
method, useful for identifying zones of significant
groundwater flow potential. They are also useful for
sampling pore waters in cohesionless sediments of
seepage zones. These samplers are installed simply by
driving a 1/2 inch (nominal) steel pipe to the desired
depth (4 m maximum), inserting the plastic sampling
tube(s), and withdrawing the pipe (Lee, et al. 1980). As
shown in Figure 2, zones of upward flow can be
identified once equilibrium water levels are reached.
Point dilution measurements of groundwater flow
- A critical review of this method was given by Halevy,
et al. (1966). The technique consists of labeling the
water in a well screen with a tracer and observing its
rate of dilution. If the tracer solution is well mixed, the
slope of the dilution curve (log concentration vs time)
gives the rate of apparent groundwater flow.
The speed of the groundwater can be related to the
rate of dilution through the equation:
Vi =
V
crFt
In C/Co
-------
SPECIAL TOPICS
483
where v is the volumetric flux of the water through the
screen
v the dilution volume
F the cross section of the well screen
t the time from the beginning of measurement
C the original concentration
C the observed concentration at t and
a correction factor for distortion of flow by the well
screen.
The value of the point dilution technique can be
illustrated by noting that apparently homogeneous
sands can conduct groundwater flow at rates that vary
by a factor of 5 (Pickens, et al. 1977). In most cases the
methods for measuring hydraulic conductivity are not
as sensitive as the point dilution technique.
Table 2. — Perch Lake water budget for 1970-77.
Source
Surface streams
tt 1
#2
f 3
f 4
f 5
Surface stream inflow
Precipitation
TOTAL IN
Surface stream outflow
Evaporation
TOTAL OUT
Net ground water
(TOTAL IN — TOTAL OUT)
Volume (x105m3 • yr"1)
2.23
9.70
1.51
0.84
0.29
14.57
3.63 (s 0.807 cm)
18.20
17.95
3.14 (= 0.698cm)
21.09
2.89
Note: Evaporation measurements are from energy-budget
calculations (Barry, et al. 1979).
Clearly this method and the seepage-meter method
obviate the necessity of separately determining the
hydraulic conductivity and the gradient. The point
dilution method cannot distinguish readily the vertical
and horizontal components of flow, nor does it indicate
flow direction.
The fundamental limitation of point methods is the
need to interpolate between bore holes or measure-
ment locations on the lakebed since there are practical
limitations on density of sampling points. In many lake
settings, successful application of point methods in
estimating groundwater inflow or outflow will probably
require methods for characterizing aquifers by rapid
remote sensing techniques.
COMPARISON OF METHODS
Table 3 provides a basis for comparing groundwater
inflow estimates for the same 4-month period. The
point dilution and Darcy estimates are constant
because the calculations were based simply on average
annual or representative values. These two methods
were used on the northern side of Perch Lake, not the
whole lake perimeter, so the estimate is expected to be
low. Both the water budget and the stable isotope
methods appear to agree. However, the uncertainties
in the stable isotope method are large (±50 percent)
relative to those of the water budget method (±10
percept) so the agreement may be coincidental.
Because the northern side of the lake is soft peat, it
would have been difficult to employ seepage meters
there.
Table 3. — Comparison of estimates of groundwater inflow to
the lake for the period May through September, 1973.
Groundwater inflow to the lake in mm-da"1*
Method
May June July August September
Water budget
Stable isotope
Point dilution
Darcy0
2.0
1.0
0.9
5.5
6.4
1.0
0.9
1.6
0.35
1.0
0.9
1.2
1.2
1.0
0.9
0.43
1.0
0.9
'Values are daily volumetric water flow into the lake divided by lake
surface area
"The groundwater inflow from the Darcy method is 0.2 or 0.9 for the
lower and upper estimates of hydraulic conductivity.
CONCLUSION
There are essentially two types of lake/groundwater
flux methods: Gross measurements which, by their
nature, are averaged over large areas (the stable
isotope and surface hydrology methods), and syntheses
made from many point measurements (the seepage-
meter/mini-piezometer, the classical hydrologic, and
the point dilution methods). The major limitation of
point measurements is geologic complexity which
contributes to a wide spatial variation in flow rate. A
most promising technique to allow interpolations
between sampling points is the use of ground-probing
radar that "sees" into the subsurface, particularly in
coarse-grained soils (Annan and Davis, 1976).
The choice of methods will depend on:
1. The type of information needed;
2. The features of the study site, e.g., the presence of
access roads or stream gaging structures; and
3. The manpower, skills, and equipment available.
All methods for groundwater measurement are
expensive but it is unrealistic to appraise any costs
until the study area and requirements are known. All
methods require long times to get a stable mean flux.
Precipitation, for example, varies widely from year to
year. Changes in storage of energy and water in lakes
are subject to considerable error on the short term (1 or
2 days).
One of the lessons at Perch Lake has been that it was
necessary to compare methods if we were to avoid
being deluded into thinking that one method gave a
correct measurement. It was often necessary to go
through an iterative process of checking one method
against the other. The significance of ground water to
lake processes should be studied by as many methods
as feasible. With an environmental variable as complex
as ground water, it could be quite misleading to put
complete faith in any single method. Internal con-
sistency among various methods has provided a basis
for confidence. But none of the methods addresses the
problem of nutrient fluxes due to ground water/surface
water interaction.
-------
484 RESTORATION OF LAKES AND INLAND WATERS
REFERENCES
Annan, A. P., and J. L. Davis. 1976. Impulse radar sounding in
permafrost. Radio Sci. 11:383.
Barry, P J., ed. 1975. Hydrological studies on a small basin
on the Canadian Shield. Atomic Energy Can. Ltd., Publ.
5041/1, II.
Barry, P. J., et al. 1979. Water and tritium budgets for Perch
Lake, 1970-1977. In P. J. Barry, ed. Hydrological and
geochemical studies in the Perch Lake basin: A second
report of progress. Atomic Energy Can. Ltd., Publ 6404.
Bruce, J. P., and B. Weisman. 1967. Provisional evaporation
maps of Canada. Can. Dep. Transport. Meterolog. Branch,
Circ. 4531. Toronto, Ontario
Cartwnght, K., et al. 1979. Hydraulic potential in Lake
Michigan bottom sediments. Jour. Hydrol. 43.67.
Cherry, J..A., et al. 1975. Physical hydrogeology of the lower
Perch Lake basin. Pages 625-680 in P. J. Barry, ed.
Hydrological studies on a small basin on the Canadian
Shield. Atomic Energy Can. Ltd., Publ. 5040
Halevy, E., et al. 1966. Borehole dilution techniques, a critical
review. Pages 531 -562 in Isotopes in hydrology. Int. Atomic
Energy Assoc., Vienna.
Lee, D. R. 1977. A device for measuring seepage flux in lakes
and estuaries. Limnol. Oceanogr. 22:140.
Lee, D. R , J. A. Cherry, and J. F. Pickens. 1980. Groundwater
transport of a salt tracer through a sandy lakebed. Limnol.
Oceanogr. 25:45.
Pickens, J. F., et al. 1977. Field studies of dispersion in a
shallow sandy aquifer. In Proc. Invitational Well Testing
Symp. Lawrence Berkeley Lab Publ. 7027. University of
California, Berkeley.
Uttormark, P D., J. D. Chaplin, and K. M. Green. 1974.
Estimating nutrient loadings of lakes from non-point
sources. EPA-660-3-74-020. U.S. Environ. Prot. Agency.
Welhan, J. A., and P Fritz. 1977. Evaporation pan isotopic
behavior as an index of isotopic evaporation conditions
Geochim. Cosmochim. Acta 41:682.
Zimmerman, U., and D. H. Ehhalt. 1970. Stable isotopes in
the study of the water balance of Lake Neusiedl, Austria.
Proc. IAEA Symp. Isotope Hydrology, Vienna
-------
485
REHABILITATION PROJECT FOR A QUEBEC LAKE:
WATERLOO LAKE, NEAR MONTREAL
FRANCOIS J. GUIMONT
Ministry of the Environment
Quebec, Canada
ABSTRACT
Lake Waterloo, a small shallow eutrophic lake (A, 1.5 km2; Z 2.9) , js situated in the southeastern
region of the Province of Quebec. This lake, which possesses a very small watershed (31.5 km2),
has been the object of a restoration program by the Ministry of the Environment since 1976. This
program is divided into two distinct operations. The first one consists of installing an aeration
system of the diffuser type; at the present time, it has been in continuous operation for 4 years.
During this period no winterkill episodes have been observed because of the concomitant increase
of mean dissolved oxygen values under ice cover (bottom: 8.0 mg I"1).Parameters mostly affected
by aeration were total iron (mean decrease 95 percent), total manganese (mean decrease 55
percent), ammonia (mean decrease 55 percent), and total phosphorus (mean decrease 42 percent).
A study of the phytoplankton populations has shown a marked transition since 1978, from the
previous dominant Cyanophycea towards Bacillariophycae. The second part of the overall project is
a study of the total phosphorus input to the lake (1963 kg P yr~'). Industrial and urban activities
account for 50 percent of this and should be eliminated by 1982. Total restoration of this water
body needs a supplementary method to diminish residual phosphorus originating from the
watershed (884 kg P yr~1). In this particular case, dredging the lake sediment (thickness 5.8 m) is
the only way to achieve this objective even if the investment seems infeasible on a short-term
basis.
INTRODUCTION
The installation of an aeration system in 1976 at
Lake Waterloo was ecological intervention urgently
needed to eliminate further winterkills. The continua-
tion of this intervention is justified because artificial
aeration of lakes can improve water quality and extend
the vertical distribution of the biota. Numerous studies
have shown an increase of dissolved oxygen concen-
trations (Irwin, et al. 1966; Haynes 1971). Noticeable
decreases in the concentration of manganese and iron
(Wirth and Dunst 1967; Haynes J971), ammonia
(Symons, etal. 1967) and hydrogen siilfide (Irwin, etal.
1966; Leach and Harlin 1970) have been observed in
the deepest portion of such lakes. Changes in the
biological populations have been observed in artificially
mixed lakes, including a decrease of the phytoplankton
populations (blue-green algal biomass. Anon. 1971;
Malueg, et al. 1971), an extension of the vertical
distribution of zooplankton (Fast 1971), and an increase
in number and speciation of the benthic macroin-
vertebrates. The second phase of the program is
synthetic and corresponds to the estimation of the
allochtonous and autochtonous phosphorus budget.
This should allow us to determine the restoration
techniques having the highest probability of success.
METHODS
Aeration
Figure 1 shows the locations of the diffusers, the
sampling station used to interpret the physico-
chemical data, as well as certain morphometric
parameters. The samples were collected monthly
during winter and bi-monthly for the summer, at the
surface (0.5 m) and at the bottom (3.5m), and analyzed
using standard government laboratory methods. The
results of the physico-chemical parameters were
analyzed for every year of the aerator's operation. The
efficiency of the aeration device was determined by
comparing the means for 1975 (before aeration) with
those for the period 1976 to. 1979 inclusively. The
statistical significance of the differences between the
means was determined using the student t test.
Phosphorus budget
The phosphorus budget was estimated by three
different methods. The first was an indirect estimation
of the phosphorus inputs using available data from
existing land use maps. The choice of the phosphorus
exportation coefficients is in agreement with the
literature (e.g; urban zones — 105 kg P krrr2 yr1, Potvin
1976; swamps — 25 kg P km-2 yr-1, Uttomark, et al.
1978) and permits a preliminary quantification of this
watershed nutrient output. The second is more direct
because it evaluates in situ the total phosphorus load
from the tributaries discharging into the lake as well as
anthropogenic point emissions in the vicinity of the
lake. The data were collected bimonthly between
September 1975 and September 1976 and permitted a
more precise evaluation of the phosphorus load. Lastly,
the autochtonous phosphorus input from the oxygen
deficient (< mg I ') sediments was established using a
releasing coefficient of 8.0 mg P m"2 day~1 (Kamp-
Nielsen, 1974; Fekete, et al. 1976).
-------
486
RESTORATION OF LAKES AND INLAND WATERS
RESULTS AND DISCUSSION
Aeration
Diffuser type aeration systems normally produce
convection currents which destratify water bodies. This
mixing of the water column may provoke an increase in
temperature. From Table 1 we can see that such a
phenomenon has not occurred; on the contrary a
significant cooling is observed (Table 2). The dissolved
oxygen concentration has increased by 20 percent near
the bottom and no oxygen deficit has been detected
since the winter of 1977, eliminating winterkill
episodes. The oxygen saturation levels are similar to
the dissolved oxygen values (Tables 1 and 2). The
transparency of the water column did not change as
expected, because of the mixing by the diffusers. It
would seem that this parameter is influenced by
biological populations such as phytoplankton and
zooplankton. The lowering of the pH is significant for
1976-1979 period and has occurred in the entire water
column (Tables 1 and 2).
These results were predictable if we consider the fact
that fermentation processes have been replaced by
heterotrophic oxidation, producing COs. The increase in
the concentration of the CCh is attributable to the
nitrification of ammonia and the oxidation of sulfates.
This phenomenon reduces pH values which is less
perceptible within the upper water layer because of its
bioassimilation by the phytoplankton. The soluble iron
concentration increased for the aeration period (Table
1) but these differences are statistically non-significant
(Table 2)., The change of the soluble ferrous ions
previously released from anoxic sediments into an
insoluble ferric hydroxide (Fe(OH3» has resulted in a
marked decrease (95 percent) of the total iron
concentration of the water-sediment interface. Low
manganese values be they soluble or total are
significant (Table 2).
As in the case of iron and sulfates an increase in the
redox potential (En) has caused the precipitation of
compounds such as manganese carbonate (MnCOa),
manganese sulfide (Mn S), and manganese hydroxides
(Mn (OH2). Magnesium did not show major fluctuations
which is understandable since it rarely precipitates out.
The organic phosphorus concentration remained
constant during the aeration period. This is interesting
since the transition from fermentation to oxydation of
the organic matter did not increase the concentration
of this labile substance.
Table 1 indicates that before aeration an active
unidirectional flux of inorganic phosphorus resulted in
a continuous enrichment of this water body. The
inorganic phosphorus flux was stopped and probably
reversed with the regeneration of an oxidized micro-
zone at the water-sediment interface. The formation of
[f>]
x,VILCM
[P]XVILCM * [P],
1-K/Tw
Cr*] i i/ii rM - PREDICTED CONCENTRATION
X , VILLH - OF TOTAL PHOSPHORUS
[ P ] , = PHOSPHORUS INFLOW CONCENTRATION
To» = WATER RESIDENCE TIME
Figure 2. — Probability of a prediction falling within a
particular trophic class.
Diffu»ers
Sampling station
Area' 1,5km2
Perimeter' 9,TO km
Max length > 2,9 km
th< 1.13 km
Volume 4,35* I06m3
Max Depth' 4.90m
Mean Depth 2,9m
Water residence time ,!
Figure 3. — Probabilistic loading plot
showing the logarithm of the predicted
'inflow concentration as a function of the
water residence time. Percentages represent
the certainty of the effectiveness of the inflow
concentration achieving the expected
trophic state.
Figure 1. — Lake Waterloo: Sampling station and morphological data.
-------
SPECIAL TOPICS
487
co-precipitates of phosphates with iron, manganese,
and carbonates resulted in a significant (Table 2)
reduction in the concentration of inorganic phosphorus
for the entire water column. The means (Table 1) for
total phosphorus reflect the pathways controlling the
inorganic and organic phosphorus concentrations.
When anoxic zones were detectable (before 1975),
bacterial nitrification by which ammonia is progress-
ively oxidized into nitrites and nitrates was inhibited.
The concomitant decrease in redox potential resulted in
the accumulation of NH
-------
488
RESTORATION OF LAKES AND INLAND WATERS
inorganic carbon concentration is closely linked to the
increase in dissolved oxygen and has been discussed
previously. As for organic carbon, the small increase of
the surface concentration may be attributed to a
plankton biomass increase. The phytoplankton biomass
shows a non-significant (Table 2) increase for the
aeration period which is not concomitant to the
chlorophyll a values. Since the cellular concentration of
chlorophyll a varies from species to species (Wetzel,
1975) it is normal to observe such results because a
species shift in the phytoplankton population has been
occurring since 1978 (Cyanophycae towards Bacil-
lariophycae, Choquette, 1979).
Phosphorus budget
If we refer to Table 3 we notice that the direct and
indirect methods for determining annual allochtonous
phosphorus inputs are comparable and represent a
load of 1,963kgPyr1 Estimation of the annual
autochtonous phosphorus inputs coming from the
sediments corresponds to approximately 35 percent of
the total load originating from the watershed. The
aeration of Lake Waterloo has theoretically inhibited
nutrient flux from the sediments. Figures 2 and 3 show
the actual trophic state following the probabilistic
expression of Chapra and Reckhow(1979). By 1982,55
percent of the total input (1,079 kg P yr1) will be
eliminated. This cutback of 55 percent will bring the
mean phosphorus concentration into the lake to 45 mg
m-3 which should not produce marked modifications in
the visual aspect of the lake (fig. 2) The penultimate
solution seems to be a lake deepening operation that
could effectively buffer the residual phosphorus
loading (884 kg P yr-1). Dredging 3,000 m3 by suction
would cost approximately $3,000,000 and would
augment the water volume by 23 percent, but this
would not assure a defirntive restoration of Lake
Waterloo (Figure 3: 1 + \/tw = 1.5). Other dredging
techniques are presently being studied to lower the
cost of sediment extraction (e.g.; bulldozer, etc.)
Table 3. — Phosphorus budget in Lake Waterloo.
ASSESSMENT
METHODS*
Loading
kg P yr"1
Indirect Estimation
Direct Evaluation
Sediments
1963
1991
686
REFERENCES
Anonymous. 1971. Artificial destratification in reservoirs, a
committee report. Jour. Am. Water Works Assoc. 63:597.
Chapra, C. S., and K. H. Reckhow. 1979. Expressing the
phosphorus loading concept in probabilistic terms. Jour.
Fish. Res. Board Can. 36:225.
Choquette, S. 1979. Etude des populations phytoplancto-
niques du lac Waterloo depuis la mise en marche du
systeme d'aeration. Gouvernement due Quebec, Ministere
de I'Environnement. Rapport interne.
Fast, A. W. 1971. The effects of artificial aeration on lake
ecology. Ph.D. Thesis. Michigan State University, East
Lansing.
Fekete, D. N., et al. 1976. A bioassay using common
duckweed to evaluate the release of available phosphorus
from pond sediments. Jour. Aquatic Plant Manage. 14:19.
Haynes, R. 1971. Some ecological effects of artificial
circulation on a small eutrophic New Hampshire lake. Ph.D.
Thesis. Unversity of New Hampshire, Durham.
Irwin, E. W., J. M. Symons, and G. G. Robeck. 1966.
Impoundment destratification by mechanical pumping Jour.
San. Eng. Div. 92:21
Kamp-Nielsen, L. 1974. Mud-water exchange of phosphate
and other ions in undisturbed sediment cores and factors
affecting the exchange rates. Arch. Hydrobiol. 73:218.
Leach, L. E., and C. C. Harlin, Jr. 1970. Induced aeration of
small mountain lakes. Natl. Water Quality Control Res.
Program, Region VI. Office of Water Qual. U.S. Environ.
Prot. Agency.
Malueg, K., et al. 1971. Effects of induced aeration upon
stratification and eutrophication processes in an Oregon
Farm pond. Int. Symp. Manmade Lakes, Knoxville, Tenn.
May.
Potvin, P. 1976. Relation entre I'etat trophique d'un lac et
I'utilsation du territoire dans son bassin versant. These de
maitrise. INRS-Eau.
Provencher, M., B. Belanger and H. Durocher. 1979.
Caracterisation de la qualite de I'eau de la riviere Yamaska-
Nord: Rapport complementaire. QE. - 41.
Symons, J. J., W. H. Irwin, and G. G. Robeck. 1967.
Impoundment water quality changes caused by mixing.
Jour. San. Eng. Div. Proc. Am. Soc. Civ. Eng. 93:1-20.
Uttomark, P. D., J. D. Chapin, and K. M. Green. 1974.
Estimating nutrient loading of lakes from nonpoint sources.
EPA, 669/3-74-020. U.S. Environ. Prot. Agency.
Wetzel, G. R. 1975. Limnology. W. B. Saunders Co. New York.
Wirth, T. L., and R. C. Dunst. 1967. Limnological changes
resulting from artificial destratification and aeration of an
impoundment. Wis. Conserv. Dep. Res. Rep. 22.
SOURCES
IMPORTANCE
%
Industrial
Stockbreeding
Fertilizer
Domestic uses
Forests & Rain
21
18
3
37
21
Provencher et al., 1979.
-------
489
QUANTIFICATION OF ALLOCHTHONOUS ORGANIC INPUT
TO CHEROKEE RESERVOIR: IMPLICATIONS FOR
HYPOLIMNETIC OXYGEN DEPLETIONS
RICHARD C. YOUNG
W. MICHAEL DENNIS
NEIL E. CARRIKER
Division of Water Resources
Tennessee Valley Authority
Muscle Shoals, Alabama
ABSTRACT
Cherokee Reservoir was created by the Tennessee Valley Authority in 1947 as a multipurpose
reservoir to provide flood control, power generation, and recreation. It is the largest of five TVA
impoundments in the Holston River Basin of upper East Tennessee. Water releases from Cherokee
Reservoir have been documented to be low in dissolved oxygen content since 1950. From 1970 to
1978, water released during power generation had less than 5.0 mg/1 DO, an average of 149 days
per year, and less than 1.0 mg/l an average of 45 days per year. Previous studies strongly suggest
that inputs of allochthonous organic material originating in the highly productive aquatic
macrophyte beds in the Holston River below Kingsport, Tenn. adversely impact hypolimnetic DO
concentrations in Cherokee Reservoir. The present study indicates that the annual average net
primary production of aquatic plants in the Holston River above Cherokee Reservoir is 16.6 metric
tons/ha/year (dry weight), a rate much higher than reported for rivers in the temperate regions of
North America. Biomass contribution from this reach of river is estimated at 4,570 metric tons dry
weight to Cherokee Reservoir annually. Deposition, subsequent decomposition, and nutrient
release from this large amount of allochthonous aquatic macrophyte input represents 94 metric
tons nitrogen, 10 metric tons phosphorus, and 4,570 metric tons biochemical oxygen demand, all
significant factors in hypolimnetic oxygen depletion in Cherokee Reservoir.
INTRODUCTION
Cherokee Reservoir was created in 1941 as a
multipurpose reservoir to provide flood control, power
generation, and recreation. It is the largest of five
Tennessee Valley Authority impoundments in the
Holston River Basin northeast of Knoxville, Tenn.
(Figure 1). The reservoir has a surface area of 131 km2,
an average depth of 15 meters and a mean hydraulic
retention time of 178 days. Thermal stratification
begins in mid-April and the use of hypolimnetic waters
for power generation results in the reservoir becoming
isothermal (24°C) by late summer. During early to
midsummer, the reservoir characteristically has a
shallow, highly productive epiliminion and a thick,
oxygen deficient hypolimnion. Consequently, water
released for power generation has less than 5.0 mg/l
dissolved oxygen an average of 149 days per year and
less than 1.0 mg/l an average of 45 days per year
(1970 to 1978).
The problem of low DO content in water released for
power generation became evident in 1950 (9 years
after closure), when the reservoir discharged water
with a DO content
-------
490
RESTORATION OF LUKES AND INLAND WATERS
Gordon (1971) investigated several different mech-
anisms of oxygen depletion in Cherokee Reservoir. He
concluded that nitrification caused over 50 percent of
the oxygen loss in the hypolimnion for the 1967-1970
period. While these studies did not directly address the
impact of allochthonous organic matter on Cherokee
Reservoir, several observations support the premise
that aquatic macrophytes are a significant factor in
hypolimnetic DO depletion.
Using 1970 data and a computerized reservoir
hydrodynamics model from the Massachusetts Insti-
tute of Technology, Gordon (1971) also determined that
flow to Cherokee Reservoir enters as an interflow;
about 80 percent of the time, from mid-April to late
September. In addition to in situ DO depletion
mechanisms, the well-oxygenated water initially
trapped under the thermocline in the deeper end of the
reservoir at the onset of stratification eventually is
discharged through the power turbines and is replaced
by poorly oxygenated water from upstream reaches and
interflow, thus reservoir hydraulics are a major factor
contributing to the low hypolimnetic DO values in the
lower end of Cherokee Reservoir. Gordon also
demonstrated that oxygen depletion first begins and is
most rapid between river miles 70 and 95 at the upper
end of the reservoir, and that ammonia increased in the
hypolimnion (after DO depletion) as a result of
anaerobic deamination of the highly organic sedi-
ments.
Gordon's conclusions concerning interflow validate
the hypothesis of Churchill and Nicholas who noted "..
.it seems likely from observed river and reservoir
temperature data that some of the intermittently colder
masses of inflowing waters have entered the head of
Cherokee pool as interflows, some possibly entering
below the thermocline." The drift of aquatic plants into
Cherokee Reservoir was related to this thermal
discontinuity by Hall (1966) who observed that
"Between Cherokee boat dock (HRM 93) and the
powerline crossing downstream (HRM 90), the de-
tached, floating plants were lined up more or less
perpendicularly to the axis of the old river or, in other
words, formed more or less a line across the reservoir.
It is wondered if the transverse accumulation of
floating plants represents the approximate location at
which the Holston River 'dives' under Cherokee
Reservoir." Gordon's data and these observations
indicate that the earliest and most rapid DO depletion
occurs in the same area where the thermal disconti-
nuity between inflow waters and Cherokee Reservoir
would allow the deposition of allochthonous organic
material.
The study by Iwanski, et al. (1980) concluded that the
major causes of DO depletion and eutrophication in
Cherokee Reservoir are inflows of phosphorus, nitro-
gen, BOD5, and volatile suspended solids. Using the
Water Quality River-Reservoir Systems (WQRRS)
model (U.S. Corps of Engineers, 1977) anti 1978 data,
Iwanski, et al. conducted a sensitivity analysis and
simulated the effect of these key factors on DO
depletion in Cherokee Reservoir. This analysis showed
that 37.5 percent of the annual DO depletion could
result from inflow detritus (volatile suspended solids)
alone and when combined with the temperature effect
could account for over 60 percent of the annual DO
depletion. Dissolved organic carbon (BOL>5), total
dissolved nitrogen, and total dissolved phosphorus
accounted for 15.0, 6.4, and 5.0 percent respectively,
of the annual DO depletion, according to this model.
These studies strongly suggest that inputs of
allochthonous organic material originating in the
highly productive aquatic macrophyte beds in the
Holston River adversely impact hypolimnetic DO
concentrations in Cherokee Reservoir. However, lack of
adequate data on aquatic macrophyte productivity and
drift characteristics in the Holston River have precluded
accurate assessment of the impact of aquatic macro-
phytes on DO regimes in Cherokee Reservoir. This
paper reports the results of work in progress designed
to obtain this data.
METHODS AND MATERIALS
Study Area
The Holston River is formed by the confluence of the
North and South Fork Holston Rivers at Kingsport,
Tenn. Flow rate in the Holston River is partially
regulated by Fort Patrick Henry Dam, located on the
South Fork Holston River approximately 5 miles above
its confluence with the North Fork. In accordance with
an agreement with the Tennessee Eastman Company,
TVA releases a minimum daily average flow of
21.2m3/sec from the dam to maintain an adequate
water supply for the company. Bihourly flow records
(1978) for the U.S. Geological Survey station at HRM
118.4 show the total flow varies from a minimum of
23.8 mVsec in October and December to a maximum
of 1,209.1 mVsec in March. Instantaneous hourly
flows during late summer and fall can vary from zero to
170 mVsec depending on power generation schedules
and the flow from the unregulated North Fork Holston
River. According to Iwanski, et al. (1980), total waste
loads to the Holston just above Cherokee Reservoir
(HRM 103.4) were measured to be 14,746 kg/day total
nitrogen, 1,145 kg/day total phosphorus, and 10,024
kg/day BODs. Point source discharges accounted for
18 percent of the total nitrogen load, 65 percent of the
total phosphorus load. Land use in the watershed of
Cherokee Reservoir consists of 55 percent forest, 5
percent urban, 37 percent agriculture, and 3 percent
other uses.
The primary study area was from HRM 141.2
(confluence of North and South Fork Holston Rivers) to
HRM 109.1 (backwaters of Cherokee Reservoir). In this
reach the Holston River has a gradient of 0.6
meters/km, surface area of 497.2 hectares, water
depth of 0.3 to 3.5 meters, and a width of 106 to 203
meters. Substrate composition varies from solid rock to
rocky cobble/sand/silt composition with rocky cobble
comprising the major fraction of the substrate at most
locations (usually > 80 percent).
The oxygen content of the Holston River is usually
greater than 5.0 mg/l but may vary as much as 8.0
mg/l diurnally during summer months. Water quality
parameters characterisitic of the study area during
1978 are given in Table 1.
During the growing season (March-October), the
study area is colonized by aquatic macrophytes
including sago pondweed (Potamogeton pectinatus L.),
-------
SPECIAL TOPICS
491
American pondweed (P. nodosus Poir.), curlyleaf
pondweed (P. crispus L), water stargrass (Heteran-
thera dubia, Jacquin, MacM.), eel grass (Vallisneria
americana Michx.), Canadian elodea (Elodea cana-
densis Michx.), and the aquatic mosses (Fissidens
fontanus (B-Pyl.) Steud. and Leptodictyum riparium
(Hedw.) Watnst).
Macrophyte Productivity
In this study incremental change in biomass through
the growing season was determined to estimate
annual net primary production using the assumptions
and model of Fisher and Carpenter (1976). This method
was selected because extensive cropping occurs in the
study area because of large daily variations in stream
flow controlled by the upstream power generation
facility, and the fact that the Fisher and Carpenter
model includes an estimate of mortality prior to
maximum biomass and net production after maximum
biomass. Both of these values are important when
estimating annual net productivity in systems ex-
periencing extensive cropping.
Five sampling stations were selected at areas
encompassing morphological variations in the 51
kilometer study area. These stations were selected
based on interpretations of low altitude aerial
photographs (color infrared) taken in 1977 and field
inspections during 1979. Each station was perm-
anently marked (10 x 10 cm posts or lead marker
weights) and a 20 m x 30 m sampling plot selected. The
plot was graphically divided into 600 potential 1 square
meter sampling points. Thirty 0.1 m2 quadrants were
sampled monthly at each station from April through
August. The sampling points were randomly selected
and located in the field by means of a vector board and
meter tape. Sampling consisted of removing by hand all
macrophytes (including roots) rooted within a square
metal frame having an area of 0.1 m2. The 0.1 m2
samples were placed on ice and returned to the lab
-------
CO
to
Figure 2. — Historical trend in dissolved oxygen cocentrations mg/l from Cherokee Dam.
-------
SPECIAL TOPICS
493
the curve, i.e., September-December, was estimated
based on biomass data for station 3 taken during the
previous fall and winter). The biomass curve was
converted to a rate curve from which the average
annual net productivity was calculated using the
procedure of Fisher and Carpenter (1976).
Using this approach, the cumulative annual net
production for aquatic macrophytes in the 51 kilometer
reach of the Holston River above the Cherokee pool is
estimated to be 16.60 metric tons/hectare/year. This
value is 3.9 times the maximum biomass of 4.27 metric
tons/ha. This indicates that on an annual basis, the
Holston River contributes nearly four times its peak
biomass of .aquatic macrophytes to Cherokee Reservoir.
This turnover rate is greater than the estimate of 1 to 3
times the peak biomass suggested by Westlake(1963)
for submersed aquatic macrophyte communites but
falls within the 0.5 to 5.0 range reported by Rich, et al.
(1971). Considering that the Holston River is nutrient
rich, has clear, shallow waters, and is subjected to
extensive cropping because of fluctuating flows and
velocities associated with operation of the hydro-
electric facility at Fort Patrick Henry Dam, a turnover
rate of 3.9 crops/year seems reasonable.
Annual macrophyte net productivity (16.6 metric
tons/ha/yr) converted to total river biomass (16.6
metric tons/ha/yr x 323.18 ha) is 5,365 metric tons
dry weight. EPA (1978) reported plants from the
Holston River contained 2.06 percent nitrogen and
0.22 percent phosphorus on a dry weight basis.
Assuming these percentages, the average total
nitrogen and total phosphorus bound by aquatic plants
annually is 110 metric tons and 12 metric tons,
respectively. Total organic carbon is estimated to be
2,143 metric tons (AFDW being about 85 percent of
DW and organic carbon being 47 percent of AFDW
(Westlake, 1966). The potential chemical oxygen
demand is estimated to be 5,365 metric tons assuming
that 1 gram DW of plant material equals 1 gram BOD5
(Jewell, 1971).
Discussions
The annual net primary production (16.6 metric
tons/ha/yr) of the Holston River above Cherokee
Reservoir greatly exceeds the 6 metric tons/ha/yr
value for freshwater submersed macrophytes in
temperate climates suggested by Westlake (1963) and
more closely approximates the 17 metric tons/ha/yr
value given for freshwater submerged macrophytes in
a tropical system. The average maximum biomass (427
g/m2) agrees with the data of Peltier and Welch (1968)
who reported an average maximum biomass of 457
g/m2 and 420 g/m2 for two stations located in the
current study area. However, Peltier and Welch (1968)
estimated that the areal plant coverage was only 20
percent. EPA (1972, 1978) estimated the maximum
biomass to be 155 and 200 metric tons, respectively,
for the same reach of the Holston River. This is
approximately an order of magnitude less than the
1,378 metric tons (4.27 metric tons/ha x 323.18 ha)
average maximum biomass reported in this study.
Jewell (1971) reported the average rate of decay of
aquatic weeds to be on the order of 0.086/day at 18°C
(variation 0.05 to 0.19). The average time of travel
between the head of the study reach (HRM 141.1) and
Cherokee pool is 2 to 4 days during the growing season
depending on the flow from the North Fork of the
Holston River and power generation schedules at Fort
Patrick Henry Dam (Ruane and Krenkel, 1978).
Therefore, plant material lost to cropping and floating
unrestricted from the head of the reach would be
reduced by 17 to 34 percent prior to entering the
Cherokee pool. This factor would represent a maximum
since those plants further downstream would ex-
perience shorter times of travel and consequently,
would undergo less decay before reaching Cherokee
Reservoir.
Extensive decay of plant structures within the river
system does occur when plants become impinged on
stumps, tree limbs, bridge pilings, etc. No data were
collected to quantify the amount of plant material
impinged within the river, but from observations, it is
believed to amount to only a small percentage of that
floating in the river at a given time. Dennis (1976)
reports that 88 percent of the detritus in the water
column trapped by 0.1 millimeter mesh screens was
within 20 centimeters of the surface, indicating that 15
to 30 cm/sec velocities are sufficient to keep the plant
material suspended. Therefore, a majority of the
detached aquatic plants probably reach Cherokee pool
without undergoing significant decay.
Previous studies strongly suggest that inputs of
allochthonous organic material originating in the
highly productive aquatic macrophage beds in the
Holston River below Kingsport, Tenn., adversely impact
hypolimnetic DO concentrations in Cherokee Reser-
voir. The present study indicates that the annual
average net primary production of aquatic plants in the
Holston River above Cherokee Reservoir is 16.6 metric
tons/ha/yr, a rate much higher than reported for rivers
in the temperate regions of North America. Assuming
15 percent of the plant material is lost to impingement
and decay, the biomass contribution from this reach of
river is estimated at 4,570 metric tons dry weight
(5,365 x 0.85) to Cherokee Reservoir annually.
Deposition, subsequent decomposition, and nutrient
release from this large amount of allochthonous
aquatic macrophyte input represents 94 metric tons
nitrogen, 10 metric tons phosphorus, 4,570 metric tons
biochemical oxygen demand, all significant factors in
hypolimnetic oxygen depletion in Cherokee Reservoir.
Table 1. — Water quality characteristics of the Holston River above
Cherokee Reservoir (1978).
Parameter
pH (units)
Temperature (°C)
DO (mg/l)
BOD5 (mg/l)
Turbidity (JTU)
Specific conductance (^mohs at 20°C)
Alkalinity (mg/l CaCOs)
Nitrogen, NO2 + NO3 (mg/l)
Nitrogen, NH3 + NH< (mg/l)
Nitrogen, organic (mg/l)
Phosphorus, total (mg/l)
Calcium, total (mg/l)
Magnesium, total (mg/l)
Sodium, total (mg/l)
Potassium, total (mg/l)
Sulfate, dissolved (mg/l)
Chloride, dissolved (mg/l)
Total dissolved residue (mg/l)
Average
7.6
15.1
9.2
2.2
9.0
276.0
84.0
0.79
0.09
0.17
0.06
33.0
7.5
12.3
1.7
28.0
19.0
162.0
Minimum
7.3
3.3
5.4
1.0
3.0
200.0
72.0
0.60
0.02
0.06
0.04
28.0
6.3
7.7
1.4
18.0
10.0
140.0
Maximum
8.5
23.9
13.4
3.6
20.0
370.0
100.0
1.00
0.30
0.36
0.10
40.0
9.1
19.0
2.3
37.0
31.0
220.0
-------
494
RESTORATION OF LAKES AND INLAND WATERS
Table 2. — Average monthly biomass and 95 percent
confidence limits for aquatic macrophytes at each station*
(g/m2 DW).
Station
1
2
3
4
5
April 17
12 ± 5
17± 6
60+ 23
240 ± 88
2 ± 1
May 13
210+ 78
160 ± 54
97 + 36
76 ± 24
54+ 21
June 18
453 ± 249
620 ± 370
447 ± 181
432 ± 164
227 + 84
July 12
321 ± 118
374+ 139
380+ 150
491 ± 182
605 ± 225
* Large variations in 95 percent confidence limits due to the
natural structure of the macrophyte beds which sometimes
resulted in samples having no rooted plants.
Table 3. — Average biomass and 95 percent limits of aquatic
macrophytes in the Holston River above Cherokee Reservoir
(pooled data).
Date
4-17-80
5-13-80
6-18-80
7-12-80
Mean biomass 95% confidence limits
(g/m2 DW) Upper Lower
66
119
400
427
77
139
475
495
55
100
326
358
Ruane, R. J., and P. A. Krenkel. 1978. Nitrification and other
factors affecting nitrogen in the Holston River. Jour. Water
Pollut. Control Fed. 50:1885.
U.S. Army Corps of Engineers. 1977. Water quality for river -
reservoir systems. Hydrologic Eng. Center. (Draft.)
1972. Water quality and waste treatment
requirements on the Upper Holston River, Kingsport, Tenn.
to Cherokee Reservoir. Tech. Study TS-03-71-208-07.
Surveill. Anal. Div. Region IV, Athens, Ga.
U.S. Environmental Protection Agency. 1978. Holston River
study. EPA 904/9-78-019. Surveill. Anal. Div., Region IV,
Athens Ga.
Westlake, D. F. 1963. Comparisons of plant productivity Biol
Rev. 38:385.
1966. The biomass and productivity of Glyceria
maxima. I. Seasonal changes in biomass. Jour. Ecol
54:745.
ACKNOWLEDGEMENTS
The authors wish to thank Dr. Paula Collier, Dr. David Webb,
and Douglas Murphy for assistance in field collections;
Jennifer Neill for sample processing and data tabulations;
Leon Bates and Billy Isom for review of the manuscript; and
Albert Price for figure preparation.
REFERENCES
Churchill, M. A., and W. R. Nicholas. 1966. Effects of
impoundments on water quality. Presented at the Natl.
Symp. Quality Standards for Natural Waters, Ann Arbor,
Mich., July 19-22.
Dennis, W. M. 1976. Determination of physical characteris-
tics and amount of organic debris in the vicinity of the
Phipps Bend Nuclear Plant water intake Tennessee Valley
Authority, Muscle Shoals, Ala.
Fisher, S. G., and S. R. Carpenter. 1976. Ecosystem and
macrophyte primary production of the Fort River, Mass.
Hydrobiologia 47:175.
Gordon, J. A. 1971. Effects of impoundments on water
quality, report of research conducted at Cherokee Reservoir
from 1966-1970. Tennessee Valley Authority, Chattanooga,
Tenn.
Hall, T. F. 1966. Field inspections of segment of the Holston
River for submersed aquatic plants. Tennessee Valley
Authority, Muscle Shoals, Ala.
Higgins, J. M. 1978. Water quality progress in the Holston
River Basin. TVA/EP-78-08. Tennessee Valley Authority,
Chattanooga, Tenn.
Jewell, W. J. 1971. Aquatic weed decay: Dissolved oxygen
utilization and nitrogen and phosphorus regeneration. Jour.
Water Pollut. Control Fed. 43:1457.
Iwanski, M. L. 1978. Water quality in Cherokee Reservoir.
Tennessee Valley Authority, Chattanooga, Tenn.
Iwanski, M. L., J. M. Higgins, and R. C. Young. 1980. Factors
affecting water quality in Cherokee Reservoir. Tennessee
Valley Authority, Chattanooga, Tenn.
Odum, H. T. 1956. Primary production in flowing waters.
Limnol. Oceanogr. 1:102.
Peltier, W. H., and E. B. Welch. 1969. Factors affecting
growth of rooted aquatics in a river. Weed Sci. 17:412.
Rich, P H., R. G. Wetzel, and N. V. Thoy. 1971. Distribution,
production, and role of aquatic macrophytes in a southern
Michigan marl lake. Freshw. Biol. 1:3.
-------
495
LAKE RESTORATION METHODS DEVELOPED
AND USED IN SWEDEN
WILHELM RIPL
Institute of Ecology
Technical University of Berlin
Hellriegelstr, Berlin
ABSTRACT
Lake restoration research has been carried out at the Institute of Limnology, University of Lund
since 1966. Lakes damaged by excessive eutrophication, acidification, and lowering of water
levels, are the objects of research. In many eutrophicated lakes, phosphorous reduction either by
diversion or treatment of effluents did not improve water quality and oxygen conditions as
expected. This continuing phosphorus concentration was caused by intensive nutrient recycling
from the lake sediments. Lake Trummen, Sweden, improved permanently following removal of the
upper nutrient-rich sediment by suction dredging. In another Swedish lake the sediments were
oxidized by induced denitrification. An apparatus for injecting chemicals into the lake sediments
will be used in a pilot project to convert humic acids in the sediments to sodium humateswith ionic
exchange properties. Techniques were developed to restore lakes damaged by lowering of the
water level and overgrown by dense reeds. The commercially available Limno device was
developed to improve lakes' receiving efficiency. Systems combining wastewater treatment and
biomanipulation were developed.
INTRODUCTION
In the last century especially, water bodies close to
densely populated areas or near intensively used rural
areas have shown dramatic changes in water quality.
Although initial experiments in lake restoration were
carried out early in this century with Naumann
publishing in 1915 results of the restoration of Berlin's
Lietzensee, the need to restore lakes first became clear
in the mid-20th century when industry seriously began
to affect the environment.
In Sweden, practically all lakes close to settlements
were highly polluted because they were used as
receivers. After improved wastewater treatment meth-
ods made it possible to reduce nutrient input, people
wanted these environments restored for recreational
purposes. The diversion of sewage water frequently did
not immediately improve conditions except in lakes
which were not excessively eutrophic. The storage of
sapropelic mud on top of sediments deposited during
oligotrophic conditions usually delayed response.
Methods therefore had to be developed to restore the
sediment function of unpolluted lakes; that is, to act as
nutrient sinks and to recycle as few nutrients as
possible.
Not only these polluted lakes were objects of
restoration. In a large number of lakes, the water level
had been lowered to increase fertile areas for
agricultural purposes. After brief usage, most of these
drained areas did not produce good crops and were
abandoned. However, the lakes, or whatever was left of
them, were in many cases irreversibly damaged.
Macrophytes such as reeds and sedges had overgrown
large parts of these lakes. Even when attempts were
made to raise the water level, often the root felt of
these reedbeds was lifted to the surface because of
intense methane production beneath them. Even for
these lakes restoration methods had to be developed.
A third attack on the Swedish environment was first
sustained in the last decade, when it became evident
that more and more lakes were being damaged by
acidification processes induced by the excessive
burning of sulfur-containing fuels. These damages to
the lake ecosystems are probably the first indications of
large scale processes in the soils, possibly leading to
decreased forest production. Acidification problems
are, of course, not controllable by simple restoration
techniques. However, since large numbers of animals
and fish usually found in these normally oligotrophic
and dystrophic lakes have already vanished, measures
had to be taken to save some of these sensitive
environments.
During the last 15 years, a group of workers at the
Institute of Limnology in Lund led by Professor S. Bjork
have developed restoration methods and carried out
restoration activities. They have cooperated closely
with technicians to develop new equipment and with
authorities to obtain tailor-made solutions for certain
lake ecosystems in densely populated areas. Another
purpose of this work was to study specific entities of
aquatic ecosystems by evaluating the responses of the
systems after restoration.
THE RESTORATION OF LAKE TRUMMEN
Lake Trummen, situated in the South Swedish
uplands close to the town of Vaxjo, was polluted by
municipal sewage and effluents from a textile industry
for a period of less than 50 years. When the sewage
was diverted to the next lake in the lake system it was
-------
496
RESTORATION OF LAKES AND INLAND WATERS
expected that this lake which had a relatively short
water renewal time of about 3 months, would recover
quickly. This was not the case, and after a period of
more than 10 years with heavy algal blooms and
frequent fish kills it was decided to restore the lake. The
preinvestigations showed that the upper 50 centi-
meters of the sediment were deposited during the
pollution period (Digerfeldt, 1972).
The restoration plan provided for removing the upper
sediment layer containing the excessive amounts of
nutrients deposited during the pollution period. The
restoration was carried out during the summer months
of 1970 and 1971; about 300,000 m3 mud were
rerfioved. The sediment was deposited in dewatering
ponds and the backwater from these ponds was
reduced in phosphorus by a small treatment plant with
P-precipitation (Figure 1).
The conditions in the lake improved practically
instantaneously. Microcystis, dominant before restora-
tion from early spring until autumn, vanished and
nanoplanktic species appeared (Gelin and Ripl, 1978;
Cronberg, 1980). Nutrient concentrations decreased
drastically and good oxygen conditions were main-
tained during the whole year (Bjork, et al. 1979). The
lake is now used for recreational purposes such as
fishing and bathing. The experiences from restoring
Lake Trummen showed clearly that the internal
processes were controlled by microbially mediated
exchange processes at the sediment water interface.
1. = during treatment
2. - upper sapropelic sediment
3. - mud suction pipe
4. =* sedimentation pond
5. = mud for fertilisation
6. = automatic dosage of P-precipitant
12
7. - reaction and aeration
8. = precipitation pond
9. = sludge from P-precipitation
10 = return water
11, = consolidated sediment
12. = after treatment
Figure 1. — Principal scheme for treatment of Lake Trummen
(Sweden).
THE BIOCHEMICAL OXIDATION OF THE
SEDIMENT IN SITU
As experiments with different sediment cores of
various properties showed, release of phosphorus was
mainly enhanced in reduced sediments loaded with
fresh organic material because of the intense microbial
activity exerted by anaerobic bacteria. The intensity of
the bacterial processes was shown to be a function of
the quality of the organic substance, the presence of a
suitable electron acceptor, and temperature. Experi-
ments where nitrate was added to the sediments as an
electron acceptor showed that it was possible to oxidize
not only easily degradable organic matter by induced
denitrification, but also to oxidize the inorganic
environment as sulfur and iron species. By this
oxidation the sediment became again phosphorus-
sorbing and the phosphorus concentrations in the
sediment interstitial water decreased drastically.
A tentative restoration was carried out in spring
1975 in a small Swedish lake, Lake Lillesjon close to
Varnamo. A harrow to distribute the chemicals was
developed in cooperation with the Atlas Copco Co.
(Figures 2 and 3). The restoration procedure took 3
weeks. Three chemicals, 13 tons FeCI3 (146 g Fe/m2),
5 tons slaked lime(180gCa/m2),and12 tonsCa(NO3)2
(141 g N/m2) were injected to an area of 1.2 hectares
of reducing sediments. All nitrate was denitrified
during 1.5 months. Since that time the lake has
stabilized at lowered trophic conditions. The internal
nutrient recycling with respect to phosphorus and
nitrogen decreased immediately to only 1/6 of the
original values. Dense duckweed development during
summer stratification over the whole surface of this
4.2-hectare lake was replaced by phytoplankton with
summer transparencies of 1.5 to 2.5 meters (Ripl,
1976, 1978).
Atter this tentative treatment, restoration measures
were planned for the LakeTrekanten in Stockholm (Ripl
and Lundquist, 1977). This restoration was carried out
in May 1980 with a newly designed application harrow
(Figure 4). The method is now offered commercially by
LflKE LILLESJOEN
HREH -
VOLUME -
MHX DEPTH =
MEHN DEPTH =
WRTER RENEWRL -
42 000 m2
86 000 m3
4.2 m
2.0 m
LJ . 3 Months
remaining reeds
TI area o-f sediment
t re atment
T2 area of vegeta=
tion t re atment
max!mum depth
Figure 2. — Map showing treated and morphometric data of
Lake Lillesjoen (Sweden).
-------
SPECIAL TOPICS
497
Figure 3. — Principal scheme for the RIPLOX treatment.
Figure 4. — The sediment harrow has three sections and a total
width of 10 meters. Each section has five rows of supply tubes.
The three rows in the middle inject air to suspend the
sediments. The rear row (depending on direction of movement)
applies the chemicals.
Atlas Copco under the trademark "Riplox". So far the
results from this recent restoration of Lake Trekanten
seem to be the same as in the tentative treatment.
Unlike Lake Lillesjon whose sediments were very low
in iron compounds because of prolonged periods of
anoxic conditions, Lake Trekanten had plenty of natural
iron compounds, especially iron sulfides, present in its
sediments; injection of iron and lime was therefore
unnecessary at Trekanten.
The induced denitrification process was obtained
after a short lag period (about 1 week) and at the end of
July 1980, 70 percent of the injected nitrate had been
denitrified. The oxygen demand of the sediments had
reduced drastically; oxygen was found during the
summer stratification until the end of July in the whole
hypolimnetic zone. The sediments had become phos-
phorus sorbing and the phosphate concentrations in
the interstitial water had decreased from 2 to 4 mg
PCVP/I to values between 0.01 and 0.3 mg/l in the
most reducing sediments at maximum depth. Despite
the high denitrification activity followed by vigorous
emanation of gas from the sediments, the stratification
was preserved and the water in the euphotic zone was
never reached by nitrate concentrations higher than
about 0.5 to 1 mg NO3-N/I.
The lake has, of course, not stabilized yet and will
probably be labile with respect to planktonic and fish
populations. About 20 to 25 centimeters of the upper
sediment layers have been oxidized by this induced
denitrification process, enabling the benthic fauna to
recolonize large areas of the sediments and the lake
ecosystem to reach a new steady state with improved
loading conditions. The total costs for the restoration of
Lake Trekanten were $170,000 or $1.3 per m2.
THE RESTORATION OF
ACIDIFIED LAKES
Many lakes in large areas of Scandinavia, as well as
Canadian lakes, are suffering from excessive loading
with hydrogen ions, produced by the extensive use of
fossil fuel containing sulfur compounds. These acid
rains have already partially sterilized thousands of
lakes. Until now the only measures that have been
taken to save some sensitive fish and crayfish species
were liming; however, acid precipitation has caused
most humic substances in these lakes to sediment. The
addition of lime instantly increases pH values, if this
lime is applied from the lake surface in the form of
calcium hydroxide. But after a short period the effect of
lime is reduced because calcium humates precipitate
from the water and the lime reacts with humic
substances in the sediments to become insoluble
calcium humates (Figures 5 and 6). Another way of
lime inactivation is to coat lime particles with humic
substances; this reduces the potential for neutralizing
acid rain.
In laboratory experiments these effects were in-
vestigated and it could be shown that injecting soda
solutions directly to the sediments neutralizes the
acidic groups, and the sodium humates which are
partially soluble react like ionic exchange resins. The
gradually introduced acidic rain just exchanges the
sodium ions. This treatment is about five to seven times
as efficient as adding lime on an equivalent basis. This
means that although the chemicals are about three
times as expensive, the treatment is more long lasting.
The preinvestigations showed that this sediment
treatment with soda probably will be competitive with
lime treatment when the longer lasting effect and
chemical costs are considered. But an even more
pronounced positive effect of the soda-sediment
treatment is the natural aluminum phosphate precipi-
tation of extremely nutrient impoverished lakes. The
increased exchange processes between water and
sediment, after the sediments have been treated with
soda, not only increase alkalinity, but also nutrients,
leading to primary production and a self-maintaining
recycling of nutrients. A certain primary production is a
prerequisite for maintaining a fish population.
Another advantage of this treatment is that the pH is
not as affected as with frequently conducted liming
measures, thus producing more stable physical-
chemical conditions suitable for the populations
characteristic of ecosystems.
A tentative treatment in the acidified Lake Lilla
Galtsjon in Blekinge, Sweden will be carried out this
year by the restoration team in Lund, Atlas Copco, and
the author. The pretreatment studies in this lake were
mainly concerned with evaluating the sediment
-------
498
RESTORATION OF LAKES AND INLAND WATERS
THE CONTRRCID METHOD
STRUCTURE OF HUMIC RCIDS
BRIDGE
RERCTIVE GROUPS:
RCIDIC RLKRLINE
DISSOCIRTION
— OH
— COOH
0 H
N H
COO H
Figure 5. — Structure of humic aids.
THE CONTRRCID METHOD
RERCTION OF HUMIC RCIDS HITH
1 LIME PRODUCTS 2 SODR PRODUCTS
CaCOH)_
CaO
CaCO
NaOH
Our purpose is to obtain the most favorable and
stable conditions combined with a long lasting effect.
The procedure will be suited for strongly acidified lakes
and will be available as "Contracid" method. The
necessary restoration parameters, however, have to be
evaluated in advance by experimental and limnological
field work (Ripl, 1978).
THE RESTORATION OF OVERGROWN
LAKES AND WETLANDS
A considerable number of lakes in Sweden were
damaged by lowering the water level and the resulting
expansion of macrophytes. Some lakes which were of
great importance for the reproduction of water fowl, or
important stations for migrant birds such as cranes, are
now restoration objects. One of considerable size —
and the largest restoration project in Sweden— is the
famous Lake Hornborga. Since it is not possible to fill
overgrown lakes again with water without first
preparing the lake area which had been overgrown by
reed and sedge vegetation, methods had to be
developed for such restoration. The usually very
resistant root felts had to be cut and removed by
amphibious machines and large amounts of accumu-
lated biomass had to be removed and burned; this was
done mainly during winter when the ice cover made
the use of heavier machines possible. The water level
could then be increased, leading almost instantly to the
development of underwater vegetation.
The restoration plan for Lake Hornborga includes
raising the water level to a maximum depth of 2.4
meters. It should take only one spring to fill the lake
with water, as Lake Hornborga is flooded every year
after snow melt. The project goal for Lake Hornborga is
restoration of an open water area of 11 km2. In 1977
the Swedish government decided to spend about $7
million on this restoration (Bjork et al. 1979) (Figure 7).
R= HUMIC RESIDUE
R 0 Ca
R = N Ca
R - COO Ca -
INSOLUBLE
PRODUCTS
R - 0~
R N~
Na1
Na +
R COO Na
REVERSIBLE IONIC-
EXCHRNGE
SOLUBLE PRODUCTS
Figure 6. — Reaction of humic acids with lime and soda
products.
properties, the diffusion of various soda solutions into
the sediment, the optimal area to be treated, and the
extent of the eutrophication caused by nutrient
exchange between sediment and water.
nred fay common nw>d until 1967
ATM: Ca 1 km1 fi natty prepared
deposition
decomposition
Bottom fount:
ChlronomkJae Ind/m1 x 1200
BMi (pain x 10):
Podlcep* auritui, Homed Grebe
Avthva fuHgula, Tufted Duck
Aythya farina. Pochard
Figure 7. — Lake Hornborga before and after experiments for
directing the primary production from emergent to submerged
vegetation. Water level not yet raised. Comparison between
conditions in 1965 and 1971 (Bjork, 1972).
-------
SPECIAL TOPICS
499
Draining large areas in Sweden for agricultural
purposes has destroyed many wetlands. New eco-
logical insight has led to projects to restore wetlands.
The reestablishment of wetlands in drained lakes and
peat pits implies the production of energy reeds,
potential habitats for waterfowl, fish, and wildlife, and
shallow water reservoirs for the recessive amphibious
fauna (Bjork and Graneli, 1978).
AERATION METHODS
Eutrophic lakes of a certain depth, especially
receivers of polluted effluents, suffer during stagnation
periods of insufficient oxygenation of the hypolimnetic
zone. Furthermore, lagoons used for storage and the
breakdown of a heavy load of industrial oxygen-
demanding effluents prior to their discharge need
additional oxygen. In drinking water reservoirs the
addition of oxygen will prevent the dissolution of iron
and manganese compounds from the bottom areas and
thereby avoid the relatively expensive water treatment
in a treatment plant. For this reason different aeration
devices have been developed and used. In Sweden the
original idea proposed by Bernhardt and Hotter (1967)
to achieve aeration with an air lift was further
developed by Atlas Copco and resulted in the Limno
device. Thorough limnological studies in connection
with aeration measures showed that it is possible to
control oxygen abundance and thereby improve
aerobical breakdown of autochthonous and alloch-
thonous organic material in aquatic environments still
overloaded by organic matter, or by excessive
nutrients.
*fc^K*4^ftr ....^ajKv^,.:^ . -s jt'' L '^^^^^^j^i^^S^^^^^f^^^i^^^[{
^^5^°^ooip^-p%^|;f^if^£f?^g%-:?5^^
Figure 8. — Cutaway sketch of polyester plastic limno unit.
Examples for the application of these aeration
devices are now numerous in Sweden and abroad. One
of Sweden's mining companies has installed five
Limno units with a capacity of about 250 kg 0 /day
each in several basins receiving effluents from an ore
flotation process using organic flotation chemicals.
Another example is the emergency drinking water
reservoir for the town of Brussels where a Limno unit
was installed to prevent anaerobic conditions. Other
lakes are maintained by aeration until final solutions to
divert pollutants such as phosphorus precipitation are
installed. In West Berlin one of the Havel lakes, the
Tegeler See, used by thousands of people for
swimming and other recreational purposes is equipped
with three Limno units; it will eventually be equipped
with eight or nine more units, each delivering 350 kg
Oz/day, until the phosphorus elimination plant is built
(Figure 8).
BIOMANIPULATION AND
OPTIMIZATION OF THE COMPLEX
TREATMENT PLANT RECEIVER
Simultaneous to the development of the different
restoration methods, changes in structure and function
of the treated ecosystems were analyzed. Andersson
(1977) investigated the relationships between phyto-
plankton, zooplankton, and various species of fish. He
was able to show in both limnocoral experiments and
whole lake studies that the various fish species by
selecting their diet and thereby controlling the
abundance of filter feeders had a more or less
eutrophicating effect. Selective fishing in Lake Trum-
men for roach and bream resulted in increased
transparency, lowered nutrient level, and reduced
biomass. Experiments were simultaneously conducted
in enclosures in this lake. The results from these
experiments showed even more pronounced effects
(Andersson, 1979).
Ripl (1978) proposed experiments to show that a
nitrification step in the advanced sewage treatment
plants with phosphorus reduction, and a direction of
the plant effluents to the reducing sediment areas
would oxidize the reducing sediments, denitrify
excessive nitrogen, and increase the phosphorus-
sorbing capacity of the sediments. Part of the nitrified
effluents would, of course, serve as a nitrogen source
for algae and probably induce a succession of nitrogen-
fixing blue-green algae. These experiments were
partially carried out in 1978 (Ripl, et al. 1979) and
showed that most of the added nitrate nitrogen was
denitrified in the sediments of even shallow systems.
Further, the almost monoculture of the nitrogen-fixing
species Anabaena flos aquae broke down and was
replaced by green algae. Transparency increased since
the green algal population was controlled by filter
feeders. There are now plans for some receiver-
ecosystems to try this concept in Lake Finjasjon in
Scania, Sweden, and in the large German fjord Schlei
(Figure 9).
Theoretical knowledge of aquatic ecosystems is
growing stronger at the limnological institutes. This
means that the management of lake ecosystems as
well as their maintenance and in some cases their
restoration becomes safer. However, since lake
-------
500
RESTORATION OF LAKES AND INLAND WATERS
ecosystems are individuals, there will never be a best
method which can be applied to every lake ecosystem.
Only knowledge of the structure and function of each
individual system makes it possible to develop suitable
methods to reach a new equilibrium with changed and
improved conditions.
Figure 9a. —
STRATIFIED RECIPIENT
Epulmnkxl Hl9h BkMtun
Figure 9b. —
SHALLOW RECIPIENT
Figure 9a, b. — Schematic diagrams of optimized and traditional models of the treatment plant/recipient system (stratified
recipient). WTP wastewater treatment plant with biological treatment (Bio) and chemical precipitation (Che). Nitr
nitrification process. Denitr = denitrification process. (Ripl, et al. 1979).
REFERENCES
Andersson, G. 1979. Fiskens inverkan pa trofiforhallandena i
eutrofa sjoar. Limnologiska institutionen, LundsUniversitet.
Coden Lunbds (NBLI-3024)/!.
Andersson, G., et al. 1978. Effects of planktivorous and
benthivorous fish on organisms and water chemistry in
eutrophic lakes. Hydrobiologia 59:9.
Bengtsson, L, et al. 1972. Restaurering av sjoar med
kulturbetingat hypolimniskt syrgasdeficit. Limnologiska
institutionen, Lunds Universitet, Centrala fysiklaboratoriet.
Atlas Copco AB.
Bernhardt, H., and G. Hotter. 1967. Moglichkeiten zur
Verhinderung anaerober Verhaltnisse in einerTrinkwasser-
talsperre wahrend der Sommerstagnation. Arch. Hydrobiol.
63:404.
Bjork, S. 1972. Bringing sick lakes back to health. Teknisk
Tidskrift. 102:11:93.
Bjork, S., and W. Graneli. 1978. Energy needs and the
environment. Ambio 7:150.
Bjork, S., etal. 1979. Lake management. Studies and results
at the Institute of Limnology, University in Lund. Arch.
Hydrobiol. Beih. Ergebn. Limnol. 13:31.
Cronberg, G. 1980. Phytoplankton changes in LakeTrummen
induced by restoration. Limnologiska institutionen, Lunds
Universitet. Coden Lunbds (NBLI-1005)/1.
Digerfeldt, G. 1972. The postglacial development of Lake
Trummen. Regional vegetation history, water level changes
and paleolimnology. Folia Limnol. Scandinav. 16:1.
Gelin, C., and W. Ripl. 1978. Nutrient decrease and response
of various phytoplankton size fractions following the
restoration of Lake Trummen, Sweden. Arch. Hydrobiol.
81:339.
Leonardson, L., and W. Ripl. 1980. Control of undesirable
algae and induction of algal succession in hypertrophic lake
ecosystems. Proc. Workshop on Hypertrophic Lake Eco-
systems, Vaxjo, Sweden, Sept. 10-14. 1979. (In press).
Naumann, E. 1915. Lietzensee vid Berlin. Skr. Sod. Sver.
Fiskfor. 13:1.
Ripl, W. 1976. ProzeBsteuerung in geschadigten See-
Okosystemen. Vjschr. naturf. Ges. Zurich 121:301.
1978. Oxidation of lake sediments with nitrate. A
restoration method for former recipients. Institut of
Limnology, University of Lund. Coden Lunbds (NBLI
1001 )/1.
Ripl, W., and I. Lundqvist. 1977. Forslag till restauering av
sjoar inom Stockholms kommun. Limnologiska institu-
tionen, Lunds Universitet. Rapport.
Ripl, W., et al. 1979. Optimering av reningsverk/recipient-
system. Vatten 2:96.
-------
Appendix A
SUMMARY OF CLEAN LAKES PROJECTS
501
Name: Albert Lea
Location: Freeborn County, Minn.
Problem: High phosphorus content, depleted D.O.,
turbid water, and severe algal bloom.
Project Objectives: To restore the water quality
and to restore recreation activities.
Restorative Techniques Used: Relocate treatment
plant effluent by implementing a 201 project. Filter
stormwater runoff through native soil filter media. The
feasibility of dredging and other in-lake activities will
be determined following diversion of the effluent.
Project Progress: Fountain Lake soil filters have been
completed; Albert Lea Lake 201 project has been
approved.
Implementation Problems: None.
Name: Allentown
Location: Allentown Borough, N.J.
Problem: Pollution inputs have resulted in a hyper-
eutrophic condition which is exemplified by nuisance
algal blooms and aquatic weeds plus sedimentation.
Project Objectives: Reduce pollutant inputs from
both the watershed and the lake sediments; remove
sediments.
Restorative Techniques Used: Dredge sediments,
stabilize the shoreline, divert stormwater pollution.
Project Progress: Project implementation expected to
begin in 1981.
Implementation Problems: Delays caused by lengthy
404 permit process.
Name: Ann Lee
Location: Albany County, N.Y.
Problem: Nuisance macrophytes, accelerated eutro-
phication, shallowness, runoff from agricultural, resi-
dential, and commercial areas.
Project Objectives: Remove accumulated sediments
and reduce sediment and nutrient inputs.
Restorative Techniques Used: Dredge sediments;
treat runoff, and stabilize shoreline.
Project Progress: Dredging is underway; construc-
tion of retention pond in design phase.
Implementation Problems: None.
Name: Apopka
Location: Orange and Lake Counties, Fla.
Problem: Fish kills, algal blooms, water hyacinths,
hypereutrophication.
Project Objectives: Restore water quality of lake by
reducing internal nutrient loading from unconsolidated
sediments.
Restorative Techniques Used: Drawdown lake to
consolidate bottom sedimentation and reduce internal
nutrient loading.
Project Progress: Preliminary monitoring and engi-
neering design study have been completed as has an
Environme'ntal Impact Statement.
Implementation Problems: Costs of the project, due
to environmental constraints, turned out to be prohibi-
tively expensive. Project was terminated in 1980.
Name: Ballinger
Location: King-Snohomish Counties, Wash.
Problem: Excessive algal and macrophyte growths
and turbidity affecting swimming, boating, and picnick-
ing.
Project Objectives: Control external sediment/nutrient
sources and reduce in-lake nutrient buildup.
Restorative Techniques Used: Tributary sedimenta-
tion basins; reduce lake level fluctuations; remove
polluted hypolimnetic water.
Project Progress: Sedimentation basins complete;
monitoring and evaluation continuing.
Implementation Problems: Unclear responsibility for
lake level control; interjurisdictional disagreement on
lake project operation and impacts.
Name: Bantam
Location: Litchfield County, Conn.
Problem: Lakewide phytoplankton blooms and
macrophyte beds in the extensive littoral zones which
cover as much as 20 percent of the lake's 371 surface
hectares.
Project Objectives: To deepen the lake and reduce
aquatic macrophytes, and to improve recreational
opportunities.
Restorative Techniques Used: Selective dredging of
343,970 cu. meters of sediment from those areas where
sufficient organic sediment exists to promote growth of
aquatic macrophytes; nonpoint source loading abate-
ment program for the lake watershed.
Project Progress: Watershed study work plan develop-
ment is in progress.
Implementation Problems: None.
Name: Big Alum
Location: Worcester County, Mass.
Problem: Septic tank leachate from shoreline resi-
dences and erosion are causing high concentrations of
phosphorus in the lake, pointing to eventual eutrophic
conditions.
-------
502
RESTORATION OF LAKES AND INLAND WATERS
Project Objectives: Preserve the present water quality
of the lake.
Restorative Techniques Used: Develop plan which
may include: sedimentation basins, composting toilets,
and modified septic systems; public participation and
education program; watershed management; purchase
and management of wetlands areas.
Project Progress: Engineering design study is under-
way and should be completed in spring 1981.
Implementation Problems: Numerous delays in
awarding contract.
Name: Blue
Location: Monona County, Iowa
Problem: Heavy siltation; low water levels; dense
growth of macrophytes; and decreasing lake usage.
Project Objectives: Restore water quality and deepen
lake.
Restorative Techniques Used: Dredge approximately
36 percent of lake to remove aquatic vegetation and
nutrient-enriched bottom sediment; form sanitary dis-
trict to ensure proper construction and operation of
treatment facilities in the future.
Project Progress: Dredging has been completed;
water quality assessment underway.
Implementation Problems: None.
Name: Bomoseen
Location: Rutland County, Vt.
Problem: High nutrient concentrations have resulted
in heavy growth of aquatic macrophytes and blue-green
algae, which interfere with recreational activities.
Project Objectives: Since nutrient sources entering
the lake have been controlled, the goal is to remove the
in-lake nutrient source, which is recycling from existing
aquatic vegetation.
Restorative Techniques Used: Harvesting 73 hectares
of the lake each year for 3 years will remove excessive
nutrient levels, thereby reducing aquatic plant growth and
increasing public access and use of the lake.
Project Progress: 3 years of harvesting has been com-
pleted. To date, harvesting has limited plant growth as
indicated by the decrease in pounds of aquatic macro-
phytes removed from harvested areas. Final project
assessment is now underway.
Implementation Problems: None.
Name: Broadway
Location: Anderson County, S.C.
Problem: Siltation
Project Objectives: Remove lake sediments and re-
duce erosion and sedimentation in the watershed above
the lake.
Restorative Techniques Used: Best management
practices, dredging, and roadbank stabilization.
Project Progress: Roadbank stabilization in progress;
archeological study in progress; pre-monitoring program
complete.
Implementation Problems: Dam safety issue has pre-
vented release of funds for sediment dredging. The COE
will not commit to dredging in the lake until the dam is
repaired and passes a safety inspection.
Name: Buckingham
Location: Albany County, N.Y.
Problem: Extensive aquatic plant growth; lake bot-
tom covered with organic debris and silt.
Project Objectives: Improve overall lake water quality.
Restorative Techniques Used: Drain lake and remove
the accumulated silt and muck by excavation.
Project Progress: Removal of accumulated sediment
did not substantially reduce the eutrophic state of the
lake.
Implementation Problems: None.
Name: Bugle
Location: Trempealeau County, Wis.
Problem: Sediment infilling has reduced the recrea-
tional value of the lake.
Project Objectives: To remove most of the sediment
in the lake and to enhance streambank stabilization
upstream.
Restorative Techniques Used: Hydraulic dredging,
riprapping, and limited sloping and seeding.
Project Progress: Project is 70 percent complete.
Some dredging and watershed work will have to be
carried over to summer 1981.
Implementation Problems: Heavy rains in August
1980, plus farmers' reluctance to allow heavy machinery
into their fields slowed the project.
Name: Charles River
Location: Suffolk County, Mass.
Problem: Saltwater stratification prevents vertical
mixing, and decomposition of organic materials results
in complete oxygen depletion in the deeper zones result-
ing in odor from hydrogen sulfide production.
Project Objectives: Destratify the basin and improve
water quality.
Restorative Techniques Used: Destratification of the
Charles River lower basin by induced circulation using
compressed air.
-------
503
Project Progress: Austin equipment in place. Destrati-
fication has been effective. Final water quality assess-
ment is performed.
Implementation Problems: None.
Name: City Park Lakes
Location: East Baton Rouge Parish, La.
Problem: Heavy metals sedimentation, hypereutro-
phication and agricultural runoff.
Project Objectives: Restore lakes' water quality by
reducing nutrient inputs and sediment removal.
Restorative Techniques Used: Dredging to remove
sediments; control of urban stormwater runoff; rehabil-
itation of sewer lines; institute agricultural BMP's; and
divert water to aid in faster stormwater dissipation.
Project Progress: Completed preliminary feasibility
work and are ready to bid the dredging work.
Implementation Problems: Multiple coordination has
caused project delays.
Name: Clear
Location: Waseca County, Minn.
Problem: High nutrient content and algal bloom in
summer.
Project Objectives: Restore water quality of Clear
Lake.
Restorative Techniques Used: Diversion of stormwater
through a marsh filter system.
Project Progress: Approximately 60 percent com-
pleted.
Implementation Problems: Delays have been caused
by adverse weather.
Name: Clearwater River Chain of Lakes (Clearwater,
Augusta, Caroline, Marie, Louisa, Scott, Betsy Lakes)
Location: Wright, Stearns, and Meeker Counties, Minn.
Problem: Excessive nutrient loading, algal blooms,
and proliferation of macrophytes.
Project Objectives: Restore the recreational, aesthetic,
and water quality of the Chain of Lakes.
Restorative Techniques Used: Wetland treatment sys-
tems for stormwater runoff from tributary streams will
be used. On-land disposal of three community effluents
is being done now. Hypolimnetic alum treatment of
Augusta Lake will be undertaken.
Project Progress: Work plan is being finalized.
Implementation Problems: None.
Name: Cobbossee Watershed District I
Location: Kennebec County, Maine
Problem: Three lakes comprising the watershed
(Annabessacook, Cobbossee, and Pleasant Pond) are
eutrophic and suffer from excessive phosphorus enrich-
ment and dense algal blooms.
Project Objectives: Reduce phosphorus loading to the
lakes.
Restorative Techniques Used: Hypolimnetic aeration
to control internal nutrient cycling; chemical addition
(alum) to bind or absorb soluble phosphorus; construc-
tion of manure storage facilities to control phosphorus
runoff; diversion of runoff; and livestock exclusion
from streams.
Project Progress: Watershed nutrient controls have
been almost completed. In-lake work has finished. Mon-
itoring is ongoing to assess the project results.
Implementation Problems: Reluctance of farmers to
initially participate in the program because of unproven
benefits.
Name: Cobbossee Watershed District II
Location: Kennebec County, Maine
Problem: A dozen lakes in the Cobbossee watershed
are being threatened by agricultural runoff.
Project Objectives: Protect these lakes in this water-
shed by implementing agricultural BMP's.
Restorative Techniques Used: Implement agricultural
BMP's to reduce nutrient inputs to lakes. Control meas-
ures will include manure storage facilities, diversion of
barnyard runoff.
Project Progress: The work plan has been finalized
and implementation is awaiting assessment of agricul-
tural BMP's from the Cobbossee I project.
Implementation Problems: None.
Name: Cochituate
Location: Suffolk County, Mass.
Problem: Eutrophication is resulting in excessive
blue-green algal production, odor problems, oxygen de-
pletion, and possible loss of cold water fishery.
Project Objectives: Reduce influx of nutrients from
surface water runoff and septic tank seepage.
Restorative Techniques Used: Purification of tribu-
tary water by natural sand filter beds; dredging of three
settling ponds and installation of an automatic nutrient
inactivation system in first settling pond; public aware-
ness program; drawdown; and harvesting of rough fish
for nutrient removal.
Project Progress: Filter beds have been evaluated;
nutrient budgets have been computed; and a technical
memorandum on the methodologies, costs, and impact
of dredging has been completed. Reassessment of
restoration alternatives is underway.
Implementation Problems: Cost effectiveness of
-------
504
RESTORATION OF LAKES AND INLAND WATERS
project has been questioned. More engineering work
needs to be done.
Name: Cochrane
Location: Duel County, S.D.
Problem: Blue-green algal blooms caused by nutrient
influx from agricultural runoff.
Project Objectives: Reduce nutrient input to lake.
Restorative Techniques Used: Construct three sedi-
ment control dams to intercept runoff and construct
settling basins behind the dams to catch sediments and
nutrients.
Project Progress: Sediment traps have been developed
and preliminary evidence suggests that the influx of sus-
pended solids has been greatly reduced.
Implementation Problems: None.
Name: Collins Park
Location: Schenectady County, IM.Y.
Problem: Sediment and nutrient loadings from a
storm sewer outfall are causing sediment buildup and
aquatic vegetation growths.
Project Objectives: Reduce sediment and nutrient
loadings so lake might be used for recreational activities.
Restorative Techniques Used: Dredging; planting of
macrophytes to act as a nutrient trap; and removal of
snow and cut vegetation (which were dumped in or near
the lake) from the lake drainage area.
Project Progress: Dredging has been completed.
Implementation Problems: None.
Name: Commonwealth
Location: Washington County, Ore.
Problem: Siltation and excessive algal growths pre-
venting use of the lake.
Project Objectives: Identify silt and nutrient sources;
develop and implement corrective measures.
Restorative Techniques Used: Dredging, dilution,
riprap, and revegetation.
Project Progress: Project completed-achieved greater
depth and clarity of lake water. Developed attractive
lake setting and facility for fishing, boating, and picnick-
ing.
Implementation Problems: None.
Name: Covell
Location: Minnehaha County, S.D.
Problem: Eutrophication and sediment loading to
the lake.
Project Objectives: Improve water quality, develop
better fisheries habitat.
Restorative Techniques Used: Dredging, modification
of outlet structure, and construction of sediment reten-
tion pond.
Project Progress: Engineering design completed and
dredging to begin shortly.
Implementation Problems: 404/402 permit require-
ments for discharge of dredge elutriate have delayed the
project.
Name: Creve Coeur
Location: St. Louis County, Mo.
Problem: Sedimentation is resulting in decreasing
surface area and depth.
Project Objectives: Increase surface area and depth of
Creve Coeur Lake and improve recreational opportunities.
Restorative Techniques Used: Dredge 121 hectares to
a depth of 3 meters; dredged spoils are to be deposited
in the area surrounding the lake and used for lake devel-
opment.
Project Progress: Dredging to begin in early 1981.
Implementation Problems: Coordination problems
between region and locals have caused several delays in
project implementation.
Name: Decorah
Location: Juneau County, Wis.
Problem: Excessive sediment.
Project Objectives: Remove and transport sediment
and improve water quality.
Restorative Techniques Used: Hydraulic and mechan-
ical dredging.
Project Progress: Work plan has been completed, but
project work has not been implemented.
Implementation Problems: A lawsuit challenging the
legality of the formation of the lake district is in court.
Until this is resolved, no work is being done.
Name: Delaware Park
Location: Erie County, N.Y.
Problem: Floating debris, siltation, and sewage de-
posits from Scajaquada Creek resulted in the lake being
closed for public use.
Project Objectives: Reduce and/or remove pollution
entering the lake from Scajaquada Creek.
Restorative Techniques: Install stormwater intercep-
tors; detour Scajaquada Creek around the lake through a
closed underground conduit; finally, dewater and dredge
the lake. Refill the lake with clean spring water.
Project Progress: Stream diversion conduit has been
-------
505
completed. Dewatering and dredging are scheduled to
begin in 1981.
Implementation Problems: None.
Name: Ellis
Location: Yuba County, Calif.
Problem: Urban runoff has caused excessive nutrients
and sedimentation and growths of the nuisance macro-
phyte, Hydrilla.
Project Objectives: To rehabilitate lake by eliminating
excessive growths of hydrilla, and diverting stormwater
flow from the lake.
Restorative Techniques Used: Nonpoint source con-
trol program, removal of sediments, application of herb-
icides.
Project Progress: Sediments have been removed and
stormwater interceptor system has been constructed.
Implementation Problems: Project costs have been
high.
Name: Ellis Brett Pond
Location: Plymouth County, Mass.
Problem: Pond is eutrophic and nonpoint source
pollution including stormwater runoff from a regional
shopping center has made the pond unsafe for swimming.
Project Objectives: Reduce impact of nonpoint source
pollution and remove accumulated sediments and prob-
lem aquatic plants.
Restorative Techniques Used: Streetsweeping; instal-
lation of filters and oil traps on parking lot drains; con-
struction of catch basins; and dredging.
Project Progress: Engineering study showed costs
for dredging and catch basins would be extremely
expensive.
Implementation Problems: Project cancelled because
it was judged not to be cost effective by grantee.
Name: Eola
Location: Orange County, Fla.
Problem: Urban runoff resulting in hypereutrophi-
cation.
Project Objectives: Reduce urban runoff into the lake
and restore it for increased public usage.
Restorative Techniques Used: Parking lot diversion
into percolation ponds; street inlet modification for
percolation; and diversion of runoff through natural
areas.
Project Progress: Water quality monitoring program
in progress; designs for the percolation and lake natural
systems are complete; pilot percolation basin completed;
construction to begin in early 1981.
Implementation Problems: None
Name: Lake Fenwick
Location: King County, Wash.
Problem: Turbidity and sediment interfering with
park development and lake use for boating, fishing, and
swimming.
Project Objectives: Control stormwater and erosion
of banks and bottom of inlet stream.
Restorative Techniques Used: Enforce clearing and
grading ordinance; divert peak flow of inlet stream; and
provide detention for inlet stream water.
Project Progress: Clearing and grading ordinance
being enforced; inlet stream diversion system complete;
detention basin and outlet stream to lake complete;
revegetation plan complete and work scheduled; mon-
itoring and evaluation continuing.
Implementation Problems: Inability to obtain ease-
ment for better diversion pipeline route resulting in in-
stream pipeline.
Name: 59th Street Pond
Location: New York, N.Y.
Problem: i ne pond is stagnant and turbid with
excessive growths of algae and grasses; substantial re-
duction of water depth from siltation; high color levels;
and high coliform content.
Project Objectives: Restore quality of pond to in-
•crease its value as a passive recreational source for
tourists and local residents.
Restorative Techniques Used: The pond will be
drained and dredged. The bottom of the pond will be
made impenetrable to prevent remaining bottom nu-
trients from entering the water column. Pond bank
riprap will be repaired, and clogged stormwater drainage
pipes will be cleaned.
Project Progress: Construction and dredging are
completed. Post-construction monitoring is underway.
Preliminary results indicate a marked improvement in
the appearance of the pond.
Implementation Problems: The total project costs
have been high ($250,000/acre).
Name: Finger Lakes (12 lakes)
Location: Boone Countv Mo.
Problem: All of the lakes are acidic as a result of acid
mine drainage caused by exposed sulfurous spoil areas.
Project Objectives: Improve water quality of the lakes
by eliminating acid sources.
Restorative Techniques Used: Connect 12 separate
lakes by construction of five small earthen dams and two
canals to form a single lake of 17 hectares; divert to pro-
ject lakes the drainage of 405-hectare rural watershed
not disturbed by mining.
Project Progress: Construction completed and final
-------
506
RESTORATION OF LAKES AND INLAND WATERS
assessment underway.
Implementation Problems: None.
Name: Frank Molten Lakes
Location: St. Clair County, III.
Problem: Accumulated silt deposits and nutrients
. caused by runoff have degraded water quality of all
three lakes.
Project Objectives: Restoration of lakes to suitable
depth and rehabilitation of fish population. Relocation
of Harding Ditch, a major source of pollutants.
Restorative Techniques Used: Dredging, relocation of
Harding Ditch, and construction of inverted siphon.
Project Progress: Dredging projects for lakes 1 and 2
are expected to be in engineering design during the win-
ter of 1980-81.
Implementation Problems: Administrative delay and
budgetary problems necessitated an extension of the
project and budget periods.
Name: Gibralter
Location: Santa Barbara County, Calif.
Problem: Lake is filling in with sediments. Some of
the sediment is contaminated with mercury from past
mining activities.
Project Objectives: Remove contaminated sediments.
Restorative Techniques Used: Dredge sediments using
a "Pneuma" pump method.
Project Progress: Planning completed. Dredging
should begin in winter 1980.
Implementation Problems: Delays in planning due to
project complexity. Dispute about patents and rights
to the Pneuma pump.
Name: Green Valley
Location: Union County, Iowa
Problem: Shallowness, excessive sedimentation and
runoff.
Project Objectives: Reduce runoff and sedimentation
and deepen the lake.
Restorative Techniques Used: Dredging to remove
accumulated sediments, agricultural BMP's.
Project Progress: Work plan being developed.
Implementation Problems: None.
Name: Half Moon
Location: Eau Claire County, Wis.
Problem: High phosphorus loading is a major cause
of abundant nuisance algae and indirectly creates ex-
cessive oxygen demands with resultant fish winterkill.
Project Objectives: To reduce phosphorus loading.
Restorative Techniques Used: Storm sewer diversion
and installation of supplemental wells.
Project Progress: Storm sewer work is underway,
collectors are being installed, and project is 80 percent
complete.
Implementation Problems: The makeup wells did not
provide sufficient water. Collectors are being extended
frcnan the less porous ground near the wells to the gravels
associated with the nearby Chippewa River.
Name: Hampton Manor
Location: Rensselaer County, N.Y.
Problem: The lake has a eutrophic condition with
algal blooms in summer months, and the encroachment
of rootea macrophytes is threatening recreational
activities.
Project Objectives: Restore lake by removing sedi-
ments; oxygenate the bottom waters.
Restorative Techniques Used: Drawdown of lake;
consolidation and removal (dredging) of sediments.
Placement of aeration system.
Project Progress: Turbidity and algal blooms have
been reduced; transparency has been improved.
Implementation Problems: None.
Name: Lake Harriet/Lake of the Isles
Location: Hennepin County, Minn.
Problem: Urban stormwater runoff.
Project Objectives: To improve the water quality
of Lake Harriet and Lake of the Isles.
Restorative Techniques Used: Vacuum sweep streets
that drain into Lake Harriet, and install first flush diver-
ters in the Lake of the Isles drainage area.
Project Progress: Diverters are completed and are
monitored on a storm basis. Vacuum sweeping continu-
ing on project.
Implementation Problems: Difficulties in scheduling
street vacuuming.
Name: Henry
Location: Trempealeau County, Wis.
Problem: Excessive sedimentation.
Project Objectives: Increase water depth, reduce sedi-
mentation, and reduce nutrient inflow.
Restorative Techniques Used: Hydraulic dredging,
streambank stabilization by rock riprapping, sloping and
seeding, and selective fencing.
Project Progress: Project completed. The lake was
dredged, streambank stabilization was achieved, and
-------
507
runoff from barnyards upstream was diverted. Project
assessment is underway.
Implementation Problems: None.
Name: Herman
Location: Lake County, S.D.
Problem: Advanced eutrophication, algal blooms, low
D.O., occasional fish kills.
Project Objectives: Reduce sediment and nutrient
loadings to lake.
Restorative Techniques Used: BMP's and sediment
control structures in the watershed.
Project Progress: The project is half complete. BMP's
and sediment control structures are almost in place.
Plans are being formulated for additional in-lake restora-
tive work.
Implementation Problems: None.
Name: Hyde Park
Location: Niagara County, N.Y.
Problem: Deteriorating quality due to increased pol-
lution loading from housing developments, a sanitary
landfill, accidental oil spills from a railroad yard, and
sedimentation.
Project Objectives: Improve overall quality of lake by
reducing pollutant loadings and removing sediment.
Restorative Techniques Used: Drain and dredge
lake; augment flow to lake; plant native vegetation
along streambank to retard erosion; construct siltation
pond; install oil boom system downstream from siltation
pond; and carry out limnological monitoring program.
Project Progress: Watershed measures are being imple-
mented including sewering 900 homes, proper landfill
management, and control of pollutants from the railway
yard. Dredging is underway and should be completed in
1981. Sedimentation pond construction is underway.
Implementation Problems: None.
Name: Hyland
Location: Hennepin County, Minn.
Problem: High phosphorus content, algal blooms, and
turbid water.
Project Objectives: Restoration of lake water quality.
Restorative Techniques Used: Lake drawdown, treat
bottom sediments for phosphorus removal, build storm-
water settling ponds, and drill wells for flow augmenta-
tion.
Project Progress: All implementation work completed.
Implementation Problems: After lake was drained, an
enormous growth of smartweed had to be harvested so
that nutrients would not be reintroduced during flooding.
Name: Jackson
Location: Leon County, Fla.
Problem: Nonpoint source pollution, sediment and
nutrient loading into the lake.
Project Objectives: To reduce sediment and nutrient
load entering the lake from nonpoint sources.
Restorative Techniques Used: A filtration impound-
ment system coupled to a marsh to reduce nutrient and
sediment loading.
Project Progress: All land purchased, lagoons and
marsh filtration system is underway.
Implementation Problems: Land acquisition has been
delayed several times by high appraised values but all
property has been purchased.
Name: Kampeska
Location: Codington County, S.D.
Problem: Shoreline erosion and high sediment load-
ing.
Project Objectives: Reduce shoreline erosion and con-
trol input of nutrients to the lake.
Restorative Techniques Used: Riprapping shoreline
areas.
Project Progress: Riprapping is near completion and
sediment loading rates have been developed.
Implementation Problems: One of the riprap areas
failed due fo the steep slope. The slope could not be
modified prior to riprapping due to its historic nature.
Name: Lafayette
Location: Alameda and Contra Costa Counties, Calif.
Problem: Excessive growth of blue-green and other
algal types creates taste and odor problems and clogs the
filter of the nearly completed water treatment plant; low
oxygen concentration in the hypolimnion.
Project Objectives: Restore the recreational, aesthetic,
and economic values of Lafayette Reservoir.
Restorative Techniques Used: Hypolimnetic aerations
and nutrient inactivation.
Project Progress: Project has not been implemented.
Only water quality monitoring program was undertaken.
Implementation Problems: Project was not imple-
mented because of cost increases and problems securing
additional local funds.
Name. Lansing
Location: Ingham County, Mich.
Problem: The shallowness of the lake has allowed for
extensive macrophyte growth and has resulted in recrea-
tional impairment.
Project Objectives: To restore recreational use, espe-
-------
508
RESTORATION OF LAKES AND INLAND WATERS
cialiy boating, and to improve aesthetics and fish popula-
tion.
Restorative Techniques Used: Hydraulic dredging of
lake bottom, and beach nourishment through depositing
of dredged sand on selected beaches.
Project Progress: More than 30 percent of dredging
has been done.
Implementation Problems: Implementation has been
hindered by controversial actions from the Township to
the Federal level. Delays have been caused by court
battles. The delays have contributed to cost increases,
including the general factor of inflation and the par-
ticular factor of greatly increased fuel costs.
Name: Lenox Reservoir
Location: Taylor County, Iowa
Problem: Eutrophic, highly turbid with odor and
taste problems; extensive siltation; increasing macro-
phyte growth.
Project Objectives: Restore overall water quality of
Lenox Reservoir by deepening the lake and removing
vegetation.
Restorative Techniques Used: Dredging and some
dike construction to insure that dredged material does
not return to the lake.
Project Progress: Project has been completed and
water quality goals have been accomplished.
Implementation Problems: None.
Name: Liberty
Location: Spokane County, Wash.
Problem: Excessive blue-green algal growth reducing
boating, swimming, and aesthetic values.
Project Objectives: Reduce external and internal
nutrient sources; inactivate phosphorus and provide sed-
iment release barrier.
Restorative Techniques Used: Discontinue septic
tanks implementing 201 program, marsh water manipu-
lation/diversion, selective dredging, alum sulfate treat-
ment, stormwater management program.
Project Progress: Marsh water control completed;
alum treatment-dredging underway; monitoring and
evaluation continuing.
Implementation Problems: Coordination with State
game department; locating dredge spoils disposal area.
Name: Lilly
Location: Kenosha County, Wis.
Problem: In-filling with accumulated organic mater-
ials, rough fish, and reduced recreational opportunities.
Project Objectives: Restore lake fisheries and deepen
to prohibit winter fish kills.
Restorative Techniques Used: Dredging with cutter-
head hydraulic dredge.
Project Progress: Dredging completed, dredging
equipment removed, booster pumps removed, and dikes
around spoils removed. Final landscaping will be finished
in 1981.
Implementation Problems: Wet summer during 1980
prevented final spoil incorporation into the soils and
landscaping.
Name: Little Muskego
Location: Waukesha County, Wis.
Problem: Severe infilling and rooted emergents in the
near-shore area affect approximately 40 percent of the
lake.
Project Objectives: Increase recreational opportunities
and improve water quality.
Restorative Techniques Used: Dredging the lake.
Project Progress: Preliminary studies completed.
Implementation Problems: Potential arsenic contam-
ination of local groundwater supplies; State-required
EIS; local dissent to proposed disposal sites; and spiral-
ing costs.
Name: Little Pond
Location: Lincoln County, Maine
Problem: Heavy growth of zooplankton was causing
taste and odor problems in water distribution lines.
Project Objectives: Alleviate taste and odor problems.
Restorative Techniques Used: Introduce alewives to
control zooplankton population.
Project Progress: Plankton populations were reduced
and potability of Little Pond water was increased. Pro-
ject was successful.
Implementation Problems: None.
Name: Loch Raven
Location: Baltimore County, Md.
Problem: Excessive seasonal algal blooms, and high
manganese levels during every fall reservoir turnover.
Project Objectives: Insure potable water in the Balti-
more area is of high quality and free from objection-
able tastes and odors.
Restorative Techniques Used: Install a diffusive
aeration system for the purpose of destratifying the
reservoir.
Project Progress: Monitoring of reservoir water qual-
ity. Workplan development to assess different aeration
systems including wind-driven aerators and hypolimnetic
aeration is presently underway.
-------
509
Implementation Problems: Numerous procedural
delays in carrying out the project; all bids for installing
aeration system significantly exceeded the budgeted
amounts.
Name: Lone Star
Location: Douglas County, Kans.
Problem: Shallowness, excessive sedimentation, poor
water quality.
Project Objectives: Deepen lake and reduce sedimen-
tation.
Restorative Techniques Used: Control erosion by
shoreline stabilization; dredge to remove excessive sedi-
ments.
Project Progress: Work plan being developed.
Implementation Problems: None.
Name: Long
Location: Kitsap County, Wash.
Problem: Excessive algal and weed growths interfer-
ing with boating, swimming, and fishing.
Project Objectives: Reduce external and internal
sources; inactivate phosphorus and provide sediment
release barrier.
Restorative Techniques Used: Septic tank zoning and
clearing/grading ordinances; outlet area dredging; draw-
down and beach renovation; aluminum sulfate treatment.
Project Progress: Considerable improvement in water
clarity; beach improvement completed (private beach
owners) and macrophyte reduction achieved. Monitor-
ing and evaluation continuing.
Implementation Problems: Inability to obtain dredging
contractor delayed project 1 year; dredge temporar-
ily shut down due to high disposal area turbidity.
Name: Long Lake Chain of Lakes
Location: Ramsey County, Minn.
Problem: High phosphorus content in Chain of Lakes,
algal blooms severe, turbid water during storms, and
stormwater runoff.
Project Objectives: Prevent, remove, reduce, and elim-
inate pollution of Long Lake Chain of Lakes.
Restorative Techniques Used: Sedimentation basins,
channel repairs, upstream BMP's, wetlands treatment
systems, and dredging.
Project Progress: Total project is approximately 65
percent completed. Sedimentation basins, channel re-
pairs, and wetland treatment systems have been con-
structed. Dredging has not been started.
Implementation Problems: Keeping contractors on
schedule because of delays caused by weather conditions.
Name: Lower Mystic
Location: Suffolk County, Mass.
Problem: Construction of a dam in 1909 resulted in
the entrapment of 946 million liters of saltwater in
two deep kettle holes in the lake. The anoxic zone has
generated high concentrations of sulfides, ammonia, and
phosphorus.
Project Objectives: Remove salt water; aerate bottom
waters; and reduce sulfide concentrations.
Restorative Techniques Used: Pump saline water
from the lake; remove hydrogen sulfide by precipitation
with ferric chloride; and aerate bottom waters.
Project Progress: Work plan has been completed and
construction has begun.
Implementation Problems: None.
Name: Manawa
Location: Pottawattamie County, Iowa
Problem: Excessive sedimentation and aquatic
macrophyte growth.
Project Objectives: Improve water quality, deepen the
lake, and improve fishing.
Restorative Techniques Used: Dredging to remove
accumulated sediments.
Project Progress: Work plan has been accepted and
dredging is scheduled to start in late 1980.
Implementation Problems: None.
Name: Marinuka
Location: Trempealeau County, Wis.
Problem: Excessive sedimentation with consequent
large growth of nuisance aquatic plants.
Project Objectives: Removal of sediments.
Restorative Techniques Used: Dredge 653,937 cubic
meters of sediment and stabilize upstream banj
-------
510
RESTORATION OF LAKES AND INLAND WATERS
algal blooms; and occasional anaerobic conditions in the
hypolimnion.
Project Objectives: Restore water quality by removing
sediments.
Restorative Techniques Used: Dredging to increase
depth of the lake and to prevent macrophyte growth.
Project Progress: Dredging is about to begin.
Implementation Problems: Project delay due to bid
for dredging being $100,000 over planned amount.
Name: Medical
Location: Spokane County, Wash.
Problem: Excessive blue-green algae and low dissolved
oxygen preventing fish survival, boating, and swimming.
Project Objectives: Inactivate phosphorus and provide
a sediment release barrier.
Restorative Techniques Used: Aluminum sulfate
treatment.
Project Progress: Phosphorus and chlorophyl are con-
siderably reduced and blue-green algae under control.
Lake returned to high level of boating, water skiing,
swimming, picnicking, and fishing. Report available.
implementation Problems: Wind effect on alum dis-
tribution barges. Impurities in liquid alum supply.
Name: Mirror/Shadow
Location: Waupaca County, Wis.
Problem: Advanced eutrophication due tostormwater
has caused algal blooms, high phosphorus concentrations,
and fish winterkills.
Project Objectives: Divert stormwater discharges,
immobilize phosphorus in the bottom sediments, and
increase winter dissolved oxygen concentrations.
Restorative Techniques Used: Construction of new
storm sewers away from lake; application of aluminum
sulfite to precipitate phosphorus and to seal bottom; and
installation of aeration system in Mirror Lake.
Project Progress: Project completed. External phos-
phorus loading rates were reduced 65 percent, and in-
ternal phosphorus rates were also reduced. Aeration has
increased dissolved oxygen.
Implementation Problems: None.
Name: Moore
Location: Ramsey County, Minn.
Problem: Moore Lake is a shallow, eutrophic lake
maintained primarily by stormwater runoff. It has 1
meter of organic muck on top of a firm clay bottom.
Project Objectives: Halt the eutrophication of Moore
Lake.
Restorative Techniques Used: Elimination of external
phosphorus sources by diversion or treatment; inacti-
vation of nutrient in sediments; and dredging of sediment
deltas.
Project Progress: Work plan is being developed.
Implementation Problems: None.
Name: Morse Pond
Location: Norfolk County, Mass.
Problem: High nutrient loading from urban runoff
and sediments has resulted in blue-green algal blooms,
and high organic loading from deciduous leaves has
resulted in color problems.
Project Objectives: Control algae and nutrient and
organic loadings.
Restorative Techniques Used: Chemical treatment for
iron and colloidal particle removal; harvesting; dredging;
public education; and replacing deciduous trees with
evergreens.
Project Progress: Seminars have been conducted and
newspaper articles written in compliance with the educa-
tional program activities. One of two wetland areas
around the lake has been purchased as a buffer zone.
Chemical treatment has been applied to the lake. All
dredging has been completed. Project assessment is
underway.
Implementation Problems: None.
Name: Moses
Location: Grant County, Wash.
Problem: Excessive algal growths interfering with
boating, swimming, and fishing.
Project Objectives: Identify implementable agricul-
tural BMP's; have sewage treatment plant discharge
removed from lake; determine effective lake dilution
rates and volumes; implement lake dilution system using
Columbia River water.
Restorative Techniques Used: Lake dilution.
Project Progress: Pilot dilution study complete;
State EIS complete; agricultural BMP study in progress;
monitoring and evaluation continuing.
Implementation Problems: Assurance of permanent
availability of dilution water, low State 201 funding for
removal of sewage treatment plant discharge from lake.
Name: Mystic
Location: Rutherford County, N.C.
Problem: Use of the lake has been seriously impaired
by aquatic weed growth and high turbidity, both caused
by increased sedimentation.
-------
511
Project Objectives: Renovation of the lake to provide
recreational opportunities (swimming, boating, fishing,
etc.)
Restorative Techniques Used: Dredge existing sedi-
ment deposits and use dredged material for the construc-
tion of two sediment control dams. Construct spillways
and install riprap along the shore.
Project Progress: Construction has been completed.
Water quality assessment underway.
Implementation Problems: None.
Name: Noquebay
Location: Marinette County, Wis.
Problem: Excessive aquatic vegetation has greatly
reduced open water and impaired recreational value.
Project Objectives: To harvest the aquatic nuisance
plants, and to demonstrate whether weed harvesting is
a viable technique for removing nutrients that have
accumulated in a lake.
Restorative Techniques Used: Mechanical weed har-
vesting.
Project Progress: Two seasons of harvesting have been
completed. There are some indications that harvesting
results in the development of less dense but more diverse
weed patches the year after treatment. It has not yet
been determined if the nutrients in the lake are actually
being reduced.
Implementation Problems: Problems in acquiring and
maintaining harvesting machines caused considerable
delay. There have been problems quantifying the bene-
fits resulting from the project, but two independent
studies are underway to accomplish this task.
Name: North Park
Location: Allegheny County, Pa.
Problem: Excessive siltation which has caused a re-
duction in public usage of the lake.
Project Objectives: The removal of accumulated sed-
iment and the restoration of lake water quality.
Restorative Techniques Used: Dredge approximately
130,787 cubic meters of sediment.
Implementation Problems: Costs of the project have
increased significantly and project has had to be scaled
down.
Name: Nutting
Location: Middlesex County, Mass.
Problem: High nutrient levels; blue-green algae; low
transparency; nuisance aquatic vegetation; high oxygen
demand of mucky sediments; color; and organic sedi-
ment accumulation.
Project Objectives: Improve overall quality of lake for
recreational activities.
Restorative Techniques Used: Dredging and post-
dredging flocculation; control of overland runoff inputs
by street sweeping, sediment entrapment; establishment
of buffer zones; public education; and diversion of
stormwater around the lake.
Project Progress: Detailed scope of work, including
the identification of dredged material disposal areas and
program budget, has been developed. Dredging has begun
and will continue for 2 more years.
Implementation Problems: None.
Name: Oakwoods
Location: Brookings County, S.D.
Problem: Sediment loading and unstable shoreline.
Project Objectives: Bank stabilization and improve
water quality of the lake.
Restorative Techniques Used: Riprapping of shoreline
areas.
Project Progress: Eroding shoreline banks have been
stabilized.
Implementation Problems: Archeological site within
one riprap area. This caused oroiert delavs because it
required designation as eligible tor National Register and
excavation prior to finishing of project.
Name: Oelwein
Location: Fayette County, Iowa
Problem: Excessive siltation and shallowness.
Project Objectives: Improve water quality and deepen
lake.
Restorative Techniques Used: Dredge to remove
accumulated sediments; construct sedimentation ponds.
Project Progress: Dredging has been completed and
siltation ponds have been constructed.
Implementation Problems: None.
Name: Pauls Valley
Location: Garvin County, Okla.
Problem: Excessive sedimentation.
Project Objectives: Reduce sedimentation and restore
lake's water quality.
Restorative Techniques Used: Construct flood control
structures and erosion control ponds, including BMP's
in grass planting, critical area planting, cross fencing,
rotational grazing, diversion terraces, field terraces,
pasture fertilization.
Project Progress: Work plan is being developed.
Implementation Problems: None.
-------
512
RESTORATION OF LAKES AND INLAND WATERS
Name: Penn
Location: Scott County, Minn.
Problem: Dissolved oxygen depletion and urban
stormwater runoff over gardens and lawns.
Project Objectives: Aerate lake and divert stormwater.
Restorative Techniques Used: Pump well water
over stair-step outlet and introduce to the lake, and
build filter ponds at storm outlets.
Project Progress: Work approximately 85 percent
completed.
Implementation Problems: None.
Name: Phalen
Location: Ramsey County, Minn.
Problem: High phosphorus content, algal bloom, and
stormwater runoff.
Project Objectives: To restore the water quality of
the lake.
Restorative Techniques Used: Divert runoff through
marsh filter, and address upstream BMP's. Installation of
holding ponds for storm sewers.
Project Progress: Progress has been slow and the pro-
ject, in the final design stage, is only 10 percent com-
plete.
Implementation Problems: As a result of public and
neighbors' objecting to in-lake holding pond and bottom
sealing, both of which were dropped from work program,
construction has not started.
Name: Reeds
Location: Kent County, Mich.
Problem: Eutrophication at an accelerated rate, with
filamentous algal blooms and macrophyte growth in the
littoral zone in summer, and reduction of "game fish''
populations and recreational usefulness.
Project Objectives: To improve water quality.
Restorative Techniques Used: Reduction of phosphate
in surface runoff by passage and enforcement of a debris-
bagging ordinance and the City's sale of no-phosphate
fertilizers.
Project Progress: Cooperation from citizens has been
excellent. No construction has taken place.
Implementation Problems: The City signed con-
tracts without EPA approval. A little work (mostly
of a monitoring or research nature) was accomplished
before the City was asked not to use its letter of credit.
Name: Rivanna Reservoir
Location: Albermarle County, Va.
Problems: Taste and odor problems, fish kills, heavy
blooms of blue-green algae, high nutrient loading.
Project Objectives: To calculate the efficiency and
cost effectiveness of several nutrient management and
lake restorative pilot projects.
Restorative Techniques Used: Installed grassed water-
way on crop land; constructed residential sedimentation
ponds; installed an aeration system in the reservoir.
Project Progress: Aeration system has been installed
and construction of grassed waterways and sedimenta-
tion ponds has been completed. Results of the pilot
studies are being assessed.
Implementation Problems: None.
Name: Ronkonkoma
Location: Suffolk County, IM.Y.
Problem: High coliform bacteria counts and storm-
water runoff inputs of nutrients and toxic metals.
Project Objectives: Reduction in coliform bacteria,
heavy metal and nutrient inputs. Increase public uses.
Restorative Techniques Used: Diversion of stormwater
runoff; installation of biofiltration ponds; shoreline sta-
bilization.
Project Progress: Two ponds have been installed.
Preliminary data suggest that the marsh ponds can
remove significant amounts of stormwater pollutants.
Implementation Problems: Land acquisition problems
have caused significant project implementation delays.
These delays have resulted in changes in the scope of
the project. Local administrative and management
problems have been significant.
Name: Rothwell
Location: Randolph County, Mo.
Problem: Excessive siltation and inputs of nutrients.
Project Objectives: Rehabilitate the lake's silted-in
area by removing the accumulated sediments.
Restorative Techniques Used: Dredging to remove
sediments.
Project Progress: None.
Implementation Problems: No action taken because
locals are having difficulties raising matching funds.
Name: Sebasticook
Location: Penobscot, Maine
Problem: Excessive nutrient loading has led to a
condition of hypereutrophy with classical symptoms
of chronic dense algal blooms, increased vascular plant
growth, and fish kill.
Project Objectives: To improve lake water quality
and recreational opportunities.
-------
513
Restorative Techniques Used: The proposal provides
for dam reconstruction in order to permit a 3.5 meter
drawdown of the lake. The drawdown in concert with
point and nonpoint source controls is expected to sig-
nificantly improve water quality.
Project Progress: Watershed work plan is being de-
veloped. Final negotiation is underway for dredging
work.
Implementation Problems: None.
Name: Sabattus Pond
Location: Androscoggin County, Maine
Problem: In recent years, the pond has deteriorated
due, in part, to the existence of nuisance blooms over
most of the summer. In fact, water contact recreation is
severely restricted every summer.
Project Objectives: To improve lake water quality
and recreational opportunities.
Restorative Techniques Used: The proposal provides
for dam reconstruction and outlets to permit a 3-meter
drawdown of the pond. Other work includes dredging
and nonpoint source control to improve lake water
quality.
Project Progress: Watershed work plan is being devel-
oped for agricultural lands.
Implementation Problems: None.
Name: Sacajawea
Location: Cowlitz County, Wash.
Problem: Excessive algal and macrophyte growths
and turbidity affecting swimming, boating, fishing, and
picnicking.
Project Objectives: Remove external and internal
sources of nutrients and dilute with low nutrient river
water.
Restorative Techniques Used: Intercept and divert
stormwater outfalls; dilute with low nutrient Cowlitz
River water; remove nutrient sediment and macrophytes
by dredging.
Project Progress: Stormwater diversion system com-
plete; flushing/dilution system under construction;
dredging plan in progress; monitoring and evaluation
continuing.
Implementation Problems: Mt. St. Helens' mud in
Cowlitz River preventing completion of dilution water
system and using up dredged material disposal sites in
Longview area.
Name: Sacajawea
Location: Park County, Mont.
Problem: Extremely shallow, high sediment and
nutrient loading, low in-flow.
Project Objectives: Restore water quality and fish
habitat.
Restorative Techniques Used: Sediment removal;
diversion of sediment-laden in-flow tributary; flow
augmentation.
Project Progress: Bids are being let for construction
of the in-flow line. The lake has been drained to allow
for sediment excavation.
Implementation Problems: Poor estimates on project's
cost required modification and review of project scope.
Name: Salmon
Location: Kennebec, Maine
Problem: Salmon Lake once supported a diverse cold
water fishery; however, recently only brown trout were
able to maintain themselves. Obvious signs of eutrophi-
cation are apparent with noxious algal blooms occurring
frequently.
Project Objectives: To improve lake water quality
and recreational opportunities.
Restorative Techniques Used: Modification of a dairy
farm drainage area. A 3-year construction phase during
which diversions, tiles, a storage lagoon, and irrigation
system will be built.
Project Progress: Watershed work plan being devel-
oped. Project implementation is awaiting results of
Cobbossee I project.
Implementation Problems: None.
Name: Scudders Pond
Location: Nassau, N.Y.
Problem: Excessive sedimentation, stormwater runoff,
advanced state of eutrophication, blue-green algal blooms.
Project Objectives: Remove excessive sediments;
increase public use.
Restorative Techniques Used: Dredge sediments;
control incoming sedimentation problem by construct-
ing stormwater retention basins.
Project Progress: Work plan completed, project bid
accepted, contract let for dredging.
Implementation Problems: Obtaining local matching
funds.
Name: Skinner
Location: Noble County, Ind.
Problem: Sediment and nutrient-contaminated run-
off from the lake's agricultural watershed is causing
sediment buildup at the stream outlet and weed growth
around the shallow edge of the lake.
-------
514
RESTORATION OF LAKES AND INLAND WATERS
Project Objectives: Reduce sediment and nutrient
runoff in combination with sediment settling and nutri-
ent filtering; remove accumulated sediments and use
weed removal and chemical application to eliminate
existing weed growth in lake.
Restorative Techniques Used: Control of sediment
and agricultural runoff pollutants by conservation prac-
tices, channel stabilization, and dredging.
Project Progress: Watershed conservation practices
have been implemented. Large sediment basin will be
constructed in 1981 and channel stabilization and sed-
iment dredging will also begin in 1981.
Implementation Problems: Local matching funds for
large sediment basin, dredging, and channel stabilization
were less than originally planned and progress was de-
layed. Farmers in the project area have rejected non-
structural conservation practices such as reduced tillage
and the use of cover crops in favor of structural prac-
tices, usually parallel tile outlet terraces.
mented; removal of sediment will be initiated in late
1980.
Implementation Problems: Project delays due to
drought conditions experienced during 1976-1977 that
prevented the drawdown of Stafford Lake.
Name: Steinmetz
Location: Schenectady County, N.Y.
Problem: Sediment accumulation, macrophyte prob-
lem, accelerated eutrophication.
Project Objectives: Remove sediments, improve water
quality.
Restorative Techniques Used: Dredge sediments,
place sand on beach areas, divert stormwater runoff.
Project Progress: Project has been completed. Turbid-
ity has decreased, water quality has improved, public
usage has increased.
Implementation Problems: None.
Name: Spada/Chaplain
Location: Snohomish County, Wash.
Problem: Increased turbidity preventing adequate
disinfection of raw water supply serving 200,000 people.
Project Objectives: Identify turbidity sources; develop
and select best turbidity control plan; develop and adopt
interjurisdictional basin resource management plan.
Restorative Techniques Used: Stream channel modi-
fication-riprapping; gabion construction around blue
clay outcroppings; selected slope area revegetation;
resource management plan.
Project Progress: Project completed—reduced tur-
bidity to acceptable level for simple chlorination.
Implementation Problems: Obtaining agreement
among jurisdictions, i.e., County, State (DNR & Health),
USFS, and private owners on objectives and manage-
ment practices.
Name: Stafford
Location: Marin County, Calif.
Problem: Eutrophication as evidenced by algal
blooms, high levels of organic matter associated with
lake sediment, seasonally high nutrient levels, and high
coliform bacteria concentrations.
Project Objectives: Control of organic and nutrient
inputs.
Restorative Techniques Used: Dry excavation of sedi-
ment from the lake and erosion control. Spoil to be used
in expansion of present park area.
Project Progress: Necessary property for buffer zone
has been purchased; erosion control has been imple-
Name: Summit
Location: Summit County, Ohio
Problem: Accelerated eutrophication caused by urban
nonpoint source runoff.
Project Objectives: Remove nonpoint source nutrient
inputs to the lake.
Restorative Techniques Used: Retention and/or diver-
sion of stormwater runoff and in-lake aeration have been
tentatively identified as restorative techniques.
Project Progress: Work plan developed to study storm-
water runoff abatement practices.
Implementation Problems: None.
Name: Sunset
Location: Texas County, Okla.
Problem: Sedimentation has caused water quality
problems.
Project Objectives: Stop rapid sedimentation; restore
water quality; improve fish habitat; repair dam and over-
flow pipe; and dredge lake.
Restorative Techniques Used: Draining and sediment
removal; shoreline stabilization by vegetation and soil
cement; and construction of upstream impoundments.
Project Progress: Work plan formulation underway.
Sensibility study has been implemented.
Implementation Problems: None.
Name: Swan
Location: Turner County, S.D.
Problem: High sediment loading due to shoreline
wave erosion.
-------
515
Project Objectives: Reduce sediment loading and
stabilize bank area.
Restorative Techniques Used: Riprapping of shoreline
and renovation of outlet structure.
Project Progress: Shoreline areas have been stabilized.
Implementation Problems: Inclement weather caused
numerous delays during riprapping and construction.
Name: Sylvan
Location: Custer County, S.D.
Problem: Excessive sedimentation influx.
Project Objectives: Protect lake from future sedi-
mentation; reduce erosion of surrounding areas.
Restorative Techniques Used: Construction of erosion
control structures in camping area, re-seeding and mod-
ification of parking area to redirect runoff.
Project Progress: Project is being delayed until State
legislature authorizes acceptance of Federal money.
Implementation Problems: None to date.
Implementation Problems: Source control program
has been hard to institute: lack of resources and reluc-
tance of city to regulate construction; inability to con-
trol activities in watershed.
Name: Thurston Lakes (Long, Patterson, Hick & Lois)
Location: Thurston County, Wash.
Problem: Excessive algal and weed growths interfering
with boating, swimming, fishing, and picnicking.
Project Objectives: Evaluate and establish storm and
on-site wastewater management program, agricultural
BMP's, and in-lake restorative needs.
Restorative Techniques Used: On-site wastewater
management programs; storm water management pro-
gram; agricultural BMP program; in-lake procedures
{to be determined).
Project Progress: Initial work plan being developed.
Implementation Problems: None.
Name: Tahoe
Location: Washoe and Douglas Counties, Nev.
Problem: Development has increased sediment and
nutrient loading, causing an increase in primary produc-
tivity and algal growth in near shore areas.
Project Objectives: Control sediment and nutrient
contributions in critical erosion areas.
Restorative Techniques Used: Erosion control and
slope stabilization structures including rock-lined ditches,
rock slope protection, gabion walls, and revegetation.
Project Progress: Plans and specifications approved
by EPA on July 25, 1980. Project now in construction
bid stage. Grant offer dated July 23, 1980. Offer has not
yet been accepted by State of Nevada.
Implementation Problems: The award for Kingsbury
Grade has not been accepted because of problems with
local funding.
Name: Temescal
Location: Alameda County, Calif.
Problem: Nutrients, coliforms, and sediment are main
problems.
Project Objectives: Remove sediments and implement
nutrient control program.
Restorative Techniques Used: Dredging sediments;
install retention pond; implement source control through
city regulation of grading.
Project Progress: Sediments have been removed and
retention ponds are being installed.
Name: Tivoli
Location: Albany County, N.Y.
Problem: Accumulated raw sewage sludge sediments;
stormwater runoff; siltation caused by soil erosion; and
pollutants from old, broken sewer lines.
Project Objectives: Clean up water ecosystem; stabil-
ize soil; and develop associated ponds and wetlands.
Restorative Techniques Used: Develop shallow water
areas and wetland areas upstream to retard stormwater
runoff and reduce siltation; drain and excavate main lake
to a maximum depth of 8 to 10 feet; regrade banks and
vegetate to prevent erosion; redesign and rebuild existing
earthen dike and emergency spillway.
Project Progress: Project completed; accumulated
sediments have been removed; water quality has been
improved as has public usage of the lake.
Implementation Problems: None.
Name: Upper Willow
Location: St. Croix County, Wis.
Problem: Excessive sediment and excessive growth of
emergent and submergent plants.
Project Objectives: Removal of sediments.
Restorative Techniques Used: Riprapping, sloping
and mulching, seeding, dredging, and installing a sediment
trap.
Project Progress: Work plan is being developed.
Implementation Problems: None.
-------
516
RESTORATION OF LAKES AND INLAND WATERS
Name: Vancouver
Location: Clark County, Wash.
Problem: Siltation, excessive algal growths, and high
coliform preventing summertime use of the lake.
Project Objectives: Control external sediment and
other pollution sources; remove sediments; deepen and
contour for maximum circulation; provide dilution water
system.
Restorative Techniques Used: Watershed 208 water
quality management plan; dredging; flushing (dilution)
with Columbia River water.
Project Progress: Pilot dredge study complete; NEPA-
Wetland study complete; operations plan complete;
permits applied for and bid documents being prepared
for dredging and flushing channel construction; monitor-
ing and evaluation continuing.
Implementation Problems: Approval of dredge spoils
disposal sites; flushing channel design to prevent salmon
migration into lake; State hydraulics and Corps of Engi-
neers 404 permits.
Name: Vandalia Reservoir
Location: Pike County, Mo.
Problem: Siltation from stormwater runoff has re-
duced the storage capacity of Vandalia Reservoir by
50 percent.
Project Objectives: Improvement of lake water quality
and restoration of the impoundment to its original
capacity.
Restorative Techniques Used: Dredging of 137,195
meters of bottom sediment and construction of sedi-
ment catchment basins in the watershed.
Project Progress: Dredging has been completed. Final
assessment of water quality is underway.
Implementation Problems: None.
Name: Lake Wapato
Location: Pierce County, Wash.
Problem: Excessive algal and weed growths interfering
with swimming, fishing, and other recreational uses.
Project Objectives: Reduce external and internal
sources of nutrients and provide low nutrient dilution
water.
Restorative Techniques Used: Stormwater detention
basin and diversion system; drawdown for weed control
and bottom compaction;dilution system using city Cedar
River supply.
Project Progress: Dilution experimental study com-
plete; revised plan complete; final design started; moni-
toring and evaluation continuing.
Implementation Problems: Scheduling drawdown and
diversion system construction for least impact on park
activities and lake use.
Name: Lake Waramaug
Location: Kent County, Conn.
Problem: The extensive summer and fall blue-green
algal blooms in the lake are the most obvious symptoms
of the lake's eutrophication problems. Agricultural
runoff from barnyards, feedlots, etc., is a major source
of the pollution.
Project Objectives: To improve lake water quality
and recreational opportunities.
Restorative Techniques Used: Restoration includes
implementation of conservation practices; local land use
controls; comprehensive information, education, and
public participation programs; water quality monitoring;
and project coordination.
Project Progress: Watershed work plan development is
underway.
Implementation Problems: None.
Name: Washington Park
Location: Albany County, N.Y.
Problem: Increased lake nutrient levels; reduced trans-
parency and lake depth; excessive aquatic weed growth.
Project Objectives: Improve overall lake water quality.
Restorative Techniques Used: Drain lake and remove
bottom sediment by dredging.
Project Progress: Post-restoration monitoring shows
an improvement in lake transparency and an elimination
of aquatic weed growth along the shorelines.
Implementation Problems: None.
Name: Waterford
Location: Anne Arundel County, Md.
Problem: Insufficient storm drainage system causing
erosion and water quality problems.
Project Objectives: Improve the water quality of the
lake by reducing the bank and shoreline erosion and the
resultant siltation and suspended solids problem in the
lake.
-------
517
Restorative Techniques Used: Construction of closed
storm drainage system, timber bulkheading, and gabion
shores protection.
Project Progress: Work plan has been approved and
project construction is about to begin.
Implementation Problems: None.
Name: White Clay
Location: Shawano County, Wis.
Problem: The lake has become eutrophic because of
.phosphorus loading from animal wastes.
Project Objectives: To reduce phosphorus loading
from animal wastes and from cropland runoff.
Restorative Techniques Used: Seventeen barnyard
storage facilities were built. Farmers cooperated in
spreading animal wastes onto fields when they thawed.
Grassed waterways, terraces, diversions, and reduced till-
age were instituted.
Project Progress: Project completed. Total phos-
phorus loading from all sources of animal waste is esti-
mated to have been reduced from 451 kg in 1970 to
342 kg in 1978, a reduction of 25 percent.
Implementation Problems: Landowners had to pay
for. manure storage facilities and wait for reimbursement,
resulting in some reluctance and delay in the early stages.
CLEAN LAKES PHASE II IMPLEMENTATION PROJECTS
Restoration/Preservation Techniques
In-lake Techniques
1. Dredging/Sediment Removal
2. Aeration/ De stratification
3. Flushing/Dilution
4. Nutrient Precipitation/lnactivation
5. Drawdown/Waterlevel Manipulation
6. Macrophyte Control
7,'Biomanipulation
8. Sediment Sealing
Watershed Techniques
9. Agricultural BMP's
lO.Stormwater Control
11. Erosion Control
12.Tributary Diversion/Treatment
Restoration Techniques
STATE and LAKE
California
Ellis
Gibraltar
Lafayette
Stanford
Temescal
Connecticut
Warramug
Bantam
Florida
Apopka
Eola
Jackson
Illinois
Frank Holten
Indiana
Skinner
T
X
X
projec
X
X
X
srojec
X
X
2
term
term
3
X
nate<
natec
4
prior
prior
8
o res
0 res
6
X
oratio
oratio
i
*
i. •
8
X
§
•x '
X
X
10
X
X
X
X
X
X
X
ft
x;
X
X
X
X
X
X
-------
518
RESTORATION OF LAKES AND INLAND WATERS
Restoration Techniques
STATE and LAKE
Iowa
Blue
Green Valley
Lenox
Manawa
Olwein
Kansas
Lone Star
Louisiana
City
Maine
Cobbossee I
Cobbossee II
Little
Sabasticook
Sabattus
Salmon
Maryland
Loch Raven
Waterford
Massachusetts
Big Alum
Charles River
Cochituate
Ellis Brett
Lower Mystic
Morse
Nutting
Michigan
Lansing
Reeds
Minnesota
Albert Lea
Clear
Hyland
Long Lake Chain
Penn
Phalen
Moore
Clearwater River Chain
Harriet/Isles
Missouri
Creve Coeur
Finger
Rothwell
Vandalia
Montana
Sacajawea
Nevada
Tahoe I
Tahoe II
New Jersey
Allentown
New York
Ann Lee
Buckingham
Collins
Delaware Park
Fifty-ninth Street
Hampton Manor
Hyde Park
Ronkonkoma
Scudders
Steinmetz
Tivoli
Washington Park
North Carolina
Mystic
12 3 4 5
X
X
X
X
X
X
x •' ; : : ••
X X
X
X
X
X
X
X
X X
X X
X
X
X
X X
X
X X
X
XX
X X
X
X
X
X
X
X
X
X
X X
X
X X
X
x x
X
X
X
X X
X
X
X
6 7 8 9 10 11
x x
x :• x- :
x
X
XX X
X X
XXX
X
X
X
»• : x ^f;,..
',•>, x x
"X'i ."'• X X
X X
X X
X X X
XXX
X X
X
XXX
X X
XX X
X X X
X
X
X X
X
X
X
X
X
X X
X X
X X
X X
X X
X
X X
X X
X X
x x
X X
X 1 X X
X X
X
^
\ 12
'; x
X
X
X
x
,:',' x
-------
519
Restoration Techniques
1 2 3 4 5 6 .?,, 8 9 10 It 12
STATE and LAKE '"' - 1
Ohio
Summit x
Oklahoma
Pauls Valley x x
Sunset x X
Oregon
Commonwealth
Pennsylvania
North Park
South Carolina
Broadway
South Dakota
Cochrane x .-\.. x
Covell x - ,-' x
Herman X -. • < H.
Kampeska •>" • 4 x
Oakwood , .
Swan ",;-,' . X
Sylvan x ;'^..'-4, -..?t--.- x
Texas - ,<>
McQueeney X x
Vermont
Bomoseen
Virginia „
Rivanna x ;.;,, -^, a.-.S.j>- X
Washington •'•!,'
Ballmger x X - ^ .,
Fenwick , •'* , ~ 't-«;, x
Liberty x x ,' , . ,X, - x
Long X x - ,--, ' „'%, x
Medical X '._," - t '-'
Moses X X ! -'', }' v
Sacajawea x ""•-•, \<~- x
Spada/Chaplam x X
Thurston , ' •' * x
Vancouver X ,,-"" ^. x
Wapato x X x :* \-' ...'..!>•.. x
Wisconsin -.-
Bugle x
Decorah x , ; X
Half Moon x x
Henry x X
Lilly X x ~; -
Little Muskego X x 'V
Mannuka x ,
Mirror/Shadow x x x
Noquebay X x "„; k x
Upper Willow x ",. f. " *'•*
White Clay »-'it-*!; S-X. .
-------
520
Appendix B
SYMPOSIUM PARTICIPANTS
Co-chairs:
PLENARY SESSIONS
Opening Session
Paul D. Uttormark
11 Coburn Hall
University of Maine
Orono, Maine 04469
Richard A. Vollenweider
Canada Centre for Inland Waters
Box 5050
Burlington, Ontario, Canada L7R 4A6
Modeling and Assessment of the Trophic State
Kenneth H. Reckhow
323 Natural Resources Bldg.
Michigan State University
East Lansing, Mich. 48824
Richard A. Vollenweider
Special Topics
Heinz Bernhardt
Wahnbachtalsperrenverband
Siegels Knippen D 52 Siegburg
Siegburg, Germany
William H. Funk
141 Sloan Hall
Washington State University
Pullman, Wash. 99164
The Acid Rain Problem: Mechanism and Effects
Stephen A. Norton
Boardman Hall, 110
University of Maine
Orono, Maine 04469
Brynjulf Ottar
Norwegian Institute for Air Research
P. 0. Box 130, N-200I LILLESTROM
Norway
Conclusions and Guidelines
Gerard Dorin
OECD-Environment Directorate
2 Rue Andre Pascal
Paris, France 75016
David G. Frey
Indiana University
Bloomington, Ind. 47405
WORKING SESSIONS A
Factors Influencing the Dynamics
of Eutrophication
Jurgen Clasen
Wahnbachtalsperrenverband
Siegel Knippen D52 Siegburg
Siegburg, Germany
G. Richard Marzolf
Kansas State University
Manhattan, Kan. 66506
Nutrient Loading/Trophic Response
Kenneth H. Reckhow
Richard A. Vollenweider
Public Benefit and Institutional Problems
Douglas A. Yanggen
University of Wisconsin Extension
1815 University Ave.
Madison, Wis. 53706
Lowell Klessig
University of Wisconsin Extension
1815 University Ave.
Madison, Wis. 53706
-------
521
Special Projects and Topics for Assessing
the Trophic State
Heinz Bernhardt
William H. Funk
Health-Related Problems
David E. Armstrong
University of Wisconsin
114 University Bay Drive
Madison, Wis. 53705
Michael J. Suess
WHO Regional Office for Europe
8 Scherfigsrej
Copenhagen, Denmark 2100
WORKING SESSIONS B
Dredging and Biomanipulation as
Restoration Techniques
Spencer A. Peterson
U.S. Environ. Prot. Agency
200 S.W. 35th Street
Corvallis, Ore 97330
Peter Sly
Canada Centre for Inland Waters
Glenora Fisheries Station, RR4
Picton, Ontario KDK 2TO
Canada
Aeration/Mixing and Aquatic Plant Harvesting
as Restoration Techniques
Deric Johnson
W.R.C. Medmenham Laboratory, Henley Road
Medmenham, Marlow, Buckinghamshire
SL7 2HD England, UK.
Marc Lorenzen
Tetra Tech, Inc.
1900 116th Avenue N.E.
Bellevue, Washington 98004
Peter R. Newroth
B.C. Ministry of Environment
Parliament Buildings
Victoria, British Columbia
Canada V8V 1X5
Rural Watershed Pollution Control
H. L. Golterman
Biology Station
le Sambuc, 13200 Aries, Tour du Valat
Aries, France 13200
Walter F. Rittall
U.S. Environ. Prot. Agency
401 M. Street S.W., WH-554
Washington, D. C. 20460
Urban and Point Source Pollution
Control Technology
Richard Field
U. S. Environ. Prot. Agency
Building 10, Woodbridge Ave.
Edison, N.J. 08817
Curt Forsberg
Institute of Limnology, Box 557
75122 Uppsala
Uppsala, Sweden
Nutrient Prevention and Inactivation
G. Dennis Cooke
Kent State University
Kent, Ohio 44242
Valerie May
National Herbarium of N.S.W.
Royal Botanic Gardens
Sydney, N.S.W. Australia
-------
522
RESTORATION OF LAKES AND INLAND WATERS
Speakers:
Riaz Ahmed
Center for the Environment & Man, Inc.
275 Windsor St.
Hartford, Conn. 06120
David J. Allee
Cornell University
218 Warren Hall
Ithaca, N.Y. 14853
Martin T. Auer
University of Michigan
Room 115 Engineering Bldg. 1-A
University of Michigan
Ann Arbor, Mich. 48109
Mark Brown
N. Y. State Dep.
Environ. Conserv.
50 Wolf Rd., Room 519
Albany, N.Y. 12233
Tom Brydges
Ontario Ministry of the Environment
Box 213 Rexdale
Ontario, Canada
Robert Carlson
Dep. of Biological Sciences
Kent State University
Kent. Ohio 44242
Roger Bachmann
Iowa State University
Dep. Animal Ecology
Ames, Iowa 50011
G. Barroin
I.N.R.A. Sta d'Hydrobiologie Lacustre
75, Av. de Corzent
Thonon, France 74203
A.F. Bartsch
3238 N.W. Gumwood Dr.
Corvallis, Ore. 97330
E. B. Bennett
National Water Research Institute
Canada Centre for Inland Waters
Burlington, Ontario L7R 4A6
Jay Bloomfield
N.Y. State Dep. Environ. Conserv.
Fort George Rd.
Lake George, N.Y. 12845
Nicolaas Bouwes
Agricultural Economics Dep.
University of Wisconsin
Madison, Wis. 53706
Leslie Carothers
U.S. Environ. Prot. Agency, Reg. I
JFK Federal Bldg.
Boston, Mass. 02203
Steven C. Chapra
Great Lakes Environ. Research Lab.
2300 Washtenaw Ave.
Ann Arbor, Mich. 48104
Neils Christiansen
U.S. Environ. Prot. Agency
200 S.W. 35th St.
Corvallis, Ore. 97330
David R. Dominie
Maine Dep. of Environ. Prot.
Augusta, Maine 04333
Russell Dunst
Wisconsin Dep. Natural Resources
1022 Sequoia Trail
Madison, Wis. 53713
Alan W. Elzerman
Environmental Systems Engineering
Clemson University Rhodes Center
Clemson, S.C. 29631
-------
523
Charles E. Fogg
U.S. Dept. Agric. Soil Conserv. Serv.
South Building, P. 0. Box 2890
Washington, D.C. 20013
Bruce R. Forsberg
Limnological Research Center
University of Minnesota
310Pillsbury Dr.
Minneapolis, Minn. 55455
Hansjorg Fricker
Swiss Federal Institute of Water Resources
8600 Dubendorf, Switzerland
James Galloway
Clark Hall, Environmental Sciences
University of Virginia
Charlottesville, Va. 22903
Anthony Gasperino
Battelle-Northwest
Box 999
Richland, Wash. 99352
George R. Gibson, Jr.
University of Wisconsin
Environmental Resources Unit
1815 University Ave.
Madison, Wis. 53706
Thomas U. Gordon
Cobbossee Watershed District
15 High St.
Winthrop, Maine 04364
Dennis J. Gregor
Environment Canada
P.O. Box 5050
Burlington, Ontario, Canada L7R4A6'
Herbert J. Grimshaw
Oklahoma Water Resources Board
2833 S. W. 86th St.
Oklahoma City, Okla. 73159
Francois Guimont
Gouvernement Quebec
1640, Boul. de I'Entente
Quebec, P.O. G1S 4N6 Canada
Terry A. Haines
U.S. Fish & Wildlife Service
University of Maine
Orono, Maine 04469
George Hendrey
Environmental Sciences Group
Brookhaven National Laboratory
Upton, N.Y. 11973
Frank Humenik
Biological and Agricultural Eng. Dep.
North Carolina State University
Raleigh, N.C. 27650
Mark L. Hutchins
11 Coburn Hall
University of Maine
Orono, Maine 04469
Dieter M. Imboden
EAWAG, Ueberlandstrasse 128
CH-8600 Duebendorf, Switzerland
T. A.Jackson
Freshwater Institute
501 University Crescent
Winnipeg, Manitoba
R3T 2N6, Canada
Robert Kennedy
U.S. Army Corps Engineers
Waterways Experiment Station
Vicksburg, Miss. 39180
Joseph J. Kerekes
Environment Canada
Canadian Wildlife Service
Halifax, Canada N.S. B3H4J1
Darrell L..King
Institute of Water Research
Michigan State University
East Lansing, Mich. 48824
Douglas Knauer
Wisconsin Dep. of Natural Resources
P.O. Box 7921
Madison, Wis. 53707
-------
524
RESTORATION OF LAKES AND INLAND WATERS
James R. Kramer
Dep. of Geology
McMaster University
Hamilton, Ontario L854M1
David P. Larsen
U.S. Environ. Prot. Agency
200 S.W. 35th St.
Corvallis, Ore. 97330
David R. Lee
Atomic Energy of Canada
Chalk River
Ontario, Canada KOJ IJO
Kenneth M. Mackenthun
Enwright Laboratories, Inc.
104 Tower Dr.
Greenville, S.C. 29607
Diane F. Malley
Fisheries & Oceans
501 University Crescent
Winnipeg, Manitoba, Canada R3T 2N6
Madonna F. McGrath
U.S. Environ. Prot. Agency
536 S. Clark St., Room 932
Chicago, III. 60605
Lee A. Mulkey
U.S. Environ. Prot. Agency
College Station Rd.
Athens, Ga. 30613
Robert Pastorok
Tetra Tech, Inc.
1900 116th Ave. N.E.
Bellevue, Wash. 98004
Michael A. Perkins
Dep. of Civil Engineering
University of Washington
Seattle, Wash. 98195
William C. Pisano
EDP, Inc.
257 Vassar St.
Cambridge, Maine 02139
Oscar Ravera
Commission of European Communities
Euratom J.R.C.
Ispra (Varese) Italy 21020
Wilhelm Ripl
Technical University
Hellriegelstr. 6
1000 Berlin 33
West Germany
Steve Schatzow
U. S. Environ. Prot. Agency
401 M S.W.
Washington, D.C. 20460
C. L. Schelske
Great Lakes Res. Div.
University of Michigan
Ann Arbor, Mich. 48109
Lynn R. Schuyler
U.S. Environ. Prot. Agency
Robert S. Kerr Environ. Research Lab.
Ada, Okla. 74820
Val H. Smith
Limnological Research Center
University of Minnesota
Minneapolis, Minn. 55455
V. Michael Stallard
Metcalf & Eddy, Inc.
106 K St. Suite 200
Sacramento, Calif. 95814
Robert E. Stauffer
Water Chemistry Laboratory
University of Wisconsin
Madison, Wis. 53706
Heinz G. Stefan
University of Minnesota
Mississippi River at 3rd Ave. S.E.
Minneapolis, Minn. 55414
Project Manager
Robert J. Johnson
Editor
Judith F Taggart
-------
Appendix C
SYMPOSIUM ATTENDEES
525
Alva Achorn
Maine Dep. Environ. Prot.
State House, Station 17
Augusta, Maine 04333
Michael T. Ackerman
Mass. Div. Water Pollution Control
Box 545
Westborough, Mass. 01581
David F. Aitkens
Ontario Ministry of Environment
Southeastern Region
133 Dalton St.
Kingston, Ontario, Canada K7L4X6
Marshall K. Akers
Air Products & Chemicals, Inc.
P.O. Box 538
Allentown, Penn. 18105
Peter J. Alexander
East Bay Regional Pk. Dist.
11500 Skyline Blvd.
Oakland, Calif. 94619
W.J.R. Alexander
Dept. of Water Affairs, Forestry & Environ. Conserv.
c/o S. African Embassy
Suite 300, 2555 M St. N.W.
Washington, D.C. 20037
Edna Allen
Menardi-Southern
West Lake Road
Cossayuna, N.Y. 12823
Robert A. Allen
Menardi-Southern
West Lake Road Box 145
Cossayuna, N.Y. 12823
Dale E. Anderson
URS Company
Fourth and Vine Building
Seattle, Wash. 98121
J.O. Anderson
Bantam Lake Rehab.
Morris Ct.
Litchfield, Conn. 06759
Norman Anderson
Bureau of Air Quality Control
Dept. Environ. Prot.
State House, Station 17
Augusta, Maine 04333
Terry P. Anderson
Kentucky Division of Water Quality
1065 U.S. 127 Bypass South
Century Plaza
Frankfort, Ky. 40601
Desmond D. Anthony
Nipissing University
P.O. Box 5002, Gormanville Rd.
North Bay, Ontario, Canada P1B 8L7
James M. Arnold
Mass. Div. Water Pollution Control
P.O. Box 545
Westboro, Mass. 01581
F.M. Atton
Sask. Government Cam.
30 Campus Dr.
Saskatoon, Sask., Canada S7N 0X1
Donald B. Aulenbach, Ph.D.
R.P.I.
Rensselaer Polytechnic Institute
Ricketts Building 102
Troy, N.Y. 12181
Don Aurand
The MITRE Corporation
1820 Dolley Madison Ave.
(Mail Stop W263)
McLean, Va. 22102
Apley Austin, Jr.
Bantam Lake Rehab.
E. Shore Road
Morris, Conn. 06763
-------
526
RESTORATION OF LAKES AND INLAND WATERS
Benjamin K. Ayers, Jr.
L.R.P.C.
Route 62, Box 435
Center Harbor, N.H. 03226
John Bailey
Maine Legislative Staff
State House, Station 13
Augusta, Maine 04333
OrvilleP. Ball
Orville P. Ball & Assoc.
8755 Vista del Verde
El Cajon, Calif. 92021
R.A. Bannink
Dutch Ministry of Pub. Works & Transport.
Duunmede 38
Middelburg, Netherlands
David A. Bare
Fla. Sugar Cane League, Inc.
115 South Lopez
P.O. Box 1148
Clewiston, Fla. 33440
James L. Barker
U.S. Geological Survey
Federal Building
Box 1107
Harrisburg, Penn. 17108
John W. Barko
WES-Environmental Lab
P.O. Box 631
Vicksburg, Miss. 39180
John Barten
City of Waseca
508 S. State St.
Waseca, Minn. 56093
Gerald Bates
Bureau of Health
Augusta, Maine 04333
Ralph Bautz
Key Engineers
80 S. White Horse Pike
Berlin, N.J. 08009
Susan G. Beck
Iowa State University
Dept. of Animal Ecology
124 Science II
Ames, Iowa 50011
Richard J. Benoit, Ph.D.
Mohegan College
Mahan Drive
Norwich, Conn. 06360
William J. Bergstresser
c/o Penn. Power & Light Co.
Route 4
Honesdale, Penn. 18431
Mark Bernard
City of Dartmouth
P.O. Box 817
Dartmouth, Nova Scotia B2Y3Z3
Robert T. Berrisford
USDA Forest Service
Chippewa National Forest
Supervisors Office
Cass Lake, Minn. 56633
Wendell Berry, Jr.
N.H. Association of Conservation Commissions
45 Sherwood Dr.
Hooksett, N.H. 03106
Douglas E. Bertrand
Central Maine Power Co.
P.O. Box 196
Searsport, Maine 04974
Paul H. Bewick
Ontario Ministry of Natural Resources
Postal Bag 2002
Kemptville, Ontario, Canada KOG 1JO
Harrison Bispham
Maine Dept. of Environ. Prot.
State House, Station 17
Augusta, Maine 04333
Frederic C. Blanc
Dept. of Civil Eng.
Northeastern University
Boston, Mass. 02115
Jeff Bode
Wisconsin Dn. R.
11611 W. North Ave.
Milwaukee, Wise. 53226
Edwin O. Boebel
Wise. Dept. of Nat. Res.
Box 7921
Madison, Wis. 53707
-------
527
Phillippe Boissonneault
Portland Water District
225 Douglass St.
Portland, Maine 04104
Fred Bonner
Del. Div. of Fish & Wildlife
Box 425
Little Creek, Del. 19961
Warner P. Boortz
Nekoosa Papers Inc.
100 Wisconsin River Dr.
Port Edwards, Wis. 54469
John S. Brasino
University of Vermont
Dept. of Zoology
Burlington, Vermont 05405
Benjamin W. Breedlove
Breedlove Associates Inc.
618N.W. 13th Ave.
Gainesville, Fla. 3260I
Jennie E. Bridge
N.E.I.W.P.C.C.
607 Boylston St.
Boston, Mass. 02116
Charles L. Boothby
Natl. Assn. of Conservation Districts
1025 Vermont Ave., N.W., Room 730
Washington, D.C. 20005
Daryll C. Borst
Quinnipiac College
Dept. of Biological Sciences
Hamden, Conn. 06518
Mario M. Boschetti
Mass. Dept. of Environ. Qual. Eng.
100 Nashua St., Room 532
Boston, Mass. 02114
L.D. Bowen
Sydney Water Board
P.O. Box A53
Sydney, South NSW 2000
Sydney, NSW Australia
Sylvia Bradeen
University of Maine at Orono
50 Bosworth St.
Old Town, Maine
Norris D. Braley
Time & Tide RC & D
Route 1
Waldoboro, Maine 04572
Norman Brandel
U.S. Environ. Prot. Agency
401 M St. S.W.
Washington, D.C. 20460
Jeff Brandow
Environmental Eng.
University of Maine
451 Aubert Hall
Orono, Maine 04469
Douglas L. Britt
International Research & Technology Corp.
7655 Old Springhouse Rd.
McLean, Va. 22102
Richard R. Bronaugh
Kansas Dept. of Health & Environ.
Building 740, Forbes Field
Topeka, Kans. 66620
J. Willcox Brown
N.H. Acid Rain C.C.
Dunbarton, N.H. 03301
William E. Brown
Wright-Pierce Engineers
99 Main St.
Topsham, Maine 04086
Frank Browne
F.X. Browne Associates
P.O. Box 401
Lansdale, Penn. 19446
William F. Brutsaert
University of Maine at Orono
Dept. of Civil Engineering
Orono, Maine 04473
Nancy Bryant
Vt. Dept. of Water Res.
Box 28
Jericho Court, Vt. 05465
John Brzozowski
New Jersey Dept. Div. Water Resources
P.O. Box CN-029
Trenton, N.J. 08625
Robert C. Bubeck
U.S. Environ. Prot. Agency, Region III Lab.
839 Bestgate Road
Annapolis, Maryland 21401
-------
528
RESTORATION OF LAKES AND INLAND WATERS
James H. Buckler
City Water Light & Power
Municipal Building
7th & Monroe
Springfield, III. 62757
Ken Burger
East Bay Regional Pk Dist.
11500 Skyline Blvd.
Oakland, Calif. 94619
David Burmaster
Council on Environmental Quality
722 Jackson Place, N.W.
Washington, D.C. 20006
Douglas Burnham
Dept. Water Resources
State Office Building
Montpelier, Vt. 05602
Dan Burrows
U.S. Environ. Prot. Agency
40! M St. S.W.
Washington, D.C. 20460
Anthony D. Burton
N/E Mich. Council of Govts.
P.O. Box 457
408 W. Main St.
Gaylord, Mich. 49735
R. Edward Burton
E.B.C. Company
222 Franklin St.
Willits, Calif. 95490
Richard S. Burton
Monroe Co. Health Dept.
111 Westfall Rd.
Rochester, N.Y. 14602
Don Bushey
E.L. Jordan Co.
562 Congress St.
Portland, Maine 04110
Don Buso
Cornell University
U.S.F.S. Station
West Thornton, N.H. 03285
William J. Butler
U.S. Environ. Prot. Agency
JFK Federal Building
Boston, Mass. 02203
Gordon L. Byers
WTR Resource Research Ctr.
University of New Hampshire
Pettee Hall, Room 108
Durham, N.H. 03824
Edward Callender
U.S. Geological Survey
National Center, MS 432
12001 Sunrise Valley Dr.
Reston, Va. 22092
Paul Campanella
J.R.B. Associates, Inc.
8400 Westpark Dr.
McLean, Va. 22102
Daniel E. Canfield, Jr.
University of Florida
Aquatic Plant Research Center, Bldg. 737
Gainesville, Fla. 33611
Italo Carcich
New York State Environ. Conserv. Dept.
Bureau of Water Research
50 Wolf Road
Albany, N.Y. 12233
Carlos Carranza, Ph.D.
Springfield College
Box 1841
Springfield, Mass. 01109
Terrance Carter
Water Quality Control—Colorado
4210 E. 11th Ave.
Denver, Colo. 80220
Leighton Carver
Bureau of Air Quality Control
State House, Station 17
Augusta, Maine 04333
Charles Chakoumakos
Dept. of Chemistry
University of Maine
Farmington, Maine 04938
Alice Chamberlin
New Hampshire Lung Assoc.
456 Beech St.
Manchester, N.H. 03105
Joanne Chance
Washington State Dept. of Ecology
Mail Stop PV-11
Olympia, Wash. 98504
-------
529
John Chandler
Bureau of Air Quality Control
State House, Station 17
Augusta, Maine 04333
Murray N. Charlton
NWRI Environment—Canada
Canada Centre/Inland Waters
P.O. Box 5050
Burlington, Ontario, Canada L7R 4A6
Bernard X. Chenette
Dufresne-Henry Inc.
162 BarreSt.
Montpelier, Vt. 05602
Clair Chesley
Bureau of Air Quality Control
State House, Station 17
Augusta, Maine 04333
Jerry Choate
Environment—New Brunswick
P.O. Box 6000
Fredericton, New Brunswick, Canada E38 5H1
P.M. Chutter
Council for Sci. & Ind. Res., Pretoria
c/o S. African Embassy, Suite 300
2555 M St. N.W.
Washington, D.C. 20037
Robert H. Ciullo
Nasson College
Water Quality Laboratory
Springvale, Maine 04083
Gordon Clark
Springfield College
Box 184
Springfield, Maine 01109
Ann N. Clarke
AWARE, Inc.
P.O. Box 40284
Nashville, Tenn. 37204
Nicholas L. Clesceri
R.P.I. Rensselaer Polytechnic
Ricketts Building, Room 102
Troy, N.Y. 12181
John F. Cobianchi
J.P. Stevens & Co., Inc.
1185 Ave. of the Americas
New York, N.Y. 10036
Jacqueline Cohen
Greater Portland Council of Govts.
316 Ocean Ave.
Portland, Maine 04103
William E. Colby
Northeast Laboratory Serv.
Box 788
China Road
Waterville, Maine 04901
Carol R. Collier
Betz Converse Murdoch, Inc.
1 Plymouth Meeting Hall
Plymouth Meeting, Pa. 19462
Arthur J. Conden
Edward C. Jordan Co.
P. O. Box 7050
Downtown Station
Portland, Maine 04112
Jody Connor
N.H. Water Supply and Pollution Control Comm.
Concord, N.H. 03301
Michael F. Conway
University of Connecticut
Dept. of Civil Engineering
Box U-37
Storrs, Conn. 06268
Pat Cooper
Los Alamos Scientific Lab.
MS 603, P.O. Box 1663
Los Alamos, N. M. 87545
Roger Copp
Huron River Watershed Council
415 W.Washington
Ann Arbor, Mich. 48103
James J. Corbalis, Jr.
Fairfax County Water Auth.
P. 0. Box 1500
Merrifield, Va. 22116
D. M. Cote
E. C. Jordan Co.
562 Congress St.
Portland, Maine 04110
-------
530
RESTORATION OF LAKES AND INLAND WATERS
David Courtemauch
Maine Dept. of Environ. Prot.
State House
Augusta, Maine 40333
Charles M. Courtney
Applied Environmental Svc.
990 N. Barfield Dr.
Marco Island, Fla. 33937
Bruce C. Cowell
Department of Biology
University of S. Florida
Tampa, Fla. 33620
David Cowgill
U.S. Army Corps of Engineers
536 S. Clark St.
Chicago, III. 60605
Richard Cox
Cullinan Eng. Co., Inc.
200 Auburn St.
Auburn, Mass. 01501
Richard Crabtree
Lee Pare and Associates
105 Whipple Street
Providence, R.I. 02908
John R. Craig
Michigan State University
Limnological Research Lab.
Dept. of Fisheries & Wildlife
East Lansing, Mich. 48824
Steve Cringan
Kansas Dept. of Health & Environ.
Building 740, Forbes Field
Topeka, Kans. 66620
Richard L. Crocker, Jr.
Portland Water Dist.
225 Douglass St.
P.O. Box 3553
Portland, Maine 04104
Richard J. Croft
USDA-Soil Conserv. Serv.
1 Burlington Square, Room 205
Burlington, Vt. 05401
David Crouthamel
The Farm
Mildram Road
Wells, Maine 04090
Robert Culver
Camp Dresser & McKee
One Center Plaza
Boston, Mass. 02108
Bob Cummings
Portland Press
390 Congress
Portland, Maine 04204
Michael D. Curtis
State of Connecticut
32 Grandview Terrace
Portland, Conn. 06480
Lyn Dabagian
Sussex Co. Planning Dept.
55-57 High Street
Newton, N.J. 07860
Henry Damegeila
Bechmen Instruments
2500 N. Harbor Blvd.
Fullerton, Calif. 92634
Luisa Damia
Ministry of Environmental
C/114 No. 105-89
Urbanizarion Campo Alegre
Valencia, Carabobo, Venezuela
Cluis Daniel
Universite du Quebec
INRS-Eau CP7500
Ste-Foy
Quebec, Canada G1V 4C7
Robin Davidov
Maryland Dept. of Natural Res.
Tawes Office Building, C-4
Annapolis, Md. 21401
Arlene Davis
Maine Dept. Environ. Prot.
State House, Station 17
Augusta, Maine 04333
Fred Davis
S. Fla. Water Mgmt. Dist.
P.O. Box V
West Palm Beach, Fla. 33462
Howard Davis
U.S. Environ. Prot. Agency-WERL
60 Westview St.
Lexington, Mass. 02173
-------
531
Joanne Davis
Municipality of Metro. Seattle
821 2nd Ave.
Seattle, Wash. 98104
Paul de Graauw
Organisation for Econ. Coop. & Dev.
Environment Directorate
2 Rue Andre-Pascal
Paris, France 75016
Warren Dean
Bechmon Inst.
599 North Ave.
Wakefield, Mass. 01880
Wayne 0. Deason
Office of Water Research and Tech.
18th &C St., N.W.
Washington, D.C. 20240
P. Dehavay
I.H.E.
14 WE J Wytsmanstratt
1050 Brussel, Belgium
Gill Delong
N.B. Dept. Environment
Fredericton Province
P.O. Box 6000
New Brunswick, Canada E3B 5H1
Frank Deluca
Environmental Devices
Tower Building
Marion, Mass. 02738
Francesca C. Demgen
Demgen Aquatic Biology
118 Mississippi St.
Vallejo, Calif. 94590
Douglas Denison
Applied Environ. Research
444 South Main
Ann Arbor, Mich. 48104
Jeffrey Dennis
Maine Dept. Environ. Prot.
Hospital St.
Augusta, Maine 04333
Ronald E. Despres
Town of Wellesley BPW
56 Woodlawn Ave.
Wellesley Hills, Mass. 02181
Dante A. DiDomenico
Dept Environ. Regulation
2600 Blair Stone Road
Tallahassee, Fla. 32301
Shirley Dilg
Boston University
11 Springdale Ave.
Wellesley, Mass. 02181
Norman P. Dion
U.S. Geol. Survey
1201 Pacific Ave., Suite 600
Tacoma, Wash. 98402
Laureen Dolan
Florida Institute of Tech.
1011 Martin Blvd.
Jensen Beach, Fla. 33457
Edward E. Donahue
Associated Engineers
2387 West Monroe
Springfield, III. 62704
Bill Doolittle
University of Tennessee
Zoology Dept.
Knoxville, Tenn. 37916
Bonnie Dovenmuehle
USDA Forest Service
P. O. Box 338, Federal Bldg.
Duluth, Minn. 55803
Bruce R. Dreisinger
Inco Metals Company
General Engineering Bldg.
Copper Cliff, Ontario, Canada POM 1NO
Dianne Dumanoski
The Boston Globe
Boston, Mass. 02107
Edward F. Dunn
Ark. Dept. of Pollution Control & Ecology
8001 National Dr.
Little Rock, Ark. 72209
Arthur W. Dutton
Maine Dept. of Environ. Prot.
State House, Station 17
Augusta, Maine 04333
-------
532
RESTORATION OF LAKES AND INLAND WATERS
Richard Dwinell
Mass. House of Rep.
State House
Boston, Mass.
Craig W. Dye
Fla. Dept. Environ. Rep.
2600 Blair Stone Rd.
Tallahassee, Fla. 32301
Ute J. Dymon
Clark University
232 Bullard St.
Holden, Mass. 01520
Barry R. Edgerton
USDA Forest Service
Ottawa National Forest
East U.S. Route 2
Ironwood, Mich. 49938
Paul D. Eiler
University of Maine
Dept. of Entomology
304 Deering Hall
Orono, Maine 04469
Steven J. Eisenreich
University of Minnesota
Environ. Eng. Prog.
103 Exp. Eng. Bldg.
Minneapolis, Minn. 55455
David Eisentrout
Pennwalt Corp.
2753 Wildwood Dr.
Columbus, Ohio 43229
Nathan Emerson
Weston & Samson Engineers
10 High Street
Boston, Mass. 02110
Katherine Enright
Dept. Water Resources
State Office Bldg.
Montpelier, Vt. 05602
John Erdmann
E.Z. Hickok & Assoc.
545 Indian Mound
Wayzata, Minn. 55391
Gary Erickson
III. Dept. of Conservation
110 James Road
Spring Grove, III. 60081
Robert H. Estabrook
N. H. Water Pollution Comm.
State Lab., Hazen Dr.
Concord, N.H. 03301
Timothy E. Fannin
Wyoming Game & Fish Water Lab.
360 Buena Vista
Lander, Wyo. 82520
Everett Fee
Freshwater Institute
501 University Crescent
Winnipeg, Manitoba, Canada R3T 2N6
L. Ronny Ferm
Statens Naturvardsverk
Solna, Sweden S-17125
Herbert Ferran
Nasson College
Dept. of Chemistry
Springvale, Maine
Randy Ferrin
USDA Forest Service
P.O. Box 638
Laconia, N.H. 03246
Monty Fischer
New England River Basin Comm.
177 Battery St.
Burlington, Vt. 05401
John Fitch
Mass. Audabon Society
Lincoln, Mass. 01773
Jo Fitzpatrick
Bureau of National Affairs
Washington, D.C.
Richard A. Flanders
N.H. Water Supply & Pollution Control
Hazen Drive P.O. Box 95
Concord, N.H. 0330I
Myra Flowe
Duke Power Co.
Route 4, Box 531
Huntersville, N.C. 28078
Louis Fontaine
Maine Dept. of Environ. Prot.
State House
Augusta, Maine 04333
-------
533
Nancy E. Forrester
E.G. Jordan Co.
562 Congress St.
Portland, Maine 04110
Mark Fouhy
Cape Cod Ping. Commission
1st District Courthouse
Barnstable, Mass. 02630
Richard A. Fralick
Plymouth State College
Plymouth, N.H. 03264
Charles G. Fredette
Conn. Dept. Environ. Prot.
165 Capitol Ave.
Hartford, Conn. 06115
Gunther Friedrich
Federal Inst/Water & Waste
AM Walowinkel 70
D4150 Krefeld-Hulserberg
West Germany
Terry L. Fung
Water Resources, Vermont
1 Corrine St.
Winooski, Vt. 05404
Wally Fusilier
University of Michigan
9200 Dexter Chelsea Rd.
Dexter, Mich. 48130
Thomas J. Gallagher
P.O.Box 1207
Watertown, S.D. 57201
Richard A. Gallo
USDA Soil Conserv. Serv.
1 Burlington S., Suite 205
Burlington, Vt. 05401
Eugene P. Galvagni, Jr.
County of Berkshire
Court House
Engineering Dept.
Pittsfield, Mass. 01201
Jacques Garancher
Ministere de I'Environnement
30 Allee de la Pepiniere
Suzesnes, France 92150
Paul Garrett
Deer Lodge & Granite
8753 N. Montana Ave.
Helena, Mont. 59601
Paul J. Garrison
Wis. DNR
3911 Fish Hatchery Rd.
Madison, Wis. 53711
Virginia Garrison
Dept. Water Resources
State Office Bldg.
Montpelier, Vt. 05602
David Gates
E.C. Jordan Co.
P.O. Box 7050
Portland, Maine 04112
Stephen E. Gatewood
Kissimmee Coordinating Council
2600 Blair Stone Rd., Room 538
Tallahassee, Fla. 32308
Richard Gelpke
Boston State College
625 Huntington Ave.
Boston, Mass. 02115
Richard S. Geney
Atlas Copco
70 Demarest Dr.
Wayne, N.J. 07470
Harry L. Gibbons, Jr.
Washington State University
C & Environ. Eng.
Sloan Hall 141
Pullman, Wash. 99164
K. E. Gibbs
Dept. of Entomology
University of Maine at Orono
Orono, Maine 04469
Paul J. Godfrey
Water Resources Research Center
Graduate Research Center, Room A-211
Amherst, Mass. 01003
Alan L. Goldstein
South Florida Water Management
111-113 S. West Park St.
Okeechobee, Fla. 33472
-------
534
RESTORATION OF LAKES AND INLAND WATERS
Gareth A. Goodchild
Ontario Ministry of Nat. Res./Fisheries
Whitney Block Queen's Pk.
Toronto, Ontario, Canada M7A 1W3
Cassie Ann Gosselin
University of Massachusetts
Gray Road
Templeton, Mass. 01468
Dennis J. Gregor
Environment Canada
P.O. Box 5050
Burlington, Ontario, Canada L7R 4A6
Jean W. Gregory
Va. St. Water Control Board
P.O. Box 11143
Richmond, Va. 23230
Richard Griffin
E.G. Jordan Co.
562 Congress St.
Portland, Maine 04110
Thomas Griffin
Dept. of Civil Engineering
Princeton University
Princeton, N.J. 08544
Donald Groff
Merwin Brook Road
Brookfield Center, Conn. 06805
Galen Groff
Dept. Environ. Prot.
State House
Augusta, Maine 04333
Peter Groth
Harzwasserwerke Granetalsperre
3394 Langelsheim
West Germany
Stephen Groves
Dept. Environ. Prot.
State House, Station 17
Augusta, Maine 04333
David A. Gruber
Mil. Metro Sewerage Dist.
735 North Water St.
Milwaukee, Wis. 53202
H. Christopher Grundler
University of Michigan
Engineering Bldg. 1-A, Room 117
Ann Arbor, Mich. 48109
Karla I. Gustafson-Marjanen
Dept. of Zoology
University of Maine
Orono, Maine 04469
J. W. Habraken
City of Akron
1570 Ravenna Road
Kent, Ohio 44240
Bonny L. Hadiaris
Maine Dept. of Environ. Prot.
Assistant Engineer
State House, Station 17
Augusta, Maine 04333
Bart Hague
U.S. Environ. Prot. Agency—Reg. 1
JFK Federal Building
Boston, Mass. 02203
James Hall
Water Pollution Comm.
Hazen Dr.
Concord, N.H. 03301
Ethel Hammer
Hillsborough County Plann. Comm.
Suite 288, Courthouse
Tampa, Fla. 33601
Kenneth Spencer Hanks
Arizona Game & Fish Dept.
2222 W. Greenway Rd.
Phoenix, Ariz. 85023
John W. Hannah
Brevard County Fla.
2575 N. Courtenay Parkway
Merritt Island, Fla. 32952
H. H. Hannan
Southwest Texas State Univ.
Biology Dept.
San Marcos, Texas 78666
T. A. Hannula
University of Maine at Orono
438 English/Math
Orono, Maine 04469
Mark Hanson
City of Fairmont
114 E. First
Fairmont, Minn. 56031
-------
535
Fred W. Hardt
Enviromed Associates
133 Saratoga Road
Scotia, N. Y. 12302
Jack R. Hargis
University of Minnesota-Duluth
Department of Biology
Duluth, Minn. 55812
Elaine M. Hartman
University of Mass.
Dept. Environ. Science
Marshall Hall
Amherst, Mass. 01003
Richard T. Hartman
University of Pittsburgh
Dept. of Biological Sciences
Pittsburgh, Pa. 15260
David L. Haselow
Ecol Sciences, inc.
735 N. Water St., Suite 715
Milwaukee, Wis. 53202
W. Hattingh
Dept. of Water Affairs, Forestry & Environ.
c/o S. African Embassy, Suite 300
2555 M St., N.W.
Washington, D.C. 20037
H. Heida
Environ. Health Lab.
Amstelveenseweg 88
Amsterdam,
The Netherlands
Carson O. Helfrich
Lake Wallenpaupock Watershed Mgt. Dist.
511 Broad St.
Milford, Pa. 18337
D. Dickinson Henry, Jr.
Northwestern Ct. Regional Planning Agency
Lake Waramaug Task Force
P.O. Box 30
Warren, Conn. 06777
Charles E. Herdendorf
Center for Lake Erie Res.
Ohio State University
484 W. 12th Ave.
Columbus, Ohio 43210
Maj. John E. Hesson
Science Research Lab.
U.S. Military Academy
West Point, N.Y. 10996
Patrick Hickey
Montachusett Regional Planning Comm.
150 Main Street
Fitchburg, Mass. 01420
I. Sam Higuchi, Jr.
Minnesota Pollution Control Agency
13520 Excelsior Blvd.
Minnetonka, Minn. 55343
Barbara Hoag
City of East Grand Rapids
1059 Eastwood, S.E.
Grand Rapids, Mich. 49506
George E. Hoag
University of Connecticut
Box U-37
Storrs, Conn. 06278
R.A. Hoare
Ministry of Works & Dev.
MWD Private Bag
Hamilton, New Zealand
Lynn M. Hodgson
University of Florida
c/o Dept. of Botany
Gainesville, Fla. 32611
Diane Hoffman
Executive Office of Environ. Affairs
100 Cambridge St.
Boston, Mass. 02202
G.C. Holdren, Jr.
University of Louisville
Water Resources Lab.
Louisville, Ky. 40292
Robert Holman
State of North Carolina
Route 2, Box 369A
Edenton, N.C. 27932
Hans Holtan
Norwegian Institute for Water Research
Postbox 333
Oslo, Norway
-------
536
RESTORATION OF LAKES AND INLAND WATERS
Steve Holtman
Louisiana State University
Baton Rouge, La. 70803
Abe Horpestad
WQB-State of Montana
WQB-DHES Rm A206
Cogswell Building
Helena, Mont. 59601
Barbara Hosper-Doop
S. Harry Hosper
Rijkswaterstaat
Jol 12-17
Lelystad, Netherlands 8243EA
John Houlihan
University of Southern Maine
Route 2
Gorham, Maine 04038
Jeff Hovis
NCASI, Tufts University
Anderson Hall
Medford, Mass. 02155
David W. Howard
Kleinschmidt-Duttig 22
73 Main Street
Pittsfield, Maine 04967
Warren Howard
U.S. Environ. Prot. Agency, Reg. 1
JFK Federal Bldg.
Boston, Mass. 02203
William N. Howard
Dept. of Natural Resources
Province of Manitoba
1129 Queens Ave.
Brandon, Canada R7A 1L9
Janet Hren
U.S. Geological Survey
975 W. Third Ave.
Columbus, Ohio 43212
Geoffrey Hughes
City of East Grand Rapids
260Hodenpyl Dr. S.E.
Grand Rapids, Mich. 49506
Craig Hull
City of Pine Bluff
200 E. Eighth Ave.
Pine Bluff, Ark. 71601
Robert Humphrey
MUD Cat. Div.
P.O. Box 451
East Longmeadow, Mass. 01028
Malcolm Hunter
School of Forestry
University of Maine
Orono, Maine 04469
James Huson
Center for Natural Areas
Box 98
South Gardiner, Maine 04359
Anders Hustredt
Williams & Work
4327 Northgate N.E.
Grand Rapids, Mich. 49505
Byron J. Israelson
The New York Times
New York, N.Y
Russell I. James
Ecoscience, Inc.
517 S. Main St.
Old Forge, Pa. 18518
Lorraine Janus
Phycologist
518 Indian Rd., Apt. 4
Burlington, Ontario, Canada L7R 3T3
Karen Jeffrey
Journal Tribune
201 A Main St.
Sanford, Maine 04073
Mark Johnson
Iowa Dept. Environ. Quality
Henry Wallace Building
Des Moines, Iowa 50319
Robert Johnson
U.S. Environ. Prot. Agency
401 M St., S.W.
Washington, D.C. 20460
Eric Johnston
Albright & Wilson, Ltd.
Marchon Works
Whitehaven Cumbria, England CA289QQ
Bill Jones
400 E. 7th St.
Bloomington, Ind. 47405
-------
537
Brad Jones
Iowa State University
Dept. of Animal Ecology
Ames, Iowa 50011
Chris Jones
USDA Soil Conserv. Serv.
7 High St.
Skowhegan, Maine 04976
John R. Jones
University of Missouri
112 Stephens Hal I
Columbia, Missouri 65211
Kenneth Jones
Maine Dept. of Environ. Prot.
Div. of Operation & Maint.
State House, Station 17
Augusta, Maine 04333
James Jowett
U.S. Environ. Prot. Agency
401 M St. S.W.
Washington, D.C. 20460
Michael Kachur
Acad. Natl. Sciences
19th & Park way
Philadelphia, Pa. 19103
Michael Karolle
City of East Grand Rapids
2660 Albert Dr., S.E.
Grand Rapids, Mich. 49506
Susan Kaufman
Lake Waramaug TF
P.O. Box 30
Warren, Conn. 06754
Brian Kelso
Environ. Prot. Service
Kapiland 100 Park Royal
West Vancouver, B.C., Canada V7T1A2
Dennis L. Keschl, Ess. II
Maine Dept. of Environ. Prot.
State House, Station 17
Augusta, Maine 04917
Kenneth D. Kimball
University of New Hampshire
Botany & Plant Pathology
Durham, N.H. 03824
Robert Kimball
Camp Dresser & McKee
One Center Plaza
Boston, Mass. 02108
Sarah F. Kimball
Brewster Academy Science Dept.
Wolfboro, N.H. 03894
Terry Kimball
Maryland Dept. of Natural Resources
Tawes Office Bldg., C-4
Annapolis, Md. 21401
C. Dexter Kimsey, Jr.
S.C. Dept. of Health & Environ. Con.
2600 Bull St.
Columbia, S.C. 29201
Kathie Kimsey
S.C. Dept. of Health & Environ. Con.
2600 Bull St.
Columbia, S.C. 29201
Fredric W. King
Billerica Conservation Comm.
Town Hall, Concord Rd.
Billerica, Mass. 01821
Wendy L King
Cobbossee Watershed Dist.
15 High St.
Winthrop, Maine 04364
Viggo Kismul
State Pollution Control
P.O. Box 8100 Dep.
Oslo, Norway
Tom G. Kizis
Planning & Community Dev.
531 Main St., Rm 203
Worcester, Mass. 01608
Andrew Klemer
Suny-Purchase
Div. of Natural Sciences
Purchase, N.Y. 10577
Ron Knaus
Louisiana State University
Dept. of Nuclear Science
Baton Rouge, La. 70803
John C. Knight
Duke Power Co.
Route 4, Box 531
Huntersville, N.C. 28078
-------
538
RESTORATION OF LAKES AND INLAND WATERS
Nicholas P Kobriger
Rexnord Environ. Research Center
5103W. Beloit Rd.
Milwaukee, Wis. 53214
Robert Koch
N/E Mich. Council of Govt.
P. 0. Box 457
408 W. Main St.
Gaylord, Mich. 49735
Brian D. Kooiker
Vermont Dept. of Water Res.
Route 1, Box 6A
East Calais, Vt. 05650
Carol Koscik
4703 Winnequah Rd.
Monona, Wis. 53716
Kenneth J. Koscik
Dane County
210 Monona Ave.
Madison, Wis. 53709
Robert J. Kotch
New Jersey Dept. Div. Water Res.
P. 0. Box CN-029
Trenton, N.J. 08625
J.N. Krider
USDA Soil Conserv. Serv.
1974Sproul Rd.
Broomall, Pa. 19008
Joseph A. Krivak
U.S. Environ. Prot. Agency
401 M St. S.W.
Washington, D.C. 20460
Howard F Krosch
Minn. Dept. Natural Resources
Box 25, Centennial Bldg.
St. Paul, Minn. 55155
Andy Kuether
Lake Wanamaug
P.O. Box 30
Warren, Conn. 06754
Jochen Kuhner
Meta Systems
14 Hillside Rd.
Newton, Mass. 02161
Jan-Tai Kuo
AWARE, Inc.
P.O. Box 40284
Nashville, Tenn. 37204
Venkatakasi Kurmala
RPI3
Ricketts Bldg., Room 102
Troy, N. Y. 12181
James W. LaBaugh
U.S. Geological Survey
Water Resources Div.
Mailstop413DFC
Denver, Colo. 80225
Eleanor Lacombe
Research Dept.
Maine Medical Center
Portland, Maine 04101
Aimlee D. Laderman
Marine Biology Lab
Woods Hole, Mass.
P. Lambert
UCB S.A.
4 Chaussee de Charberci
B-1060 Brussels, Belguim
D. J. Lane
Engineering & Water Supp. Dept.
Private Mail Bag
Salisbury, S. Australia 5108
Richard Langdon
Vermont Dept. of Water Res.
State Office Bldg.
Montpelier, Vt. 05602
David P Larsen
U.S. Environ. Prot. Agency
200 S.W. 35th St.
Corvallis, Ore. 97330
GiHes LaRoche
Marine Sciences Centre
McGill University
3620 University St.
Montreal, Canada H3A 282
Fred Lavallee
Maine Dept. of Environ. Prot.
State House, Station 17
Augusta, Maine 04333
-------
539
David Leake
Maine Dept. of Environ. Prot.
State House, Station 17
Augusta, Maine 04333
Lillian C. Lee
Lee Quality
6 Kirkwood Cir.
Scarboro, Maine 04074
Mary Leslie
Jones, Edmund & Assoc.
730 N.Waldo Rd.
Gainesville, Fla. 32601
C. Kwei Lin
Great Lakes Research Div.
University of Michigan
Ann Arbor, Mich. 48109
Owen T. Lind
Biology & Environ. Studies
Baylor University
Waco, Texas 76703
Robert A. Lohnes
Iowa State University
Civil Engineering Dept.
Ames, Iowa 50011
Lauren H. Long
USDA Soil Conserv. Serv.
University of Maine at Orono
Orono, Maine 04473
Raymond Loveless
Mountainland Assoc. of Govts.
160 East Center St.
Provo, Utah 84601
Stuart D. Ludlam
University of Massachusetts
Box 48
Whatley, Mass. 01093
John Lutz
Metro Planning Comm.
City/County Bldg., Suite 403
Knoxville, Tenn. 37902
J. Gualberto Limcn Macias
S.A.R.H. Mexico
Calle Dia 2528
Jard del Bosque
Guadalajara Jalisco, Mexico
Larry MacMillan
U.S. Env. Prot. Agency, Reg. 1
JFK Federal Bldg.
Boston, Mass 02203
Ben L. Magee
Tenn. Div. of Water Quality Control
Cordell Hull Bldg., Room 630
Nashville, Tenn. 37219
Robert Magnien
Dartmouth College
Dept. of Biological Sciences
Hanover, N.H. 03755
Tony Mais
Intl. Atlantic Salmon Foundation
P. 0. Box 429
St. Andrews, New Brunswick, Canada EOG 2X0
John C. Malley
USDA Soil Conserv. Serv.
587 Sperry St.
Westbrook, Maine 04092
Ron Malone
Louisiana State University
Civil Eng. Dept.
Baton Rouge, La. 70803
Enrique F. Mandelli
United Nations-UNESCO
Av. Revolucion 1909-4C
Distrito Federal Piso
Mexico City 20
Ronald G. Manfredonia
U.S. Environ. Prot. Agency
JFK Federal Bldg.
Boston, Mass. 02203
Audrey E. Manzer
Dartmouth Lakes
Advisory Board
35 Clearview Crescent
Dartmouth, Nova Scotia, Canada B3A 2M9
Dan Martin
U.S. Fish & Wildlife Service
P.O. Box 139
Yankton, S. Dakota 57078
Samuel R. Martin
Regional Planning Council
2225 N. Charles St.
Baltimore, Md. 21218
-------
540
RESTORATION OF LAKES AND INLAND WATERS
Luis Martinez-Avalos
Embassy of Guatemala
2220 R St. N.W.
Washington, D.C. 20008
Peter Mason
Grand River Conserv. Auth.
400 Clyde Rd.
Cambridge, Ontario, Canada
Robert J. Massarelli
Brevard Co. Water Res. Dept.
2575 N. Courtney Parkway
Merritt Island, Fla. 32952
Paul Mathieu
Mudcat Division
P.O.Box 16247
St. Louis Park, Minn. 55416
Edmond A. Mayhew
Gainesville Junior College
Gainesville, Ga. 30501
Barbara Huss Mazur
Missouri Dept. of Natural Res.
P.O. Box 1368
Jefferson City, Mo. 65102
Daniel J. Mazur
Missouri Dept. of Natural Res.
2010 Missouri Blvd.
P.O.Box 1368
Jefferson City, Mo. 65102
Patrick M. McCaffrey
Kissimee River Coor. Council
2600 Blair Stone Rd.
Tallahassee, Fla. 32301
Kenneth M. McCoig
Orange City Sewer Water
2450 W. 33rd St.
Orlando, Fla. 32806
Steve McCullers
Mayes, Sudderth & Etheridge
1775 The Exchange
Atlanta, Ga. 30339
Pat McCullough
Entrance Engineers
100- 116th Ave.S.E.
Bellevue, Wash. 98004
HankMcKellar
Dept. Environ. Health Sciences
University of South Carolina
Columbia, S.C. 29208
Glenn McKenna
Louisiana State University
Dept. of Civil Engineering
Baton Rouge, La. 70803
Paula McKenzie
Mayes, Sudderth & Etheridge
1775 The Exchange
Atlanta, Ga. 30339
Wallace M. McLean
State of Vermont
Dept. of Water Resources
State Office Building
Montpelier, Vt. 05602
Jim McMahon
Federal Building
151 Forrest Ave.
Portland, Maine 04101
Jim McMillan
University of Maine
Dept. of Environ. Engineering
Aubert Hall, Room 451
Orono, Maine 04473
Jeffrey L. McNelly
Camdent Rockland Water Co.
P.O. Box 689
Rockland, Maine 04841
Edward K. McSweeney
U.S. Environ. Prot. Agency
JFK Federal Building
Boston, Mass. 02203
Richard S. McVoy
Mass. Div. Water Poll. Control
Box 545
Westborough, Mass. 01581
W. Ross McWilliams
Dept. Environ. Reg.
2600 Blair Stone Rd.
Tallahassee, Fla. 32301
Don Meals
Vt. Water Resources Research Ctr.
University of Vermont
601 Main St.
Burlington, Vt. 05405
-------
541
Steven Medlar
E.G. Jordan Co.
P.O. Box 7050
Downtown Station
Portland, Maine, 04112
Dennis Merrill
Maine Dept. of Environ. Prot.
State House
Augusta, Maine 04333
LaVere B. Merritt
Brigham Young University
368-R CB BYU
Provo, Utah 84602
Michael Michalski
Hough Stansbury Michalski, Ltd.
63 Galaxy Blvd., Unit 1
Rexdale, Ontario, Canada M9W 5R7
Richard Micheal
Dufresne-Henry, Inc.
1321 Washington Ave.
Portland, Maine 04103
Gerald Mikol
NYS Dept. Environ. Conserv.
50 Wolf Rd., Room 519
Albany, N.Y. 12065
Richard Mikula
Michigan Dept. of Nat. Res.
P.O. Box 30028
Lansing, Mich. 48909
Richard Milbrodt
City of South Lake Tahoe
P.O. Box 1210
South Lake Tahoe, Calif. 95731
Don Miller
University of New Hampshire
INER, Rm. 8, Pettee Hall
Durham, N.H. 03824
Donald Miller
Town of Morris
E. Shore Rd.
Lakeside, Conn. 06758
Richard H. Millest
Canadian Section Intl. Joint Comm.
100 Metcalfe St. 18th Floor
Ottawa, Ontario, Canada K1P 5M1
Patricia A. Mitchell
Alberta Environ. Water Quality
9820- 106 St.
Edmonton, Alberta, Canada T5K 2J6
Paul Mitnik
Maine Dept. of Environ. Prot.
State House, Station 17
Augusta, Maine 04333
Barbara Montague-Wilkie
Plymouth State College
Plymouth, N.H. 03264
Karen E. Moore
Springfield College
Rt. 1, Mountain Rd.
Stowe, Vt. 05672
Sharon Moore
U.S. Environ. Prot. Agency, Reg. 1
JFK Federal Bldg.
Boston, Mass. 02203
Elizabeth Moran
Cornell University
Dept. of Agronomy
Bradfield Hall
Ithaca, N.Y. 14853
Judith A. Morrison
Westfield State College
257 Cordaville Rd.
P.O. Box 160
Southboro, Mass. 01772
Denis Morrissette
New Brunswick Govt. Dept. of Env.
P.O. Box 6000
Fredericton, N.B., Canada E3B 5H1
James W. Morse 11
Dept. Water Resources
State Office Bldg.
Montpelier, Vt. 05602
William B. Morton
NYS Dept. Environ. Conserv.
50 Wolf Rd.
Albany, N.Y. 12233
G. Mourkides
Aristotelian University
Thessaloniki, Greece
-------
542
RESTORATION OF LAKES AND INLAND WATERS
Barry Mower
Maine Dept. of Environ. Prot.
RFD5
Augusta, Maine 04330
Anne V. Mullen
Dept. Water Resources
State Off ice Bldg.
Montpelier, Vt. 05602
Michael W. Mullen
Engineering Analysis, Inc.
2109 Clinton Ave., W, Suite 432
Huntsville, Ala. 35805
Jim Murphy
Springfield College
Box 1125
Springfield, Mass. 01109
Michael Murphy
E.G. Jordan Co.
25 Birch Rd.
South Portland, Maine 04106
Declan A. Murray
University-College of Dublin University
Belfield, Dublin 4, Ireland
Robert Muylle
c/o Societe de Traction et O'Electricite
31 Rue de la Science
1040 Bruxelles, Belgium
Vernon Myers
U.S. Environ. Prot. Agency
401 M St. S.W.
Washington, D.C. 20460
P. A. Neame
Montreal Engineering Co.
1259th Ave., S.E.
Calgary, Alberta, Canada T2G OP6
John Olaf Nelson
N. Marin Co. Water Dist.
P.O.Box 146
Novato, Calif 94947
Richard D. Newman
Proctor & Gamble Co.
Ivorydale Technical Center
Cincinnati, Ohio 45217
Stanley A. Nichols
University of Wis.—Ext.
1815 University Ave.
Madison, Wis. 53706
William J.Nichols, Jr.
U.S. Geological Survey
26 Ganneston Dr.
Augusta, Maine 04330
Peter M. Nolan
U.S. Environ. Prot. Agency, Reg. 1
60 Westview St.
Lexington, Mass. 02173
Terry Noonan
Ramsey County, Minnesota
3377 North Rice St.
St. Paul, Minn. 55112
W. A. Norvell
Conn. Agric. Exp. Sta.
P.O. Box 1106
New Haven, Conn. 06504
Barbara R. Notini
Mass. Water Poll. Control
P.O. Box 545
Westboro, Mass. 01581
Richard P. Novitzki
U.S. Geological Survey
521 West Seneca St.
Ithaca, N.Y. 14850
Charles W. Noxon
Menardi Southern Corp.
3908 Colgate St.
Houston, Tex. 77087
Carlton L. Noyes
Jason M. Cortell & Assoc.
244 Second Ave.
Waltham, Mass. 02154
Robert Nuzzo
22 Forest Street
Cambridge, Mass. 02140
William Nuzzo
U.S. Environ. Prot. Agency, Reg. 1
JFK Building
Boston, Mass. 02203
Paul H. Oakland
State of New Hampshire
Water Supply & Pollution Control
Health & Welfare Bldg.
P.O. Box 95
Concord, N.H. 03301
-------
54.3
Barbara A. Obeda
E.I.S., Inc.
Environ. Impact Service
35 Sunset Hill Rd.
Brookfield Center, Conn. 06805
George O'Carrol I
Middlesex County Mosquito Comm.
200 Parsonage Rd.
Edison, N.Y. 08817
Joe O'Connor
U.S. Environ. Prot. Agency, Reg. IV
345 Courtland St., N.E.
Atlanta, Ga. 30365
James C. O'Shaughnessy
Dept. Civil Engineering
Northeastern University
Boston, Mass. 02115
Adam W. Olivieri
Calif. Regional Water Qual. Control Bd.
1111 Jackson St.
Oakland, Calif. 94607
Richard Osgood
Metropolitan Council
300 Metro Square Bldg.
St. Paul, Minn. 55101
David Osmond
Gartner Lee Assoc., Ltd.
Toronto-Buttonville Airport
Markham, Ontario, Canada L3P 3J9
Donna Palmer
North Carolina Div. of Env. Mgmt.
P.O. Box 27687
Raleigh, N.C. 27611
Cindy Parks
Vt. Dept. of Water Res.
River St.
Montpelier, Vt. 05602
Scott Parrish
Poplars Res. Center
400 E. 7th St., Room 426
Bloomington, Ind. 47405
Harry Parrott
USDA-Forest Service
633 W. Wisconsin Ave.
Milwaukee, Wis. 53206
Susan T. Paschall
Springfield College
272 Middlesex St.
Springfield, Mass. 01109
Dave Paschke
Applied Biochemists
5300 W. County Line Rd.
Mequon, Wis. 53092
Clay Patmont
Harper-Owes
301 Commuter Bldg.
65 Marion St.
Seattle, Wash. 98104
A.G. Payne
Proctor & Gamble Co.
I.T.C.
Cincinnati, Ohio 45217
Bruce S. Peachey
R.P.I., Dept. of Chem. and Environ. Eng.
Troy, N.Y. 12181
Frank E. Perkins, Sr.
Maine Dept. Environ. Prot.
Bureau/Water Quality Control
State House, Station 17
Augusta, Maine 04333
Mike Peters
Guadalupe-Blanco Auth.
P.O. Box 271
Seguin, Tex. 78155
Jim Peterson
University of Wisconsin
Environ. Resources Unit
1815 University Ave.
Madison, Wis. 53706
Francis J. Philbert
Environment Canada
867 Lakeshore Rd.
Burlington, Ontario, Canada L7R 4A6
Jon Phillippe
GKY & Associates, Inc.
5411-EBacklickRd.
Springfield, Va. 22151
Bruce Phillips
Box 31
New Gloucester, Maine 04260
-------
544
RESTORATION OF LAKES AND INLAND WATERS
Sonny Pierce
Maine Inland Fish & Wildlife
Route 1, Box 570
Scarboro, Maine 04074
Don Pievson
Trent University
Dept. Biology-Geography
Peterborough, Ontario, Canada K9J 7B8
Frank Piveronas
E.G. Jordan Co.
562 Congress St., Box 7050
Portland, Maine 04112
David Platt
Bangor Daily News
Bangor, Maine 04401
Paul E. Plekavich
8 Menotomy Rd., No. 10
Arlington, Mass. 02174
Gerald M. Pollis
U.S. Environ. Prot. Agency, Reg. Ill
6th & Walnut Sts.
Philadelphia, Pa. 19106
Don Porcella
Tetra Tech
3746 Mount Diablo Blvd., Suite 300
Lafayette, Calif. 94549
Thomas Porucznik
U.S. Environ. Prot. Agency, Reg. II
26 Federal Plaza
New York, N.Y. 10278
Milton Potash
University of Vermont
Dept. of Zoology, UVM
Burlington, Vt. 05405
Chris P. Potos
U.S. Environ. Prot. Agency/COE
536 S. Clark St.
Chicago, III. 60605
Thomas Potter
Maine Dept. of Environ. Prot.
State House, Station 17
Augusta, Maine 04333
Judy Potvin
Maine Dept. of Environ. Prot.
Pleasant Hill Rd.
Augusta, Maine 04330
Waldo E. Pray
Maine Dept. of Environ. Prot.
State House, Station 17
Augusta, Maine 04333
Mike Pruitt
439 Congress
Portland, Maine 04104
Al Prysunka
Maine Dept. of Environ. Prot.
State House, Station 17
Augusta, Maine 04333
Joseph J. Przywara
Ocean County Health Dept.
C.N. 2191 Sunset Ave.
Toms River, N.J. 08753
Gordon R. Pyper
Dufresne-Henry, Inc.
Precision Park
N. Springfield, Vt. 05150
Todd A. Rathkamp
Aquamarine-Corp.
P.O. Box 616
Waukesha, Wis. 53186
Jeff Raymond
Applied Biochemists
5300 W. County Line Rd.
Mequon, Wis. 63092
Garth W. Redfield
NUSAC, Inc.
7926 Jones Branch Dr.
McLean, Va. 22102
Susan Redfield
NUSAC, Inc.
2405 Earlsgate Ct.
Reston, Va. 22901
John Reed
Malcolm-Pirnie, Inc.
2 Corporate Park Dr.
White Plains, N.Y. 10602
Joel Rekas
Antioch/New England Grad. School
Environ. Studies Dept.
Keene, N.H.
Charles D. Rhodehamel
Columbia Association
5829 Banneker Rd.
Columbia, Md. 21044
-------
545
F. Brandt Richardson
Minn. Water Planning Bd.
American Ctr. Bldg., Room 600
150E. Kellogg Blvd.
St. Paul, Minn. 55101
Paul D. Ring
New Harbor Water Co.
P.O.Box 11
New Harbor, Maine 04554
Dorothy Risen
AWARE, Inc.
P.O. Box 40284
Nashville, Tenn. 37204
John R. Ritter
U.S. Geological Survey
P.O.Box 1107
Harrisburg, Pa. 17108
William F. Ritter
University of Delaware
Agricultural Eng. Dept.
Newark, Del. 19711
Tomas Rivera
Environ. Quality Bd.
P.O.Box 11488
Santurce, Puerto Rico 00910
Debbie Roberts
RPI Freshwater Institute
51 Gull Bay Rd.
Putnam Station, N.Y. 12861
Thomas E. Robertson
University of Maine
Dept. of Botany
Orono, Maine 04469
Glenn Robinson
Ministry of the Environment
Water Resources Branch
Box 213
Rexdale, Ontario, Canada M9W 5L1
Chet A. Rock
University of Maine
457 Aubert Hall
Dept. of Civil Eng.
Orono, Maine 04469
Fanny Rodriguez
Ministerio del Ambiente
Calle Momrrique No. 106-22
Valencia, Carabobo, Venezuela
Larry A. Roesner
Water Resources Engineers
8001 Forbes Place, Suite 312
Springfield, Va. 22151
Alice M. Rojko
Div. of Water Pollution Control
Lyman School
Westboro, Mass. 01581
Paul Roland
LIFE-New York Limnology Information
Freshwater Ecology, Inc.
Ponderosa Rd.
Carmel,N.Y. 10512
Eric Root
CPCOG
331 Veranda St.
Portland, Maine 04103
Kenneth Rose
UNI NewHamp.
200 S. Main St.
New Market, N.H. 03857
Richard J. Ross
Nekoosa Papers, Inc.
100 Wisconsin River Dr.
Port Edwards, Wis. 54469
K.R. Gina Rothe
California State University
Chico, Calif. 95929
Douglass Rothermel
Ocean County Health Dept.
CN 2191 Sunset Ave.
Toms River, N.J. 08753
Michael W. Roughton
24C University Pk.
Orono, Maine
Bruce W. Rummel
Great Water Assoc.
5340 E 26 Ave., Suite 56
Anchorage, Alaska 99504
Hal Runke
Environ. Research Group
4663 Chats worth St.
St. Paul, Minn. 55112
Katherine J. Sage
Cobbossee Watershed Dist.
15 High St.
Winthrop, Maine 04364
-------
546
RESTORATION OF LAKES AND INLAND WATERS
Mitsuru Sakamoto
Water Research Institute
Nagoya University
Chikusa-Ku, Nagoya, Japan 464
James D. Scerra
U.S. Geological Survey
26 Gannestone Dr.
Augusta, Maine 04330
Daniel Schacht
Ramsey County
3377 North Rice St.
St. Paul, Minn. 55112
Eliza Schacht
306 Estabrooke Hall
University of Maine
Orono, Maine
Joan Schieber, Repr.
N.H. Gen. Court
Concord, N.H.
Richard Schiller
Center for the Environ.
275 Windsor St.
Hartford, Conn. 06120
Joel G. Schilling
Minn. Pollution Control Ag.
1935W. Co. Rd. B-2
Roseville, Minn. 55113
Leah Ann Schirle
University of Michigan
852 Canterbury Crescent
Bloomfield Hills, Mich. 48013
Marcel Schmid
Baudepartement des Kantons
Aargau Abt Gewasserschutz Obere
Vorstadt 40 CH 5001 Aarau, Switzerland
Steve Schreiner
Clemson University
Zoology Dept.
Clemson, S.C. 29631
Donna F. Sefton
Illinois Environ. Prot. Agency
2200 Churchill Rd.
Springfield, III. 62706
Jeff R. Sell
Denver Regional Plan COG-DRCOG
2480 W. 26th Ave.
Denver, Col. 80211
W. Herbert Senft
Ball State University
Dept. of Biology
Muncie, Ind. 47306
Joel C. Settles
Hennepin Soil & Water
250 N. Central Ave., Suite 16
Wayzata, Minn. 55391
James R. Seyfer
S.D. Dept. Water and Natural Res.
Joe Foss Building
Pierre, S. Dak. 57501
Earl E. Shannon
Canviro Consultants, Ltd.
279 Weber St., N.
Waterloo, Ontario, Canada N2J3H8
J. Shapiro
University of Minnesota
310Pillsbury Dr., S.E.
220 Pillsbury Hall
Minneapolis, Minn. 55455
Ron Sharpin
META Systems, Inc.
10 Hoi worthy St.
Cambridge, Mass. 02138
Byron H. Shaw
University of Wisconsin
College of Nat. Res.
Stevens Point, Wis. 54482
Robert Shaw
Ontario Ministry of the Environ.
150 Ferrand Dr.
Don Mills, Ontario, Canada M3C 3C3
Gary Shearer
E. C. Jordan Co.
Portland, Maine 04112
Louis E. Shenman
Mudcat Div.
2337 Lemoine Ave.
Fort Lee, N.J. 07204
Catherine Shirvell
P.O. Box 550
Halifax, Nova Scotia, Canada B3J 2S7
Cole Shirvell
P.O. Box 550
Halifax, Nova Scotia, Canada B3J 2S7
-------
547
S. Ram Shrivastava
Larsen Engineers
444 Saginaw Dr.
Rochester, N.Y. 14623
Clifford A. Siegfried
New York Museum & Sci. Serv.
Biol. Survey
Albany, N.Y. 12230
Robert Singer
Colgate University
Dept. of Biology
Hamilton, N.Y. 13346
Jack Skrypek
Minn. Dept. Nat. Resources
Centennial Office Bldg., Box 12
St. Paul, Minn. 55255
Robert Sliwinski
RPI3
Ricketts Bldg., Room 102
Troy, N.Y. 12181
Eric Smeltzer
Vermont Dept. of Water Resources
Montpelier, Vt. 05602
Don Smith
U.S. Environ Prot. Agency, Reg. I
JFK Federal Bldg.
Boston, Mass. 02203
Douglas L. Smith
FHWA Office of Research
400 7th St. SWHRS-42
Washington, D.C. 20590
Gerald N. Smith
Aquatic Control Tech.
534 Boston Post Rd.
Wayland, Mass. 01778
Michael R. Smith
Maine Fish & Wildlife Dept.
Box 66
Enfield, Maine 04433
Phillip D. Snow
Civil Engineering Dept.
Union College
Schenectady, N.Y. 12308
Raymond A. Soltero
Eastern Washington University
Dept. of Biology
Cheney, Wash. 99004
Paul F. Sommer
Boston University
31 Champney St.
Brighton, Mass. 02135
Patrick W. Sorge
Iowa State University
Dept. of Animal Ecology
124 Sciences Hal I II
Ames, Iowa 50011
Michael Soukup
National Park Service
15 State St.
Boston, Mass 02109
John Sowles
Maine Dept. Environ. Prot.
State House
Augusta, Maine 04333
D. M. Spence
E. C. Jordan Co.
562 Congress St.
Portland, Maine 04110
Larry T. Spencer
Natural Science Dept.
Plymouth State College of the Univ. of N.H.
Plymouth, N. Y. 03264
Ronald C. Spong
City of Bloomington
1772 Ashland Ave.
St. Paul, Minn. 55104
Mark St. Cyr
Springfield College
46 Cranberry Lane
Holliston, Mass. 01746
William P. Stack
Baltimore City Water Quality Mgt.
Municipal Bldg., Room 305
200 N. Holliday St.
Baltimore, Md. 21202
William Staddard
Maine Dept. Environ. Prot.
State House, Station 17
Augusta, Maine 04333
-------
548
RESTORATION OF LAKES AND INLAND WATERS
Pius Stadelmann
Water Pollution Control of Cantone Lucerne
Klosterstr. 31
6002 Luzern, Switzerland
Bill Stallings
Oklahoma State Health
NE 10th & Stonewall
P.O. Box 53551
Oklahoma City, Okla. 73105
Cynthia A. Stanhope
Portland Water District
225 Douglass St.
Portland, Maine 04104
Jon G. Stanley
Dept. of Zoology
University of Maine
313 Murray Hall
Orono, Maine 04469
Kenneth Stewart
SEA Consultants
Charles Street
Rochester, N.H.
Ann Stroup
Los Alamos Scientific Lab.
MS 603 P.O. Box 1663
Los Alamos, New Mexico 87545
Anne Sulides
University of Maine at Orono
82Stillwater Village
Orono, Maine 04473
Jeffrey C. Sutherland
Williams & Works
611 Cascade West Parkway
Grand Rapids, Mich. 49506
Richard Swasey
Maine Dept. Environ. Prot.
State House, Station 17
Augusta, Maine 04333
David A. Sweet
KK&W Water District
Drawer 88
Kennebunk, Maine 04043
C.T. Taggart
McGill University
Dept. Biology
Montreal, Quebec, Canada
Judy Taggart
U.S. Environ. Prot. Agency
401 M St. S.W. (WH-585)
Washington, D.C. 20460
Doug Tawes
Compu-Chem
P.O.Box 12652
Research Triangle Park, N.C. 27709
Craig TenBroeck
Maine Dept. Conserv.
State House, Sta. 22
Augusta, Maine 04333
G.J. Thabaraj
State of Florida Dept. of Env. Reg.
2600 Blairstone Rd.
Tallahassee, Fla. 32301
Eberhard Thiele
University of Maine
Ft. Kent, Maine 04743
Dan Thirumwrthi
Nova Scotia Tech College
P.O.Box 1000
Halifax, N.S. Canada B3J 2X4
Fred H. Tholen
City of E. Grand Rapids
750 Lakeside Dr. S.E.
Grand Rapids, Minn. 49506
Craig H. Thomas
Bear Lake Regional Comm.
On 89 at Stateline
Fish Haven, Idaho 83261
Nelson Thomas
U.S. Environ. Prot. Agency
Lge Lakes Research St.
9311 Groh Rd.
Grosse Me, Mich. 48128
Ronald F Thomas
LMS Engineers
One Blue Hill Plaza
Pearl River, N.Y. 10965
Robert Thompson
Androscoggin Valley RPC
70 Court Street
Auburn, Maine 04210
-------
549
Kent Thornton
U.S. A.E. WES
P. 0. Box 631
Vicksburg, Miss. 39180
Laurence Tilly
CEER/UPR
Caparra Heights Station
San Juan, Puerto Rico 00935
Steven A. Timpano
Maine Dept. of Inland Fisheries & Wildlife
P. 0. Box 66
Enfield-, Maine 04433
Susan Titus
The Bionetics Corp.
P.O.Box 1575
Unit Hill Farms Station
Warrenton, Va. 22186
Susan Meyer Torrans
1000 NE 10th & Stonewall
Oklahoma City, Okla. 73105
Ronald E. Towne
N. H. Water Poll. Comm.
Box 95
Health & Welfare Bldg.
Loudon Rd.
Concord, N.H. 03301
Francesco B. Trama
Rutgers University
Dept. of Zoology
Nelson Bio. Lab.
P.O. Box 1059
Piscataway, N.J. 08854
John A. Tranquilli
III. Natural History Survey
172 Natural Resources Bldg.
Urbana, 111.61801
Nancy Morton Trautmann
235 Forest Home Dr.
Ithaca, N. Y. 14850
Joanne Tremper
Hartford Co.
Dept. of Public Works
23 N. Main St.
Bel Air, Md. 21014
David 0. Trew
Alberta Environment
Water Quality Control Bd.
9820 106 St.
Edmonton, Alberta, Canada T5K 2J6
Joan G. Trial
University of Maine
312Deering Hall
Orono, Maine 04469
David Troubridge
James F. MacLaren, Ltd.
1220 Sheppard Ave., East
Toronto, Ontario, Canada M2K 2T8
Lauren Tucker
Center for Natural Areas
Box 98
South Gardiner, Maine 04359
Judy Tumosa
USDA Soil Conserv. Serv.
Federal Building
Durham, N.H. 03824
Harold F. Udell
Conservation & Waterways
Lido Blvd.
Pt. Lookout, N.Y. 11569
Ants Uiga
P.O. Box 221
Palo Alto, Calif. 94302
Renita D. Uiga
P.O. Box 221
Palo Alto, Calif. 94302
James T. Ulanoski
Pa. Dept. of Environ. Res.
P.O.. Box 2063
Harrisburg, Pa. 17120
John K. Underwood
Nova Scotia Dept. of Environ.
P.O. Box 2107
Halifax, Nova Scotia, Canada B3J 3B7
Paul R. Vachon
New Eng. River Basin Comm.
177 Battery St.
Ice House
Burlington, Vt. 05401
Alan Van Arsdale
Mass. Dept. Environ. Quality Eng.
100 Cambridge St.
Boston, Mass. 02202
John Van Benschoten
Dept. Water Resources
State Office Bldg.
Montpelier, Vt. 05602
-------
550
RESTORATION OF LAKES AND INLAND WATERS
Janet Vance
5510 Country Dr. No. 24
Nashville, Tenn. 37211
Mary Vanderlaan
Mich. Dept. of Nat. Res.
Land Resource Prog.
P.O. Box 30028
Lansing, Mich. 48909
Douglas S. Vaughan
Oak Ridge National Lab.
Environmental Sciences
P.O. Box X
Oak Ridge, Tenn. 37830
Bo Verner
Atlas Copco
70 Demarest Dr.
Wayne, N.J. 07405
V.D. Vlugt
Dorpsstraat 53
Koudekerkaanner Ryn
Netherlands
Jean S. Wagener
Dartmouth Lakes
Advisory Board
36 Mount Pleasant Ave.
Dartmouth, Nova Scotia, Canada B3A 3T4
Gustov Wagner
267 Boston Rd.
N. Billerica, Mass. 01862
John Wagner
Dept. of Environ. Quality
Water Qual. Div.
401 W. 19th St.
Cheyenne, Wyo. 82002
Kenneth J. Wagner
New Jersey Dept. Environ. Prot.
Coleman Lane
Titusville, N.J. 08560
Charles Walbourn
Beckman Instruments
Microbics Operations
6200 El Camino Real
Carlsbad, Calif. 92008
Marcus C. Waldron
Clemson University
336 Long Hall
Clemson, S.C. 29631
Mary Veal Waldron
Okla. Dept. of Poll. Contr.
P.O. Box 53504
NE 10th & Stonewall
Oklahoma City, Okla. 73152
James E. Walsh
BEC Inc.
39 Maple St.
East Longmeadow, Mass. 01028
Thomas E. Walton, III
Jaca Corp.
550 Pinetown Rd.
Ft. Washington, Pa. 19034
Ming-Pin Wang
Mass. Inst. of Technology
305 Memorial Dr., Room 617A
Cambridge, Mass. 02139
Walton D. Watt
Canadian Dept./Fisheries & Marine Serv.
P.O. Box 550
Halifax, Nova Scotia, Canada B3J 257
Barron L. Weand
Virginia Tech
P.O. Box 784
Manassas, Va. 22110
W. A. L.Webber
Dept. of Local Govt.
Treasury Bldg.
Queen St., Brisbane 4000
Queensland, Australia
Richad Wedepahl
Wis. DNR
P.O. Box 7921
Madison, Wis. 53707
William & Diana Wegener
Fla. Game & Fish Comm.
207 West Carroll St.
Kissimmee, Fla, 32741
Irvine W. Wei
Northeastern University
21 Fairbanks Rd.
Lexington, Mass. 02173
Jane Weidman
Lee Pare & Assoc.
105 WhippleSt.
Providence, R.I. 02908
-------
551
Barbara Welch
Maine Dept. Environ. Prot.
State House
Augusta, Maine 04333
Robert J. Wengrzynek
USDA Soil Conserv. Serv.
Orono, Maine 04473
Michael Westphal
Maine Dept. of Environ. Prot.
State House, Station 17
Augusta, Maine 04333
Carolyn Wheeler
Bureau/Air Quality Control
State House, Station 17
Augusta, Maine 04333
W.S. White
Argonne National Lab.
9700 South Cass Ave.
Argonne, III. 60439
James R. Whitley
Missouri Dept. Conserv.
1110 College Ave.
Columbia, Mo. 65201
Gayle Whittaker
Mass. Water Pollution Control
P.O. Box 545
Westboro, Mass. 01581
Don Widmann
Nalco-Chemical
RouteS, Box 1328 E
Leesburg, Fla. 32748
H. Wiechers
Water Research Comm.
c/o S. African Embassy, Suite 300
2555 M St. N.W.
Washington, D.C. 20037
Jerry Wilhm
Oklahoma State University
School of Biol. Sciences
Stillwater, Okla. 74074
Doug Williams
U.S. Environ. Prot. Agency, Reg. V
26 W. St., Clair St.
Cincinnati, Ohio 45268
Scott Williams
RFD No. 1, Box 250
S. Paris, Maine 04281
Todd N. Williamson
Springfield College
Springfield, Mass. 01104
Ann Seaton Witzig
Louisiana State University
Center for Wetland Res.
Baton Rouge, La. 70803
Edward Woo
U.S. Environ. Prot. Agency
JFK Federal Bldg.
Boston, Mass. 02203
Lindsay W. Wood
Div. Labs and Research
Empire State Plaza
Albany, N.Y. 12201
Paul F. Woods
U.S. Geological Survey
Federal Bldg., Room 428
301 S. Park Ave.
Helena, Mont. 59601
Marie Wooster
COLA-Maine
P.O. Box 441
Rockland, Maine 04841
Norman Yan
Ontario Ministry of the Environ.
P.O. Box 213
Rexdale, Ontario, Canada
Janet Young
Portland Water Dist.
225 Douglass St.
P.O. Box 3553
Portland, Maine 04104
Tom Young
Clarkson College
Dept. of Civil & Env. Eng.
Potsdam, N.Y. 13676
John Zahradnik
University of British Columbia
Vancouver, B.C., Canada V6T 1W5
Fred Ziegler
AWARE, Inc.
P.O. Box 40284
Nashville, Tenn. 37204
-------
552 RESTORATION OF LAKES AND INLAND WATERS
Janet Zuckerman
University of Michigan
130 Audubon Rd.
Teaneck, NJ. 07666
Michael P. Zulzouski
Purcell Assoc.
90 National Dr.
Glastonbury, Conn. 06101
Fred Zwick
County of Westchester
Box 90, RR 2
Pound Ridge, N.Y. 10576
•U.S. GOVERNMENT PRIMTING OFFICE! 198I-O-72O-OI6/5996
-------
|