*EPA
United States Office of Water EPA 440/5-83-001
Environmental Protection Regulations and Standards
Agency Washington, D.C. 20460
Water
Lake Restoration,
Protection and
Management
'~
-------
LAKE RESTORATION,
PROTECTION, AND
MANAGEMENT
Proceedings of the
Second Annual Conference
North American Lake
Management Society
October 26-29, 1982
Vancouver, British Columbia
U.S. Environmental Protection Agency
Washington, D.C.
1983
-------
REVIEW NOTICE
This report has been reviewed by the U.S. En-
vironmental Protection Agency and approved for
publication. Approval does not signify that the
contents necessarily reflect the views and
policies of the Environmental Protection Agen-
cy, nor does mention of trade names or com-
mercial products constitute endorsement or
recommendations for use.
EPA 440/5-83-001
U.S. Environmental Protection
Agency
Office of Water Regulations
and Standards
Washington, D.C. 20460
-------
FOREWORD
Partnership — whether it be public/private or Federal/State/local — partnership
is the key to making governmental programs work today. This proceedings em-
bodies that concept: it is the product of a cooperative effort between a private
organization — the North American Lake Management Society — and a
Federal agency, the U.S. Environmental Protection Agency.
Those familiar with the history of the EPA Clean Lakes Program will
recognize this volume as the third biennial report on the protection and restora-
tion of this Nation's freshwater lakes. The first two (Lake Restoration in 1979
and the Restoration of Lakes and Inland Waters (1981)) were proceedings of
EPA conferences. This volume is the proceedings of a conference of the North
American Lake Management Society. EPA supported the publication, but had
no hand in the conference itself: a testimony to the fact that the private sector
can and will undertake worthwhile programs.
This Society represents the best expression of a cooperative approach to
solving environmental problems between the private sector and the government.
In its brief 2-year existence, NALMS has become the voice of thousands of
Americans who understand that their lakes need help — and who are willing to
put up the dollars to help them.
Those who organized the North American Lake Management Society were
the same people who helped put together EPA's Clean Lakes Program. They
built it as a partnership between the Federal government and local com-
munities, on a 50/50 cost-sharing basis.
In the seven years the program has been around, citizens have matched
almost $40 million of their own money with Federal funds to work on well over
200 lakes.
Most of these dollars — 87 percent, in fact — have been used to actually
work on lakes. Less than 8 cents per dollar have been spent on diagnostic
feasibility studies, and only 5.5 cents per dollar on the State classification
studies. With most of the money being used to clean up lakes, approximately
86 projects are either complete or nearing completion.
In 1981, EPA matched $11.5 million to work on lakes in 39 States. These
Federal awards ranged from $4,500 to over $11/4 million apiece. They were
about equally distributed between Phase I feasibility studies and implementation
projects.
In 1982, the Federal government matched $9 million in local dollars to com-
plete work on 23 lakes. All those awards were for existing Phase II projects and
were limited to the five criteria for grant awards directed-by Congress.
Congress appropriated $3 million in grant funds for Clean Lakes in 1983.
Completion of Phase II projects is of highest priority.
Americans continue to demonstrate their concern about the condition of their
lakes. They are saying yes to bond referenda to clean up their lakes; and they
are raising money in their own communities to help get rid of silt and weeds,
and reduce eutrophication of their lakes.
The Clean Lakes Program was built on the premise that the citizens of this
country, if informed of the facts, would act in their own best interests, and pay
their own way. They are. And the Federal government has just about fulfilled its
mission. EPA has supported the development and testing of various techniques
on this Nation's lakes and helped fund their use in cleaning up lakes. The pro-
gram is now in the operational phase and the future lies with the local citizens
and their desire to continue this successful program for their own lakes.
Those who led the way with the Clean Lakes Program are now among the
members of the North American Lake Management Society, an association that
-------
is fast proving it can establish a national conscience that will ensure our Na-
tion's lakes are protected. This third biennial report (as required by Section
304(j) of the 1977 Amendments to the Clean Water Act) on "methods, pro-
cedures, and processes as may be appropriate to restore and enhance the
quality of the Nation's publicly-owned freshwater lakes" resulted from a con-
ference of the North American Lake Management Society. We are proud to
have supported its publication. We believe this proceedings establishes a stan-
dard of excellence for the private sector, and certainly a pattern of cooperation
between private association and the Federal government.
Patrick Tobin
Acting Director
Criteria & Standards Division
IV
-------
PREFACE
All lakes are changing their quality, some faster than others and in ways that
we as human beings find unacceptable. Lakes resources are irreplaceable.
Once they are gone, they are gone. They also seem to have a threshold
beyond which it is no longer possible to technologically or economically return
them to a productive quality. How many can we lose before we affect the
essential ecological function they perform?
It would be nice to protect and improve the quality of all our lake resources,
but we cannot. There just isn't enough money and technological know-how to
do the job. Conventional pollution controls are frequently insufficient to protect
lakes or correct their degraded condition. Nonpoint sources of pollution carrying
toxics, pesticides, heavy metals, sediment, and urban drainage often contribute
greater than 50 percent of the pollution loading. Controlling these sources will
require further development and implementation of unconventional controls, par-
ticularly best management practices.
Citizen initiative is the greatest moderator in defining lake restoration quality.
We truly are "a government of the people, by the people, and for the people."
This is the fundamental principle behind the Clean Lakes Program; less
elegantly, Clean Lakes is grassroots participation. I firmly believe that given suf-
ficient and accurate information to deal with a problem and encouragement and
assistance in handling it, people will help themselves. And I have found they
really want to in the case of lake restoration. I have also found that it takes
government, at all levels, to provide that critical information and assistance.
The citizen must insist on his government's attention, though. In lake restora-
tion, he has to make his elected officials know what his lake quality problems
are and that he needs their support to solve them. He also needs to participate
actively in designing appropriate programs for lake pollution control and restora-
tion — participating in city council and commission meetings, working with
State legislators, advising Federal representatives.
The citizen must (1) help design local ordinances and zoning regulations, (2)
encourage the use of greenways and nonstructural measures to control and
treat nonpoint sources of pollution, (3) ensure that new urban drainage outfalls
and refurbished old ones adequately treat urban runoff before it is discharged
to streams and lakes, and (4) ensure that best management practices are re-
quired and used where applicable. Finally, citizens must insist that laws are
enacted to authorize local, State, and Federal jurisdictions to handle lake pollu-
tion control and restoration programs. And after they are enacted, they must
make sure these laws remain viable.
The goal for the U.S. program has not been achieved; only a handful of
States have developed a State/local operational program that addresses lake
quality problems. Others urgently need continued assistance as noted in a 1982
survey of the States by NALMS. We need to make this happen in all the States
that have lake resources. And to support similar Federal/provincial/local pro-
grams in Canada and Mexico.
EPA prepared a number of publications from 1978 through 1980 that address
where we are with lake restoration and what remains to be done. This volume
is the third in the series of proceedings documents on lake restoration that
-------
discuss management approaches, the science of lakes, modeling capabilities,
and engineering techniques. The Clean Lakes Program Guidance Manual
(EPA-440/5-81-003) introduced to local and State interests the best procedure to
address lake quality problems, specifying measurement protocols for lake resto-
ration projects conducted under the auspices of the Clean Lakes Program. The
goal was to develop standardized information on restorative techniques so their
effectiveness might be assessed and improved. Standardized coding forms and
reporting procedures for entering this information into EPA's Grants Information
System and the Water Quality Data System (STORET) are contained in the
Manual as well as in a recent publication (1982) by the JACA Corp., STORET
Guidance Manual, an Overview of the STORET System and the Clean Lakes
Program.
Without these comparative data, it may be impossible to improve lake
restoration capabilities. All who manage, conduct or support lake restoration
projects should use these protocols. Those who control the funds for lake work
should demand that this standardized gathering and recording of information be
continued.
We have come a long way in making lake quality a national priority in the
United States, but there is much to do! We took a substantial step on Sept. 10,
1980, by forming the North American Lake Management Society to coordinate
the efforts of all those who recognize the importance of lakes and the problems
they face. Our goal is to channel their concerns into a force that will (1) en-
courage governments to understand the importance of lakes and their pro-
blems; (2) make the protection and enhancement of lake resources a national,
State, provincial, and local priority; (3) ensure appropriate laws and regulations
are available to control pollution to these resources; and (4) ensure that ade-
quate financial and technological resources are available to control pollution to
our lakes, and upgrade those particularly needed and wanted by the general
public.
The challenge is clearly there. We can meet this challenge and we must
meet it head on, by working together to make lake restoration and protection a
reality on this continent.
Robert J. Johnson
President
North American Lake Management Society
-------
CONTENTS
Foreword in
Preface v
Conference Participants vii
PERSPECTIVES ON LAKE MANAGEMENT
From Water Quality to Ecosystem Management in the
Great Lakes Basin 1
J. R, Vallerttyne
A Few Pacific Northwest Examples of Short-term
Lake Restoration Successes and Potential Problems
With Some Techniques 4
Harry L. Gibbons, Jr., William H. Funk
Lake Sediment for Land Use Improvements 8
Glenn £ Stout, Michael J. Barcelona
RESTORATION AND PROTECTION TECHNIQUES
Restoration of Lake Eota 13
Harvey H. Harper, Martin P. Wanie/ista, Yousef A. Yousef
Nutrient Removal from Urban Stormwater by Wetland
Filtration: The Clear Lake Restroation Project 23
John Barton
Hypolimnetic Aeration and Functional Components of the
Lake Ecosystem: Phytoplankton and Zooplankton
Effects 31
Kenneth Ian Ashley
Lake Deepening Using In Situ Techniques 41
Richard E. Wedepohl, Adrian T. Hanson, Joseph E. Szewczyk
Hypolimnetic Withdrawal: Restoration of Lake
Wononscopomuc, Connecticut 46
ft W. Kortmann, £ Daws, C. ft. Frink, D. D. Henry
SOCIOECONOMIC BENEFITS OF LAKES AND RESTORATION
Hierarchy Theory: Theoretical and Methodological
Implications for Sociological Impact Evaluations 57
J. Lynn England
Keys to Lake Water Quality: Lake Quality Standards and
Point/Nonpoint Source Abatement Tradeoffs 62
Alfred M. Duda, Robert J, Johnson
Rehabilitation Plan for Lake Cajititlan: an Endangered,
Shallow Lake , , 69
J. G. Limon M., J. J. Amezcua C,, V. Bast/das ft.
Lake Water Quality and Beach Use in the Okanagan Basin,
British Columbia 73
Roger McNeill
An Evaluation of the Role Regulations for
Lake Protection and Management 77
D. A, Yanggen
ACID RAIN
Acidification of Headwater Lakes and Streams
in New England 83
Terry A. Haines, John J. Aklelaszek
Acid Sensitivity of Reservoirs in the Southern Blue
Ridge Province 88
AHda ft. Lewis. Harvey O/em
Ontario's Experimental Neutralization Program 92
Gareth Goodchild, James G. Hamilton
Effectiveness and Uncertainties Associated with the
Chemical Neutralization of Acidified Surface Waters 96
Douglas L. Britt, James £ Fraser
INTERNAL NUTRIENT LOADING DETECTION AND CONTROL
Submersed Macrophyte Community Structure and
Internal Loading: Relationship to Lake Ecosystem
Productivity and Succession 105
Stephen ft. Carpenter
Control of Internal Phosphorus Loading in a
Shallow Lake by Drawdown and Alum 112
J. M. Jacoby, E. 8. Welch, J. P. Michaud
Macrophyte Dieback: Effects on Nutrients and
Phytoplankton Dynamics 119
Dixon H. Landers, Elizabeth Lottes
Evaluation of Internal Phosphorus Loading
from Anaerobic Sediments 123
Steven B, Lazoff
Estimation of Internal Nutrient Loading in
Laguna Lake 127
G. C. Holdren, Jr.
FISCAL AND INSTITUTIONAL SUPPORT FOR LAKES
The State of Washington Lake Restoration Program:
Some Good News—Some Bad News
Roland E. Pine, Raymond A. So/tero, Richard D. Riley
Liming to Mitigate Surface Water Acidification:
International Programs, Strategies, and Economic
Conditions
James £ Fraser, Douglas L. Britt
The Role of Conservation Easements in
Lake Management
William Pray O'Connor, Gordon Chesters
Lake Protection by Watershed Management
for Wisconsin Lake Districts
George R, Gibson, Jr., Richard E, Wedepohl,
Doug/as ft. Knauer
The Massachusetts Clean Lakes Program
Eben Chesebrough
135
141
148
. 154
158
PRE- AND POST-RESTORATION ASSESSMENT DATA
AND TECHNIQUES
Experiences in Developing a Chlorophyll a Standard
in the Southeast to Protect Lakes, Reservoirs,
and Estuaries 163
ft. F. McGftee
Changes in Water Quality of Skaha Lake, British
Columbia, Following Reduction in Phosphorus Loading
Richard N. Nordin
., . 166
. . 171
Determination of Nutrient Sources for McNeely Lake ,.
G. C, Holdren, Jr.
A Method for Information Pooling to Reduce Lake
Model Prediction Error 177
Kenneth H, Recknow
PHYSICAL, CHEMICAL, AND BIOLOGICAL CONTROL OF
AQUATIC MACROPHYTES
Review of Management Tactics for Integrated Aquatic
Weed Management of Eurasian Water Milfoil
(Myriophyllum spicatum), Curly Leaf Pondweed
(Potamogeton crispus), and Elodea (Bodea canadensis).... 181
Stanley A. Nichols, Byron H. Shaw
Winter Drawdown for the Control of Eurasian Water
Milfoil in an Oregon Oxbow Lake (Blue Lake,
Multnomah County) 193
N. Stan Geiger
VII
-------
Milfoil Control in Seattle and the King County Region:
Metro's Harvesting Program 198
W. Terra Prodan
Guidelines for the Use of the Herbicide 2,4-D to
Control Eurasian Water Milfoil in British Columbia 200
Robert W, Adams
Observations on Herbivorous Insects That Feed on
Myriophyllum sp/catum in British Columbia 214
Bengt J. Kangasniemi
CAUSES AND CONTROL OF BLUE-GREENS: ALTERNATIVES
TO NUTRIENT CONTROL
Blue-Green Dominance in Lakes: the Role and
Management Significance of pH and CO2 219
Joseph Shapiro
Controlling Blue-Green Algae by Zooplankton Grazing .... 228
Robert E. Carlson, Steven A. Schoenberg
Cyanobaeterial Buoyancy Regulation and Blooms 234
Andrew R. Kiemer
The Nitrogen and Phosphorus Dependence of Blue-
Green Algal Dominance in Lakes
Val H. Smith
237
242
Cyanophage; History and Likelihood as a Control....
Paul R. Desjardins
Predatory Myxobacteria: Lytic Mechanisms and
Prospects as Biological Control Agents for
Cyanobacteria (Blue-Green Algae) 249
Jeffrey C, Burnham, Peter C. Fraieigh
PROBLEMS IN LAKE RESTORATION
Review of Lake Restoration Techniques and an
Evaluation of Harvesting and Herbicides 257
G. Dennis Cooks
Lake Restoration Criteria: the Limnologist's View
Versus Public Perception 267
Lowell L Klessig, Nicolaas W. Bouwes, Sr.
Impact of Watershed on Lake Quality
Yousef A. Yousef, Martin P. Wartietista, Harvey H. Harper
Urban Lakes: Their Benefits and Unique Problems ....
D, B, Porce/la
CONTRIBUTED PAPERS
Dredging Liberty Lake and Its Immediate
Environmental Impact
Stephen A. Breithaupt, David S. Lamb
Vancouver Lake: Pre-Restoration Status and
Restoration Progress Report
Richard B. Raymond, Fredrick C, Cooper
An Evaluation of 15 Lakes in King County, with
Projections of Future Quality
Joanne Davis, Robert G. Swartz
Lake Hicks Restoration Analysis
D. R. Heinle, H. A. Amowltz
Effects of Aluminum Sulfate to Midge Larvae
(Diptera: Chironomidae) and Rainbow Trout
(Sa/mo gaUneri)
David S. Lamb, Gary C. Bailey
Water Quality Management Strategy for Lake
Okeechobee, Florida
Frederick £ Davis, J, Steve Reel
A Comparison of the Costs of Harvesting and
Herbicides and Their Effectiveness in Nutrient
Removal and Control of Macrophyte Biomass...
Diane L Conyers, G. Dennis Cooke
An Annual Increment Lake Acidification-Fisheries
Response Model to Geologically Sensitive Areas
in Ontario, Canada
J. £ Hanna, M. F. P. Michalski
271
276
279
284
293
299
307
313
317
322
VIII
-------
North American Lake Management Society
Second Annual Conference
October 26-29, 1982
The Inn at Denman Place
Vancouver, British Columbia, Canada
President
Thomas U. Gordon
Cobbossee Watershed District
Winthrop, Maine
Conference Committee Chairman
Eugene B. Welch
University of Washington
Seattle, Washington
Local Arrangements Chairman
Peter R. Newroth
Ministry of Environment
Victoria, British Columbia
Exhibits Chairman
Gareth A. Goodchild
Ministry of Natural Resources
Toronto, Ontario
Session Chairmen
Perspectives on Lake Management: William Funk, Washington State Water Research Center, Pullman, Washington
Restoration and Protection Techniques: Joel Schilling, Minnesota Department of Natural Resources, St. Paul, Minnesota
Socioeconomic Benefits of Lakes & Restoration: Lowell Klessig, University of Wisconsin-Extension, Stevens Point, Wisconsin
Acid Rain: Matthew Scott, Maine Department of Environmental Protection, Augusta, Maine, and David Brakke, Western
Washington University, Bellingham, Washington
Internal Nutrient Loading Detection and Control: G. Dennis Cooke, Kent State University, Kent, Ohio
Fiscal and Institutional Support for Lakes: Thomas Gordon, Cobbossee Watershed District, Winthrop, Maine
Pre- and Post-Restoration Assessment Data and Techniques: Donald Porcella, Tetra Tech, Inc., Lafayette, California
Physical, Chemical, and Biological Control of Aquatic Macrophytes: Peter Newroth, Ministry of Environment, Victoria, British
Columbia
Causes and Controls of Blue-greens: Alternatives to Nutrient Control: Joseph Shapiro, University of Minnesota, Minneapolis,
Minnesota
Problems in Lake Restoration: Spencer Peterson, U.S. Environmental Protection Agency, Corvallis, Oregon
Contributed Papers: Ronald Raschke, U.S. Environmental Protection Agency, Athens, Georgia, and Donna Sefton, Illinois
Environmental Protection Agency, Springfield, Illinois
Editor
Judith Taggart
Lynn Moore, assistant
Production by
Stephen J. Downs, III
John M. Frazier
Cover art by Patricia J. Perry
IX
-------
Perspectives on Lake
Management
FROM WATER QUALITY TO ECOSYSTEM MANAGEMENT
IN THE GREAT LAKES BASIN
J. R. VALLENTYNE
Department of Fisheries and Oceans
Canada Centre for Inland Waters
Burlington, Ontario, Canada
INTRODUCTION*
The drainage basin of the St. Lawrence Great Lakes covers
three quarters of a million square kilometers with a 2:1 sur-
face ratio of land to water. At the beginning of the 19th cen-
tury the basin was inhabited by about 300,000 persons with
an average D-index (ratio of technological to physiological
energy consumption) probably in the range of 0,5 to 2. In
1982, the basin was inhabited by about 38,000,000 persons
with an average D-index of approximately 80 to 90, (See
Valientyne (1978) for explanation of the concept of
demophora and D-index.)
A succession of costly problems appeared in the basin
over this time interval: sickness and death from waterborne
diseases, erosion of topsoil and consequent filling in of har-
bors and channels, organic pollution from municipal and in-
dustrial wastes, persistent foams from nonbiodegradable
detergents, eutrophication from phosphates in detergents
and physiological wastes, toxic chemicals, and, most recent-
ly, acid rain. These problems first appeared in isolated
localities, later shifted from streams to harbors and em-
bayments, and in many cases spread throughout major
segments of the Great Lakes.
By the mid-1960's it was evident that human influences
had changed the chemistry and biology of the lower Great
Lakes (Beeton, 1965). Recognizing this, the governments
of the United States and Canada directed the International
Joint Commission to investigate and report on the nature
and extent of the problems, causes, and remedial measures,
The technical report to the Commission from the Interna-
tional Water Pollution Boards (1969) for Lakes Erie and On-
tario identified eutrophication as a problem of major concern.
To control eutrophication the Boards recommended reduc-
ing the concentrations of phosphate in detergents and
removing phosphate by chemical treatment at sewage treat-
ment plants.
This control program was based on a global review of the
causes and control of eutrophication prepared by R. A,
Vollenweider (1968) under the auspices of the Organization
for Economic Cooperation and Development. Vollenweider
showed that direct relationships existed between the loading
rates of phosphorus and nitrogen compounds to lakes and
the degree of eutrophication. The focus on phosphorus for
control was based on experimental and observational
evidence showing that the supply of phosphorus could be
made to limit plant growth in the lower Great Lakes by remov-
ing phosphorus from nutrient-rich sources, such as
detergents and effluents from sewage treatment plants.
* Editor's note: To illustrate the supposed benefits of constant
demophoric growth Dr. Valtentyne began the talk by pouring 1, 2,
4 and 8 shots of whiskey into four glasses, drinking one glass at the
beginning of each of the four sections. He did this to make the point
that constant growth is suicidal — whether tor whiskey In relation
to the human body or for human populations in finite ecosystems,
Dr. Valientyne declined to answer a question raised at the end of his
talk about the chemical nature of the liquid in the bottle.
-------
Lake Restoration, Protection and Management
CONTROL OF EUTROPHICATION IN THE
LOWER GREAT LAKES
Over the course of the decade 1972-82, Canada and six
of the eight States bordering the Great Lakes introduced
regulations governing permissible concentrations of
phosphate in detergents. By early 1982, only Ohio and Penn-
sylvania permitted high-phosphate detergents. Surprising-
ly, in 1982 Wisconsin allowed its detergent phosphate regula-
tions to lapse — a situation that will be interesting to monitor
for effects. By 1982, most municipalities in the Great Lakes
basin with populations exceeding 10,000 had reduced the
concentration of phosphorus in sewage effluents to levels
less than 1.0 mg P/l.
As a resuft of these measures, municipal phosphorus loads
to Lake Ontario were reduced from estimated values of 9,860
tonnes in 1972 to 2,520 tonnes in 1980; and loads to Lake
Erie from estimated values of 15,260 tonnes in 1972 to 3,580
tonnes in 1980 (Great Lakes Water Qual. Board, 1981), By
1981 it became clear that the program of phosphorus con-
trol had achieved the desired effect. Concentrations of solu-
ble reactive phosphate in Lake Ontario were approaching
levels predicted a decade earlier on the basis of a program
of phosphorus control as shown in Figure 1 (Dobson, 1981).
Similar, though less marked changes were discernible in data
for Lake Erie. There was also evidence of decreased
phytoplankton chlorophyll and increased transparency, but
the trends were not well marked.
The full story of the adjustment of the lower Great Lakes
to the control measures has yet to be told; nevertheless, It
is already beginning to show signs of becoming a success
story. Empirical evidence in conjunction with limnological
theory and governmental actions averted costly damage to
water quality that would have affected 18 percent of the
world's surficial supply of fresh water and the interests of
tens of millions of people.
On the other hand, the solution was piecemeal rather than
ecological. Recycling of phosphorus on land was practiced
at only a few locations in the basin. Also, even before
69 70 71 72 73 74 75 76 77 78 79
Year
81
Figure 1 .—Concentrations of soluble reactive phosphorus in spring
in Lake Ontario, 1969-1981. From Dobson (1981) with previously
unpublished data for 1981 also from Dobson.
eutrophication was brought under control two new and poten-
tially more serious problems appeared — toxic chemicals
in food chains and the emerging continental problem of acid
rain. This seemingly endless succession of problems caus-
ed the Great Lakes Research Advisory Board to alert the
International Joint Commission in 1977, and more com-
prehensively in 1978, to the futility of a piecemeal approach
to management.
THE ECOSYSTEM APPROACH
After reciting the lengthy succession of problems in the basin,
the Great Lakes Research Advisory Board (1978) describ-
ed the framework of an anticipatory approach to problem-
solving based on a view of the Great Lakes basin as an
ecosystem. Termed the ecosystem approach, it called for
an evaluation of social, economic, and environmental in-
terests. Human health was accented as an issue; jobs and
cars were in the ecosystem. It described the basis for politics
in an ecosystem context, rather than environment in a
political context.
Perhaps more than any other factor, toxic substances il-
lustrated the necessity of an ecosystem approach to
management. The list of known contaminants in water, fish,
plankton, sediments, and air of the Great Lakes basin
ecosystem included DDT, DDE, PCB's, PAH's, dieldrin,
Mirex, and a host of unidentified organic compounds. The
extent to which these technological chemicals were present
in human tissues was poorly known at the time, but the fact
that they were in the Great Lakes Basin ecosystem made
it only a matter of time before they came into equilibrium
with the human population. The Board warned the Commis-
sion that in the absence of an integrative and anticipatory
approach to problem-solving, new problems were bound to
emerge. The nature, time, and place of the next problem
could not be predicted, but the appearance of such a pro-
blem was certain under conditions of continuing demophoric
growth.
The first member of a class of new and unexpected pro-
blems emerged within 3 weeks of that prediction: The Love
Canal. Toxic industrial chemicals were seeping out of cor-
roded containers buried decades earlier in the naive belief
that they had been disposed of "for good." Months later the
governments of the United States and Canada introduced
major segments of the ecosystem approach into the Great
Lakes Water Quality Agreement signed in Ottawa on Nov.
22, 1978,
POLITICS IN AN ECOSYSTEM CONTEXT
fcosysfem is a flexible term used to designate a subdivi-
sion of the biosphere with boundaries arbitrarily defined to
suit particular purposes In hand. As limnologists, we are used
to defining ecosystems as lakes, segments of streams, or
drainage basins. But, we can also speak of personal
ecosystems ("me" plus everything outside me with which
"I" interact) and municipalities, States, Provinces, Nations
as ecosystems delineated by political boundaries.
The ecosystem concept was developed originally by
ecologists. Their studies were limited to tracts of land or
bodies of water small enough to be studied by single per-
sons or groups of students under the direction of a single
professor. Now, in regions of high demophoric activity,
ecosystem management has become a necessity on scales
far beyond the capability of ecologists to resolve. Multibillion
dollar problems in the Great Lakes basin such as eutrophica-
tion, toxic chemicals in human food chains, and acid rain
show this clearly. Ecosystem accounting in the Great Lakes
basin has, of necessity, fallen into the governmental domain.
New management approaches have appeared in response
to successional changes resulting from demophoric (popula-
-------
Perspectives on Lake Management
tion and technological) growth. Hundreds of years ago (and
still to some extent today!) people were too preoccupied with
an egosystem approach to be aware of ecofeedback from
their surroundings. Later, in the middle of the 19th century,
following discovery of the transmission of cholera by drink-
ing water from wells contaminated with the feces of infected
persons, a piecemeal approach to problem-solving was in-
troduced. In the 1960's and early 1970's the environmen-
tal approach to problem-solving came into effect. Now, in
the 1980's, we are shifting from an environmental to an
ecosystem approach to planning and problem-solving.
The distinction between environment and ecosysfem is
equivalent to that between house and home. House implies
something external; home, something that we see ourselves
in, even when we are not there. Politics in an ecosystem
context thus calls for management of people and their en-
vironments jointly in the context of politically shared
ecosystems as homes. To gage the extent of its develop-
ment one has only to consider the pronounced changes in
attitudes, perceptions, and behavior of people and organi-
zations today as compared to 20 years ago. We have
changed our ecosystems, and now those changed eco-
systems are changing us.
The essence of the ecosystem approach lies in relating
parts to wholes (people to ecosystems that they are in) rather
than parts to parts (people to their external environments).
It is an integrative and anticipatory approach to planning and
problem-solving that is based on knowledge of the interrela-
tionships of systems in nature, the necessity of ecological
behavior, and the desirability of adopting an ethic of respect
for other systems of nature. It is developing — and not only
in the Great Lakes basin.
The utility of the ecosystem approach to management lies
in the coming together of social, economic, and environmen-
tal interests. This makes indirect and long-term costs of
piecemeal planning more apparent than they would other-
wise be. It is more advanced in its development than
ecologists, who have traditionally studied ecosystems little
influenced by man, might suspect. If this is more obvious
in the Great Lakes basin than elsewhere, it is because pro-
blems have been so large in scale, conspicuous, and ex-
tensively studied there.
A workshop "Implementing the Ecosystem Approach" will
take place in Hiram, Ohio, March 22-24, 1983, under the
sponsorship of the International Joint Commission, the Great
Lakes Fishery Commission, the International Association for
Great Lakes Research, and Great Lakes Tomorrow, a
citizens' group. The goal of this workshop is to produce a
set of specific, pragmatic initiatives to further understanding
and implementation of the ecosystem approach in the Great
Lakes basin. These initiatives will be focused on introduc-
ing systemic changes in our institutions that will help us make
better use of the new knowledge gained through ecology
of our relationships with other systems of Nature. Organiza-
tions such as the North American Lake Management Society
may find it useful to examine these initiatives to see how
they might be transformed into wider, national contexts.
REFERENCES
Beeton, A.M. 1965. Eutrophication of the St. Lawrence Great Lakes
Limnol. Oceanogr. 10: 240-254.
Dobson, H.F.H. 1981. Trophic conditions and trends in the Laurentian
Great Lakes. Water Qual. Bull. 6:146.
Great Lakes Research Advisory Board. 1978. The ecosystem ap-
proach: scope and implications of an ecosystem approach to trans-
boundary problems in the Great Lakes Basin. Great Lakes Reg.
Off., Int. Joint Comm. Windsor, Ontario.
Great Lakes Water Quality Board. 1981. Rep. on Great Lakes
water quality. Great Lakes Reg. Off., Int. Joint Comm., Windsor,
Ontario.
International Water Pollution Boards (Lake Erie, Lake Ontario-
St. Lawrence River). 1969. Pollution of Lake Erie, Lake Ontario
and the International Section of the St. Lawrence River. Vol. I.
A rep. to the Int. Joint Comm., Ottawa and Washington.
Vallentyne, J.R. 1978. Today is yesterday's tomorrow. Verh. Int.
Verein. Limnol. 20:1-12.
Vollenweider, R.A. 1968. Scientific fundamentals of the eutrophi-
cation of lakes and flowing waters with particular reference to
nitrogen and phosphorus as factors in eutrophication. OECD Tech.
Rep. Organ. Econ. Coop. Dev. DAS/CS168.27, Paris.
-------
A FEW PACIFIC NORTHWEST EXAMPLES OF SHORT-TERM LAKE
RESTORATION SUCCESSES AND POTENTIAL PROBLEMS
WITH SOME TECHNIQUES
HARRY L GIBBONS, JR.
WILLIAM H. FUNK
Civil and Environmental Engineering
Washington State University
Pullman, Washington
ABSTRACT
Short-term lake restoration successes are numerous. Three lake restoration techniques that have been
used in the Pacific Northwest are discussed. Diversion, dredging, and nutrient inactivation are very useful
methods for removing adverse environmental conditions within lakes. Although improvement in water quality
is the goal of these lake restoration procedures, the degree of improvement may be impaired if investiga-
tion of the problem, planning, and operational controls are inadequate. Some potential problems associated
with the techniques described are presented and discussed.
INTRODUCTION
There are numerous examples of successful lake restora-
tion treatments, especially when addressing short-term suc-
cesses (short term arbitrarily defined as 3 years). A con-
siderable number of external and in-lake restorative pro-
cesses have been implemented in the Pacific Northwest.
Watershed management programs that control point and
nonpoint sources of sediment and nutrients have proven to
be extremely valuable in arresting eutrophic process in lakes.
Examples of these activities are the construction of sewers
for wastewater and diversion channels to divert stormwater
runoff. In-lake restoration manipulations have included
drawdown, biomanipulation, dilution, aquatic macrophyte
controls, aeration, nutrient inactivation, and dredging.
This paper discusses diversion, dredging, and nutrient in-
activation as successful lake restoration procedures that have
taken place in the Pacific Northwest. In addition, we point
out some potential problems that may reduce their effec-
tiveness if they are not properly addressed prior to initiation.
DIVERSION
Diversion of the inflowing water has been employed when
it has been deemed to be the most cost-effective method
available for controlling nutrient and suspended solid
loadings, bacterial inputs, and other allochthonous material
inputs from surface water inlets. Near Seattle, Wash., a diver-
sion is planned for Lake Hicks to reduce the fecal bacterial
inputs from summer and fall storm runoff (Heinle and Ar-
nowitz, 1982). At Mill Creek Pond, Walla Walla, Wash., a
successful diversion of sediment-laden spring runoff waters
above a critical flow has been implemented. In that situa-
tion runoff contained an excessive amount of colloidal mat-
ter, which resulted in a turbidity problem. Previously Mill
Creek Pond was mainly a water storage reservoir, present-
ly it is a multi-use reservoir. It now supports a healthy trout
fishery and it is a popular water contact recreation area.
To reduce nutrient loading from a nutrient rich wetland,
the principal inlet to Liberty Lake, Wash., was diverted
around the periphery of the marsh through which it flowed
(Funk et al. 1982). However, the stream channels that then
bypassed the marsh became overgrown with vegetation and
filled in with sediment within 2 years after construction. As
a consequence, the wetland was again severely flooded in
the spring of 1982, permitting the transfer of substantial
amounts of nutrients from the marsh into Liberty Lake. That
nutrient input most likely contributed heavily to the mild blue-
green algal bloom that occurred in the lake in late summer
and early fall of 1982.
If a water and sediment balance had been performed prior
to the diversion, more appropriate channelization procedures
might have been considered and implemented. For instance,
building sediment traps upstream of small intermittent
tributaries to the west fork of Liberty Creek would have
averted heavy sediment deposits. To prevent heavy
vegetative growth and to protect the stream bank, the chan-
nels could be lined with large rocks.
Watershed erosion during and after storm runoff events
was causing turbidity problems in Lake Fenwick in western
Washington. To solve the turbidity problems a diversion of
stormwater runoff was undertaken to reduce suspended
solids loading into the lake (URS, 1982). The high storm-
water flows are diverted to a preexisting pond that serves
as a sedimentation basin, then the clarified water Is al-
lowed to flow into the lake (Fig. 1). The results have been
fairly good. Suspended solids loading was reduced by 54
to 70 percent and, in an added benefit, total phosphorus
loading was reduced by 17 to 40 percent (URS, 1982). Water
clarity of the lake dramatically increased. However, excessive
algal growth did occur more often than before restoration,
probably as a result of increased light transmission.
In the planning of the Lake Fenwick restoration more ef-
fort should have been given to the internal loading of
nutrients and the potential for an algal response, especially
since the primary objective of the restoration was to decrease
turbidity. Because of political pressure to solve the turbidity
problem as soon as possible the possibility that in-lake
restorative methods might be needed were not addressed
as they should have been (Anderson et al. 1982). Now it
seems that Lake Fenwick may need further restoration ef-
forts to limit algal growth.
-------
Perspectives on Lake Management
DIVERSION LINE
PATH OF ORIGINAL STREAM BED
Figure 1 .—Location of Fenwick Lake showing diversion path and
sedimentation pond.
DREDGING
One of the most potentially effective in-lake restoration tech-
niques is dredging. There are four major reasons for remov-
ing lake sediments: deepening the lake basin, nutrient con-
trol, aquatic weed control, and extraction of toxic substances
(Peterson, 1981). One of the largest lake-deepening projects
is currently underway at Vancouver Lake, Wash. There is
every indication that the dredging effort will achieve the
restoration objectives.
At Long Lake, Wash., dredging was part of the restorative
process. The dredging was designed to remove some of the
nutrient rich sediment and to deepen that part of the lake
basin. A very practical result of the dredging was to facilitate
the drawdown of Long Lake, another phase of the restora-
tion. The hydraulic dredging took 60 days from July to
September 1978 to remove 60,000 cubic yards of sediment
(Entrance Engineers, 1980). The removal of sediment did
allow increased recreational use of that part of the lake. In
addition, the sediment removal was a necessary prerequisite
to drawdown. The effect of the dredging activity on the water
quality of the lake is hard to assess, because only 5 per-
cent of the lake bottom sediments were dredged. Also, other
lake restoration techniques (Entrance Engineers, 1980;
Jacoby et al. 1982) were implemented at the lake following
the dredging, i.e., drawdown and alum treatment.
One of the major problems associated with dredging is
the disposal of the dredged materials (spoils). Perhaps future
dredging projects will not have toxic materials within the
spoils, and properly disposed spoils will become an asset
to the disposal environment. However, today, unless ade-
quate planning is employed prior to dredging and enough
money is set aside for the properly planned disposal
methodology, the spoils may degrade that area by limiting
its use.
There are two basic objectives in handling the spoils. First,
a containment area must be designed to hold and dewater
the dredged materials as soon as possible. Second, that con-
tainment area should be available for reuse (Walsh and
Bomben, 1981; Walsh and Carranza, 1981) or for another
use once dredging has stopped. The second objective is
easily satisfied by locating the disposal site where the
characteristics of the spoils enhance future land use.
However, most problems result from improper design of the
first objective, especially when hydraulic dredging is
employed. Frequently the time needed to dewater the spoils
is underestimated, hence the storage area is undersized,
causing delays and unexpected costs. By performing small-
scale tests prior to implementation, these problems can be
avoided.
The southern end of Liberty Lake was dredged as part
of a multiphase restoration of that lake from 1980 through
1981. The purpose of dredging was to remove nutrient-laden
sediments (Peat substrate: Fig. 2) and thus reduce sedi-
ment/water nutrient exchange. Since the targeted sediments
are the substrate for dense aquatic macrophyte beds,
removal of those sediments would provide a degree of
aquatic macrophyte control (Funk et al. 1982; Funk and Gib-
bons, 1980). There were two major problems with the dredg-
ing operation at Liberty Lake (Funk and Gibbons, unpubl.).
As discussed earlier, a problem developed in the rate of
dewatering the spoils. That delayed the dredging applica-
tion, caused contractor disputes, and mandated redesign of
the spoils disposal handling methodology.
- OUTLET
• TRANSITION ZONE
El SAND SUBSTRATE
D SILT SUBSTRATE
M PEAT SUBSTRATE
LAKE SURFACE ELEVATION = 624.28 METERS
EAST FORK
LIBERTY CREEK
Figure 2.— Sediment substrate map of Liberty Lake, Washington.
The dredging of the southern end of the lake took place within the
area of the peat substrate.
However, the most serious problem in terms of its effect
on the success of restoration effort was in the dredging
operation itself. Although the hydraulic dredge performed
adequately, the path of the cutter head did not overlap
because of operator error. This resulted in an uneven
coverage of the bottom sediments (Fig. 3b). Within a year
the mounds and trenches merged into one another through
the process of slumping. The nutrient rich sediment still
covered the bottom of the lake, although to a lesser depth
(Fig. 3c). Hence, neither the nutrient/water exchange rate
nor the aquatic macrophyte growth were reduced. In fact,
because of the exposed sediment surface the nutrient sedi-
ment/water exchange rate probably increased, and the
aquatic macrophytes have recovered to their predredging
density in the area that was improperly dredged. Of the 50
acres dredged, only about 20 acres were dredged efficient-
ly (Funk and Gibbons, unpubl.). Figure 3 illustrates this pro-
blem. That situation can be prevented by checking the move-
ment and location of the dredge during operation.
-------
Lake Restoration, Protection and Management
Figure 3.—Illustration of the net results of uneven coverage in the
sediment removal process, (a) outline of the target sediment to be
removed, (b) cross section of sediment immediately after dredging,
(c) cross section of sediment redistribution.
As shown in Figure 4, when sediment is removed to
deepen a lake, a problem may develop if the sediment is
not removed completely. In a restoration effort to deepen
a lake, first the amount of sediment to be removed must be
identified (Figure 4a). For an example, assume 90 percent
of the target sediment can be effectively dredged within cost
restraints (Figure 4b). If the dredging is carried out without
regard for the critical angle of repose of the remaining
sediments, taking into account the force of water movements
on that angle, slumping will occur after dredging (Figure 4c).
Not only will the overall effectiveness of the dredging be
reduced, but water quality problems could develop. For in-
stance, if the sediment that becomes exposed and dis-
placed is high in nutrients internal nutrient loading could
significantly increase, resulting in excessive algal growths.
To prevent that from happening research is necessary at
the site to determine the proper tailing angle to leave a stable
sediment.
Another important point about dredging is that immediately
following the dredging operation a nutrient inactivation tech-
nique should be undertaken. No matter how efficient the
dredging is, there is newly-exposed sediment that has not
obtained an equilibrium with the overlying water in relation
to nutrients, particularly phosphorus. Also, dredging in-
creases nutrient concentrations in lake water. In this case,
they may be neutralized, for example, by applying aluminum
sulfate to the surface waters. Not only will that treatment
reduce nutrient concentrations in the water column but it will
also reduce any turbidity that has been generated.
NUTRIENT INACTIVATION
Nutrient inactivation has been a very successful in-lake
restoration technique, especially when considering the short-
term effects on the planktonic primary producers. The most
commonly employed technique uses aluminum compounds
to precipitate phosphorus (Cooke and Kennedy, 1981).
However, in Cline's Pond, Ore., zirconium chloride was ap-
plied in 1974 (Funk and Gibbons, 1979), and was very suc-
cessful in reducing algal growth. Hence, although aluminum
compounds are very popular they are not the only viable
alternatives.
An example of a successful hypolimnetic and surface ap-
plication of alum is Medical Lake (Gasperino et al. 1980).
In 1977 over a 5-week period alum was applied at the sur-
face and at a depth of 4.5 m. The treatment dramatically
decreased chlorophyll a and phosphorus concentrations.
Figure 4.—Cross section of a hypothetical lake basin with sediment
targetted for removal (4a), sediment remaining immediately after
dredging (4b), showing sediment slump after equilibrium is estab-
lished (4c).
-------
Perspectives on Lake Management
LAKE SURFACE ELEVATION = 624.26 METERS
LUKE DEPTHS IN METERS
JAST FORK/
LIBERTY CREEK
Figure 5.—Areas treated with alum in the Fall of 1980 (north end)
and Spring of 1981 (south end) in Liberty Lake.
Liberty Lake has had three separate surface alum
treatments. The first, a whole lake treatment in 1974, was
very successful. It prevented blue-green algae blooms for
3 years as predicted by Funk et al. (1975). The second alum
treatment covered approximately 70 percent of the surface
area in the northern end of the lake in the fall of 1980 (Fig.
5). In the spring of 1981 after completion of the dredging,
alum was again applied to Liberty Lake in the southern end
of the lake (Fig. 5).
Most likely the last two alum treatments largely contributed
to the lack of a major blue-green algal bloom in the sum-
mers of 1981 and 1982. Liberty Lake is a good example of
the successful use of alum as a restoration technique in the
Pacific Northwest.
With proper pretreatment evaluation of the aluminum dose
and type of compounds to be used as well as field testing
to determine what conditions have to be satisfied to obtain
the desired floe behavior, many lakes may be able to use
nutrient inactivation as a short-term restoration technique.
In summary, lake restoration can accomplish the overall
objective of enhancing water quality, but only when all the
phases of restoration are adequately planned and properly
controlled during the implementation of that plan. In most
instances, we can improve the water quality of our lakes from
extreme eutrophic conditions to much more valuable bodies
of water. However, more research into the methods of
restoration is needed to improve our short-term success
percentage and increase the long-term benefits of lake
restoration.
REFERENCES
Anderson, D.E., D.W. Browns, and E.A. Quinlea. 1982. Personal
communication. Metro, Seattle, Wash. URS Co., Seattle, Wash.
Cooke, G.D., and R.H. Kennedy, 1981. Precipitation and inactivation
of phosphorus as a lake restoration technique. EPA 600/3-81-012.
U.S. Environ. Prot. Agency, Washington, D.C.
Entrance Engineers. 1980. Long Lake Restoration Project. Final rep.
Prepared for Kitsap County, Wash.
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, Washington, D.C.
1980. Lake assessment in preparation for a multiphase res-
toration treatment. Pages 226-237 in Restoration of Lakes and
Inland Waters. EPA 440/5-81-010. U.S. Environ. Prot. Agency,
Washington, D.C.
1982. Unpubl. data.
Funk, W.H., et al. 1975. Determination, extent and nature of non-
point source enrichment of Liberty Lake and possible treatment.
State of Wash. Environ. Res. Center Rep. No. 23. Washington
State Univ., Pullman.
. 1982. Preliminary assessment of multiphase restoration
efforts at Liberty Lake, Wash. Wash. Water Res. Center, Rep.
43. Washington State Univ. Pullman.
Gasperino, A.F., et al. 1980. Medical Lake improvement project:
success story. Pages 424-428 in Restoration of Lakes and Inland
Waters. EPA 440/5-81-010. U.S. Environ. Prot. Agency,
Washington, D.C.
Heinle, D.R., and M.A. Arnowitz. 1982. Lake Hicks restoration
analysis. Pres. at PNPCA and BCWWA Conf. November. Van-
couver, Br. Col.
Jacoby, J.M., D.D. Lynch, E.B. Welch, and M.A. Perkins. 1982.
Internal phosphorus loading in a shallow eutrophic lake. Water
Res. 16:911-919.
Peterson, S.A. 1981. Sediment removal as a lake restoration tech-
nique. EPA 600/3-81-013. U.S. Environ. Prot. Agency, Washington,
D.C.
URS Co. 1982. Final Rep. Lake Fenwick restoration project. Subm.
to U.S. Environ. Prot. Agency, U.S. Dep. Energy.
Walsh, J.E., and S.M. Bomben. 1981. Containment area design
for dredged lake materials. Presented at Am. Soc. Chem. Eng.
Water Forum, 1981. San Francisco, Calif.
Walsh, J.E., and C. Carranza. 1981. Netting Lake restoration pro-
ject. A discussion of alternative designs for dredged material con-
tainment area. Presented at Natl. Assoc. Environ. Prof. Conf.
Baltimore, Md. October.
-------
LAKE SEDIMENT FOR LAND USE IMPROVEMENTS
GLENN E. STOUT
Water Resources Center
University of Illinois at Urbana-Champaign
Urbana, Illinois
MICHAEL J. BARCELONA
State Water Survey
Division of the Illinois Department of Energy and Natural Resources
Champaign, Illinois
ABSTRACT
Currently, the midwestern United States is experiencing problems of erosion, sedimentation, and lost lake
and reservoir impoundment capacities. This paper documents research being coordinated by the Univer-
sity of Illinois at Urbana-Champaign demonstrating the benefits of lake reclamation, the value of dredged
sediment in agricultural watersheds, and efficient means of restoring dredged sediment to agricultural land
as a soil amendment. Research has shown that sediment deposits in the midwestern agricultural regions
are high in nutrient elements and low in toxic metals and organic compounds, A demonstration has been
carried out to evaluate methods for placing lake-bed sediment on agricultural land in such a manner as
to prevent the loss of nutrients and sediment through further erosion and to evaluate the agricultural pro-
ductivity of restored sediments. Terraces with rapid dewatering systems have been built on steep slopes
so that the land is productive after 1 year of rebuilding. Further tests are proposed to determine the poten-
tial of furrow and spray irrigation in applying the silt on the farm lands during the growing season in order
to refurbish soils that have lost their high productive capacity. The water will optimize crop production.
Current results at Lake Paradise in Coles County near Mattoon, III. are encouraging.
INTRODUCTION
Approximately 30 percent of Illinois' 84 municipal water sup-
ply reservoirs are experiencing severe sedimentation pro-
blems, with capacity losses of 1 percent or more each year,
according to the Illinois Environmental Protection Agency
(IEPA) (Sefton, pers. comm.). Furthermore, 56 percent of the
lakes and reservoirs are experiencing moderate to severe
sedimentation problems with capacity losses of 0.5 percent
or more each year. All Illinois municipal reservoirs are fac-
ing some capacity losses resulting from sedimentation; this
could cause supply problems some time in the future. Even-
tually, these lakes become wetlands, and new lakes have
to be developed. Lakes less than 50 years old have already
disappeared.
Moreover, most of Illinois' other lakes and reservoirs
used for recreation, flood control, electric power generation,
and so forth are experiencing similar problems. lEPA's 1978
Assessment and Classification of Illinois' Lakes reports that
of 353 Illinois lakes, 29 percent were in "poor condition,"
and 68 percent exhibited a "high problem potential."
Nevertheless, the demand for high quality Illinois lakes has
increased because of expanding demands of municipal
populations, industry, and energy needs, along with an in-
creased demand for local water recreational opportunities,
resulting partially from increased transportation costs.
Because of these increasing problems and demands, it
has become evident that the problems of our lakes must be
researched and that methodologies for their management
and restoration must be developed. The Lake Paradise
restoration project at Mattoon, III. has become an important
pilot project in identifying problems facing most of Illinois'
lakes and developing needed methodologies.
The Lake Paradise studies indicate that the sediment there
can be dredged to increase the storage capacity of the lake
and that the sediments are high in nutrients. These findings
also indicate that the sediment can probably be restored to
the surrounding farmland in a manner that will increase the
productivity of the soil. Furthermore, the studies on the
recreational potential and the economics of reclamation
(Stout, et al. 1982) indicate that recreation may be an im-
portant benefit to lake restoration and that lake restoration
may be cost efficient compared to other alternatives, such
as purchasing more land for additional water-supply lakes.
BACKGROUND
Lake Paradise is a small, 70.4 ha (176-acre) artificial im-
poundment approximately 7.5 km (4 1/2 miles) southwest
of Mattoon. It was originally constructed in 1907 to supply
water to a rapidly expanding railroad industry. Lake Paradise
sits at the head of the Little Wabash River basin and has
a drainage basin of some 4,800 ha (12,000 acres). Twice
prior to 1944, the lake was expanded in size, achieving peak
total surface area of 90 ha (225 acres). After World War II,
Mattoon began to attract industry, and its population grew
until Lake Paradise was unable to meet the area's water
needs, especially during periods of low rainfall. In 1958, Lake
Mattoon with a much greater capacity was constructed 8.3
km (5 miles) downstream from Lake Paradise. The Mattoon
area's water supply seemed assured until well into the 21st
century.
An estimated 9,090 tonnes (10,000 tons) of sediment wash
into Lake Paradise each year; this has reduced the lake's
surface area by approximately 25 percent and its storage
capacity by approximately 31 percent (Bogner, 1980).
Moreover, suspended sediment necessitates additional
water treatment to satisfy the quality requirements for mu-
nicipal use. Furthermore, unabated reduction of the lake's
storage capacity through sedimentation could cause water
shortages for Mattoon by the year 2000 (Lake Land College).
-------
Perspectives on Lake Management
HISTORY
In 1976, citizens organized a group to promote the restora-
tion of Lake Paradise. In 1978, they approached the Water
Resources Center (WRC), University of Illinois at Ur-
bana-Champaign (UIUC), for assistance and guidance on
how they might prevent rapid degradation of Lake Paradise.
The center responded with a report on the preservation and
restoration of both Lake Paradise and Lake Mattoon since
an area's total water resources need to be assessed rather
than its individual units. The citizens' group became Lake
Paradise Regional Renewal, Inc., a not-for-profit corporation.
Silt Surveys
In 1979, arrangements were made with the Illinois Depart-
ment of Transportation's Division of Water Resources (DWR)
to conduct a silt survey of Lake Paradise. Much of the ac-
tual field work was done by the Illinois State Water Survey
(The Illinois State Water Survey, the Illinois State Geological
Survey, and the Illinois Natural History Survey are all divi-
sions of the Illinois Department of Energy and Natural
Resources).
The nutrient values of the sediments were determined by
the UIUC Department of Agronomy. During the winter of
1979 to 80, Illinois State Geological Survey (ISGS), studied
drilled core samples of sediment to determine the geological
history of the sedimentation and the chemistry of the various
levels. Samples of these cores were provided to the Illinois
Natural History Survey and to Southern Illinois University at
Edwardsville, which respectively determined the amounts of
chlorinated organic chemicals and made isotopic determina-
tions of the lead in the samples.
In the spring of 1980, the Illinois Department of Energy
and Natural Resources (ENR) granted the WRC funds for
a chemical study of the potential impacts of hydraulic dredg-
ing of portions of Lake Paradise, as well as funds for studies
on the recreational and economic benefits of lake restora-
tion. Michael Barcelona, ISWS, coordinated the chemical
study of the hydraulic dredging impacts.
Economic Benefits Study
Recognizing that lake reclamation is an expensive pro-
cess but one with many economic benefits, the WRC ap-
proached the UIUC Department of Agricultural Economics
to conduct a study of the economic benefits of restoring Lake
Paradise and to determine the common factors involved in
lake restoration in general. The findings indicate that the
estimated economic benefits of lake restoration are great and
that further research on the economics of restoring the lake
should be conducted.
Based upon these findings and the evaluation that Lake
Paradise's bottom silt was of a high quality containing many
nutrients, it was decided that the sediments should be
restored to the land. Lake Paradise Regional Renewal con-
tacted representatives of the State legislature and the Illinois
Department of Agriculture (IDA) for assistance on a demon-
stration project. Funds were provided in FY 81 in the IDA
budget for a demonstration project by the University of Il-
linois Agricultural Experiment Station.
WATER QUALITY OF THE RESERVOIR
Lake Paradi.se is typical of artificial impoundments in
Illinois—it is in poor condition and exhibits a high potential
for problems in optimum usage (Hite et al. 1980) unless
rehabilitation is begun soon. Although there can be no per-
manent solution to these problems without a drastic change
in watershed management practices, dredging the upper por-
tion of the lake can buy time and improve its trap efficiency
to protect Lake Mattoon.
The water quality of the lake is suprisingly good consider-
ing the quality of its tributary, the Little Wabash River. Many
of the existing problems—algal blooms and turbidity—are
directly related to the influx of sediment-associated nutrients
and flows from nearby residential septic systems (III. Environ.
Prot. Agency, 1980). Dredging could only improve the lake's
water quality, particularly if the dredged sediment is dewa-
tered at a site distant from the lake's shoreline.
SEDIMENT AS THE PRINCIPAL THREAT
TO THE LAKE AS A RESERVOIR
Between 1908 and 1979 the lake lost 31.1 percent of its
original capacity, with 24.4 percent of this loss occurring
since 1931. This translates into an average capacity loss rate
of 0.51 percent per year. Segments 9, 10, 11, and 14 (see
Fig. 1) have been the hardest hit by sediment deposition
(Bogner, 1980).
Because the segments hardest hit by deposition are in
the upper portion of the lake, the overall trap efficiency and
hence the net sedimentation accumulation rate decreases.
As the lake fills with sediment a greater portion of the flow
will spill over the dam, carrying a larger portion of the fine
silt. Lake Paradise will therefore "outlive" itself as a reser-
voir (calculated to be 2114 A.D. based on the capacity loss
rate), and Lake Mattoon will lose capacity more rapidly than
expected. In other words, long before Lake Paradise disap-
pears into wetland-prairie, Lake Mattoon will probably ex-
hibit severe sediment-related problems.
The surface sediment accumulation aptly illustrates sedi-
ment transport dynamics. Two trends are of note. The first
is clearly evident from Figure 1: wetlands are gradually ac-
cumulating at the head of the lake, as indicated by shaded
Figure 1.—Lake Paradise and cross section locations (Bogner 1980)
-------
Lake Restoration, Protection and Management
areas with 70 to 80 percent silts (0.05 to 0.002 mm), while
the remaining clays ( 0.002 m) are carried farther into the
lake, especially during flood or high flow periods (such as
the years 1953, 1966, and 1979, corresponding to surveys
of the area). These accumulations in the upper part of the
lake are very liable to resuspension. Particle size sorting
(gradations of particle size with distance downstream) is not
expected to be very well developed, although this is shown
to a degree by the available surface particle size measure-
ments (Gross and Cahill, 1980; Fehrenbacher, 1980).
The second trend is the gradual increase in total sediment
accumulation from the headwaters toward the dam. These
net accumulations are derived from the difference in the
original bottom topography to that measured during 1979
(Bogner, 1980; Barcelona et al. 1980). The average sedi-
ment column heights are noted for each segment of the lake
in Figure 2. Particle size distributions in downstream sur-
face sediments show silt sizes have decreased to 50 per-
cent, clays again making up the remainder (Fehrenbacher,
1980).
• R6 g ft.
' t • PARADISE LAKE
COLES COUNTY,
ILLINOIS
SCALE OF FEET
Figure 2.—Lake Paradise: net sediment accumulations in specific
segments (modified from Bogner 1980).
These data show the effect of sorting, in that smaller par-
ticles (clays) accumulate downstream, though the process
is not as dramatic as might be expected. This effect is pro-
bably a function of the shallow nature of the lake, making
it easier for resuspension to take place by wind, aquatic life,
and boat traffic. When smaller particle sizes are transported
farther downstream elements bound to particles by surface
effects are concentrated because smaller particles have
larger surface area to weight ratios. The effect is clearly
shown in the distribution of trace metals and organochlorine
compounds in the lake's surface sediments (Hite et al. 1980;
Gross and Cahill, 1980; Barcelona et al. 1980). These data
are shown in Table 1 using the three divisions of the lake
used by the IEPA: reach 3 includes the upper arms (Bogner
segments 8-9); 2, the middle portion; and 1, the deep area
between the original and old dams.
Of the trace elements of interest, Cu, Zn, and Cr show
a tendency to increase downlake. These results show that
the upper reach of the lake contains rather unpolluted sur-
ficial sediment. Gross and Cahill (1980) pointed this out in
comparing their results to those from other lakes in the
Midwest.
These results suggest that the lake is losing its sediment
trap efficiency and will probably not protect Lake Mattoon
very far into the future. The sediments in the upper reach
are among the least polluted in the lake and dredging will
probably be most effective and environmentally sound in that
area. The data on sediment accumulation rates there come
from various sources (Bogner, 1980; Gross and Cahill, 1980;
Brugam, 1980).
The choice of the "best" value must be made carefully.
The head waters of a lake vary significantly in velocity and
sediment load with time, yet any net sedimentation rate must
be arrived at by assuming a uniform rate. A uniform rate is
certainly not the case with Lake Paradise. From topographic
(Bogner, 1980) and particle size (Gross and Cahill, 1980;
Fehrenbacher, 1980) considerations, the net sedimentation
rate is about 1.9 to 2.5 cm/yr1 (1 inch/yr1). Because of the
variation in flow and deposition dynamics, the dating method
(Brugam, 1980), with its limiting assumptions, is probably
not applicable to a core in the upper reaches of the lake.
Dredging by hydraulic means would thus remove approx-
imately 50 cm of the sediment, or that which has ac-
cumulated in the last 20 to 40 years.
SEDIMENT CHEMISTRY: SUITABILITY AS
A SOIL AMENDMENT
The sediment in the upper reaches is physically the best
to remove. It is low in clay content (for efficient dewater-
ing), is generally nonpolluted, is located in shallow water,
and its removal would most likely prolong the capability
of the lake to protect Lake Mattoon. Chemical considera-
tions of the sediment composition will largely dictate the
ultimate disposition of the sediment, i.e., disposal or use
as a soil amendment.
From the point of view of trace metallic elements and
organochlorine compounds, the sediment is most similar
to a soil with a good balance of trace nutrient elements,
as well as nitrogen and phosphorous. Potassium is limited.
The fate and mobility of the metallic elements or pesticide
residues depend on many factors: pH, oxidation state, and
the chemical species present. These questions can be
answered only after a careful site/sediment investigation.
However, a comparison of the Lake Paradise sediment
from the upper reaches with commercial soil amendments
(dehydrated sewage sludge and cow manure), as well as
an average soil, confirm the suitability of land application
as a disposal practice. Table 2 lists the chemical composi-
tion of these solids.
In summary, careful inspection of the data in Table 2
shows that the Lake Paradise sediment compares most
favorably with a terrestrial average soil and cow manure
(although more area-specific data may be available). It is
clear that the sediment is much cleaner than the commer-
cial soil amendments. The greatest concern in the use of
sewage sludge as a soil amendment is the absorption of
Cd and Zn by the cultivar (CAST, 1980). The most com-
prehensive study to date on this subject notes that absorp-
tion is a function of soil conditions and that individual
cultivars show wide variation in results (CAST, 1980). Soil
pH has been identified as the most critical factor in plant
absorption of these metals. (Refer to T. D. Hinesly, UIUC
Agronomy Department, a member of the CAST task group
for further comment.) Barcelona (1981) believes that if the
10
-------
sediment is well worked into the soil as Fehrenbacher sug-
gests, the soil pH can be kept above 6.
SOIL REBUILDING
As a result of the previous studies, two other grants have
been obtained to further research lake reclamation at Lake
Paradise.
In 1980 Lake Paradise Regional Renewal worked with
members of the Illinois Water Resources Commission and
IDA, along with the University of Illinois Agricultural Ex-
periment Station and WRC, to obtain a $75,000 grant for
a study to demonstrate and evaluate methods for placing
lake bed sediment on agricultural land in such a manner
as to prevent the loss of nutrients and sediments through
Perspectives on Lake Management
further erosion and runoff, and to evaluate the agricultural
productivity of restored sediments.
This pilot study of the Lake Paradise sediments will be
conducted for 5 years on a farm site about one mile from
the lake. During the spring of 1981, six .004 ha (0.01-acre)
plots were covered with sediment hauled from the lake's
exposed bottom with sediment depths of 45 cm (18 in-
ches), 30 cm (12 inches) and a check plot of undisturbed
soil. During the summer of 1981, another two .4 ha
(1.0-acre) plots were covered with wet sediment 1.2 m (4
feet) in depth that was hydraulically dredged from the lake
bottom and piped to the site. Terraces and dewatering
systems were constructed so that the sediments and
nutrients settled while the water drained back to the lake.
Table 1. — Distribution of trace metals and organochlorine compounds In Lake Paradise, 1980.
Parameter (mg/kg dry weight)
Reach
Dieldrin
Cu
Cd
Pb
Zn
Cr
As
Hg
Reference
3
2
1
Increase
0.031
0.004
0.051
0.032
0.031
0.044
downstream?
19
20
24
23
26
30
28
29
Yes
1 40
40
30
1 40
1.8 45
1 40
84
93
94
110
110
120
120
Yes
19
22
87
27
28
30
30
Yes
4.3
4.6
8.6
7.4
7.6
8.0
8.6
Likely
0.26
0.17
0.10
0.10
0.14
0.08
0.12
(Hite et al. 1980)
(Gross and Cahill,
1980)
(Hite et al. 1980)
(Barcelona et al.
1980)
Table 2. — Comparison of Paradise Lake surficial sediments with commercial soil amendments and an average terrestrial
soil.
Chemical parameter
(mg/kg dry weight)
Dieldrin
PCS (polychlorinated
Lake Paradise
sediment
(Hite et al.
1980)
(INHS1)
0.020-0.003 0.031-0.004
0.0021-0.01233 <0.01
Average
soil
(Lantzy &
Mackenzie,
1979)
Cow Manure2
(Agway, Inc.)
<0.01
<0.01
Commercial soil amendments
(Furr et al. 1976)
Vertigreen (Chicago, 111.
sludge)
0.24
11. 604
Milorganite
(Milwaukee,
Wis., sludge)
0.05
2.204
biphenyls)
Nitrogen
Phosphorous
As
Cd
Cu
Cr
Fe
Hg
Mn
Pb
(Gross & Cahill,
1980)
not available
977.00
8.60
0.10
24.00
87.00
—
0.10
—
•C5.00
2,250.00
840.00
4.40
1.00
20.00
20.00
20,000.00
0.22
370.00
40.00
5.00
0.50
20.00
100.00
38,000.00
0.05
850.00
10.00
9,000.00
12,500.00
4.00
0.80
62.00
56.00
15,900.00
0.20
286.00
16.20
39,900.00
16,000.00
29.00
14.80
578.00
207.00
8,800.00
6.10
95.00
605.00
58,000.00
18,400.00
3.30
444.00
1,288.00
14,000.00
43,000.00
3.40
134.00
2,253.00
1 Pesticide Laboratory, Illinois Natural History Survey. 1980.
2 Agway, Inc. (Furr et al. 1976)
3 as Arochlor 1242
4 as Arochlor 1254
11
-------
Lake Restoration, Protection and Management
Corn and soybeans will be grown on the plots, and
harvests will be evaluated to determine the productivity of
the various levels of sediment application. In addition, soil
nutrient analyses will be conducted prior to planting and after
harvesting to determine changes in fertility of the sediment
during each growing season.
To help evaluate the methods of sediment restoration, daily
samples of sediment and chemical nutrients from the return
flow to the lake were analyzed. Dredge discharge also was
sampled daily and analyzed for particle size distribution and
changes in nutrient levels. During storms, runoff samples
were taken at 30-minute intervals.
By October 1981, Lembke (in prep.) reported that the
preliminary results of the sediment-dewatering project were
very promising. Dewatering of the wet sediment occurred
much faster than expected. Furthermore, in October 1981,
harvests of the corn grown on the 0.01 A dry-dredged sedi-
ment resulted in a 29-bushel increase on the plots with an
18-inch sediment application as compared to control plots
with no application. Because of heavy spring and early sum-
mer rains, as well as lack of funds for evaluation, no crop
was planted in 1982.
GENERAL RECOMMENDATIONS AND
CONCLUSIONS
Efficient management of our resources requires comprehen-
sive evaluation of the uses, costs, and benefits of the
resource, of alternative resources, and of the users' percep-
tions of the value of the resource as well as their ability to
pay for it.
The Lake Paradise project suggests there may be many
instances in Illinois where the reclamation of a lake may be
a valuable and efficient management strategy if the reclama-
tion project is integrated with a watershed management pro-
gram. The Lake Paradise case study also provides one
model for evaluating the potential of restoring these lakes.
The Lake Paradise studies on dredging feasibility suggest
that dredging can be an environmentally safe element of a
lake reclamation program. In light of recent IEPA studies of
the sediments of Illinois lakes, it appears that the sediments
of most Illinois lakes located in agricultural watersheds can
be safely dredged and restored as soil amendments to
agricultural land in the near vicinity of the lakes.
Although the economic benefits of reclaiming a lake are
very difficult to quantify, Deo's study (Stout et al. 1982) of
the economic benefits of reclaiming Lake Paradise suggests
that dredging, im many instances, may be a cost-efficient
means for restoring the capacity of a community's water
supply and for increasing the water quality of the supply.
This efficiency depends upon many factors, however, such
as the availability and cost of land that might be used to con-
struct a new reservoir.
Based upon the results of these studies of Lake Paradise,
it is recommended that midwestern communities experien-
cing problems with their lakes caused by sediment suspen-
sion and accumulation should consider lake reclamation as
one possible element of a water resources management pro-
gram. In evaluating the potential of lake reclamation, com-
munities should conduct studies to determine the feasibility
of dredging and the uses of the dredged sediments, the
economic benefits of reclamation as compared to other alter-
natives, and local residents' perceptions of the benefits of
reclaiming a lake.
REFERENCES
Barcelona, M. J. 1981. Paradise Lake dredging feasibility study:
an annotated overview. III. State Water Surv., Urbana.
Barcelona, M. J., H. S. Chiang, and S. R. Heffelfinger.
1980. Lake Paradise wet-dredging operations — feasibility study.
III. State Water Surv., Urbana.
Bogner, W.C. 1980. Sediment survey of Paradise Lake, Mattoon,
III. Ill State Water Surv., Urbana.
Brugam, R. B. 1980. Lead-210 analysis of a sediment core from
Paradise Lake, Coles Co., III. Dep. Biolog. Sci., Southern Illinois
Univ., Edwardsville.
CAST. 1980., Effects of sewage sludge on the cadmium and
zinc content of crops. Rep. 83. News from CAST 7:80-82.
Fehrenbacher, J. B. 1980. Particle size and nutrient analysis
of Lake Paradise sediment from Core LP-1. Memo, May 1.
Furr, A. K., et al. 1976. Multielement and chlorinated hydrocarbon
analysis of municipal sewage sludges of American cities. Environ.
Sci. Technol. 10:683-687.
Gross, D. L, and R. C. Cahill. 1980. Geologic and chemical
studies of sediment in Lake Paradise. III. State Geolog. Surv.,
Urbana.
Hite, R. L., M. H. Kelly, and M. M. King. 1980. "Limnology of
Paradise Lake, June - October 1979. Monitor. Unit, III. Environ.
Prot. Agency, Springfield.
Illinois Environmental Protection Agency. 1978. Assessment and
Classification of Illinois Lakes. Vol. I. Springfield.
Lake Land College. 1977. Project Renewal: A proposal for the
reclamation of Lake Paradise, Coles County, Illinois. Mattoon, III.
Lantzy, R. J., and F. T. Mackenzie. 1979. Atmospheric trace metals:
global cycles and assessment of man's input. Geochim.
Cosmochim. Acta 43:511-525.
Lembke, W. D., et al. Lake renewal and watershed management:
Lake Paradise. Univ. Illinois Water Resour. Center, draft rep. in
prep.
Sefton, Donna. 1981.
Pers. comm.
Environ. Prot. Agency, Springfield.
Stout, G. E., et al. 1982. The feasibility and benefits of reclaiming
a man-made Lake: a case study of Lake Paradise, Mattoon, III.
Univ. III. Water Resour. Center, Urbana.
12
-------
Restoration and Protection
Techniques
RESTORATION OF LAKE EOLA
HARVEY H. HARPER
MARTIN P. WANIELISTA
YOUSEF A. YOUSEF
Department of Civil Engineering and Environmental Sciences
University of Central Florida
Orlando, Florida
ABSTRACT
Lake Eola is a small land-locked lake in downtown Orlando, Fla. Continual urban stormwater imputs
have caused significant deterioration of lake water quality. The lake is characterized by high rates
of algal production, a stratified and anaerobic hypolimnion, and periodic fish and duck kills. As a result,
a restoration project was begun to restore Lake Eola. Analysis of the contributing watershed indicated
that a first-flush effect was present with approximately 90 percent of stormwater pollutant mass con-
tained in the first 1.3 centimeters of excess rainfall. Since an area was not available within the highly
urbanized watershed for conventional stormwater management practices, an underground exfiltra-
tion system was developed. Stormwater management facilities were constructed and pollutant mass
efficiencies estimated. Treatment of stormwater by underground exfiltration was shown to be an at-
tractive economic alternative when compared to traditional forms. An aluminum sulfate-based waste
sludge from a drinking water treatment process was shown to be effective in both inactivating anaerobic
phosphorus release and in filtrating stormwater to remove phosphorus and heavy metals. Anaerobic
phosphorus release was shown to be as high as 1.5 mg-P/m2/day in Lake Eola sediments. A sedi-
ment application of alum sludge at a dose of 7 metric tons/hectare was applied to Lake Eola to minimize
this release. Also, two in-line filters were constructed for filtration of stormwater through alum sludge.
Actual field removal percentages were estimated for these filters. Monitoring of water quality in Lake
Eola will continue to document the success of this restoration project.
INTRODUCTION
Lake Eola is a small land-locked lake located in the heart
of downtown Orlando, Florida. The lake receives direct storm-
water runoff by way of storm sewers from a watershed of
approximately 65 hectares of dense commercial and residen-
tial areas surrounding the lake. There are currently 11 ac-
tive street drains that drain stormwater into the lake with no
treatment of any kind. The natural shoreline of the lake has
been replaced with a stone wall to prevent flooding of the
adjacent parkland, and numerous small patches of rooted
emergent macrophytes exist along this wall. The level of the
lake is controlled by a drainage well that drains into an
underlying artesian aquifer. Physical characteristics of Lake
Eola are listed in Table 1.
A summary of stormwater loading rates into Lake Eola for
various parameters is listed in Table 2. These continual
stormwater inputs have caused a significant deterioration in
water quality. With the exception of areas near the shoreline,
the bottom of the lake has become covered with an ac-
13
-------
Lake Restoration, Protection and Management
Table 1. — Physical characteristics of Lake Eola, Florida.
(Harper, 1979)
Parameter
Quantity
Average surface area 10.92 hectares
Average volume 3.30 x io5m3
Mean depth 3.02 m
Maximum depth 6.8 m
Length of shoreline 1417 m
Shoreline development 1.21
Volume development 1.72
Average height above sea level 26.8 m
Table 2. — Summary of stormwater loading rates into
Lake Eola (Wanielista et at. 1981)
Parameter
Suspended solids
BOD5
COD
TOC
TKN
NH3-N
Orthophosphorus
NO3-N
Total P
Zinc
Nickel
Copper
Iron
Lead
Average stormwater
Average loading concentration
(kg/yr) (mg/l)
54,505
5,390
39,105
52,030
1,760
226
211
816
264
204
15.4
37.4
524
234
131
13
74
99
3.3
0.43
0.51
1.96
0.48
0.38
0.03
0.07
0.99
0.44
cumulation of loose, flocculent, partially decomposed organic
matter that is easily disturbed. The loose nature of this
material makes it difficult for rooted submergent plants to
exist, and no rooted submergent plants of any kind have
been seen in Lake Eola. As a result of these decomposition
processes, concentrations of dissolved oxygen, although
usually at or above saturation near the water surface, drop
periodically during the spring and summer months to 1 mg/l
or less at depths of 4 meters or greater. In areas near the
center of the lake this organic matter, subjected to long
periods of anoxic and reducing conditions, has formed into
sapropel, complete with the characteristic hydrogen sulfide
smell. Floating masses of dead algae and fish and their ac-
companying odor are a common occurrence in Lake Eola,
and the lake has been classified as eutrophic by both the
Shannon-Brezonik and Vollenweider trophic state models.
In addition, Salmonella, Shigella and Clostridium botulinum
have been isolated from the water and shoreline sediments
in Lake Eola.
Vertical analyses in Lake Eola indicate that two distinct
periods can be observed in terms of water quality. One of
these periods is an unstratified condition that occurs mainly
during the winter and spring and is characterized for the most
part by isograde curves of dissolved oxygen, temperature,
pH, ammonia, and ORP with increasing depth (Fig. 1). The
stratified period, which occurs in summer and fall, is
characterized by decreasing pH, temperature, dissolved ox-
ygen, and ORP with increasing depth. Concentrations of am-
monia, total phosphorus, and alkalinity increase significant-
ly near the bottom (Fig. 2).
Lake Eola is a focal point and tourist attraction in
downtown Orlando. The surrounding parkland is beautifully
landscaped and is used as a site for many social and cultural
events throughout the year. Therefore, a restoration project
was initiated in 1978 to restore Lake Eola. Estimated benefits
of this restoration project are listed in Table 3 (Wanielista
et al. 1981).
(mg/m3) .
Chy"a" ,10
P (mv)QRP -577-505-433 -361 -289 -217
(a) (b)
Figure 1 .—Physical-chemical depth profiles of Lake Eola on 2/18/82.
14
-------
PHILOSOPHY OF RESTORATION
TECHNIQUES
Since stormwater runoff was determined to be the primary
source of pollution entering Lake Eola, management of this
source was considered to be essential in restoration of this
lake. It was determined through a series of laboratory alga
bioassays that phosphorus is the limiting nutrient in Lake
Eola during much of the year, and restoration techniques
were centered around management of this nutrient. It was
determined through analysis of the Shannon-Brezonik and
Vollenweider trophic state indices that reducing phosphorus
loadings approximately 80 percent would reduce Lake Eola
from eutrophic status to a borderline oligotrophic/mesotrophic
state.
Restoration and Protection Techniques
Analysis of the phosphorus budget in Lake Eola indicated
that the two main sources of this element were from storm-
water runoff and internal recycling of phosphorus from
anaerobic sediments. As seen in Table 2, stormwater runoff
contributes an average of 264 kg of phosphorus per year
into Lake Eola. Marshall (1980) determined that recycling
of phosphorus released from bottom sediments accounted
for as much as 25 percent of the phosphorus concentration
in the lake at any given time. Atmospheric fallout and ground-
water infiltration were found to be insignificant in terms of
phosphorus inputs when compared to the previous two
sources. Therefore, it became obvious that treatment of both
the stormwater runoff entering the lake as well as the bot-
tom sediment would be necessary to begin a restoration of
Lake Eola.
Table 3. — Estimated benefits of restoration of Lake Eola.
Activity
Music concerts
Arts/crafts
Tourist visits
Fish-a-thons
Food concessions
Paddle boats
Children's park
Relaxation/aesthetics
Jogging
Land value
*1
2
3
1
2
2
1
1
1
4
Approximate
frequency/year
35
3
Constant
3
Constant
Constant
Constant
Constant
Constant
Constant
Approximate
people-visits/yr
87,500
60,000
180,000
3,000
—
5,000
125,000
200,000
50,000
$/yr
262,500
300,000
90,000
9,000
100,000
20,000
187,500
600,000
150,000
600,000
TOTALS:
710,500
2,319,000
1 Based on estimated attendance and an expenditure of $3 per person per visit
2 Based on concession money received by the city of Orlando and an estimated attendance
3 Grey-line of Orlando estimated visits as a portion of a larger tour
« Based on lake-front vs. non-lake-front property tax
(mg/nAchyV
02(%)
10
pH
ecc)
0
5.5
ri
0
0.6
1.0-
1.6-
2.0
2.5-
3.0-
„ 3.5-
E
I <-
H
Si 4.6-
o
5.0-
5.5-
6.0-
6.5-
7.0-
7.6-
20
_¥_
65
I
8.0
40
ao
i
100
120
_70J (mg/l) TP
~7°(mfl/l)NH3
7.0
8.0
21
24
27
(mg/l)ALK
1 (mv)ORP-S
Figure 2.—Physical-chemical depth profiles of Lake Eola on 6/10/80.
15
-------
Lake Restoration, Protection and Management
TREATMENT BY UNDERGROUND
PERCOLATION
Since stormwater runoff was determined to be the primary
source of pollution entering Lake Eola, management of this
source is essential in restoring this lake. An extensive in-
vestigation was conducted of each contributing sub-
watershed and the loading characteristics of each were defin-
ed. During an earlier study (Wanielista et al. 1981), it was
determined that a first-flush effect was exhibited by these
watersheds, and a large percentage of pollutant mass could
be removed by diversion and subsequent retention of first-
flush waters. The efficiency of diversion for retention was
studied by Wanielista (1977) through extensive simulations
of yearly rainfall/runoff events on the Orlando, Florida, area,
and the results are listed in Table 4.
Table 4. — Efficiencies of stormwater pollutant mass
removal and corresponding diversion volumes for areas
exhibiting first-flush effects.
Average annual percent of pollutant Treatment "diversion"
mass removed (%)* volume (centimeters)
15cm CONCRETE SLAB
•} \
99
95
90
80
2.60
1.95
1.30
0.65
* Average of suspended solids, BOD5, total nitrogen, and total phosphorus.
Therefore, to achieve the desired 80 percent reduction in
phosphorus loadings, a diversion of at least the first
0.65 cm of excess rainfall would be necessary. Since the
area surrounding Lake Eola is composed of high-density
residential and commercial areas, no land area was available
for construction of traditional stormwater management struc-
tures such as above ground retention basins. As a result,
alternative stormwater treatment techniques were required.
After considerable literature research into existing storm-
water management methods, it was decided to divert storm-
water runoff to underground percolation systems. Perforated
aluminum pipe 130 centimeters in diameter with 0.65 cm
openings was chosen for percolation. The pipe could be
buried several feet under the roadway with a riser pipe ex-
tending to a stormwater inlet along the curb. The bottom
elevation of the percolation pipe should not be lower than
the average anticipated high water table level. Stormwater
runoff would be allowed to flow off parking lots and residen-
tial areas by gravity sheet flow into the street gutter and be
intercepted by the stormwater inlet above the percolation
tank. If the volume of rainfall runoff is sufficient to fill the per-
colation tank, the excess will flow over the inlet, along the
gutter, until it intercepts the regular stormwater system for
discharge into Lake Eola. Exfiltration into the surrounding
soil would allow the system to drain between storm events.
The t/olume of the tank could then be sized to achieve the
desired pollutant removal efficiency.
A typical cross section of the percolation design is shown
in Figure 3. The aluminum pipe is surrounded on the sides
and ends by 30.5 centimeters of gravel (DOT #11 or
equivalent) which increases the effective percolation volume.
This gravel was characterized as "slag" rock. A filter fabric
with a coefficient of permeability = 3.2 x 10~2 cm/sec and
filtration rate = 107 liters m2/min is also wrapped around the
gravel on all sides and ends and sealed at the seams with
roofing tar to prevent soil from washing into the pipe and
possibly causing structural or stability problems for the street
above. The excavation hole is then filled with clean building
sand and the gutter and roadway replaced.
SCALE 1cm = 24cm
Figure 3.—Typical section of underground percolation system.
Table 5. — Treatment levels and design equations.
Treatment
volume
(centimeters)
0.65
1.30
1.95
2.60
3.25
Design equations
V = 0.016
V = 0.046
V-0.09
V-0.14
V = 0.20
(A)1-28
(A)1-18
(A)'-"
(A)'-"?
(A)104
where: V = volume of treatment for 100% impervious area (acre-feet)
A = watershed area (acres)
The required volume of the percolation system was
calculated from equations developed by Wanielista (1979)
through computer simulation of rainfall/runoff/infiltration on
watersheds up to 22 hectares. These equations for various
diversion volumes and soil types are listed in Table 5.
A pilot structure using underground percolation (exfiltra-
tion) was constructed in the downtown Orlando area and was
designed to contain the first 1.3 centimeters of stormwater
runoff from a 0.4 hectare parking lot and street in an area
within sandy soils, classified by the Soil Conservation Ser-
vice as hydrologic Type A. The aluminum pipe was 130 cen-
timeters in diameter and 12.2 meters in length and was
assumed to operate at a minimum exfiltration rate of 2.5
cm/hour. Construction was performed by the city of Orlan-
do at a cost, including labor and materials, of almost $10,000.
Observation of storm events after construction indicated
that the actual exfiltration rate was considerably higher than
2.5 cm/hour and that a rainfall excess of approximately 2.8
centimeters over a 3-hour period was required to fill the per-
colation tank before excess runoff was routed into the ex-
isting curb and gutter system. A chemical analysis for this
rainfall event comparing stormwater runoff entering the per-
colation tank and stormwater runoff that flowed over the in-
let after the tank had filled is listed in Table 6. Significant
reductions were achieved in concentrations of alkalinity,
TOC, BOD, NO3-N, TKN, orthophosphorus, total
phosphorus, and suspended solids.
The success of this pilot installation led to the conclusion
that in the areas within the Lake Eola watershed where the
groundwater elevation permitted percolation, this was the
preferred method of pollution control. However, analysis of
the watershed indicated at least 36 contributing parking areas
would require diversion and treatment of stormwater runoff.
At an estimated cost of $10,000 per unit, the treatment cost
would be approximately $360,000 and would require over
a year of construction for just parking areas, leaving large
portions of streets and driveways untreated.
Further refinement and evolution of treatment techniques
led to the conclusion that the optimum solution would be
16
-------
Restoration and Protection Techniques
Table 6. — Average chemical analysis of stormwater runoff entering the underground percolation tank compared
with the overflow.
Parameter
PH
Turbidity (JTU)
Alkalinity (mg/l)
TOC (mg/l)
NO3-N (mg/l)
TKN (mg/l)
Dissolved PO4-P (mg/l)
Total P (mg/l)
SS (mg/l)
VSS(mg/l)
BOD (mg/l)
Stormwater entering
the percolation tank
8.30
28.0
96.0
218.0
8.9
5.37
0.530
0.650
27.0
26.0
18.3
Stormwater
overflow
7.63
12.0
28.7
14.2
0.43
0.41
0.020
0.158
4.2
2.9
1.0
Percent
change
- 8.1
-57.2
-70.1
-93.5
-95.2
-92.4
-96.2
-75.7
-84.4
-88.9
-94.5
to intercept, wherever possible, the main storm sewer line
before it enters the lake, allowing the stormwater from all
contributing parking lots, streets, and driveways to be treated.
Nine key points of interception were defined in areas where
percolation was about 25 cm/hour, and exfiltration systems
ranging in length from 12 meters to 55 meters with diameters
of 91.5 to 130 centimeters were designed at a total projected
capital cost of $225,000. A cross section of a typical installa-
tion is shown in Figure 4. These systems will treat approx-
imately the first 1.3 centimeters of stormwater runoff from
almost 40 hectares or 26 impervious hectares of the Lake
Eola watershed. If an average annual pollutant removal
percentage of 90 percent is assumed, the capital cost of treat-
ment becomes $94/hectare-annual-percent removal which
ro.l.gm DIA.
M.H 1
90amHIQH
DIVERSION
WALL INSIDE,
M.H.^X
7 \
i t .
j
° ^
EXIST. STORMWJ
*
\| I^-EXIST. CURB
"~
BOomDIA/
TER
CULVERT
>« — -~-
^-IBom CONCMTI
7 * ^
\.^/
•¥*.
FILTER FABRIC
WRAP
^ 1 J» DIA. PMF.
^CRUSHED POCK
DOT *11
Figure 4.—Underground percolation system section.
is competitive with current capital costs of diversion/percola-
tion systems. An economic comparison of various stormwater
management alternatives including underground percolation
is listed in Table 7.
Of interest in Table 7 is the exclusion of land costs. When
land costs are added to each management practice, except
filtration which can be constructed under the right-of-way
and would not require land purchases, exfiltration may be
more cost effective. As an example of land cost, a $250,000/
hectare land cost would add approximately $20-25/hectare-
percent removal to the prior estimates. In addition, if
sedimentation were used, only 50 percent maximum efficien-
cy would be achieved.
INACTIVATION OF PHOSPHORUS
RELEASE FROM BOTTOM SEDIMENTS
Treatment will be required in lakes experiencing significant
internal loading of phosphorus. Nutrient inactivation through
chemical precipitants such as aluminum, iron, or calcium has
been reported by many researchers, and several other
treatments, including fly ash, for nutrient inactivation have
been reported in the literature (Funk and Gibbons, 1979).
However, these treatment processes often involve a substan-
tial investment in chemicals, especially if the area to be
treated is large. Using waste products (such as water treat-
ment sludges) that contain a large amount of chemical
precipitants to treat natural systems such as lakes, reten-
Table 7. — Economic comparisons of various stormwater treatment alternatives. (Data is listed in terms of Impervious areas
treated, and land costs are not included; Wanielista, 1979.)
Management practice
Surface pond diversion/
percolationb
Percolation pond0
Swales and percolationd
Residential swalesd
Sedimentation9
Underground exfiltration'
Underground exfiltration'
Underground exfiltration'
Overall (%)
efficiency8
99
99+
92
80
50
80
90
95
ORM
$/ha/month
39.50
86.50
74.00
49.50
71.50
5.00
5.00
5.00
Average capital
cost
($/ha/% removal)
61.80
89.70
70.20
64.40
47.40
74.10
93.90
118.60
a Yearly average of BODS. N, P, and SS not discharged to surface waters
b Designed tor 2.5 cm of runnolt diversion
0 Designed tor 10 cm of runoff diversion
d 80% of the rainwater percolates
8 Designed tor 1.85 cm of runoff water
' Designed tor Type A Sons, 25 cnVhour percolation rate
17
-------
Lake Restoration, Protection and Management
tion or detention ponds, and percolation systems could
become a very attractive alternative. The possibility of us-
ing these waste products to inactivate anaerobic phosphorus
release in Lake Eola was investigated in a series of in situ
experiments.
In situ experiments designed to simulate anaerobic con-
ditions were conducted using isolation chambers constructed
from heavy duty 200-liter polyethylene containers as shown
in Figure 5. All tanks were painted on the outside with a semi-
gloss black alky-enamel to prevent light penetration and
subsequent photosynthetic activity. Chambers were inverted
on the lake bottom, isolating 0.25 square meters of surface
area. Chemicals were added through a 1.9 cm diameter
tygon tube that extended from the top of each tank to the
LAKE EOLA
WATER SURFACE
STVROFOAM
FLOATS "
NALGENE
CHAMBER'
CONCRETE
BLOCK —•
GAS RELEASE
VALVEi
hsctn
15cm
I5cm
LAKE EOLA SEDIMENT
Figure 5.—Typical isolation chamber.
water surface in the lake. The tubing was connected on the
top of the tank to an inverted 28-cm diameter polyethylene
funnel, as shown in Figure 5, to insure even distribution of
the chemical inactivants over the sediment bottom.
Two different water treatment plant sludges were in-
vestigated in Lake Eola for possible inactivation of bottom
sediments. A calcium carbonate-based water treatment plant
sludge was obtained from the Clyde Doyle water treatment
plant in Cocoa, and an alum-based water treatment plant
sludge was obtained from a treatment plant in Tampa,
Florida. The moisture content and percent loss on ignition
averaged 0.4 and 2.9 percent for the Cocoa sludge. Similarly,
Tampa sludge exhibited 89 percent moisture content and
45.8 percent loss on ignition. The calcium sludge showed
very little moisture content since it was air-dried in the
laboratory after collection.
Chemical analysis of the water treatment sludges used
as chemical inactivants is presented in Table 8. The data
shown in this table represent an average of five samples
tested.
Experimental concentrations of inactivants and corre-
spondding isolation chamber designations are listed in Table
9. Dosages were selected to provide a fairly uniform floe.
Water samples were collected in each tank approximately
24 to 48 hours after addition of the inactivant after which
they were then collected at approximately 1-week intervals
for 1 month with samples collected at 2 to 3-week intervals
thereafter.
Table 8. — Chemical analysis of water treatment plant
sludges.
Element
Average concentration tig/g oven dry weight
Cocoa plant
Tampa plant
Cd
Zn
Cu
Fe
Pb
Ni
Cr
Al
Mg
Ca
P
Mn
Ba
1
6
g
256
226
14
29
379
11,400
207,200
233
36
77
1
56
44
8,770
70
10
254
206,700
3,140
10,400
351
193
31
Table 9. — Experimental concentrations of chemical
inactivants used in Lake Eola.
Chemical
inactivant
Inactivant dosage
(wet weight)
(g)
(g/m2)
CaC03 sludge
Alum sludge
None (control)
350
2000
None
1400
8000
The release of phosphorus under anaerobic conditions us-
ing various chemical inactivants is shown in Figure 6. The
results obtained in these experiments indicated that alum
sludge was able to inhibit the release of phosphorus while
calcium carbonate sludge did not when compared with a
control tank that contained no chemical inactivants. Alum
sludge was found to reduce anaerobic phosphorus release
from the bottom sediments of Lake Eola into the overlying
water. The dosage used in these experiments, however, was
200 grams of wet alum sludge which is equivalent to 880
grams of dry sludge per square meter since this sludge con-
tained 80 percent moisture. This dosage was admittedly ex-,
cessive, but was used to illustrate the usefulness of alum
sludge. The data do not suggest the release of heavy metals
from water treatment sludges to the overlying water column.
Jellerson (1981) found no significant difference between ben-
thic populations in alum sludge-treated tanks and control
tanks.
WET ALUM SLUDOC
- WEI A
.V—.A
Incubation Period (dayg)
Figure 6.—Orthophosphorus released from treated Lake Eola bot-
tom sediments with incubation time under anaerobic environment.
18
-------
Restoration and Protection Techniques
To determine an optimum dose required to inhibit
anaerobic sediment release of phosphorus, another series
of experiments was designed using various doses of alum
sludge inside isolation chambers. Six isolation chambers
were dosed with various concentrations of wet sludges as
shown in Table 10.
Table 10. — Experimental concentrations of alum sludge
used in optimization studies.
METRIC TONS/HECTACRE OF BOTTOM SEDIMENT
Inactivant dosage
Chamber
designation
Alum sludge #1
Alum sludge #2
Alum sludge #3
Alum sludge #4
Alum sludge #5
None
(wet
(g)
27.5
55
110
220
440
None
weight)
(g/m2)
110
220
440
880
1760
This experiment was run for 62 days and changes noted
in water quality parameters inside the anaerobic isolation
chambers with various dosages of alum sludge. The data
showed the general trends observed in the previous ex-
periments. pH values and turbidity measurements general-
ly decreased, while specific conductance and ammonia
nitrogen increased with incubation time under anoxic en-
vironments. Phosphorus concentrations increased gradual-
ly in the control chamber and declined in isolation chambers
treated with alum sludge. It was apparent that when used
in large concentrations alum sludge retained the phosphorus
released from the bottom sediments and could also remove
phosphorus associated with the overlying water.
Changes in optimum orthophosphorus released from bot-
tom sediments treated with various concentrations of alum
sludge under an anaerobic environment for 2 months are
presented in Figure 7. The orthophosphorus released was
calculated by multiplying the difference between the initial
phosphorus concentration in the water column beneath the
isolation chamber and the maximum concentration reach-
ed during the study period with the water volume of the
isolated column, and dividing by the bottom sediment area.
It was interesting to see that the smooth curve shown in
Figure 7 intercepted the zero optimum orthophosphorus
released in mg-P/sq m at alum sludge dosages between 110
and 220 grams which were equivalent to 2.0 to 4.0 metric
tons of wet sludge per acre of bottom sediments. Above
these dosages, not only was the release of orthophosphorus
from bottom sediments inhibited, but the orthophosphorus
content in the overlying water column was also reduced.
APPLICATION OF ALUM SLUDGE TO
LAKE EOLA
In view of the impressive phosphorus adsorption and reten-
tion capability of alum sludge it was decided, after much
discussion, to employ this technique to inactivate anaerobic
sediment phosphorus release in Lake Eola. It was estimated
that approximately 8 hectares of Lake Eola would be sub-
ject to anaerobic conditions during the year and would re-
quire treatment. Based on the optimization experiments, a
sludge dose of approximately 7 metric tons per hectare of
a wet sludge with 89 percent moisture, was selected since
this dose was shown to reduce phosphorus release in isola-
tion chambers to virtually zero.
To supply a sufficient amount of energy to return the
sludge into solution so that it could be spread easily, it was
decided that a side-entering large impeller system operated
D-
9
E
O
Ul
in
LJ
rr
in
ID
cr
o
x
D_
in
o
X
a
o
X
t-
a.
o
D-
O
IOOO
600
600
400
ZOO
2.0
4.0
6.0
6.0
-200
110 220
WET ALUM SLUDGE
330 440
DOSAGE (grams)
Figure 7.—Effect of alum sludge on release of orthophosphorus
from Lake Eola sediments under anaerobic conditions.
at several thousand r.p.m. would be necessary. A side-
entering belt-driven agitator with three impellers was ordered
from Process Equipment Corp., of Belding, Michigan. This
impeller was coupled with a 3.5-horsepower, 240-volt,
3-phase electric motor with a 1:1.5 reducer that produced
an agitator speed of approximately 2,500 r.p.m. A sturdy sup-
port bench was constructed to accommodate the 550 liter
rewatering tank and to suppress the expected vibrations. The
bench motor mounts were designed to facilitate the anti-
cipated daily removals for security reasons. A schematic of
the land-based sludge rewatering system is given in Figure 8.
The sludge injection system proved to be the most dif-
ficult to design. The system was designed to be as
mechanically simple as possible to minimize the probability
of field failures. Another 550 liter container (holding tank) was
fastened on a flat deck, 5.8 meter utility boat powered by
a 90-horsepower outboard. The holding tank, mounted on
a wooden platform, would contain the sludge slurry during
spreading operations and was designed to evenly distribute
its weight when full (approximately 550 kilograms). A high
capacity sludge pump powered by a 1,200-watt gasoline
generator on board the boat was used to pump the slurry.
The exit pipe coupled to this pump extended to the front of
the boat and connected to a flexible joint (automatic radiator
hose). A 5 centimeter PVC inverted "T" bar was attached
to the joint. After drilling 0.6 cm holes into the inverted T,
it was used as a spreading mechanism (Fig. 9).
An average 180 kilograms of alum sludge, based on a wet
•weight basis, were placed into the rewatering tank for each
mixture. The team loaded 10 buckets of sludge, each
weighing 18 kilograms, over a 15-minute period. This timed
addition allowed the impeller to break down large clumps
of sludge without excessively straining the motor. The sludge
mixture was agitated for about 20 minutes. The consisten-
cy of the rewatered sludge, when correctly mixed, was com-
19
-------
Lake Restoration, Protection and Management
s DRY SLUDGE INPUT
FLOW VALVE
SLUDGE
SLURRY TO
BOAT
TANK
LAKE BOTTOM
550 LITER SLUDGE
REACTOR
IMPELLER -,
Figure 8.—Schematic diagram of land-based sludge rewatering
system.
parable to that of cooking oil. Barring mechanical failures,
the crew could spread approximately 12 loads of sludge per
day. At this rate, which corresponds to 2.2 metric tons of
sludge per day, it required approximately 25 days to cover
the 8 hectares of Lake Eola that were treated.
FILTRATION OF STORMWATER THROUGH
ALUM SLUDGE
Of the total 65 hectares in the Lake Eola watershed, approx-
imately 25 hectares are located in low areas near the lake
in which the water table is too high to allow for efficient per-
colation by underground exfiltration. Since treatment of
stormwater runoff generated in these areas was thought to
be an essential portion of the overall restoration process, in-
vestigations were conducted to develop treatment techniques
for these areas.
In view of the impressive phosphorus adsorption
characteristics of alum sludge when used as a sediment in-
activant the possibility of using finely ground dried alum
sludge as a filter media was investigated. Initial laboratory
column studies were conducted in which various mixtures
of alum sludge along with sand and gravel were tested for
filtration rates and pollutant removal efficiencies. The objec-
tive of these investigations was to optimize filtration rates
without sacrificing efficiency.
Early experiments indicated that crushed alum sludge itself
could not be used as a filtration media without mixing with
sand or gravel. Once the sludge became wet, it tended to
form into a dense cake, especially if compacted, and filtra-
tion would stop. After much investigation, the optimum solu-
tion seemed to be a 50-50 mixture of sludge and coarse
building or silica sand. Filtration rates ranging between 20
and 200 centimeters per hour were obtained depending upon
compaction of the media, sludge moisture content at time
of crushing, and particle size. It was also determined that
the best results were achieved when the alum sludge was
crushed into particles approximately 1 millimeter in diameter
and mixed with sand just before it completely dried and when
compaction of the media was held to a minimum.
Laboratory column studies were conducted to investigate
the adsorption characteristics of this media for phosphor-
us and heavy metals in stormwater runoff. The results of this
investigation are listed in Table 11. Although it was not possi-
ble to determine adsorption capacities for each of the heavy
metals tested, it was determined that the maximum capaci-
ty for phosphorus was approximately 140 gP/m3 of sludge-
sand mixture.
AGITATORS
CONTROL
VALVE PUMP_
$~
SECONDARY REACTOR
w.s.
Figure 9.—Schematic diagram of alum sludge spreading operation.
20
-------
Restoration and Protection Techniques
Table 11. — Removal percentages of phosphorus and heavy
metals In stormwater runoff filtered through 20 cm of a
50-50 mixture of alum sludge and sand
(Filtration rate = 25 cm/hour)
Parameter
Orthophosphorus
Total P
Total Cadmium
Total Zinc
Total Copper
Total Iron
Total Lead
Total Nickel
Total Chromium
Average
influent
(mg/liter)
791
993
258
695
596
708
880
323
83
Average
effluent
(Mg/liter)
90
117
5
15
47
28
88
39
15
Percent
removal
(%)
81
88
98
98
92
96
90
88
82
In view of the laboratory results obtained, it was decided
that this media would be used in an in-line filter on two large
storm sewer lines before discharge into Lake Eola. One of
the storm sewers was already equipped with an underground
concrete settling basin which was modified to act as a
downflow filter with underdrains. Stormwater runoff is col-
lected in the basin, filtered through the sand and sludge mix-
ture, collected in the underdrains, and then discharged into
the lake. A diagram of this structure is shown in Figure 10.
The other alum sludge filtration system was constructed
above ground in a 1.5 meter deep excavation approximate-
ly 20 meters in diameter. This unit was also designed to act
as a downflow filter with underdrains leading to the lake. The
sludge-sand mixture was covered with decorative rock, and
the sides of the excavation were sloped and landscaped to
resemble a Japanese garden. During a rain event, storm-
water first enters an entrance structure which diverts the
water onto the top of the rock layer, flooding the excavated
basin. The stormwater then percolates through the rocks and
media and into the underdrains. A cross-section of this unit
is shown in Figure 11. Chemical analysis of the influent and
effluent from an actual storm event is listed in Table 12.
SERVICE
ACCESS)
SUMMARY AND CONCLUSIONS
During this project, periodic water quality analyses were con-
ducted in Lake Eola to determine the effects of stormwater
runoff on water quality and to establish pre-restoration
background water quality. The feasibility of using water treat-
ment sludges to inactivate anaerobic phosphorus release
from bottom sediments and infiltration of stormwater was
studied. From the information obtained in this research, the
following conclusions were reached:
1. Stormwater runoff was determined to be the primary
source of pollution in Lake Eola. Continual stormwater in-
puts have degraded water quality as typified by high algal
production, a stratified hypolimnion, and periodic fish and
duck kills. Bottom sediments have become covered with a
layer of loose flocculant material, and anoxic conditions ex-
ist in areas more than 4 meters deep during the spring and
summer. As a result, a restoration project was begun to
restore Lake Eola.
Table 12. — Removal percentages of selected parameters
in stormwater runoff passing through the garden
alum sludge filter.
Parameter Average Average Percent
influent effluent removed (%)
PH
Turbidity (JTU)
Spec. cond. (^mhos)
Alkalinity (mg/liter)
Organic carbon (mg/l)
NH3-N fcg/l)
NO2-N ^g/l)
NO3-N (M9/I)
Organic N (/jg/l)
Diss. Orthophosphorus (ng/l)
Total P (^g/l)
S.S. (mg/l)
V.S.S. (mg/l)
6.20
3.0
208
95.0
2.7
87
3.3
204
563
113
382
128
61.0
6.10
1.5
200
99.0
1.1
129
3.8
212
184
12
68
28.1
14.4
1.6
50.0
3.8
4.2*
59.3*
48.3*
15.2*
3.9*
67.3*
89.4
82.2
78.0
76.4
' Denotes increase
SAND + ALUM SLUDGE
smPERFORATED ALUM. PIPE
EXIT
30cm 30cm 20cm
Figure 10.—Underground alum sludge filtration system section.
21
-------
Lake Restoration, Protection and Management
t~- OU-P°-"°eBP — " « -a o n—j'-n-Tv^-ai-Tnr"———-~— -iriiinir - . . — — . . __.
UNDERGROUND FILTER (SAND. GRAVEL. ALUM. SLUDGE)
MINIMUM
WATER
TABLE
5:16cm PERFORATED PIPE
EACH ABOUT 12m LONG IMPERMEABLE MEMBRANE
Figure 11 .—Garden walk filter - section.
2. Analysis of the contributing watershed indicated that
a first-flush effect was present with approximately 90 per-
cent of the stormwater pollutant mass contained in the first
1.3 centimeters of rainfall excess. Since an area was not
available within the highly urbanized watershed for conven-
tional stormwater management practices, an underground
exfiltration system was developed. Stormwater practices
were constructed and lake modifications were made. A
preliminary estimate of the efficiency of pollutant mass
removal was determined for this system.
3. Stormwater management by diversion to underground
exfiltration systems is an alternative to traditional surface
diversion/retention structures. Pollutant removal appears to
be possible in areas where a first-flush effect exists by selec-
ting the proper exfiltration volumes. The most cost-effective
method of treatment by underground exfiltration was to in-
tercept the storm sewer line directly before it discharged in
to the lake.
4. Alum sludge was shown to be effective in the inactiva-
tion of anaerobic release of phosphorus from bottom
sediments. The orthophosphorus and total phosphorus
released from bottom sediments to the contained lake water
column inside the isolation chambers treated with various
dosages of alum sludge decreased with increasing sludge
dosage. In some cases, the sludge inhibited the release of
phosphorus from the bottom sediments and also reduced
the phosphorus content in the overlying water column. The
control isolation chambers showed a gradual increase in or-
thophosphorus concentrations reaching maximum values
within 2 to 3 months of incubation time. The maximum or-
thophosphorus released was 136 mg P/m2 for the 3-month
period or 1.5 mg P/m2/day. Generally, all phosphorus re-
leased appeared to be in the orthophosphorus form. Based
on these findings, alum sludge at a dose of 7 metric tons/hec-
tare was applied to Lake Eola sediments.
5. Alum sludge was found to be effective as a filtration
media for stormwater runoff. Removal percentages between
80 and 99 percent were obtained in laboratory column
studies for various heavy metals and phosphorus. Two in-
line stormwater filters were constructed and removal efficien-
cies monitored.
6. Monitoring of lake water quality will continue to deter-
mine the effects of the various pollution removal systems.
REFERENCES
Funk, W.H., and H.L. Gibbons. 1979. Lake restoration by nutrient
inactivation. Pages 141-151 in Proc. Natl. Conf. Lake Restoration.
Aug. 22-24, 1978, Minneapolis, Minn. EPA 440/5-79-001. U.S. En-
viron. Prot. Agency, Washington, D.C.
Harper, H.H., Y.A. Yousef, and M.P. Wanielista. 1979. Produc-
tivity responses of Lake Eola water to urban runoff. Proc. Natl.
Conf. Urban Stormwater Combined Sewer Overflow Impacts on
Receiving Water Bodies. Orlando, Fla. Nov. 26-28.
Jellerson, D.B. 1981. Impacts of alum sludge on lake sediment
phosphorus release and benthic communities. Master's Thesis.
Univ. Central Florida, Orlando.
Marshall, F.E. 1980. Phosphorus dynamics of Lake Eola sediments.
Master's Thesis. Univ. Central Florida, Orlando.
Wanielista, M.P. 1977. Manual of stormwater management practices.
Draft rep. submitted to East Central Florida. Reg. Plann. Counc.
Winter Park.
, 1979. Stormwater Management: Quantity and Quality.
Ann Arbor Science Publ. Ann Arbor, Mich.
Wanielista, M.P., Y.A. Yousef, and J.S. Taylor. 1981. Stormwater
Management to Improve Lake Water Quality. Final rep. EPA Grant
R-8055800. Univ. Central Florida, Orlando.
22
-------
NUTRIENT REMOVAL FROM URBAN STORMWATER BY WETLAND
FILTRATION; THE CLEAR LAKE RESTORATION PROJECT
JOHN BARTEN
Limnologist
Waseca, Minnesota
ABSTRACT
Clear Lake is a 257-ha body of water located in southcentral Minnesota, It is a heavily used recreational
lake that has become severely eutrophic because of the inflow of nutrient-rich urban runoff from the adja-
cent city of Waseca, Minnesota. In 1981,50 percent of the hydraulic load and 55 percent of the phosphorus
load to the lake was diverted into a 21.4-ha peat marsh on the northwest corner of Clear Lake. A series
of ditches and dikes was constructed in the marsh to retain stormwater until the phosphorus could be
removed by percolation through the peat. The filtered water was then pumped into Clear Lake. In 1981,
73.3 x ICH ma of water were filtered through the system and 258.6 kg of phosphorus removed. In 1982,
89.6 x 10* m3 of water were filtered and 526.7 kg of phosphorus removed. The total quantity of phosphorus
removed in 1982 amounts to 40 percent of the average annual load to Clear Lake. Mean orthophosphorus,
total phosphorus, and chlorophyll a concentrations in Clear Lake decreased significantly following diver-
sion of stormwater to the marsh. The reduction in the frequency and Intensity of algal blooms in Clear
Lake have increased the recreational use of the lake. Local residents are very optimistic about the future
of Clear Lake.
INTRODUCTION
Stormwater runoff in urban areas is believed to be a signifi-
cant source of nutrients to adjacent streams and lakes.
Wanlelista and Yousef (1980) have shown that urban runoff
may be the major cause of accelerating eutrophication in
urban lakes. Removal of stormwater nutrient toads before
contact with the receiving water has become a major con-
cern of urban lake managers. Because wetlands have long
been regarded as sinks for nutrients entering them, atten-
tion has recently focused on the use of wetlands as treat-
ment areas for urban runoff. Wenck (1981) found that 77
percent of the influent phosphorus load in urban ruiioff could
be retained in a wetland. Wile, Palmateer, and Miller (1981)
found that 75 percent of influent phosphorus and 80 per-
cent of influent nitrogen could be removed from secondary
sewage effluent during the growing season by an artificially
constructed wetland.
On December 23,1980, Waseca, Minnesota, completed
a stormwater diversion and treatment system using artificially
constructed basins in an existing natural wetland to filter
nutrient-rich urban runoff water prior to its release to Clear
Lake. The project was the result of 30 years of decreasing
water quality in Clear Lake which reduced the recreational
use of the lake and threatened the tourist industry depen-
dent on it.
Located In a population center in a highly agricultural area,
Clear Lake is the focal point for the recreational activities
of the surrounding area, supporting multiple uses including
fishing, swimming, water skiing, and sailing. A privately own-
ed campground and a boathouse-restaurant are located on
the lake shoreline along with four city parks. Approximately
80 percent of the shoreline is urban developed.
In 1963, the lake was treated with toxaphene to remove
large rough fish populations and improve the lake water
quality. The effect of the treatment was shortlived, and by
the late 1960's recurring algae blooms again reduced the
recreational potential of Clear Lake and prompted the
residents to seek a long-term solution to their problem.
National Biocentric Inc. (1974) determined that urban
runoff from Waseca provided 77 percent of the phosphorus,
80 percent of the nitrogen, and 76 percent of the hydrologie
loading to the lake. The report also concluded that high
phosphorus concentrations in Clear Lake were responsible
for the frequent algae blooms.
Based on this data, Waseca requested, and in 1976 was
granted, SO percent funding by the U.S. Environmental Pro-
tection Agency to construct a filtration system for stormwater
entering Clear Lake. The Minnesota Pollution Control Agency
provided 25 percent funding for the project and the city of
Waseca provided the remainder.
This paper presents the results of the project through
August 1982.
METHODS
Project Description
Clear Lake is a 257-ha body of water located in southcen-
tral Minnesota, adjacent to and within the corporate limits
of the city of Waseca (Fig. 1). The hydrologic and mor-
phologic characteristics of the lake are shown in Table 1.
Waseca constitutes only 27 percent of the 1,518-ha water-
Table 1. — Hydrologic and morphologic characteristics of
Clear Lake,
Surface area (ha)
Maximum depth (m)
Mean depth (m)
Length (m)
Width (m)
Volume (m3)
Outflow volume (m3)*
Hydraulic retention time (yr)
Drainage area (ha)
Urban developed
Agricultural
Other
257
9.9
4.1
1920
884
10.0 x 1Q6
3.25 x 1Q8
3.08
1518
337
405
776
" Value determined tan 1979 flow volume normalized to average pmdpteilbn.
23
-------
Lake Restoration, Protection and Management
— v-pr: J( Clear Lake
iri::^..
Kl
29 m
^_~^
Figure 1 .—Clear Lake watershed boundary and location map.
shed but provides most of the hydrologic and nutrient loading
to the lake. A 1.47 meter storm sewer discharging into the
east side of Clear Lake transports 50 percent of the average
annual water volume, 55 percent of the annual phosphorus
load, and 45 percent of the annual nitrogen load to the lake
(Natl. Biocentric Inc., 1974; Environ. Res. Group, 1981). This
storm sewer also carries the overflow from Loon Lake.
As designed, the Clear Lake Restoration Project diverted
urban runoff in the 1.47-m storm sewer from Clear Lake to
a 21.4-ha marsh on the northwest corner of the lake (Fig.
1). The marsh would also receive runoff water from a 1.07-m
storm sewer draining the northeast section of the city. The
marsh soils are composed of Fibrist peat underlain by clay
loam soils (Biggar, unpub. rep.).
A major portion of the stormwater nutrient and suspend-
ed solids load was to be removed by percolating the water
through the peat. To accomplish this, the marsh was divid-
ed into five basins, called cells, by ditches, dikes, and natural
elevation (Fig. 2). Water flow to each cell could be con-
trolled by opening and closing the five gates along the
distribution ditch. The dikes surrounding each cell would pre-
vent the stormwater from flowing overland into the collec-
tion ditches. Two 3,000 GPM (11.4 m3/min) lift pumps would
pump the filtered water to Clear Lake.
The harvesting of Phalaris arundinacea three times a year
was to remove nutrients from the marsh and maintain the
phosphorus adsorption potential of the peat. Phalaris arun-
dinacea was seeded because it has the ability to withstand
alternating dry and wet conditions and has a market value
as cattle fodder.
CLEAR LAKE TREATMENT MARSH
SCALE 3 cm -loom
Figure 2.—Clear Lake treatment marsh.
An overflow structure on the north end of the marsh
prevented flooding of adjacent residential property and allow-
ed water to be diverted through the marsh during the winter
months.
Sampling Procedure
Marsh influent samples were collected at 3-week intervals
and during rainfall at the discharge end of the 1.47-meter
24
-------
Restoration and Protection Techniques
storm sewer with an ISCO Model 1680 automatic sampler.
Dry weather samples were collected at 1-hour Intervals for
24 hours and composited. Stormwater runoff samples were
collected for six rainstorm events in 1982 at 15-minute in-
tervals from the start until 3 hours after the rainfall ended.
The samples were composited relative to the runoff
hydrograph. Only grab samples of rainfall events were col-
lected in 1981,
Influent discharge was calculated from design discharge
curves for round concrete pipes and continuous stage
records measured with a Badger Series 200 Ultrasonic
Height Gauge and a Model 68 Leopold and Stevens
Recorder for the 1.47-m and the 1.07-m storm sewer,
respectively.
Marsh effluent samples were collected 75 meters
upstream from the pumping station at 3-week intervals and
when water was released from the marsh cells. Samples
were collected at 1-hour intervals for 24 to 72 hours and com-
posited. Effluent discharge was calculated from the hours
of operation of each pump. Since the pumps lack backflow
check valves, they were assumed to be only 90 percent
efficient.
Clear Lake water samples were collected biweekly dur-
ing the spring overturn period, at 3-week intervals during the
growing season, and monthly during the winter. Epilimnetic
(surface to 1.84 meters) composite samples were collected
at three sites on the lake (Fig. 1). At site 1, the water
temperature, dissolved oxygen, and conductivity were
measured at 1.5-meter intervals from the surface to the lake
bottom. Water samples were also collected at these stations
and analyzed for their total phosphorus concentration.
The lake, marsh influent, and marsh effluent composite
samples were analyzed for the following parameters:
Total phosphorus
Dissolved phosphorus
Orthophosphorus
Ammonia nitrogen
Nitrate & nitrite nitrogen
Total Kjeldahl nitrogen
pH
In addition, the chlorophyll a concentration and alkalinity
of the lake samples and the suspended solids concentra-
tion of the marsh samples were measured. Chemical analysis
procedures follow Standard Methods (1975).
The lake sampling procedure was similar to that outlined
by National Biocentric Inc. (1974, 1978, 1979) and En-
vironmental Research Group (1980, 1981 b) and allows a
good pre- versus post-stormwater diversion comparison of
Clear Lake water quality.
RESULTS
Treatment Marsh
The diversion of stormwater into the treatment marsh began
June 26, 1981. The slow germination and growth of the
Phalaris a/wjcffnacea in the spring of 1981 delayed the diver-
sion. Only "first flush" runoff water was channeled into the
marsh in July and early August 1981, because one of the
discharge pumps was broken.
The filtration system significantly reduced the nutrient con-
centrations of the influent water except for ammonia (Table
2). The marsh had the greatest impact on the suspended
solids and the total phosphorus concentrations and the least
effect on the TKN concentration. Total phosphorus and
suspended solids removal rate in 1982, 70 percent and 90
percent respectively, were higher than in 1981, 51 percent
and 35 percent respectively. The Total Kjeldahl nitrogen
removal rate, 13 percent in 1981, did not change significantly
in 1982. The ammonia concentration increased an average
of 420 percent in 1981 and 1982.
In 1981, 73.3 x to4 m3 of water were filtered through the
system and 258.6 kilograms of influent phosphorus re-
moved (Table 3). In 1982,89.6 x 104 m3 of water were filtered
and 526.7 kilograms of phosphorus removed. The higher in-
fluent phosphorus concentration and greater flow volume In
1982 contributed to the 200 percent increase in kilograms
of phosphorus remove d with a 27 percent increase in
removal efficiency.
Two modifications in the treatment marsh during the winter
of 1981-82 increased nutrient removal in 1982. A 200 GPM
(0.76 m3/min.) lift pump was installed in the marsh in
December 1981. The pump, by lowering the collection ditch
water level 20 cm, increased the hydraulic gradient in the
system thus increasing the percolation rate of water Into the
collection ditches. The percolation rate in 1981 was 0.5 cm
water/unit surface area/day, 13 percent of the design rate
of 3.7 cm water/unit surface area/day. This low percolation
rate resulted in a buildup of water in the marsh during dry
weather conditions and overtopping of the dikes during rain-
fall events. (It should be noted here that dry weather inflows
of approximately 3.79 x 10s m3/day were caused by the
Table 2. — Mean and standard deviation for treatment marsh influent and effluent water parameters for 1981 and 1982
(In mg/1). The number in parenthesis is the standard deviation.
Year
1981
Parameter
Total phosphorus
Dissolved phosphorus
Orthophosphorus
Total Kjeldahl nitrogen
Nitrate + nitrite nitrogen
Ammonia nitrogen
Suspended solids
Influent
0.588 (0.2)
0.312 (0.2)
0.229 (0.1)
2.742 (1.7)
0.356 (0.4)
0.230 (0.2)
46.41 (31)
Effluent
0.307 (0.1)
0.152 (0.1)
0.137 (0.1)
2.390 (1.2)
0.176 (0.1)
1.061 (0.5)
30.33 (8.5)
% reduction
from inlet
51
51
40
13
51
-361
35
1982
Total phosphorus
Dissolved phosphorus
Orthophosphorus
Total KJeldaht nitrogen
Nitrate + nitrite nitrogen
Ammonia nitrogen
Suspended solids
0.763 (0.7)
0.173 (0.1)
0.162 (0.1)
2.510 (1.3)
0.455 (0.3)
0.217 (0.2)
43.79 (58)
0.242 (0.1)
0.121 (0.1)
0.071 (0.1)
2.163 (0.9)
0.227 (0.2)
1.273 (1.1)
4.522 (3.1)
70
30
56
14
50
-487
90
25
-------
Lake Restoration, Protection and Management
Table 3. — Effect of the treatment marsh on phosphorus quantities in runoff water in 1981 and 1982.
Year
Month
Water volume (m4 x 104)
Influent1 Effluent
Influent
Total phosphorus (Kg) Percent
Effluent Retained removed
1981
June
July2
August2
September
October3
November3
December3
Total
4.75
30.39
26.80
16.45
8.08
0.0
0.0
86.47
6.14
13.15
18.77
17.36
12.74
2.78
2.32
73.26
26.5
142.7
140.8
96.7
47.5
0.0
0.0
454.2
18.8
40.4
57.6
53.3
39.1
8.5
7.1
224.8
7.7
102.3
83.2
43.4
4.1
0.0
0.0
240.7
52
1982
April
May
June
July
August
Total
19.55
21.95
15.02
17.21
23.88
97.61
11.56
15.73
13.22
18.39
30.71
89.61
151.4
225.2
107.8
79.1
181.0
744.5
23.1
61.6
36.8
36.0
60.3
217.8
128.3
163.6
71.0
43.1
120.7
526.7
70
' Influent volume includes precipitation on maish.
2 An unknown volume of influent water discharged to Rica Lake when the overflow structure
at the north end of the marsh washed out.
3 Stormsewer water was discharged to Clear Lake.
discharge of cooling water into Loon Lake by a local
vegetable processing plant). The stormwater had a residence
time of less than 24 hours in the marsh. In 1982, the per-
colation rate increased to 1.5 cm/unit surface area/day, 40
percent of the design rate.
The installation of sluice gates in the dikes in cells 1, 3,
and 5 in February 1982, allowed the marsh cells to be par-
tially drained between rainstorms, thus reducing the over-
topping of the dikes (Fig. 2). When possible, the stormwater
was detained for 5 days before the sluice gates were open-
ed and water released. The combination of the sluice gates
and the improved percolation rate increased the stormwater
detention time to 48 hours in 1982 and increased the removal
of suspended solids. The difference between the total
phosphorus and dissolved phosphorus concentrations (Table
2), shows that 75 percent of influent phosphorus is in the
particulate form and is removed by sedimentation. Overtop-
ping of the dikes continued to occur with rainfall events at
0.75 cm/hour intensity or greater in 1982. Figure 3 shows
that most rainstorms were followed immediately by high ef-
fluent discharge.
The watershed received 57.7 cm of rainfall in the sum-
mer and fall of 1981 and 57.5 cm in April through August
1982. More water was treated in 1982 because all runoff was
diverted to the marsh and less water overtopped the overflow
structure and discharged to Rice Lake. The loss of unknown
quantities of water to Rice Lake prevented the estimation
of the evapotranspiration which occurred in the marsh. The
snowmelt and runoff water in March of 1982 were diverted
through the marsh and discharged to Rice Lake because
the frozen soil was unable to filter the water.
The bulk of the phosphorus loading to the marsh oc-
curred during rainfall (Fig. 3). Total phosphorus concentra-
tions were one to two orders of magnitude higher than in
dry weather.
The Philaris arundinacea which was seeded in the fall of
1980 has been overgrown by cattails (Typha spp) in most
of the marsh. In other areas, especially cell 5, large areas
of open water exist with no emergent vegetation. These
areas supported profuse growths of Spirogira in the spring
of 1982 and duckweed (Lemna spp) and blue-green algae
in the summer. Marsh vegetation was not harvested in 1981
or 1982 because of high water levels in the marsh cells.
Clear Lake Water Quality
The 1981 and 1982 lake water quality data show that the
treatment marsh has reduced nutrient concentrations in Clear
iOO
zoo
IOO
o
IS
Ilk
-i IT, rv.
1
1
A M J J A
E*
A M
KEY: INFLUENT
EFFLUENT
Figure 3.—Treatment marsh influent and effluent daily volume and
total phosphorus quantity and monthly total phosphorus quantity for
1982.
26
-------
Restoration and Protection Techniques
Lake. Confidence intervals determined at the (P > 0.05)
significant level using the Students-t test for 1977 to 1980
mean concentrations of selected parameters show a signifi-
cant decrease in the 1981 and 1982 orthophosphorus and
total phosphorus concentrations and the 1981 chlorophyll
a concentration (Fig. 4, 5, and 6). The chlorophyll a concen-
tration increased significantly in 1982. The Total Kjeldahl
nitrogen concentration decreased in 1981 and increased in
1982, but not significantly for most of the growing season
(Rg. 7). The TN:TP ratio increased from 10:1 in 1978 to 1980
to 16:1 in 1981 and 18:1 in 1982 (Fig. 9). There was no signifi-
cant change in the Secchi disc depth in either year (Fig. 8).
Actual values for the parameters measured have been
previously published in National Biocentric Inc. (1974,1978,
1979), Environmental Research Group (1980, 1981b) and
Barten (1982).
Thermal stratification in Clear Lake occurred from early
June until late July 1981. The thermocline formed at a depth
of 2.4 meters and covered 70 percent of the lake. Long-term
stratification of the lake is a phenomenon not observed prior
to 1981. Stratification in 1982 reverted to the normal pat-
tern, becoming established at a depth of 6.1 meters from
late June to early July and then again from late July to early
August. Anaerobic conditions prevailed in the hypolimnion
within a week of stratification, and phosphorus concentra-
tions became two to three times epilimnetic concentrations.
Extremely low chlorophyll a concentrations were measured
in Clear Lake in January through March 1982 as a result
of heavy snowcover that blanketed the lake and prevented
sunlight penetration (Fig. 6). No zooplankton were ob-
served in the lake during this period. The lake became anoxic
below 4.6 meters from February 1982, until ice out in March.
Surface dissolved oxygen concentrations decreased to 6.2
mg/l during this period.
No blue-green algal blooms occurred in Clear Lake in 1981
until August 20, approximately 20 days later than the for-
mation of blooms in 1979 and 1980. Blue-green algae
became dominant in the lake in early June 1982, but blooms
formed only intermittently throughout the summer and did
not last longer than 2 days. Surface algal scums were visi-
ble for only 6 days in 1982.
UJ
5
i
a.
i
-------
Lake Restoration, Protection and Management
DISCUSSION
Treatment Marsh
The treatment marsh system was less efficient than an-
ticipated in removing nutrients in the 2 years of operation.
A pilot plant constructed in the marsh in 1977 indicated that
phosphorus removal rates greater than 80 percent could be
expected (Nat. Biocentric Inc., unpubl. rep.). The observed
removal rates of 51 percent and 70 percent in 1981 and
1982, respectively, resulted from the short detention time of
stormwater in the marsh caused by the slow percolation rate.
Ahern, Stanforth, and Armstrong (1980) found that a 5-day
detention time is required to remove the majority of par-
ticulate phosphorus in runoff water. Since the bulk of
phosphorus loading to the marsh occurs during rainstorms
and 75 percent of influent phosphorus is in the particulate
form, increasing the detention time of rainfall runoff is
necessary to achieve predicted removal rates.
The sluice gates, although partially successful in increas-
ing the detention time, were installed at too high an eleva-
tion to completely drajn the marsh cells, and the storage
capacity of the system was still not adequate to contain most
rainstorm runoff. During winter 1982-83, five 30-cm cor-
regated metal pipes will be installed in the marsh, one in
each cell, with a flow line elevation 0.4 meters below the
sluice gate flow lines. The pipes will be equipped with gates
to hold or release water as needed and should allow for com-
plete drainage of the marsh.
Phosphorus removal rates increased to 88 percent of in-
fluent concentration during dry periods when the detention
time increased to 5 to 7 days. Phosphorus concentrations
decreased to less than 0.1 mg/1 when profuse growths of
Spirogira were present in the water. Increased detention time
should allow more biological utilization of available
phosphorus and increase the nutrient removal capability of
the system.
Hill (1979), Barko and Smart (1980), and Perry, Armstrong,
and Huff (1981) have found that nutrients accumulated by
-S I977-I980 SECCHI DISK DEPTH
S * I98I SECCHI DISK DEPTH
S S I982 SECCHI DISK DEPTH
Figure 8.—Mean Secchi disc depths in Clear Lake in 1982,1981,
and 1977 to 1980. Vertical lines represent 95 percent confidence
intervals for the means.
194
I7O-
IfiO-
iao-
WO-
LJO-
1-20
LIO-
\ / U'TX//
K K 1977- I960 TOTAL KJELDAHL NITROGEN CONCENTRATION
K K I9JI TOTAL KJELDAHL NITROGEN CONCENTRATION
K K 662 TOTAL KJELDAHL NITROGEN CONCENTRATION
Rgure 7.—Mean TKN concentrations in Clear Lake in 1982,1981,
and 1977 to 1980. Vertical lines represent 95 percent confidence
intervals for the means.
IW-
I7M-
!*»•
O 12-1 •
< 11:1
-------
Restoration and Protection Techniques
aquatic macrophytes during growth periods can be remobiliz-
ed during tissue decay. This phenomenon occurred in mid-
September in the treatment marsh. The phosphorus con-
centration in the effluent water increased to 0.4 mg/1 follow-
ing a light frost. Pumping to Clear Lake was stopped and
the effluent water was discharged to Rice Lake.
Because of the slow percolation rate and the lack of
separation between the marsh cells, constant standing water
throughout the marsh prevented harvesting of the vegeta-
tion in 1981 and 1982. This is of some concern because the
phosphorus assimilation potential of natural aquatic systems
has been found to be limited (Kadlec, 1981). Harvesting the
vegetation is necessary to ensure long-term phosphorus ad-
sorption by the peat. In July 1982, fill was placed between
the distribution ditch and No. 3 collection ditch dike to
separate the marsh into two units. By closing the proper
gates, the two units can be successively drained and pro-
bably harvested.
The development of Typha spp as the dominant
macrophyte was caused in part by the continual flooding of
the marsh during the entire growing seasons in 1981 and
1982. Phalaris arundinacea cannot withstand long periods
of flooding, a condition that is ideal for Typha. In addition,
Bevis and Kadlek (1979) reported that Typha spp can use
high nutrient concentrations more effectively than other
aquatic macrophytes.
Despite the problems experienced with the operation of
the system, the marsh in April through August 1982, removed
22 percent of the calculated annual phosphorus load of 2,439
kilograms (Environ. Res. Group 1981 a). Adjusting the
removal rate for the March to December runoff season shows
that the marsh is removing 40 percent of the total phosphorus
load to Clear Lake.
Clear Lake Water Quality
The water quality of Clear Lake improved significantly in 1981
and 1982 compared to previous years. Mean total
phosphorus concentrations declined 50 and 30 percent in
1981 and 1982, respectively, during most of the growing
season. Orthophosphorus concentrations decreased to
undetectable levels throughout the summer of both years.
The increase in the total phosphorus concentration in Clear
Lake in 1982 over 1981 levels is probably due to internal
loading. Thermal stratification resulting in mobilization of sedi-
ment phosphorus occurred only once in 1981 but at least
twice in 1982. The anoxic conditions in the lake during the
winter of 1982 also resulted in recycling of sediment
phosphorus. The Total Kjeldahl Nitrogen concentration
decreased in 1981 and increased in 1982 but the change
was not significant in either year.
The TN:TP ratio increased 36 and 45 percent in 1981 and
1982 respectively. Previous studies indicated that Clear Lake
oscillated between phosphorus- and nitrogen-limiting con-
ditions when the TP:TN ratio was less than 15:1 (Environ.
Res. Group, 1980). Schindler (1978) indicated that when
TN:TP ratios were above 10:1, phosphorus was the more
limiting nutrient. The low Orthophosphorus concentrations in
1981 and 1982 coinciding with increased TN:TP ratios in-
dicate that Clear Lake has become phosphorus limiting.
Mean monthly chlorophyll a concentrations decreased 50
percent in 1981, but increased significantly in 1982. The high
1982 concentrations occurring despite significantly lower total
phosphorus and Orthophosphorus concentrations are be-
lieved to result from two factors. The absence of zooplankton
in January to March 1982 lake samples suggest that initial
grazing of the phytoplankton population, which increased
dramatically after ice-out on April 3, was limited because of
the small numbers of zooplankton involved. Zooplankton
populations increased until late May of 1982 when 810
organisms/I were counted. The population apparently
crashed in early June and only 6 organisms/I were ob-
served. The lack of significant grazing on the phytoplankton
community by reduced zooplankton populations may be par-
tially responsible for high chlorophyll a concentrations in
1982.
The distribution of phytoplankton in the water column may
be the other factor influencing the observed chlorophyll a
levels. In previous years, high blue-green algae concentra-
tions were characterized by floating scums of algae on the
lake surface with lower concentrations in the underlying
water. Compositing a 1.84-meter epilimnetic sample would
dilute the effect of the surface scum. In 1982 the
phytoplankton were distributed fairly evenly throughout the
photic zone, resulting in a larger number of algae being col-
lected in the epilimnetic sample even though the mean
biomass may have been less than in 1977 to 1980. Surface
scums were present for only 6 days in 1982. The observed
decrease in the 1977-1980 and 1982 chlorophyll a concen-
tration in August, when algae blooms are the most frequent
and intense, support this contention.
Although the Secchi disc depth did not change significantly
in 1981 or 1982, the reduction in algae blooms was inter-
preted by most residents as an indication of water quality
improvement. Algal booms in 1981 did not form until after
the swimming season had ended. Blooms in 1982 were in-
frequent and of short duration and affected the swimming
beach for only 3 days of the season. The use of the swim-
ming beach as well as the rest of the lake increased con-
siderably in 1981 and 1982.
The city of Waseca is currently developing and implemen-
ting plans to further decrease phosphorus loading to Clear
Lake. Modification of Gaiter Marsh, which receives runoff
water from the southeast sector of the city, will allow water
level management in the same manner as the treatment
marsh and should reduce nutrient concentrations in the ef-
fluent water. Gaiter Marsh currently contributes the largest
phosphorus load to the lake.
A recently formed Clear Lake Restoration Advisory Com-
mittee is educating residents on the proper use of lawn fer-
tilizers. The committee has also begun an anti-litter campaign
on Clear Lake to reduce the amount of garbage that winter
fishermen leave on the ice. The Waseca City Council along
with the Lake Restoration Committee is developing a lake
shoreline improvement plan designed to reduce shoreline
erosion and improve the aesthetics of the lake. The com-
munity support for and participation in efforts to improve the
lake are evidence of the optimism most city residents have
regarding the future of Clear Lake.
REFERENCES
Ahern, J., R. Stanforth, and E. E. Armstrong. 1980. Phosphorus con-
trol in urban runoff by sedimentation. Pages 1012-1021 in Proc.
Symp. Surface Water Impoundments. Am. Soc. Civil Eng., New
York.
Barko, J. W. and R. M. Smart. 1980. Mobilization of sediment
phosphorus by submersed freshwater macrophytes. Freshw. Biol.
10:227.
Barten, J. M. 1982. 1981 Clear Lake water quality and treatment
marsh assessment. Final rep. for Clear Lake restoration proj. EPA
grant no. S804691-01-0.
Bevis, F. B., and R. H. Kadlec. 1979. Effect of long-term discharge
of wastewater on a northern Michigan wetland. In Wetlands utiliza-
tion for management of community wastewater, J. C. Sutherland
and R. H. Kadlec, coordinators. Abst. Conf. held 10-12 July 1979
at Higgins Lake, Mich.
Biggar, Charles. Unpubl. rep. 1978. Agricultural specialist
Waseca, Minn.
29
-------
Lake Restoration, Protection and Management
Environmental Research Group, Inc. 1980.1979 lake water quality
and 1979 sewer sampling program. Rep. prep, for City of Waseca,
Minnesota; Clear Lake restoration proj. under EPA grant no.
S804691-01-0.
1981a. Hydrologic budget and total phosphorus loadings
for Clear Lake near Waseca, Minnesota. Rep. prep, for City of
Waseca, Minnesota; Clear Lake restoration proj. under EPA grant
no. S804691-01-0.
1981b. 1980 lake water quality and 1980 sewer sampling
program. Rep. prep, for City of Waseca, Minnesota; Clear Lake
restoration proj. under EPA grant no. S804691-01-0.
Hill, B. H. 1979. Uptake and release of nutrients by aquatic macro-
phytes. Aquat. Bot. 7:87.
Kadlec, R. H. 1981. How natural wetlands treat wastewater.Pages
241-254 in Proc. Midw. Conf. Wetland Values and Management.
Freshw. Soc., Navarre, Minn.
National Biocentric, Inc. 1974. Waseca lake study. Rep. prep, for
City of Waseca, Minn.
_. Letter to City of Waseca re pilot plant results, 1978.
Unpubl. rep.
1978.1977 sewer sampling program and 1977 lake water
quality monitoring program. Rep. prep, for City of Waseca, Min-
nesota; Clear Lake restoration proj. under EPA grant no.
S804691-01-0.
1979.1978 sewer sampling program and 1978 lake water
quality monitoring program. Rep. prep, for City of Waseca, Min-
nesota. Clear Lake restoration proj. under EPA grant
nesota. Clear Lake
S804691-01-0.
no.
Perry, J. J., D. E. Armstrong, and D. D. Huff. 1981. Phosphorus
fluxes in an urban marsh during runoff. Pages 199-211 in Proc.
Midw. Conf. on Wetland Values and Management. Freshw. Soc.,
Navarre, Minn.
Schindler, D. W. 1978. Factors regulating phytoplankton production
and standing crop in world's freshwaters. Limnol. Oceanog.
23:478.
Standard Methods for the Examination of Water and Wastewater.
1975. 14th ed. Am. Pub. Health Assn., Washington, D. C.
Wanielista, M. P., and Y. A. Yousef. 1980. An example of urban
watershed management for improving lakewater quality. Pages
307-311 in Proc. Int. Symp. on Inland Waters and Lake Restora-
tion. EPA 44015-81-010. U. S. Environ. Prat. Agency, Washington,
D.C.
Wenck, N. C. 1981. Wetlands and organic soils for the control
of urban stormwater. Pages 227-237 in Proc. Midw. Conf. on
Wetland Values and Management. Fresh. Soc., Navarre, Minn.
Wile, I., G. Palmateer, and G. Miller. 1981. Use of artificial wetlands
for wastewater treatment. Pages 225-271 it) Proc. Midw. Conf.
on Wetland Values and Management. Freshw. Soc., Navarre,
Minn.
30
-------
HYPOLIMNETIC AERATION AND FUNCTIONAL COMPONENTS OF
THE LAKE ECOSYSTEM: PHYTOPLANKTON AND
ZOOPLANKTON EFFECTS
KENNETH IAN ASHLEY
Fish and Wildlife Branch, Ministry of the Environment
University of British Columbia
Vancouver, British Columbia
ABSTRACT
The whole-lake experimental approach was used to examine the effect of hypolimnetic aeration on
selected biological components of the lake ecosystem. These included phytoplankton and the limnetic
macrozooplankton community. A small (3.9 ha, Z max. = 9.0 m), naturally eutrophic lake was divided
into experimental and control halves by a sea curtain, and a hypolimnetic aerator installed in the ex-
perimental half and operated from April 1978 to March 1979. Hypolimnetic aeration exerted a minimal
effect on phytoplankton abundance and species composition because of an apparent shortage of
micronutrients and confinement of circulation currents to the hypolimnion during thermal stratifica-
tion. Daphnia pulex was not influenced by the aeration treatment because of long-term adaptation
to low oxygen levels. The winter population of Keratella quadrata was larger on the aerated side as
aeration currents enhanced its food supply. Cyclops bicuspidatus and Diaptomus leptopus were generally
more abundant after several months' aeration as higher oxygen levels increased adult and juvenile
survival.
INTRODUCTION
Excessive fertilization of natural waters is one of the most
serious water quality problems in the world today (Dunst
et al. 1974; Natl. Acad. Sci., 1969). Cultural eutrophica-
tion is caused by excessive nutrients such as phosphorus
and nitrogen in lakes, streams, rivers, estuaries, and
coastal waters (Wetzel, 1975). In lakes, these nutrients in-
crease aquatic plant growth, cause undesirable changes
in species composition, deplete oxygen, kill fish, and
lessen water quality for domestic, recreational, and in-
dustrial use (Lee, 1970).
Following the limiting-nutrient controversy of the late
1960's, attention in the 1970's focused on reducing
nutrient inputs and rehabilitating culturally eutrophied
lakes. Certain lakes recovered from excessive nutrient
loading after nutrient diversion, e.g. Lake Washington (Ed-
mondson, 1972); however, nearby Lake Sammamish did
not respond similarly (Rock, 1974). Lake Trummen in
Sweden is another lake in which the eutrophic status re-
mained unchanged following nutrient diversion (Bjork et
al. 1972). Lakes of this type were sufficiently eutrophic to
maintain their present state via internal nutrient recycling
long after external nutrient sources were removed.
As a result, the field of lake restoration came into ex-
istence as limnologists and engineers attempted to
develop methods for restoring eutrophic lakes. Lake
restoration refers to ". . . the manipulation of a lake
ecosystem to effect an in-lake improvement in degraded,
or undesirable conditions" (Dunst et al. 1974). Artificial
aeration is one technique used in restoring eutrophic lakes
(Lorenzen and Fast, 1977). Artificial aeration reoxygenates
depleted hypolimnetic waters and technically creates
oligotrophic oxygen conditions in eutrophic lakes.
However, as is often the case with new technology, ar-
tificial aeration as a lake restoration technique was initial-
ly applied with little understanding of its ecological impact
(e.g. Patriarche, 1961). Shapiro (1978) stated "Lake
restoration is not a science yet. It is still in need of
research If only 5 percent of the moneys allocated for
doing were to be diverted to understanding, the returns
would be substantial." Fast (1975) eloquently suggested
"... lake medicine is still in the 15th century. Lake doc-
tors with their bags of toxicants, dredges, coagulants, and
other devices are still in the medical equivalent of the
bloodletting stage." Clearly, the field of lake restoration
requires increased ecological awareness.
The purpose of this experiment was to examine the ef-
fects of hypolimnetic aeration on functional components
of the lake ecosystem using a whole lake perturbation ap-
proach. Specifically, its objectives were to investigate
biological components of the lake ecosystem which I felt,
after extensive literature review, were poorly understood
in terms of their response to hypolimnetic aeration. These
components were as follows:
1. Phytoplankton. The effect of hypolimnetic aeration
on the phytoplankton community is largely unknown. Bern-
hardt (1967) documented physical redistribution of
hypolimnetic algae by aeration currents; however, few
studies have focused on epilimnetic algae. Fast's (1971)
results are questionable as leaks in the aeration tower
stimulated dense algal blooms. Generating hypotheses
about algal response is difficult because of the number
of impinging variables, many of which are poorly
understood. These include iron availability (Murphy et al.
1980), nutrient release (Fast, 1975), pH shifts (Shapiro,
1978) and turbidity changes (Fast, 1971).
I would not expect hypolimnetic aeration to immediate-
ly affect the phytoplankton community as circulation cur-
rents are generally confined to the hypolimnion. Long-term
changes in species composition (f.e. fewer blue-greens)
should occur after several circulation periods as
phosphorus levels are gradually reduced and zooplankton
numbers increase.
2. Zooplankton. The zooplankton community has also
received little attention in hypolimnetic aeration ex-
31
-------
Lake Restoration, Protection and Management
periments because of its relative isolation from the initial
objective of raising oxygen levels. Hypolimnetic aeration,
by virtue of its ability to modify the physical, chemical, and
biological environment, has the potential to alter
zooplankton distribution, abundance, and species com-
position. Shapiro (1978) demonstrated that pH shifts could
affect zooplankton by changing the palatability of their food
source. Kitchell and Kitchell (1980) elegantly demonstrated
how oxygen stratification can modify zooplankton com-
munity structure. I would expect hypolimnetic aeration to
significantly increase zooplankton vertical distribution and
population size by increasing oxygen levels, venting tox-
ic gases (HzS, NH3), and oxidizing reduced metals
(Fe+2,Mn+2) in the hypolimnion.
The whole-lake experimental approach was selected for
this research project. Experimental manipulation of small
lakes provides a realistic setting for experimental investiga-
tion while avoiding extrapolation errors of lab-
oratory/enclosure experiments and economic/logistic pro-
blems associated with manipulating large lakes. Several
strategies are currently available within the realm of whole-
lake experiments. Among these, the "split-lake" design,
similar to Schindler's Lake 226 project (Schindler, 1974)
was chosen for this experiment as it allows simultaneous
experimental and control treatments within a single lake
basin.
STUDY AREA
Black Lake lies at an elevation of 750 m near the division
between Keremeos Creek and the Marron Valley in the
Southern Interior Plateau limnological region of British Col-
umbia (Northcote and Larkin, 1956). Black Lake originated
as part of nearby Yellow Lake, whose basin was cut into
volcanic rock by a major meltwater outflow that drained
the Kaledon tongue of the main Okanagan ice lobe
(Nasmith, 1962). Following deglaciation, approximately
8,900 years B.P., Black Lake became isolated from Yellow
Lake by gradual erosion and alluvial deposits from Yellow
Lake Creek.
The Black Lake watershed faces south, at a low eleva-
tion (1,524 m max.) and has poor water retention capaci-
ty (Botham, pers. comm.). Although its size (1,532 ha) ap-
pears large in relation to Black Lake's surface area (3.95
ha), very little surface runoff actually reaches the lake via
one small creek (Yellow Lake Creek). Mean annual runoff
is estimated at 268,000 m3; however, up to 75 percent is
lost via evaporation (Reksten, pers. comm.). Yellow Lake
Creek flowed from April to June 1978, peaking in mid-May
at 3-5 m3/min. Drainage basin lithology is porous Tertiary
volcanic rock, mainly basalt and andesite (Learning, 1973),
forming rolling ponderosa pine-sagebrush uplands
characteristic of British Columbia's dry interior zone
(Lyons, 1952).
Lake Description and Morphometry
Black Lake (Fig. 1) is a naturally eutrophic dimictic lake
with marked thermal stratification in summer, inverse
stratification in winter, and historically low levels of dis-
solved oxygen (Halsey and MacDonald, 1971). Aquatic
vegetation is confined to the shallow west and southeast
ends of the lake, consisting mainly of Ceratophyllum sp.
and Potamogeton sp. (Finder-Moss, pers. comm.).
Algal blooms usually reduced transparency and Secchi
disk readings averaged (over the ice-free season) 3.7 m.
Sediments were loosely compacted in shallow water and
deep-water samples were highly organic and gelatinous
with a strong H2S odor.
Important morphometric features of Black Lake are
listed in Table 1. The contour map was based on echo-
sounder transects applied to an aerial photograph of the
lake outline.
Figure 1.—Black Lake showing depth contours, compressor site, curtain position, aerator location, and sampling sites.
32
-------
Restoration and Protection Techniques
Table 1. — Morphometric features of Black Lake
1. Location: Lat. N 49° 20' 30" Long. W. 199° 44' 5"
2. Elevation: 750 m
3. Area: 3.95 ha
4. Volume: 178, 543 m3
5. Max. depth: 9.0 m
6. Mean depth: 4.52 m
7. Shoreline development: 1.32
8. Shoreline length: 927 m
9. Max. length and orientation: 363 m, NW-SE
10. Max. width and orientation: 134 m, N-S
11. Max. level change (spring-fall): 0.73 m
12. Drainage area: 1,532 ha
13. Ice off: March 31, 1978
14. Ice on: Nov. 14, 1978
15. Max. ice thickness: 0.38 m
Whole lake
Experimental
Control
z
0
1
2
3
4
5
6
7
8
9
Area (m2)
39456
35506
30279
25114
21350
18020
14141
9920
4399
1348
Strata
0-1
1-2
2-3
3-4
4-5
5-6
6-7
7-8
8-9
Vol. (m3)
37464
32850
27652
23206
19662
16041
11967
6975
2726
Area
20204
17912
14094
12445
10400
8650
6962
5065
2316
651
Vol.
19047
15965
13261
11407
9512
7791
5988
3602
1398
Area
19252
17594
16185
12669
10950
9370
7179
4855
2083
697
Vol.
18417
16885
14391
11799
10150
8250
5979
3373
1328
Total 178,543
87,971
90,572
MATERIALS AND METHODS
Aeration System
The hypolimnetic aeration system used in this study was
based on systems described by Bernhardt and Wilhelms
(1975), Hess (1975), and Smith et al. (1975) (Fig. 2). The
aerator consisted of an insulated open box (2.4 m x 1.2
m x 0.9 m) constructed of 19 mm plywood and 5 cm x 10
cm framing. Styrofoam-filled pontoons (0.3 m x 0.3 m x 2.4
m) were attached to both sides of the box and provided
360 kg of positive buoyancy. Two 0.76 m circular holes
were cut in the floor through which 0.76 m x 7.3 m
galvanized steel pipes were fitted. An air diffusor was in-
stalled 0.3 m inside the bottom of the intake pipe. The dif-
fusor consisted of four 0.38 m iron pipes (3.81 cm ID) con-
nected to a common center, and drilled with ten 1.5 mm
air release holes per arm. The outlet pipe was fitted with
a 45-degree elbow to prevent recirculation of aerated
water.
A concrete pad (3 m x 3 m x 0.3 m) was poured at the
lake's west end, 18m from the shoreline, and a plywood
shed erected to house the compressor and provide work-
ing space. A new 7.5 kw rotary vane compressor
(Hydrovane SR 4000 rated 1.13 m3/min. @ 7.0 kg/cm2)
was purchased and installed in the shed. A weighted air
line (106 m x 1.9 cm I.D.) fitted with one-way valves was
laid out on the ice surface connecting the compressor and
aerator and allowed to sink into position at ice-off.
Curtain
The lake was divided into approximately two equal sec-
tions by a plastic curtain (Fig. 1 and Table 1) designed
by the author and manufactured in Vancouver (False
Creek Industries Ltd.). The main section (103 m x 10.4 m)
was composed of woven polyolefin (Dupont Fabrene Type
P) which transmits 80 to 85 percent of visible light (Eadie,
pers. comm.). A double collar (103 m x 0.3 m) of ultra-violet
resistant black woven polyolefin (Dupont Fabrene Type
TM) was attached to the top of the main section. A rope
stretched across the ice surface marked the curtain in-
stallation position and the entire distance (103 m) was cut
open by chain saws. The curtain was unfolded along the
slot, 113 kg of lead rope attached to its lower edge and
a floatline strung through the surface collar. The curtain
was then pushed into place and sunk into position. Rocks
were piled on the near-shore area to ensure a snug fit and
scuba observation confirmed the lower edge was well
sealed into the sediment. The floation collar floated 10 cm
above the water surface and minimized surface water
exchange.
Operation
The west end of the lake contained the aerator and was
designated as the experimental side while the east end
served as the simultaneous control side. The compressor
and experimental period started April 11,1978, 111 days
after ice-off, and ran continuously until March 6, 1979, a
period of 329 days.
Sampling
All samples were collected from permanent sampling sta-
tions located near the center of the lake approximately 20
meters apart on either side of the curtain (Fig. 1). Max-
imum station depth ranged from 8.27 m to 9.0 m during
the experimental period. The west (experimental) station
was situated at right angles to the aerator outlet tube to
avoid sampling directly in the plume of aerated water.
Replicate oxygen-temperature profiles taken prior to the
INFLOW TUBE**
COMPRESSED
t
•t-
ol °
o 0 o o
c
O O o
EPILIUNION
• OUTFLOW TUBE
MVPOLIIMION
IATCR INLET
WATER OUTLET
Figure 2.—A schematic diagram of the Black Lake hypolimnetic
aerator..
33
-------
Lake Restoration, Protection and Management
experiment indicated the two central stations were
representative sites. Samples were taken every two weeks
from April to October 1978 and at three week intervals from
November 1978 through March 1979. Samples were
usually collected between 1000 and 1400 hrs.
Oxygen
Dissolved oxygen was measured at 1-meter intervals with
an oxygen meter (YSI54 ARC). Two replicate Winkler titra-
tions (Azide modification) were used to calibrate the meter
during each sampling period.
Chlorophyll a
Chlorophyll a samples were taken at 2-meter intervals and
stored in light-proof coolers prior to filtering. Two replicate
sample volumes were vacuum filtered at two-thirds at-
mosphere, preserved with magnesium carbonate solution,
stored at -73°C in dark bottles containing a silica desic-
cant, and delivered to the Environmental Lab within 48
hours. Chlorophyll a and phaeophyton a concentrations
were determined colorimetrically after extraction in 90 per-
cent acetone (Strickland and Parsons, 1968).
Phytoplankton
Phytoplankton were collected at 2-meter intervals with a
3-litre Van Dorn water bottle, combined into one sample,
and preserved with Lugol's solution (Lind, 1979). For
analysis, 100-ml subsamples were pipetted into graduated
cylinders and allowed to settle overnight. The superna-
tant liquid was then decanted and the remaining 25 mis
rinsed into sedimentation chambers and allowed to resettle
for 24 hours. Plankton were counted at 400x using an in-
verted microscope with a 300-pi diameter field of view, and
results were expressed as numbers of cells/ml. Plankton
genera were partioned into four major phyla with the re-
maining genera lumped into one group. Because of the
qualitative nature of the phytoplankton data only general
trends were noted and no quantitative analysis was under-
taken. Plankton identification and counting was perform-
ed by Mr. Al Redenback at the University of British
Columbia.
Zooplankton
Zooplankton were collected at 1-meter intervals with a
27-litre Schindler-Patalas trap (Schindler, 1969) using 84
ju Nytex mesh. During the first five sampling trips one trap
set per meter was used, after which two replicate sets (54
litres total) were combined for each depth interval.
Samples were rinsed into plastic vials and preserved with
4 percent formaldehyde/sucrose solution (Haney and Hall,
1973). Plankton were subsampled with a 1-ml Stempel
pipette into 10-ml Sedgwick-Rafter cells and counted
under a dissecting microscope (Wild M5) at 25x. Subsam-
ple size was adjusted to obtain at least 100 counts for any
given species up to a maximum of 10 mis per subsam-
ple, after which the entire sample was counted. Sample
counts were expressed as numbers/m2 and tabulated on
a computer program designed for the experiment (Steer,
pers. comm.).
Aerator
Water flow rates were measured with a recently calibrated
flowmeter (General Oceanics No. 2035) situated 2 meters
below the surface in the downflow tube. Air flow was
calculated from nomograms and tables for various air
pressures and orifice diameters (Atlas Copco, 1978). Ox-
ygen was measured in the down flow tube at a depth of
5 meters with a Winkler calibrated oxygen meter (YSI 54
ARC).
Statistics
The statistical test used to analyze Black Lake experimen-
tal data is a two-way analysis of variance program written
by N.E. Gilbert of the University of British Columbia. This
particular method of analysis was recommended by N.E.
Gilbert and Dr. P.A. Larkin and is outlined in further detail
by Gilbert (1972).
RESULTS
Circulation
The compressor was started on April 11,1978 immediately
after pre-aeration data were collected from both sides of the
lake. Air bubbles rising up the inflow tube acted as an air-
lift pump and generated a large volume-low velocity water
flow. The air-water plume rose 10 cm above the water sur-
face in the separator box while degassing, then flowed across
the box toward the downflow tube. Aerated water then
entered the downflow tube and discharged back into the
hypolimnion.
A strong odor of H2S was immediately released from
upwelling water and could be detected several meters away
from the separator box. No H2S odor was evident during the
next sampling period; however, a characteristic musty odor
persisted for the remainder of the experiment. Water in the
aerator remained clear during initial startup and regular
operation indicating the sediment was not being disrupted.
The aerator was examined under several water flow re-
gimes to determine an efficient operational setting. The final
setting chosen for the experiment was 10.67 m3/min which
generated a daily flow rate of 15,365 m3/day. Theoretical
hypolimnetic circulation time (hypolimnion volume = 18,779
m3) at this setting was 1.2 days, and 5.7 days were required
to circulate the entire experimental side (87,971 m3) under
ice cover.
Oxygen
The aerator increased dissolved oxygen by an average of
0.7 mg/l on each cycle through the system. Water flow was
constant at 15,365 m3/day, therefore a total of 10.76 kg
Oz/day was added to the experimental hypolimnion (5-9 rn).
As a result, hypolimnetic oxygen concentrations increased
from 0.2 to 2.0-2.5 mg/l after 13 days' aeration while control
values remained at 0.1 to 0.3 mg/l (Fig. 3). Hypolimnetic ox-
ygen levels remained below saturation all summer; however,
near bottom values (9 m) in the aerated portion were con-
sistently higher (2.7 mg/l av.) than in the control portion (0.2
mg/l av.). Surface values were similar on both sides and
averaged 8.5 mg/l from May until August.
Oxygen stratification was disrupted by fall circulation and
by October 24 surface and bottom values differed by less
than 0.8 mg/l. Fall circulation was vigorous and breaking
waves were observed over the whole lake surface for the
first time. Black Lake froze completely on November 14 and
entered winter with oxygen levels of 7.2 (9 m) to 7.6 (0 m)
mg/l (experimental) and 5.5 (9 m) to 7.1 (0 m) mg/l (control).
Aeration continued throughout winter, circulating the en-
tire experimental side under ice cover. Experimental oxygen
levels remained above 4.9 mg/l all winter and by March 6
the lowest recorded value was 6.0 mg/l. In contrast, control
oxygen concentrations declined markedly during winter and
less than 1.0 mg/l was present in the lowermost strata (8-9
m) from January 23 to March 6.
34
-------
Restoration and Protection Techniques
Figure 3.— Oxygen isopleths for experimental (west) and control (east)
sides.
Phytoplankton
Biomass (chlorophyll a)
The spring phytoplankton bloom was well underway when
sampling started in early April 1978, As a result, April
chlorophyll values were among the highest recorded during
the experiment. Control chlorophyll reached a maximum of
54 (jg/l; experimental, a 43 pg/l maximum. After the spring
bloom, chlorophyll on both sides declined to low levels
(Fig, 4). Average surface values during summer months
(May-August) were 4.2 ^g/l (experimental) and 4.5 fig/I
(control). Bottom values (9 m) were also low, averaging
5.5 fig/l (experimental) and 4.5 ^g/l (control).
The fall phytoplankton bloom occurred in September
and started two weeks earlier on the experimental side.
BLfCK-LflKE-VEST-CHlOROPHYU : R-yUG/L
LfiCK-LRKE-ERST-CHLOROPHYlUH-UG/L
Figure 4.—Chlorophyll a isopleths for experimental (west) and con-
trol (east) sides.
As a result, fall chlorophyll concentrations were higher on
the experimental side. Chlorophyll then declined to low
levels (4 to 8 j^g/l) with the onset of fall circulation and re-
mained low until ice formation.
A large under-ice bloom (55 ^g/l max. experimental; 85
/^g/l max. control) occurred in December-January and was
generally confined to the surface (0 to 2 m) layers. Chlor-
ophyll concentrations declined after the winter bloom and
remained below 10 ^g/l for the duration of the experiment.
The seasonal concentrations of chlorophyll were
significantly (2 way ANOVA, p<: .01) different as control
side values were higher during the spring bloom and lower
during the fall bloom. Phaeophytin was detected during
the spring bloom and remained below detection levels for
the remainder of the year.
Composition
Cyanophyta
Blue-green algae in Black Lake were represented by
several taxa: colonial Merismopedlum sp., Aphanizomenon
sp. and a small (2 to 4 p dia.) unidentified coccoid form
that dominated numerically. Merismopedium sp. and the
coccoid were present throughout the entire year whereas
Aphanizomenon sp. appeared only during the fall bloom.
In addition, Anabaena sp. briefly appeared in August
samples.
Blue-greens appeared in large numbers in late spring
(4,000 to 6,000 cells/ml) before declining to low levels
(1,760 cells/ml) in June and July (Fig. 5). Their abundance
increased again in early autumn and reached a fall peak
on October 10 at 6,775 cells/ml (experimental) and 5,135
cells/mi (control). Blue-green numbers then declined
steadily for the remainder of the experiment. The seasonal
occurrence of blue-greens was similar on both sides,
Cryptophyta
Chroomonas sp. and Cryptomonas sp. were the two Cryp-
tophyta identified in Black Lake samples. Cryptophyta
were most abundant during spring and fall months and
less prevalent through summer and winter. The seasonal
abundance of Cryptophyta was similar on both sides (Fig.
5).
The spring bloom peaked on May 9 (experimental) and
April 24 (control), then diminished over summer as Cryp-
tomonas sp. declined to undetectable levels. After August
15 both sides increased because Cryptomonas sp. reap-
peared and existing Chroomonas sp. were enhanced. The
fall bloom peaked on November 14 (experimental) and
December 5 (control) at 376 and 340 cells/ml. Total
numbers of Cryptophyta then dwindled as winter
progressed.
Chlorophyta
Two genera of Chlorophyta, Chlamydomonas sp, and
Schroderia sp,, were identified in Black Lake samples.
Chlamydomonas sp. was initially present in samples but
declined to low levels after a few weeks aeration (Fig. 5),
Schroderia sp, then appeared and became the dominant
Chlorophycean. The seasonal abundance of Chlorophyta
fluctuated throughout the summer months. Experimental
side Chlorophyta dipped to undetectable levels in late
August while control side numbers remained at 90 cells/ml.
Both sides increased to 360 cells/ml in early October as
Chlamydomonas sp. reappeared in the water column. The
fall bloom persisted until December 5, then gradually
dwindled over the remaining winter months. Both sides
35
-------
Lake Restoration, Protection and Management
BLOCK LflKE PHTTOPLflNKTON
UEST-CELLS/ML +
CAST-CEULS/ML x
BLOCK LUKE PHrrOPLflNKTON
CHLORopHrrn w.sr-cELLs/nt +
CHLOROPHTIH EBST-CEU.S/HL X
01
juftg jui? *uc AW aerr
act MOV dec 4AN FEB MARCH
1979
yEST-CELLS/ML 4
CRTPTOPHrTR EHST-CELLS/ML X
SWHL MAY JUNI JtltY W« *OQ 8feW OCT
WB
BflCILLflRIOPHrTR VEST-CEUS/MU +
BfiCILLRRIOPHTTfl EflST-CELLS/ML X
to
_J
UJ
(—1
_J
Rgure 5.— Phytoplankton composition in experimental (wast) and control (east) sides.
exhibited similar seasonal trends with the exception of a
late December-January bloom on the control side.
Bacillariophyceae
Bacillariophyta in Black Lake were represented by the
orders Centrales and Pennales. A small spring pulse of
diatoms developed on the control side and peaked on May
9; however, a spring pulse was not observed on the ex-
perimental side (Fig. 5). Experimental side diatoms even-
tually appeared on June 13.
The abundance of diatoms fluctuated during the fall
bloom; however, both sides converged at 54 cells/ml on
October 10. The winter pulse of diatoms peaked on
January 23 (experimental) and November 14 (control) at
409 and 161 cells/ml. Diatoms remained abundant for the
duration of the experiment. In general, experimental side
diatoms were more numerous through summer, fall, and
winter months whereas control side diatoms dominated
in spring.
Zooplankton
The limnetic macrozooplankton community in Black Lake
was quite simple, consisting of a calanoid copepod, a
cyclopoid copepod, one Daphnia species, and a single
species of rotifer.
Zooplankton data were analyzed for both vertical and
seasonal differences in distribution and abundance. The
vertical distribution of zooplankton was not significantly
different (2 way ANOVA, p > .01) between control and
experimental sides for any species; therefore, zooplankton
abundance was converted to an areal basis and ex-
pressed as no,/m2.
Total Zooplankton
Total zooplankton (Daphnia pulex, Keratella quadrats,
Cyclops bicuspidatus, and Dlaptomus leptopus) numbers
were statistically different (2 way ANOVA, p •< .01) be-
tween sides. This was primarily caused by two- to four-
fold greater numbers on the control side during the spring
bloom (Fig. 6). Total numbers were similar during sum-
mer; however, experimental side numbers were general-
ly higher during fall and winter months.
Daphnia pulex
The seasonal abundance of Daphnia pulex was not
significantly (2 way ANOVA, p •< .01) influenced by
hypolimnetic aeration as nearly identical seasonal trends
occurred on both sides of the lake (Fig. 6). Daphnia were
present in low numbers (< 600 m2) at the start of the ex-
periment and first appeared in appreciable numbers
(113,800/m2 experimental; 80,500/m2 control) on May 9.
The population then rapidly expanded and remained
above 60,000/m2 throughout the summer and early fall
months. Maximum abundance occurred during mid-June
to mid-August with peak numbers of 313,800/m2 (ex-
perimental) and 262,200/mz (control).
Daphnia increased briefly during the first two weeks of
fall circulation; however, by mid-November Daphnia had
declined considerably and continued decreasing
throughout the winter.
Keratella quadrata
The seasonal abundance of Keratella quadrata was
statistically higher (2 way ANOVA, p <:.01) on the ex-
perimental side, particularly during fall and winter months.
Keratella was initially detected in late April to early May
but remained scarce for several weeks. Rotifer abundance
36
-------
Restoration and Protection Techniques
a: i
UJ
a
CD
-~ 4320CIXI
81RCX LflSE IQOPIR'JKION
QflPHNIR WfST-NO /SO H -t-
QSPHN^ EqST-NO /SO H X
BLOCK LUKE zaopiwwrtw
KERSTFLlfi ME5T-SO./SG fi 1-
KERflfEUH E«ST-SD /SO.M X
Figure 6.—Total zooplankton, Daphnia pu/ex and Keratel/a quadrata
(numbers/m2) in the experimental (west) and control (east) sides.
then increased above 10,000/m2 on June 13 (Fig. 6). Con-
trol side rotifers were generally more abundant than their
experimental side counterparts throughout the summer
and early fall.
As fall circulation commenced Keratella abundance in-
creased two- to threefold over late September values and
continued increasing throughout fall and winter. Ex-
perimental side rotifers were now more numerous and re-
mained higher for the rest of the experiment. The fall-
winter population peaked on January 23 at 898,300/m2 (ex-
perimental) and 745,700/m2 (control), then started declin-
ing. Control side rotifers decreased more rapidly as only
81,400/m2 were present on March 6 as compared to
560,300/m2 on the experimental side,
Cyclops bicuspidatus
The population of Cyclops bicuspidatus inhabiting Black
Lake during the experimental period was multivoltine, ex-
hibiting a large spring pulse and a smaller fall peak in
abundance. Most Cyclops overwintered in the form of
planktonic late-stage copepodites. Each developmental
stage investigated (nauplii, copepodites, and adults) was
similar in vertical distribution on both sides; however, each
stage differed significantly (2 way ANOVA, p <.01) in
terms of seasonal abundance.
In general, each stage was more abundant on the con-
trol side during the spring months and similar on both sides
through summer (Fig. 7). This large spring population was
responsible for the two- to fourfold difference observed in
spring total zooplankton numbers (Fig. 6). The experimen-
tal side was slightly more populated in fall and both sides
varied during winter months. The development of Cyclops
bicuspidatus from nauplii to copepodites to adults was well
defined and easily followed through both spring and fall
generations.
Diaptomus leptopus
Diaptomus leptopus was the only limnetic calanoid
copepod present in Black Lake during the experimental
CYCLOPS NflUPLI!
CrtLOs5 CEPEPOOIFFS l-S .ifST-NO MO /5S 'iflfR 4
CTCLOPS ro^tpaai'fs 1-5 fusi-tco NO /^ HrrtR x
LU
i—
LU
JUNE JULY *IK5
SEPT OCT
SVflW LfKE ZOQfLtKWIOM
craors flou.is yfsr-ra /so n
CTCLOPS QGUUS EHS'-NQ '53 M
Figure 7.—Cyclops bicuspidatus nauplii, copepodites, and adults
(numbers/m*) in the experimental (west) and control (east) sides.
37
-------
Lake Restoration, Protection and Management
BUCK IRKE ZODPLHNKTON
QrRPTQHUS NSUPLII WEST- N9 ';- HETER 4-
oispimos
o
to
SLflCK LSKE ZOOPLflNKTQN
OIFSPTGK'JS COPEPQQITE C1-C5 YEST-NQ /SO.METER -t-
Diflpronus copEpoone o-cs EUST-NO /SOFTER x
UJ
I—
UJ
o
BLOCK LUKE 200PLRNKTON
ronus flouir VESI-NS /so H +
10ULT EQSl-fiO /SO H X
Figure 8.—0/aptomus leptopus nauplii, copepodites, and adults
(numbers/m2) in the experimental (west) and control (east) sides.
period. In contrast to Cyclops blcuspidatus, Diaptomus lep-
topus was univoltlne and produced a brief midsummer
generation of nauplii and copepodites (Fig. 8). Adult Diap-
tomus were present in the water column from June 1978
to March 1979. Diaptomus overwintered in the form of
resting eggs produced during fall circulation.
Copepodite stages of Diaptomus leptopus were not in-
fluenced by hypollmnetic aeration; however, nauplii and
adults showed significant (2 way ANOVA, p -<.01) dif-
ferences in seasonal abundance. In general, control side
nauplii were more abundant in early summer and similar
for the rest of the experiment. Experimental side adults
were more abundant during late summer to fall months
and variable at other times.
DISCUSSION
The phytoplankton community in Black Lake was
characterized by seasonal blooms of Chlorophyta, Cryp-
tophyta, and diatoms superimposed against a relatively
large background population of blue-green algae. Slight
differences did exist between control-and experimental
sides; however, these were generally short-lived and the
overall pattern of seasonal abundance was similar except
for the December-January period.
During this period, diatom abundance was up to ten-
fold higher on the experimental side and Chlorophyta
numbers were up to fourfold greater on the control por-
tion. I believe this difference was related to aeration cur-
rents circulating the experimental side under ice cover.
Diatoms would be favored in this environment as they often
occur naturally in turbulent conditions such as spring and
fall circulation. Chlorophyta would be favored on the con-
trol side as quiescent conditions would allow them to
stratify in the shallow photic zone immediately beneath
the ice cover (Wetzel, 1975).
Black Lake chlorophyll a values were high during spring,
fall, and midwinter; however, midsummer levels resembled
oligotrophic values despite large reserves of or-
thophosphate (Schindler, 1974). Nitrate levels were
generally less than 0.02 mg/l, therefore I expected large
blooms of N-fixing blue-green algae. Their absence and
low numbers of other species suggest midsummer
phytoplankton in Black Lake were restricted by some form
of micronutrient. Murphy et al. (1976) and Goldman (1966)
have both demonstrated the importance of micronutrients
in stimulating phytoplankton growth and N fixation in blue-
green algae. Molybdenum levels were not determined;
however, iron was consistently below 0.1 mg/l during sum-
mer months. This may explain the absence of large blooms
of N-fixing algae.
Epilimnetic (0 to 4 m) chlorophyll a values were lower
on the experimental side in early May. I believe this
resulted from stream flushing and dilution by Yellow Lake
Creek which peaked in mid-May at 3 to 5 m3/min. Algai
composition data were less affected by stream flushing
as they were averaged over several depth intervals.
Chlorophyll a data also indicated the fall phytoplankton
bloom started two weeks earlier and was larger on the
aerated side. This result was probably caused by
micronutrients that had accumulated in the experimental
hypolimnion throughout the summer. At fall turnover these
nutrients were mixed throughout the water column and
stimulated an earlier fall bloom. Murphy et al, (1980)
discovered higher iron values on the aerated side of Black
Lake, supporting this hypothesis.
Long-term effects of hypolimnetic aeration on the
phytoplankton community should become more apparent
after several annual circulation periods. It is difficult to
predict exactly which changes will occur as the
micronutrients that limited algal growth were not identified.
I suspect one of the few ways to significantly alter the
phytoplankton community in Black Lake would be via total
destratification or iron and nitrogen additions (e.g. Barica
et al. 1980).
Zooplankton
The zooplankton community in Black Lake did not respond
as expected to the experimental treatment. The reasons
for this are twofold. First, the 5 to 7 m strata in the control
hypolimnion was aerobic throughout the entire experiment.
This reduced the effect of hypolimnetic aeration on
zooplankton vertical distribution as the control hypolim-
nion was already partially aerobic. Therefore, as far as the
zooplankton were concerned, the two sides of Black Lake
were relatively alike, hence the similarity in vertical
distribution.
Second, stream flushing in spring 1978 reduced juvenile
Cyclops numbers on the experimental side. Yellow Lake
38
-------
Restoration and Protection Techniques
Creek inflow peaked in early May, coinciding with the max-
imum difference between Cyclops numbers on the con-
trol and experimental sides. This theory is supported by
organic N, chlorophyll a, dissolved organic P and par-
ticulate P levels (Ashley, 1981) that were also unexpectedly
lower on the experimental side at this time. Therefore,
zooplankton data must be interpreted with these confound-
ing factors in mind.
Hypolimnetic aeration did not affect the seasonal abun-
dance of Daphnia pulex. This is an interesting result since
low oxygen levels are known to reduce Daphnia filtering
and respiration rates (Heisey and Porter, 1977) and even
cause mass die-offs (Nicholls et al. 1980). One possible
explanation is long-term exposure to low oxygen levels.
Kring and O'Brien (1976) observed low oxygen levels (1
to 3 mg/l) initially depressed filtering rates in Daphnia
pulex; however, prolonged exposure (8 to 12 hrs.)
stimulated hemoglobin production and enabled Daphnia
to resume its initial high filtering rates. Many Daphnia col-
lected during spring and summer months were noticeably
red-stained, possibly representing hemoglobin synthesized
in response to low oxygen levels. This would allow Daphnia
to remain in the control hypolimnion and explain the
similarity in vertical and seasonal distribution.
Keratella quadrata inhabiting Black Lake reached max-
imum population in late autumn to winter, thus conform-
ing to a cold stenothermal type of seasonal distribution
(Hutchinson, 1967). The seasonal abundance was similar
on both sides through spring, summer, and early fall;
however, late fall and winter numbers were considerably
higher on the experimental side. Low oxygen levels in the
control hypolimnion (8+9 m) did not influence rotifer ver-
tical distribution, and this response may reflect behavioral
or metabolic adaptations to low oxygen conditions
(Ruttner-Kolisko, 1975).
Seasonal population changes in rotifers are poorly
understood and quite variable (Wetzel, 1975). Planktonic
rotifers feed mainly on sedimenting seston, and I believe
circulation currents generated by the aerator enhanced
the aerated side food supply by decreasing settling rates
of seston during the critical winter period when
autochthonous production was low and allochthonous in-
puts negligible. In addition, long-term exposure to higher
oxygen levels may improve some aspect of rotifer growth
or reproduction, thus explaining higher experimental rotifer
numbers during fall and winter months.
The vertical distribution of Cyclops bicuspidatus was
similar on both sides despite hypolimnetic (8+9 m) oxygen
depletion in the control portion. This result was not en-
tirely unexpected as Cyclops sp. are capable of briefly
undergoing anaerobic metabolism (Chaston, 1969).
Cyclops may have migrated vertically during part of the
day to escape low oxygen conditions.
Cyclops nauplii and copepodites were more abundant
on the control side during spring and on the experimen-
tal side during fall. I believe spring differences in abun-
dance resulted from stream flushing on the experimental
side. This reduced juvenile Cyclops numbers; however,
other species were not affected because they were not
present at this time (e.g. Diaptomus) or were in their adult
form and able to resist flushing currents (e.g. Daphnia).
Fall differences were caused by an increased food supp-
ly on the experimental side resulting from an earlier and
larger fall bloom.
In addition, long-term exposure to low oxygen levels
reduces Cyclops bicuspidatus abundance through adult
mortality and diapause of copepodite stages (Heberger
and Reynolds, 1977). Higher experimental side oxygen
levels may have reduced the number of Cyclops mortalities
and diapausing copepodites. As a result, more Cyclops
were present in the water column during late summer-fall
despite initially higher control side numbers. Unfortunately,
both sides experienced vigorous wind-driven circulation
in October which minimized side to side differences for
the remainder of the experiment.
The seasonal distribution of Diaptomus leptopus was
significantly influenced by the experimental treatment.
Control nauplii were more abundant in early summer and
experimental adults more numerous in late summer to fall.
Copepodites' stages, while not significantly different, were
slightly more abundant on the aerated side.
I believe seasonal differences between sides were
related to resting egg development and adult survival
under aerobic/anaerobic conditions. Brewer (1964) work-
ed extensively with Diaptomus stagnalis resting eggs and
concluded anoxic conditions were necessary to suc-
cessfully hatch resting eggs. Diaptomus leptopus in Black
Lake overwintered as resting eggs deposited during fall
circulation of the previous year. Hypolimnetic aeration of
the experimental side may have delayed hatching of
resting eggs by reducing their exposure to low oxygen con-
ditions. This would explain initially higher control numbers.
However, once experimental resting eggs hatched higher
oxygen levels may have enhanced their survival as low
oxygen levels are toxic to Diaptomus sp. (Cooley, 1971).
Fall circulation increased oxygen on both sides and
minimized any remaining side to side differences.
On an overall basis the zooplankton community was not
greatly altered by one year of hypolimnetic aeration.
However, this does not eliminate the possibility of long-
term changes that could occur after several years. Most
literature surveyed (Chaston, 1969; Cooley, 1971;
Heberger and Reynolds, 1977; Heisey and Porter, 1977)
indicates limnetic zooplankton cannot tolerate anoxia or
the toxic products (H S, Mn+2) which accumulate in an
anoxic hypolimnia for extended periods of time. Therefore,
after several years I would expect zooplankton on the
aerated side to significantly expand their vertical range
and experience fewer anoxia-related mortalities than their
control side counterparts.
The introduction of planktivorous fish (which usually
follows successful aeration) would intensify these dif-
ferences by providing experimental zooplankton with a
predation-free refuge (Shapiro, 1978). Increased light
transmission may offset this advantage in certain lakes
(e.g. Kitchell and Kitchell, 1980); however, this would not
occur in Black Lake because of increased hypolimnetic
turbidity. This demonstrates the importance of oxygen
stratification in structuring the limnetic macrozooplankton
community.
ACKNOWLEDGEMENTS: A project of this magnitude would not
have been possible without the assistance of several individuals.
I am especially grateful to F. A. Ashley and G. T. Ozone
Sutherland whose mechanical intuition and construction skills en-
sured the successful design and installation of experimental
equipment. This study was supported by the Fisheries Research
Section of the British Columbia Fish and Wildlife Branch (Ministry
of Environment) and an NRC grant (67-3454) to Dr. T. G.
Northcote.
REFERENCES
Ashley, K.I. 1981. Effects of hypolimnetic aeration on function-
al components of the lake ecosystem. M.S. Thesis, Department
of Zoology, University of British Columbia.
Atlas Copco Manual. 1978. 3rd ed. Stockholm, Sweden.
Barica, J., H. Kling and J. Gibson. 1980. Experimental manip-
ulation of algal bloom composition by nitrogen addition. Can.
J. Fish. Aquat. Sci. 37:1175-1183.
39
-------
Lake Restoration, Protection and Management
Bernhardt, H. 1967. Aeration of Wahnbach Reservoir without
changing the temperature profile. J. Am. Water Works Assoc.
59:943-964.
Bernhardt, H. and A. Wilhelms. 1975. Hypolimnetic aeration
as a means of controlling redox processes on the bottom of
a eutrophic reservoir. Verh. Int. Verein. Limnol. 19:1957-1959.
Bjork, S., et al. 1972. Ecosystem studies in connection with the
restoration of lakes. Verh. Int. Verein. Limnol. 18:379-387.
Botham, J. Personal comm. Water Rights Branch, Ministry
of the Environment, Canada.
Brewer, R.H. 1964. The phenology of Diaptomus stagnalis
(Copepoda:Calanoida). The development and the hatching of
the egg stage. Physiol. Zool. 37:1-20.
Chaston, I. 1969. Anaerobiosis in Cyclops varicans. Limnol.
Oceanogr. 14:298-300.
Cooley, J.M. 1971. The effect of temperature on the develop-
ment of resting eggs of Diaptomus oregonensis. Limnol.
Oceanogr. 16:921-926.
Dunst, R.C., et al. 1974. Survey of lake restoration techniques
and experiences. Wisconsin Dep. Nat. Res. Tech. Bull. No. 75.
Eadie, W.R. 1978. Personal comm. DuPont Ltd., Montreal,
Canada.
Edmondson, W.T. 1972. Nutrients and phytoplankton in Lake
Washington. Spec. Symp. Nutrients and Eutrophication. Am.
Soc. Limnol. Oceanogr. 1:172-193.
Fast, A.W. 1971. The effects of artificial aeration on lake eco-
logy. EPA Water Pollut. Control Res. Ser. 16010 EXE 12/71.
U.S. Environ. Prot. Agency, Washington, D.C.
1975. Artificial aeration and oxygenation of lakes as
a restoration technique. In Cairnes et al., eds. Recovery and
Restoration of Damaged Ecosystems. The University Press of
Virginia, Charlottesville.
Gilbert, N.E. 1972. Biometrical Interpretation. Oxford Uni-
sity Press, England.
Goldman, C.R. 1966. Molybdenum as an essential micronutrient
and useful watermass marker in Castle Lake, California. In H.L.
Golterman and R.S. Clymo, eds. Chemical Environment in the
Aquatic Habitat. I.B.P. Symposium. Amsterdam.
Halsey, T.G., and S.J. MacDonald. 1971. Experimental trout intro-
duction and artificial circulation of Yellow Lake, British Columbia.
Rsh. Manage. Rep. No. 63. British Columbia Fish Wildl. Br.
Haney, J.F., and D.J. Hall. 1973. Sugar-coated Daphnia: A preserva-
tion technique for Cladocera. Limnol. Oceanogr. 18:331-332.
Heberger, R.F., and J.B. Reynolds. 1977. Abundance, composition
and distribution of crustacean zooplankton in relation to
hypolimnetic oxygen depletion in west-central Lake Erie. U.S. Fish
Wildl. Serv. Tech. Pap. No. 93.
Heisey, D., and K.G. Porter. 1977. The effect of ambient oxygen
concentration on filtering and respiration rates of Daphnia galeata
mendotae and Daphnia magna. Limnol. Oceanogr. 22:839-845.
Hess, L. 1975. The effect of the first year of artificial hypolimnetic
aeration on oxygen, temperature and depth distribution of rain-
bow trout Salmo gairdneri in Spruce Knob Lake. W.Va. Dep. Nat.
Res. F-19-R-3.
Hutchinson, G.E. 1967. A Treatise on Limnology. Vol. 2. Introduction
to Lake Biology and the Limnoplankton. John Wiley and Sons,
New York.
Kitchell, J.A., and J.F. Kitchell. 1980. Size-selective predation, light
transmission and oxygen stratification: Evidence from the recent
sediments of manipulated lakes. Limnol. Oceanogr. 25:389-402.
Kring, R.L., and W.J. O'Brien. 1976. Effect of varying oxygen concen-
trations on the filtering rate of Daphnia pulex. Ecology 57:808-814.
Learning, S. 1973. Rock and mineral collecting in British Columbia.
Geolog. Surv. Canada. Pap. 72-53.
Lee, G.F. 1970. Eutrophication. Univ. Wisconsin Water Resour. Cen-
ter, Eutrophication Inf. Progr. Occasional Pap. No. 2. Madison.
Lind, O.T. 1979. Handbook of Common Methods in Limnology. 2nd
ed. C.V. Mosby, St. Louis.
Lorenzen, M. and A.W. Fast. 1977. A guide to aeration/circulation
techniques for lake management. EPA-600/3-77-004. U.S. Environ.
Prot. Agency, Washington, D.C.
Lyons, C.P. 1952. Trees, Shrubs and Flowers to Know in British
Columbia J.M. Dent and Sons, Oshawa, Ontario.
Murphy, T.P., D.R.S. Lean, and C. Nalewajko. 1976. Blue-green
algae: Their excretion of iron-selective chelators enables them to
dominate other algae. Science 192:900-902.
Murphy, T.P., K. Hall, and I. Yesaki. 1980. Iron requirements of blue-
green algae in a naturally eutrophic lake. Abst. 1980 Int. Limnol.
Soc. Japan.
Nasmith, H. 1962. Late glacial history and surficial deposits of the
Okanagan Valley, British Columbia. B.C. Dep. Mines Petrol.
Resour. Bull. No. 46.
National Academy of Sciences. 1969. Eutrophication: Causes, Con-
sequences and Correctives. Washington, D.C.
Nicholls, K.H., W. Kennedy, and C. Hammett. 1980. A fish kill in
Heart Lake, Ontario, associated with the collapse of a massive
population of Ceratium hirundinella (Dinophyceae). Freshw. Biol.
10:553-561.
Northcote, T.G., and P.A. Larkin. 1956. Indices of productivity in
British Columbia lakes. J. Fish. Res. Board Can. 13:515-540.
Patriarche, M.H., 1961. Air induced circulation of two shallow Michi-
gan lakes. J. Wildl. Manage. 25:282-289.
Pinder-Moss, J. Personal comm. Curator, Herbarium Univ. British
Columbia, Vancouver.
Reksten, D.E. 1978. Pers. comm. Water Investigations Branch.
Ministry of the Environment, Canada.
Rock, C.A. 1974. Problems in restoring a mesotrophic lake using nu-
trient diversion Pap. presented at 37th annu. meet. Am. Soc. Lim-
nol. Oceanogr. June 23-28, 1974.
Ruttner-Kolisko, A. 1975. The vertical distribution of plankton rotifers
in a small alpine lake with a sharp oxygen depletion (Lunzer
Obersee). Verh. Int. Verein. Limnol. 19:1286-1294.
Schindler, D.W. 1969. Two useful devices for vertical plankton and
water sampling. J. Fish. Res. Board Can. 26:1948-1955.
. 1974. Eutrophication and recovery in experimental lakes:
Implications for lake management. Science 184:897-898.
Shapiro, J. 1978. The need for more biology in lake restoration. EPA
440/5-79-001. U.S. Environ. Prot. Agency, Washington, D.C.
Smith, S.A., D.R. Knauer, and T. Wirth. 1975. Aeration as a lake
management technique. Wisconsin Dep. Nat. Res. Tech. Bull.
No. 87.
Steer, G.J. Pers. comm. Graduate student. San Francisco Univ.
Strickland, J.D.H, and T.R. Parsons. 1968. A Practical Handbook
of Seawater Analysis. Bull. 167. Fish. Res. Board Canada.
Wetzel, R.G. 1975. Limnology. W.B. Saunders Co. Philadelphia.
40
-------
LAKE DEEPENING USING IN SITU TECHNIQUES
RICHARD E. WEDEPOHL
ADRIAN T. HANSON
JOSEPH E. SZEWCZYK
Wisconsin Department of Natural Resources
Madison, Wisconsin
ABSTRACT
Studies were conducted to evaluate alternatives to dredging, both for lake deepening and sediment treat-
ment. Sediment consolidation by drawdown is the physical lake deepening technique that was studied.
Although useful In many instances, it is often limited by physical problems associated with the dewater-
ing, whether it be by surface pumping or by aquifer dewatering. Sediment digestion using aeration, nitrate,
hydrogen peroxide, ozone, and proprietary microorganisms was evaluated in a laboratory scale study.
Although hydrogen peroxide and ozone have the most dramatic effects on sediment volume reduction,
degradation of the overlying water column may limit their usefulness. The lakes most susceptible to in
situ treatments are those having highly organic sediments.
INTRODUCTION
Decreased lake depth resulting from organic and inorganic
sedimentation is one of the common sequences of lake ag-
ing. Shallow lakes often contain excessive growths of rooted
aquatic plants and are more prone to fish winterkill than deep
lakes. Increasing a lake's depth can reduce the area of the
littoral zone, provide a substantial increase in lake water
volume, and reduce the sediment surface area to lake
volume ratio. Removal or stabilization of highly organic
sediments can reduce oxygen demand and plant produc-
tion, thereby lowering a lake's trophic level.
Reversing or delaying this aging process by dredging has
proven to be the most popular method of lake restoration
in the past. Of 104 lake projects funded by the U. S. En-
vironmental Protection Agency, 61 or 59 percent used this
technique.
Dredging, although a very effective method of deepening
lakes, also has some severe drawbacks. It is often beyond
the financial capabilities of lake communities. Peterson (197i)
estimated dredging costs by geographical region, reporting
that mean costs ranged between $1.34/ma in the Great Lakes
area to $5.63/m3 in the Northeast. Environmental concern
has also been expressed regarding the disposal of dredge
spoils.
Several alternatives to dredging have been evaluated in
Wisconsin. These can be divided into three categories: (1)
sediment consolidation through use of physical methods, (2)
sediment oxidation using biochemical agents, and (3) sedi-
ment oxidation using chemical agents.
SEDIMENT CHARACTERISTICS
Preliminary screening of the applicability of In situ techni-
ques has focused on sediments normally found in glacially
formed Wisconsin lakes. Although the limnological
characteristics of manmade lakes (Impoundments) versus
natural lakes are often not distinct, useful differences can
often be found in their sediments. Two factors of concern
in evaluating a sediment's suitability for in situ reduction are
infilling rate and organic content.
Sediment infilling rates can be used to estimate the
longevity of a lake deepening project and may also indicate
the sediment's composition. Figure 1 illustrates sediment in-
filling rates found in several Wisconsin lakes. Both lead-210
techniques as described by Koide et al. (1973) and
cesium-173 (Pennington et al. 1973) have been used to
define these rates. The variation of infilling rates observed
in a single impoundment typically results from sampling loca-
tion differences. Rates are higher near the inlets than near
the outlets. Similarly, grain sizes and organic content of the
sediments vary within the impoundments, depending upon
sample location. Although the high infilling rates found in
impoundments would negatively affect the longevity of any
INFILLING RATE (cm/yr)
Mirror
Noquebay
Trout
Mendota
Wingra
White Clay
& Black Otter
•fc Honey
« Bugle
-S Leota
«Elk Creek
O O.O6.O.3
DO.16
Zl 0.23
1 0.6
~ 1 0.68
1 0.66
0.5
//A 1 2.5
1.2
/y/7///X | 2.8
L/"/////'///////////I/J 3 4
1.5
ZXWXX///1 1 2.7
/ //////S///SfS/\ 2 . S
3 cm
^Impoundment
1///SA I
Range
Figure 1.—Some infilling rates found in glacial lakes and
impoundments.
41
-------
Lake Restoration, Protection and Management
lake deepening project, a more important factor in evaluating
whether to apply in situ techniques is the organic content
of the sediments.
Wisconsin's glacial lakes generally exhibit higher levels
of organic content in their sediments than do manmade
lakes. Most of the data in Figure 2 were collected from lakes
using a three-foot piston corer. Samples were composited
% ORGANIC MATTER I DRY WT)
tical method of applying the load is to lower the lake level,
and subsequently the water table, below the sediment sur-
face. This process removes the buoyant support of water
from the sediments and consolidates them under their own
weight. This process of consolidation is irreversible, and the
sediments do not rebound upon reflooding.
In a laboratory study Smith (1973) used a device he termed
a consolidometer (Fig. 3) to determine the consolidation
potential of sediments from several Wisconsin glacial lakes.
25
50
75
100
Big Round
Lakes
30
Figure 2.—Percent organic matter found in some glacial lakes and
impoundments.
and then analyzed for percent water and organic content.
Factors other than simple sedimentation differences were
present in those instances where the organic content of the
impoundment sediments was higher than 10 percent.
Typically when sediments with high organic content are
found in impoundments, it is because the lake was initially
formed by the flooding of a marsh.
Even within glacial lake sediments as a group a large
variation is found. As described by Wetzel (1975) lake
sediments (as opposed to sediments of running waters) con-
sist of three primary components: (1) organic matter in
various stages of decomposition, (2) paniculate mineral mat-
ter, and (3) inorganic compounds of biogenic oxygen, e.g.,
diatom frustules and calcium carbonate. The organic material
can also be divided into two main types: acid humus (dy)
such as found in peat bog systems and neutral humus (gyt-
ta), much of which is formed under anaerobic conditions.
The lake sediments evaluated for possible in situ sediment
reductions were so different that broad conclusions cannot
be made about expected results based on lake origin.
However, the very thick, organic sediments found in natural
lakes appear to hold the greatest promise for consolidation,
while artificial impoundments have much less potential. Field
and laboratory studies evaluating In situ deepening tech-
niques have therefore concentrated on glacial lakes.
PHYSICAL CONSOLIDATION
Lake deepening by bottom sediment consolidation offers a
possible alternative to dredging. Consolidation, as de-
scribed by soil mechanisms, refers to a gradual decrease
in the water content of a saturated soil under load, with an
associated rearrangement of the soil structure and a reduc-
tion in volume. In the case of lake sediments, the most prac-
Lead Weight
V HTJ _ Y
!*-<•'• {i |
•Core Tube
•Water Surface
•Sediment Sample
Figure 3.—A consolidometer for measuring potential compaction of
sediments.
He developed equation 1 to describe the degree of consolida-
tion expected in lakes having sediments of very high organic
content. This equation appeared to be quite accurate for
predicting the consolidation caused by lowering the water
table in the sediments between 1 and 2 meters.
Equation 1
= 0.10 logt
100
where
C% = percent consolidation
t = time in minutes
0.10 = empirical constant
More recently the U. S. Army Corps of Engineers, as part
of the Dredged Material Research Program, defined other,
but similar relationships to describe consolidation potential
of fine-grained sediments. Data, as presented by Halibur-
ton (1978), may apply to lake sediments as well.
Several field tests have been conducted in Wisconsin
(Dunst et al. 1974) in which drawdown and sediment con-
solidation have been studied. One field test was conducted
on Jyme Lake, a natural bog lake having a surface area of
0.5 hectares and a maximum depth of 4 meters. Smith et
al. (1972) found that an appreciable amount of consolida-
tion occurred in the 4-hectare bog area surrounding the lake
after 25 days of pumping had lowered the lake level 2.5
meters. However, difficulties in pumping and sediment/bog
migrations precluded a complete dewatering of the lake and
42
-------
Restoration and Protection Techniques
only limited amounts of consolidation appeared to occur in
the lake mud.
In another field evaluation sediment consolidation poten-
tial was studied at Horsehead Lake, Wis. As schematically
shown in Rgure 4, sediment consolidation was proposed by
dewatering the underlying aquifer. This lake's sediment had
Figure 4.—Schematic of sediment consolidation by aquifier
dewatering.
an organic content of 60 percent and a water content of 95
percent. It appeared to be a good candidate for a consolida-
tion project as it had over 30 feet of low density sediment
in some areas. A joint Wisconsin Department of Natural
Resources-U. S. Geological Survey test pumping was con-
ducted to define aquifer characteristics. Three high-capacity
wells were placed in representative locations around the lake
along with shallow observation wells. From data obtained
during the short-term pumping test, a two-layer, three-
dimensional digital groundwater model was developed. It was
determined that at least 200 days of pumping would be re-
quired to sufficiently lower the water table in the sediments
for consolidation to begin. Unfortunately, as shown in Figure
5, the predicted regional aquifer drawdown was quite ex-
tensive. Not only would the nearby shallow wells used for
drinking water be dewatered but also a full scale pumping
would have lowered the water level of a nearby, recreationally
important lake, by over 1.5 meters—an unacceptable
amount. No further evaluation of using this technique on this
lake was made.
CHEMICAL AND BIOCHEMICAL SEDIMENT
CONSOLIDATION/OXIDATION
Little has been published on in situ chemical or biological
lake deepening or sediment oxidation. Ripl (1976) has
reported success in applying nitrate to lake sediments. He
found that sediment oxygen demand and internal recycling
of nutrients were greatly reduced at Lakes Lilliesjon and
Trekanten, Sweden. Although there have been reports of
sediment volume reduction using aerators, this does not ap-
pear to have been substantiated through carefully con-
trolled studies.
Preliminary results of a University of Wiscon-
sin-Madison/DNR screening study evaluating in situ sedi-
ment volume reduction techniques has recently been
reported (Hanson, 1981). Table 1 presents some of the
treatments tested as part of this study. Sediments from four
glacial lakes having highly organic sediments were selcted
for this study (Fig. 6). The key sediment variables monitored
were:
1. Sediment volume
2. Sediment volatile solids (VS)
3. Sediment chemical oxygen demand (COD)
The supporting variables monitored in the water column
included:
1. Chemical oxygen demand (COD)
2. Total dissolved solids (TDS)
3. Volatile dissolved solids (VDS)
4. Total Kjeldahl nitrogen (TKN-N)
5. Total phosphate phosphorus (PO4-P)
6. pH
7. Sulfates (SO4)
8. Biochemical oxygen demand (BODS)
9. Organic acids.
The lake sediments were collected using an Eckman
dredge and transported to the laboratory. There 4 liters of
sediment were placed in 10-liter jugs along with 4 liters of
lake water. The standard test period for this study was 7
weeks. Final sediment and water column analyses were run
after a 24-hour settling period. Figures 7 and 8 illustrate the
effectiveness the various treatments had on two of the lake
sediments. The treatment using the strong oxidizing agents
showed the most dramatic results. These treatments also
produced the most significant changes in the quality of the
overlying water. For example, pH dropped to as low as 3.0
Table 1. — Some in situ digestion treatments studied by
Hanson (1981).
Treatment
1) Control
2) Nitrogen gas
3) Air
4) Ozone
5) Air, ozone
6) Air, hydrogen peroxide
7) Nitrate
8) Air, nitrate
9) Air, microorganisms
10) Air, microorganisms
Procedure
No treatment was made.
Nitrogen gas was bubbled through
the sediments.
Air was bubbled through the
sediments.
Ozone was bubbled through the
sediments.
Ozone was bubbled through the
sediments following a period of
water column aeration.
Hydrogen peroxide was distributed
through the sediments following a
period of water column aeration.
Calcium nitrate was distributed
through the sediments.
Calcium nitrate was distributed
through the sediments following a
period of water column aeration.
Microorganisms were added to the
water column. The water column
was then aerated.
Microorganisms were aded to the
water column. The sediment/water
column mixture was then aerated.
43
-------
Lake Restoration, Protection and Management
Figure 5—Predicted aquifier drawdown around Horsehead Lake for
a proposed consolidation project, calculated in feet.
Figure 6.—Location of lakes whose sediment was used for the
screening study.
pH units in two tests, one using ozone and one using
hydrogen peroxide. Barroin (pers. comm.) reported similar
results in a parallel study he was involved in in France work-
ing with H2O?. Further work is still being conducted on,
evaluating optimal treatment dosages and possible methods
to mitigate adverse effects on the overlying water column
(Boyle, pers. comm.).
CONCLUSIONS
Techniques other than dredging are potentially most effec-
tive on highly organic sediments most often associated with
glacial lake sediments. Sediment consolidation by drawdown
and sediment treatment using hydrogen peroxide appear to
have the most promise for whole lake in situ deepening.
However, in many circumstances adverse side effects may
prevent using these alternative techniques.
Alternative deepening techniques must still compete with
the more commonly used rehabilitation techniques. For ex-
ample, each of the four lakes whose sediment was used for
the laboratory oxidation testing has implemented a more con-
ventional lake rehabilitation plan. Lilly Lake in Kenosha Coun-
ty in southeastern Wisconsin proceeded with a 665,000 m3
dredging project. At Lilly Lake, Marathon County, Wis., a
30-hectare lake having an average depth of less than 1
meter, a 40,000-cubic meter dredging project was im-
plemented. This project was designed to improve the lake's
usability adjacent to a park. Horsehead and Little Elkhart
lakes both installed aeration systems designed to prevent
fish winterkill. Additionally, Little Elkhart opted for a
macrophyte harvesting program following a piscicide treat-
ment for fish eradication. However, all these lakes are still
possible candidates for lake deepening should an accep-
table methodology prove to be economically and en-
vironmentally acceptable.
REFERENCES
Barroin, G. 1982. Pers. comm. Station d'Hydrobiologie Lacustre,
Inst. Natl. de la Recherche Agronomique. Thonon les Bains,
France.
Boyle, W.C. 1982. Pers. comm. Professor, Dep. Civil Environ. Eng.
Univ. Wisconsin, Madison.
Dunst, B.C. et al. 1974. Survey of lake rehabilitation techniques and
experiences. Tech. Bull. 75. Dep. Nat. Resour. Madison, Wis.
Haliburton, T.A. 1978. Guidelines for dewatering/densifying confined
dredged material. Tech. Rep. DS-78-11. U.S. Army Eng. Water-
ways Exp. Sta. Vicksburg, Miss.
Hanson, A.T. 1981. A screening test for eight proposed in situ sedi-
ment digestion techniques. Masters Thesis, Civil Environ. Eng.
Madison, Wis.
300
700
600
900
400
JOO
200
100
0
LE
40
20
0
-20
-40
-60
-90
-100
LLK
700
500
JOO
400
500
200
lOQ
0
A.r CF-2 CF-4
Oj HjOj
-dO
-100
Oj H;.0Z
NO 3
NOj
Figures 7 & 8.—Comparison of sediment COD and volume following various treatments.
44
-------
Restoration and Protection Techniques
Koide, M.K. et al. 1973. Th-228fTh-232 and Pb-210 geochronologies
in marine and lake sediments. Geochim. Cosmochim. Acta
37:1171-1187.
Pennington, W. et al. 1973. Observation on lake sediments using fall-
out Cs-137 as a tracer. Nature 242:324-326.
Peterson, S.A. 1979. Dredging and lake restoration. Pages 105-1 14
in Lake Restoration. EPA 400/5-79-100. U.S. Environ. Prat. Agen-
cy, Washington, D.C.
Ripl, W. 1976. Biochemical oxidation of polluted lake sediments
with nitrate - a new lake restoration method. Ambio 5 (3),
Smith, S.A. 1973. Lake deepening by sediment consolidation.
Master's Thesis, Geol. and Geophys., Madison, Wis.
Smrth| s A et a, 1972. L^ deepening by sediment consolidation-
Lalf 'nlandu!;ake Demon- Proj' Upper Great Lakes Re9'
- Madison> Wls-
Wetzel, R.G. 1975. Limnology. Saunders, Philadephia.
45
-------
HYPOLIMNETIC WITHDRAWAL: RESTORATION OF LAKE
WONONSCOPOMUC, CONNECTICUT
R. W. KORTMANN
Ecosystem Consulting Service, Inc.
Coventry, Connecticut
E. DAVIS
Hotchkiss School
Salisbury, Connecticut
C. R. FRINK
Connecticut Agricultural Experiment Station
New Haven, Connecticut
D. D. HENRY
Lake Waramaug Task Force
Warren, Connecticut
ABSTRACT
The total phosphorus concentration in surface waters of Lake Wononscopomuc has increased from 9 to
29 ppb from 1937 to 1973. The rate of oxygen depletion in the hypolimnion has changed from 31 to 68
mg/m2/day during the same 36-year interval. The relationship between the small (15 meter deep) and large
(30 meter deep) internal basins, separated by a submerged ridge immediately below the metalimnion (ca.
11 meters deep), presumably results in the observed metalimnetic blooms of Oscil/atoria rubescens due,
primarily, to internal nutrient loading from the shallow basin. Benthic detrital electron flux (BDEF) and subse-
quent photolithotrophy have been implicated as major factors controlling metalimnetic blooms of Oscillatoria
rubescens. These metalimnetic blooms have resulted in increased organic loading of the hypolimnion,
reduced light penetration, subsequent increases in the internal phosphorus loading from the entire
hypolimnetic sediment surface, and have threatened the coldwater fishery habitat of Lake Wononscopomuc.
Implementation of hypolimnetic withdrawal from the small basin (ca. 15 meters) at a rate of 400,000 gal/day
with downstream discharge (following treatment to meet NPDES permit requirements) has resulted in in-
creased nutrient export, decreased hypolimnetic oxygen depletion, and reduced internal loading directly
(from the small basin) without altering lake level or thermal structure significantly. The combined water-
shed control efforts and selective (hypolimnetic) withdrawal have eliminated metalimnetic Oscillatoria
rubescens, increased light penetration, and restored a suitable coldwater fishery habitat.
INTRODUCTION/BACKGROUND
Lake Wononscopomuc is a dimictic, 349 acre, glacial lake
with a maximum depth of 120 feet. It is located in the
township of Salisbury, Litchfield County, Connecticut. For
the past 10 years Lake Wononscopomuc has been
monitored on a regular basis by the Science Department of
the Hotchkiss School and since 1971 has been investigated
by a variety of organizations (Tables 1-5) at the request of
the town of Salisbury. Several State agencies have done spot
checks on the lake for the past 30 to 40 years because of
its importance as a coldwater fishery and because it is the
deepest lake in Connecticut.
The local residents began to notice a decline in water clari-
ty about 20 years ago. This loss of water clarity was most
evident during the summer stratification period. The studies
that have been conducted since 1968 have indicated a
relatively rapid change in trophic condition with the occur-
rence of a dense layer of Oscillatoria rubescens in a
metalimnetic layer. The lower hypolimnion began to become
anoxic in early September 1972. In subsequent years anoxic
conditions were evident earlier in the season and higher in
the water column. This rapid decline in water quality and
apparent rapid eutrophication rate caused concern. The large
lake trout and brown trout became very scarce in the early
1970's while weed communities and the metalimnetic algal
layer flourished.
A selective withdrawal system was installed in the shallow
lake basin to increase the nutrient export from the hypolim-
nion. This system removed about 30,000 gallons per day
from the bottom of the shallow basin in 1980. Following this
pilot study and design of downstream treatment facilities, the
withdrawal rate was increased to about 400,000 gpd during
1981 and 1982. A similar siphon system was used in a
Swedish lake with much higher flowrates relative to lake size
(Gachter, 1976). That project resulted in a dramatic decline
in the biomass of Oscillatoria rubescens which was attributed
to the nutrient export characteristics of the siphon system,
as well as stabilization of redox potential, an induced
downward current (very small), decreased hypolimnetic
residence time, and changes in the thermal and metabolic
characteristics of the lake.
Lake Wononscopomuc has been studied by several State
agencies and private corporations in past years and these
data are summarized in Tables 1-5. Hypolimnetic accumula-
tion nitrogen and phosphorus were found during the early
1970's (Table 1). A peak in phosphorus and nitrogen con-
tent was also found at 10 meters (approximate inter-basin
connection depth) during July and September 1974 (Table
1). Nutrient accumulation in the shallow basin appeared to
reach the inter-basin connection depth earlier than in the
deep basin. This could have a dramatic effect on the entire
46
-------
Restoration and Protection Techniques
lake hypolimnion by enhancing metalimnetic layers of algae,
increasing the organic load, and restricting light penetration.
The nutrient gradients between hypolimnetic and
epilimnetic waters have been steep for the past 10 years
(Table 2). The high nutrient content of the mid-depth layer
was again shown in 1977 (Table 3). Watershed inputs of
phosphorus from the two major tributaries and outflow con-
centrations show that the lake acts as a phosphorus sink
(Table 3). A description of soils in the watershed is shown
in Table 4. The organic content of sediments is approximately
20 to 25 percent and has a high content of phosphorus and
iron (Table 5).
Methods and Approach
Temperature and dissolved oxygen were measured through
the water column using a YSI Model 30 probe and occa-
sional Winkler samples (Standard Methods, 1971). Secchi
disk transparency was measured to provide a simple
measure of light penetration. Total dissolved and particulate
phosphorus content was determined by molybdate col-
orimetry following persulfate digestion (Standard Methods,
1971). Total nitrogen (dissolved and particulate), ammonia,
and nitrate nitrogen were analyzed using the Kjeldahl,
Nesslerization, and cadmium reduction methods, respective-
ly. Total sulfide was determined using the methylene blue
method (Standard Methods, 1971). The temperature and ox-
ygen data were printed and plotted by a limnological soft-
ware program for an Apple II minicomputer ("JOE," Kuether,
1980). Mass balances of heat and oxygen were performed
on a IBM minicomputer using VISICALC (Visi Corp., 1981).
Morphometry
Lake Wononscopomuc consists of a shallow (15 meters)
basin and a deep (>-30 meters) basin connected to a depth
of about 11 meters (Fig. 1). A pump has removed water from
the bottom of the shallow basin according to the following
pump rate schedule:
GPP x 1Q3
July-November 1980 ca. 30-40
June 15-November 1981 ca. 330
May 15-November 1982 ca. 330
Withdrawn hypolimnetic water is aerated and retained for
several hours prior to discharge downstream.
The two-basin configuration of Lake Wononscopomuc is
the dominant feature of the lake ecosystem structure. The
percent of bottom surface area occurring below the epilim-
nion of the shallow and deep basins is 60 percent and 53
percent, respectively. The metalimnion covers approximately
10 percent of the bottom area of each basin (Fig. 2). Ther-
mal stratification is stable with some intermittent surface layer
partitioning caused by surface heating (Fig. 3). The
metalimnetic boundaries are relatively stable through the
summer season at about 5 and 10 meters (Fig. 3). The stabili-
ty of the upper metalimnetic boundary suggests that steady-
Table 1. — Wononscopomuc Lake data summary from CAES study.
Date
Iran
Depth
Alka
Ohio
Sol P
Tot P
NH4-N
NO3-N
(Connecticut Agricultural Experiment Station, Oct. 10, 1975)
Sol N
Tot N
10/17/73 4.3
5/7/74 1 .0
7/11/74 7.3
9/4/74 8.2
m mg/l
.2 2.00 —
5.0 — —
10.0 — —
15.0 — —
25.0 — —
.2 2.50 —
5.0 — —
10.0 — —
15.0 — —
25.0 — —
0-3 2.18 0.7
5.0 — —
10.0 — —
15.0 — —
20.0 — —
25.0 — —
0 - 3 2.07 2.4
5.0 — —
10.0 — 14.6
15.0 — —
20.0 — —
25.0 — —
7
8
8
12
350
5
3
4
7
7
2
1
5
4
10
—
7
26
5
5
62
276
22
18
107
30
355
50
52
44
40
29
12
13
123
31
18
192
16
54
159
25
74
296
ppb
10
10
30
50
1450
30
10
30
20
30
20
20
0
10
130
530
40
240
0
0
390
870
30
.10
10
10
60
50
40
40
40
40
40
40
30
40
40
50
10
10
10
10
10
0
320
330
290
330
1610
310
330
240
300
390
360
380
190
400
400
—
270
490
350
240
640
1260
360
400
780
390
1810
740
810
740
600
520
400
460
1160
560
490
990
620
700
1400
300
610
1380
NOTE:
Iran
Depth
Alka
Chlo
Sol P
Tot P
Sol N
Tol N
- Transparency
- Sample depth
= Alkalinity
= Chlorophyll-a
= Soluble P
= Total P
•= Soluble N
= Total N
47
-------
Lake Restoration, Protection and Management
LAKE WONONSCOPOMUC
contours in feet
600' '
Figure 1.—Lake Wononscopomuc bathymetry.
state diffusion rather than event-related mixing is controll-
ing vertical transport. The two basins are connected to a
depth of about 11 to 12 meters (approximately 1 to 1.5
meters into the hypolimnion (Fig. 4)). Because of this uni-
que arrangement of stratification boundaries and mor-
phometric characteristics the shallow basin may play a role
in promoting metalimnetic layers of algae and bacteria
disproportionate to the size of the basin.
Light Penetration
By the early 1970's a large spring bloom had become a
regular event in Lake Wononscopomuc (Fig. 5). This early
reduction in light penetration continued through 1979 and
to a lesser extent through 1982 (Fig. 5-7). The Secchi disk
transparency in late June to early July during 1973-1979
was into the metalimnion. From July through circulation in
November Secchi disk transparency was generally within the
epilimnion (Fig. 8). Secchi disk transparency has, in general,
improved in 1981 and 1982, particularly during August. When
a Secchi depth difference between basins occurred, the
small basin generally had lower transparency (Fig. 8). The
1 percent light penetration depth was measured at 11, 13,
and 13 meters in June, July, and August of 1982, respec-
tively (Battoe, pers. comm.).
Withdrawal from the small basin of Lake Wononscopomuc
appears to have caused small changes in the heat budgets
of both the large and small basins. In the small basin heat
increased 24 percent between June 18,1979 and June 22,
1981. There was a 2.25 percent decrease in hypolimnetic
heat between June 22, 1981 and June 14,1982. In the large
basin heat increased 24 percent between 1979 and 1981
and decreased 1 percent between 1981 and 1982. These
LAKE WONONSCOPOMUC: SHALLOW BASIN ( I )
(llttr=l
22219160.00
18151340.00
23*93520.00
Z7M78«0.00
UTMttO.W
136*1290.00
)W #0.00
2279120.DO
U70BOO.DO
1 Mil.28
5391.36
22J9.12
6220,80
Sa.M
2.H
4096,36
8190.7Z
71216,00
28W5.ui*
31259.W
SURFACE AREA BELOW DEPTH
VOLUME (x1O3m3)
DEEP BASIN (H )
% SURFACE AREA BELOW DEPTH VOLUME
-------
Restoration and Protection Techniques
8/17
a/10
TEMPERATURE 1981
LARGE BASIN
7/27
5/31
5/25
5/16
5/9
4/18
4/4
Figure 3.—Thermal stratification.
2O
Table 2. — Summary of available nutrient data for Lake Wononscopomuc's epilimnion and hypolimnion.
SOURCE
Union Carbide
(Table 61
CAES
Union Carbide
(Table 6)
CAES
DEP
Union Carbide
(Table 6)
CAES
DEP
CAES
Union Carbide
(Table 6)
Union Carbide
(Table 6)
Assumptions
Month
July
August
September
October
DATE
4/6/76
5/7/74
6/29/76
7/11/74
7/28/75
8/31/76
9/4/74
10/15/75
10/17/73
11/14/75
11/16/75
based on available
Epilimnion
0.035 mg/1 P
0.8 mg/1 N
0.035 mg/1 P
0.4 mg/1 N
0.02 mg/1 P
0.6 mg/I N
0.02 mg/1 P
0.35 mg/1 N
TOTAL P
)-) c
.04
.029
.229
.192
.395
.428
.296
.290
.355
1.08
.098
nutrient
-------
Lake Restoration, Protection and Management
Table 3. — Data summaries from Union Carbide study &
DEP water compliance studies
A. Phosphorus, in-lake
The accuracy of the TP test equipment during these studies was
0.04 mg/l, so any concentration below that value could not be deter-
mined. To obtain an estimate of the quantity of recycled phosphorus,
it was assumed that the TP concentrations in the water column above
21 meters remained constant over the five month period.
Hypollmnetlc accumulation of total phosphorus
8
9
10
11
B. Total nitrogen during lake stratification
The values for TN obtained in both basins of the lake were averag-
ed for the analysis. The TN value found at the top of each layer
was used for analysis.
Layer Depth (m) Volume (m3 x 10") TN mg/l kgs TN
Depth (m)
21.33-24.38
24.38-27.43
27.43-30.48
30.48-bottom
Volume (m3)
715,575
379,898
127,057
20,135
Concentration
0.131
0.300
0.428
0.428
TOTAL
KGS
93.7
133.9
54.3
8.6
290.5
1
2
3
4
5
6
7
8
9
10
11
3.04
6.09
9.14
12.19
15.24
18.28
21.33
24.38
27.43
30.48
to bottom
3.691
3.123
2.940
2.544
2.188
1.634
1.081
0.715
0.379
0.127
0.020
0.391
0.598
0.369
1.23
0.662
1.52
0.705
1.07
1.75
2.06
2.06
TOTAL
1443.2
1867.5
1084.8
3129.1
1448.4
2483.7
762.1
765.0
663.2
261.2
41.2
13,949.8
differences may reflect the different startup dates of
withdrawal in 1981 and 1982. In the small basin heat in-
creased between May and August of 1979 and 1981 2.0 per-
cent and 3.0 percent, respectively. However, during the same
period of 1982 the hypolimnion was 14 percent warmer. The
large basin showed a similar trend with overall heating losses
between May and August of 1979 and 1981 (-1.99 percent
and -14.38 percent, respectively), and a heat gain in 1982
(7.87 percent). These differences in heat content probably
resulted from increased light penetration and withdrawal of
cold bottom water from the small basin. Thermal stratifica-
tion appears to have been unaffected by withdrawal (Fig.
9-11).
Oxygen Metabolism
Summer dissolved oxygen profiles from 1970 through 1982
are shown in Figures 9-11. As early as August 1970 a very
steep oxygen gradient occurred at the inter-basin connec-
tion depth (upper hypolimnion) (Fig. 9). During subsequent
years this sharp anoxic boundary migrated upward into the
metalimnion and occurred earlier in the summer (Fig. 9),
while Oscillatoria rubescens flourished in this layer. The steep
anoxic boundary remained in the metalimnion directly
beneath the photosynthetic peak of oxygen through the sum-
mer of 1979 (Fig. 10). During the summers of 1981 and 1982
Table 3 (Continued)
C. Stream concentrations of TP (mg/l) - April thru August
(Union Carbide p. 35)
Month Sucker Brook Belgo Hill Road Brook Outlet
April
May
June
July
August
Average
0.04
0.04
0.04
0.04
0.04
0.04
D. Ammonia nitrogen
Layer
5
6
7
8
9
10
11
Depth (m)
15.24
18.28
21.33
24.38
27.43
30.48
to bottom
0.0553
0.0400
0.0515
0.0473
0.0400
0.0468
0.040
0.035
0.040
0.030
0.030
0.035
during stratification
Volume m=»
2.188
1.634
1.081
0.715
0.379
0.127
0.020
NH3 mg/l
0.098
0.844
0.177
0.577
0.997
1.52
1.52
TOTAL
kgs NH,
214.4
1444.4
191.3
412.5
377.8
193.0
30.4
2863.8
The average concentration of ammonia nitrogen during fall circula-
tion in 1976 was 0.128 mg/liler for a total lake mass of 2361.0 kgs.
(Miller, 1976; Mason, 1977b)
Table 4. — Soil content and structure along the unsewered
watershed shoreline of Lake Wononscopomuc
The soils on the eastern side of the lake which has been sewered
to the Hotchkiss School for approximately 4 years, are excessively-
drained to welkJrained Farmington soils which are shallow to bedrock
(about 18 inches is expected). The bedrock is clearly visible to many
outcroppings and cliffs along this side of the lake. This area also
has the steepest slopes (15-30%) of the entire lake periphery.
According to the USDA soil survey, this is the only area around the
lake which has very pervious soil to bedrock. Permeability should
be in the range of 1.26 to 4.0 feet per day. Nutrient absorption on
these soils should be the least of the entire shoreline.
The portion of the lake where sanitary sewers are now proposed,
for the most part, a loam or silt loam designation (Stockbridge and
Amenia) with 3 to 15% slopes. These are well-drained to moderately
well-drained soils developed in firm till. Mottling in the Amenia silt-
loams indicates the presence of a seasonal high groundwater table
in the upper part of the substratum at a depth of about 26 inches.
These particular soils, although they are rated as having severe
limitations for subsurface sewage disposal, are better able to ab-
sorb nutrients because of the fine grained structure of the substratum.
Because of this and the absence of shallow bedrock, the nutrients
are more readily available to tree and scrub roots.
Permeablility of the upper soil layer is in the range of 1.26 to 4.0
feet per day, while in the substratum it should be less than 0.4 feet
per day. The depth of the bedrock should be greater than or equal
to 10 feet.
(Mason, 1977a)
the anoxic boundary of the shallow basin did not reach the
depth of the basin connection until late September (Fig. 10).
The pump rate of 400,000 gpd is sufficient to flush the
entire shallow basin hypolimnion in about 170 days. It ap-
pears that this withdrawal rate changes the spatial (deeper)
and temporal (later) characteristics of the anoxic boundary
in the small basin. The dramatic improvement in the oxygen
content of the deep basin (Fig. 11) suggests that the shallow
50
-------
Restoration and Protection Techniques
LAKE WONONSCOPOMUC
Figure 4.—Summer lake structure: thermal and morphometric.
Table 5. — Some chemical characteristics of lake muds in
Lake Wononscopomuc and Lake Waramaug.
SECCHI DISK DEPTH
1973-1974
Sample
source
Wo-1
Wo-2
Wo-3
Wa-1
Station
depth
10ft.
56
100
40
Ash
76.6%
77.6
77.8
79.2
P
0.875 mg/gm
1.75
8.15
2.32
FE
20.2 mg/gm
19.9
25.2
48.2
Phosphorus and Iron are based on dried mud, as is the ash content.
Relationships between phosphorus and iron in the muds.
Stoichiometric quantity of iron based on phosphorus content
Wo-1
Wo-2
Wo-3
Wa-1
NOTE:
as Fe +2
2.36 mg/gm
4.75
22.0
6.25
as Fe +3
1.5 mg/gm
3.14
14.6
4.18
Wo = Wononscopomuc
Wa = Waramaug
(BenoH, 1971)
basin controls the behavior of the deep basin when the anox-
ic boundary and steep nutrient gradient migrate above the
basin connection depth,
Estimated hypolimnetic oxygen depletion rates are shown
in Table 7. Although these are rough estimates (they do not
account for eddy diffusion or varying light penetration), these
data indicate a decline in the oxygen depletion rate in the
small basin hypolimnion following withdrawal. The largest
change in the oxygen consumption rate occurred during the
first year of pumping at > 300,000 gpd while the earlier start-
up in 1982 appears to have increased oxygen consumption.
These consumption rates probably varied as a function of
organic loads from the spring bloom and purnp-induced ef-
fects.
Thermal and Metabolic Structure:
Withdrawal Effects
The oxygen profiles immediately preceding and following im-
plementation of hypolimnetic withdrawal are shown in Rgure
12 (1978) and Figure 13 (1981). These data indicate that prior
to withdrawal from the shallow basin hypolimnion the anox-
ic boundary reached the inter-basin connection in late July,
and the deep basin became anoxic shortly thereafter (Fig.
9-12). Following withdrawal the anoxic boundary was main-
tained below the connection depth and both basins re-
tained oxygen deeper and for a longer duration. The ac-
T
bECCHl DISK DEPTH
1977
SECCHI DISK DEPTH
1976
Figure 5.—Transparency 1970-1977
Dashed lines denote sampling in the shallow basin.
SECCHI DISK DEPTH
1979
JUNC JULY JUXJWT
• - - - y - - -----
Figure 6.—Transparency 1979
Dashed lines denote sampling in the shallow basin.
cumulation rate of phosphorus content of the shallow basin
actually decreased during July 1981, presumably because
of withdrawal (Fig. 14). The total phosphorus profile in the
large basin does not show a peak at the depth of the basin
connection that was observed in previous years (Fig. 14,
Table 1). These data indicate that prior to withdrawal
phosphorus accumulated in the shallow basin and elevated
51
-------
Lake Restoration, Protection and Management
SECCHI DISK DEPTH
1981
JUNE JULY . AUOUST SEPTEMBER
Table 7. — Estimated hypolimnetlc oxygen depletion rates
5
- 10
£ «
20
25
30
40
304
I
I
JT
e
i
i
L i
j
j
a.i J
i
meunimnion
9.14
hypodmnton
deepest inter* basin connection
.124
SECCHI DISK DEPTH
1982
~ 10
I 15
2O
25
30
40
1
\l
61
914
124
i . 1
. . I . - 1 .
1
. . . . 1
1
J
fleeoest 'iter.
, |
nypoiimnion
aosin connect on
Figure 7.—Transparency 1981-1982
Dashed lines denote sampling in the shallow basin.
SECCHI DISK DEPTH : SUMMARY
1973-1979
SCPIEMBCR OCTOBER M3VEMM
11 1 1
3JX
i 4
at
.B.4
1
1
fl
!
L 1 J
I>«P.« m
ny
poUmnkm
ovmnloi
1
d
II
-^ „,„.,„„, eonn«la)
SECCHI DISK DEPTH : SUMMARY
1981-1982
t,
i; f \
61
. . 1 .
914
hypoiimmon
deepest inter- bosin connecuon
Figure 8.—Transparency: Summaries
Dashed lines denote sampling in the shallow basin.
fable 6. — Hypollmnetic heat budgets.
Shallow basin:
5/31/79-8/17/79
5/31/81-8/17/81
5/29/82-8/17/82
Deep Basin:
5/31/79-8/17/79
5/31/81-8/17/81
5/29/82-8/17/82
Change In heat content
below 9 m
+ 2
+ 3
+ 12
- 2
- 14
- 8
1970
6/23-7/14
1979
5/31-6/28
1981
5/31-6/22
1982
5/29-6/21
mMD.O./square
Shallow basin
—
5.93
2.53
5.91
meter/day
Deep basin
2.42
1.56
3.86
6.72
TEMPERATURE 'C
4 8 12 16 20 21
DISSOLVE!) OXYGF.M (mfl/IUr-i )
32 A 6 12 16 m ?*]
1972-1976
TEMPERATURE 'C DISSOLVED OXYGEN (mg/llter)
4 8 12 16 20 24 28 32 48 12 16 20 24
Figure 9.—Temperature and dissolved oxygen 1970-1976.
concentrations layered over the entire lake at the depth of
the inter-basin connection. This series of events resulted in a
metalimnetic algal bloom that decreased light penetration,
increased organic load, and produced anoxic conditions in
the entire lake hypolimnion.
The influence of the shallow basin on light penetration,
metalimnetic blooms, and oxygen consumption in the large
basin appears to have been eliminated by withdrawal.
Redox System
Decomposition of organic matter does not cease with the
depletion of oxygen. In the absence of oxygen other redox
active components (e.g., iron, sulfur) act as exogenous ter-
minal electron acceptors in anaerobic respiration. As a result
of anaeorbic respiration the ambient redox potential
decreases. This results in a decreased sediment adsorption
capacity for phosphate and release of phosphorus from
sediments. The accumulation rates of total sulfide in the
shallow and deep basins during 1980-82 are shown in
Figures 15 and 16. During 1980 sulfide accumulation in the
52
-------
shallow basin reached above the basin connection and
resulted in an observed peak in the upper hypolimnion of
the large basin (Fig. 15 and 16). The rate of accumulation
of sulfide in the shallow basin was slower in 198V and did
not reach the basin connection until late September. In 1982
1978
TEMPERATURE 'C
DISSOLVED OXYGCN (mg/llur)
a i; ie 20s 24 2B u 4 e'fti it 20 24
TEMPERATURE *C
4 B 12 16 20 24 28 32
1979
DISSOLVED OXYGEN (mg/llltr)
a 12 1C 2O 24
t Down connection
Figure 10.—Temperature and dissolved oxygen 1978-1979.
1981
TEMPERATURE 'C DISSOLVED OIYOEN (mo/llur 1
12 16 20 24 28 32 4 e 12 is 20 24
7/27
1982
TEMPERATURE "C 3^7 a/23 DISSOLVED OXTGEN (mg/litw)
4 6 12 16 2O//14 28 32 4 a 12 16 20 24
Restoration and Protection Techniques
the sulfide accumulation rate showed further decreases (Fig.
15 and 16). Metalimnetic layers of Oscillatoria rubescens were
absent in 1981 and 1982. The withdrawal system appears
to have stabilized the redox dynamics of the shallow basin
at a higher redox potential.
Implications for Ecosystem Theory
Heterotrophic uptake of glucose and acetate by several
Oscillatoria species as well as photolithotrophic production
using hydrogen sulfide as an electron donor may occur in
some lakes (Saunders, 1972). Some blue-green algae resem-
ble photosynthetic sulfur bacteria in their photosynthetic abili-
ty to use electron donors other than water (Gorlenko and
OXYGEN 1978
SMALL BASIN
6/17
9/1
Figure 12.—Oxygen concentration 1978.
OXYGEN 1981
SMALL BASIN
Figure 11.—Temperature and dissolved oxygen 1981-1982.
LARGE BASIN
Figure 13.—Oxygen concentration 1981.
53
-------
Lake Restoration, Protection and Management
Kuznetsov, 1972). Oscillatoria sp. was noted in a crater lake
in Uganda where it grew photolithotrophically in a layer
(Golterman, 1975). Photolithotrophic uptake of labeled car-
bon dioxide in the presence of hydrogen sulfide was
demonstrated in that system. It appears likely that the
metalimnetic bloom of Oscillatoria rubescens in Lake
Wononscopomuc resulted from sulfide as well as phosphorus
flux from the shallow basin hypolimnion.
The eutrophication status of this lake appears to be a func-
tion of the vertical electron flux that provides a source of
energy for photolithotrophic production (Benthic Detrital Elec-
tron Flux: BDEF; Rich and Wetzel, 1978) as well as in-lake
phosphorus loading. The immediate disappearance of
Oscillatoria rubescens following withdrawal suggests that the
benthic flux of energy (BDEF) is as important as nutrient
loading (internal and external) in the eutrophication of some
lake types. The vertical electron flux was affected by the
withdrawal of hypolimnetic water as well as reduced organic
loading from spring blooms.
Implications for Management
The data collected from Lake Wononscopomuc suggest that
in lake systems where layers of Oscillatoria sp. occur
eutrophication may be as much a function of benthic detrital
electron flux and subsequent photolithotrophic production
as it is a function of external and internal phosphorus loading.
These results also indicate that restoration of such system?
may be achieved by withdrawal (or similar treatment) of a
relatively small volume of water. Hypolimnetic withdrawal of
hypolimnetic water tends to stabilize redox at a higher poten-
tial and reduces phosphorus accumulation rates. Similar
treatment systems would probably be effective in restoring
lakes where hypolimnetic metabolism and internal nutrient
loading are problems. The organic loading from spring
blooms and subsequent electron acceptor demand placed
on hypolimnia of deep lakes also appear to play a major role
in summer ecosystem structure.
ZS'lmg/liter)
p 0.5 1.0 1.5 2.0 DEEP BASIN:
TOTAL SULFIDE
Figure 14.—Total phosphorus 1981.
SHALLOW BASIN :
TOTAL SULFIDE
js. (mg/liter)
Q.S 1.0 1.5 2.0
Figure 15.—Total sulfide—shallow basin 1980-1982.
Figure 16.—Total sulfide—deep basin 1980-1982.
54
-------
Restoration and Protection Techniques
CONCLUSIONS
1. The occurrence of Oscillatoria rubescens in a
metalimnetic layer appears to have been a result of organic
loading from spring algal blooms, metabolic consumption of
oxygen, accumulation and vertical migration of electron
donors for photolithotrophy, and in-lake nutrient loading, as
well as watershed inputs.
2. Lake Wononscopomuc consists of a shallow eutrophic
basin that exerts a profound influence over a large deep
oligotrophic basin by affecting metalimnetic blooms, light
penetration, and hypolimnetic metabolism rates.
3. The decrease in the spring bloom and hypolimnetic
withdrawal from the shallow basin appear to have stabiliz-
ed the redox system, decreased oxygen consumption,
decreased in-lake nutrient loading, increased light penetra-
tion, eliminated the control of the shallow basin over
metalimnetic Oscillatoria rubescens blooms, and improved
the trophic condition of the lake.
4. The elevated oxygen content of deep water appears
to have enhanced the coldwater fishery habitat significantly.
ACKNOWLEDGEMENTS: The data contained in this report were
collected by Edward Davis (Hotchkiss School, Salisbury, Conn.); the
Connecticut Agricultural Experiment Station (New Haven, Conn.);
Connecticut Department of Environmental Protection—Water Com-
pliance Unit; Union Carbide; and the University of Connecticut. The
Jessie Smith Noyes Foundation provided stipend support for Dr. Kort-
mann to analyze part of the data contained in this report.
REFERENCES
Battoe, 1982. Pers. comm. Univ. Conn.
Benoit, Richard J., 1971. Mud analysis done on Lake Wonon-
scopomuc and Waramaug by Environmental Research and Ap-
plications, Inc., Norwich, Conn. (Unpubl.)
Connecticut Agricultural Experiment Station. 1975. Water chemistry
and fertility of twenty-three Connecticut lakes. W.A. Norvell and
C.R. Frink. Bull. 759. New Haven, Conn.
Fredette, Charles. 1977. Summary data re: Lake Wononscopomuc
hypolimnetic drawoff calculations and destratificafon costs. Conn.
Dep. Environ. Prot. Water Compl. Unit. Hartford. (Unpubl.)
Gachter, R. 1976. Die tJefenwasseralbleitung, ein Weg zur Sanierung
von Seen. Schweis Z. Hydrol. 38: 1-29.
Golterman, H. L. 1975. Physiological Limnology: An Approach to
the Physiology of Lake Ecosystems. Elsevier, Oxford.
Gorienko, W. M., and S. I. Kuznetsov. Uber die photoshythesierende
Bakterien des Konojer Sees. Arch. Hydrobiol. 70: 1-13.
Kuether, A. 1980. "JOE" a limnolocial software program. (Unpub.)
Mason, Richard D. 1977a. Lake Wononscopomuc sanitary sewer
project. Conn. Dep. Environ. Prot. Water Compl. Unit. Hartford.
(Unpubl.)
1977b. Wononscopomuc Lake—nutrient recycle quantity
calculations. Conn. Dep. Environ. Prot. Water Compl. Unit. Hart-
ford. (Unpub.)
Miller, Richard. 1976. Hydrologic/nutrient budget study of Lake
Wononscopomuc. Union Carbide Corp. Div. Aquat. Environ. Sci.
Tarn/town, N.Y.
Rich, P. H., and R. G. Wetzel. 1978. Detritus in the lake ecosytem.
Am. Nat. 112: 57-71.
Saunders, G. W. 1972. Potential heterotrophy in a natural population
of Oscillatoria agardhiivar. Isothrix Skuja. Limnol. Oceanogr. 17:
704-711.
Standard Methods for the Examination of Water and Wastewater.
1971. 13th ed. Am. Publ. Health Assoc. Washington, D.C.
Visi Corp. 1981. VisiCorp Personal Software, VisiCalc. San Jose, Calif.
55
-------
Socioeconomic Benefits of
Lakes and Restoration
HIERARCHY THEORY: THEORETICAL AND METHODOLOGICAL
IMPLICATIONS FOR SOCIOLOGICAL IMPACT EVALUATIONS
J. LYNN ENGLAND
Department of Sociology
Brigham Young University
Provo, Utah
One of the most perplexing and difficult problems I have dealt
with in analyzing social impacts is the incongruity between
my viewpoints at different stages. As a social scientist do-
ing an impact study itself the assumptions I make call for
different efforts than the efforts I will later make using new
assumptions in my role as a consultant to the decisionmakers
who will use my analysis to select policies and procedures
to attain desired outcomes.
For example, one of my earliest social impact studies in-
volved a large proposed ski resort on the outskirts of a stan-
dard metropolitan statistical area of about 100,000 in-
habitants. The project was funded by the U.S. Forest Ser-
vice, a multistate region, a county government, and a city
government.
Research compared organizations, cultures, and interac-
tion patterns in communities with and without ski resorts;
it also compared values and behaviors of skiers with
residents of the SMSA. From this, the analysis was to pro-
ject impacts of ski resorts on their host communities by
causal modeling techniques. The assumption underlying the
methodology was that ski resorts lead to certain conse-
quences for communities, their organizations, and their
residents.
When we met with the planners in each of the sponsor-
ing agencies, the assumptions and flavor of the analysis
changed dramatically to ways in which the projected impacts
could be mitigated or enhanced. Now, the assumptions were
that the projected changes were not inevitable, that purposive
actions could accomplish the desired outcomes. It is impor-
tant to recall that these planners were part of the system
being studied and acted within it. They were not disinterested
scientists who remain outside of the system under
consideration.
A second example further explicates this incongruity. A
large coal-fifed electrical power generating plant is to be built
in west central Utah. It will be located in a community whose
population has declined 18.9 percent since 1970 to 90 in-
habitants. The largest community within 25 miles has less
than 2,000 residents.
The consortium of power companies funding the project
bought most of the water rights in the area from local farmers
and ranchers. The U.S. Forest Service believes that the
move threatens the livestock industry and the traditional
agriculture of that part of Utah. Believing this to be
undesirable, the Service began a rangeland development
program in the mountains near the affected communities,
improving rangeland on Federal properties and convincing
private land owners to participate with the Forest Service
in improving theirs as well.
We could research this project by assessing the social con-
sequences on the assumption that the changes would fit a
natural-order process just as the range biologists assume for
botanical and zoological consequences of the same
rangeland improvements. In this natural-order view, the deci-
sion and control processes are external to the system. This
is the most common approach in social impact analyses and
fits nicely with the usual approaches in the discipline. But
not flawlessly.
A natural-order process implies decision-and-choice
located outside the system. It excludes a significant collec-
tion of activities from study—in particular, science and plan-
ning. Sociologists such as Smart (1982), Dawe (1979) and
57
-------
Lake Restoration, Protection and Management
Giddens (1979) have focused on the problem of recovering
the sociological subject's intentions, deelsionmaking
mechanisms and agency alongside 'the traditional
sociological assumption of society as a natural order. Social
impact analysis is a fertile area to pursue the problem
because of its close ties to actual, ongoing decision pro-
cesses. The remainder of this paper will develop hierarchy
theory as a conceptual mechanism to resolve this conflict.
HIERARCHY THEORY
One of positivism's most widely accepted and long-lived
legacies is the belief that all science is ideally part of a unified
whole. Positivism claims a fundamental continuity between
the phenomena of the social, biological and physical
sciences—each science dealing with an aspect of the natural
order in which concepts such as necessity, causation, and
scientific law are appropriate.
In contrast, several highly respected philosophers and
social scientists have suggested that there are several
discontinuities in nature that make the position untenable.
Perhaps the best developed collection of views of the discon-
tinuities has been labeled hierarchy theory. Persons who
have advocated various versions of hierarchy theory come
from philosophy (Bunge, 1977), natural sciences (Polanyi,
1969), and the social sciences (Simon, 1981 a; Boulding,
1956a).
This paper will discuss the contributions of three promi-
nent advocates of the theory—Boulding, Simon, and
Polanyi—and integrate their views into a summary statement
of hierarchy theory directly applicable to social impact
analysis and examine the methodological implications of us-
ing it.
Boulding's approach is basically taxonomic. In the Image
(Boulding, 1956a) he argues that there are several hierar-
chical levels. The first level is that of static structure and con-
sists of objects such as trees, houses, and planets. The sec-
ond level, the clockwork level, consists of entities that are
characterized by predetermined dynamic change in struc-
ture such as the solar system.
Level three consists of homeostatic control mechanisms:
thermostats. They are cybernetic devices with feedback
mechanisms. The cell or open system is a thermostat to
which has been added the capacity to maintain self and
reproduce kind. When cells come together to form societies
through the use of sexual reproduction and differentiation
of function, the plant level is achieved. The animal level dif-
fers from that of plants by the presence of mobility, sensory
awareness of the environment, sleep and wake states, and
learning mechanisms.
Humanity is an animal which is self conscious, rational,
capable of values and evaluation, symbolizing, imaging, and
building organizations. The human knows and knows that
he knows.
In another article published the same year, Boulding
(19565) added an eighth level, social organizations. He
believed that the level was difficult to separate from the
human level because the development of human-ness
depends on socialization through organizations. Social
organizations are structured around basic units which are
roles.
The taxonomic approach basically attempts to identify
hierarchic levels and argues that the significant differences
at each level require separate study.
Simon's work (1981 a) with hierarchy theory represents a
major advance over the taxonomic approach because it
begins to spell out some of the basic principles in hierarchy
theory. He assumes that hierarchies are common in all of
nature, found as cascading structures of atoms, cells, and
societies. He argues that they are common because they
speed change by allowing it to occur in selected subsystems
rather than in the total system.
One of the major features of a hierarchical system is its
near-decomposability: subsystems are distinguished because
there are strong relationships within, but only weak ties be-
tween them (Simon, 1981 a). Loose vertical coupling makes
it possible for the system to function without complete infor-
mation; only the subsystem events significant at the next
level are passed along. In addition, alternative subsystem
structures can often achieve the same goal for the system.
Loose horizontal coupling has two primary features: (1) short-
term behavior of subsystems is independent of the short-
term behavior of the other components; (2) the behavior of
any of the components is dependent on an aggregate
behavior of the other components (Simon, 1977).
Simon (1981b) also distinguishes between the sciences
of the natural and those of the artificial. He states that, "If
natural phenomena have an air of 'necessity' about them
in their subservience to natural law, artificial phenomena
have an air of 'contingency' in their malleability by environ-
ment." The basic problem for the scientist of artificial is to
show how empirical propositions are then possible. The study
of the artificial becomes the study of design; the discovery
of ways to accomplish a set of goals and purposes. Such
sciences of the artificial include engineering, business, and
painting. They may not ignore or violate natural law, but serve
human goals and purposes. When goals change, so do the
products. Thus, the contingency is partly due to the
malleability of goals.
Simon points out four major differences between the
natural and the artificial:
1. The artificial are synthesized by man,
2. The artificial may imitate nature, but will lack some
quality.
3. The artificial can be examined with respect to functions,
goals and adaptations, all of which are excluded in natural
science,
4. Artificials are discussed in terms of "oughts" as well
as "is's,"
Concerning human intelligence, Simon (1981b) adds that
only a small fraction of the alternative human behaviors are
"wired in." Examples are language, eating habits, dress,
marriage customs, and economic institutions. In all of these,
the important feature is that the science of the artificial does
not attain the necessity that is encountered in the natural.
Polanyi's formulation of hierarchy theory expands the
understanding of Simon's contributions by introducing two
closely related principles, the theory of boundary conditions
and the principle of dual control.
1. Boundary Conditions. Polanyi (1968) states that the
higher levels of life form a hierarchy, each level relying for
its workings on the principles of the levels below it but ir-
reducible to these lower principles.
Each hierarchical level's set of laws may not be violated
by the entities at other levels, but their operation is con-
trolled by principles from higher levels. For instance, the
operational principles that govern the construction of a
machine may not violate mechanical laws governing the
movement of the materials used to make it, but the principles
of the machine do either determine ways in which those
lower laws are activated, or they rely on the laws governing
the lower levels.
Polanyi (1968) warns in Science that "Irreducibility must
not be identified with the mere fact that the joining of parts
may produce features which are not observed in separate
parts...." In the Tacit Dimension (Polanyi, 1966) he refines
this principle: "A complete physical and chemical topography
of an object would not tell us whether it was a machine, and
if so, how it works and for what purposes." Any attempt to
account for a machine and its activities by relying on physical
58
-------
Socioeconomic Benefits of Lakes and Restoration
and chemical laws alone must fail, because of the impor-
tance of technology, purpose, and design.
Gelwick (1977) examines error and irreducibility in the con-
text of the rise of sentience, which is "irreducible" to
molecular action. Sentience is the first level where error
becomes possible. It makes no sense for a set of molecules
to be wrong. However, sensations may misrepresent the sub-
ject matter of sensation. Subsequently, each higher level
bears an increase in the chance for error and failure.
The argument for irreducibility turns on the presence of
something in each hierarchical level that is qualitatively dif-
ferent from the lower levels. Atoms, molecules, and rocks
do not err, and not even sentient creatures fail because
failure entails yet another hierarchy—consciousness, which
brings in purpose and goals.
2. Dual Control. Basically, dual control asserts that each
hierarchical level is subject to control in accordance with the
laws applying at its own level and in accordance to the laws
of the next higher level. Boundary conditions' role in dual
control is the means by which the higher principle harnesses
the lower one. This principle assumes a degree of openness
at each hierarchical level (Polanyi, 1969). As an example of
the principle the laws of phonetics do not determine utterance
because language uses all the phonetically possible sounds;
rules of word-forming also control utterance.
In many of his writings Polanyi indicates that the higher
level gives meaning to the lower level. For example, in talk-
ing of the relationship of raw materials to machines he states
that the physico-chemical properties represented the con-
ditions for the successful operation of the machine. They are
needed to detect the causes of breakdowns. However, the
principles of engineering at the machine level are basic.
Polanyi suggests that control also travels upward not simp-
ly as a constraint (Polanyi, 1968).
Possibly, as with evolution, the operation of the laws of
a lower level may not only produce conditions necessary for
the development of higher level principles, but also facilitate
the emergence of new higher level principles, or even new
higher levels.
The merging of dual control with evolutionary theory is
made possible by assuming that the control exercised by
the boundary conditions is a variable whose value changes
overtime and which may even be zero. The Coleman (1974)
writings on the emergence of corporate actors and the
organizational revolution may provide interesting examples
of this in recent human history.
HIERARCHY THEORY AND SOCIOLOGY
The components of hierarchy theory suggest ways in which
to understand the place of sociology within the sciences, the
place of social impact analysis within sociology, and the
methodological implications of such a view.
The typologies of nature suggest several continuities and
discontinuities between sociology and other sciences. A
general hierarchy is presented in Figure 1. Rifts are intro-
duced between some hierarchical levels because the
systems above the rift have characteristics and potentialities
so unlike those found below the rift that the lower systems
provide no reliable theoretical or methodological precedents
for those above it.
Rift 1 occurs as a consequence of the shift from inanimate
to animate objects of study. At least four major distinctions
emerge when the objects are animate. One is reversibility.
At the inanimate level a macro object can be decomposed
into its elements. The elements can then be recombined to
obtain another macro-object with the same properties. Above
rift 1 the reversibility is lost. An animal, or plant, that is
dissected and, consequently, killed cannot be brought back
Social
organizational
human
Rift 2
Biological
animal
plant
Rift 1
Physical
matter
(inanimate)
macro-world of large objects
micro-heisenberg's world
Figure 1.—General hierarchy.
to life simply by putting the dissected parts back in the "right"
place and adding some energy form.
A second feature is the introduction of morphogenesis at
the animate level. Monod (1971) argues that "a living be-
ing's structure results from a totally different process in that
it owes almost nothing to the action of outside forces, but
everything, from its overall shape to its tiniest detail, to mor-
phogenetic interactions within the object itself."
The third feature is reproduction—and their "ability .
to transmit ne varieturfae information corresponding to their
own structure" (Monod, 1971).
Finally, the major process is evolution whose theoretical
formulation is quite different from that of physical theories.
While the laws of physics and chemistry may be harnessed
to serve evolution, they do not determine the evolutionary
sequence. The lack of predictibility is not merely a range
of alternatives with a specifiable probability distribution: No
list of alternatives is attainable.
Rift 2 poses a set of significant differences between the
animate and human sciences. One difference is the degree
to which entities at the human level are preprogrammed or
"wired" at birth for the major activities they will engage in.
Simon (1981 b) stresses that while human beings are sub-
ject to both physical and biological laws, and limits, the lack
of wiring and preprogramming has significant implications.
For example, there are physiological limits to our memory
structure, especially in terms of the time it takes to move
information from short-term to long-term memory and the
limited capacity of short-term memory. However, these limits
themselves have been overcome by human invention (paper,
pencils and calculators). The development of learning
strategies, such as serialization, means ends analysis, and
simplification, are also devices for overcoming such limits.
Even within the human sciences the distinction between
discovering strategies to overcome limits and learning a
known repertoire of such strategies and aids poses major
differences in the nature of their respective sciences. Both
human discovery and human learning require a science that
is distinct in theory and method from the science of rat and
earthworm learning.
The development of strategies to achieve a desired goal,
the use of learning strategies, and the creation of organiza-
tional forms to achieve a specified goal all fit within Simon's
concept of science of the artificial. A science of the artificial,
either as the study of existing strategies or the creation of
new strategies, calls for approaches that do not fit the usual
hypothesis testing mode of natural science. The key dif-
ference is that causal necessity is replaced by empirical
possibility and situational appropriateness. Many of the most
highly valued classes of human activities (science, music,
art, sport, literature, business, and politics) are artificial in
Simon's sense.
59
-------
Lake Restoration, Protection and Management
A third difference between human and the other sciences
arises from the nature of dual control. At the nonhuman level,
a system is constrained by the laws of lower levels. But the
relationship is irreflexive in the sense that while laws govern-
ing atomic structure constrain the stable molecules that may
form, the laws governing molecular structure are inapplicable
to atomic structure. The hierarchy is either pyramidal or
linear. At the human level the hierarchy must be represented
as a closed or circular figure. Organizations and individuals
with their lack of preprogramming interlock so that rules
governing human activity constrain organizations while the
rules governing organizational activity simultaneously con-
strain individuals.
Finally, entities at the nonhuman levels do not change their
hierarchical level. Molecules are molecules. A molecule
system cannot change levels to become an atom or an organ
and still be a molecule system. However, human systems
may shift hierarchical levels and maintain system integrity.
Polanyi (1966) defines human hierarchical levels by the
possession of certain capacities such as perception, cen-
trally controlled motor activity, intellectual action, moral ac-
tion. This suggests that common human activities may
operate in alternative hierarchies. For instance, some peo-
ple, such as those raised in highly authoritarian societies,
in homogeneous communities, or in situations in which they
are never taught to reason, may be best understood by
theories developed for organisms at the animal level.
Behavioral psychology, a la George Homans, may be most
appropriate. However, a person taught to use strategies and
to understand conditioning patterns may cease to be a Homo
Homansis and move up a level to that of strategic actor.
Subsequently, the person may be taught not only to use
strategies, but to formulate them. In such a case, another
level is attained as he or she moves from strategizing man
to creative man. If this is the case, it may well be that disputes
between schools of sociological thought such as Homans'
followers and those of Habermas may not require resolu-
tion in favor of one or the other. They apply to human be-
ings and organizations at different hierarchical levels.
In summary, hierarchy theory provides some insights in-
to the continuities and discontinuities between sociology and
other sciences. The sociological enterprise must take into
account the constraints placed on people and organizations
by the laws regulating their subsystems. However, it must
also account for the openness and indeterminacy at their
own level. Such a conclusion poses serious implications for
the methodologies and theories appropriate for social im-
pact analysis; it obviously questions assumptions about
human society as a natural-order process.
METHODOLOGICAL IMPLICATIONS FOR
SOCIAL IMPACT ANALYSIS
If the discussion presented here is accurate in its statement
of differences between human systems and those located
below rift 2, then the person engaged in social impact
analysis must adopt a set of methodological strategies that
differ substantially from those taught in the majority of
methods texts available to the discipline. First, it is essen-
tial that the hierarchical level of the system of interest be
identified. Shifts in hierarchical level (within the human
sciences) produce changes in the appropriate
methodologies. A methodology that can deal with conditioned
learning is fundamentally inappropriate for the study of
strategy learning and selection. As the processes for learn-
ing strategies and selecting among alternative strategies is
made more and more conscious and systematic, the use
of natural science methods gives way to those of the
sciences of the artificial. Questions concerning maximiza-
tion, meaning, and choice replace questions concerning cor-
relation, causation, and natural law. The methodology that
is well adapted to modeling the causal determinants of quality
of life is incapable of producing an understanding of the ef-
forts of a group of cattlemen to organize and develop
strategies to protect their preferred lifestyle.
Second, the possibility that the system of interest may
change hierarchical levels during the process of the study
may require changes in methodology in the middle of an
impact analysis. For example, a rural community may func-
tion with relatively little use of systematic exploration of alter-
native futures, planning or efforts to discover effective
strategies to achieve community goals. A study of the im-
pacts of a Forest Service project may begin to evaluate ef-
fects on the community and systematically train the residents
to explore futures, plan to achieve desired futures and teach
skills in strategy development and implementation.
A third methodological consequence arises from the nature
of relationships between hierarchical levels at the human
level. Those levels characterized by a reflexive relationship
in which the principles of each serve as the boundary con-
ditions for the other must be treated very differently from
those representing the irreflexive relationship encountered
in most of the hierarchical levels encountered in science.
The theoretical and logical consequences require much more
elaboration than has been accomplished to this point, but
it seems reasonable to expect that they are substantial.
There are two specific applications of these implications
that indicate the magnitude of the changes in methodology
that are required as certain hierarchical levels are attained.
One of the principal illustrations of the shift in methodological
orientation required by the use of hierarchy theory comes
from the changing role of prediction. In the natural sciences
prediction plays a central role because failed predictions are
taken as signals that the theory being tested must either be
rejected or revised: it signals that there are inadequacies in
the theory. However, the failure of prediction in the sciences
of the artificial may signal any of a large number of possible
states, some of which have no implications for the adequacy
of the theory being used. One possibility is that the resear-
cher located the system at the wrong hierarchical level to
begin with. A second is that correct initial identification oc-
curred, but the system has changed levels. A third is that
the system is functioning on a level where prediction itself
is not a relevant criterion. The system is endowed with
mechanisms to create new strategies on structures which
are hoped to be better.
Jergen Habermas (1971) discusses a particularly signifi-
cant variation of the use of prediction in social science. In
discussing psychoanalysis, he argues that the therapist-
scientist may advance interpretations and predictions con-
cerning the actions of the person in therapy. The interpreta-
tion is not predictive unless the person in therapy accepts
it as accurate, makes it part of his or her beliefs, and uses
it to deal with repressed experiences. In other words, the
scientist makes predictions, but their value arises only in so
far as the object of investigation accepts them as accurate
and uses them in restructuring itself to make them come
true. Even then, the prediction may not come to pass
because of some failure in the implementation. Failure of
predictions to become true is not the criterion used to fault
the theoretical system. The possibilities are fascinating if the
social impact analyst is used in place of therapist and the
system replaces the person in therapy.
A second illustration of a methodological shift that is en-
couraged by hierarchy theory is a technique based on the
assumption that the process of interest (whether it is in-
dustrialization, a sudden flood, or rangeland development)
and the beliefs, values, and existing structure of the com-
munity, interact in such a way that there is a "logic" to subse-
60
-------
Socioeconomic Benefits of Lakes and Restoration
quent changes. For example, the process of industrializa-
tion requires a mobile work force that will move to the fac-
tory sites. There is a logic to the contention that when that
mobility is achieved, it will be impossible for the extended
family to retain its traditional importance. Given the demands
of industrialization, it is illogical that the extended family will
retain its traditional closeness. The "logic" replaces a tradi-
tional methodology which introduces control or comparison
groups or communities. The use of the "logic" is usually
argued for in those settings where control or comparison
communities cannot be obtained because each community
represents a collection of many diverse changes. It is ex-
tremely rare for any pair of communities to reflect the same
blend of energy development, well-being of agriculture, and
the traditional cultures. In addition, comparison communities
almost all represent some other unique blend of changes.
Such a situation makes the attribution of consequences of
any development based on comparisons highly tenuous.
Ultimately, the "logic" of the process becomes the basis of
impact attribution, not the empirical correlations encountered
in the data.
CONCLUSIONS
The implications of the version of hierarchy theory presented
here for social impact analysis are far-reaching. They ex-
plain the frustrations of the impact analyst described in the
opening paragraphs of the paper. The research assumes
a system on one hierarchical level. The use of the research
by the system assumes a completely distinct level with cor-
respondingly different governing principles.
The theoretical and methodological implications discuss-
ed in the body of the paper are, I hope, suggestive of ways
to begin to deal with the various levels. However, there are
many problems and obscurities that remain to be worked
out before the approach is more than suggestive. A typology
needs to be developed for both persons and organizations.
The criteria for dividing the subject matter into diverse hierar-
chies need to be explicated. The alternatives to prediction
and the use of "logics" of the processes need to be made
more explicit and precise. In summation, hierarchy theory
offers a promise for social impact analysis, but the details
and usefulness of the promise remain to be completed.
REFERENCES
Boulding, K. 1956a. The Image. Univ. Michigan Press, Ann Arbor.
. 1956b. General systems theory: the skeleton of science.
Manage. Sci. 2:197-208.
Bunge, M. 1977. General systems and holism. Pages 87-90 in Gen-
eral Systems. Vol. XXII.
Coleman, J. 1974. Power and the Structure of Society. Norton,
New York.
Dawe, A. 1979. Theories of social action. In Bottomore and Nisbet.
A History of Sociological Analysis. Heineman.
Gelwick, R. 1977. The Way of Discovery. Oxford Univ. Press,
New York.
Giddens, A. 1979. Central Problems in Social Theory. Macmillan,
New York.
Habermas, J. 1971. Knowledge and Human Interests. Beacon Press,
Boston.
Monod, J. 1971. On the molecular theory of evolution. Pages 11-24
in Rom Harri, ed. Problems of Scientific Revolutions. Clarendon
Press, Oxford.
Polanyi M. 1966. The Tacit Dimension. London: Routledge and
Kegan Paul, Ltd., London.
. 1967. Science and reality. Brj. Phil. Sci. 18:177-196.
. 1968. Life's irreducible structure. Science 1960:1308-1312.
1969. Knowing and Being. Univ. Chicago Press, Chicago.
Simon, H. 1977. Models of Discovery. D. Reidel Publishing, Boston.
_. 1981a. The Sciences of the Artificial. MIT Press, Cambridge,
Mass.
. 1981b. Studying human intelligence by creating artificial
intelligence. 69:300-309.
Smart, B. 1982. Foucault, sociology, and the problem of human
agency.
61
-------
KEYS TO LAKE WATER QUALITY: LAKE QUALITY STANDARDS
AND POINT/NONPOINT SOURCE ABATEMENT TRADEOFFS
ALFRED M. DUDA
ROBERT J. JOHNSON
Office of Natural Resources
Tennessee Valley Authority
Knoxville, Tennessee
ABSTRACT
During the last year, many issues of uncertainty have been raised about water quality man-
agement strategies in the United States. These issues directly affect efforts to restore acceptable quality
water to our Nation's lakes and to maintain lake-related public benefits for future generations. Based on
an examination of the nature of these uncertainties, several activities are identified as key tools for restor-
ing and protecting lake water quality. The development of scientifically defensible lake (or impoundment)
water quality criteria/standards is one of these activities. An example of such an approach adopted by
the State of North Carolina to resolve eutrophication problems is highlighted. The implementation of point/non-
point source abatement tradeoffs is the other key activity needed for protecting lake quality. Such deter-
minations of tradeoffs in controlling point and nonpoint source pollutants are essential for achieving cost-
effective improvements in lake quality.
INTRODUCTION
This year has been marked by uncertainty in the field of wa-
ter quality management in the United States. The Nation's
fundamental approach to pollution control was challenged
by proposed new approaches for establishing water quality
standards. Proposals to withdraw effluent limitations and
pretreatment regulations have been made by the U. S. En-
vironmental Protection Agency (EPA). Amendments to the
Clean Water Act were proposed, and drastic Federal fun-
ding cuts in Federal and State water quality programs made
them difficult, if not impossible, to implement. In particular,
proposed Federal and State budget cuts have seriously af-
fected lake restoration and protection efforts.
This uncertainty comes at a time when lake restoration
efforts should be accelerated, not stopped. The general
public is entitled to a whole host of socioeconomic benefits
from lakes—benefits that involve not only fishing, recreation,
and aesthetic enjoyment, but also public water supplies, in-
dustrial and agricultural water supplies, flood control,
wastewater assimilation, and economic development. But
these public benefits are now in jeopardy.
The U. S. Council on Environmental Quality noted that
more than 80 percent of urban lakes are degraded and in
need of restoration and fully one third of all lakes and im-
poundments are adversely impacted by pollution (Counc. En-
viron, dual., 1980). When lakes that are threatened by acidifi-
cation are included in this assessment, the national extent
of this pressing lake quality problem comes into focus. The
economic impact of these problems may be staggering. For
instance, in the Tennessee River Valley, eutrophication is
aggravating a low dissolved oxygen problem in lakes and
rivers of the Tennessee River system, posing a very real hin-
drance to economic development (Tenn. Valley Auth., 1980).
This paper examines the nature of some of the uncertain-
ties that have recently been raised about water quality man-
agement in the United States. It appears these uncertain-
ties are meant to challenge the course set by Congress
and to delay water quality improvement even longer. Bas-
ed on this examination, two activities are identified as key
tools for restoring and protecting the socioeconomic benefits
that the public is entitled to from our Nation's lakes. They
are the development of lake (or impoundment) water quali-
ty criteria/standards and the implementation of point/nonpoint
source pollution abatement tradeoffs. Legislation is not need-
ed to authorize these two key activities; they can be ac-
complished under existing provisions of the Clean Water Act.
HISTORICAL PERSPECTIVE
To better appreciate the significance of questions currently
being raised about water quality management strategies in
the United States, one must understand past pollution con-
trol efforts. The approach chosen by Congress in 1972 to
direct the Nation's water quality improvement effort resulted
from 40 years of experience in trying to solve pollution pro-
blems. Until 1948, State and local governments were primarily
responsible for pollution control. While numerous bills were
introduced into Congress in the 1930s to authorize Federal
pollution control planning and enforcement, the issues of
States' rights and financial impacts were used to defeat the
legislation.
In 1948 the first major Federal legislation was passed, but
it proved unworkable because State consent was required
before enforcement could be initiated. By 1965 Congress
became discontented with the slow pace of pollution con-
trol and passed the Water Quality Act of 1965. This com-
promise statute used violations of water quality standards
as the basis for enforcement, but States were hard pressed
to demonstrate in court that a certain discharger was respon-
sible for violations of standards. Very few States succeed-
ed with the water quality standards approach, and Congress
recognized that an effective nationwide approach would have
to involve national minimum effluent criteria. This new ap-
proach was established in the 1972 amendments (P. L.
92-500) and was fine-tuned in the 1977 Clean Water Act.
Consequently, it is because of this very slow progress over
the last half century that Congress mandated national mini-
mum standards for point source treatment (technology-based
effluent limitations and categorical pretreatment standards)
and nonpoint source controls (best management practices
(BMPs)). This strategy is designed so that economic advan-
62
-------
Socioeconomic Benefits of Lakes and Restoration
tages are not created among different industries, different
municipalities, or different regions of the Nation. And because
Congress specified a phased approach to pollution control,
many types of controls have not yet been implemented. Con-
ventional pollutants have been addressed and some pro-
gress has been made. Toxics and nonconventional polutants
(such as nutrients) remain to be addressed in the coming
years and are of utmost importance to lake quality
management.
TECHNICAL UNCERTAINTY
With the Clean Water Act scheduled for reauthorization dur-
ing 1982, special interest groups have raised numerous
technical questions about the scientific adequacy of water
quality management strategies in the United States. Among
these uncertainties are questions about: unnecessary treat-
ment associated with technology-based effluent standards;
scientific inadequacy of water quality criteria and standards;
the controllability of nonpoint source pollutants; and the lack
of cause/effect predictive relationships between pollution con-
trol and water quality improvement. Each issue merits a short
discussion.
once again raise uncertainty. A good example involves highly
eutrophic coastal rivers in eastern North Carolina that drain
extensive amounts of agricultural land. Agricultural groups
continue to raise issues of uncertainty about agricultural
nutrient inputs and question whether nutrient control
measures could really improve downstream water quality
(Gilliam et al. 1978; Humenik, 1980).lt took a very costly, in-
tensive monitoring investigation to produce scientifically
defensible data that implicated agricultural runoff as a ma-
jor contributor to the eutrophication problems in eastern North
Carolina (Duda, 1982).
It is clear that sufficient monitoring strategies need to be
employed to accurately assess the contribution of both point
source and nonpoint source pollutants. Unfortunately, ade-
quate water quality monitoring strategies are rarely used. In
fact, the U. S. General Accounting Office (1981) recently criti-
cized the fixed station/fixed sampling interval strategy used
virtually around the country as grossly insufficient for deter-
mining true water quality conditions. If inadequate monitor-
ing and analysis strategies continue to be used just because
they are less costly, the result will be the familiar old refrain
that more data are needed before widespread implementa-
tion of best management practices can be required.
Technology-based Standards
Some groups may point out that the effectiveness of categori-
cal technology-based standards for controlling toxics and
nonconventional pollutants is not well documented. They
may feel that implementation of such controls will result in
unnecessary expenditures and treatment for treatment's
sake. Since the need for using such treatment has not been
proven in every case across the country, the proposed solu-
tion is to remove the requirements and allow States to set
effluent limitations based on site-specific water quality con-
ditions and local needs.
Water Quality Standards
Other groups criticize Federal water quality criteria and the
resulting State standards for their lack of scientific adequacy.
Cox (1980) presented a spirited discussion of this view. The
criteria are alleged to be too stringent, too uncertain, too cost-
ly to meet, and based on too primitive a science (aquatic
toxicology). In addition, criticisms are leveled at the margin
of safety associated with the use of worst case conditions
and at laboratory development of criteria for use in ambient
waters (Lee et al. 1982). The proposed solution to such
uncertainty is once again a return to State flexibility to set
standards that reflect local interests, goals, needs, condi-
tions, etc., rather than using the best available data assembl-
ed on a nationwide basis. Other suggestions are being made
to ignore criteria and standards for protecting water-related
public benefits. In place of standards, state officials may be
given authority to subjectively judge whether or not uses of
public waters are "significantly impaired."
Nonpoint Source Pollution
In a recent study, the U. S. General Accounting Office (1982)
identified nonpoint source pollution as a major contributor
to the overall lack of significant water quality improvement
in the country. The findings noted that the extent of the non-
point pollution problem is unknown, data on its effects are
inadequate, and funding is now sadly lacking (since the ill-
fated section 208 planning has concluded).
Some groups logically ask why point source controls
should be implemented when nonpoint source controls are
voluntary in nature and, indeed, may overshadow point
source inputs. However, when controls are suggested for
nonpoint sources, others voice opposing arguments and
Cause and Effect Relationships
Closely related to the problem of inadequate data is an
imperfect knowledge of cause and effect relationships be-
tween pollution control measures and water quality im-
provement. While some professionals may believe that
such relationships must be thoroughly documented before
spending money for pollution controls, others may ques-
tion the government's ability to shoulder this burden. Some
may even question whether scientists can ever know the
complex ecosystem impacts that occur from wastewater
discharges.
A good example involves eutrophication in Cherokee
Reservoir, an impoundment of the Holston River in east
Tennessee. The Holston drains the most industrialized
area of Tennessee, and Cherokee Reservoir—with its
9,000 km2 drainage area—is located directly downstream
of many wastewater discharges. Numerous water quality
problems were noted in Cherokee Reservoir in a report
by the Tennessee Valley Authority (Iwanski et al. 1980):
interference with public water supplies from the reservoir
(taste/odor and iron/manganese); nuisance growths of
algae (up to 30 million cells per liter dominated by the blue-
greens Oscillatoria and Anacystis); fishkills; diseased fish;
and hypolimnetic oxygen depletion that resulted in dis-
solved oxygen below the 5 mg/l standard for more than
50 miles downstream from the reservoir.
Some professionals might point toward the water quality
improvement that has occurred in the Holston River over
the last decade (as controls for conventional pollutants
were implemented) as proof that very little further improve-
ment can be made. Figure 1 illustrates the reduction in
biochemical oxygen demand (BOD5), total nitrogen (TN),
and total phosphorus (TP) over a number of years in the
river as it flows into Cherokee Reservoir. Note that almost
70 percent reductions in BOD5 and 50 percent reductions
in TN and TP have occurred. Some improvement in hypo-
limnetic dissolved oxygen levels was noted for the days
with dissolved oxygen less than one mg/l. Despite these
improvements, the problems listed in the previous para-
graph still occurred.
The Tennessee Valley Authority undertook an intensive
investigation of the eutrophication problem in Cherokee
Reservoir. Point and nonpoint source nutrient loadings
were estimated (more than 50 percent of annual TP
loading was attributed to point sources) and a state-of-the-
63
-------
Lake Restoration, Protection and Management
58 60 62 64 66 68 70 72 74 76 78 79
58 60 62 64 66 68 70 72 74 76 78 79
Figure 1 .—Trends of water quality flowing into Cherokee Reservoir
and dissolved oxygen depletion in the reservoir releases.
art simulation model was used to predict improvement
following implementation of control measures (Iwanski et
al. 1980). Yet, control measures were not implemented '
because the modeling exercise only raised more questions
about cause and effect relationships: Is a two or three
dimensional model needed to improve predictions? While
history can be simulated using calibrated model coeffi-
cients and literature values for rate constraints, are predic-
tions of improvement following waste load reductions
valid? Do nutrient and organic matter inputs from exten-
sive upstream macrophyte beds greatly influence trophic
conditions? What influence do bottom-feeding fish or
zooplankton have on nutrient dynamics? What is the
significance of sediment oxygen demand, anaerobic pro-
cesses, or internal nutrient cycling?
There will always be questions about cause and effect
relationships because monitoring resources are limited,
as is our ability to understand natural processes. Conse-
quently, some professionals such as Beasley et al. (1982)
conclude that state-of-the-art simulation models should be
accepted for use in basing environmental management
decisions,
But simulation models have become more and more com-
plicated. While some modelers pile complex algorithm on
algorithm and rate constant upon rate constant in an attempt
to force complex models to mimic calibration data, they invar-
iably add a disclaimer proclaiming the model to be valid on-
ly for the specific modeled condition. If such models can't
be used for making predictions and decisions, they become
only useless intellectual toys as suggested by Baski (1979)
for entertaining intellectuals.
Based on uncertainties in nature, should pollution control
efforts stop until scientists agree on the usefulness of predic-
tive tools? Can different scientists ever agree on the scien-
tific acceptability of tools for predicting the impact of control
measures? In 1972, Congress decided that water quality
improvement could not wait until—or if ever—scientists
resolve such uncertainties. The public's use of the common
property resource known as water was in jeopardy, and the
Nation must act to restore these uses,
COMMON PROPERTY RESOURCES
Common property resources are those held in common by
the public—water, air, fisheries, and government-owned land
resources. It is government's job to protect the public's inter-
est in these resources. Competition in the free marketplace
ends up in exploitation of common property resources as
industry attempts to externalize production costs. It is pro-
fitable to pollute because individual control costs are usual-
ly much larger than the polluter's proportionate share of
readily quantifiable damage to common property.
A good example of this situation involves a Tennessee
Valley Authority impoundment on the Nolichucky River in
Tennessee. The reservoir has filled with sediment associated
primarily with feldspar and mica mining in North Carolina.
In addition to interfering with public water supply intakes and
treatment facilities, destroying game fishery and other aquatic
life in 110 miles of river, and Increasing flooding, the sedi-
ment forced abandonment of the electric generating facility
(Tenn. Valley Auth., 1981).
Not only has the public had to forego benefits of the river
and the lake, but they now must pay much more for replace-
ment electricity. Assuming that the hydropower facility gen-
erated 75 percent of the time, the cost of replacing this lost
electricity over 10 years with more expensive coal power
would be at least $15 million over the cost of hydropower;
actual replacement or dredging of the reservoir would cost
millions more.
This inevitable situation involving the abuse of common
property resources was termed the "Tragedy of the Com-
mons" by Garrett Hardin (1968), He likened the situation to
a "commons" or pasture used by a community for grazing.
In self interest, herdsmen would keep adding cattle to graze
on the commons until the numbers exceeded the carrying
capacity. As a result of each herdsman trying to maximize
his gain, the commons was overexploited and erosion and
weed dominance eventually ruined the resource.
This situation arises because real costs of producing a
product are not always being incorporated into the
marketplace and the external costs are being borne by the
public. Hardin concluded that the only way to avoid this
dilemma of abused resources was to enact laws so that there
would be "mutual coercion mutually agreed upon." When
laws are enacted, a danger still exists because enforcement
is delegated to bureaus that might not follow the intent of
the law. In the words of Hardin (1968):
The result is administrative law, which is rightly feared for an
ancient reason—Quis custodiet ipsos custodes?—"Who shall
watch the watchers themselves?" John Adams said that we must
have "a government of laws and not men." Bureau ad-
ministrators, trying to evaluate the morality of acts in the total
system, are singularly liable to corruption, producing a govern-
ment by men, not laws.
In passing the 1972 and 1977 clean water statutes, Con-
gress recognized that water quality is a common property
resource and that the public should not bear the social costs
and risks of polluted water. Congress also recognized that
the government (as custodian of this public property) can-
not bear the unreasonable, impossible burden of proving that
a certain source is responsible for a certain pollution problem,
Congress decided that the mutual coercion that Hardin
spoke of should take the form of national minimum pollu-
tion control standards for all sources for the first round of
water quality improvement. For those waters still dirty after
implementing the minimum controls, an additional levei of
control would be required based on violations of water quality
standards and on the benefits and costs of improvement,
These site-specific determinations would take place under
section 302 of the act, placing the burden of proof for deter-
mining that controls are not needed on dischargers.
While we engineers often pride ourselves on our model-
ing prowess and our wasteload allocations, we do not have
the resources, time, or technical knowledge to make sure
that just the right allocation of toxic chemicals or other
64
-------
Socioeconomic Benefits of Lakes and Restoration
pollutants is made all across the country. We have major
problems in determining actual impacts on individual species
let alone ecosystem impacts and we cannot even detect the
significance of many pollutants. Congress understood this
when it changed the course of the Nation's water pollution
control effort in 1972. The challenge facing water quality
management professionals today is to use the tools that Con-
gress has provided in a cost-effective manner to restore and
maintain public benefits from this common property resource.
KEYS TO LAKE QUALITY MANAGEMENT
The Clean Water Act contains many provisions that are es-
sential for restoring acceptable water quality to our Nation's
lakes. Certainly the promulgation of effluent guidelines for
toxic substances and nonconventional pollutants (such as
nutrients) in the coming years will improve the point source
abatement program. In addition, lake waters requiring con-
trols beyond technology-based limits to restore acceptable
quality will benefit from proper implementation of section 302
of the act. While these provisions may improve influent water
quality, section 314 Clean Lakes Program must continue to
be emphasized to deal with in-place pollutants that have
accumulated for decades and still influence lake quality. Ap-
propriate levels of Federal and State funding are needed to
accomplish these programs. Beyond these efforts, two ac-
tivities authorized by the Clean Water Act—but not effec-
tively implemented nationwide—are necessary for lake quali-
ty improvement. These key activities involve criteria/stan-
dards for protecting lake and impoundment water quality and
effective implementation of point/nonpoint source abatement
tradeoffs.
Lake Quality Standards
A recent survey by the National Governors' Association
(1982) found that almost 50 percent of all State water quali-
ty program budgets are supported by Federal funds. With
funding cuts proposed on both the Federal and State levels,
it is likely that priorities will have to be established for ad-
dressing water quality matters. While local residents may feel
their lake is degraded, unless a water quality standard has
been violated, such concerns may generate little government
attention to do something about the problem.
Criteria must be established to protect public uses of lakes
and impoundments. This will allow State water quality stan-
dards to be used to maintain clean waters and provide a
priority to address dirty waters. These criteria will vary across
the country, from natural lakes in Maine to those in Florida
and from impoundments in the West to those in the
Southeast.
Many States have not adopted the necessary criteria to
protect public uses in the variety of lakes and impoundments
that exist. Criteria developed for flowing waters may not be
appropriate for still waters. Likewise, criteria for natural lakes
may be inappropriate for impounded waters with less hydrau-
lic retention time. Since pollution problems are most often
biologically related, it makes sense that these criteria involve
biological parameters rather than just measurements of water
chemistry; aesthetic considerations may also be appropriate.
In the future, as funding resources diminish, authorities may
overlook water quality problems in lakes unless a water quali-
ty standard has been violated.
A good example of the use of such a standard for
determining water quality improvement priorities is the ap-
proach used by the State of North Carolina in dealing with
eutrophication. The State adopted a chlorophyll a water quali-
ty standard for lakes and reservoirs to indicate whether ac-
celerated eutrophication may be occurring. This standard is
a simple growing season maximum limit of 40 ng/\ for warm
waters and 15 ^g/l for cold waters. If this standard is exceed-
ed, the State begins an analysis of seasonal algal diversity
to determine whether desirable algae (indicative of a bal-
anced, highly productive system) or noxious algae (indicative
of nutrient overenrichment) predominate. If noxious algae
are impairing use, the entire watershed upstream of the pro-
blem water may be declared nutrient sensitive, and scien-
tific investigations conducted to determine major sources of
nutrient loading and to identify cost-effective abatement
actions.
The key to this approach is the adoption of the chlorophyll
a water quality standard so that the eutrophication problem
can be declared a priority problem. However, the State must
also have authority to require abatement of the problem. In
North Carolina, the establishment of a "nutrient sensitive wa-
ters" classification gives broad authority to the Environmental
Management Commission to correct nutrient-related pro-
blems. This regulation (included as Appendix A to this paper)
is activated by a violation of the chlorophyll a standard and
provides a means for controlling both point source and non-
point source inputs of nutrients in a nutrient sensitive river
basin.
While some professionals may agree that chlorophyll a
contents of different types of algae vary widely (and therefore
make it an inappropriate tool), the criteria value of 40 fjg/l
represents a very large bloom of algae that would certainly
merit further investigation. This practical approach is quite
simplistic, but it is also scientifically defensible because of
the three-step process involved: a violation of a standard oc-
curs, a scientific investigation into the significance of the pro-
blem is conducted, and if evidence merits it the waters are
declared nutrient sensitive.
The establishment of a lake quality standard is a key not
only for dealing with eutrophication but also for other lake
quality problems. Lakes are more sensitive to accumulation
of toxic substances than are flowing waters. Lake quality
standards, based on criteria that may differ from those in
rivers, will be necessary to avoid bioaccumulation of toxic
substances similar to the accumulation of selenium in Belews
Lake, N. C., that decimated the resident fish population
(Cumbie, 1978). A cooperative effort between the States,
EPA, and the research community will be needed to establish
such criteria for different types of lakes and impoundments
on a regional basis in order to pool resources and eliminate
duplication of effort. It is up to those professionals interested
in restoring and maintaining public benefits from lakes and
impoundments to see to it that these criteria and subsequent
standards are established.
Point/Nonpoint Source Abatement Tradeoffs
Once violations of a water quality standard occur, threaten-
ing or impairing the use of lake waters, the key to cost-
effective water quality improvement is the implementation
of point source and/or nonpoint source control measures
necessary to correct the problem. Determination of these
tradeoffs is often an exceedingly difficult task. Section 208
of the Clean Water Act contains the provisions for determin-
ing tradeoffs between point source and nonpoint source
abatement strategies. Unfortunately, the 208 planning pro-
cess evolved into something different than an examination
of tradeoffs; for the reasons explained by Duda et al. (1982)
the program did not fulfill its expectations and funding has
been eliminated.
It is clear that such tradeoffs need to be considered to
provide cost-effective water quality improvement. A good
example involves a highly eutrophic reservoir in northern
Virginia, the Occoquan Reservoir. An extensive monitoring
study was undertaken to determine the significance of dif-
ferent nutrient sources in the watershed. Randall et al. (1978)
65
-------
Lake Restoration, Protection and Management
reported the investigation found only 5 percent of annual
nitrogen loading and 7 percent of annual phosphorus loading
to the reservoir came from point source discharges. In this
case, strategies to control nonpoint source contributions
would be most appropriate.
Obviously, a strong monitoring program is essential to ac-
complish such a determination. In addition, the monitoring,
analysis, and simulation investigations should be conducted
in cooperation with major dischargers and special interest
groups in the area. Whatever techniques are employed, they
need to be discussed with all parties before the investiga-
tions begin to ensure that state-of-the-art techniques are used
and to resolve reasonable questions of uncertainty.
An example of the importance of implementing point/non-
point source abatement tradeoffs can be seen with the eutro-
phication problem in the Chowan River basin of North
Carolina. While an extensive description of the problem has
been given elsewhere (Duda, 1982), a short discussion will
be included here. For the last decade noxious blooms of
blue-green algae have depressed recreation activities,
caused odor problems, and been suspected of contributing
to a drastic decline in commercial fishing in slow-moving
waters of the basin. Fish populations have been devastated
by both red sore disease and fishkills from oxygen deple-
tion following bloom collapse. Aphanizomenon f/os-aquae,
Microcystis firma, and Anabaena circina/is are the blue-green
algae of concern. While the State of North Carolina water
quality standard for chlorophyll a (40 /xj/l) is commonly ex-
ceeded during the growing season, massive blooms with
chlorophyll a values exceeding 200 \JQ!\ are not uncommon.
Extensive water quality investigations as described by
Duda (1982) were conducted in the late 1970s. While 130
point source dischargers were identified in the 13,000 km2
watershed of Virginia and North Carolina, the basin is mostly
rural-agricultural in nature and only four of the point sources
discharge more than one million gallons a day. The total load-
ing of nutrients in the main river was determined as was the
contribution of point source dischargers. The result is given
in Figure 2 in terms of average daily loading of phosphorus
and nitrogen. For the 1979 water year, point source dis-
SNJMMJ SNJM
Figure 2.—Average daily loading of nutrients to the Chowan River
at Holiday Island on a monthly basis.
chargers were estimated to contribute only 3 percent of total
nitrogen loading and 6 percent of total phosphorus loading.
Concurrent with determination of river nutrient loading, au-
tomated samplers were placed in small agricultural water-
sheds (1.6-13 km2 in area) to provide estimates of relative
significance of nutrient contributions from agricultural water-
sheds in different physiographic areas, with different amounts
of livestock and different drainage characteristics. Nutrient
levels in these watersheds were compared with nutrient
levels monitored in a concurrent study by the U. S.
Geological Survey of forested watersheds in North Carolina
(Simmons and Heath, 1979). Once again, details of the com-
parison may befound elsewhere (Duda, 1982).
One way to represent the results is by plotting the varia-
tion of average stormflow nutrient concentration recorded
in the watershed with the number of livestock in the water-
shed. Figure 3 gives an example of such a relationship for
STORM-EVENT
C - Coaatal Plain watafahad
P - PMmont watarahad
(£)- ChaiwiaH»d watershed
F - FofaaUd waterahada
01 10 100 1000
ESTIMATED LIVESTOCK DENSITY, kg/ha
Figure 3.—Variation of mean stormflow total phosphorus concen-
trations with estimated livestock density recorded in small North
Carolina agricultural watersheds.
total phosphorus. Note that as more livestock are present,
average concentrations greatly increase over levels record-
ed in forested watersheds. In addition, much greater con-
centrations arefound in the watershed that has extensive
drainage improvements such as channelization.
This exercise for determining tradeoffs in controlling point
and nonpoint source pollutants was very valuable. It suggest-
ed that agricultural activities contributed much to the
eutrophication problem. To reduce nutrient loading, exten-
sive point source controls would not be effective. Rather,
nutrient control measures for agricultural activities would be
more appropriate and should be implemented in the follow-
ing priority: controls for livestock waste, controls for nutrients
from cropland in watersheds with extensive drainage im-
provements, and controls for nutrients from cropland drain-
ing directly to watercourses (without buffer zones).
Implementation of those control measures should first be
attempted through voluntary means. Multiagency,
cooperative projects are appropriate for seeking the use of
such control measures on a watershed-wide basis. If these
efforts prove futile, State and Federal regulatory actions
should be taken under section 302 of the Clean Water Act
or under State authority (such as the nutrient sensitive waters
regulation). The challenge for regulatory agencies is to use
the tools they already have to achieve cost-effective cleanup
and protection of our Nation's lakes.
CONCLUSIONS
Improving lake water quality will be a challenging task in the
years ahead. Professionals will face crippling budget cuts,
changes in regulations, and the nagging cry of technical un-
certainty from opponents of clean water. Attempts will be
made to ignore criteria and standards and to replace them
with subjective judgements whether uses of public waters
are significantly impaired. While changes in legislation are
called for in the name of economic and procedural efficien-
cy, and scientific uncertainty, many professionals recognize
66
-------
Socioeconomic Benefits of Lakes and Restoration
that such reasons have been used in the past to delay water
quality improvement. By enacting the 1972 statute, Congress
rejected the old vagaries of administrative law that had
failed to protect the public's common property water
resources. Instead, Congress established a local-State-
Federal partnership that is quite cumbersome and quite com-
plex, yet it represents the only strategy that will work to clean
up public waters. The existing Clean Water Act is workable
and necessary; it is essential to public welfare and to the
restoration and protection of suitable lake quality.
The real challenge facing those concerned about lake
quality is to ensure that the tools Congress has provided to
cost-effectively restore and protect lake waters are properly
used by water quality management professionals. This paper
has identified two of these key tools: lake quality criteria/stan-
dards and point/nonpoint source abatement tradeoffs.
With pending budget cuts, priorities will be set in dealing
with pollution problems. Without sufficient criteria and stan-
dards to protect the use of lake waters, public complaints
may fall on deaf ears. A nationwide cooperative effort be-
tween the States, EPA, and the regulated community is
needed to establish defensible standards on a regional basis
to protect public uses of different types of lakes and im-
poundments.
Once a water quality standard is violated and use of lake
waters is threatened or impaired, the key to cost-effective
water quality improvement is the implementation of point
source and/or nonpoint source control measures necessary
to correct the problem. While this is often a difficult task, it
is not impossible if sufficient resources are devoted to col-
lecting the needed information and to conducting scientifically
defensible analyses. An example was given to illustrate how
such tradeoffs were determined for a eutrophication problem
in North Carolina.
For implementation of these tradeoffs in control strategies,
point source controls can be readily specified under section
402 of the Clean Water Act. If nonpoint source controls are
needed, many States do not have appropriate authority to
require implementation. In this case, multiagency,
cooperative projects involving all levels of government and
the private sector need to be undertaken in the watershed
upstream of the lake or impoundment to address the
pollutants of concern. Those interested in protecting lake
quality need to assure that sufficient funds are appropriated
to support such cooperative, watershed-wide projects, or
alternatively that authority is established to require implemen-
tation of needed controls.
REFERENCES
Baski, H. A. 1979. Ground-water computer models—intellectual toys.
Ground Water 17:177.
Beasley, D. B., L. F. Muggins, and E. J. Monke. 1982. A monitoring/
modeling strategy for 208 implementation. Trans. Am. Soc. Agric.
25:654.
Council on Environmental Quality. 1980. Environmental quality, 11th
annual report. Washington, D. C.
Cox, G. V., 1980. Water quality criteria and standards—a conflict be-
tween science and policy. Pages 73-84 in Proc. Symp. Dev., Use,
and Value of Water Qual. Criteria and Stand. U. S. Environ. Prat.
Agency, Washington, D. C.
Cumbie, P. M. 1978. Belews Lake environmental study report: se-
lenium and arsenic accumulation. Duke Power Co., Charlotte, N.
C.
Duda, A. M. 1982. Municipal point source and agricultural nonpoint
source contributions to coastal eutrophication. Water Resour. Bull.
18:397.
Duda, A. M., D. R. Lenat, and D. L. Penrose. 1982. Water quality
in urban streams—what we can expect. J. Water Pollut. Control
Fed. 54:1139.
Gilliam, J. W., R. W. Skaggs, and S. B. Weed. 1978. An evaluation
of the potential for using drainage control to reduce nitrate loss
from agricultural fields to surface waters. Rep. No. 128, N. C.
Water Resour. Res. Inst., Raleigh, N. C.
Hardin, G. 1968. The tragedy of the commons. Science 162:1243.
Humenik, F. J., 1980. Relationships between agricultural practices
and receiving water quality. Pages 249-256 in Proc. Int. Symp.
Inland Waters and Lake Restoration. U. S. Environ. Prot. Agen-
cy, Washington, D. C.
Iwanski, M. L., J. M. Higgins, B. R. Kim, and R. C. Young. 1980.
Factors affecting water quality in Cherokee Reservoir. Tenn. Valley
Auth., Chattanooga, Tenn.
Lee, G. F., R. A. Jones, and B. W. Newbry. 1982. Water quality
standards and water quality. J. Water Pollut. Control Fed. 54:1131
National Governors' Association. 1982. The state of the States:
management of environmental programs in the 1980s.
Washington, D. C.
Randall, C. W., T. J. Grizzard, and R. C. Hoehn. 1978. Effect of up-
stream control on a water supply reservoir. J. Water Pollut. Con-
trol Fed. 50:2687.
Simmons, C. E., and R. C. Heath. 1979. Water quality characteristics
of streams in forested and rural areas of North Carolina. Water
Resour. Invest. 79-108. U. S. Geological Survey, Raleigh, N. C.
Tennessee Valley Authority. 1980. Is the water getting cleaner? Off.
Nat. Resour., Chattanooga, Tenn.
1981. Surface mining and sedimentation. Impact—TVA
Natural Resources and the Environment 4:2. Off. Nat. Resour.,
Chattanooga, Tenn.
U. S. General Accounting Office. 1981. Better monitoring techniques
are needed to assess the quality of rivers and streams. CED-81-30.
Washington, D. C.
_. 1982. Environmental protection: agenda for the 1980s.
CED-82-73. Washington, D. C.
APPENDIX 1
Regulation 15 North Carolina Administrative Code 2B .0214;
NUTRIENT SENSITIVE WATERS; has been adopted and
reads as follows:
.0214 NUTRIENT SENSITIVE WATERS
(a) In addition to existing classifications, the commission
may classify any surface waters of the State as nutrient sen-
sitive waters (NSW) upon a finding that such waters are ex-
periencing or are subject to excessive growths of microscopic
or macroscopic vegetation. Excessive growths are growths
which the commission in its discretion finds to substantially
impair the use of the water for its best usage as determined
by the classification applied to such waters.
(b) NSW may include any or all waters within a particular
river basin as the commission deems necessary to effec-
tively control excessive growths of microscopic or
macroscopic vegetation.
(c) For the purpose of this rule, the term "nutrients" shall
mean phosphorous and/or nitrogen. When considering the
assignment of this classification the commission may specify
as a "nutrient" any other chemical parameter or combina-
tion of parameters which it determines to be essential for
the growth of microscopic and macroscopic vegetation.
(d) Those waters additionally classified as nutrient sen-
sitive shall be identified in the appropriate schedule of
classifications as referenced in Section .0300 of this
Subchapter.
67
-------
Lake Restoration, Protection and Management
(e) For the purpose of this rule the term "background (1) is the result of natural variations, or
levels" shall mean the concentration(s), taking into account (2) will not endanger human health, safety or welfare
seasonal variations, of the specific nutrient or nutrients and that preventing the increase would cause a
upstream of a nutrient source. serious economic hardship without equal or greater
(f) Quality Standards applicable to NSW: No increase in benefit to the public.
nutrients over background levels unless it is shown to the
satisfaction of the director that the increase Historical Note: statutory Authority, G. s. 14*214.1; Eft. way 10,1979.
68
-------
REHABILITATION PLAN FOR LAKE CAJITITLAN:
AN ENDANGERED SHALLOW LAKE
J. G. LIMON M.
J. J. AMEZCUA C.
V. BASTIDAS R.
General Directorate for Water Uses
and Pollution Prevention
Secretariat for Agriculture and
Hydraulic Resources.
Guadalajara, Jal, Mexico
ABSTRACT
Lake Cajititlan Is a shallow lake (volume = 45, 6 x iQ8m3, mean depth = 2.2 m) located in the central
part of the Mexican state of Jalisco. In recent years, this lake has experienced serious problems. A rehabilita-
tion plan has been prepared by the Secretariat of Agriculture and Hydraulic Resources (SARH) following
the request of the government of the State of Jalisco. The lake has shown low water levels, waterweed
infestation (mostly £ crassipss and Typha sp.) and water quality deterioration. Additionally, sedimentation
may be significant. The rehabilitation project proposes water import from a neighboring watershed. It also
discusses the value of mechanical control of water hyacinth and several water quality policies, both in
the short and in the long term.
INTRODUCTION
Lake Cajititlan is the second largest lake in the Mexican state
of Jalisco. This freshwater body has experienced severe pro-
blems in recent years. Following a request from the govern-
ment of the State of Jalisco, the Secretariat for Agriculture
and Hydraulic Resources has prepared a general plan to
preserve and maximize use of the lake.
This paper presents the available information on the main
features of both the lake and its watershed. It also identifies
its main problems, i.e., water deficit, aquatic plant infesta-
tion, and water quality deficiencies. Sedimentation might also
be significant.
The restoration measures are briefly discussed and the
derived benefits outlined.
DESCRIPTION OF THE AREA
Lake Cajititlan is located in the central part of Jalisco (see
Fig, 1), 33 km south of Guadalajara. Table 1 includes signifi-
cant features of the area, such as location, morphology,
climatology, water balance, land use, and vegetation. From
its characteristics, Lake Cajititlan can be considered a
shallow, subtropical lake.
Rain falls in the area from early June to late September,
and the weather remains mostly dry the rest of the year.
From Table 1, it can be seen that Lake Cajititlan is quite
shallow, with a capacity of 45.6 x io6m3, corresponding to
an area of 2,047 ha and a mean depth of 2.2 m.
The present population of the watershed is estimated to
be 14,795 inhabitants, with 50 percent of them living in
CANAL EL GUAWBO LA CALERA
Watershed lake Cajititlan
186.5 Km*
P. U)S 8NMNTES / STA. HAHIA
CUESCOKOTITl*
CEDROS
OERIVADORA CEOROS
Watershed arrovo los sablnos
ares 107.62 km*
Figure 1.—Location of Lake Cajititlan.
69
-------
Lake Restoration, Protection and Management
Table 1. — General features of Lake Cajititlan.
A. Location
latitude
longitude
elevation
B. Morphology1
volume
area
shoreline
maximum
maximum
length
length
width
mean depth
maximum depth
20°25'
103°20'
1551.5
45.6 x 106
2,047
23.5
9.8
3.2
2.2
3.5
N
W
masl
m3
ha
km
km
km
m
m
C. Climatology (annual average, 1972-1980)2
rainfall 828
evaporation 1,564
air temperature 20.5
D. Water balance (annual average, 1972-1980)2
total inflow 25,705.5
rain 10,268.3
others 15,438.2
total outflow 29,550.3
evaporation 21,142.8
ext. for irrigation 5,078.4
others 3,329.1
E. Watershed: population, land use, and vegetation3'4
watershed area 186.5
watershed population 14,795
forest land use
agricultural land:
crops
agricultural land:
pasture and grazing
shrub-like vegetation
urban and semi
urban
13.9
77.9
57.4
34.2
3.1
mm
mm
°C
103m3
103m3
103m3
103m3
103m3
103m3
103m3
km2
inha-
b.
km2
km2
km2
km2
km2
1 Secretar. Recur. Hidral., 1948
2Div. Hidromet. 1981
3Dep. Program. Desar., 1982
• DETENAL. 1978
Tlajomulco de Zuniga (Dep. de Program, y Desar., 1982).
The villages located near the lake have a total population
of 6,769 inhabitants. The economic activities are mostly ag-
riculture and cattle breeding.
The land uses of the watershed are mostly agricultural,
with 41.8 percent used for crops and 30.8 percent for pasture
and grazing. The remaining portion consists of shrub-like
vegetation (18.3 percent), forest (7.5 percent), and urban and
semi-urban areas (1.6 percent) (DETENAL, 1978).
In the past years the uses of this lake included commer-
cial fisheries, irrigation of 2,500 ha, aquatic sports, and
recreation. However, these beneficial uses have been reduc-
ed to a minimum level by the problems affecting Lake
Cajititlan.
A recent limnological survey (Centra Estud. Limnol. 1982)
has revealed high concentrations of total inorganic
phosphorus (0.494 mg/l as P) and total inorganic nitrogen
(0.424 mg/l as N). It has also shown low Secchi disc
transparency (0.13 m), slightly alkaline pH value (8.6), and
moderately high total dissolved solids (868 mg/l).
During the period of the survey, the water volume de-
creased from 12.4 x I06m3 late in September 1981 to only
0.9 x I06m3 by the end of June 1982. The water loss
resulted from evaporation and evapotranspiration. This was
reflected by an increase in total dissolved solids, from 455
mg/l to 1,486 mg/l in the same period.
The algal plankton community, as evaluated by net
plankton (mesh 52 ^m), is dominated by Pediastrum sp.,
followed by Euglena sp., Phacus sp. and Ceratium hirudinella,
in this order. The most abundant zooplanktonic organisms
are Codonella cratera and n.i. copepod larvae.
The aquatic vascular plant community is mainly formed
by Eichhomia crassipes and Typha sp. Both plants show
levels so high they affect beneficial uses of the freshwater
body. This subject is discussed later.
IDENTIFICATION OF PROBLEMS
Available information suggests that the three most signifi-
cant problems of this lake are water deficit, aquatic
macrophyte infestation, and water quality deficiencies. The
significance of sedimentation needs to be studied further.
Water Deficit
The water balance, or rather the water imbalance, is il-
lustrated in Figure 2, where a marked trend toward a
decrease in stored capacity is shown.
i
^ ao
Figure 2.—Variation in maximum (—) and minimum (- - -) stored
volume in Lake Cajititlan, 1972-82.
The water entering Lake Cajititlan comes from direct rain-
fall over the lake surface, runoff from the lake watershed,
and from Canal Cedros. This canal collects water from a
neighboring watershed (arroyo los Sabinos) and brings into
the lake the "surplus" volume after having filled a system
of small dams (el Aniego, el Carnero, San Francisco y la
Arena).
The present water deficit results from several factors:
1. Reduction in rainfall since the 1976-77 cycle.
2. Insufficient derivation works at Cedros, which
leads to lower volumes exported to Cajititlan.
3. Reduced carrying capacity, both of Canal
Cedros and of the three canals of the west
side of the lake.
To adequately preserve Lake Cajititlan it is necessary to
guarantee sufficient volumes of water entering the lake.
Vascular Aquatic Plant Infestation
Water hyacinth (E. crassipes) and cattail (Typha sp.) have
proliferated during the last few years, causing problems for
recreation and fisheries.
According to its distribution, E. crassipes can be classified
as compact, free-floating, and marginal. Table 2 shows the
wet density, covered area and wet biomass of these three
forms of water hyacinth. At the time of the determination the
total area covered was 355 ha, representing a wet biomass
of ~ 100,000 tons. The plant size was between 0.62 and 0.96
70
-------
Table 2. — Features of Lake Cajititlan coverage by water
hyacinth (August 4, 1981).
Distribution
Feature
Free
Compact floating Marginal Total
Density, ton/ha
Area covered, ha
Wet biomass, ton
322
282
91,061
230
30
6,949
45
43
1,961
—
355
99,971
m. This information is consistent with data recently reported
in the literature (O'Brien, 1981).
The water hyacinth infestation impairs the ecosystem pro-
cesses because water hyacinth plants shade lake water;
evapotranspiration increases water losses at the air-water
interface; and old plant roots and leaves sink to the bottom.
On the other hand, Typha sp. covers 60 percent of the
lake shoreline.
Water Quality Deficiencies
The main water quality parameters of the lake have been
determined, and the corresponding values are presented in
Table 3, which also includes the values suggested by the
legislation in force (Reg. Prev. Control Contam. Aguas, 1973),
according to the lake uses.
From an examination of Table 3 it can be concluded that
the coliform bacteria values do not fulfill the quality standards.
Moreover, the nutrient concentrations are so high they pro-
duce hyperfertilization, thereby not fulfilling the requirements.
The heavy metals concentrations are adequate.
P and N concentrations indicate highly eutrophic condi-
tions for Lake Cajititlan, according to international informa-
tion coming predominantly from temperate lakes
(Vollenweider and Kerekes, 1981; Jones and Lee, 1982).
GENERAL MEASURES
The restoration plan for Lake Cajititlan has been orientated
towards its conservation and optimal use. This could be
achieved by these goals:
Socioeconomic Benefits of Lakes and Restoration
1. Guarantee of an adequate water balance that main-
tains suitable ecologic conditions.
2. Development of a program of aquatic plant
management.
3. Implementation of water quality policies in the water-
shed, and particularly in the villages.
4. Elaboration of a program for promoting recreation and
tourism in the area.
Water Quantity
During the last few years, the stored volume in the lake has
progressively decreased. If the lake is to be preserved, it is
necessary to change this tendency. This can be done by
importing water from a neighboring watershed, los Sabinos,
with an area of nearly 108 ha. According to the topography,
there is no need for pumping. This watershed can supply
from 11.7 x I06m3 to 21.7 x I06m3 per year depending on
the rainfall. A small derivation dam and a canal already ex-
ist there. To fully use this water in Cajititlan, it is necessary
to remodel the dam, to continue the existing canal for 2 km
up to the lake, and to clean the canal. These works repre-
sent priority for conserving the lake.
Aquatic Macrophyte Management
The proposed approach is to use mechanical control to
reduce the existing hyacinth biomass, and subsequently to
apply biological control methods to keep the remaining
biomass at levels acceptable to such uses as recreation and
fisheries.
In a given water body, with an area (A, ha) covered by
water hyacinth and under a rate of harvesting (R, ha/d), the
change in area covered with respect to time can be ex-
pressed as:
dt
R
(1)
where k is the specific growth rate of water hyacinth at
the lake conditions.
Equation 1 can be integrated, assuming that at t = 0,
- A = A0
A = £ (1 - ekt) +
(2)
Table 3. — Water quality characteristics of Lake Cajititlan
Parameter
pH
Dissolved oxygen
(mg/l)
Coliform bacteria
(NMP/100ml)
Range
7.6-9.0
2.7-6.8
11-750,000
Mean1
8.6
5.5
1.3263
Suggested value2
6.0-9.0
^4.0
mean 10,000 total
no 20,000 total
3-750,000
2683
mean 2,000 faecal
no 4,000 faecal
Grease and oil
(mg/l)
Turbidity (J.T.U.)
Nutrients
Tot inorg N(mg/l)
Tot inorg P(mg/l)
Heavy metals (mg/l)
Cr6+
Hg
Cu
Absence of visible
film
25-700
0.08-1.18
0.19-1.9
<0.01
< 0.005
•< 0.05-0.30
Absence of visible
film
195
Hyperfertilized
0.424
0.494
<0.01
< 0.005
<0.05
Absence of visible
film
Natural conditions
Not hyperfertilized
<0.1
<0.01
<0.1
1 Arithmetic mean Irom 16 sampling visits Irom Sept. 1981 to Sept. 1982. unless otherwise stated.
•' Class D II recreation, conservation ol flora and fauna, and industrial uses
J Log mean Irom 16 sampling visits
71
-------
Lake Restoration, Protection and Management
Figure 3.—Estimation of the time of the water hyacinth-covered area
in Lake Cajititlan under a harvesting program, with AQ = 355 ha,
k = 0.006 d~1 and R = 2.5 ha/d.
Figure 3 shows the variation in area covered under a pro-
gram of mechanical control (R = 2.5 ha/d) and assuming
k = 0.006 d~1. It is estimated that the cleaning period
would be nearly 300 days.
Mechanical harvesting of aquatic plants implies dispos-
ing of large amounts of collected plants. However, this cost
can be reduced by using the plants as a soil conditioner (Natl.
Acad. Sci., 1976) in the nearby agricultural land.
Water Quality Management
As discussed earlier, the significant water quality problems
are fecal pollution and hyperfertilization.
The most important pollution sources in the watershed in-
clude wastewater from the villages (mainly Tlajomulco),
agricultural solid wastes that can be eventually taken to the
lake, and surface runoff from agricultural land. The first one
includes point sources, whereas the last two are nonpoint
sources.
The immediate water quality management measures are
described in the following paragraphs, and the detailed long-
term policies require a complete nutrients balance, which
will be carried out in the near future.
It is considered that fecal pollution can be controlled by
adequately disposing of cattle solid wastes and secondary
treatment of the wastewater from the larger villages.
An aerated lagoons system is under construction at Tla-
jomulco. The process consists of screening, aereation ponds,
sedimentation ponds, and disinfection. Further, the nutrient
load can be significantly reduced at a reasonable cost by
biofiltration using water hyacinths and reusing the effluent
for irrigation.
The sewerage collection system of the village of Cajititlan
is almost complete. To avoid affecting the lake, secondary
wastewater treatment and reuse for irrigation are recom-
mended. Alternatively, exportation of this effluent could be
considered.
Socioeconomic Benefits
Lake Cajititlan is the nearest lake to the Guadalajara
metropolitan area and its nearly 2.75 * 106 inhabitants. In
addition to conserving this lake, the restoration program
would produce a recreation zone quite attractive to the fast-
growing city of Guadalajara. Moreover, this would increase
land property value considerably, meet the water quota for
irrigation, and restore the local fisheries.
The restoration program would need the cooperation of
several public and private agencies. This important aspect
is currently being evaluated.
REFERENCES
Centra de Estudios Limnologicos. 1982. Limnological survey of lake
Cajititlan. Secretaria de Agricultura y Recursos Hidraulicos,
Representacion General en el Esta do de Jalisco, Mexico. (Rep.
in prepar).
Departamento de Programacion y Desarrollo. 1982. Cedula de in-
formacion municipal: Tlajomulco de Zuniga. Guadalajara, Jal.
Mexico.
DETENAL. 1978. Cartas de uso de Suelo, F13-D-75 y 76. Mexico,
D. F. Mexico.
Division Hidrometrica. 1981. Registro de balance de agua laguna
de Cajititlan. Secretaria de Agricultura y Recursos Hidraulicos,
Representacion General en el Esta do de Jalisco. Mexico.
Jones, R. A., and G. F. Lee. 1982. Recent advances in assessing
impact of phosphorus loads on eutrophication related water quality.
Water Res. 16:503-515.
National Academy of Sciences. 1976. Making aquatic weeds useful:
some perspectives for developing countries. Washington, D. C..
O'Brien, W. J. 1981. Use of aquatic macrophytes for waste water
treatment. J. Environ. Div. Proc. Am. Soc. Civil Eng. 107: 681-698.
Reglamento para la Prevencion y Control de la Contaminacion de
las Aguas. 1973. Mexico, D. F. Mexico.
Secretaria de Recursos Hidraulicos. 1948. Levantamiento batimetrico
Laguna Cajititlan. Direccion General de Estudios y Proyectos.
Distrito de Riego del Bajo Lerma, Jalisco y Michoacan. Mexico.
Vollenweider, R. A., and J. J. Kerekes. 1981. Background and sum-
mary results of the OECD cooperative program on eutrophica-
tion. Pages 25-36 in Restoration of Lakes and Inland Waters. EPA
440-15-81-010. U. S. Environ. Prot. Agency, Washington, D. C.
72
-------
LAKE WATER QUALITY AND BEACH USE IN THE
OKANAGAN BASIN, BRITISH COLUMBIA
ROGER McNEILL
Economic Planning Officer
Environment Canada
Vancouver, British Columbia
ABSTRACT
Water quality has generally been good in the five lakes of the Okanagan River Basin in British Columbia.
However, cultural sources of nutrients entering the lakes increased rapidly during the 1960's, at the same
time demand for water-based recreation was also increasing. The Okanagan Basin Study recognized that
water quality problems were arising and recommended a phosphorous management program. By 1983,
all the major urban centers in the Basin will be removing at least 80 percent phosphorous from treated waste.
INTRODUCTION
The Okanagan River Basin, situated in the dry southern in-
terior of British Columbia, supports an important tourist in-
dustry. Recent surveys (Phipps and James, 1980) indicate
more than a million tourists visit the area each year and that
most are attracted by the opportunities for water-based
recreation. The five lakes of the main valley generally have
good water quality and attractive, well-serviced beaches. In
the summer, the beaches are heavily used by residents and
tourists. Total annual beach days (one visit to the beach by
one person during the day) were estimated at 7.15 million
in 1980 (Canada-Br. Colum., 1982), and are projected to in-
crease at a steady rate.
Cultural sources of nutrients entering the lakes increas-
ed rapidly during the 1960's, while at the same time demands
for water-based recreation also increased. By the late 1960's,
occasional algal blooms were occurring in the summer, and
residents were becoming concerned about the water quali-
ty. The Okanagan Basin Study (Canada-Br. Colum., 1974)
recognized the problems with water quality that were aris-
ing in the Basin. The study made recommendations for a
phosphorous management program aimed at reducing
loadings from diffuse sources and removing 80 percent of
the phosphorous from point source loadings. By 1983, all
of the major urban centers in the Basin will have im-
plemented at least 80 percent phosphorous removal from
treated waste. Data are not available on the costs of the
waste treatment measures but it is estimated that total capital
costs will be around $90 million.
As part of the Okanagan Basin Implementation Agreement
(Canada-Br. Colum., 1976), a review of the water manage-
ment plan for the Basin took place in 1980-82. One of the
objectives of this review was to see if the public's needs and
perceptions of water quality had changed since the early
1970's. A survey of Okanagan residents (Collins, 1981) in-
dicated that a significant proportion of the local population
had recently encountered a number of water quality problems
related to eutrophication, including algae growth, discolored
water, and bad-tasting water. There was almost unanimous
agreement in the survey that higher standards for nutrient
control should be enforced. The survey did not ask people
if water quality problems actually affected their use of the
lakes or the value they placed upon this use.
It was felt that further work was needed to relate actual
participation in water-based recreation to levels of water quali-
ty. In the Okanagan Basin, even the most eutrophic lake has
better water quality than many heavily used recreational
areas in other parts of the world. It was questionable whether
the relatively narrow range of water quality experienced in
the Basin would have significant effects on participation in
water-based recreation. A statistical study was therefore car-
ried out to see if a relationship did exist between water quality
and levels of recreational use. The objectives, data,
methodology, and results of this study are described in the
rest of this paper.
OBJECTIVES OF THE STUDY
It was necessary to restrict the study to existing data because
of time limitations. Given this restriction, the objectives of
the study were:
1. Estimate the relationship between the number of beach-
users and the water quality over a cross section of beaches
in the Okanagan Basin.
2. Calculate the value of a change in water quality in
selected beach areas.
Primary emphasis was placed on the first objective of
estimating the relationship between use and water quality.
With the existing data base, it was not possible to obtain
firm estimates of the value of water quality. However, with
some assumptions, the estimated model could be used to
indicate the costs or benefits associated with a change in
water quality.
DATA SOURCES
Counts of the number of beach users at 20 beaches in the
Basin were made in a separate study (Phipps and James,
1980). The study estimated the total number of beach days
at each beach over the summer session of 1980, based on
sampling done throughout the season.
The estimates of water quality at each beach were made
by biologists familiar with the area. The biologists were
reasonably confident in their estimates of the trophic status
of the lakes as a whole, but data specific to all of the beach
areas were not available. They assumed that the trophic
status of a lake as a whole could be applied to beach areas
on the lake with some modification reflecting local sources
of pollution and water depth. Water quality was rated on a
simple scale from oligotrophic to meso-eutrophic and a
numerical index was assigned to each rating as follows.
73
-------
Lake Restoration, Protection and Management
Status
Oligotrophic
Oligomesotrophic
Mesotrophic
Mesoeutrophic
Index
4
3
2
1
PRICE
P
A short field survey was carried out to obtain data on
physical features of the beach including length, direction,
and parking availability. Data on population were obtained
from Schultz International (1981) and Statistics Canada
(1976). Information on tourist accommodation was obtain-
ed from accommodation directories (Br. Colum., 1980).
Information was also available from previous studies on
the value of water-based recreation. In the Okanagan Basin
Study (Canada-Br. Colum., 1974), a survey of beach users
was undertaken that showed an average willingness to pay
of $5 (1974 dollars) per beach day in the Okanagan Basin.
The respondents were not asked to consider specific features
of the beach they were on or the availability of alternative
beaches in the Okanagan when deciding on their willingness
to pay, so the $5 is interpreted as a general average value
for a beach day in the Okanagan.
THEORETICAL MODEL
A model is hypothesized where the number of beach users
is a function of water quality, physical features of the beach,
distance to and size of population centers and tourist ac-
commodation, and the supply of other beaches in the area:
(1) Annual number of
beach days at
beach (i)
= f (water quality at beach (i),
length of beach (i), direction
of beach (i), resident popula-
tion, tourist population,
distance to tourist accom-
modation, distance to resi-
dent population, supply of
alternative beaches, water
quality at alternative
beaches).
The right-hand side variables are factors that should af-
fect the demand for beach recreation by tourists and
residents. No weather variables were included since the
weather is fairly consistent in the main valley during the sum-
mer. The direction of the beach was included because south-
and west-facing beaches receive sunlight for a longer period
of time in the afternoon than east- and north-facing beaches
allowing for greater use in the late afternoons.
Statistical estimation of equation (1) will satisfy the first ob-
jective of the paper which was to estimate the relationship
between water quality and beach use. For a given change
in water quality at beach (i), equation (1) should be able to
predict the change in use at beach (i), the change in use
at alternative beaches, and the beach days gained or lost
in the Okanagan Basin as a whole.
To place a value on a change in water quality, we assume
the existence of a composite demand function for beach days
in the Okanagan Basin. This demand function would be the
aggregate of the demand curves for beach days at individual
beaches in the Basin given existing water quality, physical
features of beaches, and resident and tourist population. This
demand curve is represented by the line BP in Rgure 1. The
line B1P1 represents demand at some lower level of water
quality.
The total value of beach recreation is represented by the
area under the line BP This area can be estimated using
the average beach day value of $5 in 1974 dollars or $10.65
in current dollars. Multiplying $10.65 per day by 7.15 million
beach days gives a total value of $76.1 million per year
ANNUAL
BEACH DAYS
Figure 1 .—Demand for beach recreation.
Given a reduction in water quality at a particular beach,
the model in equation (1) can predict the drop in use at the
beach and the gain in use at substitute beaches. The net
loss in beach days to the Okanagan Basin is the difference
between the decrease at the one beach where water quali-
ty was reduced and the gain at substitute beaches. The net
loss in beach days is represented by the distance from B
to B1 in Figure 1. The shaded area between the two de-
mand curves can be used as an approximation of the value
associated with the loss in beach days. This area can be
estimated by multiplying the average beach day value of
$10.65 by the drop in beach days represented by the
distance from B to B1.
One drawback in calculating this value as described, is
that the shaded area in Figure 1 does not account for the
welfare losses of beach users who would substitute different
beaches for their present beach if water quality were reduc-
ed. These losses would occur because of additional travel
costs, inconvenience, and congestion. In this regard the
method described would underestimate the value lost
because of a change in water quality at a particular beach.
A second drawback to the method of valuation is using
the average consumer surplus value of $10.65 in calculating
the area between the two demand curves. This value would
be appropriate if the demand curve shifts inward with no
change in slope. However, the slope may change if water
quality were reduced at a particular beach, since the loss
in beach days would likely be attributable to recreationists
with the lowest utility of beach use. The beach users with
higher utility would tend to travel to other beaches if water
quality was reduced at beaches they presently use. Thus
using an average consumer surplus value per net beach day
lost may be too high, resulting in an overestimate of the value
lost because of a change in water quality.
The two problems with the valuation method will offset
each other to a certain extent but it is not known what their
net effect on the valuation method will be. Thus values
calculated with this model should only be considered as ap-
proximations of the values associated with changes in water
quality.
RESULTS OF ESTIMATION
The model was reformulated as in equation (2) and estimated
by ordinary least squares regression:
(2) Density = - 6.58 + 51.9 WQ + 100.50 D
(4.89) (5.86)
R2 = .83 - 2840200 T - .014 S
20 observations (- 2.26) (2.74)
where density is the number of beach days in a season divid-
74
-------
Socioeconomic Benefits of Lakes and Restoration
ed by the length of the beach, WQ is the natural logarithm
of the water quality index, D is an index representing direc-
tion of the beach, T is an index representing distance to
tourist accommodation, and S is an index of substitute
beaches. The T-statistics are shown in parentheses.
The index of direction was constructed by assigning a
value of 1.0 to south-facing beaches, .9 to southwest-facing
beaches, .8 to southeast-facing beaches, .6 to west-facing
beaches, .4 to east-facing beaches, .2 to northwest-facing
beaches, and .1 to north- and northeast-facing beaches. The
indices of distance to tourist accommodation and substitute
beaches considered accommodation and alternative
beaches within 20 miles of the beach. The index of distance
to accommodation was constructed by summing the distance
to each accommodation site divided by its accommodation
capacity. The index for substitute beaches was constructed
by summing the length of each alternative beach divided by
its distance and weighted by its water quality.
The fit of the equation is good and water quality is
statistically significant. The index of substitute beaches
weighted by water quality is also significant. Preliminary
estimations showed that parking and resident population
were not significant and that coefficients of the other variables
were not appreciably affected when parking and resident
population were not entered in the estimated equation.
As a further test of the significance of water quality, the
model was re-estimated without quality as a variable and
without including water quality in the index of substitute
beaches. The results are shown in equation (3).
(3) Density =
R2= .83
20 observations
- .16 + 89.70 D
(3.51)
- 7,870,800 T + .004 S
(-.41) (.21)
Without water quality in the model, the R2 of the equation
is substantially lower and the significance of other variables
is less.
The results indicate that water quality differences between
beaches in the Okanagan Basin have significant effects on
the number of beach users. The impact of water quality on
beach use can be illustrated by examining the effects of a
change in water quality at a beach with average use and
average water quality. If water quality drops from the average
of 2.8 on our scale to 1.8 on our scale, then the density of
69 at the average beach would drop to 46, a decrease of
33 percent.
The statistical evidence also indicates that beach use is
more closely related to the logarithm of water quality, rather
than the linear index. This means that there are decreasing
returns to improved water quality. For example, improving
water quality from 3 to 4 (meso-oligotrophic) has a lesser
effect on usage than improving water quality from 1 to 2
(meso-eutrophic to mesotrophic). A third indication from the
data is that the number of users on a particular beach is
also affected by the water quality of alternative beaches in
the area.
VALUE OF A CHANGE IN WATER QUALITY
If we consider a decrease in water quality at beach (i), the
valueypsj can be estimated as follows. First, equation (2)
is usedro predict the decrease in beach days at beach (i),
given the change in the water quality index. Then equation
(2) is used to predict the increase in beach days that would
occur at substitute beaches. This is done by adjusting the
index of substitute beaches to reflect the lower water quali-
ty at beach (i) and then solving equation (2) for all other
beaches. The net change in beach days for the Basin as
a whole is then calculated by subtracting the increase in
beach days at substitute beaches from the decrease in beach
days at beach (i). The value lost is then obtained by multiply-
ing the net beach days lost by the average beach day value
of $10.65.
We do not have any firm predictions on how water quali-
ty will change in the Okanagan Basin, so someJIIustrative
cases are presented. These cases present some conser-
vative scenarios where water quality declines or improves
gradually over a 10-year period. In each case an increase
in general demand for beach days of 1.8 percent per year is
assumed, based on previous studies (Phipps and James,
1980) and a real discount rate of 10 percent is used.
Case One—Water quality in the Central Okanagan will
decline slightly from oligo-mesotrophic to mesotrophic over
the next 10 years. Two major public beaches, two smaller
beaches, and some private areas in the region will be af-
fected. After 10 years the annual beach day loss will be
196,000 with a value of $2.08 million. The present value of
these losses over a 25-year period is $15.5 million.
Case Two—Water quality in Kalamalka Lake declines
from the present oligotrophic state to mesotrophic state over
10 years. Two major public beaches and some private areas
are affected. After 10 years the annual beach day loss will
be 277,000 with a value of $2.95 million. The present value
of these losses is $21.85 million.
Case Three—Water quality improves in Osoyoos Lake
from the present meso-eutrophic state to mesotrophic over
a 10-year period. A major public beach, two smaller public
beaches, and several commercial beach areas will be af-
fected. After 10 years the annual gain in beach days will be
570,000 with a value of $6.07 million. The present value of
this gain is $44.65 million.
SUMMARY AND CONCLUSIONS
This study indicates a significant relationship between beach
use and water quality in the Okanagan Basin, even over the
relatively narrow range of water quality observed in the area.
The estimated relationship was used to calculate values lost
or gained with some hypothetical water quality changes.
These illustrative cases of water quality changes resulted
in fairly substantial values being gained or lost.
It is believed that the values determined in this study
should be used only for a preliminary indication of values
associated with changes in water qualilty since the analysis
suffered from some weaknesses. The major weakness was
that the analysis was based on the concept of an aggregate
demand for beach days in the Okanagan Basin. This ag-
gregate demand function can be used only to approximate
values lost or gained when water quality changes at specific
beaches. A more direct estimate of the value of changes
in water quality could be obtained from disaggregated de-
mand curves for specific lake areas or beaches in the Basin.
Such estimates would require considerably more data than
are currently available. Given the significance of water quality
indicated in the present study, this extra research effort pro-
bably would be justified.
ACKNOWLEDGEMENTS: This study was based on work carried
out under the Okanagan Basin Implementation Agreement between
Canada and British Columbia. The views, conclusions and recom-
mendations are those of the Author and not necessarily those of
Environment Canada.
REFERENCES
British Columbia Ministry of Tourism. 1980. Accommodation and
Campground Directory. British Columbia, Canada.
75
-------
Lake Restoration, Protection and Management
Canada-British Columbia. 1974. Main report of the Consultative
Board, Okanagan Basin Agreement. Environ. Can. Br. Colum.
Ministry Environ.
1976. Okanagan Basin Implementation Agreement.
1982. Report on the Okanagan Basin Implementation Agree-
ment. Environ. Can. Br. Colum. Ministry Environ, (forthcoming).
Collins, John B. 1981. Public opinion and attitude survey. Prepared
for the Okanagan Basin Implementation Board.
Phipps. S. A., and S. A. James. 1980. Water Based Recreation in
the Okanagan Basin, 1980 Rev. Prepared for the Okanagan Basin
Implementation Board.
Schultz International Ltd. 1981. Population Projections by Economic
Region and Sub-basin of the Okanagan River Basin. Prepared
for Environ. Can. Inland Waters Direct., Vancouver.
Statistics Canada. 1976. Census of Canada. Microfilm records of
population by electoral area.
76
-------
AN EVALUATION OF THE ROLE OF LOCAL REGULATIONS FOR
LAKE PROTECTION AND MANAGEMENT
D. A. YANGGEN
University of Wisconsin-Extension
Madison, Wisconsin
ABSTRACT
Regulators can play an important role in lake management. The water quality of lakes and the amenity
values of lake shores can be protected through zoning ordinances, subdivision regulations, sanitary codes,
erosion control ordinances and boating regulations. Density of development, preservation of shore cover,
water quality, erosion control, and conflicting lake uses are among the problems regulations can address.
Key features of the particular lake environment should be reflected in the regulations.
INTRODUCTION
The best solution to lake problems is to try to prevent them
from happening in the first place. A less satisfactory way is
to attempt to deal with them after they occur, and it is possi-
ble to restore deteriorated lakes in certain circumstances.
Appropriate regulations can be an important part of the
broader lake management package. They can be used to
protect existing lake quality as well as to preserve the im-
proved status where rehabilitation projects have been under-
taken. Governmental regulations may be adopted with three
general lake protection purposes in mind: (1) protect water
quality by regulating land use activities that cause pollution,
sedimentation, and excess enrichment; (2) control develop-
ment to preserve the amenity value of the shoreland; and
(3) regulate lake use to reduce conflicts among various water
users such as swimmers, water skiers, skin divers, boaters,
and fishermen (Born and Yanggen, 1974).
The purpose of this paper is to explore the role that local
regulations can play in lake management. It is written from
the perspective of a planner-attorney rather than a natural
scientist. The paper describes the various regulatory mech-
anisms and how they can function to address common lake
problems. It discusses the basic tests of legal validity used
by courts to evaluate regulations. The various phases of a
regulatory program are described and mechanisms to help
fit general rules to specific situations are considered.
REGULATORY TOOLS AND THEIR
APPLICATION IN A LAKE ENVIRONMENT
Zoning
Zoning developed in an urban context and its traditional func-
tion was to prevent conflicts between incompatible land uses.
A zoning ordinance establishes use districts that designate
areas in which specific uses are permitted as a matter of
right and others are conditionally permitted. Zoning is also
concerned with the dimensions of lots, the dimensions of
structures, and the location of structures on the lot. In a lake
environment, the manner in which shoreland uses are de-
veloped is a more likely threat than the encroachment of
incompatible land uses. Special provisions like the follow-
ing can be added to zoning to adapt it for lake protection
purposes.
Setbacks: The concept of street setbacks typical of con-
ventional zoning can be modified to require additionally
that buildings and septic tanks be set back a specified
distance from the water. Piers, wharves, and boathouses,
if the latter were permitted, would of course be exemp-
ted. Setbacks create a shoreland buffer that helps
preserve shore cover, natural beauty, and wildlife.
Control of removal of vegetation: Regulations requiring
retention of shoreland vegetation can be designed to pro-
tect the aesthetic appeal of a natural shoreline while still
allowing a view of the water from the lot. In addition to
scenic beauty, retaining shoreland vegetation makes land
less vulnerable to erosion and can trap sediment before
it enters the water. Shoreland vegetation also intercepts
nutrients contained in effluent and fertilizers when it uses
them as food.
Control of earth moving: Provisions regulating filling and
grading reduce erosion and sedimentation from raw soil.
Controls can direct that the smallest amount of bare
ground be exposed for as short a time as feasible. Tem-
porary and permanent ground cover can be mandated and
diversions, barriers, desilting basins, and other methods
to trap sediment required where necessary. Control of fill-
ing also reduces erosion and can be one method to pro-
tect wetlands. A special permit with technical review can
be required, depending on the size of exposed area and
the steepness of the slope.
Lagooning and dredging provisions: These controls
have several purposes: they protect wetlands from ill-
advised attempts to improve on nature, set standards to
prevent slumping of excavated areas, and avoid improper-
ly constructed lagoons where oxygen depletion may cause
fish kills.
Wetland conservancy zoning: Wetlands can be pro-
tected by establishing a special wetland conservancy
district where alteration of the natural condition of the
wetland is prohibited or, in the alternative, filling and dredg-
ing are allowable only upon issuance of a special permit
to which conditions may be attached that minimize adverse
effects on the wetland's functions.
Among the many functions performed by wetlands are
several that may be key to maintaining the long-term health
of a lake environment. Wetlands bordering lakes and
streams provide spawning grounds for northern pike,
muskellunge, and walleye. Waterfowl including mallards,
teal, and wood ducks rear their young in wetlands. Many
other species'of mammals, birds, reptiles, and amphibians
77
-------
Lake Restoration, Protection and Management
depend entirely upon wetlands for their survival at one time
or another during their life cycle.
Wetlands can also be natural treatment plants as they
trap and store nutrients and sediments that might other-
wise lower the water quality. Wetland plants take up ni-
trogen and phosphorus, reducing the nutrient input that
can overfertilize a lake, bringing rampant weed growth and
algal blooms. Water flows through wetlands slowly, allow-
ing time for sediment to settle out. Filling or draining
wetlands destroys their filtering action. When wetlands are
drained, nutrients stored in peat or muck are often dis-
charged into the adjacent lake (Yanggen et at. 1979).
Density standards and carrying capacity: Dimensional
standards, including minimum lot area and width, and the
maximum percentage of a lot that may be covered by
structures are typical of conventional zoning. Their basic
purpose is to provide open space and control density.
Dimensional standards for lakes should take into account
several key factors.
The size and shape of a lake are important considera-
tions in setting density standards. Assume two perfectly
round lakes, one with 50 acres of surface water and the
other with 200. While the 200-acre lake has four times the
surface water acreage, its shore line is only twice the
length of the 50-acre lake. Next, assume two 50-acre lakes,
one with a circular pattern (Round Lake) and the other with
a lineal form (Long Lake). The shoreline of Long Lake is
substantially greater than that of Round Lake and the
amount of surface water per lot is correspondingly less.
More surface water per lot, other things being equal,
means the lake is less vulnerable to overcrowding from
shoreland-generated recreational use and that the visual
impact of development is also less. Thus, large, regularly
shaped lakes have a greater carrying capacity in these
respects than do small, irregularly shaped lakes. The
dimensional standards on small, irregularly shaped lakes
should thus be more restrictive than for large, regularly
shaped lakes. Water volume and flush potential are two
other key physical parameters that determine the lake's
absorptive capacity in terms of its ability to assimilate
nutrients.
Once the potential vulnerability to overcrowding and
eutrophication has been estimated, the next step is to
relate it to regulatory standards. The two basic regula-
tory variables are density, i.e., the number of structures al-
lowed; and the area, width, and coverage of the lots and
setbacks, i.e., the distance structures must be located from
the water's edge. The former focuses on the carrying
capacity of the lake, the latter on the mitigating effect of
the shoreline buffer. Lakes are such complex and dynamic
natural systems that it is impossible to state with any
degree of precision the amount of protection that an ad-
ditional unit of lot width or setback will provide. It is possi-
ble, however, to group lakes based upon their vulnerability
and quality. A classification of this type using readily
available information has been developed by William
Lontz, University of Wisconsin-Extension, Hayward, Wis.
In a jurisdiction with a large number of lakes, it is thus
possible to classify lakes in a way that the vulnerable high
quality lakes are subject to more stringent development
standards than less vulnerable lakes of lower quality
(Lontz, 1980).
If the resources are available, a preferable approach is
to prepare individual lake water quality management plans
within a broader regional context. An example is the plan
for Lac LaBelle prepared by the Southeastern Wisconsin
Regional Planning Commission. The objectives of the plan
were to provide a level of water quality suitable for the
maintenance of a healthy warmwater fishery, to reduce
excessive algal growth, and to improve opportunities for
water-based recreation. The report determined the trophic
status of the lake (mesotrophic), identified the amount of
nutrient loading to the lake by source, and projected the
nutrient loading under conditions expected 20 years in the
future. Management measures necessary to meet the
water use objectives for the lake were identified. Those
included actions necessary to effect a 30 percent reduc-
tion in nonpoint source phosphorus and an evaluation of
alternative lake rehabilitation and in-lake management
techniques. Among the specific measures called for were
modification of the zoning ordinance; provision of sanitary
sewer service to a portion of the watershed; implementa-
tion of nonpoint source controls in urban and rural areas;
revision of the sanitary ordinance; implementation of a con-
struction erosion control ordinance; aeration, dredging, sedi-
ment covering, and weed harvesting; and establishment of
a lake protection district (Southeast. Wis. Reg. Plann.
Comm.).
Subdivision Regulations
To ensure proper and orderly development, subdivision
regulations control the division of land into lots for sale or
building. Subdivides are required to prepare plats, i.e., de-
tailed maps of the land proposed to be subdivided. The plats
must be approved by local regulating agencies before the
plat can be recorded and the lots sold. Plats are usually
reviewed to ensure the suitability of the area for a subdivi-
sion; the adequacy of the street system; proper dimensions
and layout of lots; sufficiency of water supply and waste
disposal systems; proper stormwater management; control
of erosion and sedimentation; adequate open space; and
safety from physical hazards.
Planned unit development provisions in zoning and sub-
division regulations can be adapted to a lake environment.
These provisions can allow a developer to arrange lots in
off-shore clusters rather than in long strips along the shore.
It's possible to increase the amount of usable open space
while still maintaining the same overall density. This can be
done by permitting a reduction of the minimum lot size for
each dwelling if an equivalent area in the development is
restricted to permanent open space.
A subdivision in northern Wisconsin shows how planned
development provisions can permit a thoughtful layout that
protects the lake environment in an economically feasible
way. In this instance, all residential development is laid out
in off-shore clusters. An undeveloped buffer strip extends
back 200 feet from the shoreline, providing a substantial pro-
tective buffer. This shoreline strip is owned by purchasers
of the individual lots as tenants-in-common. The residential
clusters, in turn, are tied together by the shoreland buffer
and a system of other commonly-owned greenways, giving
the lot owners a much larger area in which to roam than
even the largest size lot could provide. The increased value
of the more numerous off-shore lots more than compensates
for the value foregone by not developing the shoreline strip.
Under this arrangement the developer, the lot purchasers,
and the lake all benefit.
Subdivision regulations and zoning complement each
other. Zoning focuses primarily on the use of land, the dimen-
sion of lots, and the location of structures on the lot. Zoning
can also control grading, filling, vegetation removal, and other
activities that accelerate erosion. These activities can be
made conditional uses to require that they be undertaken
in a manner that avoids adverse effects. Subdivision regula-
tions focus on the process of dividing larger tracts of land
into lots for purposes of sale or building. For undeveloped
shoreland areas that have not been divided into lots, subdi-
vision regulations have particular promise. The larger size
of the parcel involved permits more flexible and environmen-
78
-------
Socioeconomic Benefits of Lakes and Restoration
tally protective lot layout, and makes it more likely that eco-
nomically feasible engineering solutions can be found to ero-
sion control, stormwater management, and waste disposal.
On a more mundane basis, subdivision regulations can
help ensure the suitability of soils for septic tank soil absorp-
tion systems where the development is outside the range
of public sewerage systems. Where a septic tank system fails,
disease-bearing effluent can cause health hazards. Septic
tank systems may contribute a more subtle type of pollu-
tion when nutrients in the effluent enter the lake via the
groundwater flow. Excess enrichment of the lake can cause
rampant weed growth and nuisance algae blooms. The soil
in a septic tank system is an integral part of the treatment
process.
Physical features that affect the operation of a soil absorp-
tion system are permeability, slope, groundwater, bedrock
and flooding. An assessment of these factors can be required
to determine whether the subdivision's soils are suitable for
onsite waste disposal and, if so, what is the appropriate
average minimum lot size. Subdivision regulations thus at-
tempt to ensure that the lots have an adequate area of
suitable soils. Sanitary regulations are designed to more
precisely control the systems.
Sanitary regulations
Sanitary regulations set construction and dimensional stan-
dards for septic tanks. Percolation and soil boring tests are
performed on each lot to determine the location and size
of the soil absorption system. Minimum separating distances
are specified between the septic tank system, the house,
well, and lake edge. In some States that use a lake classifica-
tion system, the distance the septic tank must be set back
from the water's edge depends upon the class to which the
lake is assigned.
Erosion control ordinances
In some States separate erosion control regulations exist on
the State and local level. The most common form regulates
runoff and erosion from land development activities in much
the same way as they would be controlled in zoning and
subdivision regulations. Agricultural activities may also add
nutrients and sediment to lakes. Regulation of farming opera-
tions such as winter spreading of manure or eroding
cropland, however, is much less frequent.
Surface water regulations
A lake's surface has a limited carrying capacity. When this
capacity is exceeded, conflicts arise among various types
of water users: swimmers, water skiers, fishermen, and
boaters. These conflicts are both physical and psychological.
Water skiers and fishermen cannot both occupy the same
water at the same time. The noise of a motorboat at night
may disturb the tranquility of the solitude seeker. These con-
flicts increase with more lake-lot owners and heavier use of
lakes by the general public. Surface water regulations can
help reduce conflicts, protect the lake, and make aquatic
recreation safer for lake users. Most States set general stan-
dards for boat equipment, traffic rules, and negligent opera-
tion. However, in some cases, more detailed regulations may
be needed to prohibit incompatible mixing of uses, allocate
parts of lakes to their most appropriate use, and lessen the
intensity of use.
Surface water regulations may take several different forms:
1. Fixed area zoning like land use zoning restricts
specified water uses to designated areas. These areas may
be delineated by general description, e.g., within 100 feet
of the shore, be marked on a map of the lake, or indicated
by buoys or similar markers;
2. Minimum separating distances may be established that
require that one use stay a specified distance away from
other uses wherever they are located, e.g., a power boat
must be operated at least 100 feet from an anchored fishing
boat, swimmer, or skin diver;
3. Time zoning limits certain uses to designated times,
e.g., water skiing to between 10 a.m. and 6 p.m., thus reduc-
ing conflict with anglers during the best fishing hours;
4. Exclusion of some uses entirely, e.g., motors on lakes
less than 50 acres in size. A circular lake of 50 acres is only
one third of a mile wide. It takes just 5 minutes to row across
it at a speed of 4 miles per hour. By contrast, large lakes,
even though under use pressures, might tolerate all uses
somewhere on the lake with controls maintained by a com-
bination of speed limits, fixed area zoning, minimum sepa-
rating distances, and time zoning.
GENERAL TESTS OF THE LEGAL
VALIDITY OF REGULATIONS
In devising the regulatory portion of a lake management pro-
gram, it is important to keep in mind the fact that the validi-
ty of the regulations may be challenged in court. The follow-
ing discussion highlights the most common legal challenges
to local regulations.
The Governmental Unit Lacks the
Authority to Adopt the Regulations
Local governmental units are legally creatures of the State.
Cities and villages typically have a broader legislative man-
date than do rural forms of government such as towns or
counties. This broader mandate may take the form of charter
ordinances or statutory or State constitutional home rule
powers. Essentially, it is within the power of these munici-
palities to regulate matters of local interest as opposed to
matters of statewide concern. Rural governmental units
typically operate on the basis of enabling legislation that
specifically authorizes the exercise of the particular power
or at least it must be possible to reasonably infer a delega-
tion of such authority on the basis of the statutory language.
Urban and rural general purpose governmental units are
almost always authorized to adopt zoning and subdivision
regulations. Free-standing erosion control ordinances are
likely authorized by home rule power. It is less clear whether
surface water regulations are within the authority of local
government unless authorized by statute.
The issue is whether regulation of lake use is a matter
of local interest as opposed to a matter of statewide con-
cern. The thousands of lakes in States such as Minnesota
and Wisconsin are the backbone of the outdoor recreation
industry. Here the concern is within the uniformity of boating
regulations so that the recreationist is not faced with a
bewildering array of differing boating restrictions. Statutes
in both States require State review of local ordinances. Even
in the absence of statutory provisions, it is possible that a
court would find that surface water regulations are a matter
of statewide concern rather than local interest. Special pur-
pose units of government such as lake management districts
typically are not authorized to adopt regulations.
Valid Objectives
Regulations that restrict the right to use private property must
serve valid public objectives that promote the public health,
safety, and general welfare. In Just v. Marinette County (Wis.
1972) 201 NW 2d 761, a case that upheld the validity of
shoreland regulations requiring a conditional use permit for
filling a wetland, the court noted:
79
-------
Lake Restoration, Protection and Management
The active public trust duty of the State of Wisconsin in respect
to navigable waters requires the State not only to promote naviga-
tion, but also to protect and preserve those waters for fishing,
recreation and scenic beauty (citations omitted). To further this
duty, the legislature may delegate authority to local units of govern-
ment, which the State did by requiring counties to pass shoreland
zoning ordinances (citations omitted).
Reasonable Means
The regulatory provisions must be a reasonable means
to achieve these objectives. In County of Pine v. Sfafe
Department of Natural Resources (Minn. 1979) 280 NW
2d 625, the court upheld an ordinance that established
special regulations along a river corridor. Among other
things, the ordinance established minimum lot sizes and
minimum lot width requirements at the shoreline and at
the building line, established setbacks from the normal
highwater mark and from the bluff line and regulated the
cutting of timber and vegetation. The court made the
following comment in regard to the way the ordinance ad-
vanced legitimate public objectives.
Besides promoting aesthetics, the Kettle River ordinance pro-
motes other valid zoning considerations. For instance, the bluff
setback requirement reflects a concern that construction too
near the edge of a bluff may be structurally unsafe and may
create a potential for harmful pollution from erosion of the bluff.
The minimum lot size requirements will limit the density of new
construction thereby limiting septic tank installation and other
adverse environmental factors.
Taking
There must be no taking of private property without just
compensation. In Just v. Marinette County, the court began
its discussion of the taking issue by stating traditional tests
such as (1) the degree of damage to the property owner
(decrease in value), (2) whether the land is rendered use-
less for all reasonable purposes, and (3) whether the re-
strictions create a public benefit rather than prevent a
public harm. The court then went on to point out that
special considerations apply to wetlands and other en-
vironmentally sensitive lands.
The Justs argue their property has been severely depreciated
in value. But this depreciation of value is not based upon the
use of the land in its natural state, but on what the land could
be worth if it could be filled and used for the location of a dwell-
ing. While the loss of value is to be considered in determining
whether a restriction is a constructive taking, value based upon
changing the character of the land at the expense of harm to
public rights is not an essential factor or controlling.
Adequate Standards for Flexible
Regulations
Flexible regulations must contain adequate standards.
Zoning districts often list two categories of allowable uses,
i.e., permitted uses and conditional uses. Permitted uses
are allowed as a matter of right if they meet the specified
dimensional requirements. Conditional uses, on the other
hand, may create special problems and hazards if allow-
ed as a matter of right. Whether they are appropriate for
a particular location depends upon an evaluation of the
characteristics of the proposed use and the proposed site.
The conditions that may be attached to development
permission can, in some instances, avoid potential adverse
effects on adjoining land or the public welfare. The flex-
ibility achieved through conditional uses and similar
devices that control development in light of the facts and
circumstances of the particular case raises the question
of whether the ordinance contains sufficient standards to
avoid the arbitrary exercise of discretionary power. In the
language of the Wisconsin Court. ". . . some standards
must be prescribed for the guidance of the board in
exercising the discretion and judgement with which it
is vested. Where no such definite standards are written
into the ordinance, the door is open to favoritism and
discrimination . . ."
In some States, zoning must be based upon a plan and
zoning actions are required to be consistent with the plan.
The existence of a plan or other factual underpinning great-
ly strengthens the legal defensibility of regulations. It is not
possible to make blanket generalizations concerning the
validity of various land use regulations since the determina-
tion may vary with the circumstances of particular cases. It
is clear, however, that the factual basis of land use controls
is very important. "The absence of clear theoretical
guidelines makes the facts become much more important
than the laws. What goes into the balance is more impor-
tant than the balancing. Both in drafting and defending land
use regulations, careful factual preparation is called, for."
(Bosselman, Callies and Banta, 1973.)
OTHER PHASES OF A REGULATORY
PROGRAM
Adopting legally defensible ordinances does not in itself con-
stitute an effective regulatory program. Other essential
phases are information, training, administration, and enforce-
ment. Information to and input from the general public is
necessary if the controls are to reflect public sentiment and
support. Special information to those directly affected is
essential if they are to comply with the controls.
Training the people who will administer the controls is
another essential step. Technical expertise from State and
Federal agencies is sometimes available to train people from
the local administering agencies. This expertise may also
be available in the form of advisory review to local govern-
ment and technical guidance to property owners. Without
this type of technical input, local governments may be limited
in the type of regulatory standards they can adopt. The more
readily technical expertise is available to the administering
agency, the more feasible it is to have flexible regulatory
devices such as conditional uses and planned unit develop-
ment provisions.
Technical review also permits more substitution of per-
formance standards for specification standards. Performance
standards describe the particular regulatory objective to be
achieved, but leave the selection of the means to the
developer. Specification standards on the other hand require
that particular practices be followed. In either case, onsite
inspection is required. It is not enough to issue a permit or
approve plans subject to certain conditions. What is impor-
tant is that the actual development take place in conformity
with the requirements. This means field inspection of the
development. It also means enforcement when regulations
are violated. Corrective action is best achieved through volun-
tary compliance, but in some instances, this does not work.
Then, the ultimate enforcement mechanism—a lawsuit re-
sulting in a judicial order of compliance—must be invoked.
If government does not take this step, the regulatory pro-
gram loses credibility.
SUMMARY AND CONCLUSIONS
Regulations can play an important role in a program of lake
protection and management. The particular contribution they
make depends upon a number of factors, but some general
conclusions are warranted. Land use controls have tradi-
tionally been used to implement policies designed to achieve
attractive and orderly land development. Zoning, subdivision
regulations, and sanitary ordinances can be modified by pro-
visions sensitive to the special characteristics of a lake en-
80
-------
Socioeconomic Benefits of Lakes and Restoration
vironment. In this respect, they can control development to
preserve the amenity values of the shoreland. These several
regulatory devices complement each other and apply in dif-
ferent phases of the development process.
Land use controls are by their nature largely prospective,
i.e., they apply to new activities undertaken during the de-
velopment process. If the lake environment is largely devel-
oped, this type of regulation will be less important. Surface
water regulations on the other hand can be applied to a com-
pletely developed lake to perform a remedial function. Reduc-
ing conflicts among surface water uses is best achieved by
combining regulatory measures, i.e., surface water regula-
tions to control lake use with nonregulatory measures such
as access policy.
How many people use a lake depends largely on how easy
it is to get onto the lake. Public pressures on lakes can be
affected by the size and location of launching facilities. There
are some real advantages to management programs that
operate on broader geographic lines. On "wilderness" lakes,
walk-in access may limit use to those willing to carry in a
small boat or canoe. At the other extreme, public facilities
can be enlarged to accommodate more people and larger
watercraft on lakes big enough to sustain the impact. This
disperses pressures over a large number of lakes reducing
intensity of use on particular lakes. Several midwestern
States have established regulatory programs that operate
on a Statewide basis in the sense that they set minimum
State standards for development in shoreland areas that
must be reflected in total land use controls (Yanggen, 1973).
Regulations can also help to protect water quality by
controlling activities that cause pollution, sedimentation, and
excess enrichment. Potential measures for water quality
management include control of nonpoint sources and lake
rehabilitation techniques. Regulations can be a part of non-
point pollution control, but engineering solutions in the form
of sanitary sewers and stormwater management may be re-
quired. In many instances, corrective rather than preventive
measures are needed; then the various lake rehabilitation
techniques are called for. Regulation can be combined with
corrective measures to protect the improved status of the
lake.
REFERENCES
Born, S., and D. Yanggen. 1972. Understanding lakes and lake
problems. Publ. G2411. University of Wisconsin-Extension,
Madison,
Bosselman, F., D. Cailies, and J. Banta. 1973. The taking issue.
U.S. Gov. Printing Off., Washington, D.C.
Lontz, W, 1980. The local resource-regulation connection. Hayward,
Wis.
Southeastern Wisconsin Regional Planning Commission. 1980. A
water quality management plan for Lac LaBelle. Waukesha, Wis.
Yanggen, D. 1973. Wisconsin's shoreland protection program: a
state/local regulatory approach to natural resource protection.
Pages 354-376 in C, R, Goldman, J. McEvoy III and P. J, Richer-
son, eds. Environmental Quality and Water Development. W. R.
Freeman, San Francisco.
Yanggen, D. ef a/. 1979. Wisconsin wetlands. Publ. G2818, Univer-
sity of Wisconsin-Extension, Madison.
81
-------
Acid Rain
ACIDIFICATION OF HEADWATER LAKES
AND STREAMS IN NEW ENGLAND
TERRY A. HAINES
JOHN J. AKIELASZEK
U.S. Fish and Wildlife Service
Columbia National Fisheries Research Laboratory
Field Research Station
University of Maine
Orono, Maine
ABSTRACT
A survey was conducted of 226 headwater lakes and streams in the six New England States to determine
the status of surface waters in this region with respect to acidification from atmospheric deposition. The
waters surveyed had physical and chemical conditions representative of those of lakes and streams in
the region, and watersheds relatively undisturbed by direct human activity. Each site was visited, the visi-
ble condition of the watershed and water was noted, and water samples were collected for chemical analysis.
Analyses included pH, alkalinity, specific conductance, color, calcium, and aluminum. Acidic (pH 5) surface
waters were found in every State and made up about 8 percent of the total number of waters surveyed.
Acidity was not correlated with color, an index of organic acids. Approximately half of the waters surveyed
were judged to be vulnerable to acidification based on present acid neutralizing capacity, with 53 percent
of the waters having alkalinities of 200^eq/l or less. Aluminum concentration was correlated with hydrogen
ion concentration, and the linear regression was similar to that obtained from other regional surveys. Of
the 226 waters surveyed, 13 had aluminum concentrations of 200jjg/l or more and a pH of 5.5 or less,
conditions that are likely to be toxic to sensitive fish species. Historical pH and alkalinity data indicate
that waters located in areas where acid neutralizing capacity is low have been acidified. Approximately
70 percent of the lakes in this group have reduced pH and/or alkalinity. The relation between pH and
calcium concentration (Henriksen Nomograph) predicted that 57 percent of the waters surveyed had been
acidified, a value that agrees with historical comparisons.
INTRODUCTION
Acidic precipitation has been identified as the cause of tion, based on indirect evidence such as bedrock geology
acidification of surface waters in Scandinavia (Aimer et al. (Galloway and Cowling, 1978; Hendrey et al. 1980), and this
1978; Muniz and Leivestad, 1980), Ontario, Canada (Harvey, region receives precipitation with an annual volume-weighted
1980), and New York (Pfeiffer and Festa, 1980). Several mean pH of 4.2 to 4.4 (Natl. Atmos. Depos. Progr. 1981).
studies have predicted that substantial portions of the waters Omernik and Powers (1982) used existing surface water
of the northeastern United States are vulnerable to acidifica- alkalinity data to locate areas of the United States that are
83
-------
Lake Restoration, Protection and Management
vulnerable to acidification. However, most water chemistry
data are from large watersheds and from areas affected by
human land use, and are incomplete because they contain
data for only a few of the major ions.
In this study we collected water chemistry data for major
ions from 226 relatively undisturbed lakes and streams in
the six New England States. Sample sites were selected on
the basis of geographical coverage and bedrock geology
type. The results of this study provide a measure of the acidi-
ty status of surface waters and provide a water chemistry
data base for assessment of the future effects of acidic
precipitation in this region.
METHODS
The surface waters included in this survey were primarily
lakes (193 of 226; 85 percent). Streams were sampled only
when a suitable lake was not available. Sites were selected
to represent all classes of bedrock buffering capacity (Hen-
drey et al. 1980) and soil cation exchange capacity (McFee,
1980) present in the region in proportion to the geographical
area included in each class. Sites with pronounced human
land use activities in the watershed (e.g., logging, agriculture)
were excluded. Most sites were not accessible by road and
were reached either on foot or by aircraft.
Historical water chemistry data for about half (95 of 226)
of the sample sites were obtained from State fish and game
or water quality departments, or Federal agencies such as
the U.S. Forest Service. Historical data were carefully ex-
amined for accuracy and reliability; results that were incon-
sistent with the methodology used were rejected.
Each water body included in the survey was visited by a
survey team in 1980. The team noted conditions in the water-
shed and collected water samples for chemical analysis.
Samples were collected at 0.1 m below the surface and near
the bottom over the deepest portion of the basin, using a
plastic Van Dorn type water sampler. Water temperature was
measured at the time of collection with a pocket ther-
mometer. Water samples were placed in two acid-washed,
distilled water-rinsed linear polyethylene bottles. One bottle
of each set was acidified with nitric acid for cation analysis.
Analyses of pH, alkalinity, and color were performed in
the field, within 8 hours after sample collection. The pH was
measured with a meter (Fisher model 107 or Cole Farmer
DigiSense) equipped with a Fisher plastic body gel-filled com-
bination electrode. The meter was standardized with pH 7
and 4 buffers, and electrode response was verified by
measuring the pH of dilute sulfuric acid solutions of
theoretical pH 4.
Alkalinity was determined by titrating a 100 ml aliquot with
0.0200 N sulfuric acid. Total inflection point alkalinity was
determined by the Gran method (Stumm and Morgan, 1981),
and fixed endpoint alkalinity (pH 4.5) was calculated accor-
ding to Standard Methods (1975). Apparent color was deter-
mined by comparing unfiltered samples with platinum cobalt
standard solution (LaMotte Chemical Co.).
In the laboratory, calcium was determined by nitrous oxide-
acetylene flame atomic absorption spectrophotometry (MS),
and total aluminum by graphite furnace MS (Perkin Elmer
model 703). The spectrophotometer was operated according
to the manufacturer's directions, and accuracy was assured
by analysis of Environmental Protection Agency quality
assurance standards for the elements in the concentration
ranges encountered.
RESULTS AND DISCUSSION
The 226 lakes and streams sampled were distributed
throughout the six New England States (Fig. 1). Surface
waters of pH<=5 were found in all six States (Fig. 2). Nine-
teen sites (8 percent) had pH<5, and 46 (20 percent) had
+ -»• LflKE
D D D STREflH
• • • IMPOUNDMENT
Figure 1.—Distribution of lakes, streams, and impoundments sampl-
ed in the six New England states.
Figure 2.—Distribution of sample sites by pH range.
84
-------
Acid Ram
120-
-
110-
100-^
90-
.
C 80-
0 :
L 70-
0 :
a so-;
.
U 50-
N :
I 10-
T :
S 30^
20^
10-
0-
O
a
_
O D D- O
a
D
D
O
D D
O ODD
D a a a
a a DD a D
a OB dP °OD
D OCB n DD GQDJ3 Ga
_ DO O
D D O D O O OD O cP
D an a DfBpP o CL a
D D DD D D nnrn npgnn OSB OPQU D m o
D DD 0 D WDO D US D
D O D DD n OP. p rwn o t3KOpB33 OH33 O ; ^i OO
U DU DiP D DDO DOOIID FTP
f r , , . TT-T-. , | . , , . 1 r. .,,.,..,,,., | , n ,, r ...,.,,.,.,. Tj . r.-TT . . - . | | . . |
1.0 14.5 5.0 5.5 6.0 6.5 7.0 7.5 8.
PH
Figure 3.—Relation between water color and pH.
pH of 5 to 6. Most of the lakes and streams sampled were
low in color (an index of organic acids), and color was not
related to pH (Fig. 3), indicating that acidity of these lakes
was not the result of organic acids.
Other regional water chemistry surveys have shown that
surface waters low in color and with a pH less than 5 are
not found in areas where average annual precipitation pH
exceeds 4,6 (Glass and Loucks, 1980; Wright and Gjess-
ing, 1976). The proportion of surface waters with pH less
than 5 that has been recorded in areas where average an-
nual precipitation pH is less than 4.6 is highly variable, rang-
ing from 2 percent (Malmer, 1975) to 64 percent (Wright and
Snekvik, 1978). The proportion of acidic lakes found in the
present study was lower than in most surveys in Scandinavia
(Aimer et al. 1974; Wright and Snekvik, 1978), Canada
(Beamish and Harvey, 1972; Watt et al. 1979), and New York
State (Pfeiffer and Festa, 1980). However, the pH distribu-
tion of lakes in our study was similar to that found in central
Ontario (Scheider et al. 1979), an area of comparable geology
and precipitation chemistry.
A large portion of the sites were low in inflection point
alkalinity (Fig. 4). There were 53 sites (24 percent) with
alkalinity of 20 ^eq/l or less, 40 (18 percent) greater than 20
but equal to or less than 100 (jeq/l, and 27 (12 percent)
greater than 100 but equal to or less than 200 ^eq/l. Alkalinity
is a generally accepted measure of acid neutralizing capacity
and as such is an indicator of vulnerability of a water body
to acidification (Omernik and Powers, 1982). No one alkalinity
level is generally accepted as indicative of extreme acid sen-
sitivity; suggested values range from 40 to 300 fieqll. In most
other regional surveys alkalinity was not measured. However,
in two surveys conducted in areas where the average an-
nual pH of precipitation was greater than 4.6, no surface
waters with alkalinity of 20 (jeq/l or less were found (Glass
and Loucks, 1980; Lillie and Mason, 1980). In areas where
the average annual pH of precipitation was less than 4.6,
the proportion of surface waters with alkalinity of 20 ^eq/\
or less ranged from 3 percent in south Norway (Wright et
al. 1977) to 71 percent in western Nova Scotia (Watt et al.
1979). The proportion found in our study (23 percent) is about
twice that found in central Ontario (12 percent, Scheider et
al. 1979) and half that found in New York (41 percent, Pfeif-
fer and Festa, 1980).
Aluminum concentration ranged from below detection
(-C10 p«g/l) to 630 nQ/l. Aluminum concentration exceeded
200 fig/l in 18 of the 226 lakes and streams surveyed, and
in 13 of these sites the pH was 5.5 or less, a combination
of conditions likely to be toxic to sensitive fish species (Baker
and Schofield, 1980). Total aluminum concentration was cor-
related with pH of the water (Fig. 5) as is the case in several
• » * <20
D o o 20-200
>200
Figure 4.—Distribution of sample sites by alkalinity range.
other surveys (see Table 1). The slopes and intercepts of
the regression lines from all surveys were very similar (Table
1), suggesting that the relation is an expression of the ther-
modynamic equilibrium of aluminum and hydrogen ion.
Usable historical pH data had been collected at 95 lakes
for years between 1937 and 1978. The historical pH is com-
pared with the mean surface pH obtained for the same lake
in our study (Fig. 6). If there is no difference between the
two measurements, the point will fall on a line through the
origin at a 45° angle. If the historical pH is lower than pre-
sent pH the point will fall above the line, and if historical pH
2.5-
2.0-
1.5-
1.0-
0.5-
L 0.0-
-0.5-
-1.0-
3.93 - 0.36X
2= 0.37!
1.0 li.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0
PH
Figure 5.—Relation between aluminum concentration and pH.
85
-------
Lake Restoration, Protection and Management
Table 1. — Values for the intercept (A) and slope (B) for the
linear regression of total aluminum and pH. Logi0AI (jjg/l) =
A + B (pH), from regional water chemistry surveys.
Location
Intercept Slope
Reference
New England
Sweden
Sweden
Norway
New York
226
37
73
47
214
3.93
3.85
3.94
3.85
4.14
-0.36
-0.35
-0.38
-0.41
-0.33
Present study
Dickson, 1975
Dickson, 1980
Wright & Gjessing, 1976
Schofield, 1982
1-
c
u
R
R
E 6
N
T
P
H
5-
H-
is in reasonably good agreement with the 64 percent de-
rived from the historical pH comparisons.
The present survey showed that a substantial proportion
of the headwater lakes and streams in New England is highly
vulnerable to acidification, and that a significant number are
now acidic. The proportion of acidic waters is much higher
than that in geographically comparable areas that do not
receive acidic precipitation, and cannot be attributed to
organic acids. Both comparison of recent to historical water
chemistry data and application of chemical models indicate
that about 60 percent of the water bodies surveyed have
recently been acidified, presumably by acidic precipitation.
1.0-
14.5-
5.0-
5.S-
6.0-
6.5-
7.0-
7.5-
8.0-
100 200 300 100
NBN-MBBINE CB UJEQ/L)
Figure 7.—Relationship between calcium concentration and pH for
water bodies surveyed. Curve is the empirical curve derived by
Henriksen (1979).
HISTORICRL PH
Figure 6.—Comparison of recent and historical pH data. Solid line
is the line of no change. Points above the line indicate lakes that
have become less acidic; points below the line indicate lakes that
have become more acidic.
is higher than present pH the point will fall below the line.
In this comparison 34 points (36 percent) are on or above
the line and 61 (64 percent) are below the line, indicating
that almost two thirds of the lakes for which historical data
were available were lower in pH in 1980 than when originally
surveyed. If variation were random one would expect the
number of sites where pH increased and decreased to be
about equal. Generally only one historical measurement was
made for each lake, and the time between historical and re-
cent measurements was highly variable. This variation
precluded an analysis of rate of change in pH over time.
Henriksen (1979) proposed that the relation between pH
and calcium could be used to identify lakes that have been
acidified by atmospheric deposition. In unacidified waters,
bicarbonate alkalinity is present in approximately equivalent
concentrations as the sum of non-marine calcium and
magnesium, the ratio of calcium to magnesium is relatively
constant, and alkalinity is related to pH. As lakes are acidified
bicarbonate alkalinity decreases and is replaced primarily
by sulfate (Henriksen, 1979). An empirical curve was de-
rived from the relationship of lake pH to calcium concentra-
tion, so that points falling on or below the curve reflect non-
acidified lakes and points falling above the line reflect
acidified lakes. Application of this relation to data from our
survey (Fig. 7) for lakes with calcium content of 500 ^eq/l
or less indicated that 80 of 140 lakes (57 percent) fell above
the curve and were presumably acidified. This percentage
REFERENCES
Aimer, B., W. Dickson, C. Ekstrom, and E. Hornstrom. 1978. Sulfur
pollution and the aquatic ecosystem. Pages 273-311 In J. Nriagu,
editor. Sulfur in the Environment Part II: Ecological Impacts. John
Wiley and Sons, New York.
Aimer, B., et al. 1974. Effects of acidification on Swedish lakes.
Ambio 3: 30-36.
Baker, J., and C. Schofield. 1980. Aluminum toxicity to fish as related
to acid precipitation and Adirondack surface water quality. Pages
292-293 in D. Drablos and A. Tollan, ed. Ecological Impact of Acid
Precipitation: Proc. Int. Symp., Sandefjord, Norway. Acid Rain —
Effects on Forest and Fish Project. Aas, Norway.
Beamish, R., and H. Harvey. 1972. Acidification of the La Cloche
Mountain lakes, Ontario and resulting fish mortalities. J. Fish. Res.
Board Can. 29: 1131-1143.
Dickson, W. 1975. The acidification of Swedish lakes. Inst. Freshw.
Res. Drottningholm Rep. 54: 8-20.
1980. Properties of acidified waters. Pages 75-83 in D.
Drablos and A. Tollan, ed. Ecological Impact of Acid Precipita-
tion: Proc. Int. Symp., Sandefjord, Norway. Acid Precipitation-
Effects on Forest and Fish Project. Aas, Norway.
Galloway, J., and E. Cowling. 1978. The effects of precipitation
on aquatic and terrestrial ecosystems: a proposed precipitation
chemistry network. J. Air Pollut. Control Assoc. 28: 229-235.
Glass, G., and O. Loucks. 1980. Impacts of air pollutants on wilder-
ness areas of northern Minnesota. EPA-600/3-80-044. U.S. En-
viron. Res. Lab., Duluth, Minn.
Harvey, H. 1980. Widespread and diverse changes in the biota of
North American lakes and rivers coincident with acidification.
Pages 93-98 in D. Drablos and A. Tollan, ed. Ecological Impact
of Acid Precipitation: Proc. Int. Symp., Sandefjord, Norway. Acid
Rain — Effects on Forest and Fish Project. Aas, Norway.
86
-------
Acid Rain
Hendrey, G., et al. 1980. Geological and hydrochemical sensitivity
of the eastern United States to acid precipitation.
EPA-600/3-80-024. Environ. Res. Lab. U.S. Environ. Prot. Agen-
cy, Corvallis, Ore.
HendriKsen, A. 1979. A simple approach for identifying and measur-
ing acidification of freshwater. Nature (London) 278: 542-545.
Lillie, R., and J. Mason. 1980. pH and alkalinity of Wisconsin lakes —
a report to the acid deposition task force. Wis. Dep. Nat. Resour.
Rep.
Malmer, N. 1975. Inventering om sjoars forrsurning (Inventories
of lake acidification). Statens Naturvardsverk, Solna, Sweden. PM
676. (In Swedish, English summary).
McFee, W. 1980. Sensitivity of soil regions to acid precipitation.
EPA-600/3-80-013. Environ. Res. Lab. U.S. Environ. Prot. Agen-
cy, Corvallis, Ore.
Muniz, I., and H. Leivestad. 1980. Acidification: effects on freshwater
fish. Pages 84-92 in D. Drablos, and A. Tollan, ed. Ecological Im-
pact of Acid Precipitation: Proc. Int. Symp. Sandefjord, Norway.
Acid Rain — Effects on Forest and Fish Project. Aas, Norway.
National Atmospheric Deposition Program. 1981. NADP report: preci-
pitation chemistry; 4th quarter 1980. Nat. Resour. Ecol. Lab., Col-
orado State Univ., Fort Collins.
Omernik, J., and C. Powers. 1982. Total alkalinity of surface waters—
a national map. EPA-600/D-82-333. U.S. Environ. Prot. Agency,
Corvallis, Ore.
Pfeiffer, M., and P. Festa. 1980. Acidity status of lakes in the Adiron-
dack region of New York in relation to fish resources. N. Y. Dep.
Environ. Conserv. Rep. FW-P168 (10/80).
Scheider, W., D. Jeffries, and P. Dillon. 1979. Effects of acidic
precipitation on precambrian freshwaters in southern Ontario. J.
Great Lakes Res. 5: 45-51.
Schofield, C. 1982. Historical fisheries changes as related to changes
in surface water pH in the United States. Pages 57-67 in T. Haines
and R. Johnson, ed. Proc. Acid Rain/Fish. Symp. Am. Fish. Soc.
Bethesda, Md.
Standard Methods for the Examination of Water and Wastewater.
1975.14th ed. Pub. Health Assoc. Am. Water Works Assoc., Water
Pollut. Control Fed. Washington, D.C.
Stumm, W., and J. Morgan. 1981. Aquatic chemistry. 2nd ed. John
Wiley and Sons, New York.
Watt, W., D. Scott, and S. Ray. 1979. Acidification and other chemical
changes in Halifax County lakes after 21 years. Limmol. Oceanogr.
24: 1154-1161.
Wright, R., and E. Gjessing. 1976. Acid precipitation: changes in
the chemical composition of lakes. Ambio 5: 219-223.
Wright, R., and E. Snekvik. 1978. Acid precipitation: chemistry of
fish populations in 700 lakes in southernmost Nor-way. Int. Verein.
Theoret. Angew. Limnol. Verh. 20: 765-775.
Wright, R., et al. 1977. Regional surveys of small Norwegian lakes.
Acid Rain — Effects on Forest and Fish Project, Report IR33/77.
Aas, Norway.
87
-------
ACID SENSITIVITY OF RESERVOIRS IN THE SOUTHERN
BLUE RIDGE PROVINCE
ALICIA R. LEWIS
HARVEY OLEM
Division of Air and Water Resources
Tennessee Valley Authority
Chattanooga, Tennessee
ABSTRACT
All but two of 36 reservoirs recently sampled in the eastern Tennessee Valley were found to have limited
acid neutralizing capacity. All these reservoirs are located in the Blue Ridge Province of the Southern
Appalachians in Tennessee, Georgia, and North Carolina. Existing water quality data were evaluated as
a first step toward understanding the effect of acidic deposition processes in the Southern Blue Ridge
Province. Studies are currently being performed to evaluate the present water quality conditions in 70
reservoirs in the region. Results of these surveys will be used to select a representative group of reser-
voirs for long-term monitoring.
INTRODUCTION
Acid deposition is one of the most significant environmental
issues currently facing parts of the southeastern United
States. Analyses by various organizations have shown that
acid deposition is occurring in the southeast region and
aquatic resources may be affected now or in the future.
To date, most investigations of the effects of acid deposi-
tion on aquatic resources have focused on glaciated regions
in the northeastern United States, southeastern Canada, and
northern Europe. Yet certain conditions known to exist in
the southeastern United States—particularly the Southern
Blue Ridge Province—indicate that this region too may be
vulnerable to acid deposition.
The Southern Blue Ridge Province (Fig. 1) is located in
the eastern portion of the Tennessee Valley and includes
sections of Georgia, North Carolina, Tennessee, and Virginia.
The Unaka Mountains, Iron Mountains, and Great Smoky
Mountains straddle the North Carolina-Tennessee border
on the western front and the Blue Ridge Mountains lie along
the Basin Divide on the eastern front. The region contains
numerous manmade lakes that may have limited acid
neutralizing capacity.
Figure 1.—Map of Southern Blue Ridge Province.
Certain investigators have used historical water quality
data to assess trends in pH and alkalinity of water bodies
in the region. Hendrey, et al. (1980) showed that mean pH
in 38 headwater streams in North Carolina declined
significantly from pH 6.77 to 6.51 while mean alkalinity declin-
ed significantly from 116 to 80 j^eq/l between the early 1960's-
and 1979. Meinert, et al. (1982) found declines in pH in three
of four streams above 700 meters in North Carolina for
samples collected biweekly during the 1970's. They also
compared data from the 1940's and 1970's for 30 water
bodies and found significant declines in pH from an average
of pH 6.89 to 6.59.
As a first step toward understanding the current and pro-
jected effects of acidic deposition processes on the Southern
Blue Ridge Province, the acid sensitivity of the area's reser-
voirs is being evaluated. It is anticipated that a program will
be conducted on long-term monitoring of a cluster of reser-
voirs selected as being representative of all reservoirs in the
region. These studies will provide data that eventually will
allow a more accurate determination of how changes in water
quality relate to changes in acid deposition.
This paper presents results of initial water quality evalua-
tions of approximately half of the 70 reservoirs in the region
selected on the basis of elevation, minimum surface area,
and the absence of significant local pollution sources.
ACID SENSITIVITY
A reservoir's vulnerability to acidification by acid deposition
is largely a function of the chemical composition and solubility
of the surrounding soils and underlying rocks. Rocks and
soils containing substances capable of assimilating hydrogen
ions and which are relatively soluble in water can act as buf-
fers to acidic deposition. For example, limestones (CaCO3)
and dolomites (CaMg{CO3}2) yield infinite acid neutralizing
capacity, whereas granites (i.e., quartz-SiO2 and
feldspar-KAISi3O8), and related igneous rocks, crystalline
metamorphic rocks (i.e., gneisses and schists) and non-
calcarous sandstone yield minimal buffering. Acid soils in
areas not underlain by limestones are also capable of
assimiliating hydrogen ion inputs from the atmosphere. Weak
acids in these types of soils, from either organic or mineral
88
-------
Acid Rain
origin,,may control the acidity in the soil moisture zone, and
thus control stream acidity.
The Jiydrological characteristics of the terrain also in-
fluence the degree to which a reservoir is affected by acid
deposition. Such characteristics include overland flow ver-
sus'grouhdwater flow, rate'of runoff, soil thickness?soil
permeability/porosity, residence time of water in the soil, and
precipitation patterns in the region (Hendrey et al. 1980).
Reservoir ecosystem processes have also been shown to
be very important in influencing the potential effects of acid
inputs. Brewer and Gpldman (1976) found that primary pro-
duction is very important to the internal production of alkalini-
ty in a lake's epilimnion. Also, internal processes in the
hypolimnion have been shown to either deplete or produce-
alkalinity (Harvey et al. 1981). Processes that occur in lake
sediment? may also affect the acid neutralizing capacity of
lake water.
SOUTHERN BLUE RIDGE PROVINCE
The Southern Blue Ridge Province contains two major
groups of geological formations. One group is crystalline
rocks of Precambrian age consisting mainly of gneisses and
schists cut by later series of granites, pegmatites, and basic.
intrusives. The other group is comprised of highly metamor-
phosed sedimentary rocks of the Cambrian period. These ,
are principally slates, quartzites, conglomerates, sandstones,
graywackfes, and marble. Neither of these rock formations
gives area surface waters significant buffering capacity.
The degree to which weak acids in soils provide buffer-
ing capacity to waters in the region is not very well
understood.
Reservoir Water Quality
Fifty-eight reservoirs and ponds were selected for study on
the basis of a minimum surface area of 2 hectares at an
elevation above 750 meters and the absence of significant
local pollution sources. Also included for evaluation were 12
larger reservoirs below 750 meters. Because of their small
size and difficulty of access, most of the 58 reservoirs had
little or no previous water quality data. Useful water quality
information was found only on the 12 larger reservoirs. Table
1 lists the alkalinity and pH characteristics of the water
discharged through the hydroelectric power turbines of these
12 reservoirs.
Total alkalinity has been proposed by certain researchers
as an index of sensitivity because it is a good indicator of
the acid neutralizing capacity of water-bodies and thus their
relative tolerance to acid inputs (Hendrey et al. 1980; Omer-
nik and Powers, 1982). It has been suggested that water
bodies with alkalinites below 200 ^eq/l may be considered
sensitive to acid inputs. Ten of the 12 larger reservoirs had
average alkalinity values below 200 /^eq/l. Wilbur and
Watauga Reservoirs, with average alkalinities of 510 and 500
^eq/l, respectively, do not appear to be acid'Sensitive.
Studies are now being conducted to evaluate the present
water quality conditions in all 70 selected reservoirs and
ponds. Thirty-six of these were sampled in the fall of 1982.
The remaining reservoirs are being sampled in the spring
of 1983 along with a subset of those reservoirs sampled dur-
ing fall 1982. Table 2 summarizes the water quality data col-
lected during the fall sampling. An attempt was rnade to col-
lect samples at different depths at the deepest point in each
water body. During the fall sampling, all reservoirs, except
Watauga.'were found to be completely mixed and thus data
are reported only for samples collected in the epilimnion.
Sampling and analytical procedures followed those described
in the Protocol of the Aquatic Effects Tas^ Group ,of the In-
teragency Task Force on Acid Precipitation (1982).
Variations were found between field arid laboratory data.
Most pH values were found to be higher in the laboratory
than in the field. This increase in pH could be caused by
biological activity or loss of carbon dioxide to theiatmosphere
during the time between field and laboratory measurements.
Similar reactions occurring in the samples during transit time
could have caused differences in field and laboratory alkalini-
ty and conductivity values. The differences in the alkalinity
values could also have been caused by different methods
of analysis. Alkalinity in the field was measured by a low
alkalinity method (Standard Methods, 1980), while the
laboratory used the Gran titration technique (Aquat. Effects
Task Group, 1982).
Examination of cross-correlation coefficients among the
data revealed only a few statistically significant relationships.
A correlation between elevation and color was found
(r = 0.68, p<0.001), a possible index of organic acids. This
was expected because coniferous vegetation is more abun-
dant in the higher elevations and thus there should be more
natural organic acids. Conductivity showed a high correla-
tion with alkalinity (r = 0.93, p<0.001) and, therefore, might
also be used as an index of a reservoir's sensitivity.
Table 1. — Alkalinity and pH characteristics of reservoirs in the Southern Blue Ridge Province8
pH
Reservoir
Apalachia
Blue Ridge
Calderwood
Chatuge
Cheoah
Chilhowee
Fontana
Hiwassee
Nottely
Santeetlah
Wilbur
Watauga
Mean
6.5
6.4
6.3
6.3
6.3
6.4
6.3
6.4
6.4
6.2
7.0
7.0
Range
5.5-7.2
5.6-7.0
5.2-7.4
5.2-7.0
4.4-7.3
5.4-7.2
4.1-7.1
5.3-7.1
5.6-7.1
4.3-6.7
6.2-7.5
5.8-8.0
Alkalinity, ^eq/l
Mean
150
130
120
140
110
140
140
150
150
110
510
500
Range
40-320
20-320
40-180
20-280
< 20-320
40-760
< 20-920
<20-320
60-240
< 20-220
420-640
46-660
Number
39
55
37
52
38
60
60
48
55
36
36
70
Period of record
1974-1977
1974-1980
1974-1977
1974-1980
1974-1977
1974-1980
1974-1980
1974-1980
1974-1980
1974-1977
1974-1977
1974-1980
3 Monthly samples lor water discharged through the hydroelectric turbines of each reservoir
89
-------
Lake Restoration, Protection and Management
Table 2. — Summary of present water quality conditions in selected reservoirs of the Southern Blue Ridge Province.
Parameter
Mean
Median
Minimum
Maximum
Number
Temperature, °C
Dissolved oxygen, mg/l
Field conductivity, ^mhos/cm
Lab conductivity, ^mhos/cm
Field pH
Lab pH
Field alkalinity, ^eq/l
Lab alkalinity, ^eq/l
Strong acidity, jjeq/l
Weak acidity, j^eq/l
True color, PCU
Apparent color, PCU
Fluoride, mg/l
Chloride, mg/l
Phosphate, mg/l
Nitrate, mg/l
Sulfate, mg/l
Calcium, mg/l
Magnesium, mg/l
Sodium, mg/l
Potassium, mg/l
Dissolved aluminum, ^g/l
14.3
8.0
23
25
6.1
6.68
146
147
<10
59
21
30
0.02
1.3
<0.04
0.34
1.7
1.9
0.71
1.3
0.67
<50
13.65
7.8
20
23.5
6.1
6.82
122
130
<10
60
14
24.5
0.02
0.95
<0.04
0.125
1.45
1.65
0.54
1.2
0.57
<50
7.3
5.5
9
10
5.7
5.87
10
20
<10
20
2
3
<0.02
0.6
<0.04
<0.04
0.6
0.45
0.17
0.74
0.30
<50
20.3
10.3
78
85
7.1
7.40
600
560
<10
120
90
110
0.06
2.8
<0.04
2.5
5.8
7.9
3.8
2.1
2.4
<50
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
36
Long-Term Monitoring
Of the 36 reservoirs sampled during the fall of 1982,12 were
chosen for further sampling in the spring of 1983. The
parameters used in selecting a representative group of reser-
voirs were location, size, elevation, field pH, and field alkalini-
ty. Recommendations of the survey crews were also con-
sidered. Table 3 lists the 36 reservoirs sampled along with
the parameters used in the selection process. Figure 2 shows
the locations of the 36 reservoirs.
Wilbur and Watauga were found to be very different from
the other 34 reservoirs sampled. Not only did they have
higher alkalinities, but their ionic concentrations were much
higher than those of the other reservoirs. Therefore, it was
assumed that Wilbur and Watauga are not acid-sensitive
reservoirs, and they were not considered for further study.
The 12 reservoirs chosen to cover a wide range of each
selection parameter are Lake Winfield Scott, Ravenel Lake,
Queens Lake, Big Laurel Creek Lake, Banks Lake, Unnamed
Pond, Blue Ridge Reservoir, Santeetlah Reservoir, Chatuge
Reservoir, Chilhowee Reservoir, Fontana' Reservoir, and
Hiwassee Reservoir.
In the spring of 1983 the remaining 34 of the 70 reser-
voirs will also be sampled. After examining the results of the
spring sampling effort, a final group of 12 reservoirs will be
selected as best representing all acid-sensitive reservoirs in
the Southern Blue Ridge Province. It is anticipated that long-
term monitoring of these reservoirs will be performed quarter-
ly to determine long-term trends and seasonal variations.
Sampling will be conducted in January (during the completely
mixed winter condition), April (prior to summer stratification),
July (during summer stratification), and October (about the
time of fall overturn).
At least five ongoing wet/dry deposition monitoring sta-
tions are also located in the study area. The acid deposition
data collected at these stations will be used to relate changes
in water quality to changes in the acidity of deposition. In
addition, it is anticipated that a detailed description of the
features of each watershed will be obtained.
CONCLUSIONS
1. Based on limited data, all 36 reservoirs sampled in the
fall of 1982, except Wilbur and Watauga, had alkalinities
below 300 f^eq/l, and almost all of them had alkalinities below
200 neq/l.
2. The field pH values of the 36 reservoirs sampled rang-
ed from pH 5.7 to 7.1.
3. In the spring of 1983 the remaining 34 of the 70 reser-
voirs selected for study in the region will be sampled along
with a subset of 12 of those sampled during fall 1982.
4. It is anticipated that a representative group of 12 reser-
voirs will be selected for future long-term monitoring.
Figure 2.—Location of the 36 reservoirs sampled in the Southern
Blue Ridge Province.
90
-------
Acid Rain
ACKNOWLEDGEMENTS: This study was supported with funds pro-
vided by the U.S. Environmental Protection Agency, Environmen-
tal Research Laboratory, Corvallis, Ore. under Interagency Agree-
ment No. AD-64-F-2A323, TV-60035A.
REFERENCES
Aquatic Effects Task Group. 1982. Sampling and analysis protocol
for chemical characteristics of lakes and streams sensitive to acid
deposition. Environ. Res. Lab., U.S. Environ. Prot. Agency, Cor-
vallis, Ore.
Brewer, P.G., and J.C. Goldman. 1976. Alkalinity changes generated
by phytoplankton growth. Limnol. Oceanogr. 21:108-117.
Harvey, H.H., et al. 1981. Acidification in the Canadian aquatic en-
vironment: Scientific criterion for an assessment of the effects of
acid deposition on aquatic systems. Nat. Res. Counc. Can. Rep.
No. 18475.
Hendrey, G.R., et al. 1980. Geological and nydrocnemical sensitivity
of the Eastern United States to acid precipitation. EPA
600/3-80-024. U.S. Environ. Prot. Agency, Washington, D.C.
Meinert, D.L., et al. 1982. A review of water quality data in acid
sensitive watersheds within the Tennessee Valley.
TVA/ONR/WR-82/10. Tenn. Valley Auth.
Omernik, J.M., and C.F. Powers. 1982. Total alkalinity of surface
waters—a national map. EPA-600/D-82-333. U.S. Environ. Prot.
Agency, Washington, D.C.
Standard Methods for the Examination of Water and Wastewater.
1980. 15th ed. Am. Pub. Health. Assoc., Washington, D.C.
Table 3. — Characteristics of the 36 reservoirs sampled in the Southern Blue Ridge Province8
Location Reservoir
number name
1 Lake Winfield Scottb
2 Woody Lake
3 Lake Sequoyah
4 Ravenel Lakeb
5 Mirror Lake
6 Cliffside Lake
7 Thorpe Reservoir
8 Queens Lakeb
9 Moore Creek Lake 1
10 Moore Creek Lake 2
11 Nantahala Reservoir
12 Big Laurel Creek Lakeb
13 Banks Lakeb
14 Jeffers Lake
1 5 Lake Arrowhead
16 Long Lake
17 Flat Top Lake
18 Miller Lake
19 Ripshin Lake
20 Unnamed Pondb
21 Price Lake
22 Sims Pond
23 Lance Creek Lake
24 Unnamed Lake
25 Apalachia Reservoir
26 Blue Ridge Reservoir"
27 Calderwood Reservoir
28 Chatuge Reservoirb
29 Cheoah Reservoir
30 Chilhowee Reservoirb
31 Fontana Reservoir11
32 Hiwassee Reservoir"
33 Nottely Reservoir
34 Santeetlah Reservoir"
35 Wilbur Reservoir
36 Watauga Reservoir
Elevation
(meters)
875
844
1097
1177
1109
1085
1064
914
1042
1030
917
1158
777
762
847
914
1204
1097
1067
1067
1030
1042
1122
1097
390
515
331
587
389
266
521
465
542
591
503
597
Surface area
(hectares)
10
13
32
9
16
3
591
14
4
4
638
2
2
2
2
2
4
2
24
2
16
2
2
2
445
1,332
219
2,854
256
709
4,308
2,466
1,692
1,154
29
2,603
PH
6.1
6.1
6.0
5.9
5.9
5.9
6.0
6.3
6.1
6.0
6.0
6.8
6.1
6.5
6.2
6.5
6.0
6.7
6.9
5.8
6.8
6.8
6.8
7.0
6.2
6.0
5.8
5.9
5.7
5.9
5.9
6.1
5.9
6.0
6.8
7.1
Total
alkalinity
(Meq/l)
140
124
70
84
60
44
90
250
150
160
140
100
10
150
62
80
100
220
210
20
70
100
120
180
144
110
126
110
120
156
100
150
144
210
570
600
8 Samples collected at aproximately one halt meter beneath the water surface and generally at the deepest point in the water body.
b Selected lor tuture study.
91
-------
ONTARIO'S EXPERIMENTAL NEUTRALIZATION PROGRAM
GARETH GOODCHILD
Ontario Ministry of Natural Resources
Toronto, Ontario
JAMES G. HAMILTON
Booth Aquatic Research Group Inc.
Toronto, Ontario
ABSTRACT
In July 1981 the Ontario Ministries of Natural Resources and the Environment commissioned a 5-year
study of the feasibility of lake neutralization. The experimental neutralization study has been established
as a cooperative investigation between the two ministries and is being coordinated by Booth Aquatic
Research Group Inc. of Toronto. This paper describes briefly the rationale for initiating the neutraliza-
tion experiment and the objectives of the study. The project lakes are also described as are the criteria
used for selecting the lakes. A brief description of the study design is given along with some of the
neutralization options and strategies.
INTRODUCTION
Acidic precipitation in Ontario has caused the acidification
of a number of lakes in the Sudbury area (Gorham and Gor-
don, 1960) resulting in destruction of aquatic life (Beamish
and Harvey, 1972; Beamish, 1974).
Recent studies have revealed that the "acid rain
phenomenon" was much more widespread than had
previously been believed (Dillon et al. 1978), and that
smelting operations in Sudbury were not the sole source of
acidic emissions. Precipitation events in central Ontario with
pH values< 4.0 occur regularly (Scheider et al. 1980; Dillon
et al. 1978). Furthermore, spring snowmelt can be highly
acidic causing "acid pulse" conditions, a situation highly
dangerous to aquatic life in streams and littoral areas of lakes
on the pre-Cambrian shield (Jeffries et al. 1979).
The concern that acidic precipitation was not a localized
phenomenon (Scheider et al. 1980) and that abatement of
emissions was going to take time prompted the government
of Ontario to establish a mechanism to coordinate the various
acid precipitation studies already underway and to determine
what further studies were needed. The objective of the Acid
Precipitation in Ontario Study (APIOS), is to identify the
sources and define the consequence of acid deposition in
Ontario, and to provide a basis for the development of an
effective program to protect the environment. The study will
also evaluate interim mitigative programs both to preserve
and rehabilitate stressed aquatic and terrestrial ecosystems.
Consequently, an experimental neutralization study was in-
itiated by the Ontario Ministries of Natural Resources and
of the Environment. The study is being coordinated by Booth
Aquatic Research Group Inc. of Toronto.
Why Neutralize?
To use neutralization to protect all or most of Ontario's
threatened lakes from acidic precipitation is impossible. The
majority of Ontario's lakes which are acidified or threat-
ened by acidification are not easily accessible; costs to
neutralize them would be prohibitive. For example, Swedish
estimates for neutralization are as high as $50 million an-
nually (Sverdrup, pers. comm.). Major reasons for implemen-
ting the experimental neutralization program in Ontario are:
1. Interim Mitigation—In the event of delays in abatement,
it has been deemed necessary to develop a means of slow-
ing the rate of acidification on particular lakes.
2. Protection of Gene Pools—To protect unique fish
stocks in threatened lakes.
3. Socioeconomics—There is an immense public
pressure to start constructive programs immediately to pro-
tect lakes from acidification; however, the feasibility, prac-
ticality and cost effectiveness has not been fully explored
in Ontario.
PREVIOUS ONTARIO STUDIES
In 1975 the Ministry of the Environment experimentally
neutralized four lakes in the Sudbury area using both
hydrated lime and a combination of calcite and hydrated lime.
This latter combination maintained improved water quality
for a longer period of time. Phytoplankton stocks were im-
mediately reduced by the addition of the base and did not
recover to pre-treatment levels. Zooplankton stocks were also
reduced following neutralization, and biomass failed to
recover before the lakes had reacidified. Concentration of
metals such as nickel and copper in the reclaimed waters,
however, effectively eliminated the opportunity to restore
a sport fishery in these lakes (Van and Dillon, 1981).
In 1979 the Ontario Ministries of the Environment and
Natural Resources studied adding crushed limestone to a
streambed to protect rainbow trout (Salmo gairdneri) eggs
during a spring melt (Keller and Gunn, 1982). Addition of
crushed limestone resulted in significantly increased pH,
alkalinity, conductivity, Ca and Mg ions. The limestone was
effective during periods of low flow but under high flow con-
ditions egg mortality was high, demonstrating the difficulty
in mitigating episodic events at critical periods such as spring
runoff. In subsequent studies the eggs of brook trout
(Salvelinus fontinalis), lake trout (S. namaycusti), and rain-
bow trout incubated over winter in crushed limestone showed
a substantial increase in emergence and survival when com-
pared with those in mixed noncalcareous gravel at the same
site (Gunn and Keller, 1980, 1981).
92
-------
PROGRAM OBJECTIVE
The present experimental lake neutralization study is an ex-
tension of previous Ontario studies. The study will address
fisheries concerns and resource management implications
and will investigate the changes caused by neutralization.
Emphasis will, however, be on the effects on key sport fish
species, such as lake trout
The goal is to rehabilitate an acidified lake and to protect
endangered lakes. A year will be spent studying the acidified
lake before it is neutralized in the summer of 1983, and then
the remaining 3 years will be spent studying the effects of
neutralization. Two endangered lakes are being studied and
they will not be neutralized before the spring of 1984. Until
then pre-neutralization studies will characterize the present
biological physical and chemical parameters. Following
neutrafization, time will be spent studying effects on the lakes
and associated biota.
CRITERIA FOR LAKE SELECTION
How and Why Criteria Chosen
Acidic Lake
Fish community. Due to their susceptibility to acidification,
Salmonid fisheries (particularly lake trout) are of most con-
cern to fisheries managers. The search for a study lake was
therefore directed toward a lake previously supporting a self-
sustaining lake trout population.
Chemical and physical characteristics.
1. pH—pH5 is considered to be the lowest level tolerated
by most species of fish (Schofield, 1976); some lake trout
populations are known to be eliminated at levels above pH 5.
2. AlkaTmity—A reserve alkalinity of less than 10 ^eq/l
would provide fittie buffering capacity and any episodic acidic
event such as spring melt would have an impact on the lake.
3. Aluminum—Aluminum leached from poorly buffered
watersheds represents a serious threat to fish in acidic waters
(Baker and Schofield, 1980). Additionally, an inverse rela-
AckJ Rain
tionship exists between aluminum concentration and lake
pH during periods of low pH (<7.0) (Driscoll et al. 1980).
Therefore, the aluminum problem will be investigated
because acidic lakes having aluminum concentrations of
>50 ng/l are common in the Sudbury area (Dillon et al.
1980).
4. Low concentration of other metals—The Sudbury area
of Ontario suffers extensively from fallout of local smelter
operations, with emissions high in levels of metals such as
copper and nickel. The study will not attempt neutralization
of a lake containing high metal concentrations as it is
recognized that high metal loads can interfere with re-
establishing a fishery (Van and Dillon, 1981).
5. 50-200 ha in size— Neutralization of lakes can involve
the input of large amounts of neutralizing agents. Considera-
tion of logistics therefore dictated that relatively small lakes
be chosen in spite of the fact that they may be marginal for
lake trout.
6. Retention time—Lakes with a small surface area to large
watershed area ratio flush rapidly, therefore, reducing the
long-term effectiveness of neutralization. Lakes chosen for
the experiment should therefore have a retention time that
allows some degree of re-acidification but not total re-
acidification over the length of the study.
Socioeconomic.
1. Minimal cottage development—Lakes with extensive cot-
tage development and recreational activity can make
research difficult because of interference by anglers and
boaters. These lakes were avoided in the selection.
2. Reasonable accessibility—The high level of investiga-
tion and the potential for applying large amounts (in excess
of 80 tonnes) of neutralizing agent would suggest that the
study lakes be reasonably accessible because of the high
cost of transportation.
3. Ability to close fishery—If a lake were neutralized and
a healthy trout population made available, the fish could
become too attractive to fishermen before the effects of
neutralization on fish could be fully identified. The selected
CRITERIA FOR LAKE SELECTION
Acid lake
Previous history of a self-
sustaining lake trout population.
Endangered lake
Fish community
Self-sustaining lake trout population
or lake trout/brook trout combination
present.
Chemical and physical characteristics
pH <5
Alkalinity < 10 peg/1
Aluminum >50 /jgfl
Low concentration of heavy metals
Surface area 50 - 200 ha in size
Retention time > 1 year
minimal cottage development
reasonable accessibility
ability to dose fishery for experiment
duration
a lake located in the Sudbury area
PH
Alkalinity
Aluminum
Conductivity
Color
Surface area
Retention time
<6.5 and >5.0
<20 yeg/l
<40 ^mhos/cm
•< 20 Hazen units
>50 haand<300
>1 yr and <13
Sotioeconomics
minimal cottage development
reasonable accessibility
ability to dose existing fishery for
duration of experiment
a lake located in the Muskoka-
Haliburton area.
93
-------
Lake Restoration, Protection and Management
lake therefore should be one which would draw little public
opposition to its closure to angling and one where the closure
could be easily controlled and enforced.
Endangered Lake
Fish community. Many lakes that are endangered on the
pre-Cambrian Shield area of Ontario contain either laketrout
or both lake trout and brook trout. The search for study lakes
centered on these species.
Physical characteristics. These are similar to the acidic
lake selection with the exception of conductivity and color.
Some lakes on the pfe-Cambrian shield are naturally
dystrophic because of their proximity to bog areas. It was
felt that they would not represent typical lakes because their
acidity would not have been entirely due to acidic
precipitation.
Socio-economic characteristics. These also are similar
to the acidic lake section except that the lakes are located
in the Muskoka-Haliburton area/This area represents a
prime recreational area of Southern Ontario and also is an
area most threatened by high deposition rate of acidic
precipitation. The lakes are typically sensitive to acid input.
The Study Lakes Selection
The search for suitable lakes was aided by.the,Lake Inven-
tory Data Base of the Ontario Fisheries Information System
(OFIS) described by Goodchild and Gale (In press). Key
criteria were checked against the computer file of lakes in-
ventoried by the Ministry of Natural Resources and many
lakes were identified as candidates. The extensive water
sampling program at the Ministry of the Environment also
produced several candidates as did suggestions by in-
vestigators in the field.
Acidic Study Lake: Bowland Lake (lat: N47°05';
long: W80°50')
Surface area 108 ha
Maximum depth 28.0 m
Mean depth 7.0 m
Retention time 2.0 years
pH range 4.9 - 5.3
Alkalinity range -9.0-+9.0 ^eg/l
Aluminum range 120 - 170 ^ig/l
Copper range 2 - 7 ^g/l
Nickel range 1 - 5 ^g/l
Conductivity 39.0 ^nhos/cm
Color (true) 2.7 Hazen units
The acidic study lake is located approximately 65 km north-
east of Sudbury. This lake supported lake trout until the mid-60's;
however the only known fish species present now is an abundant
population of yellow perch (Perca flavescens).
Endangered Study Lakes:
Miskokway Lake (lat: N45°35'; long: W80°10')
Surface area 237.6 ha
Maximum depth 42.~0 m
Mean depth 14.8 m
Retention time 1.77 yrs.
pH range 6.15
Alkalinity
Aluminum
Conductivity
Color (true)
20.0
23 ^mhos/cm
6 Hazen units
Miskokway Lake is a double basin lake situated approximate-
ly 40 km nofth-northwest of Parry Sound. |t contains a
natural lake trout population in addition to largemouth bass
(Micropterus salmoides), smallmbuth bass (M. ddlomieui),
yellow perch, and brown bullhead (Icta turus nebulosus). A
number of other non-game species inhabit the lake. The
fishery for lake trout is closed during the winter.
Trout Lake (lat: N45°35'; long: W80°10')
Surface area . 290.0 jia. Alkalinity^ . 20.0
Maximum depth 37.'8 rn:1 Aluminum'! 20jig/I
;Mean'deptn 11.6'rri 'Conductivity ! 26 ^mhos/cm
Retention time 2.67 years Color (true); 6 Hazen units
pH range 6.05 <: . • .
Trout Lake is a large lake situated just south of Miskokway,
approximately 35 km ;from Parry Sound. The winter lake trout
fishery is also closed. Fish species include lake trout,
smallmouth bass, yellow perch,, and a number of other norv
game species.
Study Design
Limnology
The water chemistry sarripling began during May 1982 in
the acidified lake and in September at the endangered lakes.
Bowland Lake is sampled every 2 weeks, while the en-
dangered lakes are visited monthly;
In addition, at least one of the lakes .will be intensively
sampled during the periods of fall rain and spring melt when
acid input is expected to peak.
Phytoplankton will be surveyed to document changes in
chlorophyll concentration, particle size distribution, arid tax-
onomic classification. Zooplankton communities are expected
to be less diverse in acidic waters and their importance as
a forage base for fish warrants a close investigation of
changes in species abundance and/or biqmass/ ,
Periphyton is particularly abundant in acid lakes (Jackson,
pers. comm.). These communities will be Identified and
mapped to denote changes that may be brought about by
neutralization, as will the macrophyte communities.
Fisheries .'.. ~. "_,, \ ; ' ''V. ^ ' , ,'• ;! •" '";'
A description of existing fish communities will be acquired
through extensive sampling with a variety of gear. Sport fish
populations will be estimated through mark recapture tech-
niques and index fishing, while prey fish populations will be
determined only in relative terms using index fishing
methods.
Various calcified tissues will be collected from represen-
tative species to determine age and growth and to compare
changes in growth that might occur.
Bioassay
Changes in reproductive success will be investigated through
the incubation of lake trout eggs in each of the respective
lakes prior to and following neutralization. Examination of
incubators at various stages of development will determine
sensitive life stages that must be protected from the input
of acidic water.
Additionally, caging experiments involving hatchery reared
lake trout will be used to characterize water conditions in
these study lakes.
Neutralization Strategy: Various materials and techniques
have been used to neutralize waters in both North America
and Scandinavia (Fraser and Britt, 1982).
Most commonly used for neutralization are calcite CaCO3,
slaked or hydrated lime (Ca(OH)j.), and quicklime (CaO).
Other materials used less commonly (unranked) are olivine
(MgFeSi), calcium silicate slags, cement plant bypass, waste
sludge, fly-ash, soda (Na2Cp3J; arid dolomitic lime
MgCa(Cp3)2-, When considering Choice of these materials,
various concerns must be recognized;
1. Safety in handling—For instance, quicklime and
hydrated lime can be extremely dangerous as opposed to
calcite which is less caustic and safer to handle.
94
-------
Acid Rain
2. Solubility—Solubility in water varies between materials
but this factor is most often controlled through choice of par-
ticle size (Sverdrup, 1982). Increased rate of solubility of
some materials, however, may also result in a situation refer-
red to as "pH shock" after adding the base to a water body.
The drastic effects to the biota of quickly elevated pH have
to be considered and mitigated if possible.
3. Application—Finally, the means available for applica-
tion must be considered. If the lake is accessible by road
then the costs of application and amount of neutralizing agent
chosen is of less concern. Enclosure experiments involving
the actual neutralization of columns of lake water will deter-
mine the neutralizing agent which is the most effective and
which exhibits the least detrimental effect on the biota.
Enclosures will be neutralized at various rates with agents
and particle sizes suitable for the particular lake situation.
Zooplankton and phytoplankton in the corrals will be sampled
along with regular water chemistry sampling. In addition, fish
species, including lake trout, will be exposed to determine
the immediate effects of neutralization and the water quali-
ty following neutralization.
Discussion
Data have yet to be interpreted from the first field season
so no attempt has been made to present and discuss results
in this paper. The study design presented will continue to
be adjusted as data are interpreted and new specific
research requirements are identified.
ACKNOWLEDGEMENTS: We would like to thank the Steering Com-
mittee on Experimental Lake Neutralization, the members of the
Neutralization Task Force and Booth Aquatic Research Group Inc.
for allowing us to report on this project and for reviewing this paper.
Members are as follows: Steering Committee: T. G. Brydges, D. P.
Dodge, Art Holder, Don Jeffs; Neutralization Task Force: Gail Beggs,
Gord Craig, Peter Dillon, John Gunn, Bill Keller, Jim Maclean, Bernie
Neary, Ken Nichols, Henk Rietveld (chairman), Frank Tomassinni,
Norm Van, Mike Young; Booth Aquatic Research Group Inc., Gillian
Booth and Lew Molot. Funds to travel to the NALMS conference
were provided by the Ministry of Natural Resources and Booth
Aquatic Research Group Inc. Further information about this report
is available from either T. G. Brydges of the Ontario Ministry of the
Environment or D. P. Dodge, Ontario Ministry of Natural Resources
in Toronto.
REFERENCES
Baker, J.P., and C.L. Schofield. 1980. Aluminum toxicity to fish as
related to acid precipitation and Adirondack surface water quali-
ty. In D. Drablos and T. Tollan, eds. Ecological Impact of Acid
Precipitation: Proc. Int. Conf., Sandefjord, Norway, March 11-14.
Oslo.
Beamish, R.J. 1974. The loss of fish populations from unexploited
remote lakes in Ontario, Canada as a consequence of atmospheric
fallout of acid. Water Res. 8: 85-95.
1976. Acidification of lakes in Canada by acid precipitation
and the resulting effects on fishes. Water Air Soil Pollut. 6: 501-514.
Beamish, R.J., and H.H. Harvey. 1972. Acidification of the La Cloche
Mountain Lakes, Ontario and resulting fish mortalities. J. Fish.
Res. Board Can. 29: 1131-1143.
Dillon, P.J. et al. 1978. Acidic precipitation in South-Central Ontario:
Recent observations. J. Fish. Res. Board Can. 35: 809-815.
Dillon, P.J., D.S. Jeffries, W.A. Schneider and N.D. Van. 1980. Some
aspects of acidification in Southern Ontario. In D. Drablos and
T. Tollan, eds. Ecological Impact of Acid Precipitation: Proc. Int.
Conf., Sandefjord, Norway, March 11-14. The SNSF project. Oslo.
Driscoll, C.T. Jr., J.P. Baker, J.J. Bisogni, Jr., and C.L. Schofield.
1980. Effect of aluminum speciation on fish in dilute acidified
waters. Nature, 284: 161-164.
Fraser, J.E., and D.L. Britt. 1982. Liming of acidified waters: a review
of methods and effects on aquatic ecosystems.
FWS/OBS-80/40.13. Fish Wildl. Serv. Kearneysville, W. Va.
Goodchild, G.A., and G.E. Gale. In press. Aquatic habitat inventory—
the Ontario approach to lake surveys. In Symp. Acquisition and
Utilization of Aquatic Habitat Inventory Information. Portland, Ore.,
Oct. 28-30, 1981.
Gorham, E., and A.G. Gordon. 1960. The influence of smelter fumes
upon the chemical composition of lake waters near Sudbury, On-
tario and upon surrounding vegetation. Can. J. Bot. 38: 477-487
Gunn, J.M., and W. Keller. 1980. Enhancement of the survival of
rainbow trout (Sa/mo gairdneri) eggs and fry in an acidic lake
through incubation in limestone. Can. J. Fish. Aquat. Sci. 37:
1522-1530.
. 1981. Emergence and survival of lake trout (Salvelinus
namaycush) and brook trout (S. fontinalis) from artificial substrates
in an acid lake. Ont. Fish. Tech. Rep. Ser. No. 1
Jackson, M.B. 1982. Personal comm. with Jim Hamilton, Booth
Aquatic Research Group Inc.
Jeffries, D.S., C.M. Cox and P.J. Dillon. 1979. Depression of pH
in lakes and streams in central Ontario during snowmelt. J. Fish.
Res. Board Can. 36: 640-646.
Keller, W., and J.M. Gunn. 1982. Experimental neutralization of a
small, seasonally acidic stream using crushed limestone. APIOS
Rep. No. 004/82 Ont. Ministry Environ. Publ.
Scheider, W.A., D.S. Jeffries, and P.J. Dillon. 1980. Bulk deposition
in the Sudbury and Muskoka— Haliburton areas of Ontario dur-
ing the shutdown of Inco Ltd. in Sudbury. Ont. Ministry Environ.
Publ.
Schofield, C.L. 1976. Acid precipitation: effects on fish. Ambio
5: 228-230.
Sverdrup, H.U. 1982. Pers. comm. with Jim Hamilton, Booth Aquatic
Research Group Inc.
Sverdrup, H., and I. Bjerle. 1982. Dissolution of calcite and other
related minerals in acidic aqueous solution in a pH stat. Vatten
38: 59-73.
Van, D., and P.J Dillon. 1981. Neutralization and studies of lakes
and watersheds near Sudbury, Ontario. Ontario Ministry Environ.
(Unpubl. rep.).
95
-------
EFFECTIVENESS AND UNCERTAINTIES ASSOCIATED WITH THE
CHEMICAL NEUTRALIZATION OF ACIDIFIED SURFACE WATERS
DOUGLAS L. BRITT
Director, Environmental Sciences
JAMES E. FRASER
General Research Corp.
McLean, Virginia
ABSTRACT
The chemical neutralization of acidified surface waters is currently being practiced in Scandinavia and
North America. Based upon the reported biological responses to some of these restoration projects, the
application of neutralizing chemicals appears to be effective; however, some problems have been reported.
This presentation describes the various types of liming techniques that have been used to neutralize acidified
surface waters. Characteristics of surface waters and watersheds that may affect the potential success
of liming operations are described. Some criteria for selecting liming as a mitigation option are discussed.
Several critical uncertainties associated with liming are summarized, and a preliminary set of research
recommendations to resolve these uncertainties is outlined.
INTRODUCTION
Acid deposition processes, the combination of acidic
precipitation and acidic dry deposition, have directly and in-
directly affected aquatic resources throughout many areas
of Scandinavia and North America. Among the most impor-
tant consequences of surface water acidification is the loss
of important recreational and commercial fisheries. The direct
effects of acidification on fish include acute mortality (Haines
and Schofield, 1980); reproductive failure (Beamish, 1976);
avoidance behavior (Muniz and Leivestad, 1980); altered
growth rates (Ryan and Harvey, 1977); and various chronic
impairments to body organs, muscle tissue.and skeletal
structures (Daye, 1981; Beamish et al. 1975). Indirect effects
include habitat degradation (Fritz, 1980); changes in the
population structure of preferred food organisms (Tollan,
1980); and the increased bioavailability of aluminum and
other toxic metals (Haines and Schofield, 1980; Wood, 1980).
Acidic deposition has been implicated in the widespread
elimination of entire fish communities in some of the lakes
of the Adirondack Mountains of New York (Schofield, 1976),
LaCloche Mountains of Ontario (Beamish, 1975), and in
some of the coastal rivers of Nova Scotia (Farmer et al. 1980;
Watt, 1980). The situation is even more serious in Scan-
dinavia, where thousands of Swedish and Norwegian lakes
and streams have been affected adversely (Tollan, 1980).
Furthermore, acidic deposition is not expected to decrease
significantly in the near future; therefore, these aquatic
systems probably will realize additional acidic stresses. While
it is generally believed that restricting fossil fuel combustion
emissions of SO2 and NOX at their sources may help
alleviate the acidic deposition problem, it is not certain
whether such emission control actions will occur in the near
term. As a result, other techniques and strategies designed
to mitigate effects of atmospheric acid deposition are need-
ed. One potential approach involves applying lime or other
alkaline material at the receptor site (e.g., lakes and streams)
to neutralize incoming acids.
Liming has been used with apparent effectiveness to
counteract surface water acidification in parts of Scandinavia,
Canada, and the United States (Fraser et al. 1982; Fraser
and Britt, 1982). In Sweden, a major objective of the pro-
gram is to save waters of special value for fishery and nature
conservancy use (Bengtsson, 1982). Similarly, in Canada
(Fraser et al. 1982) and the United States (Blake, 1981), most
liming projects are designed to protect important fisheries.
Liming is also employed to restore fisheries to lakes and
streams that have completely lost their fish populations (in
contrast to protecting surviving populations).
Although liming is only a temporary solution to a long-term
problem, it can be used to reduce the loss of important
fishery resources while longer-term solutions are developed
and implemented. In general, liming has been shown to im-
prove physical and chemical conditions and enhance the
biological recovery of aquatic ecosystems affected by
acidification. At the present time, Scandinavia and Canadian
researchers have partially identified some of the key factors
responsible for the successful neutralization of surface
waters. However, some potentially adverse consequences
of liming have been observed (Scheider et al. 1975a; Driscoll
et al. 1982), and several conflicting results of liming projects
also have been reported (Frazer et al. 1982; Fraser and Britt,
1982; Sverdrup, 1982a). These unresolved problems and
conflicting results have been a major focus of our recent
studies; moreover, they indicate the need for additional
research to improve the ability of scientists and resource
managers to select effective and efficient liming strategies.
LIMING MATERIALS
Many alkaline materials have been used to neutralize
acidified surface waters. These range from industrial
byproducts and wastes such as cement plant bypass dusts,
water softening sludge from water treatment facilities, and
fly ash, to the more common reagents (see Table 1) such
as olivine, lye, soda, calcitic and dolomitic limestone, calcitic
and dolomitic lime, lime slags, and lime suspensions (Grahn
and Hultberg, 1975; Scheider et al. 1975a; Edzwald and
DePinto, 1978; Bengtsson et al. 1980; Hultberg and
Andersson, 1981; Theiss, 1981; Bengtsson, 1982; Boyd
1982a).
For most liming situations, limestone (CaCO3), slaked or
hydrated lime [Ca(OH)2], and unslaked lime or quicklime
(CaO) represent the most readily available and chemically
96
-------
Acid Rain
Table 1. — Common reagents for acid neutralization.
Reagent
Empirical formula
Chemical description
Caustic lye
Soda
High calcium limestone
Dolomitic limestone
High calcium quicklime
High calcium hydrated lime
Dolomitic quicklime
Dolomitic hydrated lime
Highly hydrated dolomitic lime
divine
NaOH
Na2C03
CaCO3
CaCO3 • MgCO3
CaO
Ca(OH)2
CaO • MgO
Ca(OH)2-Mg(OH)2
Ca(OH)2 • Mg(OH)2
(Mg, Fe)2SiO4
Sodium hydroxide
Sodium carbonate
Calcium carbonate
Calcium-magnesium carbonate
Calcium oxide
Calcium hydroxide
Calcium-magnesium oxide
Normal dolomitic hydrate
Special dolomitic hydrate
Magnesium iron silicate
Source: Fraser and Britt, 1982.
consistent materials. Although lower in calcium content,
dolomite, dolomitic hydrated lime, and dolomitic quicklime
(each exceeding a 35 percent magnesium content) also are
frequently available; moreover, they have a slightly higher
neutralizing value (i.e., they can neutralize more acidity) than
equivalent weights of their high-calcium counterparts (Lewis
and Boynton, 1976). However, limestones containing more
than 10 percent magnesium carbonate dissolve too slowly
in acid systems to be of practical use in neutralizing most
acidified surface waters (Lewis and Boynton, 1976; Bernhoff,
1979). For comparison of neutralizing values, if pure CaCO3
is assigned a neutralization value of 100 percent, then the
relative neutralizing values of other pure compounds are:
CaO, 179 percent; Ca(OH)2, 136 percent; and CaMg(CO3)2
(dolomite), 109 percent (Boyd, 1982b). Because commercial
limestone is rarely pure CaCO3 the neutralizing value of most
agricultural limestone varies from 85 to 95 percent (Rafaill
and Vogel, 1978).
Although agricultural limestone is not as effective in
neutralizing acidity as quicklime or hydrated lime, aglime has
been the most commonly used material in recent years. This
is undoubtedly because of its several obvious advantages:
It is easy to handle and is noncaustic.
It is relatively inexpensive.
Its rate of dissolution can be reasonably controlled •
It does not produce harmful alkaline conditions.
It is relatively free from harmful contaminants.
It is a natural component of surface water buffering
systems.
Unlike many alkaline materials, aglime can be obtained
in a range of particle sizes. Since small particles react much
faster than larger particles, and since the dissolution rate
of limestone depends upon the pH of the water (the dissolu-
tion rate increases as pH decreases), finer ground aglime
can compensate for slower dissolution rates, especially in
surface waters with pH values greater than 5.7 (Sverdrup,
1982a). Also, aglime can be easily applied either as a powder
or water-based slurry.
APPLICATION STRATEGIES
Deciding how, when, where, and how often to apply liming
materials for effective neutralization are much more difficult
issues than selecting an adequate neutralizing agent.
At the present time several techniques are available for
distributing liming materials in lentic and lotic environments
(Fig. 1). Lime application techniques for lakes include using
trucks, boats, aircraft, and sediment injection systems. (Sedi-
ment injection systems involve the injection of chemicals into
lake sediments to enhance cation exchange with hydrogen
ions in the overlying water column). Techniques used in
rapidly flowing water include aircraft, automated silos, stream
barriers and beds, rotating limestone drums, and limestone
LIMING
MATERIALS
APPLICATION
EQUIPMENT
TARGET
AREAS
• COMMONLY USED OPTION
OCCASIONALLY USED OPTION
Figure 1 .—Summary of techniques used in the application of lime
to aquatic systems and their watersheds.
Source: Adapted from Fraser et al. 1982.
diversion wells. Trucks with compressed air blowers also
have been used to apply aglime to watershed soils surround-
ing acidified surface waters. Many of these application tech-
niques were originally developed to mitigate acid mine
drainage from abandoned coal mines (Fraser and Britt, 1982;
Rafaill and Vogel, 1978).
All of the techniques applicable to lakes and ponds have
proven effective; however, the general experience jn
neutralizing flowing waters has not been encouraging (Fraser
and Britt, 1982). It is difficult to lime streams from aircraft.
Dosing apparatuses, such as automated silos, have proven
to be mechanically unreliable. Stream barriers made of
limestone have been ineffective for several reasons: the
limestone is washed away during spates or it is rapidly deac-
tivated by silt and metal precipitates, and the dissolution rate
of coarse limestone gravel is quite slow under dilute acidic
conditions. Rotary drums are currently employed only on a
pilot scale.
To date the most effective stream liming technique has
been the use of limestone diversion wells in Scandinavia.
Limestone diversion wells are cylindrical concrete wells fill-
97
-------
Lake Restoration, Protection and Management
ed with crushed limestone. They are usually placed in the
stream bed or bank. Water is diverted from the stream into
the bottom of the well. The conveyed water flows upward
and agitates the crushed limestone, leading to abrasion of
limestone particles and neutralization of the outflowing water.
To be effective, this technique requires diversion of approx-
imately 25 percent of the streamflow. It also is subject to
freezing, and consequently deactivation, in the winter and
early spring (Johannassen and Wright, 1982). At the pre-
sent time, most streams subject to substantial seasonal varia-
tions in flow are difficult, if not impossible, to lime effectively
with any existing techniques.
The proper timing and placement of neutralizing agents
are not well understood, but they frequently depend upon
two major factors: (1) the time and location of acidic episodic
events (e.g., snowmelt or autumnal rains); and (2) the tem-
poral and spatial relationships between acidification events
and the critical life stages of important aquatic biota. In dimic-
tic lakes, liming frequently is practiced during the spring over-
turn to take advantage of lake circulation patterns to enhance
mixing and distribution of neutralizing agents; however,
neutralization may occur too late to prevent embryo and fry
mortality in numerous fish species during the potentially
serious pH reductions associated with melting snow (Jeffries
et al. 1979; Johannessen et al. 1980). Spring acidic snowmelt
also creates another problem at this time: the colder
snowmelt water may be less dense than deeper lake water,
and mixing with previously neutralized lake water may be
inhibited to the detriment of aquatic biota inhabiting the top
centimeters of the littoral zone.
The effectiveness of liming is also influenced by the se-
quential relationship of frozen ground-snowfall-snowmelt.
If the ground freezes before snow cover develops, and the
ground remains frozen during snowmelt, then the spring melt
water interacts only slightly with any potentially buffered soils
that may neutralize some of the acidity (Fraser et al. 1982).
To counteract this acidified meltwater, the Swedish govern-
ment recommends overdosing of liming materials applied
directly to surface waters, or applying liming materials to the
snowpack annually (Bengtsson et al. 1980). Another option
that has been infrequently reported is the application of lime
during the autumnal overturn, in combination with the winter
liming of snowpack or lake ice.
Distribution of liming material over an entire lake surface
may be the ideal placement (Sverdrup, 1982b); however,
because of time and resources, this may be infeasible. As
an alternative, lime may be deposited over the deepest por-
tion of the lake, thus allowing the larger particles of CaCO3
more time to react while passing through the water column.
Another alternative, reported by Blake (1981) and Driscoll
et al. (1982), is to distribute limestone in the shallow littoral
zones where maximum turbulence caused by wave action
may enhance the dissolution process. Observations in
several Swedish lakes, however, indicate water movements
in the littoral zone do not increase the dissolution rate or
erode the CaCO3 particles (Sverdrup, 1982b).
Some researchers suggest applications in areas with sand-
y bottoms rather than muddy (heavy loams or clay) bottoms,
because the former will minimize the loss of limestone
through chemical bonding with organic sediments (Boyd,
1982a; Driscoll et al. 1982). If lake size and/or economics
limit a liming program to a small portion of a lake (e.g., a
shallow cove), fish survival still may be enhanced. Hall et
al. (1980) and Muniz and Leivestad (1980) indicate certain
fish species will seek out "havens" of better water quality
LIMING EFFECTS
Although the chemical neutralization of acidified surface
waters generally improves water quality, some chemical
changes may be of potential concern. The fate of toxic
metals, especially aluminum, after liming is of special im-
portance. Aluminum and many other toxic metals are re-
moved from the water column as surface water pH is
raised after base addition (Van and Dillon, 1981). Aluminum
reactions with hydroxides may produce gill damage to fish
as pH values increase from 4.4 to 5.2 (Schofield and Tro-
jnar, 1980). Other studies indicate that inorganic aluminum
is most toxic to fish between pH 5.2 and 5.4 (Baker, 1981;
Baker and Schofield,'1981). Aluminum toxicity during this
period of transition from acidic to circumneutral conditions
has been linked to the mortality of salmon and trout
stocked in lakes immediately after liming (Bengtsson, 1982).
It is advisable, therefore, not to begin stocking programs until
the limed body of water has chemically stabilized for several
weeks.
Perhaps of greater concern is the fate of aluminum that
continues to be leached into neutralized lakes from acidified
watersheds. Theoretically, the sediments of limed lakes may
become environmental sinks for aluminum and other metals
mobilized in the watershed. If a lake is maintained in a cir-
cumneutral condition by liming for many years and then is
allowed to reacidify, the ensuing remobilization of ac-
cumulated toxic metals from the sediments may impose
serious biological consequences.
Aluminum's complexation with organic materials is another
area of concern. Driscoll (1980) and Johnson et al. (1981)
specify that significant correlations have been observed be-
tween organically chelated aluminum and dissolved organic
carbon concentrations. Driscoll et al. (1982) report that coin-
cident with the hydrolysis and precipitation of aluminum
following liming, dissolved organic carbon (DOC) concen-
trations dramatically decrease. These authors suggest that
even if reacidification occurs, organic carbon lost from the
water column may be retained (or metabolized) within the
sediments. Therefore, neutralization caused by the addition
of alkaline materials, followed by future acid loading (i.e.,
reacidification), may result in continuously very low concen-
trations of dissolved organic carbon, thus allowing any
resolubilized aluminum to remain in a toxic inorganic form
(Driscoll et al. 1982).
Nitrogen and especially phosphorus are usually limiting
factors to primary productivity in oligotrophic lakes, including
those that have been acidified or limed. The long-term con-
centrations of these two nutrients are therefore of special
interest after lime treatment. Previous studies of total
phosphorus (TP) concentrations after the application of
alkaline materials have produced contradictory results. In-
creased total phosphorus concentrations following the ad-
dition of lime were reported by Hasler et al. (1951), Waters
(1956), and Wilander and Ahl (1972). However, the lakes
studied by Hasler et al. (1951) and Waters (1956) were high
in humic content, and the precipitation and subsequent
breakdown of the organic matter could have been the source
of phosphorus (Scheider et al. 1975b). Van and Dillon (1981)
report no significant changes in concentration of TP after
liming four lakes near Sudbury, Ontario.
Data from four lakes in Sweden (two limed clearwater
lakes, one limed humic lake, and one acidified control lake)
indicate that phosphorus concentrations were reduced by
approximately 50 percent in all the lakes, limed or acidic
(Hultberg and Andersson, 1981,1982). Increased retention
of phosphorus in surrounding watershed soils, probably as
a result of increased leaching of aluminum, and precipita-
tion of the phosphorus by aluminum in the lake water, are
suggested as explanations for the observed decreases.
Because of this potential problem, Hultberg and
Andersson (1981, 1982) suggest that liming lakes and
streams exclusively, without adding lime to the watershed
and/or phosphorus to the bodies of water, may not restore
aquatic ecosystems to conditions similaMo those existing
98
-------
Acid Rain
before acidification. Bengtsson (1982), however, recently has
reported that although the availability of phosphorus in lakes
may be initially lowered after liming, it eventually increases
again over a period of time. This initial decrease in
phosphorus concentrations after pH elevation probably
results from precipitation with aluminum (Dickson, 1978).
during episodic pH reductions. The probability of finding
chemically buffered havens with elevated acid neutralizing
capacity will increase with lake size and volume regardless
of the fact that volume increased by dilution will dampen the
fluctuations in water chemistry (Muniz and Leivestad, 1980).
Liming to create less acidic or neutral havens has been con-
ducted successfully in Scandinavia (Fraser et al. 1982).
Another liming strategy involves neutralizing acidified
aquatic ecosystems by chemically treating their associated
watersheds. Sweden has actually applied 40 percent of its
total tonnage of liming materials on land adjacent fo the af-
fected surface waters (Bengtsson, 1982). Whether the
acidification of aquatic ecosystems may be ameliorated more
effectively by liming drainage basins or the lakes and streams
themselves, has yet to be determined (Gorham and McFee,
1980); however, it has been suggested that watershed lim-
ing may be the most effective way to counter lake acidifica-
tion from spring and autumnal acid spates (Swedish Ministry
Agric. Environ. Comm. 1982). It has been suggested that
combining surface water with watershed liming may be most
effective where watershed soils are especially susceptible
to the leaching of aluminum (Hultberg and Andersson, 1981).
The major disadvantage of application on land is that the
alkaline dosage must be a hundred times greater than the
dosage required for direct lime application to water to achieve
a comparable acid neutralizing capacity in the surface water
(Bengtsson et al. 1980). Based solely on economics, lime
application directly into the water appears to be a more effi-
cient neutralization technique (Bengtsson et al. 1980).
It is quite important to use liming resources as efficiently
as possible to minimize total costs of a liming project. It is
necessary, of course, to dissolve at least enough lime to
neutralize acidic surface waters. Several methods have been
developed to estimate lime requirements for surface water
neutralization projects (Natl. Fish. Board Sweden, 1982; Sver-
drup, 1982a, b; Driscoll et al. 1982; Pearson and McDon-
nell, 1975; Scheider et al. 1975b; Boyd, 1982a). These tech-
niques vary substantiality in their degree of complexity: they
range from methods based only upon acid-base titrations,
to more sophisticated, computer-assisted, chemical equilibria
models. Factors considered important to the quantification
of base requirements include: H+, alkalinity, aluminum,
humic materials, base saturation of the sediments, water
volume, annual runoff, retention time or flow rates, and water
depth (Fraser and Brittt, 1982).
When lime is applied to surface waters, some fraction of
the lime dissolves immediately and affects neutralization. The
remainder settles to the bottom. If this residual material does
not contribute significantly to the neutralization process im-
mediately, or at a later time, then it is lost or wasted. The
initial dosing of most water bodies is calculated to last several
years. Although some lakes have maintained acid neutraliz-
ing capacity for many years after lime application (Hultberg
and Andersson, 1982), the long-term dissolution of lime has
not been observed in most lakes (Sverdrup, 1982a). Applica-
tions more frequent than initially projected are often need-
ed to maintain acceptable water quality.
Sverdrup (1982a, b) has evaluated theoretical lime dissolu-
tion rates and their relationship to CaCO3 particle size and
pH, to project the amount of alkaline material immediately
available for neutralization. He also has assessed data from
Swedish and Norwegian lake liming projects to determine
the dissolution rate of residual CaCO3 and its long-term ef-
fects on alkalinity. He concludes that residual CaCO3 is not
efficient in neutralizing surface waters after liming because
sedimentation processes and inhibition by humic materials
will stop dissolution completely within a relatively short time.
The presence of aluminum and iron also has been reported
to inhibit CaCO3 dissolution through the formation of metal
hydroxide coatings (Pearson and McDonnell, 1975).
The reported effects of acid neutralization on total nitrogen
(TN) concentrations are less discordant in the literature and
do not indicate any potential limiting effects on primary pro-
ductivity (Van and Dillon, 1981; Hultberg and Andersson,
1981).
The short-term biological consequences of liming are
generally favorable. Although conclusive data are lacking,
it has been suggested that liming may facilitate the decom-
position of organic matter and enhance the overall rate of
nutrient cycling by increasing the populations of aerobic
heterotrophic bacteria (Scheider and Dillon, 1976; Bengtsson,
1982). The diversity of planktonic algae increase after
neutralization, with a concurrent decrease in acidophilic algae
and mosses (Grahn et al. 1974; Hultberg and Andersson,
1982).
The decrease in mats of acidophilic algae and benthic
mosses such as Sphagnum are believed to improve fish
growth and increase the availability of nutrients (Hultberg and
Grahn, 1975; Hultberg and Andersson, 1981).
Population changes in invertebrate fauna after neutraliza-
tion of acidic waters are somewhat similar to the changes
affecting phytoplankton standing stock. Following the addi-
tion of alkaline material, the biomass of zooplankton usual-
ly declines immediately, probably because of the rapid rise
in pH, followed by an eventual recovery up to less than pre-
neutralization levels (Van and Dillon, 1981) or occasionally
greater than pre-neutralization levels (Bengtsson et al. 1980;
Hultberg and Andersson 1981, 1982). However, recovery of
zooplankton biomass, when compared to phytoplankton,
often takes several years instead of several months
(Bengtsson et al. 1980; Van and Dillon, 1981), probably
because of the slower rates of reproduction in zooplankton
(Scheider and Dillon, 1976).
Changes in taxonomic composition of zooplankton also
occur: an initial predominance of cladocerans usually
changes to a predominance of copepods after neutraliza-
tion (Hultberg and Andersson, 1981; Van and Dillon, 1981).
Van and Dillon (1981) report that fertilization of two Sudbury
lakes (addition of 5.0 to 7.0 mg/m3 of phosphorus), follow-
ing treatment with liming materials, enhanced zooplankton
biomass recovery. They also noted that the zooplankton stan-
ding stock of less acidic Nelson Lake remained relatively con-
stant after lime treatment with no phosphorus addition.
An immediate and substantial reduction in the biomass
of benthic macroinvertebrates has been observed after lim-
ing acidified lakes (Scheider and Dillon, 1976; Hultberg and
Andersson, 1981). Acid-tolerant chironomids are replaced
over a period of 1 to 2 years by proportional increases of
less acid tolerant species found in normal circumneutral lakes
(Hultberg and Andersson, 1981). Bengtsson et al. (1980) in-
dicate that benthic communities associated with pre-acidic
conditions will eventually return over time if immigration is
possible. Hultberg and Andersson (1981) note that benthic
invaders could successfully establish good populations within
2 years after liming.
Because some benthic organisms are completely
eradicated by acidification, the reintroductton of invertebrates
to several limed lakes has been attempted in Sweden,
generally with success (Hultberg and Andersson, 1981,
1982). Survival and growth of crayfish have been especially
successful in Sweden, with large areas of limed waters now
being productive (Bengtsson, 1982). In addition to a
replenished supply of fish food, Hultberg and Andersson
(1981) report that the reintroduction of invertebrate species
99
-------
Lake Restoration, Protection and Management
may increase diversity and thus may create a more stable
ecosystem and enhance the positive effects of liming in
special cases.
Aside from occasional fish mortality caused by metal tox-
icity, the usual response of fish to liming is that most species
affected by acidification respond positively and directly to
liming (Table 2). The potential problems of gradual episodic
reacidification (e.g., from snowmelt) and sudden remobiliza-
tton of toxic metals, however, must be prevented. The means
to ensure such protection in flowing waters and lakes with
short retention times does not yet exist.
SPECIAL CONSIDERATIONS
Currently, it does not seem feasible, scientifically sound, or
economically justifiable to lime all acid-sensitive aquatic
ecosystems within a region experiencing significant acid
deposition; therefore, careful consideration should be given
to several environmental characteristics of sensitive or af-
fected surface waters prior to the selection of liming targets.
Such environmental characteristics usually include biological
factors, (e.g., the current or historic ability of the system to
support a viable and important fishery); sociological factors
(e.g., recreational area and public access to the waters);
chemical factors (e.g., present pH, alkalinity, and acid
loadings); and physical factors that influence the economics
of a liming operation (e.g., geographic location, type of lim-
ing material available, quantities required, and reapplication
rates). These general characteristics along with other poten-
tially confounding factors need to be considered on a site-
specific basis before the feasibility of liming can be adequate-
ly assessed.
Furthermore, naturally acidic aquatic ecosystems often
contain biological communities that are stable, well adapted,
Table 2. — Examples of fish populations reported to be affected favorably by liming.
Family and species
Observed effect after liming treatment
References
SALMONIDAE
Lake trout (Salvelinus namaycush)
Brook trout (Salvelinus fontinalis)
Arctic char (Salvelinus alpinus)
Rainbow trout (Sa/mo gairdneri)
Brown trout (Sa/mo trutta)
Atlantic salmon (Sa/mo salar)
Cisco (Coregonus albula)
ESOCIDAE
Northern pike (Esox lucius)
CENIRARCHIDAE
Smallmouth bass (Micropterus
dolomieul)
PERCIDAE
European perch (Perca fluviatilis)
CYPRINIDAE
Roach (Rutilus rutilus)
Others (unidentified)
Successful hatching of eyed eggs in limed
hatching boxes in George Lake, Ontario
Enhanced survival of natural populations in
Nelson Lake, Ontario
Successful hatching of eyed eggs in limed
hatching boxes in George Lake, Ontario
Enhanced survival of natural populations in
Adirondack lakes of New York State
Successful reintroduction, spawning, and
enhanced survival in Swedish lakes
Enhanced survival of natural populations in
Lake Nedsjon and Lake Ostra, Skalsjon,
Sweden
Successful reintroduction and spawning in
Swedish lakes
Successful hatching of eyed eggs in limed
hatching boxes in George Lake, Ontario
Successful introduction in Swedish lakes,
and enhanced survival and spawning in
natural populations of anadromous strains in
Swedish rivers
Successful introduction in Lake Howatn,
Norway
Enhanced survival of natural populations in
Hogvadsan River system, Sweden
Enhanced survival of parr in Mersey Fish
Cultural Station, Nova Scotia, after installa-
tion of limestone filters
Enhanced survival of n; tural populations in
Lake Nedsjon, Sweden
Successful reproduction and enhanced sur-
vival of natural populations in Swedish lakes
Successful spawning of reintroduced fish in
Nelson Lake, Ontario
Successful reproduction and enhanced sur-
vival of natural populations in Swedish lakes
Successful reproduction and enhanced sur-
vival of natural populations in Swedish lakes
Successful reproduction and enhanced sur-
vival of natural populations in Sweden
Gunn and Keller, 1981
Kelso and Gunn, 1982
Gunn and Keller, 1981
Blake, 1981; Haines, 1981
Hultberg and Andersson, 1981, 1982
Bernhoff, 1979; Bengtsson et al. 1980
Bengtsson, 1982
Gunn and Keller, 1980
Hultberg and Andersson, 1981
Fraser et al. 1982
Bengtsson et al. 1980
Farmer et al. 1980; Goff et al. 1981
Bernhoff, 1979
Hultberg and Andersson, 1981
Keller et al. 1980; Van and
Dillon, 1981
Bengtsson et al. 1980; Hultberg
and Andersson, 1981
Haines, 1981; Hultberg and
Andersson, 1981; Bengtsson, 1982
Bengtsson, 1982
Source: Fraser and Britt, 1982.
100
-------
Acid Rain
and sometimes endemic to their environment. Such systems
can be significantly altered by adding alkaline material;
therefore, the extensive liming of such systems may not be
warranted unless significant public health or economic
benefits will be attained. A generalized decision path for
evaluating a body of water as a candidate for liming is
presented in Figure 2.
LOW ALKALINITY AND
HIGH ACIDIC INPUT
OR LOWVpH
CONTAINS OR CONTAINED
IMPORTANT FOOD OR
GAME FISH
REPRESENTS IMPORTANT
RECREATIONAL RESOURCE
FEW CONFOUNDING
PHYSICAL AND
CHEMICAL
FACTORS"
LOW COST OF
LIMING RELATIVE
TO OTHER OPTIONS
''CONFOUNDING PHYSICAL AND CHEMICAL FACTORS INCLUDE HIGH FLUSHING RATES
HIGH AOUEOUS METAL CONCENTRATIONS EXTENSIVE ACCUMULATIONS OF SNOW PACK
RAPID Al LEACH HATES IN WATERSHED SOIL PREDOMINANCE Of SOFT SEDIMENTS AND
PRESENCE OF HUMIC SUBSTANCES
Figure 2.—Generalized decision path for evaluating a body of water
as a candidate for liming.
Source: Fraser and Britt, 1982.
Once a surface water body has been selected as a poten-
tial candidate for chemical neutralization, certain design
aspects and problems associated with liming require careful
consideration before initiating the project:
• Liming should be planned to neutralize worst case con-
ditions, such as periodic surges of acid input, because this
is often when most biological damage can occur.
• Lakes with short retention times (less than 1 year), or
running waters with great variations in flow are very difficult,
if not impossible, to lime effectively.
• Toxic aluminum may continue to leach into the lake if
watershed soils remain acidic.
• After liming there will be a short transition period when
precipitation of hydrolized aluminum may be toxic to fish.
• Limed bodies of water should be continually monitored
and relimed when pH falls below 6 because accumulated
metals that have precipitated over a period of time may again
become soluble and create a serious toxic condition.
• Liming in surface waters containing high concentrations
of humic materials may require more lime because humus
can precipitate onto the lime and deactivate it.
• Acidic sediments, especially those high in clay and
organic material, may use up much or all of the potential
buffering capacity of residual liming materials resting on lake
bottoms.
• None of the existing dose estimating techniques is com-
pletely reliable.
• The design of the project must reflect the unique set
of physical, chemical, and biological characteristics of the
water body and watershed.
• Long-term ecological effects are uncertain.
RESEARCH NEEDS
At the present time the data bases for lake and stream lim-
ing activities are too incomplete to resolve all of the uncer-
tainties, alternative hypotheses, and technical problems cur-
rently associated with the chemical neutralization of acidified
surface waters. If liming activities are anticipated to increase
in the United States, then a research program designed to
resolve some of the present uncertainties and problems will
be needed to evaluate the long-term implications of liming
and to enhance the efficiency and effectiveness of any future
liming projects. Some of the most critical technological and
ecological information needs related to the chemical
neutralization of surface waters are:
• More reliable dose-estimating techniques for base
addition.
• Better methods for neutralizing episodic acid surges
such as those that occur during snowmelt.
• Effective methods for neutralizing lotic environments.
• Better understanding of mechanisms involved with the
deactivation of residual liming materials.
• Data on the importance and efficiency of liming water-
sheds relative to the prevention of aluminum leaching.
• Better understanding of the toxicity of aluminum and
other metals, and their pH-dependent solubility.
• Evaluation of the effects of liming on carbon,
phosphorus, and other nutrient cycles
• Data on organic complexing agents that may chelate
toxic metals.
• Determination of optimum restocking strategies.
• Better understanding of the optimum timing and place-
ment of liming materials with respect to critical life stages
of aquatic biota.
Although the chemical neutralization of acidified surface
waters is not advocated by its major practitioners as a solu-
tion to the problem of acid precipitation, it may, if used wisely
and selectively, represent a viable option for the manage-
ment of affected fisheries, from a short-term biological
standpoint.
REFERENCES
Baker, J.P. 1981. Effects on fish of metals associated with acidifi-
cation. Presented at Acid Rain/Fisheries Sympo. Ithaca, NY. U.S.
Fish Wildl. Serv., Fish. Oceans Can, .Aug. 2-5.
Baker, J.P., and C.L. Schofield. 1981. Aluminum toxicity to fish in
acidic waters. Presented at AMS/CMOS Conf. Long-range
Transport of Airborne Pollutants. Albany, N.Y. Am. Meteorolog.
Soc. Can. Meteorolog. Oceanog. Soc. April 27-30.
Beamish, R. 1975. In Proc. 1st Int. Symp. Acid Precipitation and
the Forest Ecosystem. Columbus, Ohio: Ohio State Univ. Atmos.
Sci. Prog. U.S. Dep. Agric. Forest Serv.
. 1976. Acidification of lakes in Canada by acid precipitation
and the resulting effects on fishes. Water Air Soil Pollut 6:501-514.
Beamish. R., W. Lockhart, J. Van Loon, and H. Harvey. 1975. Long-
term acidification of a lake and resulting effects on fishes. Ambio
4:98-102.
Bengtsson, B. 1982. Acid rain mitigation strategies in Sweden. Water
Air Soil Pollut. (submitted).
Bengtsson, B., W. Dickson, and P. Nyberg. 1980. Liming acid lakes
in Sweden. Ambio 9:34-36.
Bernhoff, B. 1979. Acidification and lake liming. Cementa AB, Lime
Dep. Malmo, Sweden.
Blake, L. 1981. Liming acid ponds in New York. N.Y. Fish Game J.
28:208-214.
Boyd, C.E. 1982a. Liming fish ponds. J. Soil Water Conserv
37:86-88.
.. 1982b. Water quality management for pond fish culture.
Elsevier Scientific Publ. Co., Amsterdam, Netherlands.
101
-------
Lake Restoration, Protection and Management
Daye, P.G. 1981. The impact of acid precipitation on the physiology
and toxicology of fish. Pages 29-34 in L. Sochasky.ed. Acid rain
and the Atlantic salmon: Proc. Conf. Acid Rain and the Atlantic
Salmon, Nov. 22-23, 1980. Portland, Maine. Int. Atlantic Salmon
Found. No. 10.
Dickson, W. 1978. Pages 37-41 in G. Hendrey, ed. Limnological
aspects of acid precipitation: Proc. Int. Workshop. G. Hendrey,
ed. Brookhaven Natl. Lab. Rep., BNL-51074.
Driscoll, C.T. 1980. Chemical characterization of some dilute acidified
lakes and streams in the Adirondack region of New York State.
Dissertation. Cornell Univ., Ithaca, N.Y.
Driscoll, C.T., J.R. White, G.C. Schafran, and J.D. Rendall. 1982.
CaCO3 neutralization of acidified surface waters. Am. Soc. Civ.
Eng. J. Environ. Eng. Div. 12:(in press).
Edzwald, J., and J. DePinto. 1978. Recovery of Adirondack acid
lakes with fly ash treatment. Eng. Found. Grant No. RC-A076-4.
New York.
Farmer, G., T. Goff, D. Ashfield, and H. Samant. 1980. Some
effects of the acidification of Atlantic salmon rivers in Nova Scotia.
Can. Tech. Rep. Fish. Aquat. Soc. No. 972.
Fraser, J.E., and D.L. Britt. 1982. Liming of acidified waters: a review
of methods and effects on aquatic ecosystems.
FWS/OBS-80/40.13. U.S. Fish Wildl. Serv.
Fraser, J.E., et al. 1982. Feasibility study to utilize liming as a tech-
nique to mitigate surface water acidification. EPRI EA-2362. Elec-
tric Power Research Institute, Palo Alto, Calif.
Fritz, E.S. 1980. Potential impacts of low pH on fish and fish popula-
tions. FWS/OBS-80/40.2: U.S. Fish Wildl. Serv., Biolog. Serv. Prog.
Natl. Power Plant Team. Ann Arbor, Mich.
Goff, T.R., F.S. Baker, and D.L. MacDonald. 1981. Improved survival
of Atlantic salmon parr by limestone filtration of an acidic hatchery
water supply. Fish. Oceans Can. Halifax, Nova Scotia.
Gorham, E., and W. McFee. 1980. Effects of acid deposition upon
outputs from terrestrial to aquatic ecosystems. Pages 465-480 in
J.C. Hutchinson and M. Havas, eds. Effects of Acid Precipitation
on Terrestrial Ecosystems. Plenum Press, New York.
Grahn,. O., and H. Hultberg. 1975. The neutralization capacity of
12 different lime products used for pH-adjustment of acid water.
Vatten 2:120-132.
Grahn, O., H. Hultberg, and L. Landner. 1974. Oligotrophication —
a self-accelerating process in lakes subjected to excessive sup-
ply of acid substances. Ambio 3:93-94.
Gunn, J.M., and W. Keller. 1980. Enhancement of the survival of
rainbow trout (Salmo gairdneri) egs and fry in an acid lake through
incubation in limestone. Can. J. Fish. Aquat. Soc. 37:1522-1530.
. 1981. Emergence and survival of lake trout (Salvelinus
namaycush) and brook trout (S. fontinalis) from artificial substrates
in an acid lake. Ontario Ministry Nat. Resour. and Ministry En-
viron. Sudbury, Ontario.
Haines, T.A. 1981. Acidic precipitation and its consequences for
aquatic ecosystems: a review. Trans. Am. Fish. Soc. 110:669-706.
Haines, T.A., and C.L. Scholfield. 1980. Responses of fish to acidifica-
tion of streams and lakes in eastern North America. Pages 467-471
in Restoration of Lakes and Inland Waters: Proc. Int. Symp. EPA
440/5-81-010. U.S. Environ. Prot. Agency, Washington, D.C.
Hall, R., G. Likens, S. Fiance, and G. Hendrey. 1980. Experimental
acidification of a stream in the Hubbard Brook Experimental Forest,
N.H. Ecology 61:976-989.
Hasler, A.D., O.M. Brynildson, and W.T. Helm. 1951. Improving
conditions for fish in brown-water bog lakes by alkalization. J. Wildl.
Manage. 15:347-352.
Hultberg, H., and I.B. Andersson. 1981. Liming of acidified lakes
and streams — perspectives on induced physical-chemical and
biological changes. Presented at AMS/CMOS Conf. Long-range
Transport of Airborne Pollutants. Albany, N.Y. Am. Meteorolog.
Soc. Can. Meteorolog. Oceanogr. Soc. April 27-30.
_. 1982. Liming of acidified lakes: induced long-term changes.
Water Air Soil Pollut. 18:311-331.
Hultberg, H., and O. Grahn. 1975. Effects of acid precipitation on
macrophytes in oligotrophic lakes. Proc. 1st Spec. Symp. on At-
mospheric Contribution to the Chemistry of Lake Waters. Sept.
28 - Oct. 1 Int. Assoc. Great Lakes Res.
Jeffries, D., C. Cox, and P. Dillon. 1979. Depression of pH in lakes
and streams in central Ontario during snowmelt. J. Fish. Res.
Board Can. 36:640-646.
Johannessen, M., and R. Wright. 1982. Liming in Norway. Presented
at Int. Liming Workshop. Friday Harbor Lab. Wash. Sept. 19-25.
Johannessen, M., A. Skartveit, and R. Wright. 1980. Streamwater
chemistry before, during and after snowmelt. Pages 224-225 in
D. Drablos and A. Tollan, eds. Ecological Impact of Acid Deposi-
tion: Proc. Int. Conf. March 11-14. Sandefjord, Norway. SNSF Proj.
Oslo-As, Norway.
Johnson, N.M., et al. 1981. Acid rain, dissolved aluminum and
chemical weathering at the Hubbard Brook Experimental Forest,
N.H. Geochem. Cosmochim. Ada 45:1421-1437.
Keller, W., J. Gunn, and N. Conroy. 1980. Acidification impacts on
lakes in the Sudbury, Ontario, Canada area. Pages 228-229 in
D. Drablos and A. Tollan, eds. Ecological impact of acid precipita-
tion: Proc. Int. Conf. March 11-14. Sandefjord, Norway. SNSF Pro-
ject. Oslo-As, Norway.
Kelso, J.R.M., and J. Gunn. 1982. Response of fish communities
to acidic environments. Presented at 185th Natl. Meet. Am. Chem.
Soc. Div. Environ. Chem. Las Vegas, Nev. March 28 - April 2.
ACS Preprint Extended Abstr. 22:431-437.
Lewis, C., and R. Boynton. 1976. Acid neutralization with lime for
environmental control and manufacturing processes. Natl. Lime
Assoc. Bull. No. 216. Washington, D.C.
Muniz, I.P., and H. Leivestad. 1980. Acidification effects on fresh-
water fish. Pages 84-92 in D. Drablos and A. Tollan, eds. Ecological
Impact of Acid Precipitation: Proc. Int. Conf. March 11-14: Sandef-
jord, Norway. SNSF Project. Oslo-As, Norway.
National Fisheries Board of Sweden. 1982. Rad och riktlinjer for
kalkning av sjoar och vattendrag. Rep. No. 1.
Pearson, F.H., and A.J. McDonnell. 1975. Limestone barriers to
neutralize acidic streams. Am. Soc. Civil Eng. J. Environ. Eng.
Div. 101:425-440.
Rafaill, B.L., and W.G. Vogel. 1978. A guide for vegetating surface-
mined lands for wildlife in eastern Kentucky and West Virginia.
FWS/OBS-78-84. U.S. Wildl. Serv. Eastern Energy and Land Use
Team. Harper's Ferry, W. Va.
Ryan, P., and H. Harvey. 1977. Growth of rock bass, Ambloplites
rupestris in relationship to the morphoedaphic index as an indica-
tion of environmental stress. J. Fish. Res. Board Can.
34:2079-2088.
Scheider, W., and P.J. Dillon. 1976. Neutralization and fertilization
of acidified lakes near Sudbury, Ontario. Water Pollut. Res. Can.
11:93-100.
Scheider, W., J. Adamski, and M. Paylor. 1975a. Reclamation of
acidified lakes near Sudbury, Ontario. Ontario Ministry Environ.
Rep. Rexdale, Ontario, Canada.
Scheider, W., B. Cave, and J. Jones. 1975b. Reclamation of acidified
lakes near Sudbury, Ontario by neutralization and fertilization. On-
tario Ministry Environ. Rep. Rexdale, Ontario, Canada.
Schofield, C.L. 1976. Acid precipitation: effects on fish. Ambio
5:228-230.
1980. Effects of acid rain on lakes. Pages 55-69 in C.G.
Gunnerson and B.E. Willard, eds. Proc. Am. Soc. Civil Eng. Tech.
Counc. Res. April 2-6, 1979. Boston, Mass. Am. Soc. Civ. Eng.
New York.
Schofield, C., and J. Trojnar. 1980. Aluminum toxicity to fish in
acidified waters. Pages 341-366 in T. Toribara, M. Miller, and P.
Morrow, eds. Polluted Rain. Plenum Press, New York.
Sverdrup, H.U. 1982a. Lake liming: the calcite utilization efficiency
and long term effect on alkalinity in several Swedish lake liming
projects. Vatten 38(4): (in press).
1982b. Lake liming: the dissolution of a calcite powder
sinking in an acidified lake. Chem. Scripta (in press).
102
-------
Acid Rain
Swedish Ministry of Agriculture Environment Committee. 1982.
Acidification Today and Tomorrow. Risbergs Tryckeri AB, Ud-
devalla, Sweden.
Theiss, R. 1981. An evaluation of the recovery of an Adirondack
acid lake through chemical manipulation. M.S. Thesis. Clarkson
College of Technology, Dep. Environ. Eng. Potsdam, N.Y.
Tollan, A. 1980. Effects of acid precipitation on aquatic and terrestrial
ecosystems. Pages 438-445 in Restoration of Lakes and Inland
Waters: Proc. Int. Symp. EPA 440/5-81-010. U.S. Environ. Prot.
Agency, Washington, D.C.
Waters, T.F 1956. The effects of lime application to acid bog
lakes in northern Michigan. Trans. Am. Fish. Soc. 86:329-344.
Watt, W.D. 1980. Present and potential efects of acid precipitation
on the Atlantic salmon in eastern Canada. Pages 39-45 in L.
Sochasky, ed. Acid Rain and the Atlantic Salmon: Proc. Conf.
Nov. 22-23, Portland, Maine. Int. Atlantic Salmon Found. New
York.
Wilander, A., and T. Ahl. 1972. The effects of lime treatment to a
small lake in Bergslagen, Sweden. Vattan 5:431-445.
Wood, J. 1980. The role of pH and oxidation-reduction potentials
in the mobilization of heavy metals. Pages 223-238 in T. Toribara,
M. Miller, and P. Morrow, eds. Polluted Rain. Plenum Press: New
York.
Yan, N., and P. Dillon. 1982. Experimental neutralization of lakes
near Sudbury, Ont. Chapt. 6 in Studies of Lakes and Watersheds
Near Sudbury, Ontario: Final Limnol. Rep. SES009/82. Ont.
Ministry Environ., Rexdale, Ont., Can.
103
-------
Internal Nutrient Loading
Detection and Control
SUBMERSED MACROPHYTE COMMUNITY STRUCTURE AND INTERNAL
LOADING: RELATIONSHIP TO LAKE ECOSYSTEM PRODUCTIVITY
AND SUCCESSION
STEPHEN R. CARPENTER
Department of Biology
University of Notre Dame
Notre Dame, Indiana
ABSTRACT
The robust macrophytes of eutrophic lakes typically acquire shoot phosphorus from sediments via the
roots. Substantial biomass turnover during the growing season enriches lake water with dissolved phosphorus
and organic matter. This internal loading mechanism is ineffective in oligotrophic lakes where macrophytes
are smaller, have lower biomass turnover rates, and immobilize nutrients by oxidizing sediments. Internal
loading enhances productivity and leads to increased sediment accretion rates, which in turn increase
the surface area colonizable by macrophytes, thereby increasing internal loading. This positive feedback
loop is potentially important in shallow lakes vegetated by macrophytes that concentrate biomass near
the water surface and have rapid biomass turnover rates (e.g. Myriophyllum spicatum). Macrophyte harvesting
can reduce internal loading. Cultivation of native macrophyte species with low internal loading capability,
such as Vallisneria americana, is a novel, but as yet untested, means of controlling internal loading.
INTRODUCTION
Internal loading is the recycling of nutrients from lake
sediments to the overlying water. Most phosphorus loading
models neglect internal loading (e.g. Dillon and Rigler, 1974;
Vollenweider, 1975). Reports of internal loading are numer-
ous, and in several published cases internal loading has con-
founded lake restoration efforts (Bjork et al. 1972; Ahlgren,
1977; Cooke et al. 1977; Prentki et al. 1979; Larsen et al.
1981; Jacoby et al. 1982).
Currently emerging concern about internal loading belies
the fact that limnologists have recognized the importance
of internal loading since about 1900. In 1903 Kofoid noted
that decay of rooted macrophytes favored the growth of
phytoplankton. The prescient study by Pond (1905) used field
experiments to establish that six macrophyte species
developed root hairs, translocated nutrients from roots to
shoots, could not survive unless rooted in sediments, and
released nutrients to the water after senescing. Both Hut-
chinson (1941) and Lindeman (1942) recognized that release
of phosphorus from sediment must occur. At the same time
Mortimer (1941, 1942) established that phosphorus dia-
genesis from sediment accelerated with declining redox
potential at the sediment surface. More recent studies have
quantified internal fluxes of nutrients, permitting comparisons
of various nutrient sources and sinks within the lake
ecosystem.
This paper reviews current literature on the role of sub-
mersed aquatic plants in internal loading, and assesses the
105
-------
Lake Restoration, Protection and Management
relationship of macrophytic internal loading to trophic state,
lake succession, and management.
NUTRIENT SOURCES FOR
SUBMERSED MACROPHYTES
Pond's (1905) conclusion that macrophyte roots are effec-
tive absorptive organs is now supported by a rich and diverse
literature. Several nutrients, including nitrogen and phos-
phorus, are readily taken up by roots (McRoy and
Barsdate, 1970; Bristow and Whitcombe, 1971; DeMarte and
Hartman, 1974; Nichols and Keeney, 1976; Best and Man-
tai, 1978). Many field and laboratory studies, using various
methods and several nutrients, have demonstrated transloca-
tion from roots to shoots (Nichols and Keeney, 1976; Welsh
and Denny, 1979a,b; Denny, 1980; Barko and Smart, 1980,
1981). Typically, 90 to 100 percent of the phosphorus in
macrophyte shoots is supplied through the roots (Carignan
and Kalff, 1980). In contrast, a substantial fraction of root
and shoot potassium is absorbed by the shoots (Barko,
1982a). Macrophytes and their associated epiphytes can
remove calcium, inorganic nitrogen, phosphorus, and
potassium from the surrounding water (Mickle and Wetzel,
1978; Howard-Williams, 1981), However, most phosphorus
apparently taken up by shoots may actually be absorbed by
epiphytes (Howard-Williams and Allanson, 1981).
NUTRIENT RELEASE FROM LIVING
SHOOTS
Wetzel's (1969) initial report of dissolved organic carbon
(DOC) release by living macrophyte shoots has since been
supplemented and corroborated by field observations (Wetzel
and Manny, 1972; Hough and Wetzel, 1975). The water lily
Ntphar luteum releases phosphorus (Twilley et al. 1977).
However, Denny (1980) is skeptical of accounts of phos-
phorus release by submersed plants in laboratory ex-
periments (e.g. McRoy and Barsdate, 1970). In studies of
whole plants grown in the laboratory (Barko and Smart, 1980)
and in the field (Carignan and Kalff, 1982), minimal phos-
phorus was released from living shoots. Apparently, most
phosphorus release occurs by decay after shoots die.
LAKE ENRICHMENT BY MACROPHYTE
DECAY
Macrophyte shoot decay rates depend on several en-
vironmental factors, notably temperature, oxygen concentra-
tion, nutrient concentrations, and chemical composition of
the decaying tissue (Godshalk and Wetzel, 1978b; Carpenter
and Adams, 1979; Carpenter, 1980a). The early stages of
decay that occur when the shoot Is in the water column are
most relevant to internal loading. However, collapsed macro-
phytes decaying at the sediment surface may lower the redox
potential and thereby enhance diffusion of nutrients from the
sediment.
Shoot decay rates tend to decline as decay progresses
(Godshatk and Wetzel, 1978b), so that recently dead shoots
decompose most rapidly. This rapid decomposition is ac-
companied by substantial leaching of DOC (Otsuki and
Wetzel, 1974; Godshalk and Wetzel, 1978a) and dissolved
phosphorus (Carpenter, 1980a). In contrast, nitrogen tends
to be retained or accumulated by paniculate macrophyte
detritus (Nichols and Keeney, 1973; Harrison and Mann,
1975; Godshaik and Wetzel, 1978b; Carpenter and Adams,
1979; Carpenter, 1980a; Boston and Perkins, 1982).
Leaching and processing of DOC depends on oxygen con-
centration and water temperature (Godshalk and Wetzel,
1978a). Phosphorus leaching rates are independent of
temperature but strongly dependent on shoot phosphorus
content at death (Carpenter, 1980a).
Dissolved materials in macrophyte leachate are readily
available to planktonlc bacteria and algae. Turnover rates
of leached DOC are rapid (Godshalk and Wetzel, 1978a).
Over 75 percent of leached phosphorus is in soluble reac-
tive form with molecular weight approximately identical to
that of phosphate (Carpenter and Adams, 1978; Carpenter,
1980a). Macrophyte leachate stimulates oxygen use by
sestonic bacteria (Carpenter et al. 1979) and growth of algae
(Landers, 1982).
Quantitative estimates of lakewide internal loading by
macrophytes are rare. Littoral zone nutrient cycles for Lake
Wingra, Wis., a shallow (mean depth = 2.4 m) unstratified
lake densely vegetated by Myriophyllum sptcatum, have been
summarized by Adams and Prentki (1982). Internal loading
of DOC and dissolved total phosphorus (DTP) exceeded
loading from the watershed (Table 1). Allochthonous loading
of total organic carbon and total phosphorus are 110,000
kg/y and 1,570 kg/y, respectively, the same order of
magnitude as internal loading (Adams and Prentki, 1982).
Macrophyte decay accounts for about half of the internal
loading.
Release of dissolved phosphorus by living plants in Lake
Wingra is negligible (Adams and Prentki, 1982), but living
plants may contribute to the internal DOC loading. The
balance of the internal loading presumably comes from slow
mineralization of paniculate matter in the water or directly
from the sediment by diffusion and resuspension. Adams
and Prentki (1982) consider the possibility that the littoral
zone is a net sink for paniculate materials and a net source
for dissolved materials.
Maximum and minimum estimates of internal loading can
be calculated from macrophyte net primary production and
maximum seasonal standing crop, respectively. More re-
fined estimates of internal loading must account for (1)
seasonal variations in macrophyte mortality and decay rates
(Carpenter, I980a,b), (2) the length of time senescing shoots
are suspended in the water column before sinking to the
sediments (Carpenter, 1980a), and (3) the hydrodynamics
of littoral-pelagial water exchange (Weller, 1978; Prentki et
al. 1979; Carpenter and Greenlee, 1981),
In general, internal loading by macrophyte decay should
be greater in eutrophic lakes than oligotrophic lakes, because
macrophyte biomass and biomass turnover rates are greater
in eutrophic lakes. Correlations of macrophyte community
structure with limnologteal variables show that small rosulate
species are associated with clear, acidic, ion-poor waters
while large robust species are associated with turbid, alkaline,
Table 1. —Internal and external loadings of dissolved organic carbon (DOC) and dissolved total phosphorus (DTP) to Lake
Wingra, Wis., from data of Prentki, et al. (1977, 1979) and Carpenter (1980a). All figures are kg per annum.
Material
DOC.
DTP
Internal loading by
macrophyte decay
43,000
320
Total
Internal
loading
100,000
760
Loading from
watershed
30,000
680
106
-------
Internal Nutrient Loading Detection and Control
Table 2. — Annual biomass turnover (annual net production/maximum seasonal blomass) for selected submersed vascular
macrophyte species In temperate lakes.
Lake type Species
Oligotrophic Isoetes lacustris
Isoetes echinospora
Uttorella uniflora
Lobelia dortmanna
Marl Potamogeton praelongus,
Turnover
0.5
0.5
1.5
0.69
1.5
Source
Eriksson et al. 1974
(in Moeller, 1978)
Eriksson et al. 1974
(in Moeller, 1978)
Sand-Jensen and
Sondergaard, 1978
Moeller, 1978
Rich et al. 1971
Saline
Mesotrophic
Eutrophic
P. iHinoensis
Scirpus subterminalis
Potamogeton pectinatus
Najas minor
Potamogeton lucens
Potamogeton pectinatus
Potamogeton perfoliatus
Myriophyllum spicatum
Potamogeton pectinatus
Vallisneria americana
3.0-5.0
1.2
1.57
1.55
1.00
1.60
1.84
1.7
2.0-2.6
Rich et al. 1971
Howard-Williams, 1978
Gladyshev and Kogan, 1977
Gladyshev and Kogan, 1977
Gladyshev and Kogan, 1977
Gladyshev and Kogan, 1977
Adams and McCracken, 1974
(night respiration estimate
from Carpenter, 1979)
Carpenter, 1980b
Carpenter, 1980b
ion-rich waters (Swindate and Curtis, 1957; Pip, 1979). Small
plants of Oligotrophic lakes have annual biomass turnover
rates from 0.5 to 1.5 (Table 2). Turnover rates for the much
larger plants of mesotrophic and eutrophic lakes are general-
ly higher, and range from 1.0 to 2.6.
Differences among species in the timing of shoot turnover
should also be considered. For example, Myriophyllum spica-
tum and Potamogeton pectinatus stands have substantial
biomass turnover and internal loading during June, July, and
August (Carpenter, 1980a,b) when substantial algal growth
may result from added nutrients. In contrast, mortality and
internal loading in Vallisneria americana stands occur primari-
ly in September and October (Carpenter, 1980a,b) when algal
production rates are declining. Nutrient inputs late in the
growing season may be lost to sediments or exported before
the next growing season, without contributing appreciably
to algal production.
MACROPHYTE EFFECTS ON NUTRIENT
DIFFUSION FROM SEDIMENTS
Wium-Andersen (1972) showed a correlation between
presence of rooted vascular macrophytes and oxidized sur-
face sediments. Iron and manganese are precipitated at the
base of the oxidized layer, which penetrates as deeply as
25 cm into the sediment (Tessenow and Baynes, 1975,
1978). At least some of the oxidizing activity in upper sedi-
ment layers is caused by diffusion of oxygen from
macrophyte roots (Sand-Jensen and Prahl, 1982; Sand-
Jensen et al. 1982; Elser et al. submitted). Oxygen release
by roots is more rapid in light than in darkness (Sand-Jensen
et al. 1982; Elser et al. submitted).
Effects of oxygen release from macrophyte roots on
sedimentary phosphorus profiles have not been investigated.
In oxidized sediments, phosphate is complexed with iron-
containing colloids (Mortimer, 1941,1942). Therefore, sedi-
ment oxidation by macrophytes could immobilize phosphate
and create a barrier to diffusion or mixing of phosphate
across the sediment-water interface (Fig. 1).
This hypothetical inhibition by macrophytes of internal
phosphorus loading is most likely to occur in Oligotrophic
lakes. In a comparison of eight species, the most rapid ox-
ygen release rates were found in Lobelioids and Isoetids that
grow on relatively oxidized sediments in Oligotrophic lakes
(Sand-Jensen et al. 1982). In eutrophic lake sediments, ox-
ygen release by Myriophyllum verticillatum did not affect sedi-
ment organic content and decreased chemical oxygen de-
mand of pore waters only slightly below controls (Elser et
al. submitted). These results differ from those of similar ex-
periments in which Isoetes lacustris oxidized Oligotrophic lake
sediments (Tessenow and Baynes, 1975, 1978). Sedimen-
tary phosphorus profiles from Myriophyllum spicatum beds
in an eutrophic lake show high inorganic phosphorus con-
centrations near the sediment surface (Prentki, 1979). In sum,
littoral sediment phosphorus profiles result from a balance
between oxidizing capabilities of macrophyte roots and the
reducing capacity of the sediment. This balance appears to
favor immobilization of phosphate in Oligotrophic systems and
mobilization of phosphate in eutrophic systems.
INTERNAL LOADING AND AUTOGENIC
LAKE SUCCESSION
Macrophytes in eutrophic lakes transfer phosphorus from
sediments to lake water in three steps: uptake by roots,
INCREASING Eh, P, S
a
UJ
a
/
I
Rgure 1.—Hypothetical profiles of Eh, particulate and precipitated
phosphorus (P), and soluble phosphorus (S) in sediments with (left)
and without (right) macrophytes. Sediment-water interface is at 0.
The high soluble phosphorus concentrations in surface sediments
without macrophytes (right) promote high phosphorus transfer rates
by diffusion and mixing. Transfer is slower from oxidized surface
sediments associated with macrophytes (left) because soluble
phosphorus concentrations are lower near the surface.
107
-------
Lake Restoration, Protection and Management
translocation from roots to shoots, and leaching from senes-
cent shoots. These leachates stimulate planktonic produc-
tion. Loucks and Weiler (1979) estimated that phytoplankton
production in Lake Wingra could be decreased by 18 per-
cent if macrophytic internal loading of phosphorus could be
decreased by 50 percent. DOC released by living or decay-
ing macrophytes may also enhance production by stimulating
mineral regeneration and production by sestonic hetero-
trophic microbes.
Macrophytes close a positive feedback loop that promotes
sediment accretion in eutrophic lakes (Rg. 2). Internal loading
increases phytoplankton and macrophytes increase sedimen-
tation. As sediments accumulate, the lake becomes shallow-
er, expanding the area colonizable by macrophytes. Enlarge-
ment of macrophyte beds leads to increased internal loading,
thereby continuing the feedback cycle. This positive feed-
back loop has been quantified in Lake Wingra (Carpenter,
1981 a). Wetzel (1979) discusses the role of emergent
macrophytes in a similar cycle.
Two processes can counter the effects of the positive feed-
back loop (Fig. 2):
1. Maximum depths colonized by submersed
macrophytes are often determined by light penetration
(Spence, 1982), so increased phytoplankton concentrations
could reduce colonizable area by increasing turbidity.
Macrophyte species that do not concentrate photosynthetic
biomass near the water surface could be severely inhibited
by high phytoplankton concentrations (e.g. Jupp and Spence,
1977). In contrast, Myriophyllum spicatum, the dominant
macrophyte in Lake Wingra, forms a dense canopy near the
water surface and is relatively insensitive to variations in tur-
bidity (Titus and Adams, 1979).
2. As macrophyte beds expand, the open water area con-
tracts, thereby decreasing lakewide phytoplankton produc-
tion. The effects of this decrease on sedimentation are par-
tially offset by increased sedimentation of macrophyte re-
mains. Carpenter (1981 a) incorporated these factors in an
analysis of sediment accretion rates in Lake Wingra, and
found that expansion of macrophyte beds had a net positive
effect on sediment accretion.
Unlike loading from external sources, internal loading by
macrophytes is potentially self-accelerating. In Lake Wingra,
the increase in sediment accretion rate produced by a 1 per-
cent increase in internal loading by macrophytes is about
60 times that produced by a 1 percent increase in external
loading (Fig. 3). The sediment accretion rate is a reasonable
index of lake succession rate (Lindeman, 1942) that in Lake
Wingra is strongly dependent on internal loading by
macrophytes.
HOW COMMON IS AUTOGENIC LAKE
SUCCESSION?
The Lake Wingra data clearly illustrate the contribution of
macrophytes to autogenic lake succession. However, com-
SUBMERSED
VEGETATION
WATERSHED
I
PHYTOPLANKTON
X
COLONIZABLE
AREA
/
SEDIMENT
VEG
SED
EXT
Figure 2.—Positive feedback loop that drives accretion of sediment
in eutrophic lakes. See text for discussion.
Figure 3.—Additional accretion of colonizable sediment (m2/y)
resulting from 1 percent increases in annual phosphorus loading
from submersed vegetation (VEG), sediments by diffusion and
resuspension (SED), and the watershed (EXT). Data and calcula-
tion methods appear in Carpenter (19813).
parable data sets are available for very few lakes, so it is
difficult to determine whether this mechanism is common
or rare.
Paieolimnological data show that the sequence from
phytoplankton to submersed macrophytes to floating-leaves
macrophytes to emergent macrophytes or bog vegetation
is relatively common (Table 3). Walker (1972) classified shifts
in aquatic vegetation inferred from sediment cores at 70 sites
in Great Britain. Of 52 transitions in Walker's data, only nine
reversals occurred; these were attributed to increases in
water levels. Walker noted "progressive reduction of free
water depth in the aquatic stages" in most cases of hydrarch
succession. Walker's data reveal sequences but do not com-
pare the importance of autogenic and allogenic factors.
Where autogenic factors are important, sedimentation rates
should increase with time. Published data are equivocal: In
some lakes sedimentation rates have increased (Manny et
al. 1978; Whitehead and Crisman, 1978) but in other lakes
they have decreased (Hutchinson and Cowgill, 1970).
Internal loading is likely to be an important force in suc-
cession of shallow lakes with relatively large areas coloniza-
ble by macrophytes and reduced, nutrient-rich sediments of
low to moderate organic content (ca. 10 to 15 percent). Highly
organic sediments (ca. 20 to 30 percent or more) inhibit
growth of several submersed macrophyte species (Barko,
1982b; Barko and Smart, in press). Macrophyte species such
as Myriophyllum spicatum and Potamogeton pectinatus, with
substantial shoot biomass turnover during the phytoplankton
growing season, provide greater stimulus to algal produc-
tion than species such as Vallisneria americana that senesce
in late summer or early autumn (Carpenter, 1980a). Species
that concentrate photosynthetic biomass near the water sur-
face, such as Myriophyllum spicatum, maintain relatively high
growth rates despite shading by high algal populations (Titus
and Adams, 1979) that may result from internal loading. Tran-
108
-------
Internal Nutrient Loading Detection and Control
Table 3. — Frequencies of vegetation transitions inferred from pollen diagrams and peat stratigraphy in sediment cores from
70 British hydroseres (Walker, 1972). Phytoplankton includes Walker's categories 1 and 2; emergent macrophytes or bogs
includes his categories 5 through 12.
Subsequent vegetation
Antecedent
vegetation
Phytoplankton
Submersed
macrophytes
Floating
leaved
macrophytes
Bogs or
emergent
macrophytes
Floating
Submersed leaved
Phytoplankton macrophytes macrophytes
— 3 4
1 - 4
1 1 —
1 0 5
Bogs or
emergent
macrophytes
3
8
21
—
sitions from submersed macrophytes to phytoplankton are
uncommon in natural hydroseres (Table 3) but have occur-
red in culturally eutrophic lakes (Phillips et al. 1978).
MANAGEMENT IMPLICATIONS
Robust macrophyte species with high annual biomass turn-
over, typical of eutrophic lakes, are potential sources of
nutrients. Internal loading by the plants should be balanced
against a possible role as sinks for nutrients in influent waters
(Howard-Williams, 1981). However, in some cases macro-
phyte beds are not effective barriers to incoming nutrients
(Prentki et al. 1979).
Some management techniques, such as herbicides, do
not control internal loading by macrophytes (Carpenter and
Adams, 1978; Carpenter and Greenlee, 1981). In contrast,
macrophyte harvesting removes nutrients (Carpenter and
Adams, 1977, 1978) and reduces internal loading. Possible
benefits of macrophyte control by any means must be
balanced against the risk of favoring nuisance species over
native species (Nicholson, 1981) and prolonging infestations
of nuisance species that would otherwise subside naturally
(Carpenter, 1980c, 1981b). Replacement of nuisance species
such as Myriophyllum spicatum with native species such as
Vallisneria americana that have low summer mortality may
be an innovative and effective means of limiting internal
loading during most of the phytoplankton growing season.
ACKNOWLEDGEMENTS: I thank M. L. Jaynes and R. P.
Mclntosh for comments on the manuscript.
REFERENCES
Adams, M. S., and M. D. McCracken. 1974. Seasonal production
of the Myriophyllum component of the littoral of Lake Wingra,
Wis. J. Ecol. 62: 457-467.
Adams, M. S., and R. T. Prentki. 1982. Biology, metabolism
and functions of littoral submersed weedbeds of Lake Wingra,
Wis. USA: A summary and review. Arch. Hydrobiol. (Suppl.) 62:
333-409.
Ahlgren, I. 1977. Role of sediments in the process of recovery of
an eutrophicated lake. In H. L. Golterman, ed. Interactions Be-
tween Sediments and Freshwater. Junk, The Hague.
Barko, J. W. 1982a. Influence of potassium source (sediment vs.
open water) and sediment composition on the growth and nutri-
tion of a submersed freshwater macrophyte (Hydrilla verticillata
(L. f.) Royle). Aquat. Bot. 12: 157-172.
1982b. The growth of Myriophyllum spicatum L. in relation
to selected characteristics of sediment and solution. Aquat. Bot.
14: in press.
Barko, J. W., and R. M. Smart. 1980. Mobilization of sediment
phosphorus by submersed freshwater macrophytes. Freshwat.
Biol. 10: 229-238.
. 1981. Sediment-based nutrition of submersed macro-
phytes. Aquat. Bot. 10: 339-352.
. 1983. Effects of organic matter additions to sediment on
the growth of aquatic plants. J. Ecol. 71: in press.
Best, M. D., and K. E. Mantai. 1978. Growth of Myriophyllum: sedi-
ment or lake water as the source of nitrogen and phosphorus.
Ecology 59: 1075-1080.
Bjork, S. et al. 1972. Ecosystem studies in connection with the
restoration of lakes. Verh. Int. Verein. Limnol. 18: 379-387.
Boston, H. L., and M. A. Perkins. 1982. Water column impacts
of macrophyte decomposition beneath fiberglass screens. Aquat.
Bot. 14: 15-28.
Bristow, J. M., and M. Whitcombe. 1971. The role of roots in the
nutrition of aquatic vascular plants. Am. J. Bot. 58: 8-13.
Carignan, R., and J. Kalff. 1980. Phosphorus sources for aquatic
weeds: water or sediments? Science 207: 987-989.
. 1982. Phosphorus release by submerged macrophytes:
significance to epiphyton and phytoplankton. Limnol. Oceanogr.
27: 419-427.
Carpenter, S. R. 1979. Submersed aquatic vegetation and the
process of eutrophication. Ph.D. thesis, Univ. Wisconsin-
Madison.
. 1980a. Enrichment of Lake Wingra, Wis. by submersed
macrophyte decay. Ecology 61: 1145-1155.
_. 1980b. Estimating net shoot production by a hierarchical
cohort method of herbaceous plants subject to high mortality.
Am. Midi. Nat. 104: 163-175.
. 1980c. The decline of Myriophyllum spicatum in a eutrophic
Wisconsin lake. Can. J. Bot. 58: 527-535.
1981 a. Submersed vegetation: an internal factor in lake
ecosystem succession. Am. Nat. 118: 372-383.
1981 b. Effects of control measures on pest populations
subject to regulation by parasites and pathogens. J. Theor. Biol
92: 181-184.
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: 239-255.
109
-------
Lake Restoration, Protection and Management
1978. Macrophyte control by harvesting and herbicides:
implications for phosphorus cycling in Lake Wingra, Wis. J.
Aquat. Plant Manage. 16: 20-23.
1979. Effects of nutrients and temperature on decomposi-
tion of Myriophyllum spicatum L. in a hardwater eutrophic lake.
Limnol. Oceanogr. 24: 520-528.
Carpenter, S. R., and J. K. Greenlee. 1981. Lake deoxygenation
after herbicide use: a simulation model analysis. Aquat. Bot.
11: 173-186.
Carpenter, S. R., A. Gurevitch, and M. S. Adams. 1979. Factors
causing elevated biological oxygen demand in the littoral zone
of Lake Wingra, Wis. Hydrobiologia 67: 3-9.
Cooke, G. D., M. R. McComas, D. W. Waller, and R. H. Kennedy.
1977. The occurrence of internal phosphorus loading in two
small, eutrophic, glacial lakes in northeastern Ohio.
Hydrobiologia 56: 129-135.
DeMarte, J. A., and R. T. Hartman. 1974. Studies on absorption
of 32P, 59Fe, and 45Ca by water-milfoil (Myriophyllum ex-
albescens Fernald). Ecology 55: 188-194.
Denny, P. 1980. Solute movement in submerged angiosperms.
Biol. Rev. 55: 65-92.
Dillon, P. J., and F. H. Rigler. 1974. A test of a simple nutrient
budget model predicting the phosphorus concentration in lake
water. J. Fish. Res. Board Can. 31: 1771-1778.
Elser, J. J., K. M. Olson, and S. R. Carpenter. Effects of Myrio-
phyllum verticillatum roots on sediment redox conditions.
Submitted.
Gladyshev, A. I., and S. I. Kogan. 1977. The dynamics of macro-
phyte and phytoplankton biomass in a floodplain lake of the
middle Amu-Dar'ya. Hydrobiol. J. 13: 74-81.
Godshalk, G. L., and R. G. Wetzel. 1978a. Decomposition of aqua-
tic angiosperms. I. Dissolved components. Aquat. Bot. 5:
281-300.
_. 1978b. Decomposition of aquatic angiosperms. II. Par-
ticulate components. Aquat. Bot. 5: 301-327.
Harrison, P. G., and K. H. Mann. 1975. Chemical changes dur-
ing the seasonal cycle of growth and decay in eelgrass (Zostera
marina) on the Atlantic coast of Canada. J. Fish. Res. Board
Can. 32: 615-621.
Hough, R. A., and R. G. Wetzel. 1975. The release of dissolved
organic carbon from submersed aquatic macrophytes: diel,
seasonal, and community relationships. Verh. Int. Verein. Lim-
nol. 19: 939-948.
Howard-Williams, C. 1978. Growth and production of aquatic
macrophytes in a south temperate saline lake. Verh. Int. Verein.
Limnol. 20:1153-1158.
1981. Studies on the ability of a Potamogeton pectinatus
community to remove dissolved nitrogen and phosphorus com-
pounds from lake water. J. Appl. Ecol. 18: 619-637.
Howard-Williams, C., and B. R. Allanson. 1981. Phosphorus cy-
cling in a dense Potamogeton pectinatus L. bed. Oecologia 49:
56-66.
Hutchinson, G. E. 1941. Limnological studies in Connecticut. IV.
Mechanism of intermediary metabolism in stratified lakes. Ecol.
Monogr. 11: 21-60.
Hutchinson, G. E., and U. M. Cowgill. 1970. The history of the
lake: a synthesis. Trans. Am. Phil. Soc. 60: 163-170.
Jacoby, J. M., D. D. Lynch, E. B. Welch, and M. A. Perkins. 1982.
Internal phosphorus loading in a shallow eutrophic lake. Water
Res. 16: 911-919.
Jupp, B. P., and D. H. N. Spence. 1977. Limitations on macro-
phytes in a eutrophic lake, Loch Leven. I effects of phyto-
plankton. J. Ecol. 65: 175-186.
Kofoid, C. A. 1903. The plankton of the Illinois River, 1894-1899.
I. Quantitative investigations and general results. Bull. III. Lab.
Nat. Hist. 6: 95-629.
Landers, D. H. 1982. Effects of naturally senescing aquatic macro-
phytes on nutrient chemistry and chlorophyll a of surrounding
waters. Limnol. Oceanogr. 27: 428-439.
Larsen, D. P., D. W. Schults, and K. W. Malueg. 1981. Summer
internal phosphorus supplies in Shagawa Lake, Minn. Limnol.
Oceanogr. 26: 740-753.
Lindeman, R. L. 1942. The trophic-dynamic aspect of ecology. Ecol-
ogy 23: 399-418.
Loucks, O. L., and P. R. Weiler. 1979. The effects of harvest removal
of phosphorus on remineralized P sources in a shallow lake. In
J. E. Breck, R. T. Prentki, and O. L. Loucks, eds. Aquatic Plants,
Lake Management, and Ecosystem Consequences of Lake
Harvesting. Inst. Environ. Stud., Madison, Wis.
Manny, B. A., R. G. Wetzel, and R. E. Bailey. 1978. Paleolimnological
sedimentation of organic carbon, nitrogen, phosphorus, fossil
pigments, pollen, and diatoms in a hypereutrophic hardwater lake:
a case history of eutrophication. Pol. Arch. Hydrobiol. 25: 243-267.
McRoy, C. P., and R. J. Barsdate. 1970. Phosphate adsorption in
eelgrass. Limnol. Oceanogr. 15: 6-13.
Mickle, A. M., and R. G. Wetzel. 1978. Effectiveness of submersed
angiosperm-epiphyte complexes on exchange of nutrients and
organic carbon in littoral systems. I. Inorganic nutrients. Aquat.
Bot. 4: 303-316.
Moeller, R. E. 1978. Seasonal changes in biomass, tissue chemistry,
and net production of the evergreen hydrophyte Lobelia dortman-
na. Can. J. Bot. 56:1425-1433.
Mortimer, C. H. 1941. The exchange of dissolved substances be-
tween mud and water in lakes. I, II. J. Ecol. 29: 280-329.
_. 1942. The exchange of dissolved substances between mud
and water in lakes. Ill, IV.. J. Ecol. 30: 147-201.
Nichols, D. S., and D. R. Keeney. 1973. Nitrogen and phosphorus
release from decaying water milfoil. Hydrobiologia 42: 509-525.
1976. Nitrogen nutrition of Myriophyllum spicatum: uptake
and translocation of 15N by shoots and roots. Freshwat. Biol. 6:
145-154.
Nicholson, S. A. 1981. Changes in submersed macrophytes in
Chautauqua Lake, 1937-1975. Freshwat. Biol. 11: 523-530.
Otsuki, A., and R. G. Wetzel. 1974. Release of dissolved organic
matter by autolysis of a submersed macrophyte, Scirpus subter-
minalis. Limnol. Oceanogr. 19: 842-845.
Phillips, G. L., D. Eminson, and B. Moss. 1978. A mechanism to
account for macrophyte decline in progressively eutrophicated
fresh waters. Aquat. Bot. 4: 103-126.
Pip, E. 1979. Survey of the ecology of submerged aquatic macro-
phytes in central Canada. Aquat. Bot. 7: 339-357.
Pond, R. H. 1905. The biological relation of aquatic plants to the
substratum. Rep. U. S. Fish. Comm. 29: 483-526.
Prentki, R. T. 1979. Depletion of phosphorus from sediment colonized
by Myriophyllum spicatum L. In J. E. Breck, R. T. Prentki, and
O. L. Loucks, eds. Aquatic Plants, Lake Management, and
Ecosystem Consequences of Lake Harvesting. Inst. Environ. Stud.,
Madison, Wis.
Prentki, R. T., et al. 1977. Summary tables of Lake Wingra basin
data. Inst. Environ. Stud. Rep. 85, Univ. Wisconsin-Madison.
1979. The role of submersed weedbeds in internal loading
and interception of allochthonous materials in Lake Wingra, Wis.,
U. S. A. Arch. Hydrobiol. (Suppl.) 57: 221-250.
Rich, P. H., R. G. Wetzel, and N. Van Thuy. 1971. Distribution, pro-
duction, and role of aquatic macrophytes in a southern Michigan
marl lake. Freshwat. Biol. 1: 3-21.
Sand-Jensen, K., and C. Prahl. 1982. Oxygen exchange with the
lacunae and across leaves and roots of the submerged vascular
macrophyte Lobelia dortmanna L. New Phytol. 91: 103-120.
Sand-Jensen, K., and M. Sondergaard. 1978. Growth and production
of isoetids in oligotrophic Lake Kalgaard, Denmark. Verh. Int.
Verein. Limol. 20: 659-666.
Sand-Jensen, K., C. Prahl, and H. Stokholm. 1982. Oxygen release
from roots of submerged aquatic macrophytes. Oikos 38: 349-354.
Spence, D. H. N. 1982. The zonation of plants in freshwater lakes.
Adv. Ecol. Res. 12: 37-126.
110
-------
Swindale, D. N., and J. T. Curtis. 1957. The phytosociology of the
larger submerged plants in Wisconsin lakes. Ecology 38: 397-407.
Tessenow, U., and Y. Baynes. 1975. Redox-dependent accumula-
tion of Fe and Mn in a littoral sediment supporting Isoetes lacustris
L. Naturwissenschaften 62: 342-343.
1978. Experimental effects of Isoetes lacustris L. on the dis-
tribution of Eh, pH, Fe, and Mn in lake sediment. Verh. Int. Verein.
Limnol. 20: 2358-2362.
Titus, J. E., and M. S. Adams. 1979. Coexistence and the compara-
tive light relations of the submersed macrophytes Myriophyilum
spicatum L. and Vallisneria americana Michx. Oecologia 40:
273-286.
Twilley, R. R., M. M. Brinson, and G. J. Davis. 1977. Phosphorus
absorption, translocation, and secretion in Nuphar luteum. Lim-
nol. Oceanogr. 22: 1022-1032.
Vollenweider, R. A. 1975. Input-output models with special reference
to the phosphorus loading concept in limnology. Schweiz. Zeitschr.
Hydrol. 37: 53-84.
Walker, D. 1972. Direction and rate in some British post-glacial hydro-
seres. In D. Walker and R. G. West, eds. Studies in the Vegeta-
tional History of the British Isles. Cambridge Univ. Press, Cam-
bridge, U. K.
Internal Nutrient Loading Detection and Control
Weiler, P. R. 1978. Littoral-pelagic water exchange in Lake Wingra,
Wis. as determined by a circulation model. Inst. Environ. Stud.
Rep. 100, Univ. Wisconsin-Madison.
Welsh, R. P. H., and P. Denny. 1979a. The translocation of ^P in
two aquatic angiosperm species. New Phytol. 82: 645-656.
1979b. Translocation of lead and copper in two submerged
__ .
aquatic angiosperm species. J. Exp. Bot. 30: 339-345.
Wetzel, R. G. 1969. Factors influencing photosynthesis and excretion
of dissolved organic matter by aquatic macrophytes in hard-water
lakes. Verh. Int. Verein. Limnol. 17: 72-85.
1979. The rote of the littoral zone and detritus in lake metabo-
lism. Arch. Hydrobiol. Beih. 13: 145-161.
Wetzel, R. G., and B. A. Manny. 1972. Secretion of dissolved organic
carbon and nitrogen by aquatic macrophytes. Verh. Int. Verein.
Limnol. 18: 162-170.
Whitehead, D. R., and T. L. Crisman. 1978. Paledimnological studies
of small New England (U.S.A.) ponds. I. Late-glacial and post-
glacial trophic oscillations. Pol. Arch. Hydrobiol. 25:471-481.
Wium-Andersen, S., and J. M. Andersen. 1972. The influence of
vegetation on the redox profile of the sediment of Grane Langso,
a Danish Lobelia lake. Limnol. Oceanogr. 17: 948-952.
111
-------
CONTROL OF INTERNAL PHOSPHORUS LOADING IN A SHALLOW
LAKE BY DRAWDOWN AND ALUM
J. M. JACOBY
E. B. WELCH
J. P. MICHAUD
Department of Civil Engineering
University of Washington
Seattle, Washington
ABSTRACT
Internal loading of phosphorus in shallow Long Lake, Wash. (Kitsap County) has been identified as
a major factor causing eutrophy during summer. The source of this internal phosphorus (P) loading
apparently comes from sediments in the deeper parts of the lake. A dense submersed macrophyte
crop (dry weight of 220 g rrr2), composed primarily of Elodea densa, may represent an annual source
of P through breakdown and transport of P to deep-lake sediments. Chlorophyll a often exceeded
60 ^gl~1 during three summers prior to restoration. Following lake drawdown (which exposed 30 to
40 percent of the area) sediment consolidation was minimal (0.1 m); the macrophyte crop was re-
duced by 84 percent in 1980, and internal loading of phosphorus was curtailed resulting in a mean
summer total phosphorus concentration of 36 ^g l~1 The rate of macrophyte recolonization during
the post drawdown summer was rapid and biomass had nearly recovered to pre-drawdown levels the
following summer (1981). Alum was applied during autumn of 1980 to inactivate midlake sediment.
The treatment was successful in maintaining a curtailment of internal loading during 1981 and 1982.
Mean summer total P decreased further to 29 ^igl"1 both years and water transparency increased
while the phytoplankton shifted away from blue-greens. The beneficial effect of alum has persisted
for 2 years in spite of full macrophyte recovery and an assumed increased input of particulate P from
senescent macrophytes to surficial sediments, which verifies sediment as the immediate source of
P. Macrophytes may represent the original source of P; this may eventually be demonstrated as enrich-
ment of surficial sediments continues and the beneficial effects of alum decrease.
INTRODUCTION
Identification of the sources and quantities of internal nutrient
loading in eutrophic lakes is necessary to select the most
effective restoration technique. Phosphorus inactivation with
alum has often been used in lakes where P release from
the sediments was believed to be the major internal source
(Cook and Kennedy, 1981; Knauer and Garrison, 1981;
Soltero et at. 1981). The effectiveness of this method
depends upon application of an alum dose beyond that
necessary to remove P from the water column, thereby form-
ing a floe layer on the sediment surface, which inactivates
P (Kennedy and Cooke, 1982). Successful P inactivation
treatments have previously been limited to stratified lakes;
the only other shallow lake treatment was unsuccessful
(Knauer and Garrison, 1981).
Internal loading of P in Long Lake, Wash., has been
identified as a major eutrophying factor in this shallow
(unstratified) lake. The sediments in the deeper parts of the
lake have been shown to be the immediate source of this
internal P loading (Jacoby et al. 1982). In addition, contribu-
tion of P by the dense macrophyte crop, while difficult to
assess, is believed to be substantial. Macrophytes may repre-
sent the original annual P source through senescence,
decomposition, and transport of particulate P to deep layer
sediments and as a direct source of P to lake water in the
early spring that is subsequently taken up by algal blooms.
Contribution of P internally through macrophyte senescence
has been identified in other lakes (Kistritz, 1978; Prentki,
1979; Landers, 1982).
Prior to restoration, extensive blue-green algal blooms
(chlorophyll a often greater than 60 mg rrr3), low
transparency (about 1 m), and total P (TP) concentrations
approaching or exceeding 100 mg m"3) were
characteristically observed in Long Lake during three sum-
mers (Perkins et al. 1979). The dense macrophyte crop (areal
weighted mean dry weight of about 220 g m"2) comprised
mainly of Elodea densa also limited recreational use. It was
particularly abundant in the southern area of the lake
(Hufschmidt, 1978).
As a result of the impaired recreational use, the lake was
drawn down nearly 2 m during the summer of 1979 (June-
October) exposing 30 to 40 percent of the lake area. While
sediment consolidation was minimal (0.1 m) the macrophyte
crop was reduced by 84 percent in 1980.
In September 1980, alum was applied to Long Lake as
aluminum sulfate (AI2(S04)3) at a dose of 5.5 mg AI3+1-1 to
provide an additional barrier to P release from the sediments
(Entrance, 1980). Alum has been applied in only one other
shallow, nonstratified lake (Pickerel Lake in Wisconsin) and
was largely ineffective due to resuspension and deposition
of the alum floe toward the lake center (Knauer and Gar-
rison, 1981). In Long Lake, observations of trophic state in-
dicators during 1981 and 1982 were continued to determine
the duration of alum effectiveness in reducing internal P
loading, thereby clarifying the roles of macrophytes and sedi-
ment as sources of P in this lake.
LAKE DESCRIPTION
Long Lake has an area of 137 ha, mean depth of only 2
m, and a maximum depth of 3.7 m (Fig. 1). Salmonberry
Creek is the major inlet stream, supplying approximately 60
percent of the total input, and Curley Creek is the outlet. The
112
-------
Internal Nutrient Loading Detection and Control
CU.RLEYpCREEK
t
500
1000
METERS
^WATER COLUMN SAMPLING SITES
.E.CREEK
ULVERT
Figure 1.—Water column sampling stations and morphometric
characteristics of Long Lake. N, M, S, and L refer to North, Mid,
South and Lilies.
four sampling stations (North, Mid, South and Lilies) on the
lake are also indicated in Figure 1. The lake has a rather
high flushing rate of 3.6 to 8.0 yr1. High pH levels (8 to 10)
occur during the summer as a result of the lake's low buf-
fering capacity (alkalinity ranging from 10 to 40 mg H as
CaC03) and high primary productivity rates.
METHODS
Internal Loading Calculations
Water and P budgets were calculated for each sampling in-
terval from 1976 through 1981. The P mass balance was
used to indicate the occurrence and extent of internal P
loading from 1976 to 1981. Net sedimentation of release of
P in kg for a given time period was determined from the
following mass balance equation:
S = O +AP - IS - IU - G - ASL
where S is net sedimentation or release, 0 is Curley Creek
outflow, P is change in lake P, IS is Salmonberry Creek
inflow, IU is ungaged surface inflow, G is groundwater, and
ASL is the combined input of bulk atmospheric deposition
and lakeside septic tank leachate. The time steps averaged
about 2 weeks and correspond to the intervals between
sampling trips. A detailed description of the P concentrations
and assumptions used for each of the terms in the equa-
tion are found elsewhere (Jacoby et al. 1982).
The period and extent of internal P loading in 1978 (a
typical pre-restoration year) is shown in Figure 2. Mass
balance data are available for the other pre-drawdown years
from Lynch (1982) and Jacoby et al. (1982). In view of the
previous relationship between net internal loading and lake
concentration, determination of TP concentration during the
summer of 1982 was considered adequate to indicate the
extent of internal loading. Total P was analyzed using the
ascorbic acid molybdenum blue method (Standard Methods,
1975). Other water quality constituents determined during
1982 included temperature, pH, transparency, and
chlorophyll a. Sampling was conducted on a monthly basis
except during June, July, and August when samples were
collected bimonthly.
Sediment P Fractionation
In August 1982, a P fractionation was performed in cores
collected from midlake deep sediments to assess the
available inorganic P (AIP) content relative to AIP levels deter-
mined in August 1978, prior to restoration. The cores were
collected using a gravity corer and sectioned at intervals of
0 to 2, 2 to 5, and 5 to 15 cm immediately upon return to
the laboratory. The fractionation method used was the pro-
cedure developed by Chang and Jackson (1958) and
modified by Williams et al. (1967). the NH4CI, NH4F, 0.1 N
NaOH and the citrate dithionate-bicarbonate (CDB) fractions
were all combined into a single fraction and referred to as
available inorganic P (AIP) (Table 1) as similarly performed
by Lynch (1982) in 1978 before treatment. This fraction is
believed to be comprised of P derived from a short-range
s
CL
a
N D'J F M AM J J
Figure 2.—Net flux of P in mg m 1 across the sediment-water in-
terface determined by a 2-week internal P budget during 1977-78
and its effects on the water column (from Jacoby et al. 1982).
Table 1. — Sediment fractionation procedure for available
inorganic P (modified from Williams et al. 1967)
Extracting Target P Determination
Step Reagent(s) Fraction Method of Extract
1 0.5 N NH4CI; Easily Standard Methods, 1975
30 minutes Soluble-P
2 0.5 N NH4F; pH NH4F-P
8.2; 24 hours
Standard Methods, 1975
3 0.1 N NaOH;
17 hours
1st NaOH-P Dickman &
Fraction Bray, 1940
4 Dithionite Citrate Reductant Watanabe &
Bicarbonate Soluble-P Olsen, 1961
113
-------
Lake Restoration, Protection and Management
Table 2. — Areal-weighted means of macrophyte biomass 1978-1981 in Long Lake (g dry wt/m2 ± 95 percent confidence
interval).
Station
(Percent of total lake area)
South Lilies (18%)
South (17%)
North Shore (14%)
Midlake (51%)
Lake weighted mean
6/23/78
366 ±
191 ±
104 ±
22 ±
124 ±
100
39
76
15
43
8/24/78
496 ±
240 ±
139 ±
27 ±
163 ±
60
95
83
17
49
6/26-27/80
6
29
29
19
20
±
±
+
+
+
16
25
24
13
17
8/23-24/80
62
177
74
70
87
±
+
+
+
+
72
91
56
43
58
6/23/81
150
249
153
41
112
±
±
±
±
±
65
128
111
85
92
8/24/81
141
277
116
89
126
±
+
±
+
+
46
115
100
106
96
order Fe-rich complex, possibly related to Fe(OH)2 (Williams
et al. 1971). AIP is considered that pool of P from which P
is exchanged with overlying water or taken up by rooted
aquatic plants.
RESULTS
Post-Drawdown
The drawdown primarily affected the southern shallow (<2
m) area of the lake resulting in minimal sediment compac-
tion (0.1 m) but an 84 percent reduction in macrophyte
biomass in 1980 (Table 2). Macrophyte biomass was
estimated by the sampling procedure described in Jacoby
et al. (1982). Surprisingly, during the summer of 1980 water
column mean TP content (36 mg m~3) was approximately
50 percent lower than previous summers. Based on the mass
balance calculations for any 2-week interval in this period,
internal loading of P was not evident (Fig. 3). Correspondingly
dense blue-green algal blooms did not occur. Mean sum-
mertime chlorophyll a was 15 mg rrr3, relative to 27 mg rrr3
in 1978.
Observations on the P distribution in the lake during winter
versus summer and sediment core analyses implicated
midlake sediments as the major source of internal P loading
during the summer (Jacoby et al. 1982). Our initial hypothesis
was that winter decomposition of macrophytes supplied an
initial input of P contributing to spring algal blooms as well
as enrichment of midlake sediments. Direct release of this
P from the sediments could represent a substantial contribu-
tion to internal P loading in this lake.
High water column pH has been implicated as a
mechanism by which this release occurs in Long Lake
(Jacoby et al. 1982). Under high pH conditions, Long Lake
sediments readily release P. Pre-drawdown summertime pH
levels approached 10 while 1980 levels were significantly
lower (5 percent level) (Table 3). Release via this mechanism
would thereby be lower in 1980. In addition, the initial sup-
ply of late winter P had decreased as a result of the
macrophyte reduction, decreasing the amount available to
spring algal blooms and enrichment of midlake sediments.
Therefore the reduced available P, primary productivity, and
pH may have resulted in decreased release of P from the
sediments in 1980 (Jacoby et al. 1982).
Release of P from sediment because of anaerobic condi-
tions at the sediment-water interface during calm conditions
and temperature-controlled microbial decomposition also
may have accounted for the high internal loading rates
observed prior to restoration. Vertical gradients in P concen-
tration have been occasionally observed in Long Lake and
may have resulted from sedimentation and subsequent
mineralization of P at the sediment-water interface.
Measurement of a soluble reactive P (SRP) gradient in
the midlake station water column was performed using an
in situ sampler designed to withdraw water samples 2, 5,
10, 25, and 50 cm above the sediment surface without sedi-
ment agitation (Fig. 4). Samples were collected during an
18-hour period on a calm summer night in 1979 when the
lake was drawn down (Lynch, 1982). Soluble reactive P con-
centrations were uniform from 2 to 150 cm above the sedi-
ment surface except during an especially calm period at 0100
1 N ' D
1979
A ' M
1980
H
<
tr
Figure 3.—Net flux of P in mg m"1 across the sediment-water in-
terface determined by a 2-week internal P budget during 1979-80
and its effects on the water column (from Jacoby et al. 1982).
Table 3. — Mean summer (June-August) volume weighted
TP concentration mg m~3) and median, mean and range of
pH.
Drawdown
Alum
Year
11976
1977
1978
1979
1980
1981
1982
Median
PH
9.1
9.0
8.6
7.2
8.1
8.0
7.9
Mean
pH
_
—
7.8
—
27.0
26.6
7.7
Range
PH
8.0-9.2
7.7-9.7
7.4-8.8
7.1-8.3
6.7-8.4
6.2-9.5
6.9-9.2
Mean TP
(mg m"3
58
61
76
83
36
29
29
11ncomplete data for this period — a conservative estimate of TP levels.
2 Statistically significant difference from 1978 pH levels (P< 0.005)
114
-------
Internal Nutrient Loading Detection and Control
Figure 4.—In situ sampler designed to obtain water adjacent to the
sediment-water interface without disruption (from Lynch, 1982).
hours when water collected 2 m above the sediment was
significantly (5 percent level) enriched in SRP (Fig. 5). This
experiment demonstrated that: (1) SRP does not accumulate
appreciably at the sediment-water interface during the greater
part of the day, i.e. any P release would be quickly advected
from the interface by wind-induced water movement;
however, (2) SRP can accumulate at the sediment-water in-
terface under calm conditions (Lynch, 1982).
Observations during the bimonthly field sampling indicate
that temporary stratification occasionally occurs, resulting in
vertical gradients of temperature, oxygen, and P (Table 4).
At a distance of 0.5 m above the sediment, TP was 124 mg
m"3 versus 57 mg m"3 near the water surface. Dissolved ox-
ygen was 2.2 mg I'1 at 0.5 m above the sediment versus
9.8 mg h1 0.5 m below the water surface, suggesting that
the sediment-water interface may have been totally reduc-
ed at this time. This process of P deposition, mineralization,
and reintroduction into the water column may be important
in Long Lake, accounting for the large fluctuations between
net release and sedimentation seen in the mass balance
calculations in the prerestoration summers (Fig. 6). Upward
Table 4. — Temperature, oxygen, and P gradients at the
Midlake station, 7/21/78.
Water
depth
(m)
0.5
2.5
Dissolved
oxygen
(mg/l)
9.8
2.2
Temp
(C°)
25.0
21.7
Total P
(mg/m3)
57
124
SRP
(mg/m3)
14
34
z
111
z
Q
111
en
o
z
<
K
U)
150-
120-
— 90-
60-
30-
0 -
H— .
i
^ 1 i-
/ ,
J_
l
1
i,
7 '
f
A 6/19/79 1700
• 6/20/79 0100
• 6/20/79 1 100
\..>c.
I— UL
• II v»™_ 1
V'V\"/v^';/-.r,^'-'>/''i'i" Sed i m e n I;,--"-; >"'"•', ^"-.'"-"'r1: • -\v»
6 8 10 12
SRP
( mg P/ m3 t \ s.d., n;4 J
Figure 5.—Soluble reactive P (SRP) concentrations 2, 5,10,25, 50,
150 cm from the sediment-water interface obtained with the in situ
interface sampler at the Midlake station on 6-19 and 6-20-1979 (from
Lynch, 1982).
diffusion of P to the overlying water via any of these three
mechanisms would be enhanced by the more frequent mix-
ing typical of shallow lakes. This diffusion of P in Long Lake
sediments is also normally favored by the gradient between
sediment interstitial and overlying water (Lynch, 1982).
Post-Alum Treatment
Following the alum treatment in September 1980, TP con-
tent dropped from near 40 mg nv3 to less than 10 mg nv3.
Total P levels in 1981 remained considerably below pretreat-
ment concentration in 1976 to 1978 with a mean summer
concentration of 29 mg nr3 The mean chlorophyll a level
in 1981 of 14 mg nr3 was also lower than the 1978 mean
of 27 mg nv3 (Welch et al. 1982). Primary productivity was
lower with a resulting decrease in water column pH during
the summer. There was a significant difference (5 percent
level) between pH in 1978 (mean = 7.8 pooled at H+) and
pH in 1981 (mean = 6.6) (Table 3).
In addition, the mass balance calculations indicated that
net internal P loading did not occur in 1981 as it had in the
3 prerestoration years (Welch et al. 1982) (Fig. 7). The sedi-
ment remained a complete sink to P in 1981 as in 1980
following drawdown. This and the dramatic reduction in P
content during the summer of 1981 indicate that the alum
floe remained well-incorporated in the surficial sediments,
effectively blocking the release of P. The curtailment of in-
115
-------
Lake Restoration, Protection and Management
1}
s
6
3
_i
I
o
100-
80-
60-
40 -
20-
0
-•-•- 1978
-D-D-1981
-0-0- 1982
* p y v
/ V / \ f *
/ \ J\ ' * ' \ 1 *-
120 T
80-
0
s
60-
—
••o-
bD-Q.d
N'D'J'F'M'A'MJ'J'A'S'
Figure 6.—Chlorophyll a and total phosphorus content; volume
weighted means for the whole lake prior to and following September
1980 alum treatment. Data points are represented by closed circles
in 1978, open squares in 1981 and open circles in 1982.
ternal loading was maintained in Long Lake in spite of the
near-complete recovery of the macrophyte standing crop and
presumably resumption of their enrichment of midlake sur-
ficial sediments by 1981.
Trophic State Indicators in 1982
During 1982, TP content remained low (mean = 29 mg rrT3)
compared to pretreatment (76 mg rrr3) and was the same
as the 1981 mean (Fig. 6 and Table 3). As a result, internal
loading of P during this period was also assumed to be low.
Dense blue-green algal blooms did not occur although the
blue-green alga G/eotrichia was readily noticeable in the
water column during the latter part of the summer. However,
chlorophyll a levels (mean = 7 mg rrT3) remained substan-
tially lower than 1978 levels (mean = 27 mg nr3) (Figure
6). Summer pH levels in 1982 were slightly higher than in
the 2 previous years and not statistically different from those
in 1978, before treatment.
The fractionation of midlake cores showed that AIP was
two to three times lower in cores collected in 1982 compared
to AIP levels in 1978 (Table 5). The cause of the lower AIP
in sediments collected in 1982 is difficult to ascertain. This
may indicate that alum is still effective in complexing P in
surficial sediments despite continual macrophyte enrichment.
DISCUSSION
An understanding of the mechanisms by which internal
loading of P occurs in Long Lake has been increased by
monitoring the effects of the two restorative manipulations
applied in this lake.
The alum addition in autumn 1980 lowered lake TP levels,
even further than those observed after the 1979 drawdown
in the summers of 1981 and 1982. This verifies the sedi-
ment rather than the macrophytes as the immediate source
of P. This is in agreement with our earlier findings that in-
dicated the deep midlake area was more important in terms
of internal P loading in the summer (Jacoby et al. 1982;
Lynch, 1982). If P loading was directly from the macrophytes
to the open water, alum would not have lowered internal
6
4
2
LLJ
5 o
cr
x -2
ID
ul -4
-6
NET INTERNAL LOADING
1980 -
1
UJ^
-
-
NET
1981
mr^
SEDIMENTATION
ASONDJ FMAMJ J A
TIME (MONTHS)
Figure 7.—Net flux of P in mg m~2 day"1 across the sediment-water
interface determined by a 2 week internal P budget during
1980-1981.
Table 5. — Phosphorus fractionation of cores taken from Midlake station (mg P/kg dry sediment ± 95% confidence interval).
8/14/78
Depth (cm)
NH.CI
NH.F
OINNaOH
8/17/82 (mg P/kg dry sediment ± 1 S.E. n= 2)
CDB
Total or AIP
0-2
2-5
5-15
6 ± 2
4 ± 1
3 ± 1
695 ± 31
390 ± 12
275 ± 8
461 ± 21
211 ± 12
140 ± 6
240 ± 21
265 ± 15
266 ± 18
1422 ± 26
880 ± 23
671 ± 21
Depth (cm)
0-2
2-5
5-15
NH4CI
1 ± 1
2 ± 0
1 ± 0
NH4
20 ±
46 ±
51 ±
F
1
6
12
O.INNaOH
545 ± 1
450 ± 82
64 ± 24
CDB
53 ± 1
32 ± 1
136 ± 47
Total or AIP
619 ± 72
529 ± 77
251 ± 84
116
-------
loading and TP would have again increased during the sum-
mer with the return of the macrophytes.
Laboratory and field observations have indicated several
mechanisms by which the release of P from surficial
sediments may occur. Release of P caused by high water
column pH has been implicated in other systems (Ander-
son, 1974; Lijklema, 1976). In Long Lake high summer water
column pH is commonly observed. Laboratory studies have
verified the ease by which Long Lake sediments release P
under high pH conditions (Jacoby, 1981). After drawdown
and alum treatment, 1980 and 1981 water column pH were
significantly lower than pre-restoration summers, suggesting
a possible cause for reduced internal P loading and lower
TP levels during those summers.
Two other mechanisms may be important in transporting
P from sediments to overlying water in Long Lake.
Phosphorus is believed to be quickly recycled and retained
in the system via anaerobic release of P at the sediment-
water interface and/or temperature-controlled microbial
decomposition of sedimented TP. The vertical gradients of
temperature, oxygen, and phosphorus that have occasionally
been observed in Long Lake suggest the potential
significance of these mechanisms. The significant enrich-
ment of waters immediately above the sediment surface
demonstrates the rapid mineralization of P and its subse-
quent movement to the water column, facilitated by wind-
mixed currents (Lynch, 1982).
In nonstratified lakes, this nutrient cycling is very efficient
because sedimenting P is not irretrievably lost from the mixed
layer even if it is deposited on the sediment surface. Rapid
mineralization by decomposers is enhanced by the relative-
ly high temperatures of surficial sediments (Fee, 1979). This
mineralized P may not be totally retained by the sediment
but instead released, particularly during periods of increas-
ed temperature. Tessenow (1972) demonstrated the impor-
tance of rapid recycling of P from sedimented plankton in
Ursee, Germany.
In Long Lake the alum floe apparently remained well in-
corporated in the surficial sediment and served as an effec-
tive barrier to the vertical diffusion of SRP that otherwise may
have occurred via any or all of three mentioned mechanisms
of P release. The effect of alum has persisted for 2 years
since the alum treatment, in spite of a full macrophyte
recovery and an assumed increased particulate P input from
the dying macrophytes to surficial sediments. This is in con-
trast to the previous experience with the effectiveness of alum
inactivation of P in a shallow lake (Knauer and Garrison,
1981). Alum may have continued to be effective in 1982 by
inactivating or complexing newly deposited P from
macrophytes and other sources as indicated by the lower
AIP levels present in sediment. However, this presumes that
the 1980 alum addition caused the reduced sediment AIP
observed in 1982, and was not related to increased oxida-
tion during drawdown (because of the reduced depth of
overlying water).
Macrophytes may still be considered as the original source
of internal P loading via dieback, winter decomposition, and
transport to midlake sediments. This premise is supported
by low internal loading in 1980 following drawdown and 84
percent reduction in macrophyte biomass. While it is likely
that the macrophyte reduction has the greatest impact on
internal P loading, the possibility must still be considered that
the abrupt and extensive decrease in internal loading follow-
ing drawdown may have been caused by increased oxida-
tion of surficial midlake sediments. That condition could have
resulted from the decrease in depth of the overlying water
from a maximum of 3.5 to 1.5 m which would have provid-
ed less opportunity for stratification. Nevertheless, internal
loading was observed to continue while the lake was drawn
down in 1979. If increased oxidation due to drawdown had
Internal Nutrient Loading Detection and Control
been an important impediment to internal loading such
loading would not have continued. Therefore, it is reasonable
to assume that without the alum treatment, internal P loading
and water column P levels would have returned to pre-
drawdown levels in 1981 and 1982 because of the rapid
recovery of macrophytes by the autumn of 1980.
The long-term effectiveness of alum in such a shallow lake
as this is of course an important question. However, water
quality has remained substantially improved for the last 2
years. The duration of effectiveness may not be as long as
in stratified lakes where the effect has persisted for at least
5 years (Cooke and Kennedy, 1981). As macrophytes con-
tinue to enrich the surficial sediments through senescence
and decomposition, the beneficial effect of alum may be
reduced. Nevertheless, the fact that total P content in 1982
did not increase over that in 1981, the year after treatment,
suggests that effectiveness could last several years.
There are some potential detrimental side effects,
however. The improved water column clarity alum brings
about may result in an even greater biomass of macrophytes
which could increase their importance in P transport and
sediment enrichment in this lake. This could in turn limit the
long-term effectiveness of the alum treatment. An increase
in the density and duration of Potamageton praelongus and
P. pectinatus in the deeper area (approx. 3 m) of the lake
was observed qualitatively during the summer of 1982. Dur-
ing previous summers this area of the lake was relatively
free of noticeable macrophyte abundance, possibly because
of the shading effect from the dense algal biomass. Many
recreationists also noticed this occurrence and complained
of its impact on boating and fishing activities despite the
overall improvement in water quality and lack of blue-green
algal blooms. The reduction in mean summer chlorophyll a
from 27 mg nr3 in 1978 to 14 and 7 mg nr3 in 1981 and
1982 was associated with increases in transparency from
a mean of 1.5 to 2.2 and 3.0 m, respectively. Aquatic
macrophytes could theoretically colonize deeper areas of a
lake as a result of improved water clarity.
As discussed by Welch et al. (1979,1982), other restorative
techniques should be examined for use in a system such
as Long Lake. Long-term improvements in shallow lakes may
be better attained by macrophyte harvesting and/or dredg-
ing. At least macrophyte harvesting would be able to inter-
rupt annual P cycling but may even promote sediment P
depletion by removing the macrophytes which in turn, have
removed P from the sediments. In addition, the removal of
macrophyte stands would reduce their nuisance aspects.
Dredging of the top 1/2 m of sediment deposited over the
last 50 years in Long Lake would have reduced the surficial
sediment P content to less than one half the present level
(Lynch, 1982).
SUMMARY
1. The alum addition in autumn 1980 lowered lake TP
levels even further in the summers of 1981 and 1982. This
verifies the sediments rather than maprophytes as the im-
mediate source of P
2. The effect of alum has persisted for 2 years in spite
of a full macrophyte recovery and an assumed increased
particulate P input from macrophytes to surficial sediments
between those years.
3. Macrophytes may still be considered as the original
source of internal P loading via winter decomposition and
transport to midlake surficial sediments.
4. Without alum treatment, internal loading and water col-
umn P levels should have returned to pre-drawdown levels
in 1981 and 1982 because of the rapid recovery of
macrophytes by autumn 1980.
117
-------
Lake Restoration, Protection and Management
5. The long-term effectiveness of alum in this shallow lake
is as yet unknown although no change has occurred in 2
years.
ACKNOWLEDGEMENTS; Appreciation is extended to W.T. Trial
for critical reviews of this manuscript and for sample collection and
analysis. We also would like to thank MA Perkins for his contribu-
tions to this project over the last 5 years.
REFERENCES
Anderson, J.M, 1974. Nitrogen and phosphorus budgets and the role
of sediments in six shallow Danish Lakes. Arch. Hydrobiol.
74:427-550.
Chang, S.C., and M.L. Jackson. 1958. Fractionation of soil phos-
phorus, Soil Sci. 84:133-144.
Cooke, G.D., and R.H. Kennedy. 1981, Phosphorus inactivation: A
summary of knowledge and research needs. Pages 395-399 in
Restoration of Lakes and Inland Waters. EPA 440/5-81-010. U.S.
Environ. Prot. Agency, Washington, D.C.
Dickman, S.R., and R.H. Bray. 1940. Colorimetric determination of
phosphate. Ind. Eng. Chem. 12:665-668,
Entrance. 1980. Long Lake restoration project: final rep. Entrance
Engineers, Bellevue, Wash.
Fee, E.J. 1979, A relation between lake morphometry and primary
productivity and its use in interpreting whole-lake eutrophication
experiments. Limnol. Oceanogr. 24:401-416.
Hufschmidt, P.W. 1978. Interactions between macrophytes and
phytoplankton in Long Lake. M.S. Thesis. Univ. of Washington.
Jacoby, J.M. 1981. Lake phosphorus cycling as influenced by draw-
down and alum addition. M.S. Thesis. Univ. of Washington.
Jacoby, J.M., D.D. Lynch, E.B. Welch, and MA. Perkins. 1982.
Internal phosphorus loading in a shallow eutrophic lake. Water
Res. 16:911-919.
Kennedy, R.H., and G.D. Cooke. 1982. Control of lake phosphorus
with aluminum sulfate: Dose determination and application tech-
niques. Water Res. Bull. 18:
Kistritz, R.U. 1978. Recycling of nutrients in an enclosed aquatic com-
munity of decomposing macrophytes (Myriopbyllum spicatum).
Oikos 30:561-569.
Knauer, D.R., and P.J. Garrison. 1981. A comparison of two alum
treated lakes in Wisconsin. Pages 412416 in Restoration of Lakes
and Inland Waters. EPA 440/5-81-010. U.S. Environ. Prot, Agen-
cy, Washington, D.C.
Landers, D.H. 1982. Effects of naturally senescing aquatic macro-
phytes on nutrient chemistry and chlorophyll a of surrounding
waters. Limnol. Oceanogr, 27:428-439.
Lijklema, L. 1976. The role of iron in the exchange of phosphate be-
tween water and sediments. In H.L. Golterman, ed. Interactions
Between Sediment and Freshwater. Proc. Int. Symp. Amsterdam,
Netherlands. Junk, The Netherlands.
Lynch, D.D. 1982. Internal loading and sedimentation of phosphorus
in a shallow eutrophic lake. M.S. Thesis, Univ. of Washington.
Perkins, M.A., E.B. Welch, and J.O. Gabrielson. 1979. Limnological
characteristics of Long Lake, Kitsap County, Washington. Lim-
nological and Socioeconomic Evaluation of Lake Restoration Pro-
jects; Approaches and Preliminary Results. EPA-600/3-79-005.
U.S. Environ. Prot. Agency, Washington, D.C.
Prentki, R.T. 1979. Depletion of phosphorus from sediment colonized
by Myriophyllum apfcafum. in J.E. Beck, R.T. Prentki, and O.L.
Loucks, eds. Aquatic Plants, Lake Management and Ecosystems
Consequences of Lake Harvesting. Proc. Madison, Wis.
Soltero, R.A., D.G. Nichols, A.F. Gasperino, and M.A. Beckwith.
1981. Lake restoration: Medical Lake, Washington. J. Freshw.
Ecol. 1:155-165.
Standard Methods for the Examination of Water and Wastewater.
1975. 14th ed. Am. Pub. Health Assn., Washington, D.C.
Tessenow, U. 1972. Losungs-, diffusions- und sorptionsprozesse in
de oberschict von seesedimenten. Arch. Hydrobiol. Suppl.
38:353-398.
Watanabe, F.S., and S.R. Olsen. 1961, Colorimetric determinations
of phosphorus in water extracts of soil. Soil Sci. 93:183-188.
Welch, E.B., M.A. Perkins, D.D. Lynch, and P.W. Hufschmidt. 1979.
Internal phosphorus related to rooted macrophytes in a shallow
lake. In J.E. Beck, R.T. Prentki, and O.L. Loucks, eds. Aquatic
Plants, Lake Management, and Ecosystem Consequences of Lake
Harvesting, Proc., Madison, Wis.
Welch, E.B., J.P. Michaud, and M.A. Perkins. 1982. Alum con-
trol of internal phosphorus loading in a shallow lake. Water Resour.
Bull, (in rev.)
Williams, J.D.H., J.K. Syers, and T.W. Walker. 1967. Fractiona-
tion of soil inorganic phosphate by a modification of Chang and
Jackson's procedure. Soil Sci. Soc. Am, Proc. 31:736-739.
Williams, J.D.H., J.K. Syers, D.E. Armstrong, and R.F. Harris.
1971. Characteristics of inorganic phosphate in noncalcareous lake
sediments. Soil Sci. Soc. Am, Proc. 35:556-561.
118
-------
MACROPHYTE DIEBACK: EFFECTS ON NUTRIENTS AND
PHYTOPLANKTON DYNAMICS
DIXON H. LANDERS
State University Research Center at Oswego
SUNY—Oswego
Oswego, New York
ELIZABETH LOTTES
Columbus, Ohio
ABSTRACT
Annual senescence cycles of aquatic macrophytes can be important with regard to internal loading to
freshwaters. Sediment nutrients accumulated in macrophytes as above sediment biomass are released
as decomposition progresses, resulting in algal stimulation. Dieback of Myriophyllum spicatum (Eurasian
watermilfoil) resulted in phosphorus and nitrogen increases within in situ enclosure experiments with cor-
responding elevations in phaeophytin-corrected chlorophyll a. Senescing plants caused phytoplankton
biomass increases and affected population dynamics favoring blue-greens and heterotrophic taxa. The
annual phytoplankton biomass increase attributable to senescing plants is between 16 and 59 metric tons
in the reservoir system studied.
INTRODUCTION
Investigations of the role of aquatic macrophytes as agents
of internal loading in freshwaters have followed two paths.
One line of research has been concerned with the
mechanism of active pumping of nutrients from sediment
pools into overlying waters by vascular systems of aquatic
plants (McRoy and Barsdate, 1970: DeMarte and Hartman,
1974). The degree to which macrophytes can passively or
actively pump and release nutrients is apparently species
dependent, with wide differences between various taxa. It
has been demonstrated that for Myriophyllum spicatum L.
(Eurasian watermilfoil) little physiological release of nutrients
occurs (Barko and Smart, 1980; Best and Mantai, 1978). In
contrast, the marine macrophyte Zostera may release signifi-
cant nutrients (McRoy and Barsdate, 1970).
The other direction of investigation has focused upon the
dieback effects of senescing populations of macrophytes
(Kistritz, 1978; Nichols and Keeney, 1973; Jewell, 1971). In
temperate climates, where growing seasons are rigorously
divided by annual temperature extremes, macrophytes,
especially emergent and floating leaf species, demonstrate
an annual cycle of growth, senescence, and overwintering.
During senescence, which normally occurs in late summer
or fall, biomass that has accumulated during the growing
season decomposes. As the plant tissues decompose the
potential is greatest for macrophyte influences upon the biotic
and abiotic components of the surrounding waters.
The study reported here was conceived to demonstrate
and quantify the effects of senescing macrophytes upon
water quality parameters in a large, shallow, soft-water reser-
voir in Southern Indiana. However, the situation described
is cosmopolitan to the extent that generalizations can be
drawn to similar situations wherever they may occur.
METHODS
Monroe Reservoir (39° 03'N, 86° 25'W) contains extensive
growths of the macrophyte Myriophyllum spicatum. It grows
largely in monospecific stands in water 1 to 3 m deep.
Noticeable annual fluctuations in macrophyte abundance are
linked to changing reservoir water levels that stress plants
growing at the greater depth extremes by limiting available
light during critical spring growth.
In July 1978 before signs of senescence or decay ap-
peared, six open tubular isolation enclosures (2 m diameter)
were installed in the littoral (1.4 m) of the upper basin of
Monroe Reservoir. Three replicate enclosures were located
in areas of undisturbed milfoil growth; three other replicate
enclosures were secured in a nearby area of the same depth,
previously denuded of milfoil growth. Further details of
enclosure and experimental design are given elsewhere
(Landers, 1979,1982). Comparisons of biotic and abiotic dif-
ferences between the two enclosure treatments over the 119
days of the experiment enable me to assess the importance
of rooted aquatic plants as agents of internal nutrient loading
to the reservoir.
Mid-depth water samples were routinely collected (3 to 7
days) from each enclosure and two open-water locations ad-
jacent to the enclosures. Samples were analyzed for nitrogen
and phosphorus fractions (ammonium, nitrate and nitrite,
soluble reactive phosphorus (SRP), total phosphorus (TP),
particulate phosphorus (PP), and total soluble phosphorus
(TSP) as well as other physico-chemical measurements (pH,
alkalinity, dissolved oxygen, temperature, specific conduc-
tance, turbidity, and Secchi disc). Sub-samples of the water
used for the physico-chemical measurements were split and
used to monitor phaeophytin-corrected chlorophyll a concen-
trations (see Landers, 1982 for analytical details) and for
determining phytoplankton biomass.
Aliquots for phytoplankton enumeration were preserved
in the field with acetic Lugol's solution. Counts were made
with an inverted microscope (400 X) using standard coun-
ting and settling techniques (Utermohl, 1958; Nauwerck,
1963). All algae were counted in two perpendicular transects.
Those taxa receiving more than 15 counts in the two initial
transects were counted in additional transects until they total-
ed 100. The lower detection level of this method is 811 units
per ml at 400 X.
Identification of phytoplankton to the level of genus was
aided with the following references: Pascher (1925), Prescott
119
-------
Lake Restoration, Protection and Management
(1954, 1962), Skuja (1948), Smith (1950). Depending upon
the morphology of the various taxa, final counts were ex-
pressed as either cells, trichomes, filaments, colonies, ag-
gregations, or coenobia. Mathematical formulae appropriate
to the geometric shape of the different algal forms were us-
ed to compute the biomass of each taxa.
RESULTS AND DISCUSSION
Enclosed macrophytes as well as open-water plants ap-
peared similar throughout the study as they progressed
through their annual decomposition cycle. I divided the four-
month study into five divisions based upon the visual ap-
pearance of the plants: healthy (day 0 to 21), initial decay
period (day 21 to 51), advanced decay period (day 51 to 63),
post decay period (day 63 to 96), and winter conditions (day
96 to 119). Vertical broken lines on the figures that follow
represent these major divisions.
As reported previously (Landers, 1982), significant elevated
concentrations of TP, TSP, PP, and SRP were found in the
enclosures with plants as macrophytes died. The denuded
enclosures generally were lowest in these phosphorus con-
stituents, and the open sites, showing a diluted influence
of senescing macrophytes resulting from rapid exchange of
water between plant concentrations and open areas, were
generally intermediate to the two enclosed treatments.
The analytical results for nitrogen concentrations were not
as easily interpreted as the phosphorus data, probably
because of the more complicated biogeochemical cycles of
nitrogen. Nonetheless, the influence of senescing
macrophytes in raising ammonium and nitrate concentra-
tions was apparent during peak episodes of macrophyte
decomposition.
Chlorophyll a concentrations, interpreted as an indirect
estimate of algal biomass, showed dramatic increases in the
enclosures with plants as nutrient levels increased coinci-
dent with macrophyte dieback. From a starting level similar
to the other two treatments, chlorophyll a concentrations
reached a mean level between day 31 and 91 of 61.4
mg-m~3 (SD=14.8, n= 11), 10x greater than the cor-
responding mean value for the denuded treatment (6.1
mg-rrr3, SD = 2.0, n = 11). The open situation was in-
termediate (15.7 mg-m~3, SD = 6.7, n = 11).
These data demonstrate the important influence that
senescing macrophytes can have upon chlorophyll a
estimates of phytoplankton biomass in adjacent waters. Also,
the diluted open water effect of senescing macrophytes is
evidenced through the intermediate open water value which
was 257 percent of the control treatment (denuded enclo-
sures). Periphytic algal growth was similarly highly respon-
sive to nutrient pulses originating from decomposition of
aquatic plants (Landers, 1982).
Biomass data from actual counts of preserved algal
samples provide additional information concerning the
decompositional effects of macrophytes. Not only are
biomass trends for the three treatments consistent between
the two different measurement techniques (chlorophyll a and
microscope counts), but information is obtained regarding
the temporal dynamics of phytoplankton taxa. Table 1 lists
the phytoplankton taxa identified from the enclosure experi-
ment. The relationships between the seven major groups
varied between treatments and over time. Figures 1, 2 and
3 graphically illustrate the changing composition of the four
most common groups in each of the treatments.
Chlorophyta dominated the denuded enclosures
throughout the study (Fig. 1). Only two peaks greater than
10 mg-r1 were observed: the first peak (day 11 to 16) was
comprised of Chlorophyta and Cyanophyta while the second
peak (day 56) contained mostly Euglenophyta. After day 60,
the chlorophytes generally decreased in importance as the
Phytoplonkton Biomoss - Denuded Enclosures
I II i
Figure 1 .—Total biomass and biomass of the four most abundant
phytoplankton taxa in the denuded enclosure treatment. Plotted
values are means of replicate counts and means of three replicate
enclosures (n = 6).
Cyanophytes became more abundant. At the same time,
Cryptophyceae, which had been only of transient importance
earlier, comprised approximately one third of the total
biomass and disappeared by the last sampling day.
Cyanophytes dominated the open-water treatment in the
latter half of the study with no taxa dominating the biomass
in the first half (Fig. 2). Euglenophytes were present
sporadically throughout the experiment and, largely because
of their early abundance, were the second most prevalent
taxa. The late, sustained increase in Cyanophytes probably
resulted from the influence of senescing macrophytes that
appeared in an advanced state of decay from day 51 to 63,
about the time Cyanophytes started their late season popula-
tion increase.
Phytoplankton biomass and composition fluctuated sharply
in the enclosures with plants (Fig. 3). As in the open-water
sites, Cyanophytes dominated the biomass in this treatment
but they were accompanied by a group not important in the
other two treatments, the Peridiniales. Some organisms
within this taxa are noted for heterotrophy (Round, 1965;
Phytoplonkton Biomass Open
10 20 30 40 50 60 70 80 90 IOO 110 120
DAY
Figure 2.—Total biomass and biomass of the four most abundant
phytoplankton taxa in the open water sites. Plotted values are means
of replicate counts and replicate sites (n = 4).
120
-------
Internal Nutrient Loading Detection and Control
Table 1. — List of major algal divisions and the taxa in
each Identified from counts of all treatments during the
1978 enclosure experiment.
Cyanophyta
'Anabaena circinal/s (spiroides?)
'Anabaena (levanderi?)
'Aphanizomenon sp.
'Aphanocapsa delicatissima
'Chroococcus sp.
"Coelosphaerium naegelianum
"Lyngbya limnetica
' Merismopedia minima
' Merismopedta tenuissima
'Microcystis aeruginosa
''Oscillatoria limnetica
'Oscillatoria nigra
'Oscillatoria princeps
Chlorophyta
Ankistrodesmus falcatus
Ankistrodesmus sp.
'Botryoccus sp.
Carter/a sp.
Chlamydomonas sp.
Chlorella vulgaris
Closteriopsis longissima
'Coelastrum sp.
Cosmarium sp.
"Crucegenia sp.
*Dictyosphaerium pulchellum
Gloeocystis sp.
'Golenkinia radiata
Kirchneriella obesa
"Miractinium pusillum
Oedogonium
"Oocystis sp.
"Pediastrum sp.
"Planktosphaeria gelatinosa
Scenedesmus (abundans?)
Scenedesmus armatus
Selenastrum sp.
"Sphaerocystis sp.
Staurastrum sp.
'Tetraedron gracile
* Tetraedron sp.
Treubaria setigerum
Ulothrix sp.
Westella sp.
Chrysophyceae
Chrysococcus (biporus?)
Chrysococcus sp.
Mallomonas caudata
Bacillariophyceae
Cyclotella bodanica
Cyclotella sp.
Fragilaria construens
Melosira sp.
Navicula sp.
Synedra sp.
Terpsinoe sp.
Cryptophyceae
Chroomonas acuta
Chilomonas sp.
Cryptomonus sp.
Peridiniales
Ceratium hirundella
Peridinium sp.
Euglenophyta
Euglena polymorpha
Euglena sp.
Phacus sp.
Trachelomonas armata
Trachelomonas sp.
'Indicates taxa counted as trichomes, filaments, colonies, aggregations, or
coenobla.
Droop, 1974) and their success in these enclosures may be
a direct result of the rich nutrient mixture resulting from
macrophyte decomposition.
The two largest phytoplankton peaks in this treatment coin-
cided with major episodes of macrophyte senescence. The
peak of day 46 (41 mg-r1) was late in the initial decay
period while the larger, more sustained peak (days 63 to 84)
occurred during the post-decay period.
The phytoplankton biomass differences between denud-
ed enclosures (Fig. 1) and enclosures with plants (Fig. 3)
are large and distinct with the presence of macrophytes be-
ing the outstanding variable. Nitrogen and phosphorus flux
within the enclosures with plants was probably responsible
for the gross phytoplankton biomass differences but other
unmeasured parameters could have been important in deter-
mining interactions between major taxa.
Undiluted release of vitamins and organic compounds
from decomposing macrophytes may have been responsi-
ble for the presence of heterotrophic taxa (Peridiniales) within
the enclosures with plants. Open-water sites, with their
diluted effect of macrophyte decomposition, showed no
dominance of the Peridiniales and only one early peak of
Euglenophytes; both groups are known to contain some
heterotrophs.
Table 2 summarizes the abundance of the four major
phytoplankton groups in each treatment. In all cases these
major taxanomic groups dominated total biomass. Both open
sites and denuded enclosures were dominated by one group
while the enclosures with plants had three taxa influencing
total biomass to similar degrees.
If two assumptions are accepted, the Monroe Reservoir
upper basin phytoplankton biomass directly attributable to
internal loading from macrophytes can be estimated. First,
one must assume that the difference in phytoplankton
biomass between the two enclosure treatments resulted from
the senescence and decomposition of aquatic macrophytes;
second, one has to assume that similar effects occur in other
upper basin locations inhabited by milfoil. Then, if the total
plant coverage area is known (1978 = 1.179 * 106m2) and
a mean water depth of 1.5 m is accepted, it becomes a sim-
ple task to calculate the phytoplankton biomass attributable
to macrophytes by multiplying the water volume surrounding
plant coverage (m3) by the mean phytoplankton biomass dif-
ference between the two enclosure treatments (0.00928
kg/m3) over the duration of the experiment.
Phytoplankton Biomass - Enclosure With Plants
Figure 3.—Total biomass and biomass of the four most abundant
phytoplankton taxa in the enclosure with plants. Plotted values are
means of replicate counts and replicate enclosures (n = 6).
121
-------
Lake Restoration, Protection and Management
Table 2. — Summary of the mean biomass of the four major phytoplankton groups in each of the treatments and their
percent of total biomass.
Treatment
Taxa
I03mg-m-3(n= 132)
% of Total biomass
Denuded
Denuded
Denuded
Denuded
Denuded
Chlorophyta
Cyanophyta
Cryptophyceae
Euglenophyta
Total algal biomass
2.00
1.22
0.84
0.82
5.42
36.9
22.5
15.5
Sum 90.0
Open
Open
Open
Open
Open
Cyanophyta
Euglenophyta
Chlorophyta
Cryptophyceae
Total algal biomass
2.75
1.12
0.97
0.70
6.62
41.5
16.9
14.6
10.6
Sum 83.6
Plants
Plants
Plants
Plants
Plants
Cyanophyta
Peridiniales
Chlorophyta
Cryptophyceae
Total algal biomass
3.45
3.30
2.65
1.59
14.70
23.5
22.4
18.0
10.8
Sum 74.7
The 1978 result is that 16.4 metric tons (live mass) of
planktonic algae developed in response to internal loading
by macrophytes. Parallel calculations made for 1971, when
macrophyte coverage was greater (4.24 * 106 m2), result
in an estimated 59.1 metric tons of phytoplankton (live mass).
Internal loading by macrophytes cannot be overlooked
when a critical view of lake processes is desired. It should
be noted that milfoil has been shown to obtain most of its
nutrient requirements from the sediments through its root
system rather than from the water through shoots, stems,
or leaves. Therefore, sediment nutrients that would other-
wise generally be unavailable to planktonic organisms, are
mobilized by milfoil and released via annual dieback, at a
time when growing conditions are suitable for algal response
The importance of this process to a particular lake or reser-
voir depends upon the extent of littoral rooted plant develop-
ment, the specific macrophyte species involved, and the
trophic status of the lake. In the case of mesotrophic Monroe
Reservoir, this process can be very important in determin-
ing the production and composition of late summer and
autumn phytoplankton populations, especially in the upper
basin. Consequently, any comprehensive assessment of
nutrient availability to an aquatic system should evaluate the
importance of internal loading via rooted aquatic plants.
ACKNOWLEDGEMENTS: Financial support provided by the Office
of Water Research Technology, Indiana Academy of Science, and
Indiana University Grant-ln-Aid of Graduate Research.
REFERENCES
Barko, J. W., and R. M. Smart. 1980. Mobilization of sediment phos-
phorus by submersed freshwater macrophytes. Freshw. Biol.
10:229-238.
Best, M. D., and K. M. Mantai. 1978. Growth of Myriophyllum:
sediment or lake water as the source of nitrogen and phosphorus.
Ecology 59:1075-1080.
DeMarte, J, A., and R. T. Hartman. 1974. Studies on absorption
of ^P, 59Fe, and •'SCA by water milfoil (Myriophyllum exalbescens
Fernald). Ecology 55:188-194.
Droop, M. R. 1974. Heterotrophy of carbon. Pages 530-559 in
W. D. Stewart, ed. Algal Physiology and Biochemistry. Blackwell
Scientific Publ. Oxford.
Jewell, W. J. 1971. Aquatic weed delay: dissolved oxygen utilization
and nitrogen and phosphorus regeneration. J. Water Pollut. Control
Fed. 43:1457-1467.
Kistritz, R. U. 1978. Recycling of nutrients in an enclosed aquatic
community of decomposing macrophytes (Myriophyllum spkatum).
Oikos 30:475-478.
Landers, D. H. 1979. A durable, reusable enclosure system that com-
pensates for changing water levels. Limnol. Oceanogr. 24:991-994.
1982. Effects of senescing aquatic macrophytes on nutrient
chemistry and chlorophyll a of surrounding waters. Limnol.
Oceanogr. 27:428-439.
McRoy, C.P., and R. J. Barsdate. 1970. Phosphate absorption in
eelgrass. Limnol. Oceanogr. 15:6-13.
Nauwerck, A. 1963. IV. Die Biomasse Gewichts bestimmung. Symb.
Bot. Upsal. 17:93-97.
Nichols, D.S., and D. R. Keeney. 1973. Nitrogen and phosphorus
release from decaying water milfoil. Hydrobiologia 42:509-525.
Pascher, A. 1925. Die Stisswasser-Flora Deutschlands, Osterreichs
und der Schweiz. Cyanophycea. Jena: Gustav Fischer.
Prescott, G. W. 1954. The Freshwater Algae. 2nd ed. Wm. C. Brown
Co., Iowa.
_. 1962. Algae of the Western Great Lakes Area. Rev. ed. Wm.
C. Brown Co., Iowa.
Round, F. E. 1965. The Biology of the Algae. St. Martin's Press,
New York.
Skuja, H. 1948. Taxonomie des phytoplanktons einiger Seen in
Uppland, Schweden. Symb. Bot. Upsal. 9:1-399.
Smith, G. M. 1950. The Fresh-Water Algae of the United States.
McGraw-Hill, New York
Utermohl, H. 1958. Zur Vervollkommnung der quantitativen Phyto-
plankton-methodik. Mitt. Int. Ver. Limnol. 9:1-38.
122
-------
EVALUATION OF INTERNAL PHOSPHORUS LOADING FROM
ANAEROBIC SEDIMENTS
STEVEN B. LAZOFF
Department of Civil Engineering
University of Washington
Seattle, Washington
ABSTRACT
Nutrient-rich hypolimnetic sediments are a potentially significant source of phosphorus loading during anoxic
periods in many Northwest lakes. Several techniques, including calculation of theoretical diffusion rates
from interstitial water concentrations, laboratory simulation of the sediment-water interface, and actual
measurement of hypolimnetic phosphorus buildup during stratification have been used to evaluate anaerobic
sediment phosphorus release in several Northwest lakes. Of these techniques, laboratory simulation studies
appear to give the best estimate of sediment release rates of phosphorus. It was also found that the quan-
tity of phosphorus released from the sediments was not the critical factor affecting lake productivity but
rather the physical and chemical controls on the availability of released phosphorus to the photic zone.
INTRODUCTION
Many lakes in the Pacific Northwest lowlands are monomictic
with stratification typically occurring between late spring and
early to late fall. The formation of stable stratification is often
accompanied by oxygen depletion with a concomitant
buildup in the concentrations of phosphorus and iron in the
hypolimnion. Some of this phosphorus may diffuse across
the thermocline during stratification; however, the bulk of the
hypolimnetic phosphorus remains unavailable to algal growth
until it is mixed into the photic zone during overturn. In-
vestigations of several lakes were undertaken to determine
the contribution of hypolimnetic phosphorus release to the
overall phosphorus budgets and to evaluate the physical and
chemical controls on both the sediment release and availabili-
ty of phosphorus to the photic zone.
Evaluation of sediment release of phosphorus from anoxic
sediments can be done by various methods including in situ
measurements, laboratory simulation studies, theoretical
calculation of diffusion across the mud-water interface, and
mass balance calculations.
Hypolimnetic phosphorus buildup during stratification has
been measured directly in the four study lakes. Although this
can give a reasonable estimate of potentially available
phosphorus input from internal loading, it cannot differen-
tiate between phosphorus released from anaerobic
sediments and that hypolimnetic phosphorus originating from
external inputs or remineralization of settling detritus.
This research has focused on quantifying the components
of hypolimnetic increases in phosphorus, as the relative
magnitudes of these inputs are crucial to evaluating restora-
tion alternatives. For example, alum treatment aimed at seal-
ing off bottom sediments would not be effective when
remineralization of settling detritus accounts for the bulk of
the increases in hypolimnetic phosphorus. An example of
this was provided by the impact of ashfall from the eruption
of Mt. St. Helens on Moses Lake (Welch et al. 1982). Con-
versely, a lake in which release of hypolimnetic phosphorus
by anaerobic sediments is a significant part of the
phosphorus budget may not respond to decreases in exter-
nal loading (Bjork, 1972).
From a lake management perspective, a determination of
the availability of hypolimnetic phosphorus to the photic zone
is also necessary in determining the effect of hypolimnetic
phosphorus buildup on water quality. Phosphorus availability
to phytoplankton growth is a function of both physical and
chemical processes. The physical processes controlling ver-
tical mixing of phosphorus are related to lake morphology
and the stability of the thermocline. Chemical controls are
a function of the rate at which phosphorus is removed from
oxygenated waters by coprecipitation and/or adsorption
relative to the rate of uptake by phytoplankton. The study
lakes—Meridian, Sammamish, Wilderness, and Moses-
provide systems of varying physical and chemical
characteristics where these processes can be evaluated
(Table 1).
METHODS
Anaerobic phosphorus release was evaluated by three
methods: laboratory simulation of the mud-water interface,
theoretical calculations of phosphorus diffusion based on in-
terstitial water concentrations, and in situ measurement.
Laboratory release rates of phosphorus were determined
by incubating sediment cores under anoxic conditions. Sedi-
ment cores were collected from the profundal zones of the
study lakes by either scuba divers or with a 7 kg Phleger
corer. The coring chambers were made of clear tenite
butyrate tubing of 9.57 cm internal cross-sectional area and
30 cm lengths. After collection, the height of the water col-
umns above the sediment was adjusted to 13 cm by allow-
ing the sediment to slide out of or into the tube. The tubes
were sealed airtight and stored in an incubator at lake bot-
tom water temperature in the dark to simulate lake
conditions.
Samples of overlying water were taken from replicate cores
at incubation intervals of 0, 2, 4, 8,16, 32, and 64 days after
collection of the cores. After sampling, the cores were
discarded, thus requiring 14 cores from each lake. Analysis
of filtered (.45 \i Millipore) and unfiltered samples for ortho
and total phosphorus indicated that phosphorus was primarily
in dissolved form. Phosphorus release rates were calculated
from the slopes of the curves of overlying water concentra-
tions versus time.
Interstitial water phosphorus concentrations were deter-
mined on replicate cores collected as described for incuba-
tion studies. The cores were placed in a glove box under
nitrogen atmosphere; the top cm was discarded and slices
from 2 to 4 and 9 to 11 cm were compressed in a chamber
at 90 psi for 30 minutes. The interstitial water was passed
123
-------
Lake Restoration, Protection and Management
Table 1. — Physical and chemical features of Lakes Sammamish, Meridian, Wilderness, and Moses.
Sammamish
Meridian
Wilderness
Moses
Max. depth
(m)
Mean depth
(m)
Volume
(m3)
Surface
area
(m2)
Mean winter
Total P
(M9/0
Mean summer
chla Grtj/l)
31
17
3.28 x 1Q8
1.98 x 1Q7
30
7
27 11.6
12.5 6.4
7.52 x 1Q6 1.73 x 106
6.10 x 1Q5 2.79 x 105
20 40
3 5
10.5
5.6
1.55 x 1Q8
2.75 x 10?
75
20
through a .45 ^ Millipore filter, acidified, and analyzed for
total phosphorus. Theoretical release rates of phosphorus
were calculated from Pick's Law as given by Berner (1975):
dz
Where J = diffusive flux of species i (M/L2T)
D = diffusion coefficient (L2/T)
Q = concentration of species i (M/L3)
Z = depth in sediment (L)
6 = porosity (dimensionless) =
wt of H2O
(wt of H2O + wt of sediment)
The transport of phosphorus across the sediment-water
interface results from a combination of factors including
molecular diffusion, bioturbation, physical mixing, and tor-
dCj
tuosity. The phosphorus concentration gradient __!- was
dz
assumed to be linear over a short distance, 2 cm between
interstitial and overlying phosphorus concentrations.
In Lake Sammamish, sediment release rates were also
determined in situ within columns implanted in the sediments
(McDonnell, 1975).
Sediment traps as described by Birch (1976) were plac-
ed 2 m off the bottom and just below the metalimnion in
Lakes Meridian and Sammamish. The difference in
phosphorus sedimentation between the upper and bottom
traps was used to estimate phosphorus Input to the hypolim-
nion by mineralization of settling detritus. P and Fe sedimen-
tation rates during and following overturn were used to
evaluate removal of phosphorus through adsorption and
coprecipitation with ferric iron compounds.
Phosphorus Budget Calculations
Estimation of yearly phosphorus loading is usually based on
the assumption of a steady state system where inputs are
balanced by outputs. For phosphorus
Inputs
"Aeolian
Fluvial
Non-point Runoff
Sediment Release
[•
Sedimentation
Fluvial
Over shorter time periods or during changes in lake loading,
the steady state assumption is no longer valid so that a term
must be added to account for the changes in lake
phosphorus content, giving:
Inputs = Outputs + A Lake Content
RESULTS AND DISCUSSION
Phosphorus Release From Anoxic
Sediments
Phosphorus release rates from anoxic sediments were deter-
mined in the study lakes by several alternative methods
(Table 2). Release rates determined from laboratory incuba-
tion experiments or in situ measurements were felt to better
represent in-lake phosphorus release than theoretical calcula-
tions based on Pick's law. The theoretical calculations re-
quire a number of simplifying assumptions such as a linear
diffusion path, no adsorption of diffusing ions, and a well-
mixed layer of overlying water.
Theoretical and measured released rates agreed closely
in Lakes Sammamish and Wilderness. In Moses Lake, in-
terstitial phosphorus concentrations were measured in cores
collected in mid-November when anoxic conditions no longer
persisted in the hypolimnion. This may account for the low
release rate calculated by Fick's law.
Table 1 also includes release rates based on the actual
buildup of phosphorus in the hypolimnions. These rates in-
clude phosphorus release from both anoxic sediments and
mineralization of settling detritus and therefore should be
higher than release rates based entirely on diffusion from
anoxic sediments. Lake Wilderness is the only lake studied
that violates this assumption. Lake Wilderness is a shallow
lake and therefore experiences mixing across the ther-
mocline during stratification. This was evidenced by a con-
tinuous phosphorus supply to the epilimnion throughout
stratification and also occasional increases in hypolimnetic
dissolved oxygen during stratification.
In Moses Lake, actual hypolimnetic buildup of phosphorus
indicates a release rate five times greater than that predicted
by diffusion from the sediments alone. This leads to the con-
clusion that remineralization is far more significant than sedi-
ment release as a source of hypolimnetic phosphorus. Given
the highly eutrophic nature of Moses Lake, this is a
reasonable conclusion.
Phosphorus Budgets and Internal vs.
External Loading
Table 3 shows the phosphorus loading to the study lakes
as a result of sediment release and mineralization, as well
as the measured increases in hypolimnetic phosphorus dur-
124
-------
Internal Nutrient Loading Detection and Control
Table 2. — Release rates of phosphorus from anoxic sediments as determined from laboratory incubation, Pick's Law,
hypolimnetic buildup, and in situ sediment-water columns.
Release rates (mgpm~2d~1)
Lake
Meridian (1977)
Sammamish (1974)
Wilderness (1977)
Moses (1978)
Laboratory
studies
1.4
2.8
6.7
Pick's
law
2.5
3.1
2.6
1.7
Hypolimnetic
buildup
2.4
3.8
2.0
35
In situ
studies
3.41
1 McDonnell (1975).
ing stratification. In Lakes Meridian and Sammamish, which
are deep and develop stable thermoclines, hypolimnetic
phosphorus buildup is a reasonable estimation of internal
loading. On the other hand, in Moses Lake and Lake
Wilderness, a larger portion of the bottom sediments lie
above the thermocline resulting in remineralization of settl-
ing detritus within the epilimnion. Also, the thermocline in
these relatively shallow lakes is more easily disrupted by wind
action, resulting in phosphorus transported to the epilimnion
and oxygen to the hypolimnion. The contribution and
availability of released phosphorus by anoxic sediments to
the total lake loading is therefore a much more variable
phenomenon in shallow lakes where thermocline disruption
is a regular occurrence, than in deep lakes with stable
stratification.
In Lakes Meridian and Sammamish, where independent
measures of sediment release and mineralization were
made, a close agreement was found between hypolimnetic
phosphorus buildup and the sum of these components
(Table 3). It was not feasible to estimate mineralization by
using sediment traps in the shallow lakes because of
resuspension of bottom sediments.
Yearly total phosphorus loading and relative inputs from
external and internal sources are given in Table 4. In Lakes
Meridian, Sammamish, and Wilderness total loading was
calculated as the sum of sedimentation plus outflow. In Lake
Sammamish, the estimate of external loading was reasonably
accurate as 70 percent of the external phosphorus load was
through Issaquah Creek; however, in Lakes Meridian and
Wilderness where most of the external loading was from non-
point sources, there was considerably more room for error.
Because the total, internal, and external loading estimates
were arrived at by independent means, total loading is not
equal to internal plus external loading. The difference be-
tween total loading and its components provides a means
for qualitatively assessing the accuracy of the loading
estimates. For these lakes, the differences range between
5 and 25 percent of the total.
In Moses Lake, internal loading was evaluated from a
mass balance of measured inputs, outputs, and changes in
lake content between May and September. Because the in-
ternal loading estimate was determined from the mass
balance, total loading is exactly equal to the sum of the com-
ponents. In Moses Lake, anoxic sediment release was in-
consequential compared to other forms of internal recycl-
ing such as mineralization. In Lakes Meridian, Sammamish
and Wilderness, anoxic sediment release of phosphorus ac-
counted for 25 to 45 percent of yearly loading.
Effect of Internal Phosphorus Loading
on Productivity
In Lake Wilderness and Moses Lake, internally derived
phosphorus was mixed into the trophogenic zone during
periods of wind-induced mixing. Thus, in these shallow lakes,
internally derived phosphorus is periodically made available
to phytoplankton during the summer months when produc-
tivity may be phosphorus limited.
In Lakes Sammamish and Meridian, most of the internal-
ly derived phosphorus remains in the hypolimnion until over-
turn, which occurs in mid-November to mid-December in
both lakes. At overturn, ferrous iron, which has been released
along with phosphorus from the sediments, precipitates as
ferric hydroxides. In Lake Sammamish, iron precipitation at
overturn has been observed as the formation of a reddish
brown floe (Birch, 1976). Phosphorus is removed from the
water column along with these ferric hydroxide particulates
through adsorption and/or coprecipitation and settling.
In Lake Sammamish, the iron particulates are rapidly
sedimented from the water column as evidenced by the large
sedimentation rate of phosphorus at overturn (Birch, 1976).
Birch also estimated that as much as 90 percent of the in-
Table 3. — Hypolimnetic buildup of phosphorus and the estimated contributions of sediment release and mineralization (kg).
Lake
Meridian (1977)
Sammamish (1974)
Wilderness (1977)
Moses (1978)
Hypolimnetic1
buildup
72 (46 ^g/l)
2400 (26 ^g/l)
24 (79 Mg/l)
224 (70 ng/l)
Anoxic
sediment
release
432
23003
342
742
Minerali-
zation"
27
384
Release and
mineralization
70
2684
i Measured during stratified penod.
2 Based on laboratory release rate
3 Based on in situ release rate of 3,4 mgPm~2d~' (McDonnell. 1975)
•> p sedimentation at thermocline - P sedimentation at bottom
125
-------
Lake Restoration, Protection and Management
Table 4. — Yearly phosphorus loading (kg) with external and internal components.
Lake
Meridian (1976)
Sammamish (1974)
Wilderness (1976)
Moses (1977)
Moses (1978)
Total
loading
1701
1 4,700 ' 2
761
44,320"
37,240"
External
loading
58
10.2002
38s
40.0004
37,000"
Anoxic
sediment
release
43s
2.3003
34s
2066
746
Sediment release
plus
remineralization
70
2700
May-Sept.
mass
balance
4320"
240"
1 Calculated as sedimentation {by Pb21D) plus outflow.
2 Data from Birch, 1976.
3 Based on in situ release rate (McDonnell, 1975).
* Patmont, 1980.
5 Dion, 1979.
6 Based on laboratory measured release rates.
ternally derived phosphorus was removed from the water col-
umn by this mechanism and therefore unavailable for
phytoplankton growth. Neither nitrogen nor silica were found
to limit phytoplankton growth at overturn (Birch, 1976; Lazoff,
1980).
In Lake Meridian, ferrous iron reached a concentration of
over 2 mg/l in the hypolimnion by the end of stratification.
Although iron precipitation did occur at overturn, only about
50 percent of the internally derived phosphorus was
recovered in the sediment traps. Overturn also stimulated
a phytoplankton bloom in Lake Meridian during the study
year (Davis et al. 1978).
Although Lakes Meridian and Sammamish were similar
chemically (phosphorus limited), phosphorus mixed into the
water column at overturn had little effect on productivity in
Lake Sammamish, yet stimulated a bloom in Lake Meridian.
One possible explanation is that the morphological dif-
ferences (depth/volume) between the lakes results in a
deeper mixing zone in Lake Sammamish. Light limitation in
Lake Sammamish may therefore prevent uptake of
phosphorus by phytoplankton at overturn, thus permitting
effective removal by ferric iron. The observed increases in
productivity following wind-driven mixing across the ther-
mpcline in Lake Wilderness also indicate that phytoplankton
can use phosphorus in the presence of precipitating iron
since both iron and phosphorus would be mixed into the
epilimnion.
SUMMARY
1. Several methods were used to determine phosphorus
release from anaerobic sediments. Sediment traps were used
to evaluate phosphorus input to the hypolimnion through
mineralization during stratification and sedimentation of
phosphorus with iron compounds following overturn. These
methods proved valuable in gaining an understanding of the
phosphorus dynamics in the study lakes.
2. Phosphorus release from anoxic sediments was found
to account for up to 45 percent of the total annual loading
in the study lakes.
3. In the shallow lakes, Moses and Wilderness, wind-
induced mixing during summer stratification was found to
stimulate phytoplankton growth. In Lake Wilderness, this was
attributable largely to the mixing of hypolimnetic phosphorus
into the photic zone.
4. In the deeper lakes, Sammamish and Meridian, ther-
mocline stability throughout stratification kept the bulk of the
sediment-released phosphorus from reaching the
trophogenic zone until overturn.
5. In Lake Sammamish, up to 90 percent of the released
phosphorus is rapidly removed from the water column by
adsorption and/or coprecipitation with ferric hydroxide com-
pounds and did not increase algal productivity.
6. In Lake Meridian, although hypolimnetic iron concen-
trations are greater than in Lake Sammamish, overturn
resulted in a plankton bloom during the study year.
7. It was hypothesized that the morphologic differences
between the two lakes which result in a shallower mixing
depth in Lake Meridian may account for the difference in
response between the two lakes. At overturn, light limitation
may be the controlling factor limiting phytoplankton growth
in Lake Sammamish.
REFERENCES
Berner, R. A. 1975. Diagenetic models of dissolved species in the
interstitial waters of compacting sediments. Am. J.Sci. 275:88-96.
Birch, P. B. 1976. The relationship of sedimentation and nutrient
cycling to the trophic status of four lakes in the Lake Washington
drainage. Ph.D. Dissertation. Univ. Washington, Seattle.
Bjork, S. et al. 1972. Ecosystem studies in connection with the
restoration of lakes. Verh. Int. Verein. Limnol. 18:379-87.
Davis, J. et al. 1978. A study of the trophic status and recom-
mendations for the management of Lake Meridian. Rep. Munic.
Metro. Seattle (METRO).
Dion, N. P. 1979. Environmental features, general hydrology, and
external sources of nutrients affecting Wilderness Lake, King
County, Wash. U.S. Geolog. Surv. Water-Resour. Invest. Open-
File rep. 79-63.
Lazoff, S. B. 1980. Deposition of diatoms and biogenic silica as
indicators of Lake Sammamish productivity. M. S. Thesis. Univ.
Washington, Seattle.
McDonnell, J. C. 1975. In situ phosphorus release rates from
anaerobic sediments. M. S. Thesis. Univ. Washington, Seattle.
Patmont, C. F. 1980. Phytoplankton and nutrient responses to dilu-
tion in Moses Lake. M. S. Thesis. Univ. Washington, Seattle.
Welch, E. B., M. D. Tomasek, and S. B. Lazoff. 1982. Volcanic
ashlayer effect on lake internal phosphorus loading. In Proc. Conf.
on Mt. St. Helens: Effect on Water Resources. Wash. Water Res.
Center.
126
-------
ESTIMATION OF INTERNAL NUTRIENT LOADING IN LAGUNA LAKE
G. C. HOLDREN, JR.
Water Resources Laboratory
University of Louisville
Louisville, Kentucky
ABSTRACT
Estimates of internal nutrient loading were made for Laguna Lake, a large, shallow eutrophic lake near
Manila, Philippines, to provide a basis for lake management decisions. Four methods have been used
in other cases to assess internal loading: evaluation of nutrient mass balance data, estimation of diffu-
sional nutrient release from interstitial water data, laboratory nutrient release experiments, and in situ nutrient
release studies. The mass balance approach was unfeasible and results of an in situ experiment were
inconclusive. Calculated diffusional release rates were 0.62 mg P/m2/day and 7.6 mg N/m2/day. Experimental
release rates for P in laboratory systems averaged 2.2 mg/m2/day and were dependent on both mixing
and oxygen concentrations in the overlying water. Nitrogen release rates averaged 9.9 mg/m2/day and
were independent of mixing and dissolved oxygen concentrations. Both methods may overestimate ac-
tual internal loading rates but the results indicate internal nutrient loading in Laguna Lake is likely to be
significant and may exceed external loading under certain conditions.
INTRODUCTION
Laguna Lake is a large, shallow eutrophic lake located south
and east of Manila, Philippines. Extensive limnological in-
vestigations have been conducted (SOGREAH, 1974;
Lenarz, 1978a) to provide data on its water quality. This in-
formation is being used by the Laguna Lake Development
Authority (LLDA) as the basis for lake management and plan-
ning decisions. Some important limnological characteristics
of Laguna Lake are listed in Table 1.
Table 1. — Morphometric and hydrologlc characteristics of
Laguna Lake.1
Drainage basin area
Lake surface area2
Lake volume2
Mean depth2
Average discharge
Hydraulic residence time
Estimated N loading
Estimated P loading
3.82 x 103 km2
900 km2
3.26 x 109 m3
3.6 m
3.4 * 109 m3
0.96 yr
4.0 g/m2/yr
0.87 g/m2/yr
1 From SOGREAH (1974) and Lenarz (1978).
' At annual mean lake level.
Laguna Lake is of economic importance at the present
time largely because of an extensive fishery based on the
fish pen industry. The latest survey (Lenarz, 1978b) found
2,600 hectares of fish pens in operation with an annual yield
of three to five tons of fish per hectare. Extensive algal
blooms (predominantly Mycrocystis) occurring in Laguna
Lake have been linked with periodic fish kill incidents in the
fish pens, resulting in economic losses for the industry.
In the future, Laguna Lake may serve as a source of ir-
rigation water and as a water supply for the metropolitan
Manila area. These beneficial uses are excluded at the pre-
sent time by the intrusion of seawater from Manila Bay
through the lake outlet during the dry season, which prevents
using lake water for drinking or irrigation, and by the algal
blooms, which would greatly increase water treatment costs.
A hydraulic control structure, now nearing completion, was
designed to provide flood control and prevent seawater in-
trusion. It will also eliminate a major source of nutrients. A
sewage system is planned to further limit nutrient inputs to
the lake and reduce the frequency of algal blooms, although
previous reports (SOGREAH, 1974; Lenarz, 1978a) would
indicate significant nonpoint nutrient loading would still occur.
Existing studies of Laguna Lake did not fully evaluate in-
ternal nutrient loading, although internal loading may be
significant because the lake is shallow and sediment are
subjected to wave action. This investigation was designed
to assess the potential significance of internal loading on the
nutrient budget of Laguna Lake.
Most methods used during previous investigations of in-
ternal nutrient loading can be placed into four general
categories: mass balances, diffusion models, laboratory
studies, and in situ experiments. The nutrient mass balance
approach (Burns and Ross, 1972; Cooke et al. 1977; White
et al. 1978; Larsen et al. 1979) has been used where exten-
sive monitoring data are available and external nutrient
sources and sinks can be quantified.
Several estimates of nutrient release have been made us-
ing diffusion models and interstitial water nutrient concen-
trations (Tessenow, 1972; Bannerman et al. 1974; Kamp-
Nielson, 1974; Holdren et al. 1977; Theis and McCabe,
1978). This approach requires a minimal amount of data,
but applications may be limited to undisturbed sediments.
A variety of laboratory studies have been conducted (Lee
et al. 1977; Gallepp, 1979; Frevert, 1980; Holdren and Arm-
strong, 1980 (and references cited therein)) involving the in-
cubation of sediments under conditions chosen to simulate
natural conditions. Both intact sediment cores and dredged
sediments have been used for these experiments, with
Hargrave (1975) indicating results were similar for mixed
systems and those using undisturbed cores.
Finally, different types of artificial enclosures have been
used to measure nutrient release in situ (Lack and Lund,
1974; Rowe et al. 1975; Sonzogni et al. 1977; Freedman
and Canale, 1977). Such enclosures allow a mass balance
to be made on relatively simple and well-delineated systems.
METHODS
Nutrients can enter Laguna Lake through 21 main streams
and 32 small or intermittent streams (SOGREAH, 1974), as
127
-------
Lake Restoration, Protection and Management
50 5
Scale KM
Laguna
Lake
Figure 1.—Laguna Lake sampling locations. Roman numerals refer to sites selected by Lenarz (1978a).
well as from numerous nonpoint sources and direct precipita-
tion on the lake surface. As a result, the mass balance ap-
proach could not be used to estimate internal nutrient
loading.
The concentration gradient between interstitial and overly-
ing waters was used to estimate diffusional transport of
nutrients from the sediments following the approach of Ban-
nerman et al. (1974). Sediment cores or Ekman dredge
samples were collected from six locations in Laguna Lake
(Fig. 1). Sampling locations included five sites selected from
extensive water analyses by Lenarz (1978a) and the LLDA
experimental fish pen at Looc.
Interstitial water was obtained using the centrifugation-
filtration technique (Holdren et al. 1977). The 0 to 3 cm layer
of sediment was collected in plastic bags, mixed briefly, and
transferred to glass centrifuge tubes. A smaller depth inter-
val may have produced better estimates of interstitial nutrient
concentrations near the interface, but the small diameter of
the core tubes (3.5 cm i.d.) and large centrifuge tube volume
(125 ml) necessitated use of a large interval to minimize sam-
ple handling. Centrifuge tubes were kept tightly stoppered
and cooled until returned to the laboratory. Interstitial water
was collected by centrifugation followed by pressure filtra-
tion of the supernatant under N2. Refrigeration during the
centrifugation process was not required because of the
relatively high (31 °C) sediment temperatures.
Interstitial reactive phosphorus (IRP) and interstitial am-
monia (NH3-N) concentrations were measured col-
orimetrically. Laguna Lake sediments were anaerobic below
the interface so all inorganic nitrogen was present as NH3-N.
Bottom water samples were also analyzed for soluble reac-
tive phosphorus (SRP) and NH3-N concentrations to deter-
mine the concentration gradient betweeen interstitial and
overlying waters.
Bulk sediment samples were collected with an Ekman
dredge and a plastic scoop was used to collect approximately
the top 3 cm for laboratory nutrient release experiments.
Sediments were stored in plastic bags and refrigerated un-
til needed. Incubation systems consisted of 2.5 I acid bot-
tles that had been acid-washed and then soaked in a
phosphate solution (Holdren and Armstrong, 1980) to pre-
vent adsorption of nutrients by the container walls.
Preliminary experiments indicated P was adsorbed on un-
treated bottles, but no significant adsorption or release of
anypf the nutrients inves'rtgated (SRP, NH3N, NOJ - N, and
NOi" - N) occurred in treated bottles.
Approximately 4 cm of sediment were added to each bottle
and covered with 0.45 urn filtered water; suspended
sediments were then allowed to settle overnight. Following
this initial equilibration, the overlying water was removed and
bottles were filled with fresh water to initiate the experiments.
Experimental systems were incubated at room temperature
(28 to 30°C) and kept in the dark to minimize algal growth.
Two experiments were conducted. In the first experiment,
sediment from Station I was covered with filtered lake sur-
face water. One bottle was aerated gently to mix the water
column without sediment suspension, while the other was
aerated vigorously enough to cause sediment suspension,
because previous results (Lenarz, 1978a) indicated algal
blooms occurred only when lake water turbidity was less than
128
-------
Internal Nutrient Loading Detection and Control
40 g SiO2/m3. Because nutrient concentrations in the lake
water were elevated due to seawater intrusion and results
were somewhat erratic, a second experiment was conducted
with sediment from Station IV and aged tap water. In addi-
tion to two aerated bottles, a third bottle was purged with
N2 to reduce oxygen concentrations in the overlying water.
Water samples from the laboratory incubation experiments
were analyzed for SRP, total P, NH3-N, NO"3-N, and
NO2-N) in addition to turbidity, pH, and dissolved oxygen
(D.O). Release rates for SRP and inorganic N, taken as the
sum of NH3-N, NO~3-N, and NO~2-N) concentrations, were
calculated from a mass balance on the nutrients.
A plastic enclosure was erected in the LLDA's experimen-
tal fish pen at Looc to investigate sediment nutrient release
in situ. The enclosure was constructed from a sheet of heavy
plastic approximately 3 m x 25 m attached to a square bam-
boo frame approximately 5 m on each side. The plastic was
anchored with large rocks along the bottom edge. Water
depth was initially 1 .5 m but heavy rainfall during the course
of the experiment increased it to 2 m. Water both inside and
outside the enclosure was monitored for pH, temperature,
D.O., turbidity, chlorophyll, and nutrient concentrations.
RESULTS
Calculation of Diffusional Nutrient Release
The results of the interstitial water analyses are summar-
ized in Table 2. Concentrations of SRP in the overlying water
were less than 0.01 g/m3 at all sampling locations and, ex-
cept for a concentration of 0.07 g/m3 at Station VII, NH3~N
concentrations were also less than 0.01 g/m3. The large
observed concentration gradient between interstitial and
overlying waters indicates mixing does not occur to a signifi-
cant depth in the sediments, but a large potential for diffu-
sional release of nutrients from the sediments does exist.
The diffusional flux of nutrients from the sediments was
calculated using Pick's first law of diffusion (Berner, 1980):
where Js = flux across the sediment-water interface
(mg/cm2/sec),
$ = sediment porosity=volume of interconnected
water/total sediment volume,
Ds = molecular diffusion coefficient in the sedi-
ment (cm2/sec), including effects of tortuosi-
ty, and
o r^
—^ = the concentration gradient across the
sediment-water interface (mg/cm3/cm).
3x
An estimated value of = 0.98 was obtained from the
average water content of 81.3 percent and a sediment den-
sity of 1.2 g/ml for Laguna Lake sediments. The diffusion
coefficients used in the calculations were 1.0*10~6
cm2/sec for phosphorus (Stumm and Leckie, 1971) and
3xio~6 cm2/sec for ammonia (Imboden and Lerman,
1978). The average interstitial ammonia and IRP concen-
trations for the 0 to 3 cm sediment layer (Table 2) and overly-
ing water NH3-N and SRP concentrations were used to
calculate the concentration gradient. Using these values, dif-
fusional nutrient fluxes of 7.6 mg/m2/day and 0.62 mg/m2/day
were calculated for nitrogen and phosphorus, respectively.
The nutrient fluxes calculated in this manner may
overestimate actual diffusional release of nutrients from
Laguna Lake sediments. Both interstitial ammonia and IRP
concentrations increase with depth below the interface.
Therefore, it is likely that interstitial nutrient concentrations
measured for the 0 to 3 cm sediment layer overestimate the
actual concentration gradients at the interface. No oxidized
surface layer was discernible, however, and the viscosity of
the surface sediments would indicate oxygen penetration into
the sediments may be limited.
Interstitial nutrient concentrations are known to increase
with temperature and decrease with increased mixing in the
overlying water (Holdren et al. 1977). As a result, the in-
terstitial nutrient concentrations measured during this in-
vestigation, at a time when water temperatures are near max-
imum annual levels and average wind velocities are relatively
low (Lenarz, 1978a), probably represent the maximum con-
centrations that may be expected for Laguna Lake. This in-
dicates the calculated diffusional fluxes would also repre-
sent maximum annual rates, although temporal variations
in interstitial nutrient concentrations are not expected to be
as large as those occurring in temperate zone lakes because
of the high sediment temperatures encountered throughout
the year.
Laboratory Nutrient Release Measurements
Nutrient release from Laguna Lake sediments was measured
in the laboratory by monitoring nutrient concentrations in ex-
perimental systems as a function of time. Experiments were
allowed to continue until nutrient concentrations in the overly-
Table 2. — Interstitial water nutrient concentrations.
Sampling
date
20 May 1981
21 May 1981
26 May 1981
18 June 1981
Sampling
station
I
Looc2
I
II
III
IV
VII
Looc
Mean =
Interstitial
IRP
1.15
1.20
0.63
1.32
—
0.89
1.06
1.33
1.08
nutrient concentration1
NH3-N
. .
7.96
3.13
7.33
2.60
—
3.43
3.67
3.68
4.54
Water
content
%
85.2
77.6
83.4
79.9
78.7
82.1
82.3
—
81.3
' Fof 0 lo 3 cm sediment layer
1 LLDA experimental fish pen. Compartment 3.
129
-------
Lake Restoration, Protection and Management
ing water reached apparent equilibrium. During the first ex-
periment, SRP concentrations fluctuated before reaching an
apparent equilibrium level after 9 days, while inorganic N
concentrations reached maximum levels after 2 days, then
dropped significantly and remained relatively constant for the
duration of the experiment. Both SRP and inorganic N reach-
ed equilibrium levels after 9 days in the second experiment.
Changes in inorganic N and SRP concentrations with time
during the second experiment are plotted in Figure 2. Ex-
pected diffusional release rates, based on concentration gra-
dients between interstitial nutrient concentrations at Station
IV and overlying water nutrient concentrations in the ex-
perimental systems, are also plotted for comparison.
Table 3 summarizes the results of the laboratory nutrient
release experiments. All nutrient release rates were
on
I
•2*
c
o
u
c
o
U
Incubation Time (days)
Figure 2,—Nutrient release during laboratory incubation experiments
using sediments from Station IV and aged tap water.
calculated over the 9-day incubation period required for most
of the nutrient concentrations to reach equilibrium and,
therefore, do not include the large initial release of inorganic
N noted in the first experiment. As expected, release rates
for both inorganic N and SRP are greater than calculated
diffusional release rates. Surprisingly, the increased release
rate for SRP is proportionally greater than that for inorganic
N, despite high sediment iron concentrations.
The release of inorganic N was independent of mixing and
dissolved oxygen concentrations in the overlying water. In
all cases a large increase in ammonia concentration in the
water column occurred during the first few days. Following
the initial release, inorganic N in the water column was con-
verted almost entirely to nitrate, except in the N^-purged bot-
tle where a second ammonia peak and some measurable
nitrite concentrations were detected. Ammonia concentra-
tions remained low and nitrate concentrations remained con-
stant or decreased slightly as the experiments continued.
These results indicated inorganic N was released from the
sediments as ammonia and then oxidized to nitrate in the
aerated overlying water. The observed loss of inorganic N
following the initial release during the first experiment may
have been the result of denitrification, which can represent
a significant sink for inorganic N (Rowe et at. 1975; White
et al. 1978).
The release of SRP in laboratory systems depended on
both turbidity and dissolved oxygen concentrations in the
overlying water. Release rates for total P were similar to those
observed for SRP when turbidity was low, but much higher,
because of sediment suspension, when turbidity increased.
In the first experiment, SRP was removed from the water
column in the turbid bottle but released from the sediments
in the non-turbid bottle. Release rates for SRP during the
second experiment were similar for both bottles, but it should
be noted that sediments were kept in constant suspension
in the turbid bottle. A previous study (Holdren and Armstrong,
1980) indicated losses of SRP from the water column were
associated with the settling of suspended sediments, mak-
ing the amount of sediment suspension a primary factor in
calculated SRP release rates. The generally low SRP and
high total P concentrations in the water column of Laguna
Lake (SOGREAH, 1974) may be an indication of the control
of SRP levels by suspended sediments.
Although the final dissolved oxygen concentration was a
relatively high 3.1 g/m3, the release rate for SRP in the
N2-purged bottle was approximately twice that of release
rates observed in the aerated bottles. These results are in
accord with previous studies and may be a consequence
Table 3. — Nutrient release during laboratory incubation experiments.
Treatment conditions
Nutrient release rate1
Inorganic N
Soluble reactive P
Aeration,
final turbidity = 0.5 g SiO2/m3!
Aeration,
final turbidity = 64 g SiCVm3*
Aeration, final turbidity = 9 g SiO^m3
final D.O. cone = 7.3 glm3"1
Aeration, final turbidity = 52 g SiOz/m3
final D.O. cone. = 7.2 g/m3*
N2, final turbidity = 7 g SiOs/m3
final D.O. cone = 3.1 g/m3'
„ ff|g,
1.3
2.0
13
16
17
riwoay
4.2
-1.3
2.6
2.9
6.1
1 Calculated over &
-------
Internal Nutrient Loading Detection and Control
of the high iron concentration (50 mg/g) in Laguna Lake
sediments (SOGREAH, 1974), but the results also indicate
that iron is not completely effective in trapping phosphorus
at the oxidized sediment-water interface.
Laboratory nutrient release experiments such as this may
overestimate nutrient release occurring under natural con-
ditions because nutrients are transported over short
distances, and processes such as sediment deposition that
transport nutrients to the sediments in natural systems do
not occur. Such results would, therefore, represent gross
nutrient release rates rather than the net changes resulting
from interactions between sediments and overlying water oc-
curring in natural systems. These results do indicate,
however, the potential of the sediments to release nutrients
under the appropriate environmental conditions and would
indicate that a significant amount of nutrients could be releas-
ed from Laguna Lake sediments.
Results of the In Situ Experiment
The most noticeable change in water quality during the in
situ experiment was a change in chlorophyll concentrations.
Although no initial chlorophyll measurements were taken,
chlorophyll was analyzed beginning on the ninth day of the
experiment when the development of an algal bloom inside
the plastic enclosure caused the water to be noticeably
greener than the adjacent lake water. For the duration of
the experiment, chlorophyll concentrations within the
enclosure (Fig. 3} ranged from 87 to 111 g/m3 while surroun-
ding lake water chlorophyll concentrations averaged 55 g/m3.
The reasons for the algal bloom within the enclosure are
not clear. Because water inside the enclosure was noticeably
calmer than surrounding lake water, it is possible that settl-
ing of inorganic turbidity may have increased light penetra-
tion and triggered the algal bloom. Turbidities outside the
enclosure and initial turbidities inside the enclosure were
always less than 20 g SiO2/m3, however, far less than the
40 g/m3 cited by Lenarz (I978a) as the level inhibiting algal
growth in Laguna Lake.
The release of soluble nutrients from the lake sediments,
coupled with a lack of sediment suspension, also may have
contributed to the observed results. It can be seen (Fig. 3)
that SRP concentrations increased slightly throughout the
experiment and that inorganic N concentrations exhibited
an initial increase resulting from a large increase in ammonia
concentration. The ammonia concentration dropped below
detection limits after two weeks as the algal bloom raised
the pH to 9.2 and dissolved oxygen concentration to 10.4
g/m3 inside the enclosure, compared to a pH of 8.2 and
20
Elapsed Time (days)
dissolved oxygen concentration of 7.1 g/m3 on the outside.
Overall, the concentration of inorganic N decreased by 40
percent, probably as a result of algal uptake of nitrate.
A complete mass balance of nutrients within the enclosure
used for the in situ experiment could not be calculated,
preventing a quantitative assessment of nutrient uptake or
release by the sediments. Nutrient concentrations in the at-
tached algae growing on the enclosure walls could not be
measured and sedimentation was not assessed. In addition,
water depth increased from 1.5 m to almost 2 m because
heavy rainfall caused an increase in lake level of 10 to 15
cm per week during the course of the experiment. Dilution
of initial concentrations caused by the 20 to 30 percent in-
crease in water volume or losses by sedimentation and
nutrient uptake by attached algae could account for the
overall decrease of 35 percent for total P and 25 percent
for total N noted during the experiment.
Although sediment nutrient release could not be quantified
during the in situ experiment, the results did indicate that
current cultural practices used by the fish pen industry may
contribute to the fish kill incidents. The use of double net-
ting or fine mesh netting that is easily obstructed by attach-
ed algae impedes water circulation and can produce condi-
tions in the fish pens similar to those found in the plastic
enclosure. The lack of water movement can decrease in-
organic turbidity and/or increase net nutrient release from
the sediments. The resulting increased algal populations,
coupled with high stocking densities, can result in the ox-
ygen depletion noted in the fish pens during the fish kill
incidents.
Significance of Sediment Nutrient
Release in Laguna Lake
The significance of internal nutrient loading in Laguna Lake
was assessed by comparing experimental results with ex-
isting nutrient loading models. Acceptable nutrient loadings
for phosphorus were calculated using the models of Dillon
and Rigler (1974):
(2)
(3)
and Vollenweider (1976):
Uzv = [P] (1 +4)
Figure 3.—Changes in nutrient and chlorophyll concentrations dur-
ing the in situ experiment.
where [P] = acceptable whole lake phosphorus concen-
tration (mg/m3),
L = areal phosphorus loading rate (mg/m2/yr),
R = phosphorus retention coefficient
cp = hydraulic flushing rate (yr~1), and
z = mean depth (m)
The acceptable phosphorus loading rate was calculated us-
ing the parameters in Table 1, an acceptable phosphorus
concentration of 10 mg/m3 (Vollenweider, 1968) and the
value for R of 0.5 estimated by Lenarz (1978a). The
dangerous loading rate for phosphorus was calculated us-
ing a phosphorus concentration of 20 mg/m3 and acceptable
and "dangerous" loading rates for nitrogen were calculated
assuming a 15:1 nitrogen to phosphorus ratio (Vollenweider,
1968). Both models were found to yield essentially the same
results, with acceptable nutrient loading rates calculated to
be 0.22 g P/m%fay and 3.3 g N/m2/day.
Table 4 compares nutrient loading rates calculated from
nutrient loading models, estimated external nutrient loading
rates calculated from the data of Lenarz (1978a), and sedi-
ment nutrient release rates obtained during this investiga-
tion. These results indicate that external nutrient loadings
131
-------
Lake Restoration, Protection and Management
Table 4. — Significance of internal nutrient loading in Laguna Lake.
Loading type
Loading rate
Maximum Acceptable1
"Dangerous"1
External2
Calculated diffusional release
Experimental release, laboratory system
-mg/mz/day—•
0.22
0,44
0.87
0.62
2.2
3.3
6.6
4.0
7.6
9.9
1 Calculated from Dillon and Rigler (1974) and Vollenweider (1976).
'From Lenarz (1978).
currently exceed acceptable limits and that internal loading
is likely to be significant.
Calculated diffusional nutrient release rates for both
nitrogen and phosphorus were greater than the "dangerous"
loading rates calculated from the loading models. Diffusional
release rates are especially significant when compared to
estimated external loading. Diffusional phosphorus release
was 71 percent of the external load and diffusional nitrogen
release was almost two times the external load. Experimental
release rates were even greater. For both nitrogen and
phosphorus, experimental loadings were two-and-a-half times
the external loading. Although the gross sediment nutrient
release rates calculated during this investigation prob-
ably overestimate actual release rates, these results clearly
indicate the potential significance of internal nutrient loading
for Laguna Lake.
CONCLUSIONS
At the present time, mass balance calculations provide the
only method capable of estimating net nutrient release from
lake sediments; however, in most cases sufficient data do
not exist to permit using this approach. Other methods pro-
vide estimates of gross nutrient release from the sediment
that may still be useful in lake management decisions.
Best results from diffusion models could be obtained by
using the smallest possible sediment depth interval for
measuring interstitial nutrient concentrations and concentra-
tion gradients. Spatial and temporal variations in interstitial
nutrient concentrations should be considered. Laboratory
measurements of nutrient release should encompass the ex-
pected natural range of temperature, mixing, and dissolved
oxygen concentrations. Careful use of these procedures
should enable researchers to put upper limits on expected
internal nutrient loading rates. In many cases this informa-
tion could be sufficient to determine the effectiveness of lake
restoration measures.
The results of this investigation indicated internal nutrient
loading in Laguna Lake is likely to be significant and may
actually exceed external loading under certain conditions.
While proposed measures to control sewage inputs to
Laguna Lake are desirable and may reduce the severity of
algal blooms, costly nutrient removal practices do not ap-
pear to be justified because internal nutrient loading could
continue to cause algal blooms even in the absence of ex-
ternal nutrient sources.
ACKNOWLEDGEMENTS: Funding for this research was provided
by the World Health Organization/ U.N. Development Program. The
support and guidance of T. Baguilat, general manager, LLDA; J.
Centano, manager, Environmental Protection Division; B. Adan,
LLOA consultant; and E. Lee, regional advisor in Environmental
Health, LLDA proved invaluable during this study. Special
acknowledgement is due to R. Manto, chief biologist, limnologist;
Z, Villefuerte, chief of Laboratory; M. de Guzman, chief chemist,
and the entire EPD staff for their advice and assistance.
REFERENCES
Bannerman, R, T., D, E. Armstrong, G. C. Holdren, and R. F. Harris.
1974. Phosphorus mobility in Lake Ontario sediments. Proc. 17th
Conf. Great Lakes Res.: 158-178.
Berner, R. A. 1980. Early Diagenesis: A Theoretical Approach,
Princeton Univ. Press, Princeton, N,J.
Burns, N. M., and C. Ross. 1972, Oxygen-nutrient relationships
within the central basin of Lake Erie. Pages 193-250 in H.E. Allen
and J,R. Kramer, eds. Nutrients in Natural Waters. Wiley-
Interscience, New York.
Cooke, G, D., M. R. McComas, D, W, Waller, and R. H. Kennedy.
1977. The occurrence of internal phosphorus loading in two small,
eutrophic, glacial lakes in northeastern Ohio. Hydrobiologia
56:129-135.
Dillon, P. J. and F. H. Rigler. 1974. The phosphorus-chlorophyll
relationship in lakes. Limnol. Oceanogr. 19:767-773.
Freedman, P. L. and R. P. Canale. 1977. Nutrient release from
anaerobic sediments. J. Environ. Eng. Div. Am. Soc, Civil Eng.
103:233-244.
Frevert, T. 1980. Dissolved oxygen dependent phosphorus release
from profundal sediments of Lake Constance (Obersee).
Hydrobiologia 75:17-28,
Gallepp, G. W, 1979. Chironomid influence on phosphorus release
in sediment-water microcosms. Ecology 60:547-556.
Hargrave, B. T, 1975. Stability in structure and function of the mud-
water interface. Verh. Int. Verein. Limnol. 19:1073-1079,
Holdren, G. C. Jr., D. E. Armstrong, and R. F. Harris. 1977. Inter-
stitial inorganic phosphorus concentrations in Lakes Mendota and
Wingra. Water Res. 11:1041-1047.
Holdren, G. C. Jr., and D. E, Armstrong, 1980, Factors affecting
phosphorus release from intact lake sediment cores. Environ. Sci.
Technol. 14:79-87.
Imboden, D. M. and A. Lerman. 1978. Chemical models of lakes.
Pages 341-356 in A. Lerman, ed. Lakes: Chemistry, Geology,
Physics. Springer-Verlag, New York.
Kamp-Nielson, L, 1974. Mud-water exchange of phosphate in
undisturbed sediment cores and factors affecting the exchange
rates. Arch, Hydrobiol. 73:218-237.
Lack, T. J. and J. W, G. Lund. 1974. Observations and experiments
on the phytoplankton of Blelham Tarn, English Lake District
Freshw, Biol. 4:399-415.
Larsen, D. P., J. Van Sickle, K. W. Malueg, and P. D, Smith. 1979
The effect of wastewater phosphorus removal on Shagawa Lake
Minnesota: Phosphorus supplies, lake phosphorus and chlorophyll
a. Water Res. 13:1259-1272, *
132
-------
Internal Nutrient Loading Detection and Control
Lee, G. F., W. C. Sonzogni, and R. D. Spear. 1977. Significance
of oxic vs. anoxic conditions for Lake Mendota sediment. In H.
L. Golterman, ed. Interaction Between Sediments and Fresh
Water. Dr. W. Junk B. V. Publishers, The Hague.
Lenarz, H. 1978a. Comprehensive water quality management
programme, Laguna de Bay. Vol. 4-7. Limnology of Laguna de
Bay. Rep. submitted to the World Health Organ.
1978b. Comprehensive water quality management pro-
gramme, Laguna de Bay. Vol. 8. Fisheries of Laguna de Bay. Rep.
submitted to the World Health Organ.
Rowe, G. T., C. H. Clifford, K. L. Smith, Jr., and P. L, Hamilton.
1975. Benthic nutrient regeneration and its coupling to primary
productivity in coastal waters. Nature 255:215-217.
SOGREAH. 1974. Laguna de Bay, water resources development
study, Vol. 1-3. Rep. submitted to the U. N. Develop. Progr. and
Asian Develop. Bank.
Sonzongi, W. C., D. P. Larsen, K. W. Malueg, and M. D. Schuldt.
1977. Use of large submerged chambers to measure sediment-
water interactions. Water Res. 11:461-464.
Stumm, W., and J. O. Leckie. 1971. Phosphate exchange with
sediments: Its role in the productivity of surface waters. Proc. 5th
Int. Water Pollut. Res. Conf.
Tessenow, U. 1972. Solution, diffusion and sorption in the upper
layer of lake sediments. I. A long-term experiment under aerobic
and anaerobic conditions in a steady-state system. Arch. Hydrobiol.
Suppl. 38:353-398.
Theis, T. L., and P. J. McCabe. 1978. Phosphorus dynamics in
hypereutrophic lake sediments. Water Res. 12:677-685.
Vollenweider, R. A. 1968. Scientific fundamentals of the eutrophi-
cation of lakes and flowing waters, with particular reference to
nitrogen and phosphorus as factors in eutrophication. Organ. Econ.
Dev. Rep. DAS/CS1768.27.
. 1976. Advances in defining critical loading levels for
phosphorus in lake eutrophication. Mem. 1st. Ital. Idrobiol. 33:53-83.
White, E., et al. 1978. Sediment of Lake Rotorua as sources and
sinks for plant nutrients. N. Z. J. Mar. Freshw. Res. 12:121-130.
133
-------
Fiscal and Institutional
Support for Lakes
THE STATE OF WASHINGTON LAKE RESTORATION PROGRAM:
SOME GOOD NEWS—SOME BAD NEWS
ROLAND E. PINE
Coordinator
Lake Restoration and Aquatic Plant Management Program
Washington State Department of Ecology
Olympia, Washington
RAYMOND A. SOLTERO
Department of Biology
Eastern Washington University
Cheney, Washington
RICHARD D. RILEY
Gibbs and Olson, Inc.
Longview, Washington
ABSTRACT
The State of Washington Lake Restoration Program is alive and well in spite of the demise of the Federal
Clean Lake Program under Section 314 of the Clean Water Act. The Washington effort has succeeded
primarily because of the interest and concern of the general public demonstrated by the passage of two
referenda, one in 1972 and the other in 1980, which provided $20 million and $35 million, respectively,
for lake restoration. Since the program began, approximately $15.4 million of State funds and $18.3 million
of Federal funds have been spent or obligated for the restoration of 21 Washington lakes. Two of these
projects are discussed in relative detail indicating the types of restoration techniques used and their suc-
cess, as well as some of the problems and unanticipated difficulties encountered. The Washington ex-
perience indicates that more than 1 year of post-project monitoring is required to measure success and
to identfy post-restoration problems before they become acute and irreversible. A minimum of 2 and possibly
3 years of post-restoration monitoring is suggested. Public interest in the Washington Lake Restoration
Program at the grass roots level is substantial. It is anticipated this interest will increase as local people
become more aware of the kind of funding program available.
HISTORICAL BACKGROUND
The State of Washington Lake Restoration and the Federal ment of Ecology (WDOE), began in November 1972 with
Clean Lake programs began almost simultaneoulsy in 1972. passage by the voters of Referendum 26. The Federal pro-
The State program, administered by the Washington Depart- gram was initiated under authority of Section 314 of the
135
-------
Lake Restoration, Protection and Management
Clean Water Act (P.L. 92-500), originally passed by the 95th
Congress in 1972.
Referendum 26 allowed for regulation authorizing the State
Finance Committee to issue general obligation bonds to
generate $225 million to aid local agencies in providing im-
provements for wastewater treatment and solid waste
management facilities, agricultural waste projects, and lake
restoration programs. No bonds were sold without prior ap-
propriation by the legislature. The bonds are financed
through State sales tax income and are required to be paid
off within 30 years. Under the referendum, $20 million was
earmarked for the Lake Restoration Program. Eligible lake
restoration projects under this program were cost-shared 50
percent by the Federal Clean Lakes Program, 40 percent
by State Referendum 26 funds, and 10 percent by the local
sponsor.
Referendum 26 required that the bonds be issued by Jan.
1,1980. In anticipation of this, replacement Referendum 39,
also to be administered by WDOE, was passed by State
voters in November 1980. The referendum authorized the
sale of $450 million in general obligation bonds for the plan-
ning, design, acquisition, construction, and improvement of
public waste disposal and management facilities, and for con-
tinuation of the Lake Restoration Program initiated under
Referendum 26. The bonds are to be issued prior to Jan.
1,1990 and, as in Referendum 26, will be financed with State
sales tax income. Under Referendum 39, lake restoration
projects can be funded up to 75 percent of the total cost
with 25 percent provided by the local sponsor. If Federal
Clean Lake funds are available, the mininum required con-
tribution by the local sponsor is 10 percent of the project's
total cost.
CURRENT STATUS
Approximately $13.6 million in State funds, under Referen-
dum 26, and $18.3 million in Federal funds have been spent
or obligated to 21 separate lake restoration projects in the
State of Washington. Currently, under Referendum 26, there
are five Phase I diagnostic-feasibility studies, including an
update of the State's 1974 lake classification survey, and
eight Phase II construction projects in progress.
Approximately $1.9 million of Referendum 39 funds have
been spent or obligated to seven lake restoration projects.
Approximately $1.2 million of this is being used to complete
four projects initiated under Referendum 26. The remain-
ing $600,000 provides 75 percent of the total cost of new
Phase I and Phase II lake restoration projects that do not
involve Federal Clean Lake funds.
The Department of Ecology is encouraged by the con-
tinued public interest in the State's Lake Restoration Pro-
gram in spite of the demise of the Federal Clean Lakes Pro-
gram. The WDOE has received letters of intent, actual
Referendum 39 lake restoration grant applications, or strong
indications of interest from potential local sponsors for seven
new Phase I diagnostic-feasibility studies. At the present time,
eight Phase I studies that have been completed or are in
progress are expected to develop into Phase II projects fund-
ed under Referendum 39.
One of the primary reasons the Lake Restoration Program
has been so successful in the State of Washington is
because of the high degree of public support demonstrated
by passage of the two referenda, and the fact that individual
projects are usually encouraged and initiated at the grass
roots level. Once a grant agreement is consummated, the
sponsor and local residents have already made a significant
commitment of time and resources which helps to assure
the project's success.
HOW THE PROGRAM OPERATES
The Washington program is a two-phased effort. The Phase
I diagnostic-feasibility study develops a water and nutrient
budget, identifies water quality problems and their causes,
and recommends restoration alternatives. Cost estimates for
the proposed project are developed under Phase I and an
environmental impact statement or an environmental assess-
ment, whichever is appropriate, may also be prepared. Phase
II actually implements the Phase I findings and recommen-
dations. If not prepared in Phase I, an environmental assess-
ment or impact statement is developed during the early
stages of Phase II.
For a proposed lake restoration project to quality for fun-
ding under Referendum 39, it must satisfy three general
requirements.
1. The lake must have a documented water quality pro-
blem that impairs one or more beneficial uses. Recently,
WDOE has been more liberal in applying this requirement
to allow funding of projects that are designed to prevent pro-
blems before they start, or become more serious. A Phase
I project has recently been funded that falls into this preven-
tion category.
2. The project sponsor must be a recognized public enti-
ty, such as a city or county government, local improvement
district, park district, Federal agency, etc., capable of carry-
ing the project to completion, and able to provide at least
10 percent of the total cost to both Phase I and Phase II.
In addition, the sponsor must demonstrate its ability to
operate and maintain the restoration project once it is com-
pleted and functioning.
3. The sponsor must provide public access sufficient to
allow the general public the same opportunity as lakeshore
residents to enjoy the lake's recreational benefits. Normal-
ly, this includes, as a minimum, a public park and, if boating
is allowed, public boat launching facilities. If these facilities
do not already exist, guarantees must be provided that they
will be established and operative by the time Phase II of the
lake restoration project is completed.
Application Procedures
The local sponsor initiates the lake restoration application
procedure by submitting a letter of intent describing the
nature of the water quality problem, including beneficial uses
impaired, type of public access available, source of local fun-
ding, and a preliminary cost estimate for the proposed pro-
ject. From the information provided in the letter of intent,
WDOE makes an initial determination of project eligibility.
If the project appears to be eligible, the applicant is encour-
aged to submit a grant application, along with a proposed
work plan, for Referendum 39 funds to conduct a Phase I
study. After the application is approved, a work plan is agreed
upon and a grant agreement signed.
When the Phase I study is completed and the final report
is approved, a separate grant application must be made for
Referendum 39 funds to implement Phase II. Successful
completion of Phase I does not necessarily guarantee a grant
award for Phase II. Much depends on the project's feasibili-
ty, total cost, potential for success, public acceptance, and
other considerations.
SPECIFIC PROJECT EXAMPLES
Two Washington lake restoration proejcts currently in pro-
gress are discussed in the following paragraphs, indicating
the types of restoration techniques used and their success
as well as some of the problems and unanticipated difficulties
encountered. It is our conclusion, for the experiences gain-
ed in these and other lake restoration projects in Washington,
that more than 1 year of post-restoration monitoring is re-
136
-------
quired to measure project success and identify potential pro-
blems or difficulties before they become acute and irrever-
sible. A minimum of 2, and possibly 3 years of post-
restoration monitoring is suggested.
Medical Lake
Medical Lake is located in Eastern Washington approximate-
ly 15 miles southwest of Spokane, and lies totally within the
boundaries of the city of Medical Lake. It has a maximum
depth of 18 m (60 feet) and a mean depth of 9.6 m (32 feet)..
The surface area of the lake is 63.2 hectare (ha) (158 acres)
and the volume is 6.2 x ifje m3 (5,103 acre-feet). The total
drainage area is 346 ha (864 acres). The drainage basin is
composed primarily of igneous rock, scoured from Colum-
bia River basalt ( U.S. Geolog. Surv. 1973). The principal
beneficial uses are swimming and fishing. A major city park
and swimming area are located on the southern shore of
the lake.
At the turn of the century, the waters of Medical Lake were
valued as highly therapeutic, and the thick bottom sediments
were desired for mud bath treatments. The native Indians
called the lake "skookum limechin chuch" or strong
medicine water. People traveled many miles by horse and
buggy, railroad, and electric trolley to enjoy the reported heal-
ing waters. By the early 1920's, however, the lake had lost
its popularity because of poor water quality.
Kemmerer et al. (1924) noted that Medical Lake did not
support fish life, apparently because of low dissolved oxygen
and high alkalinity. He also found significant populations of
blue-green algae and rotifers. Ketelle and Uttormark (1971)
classified the lake as highly eutrophic, with dense stands
of blue-green algae and high nutrient levels.
An independent Eastern Washington University study con-
ducted in 1974 by Bauman and Sbltero (1978) determined
that more than 60 percent of the lake's volume was
anaerobic during summer stratification. Phytoplankton
decomposition during the summer depleted oxygen, caus-
ing phosphorus to be released from the reduced sediments
and circulated from the hypolimnion to the epllimnion dur-
ing fall turnover. The resulting high phosphorus levels
throughout the water column stimulated algal growth in the
lake the following spring.
Bauman and Soltero also noted that 74.4 percent of the
estimated phytoplankton cell volume was dominated by blue-
green algal species. The Chlorophyceae and the Cryp-
tophycae were the least abundant algal classes, with both
contributing only 10 percent of the cell volume. Approximately
85 percent of the numeric standing crop of zooplankton was
composed of rotifers, with Eucopepoda and Cladocera com-
prising 11.3 and 4.5 percent of the standing crop,
respectively.
By 1976, the water quality of Medical Lake had de-
teriorated beyond any degree of acceptability. During the
summer months, thick algal surface scums and odors
associated with decaying algae commonly occurred. At fall
turnover, foul smells emigrated from the lake and permeated
the city. As a result, the city of Medical Lake applied to
WDOE for Referendum 26 funds, and to the Environmental
Protection Agency for Section 314 clean lake funds to con-
duct a lake restoration project that would improve the quali-
ty of Medical Lake and make it more recreationally usable.
Grants were made to the city of Medical Lake from WDOE
(40 percent of the total cost) and EPA (50 percent of the total
cost) for the chemical deactivation and precipitation of
phosphorus in Medical Lake with aluminum sulfate (alum),
and for a study to monitor the results.
The earlier work by Bauman and Soltero (1978), which
served to satisfy the requirements of a Phase t study, sug-
gested that alum treatment would be the most feasible and
Rscal and Institutional Support tor Lakes
cost-effective restoration technique for Medical Lake. This
was confirmed by the city of Medical Lake through its con-
tractor, Battelle Pacific Northwest Laboratories (Gasperino,
et al. 1978). Long-term success of the project was anticipated
because Medical Lake lies in a closed basin and nutrient
loadings were primarily internal. Significant external
phosphorus loads were ruled out since the lake received no
sewage effluent or agricultural runoff.
A total of 936 metric tons (1,031 tons) of liquid alum was
applied to Medical Lake between Aug. 3 and Sept. 13,1977.
The alum was distributed in four surface and five subsur-
face (3.5 m; 15 feet) applications. Total project cost, including
4 years of post-project monitoring, was $306,655, or approx-
imately $0.05 per m^ ($60 per acre-foot).
Following treatment, a monitoring program was initiated
to determine the long-term effectiveness o* *he alum applica-
tion on the water quality of Medical Lake. Figure 1 shows
that by fall overturn of 1977, mean soluble reactive
phosphorus levels were reduced approximately 90 percent
(Soltero et al. 1978,1980). However, sulfate concentrations
increased 300 percent because of the alum application.
Phosphorus
Figure 1.—Mean monthly total and soluable reactive phosphorus
concentrations before, during and following alum treatment of
Medical Lake, WA. •
Phytoplankton Btovolume
LO^TTTjJ]^^
Rgure 2.—Mean monthly phytoplankton biovolume before, during
and following alum treatment of Medical Lake, WA.
When compared to pretreatment levels, the mean annual
phytoplankton standing crop (Fig. 2), has been reduced by
90 percent and the mean annual chlorophyll a concentra-
tions (Fig. 3) by 40 percent. The cyanophytes were replac-
ed by chlorophyceans (i.e., Oocysfe) and cryptomonads (i.e.,
Rhodomonas and Cryptomon&s) as the dominant species.
Substantial changes have also occurred in the zooplankton
community with microconsurners (primarily rotifers) giving
way to macroconsumers (i.e., Daphnia putex). Of key impor-
tance here is the increased grazing efficiency of macrocon-
sumers over the previous microconsurners. On a numerical
137
-------
Lake Restoration, Protection and Management
Chlorophyll a.
Figure 3.—Mean monthly chlorophyll a concentrations before, dur-
ing and following alum treatment of Medical Lake, WA.
basis, rotifers decreased from approximately 85 percent of
the total standing crop in 1977 to 43.5 percent in 1980 (Mires
et al. 1981). By 1980, zooplankton species diversity had in-
creased, reflecting a more balanced community.
Water clarity has substantially improved since the treat-
ment because of decreased algal growth. Figure 4 shows
that the mean Secchi disk visibilities during summer stratifica-
tion increased from 3 m (9.8 ft) prior to treatment to approx-
imately 5 m (16.4 ft) following treatment.
The alum treatment improved overall water quality with
the exception of dissolved oxygen levels, which did not in-
crease as expected. Figure 5 indicates that the dissolved
Secchi Disk and Extinction Coefficient
IK
t-
-M-
4-T
-
-ll
..". L. ~~ .
flU,
6,lMKlionl_|)
LU
9W 1M1
Figure 4.—Mean monthly secchi disk visibilities and extinction coef-
ficients before, during and following alum treatment of Medical Lake,
WA.
Dissolved Oxygen
'Figure 5.—Mean monthly dissolved oxygen concentrations before,
during and following alum treatment of Medical Lake, WA.
oxygen (D.O.) concentrations in the deeper waters of Medical
Lake remained about the same, althoughjhe depth at which
oxygen occurred in the lake had increased by about 2
meters. The mean water column concentrations declined to
lows of 5.0 and 5.8 mg/l D.O. during 1978 and 1979, respec-
tively. During 1980 and 1981 mean oxygen levels of the water
column were about 7 mg/l. Anoxic conditions prevailed in
the hypolimnion below 10 meters for 7 months (March-
September) in 1978 and 5 months (May-September) in 1979.
The lack of improvement in the overall oxygen regime pro-
bably was caused by three factors: (1) the stabilization of
the deposited organic matter as a result of the treatment;
(2) incomplete mixing at spring overturn which did not allow
for the replenishment of oxygen-depleted bottom waters
following winter stagnation; and (3) the continuous inflow of
oxygen-poor ground water, which would reduce the dissolved
oxygen concentrations of the lake. Gasperino et al. (1978)
have calculated groundwater influx to be 6.7 x 10s m3 (543
acre-feet) per year.
In June 1978, the Washington Department of Game
planted approximately 14,000 fingerling rainbow trout (5-8
cm; 2-3 in.) in Medical Lake. Survival and apparent good
growth of the trout prompted additional stockings in 1979,
1980, and 1981 for a 4-year total of approximately 38,000
fish. The 1978 and 1979 plants did very well, at least through
the summer of 1980. Concern regarding these plantings was
noted, because of the possibility they may shorten the
longevity of the restoration project and cause the lake to
revert to its former eutrophic condition because of size selec-
tive predation on Daphnia pulex. Data show that in 1977 the
majority of D. pulex were 2.25 mm or greater in length declin-
ing to between 1.50 and 1.74 mm in 1978 and 0.75 to 0.99
mm in 1979 and 1980. Daphnia pulex comprised 99.3 per-
cent (88.9 percent biomass) of the diet of 1979 0+ trout.
Although D. pulex were still dominant (59.7) percent in the
food of 1979 1+ trout, they comprised only 8.1 percent of
the biomass. In 1980, D. pulex comprised 97.7 percent and
90.3 percent of the diet of 1+ and 2+ trout, respectively.
However, the biomass contributed was much less for 2+ trout
(17.3 percent) than for 1+trout (70.7 percent) (Knapp, 1981).
Evidence points to trout selectively feeding on D. pulex
(an effective algal grazer) in Medical Lake causing algal levels
in 1981 to increase threefold over 1980 levels. In addition,
Secchi disk depth (Fig. 4) had dropped from 5 m (16.4 ft.)
in 1980 to approximately 2 m (6.6 ft.) in 1981. The data in-
dicate that the increase in algal standing crop was not caus-
ed by an increase in phosphorus or by a failure of the alum
treatment. The decrease in water clarity and the resulting
increase in algal standing crop was attributed to intense trout
predation upon D. pulex.
Continuing studies funded by the city of Medical Lake and
the State's Lake Restoration Program under Referendum
39 have shown that during 1981 trout became thin and
emaciated and actually starved to death. The average weight
of trout originally stocked in 1979 dropped 300 g (0.66 Ibs.)
between June and August 1981. In September of 1981 a
scuba survey was made of the lake and large numbers of
dead or dying trout were observed. In 1982, following the
1981 fish kill and reestablishment of some ecological
balance, D. pulex reappeared in larger numbers. Algal stan-
ding crop declined as a result of increased grazing pressure,
with a resulting increase in water clarity. The remaining trout
in Medical Lake have gained weight with D. pulex as their
preferred food. Size selective predation of D. pulex
significantly affected the water quality of Medical Lake and'
nearly destroyed the stocked trout populations. ••.*>•
Under Referendum 39 and the State Lake Restoration Pro-
gram, the Washington Department of Ecology has contracted
through the city of Medical Lake the continued monitoring
of the overall water quality of Medical Lake. The purpose
of this project is to:
138
-------
Fiscal and Institutional Support for Lakes
1. Study the dynamics of the fish populations in Medical
Lake and their impact on D. pulex;
2. Determine the maximum fish populations, with special
emphasis on the number of trout the lake will support under
the improved water quality conditions provided by the restora-
tion project; and
3. Estimate the size of the trout population to provide a
baseline for establishing future stocking rates, thus ensur-
ing the longevity of Medical Lake's enhanced water quality.
Lake Sacajawea
Lake Sacajawea is located in the center of the city of
Longview (population 30,500) in Cowlitz County, Wash. The
lake is an oxbow lake that had been at one time a channel
of the Cowlitz River. In 1924, R.A. Long spent $1.2 million
to dredge the lake and create and landscape a surrounding
park. After completing the work, he gave the lake and the
park to the city of Longview.
The lake covers approximately 21.2 ha (53 acres), is 2.69
km (1.67 miles) long, and ranges from 30 m (100 feet) to
150 m (500 feet) wide. It contains four small islands with a
total area of about 1.0 ha (2.5 acres). The maximum depth
is approximately 4.5 m (15 feet) and the mean depth is 1.8
m (5.9 feet). The volume of the lake is estimated to be 0.4
x 106 m3 (325 acre-feet).
Lake Sacajawea is essentially a "window" on the ground-
water table and, being the lowest area in the city, it was on-
ly natural when the city was designed and built in 1924 to
construct the storm sewers so that the storm runoff from the
surrounding area drained into the lake. Stormwater runoff
from approximately 652 ha (1,630 acres) of residential, com-
mercial, and semirural land drained into the lake through
some 28 separate outfall lines and one major drainage ditch.
Past reports indicated that during the early years of the
lake's existence, the water was reasonably clear and was
used for all forms of recreation, incuding swimming, boating,
and fishing. Prior to restoration, however, the lake was tur-
bid, colored and abundant with algae, and the shores were
clogged with thick growths of macrophytes which, for the
most part, prevented park visitors from reaching the water's
edge.
The city of Longview owns the entire shoreline area and
has developed it into a very functional 26.8 ha (67-acre) city
park with walking and jogging trails, horseback riding paths,
picnic areas, a physical fitness trail, playgrounds, restrooms,
children's facilities, flower gardens, etc. One can find users
enjoying these facilities night and day, 365 days a year. The
lake itself, however, is used only by a few canoers and
juvenile fishermen because of the poor water quality. The
fishery has consisted almost totally of carp with a very few
bass and bluegill.
Since Lake Sacajawea is located in an urban area and
because Cowlitz County has relatively few lakes suitable for
recreation, the lake is an important natural resource. The
people of the area have demonstrated a longstanding interest
in the condition of Lake Sacajawea, as documented by
numerous reports and newspaper articles dating back to the
early 1950's, and the formation of the Lake Sacajawea Im-
provement Committee in the early 1960's. This committee
was comprised of dozens of concerned citizens from all
walks of life within the community.
Monitoring prior to implementation of the restoration pro-
ject showed that sediment had accumulated over the entire
lake bottom, varying in depth from approximately 0.15 m (1.5
feet) to 1.5 m (5 feet). This sediment was an accumulation
of material washed into the lake and organic material that
had grown and settled within the lake itself. The sediment
was a very fine black, organic material, rich in both nitrogen
and phosphorus.
Background water quality data taken near the surface of
the lake showed total phosphorus levels ranging from 10 to
870 fjg/l with the mean value of 190 ^g/l. Chlorophyll a ranged
from 1.7 to 106 ^g/l with a mean value of 12.4 ^g/l.
Secchi disk readings ranged from 0.27 to 1.04 m (.88 to
3.42 feet) with a mean value of 0.62 m (2.03 feet).
Temperature and dissolved oxygen profiles showed a sur-
prisingly strong stratification within this shallow lake from
spring through September. The stratification, of course,
prevented mixing of the vertical water column. Consequently,
bacterial activity within the organic rich sediment resulted
in a total depletion of oxygen during summer months below
a depth of 5 to 6 feet.
Other problems within the lake included:
1. Numerous algae mats created a visual and odoriferous
nuisance.
2. Rainstorms seriously degraded water clarity. At one
point the Secchi disk readings in the northernmost portion
of the lake were reduced to 0.09 m (0.35 feet) and after 2
months this reading had increased to only 0.20 m (0.67 feet).
3. Macrophyte growth was often so heavy during sum-
mer and fall months that the underlying water was almost
invisible.
The city of Longview applied to WDOE and EPA for funds
to address these problems and to restore the quality of Lake
Sacajawea and its recreational potential.
Twelve restoration techniques were studied in 1976 to
determine their applicability to the restoration of Lake Saca-
jawea. Because of the importance of this lake to the citizens
in the Longview area, techniques were examined that would
provide a long-term solution to restoring water quality and
the aesthetics of Lake Sacajawea. After considering the
various alternatives, a restoration plan was developed that
consisted of three major elements:
1. Diverting nutrient-rich and sediment-laden stormwater
away from the lake.
2. Providng a nutrient poor makeup and dilution water
from the nearby Cowlitz River.
3. Dredging most of the sediment accumulations from the
lake to remove them as a nutrient source and, at the same
time, deepen the lake and remove many of the macrophytes.
The restoration plan was divided into two phases. The first
consisted of constructing a 72-inch stormwater line along the
east side of the lake to divert about 95 percent of the storm
water. At the same time, a new water supply line and pump
station were constructed that would supply nutrient-poor
makeup and dilution water from the Cowlitz River at the rate
of 10 MGD. The second phase of the work consisted of
dredging the lake and disposing of the sediment.
In the fall of 1979 contracts on the first phase of the pro-
ject were awarded, and construction began in January 1980.
Work proceeded smoothly, and in May of 1980 the water
supply line and pump station were approximately 95 per-
cent complete when the eruption of Mount St. Helens sent
millions of cubic yards of silt and debris down the Toutle
River into the Cowlitz River. This disaster buried the water
supply intake structure under approximately 14 feet of silt
and work was halted. However, the Cowlitz River was the
only planned source of dilution water for the lake, and
because there were weeks of relatively clear water (JTUs
less than 20) during the summer of 1980, it was decided
to complete the remaining work on the pump station. In
November, the Corps of Engineers finished dredging this
portion of the Cowlitz River, and in December 1981, the
pumps for the lake water supply became operational. Heavy
runoff in January 1981, however, once again buried the in-
take structure, preventing this portion of the project from
operating as proposed.
The 72-inch, 3,900 m (13,000 feet) stormwater diversion
line was completed in the fall of 1980, and by January 1981,
139
-------
Lake Restoration, Protection and Management
all of the associated diversion structures were operating.
Hydraulic dredging of Lake Sacajawea was begun in Oc-
tober 1981 and completed in September 1982. This work
deepened the lake by about 0.68 m (2.25 feet) and increas-
ed its volume to an estimated 0.55 x 106 m3 (443 acre-feet).
At this time, the restoration project is still not completed,
mainly because of delays and additional problems resulting
from the eruption of Mount St. Helens. The estimated cost
for the project is $5.1 million, or approximately $9.27 per m3
($11,512 per acre-foot).
The eruption of Mount St. Helens and the subsequent
deterioration of the water quality in the Cowlitz River is a
problem unique to the Lake Sacajawea project. Alternative
sources of dilution water, which may be feasible and
economical to develop, are currently being investigated. Even
without the dilution water system in place, however, the water
quality seems to have improved. With the stormwater diver-
sion line in place, the clarity of the lake is no longer degrad-
ed with each heavy rainstorm, and initial Secchi disk readings
have improved. The Washington State Game Department
also treated the lake with rotenone in May 1982, and a good
crop of bass, bluegill, and some trout has apparently been
established.
The restoration of Lake Sacajawea still has potential for
success; however, because of the unique problems
associated with the Mount St. Helens eruption, it is simply
going to take longer than originally anticipated
CONCLUSIONS
Our experience in the State of Washington, illustrated by the
Medical Lake and Lake Sacajawea projects, is that efforts
to restore a lake to a more acceptable level of water quality
can be affected by factors that may or may not be con-
trollable. Naturally occurring phenomena, as in the case of
Lake Sacajawea, or the policies of other agencies, such as
in the Medical Lake situation, may seriously affect or delay
anticipated water quality improvements. The ultimate suc-
cess of a project may also be influenced by internal systems,
either biological, chemical, or physical, which take time to
develop and be recognized.
Therefore, we strongly feel that post-restoration monitor-
ing should continue for at least 2 years, and probably 3 years
after the project is completed and the contractors have left
with their dredges, backhoes, and alum applicators. Very
often, a post-restoration problem, if recognized soon enough,
can be corrected by a relatively simple change in policy, such
as altering the trout stocking date. If a problem is not
recognized and is left unchecked, the entire restoration ef-
fort could be wasted.
REFERENCES
Bauman, L.R. and R.A. Soltero. 1978. Limnological investigation
of eutrophic Medical Lake, Wash. Northw. Sci. 52(2).
Gasperino, A.F. et al. 1978. Restoration of eutrophic Medical Lake,
Wash, by treatment with aluminum sulfate: Preliminary findings.
Battelle Pacific NW Labs, Wash.
Kemmerer, G., J.F. Bovard, and W.R. Boorman. 1924. Northwestern
lakes of the United States: Biological and chemical studies with
reference to possibilities in production of fish. Bull. Bur U.S. Fish.
XXXIX. 1923-1924, Doc. No. 944.
Ketelle, M.J., and P.O. Uttormark. 1971. Problem lakes in the United
States. EPA Proj. No. 16010 E.H.R., Water Pollut. Control Res.
Ser., Washington, D.C.
Knapp, S.M. 1981. The impact of whole-lake application on the fisher-
ies potential of Medical Lake. M.S. Thesis. Eastern Washington
Univ., Cheney.
Mires, J.M., R.A. Soltero, and G.R. Keizur. 1981. Changes in the zoo-
plankton community of Medical Lake, Wash, subsequent to its
restoration by a whole-lake alum treatment and the establishment
of a trout fishery. J. Freshw. Ecol. 1: 167-178.
Soltero, R.A., D.G. Nichols, M.A. Beckwith, and G.R. Keizur. 1978.
Limnological investigation of Medical Lake, Wash, before, dur-
ing, and after a whole lake application of alum. Battelle NW Spec.
Agree: B-49803-B-H, Proj. Comple. Rep. Eastern Washington
Univ., Cheney.
Soltero, R.A., D.G. Nichols, and J.M. Mires. 1980. Limnological
investigation of Medical Lake, Wash, following a whole-lake ap-
plication of alum. Battelle Pacific NW Labs. Agree. No. B-49803-B-
H, Supp. No. 1. Comple. Rep. Eastern Washington Univ., Cheney.
U.S. Geological Survey. 1973. The channeled scablands of Eastern
Washington - the geological story of the Spokane flood. U.S.
Govern. Print. Off., Washington, D.C.
Washington State Department of Ecology. 1981. Guidelines for the
Lake Restoration Grant Program—Referendum 39. WDOE 81-17.
140
-------
LIMING TO MITIGATE SURFACE WATER ACIDIFICATION:
INTERNATIONAL PROGRAMS, STRATEGIES, AND ECONOMIC
CONSIDERATIONS
JAMES E. FRASER
DOUGLAS L. BRITT
General Research Corp.
McLean, Virginia
ABSTRACT
Liming, as a technique to mitigate surface waters already affected by acid deposition or to protect waters
that are potentially sensitive to acid deposition, has been conducted in Scandinavia and North America.
This paper describes the history of the liming programs in the various countries. The recently completed
5-year (1977-1981), government-sponsored, pilot liming program in Sweden is discussed along with pro-
jections of Sweden's future liming program. Norway's government-funded experimental/research liming
project also is described. Past and currently projected liming projects in the provinces of Ontario, Quebec,
and Nova Scotia are summarized. These project descriptions include the experimentation by both Cana-
dian Federal and provincial researchers. Although the liming of dystrophic or softwater ponds to increase
fish productivity has been performed in the United States for several decades only seven States currently
have acid-deposition related aquatic liming projects. These projects are briefly described. The costs
associated with liming materials and various lake and stream liming techniques also are reviewed and
compared.
INTRODUCTION
Currently, aquatic ecosystems in the United States, Canada,
and Scandinavia are being affected by acidification caused
by atmospheric acid deposition. Scientific documentation in-
dicates that many of these areas are losing or have already
lost fish populations and other aquatic life. Because wide-
scale source control of acidic deposition precursors will pro-
bably not occur in the near term, other options are being
investigated to mitigate these adverse effects. One poten-
tial approach involves applying lime or alkaline materials at
the receptor sites (e.g., lakes and waterways) to neutralize
incoming acids. However, increased pH and increased acid
neutralizing capacity resulting from liming (liming is a generic
term defined as the addition of any base material to neutralize
surface waters, sediments, and soils) are only temporary ef-
fects, if lakes and streams continue to be subject to at-
mospheric acid deposition.
SWEDISH LIMING PROGRAM
Several countries already have adopted policies to mitigate
surface water acidification problems through experimental
and/or operational liming programs. Liming is the major
method of protecting and renovating acidified lakes and
streams in Sweden and Norway, with Sweden having the
largest aquatic liming program in the world. Major objectives
of this program include saving (or restoring) sport fisheries,
preserving endemic fauna, and protecting ground water
(Andersson, pers. comm.).
The federally funded National Swedish Liming Project was
started early in 1977 as a result of both public and scientific
pressure on politicians. This initial 5-year test program, which
was more operational than experimental, ended in June 1982
and involved approximately 450 liming projects (in approx-
imately 1,500 lakes and waterways), with a total application
of some 260,000 metric tonnes of liming agents. The total
Federal budget for the test period was $12,480,000 (U.S.).
The program was designed to test various application
methods and liming agents, evaluate the effects of liming
on biota and water chemistry, produce optimum liming
guidelines and strategies, and investigate the costs and
organization needed for future large-scale liming throughout
Sweden (Bengtsson, 1982a). The total cost of the program
was actually greater than $12,480,000, as the Swedish
government only paid for 75 percent of most liming projects,
with the exception of a few specialized demonstration pro-
jects (e.g., $500,000 for the Hogvadsan River project) where
the government covered 100 percent of the cost (Fraser, et
al. 1982).
In Sweden, all waters are privately owned, and 25 per-
cent of the lime application costs are borne by the owners,
such as fishing clubs, local cottage communities, or other
private interest groups (Bengtsson, pers. comm.). Applica-
tions requesting 75 percent Federal financing for liming are
screened twice a year by a committee of the Swedish En-
vironmental Protection Board and National Board of
Fisheries. The applicant has to submit data demonstrating
that the water body has become acidified. These data must
include, at a minimum, measurements of pH and alkalinity
(Fraser, et al. 1982). Another aspect of the Swedish liming
program is that the 25 percent contribution from the lake/land
owners does not have to be monetary. The owners may
substitute labor hours, boat rental charges, etc., for their
financial contribution (Bengtsson, pers. comm.).
Based on the results and experiences from the initial test
program, Sweden is now embarking on a new large-scale,
long-term liming program. In addition to mitigating the
acidification of lakes and streams, research will begin to
neutralize ground water and forest and watershed soils
(Bengtsson, 1982a). Currently, the new program is funded
at about $6 million for fiscal year 1982 and is expected to
increase to approximately $10 million in FY 1983, $14 million
in FY 1984, $32 million in FY 1985, and $40 million in FY
1986 —at which time approximately 20,000 bodies of water
will have been limed (Bengtsson, 1982a; Fraser et al. 1982).
As the Federal liming budget increases, the contribution from
141
-------
Lake Restoration, Protection and Management
lake/land owners will be reduced from 25 percent to 15
percent.
NORWEGIAN LIMING PROGRAM
Although Norway experiences many of the same surface
water acidification problems as Sweden, Norway has a much
smaller government-funded liming program which is research
oriented. In late 1979 the Norwegian Ministry of the Environ-
ment initiated a 5-year project whose goals were to in-
vestigate the scientific and technical aspects of liming.
Because the Norwegian program began several years after
the Swedish liming program, Norway is attempting to build
upon and supplement the Swedish results (Johannessen and
Wright, 1982).
The Lake Howatn liming project, which began in the
spring of 1980, was the first Norwegian experimental liming
project (Fraser et al. 1982). Since then a smaller number
of field and laboratory experiments on the liming of running
waters have been conducted (Johannessen and Wright,
1982). Johannessen and Wright (1982) report that future lim-
ing in Norway will focus mainly on running waters because
of the high rainfall (over 1,000 mm/yr), high runoff (over 800
mm/yr), and rapid flushing rates (turnover times of 1 to 6
months) in the many small lakes typical of the acidified
regions of Norway. Government funding of liming research
activities in Norway is presently less than $1 million per year
and will probably continue to be modest in future years.
Norwegian authorities have adopted a conservative attitude
toward liming because of unknown, long-term, ecological ef-
fects, and fundamental ethical and political reasons (Johan-
nessen and Wright, 1982).
CANADIAN LIMING PROGRAMS
Past aquatic liming experiments in Canada involved the ap-
plication of alkaline materials in four lakes near Sudbury, On-
tario, from 1973 to 1977 by Ontario Ministry of the Environ-
ment scientists (Scheider et al. 1975a,b; Scheider and Dillon,
1976; Van et al. 1977; Van and Dillon, 1982). Another historic
project involved the liming of Lake Solitaire in the Lauren-
tian Highlands of Quebec on land privately owned by the
Seminaire de Quebec (Can. Fed./Prov. Task Force, 1982).
More current Canadian liming projects are in their initial
phases, with both Federal and provincial researchers and
a private fishing association conducting experiments.
Federal scientists in Nova Scotia recently have limed four
headwater lakes and are planning to lime a fifth one in the
near future (Watt, pers. comm.). They also have been in-
volved in fish hatchery and streambed liming experiments
(Goff et al. 1981; Fraser et al. 1982). With an approximate
budget of $85,000 per year, the Federal scientists in Nova
Scotia are attempting to create deacidified refuges to save
existing genetic stock of Atlantic salmon (Watt, pers. comm).
The latest liming experiment in Quebec (at Lake Junior, north
of La Juque) was conducted jointly by the Quebec Fish
and Game Department and a private fishing association.
The lake was limed and then restocked with trout in the
summer of 1981, with a followup experiment to be done
in 1982 (Can. Fed./Prov. Task Force, 1982).
Currently in Ontario, the Ministries of Natural Resources
and of the Environment are in the initial phases (begun
in spring of 1982) of a 5-year, $610,000 per year ex-
perimental lake neutralization project to investigate lim-
ing as a preventative measure to save genetically impor-
tant fish species or popular fishery resources (Fraser et
al. 1982). Booth Aquatic Research Group Inc. (1982)
reports that there are two experimental goals in this pro-
ject: (1) to investigate liming as a technique to rehabilitate
acidified lakes to their pre-acidification states containing
healthy fish populations; and (2) to investigate liming as
a means to protect lakes in danger of acidification.
To facilitate the Canada-U.S. negotiations on transboun-
dary air pollution abatement programs, the Canadian
Federal/Provincial Scientific Liaison Committee estab-
lished a Task Force to develop guidelines for Canada on
the issue of artificial neutralization of lakes and streams
as a means of mitigating the effects of acidic deposition.
The resultant guidelines specify that a neutralization pro-
ject may be appropriate when at least one of the follow-
ing benefits can be expected as a result of the proposed
operation:
• A danger to human health will be reduced.
• An ecologically sensitive aquatic system will be
protected
• A fishery identified as acid-sensitive or acid-stressed
will be protected (Can. Fed./Prov. Task Force, 1982).
U.S. LIMING PROGRAM
In the United States, liming has been used since the ear-
ly 1940's to enhance fish production in dystrophic ponds
and lakes, especially in Wisconsin and the upper pennin-
sula of Michigan (Hasler et al. 1951; Johnson and Master,
1954; Waters, 1956; Waters and Ball, 1957; Kitchell and
Kitchell, 1980; Jackson, 1981). Liming to improve fish pro-
duction in softwater ponds and streams in the
southeastern United States also has been practiced for
many years (Boyd, 1974; Sills, 1974; Arce and Boyd, 1975;
Boyd, 1976). The pH values in these southeastern surface
waters are usually circumneutral, although the bottom
sediments are frequently acidic (Boyd, 1974). The
chemical neutralization of acid mine drainage from aban-
doned coal mines, especially in Appalachia, also has fre-
quently involved applying limestone and other alkaline
materials (Pearson and McDonnell, 1975).
In addition to these activities, approximately 100 liming
projects have been initiated during the past 25 years to
reduce the acidity of surface waters allegedly acidified by
acidic deposition or by unknown and presumably undeter-
minable sources (Fraser and Britt, 1982). These projects
have been conducted mostly in seven States; Connecticut,
Massachusetts, New York, North Carolina, Pennsylvania,
Rhode Island, and West Virginia. Presently, there is no
federally funded U.S. acidic deposition liming program nor
any specific Federal liming subsidies to the States that
are already chemically treating acidified surface waters.
Liming projects in these seven States have been or are
being conducted and funded through State fishery depart-
ments (some of which do receive Federal monies, although
they are not targeted specifically for liming), private fishing
associations, universities, or by cooperative ventures be-
tween the various entities. There are some exceptions: the
U.S. Fish and Wildlife's Cooperative Fishery Research
Unit and the Tennessee Valley Authority have provided
assistance to small liming experiments in Pennsylvania
and North Carolina, respectively (Fraser and Britt, 1982).
Current State liming programs are conducted to restore
surface waters that have completely lost fish populations,
as well as to protect susceptible, but still viable, fisheries.
Annual liming budgets for the States of New York and
West Virginia are approximately $20,000 to $30,000
(Kretser, pers. comm.) and $1,000 to $2,000 (Phares, pers.
comm.), respectively. Other States such as Massachusetts"
and Rhode Island do not necessarily lime on an annual
basis; instead, surface waters are treated only when water
quality surveys indicate a need. For their most recent lim-
ing operations, the annual cost was approximately $15,000
to $20,000 in Massachusetts (Oatis, pers. comm.) 'and
$6,000 in Rhode Island (Guthrie, pers. comm.). Because
142
-------
Fiscal and Institutional Support for Lakes
Connecticut has not performed any recent liming opera-
tions, and the projects in Pennsylvania and North Carolina
are small, isolated experiments, no annual liming budgets
are available for these States,
New York State's Bureau of Fisheries, which conducts
the largest U.S. aquatic liming program, is in the process
of developing a liming policy to provide working guidelines
tor fisheries managers and administrators in selecting can-
didate waters for liming projects. Highest priority will be
given to:
• Waters that contain rare, threatened, or endangered
species.
• Broodstock waters that contain heritage strains of fish
species, valuable to fish management programs.
• Waters that historically have had excellent sportfish
populations but have recently been severely im-
pacted by acidic deposition.
» Preservation and maintenance of a unique fishery in
terms of quality and experience (Kretser, 1982).
The Bureau of Fisheries has also Initiated a decentral-
ized volunteer liming program thorugh the Adirondack
Conservation Council. Through this program, County Fish
and Game Federations will be able to lime several small
waters and receive guidance from Bureau of Fisheries staff
(Kretser, 1982).
COSTS OF LIMING MATERIALS AND
APPLICATION
Numerous technical options (materials and application
techniques) are available for liming aquatic ecosystems
affected by acidic deposition. They have been described
in detail by Fraser et al. (1982) and Fraser and Britt (1982).
The choice of liming material will affect both the resource
costs and requirements. The costs of the most common-
ly used alkaline materials (i.e., limestone gravel, aglime,
hydrated lime, and quicklime) are approximately $18, $46,
$84, and $92 (1982 dollars) per tonne, respectively,
delivered to site within 300 kilometers (Gutschick, pers.
comm.). A major cost governing the price of the materials
is the transportation, primarily the distance of the material
source to the liming site. This is especially true for
limestone gravels and aglime.
Data available for current operational liming projects
suggest that directly applying lime to lakes, as practiced
in the Swedish and New York State liming projects
(Bengisson et al. 1980; Blake, 1981), is more cost effec-
tive than liming watersheds or tributary streams. Specific
cost data for watershed liming projects, however, have not
been documented.
Based on a review of 27 recent liming projects that
employed a variety of techniques in the Adirondack Moun-
tains, Blake (1981) revealed that the costs for the lime and
its application (between 1975 and 1979) ranged from a low
of $75 per hectare for accessible lakes limed by boat, up
to a maximum of $730 per hectare for remote lakes limed
by helicopter. Blake, however, suggests that if helicopter
liming was practiced on a routine basis (enabling a more
efficient operation), the cost could be lowered to approx-
imately $250 per hectare (1979 dollars). Chubbuck and
Patterson (1981), in their review of liming costs, reported
a cost of approximately $116 per hectare (1981 U.S.
dollars) for the liming (Ca(QH)2 and CaCO3J of two lakes
near Sudbury, Ontario. In southern Sweden, the average
total costs for spreading 1 tonne of aglime from 1977 to
1981, have been estimated at between $50 and $70 or
approximately $20 to $28 per hectare (Bengtsson et al.
1980; Bengtsson, 1982b; Natl. Fish. Board Sweden, 1982),
A breakdown of the per tonne costs for the purchase,
Table 1. — Range of costs for limestone materials,
transport, and application in Sweden 1981.
Liming
parameters
Costs3
(per tonne)
Materials'1
Bulk limestone
Bagged limestone
Transport of limestone
Application Techniques
Truck
Pontoon boat with blower
Helicopter
By hand
$20-$30
$40
$10420
$4-$6
$10416
$30440
$2Q-$60
1981 U.S dollars
limestone (0-G.5 mrn size)
Source' Natl Fish. Board Sweden, 1982,
transport, and application of limestone (0-0.5 mm) in
Sweden during 1981 is presented in Table 1.
Table 2 summarizes the costs of liming, estimated lime
requirements, estimated rate of application, and number
of personnel used for manual labor for various liming
techniques deployed experimentally and/or operationally
in North America and Scandinavia. This table, however,
does not include reapplication rates, which would obvious-
ly affect the suitability or economic feasibilty of a techni-
que for a specific lake or stream. All cost data have been
converted to 1981 U.S. dollars. Currency conversions are
based only on foreign exchange rates and implicit price
deflators for Gross National Product.
Based on the limited liming experience in New York
State and results from water quality data on 777 of the
2,877 lakes in the Adirondack region, Menz and Driscoll
(1982) estimated that annual costs for a 5-year liming pro-
gram for the 777 lakes would range from $2 million to $4
million (1982 dollars), depending on the specific target pH
of the limed waters. This would be approximately $50 to
$75 per surface hectare for accessible lakes, and $500
per surface hectare for remote lakes. They also reported
that a 5-year treatment program for only lakes with initial
acid neutralizing capacity less than 100 ^eq/1, would cost
approximately 40 percent more annually for both the ac-
cessible and remote sites. Menz and Driscotl (1982) also
state that these cost estimates are conservative because
they do not include the costs of transporting personnel and
application equipment to the base chemical storage sites,
costs of establishing and administering the program, and
possible economic and other costs that the liming program
may impose on the affected environment.
In general, based on the cost data presented in Table
2 and the cost estimates of Menz and Driscoll (1982), ap-
proximately $100 per hectare is probably a reasonable
estimate of present day annual lake liming costs for ap-
plication by boats in small-scale, experimental projects,
if the costs for planning and monitoring the water quality
are excluded. Having the same exclusions, helicopter ap-
plication costs for remote sites would probably average
approximately $400 to $500 per hectare. Although fixed-
wing aircraft liming costs have not been reported,
hypothetical application costs using large, multi-engine,
fixed-wing aircraft have been estimated to range from $265
per hectare to $500 per hectare for liming small remote
lakes (Table 3). Although truck application costs for
specific sites also have not been reported, a hypothetical
143
-------
Table 2. — Summary of costs, manpower, estimated lime requirements, and estimated application rates
for various liming techniques utilized in North America and/or Scandinavia.
Application
technique
TRUCK
— with spreader
— with blower
BOATS
— - with slurry
pump
— dumping
bagged lime
— with com-
pressed air
blowers
— hosing aglime
off barge
AIRCRAFT
— helicopter
SILOS
»
Location
Lake Howatn,
Norway
Surtan Water-
shed, Sweden
Middle & Hannah
Lakes,
Sudbury, Ontario
Sandy Lake,
Halifax, Nova
Scotia
6 Ponds in
Adirondack
Mountains,
New York
Southwestern
Sweden
Big Moose Lake,
Adirondack
Mountains,
New York
Inlet Brook to
Char Lake,
Sweden
4 Ponds in
Adirondack
Mountains,
New York
2 silos on
Hogvadsan River
System, Sweden
Material
Aglime
Aglime
Aglime and
hydrated
lime
Aglime
Agiime and
hydrated
lime
Aglime
Aglime
Aglime
Aglime
Aglime
Total
quantity
added
(tonnes)
200b
1600
55.8
118
61
_
36.4
30
60
200/
silo/
year
Maximum Manual
estimated labor
application Estimated utilized
rate/unit lime require- in 8 hour
time (tonnes/ ments (tonnes/ workday Reported
hour) hectare) (# people) costs
— 2b — —
7 52 —
1 1.1 5 $75/hac'd
(1975 $s)
1 1.7 4 $93/hac
(1981 $3)
— 0.8 3 $132/had
(1978 $s)
60 0.4 3 $20-$28/had
(1977-1981
$s
6 0.6 4 $5/hae'f
(1981 $s)
7.5 — — —
5 1.5 8 $343/had
(1979 $s)
0.05 — 0 $20,000
construc-
tion costs/
silo (1981 $s)
Costs3
converted
to 1981 U.S.
dollars
—
—
$116/had
$93/ha
$173/had
$28/had
$5/hae'f
—
$4l5/had
$20,000
construc-
tion costs/
silo
References
Fraser et al. 1982
Fraser et al. 1982
Scheider et al. 1975a;
Chubbuck & Patterson,
1981
Fraser et al. 1982
Blake, 1981
Bengtsson et al. 1980;
Bengtsson,
pers. comm.
Driscoll et al. 1982;
Driscoll, pers. comm.
Fraser et al. 1982
Blake, 1981
Fraser et al. 1982
i1
-------
Table 2 (concluded). — Summary of costs, manpower, estimated lime requirements, and estimated application rates
for various liming techniques utilized in North America and/or Scandinavia.
Application
technique
BARRIERS
ROTARY DRUMS
DIVERSION WELL
Location
Unnamed Creek,
Killarney Pro-
vincial Park,
Ontario
Lower Great
Brook, Milton,
Nova Scotia
Gifford Run,
Pennsylvania
Constable Creek,
Adirondack
Mountains,
New York
Otter Creek,
West Va.
Southwestern
Sweden
Material
Dolomitic
limestone
Calcitic &
dolomitic
limestone
Dolomitic
limestone
Calcitic
limestone
'Calcitic
limestone
Calcitic
limestone
Maximum Manual
estimated labor
Total application Estimated utilized
quantity rate/unit lime require- in 8 hour
added time (tonnes/ ments (tonnes/ workday
(tonnes) hour) hectare) (# people)
345 Not applicable — 450 hours
labor
donated
free
18 Not applicable — 3
36 Not applicable — 5
2.6 Not applicable — 3
— 0.069 — 0
— — — 0
Reported
costs
$17/tonnec'e
(1980 $s)
$60/tonnec'd
(1981 $s)
$20/tonnee'f
(1976 $s)
$8/tonnee'f
(1981 $s)
$3000 con-
struction
cost/well
(1981 $s)
Costs3
converted
to 1981 U.S.
dollars
$22/tonnee
$60/tonned
$30/tonnee
$8/tonnee'f
.
$3000 con-
struction
cost/well
References
Fraser et al. 1982
Chubbuck & Patterson,
1981; Fraser et al.
1982
Arnold, 1981; Arnold,
pers. comm.
Driscoll et al. 1982;
Driscoll, pers. comm.
Menendez, pers. comm.
Natl. Fish.
Board Sweden, 1982
a Costs have been converted to 1981 U.S. dollars by applying only foreign currency exchange rates (when appropriate) and implicit price deflators on U.S. Gross National Product. Because labor wages
and/or material costs do not necessarily fluctuate at the same rate as national or annual average inflation, and because of inherent differences in labor costs among different countries, these reported
values represent only rough approximations.
b Represents liming in main portion of lake only; smaller quantities (40 tonnes) of aglime also were applied, at higher doses, in an embayment.
c Canadian dollars.
d Assumes cost for labor is included.
e Labor donated free.
' Cost does not include limestone transportation charges.
9 At a flow velocity of 0.14 nvVsec.
— No data available
Source: Fraser and Britt, 1982.
§
Q.
62.
VI
-------
Lake Restoration, Protection and Management
Table 3. — Cost estimate for hypothetical liming operations in Wisconsin lakes.
Application technique
Assumptions
Reported cost factors
converted to 1981
dollars
Fixed-Wing Aircraft
•B-26
B-17 or DC-4
Capable of delivering entire payload (3400-4160)
liters lime/water slurry) on 8.1 ha lake
Capable of delivering entire payload (6800-8330)
liters lime/Water slurry) on 8.1 ha lake
• $0.30/liter for slurry
• $565/hr flying cost
• $565/day aircraft
availability charges
• Total $265-$290/ha
* $0.30/liter for slurry
» $1017/hr flying cost
* $565/day aircraft
availability charges
• total $450-$500/ha
Source: Adapted from Sheffy et si. 1980.
truck liming operation for Wisconsin lakes is estimated to
cost approximately $75 per hectare (Table 3).
The annual per hectare costs of lake liming are a direct
function of the reapplication rate of the lime material. The
frequency of reapplication will depend, in part, on several
factors:
The acidity of the lake and its sediments.
The flush rate of the lake.
The magnitude of new acid input.
The type of alkaline material used.
The application techniques selected.
The management practices of those who do the ac-
tual liming.
At the present state-of-the-art, reapplication rates ap-
pear very difficult to predict with reliability.
Based on past and present stream liming experience
with limestone barriers (Table 2), the costs, excluding plan-
ning and monitoring, have typically ranged from approx-
imately $1,000 to $7,600 per project, although some pro-
jects have been conducted for as little as $20 (with
substantial donations of labor and transportation costs).
Although operational cost data are lacking for automated
silos and limestone diversion wells, construction costs for
silos are substantial ($20,000), while the construction costs
for wells ($3,000) fall within the range of costs for limestone
barriers.
To determine the total cost of restorative and preven-
tative lake and stream liming, the costs for planning and
administering the project and monitoring the water quali-
ty must be incorporated. The addition of these costs could
easily exceed the material and application costs for many
projects.
SUMMARY AND CONCLUSIONS
Liming, as a technique to mitigate surface waters already
affected by acidic deposition or to protect waters poten-
tially sensitive to acidic deposition, has been conducted
in Scandinavia and North America. None of the countries,
States, or Provinces involved in liming advocate it as a
panacea for acidic deposition; however, it is considered
a viable, short-term, temporary measure to protect aquatic
life until atmospheric pollutants can be controlled at the
source.
It is very difficult to establish the costs of acidic
deposition-related liming from historic projects because
adequate records were rarely kept, and because many pro-
jects were undertaken on an experimental basis and do
not reflect the economics of scale associated with large
commercial operations. Based upon limited Scandinavian
and North American information, costs ranging from $100
(accessible sites) to $400 to $500 (remote sites) per sur-
face hectare appear to be reasonable approximations for
small-scale, experimental lake projects, excluding plan-
ning, administering, and monitoring costs. Approximations
for small-scale stream liming projects range from $1,000
to $7,600, excluding planning, administering, and monitor-
ing costs.
Based on the limited cost data, liming of aquatic systems
appears to be an economically viable strategy for contrbll-
ing the effects of acidic deposition. However, because of
the potential adverse effects and unknown long-term
ecological impacts resulting from liming (Bengtsson et al.
1980; Driscoll et al. 1982; Fraserand Britt, 1982; Fraser
et al. 1982; Menz and Driscoll, 1982; Van and Dillon, 1982),
this receptor abatement strategy may not be as ideal as
when examined purely on economics. Menz and Driscoll
(1982) recommended that the possible negative effects of
liming should be assigned a cost value and be incor-
porated into future economic analyses. To date, economic
analyses have not incorporated this additional cost fac-
tor, nor have adequate monetary values been estimated
for the sociological and ecological benefits resulting from
restoring acidified surface waters. Also, when consider-
ing the economics of chemical neutralization of aquatic
systems, neutralization costs should not be directly com-
pared to the costs of source control because liming only
counteracts the damage to aquatic life and does not ad-
dress the potential damage to terrestrial systems and
material structures.
REFERENCES
Andersson, F. 1981. Personal communication. Swedish Uni-
versity of Agricultural Sciences, Uppsala.
Arcs, R.G., and C.E. Boyd. 1975. Effects of agricultural lime-
stone on water chemistry, phytoplankton productivity, and fish
production in soft water ponds. Trans. Am. Fish. Soc. 104:3-312. '
Arnold, D.E. 1981. An experimental limestone flow-through
device for maintenance of pH in poorly-buffered streams. Poster
session presented at Acid Rain/Fisheries Symp. Ithaca, N.Y.
U.S. Fish Wlldl. Sen/., Fish. Oceans Can. August 2-5.'
1982. Personal comm. Pennsylvania Cooperative Fishery
Research Unit, University Park.
146
-------
Fiscal and Institutional Support for Lakes
Bengtsson B. 1982a. National Swedish liming project. Paper
presented at Int. Liming Workshop. Friday Harbor Lab. Wash.
Sept. 19-25.
1982b. Acid rain mitigation strategies in Sweden. Water
Air Soil Pollut. (Submitted.)
1982. Personal comm. National Swedish Board of Fish-
eries, Goteborg.
Bengtsson, B., W. Dickson, and P. Nyberg. 1980. Liming
acid lakes in Sweden. Ambio 9:34-36.
Blake, L. 1981. Liming acid ponds in New York. New York
Fish Game J. 28:208-214.
Booth Aquatic Research Group Inc. 1982. Experimental lake
neutralization study design. Toronto, Ontario: in cooperation
with Ontario Ministry Nat. Resour. and Ontario Ministry Environ.
Boyd, C.E. 1974. Lime requirements of Alabama fish ponds.
Auburn Univ. Agric. Exp. Sta. Bull. 459. Auburn, Ala.
_. 1976. Liming farm fish ponds. Auburn, Ala. Auburn Univ.
Agric. Exp. Sta. Leaflet 91.
Canadian Federal/Provincial Task Force. 1982. Report on the
protection and rehabilitation of aquatic systems from acid rain.
ChubbuckD., and W. Patterson. 1981. An assessment of the cost
effectiveness of lake liming in Ontario. Toronto, Ontario; On-
tario Hydro, Environ. Stud. Assess. Dep. Rep. No. 81341.
Driscoll, C.T. 1982. Personal comm. Dep. Civil Eng. Syracuse
University, Syracuse, N.Y.
Driscoll, C.T., J. R. White, G.C. Schafran, and J. D. Rendall. 1982.
CaCO3 neutralization of acidified surface waters. J. Environ.
Eng. Div. Am. Soc. Civ. Eng. 12:(in press).
Fraser, J.E., and D.L. Britt. 1982. Liming of acidified waters: a
review of methods and effects on aquatic ecosystems.
Kearneysville, W. Va. U.S. Fish Wildl. Serv. Off. Biolog. Serv.
FWS/OBS-80/40.13.
Fraser, J.E., etal. 1982. Feasibility study to utilize liming as a tech-
nique to mitigate surface water acidification. Palo Alto, Calif.
Electric Power Res. Inst. EPRI EA-2362.
Goff, T.R., F.S. Baker, and D.L. MacDonald. 1981. Improved
survival of Atlantic salmon parr by limestone filtration of an
acidic hatchery water supply. Halifax, Nova Scotia. Fish.
Oceans Canada.
Guthrie, R. 1982. Personal comm. Rhode Island Div. Fish Wildl.,
West Kingston.
Gutschick, K. 1982. Personal comm. Natl. Lime Assoc. Arling-
ton, Va.
Hasler, A.D., O.M. Brynildson, and W.T. Helm. 1951. Improving
conditions for fish in brown-water bog lakes by alkalization. J.
Wildl. Manage. 15:347-352.
Jackson, R.R. 1981. Establishment of a trout fishery by liming
in a northern Wisconsin bog lake. Paper presented at 43rd
Midwest Fish and Wildlife Conf. Wichita, Kans. December 6-9.
Johannessen, M., and R. Wright. 1982. Liming in Norway. Paper
presented at Int. Liming Workshop. Friday Harbor Lab. Wash.
Sept. 19-25.
Johnson, W.E., and A.D. Hasler. 1954. Rainbow trout production
in dystrophic lakes. J. Wildl. Manage. 18:113-134
Kitchell, J.A., and J.F. Kitchell. 1980. Size-selection predation,
light transmission, and oxygen stratification: evidence from the
recent sediments on manipulated lakes. Limnol. Oceanogr.
25:389-402.
Kretser, W. 1982. Personal comm. New York State Dep. Environ.
Conserv. Bur. Fish. Ray Brook.
1982. Update: liming acidified waters of New York State.
Ray Brook, N.Y. Statement prepared by Dep. Environ. Con-
serv. for N.Y. State Assemblyman.
Menendez, R. 1982. Personal comm. West Virginia Dep. Nat.
Resour. Elkins.
Menz, F.C., and C.T. Driscoll. 1982. An economic evaluation
of limng to neutralize acidified Adirondack surface waters.
Water Resour. Res. (Submitted.)
National Fisheries Board of Sweden. 1982. Rad och riktlinjer
for kalkning av sjor och vattendrag; Rep. No. 1.
Oatis, P. 1982. Personal comm. Massachusetts Dep. Environ.
Qual. Div. Fish Wildl. Westboro.
Pearson, F. H., and A. J. McDonnell. 1975. Use of crushed lime-
stone to neutralize acid wastes. J. Environ. Eng. Div.
100:139-158.
Phares, D. 1982. Personal comm. West Virginia Dep. Nat.
Resour. Elkins.
Scheider, W., J. Adamski, and M. Paylor. 1975a. Reclamation
of acidified lakes near Sudbury, Ontario. Rexdale, Ontario.
Canada. Ontario Ministry Environ. Rep.
Scheider, W., B. Cave, and J. Jones. 1975b. Reclamation of acidi-
fied lakes near Sudbury, Ontario by neutralization and fertiliza-
tion. Rexdale, Ontario, Canada. Ontario Ministry Environ. Rep.
Scheider, W., and P.J. Dillon. 1976. Neutralization and fertiliza-
tion of acidified lakes near Sudbury, Ontario. Water Pollut. Res.
Can. 11:93-100.
Sheffy, T., et al. 1980. A review of acid deposition in Wiscon-
sin: recommendations for studying and solving the problem.
Wisconsin Dep. Nat. Resour. Madison.
Sills, J.B. 1974. A review of the literature on the use of lime
[Ca(OH)2, CaO, CaCO3] in fisheries. U.S. Fish Wildl. Serv. Div.
Pop. Reg. Res. FWS-LR-74-10. Natl. Tech. Inf. Serv. No.
PB-235-449.
Waters, T.F. 1956. The effects of lime application to acid bog
lakes in northern Michigan. Trans. Am. Fish. Soc. 86:329-344.
Waters, T.F., and R.C. Ball. 1957. Lime application to a soft-water,
unproductive lake in northern Michigan. J. Wildl. Manage.
21:385-391.
Watt, W.D. 1982. Personal comm. Fisheries and Oceans, Re-
source Branch, Halifax, Nova Scotia, Canada.
Yan, N.D., and P.J. Dillon. 1982. Experimental neutralization
of lakes near Sudbury, Ontario. Adv. Environ. Sci. Technol. (In
press.)
Yan, N., W. Scheider, and P. Dillon. 1977. Chemical and biological
changes in Nelson Lake, Ont. following experimental eleva-
tion of lake pH. Water Pollut. Res. Can. 12:213-231.
147
-------
THE ROLE OF CONSERVATION EASEMENTS IN LAKE MANAGEMENT
WILLIAM PRAY O'CONNOR
Project Coordinator
GORDON CHESTERS
Director and Professor
Water Resources Center
University of Wisconsin
Madison, Wisconsin
ABSTRACT
The discussion is designed to provide lake management organizations with information on the nature of
conservation easements and some of the legal, institutional and tax (income, estate, and property) con-
siderations involved in developing conservation easement programs. Certain aspects of conservation
easements are discussed using representative jurisdictions as examples; however, the survey is not in-
tended to be exhaustive. Instead, it serves as an introduction to a complex subject area, providing the
reader with an understanding of the issues involved and the kinds of lake management objectives that
conservation easements may help attain. Most States have enacted statutes—including the recently adopted
Uniform Conservation Easement Act—recognizing conservation easements and ensuring their long-term
enforceability. The donation of these partial interests in real property by conservation organizations is en-
couraged by specific provisions of the Internal Revenue Code permitting landowners to deduct from tax-
able income the value of qualifying easements granted to public or certain private organizations. Addi-
tional tax incentives for donation are available to landowners under Federal estate tax law and the death,
inheritance, income, and real property tax laws of many States. This device offers potential in lake manage-
ment for the enhancement and protection of water quality and for preservation of undeveloped or scenic
areas benefitting the recreational potential and aesthetic qualities of the lake and its watershed.
INTRODUCTION
A conservation easement is a legal means permitting a land-
owner to permanently limit the future use of lands by transfer-
ring the rights to specified uses of the land to a conserva-
tion organization. This device permits the owner to retain
private ownership of the land and permits the holding agency
to ensure that future uses of the land remain consistent with
its conservation objectives.
As legal interests in real property, conservation easements
can be sold and purchased and are transferred by written
deeds recorded in public registries. They may be used to
impose limitations on land use or to impose affirmative
obligations on the landowner.
The basic mechanism involved is a transfer from the owner
of part of the "bundle of rights" that constitute fee simple
(outright) land ownership. The specific rights transferred are
detailed in the written deed conveying the easement from
the landowner to the easement holder. The restrictions and
obligations transferred by the easement "run with the land,"
binding the original parties and their successors for the term
of the easement.
Easements have been used by public and private
organizations to accomplish conservation goals of various
kinds. For example, easements have been used to preserve
scenic areas, wild rivers, historic sites, and wildlife habitat;
to protect shorelands and coastal areas; to conserve pro-
ductive forest and agricultural lands; to enhance water quali-
ty; and to improve flood plain management (Table 1).
The specific terms of a conservation easement are tailored
to meet the objectives of the conservation organization and
the landowner. Some conservation easements impose
severe limitations on the use of the land, to ensure that it
will be maintained in its natural condition. Such an ease-
ment, sometimes referred to as a "preservation easement,"
might prohibit subdividing the land into smaller parcels and
restrict or prohibit cutting trees, constructing buildings or
roads, or other activities modifying the natural condition of
the land.
Unlike zoning or subdivision regulations adopted by a
government body pursuant to the police power, these restric-
tions apply only to specific parcels of land for which
easements have been granted by the landowner.
As a lake management device, conservation easements
offer potential in two broad areas. First, preservation
easements could be used to preserve undeveloped shore-
land areas or scenic views that contribute to the recreational
and aesthetic values of lake communities. An easement of
this kind typically involves transferring from the landowner
the rights to disturb vegetation or to construct buildings or
roads.
Conservation easements also could be used as a
nonstructural mechanism to reduce runoff or nonpoint
source pollution. An easement designed for this purpose
might prohibit land disturbing activities in areas of steep
slopes or highly erodable soils and further require the land-
owner to take positive steps to maintain vegetative cover
to minimize soil loss. As a nonstructural method of con-
trolling runoff, conservation easements can reduce runoff
and pollutant generation (Consad Research Corp., 1976).
Conservation easements generally have a high level of
political acceptability, because they do not rely on the ex-
ercise of the police power, but involve acquiring partial
interests in land for compensation. Because they involve
acquisition of only partial interests in land, conservation
easements usually can be obtained for substantially less
than outright land purchase. The transfer of conservation
easements also is generally acceptable to land owners,
148
-------
Fiscal and Institutional Support for Lakes
Table 1. — Examples of uses of conservation easements to
meet particular resource management objectives.
Scenic Areas
The Wisconsin Highway Commission acquired easements along the
"Great River Road," a National Parkway along the Mississippi River.
The easements restrict the density of buildings and prohibit com-
mercial development, billboards, and dumps and the removal of trees
or shrubs on the subject lands (Jordahl, 1963; Olson, 1965).
Historic Sites
The congressionally chartered National Trust for Historic Preserva-
tion has promoted efforts to preserve historic buildings and sites
(Watson, 1980). An increasingly common technique is using facade
easements to protect significant architectural features of historic
buildings by controlling modifications of a building's exterior ap-
pearance and requiring maintenance (Hambrich, 1981).
Wildlife Habitat
Region 2 of the U.S. Fish and Wildlife Service undertook the preser-
vation of nearly 300,000 acres of Texas coastal marsh in the late
1970's. The project involves extensive use of preservation
easements, designed to protect the natural wintering habitat of
migratory waterfowl (Shelton and Hart, 1977).
On Squam Lake, New Hampshire, the privately organized Squam
Lakes Association formed the Squam Lakes Conservation Founda-
tion, a tax-exempt organzation which has received donations of
easements, to protect open spaces, islands and undeveloped shore
areas, and to preserve the habitat of the lake's summering loon
populations (Taylor, 1982).
Wild Rivers
The Nature Conservancy, a private, nonprofit national organization,
has acquired donations of easements protecting 9 miles of shoreline
(and nearly 5,000 acres of land) on the unspoiled Brule River in north-
ern Wisconsin. The land provides habitats for the endangered, tall
white bog orchid and some 90 bird species including osprey and
bald eagles and includes relic stands of white pine and cedar. The
easements require the lands to be maintained in substantially natural
conditions, prohibiting expansion of existing buildings and roads.
Many of the parcels include summer homes the owners continue
to use (Nature Conserv., 1982).
Coastal Areas
The Maine Coast Heritage Trust, a private, nonprofit organization,
conducts an active program facilitating the transfer of easements
along the Atlantic Coast to the National Park Service. Some of the
easements protect buffer areas of the Acadia National Park. The
Trust has recently begun to acquire easements on its own and to
assist land conservation throughout Maine (Hartwell, 1982).
Agricultural Lands
Public and private easement programs to protect rangelands,
croplands and dairy farms from conversion to residential and com-
mercial development have been established in California, Colorado,
Massachusetts, Montana, New York, Vermont, and other States.
For example, the Massachusetts Farm and Conservation Lands
Trust, a private, nonprofit organization has established a Farmlands
Revolving Fund and bank lines of credit to finance acquisitions of
easements (called preservation restrictions) on important tracts of
agricultural land threatened with conversion to nonfarm uses. Most
of the easements will be transferred to the Commonwealth,
municipalities and local land conservation organizations (Cherington,
1982).
Forests
The Private Society for the Protection of New Hampshire Forests
has acquired more than 44 easements covering some 9,000 acres
of forest land since 1974. The easements are designed to encourage
forest management and preserve productive uses of forest land
(McClure, 1982).
Water Quality Management
The private Brandywine Conservancy was organized after severe
flooding in the 1960's to provide long-term water management in
the 350-square mile Brandywine watershed in Eastern Pennsylvania
and Delaware. The Conservancy has acquired easements on
thousands of acres of woodlands, groundwater recharge areas, shore
areas, and highly erodable lands (Sellers, 1982).
Flood Plain Management
Conservation easements have been used in flood plain manage-
ment to provide rights for public bodies to install and maintain flood
control structures and to prevent disturbance of soil-holding
vegetative cover within flood prone areas (Ralph M. Field Associates,
1981).
because they are not required to give up ownership of the
land or permit public access to it. Sometimes, however,
public access is granted as an element of the easement.
Why would a landowner want to convey an easement?
Who would receive the easement? How would the ease-
ment be enforced? What effect would an easement
transfer have on the property tax levied on the subject
land? These are some of the questions to be addressed
here.
CONSERVATION EASEMENTS:
RECOGNITION AND ENFORCEABILITY
The Anglo-American system of property law has long
recognized the fundamental utility of easements.
Easements granting rights of ingress and egress across
another's land are commonplace. Similarly, easements to
permit the installation of electrical, telephone, and other
utilities often are found in abstracts of title to property.
Although litigation concerning the extent and validity of
these easements is not uncommon, their recognition has
posed no particular difficulty. Conservation easements,
however, have raised a variety of legal concerns (Nether-
ton, 1979; Brenneman, 1967). These have chiefly involv-
ed the enforceability of "easements in gross" (those not
benefitting an adjoining parcel); the assignability of ease-
ment interests from the original holder; and the en-
forceability of easements against the successors in interest
of the landowner originally agreeing to the easement
restrictions (Kinnamon, 1980; Netherton, 1979; Roe, 1976).
A particular problem has involved the repugnance of
"novel" interests in real property to common law judges.
In the landmark 19th century English case of Keppel v.
Bailey, Lord Chancellor Brougham, in an oft-quoted dic-
tum stated that:
There are certain known incidents to property and its en-
joyment; among others, certain burthens wherewith it may
be affected, or rights which may be created and enjoyed
over it by parties other than the owner; all which incidents
are recognized by the law . . But it must not be suppos-
ed that incidents of a novel kind can be devised and at-
tached to property, at the whim and caprice of any owner.
Keppel v. Bailey, 2 Myl. & K. 517, 535, 39 Eng Rep 1049
(1834).
149
-------
Lake Restoration, Protection and Management
Although this view no longer reflects the law of England
or most of the States, it has caused property lawyers to
regard typical conservation easements with apprehension.
As a result, legislation has been introduced in many States
specifically recognizing conservation easements and pro-
tecting them from invalidation by the courts. To date, more
than 40 States have adopted some kind of statutory pro-
visions relating to conservation easements (Netherton,
1979). These statutes vary widely in (1) their definitions
of conservation easements; (2) the entities entitled to hold
them; (3) their treatment of easement recording, duration,
and enforceability; and (4) their provisions—in some
cases—for required governmental approvals prior to a
valid easement transfer.
In 1981 the National Conference of Commissioners of
Uniform State Laws adopted and recommended for
passage by all States a Uniform Conservation Easement
Act. The Act was approved by the American Bar Associa-
tion in January 1982. So far the Act has been adopted only
in Wisconsin (Ch. 261, Wisconsin Laws of 1981), but it
represents the most comprehensive effort to address the
property law issues involved and is similar in many
respects to the laws of many States. For purposes of this
discussion, the Uniform Act serves as the chief example
of conservation easement statutes, although the laws of
several States will be referred to for comparison.
CONSERVATION EASEMENTS
AND HOLDERS
The Uniform Act defines a conservation easement as:
A nonpossessory interest of a holder in real property im-
posing limitations or affirmative obligations the purposes
of which include retaining or protecting natural, scenic, or
open-space values of real property, assuring its availability
for agricultural, forest, recreational, or open-space use, pro-
tecting natural resources, maintaining or enhancing air or
water quality, or preserving the historical, architectural, or
cultural aspects of real property.
Uniform Conservation Easement Act, Sec. 1(1).
Typically, the terms of conservation easements are
negative—they impose restrictions, obligations, or limita-
tions on land uses—rather than granting affirmative rights
to the easement holder. By contrast, a typical access or
driveway easement is an affirmative easement, granting
the holder rights of access across land which would other-
wise constitute trespass. Because conservation
easements restrict uses, the holder has the important
responsibility of ensuring compliance with the restrictions.
Because this duty requires the holder to monitor use,
periodic access to the parcel (an affirmative right) usually
is included in the easement document.
Because easements most commonly are granted "in
perpetuity," the holder must logically be an entity with
"perpetual" life, such as a government body or a corpora-
tion. Under the Uniform Act, the holder of a conservation
easement must be either a government body empowered
to hold interests in real property or a nonprofit corpora-
tion authorized to acquire interests in real property and
organized for purposes that include conservation (Uniform
Conservation Easement Act, Section 1(2)).
Conservation easements have been acquired by several
Federal agencies including the National Park Service, the
U.S. Forest Service, and the Fish and Wildlife Service.
In most States, such general purpose local governments
as counties, cities, villages, and towns are authorized to
hold interests in real property. The laws of several States
specifically authorize acquisition of easement interests for
specified purposes (Netherton, 1979). Washington State,
for example, grants specific statutory authority to local
governments to "acquire lands and easements within
shorelines when necessary to the implementation of
master programs" for shorelines of statewide significance
(Ch. 90.58.240 (1), Washington Rev. Code. Ann.).
Furthermore, some special purpose local governments
can acquire easements. In Wisconsin, enabling statutes
provide for the formation of Inland Lake Protection and
Rehabilitation Districts (Section 33.11 et seq., Wis. Stats.).
These lake districts are granted the broad power to under-
take lake management activities. More specifically,
Wisconsin lake districts "may sue and be sued, make con-
tracts, accept gifts, purchase, lease, devise or otherwise
acquire, hold, and dispose of property, disburse money,
contract debt" and perform other activities to meet their
statutory purposes (Section 33.22, Wis. Stats.). Although
no Wisconsin lake district has yet acquired a conserva-
tion easement, the authority is reasonably clear.
Minnesota statutes also provide for the creation of
special purpose "lake improvement districts" (Minn. Stat.
Ann. Sec. 378.41(3)). These districts are granted powers
to acquire land for certain purposes, "except the power
to acquire property by eminent domain." (Minn. Stat. Ann.
Sec. 378.41(3)). The Wisconsin Attorney General, Bran-
son La Follette, in a letter to Representative Calvin Potter
dated November 24, 1981, concluded that the governing
body of a Wisconsin lake district "has the power to ac-
quire property by condemnation
Examples of other special purpose governments, many
of which are authorized to acquire interests in real pro-
perty, are Soil and Water Conservation Districts, Water-
shed Districts, Conservation Districts or Commissions, and
Sanitary Districts.
Of course, an examination of the legal powers of any
unit of government would be an essential precondition to
its undertaking a conservation easement program.
In addition to government bodies, numerous private
organizations are qualified to hold conservation ease-
ments. These private entities range from comparatively
informal lake associations to more sophisticated nonprofit
organizations recognized as exempt from Federal income
tax under Sec. 501 (c) (3) of the Internal Revenue Code.
At the national level, the Nature Conservancy, the National
Audubon Society, the National Trust for Historic Preser-
vation, the American Farmland Trust, and the Trust for
Public Lands have programs of conservation easement
acquisition.
Also, numerous State and regional private organizations
currently hold conservation easements, including the
Society for the Protection of New Hampshire Forests, the
Trustees of Reservations (Massachusetts), the Outta-
quechee Regional Land Trust (Vermont), the Iowa Natural
Heritage Foundation, the Maine Coast Heritage Trust, the
Virginia Outdoors Foundation, the Montana Land Re-
liance, and many others. At the local level some 450
private land trusts now exist throughout the United States,
e.g., the Geneva Lake Land Conservancy, originated to
protected the lands and waters of the Lake Geneva Water-
shed in Walworth County, Wis.
DEVELOPING A CONSERVATION
EASEMENT PROGRAM
The initial step in developing a conservation easement pro-
gram is to identify a need for long-term control of lands
significantly affecting the lake environment. An initial deter-
mination must be made that management is required for
a particular lake or watershed and steps must be taken
to organize a lake management effort. Assuming such
lands are identified, State property law must be reviewed
to determine the status of conservation easements. If State
150
-------
Fiscal and Institutional Support for Lakes
law is compatible the next step is to identify and choose
an existing or prospective institution qualified to hold con-
servation easements. Depending on local circumstances
a particular public or private agency may be best suited
to attract the support of prospective easement grantors
and the public at large.
The proposed easement holder must develop a plan for
monitoring lands subject to easements and put into place
a mechanism for ensuring that restrictions are enforced
(Abbott, 1977). Some easement holders take aerial
photographs at the time the easement is obtained and con-
firm compliance with easement restrictions by periodic on-
site inspection or additional aerial photography (McClure,
1982),
Under the Uniform Act, the obligations imposed by an
easement can be enforced by judicial remedies. The Act
also permits provision in easements granting enforcement
rights to third parties (Unif. Conserv. Easement Act, Sec.
3(a)). For example, an easement to a local conservation
organization also may provide for the right of a State or
regional agency or private organization to enforce the
easement restrictions.
Once the Institutional development is complete, negotia-
tions to acquire easements can begin. Because ease-
ments raise many complicated issues, it is essential that
the proposed holding entity obtain the services of a
negotiator thoroughly versed in the property and tax Im-
plications of an easement transfer.
VALUATION OF EASEMENTS
The valuation of conservation easement interests is not
a wholly perfected science. Easement valuations bear lit-
tle uniformity. In part this is because there are not signifi-
cant numbers of comparable property sales involving lands
subject to easement restrictions.
Dennis (1982) has stated that the value of an easement
prohibiting development of the subject land often falls in
the range of 60 to 80 percent of the fair market value of
the land, although most State and Federal agencies have
rejected fixed schedules to determine easement values
(Shelton and Hart, 1977). Some writers have suggested
that the grant of an easement may actually increase the
value of certain land by preserving its natural character,
although most appraisers and agencies concede that land
value decreases when its potential uses are limited by an
easement (Shelton and Hart, 1977; Hambrich, 1981).
There is general agreement that the valuation of an
easement should be based on the difference in land values
before and after the imposition of the easement restric-
tions (Hambrich, 1981; Shelton and Hart, 1977; Browne
and VanDorn, 1975.). However, approaches differ in deter-
mining "after" value. One approach is capitalization, e.g.,
the capitalized value of an easement on forest lands (pro-
hibiting subdivision or commercial or industrial uses and
limiting future construction) would be based on the poten-
tial income from timber (Hallet, 1982). A capitalized value
for and easement restricting cultivation of nearshore areas
could be calculated based on the value of productive
agricultural uses of the land, discounted to a present value.
A preservation easement is valued on the basis only of
permitted uses, so the value of the land for development
purposes is not included in its valuation.
A National Task Force on the Valuation of Conserva-
tion Easements was organized in 1982 under the auspices
of the Land Trust Exchange in Boston to research methods
of easement valuation and to develop and identify ways
to standardize methodology and documentation (Land
Trust Exch., 1982).
INCOME TAX CONSIDERATIONS
Most land conservation organizations rely to some extent
on gifts — rather than purchases — of conservation lands.
An outright gift of a conservation easement, or a sale at a
discounted price (a "bargain sale"), can provide significant
tax benefits to the landowner (Daugherty, 1978; Hambrich,
1981; Kinnamon, 1980; Small, 1979). These benefits result
primarily from provisions of the Internal Revenue Code pro-
viding a deduction from taxable income for donations made
to a government body or to a public charity recognized as
tax-exempt under Section 170 of the Internal Revenue Code.
This incentive often makes it possible for conservation
organizations to acquire lands or easements for substantially
less than their actual value.
Generally, Federal tax law does not permit a charitable
deduction for gifts of partial interests in real property.
However, P.L. 96-541, enacted Dec. 17,1980, makes an ex-
ception to the rule, permitting deduction of "qualified con-
servation contributions." Such a contribution must consist
of a gift of a "qualified real properly interest," to a "qualified
organization," "exclusively for conservation purposes" (I.R.C.
Section 170(h)(5)(A)).
A qualified real property interest is defined to include "a
restriction (granted in perpetuity) on the use which may be
made of the real property" (I.R.C. Section 170(h)(2)(c)).
Qualified conservation organizations include governmental
units and certain tax-exempt "public charities." Conserva-
tion purposes are limited to the following:
(i) the preservation of land areas for outdoor recreation by,
or the education of, the general public;
(ii) the protection of a relatively natural habitat of fish, wildlife,
or plants, or similar ecosystem;
(iii) the preservation of open space (including farmland and
forest land) where such preservation is . .
(I) for the scenic enjoyment of the general public,
or;
(II) pursuant to a clearly delineated federal, state, or local
governmental conservation policy, and will yield a signifi-
cant public benefit . . I.R.C. Section l70(h)(4)(A)
As of this writing, regulations interpreting these provisions
— while long overdue — have not been adopted by the
Treasury Department. The interpretations of such terms as
"relatively natural habitat," "significant public benefit," and
"clearly delineated" governmental conservation policy have
generated considerable discussion within the tax service and
among tax practitioners.
The Senate Finance Committee's report on the new
statute states that the scenic enjoyment requirement is met
by either physical access of the general public or "visual
access," such as the view of a scenic shoreland area from
a public lake. Grants of easements on lands without public
scenic values qualify for deduction if made pursuant to a
clearly delineated governmental conservation policy. Accor-
ding to the Committee, this requirement would be met by
certain Federal executive orders or by State statutes or local
ordinances establishing conservation programs or projects
that are either funded or involve a significant commitment
by the government unit.
Further, a deductible open-space easement must yield a
significant public benefit. Among the factors identified by the
Committee for use in determining the significance of public
benefits provided by a conservation easement are the "uni-
queness" of the property, the intensity of nearby develop-
ment, and the consistency of the proposed easement use
with public programs for "water supply protection, water
quality maintenance and enhancement, flood prevention and
control, [and] shoreline protection" (96th Cong. 2d Sess.,
Sen. Rep. 96-1007, 9-14).
151
-------
Lake Restoration, Protection and Management
Under the Internal Revenue Code, a taxpayer is limited
to deducting not more than 30 percent of his "charitable con-
tribution base" (generally his adjusted gross income) in a
single tax year, although a special election may be made
permitting deduction of up to 50 percent. Amounts exceeding
these limitations may be "carried over" into as many as 5
succeeding tax years (Daughterly, 1978).
In situations where the landowner makes a partial gift of
the easement, but requires some cash compensation, a
"bargain sale" may be made. A taxpayer is permitted to
deduct the value of a conservation easement equal to the
extent he was not compensated. For example, if an ease-
ment is valued at $10,000 but the seller transfers it to a con-
servation organization for $4,000, the seller may claim the
remaining $6,000 as a charitable deduction. An outright gift
of the easement would allow the landowner the full $10,000
value of the easement as a deduction.
ESTATE AND DEATH TAX
CONSIDERATIONS
The transfer of a conservation easement also affects State
and Federal taxes imposed on the transfer of property at
death. These taxes are imposed on the gross value of pro-
perty owned by the decedent, subject to certain credits and
deductions. Lands subject to conservation easements are
valued for this purpose at their reduced "after" value.
With the increase in the minimum estate subject to Federal
Estate Tax adopted in the Economic Recovery Tax Act of
1981, the significance of Federal estate tax benefits applies
only to individuals of substantial wealth. Estates valued at
less than*$600,000 will pay no Federal estate tax beginning -
in 1985. The minimum taxable amount increases from a cur-
rent $225,000 to the $600,000 level in annual increments.
However, for those estates large enough to incur estate tax
liability, the transfer of conservation easements can be of
significant benefit. The tax rate on such estates begins at
30 percent.
Although the'level of taxation is much lower, State death
or inheritance taxes are based on valuation of the de-
ceased's assets. Therefore, a parcel of land subject to a con-
servation easement usually would qualify for reduced in-
heritance tax.
PROPERTY TAX CONSIDERATIONS
The property tax implications of transferring a conservation
easement vary widely among States. Theoretically, the
assessed valuations upon which property tax levies are
based consider all circumstances influencing the value of
the parcel. Because conservation easements can impose
significant (and often permanent) restrictions on land use,
a parcel subject to such an easement may be assessed at
a lower value.
Some State statutes expressly direct property tax
assessors to take the value of conservation easements into
account when assessing lands subject to easements (Va.
Code Sec. 10-155; Mont. Code Ann. Sec. 76-6-208). Some
statutes require reduction in the assessment only if the ease-
ment is held by a government body (Section 70.32(1), Wis.
Stats.; Conn. Gen. Stat. Sec. 7-131b(b)). In other States an
easement transfer does not reduce property taxes, the land
title holder assuming the full burden of property taxation.
Where State laws do not provide property tax relief for pro-
perties subject to conservation easements, private contrac-
tual arrangements are sometimes made between easement
grantors and holders to apportion the tax levied.
Several States, although they may not provide for a reduc-
tion in the assessed valuations of lands subject to conser-
vation easements, do provide property tax reductions or
credits for certain open-space, forest or agricultural lands.
These programs typically require the landowner to agree to
keep the land in the qualified use for a term of years and
provide penalties if the lands are withdrawn from the pro-
gram before the agreed term. Under the laws of some States,
reduced property rates are available for certain types of lands
(open-space areas, forests, wetlands, or farms) or to en-
courage soil conservation practices. Under some of these
programs, a conservation easement may qualify the land for
favorable property tax treatment (Dunford, 1980; Massey and
Silver, 1981).
SUMMARY
Conservation easements provide a mechanism for per-
manently restricting land use for conservation purposes. Most
States have statutes recognizing conservation easements
and providing for their enforcement. This tool can be used
by lake management organizations seeking to preserve
relatively natural shoreland areas and to provide parcel-
specific land-use management in areas requiring control of
nonpoint sources of water pollution.
In considering a conservation easement program, a lake
management organization should consider whether conser-
vation easements are recognized under State law and
whether an entity exists or can be created to hold conser-
vation easements. Easement holding organizations require
systematic methods for monitoring land activity on land sub-
ject to easements and mechanisms to enforce compliance
with easement restrictions. Because easement programs re-
quire significant technical know how, a trained staff is a must.
To maximize the potential tax benefits to landowners mak-
ing gifts of conservation easements, a holding organization
must be a governmental unit or a qualified, tax-exempt public
charity.
A gift — or a bargain sale — of a conservation easement
produces a valuable income tax deduction for the donor. In
addition, the transfer of a conservation easement can pro-
vide significant estate and death tax benefits to the land-
owner's estate. In some States, the transfer of a conserva-
tion easement can also reduce the assessed valuation of
the land, thereby lowering the real property tax.
REFERENCES
Abbott, G. 1977. Conservation restrictions: Procedures for administra-
tion and inspection. Trustees of Reservations, Milton, Mass.
Brenneman, R. 1967. Private approaches for the preservation of
open land. Conserv. Res. Found., New London, Conn.
Browne, K., and W. Van Dorn, 1975. Charitable gifts of partial
interests in real property for conservation purposes. Tax Lawyer
29:75.
Cherington, D. 1982. Background information and operating proce-
dures. Farm Conserv. Lands Trust, Beverly, Mass.
Consad Research Corp. 1976. Evaluation of the cost-effectiveness
of non-structural pollution controls: A manual for water quality
management planning. Contr. No. 68-01-2699. U.S. Environ. Prot.
Agency, Washington, D.C.
Dennis, R. 1982. Consensus can put programs back on track.
Exchange 1(3).
Daugherty, A. 1978. The economics of Federal tax incentives for
conservation easement donations. Natl. Tax J. 30:171.
Dunford, R. 1980. A survey of property tax relief programs for the
retention of agricultural and open space lands. Gonzaga Law Rev.
15:675.
Hallet, A. 1982. Capitalization key to appraising "after" value.
Exchange 1(3).
Hambrich, K. 1981. Charitable donations of conservation easements:
Valuation, enforcement, and public benefit. Taxes 59:347.
152
-------
Rscal and Institutional Support for Lakes
Hartwell, D. 1982. Executive Director, Maine Coast Heritage Trust,
Northeast Harbor, Maine. Pers. comm.
Jordahl, H. 1963. Conservation and scenic easements: An experi-
ence resume. Land Econ. 39:343.
Kinnamon, D. 1980. Tax incentives for sensible land use through
gifts of conservation easements. Real Prop. Probate Trust J. 15:1.
Land Trust Exchange. 1982. Valuation of conservation easements.
Land Trust Exchange, Boston, Mass.
Massey, D., and M. Silver. 1981. Property tax incentives for imple-
menting soil conservation programs under constitutional taxing
limitations. Denver Law J. 59(3).
McClure, J. 1982. Program Manager, Soc. Prot. New Hampshire
Forests, Concord, N.H. Pers. comm.
Nature Conservancy. 1982. Annual report. Arlington, Va.
Netherton, R. 1979. Environmental conservation and historic preser-
vation through recorded land use agreements. Real Prop. Pro-
bate Trust J. 14:540.
Olson, J. 1965. Progress and problems in Wisconsin's scenic and
conservation easement program. Wis Law Rev. 65:352.
Ralph M. Reid Associates. 1981. State and local acquisition of
floodplains and wetlands: a handbook on the use of acquisition
in floodplain management. U.S. Water Resour. Counc.,
Washington, D.C.
Roe, C. 1976. Innovative techniques to preserve rural land re-
sources. Environ. Affairs 7:419.
Sellers, W. 1982. Executive Director. Brandywine Conservancy,
Chadds Ford, Pa., Pers. comm.
Shelton, K. and R. Hart. 1977. Suggested methods and techniques
for appraising conservation easements. Reg. 2, U.S. Rsh Wildl.
Serv., Albuquerque, N.M. (Unpubl. mss.)
Small, S. 1979. The tax benefits of donating easements in scenic
and historic property. Real Estate Law J. 7:304.
Taylor, M. 1982. President, Squam Lakes Assn., Ashland, N.H.,
Pers. comm.
Watson, E. 1980. Establishing an easement program to protect
historic, scenic, and natural resources. Inf. Sheet No. 25. National
Trust for Historic Preservation, Washington, D.C.
153
-------
LAKE PROTECTION BY WATERSHED MANAGEMENT FOR
WISCONSIN LAKE DISTRICTS
GEORGE R. GIBSON, JR.
Environmental Resources Unit
University of Wisconsin-Extension
RICHARD E. WEDEPOHL
DOUGLAS R. KNAUER
Inland Lake Renewal Section
Wisconsin Department of Natural Resources
DEFINITION OF THE CONCEPT
Protection districts are those lake districts recognized as hav-
ing high quality lake waters, low nutrient, or contaminant
loadings and few developmental or other potentially
degrading characteristics in the watershed, i.e., oligotrophic
or low mesotrophic lakes. Protection districts are also those
lake districts that have completed rehabilitation projects and
should now concern themselves with activities designed to
protect, preserve, and enhance the lake quality im-
provements achieved; in many instances these are high
mesotrophic or eutrophic lakes.
BACKGROUND REVIEW
The Inland Lake Protection and Rehabilitation Program in-
itiated by passage of Chapter 33 of the Wisconsin Statutes
in 1974, began citizen-involved lake management in Wiscon-
sin. Basically, it consists of organizing inland lake com-
munities into local, special purpose units of government
called lake management districts. These are supported by
State technical and financial assistance and University
technical assistance. The individual lake districts, however,
exercise considerable autonomy in determining project
implementation.
In the formative years, the program Advisory Council, in
providing management strategy direction, established with
the Inland Lake Renewal (ILR) Section of the State Depart-
ment of Natural Resources, a format for initially emphasiz-
ing lake rehabilitation projects. That is, placing the greatest
funding and management priorities on the visible improve-
ment of the more degraded and eutrophic lakes. The Council
advised that after 5 years of such emphasis, the program
should then shift attention to accommodate the needs of
higher quality lakes to protect and preserve their existing
condition.
That time interval has elapsed, and in recent years it has
become evident that the program has indeed evolved toward
a greater need for protection management just as projected.
This report describes this protection which emphasizes
watershed management. The concept is in the early stages
of application even as it continues to evolve.
In any community-oriented resource management pro-
gram such as this, organization and recruitment are extreme-
ly important at the outset. A sufficient number of manage-
ment clients, i.e., lake districts, must exist before the ad-
ministrative system of management technology and funding
can get rolling. Once organized, the next big hurdle is deliver-
ing the goods, developing workable investigative procedures,
meaningful data interpretations, and cogent management
recommendations. While a certain amount of applied
research continues to keep technological and management
advice current and reliable, this second phase of the pro-
gram is well-established in Wisconsin.
The program is now in a third phase—refinement and
maintenance of the system. A key component of this refine-
ment is increased attention to protection-oriented lake
districts—largely because the number of essentially
protection-motivated districts forming has grown since 1974.
These districts have been relatively neglected in favor of
more pressing, more rewarding (in terms of techniques
developed and information gained), and more expensive
management projects with eutrophic lake systems. As the
eutrophic lakes are rehabilitated, they, too, are becoming
candidates for subsequent protection-oriented assistance to
assess the success of rehabilitation and to protect gains
made—often at considerable expenditure of funds and labor.
Protection districts are receiving more emphasis from the
program not only because it is their just due as defined by
Chapter 33—the Public Inland Lake Protection and
Rehabilitation Law—and by the Advisory Council in its defin-
ed policies—but also because protection management is
economical in a time of fiscal constraints. Such attention to
low cost management options is essential to most of Wiscon-
sin's 130-plus lake districts. Further, this emphasis on lake
district initiative preserves the concept of lake resource
management through the direct involvement of the local
community.
MANAGEMENT SERVICES PROVIDED TO
PROTECTION LAKE DISTRICTS
The hallmarks of protection management are expanded
perceptions of the lake protection management philosophy
and emphasis on technical information and education. The
expanded management perception involves activities extend-
ing beyond in-lake and eutrophication abatement manage-
ment. The information-education function has had to expand
because most responsibility for protection falls to the peo-
ple of the districts themselves. We have to provide more
resource information to the district members so they in turn
may exert responsible influence within their communities to
achieve the necessary ordinances, planning commitments,
and prescribed land use conservation/lake protection
activities.
Protection Management Elements
Land Use Planning and Regulation. The district should
clearly define the lake watershed. This is usually delineated
in the initial program-sponsored feasibility study, but a con-
siderably more detailed assessment of runoff patterns, ero-
154
-------
Fiscal and Institutional Support for Lakes
sion sites, and loading factors may be needed. Assistance
in reviewing the watershed data and determining land areas
of particular significance has been provided by the local Soil
Conservation Service (SCS), the Department of Natural
Resources (DNR), the Land Conservation Committee (LCC)
and the University of Wisconsin-Extension (UWEX). ILR and
University staff familiar with the original feasibility studies
have been essential sources of assistance at this point.
Following the review and assessment of the watershed
characteristics, the district may seek additional help from the
county or regional planning office to prepare a list of land
use management priorities: specific parcels for attention, and
approaches to take, i.e., cooperative involvement with land-
owners, easements, land purchases, rentals, or proposal of
regulations and ordinances. Local and regional planning
agencies must be included at some point to coordinate water-
shed and district functions with overall regional plans and
actions, especially where zoning, construction, and lake use
ordinances are concerned.
Weed and Algae Control. Lake districts routinely deal with
this problem now, generally with sole financial responsibili-
ty for it. Prior State and Federal Policy prevented cost shar-
ing because it is viewed as cosmetic in nature. This rationale
is best defended when applied to lake districts just organiz-
ing and which are clearly in need of rehabilitation. Lakes with
a rehabilitation program already underway or those that are
best managed by protection in the first place, deserve finan-
cial assistance with such symptomatic management.
Recent amendments to Wisconsin's Inland Lake Law now
permit public cost-share grants for weed and algae control,
and a policy to implement this is being developed. When
the district is already fully engaged in abating the eutrophica-
tion problem at its source (by rehabilitation and/or protec-
tion activities), then State financial assistance to help improve
the appearance and usability of the lake is indicated. Fur-
ther, such assistance in symptomatic treatment may be
justified on the theoretical, if not yet documented, presump-
tion that weed harvesting enhances nutrient removal from
the lake.
Fishery Management. Through the cooperation of the
local DNR fish biologist, the lake fish population can be
periodically sampled. Results of these studies indicate ad-
justments to stocking, fishery effort, and habitat manipula-
tion components of a continuing fish management program.
The advice of the fish biologist is also helpful in balancing
the extent and methods of a weed and algae control pro-
gram to accommodate a healthy fishery.
In some instances, the DNR wildlife biologist may also be
involved, i.e., where shoreline or marsh environments are
elements of a lake management proposal.
Lake Water Quality Monitoring Program. Highly perti-
nent to either form of protection district is the regular monitor-
ing of lake water quality conditions. The information provid-
ed by this data is necessary to assess a completed restora-
tion project and/or to indicate developing water quality pro-
blems before they get out of hand.
The envisioned monitoring program will have to be tailored
to the characteristics of each lake, but generally is viewed
as a monthly or quarterly sampling scheme. Selected
parameters should be (a) easily measured or collected for
laboratory analyses by lake district members; (b) pertinent
to baseline data comparisons, e.g., the feasibility study, DNR
Bureau of Research data, or University investigations; and
(c) acceptable as indicators of changing lake conditions.
Measurements being considered are dissolved oxygen pro-
file(s) temperature profile(s), pH and/or alkalinity (especial-
ly in acid rain sensitive areas), Secchi disc readings, total
phosphorus, fecal coliform bacteria from beach areas dur-
ing swimming season, and lake level readings. On an infre-
quent, as-indicated basis, sediment surveys and weed
surveys might be conducted.
An important factor considered here is keeping costs low
because eventually all lake districts should become in-
volved in monitoring. Private consultant involvement should
be minimal. A standard data reporting sheet can be de-
signed for use by all districts. They would periodically mail
these in to ILR for review and assessment with an interpreta-
tion returned to participating districts. The monitoring pro-
gram will be simple, inexpensive, of value to each lake
district, and collectively should add to the statewide lake
quality information base.
Lake and Lakeshore Use Management Activities. Most
of these activities will entail primarily technical assistance
from Inland Lake Renewal and University-Extension staff with
actual performance by the district. However, in some in-
stances, financial assistance may be appropriate. Examples
of some anticipated activities include: shoreline erosion con-
trol, establishment or preservation of a vegetative buffer zone
around the lake, public access maintenance, beach manage-
ment, and lake use management or regulation.
Lake use management in particular will probably require
considerable local political involvement by the district before
much can be accomplished. Lake districts do not have
regulatory powers and such authority is a prerequisite to any
use management plan. The districts' key roles will be to in-
stigate studies of lake use problems (winter user conflicts
are becoming as much of a problem as warmer weather
recreation conflicts), propose solutions, and when regulations
or ordinances are indicated, to push for their passage by
the appropriate unit of government. To do this, the district
will need to demonstrate considerable cooperation and in-
fluence at the municipal, town and county level.
Public Health Aspects of Lake Management. Water
quality degradation by toxins and pathogenic organisms is
a growing area of concern at both the State and Federal
level. The lake districts already have at their disposal a drink-
ing water information program designed to enhance lake
community awareness of the relationships between private
land use (especially for septic tank water disposal and
residential well drinking water supplies) and lake/groundwater
quality. This program has been offered in about 30 lake com-
munities so far and applies to communities of both
oligotrophic and eutrophic lake systems. The drinking water
program is a meaningful expansion of management activities
by the individual lake districts and helps disseminate new
information to district members.
If a drinking water program in which residents voluntarily
submit samples of their well water for testing indicates
groundwater contamination, a thorough sanitary survey might
be conducted. Such a survey, specific septic tank system
investigations, and subsequent management projects are
areas being considered for State funding. However, to date,
most remedial action has been left to individual homeowners.
A related groundwater area of public health concern is
hazardous waste disposal and groundwater contamination
by landfills. Lake district members need to be made more
aware of these risks and sensitized, so they react to pro-
posed site plans and investigate existing disposal and land-
fill sites for safe operation.
Information and Education Elements
Elements of all these management options call for an infor-
mation approach to the lake districts. Usually a staff member
(ILR or UWEX) meets with the district and explains the con-
cepts and procedures for any given management approach.
Greater staff time commitments to an educational approach
are required for programs such as drinking water, waste
disposal management, or any user regulatory approaches.
155
-------
Lake Restoration, Protection and Management
The concept most applicable to protection districts is to
inform members of the district so they may evaluate technical
information, make rational decisions, and plan a suitable
course of action. They then can persuade others in their com-
munity to carry out their proposal. This is contrary to most
rehabilitation projects where the ILR specialist or hired con-
sultant is intimately involved in much decisionmaking
because of the limnological interrelationships involved.
In addition to the information roles pertinent to manage-
ment decisions, several purely educational programs should
be offered these districts. These might be viewed as public
service information programs sponsored by the district for
the lake community in general. Lake-related topics include
boating safety and regulations, water safety, first aid and
hypothermia, and residential land use practices to protect
lake quality (i.e., setbacks, lawn fertilization, leaf removal,
septic tank maintenance, erosion control, and landscaping).
This last topic is being further developed in a set of
homeowner's guides—illustrated folders containing fact
sheets and bulletins—to be provided lake districts and in-
terested individuals as a supplement and reference to their
information seminars and/or management plans.
Where possible, the information programs should be of-
fered in conjunction with elements of a proposed or ongo-
ing management activity. For example, a fact sheet and/or
talk on residential land use practices to protect lake quality
would fit in well with a soon to be conducted drinking water
project or land use regulation hearing. Specifically design-
ed education sessions might also be indicated when pro-
grams are being planned for land use zoning, waste disposal,
or nonpoint source pollution abatement. In these instances,
specialists from allied agencies might be recruited to talk
to the district members. For example, if the district is con-
sidering agricultural runoff control in the watershed, a
speaker from the SCS, LCC, or UWEX-agriculture could be
invited to address agricultural land use practices and ap-
propriate conservation techniques. The district could then
take advantage of this insight to plan its approach to the
problems.
TECHNICAL AGENCY COORDINATION IN
PROJECTS AND INFORMATION
ACTIVITIES
As any project develops, the district should be alert to other
sources of assistance. Examples of complementary sources
of technical assistance are various UW campuses, the U.S.
Geological Survey, Soil Conservation Service, Land Conser-
vation Unit of Wisconsin Department of Agriculture, Trade
and Consumer Protection, Wisconsin Geological and Natural
History Survey, other DNR bureaus, the U.S. Army Corps
of Engineers, county agencies, and regional planning
agencies.
As a note of caution, the districts are advised to exercise
care in the way they approach these various institutions for
help. The adage "too many cooks spoil the broth" bears
reflection. The districts deserve as much technical and finan-
cial help as they can get, but Chapter 33 clearly gives them
responsibility for their programs. This is one reason why all
projects require local labor and financial contributions. To
preserve their right of autonomy, the district members should
clearly understand what they want to do and the role they
expect any assisting agency to play. If the districts don't
judiciously guard their rights of self-determination, they risk
having their projects distorted and diluted to fit the objec-
tives of the agencies providing incidental assistance.
ASSESSMENT
Over time, the water quality monitoring element of this pro-
gram should provide a means of evaluating (a) the status
of individual lake projects, and (b) the relative success of
particular management techniques attempted, such as in-
fluent wetland manipulation to reduce lake sediment and
nutrient loadings.
The collective data from all participating lake districts will
,also contribute to an assessment of statewide lake quality.
Similarly, their comments about protection management ef-
forts will influence the further development of the program.
APPLIED EXAMPLES
Lake Redstone. Lake Redstone is a manmade, eutrophic
lake in south central Wisconsin of about 600 acres with an
average depth of less than 20 feet. It was created in 1965.
Management activities on this lake began at the same time
that the protection management concept was being
developed. Consequently, the Lake Redstone activities are
as much sources of this concept as they are projects of it.
In the case of Lake Redstone, the one has fed upon the other
so to speak.
The watershed has been evaluated and critical sources
of sedimentation identified. Regional SCS personnel, UWEX
agricultural specialists, LCC specialists, and the lake district
have cooperated on the site selection and design of sedi-
ment traps in the drainage basin. They are also preparing
improved land use plans for cooperating farmers.
A combined UW/DNR study of onsite waste disposal prac-
tices has identified nearshore septic tank problems. To com-
bat this, the district participates in a deed restriction man-
date requiring all property owners to pay into a community
fund that is used to ensure regular pumping of all holding
and septic tanks. If an individual property owner does not
have this service performed as scheduled, the district has
the job done by a contractor and charges it against that in-
dividual's account. Homeowners who voluntarily see to their
own on-site waste disposal management profit from the in-
terest paid them on their unused account.
In addition to these land use improvements in the water-
shed, the district, with State cost share assistance, will be
creating and maintaining an influent marsh system to fur-
ther intercept nutrient and sediment loadings to the lake.
They also have a weed harvester and have developed an
operating schedule for optimal lake use enhancement and
macrophyte removal.
As these integrated management projects proceed, a
regular water quality monitoring program will be conducted
to provide information for "before," "during," and "after"
project assessment.
Pretty and School Section Lake Districts. These two
small southeastern Wisconsin lakes (65 and 120 acres,
respectively) are in the same watershed, but Pretty Lake is
of considerably higher quality. Both districts have willingly
participated in the protection management program as a
cost-effective approach to lake management. School Sec-
tion Lake, because of its more advanced state of eutrophica-
tion, is also considering a limited dredging project.
The districts have established a joint committee to coor-
dinate their management efforts where mutual concerns are
involved. For example, both districts sponsored drinking
water information programs this year. In each case,
suspected cases of groundwater contamination were iden-
tified. A combined follow-up testing program is planned to
try to isolate sources of unusually high chloride, nitrate, and
coliform levels in the ground water. County and State health
and water resource specialists have been advised of the
study and will be further involved as subsequent informa-
tion is developed.
Both districts are working closely with county and regional
planning authorities for land use evaluations and reviews of
156
-------
Fiscal and Institutional Support for Lakes
any requests for building permits or zoning easements af-
fecting them.
They also cooperated in the initial planning and presen-
tation of a 1981 UWEX-sponsored weed and algae control
seminar for southeastern Wisconsin lake communities. The
seminar promoted discussions of innovative, cost-effective,
and cooperative approaches to weed control. It brought
together contractors, clients, and State specialists to in-
vestigate weed and algae control options of mutual benefit
to all parties. Subsequently, School Section Lake negotiated
a contract with a participating firm for macrophyte harvesting.
The success of this seminar led to a statewide lake district
conference on the same topic in the spring of 1982. From
this came the initial development by DNR of a State cost-
sharing plan for weed and algae control by lake districts
when they participate in a comprehensive watershed
management program.
The two districts are presently investigating regional land
use practices and plans affecting them to determine the ex-
tent of their involvement in any further watershed program-
ming, However, it is already evident because of the
remarkably small direct drainage basin that riparian
homeowners at Pretty Lake will be the key to their manage-
ment program. To meet this need, we have collected and
prepared a series of fact sheets and brochures encourag-
ing responsible residential land use. These materials are to
be provided homeowners by the district and will be accom-
panied by topical information sessions at their annual
meetings to reinforce the advice in the packets. Presently,
there are three draft packages of illustrated file folders and
component literature (as described earlier in this paper) be-
ing prepared. These will address septic tanks*, well water
systems, and runoff management. This element of the pro-
gram is viewed as a pilot project for refinement and later
application to other lake districts in the State.
SUMMARY
The Wisconsin Inland Lake Renewal Program is expanding
its lake management assistance to protection-oriented lake
districts. These districts are defined as those having com-
pleted lake restoration projects, as well as those having in-
itially high quality lakes in the first place.
The two primary characteristics of protection management
are (a) the expansion of the management concept to water-
shed land use and to more than in-lake techniques and
nutrient abatement and (b) a greater emphasis on informa-
tion and education programs. The moving force behind a
protection management program should be the district
members themselves. Inland Lake Renewal will remain the
major source of technical and financial assistance, and close
contacts with the staff will be preserved, especially to review
monitoring data and respond to management proposals. The
University-Extension component of Inland Lake Renewal has
assumed responsibilities extending beyond organization and
administration to the development of indicated management
information programs.
Both the information and financial aspects of protection
programs can be augmented by actively seeking the
cooperation of allied State or Federal institutions, but in so
doing care must be taken not to dilute the authority and
responsibility of the lake management districts or of the ILR
program as defined by Chapter 33.
' Adapted from an idea developed by Eckhari Dersch. Department o! Resource
Development Michigan Slate University
SEQUENCE OF STEPS IN THE LAKE
PROTECTION MANAGEMENT SCHEME
1 Request for technical assistance from district.
2. Meeting with commissioners following basic assessment
of lake conditions by ILR,
Determination of relative protection emphasis and par-
ticular interests of lake community for consideration.
3. Specifications of feasibility study sent to district with grant
offer information.
Likely to involve watershed emphasis, including detail-
ed inventory.
4. Bid process, selection, and grant arrangements with ILR.
5. Completion of feasibility study.
6. Review and recommendations for management—
examples:
• Land use zoning investigation
* Identification of particular land use problem areas
suggestions for improvement
suggestions of local technical assistance to solve
problems
• Weed control or algae control
• Shoreland, erosion management
» Residential land use practices to improve conditions
• Lake district investigations re land use practices and
regulatory suggestions or cooperation in local
community
• Lake use conflicts and possible approaches to
solutions
• Information programs pertinent to lake situation
• Water quality monitoring
• Fish survey
7 Meeting with district commissioners to discuss basic op-
tions they wish to with, priorities, technical and finan-
cial resources available. Information programs needed.
8. Coordination meeting with other local agencies invited
to participate in program.
9. Elements of basic plan to be funded by ILR, other agen-
cies, lake districts, i.e.,
Monitoring program
Preparation of ordinances
Waste disposal investigations
Weed or algae control
Land use studies
Involvement in land use improvements
Management/planning consultant fees
1 0.Implement management plan
11 .Assessment
Based on both monitoring data and possible extensive
sampling at some later date
Opinions and impressions of district commissioners, local
residents, and involved specialists
157
-------
THE MASSACHUSETTS CLEAN LAKES PROGRAM
EBEN CHESEBROUGH
Executive Office of Environmental Affairs
Dept. of Environmental Quality Engineering
Westborough, Massachusetts
INTRODUCTION
In December of 1981 the Massachusetts legislature
passed, and the governor signed, the Massachusetts Clean
Lakes and Great Ponds Act. The State now had a com-
prehensive lakes program that provided State aid for
diagnostic- feasibility studies and implementation projects.
Although the original program was structured to complement
the Federal Clean Lakes Program, a last -minute proviso
enabled it to remain viable in the absence of Federal funding.
The original source of funding for the program was a State
bond fund used to match Federal monies for the construc-
tion grants program. The Clean Lakes Program was allow-
ed up to $3 million in any one fiscal year. Unexpended funds
could be carried over to the next year with the condition that
no more than $5 million could be carried over in any one
fiscal year. Subsequent to the passage of the Clean Lakes
Program a new bond fund was created by the legislature
that earmarked a total of $30 million for the Clean Lakes
Program. The annual funding level of $3 million dollars re-
mained the same.
Table 1 explains the funding percentages authorized by
the program.
Table 1. — Program funding percentages.
With 314 funding
Federal State local
No 314 funding
Federal State local
Phase I
Phase II
70
50
30
25
0
25
0
0
70
50
30
50
The presence or absence of Federal 314 funding is
based upon its general availability. The present (FY83)
restrictions placed upon the Federal Clean Lakes Program
we interpret as being generally unavailable.
Prior to 1981 the Commonwealth had two separate lakes
programs. One managed the State's 314 program and con-
ducted a few diagnostic studies of its own. Although this was
a rather popular program it lacked authority and fiscal
resources.
The second, older program, begun in 1969, was known
as the Eutrophication and Nuisance Aquatic Vegetation Con-
trol Program. Its purpose was to fund various maintenance
projects at selected lakes and ponds in order to improve
them for public recreation. This program had (and still has)
legislative authority and was funded each year by the State
legislature.
In developing the rules and regulations for the new State
Clean Lakes Program the two former programs were
merged into a single, comprehensive program. The new pro-
gram will provide for the restoration, preservation, and
maintenance of the publicly owned lakes and ponds of the
Commonwealth for public recreation and enjoyment.
Maintenance projects (i.e., aquatic weed harvesting, nutrient
inactivation, pesticide treatment, etc.) will be considered im-
plementation projects and thus funded on the same basis.
Just as with long-term restoration or preservation projects,
however, maintenance projects must include watershed
management plans to control and reduce incoming nutrients
wherever possible.
A critical and perhaps innovative aspect of this program
is the assessment of potential success. Experience has
shown us that some lakes and ponds, because of their par-
ticular limnological characteristics, cannot be restored by
practicable, cost-effective, available technology. Therefore,
some lakes or ponds will be allowed to forego diagnostic-
feasibility studies altogether and be eligible for maintenance
program funding. This program element is intended to ex-
pedite remedial actions on those lakes and ponds where
long-term restoration or preservation is neither feasible nor
cost effective and to avoid costly delays where year-long
diagnostic-feasibility studies are inappropriate.
It is also true, however, that applications for water quality
maintenance projects will be assessed for potential long-term
solutions. We will discourage repeated resort to short-term
solutions where long-term solutions appear feasible and cost
effective.
The following rules and regulations for the Massachusetts
Clean Lakes Program are currently in effect. We began the
program in December 1982, just 1 year after the passage
of the enabling legislation. There are many similarities bet-
ween the State program and the Federal 314 program as
well as many important differences. We hope that by in-
cluding the program guidance here we may be able to help
other States develop similar clean lakes programs. Likewise,
we would appreciate anyone who finds some way our pro-
gram could be improved communicating with us.
RULES AND REGULATIONS
Definitions
The following terms shall have the following meanings:
1. Freshwater lake or pond: Any inland lake, pond, im-
poundment, or other similar body of water that exhibits no
oceanic or tidal influences and whose total dissolved solids
content is less than 1 percent.
2. Public access: An area that offers public access to the
lake or pond through publicly owned contiguous land so that
any person has the same opportunity to enjoy non-
consumptive privileges and benefits of the lake or pond as
any other person. If user fees are charged for public use
and access through State or substate operated facilities the
fees must be used for maintaining the public access and
recreational facilities of the lake or pond or other publicly
owned freshwater lakes or ponds in the State, or for improv-
ing the quality of these lakes and ponds. Any user fees
charged shall not be excessively discriminatory.
3. Publicly owned lake or pond: Any freshwater lake or
pond, its waters and the land thereunder, which is not under
158
-------
Fiscal and Institutional Support for Lakes
private ownership. Private ownership shall not include wildlife
associations or nonprofit environmental groups whose pur-
pose of ownership is resource management, conservation,
wildlife preservation, or the like. In those situations where
private ownership pertains solely to water rights to the body
of water, that body of water may be eligible when written
assurance is given that any exercise of the water rights would
not interfere with, or affect in a negative manner the lake
or pond restoration, preservation, or maintenance program.
4. Public water supply: Any body of water currently in
use as a public water supply but not secondary or emergency
water supplies which are currently used by the public for
recreational activities.
5. The department: The Department of Environmental
Quality Engineering, Division of Water Pollution Control.
6. Public entity: Any city, town, special district, the
metropolitan district commission, or other existing govern-
mental unit eligible to receive State funds.
7. The Director: The Director of the Massachusetts Divi-
sion of Water Pollution Control.
Regulations
1. Eligibility Requirements: The following conditions must
be met for any body of water to be eligible for funding under
the Clean Lakes Program.
a. the body of water under consideration must be a
freshwater lake or pond;
b. the body of water under consideration must be a
publicly owned lake or pond;
c. the body of water under consideration must have
public access; and
d. the body of water under consideration must not be
a public water supply.
First consideration shall be given to those lakes and ponds
which meet the above eligibility criteria. After processing ap-
plications for said eligible lakes and ponds, and if program
fiscal resources allow, then second consideration shall be
given to lakes and ponds which are eligible in every regard
excepting the freshwater requirement. That is, secondary
consideration shall be given to brackish, salt, or tidal lakes
and ponds only if there remain sufficient funds after process-
ing all other eligible applications.
2. Types of Assistance: The department will provide fund-
ing assistance for studies and projects as set forth below:
A. Diagnostic-Feasibility Studies
This type of study shall consist of two parts:
1. Diagnostic study—This will include gathering infor-
mation and data to identify existing or potential sources of
pollution and to determine the limnological, morphological,
demographic, and other pertinent characteristics of the lake
and its watershed. The minimum requirements for a
diagnostic study are set forth in Appendix A (1). The Appen-
dix A data requirements may be amended for certain lakes
or ponds if such action is deemed appropriate by the
department.
2. Feasibility study—This will include an analysis of the
diagnostic information to define methods and procedures for
controlling the sources of pollution; a determination of the
most cost-effective procedure to improve or preserve the
quality of the lake or pond for maximum public benefit; the
development of a technical plan and milestone schedule for
implementing watershed nutrient controls and in-lake restora-
tion procedures (if appropriate) or preservation techniques.
The minimum requirements for a feasibility study are set forth
in Appendix A (2).
3 For the studies described above where the Federal
qovernment has provided financial assistance in the amount
of 70 percent of the cost of any study the department shall
provide an amount not exceeding 30 percent of the total eligi-
ble costs. In the event that the Federal government has not
provided any opportunity for funding due to nonappropria-
tion of funds by Congress, the department shall provide an
amount not exceeding 70 percent of the total eligible costs
for such studies.
B. Implementation of long-term restoration projects or
preservation techniques.
Projects shall have as their objectives:
1. Watershed and lake or pond management plans with
methods and procedures for controlling pollution entering
the lake and for restoring the lake. Such methods and pro-
cedures must provide long-term lake restoration.
2. Watershed and/or lake and pond management plans
for the preservation of the water quality of the lake or pond.
Such plans must provide long-term preservation.
3. For said projects described above in 2(B)1.2. where
the Federal government has provided financial assistance
in the amount 50 percent of the costs of any such project,
the department shall provide an amount not exceeding 25
percent of the total eligible costs. In the event that the Federal
government has not provided any opportunity for funding due
to nonappropriation of funds by Congress, the department
shall provide an amount not exceeding 50 percent of the
total eligible costs for such studies.
C. Water Quality Maintenance Programs
1. These methods include short-term solutions for
alleviating nuisance aquatic vegetation and algae problems
in lakes and ponds. The major objectives shall be:
a. The implementation of corrective measures for the
purpose of providing suitable conditions for public recrea-
tion in lakes and ponds.
b. The implementation of watershed management
plans whose goal it will be to control and reduce incoming
nutrients wherever possible.
2. For these water quality maintenance programs the
department shall provide funding not to exceed 50 percent
of the total eligible costs as provided under Chapter 628 of
the Acts of 1981. Should the State legislature at any time
provide additional funds for maintenance projects under
Chapter 722 of the Acts of 1969 then the cost sharing shall
be as provided by said Act.
3. Application Procedure
A. Request for Assistance
To qualify for consideration for funding under the
Massachusetts Clean Lakes Program the applicant must
submit a written request for assistance. Said written request
for assistance must provide the following:
1. The legal name of the lake or pond and a locus map.
2. Statement of ownership. If the water rights are private-
ly owned, this ownership must be documented.
3. Narrative description and/or plan showing location
of the public access area(s) relative to the lake and public
roadways.
4. Description of current recreational uses. An account
of historical recreational uses, if different, also should be
included.
5. Description of the particular problems and nuisance
conditions affecting the lake or pond and the recreational
use thereof.
6. The written request for assistance must be signed
by an eligible applicant (see 3.C.1. for a description of eligi-
ble applicants).
B. Determination of Eligibility
1. Upon receipt of a written request for assistance the
department shall make a determination of eligibility for the
lake or pond. The evaluation may also require the following:
a. An on-site visit to observe the lake or pond, its
public access area(s), the watershed, or any other pertinent
characteristic.
b. The securing of legal documentation which shows
ownership of the lake or pond and/or the public access
area(s).
159
-------
Lake Restoration, Protection and Management -
c. The arrangement of a pre-application conference
to discuss the lake or pond and the request for assistance.
The department believes the pre-application conference to
be an important aspect of the application process. During
the pre-application conference the department will advise the
applicant on the most appropriate type of application to
submit.
d. Any other information which the department feels
is relevant to a determination of eligibility.
C. Applicant Eligibility
1. Any applicant for funding under the Massachusetts
Clean Lakes Program must be a public entity.
D, Application Procedure
1. Any public entity may apply to the department for
assistance under the Clean Lakes Program by completing
and filing the Clean Lakes Program Application Form as set
forth in Appendix D of these regulations.
2. The department shall review each Clean Lakes Pro-
gram Application Form submitted, and may require the ap-
plicant to provide additional information and to attend infor-
mal conferences. Failure of the applicant to provide ade-
quate, accurate, complete, or current information could result
in less favorable rating of the project and lower placement
on the priority list.
E. Application Data Requirements
1. Diagnostic-feasibility study application requirements.
The objectives of the diagnostic-feasibility study include
the data requirements of Appendix A—Data Requirements
for Diagnostic-Feasibility Studies. An applicant may not re-
quire a complete diagnostic-feasibility study if a body of re-
cent limnological data exists. It is therefore important for the
applicant to be as thorough as possible in the submission
of the requested data.
a. Using Appendix A the applicant shall submit all of
the required information which is available. All information
shall be the most current available, especially biological and
chemical data.
2. Implementation of long-term restoration projects or
preservation techniques.
a. The information and data requirements specified
in Appendix A from a diagnostic-feasibility study or its
equivalent shall be submitted.
3. Water Quality Maintenance Program
The application requirements are identical to that of E.1
above, diagnostic-feasibility study application requirements.
F. Renewal of Application
An applicant whose study or project was on the priority
list for any given year but not funded may renew the ap-
plication for the subsequent funding year by notifying the
department in writing prior to the application period deadline.
4. Assessment of Potential Success
The need for assessing the potential success of lake or
pond projects at this stage is to prevent the inefficient use
of Clean Lakes Program resources. An applicant for a
diagnostic-feasibility study or an implementation project
which is not funded after thorough analysis of the proposal
may be encouraged to revise the application and apply to
the Water Quality Maintenance Program. The purpose of the
pre-application conference is to avoid the necessity of such
revisions by advising the applicant of the most appropriate
type of application to submit.
This program element is intended to facilitate and expedite
remedial actions on those lakes and ponds where long-term
restoration or preservation is neither feasible nor cost effec-
tive and to avoid costly delays where year-long diagnostic-
feasibility studies are determined to be inappropriate.
Applications for water quality maintenance projects will be
assessed for potential long-term solutions. The department
will discourage repeated resort to short-term solutions where
long-term solutions appear feasible and cost effective. In
such cases the department may recommend revision of the
application accordingly.
A. Assessment of Potential Success Criteria
The assessment shall consist of the review and analysis
of one or more of the following:
1. Analysis of submitted or otherwise available water
quality data.
2. Analysis of the morphometric characteristics of the
lake or pond.
3. Consideration of the mode of origin of the lake or
pond.
4. Analysis of historical data including watershed ac-
tivities, past macrophyte or phytoplankton treatment, altera-
tions to the lake level, or other relevant matters.
5. On-site visit and appraisal.
6. Any other pertinent data which the department finds
relevant to the assessment decision process.
5. State Prioritization
The department shall maintain a project priority list rank-
ed from highest to lowest for the purpose of (a) setting pro-
ject priorities for restoration and/or preservation projects, or
maintenance projects, and (b) allocating the program's fiscal
resources on an annual basis. The priority list shall be review-
ed, amended, and updated at least annually.
The Massachusetts Clean Lakes Program shall provide
special consideration to ongoing restoration or preservation
projects previously funded under the Clean Water Act
(Federal Water Pollution Control Act, as Amended, 33 USC
466 et seq.), Section 314 Clean Lakes Program. Said special
consideration shall be limited to completed diagnostic-
feasiblity studies and ongoing implementation projects.
A. Prioritization Criteria
When evauluating a project, the department shall assign
priority points based on the following criteria:
Categories*
1. Recreational
Use
2. Type of
Public Access
3. Trophic Status
Rating
4. Relative
Importance
Limits
Active contact
Non-contact
Passive only
Beach and boat ramp
Beach or boat ramp
Undeveloped
Maintenance projects:
Eutrophic (13-18 pts)
Mesotrophic (7-12 pts)
Oligotroptiic (1-6 pts)
Restoration/Preservation
Project:
Oligotrophic/
Mesotrophic (1-12 pts)
Low Eutrophic (13-15
pts)
Hypereutrophic (16-18
pts)
High
Moderate
Low
Point
Value
5
3
1
5
3
1
5
3
1
* Recreational Use: This criterion shall be based on current factual use and documented
historical use of the lake or pond and shall be the ordinary general use(s) and not excep-
tional use(s).
Type of Public Access: This criterion shall be based on on-site visit(s) and observation
of public access facilities. The department shall consider an area used for public bathing
as a beach and an area suitable for launching a boat as a boat ramp.
Trophic Status Rating: This criterion shall be based on the current edition of the
Massachusetts Lake Classification System or other appropriate data.
Relative Importance: This criterion shall take into account such factors as, (a) the prox-
imity of other publicly owned lakes and ponds and their trophic state and recreational poten-
tial; (b) degree of public support for the project; (c) degree of political support for the pro-
ject; and (d) historical local efforts (monetary or otherwise) aimed toward restoration, preser-
vation or maintenance.
160
-------
Fiscal and Institutional Support for Lakes
B. Priority Point System
For each application processed the criteria system shall be
applied and priority points established by adding the priority
points for each category to determine the project's ranking.
The processed applications will then be prioritized in accor-
dance with their total priority points from highest to lowest.
In the event the department must choose between two ap-
plications having the same priority point value, the department
shall select the application having the earliest date of receipt.
6. Local Funding
The local cost share shall be the cost of a project remain-
ing after taking into account any State funding and/or Federal
funding. No more than 25 percent of the local cost share shall
be in-kind services as approved by the director.
7. Substate Agreements
A. The department may provide financial assistance under
the Clean Lakes Program to public entities by means of a
written substate agreement transferring project funds from the
state to those public entities. The agreement may be for all
or a portion of a project.
6. Reports
1. Progress Reports: Progress reports for diagnostic-
feasibility studies, implementation projects and maintenance
projects shall be submitted by the recipient in accordance
with the schedule established in the substate agreement.
These reports must include:
a. Work progress relative to the milestone schedule
and any difficulties encountered.
b. A discussion of the project findings to date.
c. A detailed report of expenditures since the previous
progress report and those anticipated during the next report-
ing period.
2. Diagnostic Feasibility Report: A final report shall be
required from a diagnostic-feasibility study. The contents of
the final report must include the information specified in Ap-
pendix A. The Appendix A data requirements may be
amended for certain lakes or ponds if such action is deem-
ed appropriate by the department. Previously existing
diagnostic- feasibility information not produced during the
course of a diagnostic- feasibility study must be included in
the final report. Certain studies may involve special data re-
quirements which may be included in the final report. Said
special conditions shall be specified in the substate
agreement.
3. Environmental Evaluation: Feasibility studies, whether
separate or in concert with a diagnostic study, must include
in the final report an environmental evaluation of the pro-
posed project which considers, at a minimum, the questions
specified in Appendix B.
4. Implementation Projects: A final report will be required
of every implementation project. The report must detail the
activities undertaken during the project and describe the
achievement of the project with respect to the stated pro-
ject purposes and objectives. Data collected during the pro-
ject implementation period must be included in the report
as well as final engineering designs and specifications. Since
every project will be different, special report requirements
may be specified in the substate agreement. The report must
conform to the format presented in the U.S. Environmental
Protection Agency manual on "Scientific and Technical
Publications."
8. Annual Calendar for Application Procedure
There shall be an annual schedule for application to the
Massachusetts Clean Lakes Program established by the
Director by administrative procedure.
9 Program Administration
The Director shall, as he deems necessary and ap-
propriate issue administrative procedures, policies and
guidance documents required to effectively carry out the pur-
pose of these regulations.
Appendix A
Diagnostic-feasibility studies (Phase I projects) shall include
in their scope of work the following requirements:
1. Diagnostic Study
a. An identification and description of the lake or pond
to be restored, preserved, or maintained. The discussion
must include a watershed description with references to size,
topography, development and lake or pond usage.
b. A description of the public access to the lake or pond
and its suitability to the recreational use of the lake or pond
by the general public.
c. A geological description of the drainage basin in-
cluding major soil types and their relation to the water quality.
d. A summary of historical lake uses, including recrea-
tional uses up to the present time. The summary should in-
clude a description of any historical watershed activities
which may have adversely affected the water quality.
e. A description of the land uses in the lake watershed,
listing each land use classification as a percentage of the
whole and discussing the amount of nonpoint source loading
produced by each category.
f. Morphometric Data (metric units)
1. surface area
2. maximum depth
3. mean depth
4. volume
5. watershed area
6. maximum length
7. maximum width
8. shoreline
9. development of shoreline
10. bathymetric map
g. Annual nutrient budget for phosphorus and nitrogen
(must include internal and external loading, groundwater,
outflow, wet and dry precipitation and storm drainage (if
appropriate)).
h. Annual hydrologic budget, including:
1. retention time
2. flushing rate
i. A discussion and analysis of historical baseline lim-
nological data and one year of current limnological data
which must include:
1. temperature profiles with 1 meter intervals
2. dissolved oxygen profiles with 1 meter intervals
3. pH
4. total alkalinity
5. suspended solids
6. dissolved solids
7. conductivity
8. chlorides
9. Kjeldahl-nitrogen
10. ammonia-nitrogen
11. nitrate-nitrogen
12. total phosphorus
13. total and fecal conform bacteria
The above parameters must be measured at the deepest
point in the lake (or the deepest point of each major basin,
should there exist more than one), the tributaries and the
outlet.
The in-lake station must be sampled in the epilimnion,
metalimnion, and hypolimnion if stratification occurs, or near
the top and near the bottom of the water column if stratifica-
tion does not occur. The coliform bacteria sample should
be collected from the surface water only.
The in-lake station(s), tributaries and outlet must be
surveyed for the data specified above (1 .i) biweekly from
March to fall turnover and monthly for the balance of the year.
j. The in-lake station(s) also must be sampled or
tested for the following during each survey:
161
-------
Lake Restoration, Protection and Management
1. Secchi disc transparency
2. phytoplankton identification (minimally to genus
level) and density count.
3. chlorophyll a Under stratified conditions the
phytoplankton and chlorophyll a samples must be depth in-
tegrated to the thermocline.
k. A macrophyton survey must be conducted during
mid-August and must be sufficient to describe the areal
coverage and dominant genera.
I. , In the event that the lake or pond has storm drains
which discharge into it, then qualitative and quantitative wet
weather surveys must be conducted during at least two
separate storm events. In both cases the "first flush" must
be sampled and the discharge sampled thereafter at ap-
propriate intervals for a minimal period of 2 hours, or 1 hour
after peak flow, whichever is a shorter time period.
Parameters measured must include, at a minimum, the
following:
1. .suspended solids
2. dissolved solids
3. Kjeldahl-nitrogen
4. ammonia-nitrogen
5. nitrate-nitrogen
6. total phosphorus
7. total and fecal coliform
8. heavy metals (chromium, manganese, iron, cop-
per, zinc, cadmium, lead) analysis on flow weighted com-
posite sample from each storm.
9. on-site volumetric measurement of rainfall.
m. At the in-lake station(s) the bottom sediments must
be sampled and analyzed for the following:
1. total nitrogen
2. total phosphorus
3. organic/inorganic fraction (loss on ignition)
4. heavy metals (chromium, manganese, iron, cop-
per, zinc, cadmium, lead)
5. other parameters as deemed necessary by the
department to meet appropriate permit requirements.
n. Sampling, sample preservation and analytical
methodology must be conducted according to Standard
Methods, EPA's Methods for Chemical Analysis of Water and
Wastes, or other equally acceptable means.
2. Feasibility Study
a. This study must include an identification and discus-
sion of the alternatives considered for pollution control,
restoration, preservation, or maintenance and identification
and justification of the selected alternative(s). The discus-
sion must include the reasons for rejecting the unacceptable
alternatives.
For the selected alternative(s) the study must include a
discussion of technical feasibility, cost-effectiveness, and an-
ticipated water quality and recreational improvements. In this
regard a Vollenweider-type analysis is recommended. For
lake restoration projects the discussion must include an
analysis of the impact the selected alternative(s) will have
on the annual nutrient budget.
For the selected alternative(s) the study must include a
detailed description specifying what activities would be
undertaken to implement the alternative. Preliminary
engineering drawings must be provided to show the con-
struction aspects of the project, if appropriate.
b. A task by task cost breakdown of the recommended
project. This analysis must include an estimate of the opera-
tion and maintenance cost over the useful life of the project.
c. A proposed milestone work schedule for completing
the project with a proposed budget and payment schedule
that is related to the milestone work schedule.
d. The study must include a discussion of the public
participation used in the development of the selected alter-
native. A minimum of two public meetings are required in
the development of the selected alternative. The discussion
must include major public comments on the proposed
alternative.
e. An implementation project (Phase II) monitoring pro-
gram must be developed. A monitoring program must be
maintained during project implementation, particularly dur-
ing construction phases or in-lake treatment to provide suf-
ficient data that will allow the project officer to redirect the
project, if necessary, to ensure desired objectives are achiev-
ed. A 3-year post-construction or post-implementation
monitoring program also must be developed to evaluate pro-
ject effectiveness. Each monitoring program must be in-
dividually tailored to a specific implementation project and
should detail the sampling stations, sampling schedule, and
parameters.
Appendix B
Environmental Evaluation
For every implementation project proposed an en-
vironmental evaluation must be completed which considers,
at a minimum, the following questions:
1. Has the local or State historical society been contacted
and apprised of the implementation project? Relevant com-
ments from said society must be included in the final report.
2. If the proposed project involves the use of in-lake
chemical treatment, what long- and short-term adverse ef-
fects can be expected from that treatment? How will the grant
recipient mitigate these effects?
3. If any type of dredging is proposed what steps and pro-
cedures will be taken to minimize any immediate and long-
term adverse effects of such activities? Where will the dredge
spoils be deposited? How will the dredge spoils be
transported to the deposition site? Will there be any problem
in securing the necessary permits and licenses for such
activities?
4. Will the proposed project have an adverse effect on fish
and wildlife or on wetlands or any other wildlife habitat? What
actions have been incorporated to mitigate habitat or wildlife
destruction? Has the Division of Fisheries and Wildlife com-
mented on the proposed project? Relevant comments from
said agency must be included in the final report. Has the
local conservation commission been contacted for review of
the project? Any comments they may have should be includ-
ed in the final report.
5. What adverse environmental effects will occur either
downstream or upstream as result of the proposed project?
What actions have been incorporated to mitigate such
adverse effects?
6. Describe other measures not discussed previously that
are necessary to mitigate adverse environmental impacts of
the implementation of the proposed project.
162
-------
Pre- and Post-Restoration
Assessment Data and
Techniques
EXPERIENCES IN DEVELOPING A CHLOROPHYLL A STANDARD IN
THE SOUTHEAST TO PROTECT LAKES, RESERVOIRS,
AND ESTUARIES
R. F. McGHEE
U.S. Environmental Protection Agency
Atlanta, Georgia
ABSTRACT
The adverse consequences associated with excessive growths of algae have been a problem of major
concern for local, State, and Federal water pollution control agencies. The regulatory framework set forth
by the Clean Water Act makes control of algae most logical through water quality standards-based ef-
fluent limitations. This paper documents the development of a chlorophyll a water quality standard by the
State of North Carolina. The purpose of the standard was to provide regulations that could be used to
limit both point and nonpoint source discharges of nutrients.
INTRODUCTION
After passage of the Federal Water Pollution Control Act
Amendment of 1972, State water pollution control agencies
received unprecedented amounts of Federal funds to be us-
ed in water quality planning functions. The first round of plan-
ning concentrated on an inventory of pollutant sources,
development of point source waste load allocations, and ex-
amination of water quality standards. These processes were
documented, for the most part, in Section 303(e) Basin Plans.
The second round of planning continued those programs
and greatly emphasized examining nonpoint sources of
pollutants. The second round of planning culminated in the
development of Section 208-financed Water Quality Manage-
ment Plans. During the development of 208 plans, the
Federal Water Pollution Control Act was amended in 1977
and 1981 and retitled the Federal Clean Water Act (the Act).
Starting with the 1972 amendments, the Act outlined a
two-pronged approach for restoring and maintaining good
quality in the Nation's waters. The Act set forth a plan to
control water pollution by establishing technology-based
minimum levels of treatment for point source discharges and
increased the levels of treatment through time. In addition,
the Act proposed to control water pollution by establishing
water quality standards. Once established, the water quali-
ty standards served as the basis for development of allowable
waste loads for both point and nonpoint source discharges.
Except in rare cases, the development of technology-
based minimum treatment levels did not limit nutrients such
as phosphorus and nitrogen. For this reason, the regulatory
framework made development of water quality standards the
most logical approach for a statewide program to control ex-
cessive growths of algae.
NORTH CAROLINA CASE
In the 303(e) Basin Planning Process, members of the State
regulatory staff became acutely aware of the actual and
potential water quality problems associated with excessive
growths of algae. Also, in 1971 and 1972 the Chowan River,
a freshwater estuary to the Albemarle Sound, experienced
blooms of blue-green algae. These blooms made the estuary
163
-------
Lake Restoration, Protection and Management
unsuitable for water contact recreation, adversely affected
both sport and commercial fisheries, and created industrial
water supply problems. Other estuaries and lakes within the
State had also encountered adverse consequences from ex-
cessive algal growth. The public reaction to these cases man-
dated that the State pollution control agency correct the
problems.
In February 1975, the State held a public hearing on
several proposed water quality standards revisions including
a narrative standard for nutrient and algae control. That stan-
dard read as follows:
In impounded or slow-moving waters which are subjected to
nutrient enrichment and in which excessive algae activity results
in or is expected to result in Interference with established water
uses, the Department of Natural and Economic Resources is
authorized to establish a stream nutrient standard appropriate
to the body of water affected.
The proposed regulation was written using a State statute
that gave the department general authority to control water
pollution.
The language of the narrative standard appeared suffi-
cient for the State to control nutrients; however, experience
had shown that the State can be more effective with less
resources when using numerical rather than narrative stan-
dards. For this reason, the State Division of Environmental
Management staff was charged with responsibility to develop,
if possible, numerical standards for controlling algae. In May
of 1977, the State staff requested the assistance of the Water
Resources Research Institute. The Institute is an organiza-
tion that coordinates water research with the Universities.
At the Division's request, the Institute organized a Water
Quality Standards Advisory Group comprised of individuals
from the Universities in North Carolina and other State agen-
cies with technical expertise in algae. Members of the Ad-
visory Group included:
Dr. James Stuart Water Resources Research Institute,
Raleigh
Dr. Peter Campbell University of North Carolina, Chapel Hill
Dr. Charles Weiss University of North Carolina, Chapel Hill
Mr. Terry Anderson University of North Carolina, Chapel Hill
Dr. Donald Stanley University of North Carolina, Chapel Hill
Dr. Edward Kuezler University of North Carolina, Chapel Hill
Dr. Gus Witherspoon North Carolina State University, Raleigh
Dr. Donald Hayne North Carolina State University, Raleigh
Dr. Mark Brinson - East Carolina University, Greenville
Mr. Scott Van Horn North Carolina Wildlife Resources Commis-
sion, Raleigh
Mr. Cape Carnes - North Carolina Wildlife Resources Commis-
sion, Raleigh
Mr. Donald Tobaben - North Carolina Division of Inland Fisheries,
Raleigh
Mr. Alan Peroutka - North Carolina Divison of Environmental
Management, Raleigh
Mr. David Park - North Carolina Division of Environmental
Management, Raleigh
Mr. R. F, McGhee - North Cariina Division of Environmental
Managment, RaJeigh
Prior to meeting with the Advisory Group, the Division of
Environmental Managment staff conducted a literature
search and developed a proposed standard for discussion
with the Advisory Group. The proposed standard read as
follows:
Chlorophyll a shall not exceed 50 ^g/l in freshwater lakes and
reservoirs, 20 /^g/l in lakes and reservoirs designated as Trout
Waters, and 100 ^g/i in all sounds, estuaries, and other slow
moving waters. The chlorophyll a concentration shall be that
concentration determined at any one time and at a depth equal
to one-half the Secehi depth.
The Advisory Group initially met June 10, 1977 and had
a final meeting on July 7,1977. The Group concluded that
a chlorophyll a standard was a good method for controlling
excessive cultural eutrophication. The Group also conclud-
ed that presently oligotrophic lakes should possibly be main-
tained at their current level. Further study on that issue was
recommended. The Advisory Group also concluded that a
statewide standard on phosphorus, nitrogen, or both would
not be technically sound because other characteristics of the
water bodies would not be taken into account,
Of utmost utility to the Advisory Group was a presenta-
tion of information by Dr. Charles M. Weiss (1976) from his
recently completed report "Trophic State of North Carolina
Lakes." His report included sampling of 69 lakes in North
Carolina and a rigorous statistical analysis of the data. The
report suggested classification of North Carolina lakes in the
following manner:
Chlorophyll a Total Phosphorus Secehi Depth
Trophic State mg/m3 mg/m3 Meters
Oligotrophic
Oligo-mesotrophic
Mesotrophic
a - Eutrophic
8 - Eutrophic
Hypereutrophic
*<2
2-6
6-15
15-40
40-100
>100
<10
10-19
20-39
40-79
80-150
:> 150
>3.0
1 .5-3.0
1.0-1.5
0.5-1.0
0.1-0.5
<0.1
Dr Weiss advised that a standard of 40 ng/l for warmwater
lakes and 15 ^g/l for trout lakes seemed more reasonable
than the values proposed by the Division of Environmental
Management. The Advisory Group concurred with these
values.
Dr. Stanley commented that chlorophyll a concentrations
exceeded 100 ^g/l in the Pamlico estuary during the winter
with no noticeable adverse effects. The final standard pro-
vided for this situation by being applicable only during the
summer months.
In October 1977, the Division of Environmental Manage-
ment obtained permission from the Environmental Manage-
ment Commission to hold public meetings on the proposed
chlorophyll a standard along with many other proposed stan-
dards revisons. The purpose of the meetings was to obtain
public input into the formulation of all water quality standards
revisions. Two public meetings were held in November 1977,
In May 1978, the Division of Environmental Management
completed a set of proposed water quality standards that
included the chlorophyll a standard. Also during this period
the Division presented and obtained concurrence with pro-
posed standards from both the 208 Technical Advisory Com-
mittee and the 208 Policy Advisory Committee. The En-
vironmental Management Commission authorized public
hearings, mandatory for rulemaking in North Carolina.
Public hearings were held at three locations in the State
during July and August 1978. Very little concern or opposi-
tion was voiced over the chlorophyll a standards except for
applying them to very small lakes such as as farm ponds.
The Environmental Management Commission members
originally voiced concern over lake size and in response, size
limitations were added to the standards.
During the summer of 1978 the Chowan River estuary ex-
perienced massive blooms of blue-green algae that greatly
heightened the interest in establishing standards to control
algae. The Chowan situation was so intense that another
regulation called "Nutrient Sensitive Waters" was drafted
by the Division of Environmental Management Staff; it pro-
hibited, with some exceptions, the discharge of nitrogen and
phosphorus above background levels in designated waters.
This regulation proceeded simultaneously with the chlorophyll
a standard.
After several delays, not related to the chlorophyll a stan-
dard, the Environmental Management Commission adopted
the following standard on August 9, 1979:
Chlorophyll a: not greater than 40 jxj/1 for lakes, sounds,
estuaries, reservoirs, and other slow-moving waters not
designated as trout waters, and not greater than 15 jjg/l for
164
-------
lakes, reservoirs, and other slow-moving water designated as
trout waters (not applicable during the months of December
through March; not applicable to lakes and reservoirs less than
10 acres in surface area).
Also on that same date, the Chowan River Basin was
designated by the commission as a Nutrient Sensitive Water.
EPA approved the chlorophyll a standard on November 9,
1979, under the provisions of Sec. 303(c) of the Act.
POST STANDARD ADOPTION
At the time of adoption of the chlorophyll a standard, the
State recognized that data and staff resources were not
available to fully apply the standard statewide. However, the
State has used the standard in making many water quality
decisions. For the most part, the State has concentrated
resources on the most severe nutrient problem areas, and
also used the standard to formulate positions on a few pro-
Pre- and Post-Restoration Assessment Data and Techniques
posed waste treatment facilities and water resource projects
such as dams (Westall, pers. comm.)
The State is pleased with the utility of the standard and
plans to retain it until a more technically complete and usable
alternative becomes available (Westall, pers. comm.)
ACKNOWLEDGEMENTS: The author expresses appreciation to the
staff of the North Carolina Division of Environmental Management
for their assistance in researching the water quality standards files.
REFERENCES
North Carolina Division of Environmental Management. Water quality
standards files. Raleigh, N. C.
Weiss, C. M., and E. J. Kuenzler. 1976. The trophic state of North
Carolina lakes. Univ. N. C. Water Resour. Res. Inst. Rep. No. 119.
Westall, Forrest. North Carolina Division of Environmental Manage-
ment. Raleigh, N. C. Pers. comm.
165
-------
CHANGES IN WATER QUALITY OF SKAHA LAKE, BRITISH
COLUMBIA, FOLLOWING REDUCTION IN PHOSPHORUS LOADING
RICHARD N. NORDIN
Ministry of Environment
Victoria, British Columbia
ABSTRACT
Poor water quality in the late 1960's (cyanophyte blooms, poor water clarity) prompted the installation
of tertiary sewage treatment facilities at the city of Penticton which discharge into Skaha Lake. The
reduction of phosphorous (11 tonnes or approximately 40 percent of the phosphorus load) took place
in 1972. In the 1967 to 1970 period relatively high concentrations were recorded (31-41 ng/l total P
spring concentrations). Lower concentrations were noted in the 1974-78 period (11 -23 ng/l) but higher
concentrations occurred in 1979-82 (22-32 jjg/l) spring total phosphorous). In spite of the increased
lake concentrations since 1979, the measured load appears not to have increased significantly since
1972. Three possible factors are considered that partially explain the transitory response that occur-
red: increase in nonpoint (unmeasured) loadings; underestimation of total loadings; and the effects
of hydrology on lake concentrations. The hydrologic effect is particularly important since the water
residence time is short (T = 1.2 yrs) and a large interannual variation (0.6-2.5 yr) can explain part
of the concentration changes observed. Hydrologic parameters appear to be a major consideration
in assessing water quality restoration projects.
INTRODUCTION
Reducing phosphorous loading to lakes has proved to be
a generally successful strategy in improving water quality
(Edmondson and Lehman, 1981; Smith and Shapiro, 1981).
However, a number of cases have been described in the
literature where phosphorus loading was significantly re-
duced but without resulting in the expected lake response
(Emery et al. 1973; Larsen et al. 1975).
Skaha Lake is located in the Okanagan Valley of British
Columbia approximately 400 km east of -Vancouver. The
valley contains a chain of lakes which flow north to south
emptying into the Columbia River System. The Okanagan
Valley has a semi-arid climate (25 to 30 cm yr1 precipita-
tion) and is the focus of a valuable tourist industry centered
around water-based recreation ($230 million tourist revenues
in 1978). _
Skaha Lake is a medium-sized lake (20.1 km2, z = 26m)
with a relatively short residence time (1.2 yr). Its watershed
contains upland forest, lowland ponderosa pine/bunchgrass
benches, with the lowland areas containing nearly all of the
population. Numerous orchards and other developments ex-
ist in the watershed, and the city of Penticton discharges
its sewage effluent into the inflow to the lake.
Complaints of poor water quality in the late 1960's promp-
ted a number of studies of water quality (Coulthard and Stein,
1968, 1969; Stein and Coulthard, 1971; Pinsent and
Stockner, 1974) and as a consequence a decision was made
to incorporate tertiary treatment facilities into the sewage
treatment plant in 1971. The removal of 80 to 90 percent
of the phosphorus from this discharge reduced the estimated
loadings to the lake by approximately 40 percent according
to the best estimates of inputs (Alexander, 1982). Studies
were carried out during the late 1970's as part of the
Okanagan Implementation Program (Truscott and Kelso,
1979; Jensen, 1981) to assess the changes in water quality
in Skaha Lake, particularly in regard to tertiary treatment at
Penticton. Subsequent monitoring by the Ministry of Environ-
ment and earlier monitoring were used to interpret the
changes.
DATA BASE
Data exist for 1967 to the present; however, there are ma-
jor deficiencies in the quality and quantity of data. The in-
formation that does exist can be used to make qualified
judgments on patterns of events. Key parameters are con-
sidered here.
Phosphorus and Nitrogen
Spring overturn phosphorus concentrations from 1967 to
1982 are shown in Figure 1. The installation of tertiary treat-
ment took place in 1971 and the pattern, on first examina-
tion, indicates a decrease in phosphorus after the reduction
in loading but increases again after 1978. This pattern is at
odds with the best understanding of loadings for the period.
Loadings have been calculated for bioavailable phosphorus
(see Gray and Kirkland, 1982) for the pre-diversion period
(22.4 tonnes for 1970) versus the post-diversion period (12.7
tonnes for 1980) (Alexander 1982). No major significant
change in point source loading occurred between 1971 and
1980 so the increase in concentration after 1978 appears
to be caused by other factors.
Figure 1 .-Spring overturn total phosphorus concentrations for Skaha
166
-------
Pre- and Post-Restoration Assessment Data and Techniques
One of the major factors affecting lake concentrations is
hydrology. Water residence time has varied considerably
even within the short period of record considered here and
the controlled nature of the Okanagan River. Extremes vary
from 0.6 years in 1972 to 2.5 years in 1977 with an overall
mean of 1.2 years.
The relationship between concentration and hydrology is
a difficult one to elucidate. The lake is unusual in that the
major inflow (the Okanagan River) has a much lower
phosphorus concentration (6-8 ^g/T1) than the lake itself
(25-30 J4J/T1 recently). As such, the inflow generally acts to
dilute the lake concentrations so that high concentrations
occur at low flow and vice versa. However, the correlation
between water residence time and phosphorus concentra-
tion (or change in phosphorus concentration) is a weak one
complicated by the regulated and irregular flow (dams on
the inflow and outflow) and the patterns of inflow water move-
ment (inflow depth, degree of mixing, circulation).
The nitrogen loading and concentration (Fig. 2) during the
period have increased more or less steadily since no reduc-
tion in nitrogen has taken place. As a consequence, the
nitrogen to phosphorus ratios which up to 1974 were relative-
ly low (5-10:1), then rose to about 20:1 (1976-77), have
stabilized at 13-14:1 for the data from 1978 to present (Table
1). This would indicate that the lake which in the early period
was likely limited by nitrogen or both nitrogen and
phosphorus, is probably limited by phosphorus at present
and these changes in N:P ratios have some direct bearing
on the biomass and species composition of the
phytoplankton over the period of record.
Figure 2.—Spring overturn total nitrogen concentrations for Skaha
Lake.
Table 1. — Nitrogen to phosphorus ratios at overturn,
Skaha Lake 1969-1981.
Algal Standing Crop
Data for algal biomass and chlorophyll a are not sufficiently
complete to perceive any detailed year to year trends;
however, it is clear that significant differences exist before
and after 1971. Cell numbers in 1967-70 were very high
(1968 mean summer concentration 39,000 cells/ml) com-
pared to later sampling (1978 mean summer concentration
2,000 cells/ml). There also appears to be some difference
between cell numbers and chlorophyll a in the 1976-78 period
and the 1979-80 data when values tend to be higher. This
pattern parallels the phosphorus concentration but since the
data base is incomplete no fair judgment can be made. There
also appears to have been some shift in species composi-
tion from a community dominated by cyanophytes (1967-70)
to diatoms and phytoflagellates (1976-78) and then a return
to cyanophytes (1979-80).
Water Clarity
The data for Secchi disc transparency (Fig. S^show reduc-
ed water clarity in 1968 corresponding to high algal stand-
ing crop, but otherwise no discernible pattern.
Figure 3.—Secchi disc data for Skaha Lake.
Oxygen Depletion Rates
Hypolimnetic oxygen depletion rates (Fig. 4) do not appear
to show any trend or any effect of the reduced P loading.
The hypolimnetic oxygen reaches minimum concentrations
of 1 to 2 mg/r1 in the deepest water by September. No in-
ternal loading appears to be taking place.
Macrophytes
Another factor that must be considered in the water quality
trends is the colonization of significant areas of the littoral
zone by Myriophyllum splcatum in 1975-77. According to the
most recent survey Myriophyllum covers about 70 ha. The
YEAR
TN
TN
N:P
1969
70
70
71
74
76
77
77
78
78
79
80
81
163
323
202
190
290
217
295
243
323
430
420
325
31.7
41.3
32.3
21
18
13
11
14.5
18
22.8
30.8
32
25
5.14
10.0
9.6
10.5
22.3
19.7
30.3
13.5
14.1.
14.0
13.1
13.0
if
TO 71 TJ 73
75 T« 7T Ifi fg
YEAR
Figure 4,—Oxygen depletion rates for Skaha Lake.
16?
-------
Lake Restoration, Protection and Management
concern was that Myriophyllum might be transferring
phosphorus from the sediments to the water column
(Carpenter, 1980). Using the transfer rates cited by
Carpenter, the estimate of transfer was about 700 kg/yr or
about 5 percent of the annual P load to the lake. As such
it appears to be a relatively minor source; however, this can-
not be confirmed by any direct measurements.
If a direct cause and effect existed between loading and
lake concentration then the loading data for phosphorus sug-
gest there should have been a substantial reduction in
phosphorus concentrations after tertiary treatment and con-
tinued lower levels of phosphorus and biological production
to the present. However present lake phosphorus concen-
trations are almost as high as those measured before ter-
tiary treatment removed approximately 40 percent of the
phosphorus load. The biological production need not
necessarily follow the changes in phosphorus concentration
because the changing N:P ratios may mean that until
phosphorus limitation occurred there may not have been any
change in production or biomass proportional to the
phosphorus concentration.
Also in a short water residence system with varying flow
there is a relatively weak link between spring overturn con-
centration and summer phosphorus supply. The spring
phosphorus concentrations are not the ones which affect the
recreational use of the water in June through August. The
water residence time is such that there is very little memory
of the spring concentration. Nutrient supply in summer is
the overriding concern for controlling algal biomass.
In considering the lack of response in phosphorus con-
centrations three plausible explanations are presented. All
are supported by some aspects of the information available
but none by itself provides a completely convincing explana-
tion for the entire data base.
The simplest explanation of the changes in phosphorus
concentration is that the low concentrations in the mid-1970's
did reflect the effect of tertiary treatment in Penticton and
the increases after 1977 resulted from an increased loading
from other (cultural, nonpoint) sources. This explanation is
easy to make but difficult to substantiate. The origin of the
additional loading could include agriculture, forest harvesting,
and residential growth, particularly in areas outside the boun-
daries of the sewage treatment plant collection system. The
possibility of nutrients originating from lake sediments, both
pelagic (aerobic internal loading) and littoral (originating from
macrophytes at a higher rate than assumed), could be
involved.
The second explanation of the changes is that the in-
fluences of hydrology are sufficiently strong to account for
some of the changes in concentration. To test the impor-
tance of water residence time in Skaha Lake, some relation-
ships from the literature relating loading, concentration, and
water residence time were used. The three relationships and
the data used to test this hypothesis are shown as Table
2. These relationships must be used with caution because
they all assume a steady state and the Vollenweider model
requires an estimate of a sedimentation coefficient, which
can be estimated only very roughly.
First, in a general sense it was necessary to test if the
hydrology could interact with the estimated loading to ac-
count for the range of concentrations that occurred in Skaha
Lake. The matrix of three loading levels and three
hydrological regimes (Table 3) indicates that hydrology is an
important factor. The Vollenweider relationship showed an
overall range from 12 to 45 pgl\ using the extremes of loading
and flow. The Reckhow and Simpson model gave very
similar results indicating that hydrology could have caused
some of the variation in concentration observed in Skaha
Lake.
If this theoretical range of concentrations based on the
Vollenweider and Reckhow/Simpson models for the range
of hydrologic variation experiences since 1968 is plotted
against the measured concentrations over that period the
pattern shown in Figure 5 results. It is apparent that the
overlap between the two periods is significant, and a wide
range of concentrations can be experienced. Because of the
wide envelopes resulting from the hydrologic variation, it ap-
pears to be difficult to isolate the effects of loading by itself.
In a specific sense, the two models that gave similar results
(Vollenweider and Reckhow/Simpson) were then used to
Table 2. — Loading/concentration/residence time relationships.
1. TP =
2 TP-
L
Z (a + 9)
L
(Vollenweider,
(Reckow s
n .e + i .2 qs
3 TP -
YEAR
1968
1969
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
Mean
.603 L
Z (.257+
T
1.6
1.0
2.3
1.4
0.6
2.6
0.7
1.1
0.8
2.5
1.0
2.0
1.5
1.2
(Canfield &
9)
Skaha Lake
V
0.63
1.0
0.43
0.71
1.67
0.38
1.42
0.91
1.25
0.40
1.0
0.5
0.67
1969)
nd Simpson, 1980)
id Bachmann, 1981)
Data
TP
32
41/32
21
18
13
11/14.5
18/23
31/23
24.5/32
25
qs
17.2
27.9
12.1
19.7
47.9
10.6
41.2
26.1
34.0
11.3
28.4
13.7
14.9
30.2
22.4
TP - total phosphorus in mg/l
L - loading in gm-m~1-yr1
Z - mean depth (26 m for Skaha)
o - phosphorus sedimentation coefficient (0.45 for Skaha)*
-------
Pro- and Post-Restoration Assessment Data and Techniques
Table 3. — Theoretical ranges of phosphorus concentrations
due to annual hydraulic variation.
A. Concentration/Loading relationship of Vollenweider (1969)
low loading1 medium loading2 high loading3
low inflow4 25 36 45
medium inflow5 19 26 33
high inflow6 12 17 21
B. Relationship of Recknow and Simpson (1980)
low loading medium loading high loading
low inflow 26 36 46
16 23 29
11 15 19
C. Relationship of Canfield and Bachmann (1981)
low loading medium loading high loading
low inflow 23 32 40
medium inflow 13 19 24
high inflow 7 9 11
1 loading of 12.7 tonnes/yr (0.63 gm/nrffyr)
2 loading of 17.6 tonnes/yr (0.88 gm/rrWyr)
1 loading of 22.4 tonnes/yr (1.11 gm/rrffyr)
' water retention time of 2.6 years (if = 0.38, % = 10.6) eg. 1973
5 water retention time of 1.2 years ( 81
YEAR
Figure 5.—Theoretical ranges for phosphorus concentrations for the
periods before and after tertiary treatment at Penticton.
determine if, using the estimated loading rates for the pre-
and post-tertiary treatment plant period (1.12 gm/mz/yr and
0.63 gm/m2/yr) and the water retention times for each year,
concentrations would result that were close to the measured
concentrations for that year. The results suggest that with
the loading rates and the hydrology pattern given, the
predicted concentrations are reasonably close to the
measured concentrations.
For each of the models used, 7 of the 11 years of data
examined have calculated spring overturn phosphorus con-
centrations within 25 percent of the measured spring over-
turn value (Fig. 6). Three years which both models did not
predict well were 1971,1979, and 1980. The 1971 estimate
substantially overestimates the measured value; this may be
caused by the late sampling data that year. The 1979 and
1980 model values underestimated the measured values;
the most likely cause would appear to be the use of a fixed
loading rate from 1973 through 1982 when it was more like-
ly that loading would be increasing, particularly from non-
point sources noted earlier.
What this analysis implies is that the hydraulic regime can
account for much of the year to year variation in phosphorus
concentration even assuming a fixed loading rate. The low
phosphorus concentrations in the mid 1970's and the rise
in the late 1970's and early 1980's is more a response to
hydrology than a change in loading. Fleming and Stockner
(1975) had predicted increasing phosphorus concentration
as a consequence of increased loading, using Fleming's
(1975) model. Underestimation of loadings (present and 1970
loadings) present a third possible explanation for the lack
of agreement between the loading change (in 1972) and the
failure of the concentration to consistently decrease. If this
was the case, a removal of 10.7 tonnes (tertiary treatment)
would be, in percentage terms, much less, and the effect
on lake concentrations would be less than expected.
These same loading/concentration/residence time relation-
ships were applied to using measured concentrations and
hydrology and solving for the loading term. This indicates
an underestimation of loads in comparison to what would
be expected from the loading/concentration/water residence
relationships. The range for Skaha (16.7 to 23.5 tonnes) ex-
ceeds significantly the 12.7 tonnes estimated for available
phosphorus.
In summary, there appear to be at least three explana-
tions (excluding sampling bias and inaccuracy) for the ap-
parent discrepancy between loading estimates and the pat-
tern of annual phosphorus concentrations in Skaha Lake.
The sampling bias problem must be addressed first. Em-
phasis in this paper has been placed on spring overturn
phosphorus concentrations; however, the timing of the
sampling has inherent problems. All of the spring sampling
was carried out in early spring when the water column was
isothermal (the lake only very rarely has winter ice cover)
but there are indications from inorganic phosphorus and
nitrogen analysis that biological activity may have been well
underway at certain of the sampling times. Also sampling
at slightly different times by different agencies can give dif-
ferent values (Fig. 1).
The other non-sampling factors that influence the pattern
of phosphorus concentration are:
1. Nonpoint source cultural loadings have been increas-
ing. Although the estimated increase in this component has
only increased an estimated 1,200 kg in 10 years, any
changes or inaccuracies in the transmission coefficients used
in deriving these estimates could result in much larger
estimates for this component.
2. Total loading may have been higher than estimated for
both the pre- and post-tertiary treatment period so that the
real reduction after tertiary treatment of the sewage treat-
ment plant effluent was much less than the 40 percent
calculated.
3. One of the major factors controlling concentration is the
flow from Okanagan Lake. When these flows are high, there
can be considerable dilution of the Okanagan River concen-
trations and subsequently epilimnetic Skaha Lake concen-
Rgure 6.—Spring overturn total phosphorus concentrations for Skaha
Lake.
Predicted (D) and Measured values (•)
169
-------
Lake Restoration, Protection and Management
trations. The effective dilution would depend on the time
period of elevated flow and the physical limnology of Skaha
Lake, However, it would be expected that the flushing rate
of the epilimnion would be much higher than the lake as a
whole, particularly in the summer. There were several high
discharge years in the mid-70's while low discharge
dominated the late 70's. This would produce lower than
average concentrations in the former period and higher in
the latter. It would appear that hydraulic factors can be im-
portant factors in examining effects of changes in loading.
Overall, the pattern of lake phosphorus concentration in
1968-81 likely reflects all of the previous factors and the in-
creases are probably a cumulative effect of all three.
REFERENCES
Alexander, D.G. 1982. Summary of nitrogen and phosphorus load-
ings to the Okanagan main valley lakes from cultural and natural
sources. Working rep. Okanagan Basin Implementation Agree-
ment, Penticton.
Canfield, D.E., and R.W. Bachmann. 1981. Prediction of total phos-
phorus concentrations, chlorophyll a, and Secchi depths in natural
and artificial lakes. Can. J. Fish. Aquat. Sci. 38:414-423.
Carpenter, S.R. 1980. Enrichment of Lake Wingra, Wisconsin, by
submersed macrophyte decay. Ecology 61:1145-1155.
Coulthard, T.L., and J.R. Stein. 1968. A report on the Okanagan
water investigation, 1967. Mlmeo. rep., Br. Col. Dep. Lands,
Forests Water Resour., Water Invest. Br., Victoria.
1969. A report on the Okanagan water investigation,
1968-69. Mimeo. rep., Br. Col. Dep. Lands, Forests Water Resour.
Water Invest. Br., Victoria.
Edmondson, W.T., and J.T. Lehman. 1981. The effect of changes
in the nutrient income on the condition of Lake Washington. Urn-
nol. Oceanogr. 26:1-29.
Emery, R.M., C.E. Moon, and E.B. Welch. 1973. Delayed recovery
of a mesotrophic lake after nutrient diversion. J. Water Pollut. Con-
trol Fed. 45:913-925.
Fleming, W.M. 1975. A model of the phosphorus cycle and phytc-
plankton growth in Skaha Lake, British Columbia. Verh, Int. Ver.
Limnol. 19:229-240.
Fleming, W.M., and J.G. Stockner. 1975. Predicting the impacts of
phosphorus management policies on the eutrophieation of Skaha
Lake, British Columbia. Verh. Int. Ver. LJmnol. 19:241-248.
Gray, C.G.J., and R.A. Kirkland. 1982. Nutrient composition and
bioavaiiability in major tributaries and interconnecting rivers of the
Okanagan Basin. Natl. Water Res. Inst, West Vancouver, B.C.
Jensen, E.V. 1981. Results of the continuing water quality monitoring
program on Okanagan lakes for years 1979-80. Rep. to Okanagan
Basin Implementation Board.
Larsen, D.P., and H.T. Mercier. 1976. Phosphorus retention capacity
of lakes. J. Fish. Res. Board Can. 33:1742-1750.
Larsen, D.P., J. Van Sickle, K.W. Malueg, and P.O. Smith. 1979. The
effect of wastewater phosphorus removal on Shagawa Lake, Min-
nesota: phosphorus, lake phosphorus and chlorophyll a.Water
Res, 13:1259-1272.
Pinsent, M.E., and J.G. Stockner, eds. 1974. Limnology of the Major
Lakes in the Okanagan Basin. Tech. Suppl. V. Canada-British
Columbia Okanagan Basin Agreement.
Reckhow, K.H., and J.T. Simpson. 1980. A procedure using model-
ling and error analysis for the prediction of lake phosphorus con-
centration from land user information. Can. J. Fish. Aquat. Sci.
37:1439-1448.
Smith, V.H., and J. Shapiro, 1981. Chlorophyll-phosphorus relations
in individual lakes. Their importance to lake restoration strategies.
Environ. Sci. Technol. 15:444-451.
Stein, J.R., and T.L. Coulthard. 1971. A report on the Okanagan
water investigation. Univ. Br. Col. Prepared for Water Invest. Br.,
Br. Col. Water Resour. Serv. Parliament Buildings, Victoria, B.C.
Truscott, S.J., and B.W. Kelso. 1979. Trophic changes in lakes
Okanagan, Skaha and Osoyoos, B.C. following implementation
of tertiary municipal waste treatment. Progr. rep., Canada-British
Columbia Okanagan Basin Implementation Agreement.
Vollenweider, R.A. 1969. Moglichkeiten and Grenzen elementares
Modelle der Stoffbilanz von Seen. Arch. Hydrobiol. 66:1-36.
170
-------
DETERMINATION OF NUTRIENT SOURCES FOR McNEELY LAKE
G. C. HOLDREN, JR.
Water Resources Laboratory
University of Louisville
Louisville, Kentucky
ABSTRACT
A diagnostic study was conducted to estimate nutrient loadings to McNeely Lake, a hypereutrophic reser-
voir near Louisville, Ky. Major nutrient sources at the present time include four small sewage treatment
plants and runoff from a large drainage basin. Nutrient loadings were estimated from bi-weekly samples
collected from the lake inlets, with additional samples collected from the lake outlet and sewage treatment
plant outfalls. All sewage effluents enter the lake through one inlet, enabling the other inlet to be used
in the estimation of nonpoint loadings. Results indicated sewage effluents contribute 95 percent of the
total phosphorus and 66 percent of the inorganic nitrogen reaching McNeely Lake at the present time.
If all sewage effluents were diverted from the drainage basin, it is likely that the lake would remain eutrophic
but water quality should improve significantly.
INTRODUCTION
McNeely Lake, located in southern Jefferson County, Ken-
tucky, was created by the impoundment of Pennsylvania Run
creek (Penn Run) in 1953. McNeely Lake is the only State
public fishing lake in the metropolitan Louisville area and
much of the adjacent land is a county park, but nuisance
blooms of algae and duckweed interfere with recreational
use of the lake during the summer months. The lake would
be classified as hypereutrophic based on existing water
quality.
Major nutrient sources were expected to include effluents
from four small sewage treatment plants (STP's), runoff from
a relatively large drainage basin, and internal loading. This
investigation estimated current nutrient loadings from these
various sources to provide the information required to eval-
uate various lake restoration alternatives.
Previous investigations (Jones and Bachman, 1978; Hig-
gins and Kim, 1981) indicated existing nutrient loading
models overestimate phosphorus concentrations in reservoirs
because phosphorus retention is higher in reservoirs than
natural lakes. An initial attempt to apply loading models to
McNeely Lake (Holdren and McCoy, 1981) encountered
similar difficulties, but Canfield and Bachman (1981) and
Jones and Lee (1982) indicate the differences arise primari-
ly from problems associated with estimating model param-
eters and are not due to inherent differences between lakes
and reservoirs. Therefore, an attempt was made to improve
estimates of model parameters used in conjunction with
the nutrient loading estimates and nutrient loading models
used to compare observed and predicted nutrient
concentrations.
SAMPLING METHODS
McNeely Lake is located in an urban area where a high in-
cidence of vandalism prevents use of continuous monitors
to measure nutrient loadings or stream flows. As a result,
nutrient loadings were estimated using bi-weekly grab
samples collected from the two lake inlets, the lake outlet,
and sewage treatment plant outfalls (Figure 1). Samples were
collected largely on random dates, but some attempt was
made to include both storm events and dry weather loadings.
Water samples were analyzed for soluble reactive phospho-
rus (SRP), total phosphorus (total P), nitrate and nitrite
+ NOi - N), ammonia (NH3 - N), and fluoride ion (F~)
600 0 600
Apple Valley STP
Maple Grove STP
CD East Inlet (22.U)
I i Direct Inflow (22.4
Figure 1.—McNeely Lake drainage basin. Circles indicate sampl-
ing stations for nutrient loading estimates. Runoff sources and
percentage of total drainage basin area contributing to each source
are also indicated.
concentrations using procedures from the U. S. Environ.
Prot. Agency (1979) and Standard Methods (1980).
Stream flows were calculated from measurements of
cross-sectional area and current velocity, measured with an
Ott current meter. Effluent flows from the Apple Valley, Maple
Grove, and Pines sewage treatment plants were obtained
from flow meters at the plants or from records provided by
the plant operator. Flows for the Pleasant Valley plant were
obtained from records at the Jefferson County Health
Department.
Monthly lake water samples were collected to provide data
on nutrient concentrations and to monitor existing water
quality. Water samples were collected at 1 m depth inter-
vals in the water column from one site in each arm of the
lake and another near the dam (Figure 2) and analyzed for
SRP, total P, NOg + NOs-N, NH3-N, total Kjeldahl
nitrogen (TKN), and several other water quality parameters.
171
-------
Lake Restoration, Protection and Management
Concentrations of N03 + N02 - N and NH3 - N were com-
bined as total inorganic nitrogen (TIN) and N03 + NOg - N
and TKN were combined as total nitrogen (total N) to facilitate
data handling. Differences in nutrient concentrations between
the sampling sites were noted to occur, with levels usually
highest in the Penn Run arm and lowest in the east arm.
These spatial variations in nutrient concentrations were
usually smaller than changes with depth in the water col-
umn, however, and concentrations from three sites were
averaged to calculate whole-lake nutrient concentrations.
Lake sediment samples for estimates of internal nutrient
loading and phosphorus deposition were obtained with a
modified KB gravity corer. Cores from the sampling site near
the dam were sectioned into 2 cm layers and the centrif-
ugation-filtration technique (Holdren et al. 1977} was used
to obtain interstitial water for nutrient loading estimates.
Nutrient concentration gradients were calculated for eight
cores from the differences between interstitial reactive
phosphorus (IRP) and interstitial ammonia concentrations,
and concentrations of SRP and NHg - N in the overlying
water. Cores were collected in the spring and fall, both before
and after the lake stratified, to provide estimates of seasonal
variations in concentration gradients.
Cores from several locations (Figure 2) were obtained for
measurements of phosphorus deposition. Cores were sec-
tioned into annual layers, clearly indicated by varves in the
sediment, for analysis of total sediment phosphorus by the
ignition method (Andersen, 1976). Details of the procedures
used were presented by Mayfield (1982).
NUTRIENT LOADING ESTIMATES
Nutrients contributed to McNeely Lake from sewage treat-
ment plant effluents enter the lake directly (Maple Grove STP)
or indirectly through the Penn Run inlet, while nutrients from
runoff from the drainage basin can enter the lake through
McNeely Lake
the Penn Run and East Inlets, or through direct runoff into
the lake. Internal nutrient loading is expected to occur
primarily into the hypolimnetic waters. Areas of the basin con-
tributing to each of these sources are shown in Figure 1,
Because the McNeely Lake drainage basin to lake surface
area ratio is large (Table 1), nutrient loading from runoff is
expected to be significant, and direct nutrient inputs from
precipitation and dry deposition on the lake surface were con-
sidered negligible.
Table 1. — Morphometric and hydrologic characteristics of
McNeely Lake
Drainage basin area
Lake surface area
Lake volume
Mean depth
Maximum depth
Average discharge
Hydraulic residence time
1.46X107 m2
2.14 x 10s m2
6.42 x 1Q5 m3
3.0 m
9.1 m
1.0 x 104 m3/day
64 days
Figure 2.—McNeely Lake sampling sites. Locations tor collection
of water samples (»), sediment cores for internal loading calcula-
tions (x), and sediment cores for phosphorus retention measurements
(o ), are indicated.
McNeely Lake usually is thermally stratified from May
through October and isothermal the rest of the year. Dif-
ferences in nutrient concentrations measured at the lake in-
lets were also noted between the summer (May to October)
and winter (November to April) months. These variations in
nutrient concentrations were expected because differences
in interception and uptake of nutrients by vegetation and in-
filtration between the summer and winter periods affect the
amount and nature of material transported to the lake.
Because nutrient inputs during different seasons will not have
the same effects on aquatic growth, and because data from
the nearest U. S. Geological Survey (USGS) stream monitor-
ing gage indicate nearly 75 percent of the average annual
discharge occurs during the winter months, nutrient loadings
were calculated separately for the summer and winter
periods and then combined to obtain annual loadings.
Nutrient loadings from various sources for total P, solu-
ble reactive P, and total inorganic N are summarized in
Tables 2 to 4, respectively. A discussion follows of the
methods used to obtain the various nutrient loading rates.
Sewage Treatment Plants
Nutrient concentrations and effluent flows from the four
sewage treatment plants varied considerably as all four plants
were evidently affected by infiltration during storm events.
Effluents from the Apple Valley, Pines, and Pleasant Valley
plants are discharged to Penn Run and its tributaries and
enter McNeely Lake through the Penn Run inlet. Discharge
from the sewage treatment plants was often greater than
measured inflow at the Penn Run inlet during low flow
periods, indicating significant losses occurred from factors
such as evaporation, evapotranspiration, and infiltration. As
a result, nutrient loadings for these three plants were Included
in the Penn Run loading and not calculated separately.
Effluent from the Maple Grove sewage treatment plant
enters McNeely Lake below the Penn Run inlet (Figure 1)
and, because of the short distance involved, most effluent
reaches the lake. Neither nutrient concentrations nor dis-
charge rate exhibited marked seasonal variations, and there
was no significant correlation between effluent discharge rate
and nutrient concentration. Therefore, nutrient loading rates
were calculated from the average effluent nutrient concen-
trations measured during this study and the average daily
discharge obtained from plant records. Average nutrient con-
centrations for total P, SRP, and TIN were 5.7, 5.0, and 8.2
g/m3, respectively, and the mean discharge was 635
m3/day.
172
-------
Pre- and Post-Restoration Assessment Data and Techniques
Nutrient Loading from Runoff
No significant correlations were observed between measured
discharge from McNeely Lake and discharge data from the
nearest USGS gage, or between inflow rates and nutrient
concentrations. As a result, all nutrient loadings for the Penn
Run and east inlets were calculated only from measurements
made during this study and did not use the larger data base
for flow readings provided by the USGS gage. In addition,
average nutrient loadings were calculated from the individual
loadings (flow multiplied by nutrient concentration) rather than
from the average nutrient concentration multiplied by the
average inflow.
Nutrient loadings at the Penn Run inlet arising from runoff
rather than sewage effluents were estimated by using a mass
balance on fluoride ion to calculate the percentage of sewage
flow. Fluoride serves as a good tracer for sewage effluents
because it behaves conservatively and is added to drinking
water at concentrations much higher than those found in
runoff. The east inlet, which receives no sewage effluent,
was used as a control for nutrient and fluoride concentra-
tions in runoff in the McNeely Lake drainage basin. The east
inlet had a relatively constant fluoride concentration of 0.10
g/m3, while fluoride concentrations in sewage treatment plant
discharges averaged 0.90 g/m3 during dry periods when ef-
fluents were not diluted with infiltration. This permitted
estimation of the sewage treatment plant contribution to the
Penn Run inflow using the formula:
QPH(FpR) = QSTP(FiTP)
where QPR = measured inflow at the Penn Run
Inlet,
QSTP = "ow originating from sewage
effluents,
QPR -QSTP = flow from runoff entering the lake at
Penn Run,
(FpR) = measured F~ concentration at the
Penn Run inlet,
(FgTp) = F~ concentration in undiluted STP
effluents = 0.90 g/m3, and
= F~ concentration in runoff entering
the east inlet = 0.10 g/m3.
By supplying measured values for QPR and (FPR), the frac-
tion of Penn Run inflow arising from sewage effluents and
nonpoint runoff was calculated.
Application of equation 1 to experimental results indicated
49 percent of the summer inflow and 66 percent of the winter
inflow at the Penn Run inlet originated from runoff from the
drainage basin. These flows were multiplied by average
nutrient concentrations at the east inlet to obtain the
estimates of nutrient loadings from runoff entering McNee-
ly Lake through the Penn Run inlet. These calculated
loadings may underestimate actual loadings for phosphorus
because the Penn Run drainage is more highly developed
than the east inlet drainage and export coefficients for P (Ut-
tormark et al 1974) are much higher for urban areas than
for rural areas Export coefficients for N are the same for
both of these land use categories, indicating N loadings
would not be similarly affected.
Direct nutrient inputs to McNeely Lake can occur during
storm events from portions of the drainage basin discharg-
ina directly into the lake (Figure 1). Nutrient loadings from
his source were estimated by multiplying the east inlet
nutrient loadings by the ratio of the drainage areas discharg-
, to direct inflow and to the east inlet.
Using nutrient export coefficients based on land use pro-
yes another means of estimating nutrient loadings from
noff Land use in the McNeely Lake drainage basin was
determined from information provided by the Jefferson Coun-
ty Planning Commission and inspection of aerial photographs
taken in April 1979 for the Planning Commission. Land usage
was divided into the four broad categories of forested,
residential, rural/agricultural, and undeveloped.
Nutrient exports to McNeely Lake based on land use
estimates were calculated from the areas in each category
and average nutrient export coefficients found by Uttormark
et al. (1974) and used by Lee and Jones (1981) in the OECD
eutrophieation program. Land use areas and export coeffi-
cients used in these calculations are summarized in Table 5.
Both N and P loadings based on land use are higher than
those calculated for runoff (Tables 2 to 4). This difference
can arise from underestimates of runoff loadings caused by
the use of east inlet nutrient concentrations for all runoff or
from the use of the average nutrient export coefficients that
may not apply to the McNeely Lake basin. Precipitation was
also below average during the study period and this could
have led to underestimates of loading from runoff. The
relative importance of these various factors cannot be readily
determined.
Internal Loading
A number of previous studies (Tessenow, 1972; Bannerman
et al. 1974; Kamp-Nielson, 1974; Holdren et al, 1977; Theis
and McCabe, 1978) indicated internal nutrient loading to
hypolimnetic waters could be estimated from the concen-
tration gradient between interstitial and overlying waters and
simple diffusion models.
Table 2, — Estimated total P loadings to
McNeely Lake
Total P Load
Source
May- November-
Annual October April
g/yr g/day g/day
Maple Grove STP 1.3 x 106 3.6 x 103 3.6 x 1Q3
East inlet 7.4 * 104 47 360
Penn Run inlet 5.2 x 106 6.8 x 103 2,2 x 10*
Penn Run, without sewage 1.6 x irjs 140 720
Direct inflow 7.4 x 10" 48 360
Internal loading 2.7 x 104 97 47
Total, present conditions 6.7 x 106 1,1 x io4 2.6 x io4
Total, without sewage 3.3 x 105 330 1.5 x 103
Total, based on land use 9.6 x 105 — —
Table 3. — Estimated soluble reactive P loadings to
McNeely Lake
SRP Load
Source
May- November-
Annual October April
g/yr g/day g/day
Maple Grove STP 1.2 x 1Q6 3,2 x 103 3.2 x in3
East inlet 2.2 x 104 14 110
Penn Run inlet 3.5x10s 6.0 x 103 1.3 x io4
Penn Run, without sewage 7.2 x io4 56 340
Direct inflow 2.2 x 1Q4 14 -no
Internal loading 2.7 x 10" 97 47
Total, present conditions 4.8xiQ6 9.3xio3 1.6xio4
Total, without sewage ' " ""
Total, based on land use1
1.4 x 1Q5 180
9.6 x 1QS _
610
1 Tolal P toad.
173
-------
Lake Restoration, Protection and Management
Table 4. — Estimated inorganic N loadings to
McNeely Lake
Inorganic N Load
Source
Maple Grove STP
East inlet
Penn Run inlet
Penn Run, without sewage
Direct inflow
Internal loading
Total, present conditions
Total, without sewage
Total, based on land use1
Annual
g/yr
1.9x 10s
7.0 x 105
9.1 xi06
1.3x 106
7.1 x 1Q5
1.6x 10s
1.4x 107
4.3 x 106
6.6 x 1Q6
May-
October
g/day
5.2 x 103
380
1.8x10"
770
390
6.1 * 103
3.0 x 10"
7.6 x 103
—
November-
April
g/day
5.2 x 1Q3
3.5 x 1Q3
3.2 x 10"
6.3 x 103
3.5x103
2.7 x 1Q3
4.7 x 10"
1.6x 10"
—
1 Total N load.
Table 5. — Land use and nutrient export coefficients
Land use
Forested
Residential
Agricultural
and undeveloped
Export coefficient1
P N
g/m2/yr
0.02 0.25
0.15 0.5
0.03 0.5
Land use area,
McNeely basin
m2
2.4 x 1Q6
4.5 x 1Q6
5.4 x 1Q6
2.3 x 10s
1 From Utlormark et al. 1974
3c
3x
Rates of nutrient release from hypolimnetic sediments
were calculated using Pick's first law of diffusion (Berner,
1980):
Js = Ds(J^), (2)
where Js = nutrient flux from the sediments
(mg/m2/day),
<)> = sediment porosity, the fraction of
sediment volume occupied by water,
Ds = molecular diffusion coefficient in the
sediment (cm2/sec), and
the concentration gradient across the
sediment water interface (g/m3/cm).
McNeely Lake surface sediments average 70 percent water
by weight with an average density of 1.1 g/cm3, yielding a
porosity of 0.77. Values for the other parameters used in the
calculations are summarized in Table 6. Diffusional nutrient
releases were assumed to occur over the hypolimnetic area
of 1.44 x io5 m2 and were calculated separately for the
summer stratification and winter periods because both in-
terstitial nutrient concentrations exhibited seasonal variations.
Internal nutrient loading at the present time is reduced by
high nutrient concentrations in the bottom water of McNee-
ly Lake, but concentration gradients and internal loading
could increase if nutrient loadings and whole-lake nutrient
concentrations were reduced.
APPLICATION OF RESULTS
Input-output models based on the nutrient loading-mean
depth relationship found by Vollenweider (1968) have been
widely used to assess lake trophic state and predict changes
in water quality resulting from changes in nutrient inputs.
The original model has been altered to include the effects
of hydraulic residence time on the response of lakes to
nutrient loading (Dillon and Rigler, 1974; Vollenweider, 1975,
1976). The models still include the suggestion by
Vollenweider (1968) that "permissible" and "dangerous"
loadings would be loadings resulting in mean phosphorus
concentrations of 10 mg/m3 and 20 mg/m3, respectively.
Although the models have been developed for phosphorus
loadings and concentrations, the 15:1 ratio of N:P required
for algal growth (Vollenweider, 1968) can be used to apply
the results to nitrogen loadings.
Results of this investigation were used in conjunction with
three existing models that have been found to be applicable
to a wide range of lakes to determine the ability of these
models to predict nutrient concentrations in McNeely Lake.
The models used were those developed by Dillon and Rigler
(1974), Vollenweider (1975 and 1976), respectively:
where
[P] =
[P] =
[P] =
[P] =
L =
R =
Q =
z =
L(1-R)/ez,
L/(10+ez), and
(3)
(4)
(5)
mean phosphorus concentration
(g/m3),
areal P loading rate (g/m2/yr),
phosphorus retention coefficient,
flushing rate (yr~1),
mean depth (m), and
hydraulic residence time (yr).
These models differ only in their treatment of internal
phosphorus losses, described by the specific sedimentation
rate, a. Equation 3 uses a phosphorus retention coefficient
to account for sedimentation losses. A value for R can be
determined experimentally or estimated from an equation
derived by Larsen and Mercier (1976) and Vollenweider
(1976):
R = 1/(1+v^). (6)
Equation 4 was derived using the relationship calculated by
Vollenweider (1975):
a = 10/z, (7)
and equation 5 can be derived by substituting equation 6
into equation 4.
Values of the model parameters for McNeely Lake were
estimated using sampling program results. The mean
nutrient concentrations were calculated from monthly water
analyses for the period September 1980 to August 1981,
which coincided with the collection of most of the runoff
samples. Nutrient loading estimates were previously describ-
Table 6. — Parameters used for internal nutrient
loading calculations
IRP/SRP
NH3-N
Parameter
Summer Winter Summer Winter
Interstitial nutrient
concentration (g/m3)
Diffusion coefficient
(cm2/sec)
Concentration gradient
(g/m3/cm)
Nutrient flux
(mg/m2/day)
2.21 1.55 26.4 10.0
1.0x10-6' 3.0x10-6*
1.0 0.50 21 9.3
0.69
0.34
44
19
Internal loading (g/day) 97
47 6.1 x 10-
3 2.7x10-
3
1 From Stumm and Leckie (1971)
! From Imboden and Lerman (1978).
174
-------
ed. A value for R was calculated from directly measuring
phosphorus deposition. Mayfield (1982) analyzed a series
of cores and found a value for average annual phosphorus
deposition of 13 g/m2/yr. This was compared to the
phosphorus loading rate of 31 g/m2/yr to calculate a value
of R = 0.42. Values for flushing rate and hydraulic residence
time were calculated from the average sum of measured and
calculated inflows rather than from measured outflow
because of larger errors in the measurement of the latter
value. Table 7 summarizes the values of the parameters us-
ed in the nutrient loading models.
A comparison of observed phosphorus concentrations and
those predicted from phosphorus loading models is
presented in Table 8. It is readily seen that all three models
overestimated total P concentrations in McNeely Lake. Bet-
ter agreement was observed between actual and predicted
SRP concentrations; equation 3, using an experimental
measurement of phosphorus retention, produced the best
results in each case. This indicates that deficiencies in
estimating internal losses are the major cause for the
discrepancies observed when loading models are applied
to reservoirs.
Because equation 3 produced the best agreement bet-
ween observed and predicted phosphorus concentrations,
it was used to calculate the "dangerous" nutrient loadings
for McNeely Lake for comparison with nutrient loading in the
absence of sewage effluents. The calculated loadings of 1.5
g/m2/yr for total P, 0.65 g/m2/yr for SRP, and 20 g/m2/yr for
TIN (Tables 2 to 4) are still greater than the "dangerous"
loading of 0.6 g P/m2/yr calculated from equation 3 and the
"dangerous" loading of 9 g N/m2/yr calculated by assum-
ing a 15:1 ratio of N:P.
Table 7. — Values for nutrient loading model parameters
Parameter
Units Value
Mean depth, z m 3.0
Hydraulic residence time, TW yr 0.175
Flushing rate, Q yr~1 5.7
Areal water load, qs m/yr 17
Phosphorus retention coefficient, R — 0.42
Nutrient loading rate, L
Total P g/m2/yr 31
SRP g/m2/yr 21
TIN g/m2/yr 70
Mean annual nutrient concentration, [P], [N]
Total P1 g/m3 0.89
SRP1 g/m3 0.71
Total N2 g/m3 2.6
TIN1 g/m3 2.3
' Average 9/80-8/81
2 Average 9/80-7/81
Pre- and Post-Restoration Assessment Data and Techniques
These results indicate considerable improvement in water
quality could be expected if all sewage effluents were remov-
ed from the drainage basin, although some algal blooms
could still be expected. Improvements may be better
than predicted, however, because the loading models may
overestimate phosphorus concentrations in McNeely Lake.
More important, most of the nutrient loadings from runoff
occur during the winter months and may not fully contribute
to algal growth during the critical summer growing period.
CONCLUSIONS
Current nutrient loadings to McNeely Lake far exceed
amounts necessary to create nuisance aquatic growth. Total
P loadings are nearly 60 times higher than suggested limits
and inorganic N loadings are over 7 times higher than the
suggested limit. Sewage effluents account for approximately
95 percent of the total P, 96 percent of the SRP, and 66
percent of the inorganic N loadings at the present time.
In the absence of sewage effluents, runoff would become
the major nutrient source, but internal loading could con-
tribute approximately 10 percent of the total P load and 40
percent of the total N load if interstitial nutrient concentra-
tions did not decline. Although a significant improvement in
water quality could be expected, it is likely that nuisance algal
blooms would continue to occur.
ACKNOWLEDGEMENTS: This project has been financed in part
with Federal funds from the Environmental Protection Agency under
grant number S004402010 administered by the Kentucky Depart-
ment for Natural Resources and Environmental Protection. Addi-
tional support was provided by the Graduate School at the Univer-
sity of Louisville. The following individuals assisted with various
phases of this project: L. E. McCoy, J. D. Mayfield, and A. J. Elpers,
University of Louisville; and T. Anderson and S. Porter, KDNREP.
The assistance of the Andriot-Davidson Co. in providing records
for the sewage treatment plants is also acknowledged.
REFERENCES
Andersen, J. M. 1976. An ignition method for determination of
total phosphorus in lake sediments. Water Res. 10:329-331.
Bannerman, R. T., D. E. Armstrong, G. C. Holdren, and R. F
Harris. 1974. Phosphorus mobility in Lake Ontario sediments.
Proc. 17th Conf. Great Lakes Res.
Berner, R. A. 1980. Early diagenesis. A theoretical approach.
Princeton University Press, Princeton, N.J.
Canfield, D. E., Jr. and R. W. Bachmann. 1981. Prediction of
total phosphorus concentrations, chlorophyll a, and secchi depth
in natural and artificial lakes. Can. J. Fish. Aquat. Sci. 38:414-423.
Dillon, P. J., and F. H. Rigler. 1974. The phosphorus-chlorophyll
relationship in lakes. Limnol. Oceanogr. 19:767-773.
Higgins, J. M., and B. R. Kim. 1981. Phosphorus retention mo-
dels for Tennessee Valley Authority reservoirs. Water Resour.
Res. 17:571-576.
Table 8. — Comparison of observed and calculated phosphorus concentrations
Observed
concentration
Calculated concentration
[P]=L(l-R)/ez1
] = L/(10 + ez)2
Total P
SRP
0.89
0.71
1.05
0.71
g/m3
1.1
0.77
1.3
0.86
< Dillon and Rigler (1974)
J Vollenweider (1975)
' Vollenweider (1976)
175
-------
Lake Restoration, Protection and Management
Holdren, G. C., D. E. Armstrong, and R. F. Harris. 1977. Inter-
stitial inorganic phosphorus concentrations in Lakes Mendota
and Wingra. Water Res. 11:1041-1047.
Holdren, G. C., and L. E. McCoy. 1981. Modification of nutri-
ent loading models for trophic state prediction in a small reser-
voir. Pages 447-454 hi H. G. Stephan, ed., Proc. Symp. Surface-
Water Impoundments. Am. Soc. Civil Eng. New York.
Imboden, D. M., and A, Lerman. 1978. Chemical models of
lakes. Pages 341-356 in A. Lerman, ed. Lakes: Chemistry,
Geology, Physics. Springer-Verlag, New York,
Jones, J. R., and R, W. Bachmann. 1978. Phosphorus removal by
sedimentation in some Iowa reservoirs. Verh. Int. Verein. Lim-
nol. 20:1576-1580.
Jones, R. A., and G. F. Lee. 1982 Recent advances in assessing
impact of phosphorus loads on eutrophication-reiated water quali-
ty. Water Res. 16:503-515.
Kamp-Nielson, L, 1974. Mud-water exchange of phosphate in
undisturbed sediment cores and factors affecting the exchange
rates. Arch. Hydrobiol. 73:218-237.
Larsen, D. P., and H. T. Mercier. 1976. Phosphorus retention
capacity of lakes. J. Fish, Res. Board Can. 33:1742-1750.
Lee, G. F., and R. A. Jones. 1981. Study program for development
of information for use of OECD modeling in water quality
management, Am. Water Works Assn. Qual. Control in Reser-
voirs Comm. Rep. Submitted.
Mayfield, J, D. 1982. Phosphorus and organic matter in the
sediments of McNeely Lake, Jefferson County, Ky. M. S. Thesis,
University of Louisville.
Standard Methods for the Examination of Water and Wastewater.
1980. 15th ed. Am. Pub. Health Assn., Washington, D. C.
Stumm, W., and J. O. Leckie. 1971. Phosphate exchange with
sediments: Its role in the productivity of surface waters. Proc.
5th Int. Water Pollut. Res. Conf.
Tessenow, U. 1972. Solution, diffusion and sorption in the upper
layer of lake sediments. I. A long-term experiment under aerobic
and anaerobic conditions in a steady-state system. Arch.
Hydrobiol. Suppl. 38:353-398.
Theis, T. L., and P. J. McCabe. 1978. Phosphorus dynamics in
hypereutrophlc lake sediments. Water. Res. 12:677-685.
U. S. Environ. Prot. Agency. 1979. Methods for chemical analysis
of water and wastes.EPA/4-79-020. Environ. Monitor. Support
Lab. , Cincinnati, Ohio.
Uttormark, P. D., J. D. Chapin, and K. M. Green. 1974. Esti-
mating nutrient loadings of lakes from nonpoint sources.
EPA-660/3-74-020. Natl. Environ. Res. Center, U. S. Environ,
Prot. Agency, Corvaliis, Ore,
Vollenweider, R. A. 1968. Scientific fundamentals of the eutro-
phieation of lakes and flowing waters, with particular reference
to nitrogen and phosphorus as factors In eutrophication. Organ.
Econ. Coop. Develop. Rep. DAS/CS1/68.27.
1975. Input-output models with special reference to
the phosphorus loading concept in limnology. Schweiz. Z, Hydrol,
37:53-84,
1976. Advances in defining critical loading levels for
phosphorus in lake eutrophication. Mem. 1st. Ital. Idrobiol,
33:53-83.
176
-------
A METHOD FOR INFORMATION POOLING TO REDUCE
LAKE MODEL PREDICTION ERROR
KENNETH H. RECKHOW
School of Forestry and Environmental Studies
Duke University
Durham, North Carolina
ABSTRACT
It is not uncommon to conduct a 1-year study of phosphorus inputs to, and outputs from, a lake as
part of the lake management planning process. Upon acquiring this information, the analyst typically
uses a model to evaluate alternative management strategies. The analyst may use the model to estimate
phosphorus trapping (e.g., retention or settling) or the analyst may use the lake input-output
measurements to estimate trapping, but never are both used. This is an inefficient use of information.
It is possible to pool the information from the (cross-sectional regression) model and the input-output
measurements to calculate a phosphorus trapping parameter that is estimated with less uncertainty
than is the model trapping parameter or the measured trapping parameter. This procedure is
demonstrated using the phosphorus retention coefficient as the phosphorus trapping parameter. The
model employed is the Kirchner-Dillon regression model, but the procedure may easily be applied
with any regional (cross-sectional) model and lake-specific input-output measurements.
INTRODUCTION
One of the major weaknesses of lake water quality models
is the magnitude of the prediction error term. Although
most applications of lake models are deterministic, this
does not hide the fact that prediction error in, for exam-
ple, eutrophication models, is ± 30 percent or more. This
fact has probably led to reluctance on the part of engineers
and planners to report error terms. As a result, lake
management decisions are often made without an ap-
preciation of the large uncertainties present that create
substantial hidden risks in these decisions.
It therefore may be said that to reduce planning risks
and perhaps to stimulate the reporting of model predic-
tion errors, methods are needed to reduce these error
terms. One approach has been suggested by Reckhow
(1981) using existing lake data and employing the model
to predict the impact of only watershed land use changes.
Under certain conditions, this method may reduce predic-
tion error by 50 percent or more.
Another approach is suggested when one considers the
strategy followed by many scientists and engineers when
a thorough lake eutrophication study is undertaken for
management purposes. Not infrequently, the investigators
measure inputs and outputs of nutrients (usually
phosphorus) over a 1-year period (a 1-year nutrient
budget). These measurements are then used to confirm
the applicability of a simple phosphorus model that is to
be applied to model the lake. The data may be used to
estimate nutrient loadings (which may not be terribly
helpful since planning implies projection of the impact of
unrealized land uses), but they rarely are applied to
estimate phosphorus trapping (retention or settling) in the
lake. Valuable information is therefore lost.
Consider the commonly used phosphorus trapping
parameter that is expressed as percent phosphorus re-
tained in lake. This phosphorus retention parameter, RP,
may be estimated on the basis of a 1-year phosphorus
budget study or from a regional regression for retention
(as presented in Kirchner and Dillon, 1975). In the past,
investigators have used one of these estimates of RP to
model a lake, but not both. Upon calculation of RP, it is
substituted into the following input-output model:
p= L
— (1 -RP) (1)
Ps
where:
P = lake phosphorus concentration [M.lT3]
L = annual areal phosphorus loading [M L-2T~1]
qs = annual areal water loading [LT~1]
Equation 1 can now be used to evaluate projected manage-
ment strategies.
AN IMPROVED RETENTION ESTIMATOR
It is possible to pool the estimates of phosphorus retention
(or of any other phosphorus trapping parameter) to deter-
mine a value for retention that will always be estimated wth
less uncertainty than are the "measured" retention or the
"regression" retention. Using the first and second moments
(mean and variance), we can estimate a mean and variance
for the '^pooled" retention parameter in the following man-
ner. For both the measurement-derived retention coefficient
and the regression retention coefficient, we can calculate
a mean and a variance as shown in the following equations.
Using these statistics, the pooled retention coefficient (Rpp)
may be expressed as a weighted function of the measured
retention coefficient (Rpm) and the regression retention coef-
ficient (Rpr):
RPP = (1 - k) Rpm + k Rpr
(2)
where k is a weight between zero and one. The expected
value of Rpp is:
= (1 - k) E (Rpm) + k E(Rpr) (3)
iance of Rpp is:
Var(Rpp) = (1 - k)2 Var (Rpm) + k* Var (Rpr) (4)
and the variance of Rpp is:
177
-------
Lake Restoration, Protection and Management
The necessary equations are found by minimizing the
variance in Equation 4:
3Var(Rpp)
3k
Solving this, the minimum variance estimate of the
weight is:
k*= Var(Rprn)
Var(Rpm) + Var(Rpr)
This results in a pooled mean for Rp of:
-2(1 - k)Var(Rpm) + 2kVar(Rpr) = 0 (5)
(6)
Rpp =
= (1 - k*)E(Rpm) + k*E(Rpr) (7)
and a pooled variance of:
Var(Rpp) =
1
-I
-1
(8)
[Var(Rpm) Var(Rpr)J
The mean and variance in Equations 7 and 8 are the mean
and variance for the pooled estimate of phosphorus reten-
tion. These statistics may be used to characterize a predic-
tive probability density function for Rpp (Reckhow and
Chapra, 1983).
AN APPLICATION
The pooling of information on lake phosphorus retention re-
quires an error analysis (Reckhow and Chapra, 1979) of the
Kirchner-Dillon model for Rp. This analysis was applied by
Reckhow and Chapra to Lake Charlevoix (an oligotrophic
lake in Michigan) using data from the EPA National
Eutrophication Survey (U.S. Environ. Prot. Agency, 1975).
The data from Lake Charlevoix (Table 1) are used in a
demonstration of the method proposed herein.
For this example, the measured Rp is estimated from a
1-year phosphorus budget (U.S. Environ. Prot. Agency, 1975)
and the regression Rp is estimated from the Kirchner-Dillon
regression model with the corresponding Reckhow-Chapra
error analysis. The measured mean may be calculated by
re-expression of the retention version of the input-output
model (Eq. 1) as:
(9)
P, qs, and L are directly measured or may be easily
calculated from the data usually obtained in a 1-year
phosphorus budget; for Lake Charlevoix these values are
presented in Table 1. Substitution of these values into Equa-
tion 9 results in a measured mean for phosphorus retention
of:
pm
(0.007 mg/l) (5.44 m/yr)
0.12 g/m2-yr
The measured variance may be calculated using first order
error analysis (Reckhow and Chapra, 1983) on Equation 9.
The error propagation equation:
Var(y) =
Var(xi) +
l^- ^r,x.(
3X; 3xi ')
i(Var(Xi))l/2(Var(Xj))1/2 (10)
is used to transform error in the independent variables (P,
qs, and L) in Equation 9 into error in the dependent variable
(Rp). The error propagation equation contains a sensitivity
factor (the partial derivative of the dependent variable, y, with
respect to each of the independent variables, x), error terms
for each independent variable (Var(x)), and correlation coef-
ficients (r) between the independent variables. Since there
is generally no objective means of estimating most of the
error and correlation terms, we rely on judgment and ex-
perience; for this example these terms are subjectively deter-
mined and are presented in Table 2.
Applying the error propagation equation to Equation Q, we
calculate the variance in the measured phosphorus reten-
tion as:
Var(Rp) =
(Var(P)) +
(Var(L))
, 3R,
(Var(qs))
+ 2-
3L
rPL(Var(P))^(Var(L))V2
rpqs(Var(P))v>(Var(qs))V2
rLqs(Var(L))1/>(Var(qs))"*
(11)
with the variance and correlation terms defined in Table 2.
Substitution of the numerical values for these terms into
Equation 9 results in a measured variance of:
Var(Rpm) = (-9s ) 2 (Var(P)) + (-f )2 (Var(L))
+ (.f-)2 (Var(qs))
2 -
s) rPL(Var(P))^(Var(L))
%
Solving:
'0.12
Table 1. — Data1 for Lake Charlevoix example
Term definition
Symbol and numerical value
Observed phosphorus concentration
Annual areal phosphorus loading
Annual areal water loading
P = 0.007 mg/l (fall mean and median)
L = 0.12g/(m2-yr)
qs = 5.44 m/yr
1 From U.S. Environ. Prot. Agency, 1975.
178
-------
Pre- and Post-Restoration Assessment Data and Techniques
Table 2. — Error terms and correlation coefficients for the Lake Charlevoix example
Term definitions
Symbol and numerical value
Standard error for mean phosphorus concentration
Standard error for phosphorus loading
Standard error for areal water loading
Correlation between P and L
Correlation between P and qs
Correlation between L and qs
sp = V Var(P) = 0.005 mg/l
SL = V Var(L) = 0,04 g/m^yr
sq = \A7ar(q^) = 1.0 m/yr
+ 2
.5.44. ,(0.007) (5.44),
V0.12'
(0.12)2
) (0.8) (0.0005) (0.04)
Var(Rpm) = 0.0046
sRpm = V Var(Rpm) = 0.068
The regression mean is obtained from the Reckhow-
Chapra version of the Kirchner-Dillon regression model
(Reckhow and Chapra):
Rp = 0.4088 exp(- 0.2899 q^ + 0.5912 exp(-0.01019
(12)
For the Lake Charlevoix example, this yields a regression
mean phosphorus retention of:
R~pr = 0.644
The regression variance is obtained from the prediction er-
ror for this model applied to Lake Charlevoix. The predic-
tion error for this regression equation is a function of model
error and parameter error. Reckhow and Chapra use first-
order error analysis to propagate the parameter error through
the model so that it can then be combined with the model
error; this is illustrated graphically and is reproduced in Figure
1 . The total prediction error for the Lake Charlevoix exam-
ple may be calculated as demonstrated in Reckhow and
Chapra or read from the Total Error line on Figure 1 ; it is:
'N/ Var(Rpr) = 0.103
Var(Rpr) = 0.0106
U.V2
0.10
008
§006
01
004
002
0 00
L
— Total Error
— Model Error
-
-
_-. Parameter Error
'<- ^^ •--"' ~~~
—
~ 1 1 1 1 1 1 1
50 100 150
qs(m/yr)
- — ._
1 1 1
200 250
Figure 1.—Error terms for the Kirchner-Dillon Model (from Reckhow
and Chapra 1979).
The measured and regression means, variances, and
standard errors are summarized in Table 3. We can com-
bine the measured and regression statistics (characterized
by the means and variances) using the equations presented
to calculate the pooled value for phosphorus retention (Table
4).
The parameters (mean and variance) for the pooled reten-
tion coefficient are calculated using Equations 7 and 8. To
solve these equations, calculate k*, which is:
Var(R.
'pm;
Var(Rpm) + Var(Rpr)
k* =
0.0046
0.0046 + 0.0106
This results in a pooled mean of:
= 0.3026
Rpp = (1-0.3026)(0.6827) + (0.3026)(0.644)
Rpp = 0.671
and pooled variance of:
Var(RPP> = [ 00
= 0.02 L
1
0046 0.0106
Var(Rpp) = 0.0032
SRPP = V Var(Rpp) = 0.057
Table 3. — Parameters for the Lake Charlevoix example
Source
Measurements
Regresssion
Mean
0.683
0.644
Variance
0.0046
0.0106
Standard error
0.068
0.103
Table 4. — Pooled parameters for the Lake Charlevoix
example
Term
Mean
Variance
Standard error
pp
0.671
0.0032
0.057
CONCLUSIONS
The pooling of measurement and regression information to
obtain an improved estimate of retention is graphically
presented in Figure 2 for a normal distribution character-
179
-------
Lake Restoration, Protection and Management
f(Rp)
0 0.2 0.4 0.6 0.8 1.0
Table 5. — A summary of phosphorus retention parameters
for the Lake Chartevoix example
Measurements
Mean
Regression
Pooled
0.683
0.644
0.671
Standard error
Measurements
Regression
Pooled
0.068
0.103
0.057
Figure 2.—Error distributions for phosphorus retention.
ized by the means and variances for each Rp. In Figure 2,
the data label refers to the measured Rp, the model label
refers to the regression Rp, and the posterior label refers
to the pooled Rp. Note that the pooled variance is smaller
than the variance for the measurements or for the regres-
sion. For the example, this means that the uncertainty in the
pooled phosphorus retention is less than the uncertainty from
the measured nutrient budget and less than the uncertainty
from the cross-sectional or regional regression model. Note
in particular the substantial improvement in standard error
from the regression model to the distribution for the pooled
retention coefficient. These error terms (and the means) are
summarized in Table 5. Pooling information from the two
sources thus has the desired effect of reducing error.
In conclusion it is useful to note that error reduction
methods can be employed a priori to assess the value of
lake-specific sampling for the reduction of prediction error.
For the method using existing lake data (Reckhbw, 1982),
additional lake sampling reduces the standard error of the
mean for the lake measurements. For the information-pooling
procedure, the analyst can use Equation 11 to determine
the error in phosphorus retention resulting from a 1-year
phosphorus budget study. The impact of this error term can
then be assessed using the method proposed herein. In this
manner the analyst may use the error reduction methods
to guide data collection and later to estimate prediction uncer-
tainty for alternative lake management strategies.
REFERENCES
Kirchner, W.B., and P.J. Dillon. 1975. An empirical method of
estimating the retention of phosphorus in lakes. Water Resour.
Res. 11:182-183.
Reckhow, K.H. 1982. A method for the reduction of lake model pre-
diction error. Water Res. In press.
Reckhow, K.H., and S.C. Chapra. 1979. A note on error analysis
for a phosphorus retention model. Water Resour. Res. 15:1643-1646.
1983. Engineering approaches for lake management, Vol.
I: Data Analysis and Empirical Modeling. Ann Arbor Science
Publishers, Inc. Ann Arbor, Mich.
U.S. Environmental Protection Agency. 1975. National Eutrophica-
tion Survey working paper on Lake Charlevoix. Corvallis, Ore.
180
-------
Physical, Chemical, and
Biological Control of Aquatic
Macrophytes
REVIEW OF MANAGEMENT TACTICS FOR INTEGRATED AQUATIC
WEED MANAGEMENT OF EURASIAN WATER MILFOIL
(MYRIOPHYLLUM SPICATUM), CURLYLEAF PONDWEED
(POTAMOGETON CRISPUS) AND ELODEA (ELODEA CANADENSIS)
STANLEY A. NICHOLS
University of Wisconsin-Extension
Madison, Wisconsin
BYRON H. SHAW
University of Wisconsin-Extension
Stevens Point, Wisconsin
ABSTRACT
Integrated aquatic weed management (IAWM), a special case of the agricultural communities' Integrated
Pest Management program, is a holistic, systems approach to solving aquatic weed problems. One princi-
ple of IAWM is to consider all available management tactics in an overall management strategy. As input
for an IAWM model, we reviewed the ecological life history of each species and the following manage-
ment tactics: chemical control; mechanical control including dredging and harvesting; habitat manipula-
tions techniques including the use of shades, dyes, bottom coverings, and drawdown; and biological con-
trols including the use of fish, shellfish, insects, disease, and competitive plants. Use, efficacy, associated
impacts, and costs were reviewed for each technique. Because of the need for brevity, only the efficacy
and associated impacts of the techniques are reviewed in this paper.
INTRODUCTION
The three species selected for this review are common
aquatic plants in the northern United States and southern
Canada. A recently completed survey by Mitre Corp. (pers.
comm.) indicates that they are three of the four most com-
mon aquatic nuisance plant species in the northeastern
United States. Hesser and Gangstad (1978), reporting on
the Columbia and Snake River watersheds, Dunst and
Nichols (1979), reporting on Wisconsin, and Mulligan et al.
(1976), reporting on New York State, all concur that
Myriophyllum splcatum, Elodea canadensis, and
Potamogeton crispus are important aquatic nuisances in their
region.
Chemical Control
A variety of chemicals including 2,4-D, Diquat, Endothal,
Simazine, Fenac, Dichlobenil, Acrolein, Fluridone, and cop-
per compounds are available for aquatic weed control.
However, Endothal, Diquat, and 2,4-D are the chemicals of
choice for controlling the three species of interest in multi-
181
-------
Lake Restoration, Protection and Management
pie use recreational waters. Their use is reviewed in Table
1. Water use restrictions resulting from toxicities or slow
breakdown, or lack of proven efficacy, limit the use of the
other chemicals.
Because of large costs associated with pesticide develop-
ment and registration, few new aquatic herbicides have been
developed in recent years. There has, therefore, been con-
siderable research into more efficient use of existing her-
bicides and herbicide combinations.
Combinations of herbicides and metal ions have been
shown to increase the efficacy of weed control, often with
lower concentrations of chemicals used. Combinations of Di-
quat and CuSO4 and Endothal and CuSO4 have been shown
to be effective for weed control at concentrations lower than
required for the herbicides alone (Sutton et al. 1970,1971;
Yeo and Dechoretz, 1977). Stanley (1974b) found synergistic
effects between 2,4-D and HgCI2, AICI3, NaCI, and NaAsO4
to be greater than with CuSO4.
The use of invert and bivert emulsions of herbicides has
been shown to effectively control weeds using only 20 per-
cent the amount of herbicides used by ordinary application
methods (Pitting, 1974; Baker et ai. 1975). The invert or bivert
adheres to the plants, putting the herbicide in direct contact
with the plant and minimizing the concentrate in the water.
Another method of herbicide application that can reduce
the amount of herbicide used, its concentration in the en-
vironment, and provide season-long control is the use of slow
release formulations of herbicides. Much of this work is still
in the development stage, but results look favorable (Har-
ris, 1973; Steward, 1981; Harris and Talukder, 1982; Hoep-
pel and Westerdahl, 1982).
Herbicides are usually most effective at water
temperatures about 15 to 18°C, in water with low turbidity,
and on young plants. Water hardness and high calcium con-
centrations have been shown to increase the efficiency of
herbicides (Parker, 1960; Stanley, 1975). After the water has
reached 18°C and before weeds develop seed is the most
effective time to apply herbicides. Regrowth later in the sum-
mer or growth of weeds resistant to the initial herbicide treat-
ment may require additional application later in the year.
Chemical weed control either directly affects the aquatic
environment through herbicide toxicity or causes secondary
effects resulting from loss of weeds (Brooker and Edwards,
1975). Toxicity of herbicides to aquatic life is extremely
variable even for different formulations of the same herbicide
(Table 1). For instance, the amine formulation of Endothal
and ester formulations of 2,4-D are much more toxic than
is dipotassium endothal or 2,4-D amines. Acute toxicity data
for pesticides have been summarized by Johnson and Finley
(1980).
Direct toxic effects on aquatic life can be minimized by
carefully selecting herbicides and applying them properly.
Some fish have been shown to exhibit an avoidance reac-
tion to herbicides (Folmar, 1976). This indicates that if an
escape route is provided the toxic effects on fish may be
avoided.
Secondary effects related to weed destruction can more
drastically affect aquatic life than the herbicide itself. The
oxygen-carbon dioxide balance is upset because of decreas-
ed photosynthesis and increased metabolism of dying
vegetation (Brooker and Edwards, 1973b). This decreases
oxygen concentrations which occasionally kills fish in lakes
having heavy weed growth and where a fast-acting herbicide
is used over a large portion of the lake (Jewell, 1970; Pokomy
et al. 1971; Way et al. 1971; Brooker, 1974; Brooker and
Edwards, 1973a,b). The toxic effect of low oxygen can be
enhanced by the associated high carbon dioxide concen-
tration (Brooker and Edwards, 1975). This problem can usual-
ly be avoided by treating only part of the lake at a time or
applying herbicides before a large plant biomass develops.
Decomposition processes also release nutrients into the
water (Kormandy, 1968; Boyd, 1970; Jewell, 1970; Nichols
and Keeney, 1973). These releases could also be partly due
to sediment nutrient release under anaerobic conditions (Mor-
timer, 1971). The released nutrients, carbon dioxide, and im-
proved light penetration resulting from weed control can be
Table 1. — Weed control, use limitations, and fish toxicity of some major herbicides used for aquatic weed control.
Weed controlled
and use required
Elodea
Myriophyllum
Potamogeton
crispus
Rate of Kill
Use Restrictions
Days After
Application
Drinking
Fishing
Swimming
Irrigation
Fish Toxicity
Ortho
Diquat1
1-4ppmw
1-2ppmw
.5-2ppmw
Rapid
10
10
10
10
20 mg/i
Endothal
Dipotassium
NC
2-3ppmw
.5-1 .Sppmw
Rapid
7
3
1
7
>-1QO mg/1
2,4-D
Esters
NC
2ppmw
NC
2-6 weeks
2
3
3
3
.6-1 mg/l
Hydrothai 191
Endothal
Amine Salts
CC
CC
—
Rapid
7-25
3
—
14
.2-1 mg/l
' Label In process of review; use restrictions may be towered for 1983.
2 No value currently available. Consult State authorities.
NC Not Controlled
CC Conditionally Controlled
Sources: Product labels, Johnson and fintey, 1%0; Anonymous, 1976; Perusal! Corp.
182
-------
Physical, Chemical, and Biological Control of Aquatic Macrophytes
quickly used for growth by nonsusceptible species. Brooker
and Edwards (1973a) and Mathews (1967) found rapid
growth of the macrophytic alga Chara globularis following
a herbicide treatment. Others have found that planktonic
algal blooms follow herbicide treatments (Newbold, 1975;
Cope et al. 1969; Way et al. 1971; Brooker and Edwards,
1973b; Carter and Hestand, 1977).
Effects of herbicide treatments on aquatic fauna are most
dramatic on the invertebrates associated with the
macrophytes destroyed (e.g. molluscans, trichopterans,
lepidopterans, chironomids) (Walker, 1962; Hilsenhoff, 1966;
Brooker and Edwards, 1974). Benthic invertebrates may,
however, increase in abundance as a result of detritus
associated with death and decay of plants (Walker, 1962;
Brooker and Edwards, 1974). Reports of the effects of weed
control on fish productivity are inconsistent, showing in-
creases in some cases and decreases in others (Bennett,
1971). Development of a weed control program should in-
clude advice from state fishery biologists to minimize impacts
on the fish populations.
Mechanical Harvesting
Mechanical harvesting is a common weed control technique
in the Northeast, Upper Midwest and West Coastal regions.
Few reports of harvesting for the species of interest are
reported for southern locations. Some of the larger and more
closely monitored harvesting projects include those on the
Dane County, Wis. lakes; Chemung Lake, Ontario; Lake
Sallie, Minn.; Lake Washington, Wash.; and the Okanagan
Valley lakes, British Columbia.
The species of interest are soft enough to be cut and will
float to the top of the water so they can be removed with
conventional harvesting equipment. Therefore, harvesting
becomes a materials handling problem. The question is what
methods, techniques, and modifications will remove a mass
of plants dispersed over an area most quickly, efficiently,
cheaply, and with the fewest problems?
The efficacy of harvesting relates to the biology of in-
dividual plant species when information about potential
biomass, regrowth rates, and methods of reproduction is
needed to determine the longevity of a harvest treatment.
Most information on regrowth was developed from studies
of undifferentiated biomasses of a variety of plants or from
populations of plants strongly dominated by Eurasian milfoil.
Therefore, little information is available on the regrowth of
curlyleaf pondweed and elodea after cutting.
Most authors agree (Nichols, 1974a; Wile, 1978; Newroth,
1980; Peverly et al. 1974) that more than one harvest is need-
ed to control milfoil regrowth over the growing season.
Nichols (1974a) recommends at least two harvests at mon-
thly intervals to reasonably control milfoil in southern Wiscon-
sin. More harvests would probably be needed in regions with
longer growing seasons.
In a review of the dynamics of Myriophyllum spicatum
biomass following harvest, Kimbel and Carpenter (1979) state
that multiple harvests are more effective than a single harvest
in reducing the amount of regrowth during a single growing
season and that recovery from a single harvest declines as
the date of the harvesting becomes progressively later.
Research in British Columbia (Anon. 1981) indicates that
harvesting may actually stimulate the growth of milfoil.
Harvesting removes the shading plant canopy and allows
light penetration down to basal shoots, thus encouraging their
development.
The long-term influence of harvesting on regrowth is less
well known. Nichols and Cottam (1972) showed, on a small
plot basis, that intensive harvesting (three times per sum-
mer) for 2 years significantly reduced the biomass in an area
the third year. For a number of reasons, including lack of
detailed maps of harvested areas and natural declines of
milfoil populations, this phenomenon was never
demonstrated over large areas in the Madison, Wis. region
using commercially available equipment.
Reviewing the long-term effects of harvesting, Kimbel and
Carpenter (1979) found that biomass the following years was
reduced by harvesting in all but one case. Thus, sustained
harvesting, even only once per year, reduced M. spicatum
production the following year. Peterson et al. (1974), also
concluded that intensive harvesting on Lake Sallie, Minn.
one year reduced plant growth a second year.
The major positive environmental effects of harvesting are:
(1) organic material removed by harvesting is no longer
available to deplete oxygen supplies upon decay; (2) nutrients
are not available for recycling upon decay of the plant; and
(3) foreign material of a chemical or biological nature is not
being introduced into the system. The benefits from remov-
ing oxygen-demanding and nutrient-containing material are
proportional to the amount of material removed and the
magnitude of the benefit depends on the limnology of the
specific system.
The negative impacts include: (1) a temporary increase
in turbidity; (2) loss of animal habitat; (3) potential of plant
spread by vegetative means; (4) increased growth by remov-
ing the canopy; (5) harvesting of animal material; and (6)
release of nutrients from cut "stumps.'These impacts all ap-
pear minof (Carpenter and Adams, 1977) or at least no worse
than conditions found after treatment by other methods, ex-
cept for the potential of spreading species by vegetative
means and increased growth after cutting.
Dredging
There is little doubt that increasing water to a depth below
the light compensation point will eliminate plant growth. The
question is how deep is deep enough? Depending on the
conditions present, dredging deep enough to eliminate the
species of question could be a nearly impossible task.
Sheldon and Boylen (1977), for instance, reported elodea
growing to a depth of 12 m in the clear water of Lake George,
N.Y.
A more realistic question is how deep does an area have
to be dredged before the plants are sufficiently below the
surface of the water so as not to cause problems? This, of
course, will depend on water clarity and the intended use
of the water body.
It appears that shallow water dredging to remove nutrient
sources for aquatic plant growth has little lasting impact on
the abundance of plants, but it can have a substantial im-
pact on the species present and on successional trends.
Dunst (pers. comm.) reports that the shallow areas of Lilly
Lake, Wis., were invaded by chara the first year after dredg-
ing and milfoil (Myriophyllum exalbescens and M. spicatum)
was common to abundant in water depths to 4 meters by
the second year after dredging. Carline and Brynildson (1977)
report that Elodea canadensis slowly recolonized Krause
Spring, Wis. In a limited area of the pond, the wet weight
of elodea increased from 1 g/m2 before dredging to 7 g/m2
after dredging. Either value is extremely low compared to
the 500 g/m2 of elodea typical in similar spring ponds.
Another technique used is a hybrid between shallow water
dredging and mechanical harvesting. It is the mechanical
removal of plants using dredging equipment. Because dredg-
ing equipment is used, the costs and environmental impacts
are similar to dredging. The technique removes the whole
plants, including subsediment structures (i.e. roots and
rhizomes). Newroth (1979,1980) indicates that diver-operated
suction dredges controlled isolated stands of M. spicatum
in the Okanagan Valley lakes of British Columbia. However,
where there is a ready source of plant material reinvasion
is rapid. Using a "rotovator" of local design, he was up to
95 percent successful at removing roots and shoots from
183
-------
Lake Restoration, Protection and Management
limited problem areas. However, plants returned to 50 per-
cent of pretreatment densities by the second growing season
following treatment.
On small plots where the plants were manually uprooted,
the results are mixed. Peverly et al. (1974) reported a
biomass of 107 g/m2 in a pond where milfoil was uprooted
the previous season as compared to 452 g/m2 in a control
pond, thus indicating a carryover of plant control by
uprooting. Nicholson (1981) reported a 25 percent reduction
in biomass 1 year after completely removing all plants in plots
dominated by milfoil in Chautauqua Lake, N.Y. In Lake
Musconetcong, N.J., Patten (1956) reported removing all M.
spicatum from a I m x 4 m plot, then introducing 15 in-
dividuals at the end of July. By the following April, the quadrat
and a buffer zone were indistinguishable from the surround-
ing area.
Many impacts can be associated with dredging; they are
reviewed by Peterson (1979).
Dyes
Early attempts to control aquatic plants with dyes date back
to Eicher (1947) when he used the commercial aniline dye,
nigrosine, to control P. crispus in hatchery ponds and in
Granite Basin Lake, Ariz. Initially, the dye was applied at the
rate of 11 kg/ha in the month of May. This application con-
trolled P. crispus in water depths over 2 meters. Surber and
Everhart (1950) applied nigrosine at the rate of 17 kg/ha in
hatchery ponds in May followed by 57 kg/ha of 10-6-4 fer-
tilizer in mid-June to stimulate plankton growth. Again, this
was an effective control method. However, when the dye was
applied to ponds in August, the spread of plants was con-
trolled, but the area they occupied was not decreased.
Levardson (1953) was not able to control elodea using
nigrosine dye applied at the rate of 17 kg/ha in a Michigan
lake. This application was not made until July, and most of
the weed growth was in less than 150 cm of water.
"Aquashade" (Aquashade), "Mariner Blue Pond Dye,"
(3-M Corp.), and "Sierra Blue" (Aquat. Sys.) were commer-
cially developed for aquatic weed control. Production of
Mariner Blue Pond Dye has been discontinued, and no in-
formation was found about the efficacy of Sierra Blue. Suc-
cess with Aquashade has been mixed. Both Peverly (pers.
comm.) in New York and Lembi (pers. comm.) in Indiana
are cautiously recommending the product for use in their
states. Lembi suggests that the product should be used when
the water is over 70 cm deep and it should be applied early
in the growing season (March or early April). Used in this
manner, Aquashade controlled a broad spectrum of aquatic
plants. Likewise, Peverly recommends early application of
dye. He applied it onto ice before the pond thawed. This
technique, at manufacturer's recommended levels, suc-
cessfully controlled M, spicatum.
Nichols (1974b) was not successful at controlling M.
spicatum var. exa/bescens using Aquashade in a Wiscon-
sin farm pond. The dye was added early in the spring at the
manufacturer's recommended level. However, an extreme-
ly wet spring caused a high turnover of water in the pond
which' probably caused the treatment to fail.
Besides the dilution of dye in a flowing system, other fac-
tors can remove dye from the water. Levardson (1963) In-
dicated that nigrosine dye could be removed from the water
by plants, by turbidity, and by a chemical reaction between
salts in the water and the dye. Likewise, Surber and Everhart
(1950) indicated that nigrosine was taken up by the plants
in some manner. The rate of removal and the total amount
removed were proportional to the amount of weeds present
Lembi (pers. comm.) indicates that Aquashade does not
degrade rapidly via microbial or photo-degradation, but ft ap-
pears to bind to clay-sized particles so it is lost from the
system In turbid waters,
Levardson (1963) reported oxygen depletion problems in
the lake he treated with nigrosine and Surber and Everhard
(1950) indicated a slight increase in water temperature in
their ponds. Other than these, none of the authors reported
any negative environmental impacts when using dyes.
However, Buglewicz and Hergenrader (1977) state that
although the toxiclty of aniline dyes to other organisms is
unknown, they are extremely toxic to humans. The aesthetics
of colored water is an additional factor to consider when us-
ing dyes.
Shades
Floating black plastic sheeting on the water surface as a
shade was first reported by Mayhew and Runkel (1962). The
technique was also used In a Wisconsin farm pond by
Nichols (1974b), in New York by Peverly et al. (1974), and
is among the standard practices recommended for control-
ling weeds in ponds in Missouri (Whitely, 1964).
Mayhew and Runkel (1962) found that all species except
Cham vulgaris and Sagittaria latlfolla were controlled within
30 days in the farm pond they were studying. However, they
were not dealing with any species of direct interest to this
review.
Nichols (1974b) found that the black plastic shade very
effectively controlled M, spicatum var. exalbescens, leaving
it brown and dead after four weeks. After the shade was
removed, there was little or no regrowth of the plants the
remainder of the summer. Peverly et al. (1974) found that
the biomass of plants (about 90 percent milfoil) under a
floating shade applied in mid-June was reduced by 90 per-
cent by mid-August (from 518 g/m2) but that this later number
Increased to 92 g/m2 by mid-September. The case for con-
trol here was not as clear-cut as in Nichols* study because
the biomass in an adjacent control plot dropped from 518
g/m2 to 83 g/m2. The following May the biomass in the treat-
ment plot was about half that of the control (244 g/m2 vs.
452 g/m2).
More recently, a commercial product, Aquascreen
(Menadri-Southern Corp.), is available for aquatic plant con-
trol. It is a vinyl-coated, fiberglass screen material with a
mesh size of 64 apertures/cm2. It is more dense than water
so that it is placed on the bottom or draped over existing
weed beds. Although the material is designed as a shade,
Perkins (1980) does not believe that shade is the primary,
or at least the only mode by which Aquascreen controls
plants. He found that, based on laboratory studies, there is
ample light under the screen for milfoil growth. He believes
that It is Important that the plants be weighted enough to
touch the bottom sediments, but he is unsure of the mode
of control.
Aquascreen is easily installed, and it should be removed
every year and cleaned. If silt builds up on the screen, plants
will grow on the silt and reduce its effectiveness. The pro-
duct is very durable so, with care, It can be used for many
seasons.
Mayer (1978), Goldsby (1980), Perkins et al. (1980), and
Engel (pers. comm.) all found that Aquascreen was effec-
tive at controlling one or all of the species of interest. The
main difference in the studies was the time it took to achieve
control. Mayer found that 95 percent of all the plant material,
including the three species of interest, was destroyed in 3
weeks,
Perkins found It took 2 months to reduce the biomass of
milfoil in Lake Washington, Wash, by 75 percent. Perkins
did increase the effectiveness of the treatment the follow-
ing year by placing the screen in water in early spring (April).
He also found that Installation was easier at this time and
184
-------
Physical, Chemical, and Biological Control of Aquatic Macrophytes
that the screens placed for 2 or 3 months from April to June
could be removed and there would be little regrowth the re-
mainder of the growing seasoir. The screening could be
transferred to another plant-infested area, hence doubling
the effective area of treatment with a single piece of material.
Dawson and Kern-Hansen (1977) recommend planting tall
grass, shrubs, or trees along the south and west banks of
streams so that dense shade is cast on the water surface
and using fertilizers to stimulate heavy phytoplankton blooms
and thus form a biological shade. This is a standard weed
control technique recommended for fish ponds in the
southern United States, but has little applicability to multi-
ple use, recreational lakes.
In one Kansas lake where fertilization was attempted to
control aquatic weeds it was not successful. Schryer (1968)
reported that he did not control milfoil in water less than 2
meters using fertilizer to stimulate plankton turbidity. He felt
that part of the problem (the poor results) was caused by
carryover of plants from the previous year. Also, the water
generally had a Secchi disk clarity greater than the half meter
recommended by Boyd (1979).
Floating shades make the water underneath them
unusable. It is conceivable that water underneath a large
shade could become stagnant and depleted of oxygen. The
shades are also relatively fragile, so they cannot withstand
high winds or pounding waves.
No negative impacts were noted for the use of
Aquascreen, but the benthic fauna and water chemistry
underneath the screen have not been studied in detail.
Fertilizing is recommended only if it is done in conjunc-
tion with increasing fish productivity, and is not recommend-
ed in multiple use waters or in regions where increasing
organic matter threatens winter kill of fish populations. White-
ly (1964) also points out that the weed problem worsens
when fertilizing stops.
Bottom Coverings
Sediment manipulation techniques include covering bottom
sediments with flyash, sand, clay, or plastic or rubber
sheeting, either alone or with sand and gravel on the top.
Cooke (1980a) reviewed the use, efficacy, and associated
impacts of most covering techniques. It appears that none
of the techniques can be used for long-term control of the
three species of interest in still water. Sediment collects rapid-
ly on top of the covering and is invaded by plants as soon
as it can support plant growth. However, coverings could
be used if they were cleaned regularly or in heavily used
and intensively managed areas such as beaches. Opera-
tional testing of common burlap is showing promise as a
relatively inexpensive bottom covering to control Eurasian
water milfoil (Newrqth, pers. comm.).
It was found, however, (Anon., 1959), that plastic liners
controlled weeds in an irrigation ditch. Perhaps water move-
ment was fast enough in the reported situation to prevent
heavy siltation. Dunst (pers. comm.) has investigated the
potential for mixing alum with sediments, thus binding
phosphorus and making it unavailable to plant roots.
However, his preliminary laboratory work was not encourag-
ing enough to continue the research. Sediment amendments
(Ca(OH)2, HeSO4, and NaCI) were tested in British Colum-
bia to alter interstitial pH and salinity, but none of the
treatments killed plants (Anon. 1981).
Drawdown
Milfoil appears to be successfully controlled by winter
drawdown. Beard (1973), Davis et al. (1964), Goldsby et al.
(1978) and Smith (1971) all report controlling milfoil with
winter drawdown. Stanley (1976) found in laboratory ex-
periments that exposing dewatered milfoil to subfreezing
temperatures (~1°C) for as little as 96 hours severely
restricted regrowth of the species. However, field experience
recommends at least a 3-week exposure period to freezing
temperatures after the plant is dewatered (Cooke, 1980b).
Based on one study (Cooke, 1980b), there is no signifi-
cant impact on P. crispus of overwinter drawdown.
The use of overwinter drawdown to control E. canaden-
sis is in question (Cooke, 1980b). Beard (1973) and Nichols
(1975a) both reported an initial decline in elodea following
a winter drawdown. However, continued or annual winter
drawdown may actually cause the species to increase
(Nichols, 1975a, b).
Cooke (1980b) cautions potential users that the technique
is species specific, that species resistant to drawdown rapidly
appear, that sediments need to be nearly completely
dewatered, and that the thallus and reproductive structures
must be rigorously exposed to temperature extremes for the
technique to be successful.
In addition to weed control, game fishing often improves
after drawdown as forage fish are more concentrated and
have less available cover. Flocculent sediments may be con-
solidated, thus reducing water turbidity (Tarver, 1980).
Drawdown can also provide riparians an opportunity to repair
docks and dams, clean and repair shorelines, and deepen
areas such as swimming beaches.
On the negative side, impoundments may lose their at-
tractiveness to waterfowl if drawdown eliminates desirable
food plants. Fishkills may occur if the volume of water does
not hold sufficient oxygen through winter or summer stress
periods (although oxygen conditions sometimes benefit dur-
ing drawdown because shallow areas with highly organic
sediments are not stressing oxygen supplies under
drawdown conditions). Algal blooms often accompany
reflooding after drawdown. The impact of drawdown on ben-
thic organisms and other components of the ecosystem is
not well known. Invasion of undesirable species such as cat-
tails (Typha spp.), willow (Salix spp.), buttonbush (Cephalan-
thus occidentalis), and cypress (Taxodium distichum) on ex-
posed mudflats can be a problem with summer drawdown.
Finally, drawdown can make vast areas of a reservoir
unusable for recreational purposes. A major problem with
drawdown is making it compatible to the multiple use
demands of a reservoir. Annual winter drawdowns of some
large lakes in the Okanagan Valley to provide for spring
runoff expose shallow shorelines, facilitating shallow tillage
of beach areas to remove Eurasian water milfoil (Newroth,
pers. comm.).
Shellfish
Snails in the genera Marisa and Pomacea and crayfish in
the genera Orcanectes, Astacus, and Cambarus have been
suggested as biological controls of aquatic plants.
Marisa comuarietis L and Pomacea australialis are tropical,
South American species whose potential as macrophyte her-
bivores is being investigated (Rushing, 1975). If environmen-
tal and efficacy questions are answered, the potential for us-
ing these species is limited to subtropical areas of the United
States.
Native snail species in temperate regions do eat
macrophytes. In Magic Lake, British Columbia, Anonymous
(1981) found that snails in the genus Physa were grazing
on Eurasian water milfoil. Specifics with regard to snail
species, macrophyte species consumed, amounts, etc., have
never been thoroughly investigated.
The suggestion that crayfish be used as a weed control
agent has usually been made as an aside to other research
projects. Magnuson et al. (1975), Abrahamson (1966), and
Taub (1972) all noted that certain species of crayfish were
voracious plant feeders and kept lakes or ponds cleared of
185
-------
Lake Restoration, Protection and Management
weeds. Dean (1969) purposely increased the population of
crayfish for the specific purpose of macrophyte control. He
found that the crayfish Orconectes causey/ fed on both
Elodea canadensis and Myriophyllum exa/bescens.As popula-
tions of this crayfish increased over the course of 2 years
in Mesa Hill Lake, N.Mex., the vegetation declined from
about 75 to 25 percent coverage of the lake. Magnuson et
al. (1975) provide a number of points that should be con-
sidered before introducing crayfish into a new environment.
Insects
As milfoil became a severe problem in the United States,
entomologists searched its native range, especially
Yugoslavia and Pakistan, for suitable insect predators.
Spencer and Lekic (1974) reported 25 insect species feeding
on Eurasian water milfoil, but many of them are polyphagous
and could not be imported into the United States. Some of
the promising species identified were Parapoynx stratiota,
P. allionealis, and Acentria nivea, all aquatic moths, and the
beetle LJtodactylus leucogaster.Other studies identified
Bagous geniculatus, B. vicmus, Triaenodes tarda, Cricotopus
sp., and LJtodactylus griseomicans as feeding on milfoil
(Baloch et al. 1972; Anon., 1981; Kangasniemi, pers. comm.).
The following discusses each species individually.
Although P. stratiota displayed potential for milfoil control
in its host range, the species was not sufficiently specific
for introduction into the United States (Buckingham, pers.
comm.).
P. allionealis is already present in the United States, and
Habeck (1974) did further research into the life history and
food preference of this species. Nothing is reported about
using it in management schemes, and natural populations
of Eurasian water milfoil, at least in Florida, are very low
(Buckingham, pers. comm.).
A, nivea (Acentropus niveus) is also found in the United
States. It is not specific to milfoil and it can feed on a broad
range of plants. Immature insects have also been collected
from P. crispus and £ canadensis. Buckingham and Ross
(1981) do not believe that this species has great potential
for controlling milfoil. It is highly vulnerable to predation and
larvae do not feed on algal-covered leaves. It does not,
therefore, build up to very high levels in a milfoil mat.
L leucogaster is native to North America, but it is not pre-
sent over the entire range of Eurasian milfoil. Its specializ-
ed biology indicates that it is well adapted to water milfoils
(Buckingham and Bennett, 1981). Under laboratory condi-
tions it causes extensive damage to milfoil spikes. It is not
known if it does the same under field conditions. Because
the milfoil flowering period is short and milfoil spreads mainly
by vegetative means, the damage the weevil might do is
moderate. Its initial use would be limited to innoculative
release on Eurasian water milfoil populations outside the pre-
sent natural range of the weevil. A closely related species,
L griseomicans, feeds on M. spicatum in British Columbia
(Kangasniemi, pers. comm.).
B. geniculatus and B. vicinus are very similar in ecological
requirements. Both species use milfoil for feeding and
oviposition (Baloch et al. 1972). Since both these species
breed only in water milfoil growing on banks and not in the
water, it is unlikely that the insects will build up in popula-
tions high enough to affect water milfoil.
In August 1978 a routine survey of Magic Lake, Fender
Island, British Columbia, revealed that several hectares of
M. spicatum had been reduced to bare stems (Anon., 1981).
The agent of this destruction was identified as the caddisf-
ly, T. tarda.S\r\ce that time, the insect has been observed
feeding on milfoil in several other British Columbia lakes.
The insect uses the plant material for both feeding and case
building. During case building, plant tissue is cut and discard-
ed. Thus, the damage inflicted by the insect is greater than
would be expected, based on feeding alone. Consequent-
ly, a moderate number of larvae can inflict considerable
damage.
Cricotopus sp. attacks mainly the newly developing tissues
in the apical meristem of milfoil. It, therefore, retards stem
elongation. It appears to have a greater impact on M.
spicatum than does Triaenodes on British Columbia milfoil
populations (Anon., 1981). This may be because it is hid-
den by the apical meristem and is less susceptible to fish
predation.
No use of insects was reported for the biological control
of P. crispus or E. canadensis.
Disease Organisms
More attention has been focused on diseases of milfoil than
on those of the other two species. In fact, P. crispus and
£ canadensis appear to be remarkably disease free (Zet-
tler and Freeman, 1972). Declines in milfoil populations in
Chesapeake Bay (Bayley et al. 1978), Chemung Lake, On-
tario (Wile et al. 1979), the Madison, Wis. lakes (Carpenter,
1980), Lake Washington, Wash. (Perkins, pers. comm.) and
the Okanagan Valley lakes (Newroth, pers. comm.) would
certainly lead one to search for a causal agent. Two
"diseases" or conditions were recognized on the declining
milfoil populations in Chesapeake Bay. These are "Lake
Venice" disease and "Northeast" disease. Conditions similar
in part to "Northeast" disease were observed in declining
milfoil populations in University Bay, Lake Mendota, Wis. (An-
drews, 1980).
The causative factors of these two syndromes have not
been clearly delineated and the ability to use the pathogenic
agent for plant control has not been attained. Bayley et al.
(1968) believe the causative agent of Northeast disease is
a virus, virus-like particle, or a toxin. Bean et al. (1973) stated
that their data do not support the hypothesis that a
phytopathogen was responsible for Northeast or Lake Venice
disease. Hayslip and Zettler (1973) noted that attempts to
establish Northeast disease in Florida's Crystal River were
unsuccessful. Steenis (1970) suggested that the two syn-
dromes are caused by pollution, siltation, autotoxins, or a
combination of factors, and Correll et al. (1978) suggested
that suspended particulates and high agricultural runoff were
at least partially responsible for aquatic macrophyte declines.
Kangasniemi (pers. comm.) notes that insect attack can pro-
duce morphological characteristics similar to those attributed
to Northeast and Lake Venice disease.
In the past 10 years, only three pathogens — Rhizoctonia
solani (Joyner and Freeman, 1973), Fusarium sporotrichoides
(Andrews and Hecht, 1981), and Acremonium curvulum (An-
drews et al. 1982) — have been found that infect milfoil.
These researchers feel that the respective organisms are
so mildly pathogenic that they are of little value for plant
management unless some other factor predisposes the plant
to disease.
Small Spikerush
Three species of small spikerush, Eleocharis coloradoensis,
£ acicularis, and £ parvula, are spreading and displacing
other aquatic plants in irrigation and drainage canals and
in reservoirs in California. The most striking example of this
occurred in the Corning Canal. Over a 10-year period, water
weeds were displaced in 31 km of the canal. £ canadensis
and P. crispus were two of the primary offending species
in this canal (Yeo, 1980a).
Spikerush plants displace rooted, submerged, aquatic
plants to depths of 1.5 m in clear water. The mechanisms
of displacement and control aren't entirely known, but it ap-
pears to be a combination of mechanical and allelopathic
186
-------
Physical, Chemical, and Biological Control of Aquatic Macrophytes
action. The small spikerushes form a dense sod with stiff,
upright bristles. This growth prevents any propagules of other
species from reaching and becoming rooted in the sediment.
There is also evidence that the dwarf spikerush releases an
allelopathic substance (Frank and Dechoretz, 1980) that
causes rooted, submerged species to be "unthrifty."P.
crispus and £. canadensis were not the test organisms in
the allelopathy studies, so their control by this means is anec-
dotal. In fact, Yeo (1980b), despite observations made in the
Corning Canal, found that elodea and curlyleaf pondweed
were not as easily displaced by spikerush as were other
species. No interaction between spikerush and milfoil was
reported.
Fish
The use of a variety of fish has been suggested for the
biological control of aquatic weeds. These include the com-
mon carp (Cyprinus carpio) and the Israeli strain of the com-
mon carp, roach (Rutilus rutilus), rudd (Scardinus erythop-
thalmus), various species of tilapia including 77/ap/a zillii and
77/ap/a mossambica, silver dollar fish (Metynnis roosevelti and
Mylossoma argenteum), white amur (Ctenopharyngodon
idella), and a variety of hybrids of the white amur, the most
promising of which appears to be the triploid hybrid, between
the white amur and the bighead carp (Hypopthalmichys
nobilis). All these species are not native to the United States,
although the common carp is naturalized.
The questions to be answered about using fish as a
biological control include: Do the fish species selectively feed
on the plant species of interest? What stocking rates and
environmental conditions are needed for given levels of con-
trol? What are the habitat limits of the fish? If exotic fish are
introduced, what is their potential for becoming a nuisance?
What are the impacts associated with plant control induced
by the fish? These questions will be addressed to the ex-
tent of the information available on a fish-by-fish basis.
Silver Dollar Fish, Roach and Rudd
Prejs and Jackowska (1978) studied the food habits of roach
and rudd in three Polish lakes. They found that both species
grazed elodea and milfoil. The contribution of macrophytes
to the diet of rudd increased with the size of fish. In fish over
16 cm in length, 90 percent of their diet consisted of plant
material. The contribution of macrophytes to the diet of roach
was not clear. Their diet varied between different lakes and
between different size classes of fish. Prejs and Jackowska
used an electivity index that demonstrated that although
elodea and milfoil were both grazed, elodea was selected
for and milfoil was selected against as food by both fish. Their
studies of the fish have not yet progressed to the point of
using the fish for management purposes, nor was there any
reference to considering the fish for use in the United States.
Yeo (1967) reported that silver dollar fish will eat both
elodea and curlyleaf pondweed. This information was
developed from experiments in aquaria and small pools and
was not subjected to field testing. However, Yeo did note
that the fish are very temperature sensitive. They eat little
below 21 °C and they begin to die below 16°C.
Common Carp
Common carp are not primarily herbivores. They control
aquatic plants by rooting them out when searching for ben-
thic organisms, roiling through weed beds when spawning,
and by greatly increasing the turbidity of the water. Both
Tryon (1954) and Threinen and Helms (1954) demonstrated
the impact carp have on aquatic vegetation by building ex-
closures in lakes that already had carp populations. Black
(1946) and Mathis (1966) stocked carp as a weed control
measure. In all cases, carp effectively controlled elodea. They
also destroyed the curlyleaf pondweed population in
Pymatuning Lake, Pa. Most research on the influence of carp
on aquatic vegetation was done before Eurasian milfoil
became a serious problem in the United States.
Carp can lead to the demise of sportfish population and
can cause water quality problems because they stir up bot-
tom sediments and cause turbidity. In many areas of North
America they are considered a pest.
Tilapia
Tilapia zillii and T. mossambica are being widely used in the
irrigation system and recreational lakes of the lower Sonoran
Desert region of California. Their food preference for elodea
and curlyleaf pondweed has not been tested and initial
feeding tests in aquaria using milfoil were not encouraging.
In fact, of four species common to the irrigation canals, milfoil
was the least preferred (Mauser et al. 1976). This caused
concern, because the fish might feed selectively in the
canals, leaving a persistent monoculture of M. spicatum (var.
exalbescens in this case). In field trials, food preference was
apparently not a problem, at least at relatively high fish den-
sities. Legner and Murray (1981) state, "Therefore, T. zillii
apparently maintained M. spicatum at a very low density in
spite of laboratory avoidance of this weed."They attribute
success for controlling milfoil in mixed stands to lack of ability
by the fish to easily discriminate between plant species and
the fact that M. spicatum provides a substrate and refuge
for a variety of organisms that could serve as suitable
nutrients for the fish.
There is extensive winter mortality of the fish in water
temperatures below 10°C. Even in southern California canals
they don't survive most winters, so there is a question
whether large fish could be maintained in adequate densities
for weed control. Presently, winterkill and predation appear
to thin populations adequately so that territoriality is not a
problem.
Dispersal poses another problem. Studies show that these
fish tend to remain near the stocking area, so they must be
manually dispersed for maxiumum effectiveness.
The people studying tilapia are not advocating widespread
use of the fish and its use appears to be limited to warm
regions. They have found the fish to be a useful weed con-
trol technique, either by itself or in conjunction with
mechanical control in southern California.
Grass Carp or White Amur
The most extensively studied herbivorous fish is the white
amur or Chinese grass carp. At least two States — Arkan-
sas and Iowa — use the fish as a standard management
tool. Other States have used them on a limited basis and
still others have banned their use. White amur have been
used in a variety of situations including lakes, reservoirs,
ponds, and canals.
Under good conditions, the fish will eat its body weight
in plants every day (Stott, 1972). Krzywosz et al. (1980)
estimate that grass carp consume only about 50 percent of
aquatic vegetation they pull out. So these researchers dou-
ble the assumed daily food ration when estimating the weed
biomass grass carp will control.
Grass carp feed on all three species of interest. A number
of papers agree with the findings of Alabaster and Stott
(1967) that elodea is a highly preferred food species and
milfoil is not. Few studies refer to P. crispus, but studies by
Willey et al. (1974) and Laal and Thakur (1977) indicate this
is a preferred species.
Basically, in field trials with the fish the preferred species
are eaten first and the less preferred species remain until
other foods become limiting. Therefore, milfoil is controlled
187
-------
Lake Restoration, Protection and Management
only if it is present in a monoculture or if fish densities are
high enough to eliminate all the preferred species first.
Recommendations for using grass carp are not highly
refined. No recommendations were found that outline
management techniques based on the combination of the
species of plant present, the biomass of plant present, the
growing season, the productivity of the plant and fish-related
factors such as age, size, stocking density, oxygen and
temperature conditions of the water, and previous feeding
history.
The environmental concerns about using white amur
center on four areas: (1) their potential for becoming a pest;
(2) their potential for recycling nutrients or causing other
water quality problems; (3) their selectivity of feeding habits;
and (4) their ability to destroy habitat for other organisms.
Proponents of using grass carp indicate there is little
potential for it becoming a pest such as the common carp.
The grass carp is not as fecund as the common carp. Its
spawning requirements are quite restrictive. Bailey (1972)
notes that the fish needs water temperatures near 21 °C to
spawn, that eggs must be moved about by current for 15
to 35 hours before hatching, and that the larva or fry are
highly susceptible to predation and silt. Other authors note
that rubble substrates and temporary rises in water level are
also needed for successful spawning. Bailey (1972) believes
that with all these environmental restrictions, the white amur
will not be fecund enough to become a problem.
Ware et al. (1975), however, believe that the spawning re-
quirements for grass carp are similar to those of the native
striped bass (Morane saxatilis) so that it is probable that it
could spawn in many Florida rivers. They also cite studies
from Taiwan and Japan where grass carp successfully
reproduced and became a predominant fish. States which
banned the import of grass carp probably feel they have con-
ditions more suitable to grass carp reproduction than those
that have embraced its use. The non-user States are not
willing to take the environmental risk that the fish will not
become a pest.
A second question relates to the influence of grass carp
on water quality. The grass carp consume large amounts
of plant material, digesting and using about half the food
intake. The remainder is egested. Stanley (1974a) reported
that large quantities of phosphorus and nitrate were excreted
back into the water column when grass carp were kept in
aquaria.
This would not be an advantage in multiple use lakes but
it appears the relationship is not as simple as aquaria studies
indicate. Hestand and Carter (1978) compared the nutrients
released in chemically treated ponds to those released in
a pond treated with grass carp. They found that total organic
nitrogen concentrations differed only slightly from those found
in the control pond and the quantities of orthophosphorus
were lower than or the same as the control throughout the
study. Stimulation of the plankton community did not occur
after vegetation removal by grass carp. Mitzner (1978)
reported similar findings for Red Haw Lake, Iowa. Nitrate
and nitrite concentrations, BOD, and turbidity were less after
planting grass carp than before. Only alkalinity showed a
significant increase in concentration. The trend of gradually
decreasing fertility was confirmed by decreasing
phytoplankton primary production.
Lembi et al. (1978) summarize their findings by stating that
"The release of nitrogen and phosphorus from vegetation
consumption has not been as great as anticipated from
aquarium and pool studies, indicating that the flow of nu-
trients into components of the aquatic ecosystem other than
water or phytoplankton is relatively rapid." In comparing
water quality from enclosures of three different sizes con-
taining white amur feeding on aquatic plants, Michewicz et
al. (1972) concluded that there was greater deviation from
normal conditions as container size decreased. Therefore,
it appears that weed control by grass carp in larger water
systems does not stimulate phytoplankton growth as much
as might be anticipated, as a result of increased nutrient
concentrations.
The two concerns with relation to other organisms relate
to whether grass carp compete for food with other organisms
and if they destroy critical habitat. Michewicz et al. (1972)
state that in the fry stage, the white amur diet consists of
algae, rotifers, and crustaceans, with occasional chironomids.
Macrophyte material is first eaten when the fish are 17 to
18 mm long; at that time, the importance of rotifers declines
and that of chironomids increases. Macrophyte material
forms the main basis of the diet at 27 mm, and from 30 mm
onward, the fish is vegetarian. Other authors (Kilgen and
Smitherman, 1970; Willey et al. 1974) note that grass carp
eat animal food at other times, especially if plant food
becomes unavailable. Therefore, it appears there is some
potential for interspecific competition for food if grass carp
are stocked with other fish species.
Fish, waterfowl, aquatic fur bearers, and other organisms
use aquatic plants for food or cover. Unregulated weed con-
trol by grass carp might have a negative impact on these
organisms. Anonymous (1981) gave this as one important
reason why grass carp were not considered as a weed
management measure in British Columbia. Krzywosz et al.
(1980) felt that large stocks of grass carp in Lake Dgal Wielki,
Poland, negatively affected the lake ichthyofauna. This oc-
curred primarily because the carp overgrazed the
macrophyte cover that provided habitat for native fish.
Finally, the carps' selectivity for plant species can cause
management problems. Fowler and Robson (1978) conclud-
ed that in mixed macrophyte communities containing milfoil
understocking must be avoided unless some other measure
is added to control unpalatable plants. If the water body is
understocked, only highly preferred plants are likely to be
selected and milfoil will spread to form a monotypic nuisance
problem. Newton et al. (1979) experienced similar problems
with the spread of milfoil when using grass carp to control
aquatic plants in irrigation canals and reservoirs in Arkansas.
One means of eliminating the potential for the prolifera-
tion of a nuisance species is to develop a sterile organism.
The triploid cross between the female grass carp and the
male bighead carp is an attempt to produce a sterile her-
bivorous fish. This is now being actively researched and
developed, so there is little information at this point on
feeding preferences, environmental impacts, etc, Sutton et
al. (1981) review the basic information about this cross.
SUMMARY AND CONCLUSIONS
A review of the literature indicates a variety of tactics have
been used effectively to manage the three species of interest.
Some have proven their efficacy over broad geographical
ranges, in a variety of water types, and under a variety of
environmental conditions. Others have a more restricted use,
but all should be considered in an overall management
strategy.
Chemical weed control has been used as an effective
management tool for many years. New application methods
and herbicide combinations have reduced impacts on
aquatic life, decreased costs, and improved longer-term con-
trol. Further development of these techniques, along with
a combination of other control measures with herbicides, can
increase weed control efficiency with minimal detrimental ef-*
fects on aquatic ecosystems.
The techniques for harvesting are well developed and it
appears that, at least for Eurasian milfoil, harvesting may
reduce growth for more than one season. Work on the
materials handling aspect of harvesting should make opera-
tions more efficient and more research relating to the
188
-------
Physical, Chemical, and Biological Control of Aquatic Macrophytes
physiology of individual species should make the technique
more effective.
For dredging to have a lasting impact, it should be done
to a depth sufficient to place the lake bottom below the photic
zone or at least to prevent plants from reaching the surface
where they interfere with water use. The exact depth to
dredge will vary from lake to lake, and shallow water dredg-
ing to remove nutrient-rich sediments appears to have little
long-term impact on plant growth.
Shades and dyes are effective, but their potential use is
limited. The use of shades is limited to relatively small areas.
Submerged shades may need cleaning to maintain their ef-
fectiveness. Commercial dyes should be used only where
the turnover of water is slow, where there is little suspend-
ed sediment in the water, and where the water is at least
70 cm deep. Dyes should be applied early in the growing
season.
Bottom coverings are not recommended unless silt ac-
cumulation on the cover is minimal or can be prevented.
It appears that milfoil can be successfully controlled where
conditions are appropriate for a winter drawdown. Drawdown
may have little impact on P. crispus and E. canaden-
sis. Drawdown is a species specific technique, so invasion
of an area by species resistant to drawdown may be rapid.
Of the biological techniques, the use of two species of
tilapia, grass carp, common carp, and dwarf spikerushes
have shown their efficacy. Of these, grass carp or grass carp
hybrids have the greatest potential for widespread use.
Elodea and P. crispus are food species preferred by grass
carp, but milfoil is not. A variety of environmental questions
need to be answered about grass carp, and more research
could be done to custom design a weed control program
based on local needs and conditions.
Crayfish are known vegetation feeders, but conditions for
crayfish growth are usually not optimal in areas that are op-
timal for weed growth. The impact of crayfish on other com-
ponents of the ecosystem can be significant, so they may
not be desirable inhabitants of a water body.
The use of insects and disease organisms have interesting
potential for control of the three target species, but, to date,
they haven't shown their efficacy. Development is likely to
be slow because an agent has to be developed that is en-
vironmentally safe, works under a variety of management
conditions, and is easily produced and distributed.
Finally, the integration of techniques so that the synergistic
potential of two or more techniques is used is a fertile area
for future research and development.
REFERENCES
3-M Corp., St. Paul, Minn.
Abrahamson, S. A. A. 1966. Dynamics of an isolated population of
crayfish Astacus asfacus.Oikos 17:96-107.
Alabaster, J. S., and B. Stott. 1967. Grass carp (Ctenopharyngodon
idella val.) for aquatic weed control. Pages 123-126 in Eur. Weed.
Res. Com. Int. Symp. Water Weeds, Oldenburg.
Andrews, J. H. 1980. Plant pathogens as agents for biological and
integrated control of aquatic plants. Univ. Wis. Water Resour.
Center Tech. Comple. Rep. 80-01. Madison.
Andrews J. H., and E. P. Hecht. 1981. Evidence for pathogenicity of
Fusarium sporotrichioides to Eurasian water milfoil, Myriophyllum
spicatum. Can. J. Bot. 59:1069-1077.
Andrews, J. H., E. P. Hecht, and S. Bashirian. 1982. Association
between the fungus Acremonium curvulum and Eurasian water-
milfoil, Myriophyllum spicatum. Can. J. Bot. (in press).
Anonymous, 1959. Water conservation and weed control. World
Crops Oct. :375.
. 1976. How to identify and control water weeds and algae.
Applied Biochemist, Inc. Mequon Wis.
. 1981. A summary of biological research on Eurasian milfoil
in British Columbia. Vol. XI: Inf. Bull. Aquat. Plant Manage. Progr.
Br. Colum. Ministr. Environ.
Aquashade, Inc. Eldred, N.Y.
Aquatic Systems. Butler, Wis.
Bailey, W. M. 1972. Arkansas evaluation of the desirability of intro-
ducing the white amur (Ctenopharyngodon idella val.) for control
of aquatic weeds. A paper prepared for the 102nd Annu. Meet.
Am. Fish. Soc. and the Int. Assoc. Game Fish Comm. Hot Springs,
Ark. Sept. 10-15.
Baker, G., et al. 1975. Improved application techniques for aquatic
herbicides. Hyacinth Control J. 13:21-24.
Baloch, G. M., A. G. Khan, and M. A. Ghani. 1972. Phenology,
biology and host specificity of some stenophagous insects attack-
ing Myriophyllum spp. in Pakistan. Hyacinth Control J. 10:13-16.
Bayley, S., H. Rabin, and C. Southwick. 1968. Recent decline in the
distribution and abundance of Eurasian milfoil in Chesapeake Bay.
Chesapeake Sci. 9:173-181.
Bayley, S., V. D. Stotts, P. F Springer, and J. Steenis. 1978.
Changes in submerged aquatic macrophyte populations at the
head of Chesapeake Bay. 1958-1975. Estuaries 1:73-84.
Bean, G. A., M. Fusco, and W. I. Klarman. 1973. Studies on the
"Lake Venice disease" of Eurasian milfoil in the Chesapeake Bay.
Chesapeake Sci. 14:279-280.
Beard, T. D. 1973. Overwinter drawdown, impact on the aquatic
vegetation in Murphy Flowage, Wis. Wis. Dep. Nat. Res. Tech.
Bull. No. 61. Madison.
Bennett, G. W. 1971. Management of Lakes and Ponds. Van
Nostrand Reinhold.
Bitting, L. E. Sr. 1974. Field use of invert emulsion. Hyacinth Con-
trol J. 12:34-35.
Black, J. D. 1946. Nature's own weed killer — the German carp. Wis.
Cons. Bull. 11:3-7.
Boyd, C. E. 1970. Losses of mineral nutrients during decomposition
of Typha /af/fo//a.Arch. Hydrobiol. 66:511-517.
. 1979. Water Quality in Warm Water Fish Ponds. Auburn
Univ. Agric. Exp. Sta. Auburn, Ala.
Brooker, M. P. 1974. The risk of deoxygenation of water in herbi-
cide application for aquatic weed control. J. Instr. Water Eng.
28:206-210.
Brooker, M. P., and R. W. Edwards. 1973a. Effects of the herbicide
paraquat on the ecology of a reservoir. I. Botanical and chemical
aspects. Freshwat. Biol. 3:157-175.
1973b. Effects of the herbicide paraquat on the ecology of
a reservoir. II. Community metabolism. Freshwat. Biol. 3:383-389.
. 1974. Effects of the herbicide paraquat on the ecology of a
reservoir. III. Fauna and general discussion. Freshwat. Biol.
4:311-333.
_. 1975. Review paper: Aquatic herbicides and the control of
water weeds. Water Res. 9:1-15.
Buckingham, G. R. Pers. Comm. Agric. Res. Serv. U.S. Dep. Agric.
Gainesville, Fla.
Buckingham, G. R., and C. A. Bennett. 1981. Laboratory biology
and behavior of LJtodactylus leucogaster, a ceutorhynchine weevil
that feeds on watermilfoils. Ann. Entomol. Soc. Am. 74:451-458.
Buckingham, G. R., and B. M. Ross. 1981. Notes on the biology and
host specificity of Acentria nivea (=Acentropus niveusj.J. Aquat.
Plant Manage. 19:32-36.
Buglewicz, E. G., and G. L. Hergenrader. 1977. The impact of artifi-
cial reduction of light on a eutrophic farm pond. Trans. Neb. Acad.
Sci. 4:23-33.
Carline, R. F., and O. Brynildson. 1977. Effects of hydraulic dredging
on the ecology of native trout populations in Wisconsin spring
ponds. Wis. Dep. Nat. Res. Tech. Bull. No. 98. Madison.
Carpenter, S. R. 1980. The decline of Myriophyllum spicatum in a
eutrophic Wisconsin lake, U.S.A. lake. Can. J. Bot. 58:527-535.
189
-------
Lake Restoration, Protection and Management
Carpenter, S. R., and M. S. Adams. 1977. Environmental impacts of
mechanical harvesting of submersed vascular plants. Inst. En-
viron. Stud. Rep. 77. Univ. Wisconsin, Madison.
Carter, C. C., and R. S. Hestand. 1977. The effects of selected herbi-
cides on phytoplankton and sulfur bacteria. J. Aquat. Plant
Manage. 15:47-56.
Cooke, G. D. 1980a. Covering bottom sediments as a lake restoration
technique. Water Resour. Bull. 16:921-926.
. I980b. Lake level drawdown as a macrophyte control tech-
nique. Water Resour. Bull. 16:317-322.
Cope, O. B., J. P. McCraren, and L. Eller. 1969. Effects of dichlo-
benil on two fish pond environments. Weed Sci. 17:158-165.
Davis, J. T., J. S. Hughes, and H. E. Schafer. 1964. Water level
fluctuation to control aquatic vegetation. Proc. 17th St. Weed Conf.
32-38.
Dawson, F. H., and U. Kern-Hansen. 1977. Aquatic weed manage-
ment in natural streams;the effects of shade by marginal vegeta-
tion. In Proc. 10th Congr. Int. Vereinigung for Theoretische und
Augewandte Limologie; Copenhagen, Denmark. 20:1451-1456.
Dean, J.L. 1969. Biology of the crayfish Orconecfes causey/ and its
use for control of aquatic weeds in trout lakes. Bur. Sport Fish.
Wildl. Tech. Pap. 24.
Dunst, R. Pers. comm. Wis. Dep. Nat. Resour. Madison.
Dunst, R., and S.A. Nichols. 1979. Macrophyte control in a lake
management program. Pages 411-418 in J.E. Breck, R.T. Pren-
tki, and O.L. Loucks, eds. Aquatic Plants, Lake Management and
Ecosystem Consequences of Lake Harvesting. Inst. Environ. Stud.
Univ. Wisconsin-Madison.
Eicher, G. 1947. Aniline dye in aquatic weed control. J. Wildl.
Manage. 11:193-197.
Engel, S. Pers. comm. Wis. Dep. Nat. Resour. Madison.
Folmar, L.C. 1976. Overt avoidance reaction of rainbow trout fry to
nine herbicides. Bull. Environ. Contam. Toxicol. 15:509-514.
Fowler, M.C., and T.O. Robson. 1978. The effects of the food pre-
ference and stocking rates of grass carp (Ctenopharyngodon idella
val.) on mixed plant communities. Aquat. Bot. 5:261-276.
Frank, P.A., and N. Dechoretz. 1980. Allelopathy in dwarf spikerush
(Eleocharis coloradoensis). Weed Sci. 28:499-505.
Goldsby, T.L. 1980. TVA testing of Aquascreen barrier for aquatic
weed control. Tenn. Valley Author. Mimeo Rep.
Goldsby, T.L., A.L. Bates, and R.A. Stanley. 1978. Effect of water
level fluctuation and herbicide on Eurasian watermilfoil in Melton
Hill reservoir. J. Aquat. Plant Manage. 16:34-38.
Habeck, D.H. 1974. Caterpillars of Parapoynx in relation to aquatic
plants in Florida. Hyacinth Control J. 12:15-18.
Harris, F.W. 1973. Controlled-release herbicide polymer formulations
for aquatic weed control. Hyacinth Control J. 11:61-63.
Harris, F.W., and M.A. Talukdar. 1982. Development of polymeric
controlled-release herbicide systems. Proc., 16th Annu. Meet.
Aquat. Plant Contr. Res. Plann. Oper. Rev. Nov. 17-19. St. Paul,
Minn. Misc. Pap. A-82-3, U.S. Army Eng. Waterways Exp. Stat.
Vicksburg, Miss.
Hauser, W.J., E.F. Legner, and R.A. Medved. 1976. Tilapia—a
management tool for biological control of aquatic weeds and in-
sects. Fisheries 1:24.
Hayslip, H.F., and F.W. Zettler. 1973. Past and current research on
disease of Eurasian watermilfoil (Myriophyllum spicatum L.).
Hyacinth Control J. 11:38-40.
Hesser, E., and E. Gangstad. 1978. Nuisance aquatic macrophyte
growth. J. Aquat. Plant Manage. 16:11-13.
Hestand, R.S., and C.C. Carter. 1978. Comparative effects of grass
carp and select herbicides on macrophyte and phytoplankton com-
munities. J. Aquat. Plant Manage. 16:43-50.
Hilsenhoff, W. 1966. Effects of diquat on aquatic insects and related
animals. J. Econ. Ent. 59:1526-1527.
Hoeppel, R.E., and H.E. Westerdahl. 1982. Evaluation of controlled
release 2,4-D formulations in Lake Seminole, Ga. Proc. 16th An-
nu. Meet. Aquat. Plant Contr. Res. Plann. Oper. Rev. Nov 17-19,
1981. St. Paul, Minn. Misc. Pap. A-82-3. U.S. Army Eng. Water-
ways Exp. Sta. Vicksburg, Miss.
Jewell, W.J. 1970. Aquatic weed decay: dissolved oxygen utilization
and nitrogen and phosphorus regeneration. A. Conf. Water Pollut.
Control Fed. 43:1-27.
Johnson, W.W., and M.T. Finley. 1980. Handbook of Acute Toxicity
of Chemicals to Fish and Aquatic Invertebrates. U.S. Dep. Inter.
Resour. Publ. 137.
Joyner, B.G., and T.E. Freeman. 1973. Pathogenicity of Rhizoctonia
solani to aquatic plants. Phytopathology 63:681-685.
Kangasniemi, B. Pers. comm. Br. Colum. Minstr. Environ. Victoria.
Kilgen, R.H., and R.O. Smitherman. 1970. Food habits of the white
amur stocked in ponds alone and in combination with other
species. Prog. Fish Cult. 33:123-127.
Kimbel, J.C., and S.R. Carpenter. 1979. The dynamics of Myrio-
phyllum spicatum biomass following harvest. In J.E. Breck, R.T.
Prentki, and O.L. Loucks, eds. Aquatic Plants, Lake Management
and Ecosystem Consequences of Lake Harvesting. Inst. Environ.
Stud. Univ. Wisconsin, Madison.
Kormandy, E. S. 1968. Weight loss of cellulose and aquatic macro-
phytes in a Carolina bog. Limnol. Oceanogr. 13:522-526.
Krzywosz, T., W. Krzywosz, and J. Radziej. 1980. The effect of grass
carp, Ctenopharyngodon idella (val.) on aquatic vegetation and
ichthyofauna in Lake Dgal Vielke. Ekol. Pol. 28:433-450.
Laal, A. K., and N. K. Thakur. 1977. Efficacy of grass carp (Cteno-
pharyngodan idella) in controlling Potamogeton crispus infestations
in a derelict pond in North Bihar. Sci. Cult. 43:391-392.
Legner, E. F., and C. A. Murray. 1981. Feeding rates and growth
of the fish 77/ap/a z/ffi (cichlidae) on Hydrilla vert/cffiafa, Potamogeton
pectinatus and Myriophyllum spicatum var. exalbescens and in-
teractions in irrigation canals of Southeastern California. Mosq.
News 41:241-250.
Lembi, C. A. Pers. comm. Dep. Bot. Plant Pathol. Purdue Univ.
W. Lafayette, Ind.
Lembi, C. A., B. G. Ritenour, E. M. Iverson, and E. G. Forss. 1978.
The effects of vegetation removal by grass carp on water chemistry
and phytoplankton in Indiana ponds. Trans. Am. Fish. Soc.
107:161-171.
Levardson, N. O. 1953. Experiments with nigrosine dye in aquatic
plant control. Prog. Fish Cult. 15:109-113.
Magnuson, J., G. Capelli, J. Lorman, and R. Stein. 1975. Considera-
tions of crayfish for macrophyte control. Pages 66-74 in P. L.
Brezonik and J. L. Fox, eds. Proc. Symp. Water Quality Manage-
ment Through Biological Control. Rep. No. ENV-07-75-1. Univ.
Florida, Gainesville.
Mathews, L. J. 1967. Further studies of spraying lake weeds. Page 76
in V. J. Chapman and C. A. Bell, eds. Roturua and Waikata, Water
Weeds Problems and Search for a Solution. Univ. Auckland, New
Zealand.
Mathis, W. P. 1966. Observations on control of vegetation in Lake
Catherine using Israeli carp and a fall and winter drawdown. Pages
197-205 in Proc. 19th Annual Conf. Southeast Assoc. Game
Comm.
Mayer, R. J. 1978. Aquatic weed management by benthic semi-bar-
riers. J. Aquat. Plant Manage. 16:31-33.
Mayhew, J. K. and S. T. Runkel. 1962. The control of nuisance aqua-
tic vegetation with black plastic sheeting. In Proc. Iowa Acad. Sci.
69:303-307.
Menadri-Southern Corp. Augusta, Ga.
Michewicz, J. E., D. L. Sutton, and R. D. Blackburn. 1972. Water
quality of small enclosures stocked with white amur. Hyacinth Con-
trol J. 10:22-25.
Mitre Corp. Pers. comm. McLean, Va.
Mitzner, L. 1978. Evaluation of biological control of nuisance aquatic
vegetation by grass carp. Trans. Am. Fish. Soc. 107:135-145.
190
-------
Physical, Chemical, and Biological Control of Aquatic Macrophytes
Mortimer, C. H. 1971, Chemical exchanges between sediments and
water in the Great Lakes—speculations on probable regulating
mechanisms. Limnol. Oceanogr, 16:387-404.
Mulligan, H. F., A. Barnowski, and R. Johnson. 1976. Nitrogen and
phosphorus fertilization of aquatic vascular plants and algae in
replicated ponds. Part I. Initial response to fertilization.
Hydrobiologia 48:109-116.
Newbold, C. 1975. Herbicides in aquatic ecosystems. Biol. Conserv,
7:97-118.
Newroth, P. R. 1979. British Columbia Aquatic Plant Management
Program, J. Aquat. Plant Manage. 17:12-19,
1980. Case study of aquatic plant management for lake
restoration and preservation in British Columbia, Canada, Pages
146-152 in Proc. Int. Symp. Restoration of Lakes and Inland
Waters. EPA-440/5-81-010. U. S. Environ. Prot. Agency,
Washington, D.C.
Pers. comm. Br. Colum. Ministr. Environ. Victoria.
Newton, S. H., J. M. Martin, J. N. Ferguson, and D. L. Gray. 1979.
Grass carp and removal of weeds in irrigation canals and reser-
voirs. Ark. Farm Res. 28:12.
Nichols, D. S., and D. R. Keeney. 1973. Nitrogen and phosphorus
release from decaying water milfoil. Hydrobiologia 42:509-525.
Nichols, S. A. 1974a. Mechanical and habitat manipulation for aqua-
tic plant management. Wis. Oep. Nat. Res. Tech. Bull. 77.
Madison.
_. 1974b. Progress Report — Richland farm project. Inland
Lake Renewal Rep, Mimeo.
. 1975a. The use of overwinter drawdown for aquatic vegeta-
tion management. Water Res. Bull. 11:1137-1148.
1975b. The impact of overwinter drawdown on the aquatic
vegetation of the Chippewa Flowage, Wis, U.S.A. Trans. Wis.
Acad. Sci. Arts Lett. 63:176-186.
Nichols, S. A., and G. Cottam. 1972. Harvesting as a control for
aquatic plants. Water Res. Bull. 8:1205-1210.
Nicholson, S. A, 1981, Effects of uprooting on Eurasian watermilfoil.
J. Aquat. Plant Manage. 19:57-59.
Parker, C. 1960. Influence of water hardness on paraquat. Nature
212:1465-1466.
Patten, B. C.Jr. 1956. Notes on the biology of Myriophyllum splcatum
L. in a New Jersey lake. Bull. Torrey Bot. Club. 83:5-18.
Penwalt Corp. The uses and properties of Endothal as an aquatic
algicide and herbicide.
Perkins, M, A, 1980. Managing aquatic plants with fiberglass screen.
Pages 245-248 in Proc. Int. Symp. Restoration of Lakes and In-
land Waters. EPA 440/5-81-010. U.S. Environ. Prot. Agency,
Washington, D.C.
. Pers. comm. Univ. Washington, Seattle.
Perkins, M. A., L. Boston, and E. F. Curren. 1980. The use of fiber-
glass screens for control of Eurasian watermilfoil. J. Aquat. Plant
Manage. 18:13-19.
Peterson, S. A. 1979. Dredging and lake restoration. Pages 105-114
in Proc. Natl. Conf. Lake Restoration, August, 1978. Minneapolis,
Minn, EPA 440/5-79-001. U.S. Environ Prot. Agency, Washington,
D.C.
Peterson, S. A., W. L, Smith, and K. W. Malueg. 1974. Full-scale
harvest of aquatic plants: nutrient removal from a eutrophic lake.
J. Water Pollut, Control Fed. 46:697-707.
Peverly, J. Pers. comm. Dep. Agron. Cornell Univ. Ithaca, N.Y.
Peverly, J., G. Miller, W. H. Brown, and R. L. Johnson. 1974. Aqua-
tic weed management in the Finger Lakes. Tech. Rep. 90. Cor-
nel) Univ. Water Resour. Mar. Sci. Center, Ithaca, N.Y.
Pokorny, J. 1971. Changes in hydrochemical and hydrobiological
relations occurring when Elodea canadensis was controlled with
paraquat. Pages 217-229 in Int. Symp. Aquat. Weeds, 3rd. Wagen-
ingen, Eur. Weed Res. Counc.
Prejs, A., and A. Jackowska. 1978. Lake macrophytes as the food
of roach (Rutilus rutilus L.) and rudd (Scardinus erythrophytha/mus
L.). I, Species composition and dominance relations in the lake
and food. Ekol. Pol. 26:429-439.
Rushing, W. N, 1975. Marisa and other snail predators of macro-
phytes. Pages 61-65 in P. L. Brezonik and J. L. Fox, eds. Proc.
Symp. Water Quality Management Through Biological Control.
Rep. No. ENV. 07-75-1. Univ. Florida, Gainesville.
Schryer, F. 1968. Determination of the effects of a fertilizer-induced
plankton turbidity of 18 to 24 inches by Secchi disc reading on
submerged aquatic plants. Job Compie. Rep, Fed. Aid in Fish.
Restor. Act. Kans. Forestry Fish Game Comrn,
Sheldon, R. B., and C. W. Boylen. 1977. Maximum depth inhabited
by aquatic vascular plants. Am, Midi. Nat. 97:248-254.
Smith, G. E. 1971. Resume of studies and control of Eurasian water-
milfoil (Myriophyllum spicatum L.) in the Tennessee Valley from
1960 through 1969. Hyacinth Control J. 9:23-25.
Spencer, N. R. and M. Lekic. 1974. Prospects for biological control
of Eurasian watermilfoil. Weed Sci. 22:401-404,
Stanley, J. G. 1974a. Unisex studies on the white amur. Pages
89-95 in Proc. Research Planning Conference on Integrated
Systems of Aquatic Plant Control. U.S. Army Corps Eng. Water-
ways Exp. Sta. Vicksburg, Miss.
Stanley, R. A. 1974b. Effects of 2,4-D and various salts on Eurasian
watermilfoii, Weed. Sci. 22:591-594.
1975. Interaction of calcium and 2,4-D on Eurasian watermil-
foil. Weed Sci. 23:182-184.
1976. Response of Eurasian watermilfoil to subfreezing tem-
perature. J. Aquat. Plant Manage. 14:36-39.
Steenis, H. 1970. Eurasian watermilfoil and associated submerged
species in the Chesapeake Bay area. Paper prepared for the Weed
Sci. Soc. Meet. Montreal, Feb. 3-5.
Steward, K. L. 1981. Improving technology for chemical control of
aquatic weeds. Tech. Rep. A-81-2. U.S. Army Corps Eng. Water-
ways Exp. Sta. Vicksburg, Miss.
Stott, B. 1972. Biological control of water weeds, Symp. Br. Ecol.
Soc. 13:233-238.
Surber, E. W., and M. H. Everhart. 1950. Biological effects of nigrc-
sine used for control of weeds in hatchery ponds. Prog. Fish Cult.
12:135-140.
Button, D. L., L W. Weldon, and R. D, Blackburn. 1970. Effect of
diquat on uptake of copper in aquatic plants. Weed Sci,
18:703-707.
Button, D. L., R. D. Blackburn, and W. C. Bonlowe. 1971. Response
of aquatic weeds to combinations of endothal and copper. Weed
Sci. 19:643-646.
Sutton, D. L., J. G. Stanley, and W. Miley. 1981. Grass carp hy-
bridization and observations of a grass carp X bighead hybrid.
J, Aquat. Plant Manage. 18:37-39.
Tarver, D. P. 1980. Water fluctuation and the aquatic flora of Lake
Miciosukee. J. Aquat. Plant Manage. 18:19-23.
Taub, S, H. 1972. Exploitation of crayfish by largemouth bass in a
small Ohio pond. Prog. Fish Cult. 34:55-58.
Threinen, D. W., and W. T. Helms. 1954, Experiments and observa-
tions designed to show carp destruction of aquatic vegetation. J,
Wildl. Manage. 18:247-251,
Tryon, C. A., Jr. 1954. The effect of carp enclosures on growth of
submerged aquatic vegetation in Pymatuming Lake, Pa. J. Wildl.
Manage. 18:251-254.
Walker, C. R. 1862. Toxicological effect of several herbicides to bot-
tom welling fish food organisms in Missouri ponds. Res. Rep. N.
Centr. Weed Control Conf. 18:104-105.
Ware, F., R. Gasaway, R. Martz, and T. Drda. 1975. Investigations
of herbivorous fishes in Florida. Pages 79-84 in P. Brezonik and
J. L. Fox, eds. Proc. Symp. Water Quality Management Through
Biological Controls. Rep. ENV-07-75-1. Univ. Florida, Gainesville.
191
-------
Lake Restoration, Protection and Management
Way, J. M,, J. F. Newman, N, W. Moore, and F. W. Knaggs. 1971,
Some ecological effects of the use of paraquat for the control of
weeds in small lakes. J. Appl. Ecol. 8:509-532.
Whitely, J.R. 1964. Control of aquatic weeds in farm fish ponds.
Proc. N. Centr. Weed Control Conf. 20:31-32. Abstr.
Wile, I. 1978. Environmental effects of mechanical harvesting. J.
Aquat. Plant Manage. 16:14-20.
Wile, I., G. Hitehin, and C. Beggs. 1979. Impact of mechanical har-
vesting on Cheming Lake. Pages 145-160 in J. E. Breck, R. T.
Prentki, and O. L. Loucks, eds. Aquatic Plants, Lake Manage-
ment and Ecosystem Consequences of Lake Harvesting. Univ.
Wisconsin Inst. Environ. Stud. Rep.
Willey, R. G., M. J. Doskocil, and C. A. Lembi. 1974. Potential of the
white amur (Ctenopharynagodon /cte//a val.Jas a biological con-
trol tor aquatic weeds in Indiana. Proc. Ind. Aead. Sci. 83:173-178.
Yeo, R. R. 1967. Silver dollar fish for biological control of submersed
aquatic weeds. Weeds 15:27-31.
1980a. Spikerush may help control waterweeds. Calif. Agric.
Exp. Sta. Berkeley. The Stat. 34:12-13.
.. 1980b. Life history and ecology of dwarf spikerush (Eleo-
charis co/oradoensfe).Weed Sci, 28:263-272.
Yeo, R.R. and N. Dechoretz. 1977. Acute toxicity of a herbicidal
combination of diquat, plus copper ion to eggs, alevins and fry
of rainbow trout and to aquatic invertebrates. J. Aquat. Plant
Manage. 14:57-60.
Zettler, F. W., and T. E. Freeman. 1972. Plant pathogens as bio-
controls of aquatic weeds. Annu. Rev. Phytopathology 10:455470.
192
-------
WINTER DRAWDOWN FOR THE CONTROL OF EURASIAN WATER
MILFOIL IN AN OREGON OXBOW LAKE (BLUE LAKE,
MULTNOMAH COUNTY)
N. STAN GEIGER
Beak Consultants Inc.
Portland, Oregon
ABSTRACT
Water milfoil (Myriophyllum spicatum L), chemically similar to the nuisance milfoil species in British Col-
umbia, was first reported in Blue Lake in 1973. Dense stands of the plant have since interfered with recrea-
tional use of the lake. U. S. EPA Clean Lake Phase I funding was used to examine the feasibility of lake
drawdown as a control technique during the winter of 1981-82. Drawdown reduced standing crop biomass
by 44 to 57 percent. Aerial photography also showed a reduction in area occupied by milfoil. Effects of
drawdown on rootcrown viability were forecast by laboratory aquaria tests. Results were mixed because
of less than desirable drawdown and meteorological conditions. However, new growth was exclusively
at the sediment surface from surviving rootcrowns, providing an opportunity for an effective lake-wide ap-
plication of granular 2, 4-D.
INTRODUCTION
Overwinter drawdown has been used to control various
species of nuisance vegetation. Beard (1973) and Nichols
(1975) reported drastic reductions of Ceratophyllum demer-
sum, Myriophyllum spp. and Potamogeton spp. in Wiscon-
sin lakes from single overwinter drawdowns. Goldsby et al.
(1978) confirmed the laboratory observations of Stanley
(1976) in their report on obtaining control of Myriophyllum
spicatum in a Tennessee reservoir by drawdown and subse-
quent dewatering and prolonged subfreezing temperatures.
Overwinter drawdown of Banks Lake in eastern Washington
1980-81 resulted in a significant decrease in area occupied
by Myriophyllum spicatum (Leonard, unpub.).
Blue Lake, Ore., in a maritime region of relatively mild
winters, was drawn down during the winter of 1981-82 to
control Eurasian water milfoil, Myriophyllum spicatum.
Changes in milfoil distribution and biomass were studied and
tests of rootcrown viability performed to determine the ef-
fect of the treatment.
STUDY AREA
Blue Lake is located in Multnomah County, Oregon, 18.5
km east of Portland and approximately 300 m south of the
Columbia River. The oxbow lake now lies behind dikes to
the north and east in a watershed about twice the 26.3 ha
surface area of the lake (Fig. 1). The lake has a shoreline
length of 3,150 m, a longest dimension of 1.32 km, a max-
imum width of 0.2 km, a volume of 8.9 x 105 m3, a max-
imum depth of 7.3 m and a mean depth of 3.4 m. Selected
water quality values are listed in Table 1 and species of at-
tached or rooted macrophytes recently reported in the lake
are listed in Table 2.
At least since 1973 Eurasian water milfoil, Myriophyllum
spicatum, has been the dominant macrophyte growing in
dense stands (as indicated on Fig. 1) under 43 percent of
the lake's total surface and generally to a depth of 2.7 m,
although isolated plants have been observed to depths of
3.7 m. Aside from occasional Potamogeton pectinatus plants
in the lake's northwest milfoil beds and scattered patches
of water lily, Nymphaea odorata, milfoil exclusively occupied
the littoral muck-sediment areas of the lake. Of nearshore
area, only the frequently shaded, gravelly, steeper sloped
shoreline on the south was relatively free of milfoil.
The density of milfoil beds has interfered with recreational
uses of the lake. Controversy over the use of 2, 4-D to con-
1000 FFET
0 300 METERS
CONTOUR INTERVAL 4 FEET
N
Figure 1 .—Map of Blue Lake showing distribution of Myriophyllum and the location of the biomass sampling stations (-h).
193
-------
Lake Restoration, Protection and Management
Table 1. — Selected water quality values for Blue Lake 1981-82 (range of vertical measurements at maximum depth
sampling station, mg/1 except as indicated).
Conductivity (^mos/cm)
Alkalinity1
Total suspended solids
Turbidity (NTU)1
Secchi disk (m)
Total nitrogen
Total phosphorus
Silica1
Dissolved oxygen
Chlorophyll a (mg/m3)1
July 16
92-106
43.0
6.40
2.25
2.4
0.207-1.175
0.044-0.242
3.84
0.8-9.0
6.6
Feb 17
76-80
41.0
10
3.1
1.0
0.628-0.838
0.066-0.076
8.04
12.3-16.0
20.5
March 17
84-91
44.5
16.8
6.2
0.75
0.772-2.918
0.082-0.574
12.81
11.2-12.5
27.0
JulyS
115-133
65.2
2.4
1.9
1.25
0.754-1.271
0.026-0.291
7.5
0.5-10.8
17.8
1 Single value obtained from subsurface (0.5 m) depths
Table 2. — Attached or rooted macrophytes in Blue Lake1
1981-82.
Chara globularis Thuill
Cyperus sp. sedge
Eleocharis sp. spike-rush
Iris pseudacorus L. iris
Myriophyllum spicatum L. Eurasian water milfoil
Nymphaea odorata Ait fragrant water lily
Potamogeton pectinatus L. Sago pondweed
Scirpus validus Vahl. American great bulrush
Typha latifolia L. cattail
1 Naming of species follows Ceska (1980) for Myriophyllum and Hitchcock and Cronquist
(1973), and Wood (1967) for the remainder.
trol plant growth in the lake prompted Multnomah County
to test lake drawdown as a nonchemical control technique
with U. S. Environmental Protection Agency Clean Lake
Phase 1 funding during the winter of 1981-82.
METHODS
The lake level was dropped 2.7 m to the base of most of
the milfoil beds by opening a weir and by pumping. From
Dec. 17 through Jan. 11, the level was dropped 0.3 m by
opening the weir, with the remaining drop by pumping oc-
curring over 38 days from Jan. 12 to Feb. 19. Pumping ceas-
ed Feb. 20 because of pump failure, heavy rains, and high
Columbia River levels.
Effects of the drawdown on milfoil were estimated through
aerial photograph, sampling of milfoil stands to determine
biomass, and viability testing of previously exposed milfoil
plants in laboratory aquaria.
Both color and infrared color photographs were taken from
150 m and 300 m on Aug. 1 and Sept. 12, 1981, and on
July 24 and Sept. 30,1982, to determine visual differences
in distribution and density.
Quantitative samples of milfoil were obtained by scuba
divers from an undisturbed milfoil stand along the north shore
(Fig. 1). A 0.1 m2 wire frame was lowered by diver from sur-
face to bottom enclosing plant stems which were removed
with roots. Replicate samples (n = 4) were obtained from
depth ranges (or equivalent depths of the drawdown) of 0
to 1.2 m, 1.2 to 2.4 m, and 2.4 to 3.7 m. Dry weights and
ash-free weights were determined by Standard Methods
(1976).
The effect of drawdown on milfoil plants was tested by
transferring intact single rootcrown, root, and sediment plugs
to aquaria to test growth at room temperatures. Four replicate
root crowns were obtained from each of two locations within
the 1.2-2.4 m depth range in sandy sediment within the
county swim center and from the fine silt muck at the sampl-
ing location on Figure 1. In addition, four plants with stems
removed were obtained from below the drawdown level in
muck sediments on the north shore near the swim center.
Samples were placed in pots which were put in aerated,
tapwater-filled aquaria. A 15/9 hour light/dark cycle of Grc-
lux fluorescent light was maintained throughout the test.
Water temperature in aquaria through the test ranged from
20 to 23°C. Exposed plants were potted Feb. 15 and
submerged plants Feb. 28; the test was terminated March
29.
RESULTS AND DISCUSSION
Lake levels and levels of the Columbia River adjacent to Blue
Lake before, during, and after drawdown are shown in Figure
2. Subsurface seepage into the lake basin was observed in
exposed weedbed areas. Observed high water retention of
the fine silt-muck sediments, the seepage probably produced
by elevated Columbia River water levels, and rainfall of 52.8
cm from mid-January to mid-April provided moisture to roots
throughout drawdown and refilling.
Approximately 45 percent of the lake bottom was exposed
at the lowest lake level obtained. Between February 4 and
10 subfreezing temperatures occurred with 27 hours of less
than -1°C but above -4°C and 5 hours of less than -4°C
but above -5.6°C temperatures. The greatest depth of freeze
(2.5 cm) was noted on bare mud flats. Freezing depth ap-
peared to be less than 1.5 cm below formerly dense milfoil
stands with dried plant matter appearing to insulate the
rootcrowns.
Water quality effects of drawdown included increase in
total suspended solids, turbidity, decrease in Secchi disk
transparency, increase in chlorophyll a, and an increase in
total nitrogen and total phosphorus concentrations (Table 1).
Decrease in transparency probably prevented establishment
of milfoil at depths lower than it had formerly occurred. Depth
distribution of milfoil in the summer of 1982 was similar to
1981. Samples of dried milfoil shoots from two locations in
the 1.2 to 2.4 m depth range showed mean dry weight values
of 123.1 grams/0.25 m2, and mean values for total
phosphorus of 450 ^g/g and for total Kjeldahl nitrogen of 14.2
mg/g. The influence on water quality from decaying milfoil
vegetation was uncertain.
The winter drawdown greatly reduced biomass of milfoil.-
Ash-free weight/m2 decreased from 44.5 to 57.0 percent
(Table 3). Aerial observations of milfoil distribution and den-
sity (Fig. 3) confirmed the decrease in quantity of milfoil at
all locations.
Tests of milfoil viability in laboratory aquaria showed a con-
sistent 75 percent rootcrown death of exposed plants at four
locations from both sandy and muck sediments (Table 4).
194
-------
Since milfoil biomass to rootcrown number varies according
to depth, sediment type, nutrients, plant vigor, and possible
crowding, no relationship between root crown survival and
biomass/m2 was expected. The reason for delayed growth
in plants from sandy sediments (Table 4) is unknown. Aerial
photographs suggest milfoil reduction may have been more
severe in the swim center where the sediments were sandy.
Since the lake did not refill to the previous year's levels,
milfoil plants in the upper 0 to 1.2 m depth range were also
exposed to drying conditions and high temperatures before
sampling occurred at the end of July 1982. In July live and
reddened milfoil plants were common along the exposed
shorelines, indicating high tolerance for these conditions in
contrast to the effects of freezing.
Exposing nearly the entire milfoil stands in the lake to the
dewatering and freezing of stems eliminated nearly all milfoil
biomass above the sediment. The only stems remaining were
in a narrow fringe of milfoil lying below the drawdown level.
Physical, Chemical, and Biological Control of Aquatic Macrophytes
Growth from submerged and exposed surviving rootcrowns
began in late March and early April. By mid-April, previous-
ly submerged plants were approximately 0.3 m tall while
previously exposed plants grew slower to only 10 cm in
height.
Since new milfoil growth was noted in all areas previous-
ly occupied and appeared vulnerable to granular 2, 4-D,
Multnomah County reconsidered its decision against using
herbicides in the lake. With the support of lakeside residents
and the lack of objection in a public hearing, the County ap-
plied granular isooctyl ester or 2, 4-D on July 24, 1982.
By the end of July when the herbicide was applied, plants
had grown to the surface of the lake, which then was 0.9
m below the level before drawdown. The requirements of
the political process had delayed the herbicide application
beyond what appeared to be the most favorable application
time in April. Application effects have yet to be determined.
Table 3. — Myriophyllum splcatum biomass at location 61 m west of county swim dock: four replicates per x before
(Sept. 13, 1981) and after (July 22, 1982) lake drawdown.
Depth (m)
0-1.2
1.2-2.4
2.4-3.7
1981
x ash-free wt (g/m2)
117.18
244.13
24.35
SD
± 29.06
±215.53
± 10.15
1982
x ash-free wt (g/m2)
61.58
135.58
10.48
SD
17.86
57.31
4.00
% reduction
in ash-free wt/m2
47.45
44.46
56.96
Figure 2.—Levels of Blue Lake and the Columbia River July 1981 - Sept. 1982.
195
-------
Lake Restoration, Protection and Management
Figure 3.—Photographs of milfoil beds from 300 m before (Sept. 12, 1982) and after (July 24,1982) lake drawdown. First pair, location
west of swim center (arrows indicate position of sampling transect) before (top) and after (bottom). Second pair, location at east end
of lake (see Fig. 1) before (top) and after (bottom).
CONCLUSIONS
The overwinter drawdown of Blue Lake greatly reduced
dense beds of milfoil that had maintained position and den-
sity in the lake relatively unchanged since 1973. Drawdown
reduced milfoil biomass by shoot-death from drying and
freezing and possibly by suppressing growth through lower
spring-summer water levels. Laboratory tests indicated
drawdown and freezing affected rootcrowns as well as
shoots.
Regrowth of new stems from surviving rootcrowns was
widespread so it does not appear that drawdown in itself,
Table 4. — Occurrence of growth of Myriophyllum in laboratory aquaria In potted plugs from five locations in Blue Lake
(n = 4; numbers are percentage of four root crowns showing growth)
Location of Plants
Swim center sandy sediment
- exposed (5 ft.)
Swim center sandy sediment
- exposed (7 ft.)
North shore muck sediment
- exposed (5 ft.)
North shore muck sediment
- exposed (7 ft.)
North shore muck sediment
- submerged (9.5 ft.)
Feb. 15
0
0
0
0
—
Feb. 22
0
0
0
0
—
Feb. 28
0
0
25
25
0
Mar. 5
0
0
25
25
25
Mar. 9
251
O1
25
252
75
Mar. 16
25
0
251'3
25
100
Mar. 19
25
25
25
25
100
Mar. 22
25
25
25
25
100
Mar. 29
25
25
25
25.
1004
1 appearance of juvenile Potamogeton pectinatus in one pot of four
2 appearance of juvenile Chare gfobu/aris in all four pots
> appearance of juvenile Cnara globu/aris in 3 of 4 pots
' appearance of juvenile Cnara gfobularis in 1 of 4 pots
196
-------
a nonchemical solution to the problem, is a sufficient control
method given the meteorological conditions encountered. In
Blue Lake, the combination of drawdown with timely her-
bicide application may have reduced milfoil growth to non-
nuisance levels for summer 1983. Proper springtime applica-
tion of herbicide may in itself be sufficient to control milfoil
to acceptable levels in this lake although this solution has
not been tested.
ACKNOWLEDGEMENTS: The author acknowledges the support
of Beak Consultants Inc. staff who provided field and laboratory
assistance and encouragement for this study, especially Bruce Ed-
dy, Krystyna Wolniakowski, Lynn Foster, and Ed Mulvihill. The
author is also indebted to the following Multnomah County staff
for their guidance and support throughout the study and for their
efforts in assembling the equipment for the drawdown: Richard
Engstrom, Leo Sorensen, Guy Schwartz, and Max Kilgore. Without
the public-spirited loan of the pumps by the Port of Portland, the
loan of dredge pipe by Reidel International Inc. and funding from
U. S. EPA, this project would not have been possible.
Physical, Chemical, and Biological Control of Aquatic Macrophytes
REFERENCES
Beard, T. D. 1973. Overwinter drawdown. Impact on the aquatic
vegetation in Murphy Flowage, Wis. Tech. Bull. No. 61. Dep.
Nat. Resour. Madison, Wis.
Ceska, Oldriska. 1980. Chromatographic analyses of milfoil
from Blue Lake, Ore. Ceska Geobotan. Res. Co. Victoria, B. C.
Letter comm., Jan. 16.
Goldsby, T. L, A. L. Bates, and R. A. Stanley. 1978. Effect of water
fluctuation and herbicide on Eurasian water milfoil in Mellon Hill
Reservoir. J. Aquat. Plant Manage. 16:34-38.
Hitchcock, C. L., and A. Cronquist. 1973. Flora of the Pacific North-
west. Univ. Washington Press, Seattle.
Leonard, R. 1982. Conservation agronomist, Bur. Reclam.
Columbia Basin Project. Euphrata, Wash. Personal comm. March 1.
Nichols, S. A. 1975. The use of overwinter drawdown for aquatic
vegetation management. Water Res. Bull. 11:1137-1148.
Standard Methods for the Examination of Water and Wastewater.
1976. 14th ed. Rand, M. C., A. E. Greenburg, and M. J. Taras,
eds. Am. Pub. Health Assoc. Washington, D. C.
Stanley, R. A. 1976. Response of Eurasian water milfoil to sub-
freezing temperature. J. Aquat. Plant Manage. 14:36-39.
Wood, R. D. 1967. Charophytes of North America. Stellas Print-
ing, West Kingston, R. I.
197
-------
MILFOIL CONTROL IN SEATTLE AND THE KING COUNTY REGION:
METRO'S HARVESTING PROGRAM
W. TERRA PRODAN
Municipality of Metropolitan Seattle
Seattle, Washington
ABSTRACT
The Municipality of Metropolitan Seattle has conducted an aquatic plant harvesting program for 3 years
to control growth of Eurasian water milfoil in public areas. Two machines harvested over 166 hectares
(400 acres) during the 3-year period. This case study discusses costs, funding, and some operational
problems.
INTRODUCTION
The aquatic plant that historically has caused the Seattle/King
County area the most problems is Eurasian water milfoil. It
clogs boat propellers and annoys swimmers and other direct
lake users. Metro field workers first documented milfoil in
the early 1970's in Lake Sammamish, one of the large lakes
in the Seattle/King County region. Because Metro is respon-
sible for areawide water quality planning, public pressure in
the late 1970's led the Metro Council, which consists of local
government representatives, to consider a series of studies
on the various chemical and non-chemical methods of
aquatic plant control, including literature reviews, laboratory,
and field studies.
Metro has also conducted surveys of the extent of milfoil
growth for 6 years. The Metro Council passed a resolution
in 1980 directing Metro staff to pursue nonchemical control
options. Current Metro policy advises against using the her-
bicides dichlobenil (or Casoron), diquat, and 2, 4-D. Non-
chemical control methods are preferred, but if all other
methods are considered not feasible, then endothall is the
chemical suggested for use.
Metro's harvesting program, which receives most of the
agency's milfoil control efforts, is funded largely by the
U. S. Army Corps of Engineers (70 percent). That money is
passed through the Washington State Department of Ecology
for costsharing. Local jurisdictions pay the remainder of pro-
gram costs, roughly 30 percent. Council restrictions prohibit
Metro from contributing funds. Currently participating are one
county government, one State agency (Washington State
Parks), three cities, and three towns. The program relates
only to public waters, particularly those that receive heavy
recreational use such as swimming beaches and boating
moorages and marinas. Metro believes the private sector
should absorb the costs of control in private waters.
CASE STUDY AND DISCUSSION
Metro began this harvesting program 3 years ago, using two
Mud Cat aquatic weed harvesters (Model H7-450). The Mud
Cat is 10.7 meters long by 3.7 m wide and 2.7 m high. Its
2.2 m wide cutting bar can cut to a variable depth of up to
1.5 m. The harvester is driven by two hydraulic paddle
wheels that are independently reversible. It has a
33-horsepower engine. Purchase price in 1980 of two Mud
Cats, a shore conveyor, and trailer was about $100,000.
Metro's crew consists of four seasonal employees, usually
college students, coordinated by one lead individual.
Because most crew members return each year, they are ex-
perienced in repairing the equipment on site. Crew work
hours are staggered, thus each machine can harvest a max-
imum of 12 hours a day.
Throughout the season, Metro keeps close track of ex-
penditures. Cost elements of the program are on-site and
off-site labor, insurance, equipment, fuel, general operating
supplies, all materials, maintenance, and administration
(tracking, environmental regulatory processes, reports) (Table
1). The cost of equipment has been a high percentage of
program costs (43 percent this year) because of the 3-year
amortization, an unusually short payback schedule. Ad-
Table 1. — Program elements.
Operations
Harvesting equipment
Labor, on-site1
Insurance
Vehicles & boat
Fuel
General operating supplies
Maintenance, materials, & labor
Administration
Contract arrangements, permits,
environmental regulatory
processes, budget preparation,
cost tracking, report writing
%of
program costs
35
16
6
1
1
1
29
12
100
11ncludes disposal
ministration costs of Metro (12 percent), a public agency,
are likely higher than for a private business because of
necessary State Environmental Policy Act and insurance re-
quirements and grant tracking. Metro keeps track of all costs,
even those that would ordinarily be considered "hidden" or
"indirect," such as insurance and administration. Each year
the program has come in below budget. The agency is also
able to give participants in the program a ceiling cost to them.
The harvesting program began on July 17,1980, with one
harvester. A second harvester was purchased and arrived
Aug. 9. The program ran for 8 weeks. Thirty-six hectares
(90 acres) were harvested, with 224 cubic meters (293 cubic
yards) of material removed from seven major locations.
198
-------
During 1981 the two Mud Cat machines harvested 69 hec-
tares (172 acres) from early July to mid-September—nearly
twice that harvested in 1980. Two hundred and seventy cubic
meters (353 cubic yards) of material were removed from area
lakes. Since Metro's dump truck holds 8 cubic yards, the
harvest produced roughly 40 dumptruck loads in a 10-week
period.
This year, Metro's program went 12 weeks and included
three additional jurisdictions. Sixty-two hectares (154 acres)
were harvested, for a total of 336 cubic meters (440 cubic
yards), or about 55 truckloads of material—more than in the
two previous years. Historically, the highest density has been
8.2 cubic meters per hectare (4.3 cubic yards per acre). This
year, however, saw a much larger volume of material, at one
site exceeding an average 16.2 cubic meters per hectare
(8.5 cubic yards per acre) for 1 week.
The harvesters made a large circuit of Lakes Washington
and Sammamish, working at nine different sites. Lake
Washington is primarily an urban-suburban lake, and Lake
Sammamish is primarily suburban.
In the two lakes as a whole, 13 percent of the total acreage
of milfoil present was harvested this year. Only public areas
are harvested; control of other areas is left up to lakefront
citizens without any public financing. In Union Bay, the bay
most heavily infested with aquatic plants, the harvesting pro-
gram concentrates on public high-use areas. It keeps those
areas clear of milfoil. However, only 3 percent of the total
milfoil area is harvested. In other bays, such as Cozy Cove,
a larger area is harvested (49 percent). It must be empha-
sized that local jurisdictions designate priority areas in terms
of public benefit.
The manufacturer of Metro's harvesting equipment
estimates a cutting rate of .2 ha (.5 acre) per hour. Metro
has found, however, that .1 ha (.25 acre) per hour may be
more accurate. The manufacturer bases its estimate on use
of the equipment in open water. Metro experience shows
that areas with obstructions such as docks can involve con-
siderably more time. Also, the Metro figure includes transport,
startup, and down time. Because two machines are involv-
ed, with an off-load site up to one-half hour's travel time from
the cutting area, travel time back and forth can be high. In
addition, communications have been difficult. Metro workers
have tried radios and walkie-talkies but have not found a
satisfactory method to communicate from the off-load site
to the harvesting site. Thus, the crew often had to travel back
to the off-load site to check in with the lead individual, caus-
ing long delays.
This year the crew experienced more down time than
previously. Incidentally, on-site repairs are preferable
because of the distance to maintenance facilities. One pro-
blem experienced is water accumulating in the pontoons.
Metro hopes to solve this problem next year by installing a
bilge pump in the pontoons. Currently, the crew has to keep
an eye on each machine. When it lists they know it's time
to pump out the pontoons, adding to on-site down time. The.
manufacturer is apparently continually working to refine
equipment design. Meanwhile, Metro mechanical crews have
also redesigned parts of the machines to make them operate
more smoothly and efficiently. For example, they painted the
front edge of the cutter bar white to enable the operator to
see the plant material more easily.
One other aspect of the nuisance nature of milfoil is that
it accumulates on the shoreline in the autumn months as
the plants begin to break down. Milfoil can reproduce by
Physical, Chemical, and Biological Control of Aquatic Macrophytes
flowers or by fragmentation. As the growing season comes
to an end, fragments broken from the plant by wind or wave
action develop roots, sink to the bottom, and take root there.
The presence of this material appears to be almost as
much a problem as the growing plant itself. Although Metro
has received many complaints from local citizens, no local
government had requested assistance until this year, when
the city of Seattle asked the Metro harvesters to pick up
floating material that had accumulated following Lake
Washington's annual hydroplane race.
Besides wind and waves, recreational boaters also frag-
ment plants. In response to questions about the harvester
increasing fragmentation, technical staff and the operations
manager believe the two harvesters do not add significantly
to the drifting material. The machines are designed to remove
all floating material from the water.
Metro receives a number of inquiries from around the
country regarding program costs: Table 2 summarizes the
3-year program. As mentioned previously, these costs in-
Table 2. — Program summary.
Acres harvested
Hectares harvested
Cubic yards of milfoil
Cubic meters
Average cost per acre
Ave. cost per hectare
Cost to locals, per acre
Cost to locals, per hectare
1980
90
36
292
224
$1052
$2630
$ 347
$ 868
1981
171
69
373
285
$ 535
$1338
$ 143
$ 358
1982
154
62
440
336
$ 630
$1575
$ 189
$ 472
elude equipment amortization over a 3-year period, in-
surance, and administration. Therefore, costs per unit area
should not be taken as an index of the costs that could be
anticipated by a private operator. Generally costs have
decreased each year through increased efficiency. Since 70
percent of the costs are reimbursed by the Corps of
Engineers through the Washington Department of Ecology,
local jurisdictions share 30 percent of the costs, this year
totaling approximately $472 per hectare, or $189 per acre.
The cost of disposal is included this year in the program
element "Labor, on-site," since the crew removed the milfoil
to acceptable out-of-the-way disposal sites as part of their
operations. In all 3 years of the program, milfoil has been
composted by various parks departments. Metro promoted
this disposal method since dumping at a local landfill costs
$18 a ton (1982 costs), plus higher transit costs, which would
increase overall costs. A study funded by the U.S. En-
vironmental Protection Agency (Swayne, 1980) showed
milfoil to be relatively high in nutrients, primarily nitrogen,
potassium, and phosphorus. Metro currently promotes the
use of milfoil as a soil amendment, mulch, or compost.
This year, for the first time, cut milfoil from the harvesting
program was made available to the general public through
advertisements in newspapers and on local radio. Persons
desiring the material, available at no cost, were directed to
one of several sites where they could pick up the amount
needed. So far, feedback on this aspect of the program has
been extremely favorable.
Metro has been very pleased with the low costs and the
effectiveness of the program. Since the equipment will be
fully paid for, next year Metro expects costs to be the lowest
yet.
199
-------
GUIDELINES FOR USE OF THE HERBICIDE 2,4-D TO CONTROL
EURASIAN WATER MILFOIL IN BRITISH COLUMBIA
ROBERT W. ADAMS
Water Management Branch
Ministry of Environment
Victoria, British Columbia
ABSTRACT
Observations on the effectiveness, persistence, and drift of 2,4-D (Aqua-Kleen) were reviewed follow-
ing 33 treatments to control Eurasian water milfoil between 1976 and 1981. Technical information
was summarized to guide implementation of future treatment programs. Nuisance growth of the aquatic
plant was eliminated for up to 3 years following treatments in some sites. Reduced effectiveness in
some areas was attributed to groundwater discharge, rocky substrate, suspension of herbicide granules
in the foliage of dense populations, or small treatment size (less than 0.2 ha). Concentrations of 2,4-D
above the drinking water standard (0.1 mg/l), persisted for a maximum of 4 to 6 days, and drift at
this concentration was detected at a maximum of 125 m from open water treatments of 12 ha or less
in large lakes. Low concentrations of herbicide (just above 0.001 mg/l) were observed to drift up to
2 km and, typically, persisted for less than 40 days in treatment areas. It was unlikely that well water
near treatment areas could be contaminated with 2,4-D, but the potential for hydraulic continuity bet-
ween wells and treated water should be considered. No significant detrimental effects on nontarget
organisms were observed although agencies responsible for wildlife resources should be contacted
about new treatment sites.
INTRODUCTION
In 1976, the herbicide 2,4-D was selected for testing by staff
of the Water Management Branch of the Ministry of Environ-
ment for control of Eurasian water milfoil. This aquatic plant
was first noted as a nuisance in Okanagan Lake, British Col-
umbia in 1971 and was observed to spread and occupy over
600 ha of lake bottom in six Okanagan Valley mainstem
lakes by 1977. Trials with 2,4-D commenced after limited
success was obtained in reducing Eurasian water milfoil
spread following tests of several mechanical technologies
and three other herbicides.
In Canada, only the butoxyethanol ester (BEE) form of 2,4-D
(Aqua-Kleen) is registered for aquatic plant control. Small
attaclay granules are impregnated with 20 percent 2,4-D BEE
active ingredient. In tests by Agriculture Canada, no diox-
ins were found in the herbicide at 1 ppb concentration or
higher, confirming that this 2,4-D formulation was manufac-
tured by a "clean process." The Canadian Drinking Water
Standards (Health Welf. Can., 1978) establish a maximum
acceptable level of 0.1 mg/l 2,4-D and an objective concen-
tration of less than 0.001 mg/l 2,4-D in domestic water sup-
plies. Testing of 2,4-D BEE for Eurasian water milfoil con-
trol was conducted following a review of project objectives
by the B.C. Aquatic Weed Advisory Committee, represen-
ting various concerned agencies including Fisheries and
Oceans Canada, Environment Canada, Agriculture Canada,
and British Columbia Ministries of Health, Environment,
Agriculture, and Lands, Parks, and Housing.
2,4-D was selected for testing in the face of considerable
controversy raised by opponents to herbicide use, who
equated 2,4-D to 2,4,5-T and "Agent Orange" which may
be contaminated with dioxins. Therefore, the B.C. Ministry
of Environment adopted a conservative approach and
established objective concentrations of less than 0.001 mg/l
2,4-D in domestic water supply intakes for even the shortest
duration. In retrospect, it is now apparent that this conser-
vative position prohibited extensive use of this herbicide for
aquatic weed control and treatments eventually performed
were too small and too late to stop the spread of Eurasian
water milfoil. However, 2,4-D BEE proved to be the most
suitable Eurasian water milfoil control method available in
some locations, and use of this herbicide by local agencies
is expected to continue for site-specific aquatic weed control.
As a result of the stringent water quality objectives imposed
during testing of 2,4-D, considerable information was col-
lected on the persistence, drift, and effectiveness of this her-
bicide from samples of water, sediment, plant tissues, and
nontarget organisms. The data were compiled in annual
reports (Lim, 1978; Lim and Lozoway, 1977; Anon. 1980;
Goddard, 1980; Rudolph and Dyer, 1981; Wallis, in prep.).
This information is summarized here to assist agencies iden-
tify when 2,4-D BEE may be most appropriate for Eurasian
water milfoil control and to guide implementation of treat-
ment programs.
EXPERIMENTAL TREATMENTS
From 1976 to 1981, 33 Aqua-Kleen applications to control
Eurasian water milfoil were monitored by Water Management
Branch staff. Treatment sizes ranged from a 0.9 m2 lab test,
to spot applications up to 1 ha (in an attempt to kill small
clumps of plants just becoming established) to large
treatments up to 12 ha (30 acres). Field treatment locations
are identified on Figure 1.
The primary nuisance of Eurasian water milfoil results from
the high density of populations which form surface mats,
precluding water use by swimmers and boaters. Air
photographs were used to compare weed densities before
and after herbicide treatments. Scuba divers observed the
condition of plants within and adjacent to treated areas from
several months before to several years following treatments.
Stem densities and/or root biomass of Eurasian water milfoil,
and in some cases cover of nontarget plant species, were
recorded in 0.25 m2 quadrants along transect lines prior to
and after selected herbicide applications.
Over 2,000 water samples were taken in treatment areas
to investigate the persistence of 2,4-D in the water column.
200
-------
Physical, Chemical, and Biological Control of Aquatic Macrophytes
Naswhito
Creek 1977
Wilsons Landing
1978
Oiler I .ike
_-Norlh Ann 19/b \'tl t 19/8
Swan I ,lke
-('•VERNON
— Norlh .-.id PI/9
-- -Lishiitn Estates 1978
-Spot treatments 19791980
'K:ilanmlk;i I .ike
-Spot IreatmtjnK 1980
-North <;nd I960
' Wood I n
-South end 19/9 1980
• PENTICTON Okanagan Valley
Northeast Skaha Lake 1977
Sk;ih;i Luke
Figure 1 .—Location and dates of 2,4-D treatments in Okanagan
Valley Lakes, 1976-1981.
Generally, a minimum of two sampling stations was
established per application site and at each station near sur-
face and near bottom water samples were taken on sche-
duled days until no herbicide was detected. Several
treatments were intensively sampled including hourly for the
first 6 hours after treatment, then daily for 10 days, then
weekly until no herbicide was detected. In some cases
samples were taken of sediment, plant stems, and
invertebrates.
Herbicide drift sampling stations were established outside
most application sites. These were located in an array around
the application site or in the direction of anticipated drift. Over
3,000 water samples were taken at daily to weekly intervals
to determine the rate of dissipation and concentration of her-
bicide at various distances from treatments.
All water samples were analyzed for 2,4-D residues at the
Ministry of Environment Laboratory in Vancouver, B.C., us-
ing a Hewlett Packard 58404 gas chromatograph equipped
with a nickle 63 electron capture detector. The analytic pro-
cedure used by the laboratory was capable of detecting 2,4-D
at a lower limit of 0.001 mg/l. At the water temperature
(17-22° C) and pH (8.2-8.7) in the treated areas, 2,4-D BEE
is rapidly hydrolyzed to 2,4-D acid and has a theoretical half
life of 5 to 16 hours (Bothwell and Daley, 1981). Therefore,
water samples were usually analyzed for the more persis-
tent 2,4-D acid. However, following several treatments,
samples were analyzed for 2,4-D BEE, 2,4-D acid, and its
breakdown product, 2,4-dichlorophenol.
In addition to observing herbicide efficacy, persistence,
and drift, a special study was made of the potential for 2,4-D
in sediments to move in groundwater and to contaminate
lake side wells (Kangasniemi and Nagpal, in prep.). Also the
B.C. Fish and Wildlife Branch studied the effect of 2,4-D on
nontarget organisms, including fish, waterfowl, and in-
vertebrates (Robinson and Morley, 1980; Robinson, 1981;
South and Robinson, in prep.).
2,4-D vs OTHER CONTROL METHODS
Since 1972, the B.C. Ministry of Environment has been
researching, developing, and testing aquatic plant manage-
ment technologies (Newroth, 1979, 1980). Several innova-
tions have been developed for special situations but no single
method has been found suitable for universal use.
Results of testing the herbicide 2,4-D indicate it is ap-
propriate as one tool among several for aquatic plant
management in specific situations; its suitability depends on
such factors as site-specific costs and potential effectiveness.
The main advantage of 2,4-D over other control methods is
that it can eliminate nuisance growth for 2 to 3 years in some
sites. This compares very favorably with harvesting, for ex-
ample, which is required twice per year in some Okanagan
Valley lake sites. The major limitation of herbicide use is
possible conflicts with water use, since if alternate water must
be supplied to lake water users until the herbicide has
dissipated, costs of treatment and monitoring may increase
significantly. Also application costs per unit area increase
for small treatments (<10 ha).
The costs and limitations of the most practical control
methodologies tested in British Columbia are compared in
Table 1. In addition to these cost comparisons, evaluation
of the herbicide 2,4-D for a specific use requires review of
a number of technical considerations on timing of applica-
tion, physical characteristics of the area to be treated, and
the implications of treatment size to drift and persistence of
herbicide.
TECHNICAL CONSIDERATIONS FOR
2,4-D TREATMENTS
Effectiveness
Considerable variation was observed in the effectiveness of
the 33 2,4-D treatments monitored between 1976 and 1981.
In most sites, 2,4-D treatment at recommended application
rates removed nuisance Eurasian water milfoil for the re-
mainder of the treatment growing season and one subse-
quent growing season. Measurements of stem and root
abundance following treatments in 1977 and 1978, indicated
complete plant removal from an enclosed 2.5 ha lagoon ad-
jacent to Okanagan Lake, while stem densities were re-
duced from 54 to 100 percent at open lake sites at the end
of the treatment growing season (Table 2). In some of the
small treatment plots (<0.5 ha) rapid regrowth occurred the
year following treatment. In three larger treatments (2.0, 4.0,
and 5.3 ha) stem densities were still 70 to 90 percent less
than surrounding untreated densities 2 years after
treatments.
Treatment conditions that may result in variable plant mor-
tality include the following:
Application rate—Recommended application rates
specified for the Aqua-Kleen formulation of 2,4-D are 22 kg
a.i./ha for light to moderate infestations or 45 kg a.i./ha for
dense infestations, where the water is more than 2.4 m deep
and where there is a large volume turnover (water move-
ment). Observation of treatments in the Okanagan Lakes
system indicated that the low application rate resulted in
more rapid regrowth than the higher application rate in open
lake sites. Thus the 44 kg a.i./ha rate is recommended for
large lakes where the herbicide can be quickly diluted.
Dissipation of the herbicide was enhanced by wind-induced
201
-------
L3K6 nestoration, rrotecuon ana Management
currents, groundwater discharge and creek inflow. In enclos-
ed water bodies where nearly all the surface area was treated
(ponds, lagoons) the lower application rate (22 kg a.Uha) was
adequate to achieve long-term control.
Density—It would appear that root contact and absorp-
tion of 2,4-D is essential to plant mortality. In dense weed
beds, Aqua-KIeen pellets tend to become trapped and
suspended in the foliage, thereby reducing the amount of
herbicide reaching the substrate and root region of the plant.
After some treatments, plants in the center of dense clumps
survived, presumably because few herbicide pellets drop-
ped through the dense foliage to release 2,4-D near the root
system. To avoid this situation, it is recommended that
treatments be carried out before dense surface mat forma-
tion in early summer or when the canopy has opened up
in late summer. Also, herbicide applications may be in-
tegrated with mechanical harvesting to reduce the dense sur-
face mat prior to granular herbicide application.
Substrate type—2,4-D treatment of Eurasian water milfoil
growing on rocky bottoms in exposed Okanagan Lake loca-
tions resulted in relatively low plant mortality. This supports
the hypothesis that 2,4-D release and persistence near roots
is important for plant mortality. Finely divided sediments ap-
pear to retain larger quantities of 2,4-D than coarse
substrates. On rocky bottoms, where roots penetrate
crevices, the herbicide granules may have less contact with
roots.
Herbicide Persistence in the Treatment
Area Water Column
Laboratory experiments have shown that 2,4-D Is released
from Aqua-KIeen granules most rapidly the first day after ap-
plication, then more slowly over several subsequent days
(Bothwell and Daley, 1981). The herbicide is released mainly
as 2,4-D BEE which is quickly hydrolyzed to 2,4-D acid at
Table 1. — Estimated operating costs and treatment rates for selected Eurasian water milfoil control methods in British
Columbia.
Method
Operating exist1
(cost/hectare)
Average rate2
(hectares/day)
Comments
1 2,4-D (BEE)
4. Rotavator
$700/ha @ 10 ha/day
to
$2500/ha @ 1 ha/day
(herbicide cost at 45 kg a.i.
per ha = $200/ha; capita!
cost/applicator unit about
$12,000)
30 ha/day Treatment area limited to 1 ha requires the same preparation
and monitoring effort as a 10 ha plot, therefore cost/ha is
much larger for smaller treatments. These operating costs do
not include supplying any alternate water, which can increase
costs considerably. Allowance for minimal monitoring is
included.
2. Hydraulic
Dredge
3. Diver Dredge
$5,000/ha
(capital cost/dredge unit
about $100,000)
$2,500/ha
to
$19,000/ha
(capital cost/dredge unit
about $12,000)
0.1-0.2 ha/day
.20 ha/day ex-
tremely sparse
.02 ha/day
moderate
.01 ha/day
very dense
Operating cost obtained from a 2.9 ha test conducted
Aquatic Studies Branch in 1975. Rate is dependent on
ment depth removed and obstacles.
by
sedi-
Treatment costs and rates are highly dependent on weed
density.
$400/ha
to
$1,200/ha
(capital cost/machine about
$60,000)
0.2-0.5 ha/day Cost and rate are dependent on the degree of difficulty in
treating the site and the number of passes necessary to
remove most of the plant material.
5. Shallow Water
Tillage
$400/ha
(capital cost/machine about
$60,000)
0.7 ha/day Cost may increase in soft substrate areas or where obstacles
are present.
6. Harvester $1,200/ha
(capital cost/machine about
$55,000)
0.4 ha/day
Rate includes shore disposal of spoils.
7. Bottom Burlap-$l2,100/ha
Barriers Polyethylene-$13,000/ha
Window Sereen-$20,000/ha
Aq uasc reen-$50,000/ha
.05 ha/day Polyethylene, window screen, and Aquascreen may be
reusable and their high cost could likely be amortized over
several years.
' Figures do not include rental or depreciation of capital costs of machinery, expenses incurred from transport/launching of machines, or administration costs.
Costs in 1982 Canadian dollars.
2 The above rales apply to an 8 hour work day and include set-up time.
202
-------
high pH and warm temperatures. The acid is dispersed in
the water column, usually over a period of several days, and
degradation occurs by various processes over several weeks
or months.
Table 3 summarizes herbicide persistence data following
the 1976-1981 treatments. The maximum concentrations of
2,4-D acid sampled in open water treatment areas were 0.42
mg/l in surface water and 3.25 mg/l in bottom water. Max-
imum 2,4-D concentrations exceeded 0.1 mg/l in samples
taken within about 40 percent of the open water treatment
sites, but rapidly declined to less than 0.1 mg/l 1 to 7 days
following these treatments. Generally, 2,4-D acid concentra-
tions declined to less than 0.01 mg/l within 15 days although
there were exceptions where infrequent samples contained
concentrations just above 0.01 mg/1 for up to 21 days after
treatment. The decline to less than detectable levels of 0.001
mg/1 in open water treatments was more prolonged, but rare-
ly longer than 40 days. Herbicide persistence was not related
to treatment size; for example, residues from spot treatments
(<1 ha) could remain detectable for over 30 days but
residues from a 5.3 ha treatment were not detectable after
15 days.
Persistence was prolonged if there was little water move-
ment (i.e., no wind or river-generated currents) to disperse
Physical, Chemical, and Biological Control of Aquatic Macrophytes
the herbicide. In an extreme example of this situation, the
water column remained at barely detectable levels (0.001
to 0.003 mg/l) for about 100 days in a corner of Wood Lake
where water movement was restricted following treatment
of 5.9 ha in 1979.
In enclosed treatment sites, concentrations as high as 2.06
mg/l were obtained in surface samples and 4.00 mg/l in bot-
tom samples. Residues generally declined more slowly than
in open water sites, taking over 190 days to decline to
nondetectable levels following one treatment. Such en-
closed sites are those where sufficient area was treated to
produce detectable concentrations in the entire water col-
umn of a lake or pond, or in an isolated arm or semi-enclosed
bay of a large lake.
The herbicide persistence data from enclosed sites in-
dicated that calculation of the theoretical concentration of
2,4-D applied in a lake epilimnion, assuming simple dilution
and complete dispersion of the herbicide, provides a rough
estimate of the size of treatment that will produce detectable
concentrations in most of the water column. For example,
in Wood Lake (930 ha), where 18.5 ha were treated in 1980,
the predicted concentration of 2,4-D in the epilimnion, assum-
ing complete dilution and dispersion, was 0.0075 mg/l. The
observed concentrations in drift water samples averaged
Table 2. — Summary of Eurasian water milfoil treatment site characteristics and 2,4-D (BEE) effectiveness at end of treatment
seasons in sample areas, 1977-1978.
Maximum daily 2,4-D residue sample means (mg/l)1
Percentage reduction in
E.W.M. abundance2
Herbicide
treatment
sites
Skaha Lake
Area A
Area B
Naswhito Creek
Area A
Area B
Area C3
Westside Cays
Area A
Area B
Area C
North Arm
Area A4
Area B
Area C
Area D*
Kelowna Foreshore
Summerland Marina
Area A
Area B3
Spot Treatments
Kalamalka Lake
Lisheen Estates
Okanagan Lake
Wilson Landing
Squally Point
Sediment
type
sand/clay
sand/clay
silt/clay
silt/clay
silt/clay
organics/silt
organics/silt
organics/silt
silt
silt/clay
silt/clay
silt
sand/silt
sand/silt
sand/silt
sand/silt
rock
rock
Water
column
0.051
0.007
0.052
0.052
0.052
0.972
0.972
0.972
0.129
0.095
0.077
0.053
0.047
0.197
0.197
0.008
0.004
NDC
Day
1
0
1
1
1
6
6
6
2
2
0
2
1
0
0
1
1
Hydrosoil
9.92
2.73
52.90
52.90
52.90
69.40*
34.11
17.00
9.15
34.61
29.73
27.55
2.13
16.89
6.92
4.95
NDC
NDC
Day
2
11
4
4
4
5
2
9
8
2
7
3
9
7
1
Stems
30.60
68.00*
51.41
51.41
51.41
39.20*
70.10*
110.10
67.00
36.18
31.50
11.51
30.64
23.79
40.68
NDC
NDC
NDC
E.W.M.
Day
1
2
1
1
1
7
2
3
2
5
4
3
6
3
3
tissue
Roots
9.35
82.00*
29.19
29.19
29.19
24.50*
241.00*
67.30
12.68
23.13
19.04
11.48
30.77
25.21
38.57
NDC
NDC
NDC
Day
1
7
21
21
21
5
6
5
5
10
2
3
6
7
3
Stem
density
moderate
moderate
94
87
93
100
100
100
100
54
77
79
98
high
high
low
moderate
low
luui iuai IUG
Root biomass
g dry wt.
NDC
NDC
NDC
NDC
NDC
NDC
NDC
NDC
NDC
NDC
NDC
NDC
94
60
53
NDC
NDC
NDC
1 Samples were averaged lor each day; maximum average values are listed
* At end ol treatment season
3 Treated mechanically several weeks before herbicide application
•" Previously treated in 1976
NDC - No data collected
• Indicates only one sample collected per day
203
-------
A. Treatment smaller
Location
N. Arm Okanagan
Lake
N. Arm Okanagan
Lake
Lisheen Estates,
Kalamalka Lake
Wilson's Landing,
Okanagan Lake
S. End Wood Lake
N. End Kalamalka
Lake
Cosens Bay,
Kalamalka Lake
Cosens Bay,
Kalamalka Lake
Coldstream Ck.,
Kalamalka Lake
Lisheen Estates,
Kalamalka Lake
S. End Kalamalka
Lake
than 1 ha in
Application
date
31/5/76
31/5/76
12/7/77
12/7/77
12/7/77
17/7/78
01/8/78
21/6/79
21/6/79
21/6/79
21/6/79
21/6/79
11/7/79
01/8/79
12/9/79
26/6/80
09/9/80
30/7/80
25/9/80
laoie ..
i. — rersisien
open water, listed by treatment
Treatment
size (ha)1
0.60
0.60
0.67
0.36
0.30
0.07
0.10
0.50
0.10
0.10
0.21
0.21
0.4 (2 sites)
0.2 (2 sites)
0.2 (3 sites)
0.5 (6 sites)
1.0 (2 sites)
0.2 (4 sites)
1 .52 (6 sites)
Treatment
rate
(kg a.i./ha)
22
22
22
22/45
33
22
45
45
45
45
22
45
45
45
45
45
45
45
45
ce or z,4-u acia
date
No. sampling
stations &
(no. samples)
3(17)
3(17)
2(69)
2(73)
2(68)
3(42)
3(36)
1 (7)
1 (7)
1 (14)
1 (18)
1 (19)
2 (92)
2(40)
3 (30)
4 (190)
2(36)
4(55)
3(42)
in treatment a
Maximum2
2,4-D Cone.
0.140
0.050
0.360
3.250
0.670
0.013
0.011
ND
ND
0.012
0.020
0.029
0.030
0.006
0.007
0.023
0.035
0.016
0.31
reas ton
>0.1
2
ND
5
2
2
ND
ND
—
—
ND
ND
ND
ND
ND
ND
ND
ND
ND
4-7
owing nero
Persistence
>0.05
2
5
2
2
ND
ND
—
—
ND
ND
ND
ND
ND
ND
ND
ND
ND
11-14
ciae appi
ications, nacb-iasi
in days3 per cone, in mg/l
>0.01
3-8
9
3
8
1
1
—
—
8-11
8-11
6
8-13"
ND
ND
4-65
5
<1
14-21
•>• 0.005 > 0.001 Comments
3-8 3-8
2 3-8
9 9 :
5 5
8 8 .
1 10
3-7 3-7
— Treatments separated by 50 m
— Treatments about 400 m apart
— Narrow band of weeds treated
along 120 m shoreline; first sampled
1 day after treatment
— First sampled 1 day after treatment
— First sampled 3 days after treatment;
suspect last station affected by drift
from 5.9 ha treatment 450 m to
south on 8/6/79
— First sampled 3 days after treatment;
suspect stations affected by drift
from 12 ha treatment 80 m away on
4/6/79
.ake Restoration, Protection and
Managemeni
43-45 43-45 — sites 100 m apart; first sampled
15-19 28
8-12 12
17 32
5 days after treatment
— sites 25 m apart; first sampled 12
days after treatment
— sites 300 m apart; first sampled 7
days after treatment
— 6 sites along 1500 m shoreline
7 >136 — sites 150 m apart
1 14
— 4 sites along 1100 m shoreline
14-21 25-28 — 6 sites along 3.6 km shoreline; first
sampled 4 days after treatment;
suspect stations affected by drift
from 8.5 ha treatment at N. end
Wood Lake on 27/9/80
-------
Table 3 (Continued)
Location
B. Treatments larger
N. Arm Okanagan
Lake
N. End Skaha
Lake
Naswhrto Ck.,
Okanagan Lake
Summer land,
Okanagan Lake
Ketowna,
Okanagan Lake
S. End Wood Lake
N. End Kalamalka
Lake
Application
date
than 1 ha in
18/9/78
23/6/77
27/6/77
14/8/78
22/6/78
08/6/79
14/6/79
C. Treatments in enclosed water
Laboratory
Experiment
18/3/76
Treatment
size (ha)1
open water, listed
2.0
2.4
3.1
4.0
5.3
5.9
11.5
Treatment
rate
(kg a.i./ha)
by treatment
45
22/45
45
45
45
22/45
22/45
No. sampling
stations &
(no. samples)
size
3(80)
6(97)
6 (156)
3(73)
3(39)
4(65)
3(72)
Maximum2 Persistence in days3 per cone, in mg/l
2,4-D Cone. 0.1
0.194 4-7
0.140 1
0.233 4
0.197 <1
0.169 1
0.027 ND
0.038 ND
0.05 0.01 0.005
4-7 9-14 17-22
344
666
<1 11-14 15
1 3 10-14
ND 21-24 41 -467
ND 20 25
0.001
17-22
5
8
17
10-14
107
25
Comments
— First sampled 1 day after treatment
— First sampled 9 days after treatment
— First sampled 5 days after treatment
bodies, listed by treatment date
0.9 m2
22
3(144)
0.32 84-166
84-166 84-166 84-166
84-166
— In this laboratory experiment, 2,4-D
was applied to Myriophyllum
Westwide Cays, 12/7/77 2.4
Okanagan Lake
11/22/33 4(199)
Walshe's Pond,
Kalamalka Lake
S. End Wood Lake
01/8/79
0.20
10/9/80 10.0
N. End Wood Lake 27/9/80 8.5
Champion Lake 11/9/81 0.56 (4 sites)
22
45
45
45
1 (9)
4(116)
4(45)
2(32)
1.22
0.62
0.015
0.288
61-;
1-4
ND
<1
spicatum growing in a Plexiglas
cylinder measuring 2.1 m high by 1.1
m in diameter.
4.00 43-59 43-59 43-59 43-59 43-59 — All but one sample was below 0.1
mg/l on day 43; all samples 0.001
mg/l on day 59 as a result of Cays
opened to mixing with Okanagan
Lake on day 45.
61-77 =>77 >-77 >-77 — First sampled 12 days after
treatment.
1-4 43-47 47-50 50-55 —
ND 26-30 30-33 30-38
<1 45-268 45-268 45-268
Cone, increased after 27/9/80 treat-
ment at N. End Wood Lake; cone, in
both areas was not detectable after
fall turnover of lake water column.
First sampled 5 days after treatment.
Not sampled after day 45 until
21/6/82 when no herbicide was
detected (<0.001).
-------
Table 3 (Continued)
I
I
3
C. Treatments in enclosed water bodies, listed by treatment date (continued)
8
en
Location
Solana Bay,
Osoyoos Lake
Treatment No. sampling
Application Treatment rate stations & Maximum2
date size (ha)1 (kg a. Una) (no. samples) 2,4-D Cone.
16/9/81 6.14 44 2(17) 2.06
Persistence in days2 per cone, in mg/l
0.1 0.05 0.01 0.005 0.001 Comments
49-190 :>190 =»190 >190 :>190 — Solana Bay was closed off from
Osoyoos Lake by an earth barrier
prior ta.treatment. Herbicide cone.
declined over winter and were below
0.1 mg/l on 25/3/82; Bay was open-
ed to Osoyoos Lake in May, 1982,
and sampling in June showed no
2,4-D residues.
ND = not detected at this concentration
1 Where several adjacent sites were treated, total area of all sites combined is given.
2 Treatment areas were first sampled on treatment day (approx. 6 hrs. after treatment) and 1 day after treatment unless noted otherwise in comments. Maximum 2,4-D cone, are not
comparable for sites first sampled 2 days or more after treatment as the highest cone, often occurred on the first day following treatment.
3 Persistence is given as the last day when herbicide in water samples at any station was greater than specified concentrations. All samples taken on the next day were below the
specified concentrations. Persistence is shown as a range where time between sample dates 1 day.
4 A sample with 2,4-D cone, of 0.012 mg/l collected on day 29 at Cosens Bay site 11/7/79 was anomalous.
5 A sample with 2,4-D cone, of 0.016 mg/l collected on day 17 at Cosens Bay site 26/6/80 was anomalous.
6 2,4-D may have persisted a few days longer at concentrations 0.005 mg/l as no samples were taken .after day 13 following the 9/9/80 Coldstream Ck. treatment.
7 A sample with 2,4-D cone, of 0.007 mg/I collected on day 96 at the S. end Wood Lake site 8/6/79 was anomalous.
(D
3
-------
Physical, Chemical, and Biological Control of Aquatic Macrophytes
0.005 mg/l for 19 to 30 days after treatments.
Following several experimental treatments in Wood and
Kalamalka Lakes, the concentrations of BEE and acid forms
of 2,4-D as well as 2,4-dichlorophenol were analyzed
separately in water samples. At the pH values of 8.2 to 8.5
and temperatures of 15 to 20°C in surface waters of these
lakes, 2,4-D BEE was rapidly converted to 2,4-D acid and
was barely detectable in water samples 48 hours after ap-
plication. 2,4-dichlorophenol is considered less stable than
2,4-D acid and was detectable in samples taken no later than
24 hours after applications. At lower water temperatures
and/or lower pH, these compounds can be expected to per-
sist for longer periods (Bothwell and Daley).
Herbicide Drift
Herbicide drift from treated areas must be considered where
environmentally sensitive areas are being treated or where
there may be conflicts with various water users in the treat-
ment vicinity. In theory, the potential drift distance increases
with the quantity of herbicide used. However, the actual drift
distances appear to depend largely upon exposure of the
treatment site to water movement. In lakes, currents are
generated principally by prevailing winds.
Observations on herbicide drift from treatment areas were
made by Water Management Branch staff following 21 ap-
plications during 1976-1981. Data are summarized in Table
4. Little herbicide drift was observed outside treatment areas
less than 1.0 ha in size (i.e., spot treatments). No 2,4-D acid
was detected at concentrations over 0.1 mg/l in water
samples taken at distances greater than 70 m from spot
treatments. Concentrations around the detection limit (0.001
mg/l) were measured at a maximum distance of 460 m from
spot treatments, but usually were not detected over 100 m
away.
Drift samples from small and large (up to 12 ha) treatments
did not contain concentrations greater than 0.1 mg/l, with
one exception. One sample, taken at a distance of 123 m
from the 5.3 ha Kelowna foreshore treatment (1979), con-
tained a concentration of 0.131 mg/l. In the many hundreds
of other drift samples taken adjacent to large and small
treatments, 2,4-D concentrations remained below 0.1 mg/l.
In most treatments greater than 4 ha in size, concentra-
tions of 2,4-D acid over 0.01 mg/l were detected at distances
greater than 500 m from the treated area, but within 4 days
of treatment, concentrations had declined to less than 0.01
mg/l at these distances. Thus, drift of the herbicide at con-
centrations over 0.01 mg/l was shortlived. The drift distance
reached by concentrations at the limit of detection (0.001
mg/l) was over 2,000 m after some large treatments. Such
drift occurred within a few days of these treatments and no
herbicide was found in samples taken 1,000 m or more
beyond the treatment sites after 5 days.
In cases where treatments are sufficiently large that the
herbicide is dispersed at detectable concentrations
throughout an entire lake, the drift persistence is likely to
be roughly the same as the persistence of herbicide in the
treatment area. For example, treatment site and herbicide
drift residues throughout the epilimnion of Wood Lake (930
ha) were recorded for a period of 30 days, following the 10.0
ha and 8.5 ha herbicide treatments in September, 1980.
Monitoring was also conducted to examine the vertical
distribution of herbicide. No 2,4-D was detected in the
hypolimnion below surface water that contained relatively
high 2,4-D concentrations. It is unlikely that herbicide drift
from a surface application will penetrate a well-developed
thermocline. In practical terms, this means that deep water
intakes that draw water from below the thermocline are
naturally protected from contamination during summer her-
bicide treatments. In a 1979 study of herbicide treatments
in Wood and Kalamalka Lakes, Bothwell and Daley (1981)
investigated the presence and transport of 2,4-D residues
in the lake surface microlayer. Low concentrations of her-
bicide in the microlayer (<0.001 mg/l) traveled considerable
distances in a short period of time, but in terms of the total
amount of herbicide movement away from treatment sites,
herbicide drift in the surface microlayer was considered to
be insignificant.
Herbicide Persistence in Sediments
and Movement in Groundwater
As root mortality is the primary objective of herbicide applica-
tions, residue penetration into the bottom sediment is con-
sidered to be critical to herbicide uptake. In treatments car-
ried out by Water Management Branch staff, 2,4-D acid was
found in sediment samples only in the treated areas where
Aqua-Kleen pellets landed. Herbicide concentrations in the
sediment varied considerably between sampling locations
and dates. Maximum concentrations of 2,4-D acid in the sedi-
ment occurred between 2 and 11 days after treatment and
ranged between 5.33 and 288.0 ppm, although mean con-
centrations were considerably lower. This high variability was
thought to be caused by the small volume of the sediment
corer (250 cm3), the granular nature of the herbicide pellets,
and their uneven dispersion. Herbicide persistence in sedi-
ments ranged from approximately 1 to 11 months. Greater
persistence was associated with the larger treatments and
higher rates of application.
A separate study was conducted to investigate the mobility
of 2,4-D and water movement in lake bottom sediments.
Analysis of sediment pore water (interstitial water) and sedi-
ment remaining after removal of the pore water indicated
that little 2,4-D was bound to the sediment; the detectable
2,4-D was in the interstitial water (Kangasniemi and Nagpal,
in prep.). In view of the potential mobility of 2,4-D in the sedi-
ment, concern has been raised about the possibility of 2,4-D
contamination of groundwaters feeding domestic or irriga-
tion water sources such as a well or sandpoint. Groundwater
could flow from a lake to recharge adjacent wells, particularly
downslope of lakes, such as at their outlet. Also, lake water
may enter groundwater where the water table is depressed
below the lake level because of continuous pumping of a
well. Each treatment area and any adjacent wells should be
assessed on a site-specific basis to determine if (1) a
hydraulic gradient exists that will move groundwater from
a treatment area towards an active well; (2) there is hydraulic
continuity; (3) the flow velocity is adequate to move the her-
bicide residues to wells before degradation occurs; and (4)
dilution is not sufficient to reduce 2,4-D concentrations below
water quality objectives. If these conditions exist, well use
should be discontinued while residues persist above objec-
tive concentrations, taking into consideration the period it
would take for groundwater to travel from the treated lake
bottom to the well.
Groundwater movement also may play a role in reducing
the effectiveness of 2,4-D applications by rapidly diluting her-
bicide concentrations in the sediment. If there was sufficient
groundwater movement from the water table to the lake (i.e.,
upwelling), residues may be flushed from the sediment.
Following treatments in exposed locations, there was a
concern that 2,4-D in the sediment could be re-released in-
to the water column sometime after lake water residues had
declined (i.e., from disturbance of the sediment by wave ac-
tion in large treatment areas). However, concentrations of
such herbicide in the water column were observed to be low.
After treatment of 5.3 ha along the Kelowna foreshore, the
herbicide was detectable for 5 days in the treatment area,
then was nondetectable when sampled on days 6 and 8.
Subsequently, concentrations of 0.001 to 0.007 mg/l were
detected in bottom water samples on day 10, probably as
207
-------
Table 4. — Drift of 2,4-D acid from herbicide treatments, 1976-1981 (Drift estimated from analyses of water samples
collected in sample grid in direction of anticipated water movement from treatment areas).
A. Treatments smaller than 1 ha in open water,
Location
N. Arm
Okanagan Lake
N. Arm
Okanagan Lake
8
00
Lisheen Estates,
Kalamalka Lake
Wilson's Landing,
Okanagan Lake
Cosens Bay,
Kalamalka Lake
Coldstream Ck.
Kalamalka Lake
Lisheen Estates
Kalamalka Lake
S. End
Kalamalka Lake
Application
date
31/5/76
31/5/76
12/7/77
12/7/77
12/7/77
17/7/78
01/8/78
26/6/80
09/9/80
30/7/80
25/9/80
Treatment
size
(ha)1
0.60
0.60
0.67
0.36
0.30
0.07
0.10
0.50
(6 sites)
1.0
(2 sites)
0.2
(4 sites)
1.52
(6 sites)
listed by treatment date
Treatment
rate
(kg a.i./ha)
22
22
22
22/45
33
22
45
45
45
45
45
No. sampling
stations &
(no. samples)
7(29)
8 (38)
6(94)
7(70)
5 (50)
15 (90)
21 (160)
2 (23)
4(72)
5(64)
6(45)
Drift at cone.
0.05 mg/l
Max. Days2
Distance (m) >50 m
ND —
ND —
ND —
ND —
27 —
ND —
ND —
ND —
ND —
ND —
ND —
Drift at cone.
> 0.01 mg/l
Max. Days2
Distance (m) ?»50 m
25 —
30 —
67 3
15 —
27 —
ND —
ND —
ND —
ND —
ND —
ND —
Drift at cone.
> 0.001 mg/l
Max. Days2
Distance (m) > 50 m
25 —
90 3-8
67 3
15 —
27 —
ND —
ND —
460 20
425 2
100 7
1100 4-7
Comments
— Stations 10-25 m from site3;
treatments 50 m apart
— Stations 25-150 m from site;
treatments 50 m apart
— Stations 1 5-88 m from site3
— Stations 15-64 m from site3;
detected last on day 2
— Stations 15-46 m from site3;
detected last on day 2
— Stations 80-360 m from site
— Stations 60-400 m from sites
— Stations at 90 and 460 m
from sites3
— Stations 190, 425 and 610 m
from sites3
— Stations 100-600 m from
sites
— Stations 1100-2100 m from
sites; suspect drift through
Oyama Canal from 8.5 ha
treatment at N. end Wood
Lake on 27/9/80
-------
Table 4 (Continued)
B, Treatment smaller than 1 ha in open water, listed by treatment size,
Drift at cone.
Treatment Treatment No. sampling 0.05 mg/1
Application size rate stations & Max. Days2
Drift at cone.
0.01 mg/l
TBIZDays2 Days2"
Drift at cone.
0.001 mg/l
Max.
Days2 Days2
Location date (ha)1 (kg a.i./ha) (no. samples) Distance (m) >-5QO m Distance (m) >100Q m >»500 rn Distance (m) >1000 m >500 m
N, Arm
Okanagan Lake
N. End
Skaha Lake
18/9/78
23/6/77
2.0
2,4
45
22/45
29 (409)
5(40)
ND
ND
Naswhito Ck., 27/6/77 3.1 45
Okanagan Lake
6 (72) ND
40
140
40
Summerland, 14/8/78 4.0 45 37 (593) 1500 <1 1500
Okanagan Lake
Kelowna, 22/6/78 5.3 45 55 (799) 400
Okanagan Lake
S. End
Wood Lake
08/6/79
5.9 22/45
15 (78)
N. End 14/6/79 11.5 22/45 9(108)
Kalamalka Lake
ND
ND
830
2000
1000
400"
140
40
1500
1450
2000
1 NS 1000
11 —
4 _
3 —
NS —
Comments
Stations 30-2500 m from
site; last detected at 400
m 17 days after treatment
Stations 10-140 m from
site3; 2,4-D was detected
in only one drift station
sample taken at 140 rn
with 0.017 mg/l on day 1
Stations 18-180 m from
site; 2,4-D detected in
drift for 3 days only
Stations 200-3500 m
from site; max. eonc. at
1500 m = 0.067 mg/l; sta-
tions beyond 1500 m add-
ed after day 13
Stations 60-5850 m
from site; max cone, at
400 m = 0,063 mg/l
Stations 500-7300 m from
site; stations beyond 2000
m added after day 43
Stations 1000-4200 m
from site
-------
C. Treatments in areas where dispersion restricted by lake size.
Table 4 (Continued)
Drift at cone.
0.05 mg/l
Location
Treatment Treatment No. sampling
Application size rate stations & Max. Days2 Max.
date (ha)1 (kg a.i./ha) (no. samples) Distance (m) >500 m Distance (m)
Drift at cone.
0.01 mg/l
Drift at cone.
0.001 mg/l
Days2 Days2 Max.
>1000 m >500 m Distance (m)
Days2 Days2
••1000 m >500 m
Comments
r\j
o
S. End
Wood Lake
N. End
Wood Lake
10/9/80
27/9/80
10.0
8.5
45
45
23 (142)
5(86)
ND —
ND —
ND
ND
Champion Lake 11/9/81
0.56
(4 sites)
45
2(20)
ND —
470
6000 :>14 >14 — Stations 500-6000 m
from sites6; most of the
930 ha lake epilimnion
had 2,4-D cone, of
6000 30-33 30-33 0.001-0.003 mg/l just
before 8.5 ha treatment
at north end on
27/9/80 which in-
creased epilimnion
cone, at all drift sta-
tions to max. of 0.0095
mg/l until lake turnover
about 28/10/80
470 — — — Stations about 270 and
470 m from major
treatment sites; much
of the 14.6 ha lake
surface had cone, of
about 0.017 mg/l 20
days after treatment.
No samples were
taken after day 45 until
21/6/82, when no her-
bicide was detected
(:>0.001 mg/l)
Note: Herbicide residue samples taken closer than 10 m to treatment site not considered
ND = not detected; NS = not sampled
1 Where several sites were treated, total area of all sites combined is given.
2 Persistence is given as the last day when herbicide concentrations in water samples were greater than specified concentrations. All samples taken on the
next day were below the specified concentrations. Persistence is shown as a range where time between sample dates>1 day.
3 At these treatment sites, an increase in the size of the sample station grid wrthin<1 day to 2 days after treatment may have resulted in detection of
herbicide further than the maximum recorded.
4 A sample with 2,4-D cone, of 0.003 mg/l collected on day 3, 2500 m from North Arm Okanagan Lake treatment site 18/9A78, was anomalous.
5 Drift stations were located along entire length of Wood Lake (6000 m) for treatments on 10/9/80 and 27/9/80.
-------
Physical, (jnemical, ana tsioiogicai oomroi or Hquauc iviaurupnyies
a result of wind on the previous day. This recurrence of the
herbicide in the treatment area water column was followed
by measurement of low levels (0.002 to 0.004 mg/l) at drift
stations up to 200 m from the treatment boundary on day
10 and 14 after 5 days of nondetection.
Alternate Water Supply and Buffer Zones
Water use surveys of the area likely to be affected by 2,4-D
drift should be conducted prior to establishing buffer zones
around the application area. Door to door surveys may be
required to identify and map both licensed and unlicensed
lake and well water use. Buffer zone size should be based
on the predicted drift of the herbicide and the maximum ac-
ceptable concentration for the particular water use in the
zone of herbicide influence. Where herbicide concentrations
might exceed these water quality standards, water use (in-
cluding domestic use, irrigation or swimming) should be
discontinued after herbicide application until sufficient time
has passed, or until monitoring indicates that residue levels
are no longer a concern.
Table 5 lists recommended buffer zone distances between
treatments and lake water use, based on the observed drift
of the herbicide in British Columbia lakes. A general zone
between lakeshores and wells, where water use should be
discontinued after herbicide treatments, cannot be specified.
The source of well water and hydraulic gradients should be
assessed where wells adjoin a proposed treatment area and
buffer zones should be established accordingly.
The buffer zone distances in Table 5 are estimates of
the maximum drift that would be anticipated under typical
treatment conditions in large British Columbia lake systems.
The most information is available on small treatments; only
seven treatments over 1 ha in size (2 to 12 ha) were
monitored for drift. A considerable distance was incorporated
in the buffer zones to allow for uncertainty, particularly where
limited data were available. The buffer zone distances app-
ly in the direction of possible water movement and could be
reduced in directions where it can be shown that little move-
ment is possible. The time during which the buffer zones
should be maintained might be reduced if monitoring in-
dicates a more rapid dispersion of herbicide than indicated
in Table 5 (which is likely following most treatments). Buffer
zones might have to be increased under unusual conditions
such as rapid movement of treated water in a confined
channel.
The provision of alternate water to those users who must
discontinue water use within buffer zones and treatment
areas may increase the cost of herbicide treatments to such
an extent that this aquatic weed control method is no longer
cost effective. A preliminary study of alternate water re-
quirements and costs should be made to determine the
feasibility of such treatments even before an application is
made for a Pesticide Use Permit. In some cases, the timing
of the proposed treatment may make it possible to avoid the
need for extensive alternate supplies. Alternate water sup-
plies can be provided in a number of ways, ranging from
local residents supplying their own needs to municipal in-
stallation of new pipelines to neighboring utilities. These ar-
rangements are site specific. During trial herbicide treatments
from 1976 to 1981, the costs associated with supply of water
for single large treatments (over 3 ha) ranged from under
$10,000 to a maximum of approximately $47,000.
Effects on Nontarget Organisms
Other macrophytes: Qualitative and quantitative observa-
tions in treated areas indicated nontarget plants were not
eliminated at the application rates used to control Eurasian
water milfoil populations (Anon. 1980c). After treatment, plant
species diversity was never reduced and cover of some non-
target species (Elodea canadensis Michx., Potamogeton
crispus L. and Potamogeton pectinatus L.) usually in-
creased. In two areas these species grew to nuisance levels
one year after treatment. The effectiveness of herbicide
treatments can be shortened in some areas by regrowth of
herbicide tolerant native species.
Pelagic organisms: A study by the International Pacific
Salmon Fisheries Commission indicated that the 96-hour
LC50 of 2,4-D BEE was approximately 0.45 mg/l for sockeye
Table 5. — Suggested buffer zones and water use closure periods for 2,4-D treatments of
Eurasian water milfoil in British Columbia
Open lake treatments1
Water quality objective
Treatment size
0.1 mg/l
0.01 mg/l
0.001 mg/l
1 ha
1 - 3 ha
3 - 5 ha
100 m (M5 days)
200 m (~6 days)
300 m (~6 days)
300 m (~20 days)
1500 m for 4 days
then 500 m (~30 days)
2500 m for 4 days
then 1000 m (^40 days)
600 m (>20 days)
2000 m for 5 days
then 800 m {^50 days)
3500 m for 5 days
then 1000 m (~50 days)
Enclosed lake pond treatments2
Water quality objective
0.01 mg/l
0.01 mg/l
0.001 mg/l
"100 days
190 days
190 days
1 Closure period in days is the longest time period that herbicide concentrations would remain above water quality objectives in mosl situations. Monitoring should verity that shorter
closure periods are required in most sites
3 Closure periods listed are lor water bodies where entire surface area was treated and would be shorter for sites where only a portion of the water body is treated Buffer zones
would be the entire lake or pond where complete dispersion of the herbicide results in a concentration greater than the water quality objective throughout the water body surface
waters BuHer zones would be smaller where dilution reduced 2,4-D concentrations below water quality objectives. Monrtonng may be required to determine closure penod.
211
-------
Lake Restoration, Protection and Management
salmon fingerlings, coho salmon fry, and pink salmon fry
(Martens et al. 1980). However, no mortality or distress was
observed among sockeye fingerlings exposed to 200 mg/l
2,4-D acid for 168 hours.
Water samples were collected by Water Management
Branch staff to observe the rate of hydrolysis of 2,4-D BEE
following a large application (10 ha) in Wood Lake in 1980.
The concentrations of 2,4-D BEE sampled in the treated area
were variable, both vertically and horizontally, indicating a
patchy herbicide distribution. Mean surface water concen-
tration 1 to 5 hours after treatment was 0.20 mg/l and mean
bottom water concentration was 0.38 mg/l. After 24 hours,
concentrations of 2,4-D BEE declined to mean levels of 0.012
mg/l at the surface and 0.016 mg/l at the bottom. The data
from Wood Lake and theoretical calculations of Bothwell and
Daley (1981) suggest that fish mortality caused by BEE in
treated areas is not likely to be significant at the rate of
hydrolysis predicted for Okanagan Valley lakes. At lower
temperature and pH, hydrolysis will be slower and 2,4-D BEE
may persist longer in treatment areas. Such persistence may
be a concern if treatments are proposed for large areas fre-
quented by valuable fish species.
In practice, no fish kills have resulted from any of the 2,4-D
treatments during trials in the Okanagan Valley lakes from
1976 to 1981. The treatment in Westside Cays, monitored
by the B.C. Fish and Wildlife Branch, was of particular
significance. In this enclosed site, where the entire surface
area was treated, concentrations of 2,4-D acid averaged
0.536 mg/l for 10 days after treatment. Fish could not escape
from the treatment area. No fish kill was observed following
treatment and samples of those fish present (squawfish,
shiners, largescale suckers, carp, longnose suckers, perch,
sculpins, and peamouth chub) did not contain detectable
2,4-D when collected 50 to 71 days after treatment (Robin-
son and Morley, 1980).
Additional studies were conducted by the Fish and Wildlife
Branch in 1978 to evaluate the effect of 2,4-D treatments
on fish (South and Robinson, in prep.). The data suggest
that some fish captured in or near treated areas might con-
tain detectable levels of the herbicide. Three fish (mountain
whitefish, largescale sucker, and peamouth chub) out of 10
netted within the treated area 3 to 4 days after the 5.3 ha
treatment on the Kelowna foreshore in 1978 had concen-
trations of 2,4-D between 0.10 and 0.72 ppm. Only the sucker
contained 2,4-D in muscle tissue, the other fish had 2,4-D
in gills, skin, liver, and/or alimentary tracts. Whole fish
samples of mixed fish species including redside shiners,
peamouth chub, prickly sculpins, northern squawfish, and
longnose and largescale suckers, netted from the treatment
area prior to application and held in cages, contained 2,4-D
levels up to 1.48 ppm when collected about 3 weeks after
herbicide application. Only the mountain whitefish is an-
gled for consumption. Most gamefish remain in cold water
below the thermocline in summer and as a result, will not
contact any herbicide.
It is unlikely that fish containing 2,4-D as a result of aquatic
weed treatments would be taken for human consumption,
although fish in shallow lakes subject to large treatments (>1
ha) might contain low 2,4-D concentrations for short time
periods. There is no indication that consumption of such fish
would risk exceeding the maximum acceptable daily intake
values for 2,4-D established by the World Health
Organization.
Zooplankton, which are important in the diet of some fish
species, can be affected by herbicide applications through
direct toxicity (Walker, 1961), or secondary effects such as
habitat loss (Brooker and Edwards, 1975). Data collected by
the Fish and Wildlife Branch at several treatment sites in
1977 indicate a reduction in zooplankton species diversity.
They concluded that the probable long-term effect of a large-
scale reduction of M. spicatum would be a corresponding
reduction in associated zooplankton populations through
habitat loss. Limnetic species and littoral zooplankton species
not associated with M. spicatum would probably be unaf-
fected (Robinson and Morley, 1980).
Benthic organisms: Brooker and Edwards (1975) stated
that the status of most benthic species does not change as
a result of herbicide use, but they also reported studies in
which benthic invertebrate density increased after treatment.
Smith and Isom (1967) concluded from their studies involv-
ing granular 2,4-D BEE that benthic populations were not
depressed by the herbicide. In a study involving the same
herbicide, Pierce (1961) observed that in no benthic in-
vertebrate samples did any one group show significant varia-
tion from the seasonal fluctuation exhibited by the control.
The study conducted by Robinson and Morley (1980) in
three control areas and three treated areas (0.3 to 0.7 ha
in size) in the North Arm of Okanagan Lake indicated that
the predominant short-term effect of the treatments was to
temporarily increase the abundance and decrease the diver-
sity of benthic invertebrates. Increased abundance may
reflect opportunities for detritus feeders, while decreased
diversity may have resulted from toxicity or loss of habitat.
Water Management Branch data indicate that 2,4-D levels
in sediments were higher and persisted longer than in the
water column. Sample concentrations were highly variable
and indicated the herbicide had a somewhat uneven distribu-
tion in sediments. However, it appears that 2,4-D could per-
sist long enough and at levels that would produce toxic ef-
fects to a variety of benthic organisms if they were exposed
to these sediments. In studies by Water Management Branch
and Fish and Wildlife Branch, clams appeared relatively
tolerant as there was no evidence of mortality in treated
areas. In theory, toxic effects could occur to various am-
phipods and insect larvae living in or on the sediments, as
determined from bioassays. However, changes in benthic
communities were short-lived following the small treatments
(<1.0 ha) monitored in Okanagan Lake. In larger treatments,
such changes in benthic populations might be a concern if
the treated areas supported a significant food source for
valued fish species.
PROGRAM IMPLEMENTATION
Laboratory and field testing of 2,4-D (Aqua-Kleen) has in-
dicated that this herbicide is one of several effective methods
for controlling nuisance Eurasian water milfoil in a number
of British Columbia lake systems. Advantages of this her-
bicide are that in some sites it can be applied relatively quick-
ly over large areas, at a low cost without risk of plant
fragmentation or encouragement of further spread. Also,
relief from nuisance aquatic plant populations may persist
for 1 to 3 years. Review of the possible use of 2,4-D in a
specific location can be approached logically on a step by
step basis as follows:
1. Review physical characteristics of proposed treatment
area and compare basic cost and overall effectiveness of
alternate control methods. 2,4-D is particularly effective in
enclosed waterbodies or embayments where there is little
water movement and where reinfestation from untreated
populations is likely to be minimal. It is also important to
determine whether native species, which may be unaffected
by 2,4-D, will proliferate and produce a nuisance weed pro-*
blem after Eurasian water milfoil has been selectively
controlled.
2. Identify aquatic resources that may be affected by 2,4-D
in the treatment area and determine whether adverse en-
vironmental effects can be minimized by optimum timing and
elimination of sensitive areas from treatment. No significant
212
-------
Physical, Chemical, and Biological Control of Aquatic Macrophytes
impacts of 2,4-D treatment on fish, waterfowl or other wildlife
were observed during tests in the Okanagan lakes system.
Agencies responsible for aquatic resources should be con-
tacted about new treatment sites.
3. Determine water use in the treatment area and areas
which may be exposed to herbicide drift. Both water license
information and door to door surveys may be required in high
use areas.
4. Establish the standards for maximum permissible levels
of herbicide in water for various use purposes. Local health
officers or Guidelines for Canadian Drinking Water Quality
should be consulted.
5. Establish the size of the buffer zone required between
the treatment area and active water intakes or other water
use in lake. This distance should be based on water quality
standards and potential herbicide drift.
6. Determine cost of supplying alternate water to those
users who must be disconnected in the treatment area and
buffer zone.
7. Apply for herbicide use permit.
It is the experience of the Water Management Branch that
the need for nuisance weed control and selection of a her-
bicide as the most cost-effective control method must be sup-
ported by the public adjacent to the problem area. Unfor-
tunately, the process to determine whether herbicide is ap-
propriate, particularly for large treatments in open lake situa-
tions, can be time-consuming and requires some technical
expertise. Of course, part of the cost of herbicide treatment
must include the research required in selecting the control
method for a particular situation. It is anticipated that govern-
ment agencies at the provincial level could assist with such
investigations.
REFERENCES
Aly, O.M., and S.D. Faust. 1964. Studies on the fate of 2,4-D and
ester derivatives in natural surface water. J. Agric. Food Chem.
12:541-546.
Anonymous. 1980a. Studies on aquatic macrophytes. Part XIX.
Eurasian water milfoil treatments with 2,4-D in the Okanagan
Valley, 1977-78. Vol. 1: Description of study area, treatment
methods and herbicide residue persistence in the water column
and water quality implications. Inv. Eng. Br. Rep. No. 2913, British
Columbia Ministry Environ.
1980b. Studies on aquatic macrophytes. Part XIX. Eura-
sian water milfoil treatments with 2,4-D in the Okanagan Valley,
1977-78. Vol. 2: Herbicide residue concentrations and persistence
in biological tissues and hydrosoils. Inv. Eng. Br. Rep. No. 2914,
British Columbia Ministry Environ.
1980c. Studies on aquatic macrophytes. Part XIX. Eurasian
water milfoil treatments with 2,4-D in the Okanagan Valley,
1977-78. Vol. 3: Herbicide application effects on Eurasian water
milfoil. Inv. Eng. Br. Rep. No. 2815, British Columbia Ministry
Environ.
Bothwell, M.L. and R.J. Daley. 1981. Selected observations on
the persistence, and transport of residues from Aqua Kleen 20
(2,4-D) treatments in Wood and Kalamalka Lakes, B.C. Environ.
Can. Inland Waters Directorate, Pacific and Yukon Region, Van-
couver, B.C..
Brooker, M.P., and R.W. Edwards. 1975. Aquatic herbicides and
control of water weeds. Weed Res. 9:1-15.
Goddard, J.M. 1980. Studies on aquatic macrophytes. Part XXX.
Control of Myriophyllum spicatum in Kalamalka and Wood Lakes
using 2,4-D butoxyethanol ester in 1979. Vol. I: Data Rep. Inv.
Eng. Br. Rep. No. 2824, British Columbia Ministry Environ.
Health and Welfare Canada. 1978. Guidelines for Canadian Drink-
ing Water Quality. Canadian Gov. Publ. Centre, Hull, Quebec.
Kangasniemi, B.J., and N. Nagpal. In prep. A study of the potential
for groundwater contamination from the use of 2,4-D for aquatic
weed control. Water Manage. Br. British Columbia Ministry
Environ.
Lim, P.O. 1978. Studies on aquatic macrophytes. Part XVI. A
laboratory experiment with granular 2,4-D for control of Eurasian
water milfoil, 1976-77. Water Invest. Br. Rep. No. 2726, British
Columbia Ministry Environ.
Lim, P.G., and K.R. Lozoway. 1977. Studies on aquatic macrophytes.
Part X. A field experiment with granular 2,4-D for control of Eura-
sian water milfoil, 1976. Water Invest. Br. Rep. No. 2613, British
Columbia Ministry Environ.
Martens, D.W., R.W. Gordon and J.A. Servizi. 1980. Toxicity of
butoxyethanol ester of 2,4-D to selected salmon and trout. Int.
Pacific Salmon Fish. Comm. Prog. Rep. No. 40.
Newroth, P.R. 1979. B.C. aquatic plant management
program. J. Aquat. Plant Manage. 17:12-19.
1980. Case studies of aquatic plant management for lake
preservation and restoration in British Columbia, Canada. Pages
146-152 in Int. Symp. Inland Waters and Lake Restoration, Sept.
8-12,1980. Portland, Maine. EPA 440/5-81-010. U.S. Environ. Prot.
Agency, Washington, D.C.
Pierce, M.E. 1961. A study of the effect of weed killer 2,4-D Aqua
granular on six experimental plots of Long Island, Dutchess Coun-
ty, N.Y. Proc. N.E. Weed Conf. 15:539-544.
Robinson, M.C. 1981. Eurasian water milfoil studies. Vol. I. The
effect of Eurasian water milfoil (Myriophyllum spicatum L.) on fish
and waterfowl in the Okanagan Valley, 1977. Rep. prep, for Aquat.
Stud. Br. by Rsh Wildl. Br. British Columbia Ministry Environ.
Assess. Plann. Div. Bull. No. 15.
Robinson, M.C., and R.L Morley. 1980. Eurasian water milfoil
studies. Vol. II. A monitoring study of the effects of 2,4-D on fish
and waterfowl as applied in lakes of the Okanagan Valley, 1977.
Rep. prep, for Inv. Eng. Br. by Fish Wildl. Br. Ministry Environ.
B.C. Inv. Eng. Br. Rep. No. 2918.
Rudolph, J.R., and C.E.W. Dyer. 1981. Studies on Aquatic Macro-
phytes. Part XXXI. Control of Myriophyllum spicatum in Kalamalka
and Wood Lakes using 2,4-D butoxyethanol ester in 1980: Data
Rep. Assess. Plann. Div. Bull. 4, British Columbia Ministry Environ.
Schultz, D.P., and P.O. Harman. 1974. Residues of 2,4-D in pond
waters, mud and fish, 1971. Pestic. Monitor. J. 8:173-179.
Smith, G.E., and B.G. Isom. 1967. Investigations of effects of large-
scale application of 2,4-D on aquatic fauna and water quality.
Pestic. Monitor. J. 1:16-21.
South, E.L., and M.C. Robinson. Eurasian water milfoil studies. Vol.
III. Monitoring the effects of 2,4-D on fish and waterfowl and the
effect of Eurasian water milfoil on fish and waterfowl in the
Okanagan Valley, 1978. Rep. prep, for Water Manage. Br. by Rsh
and Wildl. Br., British Columbia Ministry Environ. In prep.
Walker, C.R. 1961. Toxicological effects of several herbicides to
bottom dwelling fish food organisms in Missouri ponds. Proc.
Northcentral Weed Control Conf. 18:104-105.
Wallis, M.A. Studies on aquatic macrophytes. Part XXXVII. Obser-
vations on the control of Myriophyllum spicatum in Kalamalka and
Wood Lakes using 2,4-D butoxyethanol ester in 1979 and 1980.
Water Manage. Br., Ministry Environ. In prep.
213
-------
OBSERVATIONS ON HERBIVOROUS INSECTS THAT FEED ON
MYRIOPHYLLUM SPICATUM IN BRITISH COLUMBIA
BENGT J. KANGASNIEMI
Ministry of Environment
Victoria, British Columbia
ABSTRACT
Field observations from five British Columbia lakes indicate that several aquatic insects are grazing on
the nuisance aquatic plant, Myriophyllum spicatum.Certain species of caddis flies, weevils, and chironomids
have been observed to cause significant plant damage. Caddisfly and weevil damage appears restricted
to a few isolated circumstances. Ongoing field surveys suggest that a chironomid larva belonging to the
genus Cricotopus is causing the most widespread impact. The larva of this nonbiting midge feeds and
pupates in the apical meristem region of M. spicatum.Jhls chironomid has been found at all sites surveyed,
and in certain locations is abundant enough to stop or retard M. spicatum shoots from growing to the
lake surface. Details of the natural history of these insects are presented and their potential as biological
control agents is discussed.
INTRODUCTION
Myriophyllum spicatum L was introduced to Okanagan Lake,
British Columbia, about 1970, and spread rapidly through
all six mainstem lakes of the Okanagan Valley (Fig. 1) dur-
ing the following 7 years, exploiting relatively unvegetated
littoral areas and outcompeting most native aquatic plants.
The growth vigor of these new populations was phenomenal.
In many sites, dense mats of reddish flower spikes appeared
at the surface by June, and remained there until September,
annoying swimmers and boaters. By 1975, M. spicatum oc-
cupied 238 ha of the six Okanagan Valley lakes, and by 1979
this had increased to 674 ha (Hepburn, 1979). As early as
1978, there were indications that M. spicatum had reached
an equilibrium in Okanagan Lake in terms of affected area,
although it continued to spread downstream to Skaha,
Vaseux, and Osoyoos Lakes and reached two upstream
lakes (Wood and Kalamalka) about 1974.
In 1980, it was observed that many M. spicatum popula-
tions failed to reach the surface and flower. To determine
whether the reduced growth was a short-term fluctuation or
an early sign of an overall decline, eight representative study
sites in four lakes were selected for detailed study (Fig. 2).
This study was to document the condition of the selected
plant populations and monitor physical, chemical, and
biological parameters considered critical to aquatic plant
growth. This paper summarizes the observations of her-
bivorous insects associated with M. spicatum derived from
12 surveys of these sites carried out during 1980,1981, and
1982. Observations were also made in Magic Lake, located
on Render Island (Fig. 1), during occasional visits in 1979,
1980, and 1981.
The results and observations presented in this report are
largely qualitative and are based on hundreds of hours of
scuba diving and detailed study of thousands of plants under
dissecting microscopes.
Three groups of herbivorous insects are found associated
with M. spicatum: the caddisfly, Triaenodes tarda; four
species of weevils, and one chironomid, Cricotopus sp.
TRIAENODES TARDA (CADDISFLY)
Caddisflies of the genus Triaenodes are generally regarded
as herbivorous, using aquatic plants for both food and
material for case construction (Wiggins, 1977). Ross (1944)
reports T. tarda as widely distributed through the northeastern
States and southwest through into Oklahoma and Arizona,
and also British Columbia. Nimmo and Scudder (1978) report
four species of Triaenodes for British Columbia, including
T. tarda which was found in both coastal and interior habitats.
T. tarda was first observed grazing on M. spicatum in
Magic Lake in 1979. Subsequent sampling indicated this
species was probably widespread in British Columbia, and
is commonly associated with Myriophyllum species. T. tar-
da has also been observed grazing on M. spicatum
throughout the Okanagan Valley lakes.
An extremely abundant population of T. tarda in Magic
Lake virtually eliminated the standing crop of a 1 ha M.
spicatum population on one occasion in 1979. Only bare
stems and roots remained throughout the entire weed bed.
Subsequent surveys indicated heavy grazing was checking
regrowth. This dramatic impact has not been observed
elsewhere. Vaseux Lake supports an abundant population
of T. tarda that appears to exert some grazing pressure on
M. spicatum (Table 1), but not to the same degree as in
Magic Lake. Caddisfly larvae are considered prime fish food
organisms; the apparent absence of fish in Magic Lake may
explain the remarkable abundance of T. tarda.
In other locations, T. tarda tends to have a minor impact.
Generally, leaflets from any part of the plant are haphazardly
cut, having little effect on the growth of the shoot. Close
observation of T, tarda's feeding behavior in aquaria indicates
that most leaf material is cut and dropped; a relatively small
portion was used as food or for case building. T. tarda also
has been observed using Ceratophyllum demersum and
Elodea canadensis, although when given a choice in an
aquarium, this insect showed a strong preference for
Myriophyllum species such as M. spicatum and M.
verticillatum.
T. tarda constructs an elegant tapered case from living
Myriophyllum leaflets. As the larva metamorphoses through
five instar states it adds to the case in a spiral fashion, grow-
ing to about 10 mm. Unlike many other caddisflies,
Triaenodes are effective swimmers, remaining in their cases
and using their long hairy hind legs as paddles. Swimming
enables the larvae to graze an extensive area up through
the water column. Their swimming action makes the other-
wise well-camouflaged larvae conspicuous to fish.
214
-------
Most caddisflies in temperate latitudes complete one
generation each year (Wiggins, 1977). Some caddisflies use
two overlapping generations per year; one group overwinters
in the last larval instar with the case sealed in preparation
for pupation and emergence in spring; the other group over-
winters in the third or fourth instar and emerges in the sum-
Table 1. — Percentage of randomly selected M. spicatum
shoots grazed (n = 50)
Lake
Okanagan
Okanagan
Okanagan
Okanagan
Skaha
Vaseux
Osoyoos
Osoyoos
Site
Beachcomber
Corners
Newport
Mission
Kaleden
Hatfield
Mica
Stallion
June 1982
741
441
601
441
361
982
243
443
August 1982
901
101
801
601
921
922
1003
963
' Crtcofopus sp. dominant herbivore
2 TriaenoOes tarda, Crlcolopus sp and weevils all important herbivores
3 Weevils dominant herbivore
Physical, Chemical, and Biological Control of Aquatic Macrophytes
mer. The relative abundance of larvae, pupae, and adults
collected at various times from Magic Lake indicates that
7. tarda may undergo two generations per year. Ross (1944)
suggests 7. tarda may have more than one generation,
possibly two per year.
The adult is about 12 to 13 mm long, with a conspicuous
cream and brown wing coloration pattern. The antennae are
about twice the length of the body.
To determine if 7. tarda could be introduced to another
M. spicatum population and achieve a significant degree of
grazing, approximately 3,500 larvae were collected from
Magic Lake in October 1980 and transferred to Sardis Pond,
110 km east (Fig. 1). The larvae were easily collected by
divers. Transport involved overnight storage in aerated
coolers; no significant mortality resulted from the 20-hour
transfer. Subsequent surveys indicated that 7. tarda did not
reproduce and no grazing damage was observed at the new
site. It is believed that the unusually cold winter, which froze
part of the pond, and fish predation eliminated the larvae.
WEEVILS
Lekic and Mihajlovic (1971) have reported several species
of weevils associated with M. spicatum in Yugoslavia, in-
cluding Bagous longitarsus, Eubrychiopsis velatus and LJto-
dactylus leucogaster. LJtodactylus leucogaster is widespread
throughout North America and has been extensively studied
as a potential biological control agent for M. spicatum (Buck-
Figure 1.—Location of Okanagan Valley Lakes, Magic Lake and Sardis Pond.
215
-------
Lake Restoration, Protection and Management
ingham et al. 1981). Kissinger (1964) reports that Eubrychiop-
sis is associated with M. spicatum and has been collected
in Michigan, Wisconsin, Iowa, Alberta, and British Colum-
bia. Hatch (1971) has documented Eubrychiopsis lecontei as
being associated with Myriophyllum and Potamogeton in
British Columbia and Washington. Eubrychiopsis albertanus
has also been collected at Creston, B.C. (Cannings, pers.
comm.).
Surveys during 1980 and 1981 have indicated that four
species of weevils are associated with Myriophyllum spicatum
in various Okanagan Valley lakes: LJtodactylus griseomicans,
Perenthis vestitus, Eubrychiopsis albertanus, and Eubrychiop-
sis lecontei. Unfortunately, the taxonomy of this group of in-
sects is not clear, with many proposed synonyms, therefore,
the identifications presented here may be subject to revision.
A review of the taxonomy of this group is badly needed.
Weevil larvae were seen to have a significant detrimental
impact on M. spicatum in Osoyoos Lake. During a survey
in August 1982, stem mining by larvae in both Osoyoos Lake
study sites were so extensive that most of the plant canopy,
including flower spikes and lateral shoots, had broken and
drifted away. The remaining plant stems were riddled with
stem mines. Remaining stems were badly decayed where
larvae had grazed, and fragmented easily. Active weevil lar-
vae were found in many stems.
Newport
Beachcomber.
Corners
Westbank*
Otter Lake
Swan Lake
'•VERNON
)
7
Kalamalka Lake
Okanagan Valley
10 20 km
Stallion
Osoyoob Lake
Osoyoos «\\ CANADA
Figure 2.—Location of study sites in the Okanagan Valley.
The Osoyoos Lake sites have been surveyed regularly bet-
ween 1980 and 1982; weevils were first noted there in 1981
but no significant grazing impact had been observed previous
to 1982. The 1982 results (Table 1) suggest that weevil graz-
ing was more significant toward the end of summer.
M. spicatum in Skaha Lake also supports a relatively abun-
dant weevil population, but surprisingly, grazing impact is
negligible. This could be because of the remarkable vigor
of M. spicatum in this lake. Mean summer biomass of the
population is 1,236 gm/m2 dry weight, about threefold higher
than values elsewhere in the Okanagan. M. spicatum in
Skaha Lake surfaces earlier and remains at the surface
longer than at most other locations in the Okanagan Valley
lakes and has remarkably thick robust stems.
It is not yet resolved which of the four weevil species ex-
erts the most significant grazing on M. spicatum, although
to date Eubrychiopsis albertanus appears to be the most
widespread.
The feeding behavior observed for the weevils collected
in the Okanagan Valley lakes differs significantly from the
behavior reported for LJtodactylus leucogaster by Buck-
ingham et al. (1981), who indicate that L leucogaster larvae
feed mainly on the emergent flowers of M. spicatum, and
only occasionally on submersed stems or flowers.
Field observations of mainly Eubrychiopsis larvae indicate
that the weevil larvae mine within submersed stems, often
burrowing through several internodes. Entry and exit holes
are often found at each end of the channel. The stem min-
ing destroys all but the epidermal cells and part of the outer
cortex. Microbial infection soon causes the remaining stem
tissue to become dark and soft, often resulting in fragmen-
tation. It is not uncommon to find several larvae mining dif-
ferent parts of the same stem. This stem mining is most fre-
quent in the top 20 to 30 cm of the plant. Mining around
the base of the flower spike is common.
The larvae appear to be well adapted to a wholly sub-
mersed existence and adult weevils also have been collected
on plants that are still far from surfacing. Adult weevils ap-
pear to cause some damage to the plant stems in the form
of small holes or superficial epidermal damage and may
benefit from the O2 available in the extensive air-filled lacunal
system of M. spicatum stems. However, the most abundant
populations of weevils have been observed on weed popula-
tions that surface most readily, suggesting that emergent
flowers considerably benefit weevil reproduction.
It has not yet been determined whether tolerance to a
submersed condition is specific to certain species found in
the Okanagan Valley lakes. However, because of the relative-
ly brief period of emergence and the extensive area of M.
spicatum that remain submersed all year in the Okanagan
Valley, weevil species that can survive a largely submersed
life cycle will be favored.
CRICOTOPUS SP. (CHIRONOMID)
Chironomid larvae were first noted grazing on M. spicatum
apical meristems in 1980. The larvae were later identified
as belonging to the genus Cricotopus. Identification of the
chironomid to the species level proved difficult; this
chironomid may be a new species sharing many of the
characteristics of the sylvestris group. A detailed taxonomic
assessment of adults and larvae is underway to determine
the appropriate species classification.
This larva was widespread through M. spicatum popula-
tions in the Okanagan Valley lakes. Of the three herbivorous
insects discussed in this paper, Cricotopus sp. has de-
monstrated the greatest impact on M. spicatumM is more
widespread and abundant than the other insects, and ex-
erts grazing pressure throughout the growing season. In Ver-
non Arm of Okanagan Lake, for example, an average of 80
216
-------
Physical, Chemical, and Biological Control of Aquatic Macrophytes
percent of plant apices were being grazed between April and
September. An extensive survey of 25 sites carried out in
Okanagan Lake during August 1981 indicated that an
average of 52 percent (n = ==2,600) of randomly sampled
M. spicatum apices were grazed. Values at each site rang-
ed from 0 to 96 percent. This survey also showed that
unharvested populations were generally subject to greater
damage. Table 1 summarizes the proportion of plant apices
grazed by all three insect groups during 1982.
The Cricotopus sp. larvae observed in the Okanagan
Valley lakes exemplify many of the features typical of
chironomid larvae. The fourth instar achieves a maximum
length of about 5 mm. The larvae are a translucent cream
color, and possess fine lateral setae. Cricotopus sp. is
thought to pass through four larval instar stages. The larva
inhabits a silk case attached to the apical bud, or just below
on the first internode or on a nearby leaf. Occasionally, cases
also may be found below the apical bud, but usually not more
than 10 cm away.
The larvae are very mobile within the case and often ex-
tend their entire body length outside the case to graze. Also,
they will reverse their orientation to feed in both directions.
If disturbed, the larvae will abandon their cases and swim.
The larvae usually cause the most damage to the stem and
apical bud, but also damage surrounding leaves. Microbial
infections occur rapidly once epidermal cells are damaged.
It appears that microbial decay contributes significantly to
the overall impact of grazing. If the apical bud continues to
grow despite being damaged, the resulting growth is often
slow and deformed. Twisted leaves, fused leaflets, and great-
ly shortened internodal distances are common deformities.
The net effect of persistent grazing of the apical bud region
is to retard or stop stem elongation and flower formation.
The inability of many M. spicatum populations to surface and
flower has been associated with the presence of abundant
Cricotopus sp. larvae. Although this grazing pressure is con-
sidered primary, other factors being investigated include
water transparency, temperature, water level changes,
climate, nutrient availability, disease, periphyton shading,
plant competition, and population age.
Laboratory rearing and field observations indicate that
several generations occur during the growing season (May-
September). Cricotopus sp. overwinters on the plants in the
larval stage but remains inactive. Significant grazing has
been observed as early as April and continues through the
entire growing season. Adult mating behavior is not well
known.
The overall impact of Cricotopus sp. grazing is subject to
the ability of the plant to outgrow the damage by initiating
laterals or new shoots from the roots. The ability to recover
from grazing is a function of the physiological state of the
plant that varies with season, location, and other stress fac-
tors. The M. spicatum populations in Skaha Lake are so
vigorous that despite significant numbers of herbivorous
chironomids and weevils, no overall impact is apparent. In
Okanagan Lake, however, few populations have been
observed to surface and flower in recent years. Weevils and
caddisflies are present in Okanagan Lake, but Cricotopus
sp. is the dominant grazer.
M. SPICATUM DECLINE
After the initial period of invasion and vigorous growth, M.
spicatum has exploited most available habitats in the
Okanagan Valley lakes. It now appears to have reached an
equilibrium and, to a limited extent, is declining. There may
be several possible reasons for the reduced plant vigor
demonstrated by the reduction in flower production and
emergence. The grazing activities of these three groups of
herbivorous insects, particularly the chironomid larva,
Cricotopus sp., are believed to play an important role.
Thaenodes and Eubrychiopsis are believed to have been
present in the Okanagan Valley lakes prior to the introduc-
tion of M. spicatum.\\ is estimated there was a 5-year lag
period between the initial establishment of M. spicatum in
Osoyoos Lake and subsequent conspicuous weevil grazing.
Similarly, Triaenodes tarda was not observed in Vaseux Lake
until 6 years after the introduction of M. spicatum. Cricotopus
sp. activity was observed about 9 years after the introduc-
tion of M. spicatum. Until the taxonomic difficulties are resolv-
ed with respect to Cricotopus sp. it is not possible to
speculate whether this species is endemic to British Colum-
bia. Although invertebrate surveys were not initiated until
1980, it is assumed that if significant grazing had occurred
before then, it would have been noted during several ex-
tensive mapping surveys and other detailed field studies car-
ried out in the Okanagan Valley lakes by various biologists
since 1972.
BIOLOGICAL CONTROL POTENTIAL
Of the three insects discussed in this paper, Cricotopus sp.
shows the greatest potential for use as a biological control
agent. This is principally because of the larva's ability to at-
tack the apical bud and thereby achieve significant effect
within a short time period, and at much lower population den-
sity than other organisms that simply consume biomass. The
presence of grazing larvae throughout the growing season,
because of the relatively short generation time, is also
important.
Rearing and dissemination of eggs or first instars may
assist native populations achieve greater control in areas
where terrestrial or aquatic conditions are sub-optimal for
the natural development of an abundant Cricotopus sp.
population. Introducing Cricotopus sp. to new habitats us-
ing either field-collected or reared individuals could be useful.
Rearing adults from larvae using simple aquarium facilities
has proved very successful. If the proper conditions were
created to facilitate the mating of adults, large quantities of
eggs or young larvae could be produced. However, a con-
siderable amount of information concerning life history and
plant preferences is still needed, including resolution of its
taxonomic status, before Cricotopus sp. can be employed
as a biological control agent.
IMPLICATION TO PRESENT CONTROL
OPERATIONS
The tolerance of Cricotopus sp., or other herbivorous insects,
to various weed management techniques is largely unknown.
Preliminary data from the Okanagan lakes suggest that
harvesting reduces Cricotopus sp. abundance. A detrimen-
tal impact on any herbivorous insect would be expected as
a result of harvesting and removal from the lake of large
quantities of eggs, larvae, pupae, and adults. A large her-
bivorous insect population might not become established
readily in situations where most of the weed population of
any particular lake is regularly harvested. This has not been
the case in the Okanagan Valley lakes, as most weed
populations have remained untreated. Because of the in-
tolerable nuisance certain weed populations represent,
harvesting operations are essential. However, it is suggested
that harvesting be restricted to the most critical areas. Ad-
jacent untreated areas should not be viewed simply as
sources of fragments and seeds, but may be important for
establishing a natural or introduced population of beneficial
herbivorous insects.
ACKNOWLEDGEMENTS: to Dr. G. B. Wiggins (Royal Ontario
Museum, Toronto) for his valuable assistance in identification of the
217
-------
Lake Restoration, Protection and Management
caddisfly; Dr. D.E. Bright (Biosystematics Research Institute, Ottawa)
for identifying the weevils. Dr. D.R. Oliver's (Biosystematics Research
Institute, Ottawa) help in initiating a detailed taxomnomic review of
the chironomid has proved most valuable.
REFERENCES
Buckingham, G. R., C. A. Bennett, and B. M. Ross. 1981. Investiga-
tion of two insect species for control of Eurasian watermilfoil. U.S.
Army Eng. Waterways Exp. Sta., Tech. Rep. A-81-4.
Cannings, R. 1982. Pers. comm. Provincial Museum, Province of B.C.
Hatch, N. H. 1971. Beetles of the Pacific Northwest. Part 5. Univ.
of Washington Press.
Hepburn, P. H. 1979. Studies on aquatic macrophytes. Part XXVI.
Aquat. Plant Doc. Okanagan Basin, 1978. Water Invest. Br. Rep.
No. 2805.
Kissinger, D. G. 1964. Curculionidae of America North of Mexico.
A Key to the Genera. Taxonomic Publications, South Lancaster,
Mass.
Lekic, M., and Lj. Mihajlovic. 1971. Entomafauna of Myriophyllum
spicatum L. (Halorrhagidaceae), on aquatic weed on Yugoslav ter-
ritory. J. Sci. Agric. Res. 23:59074
Nimmo, A. P., and G. G. E. Scudder. 1978. An annotated checklist
of the trichoptera (insecta) of British Columbia. Syesis 11:117-134.
Ross, H. H. 1944. The Caddis Flies or Trichoptera, of Illinois. Nat.
Hist. Surv. Div. Bull. 23:1-326.
Wiggins, G. B. 1977. Larvae of the North American Caddisfly Genera
(Trichoptera). University of Toronto Press, Toronto.
218
-------
Causes and Control of
Blue-Greens: Alternatives to
Nutrient Control
BLUE-GREEN DOMINANCE IN LAKES: THE ROLE AND MANAGEMENT
SIGNIFICANCE OF pH AND CO2
JOSEPH SHAPIRO
Limnological Research Center
University of Minnesota
Minneapolis, Minnesota
ABSTRACT
Field and laboratory experiments show that when nutrients (N and P) are added to a mixed population
of algae, blue-greens predominate. However, if CO: is added, or if pM is lowered with MCI, greens
predominate. The phenomenon is reproducible and works with most lakes. Although most successful at
pM 5.5, the shift can be made to happen at pM values as high as 8.5. Most blue-greens appear to be
susceptible to the shift and Scenedesmus and Chtorella are the predominant greens. If pH is raised, the
shift is reversed. The reason for the shift is not known. It may involve competition by the algae for CO2,
but other evidence suggests that the lowered pH stimulates cyanophage production, and lysis of the blue-
greens, with release of nutrients which are then used by the greens. Analysis of results of lake circulation
data from the literature and from experiments suggests that the algal shifts resulting from circulation may
involve the same phenomena. Understanding of these phenomena should lead to predictable use of cir-
culation as a lake management tool.
INTRODUCTION
With regard to algae, eutrophication has two dimensions,
quantitative and qualitative. Not only does eutrophication
result in more algae, but the nature of the algal community
changes, generally toward one dominated by blue-greens.
In the last 15 years, our understanding of the quantitative
relationships between nutrients and algae has progressed
rapidly. Beginning with Sakamoto (1966), and continuing with
the work of Dillon and Rigler (1974) and Jones and
Bachmann (1976), the direct dependence of algal biomass
on phosphorus concentration has been amply demonstrated.
Recently it has been shown that the magnitude of this
dependence may differ in different lakes (Smith and Shapiro,
1981), and the suggestion has been put forth that these dif-
ferences in degree of response to the algal population may
be a function of the N/P ratio in the system (Smith, 1982).
We are also aware of the role of herbivorous zooplankters
in reducing algal biomass (Shapiro, 1980), and while this
aspect is not yet highly quantified, we are well on our way
toward answering the question, "How much?"
However, the question "What kind?" has proven more dif-
ficult to answer. Early investigators tended to link blue-green
dominance with conditions such as abundant organic
nitrogen, and later investigators leaned toward high nutrient
content in general as the reason. However, the problem has
remained refractory, to some extent as the result of such
peculiarities as the preponderance of blue-green algae in
Lake Washington for many years following sewage diver-
sion. That is, although the phosphorus content had been
reduced to about 10 ^g/l, blue-greens continued to be the
most abundant form (Edmondson, 1977). Interestingly, in a
219
-------
Lake Restoration, Protection and Management
recent empirical analysis of the U.S. Environmental Protec-
tion Agency National Eutrophication Survey data, Reckhow
and Simpson (1980) concluded that the ratio of blue-greens
to greens is not related to trje concentration of total
phosphorus in the system.
The picture is not entirely bleak, however, although it may
be more complex than we thought at one time. Thus,
Schindler's (1977) and other (Barica et al. 1980) suggestions
regarding the qualitative role of the N/P ratio seem to be in
the right direction, and the recent findings by Smith (1983)
increase our belief that the N/P ratio may be very important
in determining the proportion of blue-greens. Studies by
Tilman et al. (1982) strongly implicate the Si02/P ratio in struc-
turing the algal community. Zooplankton grazing has been
shown to be capable of strongly altering the algal community
structure—in both directions (Shapiro, 1980; Lynch, 1980).
Others have suggested that even such substances as
vanadium are important in determining the degree of
dominance of blue-greens (Patrick, 1976).
Undoubtedly, many factors are involved. However, from
the management point of view, an important question to ask
is: Which of these factors result from eutrophication, and are
therefore controllable? One pair of factors that I believe falls
into this category, and is of overriding importance, is pH and
carbon "dioxide. My belief sterns from a preliminary study I
did (Shapiro, 1973), based on suggestions by King (1970),
that blue-green algae are more efficient than greens at utiliz-
ing low concentration of CO2. Thus, as productivity rises with
increased nutrients, and CO2 declines as reflected in in-
creased pH, blue-greens should come to dominate. My test
of this was direct. I determined, using enclosures, that begin-
ning with a mixed population of greens and blue-greens, ad-
ding N and P alone resulted in a preponderance of blue-
greens, while adding C02 or HCI to pH 5.5, in addition to
nutrients, caused the population to shift toward greens. I also
suggested that the phenomenon might be related to the
sometimes shift of algae from blue-greens to greens during
artificial circulation of lakes (Symons et al. 1969).
Although encouraging, these studies were preliminary and
raised many questions. Accordingly, from 1972 to 1976, work
was done to resolve such issues as response of different
lakes, reproducibility, reversibility, range of pH required,
presence of resistant algae, and the mechanisms involved
in the shift. The results of these studies and other findings
continue to support my belief that pH and C02 are strong
determinants of algal community structure and that we may
be able to control that structure by controlling pH and C02.
METHODS
The methods used were similar to those in the preliminary
experiments (Shapiro, 1973). Polyethylene bags, 1 m in
diameter and 11/2 m deep, closed at their bottom ends, were
suspended from rafts in the lake. Nutrients added in solu-
tion inpluded phosphorus as KH2P04, added over the first
5 days to a final concentration of 100 ^g/l, and nitrogen as
NH4N03, similarly added to a final concentration of 700 to
1,000 ng/l. The pH was adjusted with dilute HCI, dilute NaOH,
or with 100 percent C02 gas bubbled in. Some of the bags
were stirred with compressed air. Sampling was done at the
surface after mixing the upper meter with a wooden pad-
dle. The pH was measured with a variety of field instruments.
Most of the field work was on Lake Emily, a small eutrophic
lake north of St. Paul, the site of the preliminary study. Other
experiments with Lake Emily and other lake waters were car-
ried out in incubators in the laboratory with samples con-
tained in flasks held at 22° C and 300 foot-candles' con-
tinuous light.
RESULTS
Reproducibility
Although the carbon dioxide-induced shift had been
demonstrated first in 1971 and confirmed in 1972, every time
a series of experiments was done at least one pH 5.5 bag,
continuously bubbled with C02 and with nutrients added, was
used along with an appropriate control, with nutrients only,
to confirm the continued capacity of the population to shift.
Such experiments are noted by asterisks in Table 1.
Of these 20 comparable lake experiments, using additions
of C02 and pH 5.5 and additions of N and P, 19 showed
the shift from blue-greens to greens. The sole exception was
experiment F 19 in which N and P were added 6 days after
the C02 treatment was begun. In some of the experiments
stirring with air was continuous while in others stirring oc-
curred only at sampling times. No difference in results was
noted. It appears, therefore, that as the controls never went
over to greens and the C02/nutrient treatments virtually
always did, the phenomenon is highly reproducible.
On one occasion (Table 1, J (1 )6) C02 treatment was con-
tinued only long enough (about 30 minutes) to lower the pH
to 5.5, whereupon the C02 was turned off. In this case only
a partial shift occurred.
Tests in Other Lakes
The phenomenon appears not to be limited to Lake Emily.
On several occasions (Table 1, series H, I, J2), various other
lake waters were tested by incubating them in the laboratory
with N and P and C02. Lake Emily water was treated similarly
for comparison. Not only did the shift occur in Lake Emily
under the laboratory conditions, but it took place in all 10
other lake waters tested. In three of these tests the controls,
containing N and P only, also changed. This may have been
caused by C02 leaking into the incubator from those systems
being bubbled with it. They represent the only controls that
shifted during all the studies.
Time of Year
In addition to the phenomenon being reproducible, and ap-
plying to other lake waters and their algae, it appears to ex-
tend over most of the growing season for the algae. Ex-
periments in Lake Emily have been begun as early as June
24 and as late as Sept. 24, and have gone on as late as
Oct. 25, all successfully.
Effect of HCI versus C02
Because one effect of the C02 addition is to lower the pH
to about 5.5, and because an understanding of the
mechanism of the shift necessitates knowing whether lower-
ing the pH artificially without C02 addition is effective, three
field experiments (Table 1, C(1) 5, 6; and D11) were done
in which N and P were added and the pH was lowered to
5.5 with HCI. The shift took place in all three. However, in
two laboratory tests with water from Quarry Lake the shift
did not occur (Table 1, I).
Experiments at Different pH Values
As noted, the original intent of the work had been to test
a hypothesis of King (1970) on the relative ability of green
and blue-green algae to compete under various conditions.
However, as the possibility exists that the algal shift might
have a practical use in lake restoration, and as it is obvious
that one cannot lower the pH of a whole lake to 5.5 as a
practical procedure, the extent to which the shift would oc-
cur at higher pH values was of interest. Accordingly, several
series of field experiments were done, all at Lake Emily. Elec-
trodes connected to pH stats were used to monitor the bags
220
-------
and to automatically adjust their pH twice daily, with C02 in
one set and HCI in another.
Three such graded pH experiments were done in 1974:
numbers D, E, and F, beginning on June 22, July 16, and
Aug. 21, respectively. As in the pH 5.5 experiments, N and
P were added to all of the bags. The experimental condi-
tions and results are shown in Table 2.
The results were best in experiment D, where the bags
adjusted with C02 to pH 5.5, 6.5, 7.5, and 8.5 all showed
Causes and Control of Blue-Greens: Alternatives to Nutrient Control
the shift, although it was incomplete at pH 7.5. The bags
adjusted with HCI showed the shift at pH values of 5.5, 6.5,
and 7.5. The bag adjusted to pH 8.5 with HCI retained its
blue-green population, as did the control bag, #12, where
the pH exceeded 9.5 as a result of photosynthesis.
In experiment E the bags adjusted to pH 6.5 and 7.5 with
C02 failed to shift, and the pH 8.5 C02 bag and pH 7.5 HCI
bag showed partial shifts only. The control remained blue-
green.
Table 1. — Experiments in which either CO2 or HCI was used to lower the pH to 5.5-6.0. Experiments marked
where CO2 and N & P were added to enclosures in the field.
are those
Series &
experiment
*B2
B3
C(1)!
C(1)s
C(1)6
*C(1)9
C(1)n
C(1)«
*C(2)e
*D2
D 11
*E2
*F2
*F3
*F 15
*F 16
*F 17
*F 18
*F 19
*F22
*F 13
*F 14
*G 2
H
H
H
H
H
H
H
I
I
I
I
j
J/1\2
* \(
(D4
* i
WHO
*J 7
J(2)
J(2)
J' '
J
J(2)
Date
7/29/71
7/29/71
7/23/72
7/23/72
7/23/72
7/23/72
7/23/72
7/23/72
7/23/72
8/9/72
6/22/74
6/22/74
7/16/74
8/21/74
8/21/74
8/21/74
8/21/74
8/21/74
8/21/74
8/21/74
8/21/74
8/21/74
8/21/74
9/24/74
10/74
10/74
10/74
10/74
10/74
10/74
10/74
10/74
10/74
10/74
10/74
7/28/75
7/28/75
7/28/75
7/28/75
7/28/75
7/28/75
7/76
7/76
7/76
7/76
7/76
7/76
Lake
Emily
Emily
Emily
Emily
Emily
Emily
Emily
Emily
Emily
Emily
Emily
Emily
Emily
Emily
Emily
Emily
Emily
Emily
Emily
Emily
Emily
Emily
Emily
Emily
Emily
Pleasant
Powderhorn
Como
Emily
Emily
Emily
Quarry
Quarry
Quarry
Quarry
Emily
Emily
Emily
Emily
Emily
Emily
Ardmore
Lobe
Ryan
Half Moon
Dickey
Shady Oak N.
Agent
C02
CO2
HCI
HCI
HCI
C02
C02
CO2
CO2
C02
CO2
HCI
CO2
C02
CO2
CO2
CO2
CO2
CO2
CO2
C02
C02
CO2
CO2
C02
CO2
C02
C02
C02
CO2
C02
CO2
CO2
HCI
HCI
CO2
CO2
C02
C02
C02
C02
C02
CO2
CO2
C02
CO2
C02
Nutrients Shift
+ Yes
Partial
No
+ Yes
+ Yes
+ Yes
+ Yes
Yes
Yes
+ Yes
+ Yes
+ Yes
+ Yes
+ Yes
+ Yes
+ Yes
+ Yes
+ Yes
+ (a) Yes
+ (a) No
+ Yes
+ Yes
+ Yes
+ Yes
+ (b) Yes
+ (b) Yes
+ Yes
+ (b) Yes
+ Yes
+ Yes
+ Yes
+ Yes
+ Yes
+ No
+ No
Partial
- (c) No
+ Yes
+ Yes
+ (c) Partial
+ Partial
+ Yes
+ Yes
+ Yes
+ Yes
+ Yes
+ Yes
Comments
1 X inoculum
1 X inoculum
1 X inoculum
1 X inoculum
1 X inoculum
1 X inoculum
1 X inoculum
1 X inoculum
5 X inoculum
20 X inoculum
lab
lab
lab
lab
lab 1 X inoc.
lab 2 X inoc.
lab 4 X inoc.
lab
lab
lab
lab
lab
lab
lab
lab
lab
lab
(a) - CO] begun day 1; nutrients added days 6-10.
(b) - control also shifted.
(c) - COi added only briefly a' "he beginning.
221
-------
Lake Restoration, Protection and Management
In experiment F both pH 5.5 bags shifted. Neither bag at
pH 6.5 did so. At pH 7.5 one bag remained predominantly
blue-green and the other showed a partial shift. Of the bags
at 8.5 one did not shift and the other did. However, the former
actually was at pH 7.5 and the latter at 6.5 because of poor
control by the pH stats. The control remained blue-green.
Thus the results were not clear cut and seemed to show
an element of chance. However, they do show that in nine
of 16 cases the change occurred, completely or partially, at
pH values above 5.5 and at values as high as 8.5. Thus the
possibility of shifting algal populations from blue-greens to
greens, as a practical measure, is reinforced.
Biological Details of the Shift
At this point it is worth considering the nature of the shift
in some detail. Table 3 lists the predominant algae at the
beginning of each series of experiments, and the algae mak-
ing up the bulk of the population after the shift had oc-
curred. The point of greatest interest here is that filamen-
tous and gelatinous blue-greens give way to greens of a few
restricted genera. The general effect is best portrayed by
a photograph in Figure 1. In the case shown (Exp. D2, Table
1) the shift occurred at a pH of 5.5 and resulted in mostly
Chlorella and Scenedesmus. These changes at pH 8.5 are
summarized in Figure 2.
Frequently more than one species of the dominant green
genus appeared. For example, in the pH 8.5 experiment
described, following the shift no fewer than 22 species and
subspecies of Scenedesmus alone were found.
Rate of the Shift
The rate at which the shift occurs is variable. In general, ex-
periments done in the late spring and perhaps in the fall
result in quicker shifts than do those done in mid-summer.
This may be related to temperature, but it may also be
because a larger inoculum of green algae is present earlier
DAY 2
DAYS
DAY tt
DAY
DAY is
DAY 19
Figure 1 .—Photographs showing the shift induced by addition of
N and P, and CO2 to a pH of 5.5.
Table 2. — Experiments in which pH was maintained at predetermined values. N and P added in all cases.
Series &
experiment
D 2
D4
D5
D6
D 11
D 7
D 8
D 9
D 12
E 2
E3
E5
E7
E8
E 1
F 2
F3
F5
F6
F7
F 8
F9
F 10
Date
6/22/74
6/22/74
6/22/74
6/22/74
6/22/74
6/22/74
6/22/74
6/22/74
6/22/74
7/16/74
7/16/74
7/16/74
7/16/74
7/16/74
7/16/74
8/21/74
8/21/74
8/21/74
8/21/74
8/21/74
8/21/74
8/21/74
8/21/74
pH
5.5 (a)
6.5
7.5
8.5
5.5 (b)
6.5
7.5
8.5
> 9.5 (c)
5.5 (a)
6.5
7.5
8.5
7.5
>9.5(c)
5.5 (a)
5.5 (a)
6.5 (7.0)
6.5 (7.0)
7.5
7.5
8.5 (7.5)
8.5 (6.5)
>9.5(c)
Agent
CO2
CO2
CO2
CO2
HCI
HCI
HCI
HCI
—
CO2
C02
CO2
CO2
HCI
—
CO2
C02
C02
CO2
CO2
CO2
C02
CO2
~
Shift
Yes
Yes
Partial
Yes
Yes
Yes
Yes
No
No
Yes
No
No
Partial
Partial
No
Yes (d)
Yes (d)
No(d)
No(d)
No(d)
Partly (d)
No(d)
Yes (d)
No(d)
( ) = actual pH
(a) = excess CC>2, no stat
(b) =* HCI, no stat
(c) = no stal, high pH from photosynthesis
(d) * 1 X inoculum
222
-------
Causes and Control of Blue-Greens: Alternatives, to Nutrient Control
and later. In one case (Series F, Table 2) suspecting too few
green algae to begin with, we inoculated all of the bags with
green algae (mostly Scenedesmus and Chlorella) that had
resulted from a shift in the previous experiment. The nor-
Table 3. — Algae predominating before and after the
shift. (* = shift done with chlorine Instead of pH change)
Experiment series
B
C
D
E
F
H Emily
H* Emily
H Pleasant
u*
n Pleasant
" Powderhorn
" Como
H *
n Como
J 0)
J*(1)
J{2) Ardmore
J(2) Lobe
J(2) Ryan
J(2) Halt Moon
J(2) Dickey
J(2) Shady Oak N.
Algae before shift
Oscillatoria
Anabaena
Microcystis
Oscillatoria
Anabaena
Aphanizomenon
Oscillatoria
Anabaena
Gomphospheria
Oscillatoria
Aphanizomenon
Oscillatoria
Oscillatoria
Scenedesmus
Oscillatoria
Scenedesmus
Oscillatoria
Ankistrodesmus
Oscillatoria
Ankistrodesmus
Oscillatoria
Scenedesmus
Oscillatoria
Scenedesmus
Oscillatoria
Scenedesmus
Chroococcus
Microcystis
Anabaena
Chroococcus
Microcystis
Anabaena
Aphanizomenon
Oscillatoria
Merismopedia
Anabaena
Gleocystis
Anabaena
Tribonema
Microcystis
Oscillatoria
Chroococcus
Oscillatoria
Chroococcus
Algae after shift
Scenedesmus
Chlorella
Chlorella
Dictyosphaerium
Scenedesmus
Chlamydomonas
Chlorella
Scenedesmus
Coelastrum
Scenedesmus
Chlorella
Scenedesmus
Stichococcus
Chlorella
Scenedesmus
Chlorella
Chlorococcus
Scenedesmus
Chlorella
Oscillatoria
Mougeotia
Scenedesmus
Scenedesmus
Chlorella
Scenedesmus
Chlorella
Nitzschia
Chlorococcus
Cosmarium
Selenastrum
Dictyospherium
Gleocystis
Cryptomonas
Ankistrodesmus
Selenastrum
Scenedesmus
Closteriopsis
Chorella
Chlorella
Ankistrodesmus
Chlorella
Ankistrodesmus
Ankistrodesmus
Nitzschia
Closteriopsis
Selenastrum
Scenedesmus
Sphaerocystis
mal inoculum used was about 1 volume/125 volumes. To
determine the effect of inoculum size we inoculated two other
bags (F 13 and F 14) with 5 X and 20 X this quantity, respec-
tively. The bags with 1 X and 5 X inoculum appear to have
undergone the shift after approximately 18 days treatment
with C02 and nutrients at pH 5.5, but the bag with 20 X in-
oculum had shown a clear shift by day 13.
Obviously then, it is necessary to have greens present to
grow. But not all of the greens respond. On occasion some
have not changed in abundance while others have even
decreased.
In looking at the effect of pH on the rate of the shift, the
data are equivocal. In Experiment D (Table 2) where pH
values of 5.5, 6.5, and 8.5 were achieved with C02, the shift
occurred in 11,10, and 10 days, respectively. When the pH
was maintained at 5.5, 6.5, and 7.5, with HCI, the shift re-
quired 9, 10, and 13 days, respectively — suggesting, under
these conditions, a slightly faster response at the lower pH.
One fact about the shift is that it often appears to occur
somewhat precipitously after a variable period with no ap-
parent change. For example, we have on several occasions
examined the bags on say, Friday evening, and decided that
little change had occurred, while on the following Monday
the algae seemed to be almost all Chlorella and
Scenedesmus.
Resistant Algae
Of perhaps as much interest as what species of algae
become abundant, is the question of whether any species
of blue-greens are resistant to the shift. As noted earlier, in
19 of 20 cases where nutrients were added, and pH was
lowered to 5.5 with C02, the shift occurred. However, in
several instances where other pH levels were used, the shift
did not occur. Does this mean that certain species of algae
are resistant? Probably not. For example, in experiments D2
and D6, carried out with C02 at pH 5.5 and 8.5, respective-
ly, all of the following disappeared to the point where they
were below enumeration levels: Oscillatoria, Anabaena,
Chroococcus, Gomptiosphaerium, and Aphanizomenon.
Thus, these species were vulnerable at both pH levels.
However, examination revealed that Oscillatoria was begin-
ning to reappear after a week at pH 5.5 but remained ab-
sent at pH 8.5. Apparently, as with most organisms, some
individuals are more resistant than their fellows and if in a
particular experiment the shift does not occur it probably is
for some reason other than that the algae are resistant
species.
In certain ol the experiments Phormldlum may have been present with
Osc/llatorla. We have used the name Oscillatoria alone.
Figure 2.—Changes in the algal community brought about by ad-
ding N and P, and maintaining a pH of 8.5 with CO:.
223
-------
Late Restoration, Protection and Management
Reversibility of the Shift
Several attempts were made to determine if the shift was
reversible, i.e. if once again raising the pH or allowing it to
increase by itself through photosynthesis, would bring back
the blue-green dominance. One such attempt was experi-
ment C! (6) (Table 1). This experiment was begun on July 27,
1972 with the initial algal population consisting of Mterocystis,
Osclllatoria, Anabaena, and Aphanizomenon. The enclosure
was treated with N and P and HCI, at pH 5.5, and by Aug.
3 Scenedesmus and Dictyosphaerium were abundant. By
Aug. 8 Scenedesmus was predominant. At that time the pH
was raised to 9,2 with KOH and more N and P were added.
On Aug. 16 Scenedesmus was still the dominant alga but
by Sept, 24 Anabaena, Oscillatoria, and Aphanizomenon had
regained dominance.
Perhaps the best such experiment from the technical point
of view was that begun on Aug. 21,1974. The algae in bag
E2 (Table 1) which had been treated with C02 and N and
P at a pH of 5.5, had become predominantly greens
(Scenedesmus) between July 16 and Aug. 8. From Aug. 8
to 21 the bag was allowed to sit with no further additions
of C02. It remained as it was. On Aug. 21 it was divided into
two bags. The first had C02 and N and P added as original-
ly. The other had its pH raised to 9.5 with KOH. The first
retained its dominance of Scenedesmus until the experiment
was ended on Sept. 18, although a few filaments of
Oscillatoria did remain. The second bag, with the high pH,
behaved similarly at first but began to develop more and
more filamentous blue-greens. By Sept. 19 the population,
although it still had many Scenedesmus, was once again rich
in filamentous blue-greens.
In another example, bag G^0, which had been shifted to
greens at pH 5.5 with C02 and N and P, had its pH raised
with KOH but, by chance, only to 6,7. Some reversion did
occur but the population remained about half greens and
half filamentous blue-greens.
Finally, one case exists in which reversion occurred without
adding KOH. This was experiment C^f, Treatment begun on
July 23,1972 caused the blue-greens to shift over to Chlorella
and Scenedesmus, and by Aug. 16 a notation was made
that the bag contained "all greens, no blue-greens." On Aug.
25 the C02 was shut off. By Sept, 24 Scenedesmus was still
abundant but C/voococcus, and many filamentous blue-
greens, were present. The pH had not been monitored but
undoubtedly it had risen as a result of photosynthesis.
Mechanism of the Shift
The reason for the shift is still not known. One might presume
that lowering the pH to 5.5 with C02 or HCI would be, as
suggested by Brock (1973), detrimental to the blue-green
algae and, in fact, might kill them. However, notwithstand-
ing this possibility, we have been able to culture blue-greens,
including Oscillatoria, successfully at pH 5,5. Furthermore,
the shift has been made to occur at pH values as high as
8.5, and surely this cannot be claimed to be a physiologically
damaging pH for blue-greens. The fact that the shift occurs
only sometimes at a pH value above 5.5 also suggests the
view that it is not the pH per se that is important.
Originally we believed that the shift resulted from the
superior C02 uptake kinetics of the blue-green algae and that
reversal of competition to favor the greens was brought about
by supplying C02 either directly by bubbling it in, or by lower-
ing the pH so as to decompose the native alkalinity of the
lake waters. Although Long (1976) has confirmed the ad-
vantage of blue-greens over greens in C02 uptake ability,
we do not believe that this is the sole explanation. That is,
the change does not appear to be one in which green algae
slowly replace blue-greens. In most cases where the shift
has occurred the blue-greens have rapidly decreased — as
indicated by changes in numbers and decreases in
chlorophyll — before the greens have begun growing rapid-
ly. The blue-green decrease and the green increase thus
appear to be separate phenomena.
Furthermore, phosphorus and perhaps nitrogen appear
to be involved. When the blue-greens decrease, high con-
centrations of PO4-P and NH3-N are found in the water.
These substances are not left over from the initial addition
that is usually made, but apparently have been liberated from
the blue-green cells. In control systems where the pH re-
mains high and no shift occurs, phosphate does not appear
in solution. However, where the shift occurs, phosphate and
ammonia do show up (Fig. 3). Thus, of 42 examples for
which data are available, 17 showed a positive cor-
respondence between the shift and the release of phosphate;
18 showed a correspondence between no shift and no
release of phosphate; and the remaining seven cases
demonstrated neither correspondence. However, in four of
the last seven the release began but the experiment was
terminated, and in these the shift might have taken place
if given time.
—-—-^_
10 12 14 16
DAYS
-150
-125
-tOQg;
-J
-50
-28
-0
Figure 3.—Changes in chlorophyll a, and in dissolved phosphate,
at pH 5.5.
Thus the release of nutrients from the blue-greens ap-
parently must occur before the greens grow. The nutrients
then are, of course, used by the greens. Curiously enough,
adding nutrients at the beginning of the experiment, even
when greens are rare, facilitates the shift. If N and P are
not added the shift, even at pH 5.5, is slow and partial. But
if N and P are added, even though they are absorbed rapidly
by the blue-greens, the eventual shift is more dramatic.
One hypothesis about what might be happening to cause
the shift is that the blue-greens are being destroyed by
cyanophage or bacteria as a result of the pH manipulations
(Lindmark, 1982), and that once the blue-greens are gone
the few greens initially present no longer have competition
for the nutrients and increase rapidly. The reason that
nutrients are required may be that when blue-greens
photosynthesize rapidly their cyanophages replicate (Des-
jardins, 1983).
This hypothesis is supported by certain observations. For
example, when Qscillatoria disappears as a result of lowered
pH, the filaments do break up first as though certain cells
were being lysed. Also, not all blue-greens are affected. Thus,
Raphidiopsls, for example, may continue to grow and in-
crease in numbers even though Osdllatoria is decreasing
in numbers. This suggests a species-specific agent. Also,
even when Oscillatoria has decreased it sometimes regrows
224
-------
Causes and Control of Blue-Greens: Alternatives to Nutrient Control
in the same system, (e.g. D2), indicating that it is not the
pH, which remains low, that is the important feature.
An observation that bears on the mechanism of the shift
concerns the effect of arsenic on the phenomenon. Earlier
work (Shapiro, unpubl.) had shown that whereas blue-green
algae distinguish between phosphate and arsenate, taking
up the former and leaving the latter, green algae do not make
this distinction and arsenate slows their uptake of phosphate.
Accordingly, we reasoned, if phosphate transfer from
lysed blue-greens to greens is important, would adding
arsenate slow or stop the shift: Two experiments were done
to test this. The first was begun July 16, 1974. Two bags
were set up at Lake Emily, both treated with C02 to pH 5.5,
and with N and P added as usual. However, at the beginn-
ing of the experiment 100 micrograms/liter of arsenic as
sodium arsenate was added to one of the bags, with the
other serving as a control. Both bags behaved similarly un-
til about the 14th day, i.e. chlorophyll fell to a low level by
day 10, and PO4-P was found in solution as expected.
However, the control bag shifted to greens with a maximum
of chlorophyll by day 17, while the "arsenate" bag did not
reach its peak of green algae until day 24. Thus the presence
of the arsenate caused a 1 -week delay, although the shift
did occur and the final chlorophyll levels reached were
similar.
The second experiment, begun Aug. 21, 1974, was con-
ducted in the same way as the first. Again, the control and
experimental bags behaved similarly for the first 2 weeks.
However, the control bag shifted shortly thereafter with a
significant increase in chlorophyll (to about 90
micrograms/liter) while no shift took place in the presence
of the arsenate, and the chlorophyll reached only 32
micrograms/liter. Thus, in this case the arsenate complete-
ly prevented the shift from occurring.
Finally, in considering the mechanism of the shift, the ef-
fects of chlorine must be taken into account. It had previously
been found that chlorine limited the rate of phosphate up-
take by blue-green algae more than by green algae (Shapiro,
unpubl.). Therefore, it was reasoned, if phosphate is added
to a mixture of algae in the presence of chlorine, the green
algae should have an advantage and the shift from blue-
greens to greens might occur even at a high pH.
In the first such experiment, begun July 16, 1974, 750
micrograms/l of chlorine were added to an experimental bag
along with N and P, while the control bag had only N and
P added. The two bags behaved similarly in that both
developed crops of algae containing approximately 300
micrograms/l of chlorophyll. No shift occurred in either one
although phosphate was released and later reabsorbed in
the chlorine bag. Both had pH values of 10.
The second chlorine experiment was begun on Aug. 21,
1974, and was done in duplicate. In both experimental
bags the blue-greens were greatly diminished in the
presence of chlorine, and there was a shift to greens even
though the pH was between 10 and 11. As in the previous
experiment, PO4was released in both experimental bags.
Another field experiment in Lake Emily in July 1975, and
four of five laboratory experiments on other lakes, all
demonstrated that chlorine caused the shift at high pH.
These experiments, along with the arsenate work, sup-
port the idea that phosphate uptake by greens and their
subsequent growth is a phenomenon separate from the
demise of the blue-greens. However, the reason for the
blue-green decline still remains unknown. Some ex-
perimental evidence suggests C02 competition as the
mechanism responsible for the shift and some indicates
that cyanophage may be the agent. Possibly both are
important.
Regardless of the mechanism, the phenomenon of the
blue-green to green shift is real, and other studies in ad-
dition to those already cited show the importance of C02
and pH. For example, Tailing (1976) has shown that at pH
7, the specific rate of photosynthesis of Selenastrum
capricornutum exceeds that of Anabaena f/os-aquae, but
at pH values above 8, the situation is reversed. The diatom
Asterionella behaves in a manner similar to Selenastrum,
with rates of net C02 uptake increasing sharply as pH is
lowered from 9 to 7.
RELATION TO LAKE CIRCULATION
In the earlier study (Shapiro, 1973) it was suggested that
the blue-green to green shift is reminiscent of what
sometimes happens in lakes during artificial destratifica-
tion, and that mixing a stratified lake is analogous to ad-
ding C02and nutrients. It is remarkable then that Tailing
(1966) found that in Lake Victoria, during several natural
periods of increased vertical mixing, diatoms and green
algae increased in numbers while most blue-greens de-
clined in abundance. Although Tailing suspected that the
blue-greens declined in relation to a maximum "critical
depth," it seems just as likely that the response was to
pH which decreased in near-surface waters during the mix-
ing periods.
This view is reinforced by experiences of various in-
vestigators using artificial destratification to "improve"
lakes. Shapiro and Forsberg (unpubl.) in a preliminary
survey of the literature, found that in most cases where
artificial circulation led to a blue-green-green shift, pH had
declined as a result of circulation; whereas where blue-
greens had increased following circulation, pH was shown
to have risen—probably from increased primary produc-
tion. Recently, Pastorok, etal. (1981), completed a similar
analysis with similar results. In 15 cases in which the ratio
greens/blue-greens increased, pH declined in 11, re-
mained constant in three, and rose in one. In 16 cases
where the ratio decreased, i.e. where blue-greens became
more dominant, pH declined in four, remained constant
in 11, and possibly rose in one.
It is of interest to consider why the second group dif-
fered from the first, i.e., why some circulation attempts led
to a higher proportion of blue-greens instead of resulting
in their demise. The answer may lie partly in the relative
rates of mixing of the lakes involved, but it is more com-
plex than that. That is, Lorenzen and Fast (1977) sug-
gested that as a rule of thumb, approximately 9 m3 of
air/106 m2 of surface area is required to maintain good mix-
ing. Of the lakes in the aforementioned two categories,
in those in which pH fell mixing had been applied at 0.49
to 21.5 m3/106m2, and only two were at or above the
Lorenzen/Fast criterion. The sole datum for the lake where
pH rose was 3.9 m3/106m2, below the criterion. Lakes
showing no pH change were aerated in the range of 1.0
to 9.80 m3/106m2, with two of the eight lakes exceeding
the criterion. Thus the picture is far from clear.
A more likely explanation is that the effectiveness of the
mixing rate applied to a given lake is what determines
whether or not pH declines, and therefore whether or not
the shift will occur. Included in such effectiveness would
be velocity of mixing, rate of CO2 introduction to the
euphotic zone, buffering capacity of the water, and rate
of introduction of nutrients to the euphotic zone. The last
could increase pH as a result of increased photosynthesis.
Experiments done by Forsberg and Shapiro (1982) us-
ing deep (7 m) enclosures in stratified lakes confirm that
these factors are important. In general, shallow mixing of
such enclosures was ineffectual, while slow deep mixing
resulted in higher phosphorus concentrations and higher
pH of the surface waters, along with increased proportions
of blue-greens. On the other hand, rapid deep mixing,
225
-------
Lake Restoration, Protection and Management
although it resulted in high concentrations of phosphorus
in the euphotie zone, also lowered pH and brought about
an increase in the growth rate of diatoms and green algae
relative to that of blue-greens, sometimes leading to the
blue-green to green shift. Comparison of two lakes, one
with a higher buffering capacity, showed that the rate of
mixing adequate to cause the shift in the less buffered
lake, was not adequate in the more highly buffered lake,
which initially also had less CO2 in its hypolimnion.
Thus, manipulation of pH and CO2 by artificial circula-
tion, when done properly, may lead to a decline in blue-
green dominance (Fig. 4), although if circulation is done
improperly, blue-greens are likely to become more domi-
nant and abundant (Fig. 5).
Other effects of pH and CO2 may also play a role. For
example, Booker and Walsby (1981), Klemer et at. (1982),
and Paerl and Ustach (1982) have all shown that re-
duced concentrations of C02 and/or high pH cause blue-
greens to become more vacuolate and to rise to the sur-
face. Circulation, which increases CO2and lowers pH in
upper waters, would therefore reduce surface scums at
the same time that it caused the blue-green to green shift.
This would cause an apparent diminution of biomass that,
with the increased grazing on green algae by herbivorous
zooplankton, would accentuate the clarity and aesthetics
of the lake.
CONCLUSION
CO2 concentration and pH appear to be important
variables helping to determine the proportion of blue-
greens in the algal community. Although the manner by
which C02and pH exert their effect is not known, the fact
CIRCULATION
Aerot
of Or
water
ic Resp
panics n
+ sed
Surface Recarb
* Hypolimnelic CO?
~>_ N
^\. K
"t
S
/
jtrienl
nelcs
P04
>ecies
j Rec^uOmen!
j of Heavier Algoe
ecrease
in
iperoturg
1
Inadequate
CIRCULATION
increase in proportion
of blue-green algoe
Figure 4.—Diagrammatic representation of the manner by which
whole-lake circulation might cause a shift from blue-greens to green
algae and diatoms.
Figure 5.—The probable manner by which inadequate circulation
results in a greater proportion of blue-greens.
that both can be manipulated by such measures as in-
duced circulation may allow us to control blue-green
populations in takes.
ACKNOWLEDGEMENTS, Vincent Lamarra, George Zoto, and
Bruce Forsberg participated in various aspects of the study and
in discussions about the data. Funds were provided at various
times by the National Science Foundation, the U.S. Department
of the Interior, and the U.S. Environmental Protection Agency.
Contrib. No. 259 from the Limnological Research Center, Univer-
sity of Minnesota.
REFERENCES
Barica J., H. Kling, and J. Gibson, 1980, Can. J. Fish Aquat
Sci. 37: 1175-1183.
Booker, M. J., and A. E. Walsby. 1981, Br. Phycol. J 16-
441,
Brock, T, D. 1973. Science 197: 480.
Desjardins, P. R. 1983, Cyanophage: History and likelihood as
a control. In Lake Restoration, Protection, and Management,
Proc. 2nd Annu. Conf. N. Am. Lake Manage. Soc. EPA
400/83/001, U.S. Environ. Prot, Agency, Washington, D.C.
Dillon, P. J., and F. H. Rigler. 1974. Limnol, Oceanogr. 19:
767.
Edmondson, W. T. 1977. In Trophic Equilibrium of Lake Wash-
ington, EPA-600/3-77-087. U.S. Environ. Prot Agency
Washington, D.C.
Forsberg, B., and J. Shapiro. 1982. In Biomanipuiation. EPA-
600/3-82-096. U.S. Environ. Prot. Agency, Washington, D.C.
Jones, J. R., and R. W. Bachmann. 1976. J. Water Pollut Con-
trol Fed.48: 2176.
King, D. L, 1970. J. Water Pollut. Control. Fed. 42: 2035.
226
-------
Causes and Control of Blue-Greens: Alternatives to Nutrient Control
Klemer, A. R., J. Feuillade, and M. Feuillade. 1982. Science 215:
1629.
Lindmark, G. 1979. Effects of environmental stresses on the rela-
tionship between Plectonema boryanum and cyanophage
LPP-1. In Biomanipulation. EPA-600/3-82-096. U.S. Environ.
Prot. Agency, Washington, D.C.
Long, E. B. 1976. The interaction of phytoplankton and the bicar-
bonate system. Ph.D. Thesis, Kent State Univ.
Lorenzen, M. W., and A. W. Fast. 1977. Ecol. Res. Ser.
EPA-600/3-77-004. U.S. Environ. Prot. Agency, Washington,
D.C.
Lynch, M. 1980. In W. C. Kerfoot, ed. Evolution and Ecology
of Zooplankton Communities. Spec. Symp. Vol. 3. Am. Soc.
Limnol. Oceanogr.
Paerl, H.W., and J. F. Ustach. 1982. Limnol. Oceanogr. 27: 212.
Pastorok, R. A., M. W. Lorenzen, and T. C. Ginn. 1981. Tech.
Rep. E-81. Prepared byTetraTech. Inc., Bellevue, Wash, for
U.S. Army Eng. Waterways Exp. Sta. Vicksburg, Miss.
Patrick, R, 1976. In J. Tourbier and R. W. Pierson, eds. Biological
Control of Water Pollution. Univ. Pennsylvania Press.
Reckhow, K. W., and J. T. Simpson. 1980. Proj. Comple. Rep.
OWRT A-102-MICH. Inst. Water Res. Michigan State Univ.
Sakamoto, M. 1966. Arch. Hydrobiol. 62: 1.
Schindler, D. W. 1977. Science 195: 260-262.
Shapiro, J. 1973. Science 179:382
1980. In J. Barica and L. R. Mur, eds. Hypertrophic
Ecosystems. Junk, The Hague.
Shapiro, J., and B. Forsberg. Limnol. Res. Center, University
of Minnesota (Unpubl.).
Smith, V. H. 1982. To be publ. in Limnol. Oceanogr.
1983. The nitrogen and phosphorus dependence of
blue-green algal dominance in lakes. In Lake Restoration, Pro-
tection, and Management. Proc. 2nd Annu. Conf. N. Am. Lake
Manage. Soc. EPA-400/-83-001, U.S. Environ. Prot. Agency,
Washington, D.C.
Smith, V. H., and J. Shapiro. 1981. Environ. Sci. Technol. 15: 444.
Symons, J. M. et al. 1969. Water Quality Behavior in Reservoirs.
U.S. Dep. Health Edu. Welf. Pub. Health Serv. Pub. No. 1930.
Tailing, J. F. 1966. Int. Revue ges. Hydrobiol. 51: 545.
. 1976. J. Ecol. 64: 79.
Tilman, D., S. S. Kilham, and P. Kilham. 1982. Ann. Rev. Ecol.
Syst. 13: 349.
227
-------
CONTROLLING BLUE-GREEN ALGAE BY ZOOPLANKTON GRAZING
ROBERT E. CARLSON
Department of Biological Sciences
Kent, Ohio
STEVEN A. SCHOENBERG
Department of Zoology and Institute of Ecology
University of Georgia
Athens, Georgia
ABSTRACT
The use of zooplankton grazing to control the density of blue-green algae in lakes has been seldom
attempted because of a resistance to use a non-nutrient-related means of control and because of
a general belief that blue-green algae are unavailable or toxic to grazing zooplankton. We suggest
that eutrophication is a change in the biological, not nutrient status of a lake, and any method, whether
abatement of nutrient inputs or manipulation of zooplankton grazing intensities that can be demonstrated
as a feasible method to alter that biological condition, is a viable means of controlling eutrophication.
The evidence that zooplankton cannot graze effectively on blue-green algae is largely derived from
laboratory experiments, but a critical review of these experiments reveals numerous instances where
zooplankton did graze effectively on blue-greens. In lakes, zooplankton may not only graze directly
on the blue-greens, but may cause their demise by altering the ambient pH, light environment, and
nutrient concentration.
INTRODUCTION
Blue-green algae often become the dominant algae in
eutrophic fakes. They have characteristics such as a pro-
pensity to float, to be toxic to mammals, and to impart a taste
and odor to water and fish flesh that lead to their being con-
sidered undesirable in lake ecosystems. Control of blue-
greens by chemicals is often short-lived; it is thought they
can be permanently eliminated only by removing the source
of the nutrients that caused the shift to eutrophy in the first
place.
Although nutrient removal is a proven method for restor-
ing lakes, it is often expensive, more often impossible in the
case of nonpoint nutrient sources, and in some cases, inef-
fective because of internal nutrient sources. Alternative forms
of algal control must, by necessity, be considered. This paper
concerns the possibility of enhancing the grazing effec-
tiveness of zooplankton as a means of controlling blue-
greens.
THE MEANING OF EUTROPHICATION
A major road block to the acceptance of zooplankton
manipulation as a eutrophication control mechanism lies in
the popular definition of eutrophication itself. For the past
15 to 20 years we have so effectively propagandized the role
of nutrient loading (especially phosphorus) that eutrophica-
tion has become synonymous with the increase in nutrient
inputs to lakes (Bartsch, 1969; Kratzer and Brezonik, 1982).
By implication, a eutrophic lake is a lake that has high
nutrient inputs. Because of our fixation on nutrient inputs,
we are Winded to the possibility that in some instances wtthin-
lake factors may be even more important than nutrient in-
puts in determining the conditions of the lake. Wtthin-lake
treatments are often considered to be "cosmetic," treating
the superficial appearance of eutrophication rather that its
"real" cause, i.e., nutrient inputs.
We object to a nutrient-based definition of eutrophication,
and, following in the tradition of Rodhe (Aberg arid Rodhe,
1942; Rodhe, 1969) and Vollenweider (1968), we would em-
phasize that eutrophication is a biological rather than a
nutrient condition. We differ from these authors in that we
would stress the total biological condition of the lake rather
than primary productivity alone, but we believe it is critical
to our understanding of lakes that we separate causal fac-
tors, such as nutrient inputs, from resulting conditions, i.e.,
primary productivity, algal btomass, macrophyte abundance,
and blue-green domination.
If one can separate cause from result, then it is a short
mental jump to the realization that any factor, whether
nutrients, bkrtic shifts, or basin morphometry, that increases
algal and macrophyte abundance is a causal factor of
eutrophication, and, conversely, any factor, such as
zooplankton grazing, that can reduce algal biornass, is not
simply cosmetic, but a possible alternative to nutrient reduc-
tion for decreasing the degree of eutrophication.
The preoccupation with nutrients is not surprising, as we
have been rather successful in producing nutrient loading
models that can in some cases satisfactorily predict the
amount of chlorophyll in the open water of lakes. Nutrient
inputs have of course been involved in the eutrophication
of many lakes, and the pressure to find simplistic causes
and solutions may have contributed to our emphasis on
nutrient loading as the sole cause of eutrophication.
However, once the more blatant examples of nutrient abuse
are behind us, a number of eutrophic lakes will remain in
which it will either not be economically feasible to reduce
nutrient inputs, or in which no clear relationship between
nutrient inputs and trophic state exists.
If one can accept the concept that eutrophy is the
biological state of the system, and that eutrophy is only
broadly related to nutrients, then one can objectively explore
methodologies other than nutrient reduction to control blue-
green algae in lakes. Figure 1 is a simplified description of
some of the interactions that occur in lakes during
eutrophication. We are suggesting that blue-green algal
dominance results from ambient conditions produced by high
228
-------
Causes and Control of Blue-Greens: Alternatives to Nutrient Control
INCREASED NUTRIENT LOADING
PRIMARY PRODUCTIVITY
I'
pH
BLUE-GREENS
MACROPHYTES
FORAGE FISH
COVER
GRAZING
IMPACT
ANOXIC
NUTRIENT
RELEASE
SEDIMENTATION
RATE
HYPOLIMNETIC '
0,
ZOOPLANKTIVORE
SEARCH
AREA
Figure 1 .—A conceptual model of the biotic interactions involved
in eutrophication.
algal biomass and high primary productivity. The decrease
in carbon dioxide concentration and increase in pH have
been implicated in blue-green dominance by King (1970),
Shapiro (1973), and Long (1976). The low-light environment
caused by high densities of algae has been suggested to
promote growth of some blue-greens (Brown and Richard-
son, 1968; Wall and Briand, 1979). Figure 1 also implies that
decreased grazing efficiency in eutrophic lakes may be
because of the decreased grazing on the inedible or toxic
blue-greens, but may also be the result of increased fish
predation on the herbivores, caused by increases in
macrophyte cover for zooplanktivorous fish (Kerfoot, 1974)
or by decreased zooplanktivore search volume resulting from
the depletion of hypolimnetic oxygen (Haney, 1973).
The enhancement of zooplankton grazing is an attractive
possibility for algal control. Grazing should reduce algal
biomass and change the energy pathways into the grazing
food webs, therefore both increasing sportfish productivity
and decreasing detrital export into the hypolimnion. Why then
hasn't zooplankton enhancement been used to manipulate
eutrophy? Part of the answer lies in our firm belief that
zooplankton cannot graze effectively on blue-green algae.
ZOOPLANKTON AND THE CONTROL OF
BLUE-GREENS
The common belief is that zooplankton cannot control blue-
green algae because blue-green algae themselves are too
large, indigestible, or toxic, or that the densities of
zooplankton during the summer months are too low to ef-
fect any significant control. Many of the nuisance algae are
quite large, being either long filaments or colonial. Unless
the herbivores can crush the filament or eat it from the end,
or nibble away at the cells within the gelatinous matrix, these
cells should be invulnerable to grazers. A review of the
literature, however, reveals a disturbing number of incon-
sistencies.'All too often the same blue-green species is
reported noningestible or toxic by one author, yet another
author finds it to be a totally acceptable food.
Filaments, although large, are not necessarily unavailable
as food for herbivores. Some copepods apparently can in-
gest filaments. The alimentary tract of the copepod, Diap-
tomus gracilis, was found to contain filamentous algae
(Lowndes, 1935), while cyclopoid copepods can apparently
coil filaments within their guts (Fryer, 1957). The longest fila-
ment found by Fryer was 380 ^m. Although Burns (1968)
observed that Daphnia could ingest only broken filaments
of Anabaena, Birge (1897) observed that Daphnia "greedi-
ly" ate Anabaena and Aphanizomenon. Lefevre (1950) stated
that Daphnia rejected Aphanizomenon because the filaments
were too stiff to be bent into the mouth. Gliwicz and Siedlar
(1980) found that Daphnia avoided filtering Anabaena by nar-
rowing the carapace gape width, but cited a personal com-
munication from Porter and Korstad that Anabaena has been
used successfully as the main food source for Daphnia.
Porter and Orcutt (1980), however, found that Daphnia
feeding on a toxic strain of Anabaena flos-aquae had limited
growth and a shortened life span. Arnold (1971) found A.
flos-aquae to be ingested at low rates by Daphnia magna,
but 100 percent of that ingested was assimilated. Moreover,
survivorship was high. Feeding on Lyngbya, another relative-
ly stiff filament, caused a decreased filtering rate in several
species of Cladocera (Webster and Peters, 1978). Schindler
(1971) found that Oscillatoria was ingested at moderate to
very high rates while ingestion rates were consistently low
when the animals were fed Anabaena.
The literature for the colonial blue-greens is similarly con-
fused. Anacystis, Aphanocapsa, Coebsphaerium, Gleocap-
sa, Merismopedia, and Microcystis have all been reported
ingested, although not in every case assimilated (Arnold,
1971; Sorokin, 1968; Schindler, 1971). Porter (1973) reported
Microcystis as unaffected by grazing and Chroococcus as
suppressed. Schoenberg (1980) found that the degree of
suppression of Microcystis depended on the size of the
grazer (Fig. 2). Ferguson et al. (1982) also observed ap-
100O-I Lakewater 24 hrs following zooplankton enrichment 1430/I)
600-
c
D 400-
Microcystis>20u
< 20il
Chlorophyta
Cryptophyta
Chrysophyta
'control
Bosmina Daphnia galeata mendotae
0.68 1.14 i 53 mm
Figure 2.—Phytoplankton concentrations 24 hours after additions
of Bosmina longirostris or several size classes of Daphnia galeata
mendotae mean body lengths given below X-axis). Means of two
replicates. After Schoenberg, 1980.
229
-------
Lake Restoration, Protection and Management
preciable grazing on Microcystis by Daphnia hyallna, but only
on colonies less than 60 urn. Similarly, Moriarty et al. (1973)
have measured high ingestion rates and moderate (35-58
percent) assimilation efficiencies of Microcystis by Ther-
mocye/ops hyalinus in Lake George, Uganda.
The confusion of the edibility of blue-greens is also con-
founded by contradictory evidence that the blue-greens are
toxic. Arnold (1971) found a number of blue-green species
to provide either no nutriment or to be toxic. Porter and Or-
cutt (1980) found that a strain of Anabaena toxic to mam-
mals was also toxic to Daphnia magna.Lampert (1981) found
Microcystis aerugjnosa to be highly toxic to Daphnia pulicaria,
while Aphanizomenon gracile and Synechococcus etongatus
actually stimulated the daphnid's growth when mixed with
Scnenedesmus.Conversely, De Bernard! et al. (1981) found
no evidence of toxic effect of Microcystis aeruginosa when
fed to Daphnia obtusa, D. hyatina, or D. cucuttata.
In conclusion, the evidence that blue-green algae are
either too large to be ingested, undigestible because of
gelatinous sheaths, or toxic Is Inconclusive and often con-
tradictory. The reason for the confusion may be because dif-
ferent species of zooplankton were used In the experiments,
that both toxic and non-toxic strains of algae exist, or, as
often was the case, the experimental animals were given
only a single species of food, thus excluding the possibility
that the daphnid could have survived on the alternate species
while still grazing on blue-greens.
In a lake a grazer would seldom be confronted by a pure
culture of blue-green species. Even if the blue-green is nutri-
tionally inadequate, it may still be grazed together with other,
more nutritious species. The possibility also exists that fre-
quent handling of the blue-greens in the filtering mechanisms
of the grazers may break the filaments or colonies into
smaller, more manageable pieces.
The population density of zooplankton has been ob-
served to decline during the summer. In addition, the mean
size of the zooplankton frequently also decreases. If this
decline In abundance and mean size is the result of nutri-
tional inadequacy, toxiclty, or unavailability of the blue-
greens, or the interference of the large forms in the filtering
rate of the grazers (Gliwicz, 1977), then zooplankton could
not reach densities sufficient to reduce algal biomass.
Shapiro (1980) has argued that sufficient evidence exists that
this decline is the result of fish predation on the zooplankton.
As fish predation is selective on the larger zooplankton, the
fish are selectively removing the very species that have the
greatest chance of grazing on the blue-greens (Lynch and
Shapiro, 1981).
According to Shapiro (1979, 1980), an effective method
of reducing blue-greens would be to promote the growth of
the larger zooplankton forms by decreasing the abundance
of planktivorous fish. He cites several field experiments where
fish removal or fish density manipulation has increased the
clarity of the water and decreased blue-green abundance.
The controversy then has centered on whether blue-green
algae are effectively grazed by herbivorous zooplankton. The
evidence derived from laboratory experiments, although
highly variable, does suggest that some blue-greens may
not be ingested and may be toxic. On the other hand,
evidence from field observations and experiments suggest
that, in the absence of planktivorous fish, zooplankton readily
decrease the total abundance of algae.
Schoenberg (1980) suggested an alternate explanation to
direct grazing to explain why the blue-greens are suppress-
ed when fish are removed. In an enclosure experiment in
a shallow, hypereutrophic lake, the fish and zooplankton
were initially killed by carbon dioxide added as dry ice.
Subsequently, monospecific cultures of several species of
Table 1. — Phytoplankton biomass (^m'.mh3 x io«) and relative abundance (% total biomass) In Bosmina and Daphnia-
dominated enclosures, averaged for the final three samples in the study (±S.E.). Asterisk indicates significant difference be-
tween Bosmina and Daphnia enclosures (p -c 0.05). After Schoenberg and Carlson, in prep.
Division
genus
Cyanophyta
Microcystis
Osc///ator/a
Chlorophyta
Scenedes/nus
Coelastrum
Microspora
Chlamydomonas
Cosmarium
Qocystis
Chrysophyta
Maltomoms
Euglenophyta (1 sp.)
Trachebmonas
Cryptophyta
Ciyptomonas
Rhodomonas
Pyrrhophyta (1 sp.)
Ceratium
Total cell volume
Total chlorophyll a
Narto chlorophyll a
% Nano chlorophyll a
Daphnia
45.1 (±68.7)
44,8(±68.8)
.01 (±.01)
21.8(±29.6)
13.8(124.4)
5.6(±5.6)
1.8(±2.7)
,fl(±1.1)
.44(±.88)
.12(1.22)
,64<±.89)
.16(1.40)
.13(1.17)
31.1(150.1)
30.9(±50.2)
.1(11.7)
6.6(14.4)
100.0(189.3)
31.9(±25.3)
8.5(18.5)
24.8(18.7)
Biomass
Bosmina
* 457.0(1403.6)
' 455.6(± 403.5)
.01(±.03)
8.4(113.1)
3.2(15.4)
* 0.0
.12(1.33)
2.6(16.8)
0.0
.04(1.08)
14.0(119.6)
8.7(±9.6)
* 3.2(14.0)
50.5(1134.0)
49.4(1134.4)
1.2(±1.3)
9.6(19.8)
* 542.9(1395.1)
* 96.0(171.4)
12,2(112.2)
* 13.6(15,4)
Relative
Daphnia
46.5(±29.6)
39.6(119.6)
.03(1.02)
19.3(114.3)
8.4(1 10.9)
4.8(±4.8)
4.5(19.0)
.6(11.2)
.4(1.4)
.13(1.14)
1.7(±3.0)
.08(1.20)
.12(1.59)
20.5(123.4)
20.2(124.0)
.16(1.29)
19.2(116.5)
Abundance
Bosmina
80.8(123.7)
80.7(123.7)
.003(1.0)
2.0(12.3)
.8(11.1)
0.0
.01(1.04)
.5(11.1)
* 0.0
* .01(1,02)
3.5(15.6)
* 1.4(12.2)
.66(1.71)
9.3(120.7)
8.7(120.9)
.58(177)
* 4.0(16.5)
230
-------
Causes and Control of Blue-Greens: Alternatives to Nutrient Control
Daphnia and Bosmina longirostris were introduced to
separate enclosures. The effect of the dry ice addition was
to produce a low pH and low percentages of blue-greens,
as observed also by Shapiro (1973).
Over the following weeks, the enclosures dominated by
Bosmina recovered to the pre-experiment densities of blue-
greens. The Daphnia enclosures, however, retained low den-
sities of algae and lower percentages of blue-greens (Table
1). The dominant blue-greens, Microcystis and
Aphanizomenon, were effectively suppressed.
Other than the possibility of direct grazing as the reason
for the control of the blue-greens, Schoenberg suggested
that the carbon dioxide treatment and the subsequent graz-
ing of Daphnia maintained a set of ambient environmental
conditions that were not conducive to the growth of blue-
green algae. The Daphnia enclosures had significantly
lowered primary productivity which did not promote the in-
crease in pH, nor did the density of the algae decrease the
transparency to levels that would provide the low-light en-
vironment suggested necessary to promote blue-green
growth (Table 2).
In this experiment the environmental conditions were in-
itially set by the carbon dioxide treatment, and the daphnids
apparently maintained these conditions, but Schoenberg
suggested that prior manipulation may not have been
necessary. Although the blue-greens may dominate the algal
biomass, most of the primary productivity in eutrophic lakes
is effected by the smaller algae (Rodhe, 1958; Kalff, 1972).
In addition, transparency also is largely determined by the
smaller particles in the water (Edmondson, 1980). Intense
selective grazing by herbivores on these smaller particles
should then decrease primary productivity and increase the
ambient light to the water column disproportionately to the
amount of total algal biomass consumed. If light and pH are
major determinants of blue-green dominance, then the blue-
greens should decline, even though they are not themselves
grazed (Fig. 3).
The question then, may not be whether zooplankton can
successfully graze on blue-green algae, but whether blue-
green abundance can be controlled by both the direct and
LARGE ZOOPLANKTOH -
t
BLUE-GREENS
TOTAL P v
PRIMARY I
PRODUCTIVITY
Figure 3.—A possible mechanism for the control of blue-green
algae by both direct grazing on the algae and by alteration of the
ambient environment by the grazing activities on non-blue-green
species.
the indirect effects of grazing. This question does not directly
concern itself with the edibility or even toxicity of blue-greens,
but forces a consideration of whether or not, in the presence
of abundant grazers, blue-green algae in lakes are sup-
pressed. Field observations and experiments suggest that
such control is possible.
FEASIBILITY AND LIMITS OF CONTROL
BY GRAZING
If blue-green control by grazing is possible, the next ques-
tion must be that of feasibility. Feasibility has several aspects.
The biomanipulation of grazing intensity must be
technologically possible, the results must be at least
ecologically significant, if not observable to the users, and
the effect must be either long-lasting or maintainable at a
cost less than that of alternative methods.
Techniques for enhancing zooplankton grazing involve
decreasing the pressure of the zooplanktivorous fish on the
zooplankton community. Removal of fish predation increases
Table 2. — Physicochemical characteristics of Crystal Lake, Akron, Ohio, and the control, Daphnia, and Bosmina
enclosures, averaged of the final three samples of the study (±S.E.). Asterisk indicates significant
difference between Daphnia and Bosmina enclosures (p < 0.05). After Schoenberg and Carlson, in prep.
Daphnia
Bosmina
Control
Lake
Secchi depth (cm)
PH
2 meter dissolved O2 (mg liter ~1)
total phosphorus (^g P liter-')
paniculate P (^g P liter1)
soluble unreactive P (ng P liter1)
soluble reactive P fag PO4 liter-1)
% of total P as soluble P
(reactive + unreactive
% of total P as soluble reactive P
118.8
±66.5
6.6
±0.3
4.1
±2.2
98.9
±40.5
41.2
±33.9
36.7
±11.7
18.7
±21.0
52.3
±27.0
15.9
±15.9
27.4
±23.0
8.1
±1.2
1.5
±1.3
234.9
±100.3
181.0
±99.6
43.2
±8.1
11.9
±3.6
26.9
±11.2
5.9
±3.0
32.7
±4.9
9.1
±0.3
2.7
±2.0
221.3
±64.5
136.8
±55.9
60.9
±10.0
16.9
±8.0
37.2
±12.9
7.8
±2.6
38.7
±9.2
9.1
±0.3
1.5
±0.5
164.6
±11.9
100.7
±10.3
41.9
±17.7
22.0
±15.1
38.8
±4.2
13.0
±8.2
231
-------
Lake Restoration, Protection and Management
both zooplankton abundance and the mean size of the
zooplankton (Shapiro and Wright, unpubl). Techniques to
reduce the predation pressure of the zooplanktivores might
include using fish poisons such as rotenone (Anderson, 1970;
Henrikson et al. 1980; Shapiro and Wright, unpubl.) or tox-
aphene (Galbraith, 1967), and the subsequent reestablish-
ment of a fish community that does not use primarily
zooplankton as food, or the addition of carnivorous fish such
as walleye or northern pike. Shapiro (1979) suggested that
hypolimnetic destratification by aeration would increase the
volume available for zooplankters and decrease the impact
of fish predation (Haney, 1973). If macrophytes serve as
refuges for zooplanktivores as suggested by Kerfoot (1974),
then reduction in the extent of macrophytes might also
decrease the abundance and impact of zooplanktivores.
Another possibility that has not as yet been explored is
the introduction of larger forms of zooplankton. Such a
manipulation would be predicated on, first, the reduction of
fish predation and, second, the demonstrated absence of
larger forms within the lake. Often the larger zooplankton
are extremely rare in the presence of intense fish predation,
but will increase in abundance when that predation is removp
ed (Schoenberg, 1980). However, such introductions may
not be sucessful. Schoenberg (1980) attempted to introduce
monospecific cultures of Daphnia magna, D. pulex, and D.
galeata mendotae to enclosures in a Sosm/na-dominated
lake. Neither D. magna nor D. pulex survived, but D. galeata,
which was rare in the lake (only one specimen was found
in tows of the lake during the entire summer), reached high
densities within some enclosures.
For grazing manipulation to be used, there must be some
criterion for success. Certainly one might wish that grazers
could give all lakes the transparency of Lake Superior, but
a more realistic criterion might be to gage success in terms
of the expense of biomanipulation relative to nutrient diver-
sion or nutrient inactivation. To halve the nutrient input to
a lake should decrease chlorophyll by approximately 50 per-
cent and double transparency, assuming transparency is
largely a function of chlorophyll. If zooplankton manipula-
tion could bring about equivalent effects for less money than
nutrient diversion, then the technique should be considered
successful. The results of several studies (Shapiro, 1980;
Shapiro and Wright, unpubl.) suggest that it is not unrea-
sonable to expect the transparency to double.
Finally, the permanency of zooplankton manipulation must
be addressed. The zooplankton must be able to continue
to control the algae for more than short periods during the
summer. It is possible that shifts in the algal composition
brought about by selective zooplankton grazing may bring
about an algal community that cannot be grazed. There is
no doubt that some algae are too large to be grazed, or are
sheathed, or distasteful. These species may not be blue-
greens, as other algal groups besides blue-greens also have
these forms, but the possible effect of decreased grazing
efficiency is the same. Some sheathed algae, such as the
green alga, Gleocystis, may actually be stimulated by graz-
ing (Porter, 1976). Lynch (1980) and Shapiro (1980) have
documented cases where the large flake-like forms of
Aphanizomenon appear during heavy grazing by Daphnia.
It is possible that intense grazing would shift algal composi-
tion entirely to non-ingestible, distasteful, and sheathed forms
(Fig. 4), although these forms may not be blue-greens.
If algal compositional shifts are possible with selective
grazing, the effectiveness of zooplankton control of algae
may depend on the efficiency with which the grazers can
reduce nutrients limiting algal growth. Nutrients excreted by
zooplankton will again be available for uptake by algae and
will stimulate algal productivity (Gliwicz, 1975). With selec-
tive grazing, the inedible forms of algae would predominate.
However, in some instances, the amount of nutrients in the
open water has been observed to decline in the presence of
intense grazing (see Table 2, and Shapiro and Wright, un-
publ.). A certain amount of nutrients will be sequestered in
the bodies of the zooplankton themselves, and an increase
in mean body size may increase the turnover time of
nutrients, as was suggested by Henrikson et al. (1980).
Another explanation for the reduction of nutrients with graz-
ing may be their translocation out of the epilimnion by either
sedimentation of fecal material or by excretion of nutrients
in the metalimnion by diurnally migrating Daphnia (Shapiro
and Wright, unpubl.). In shallow lakes the sediments may
SIZES TOO LARGE
OR TOO SMALL
TO BE FILTERED
ASSIMILATION
PARTICIPATE MATTER
IN WATER
ALGAE
BACTERIA
DETRITUS
AMOUNT
FILTERED
INGESTION
PARTICLES REJECTED
BECAUSE OF SIZE
OR TASTE
EGESTION
DETRITUS
UNDIGESTED
' ALGAE
Figure 4.—Biotic feedback mechanisms in grazing, illustrating the
possible accumulation of nonfilterable, distasteful, indigestible algal
forms in the water during intense grazing.
sorb the excreted phosphorus, in effect, competing with the
algae for phosphorus (Carlson, 1976). Large zooplankton
may facilitate this process by keeping the ratio of soluble
reactive phosphorus to total phosphorus high (Table 2).
Controlling blue-green algae by zooplankton grazing ap-
pears to be a promising alternative to nutrient-based con-
trol techniques. It should be less expensive than construc-
ting and maintaining a tertiary sewage treatment plant, and,
as it diverts energy from the detrital chain into the grazing
chain, it may both decrease hypolimnetic oxygen depletion
and increase sport fishing. As zooplankton species composi-
tion appears to change with hypolimnetic aeration or weed
harvesting, conscious manipulation of the fish species in con-
junction with either of these techniques may enhance their
benefits. Certainly we still do not know the exact mechanisms
by which control by grazing works or the extent to which
its success can be predicted in any given lakes, but this in-
formation can be obtained only as the technique is applied.
It is our hope that repeated demonstration of the success
of the technique will eventually break down the conceptual
biases that have obstructed its widespread use.
REFERENCES
Aberg, B., and W. Rodhe. 1942. Uber die Milieufaktoren in einigen
sudschwedischen Seen. Symb. Bot. Ups. 5:1-256.
Anderson, R. S. 1970. Effects of rotenone on zooplankton communi-
ties and a study of their recovery patterns in two mountain lakes
in Alberta. J. Fish. Res. Board Can. 27:1335-1356.
Arnold, D. E. 1971. Ingestion, assimilation, survival, and reproduc-
tion by Daphnia pulex fed seven species of blue-green algae Lim-
nol. Oceanogr. 16:906-920.
Bartsch, A. F. 1969. Discussion. Page 117 in Modeling the Eutro-
phication Process. Proc. Symp. U.S. FWQA and Dep. Environ
Eng., Univ. Florida, Gainesville.
232
-------
Causes and Control of Blue-Greens: Alternatives to Nutrient Control
Birge, E. A. 1897. Plankton studies on Lake Mendota. II. Crustacea of
plankton from July 1894 to December 1896. Trans. Wis. Acad.
Arts Sci. 11:274-448.
Brown, T. E., and F. L. Richardson. 1968. The effect of growth en-
vironment on the physiology of algae: light intensity. J. Phycol.
4:38-54.
Burns, C. W. 1968. Direct observations of mechanisms regulating
feeding behavior of Daphnia in lakewater. Int. Rev. Gesamten
Hydrobiol. 53:83-100.
Carlson, R. E. 1976. Phosphorus cycling in a shallow eutrophic lake
in southwestern Minnesota. Ph.D. Dissertation, Univ. Minnesota.
De Bernardi, R., G. Giussani, and E. L. Pedretti. 1981. The signifi-
cance of blue-green algae as food for filter feeding zooplankton:
experimental studies on Daphnia spp. fed by Microcystis
aeaig/nosa.Verh. Int. Verein. Limnol. 21: 477-483.
Edmondson, W. T. 1980. Secchi disk and chlorophyll. Limnol.
Oceanogr. 25:378-379.
Ferguson, A. J. D., J. M. Thompson, and C. S. Reynolds. 1982.
Structure and dynamics of zooplankton communities maintained
in closed systems with special reference to the algal food supp-
ly. J. Plankton Res. 4:523-543.
Fryer, G. 1957. The food of some cyclopoid freshwater copepods
and its ecological significance. J. Anim. Ecol. 26:263-286.
Galbraith, M. G. 1967. Size-selective predation on Daphnia by rain-
bow trout and yellow perch. Trans. Am. Fish. Soc. 96:1-16.
Gliwicz, Z. M. 1975. Effect of zooplankton grazing on photosynthetic
activity and composition of phytoplankton. Verh. Int. Verein. Lim-
nol. 19:1490-1497.
. 1977. Food size selection and seasonal succession of filter
feeding zooplankton in an eutrophic lake. Ekol. Pol. 25:179-225.
Gliwicz, Z. M., and E. Siedlar. 1980. Food size limitation and algae
interfacing with food collection in Daphnia. Arch. Hydrobiol.
88:155-177.
Haney, J. F. 1973. An in situ examination of the grazing activities
of natural zoopiankton communities. Arch. Hydrobiol. 72:87-132.
Henrikson, L., H. G. Nyman, H. G. Oscarson and J. A. E. Stenson.
1980. Trophic changes, without changes in external nutrient
loading. Hydrobiologia 68:257-263.
Kalff, J. 1972. Net plankton and nanoplankton production and bio-
mass in a north temperate zone lake. Limnol. Oceanogr.
17:712-720.
Kerfoot, C.W. 1974. Net accumulation rates and the history of clado-
ceran communities. Ecology 55:51-61.
King, D.L. 1970. The role of carbon in eutrophication. J. Water Pollut.
Control Fed. 42:2035-2051.
Kratzer, C. R. and P. L. Brezonik. 1982. Reply to discussion by
Richard A. Osgood, A Carlson-type trophic state index for nitrogen
in Florida lakes. Water. Res. Bull. 18:543-548.
Lampert, W. 1981. Inhibitory and toxic effects of blue-green algae on
Daphnia. Int. Rev. ges. Hydrobiol. 66:285-298.
Lefevre, M. 1950. Aphanizomenon gracile Lem. Cyanophyte de'-
favorable au zooplankton. Ann. de la Stat. Cent. d'Hydrobiol. Ap-
plique. 3:205-208.
Long, E. B. 1976. The interaction of phytoplankton and the bicar-
bonate system. Ph.D. Dissertation. Kent State Univ., Kent, Ohio.
Lowndes, A. C. 1935. The swimming and feeding of certain calanoid
copepods. Proc. Zool. Soc. London 1935:687.
Lynch, M. 1980. Aphanizomenon blooms: alternate control and culti-
vation by Daphnia pulex. Pages 299-304 in W.C. Kerfoot, ed. Evolu-
tion and Ecology of Zooplankton Communities. ASLO Spec. Symp.
Vol 3. Univ. Press of New England, Hanover, N.H.
Lynch, M., and J. Shapiro. 1981. Predation, enrichment, and phyto-
plankton community structure. Limnol. Oceanogr. 26:86-102.
Moriarty, D. J. W., et al. 1973. Feeding and grazing in Lake George,
Uganda. Proc. R. Soc. Lond. B. 184:227-346.
Porter, K.G. 1973. Selective grazing and differential digestion of algae
by zooplankton. Nature 244:174-180.
1976. Enhancement of algal growth and productivity by
grazing zooplankton. Science 192:1332-34.
Porter, K. G., and J. D. Orcutt. 1980. Nutritional adequacy, manage-
ability, and toxicity as factors that determine the food quality of
green and blue-green algae for Daphnia. In W.C. Kerfoot, ed.
Evolution and Ecology of Zooplankton Communities. ASLO Spec.
Symp. Vol 3. Univ. Press of New England, Hanover, N.H.
Rodhe, W. 1958. The primary production in lakes: some results and
restrictions of the C-14 method. Rapp. Proc. Verb. Cons. Perm.
Int. Explor. Mer. 14:122-128.
. 1969. Crystallization of eutrophication concepts in northern
Europe. In Eutrophication: Causes, Consequences and Correc-
tives. Natl. Acad. Sci., Washington, D.C.
Schindler, J. E. 1971. Food quality and zooplankton nutrition. J. Anim.
Ecol. 40:589-595.
Schoenberg, S. A. 1980. The effects of zooplankton species manipu-
lation on the phytoplankton of a hypereutrophic lake. M.S. Thesis,
Kent State Univ., Kent, Ohio.
Schoenberg, S. A., and R. E. Carlson. In prep. Kent State Univ.,
Kent, Ohio.
Shapiro, J. 1973. Blue-green algae: why they become dominant.
Science 179:382-384.
1979. The need for more biology in lake restoration. Pages
161-167 in Lake Restoration, Proc. Natl. Conf., Aug. 22-24, 1978.
EPA 440/5-79-001 U.S. Environ. Prot. Agency, Washington, D.C.
. 1980. The importance of trophic-level interactions to the
abundance and species composition of algae in lakes. Pages
105-116 in J. Barica and LR. Mur, eds. Hypertrophic Ecosystems.
W. Junk, The Hague.
Shapiro, J., and D. I. Wright. 1982. Lake restoration by biomanipula-
tion: Round Lake, Minn. Unpubl.
Sorokin, S. T. 1968. Carbon-14 method in the study of the nutrition
of aquatic animals. Mitt. Int. Verein. Limnol. 16:1-41.
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. Devel., Paris.
Wall, D., and F. Briand. 1979. Response of lake phytoplankton com-
munities to in situ manipulations of light intensity and colour. J.
Plankton Res. 1:103-112.
Webster, K. E., and R. H. Peters. 1978. Some size-dependent inhibi-
tions of larger cladoceran filterers in filamentous suspensions. Lim-
nol. Oceanogr. 23:1238-1245.
233
-------
Lake Restoration, Protection and Management
CYANOBACTERIAL BUOYANCY REGULATION AND BLOOMS
ANDREW R. KLEMER
Division of Natural Sciences
State University of New York-Purchase
Purchase, New York
ABSTRACT
Cyanobacterial blooms are associated with high degrees of relative gas vacuoiation (RGV). RGV is sub-
ject to regulation and surface scums may result from an impairment of the regulatory mechanism(s). The
regulatory mechanism of at least some major bloom-formers involves an interaction between light and
certain limiting nutrients. The effects of limiting nutrients on buoyancy depend on the nutrients involved.
Surface accumulations or blooms of blue-green algae oc-
cur as a result of excessive buoyancy, buoyancy resulting
largely from a high degree of relative gas vacuoiation (RGV).
Other cell constituents such as reserves of carbohydrates
may have significant effects on the specific gravity of cells,
but there is clear evidence that the degree of RGV can be
the decisive or overriding factor. Figure 1 shows the distribu-
tion of filaments of Oscillatoria agardhii var. isothrix in test
tubes after having different proportions of their gas vesicles
collapsed. With none of their gas vesicles collapsed, most
of the filaments remained in the top third of the test tube
after 20 hours in the dark at 12° C. But with 25 percent of
their gas vesicles collapsed, most of the filaments occurred
in the bottom third of the tube, and with 100 percent col-
lapsed, 98 percent were in the bottom and only 0.1 percent
in the top third.
20
80
30 40 SO 60 70
GAS VACUOLES
COLLAPSED, % OF INITIAL LEVEL
Figure 1 .—The distribution of filaments of Oscillatoria agardhii var.
isothrix in test tubes after the collapse of different proportions of
their gas vesicles. The test tubes were partitioned after 20 hours
in darkness at 12 C.
BUOYANCY REGULATION
Blue-green algae use their gas vacuoles (aggregates of the
minute protein capsules called gas vesicles) not only to float
to the surface, but, in some cases, also to accumulate at
some depth in the water column. Studies of the latter
phenomenon indicate a high degree of buoyancy regulation.
For example, Walsby and Klemer (1974) took O. agardhii var.
isothrix from a metalimnetic layer in Deming Lake, Minn.,
and resuspended the alga in 1-liter bottles at 1-meter inter-
vals with and without additional nitrogen (1mgNH4-N/L). After
2 days, buoyancy status determinations indicated a buoyancy
decrease in alga suspended above the metalimnetic layer
and, at least in N-enriched alga, a buoyancy increase in those
suspended below the layer. At every depth, the alga was
more buoyant with added nitrogen.
MECHANISMS OF BUOYANCY
REGULATION
Buoyancy regulation in blue-green algae has been related
to changes in turgor pressure and to changes in relative rates
of growth and gas-vesicle synthesis. The first environmen-
tal factor to be implicated in such changes was light inten-
sity. Walsby (1969) found Anabaena flos-aquae became more
buoyant at low light intensities. Low light intensities yield low
rates of photosynthesis and growth, physiological conditions
associated with gas vesicle accumulation and buoyancy in-
creases (Walsby, 1970). Light intensities, appropriate for such
buoyancy increases, occur at the base of the photic zone
in the stable portion of a water column (i.e., in the metalim-
nion) or in the mixed portion of a water column if the mean
light intensity is low enough. Once positively buoyant in a
stable water column, an alga could move up into higher light
intensities that would promote higher rates of photosynthesis
and a loss of buoyancy.
Walsby (1970) considered two possible mechanisms by
which RGV would decrease with increased light intensity.
The alga's gas vesicles might be diluted out by rapid growth,
and/or weaker gas vesicles might be collapsed by turgor in-
creases resulting from "excess" photosynthesis (photosyn-
thesis in excess of nutrient-limited growth) (Walsby, 1971;
Dinsdale and Walsby, 1972).
Later, Grant and Walsby (1977) demonstrated that the ac-
cumulation of osmotically-active photosynthate contributes
to the rise in turgor pressure, and Allison and Walsby (1981)
discovered that potassium accumulation is also a factor. This
mechanism allows for light-nutrient interactions in that the
availability of limiting nutrients would determine the rate of
photosynthate assimilation and, hence, the light level at
which photosynthesis became excessive and photosynthate
began to accumulate.
LIGHT, LIMITING NUTRIENTS, AND
BLOOM FORMATION
Given these mechanisms, how could an alga remain at the
surface? One possibility is that the mechanisms could fail.
If photosynthesis ceased or perhaps failed to increase with
light intensity as the alga neared the surface, a reduction
in RGV by increased turgor pressure or growth rate would
not occur and the alga could become lodged at the surface.
Photoinhibition of photosynthesis might do this, but so might
nutrient limitation. The role of nutrient limitation in this pro-
cess is not as unambiguous as that of light appears to be.
C, N, and P have all been suspected of limiting growth or
photosynthesis in such a way that RGV either failed to
234
-------
Causes and Control of blue-ureens: Alternatives to Nutrient control
decrease or actually increased (Reynolds and Walsby, 1975).
In so doing, such nutrient limitation would allow an alga to
remain buoyant despite the increases in light intensity it
would experience on its way to the surface. In fact, the light
effects on buoyancy in blue-green algae depend upon the
nutritional status of the alga. When N-limited, O. rubescens
did not show a positive buoyancy response to reduced light
intensity (Klemer, 1978); and when COrdeprived, A. f/os-
aquae showed no reduction in RGV upon exposure to higher
light intensity (Dinsdale and Walsby, 1972).
NUTRIENT-SPECIFIC EFFECTS ON
BUOYANCY
The effect of nutrient limitation on buoyancy depends on the
nutrient involved. RGV in O. rubescens grown in N-limiting
chemostats decreased with transitions to more severe N
limitation (Klemer, 1978). In subsequent work with O.
rubescens in N-limiting cyclostats (see Fig. 2), we observed
higher RGV at higher growth (dilution) rates (Klemer,
Feuillade and Feuillade, unpubl.). The latter work involved
continuous cultures in steady state, not in transition, and the
increase in RGV with dilution rate and N availability suggests
that, in Oscillatoria, RGV is more a function of nutritional
status than of absolute growth rate. In other words, the
growth-rate hypothesis of gas vacuole regulation does not
seem to apply to N-limited Osc///atora.This conclusion is sup-
ported by the positive buoyancy responses of natural popula-
tions of O. agardhii var. isothrix to N enrichment (Klemer,
1976; Walsby and Klemer, 1974).
On the other hand, in cyclostat cultures of O. rubescens,
RGV did increase with transitions from N-limiting to
1.7
1.6
K 1-5
< -1
O . 1.4
So ,.,
<< 1.2
111 D
EC O 1.1
0.10 0.12 0.14 0.16 0.18 0.20 0.22 0.24 0.26 0.28 0.30 0.32 0.34 0.36
DILUTION RATE PER DAY
Figure 2.—Relative gas vaculation of Oscillatoria rubescens in
steady state cultures at different dilution rates.
CO2-limiting conditions (Klemer, Feuillade and Feuillade,
1982). Whereas, Dinsdale and Walsby (1972) had
demonstrated that CO2 deprivation could impair the turgor-
collapse mechanism and prevent a light-induced reduction
in the RGV of A. flos-aquae, our work with O, rubescens
showed that RGV could actually increase with CO2 depriva-
tion and could do so despite increases in mean light inten-
sity as the cultures thinned out. Transitions from nitrogen
to CO2-limiting conditions permit increases in relative pro-
tein synthesis, including gas vesicle protein synthesis.
INORGANIC CARBON AVAILABILITY AND
BLOOM FORMATION
As early as 1970, Walsby had indicated the possible
significance of the fact that bloom-forming algae have the
advantage of better access to atmospheric CO2. Recently,
Paerl and Ustach (1982) showed that Aphanizomenon flos-
aquae and Anabaena circinalis photosynthesize faster with
CO2 than with other forms of inorganic carbon and suggested
that surface scum formation is a means of assuring access
to the preferred form. Klemer and Graver (unpubl.) compared
the effects of HCOl and atmospheric CO2 on RGV in
COj-deprived O. rubescens and observed reductions that dif-
fered only in the speed of the response. Alga with access
to atmospheric CO2 showed decreases in RGV within 6
hours whereas alga supplied with HCOs (50 mg C/l) in
capped vials achieved slightly greater reductions in the in-
terval between 7 and 29 hours. We are currently preparing
to obtain more accurate estimates of the response times.
In experiments with Anabaena flos-aquae, Booker and
Walsby (1981) found that NaCO3 enrichment reduced the
buoyancy of the alga in stoppered bottles as well as the
amount of surface accumulation in a laboratory water col-
umn. In lake experiments, algal responses to HCO~3 enrich-
ment have varied. A natura[ bloom of Anabaena spp. was
much more reduced in HCGvenriched (40 mg C/l) columns
than in control columns of lake water (Klemer and Brasino,
in prep.). In contrast, the combination of HCO~3, N and P in-
duced more intense blooms of O. agardhii var. isothrix than
did the combination of N and P. In the latter study (Klemer,
Pierson and Whiteside, in prep.) we initially induced surface
blooms by enriching only the metalimnetic portion of isolated
water columns with N and P (10 mg NH4-N and 1 mg
PO4-P/I). In those experiments, Oscillatoria began to move
up from the metalimnion before the transparency of the sur-
face water was significantly reduced. This tends to rule out
a buoyancy response to reduced light intensity, at least as
a primary effect.
We considered the alternative possibility that N and P
enrichment induced C limitation in these metalimnetic
populations and, thereby, enabled the alga to remain buoyant
as it approached the surface and higher light intensities. If
this were the case and if Oscillatoria were to respond as Ana-
baena had responded, HCOl enrichment would, so we
thought, deter bloom formation. In an early experiment in
which N and P were added only to the metalimnetic portion
of treated water columns, HCOs enrichment (50 mg C/l) at
the surface was followed by a more rapid collapse of the
treated bloom than of the bloom in a column with no C add-
ed. However, collapse occurred only after the C-enriched
bloom had intensified.
In a subsequent experiment in which N was provided
throughout water columns while P was added only to
metalimnetic layers, HCOs enrichment increased N utiliza-
tion and intensified blooms instead of deterring their forma-
tion or causing their collapse. This study is still in progress,
but the data in hand seem to warrant some tentative con-
clusions. The amount of N and P available may not only
determine the likelihood of C limitation of photosynthetic
rates, but it may also determine the nature and magnitude
of buoyancy responses by C-limited blue-green algae to in-
creased C availability. If N and P availability is marginal, in-
creased C availability will probably elicit a relatively rapid
reduction in buoyancy. However, if N and P are present in
excess, increases in C will probably increase N and P utiliza-
tion and intensify blooms, perhaps to the point that self-
shading becomes a major factor in maintaining buoyancy.
This sort of interaction would have effects on buoyancy
similar to those of the light-limiting nutrient interaction
predicted by the turgor-collapse mechanism (Dinsdale and
Walsby, 1972; Walsby and Klemer, 1974). Of course, even
species of freshwater blue-green algae can be expected to
show some variation in their mechanisms of buoyancy
regulation.
As for the best way to deal with water blooms, I still favor
reversing the eutrophication process in some fundamental,
long-term manner (e.g., by reducing N and P loading). Where
a reduction in overall nutrient loading is not feasible, my bias
is towards some form of biomanipulation, either to reduce
the utilization of key nutrients or to direct the course of suc-
cession away from dominance by noxious bloom-formers.
235
-------
Lake Restoration, Protection and Management
In addition to its effects on pH and the available form of car-
bon, artificial destratificatlon may discourage blue-green
algae because it reduces not only the competitive advan-
tage with respect to light and nutrients that buoyancy regula-
tion confers under stratified conditions, but also the actual
buoyancy of blue-greens that might otherwise be COz-limited.
However, even if the blue-greens are CO2-limited, the
buoyancy response to COj enrichment may well depend on
how much N and P are available,
REFERENCES
Allison, EM, and A.E. Walsby. 1981. The role of potassium in the
control of turgor pressure in a gas-vacuolate blue-green alga, J,
Exp, Bot. 32: 241-249.
Booker, M.J., and A.E. Walsby, 1981. Bloom formation and strati-
fication by a planktonic blue-green alga in an experimental water
column, Br. Phycol. J. 16: 411-421.
Dinsdale, M.T., and A.E. Walsby. 1972. The interrelations of cell
turgor pressure, gas-vacuolation, and buoyancy in a blue-green
alga. J. Exp. Bot. 23: 561-570.
Grant, N.G., and A.E. Walsby. 1977, The contribution of photosyn-
thate to turgor pressure in the planktonic blue-green alga Ana-
baena flos-aquaa.J, Exp. Bot. 28: 409-415,
Klemer, A,R. 1976. The vertical distribution of Oscillator/a agardhii
var. isothfix. Arch. Hydrobiol. 78: 343-362.
1978. Nitrogen limitation of growth and gas vacuolation in
Oscillatoria rabescens.Verh. Int. Ver. Theor. Angew. Limnol. 20:
2293-2297.
Klemer, A.R., and T.L. Brasino. Carbon enrichment hastens the col-
lapse of isolated portions of an Anabaena bloom. (In prep.)
Klemer, A.R., J. Feuillade, and M. Feuillade. 1982. Cyanobacterial
blooms: carbon and nitrogen limitation have opposite effects on
the buoyancy of Oscillatoria. Science 215: 1629-1631.
Klemer, A.R., D. Pierson, and M.D. Whiteside. Limiting nutrients in
the vertical distribution of Oscillatoria. (In prep.)
Paerl, H.W., and J.F. Ustacn. 1982. Blue-green algal scums: An
explanation for their occurrence during freshwater blooms. Lim-
nol. Oceanogr. 27: 212-217.
Reynolds, C.S., and A.E. Walsby. 1975. Water-blooms. Biol. Rev.
50: 437-481.
Walsby, A.E. 1969. The permeability of the blue-green algal gas-
vacuole membrane to gas. Proc. R. Soc. B. 173: 235-255.
1970. The nuisance algae: Curiosities in the biology of plank-
tonic blue-green algae. Water Treat. Exam. 19: 359-373.
_. 1971. The pressure relationship of gas vacuoles. Proc. R.
Soc. B. 178: 301-326.
Waisby, A. E.. and A.R. Klemer. 1974. The role of gas vacuoles
in the microstratification of a population of Oscillatoria agardhii var.
isothrix in Doming Lake, Minn. Arch. Hydrobiol. 74: 375-392.
236
-------
THE NITROGEN AND PHOSPHORUS DEPENDENCE OF BLUE-GREEN
ALGAL DOMINANCE IN LAKES
VAL H. SMITH
Department of Biology
McGill University
Montreal, Quebec, Canada
ABSTRACT
The relative proportion of blue-green algae in the epilimnetic phytoplankton of 15 natural lakes was found
to depend on the total nitrogen:total phosphorus (TN:TP) ratio in the lake water. In general, blue-greens
tended to dominate in lakes where TN:TP-<29, and to be rare when the TN:TP ratio exceeded this value.
These data support the hypothesis that nutrient competition is a major factor structuring natural phytoplankton
communities.
Our ability to manage algal biomass and transparency in
lakes has been greatly improved by the recent development
of phosphorus loading models of eutrophication (e.g.
Vollenweider 1968,1976; Dillon and Rigler 1974,1975; Jones
and Bachmann, 1976). However, our ability to predict
changes in the plankton that accompany changes in nutrient
loading is still minimal. In particular, it is not yet possible to
predict whether a given lake restoration technique will suc-
cessfully reduce the proportion of bloom-forming blue-green
algae (Cyanophyta) in the epilimnetic phytoplankton.
The N:P ratio of the water has been discussed as one im-
portant determinant of algal species composition since the
early work of Pearsall (1932) and Hutchinson (1944).
Schindler (1977) has shown that low N:P ratios in the nutrient
supply of lakes should favor nitrogen-fixing blue-green algae,
and subsequent field studies have supported this hypothesis
(Barica et al. 1980; Flett et al. 1980; Leonardson and Ripl,
1980). Similarly, Niemi (1979) has suggested that low N:P
ratios are responsible for blue-green blooms in the Baltic Sea.
Furthermore, a recent analysis of 90 north temperate lakes
by Reckhow and Simpson (1980) suggested that the con-
centration of inorganic nitrogen is an important control
variable in determining the dominance of blue-greens in high
alkalinity, eutrophic lakes. In this paper, I report a dramatic
tendency for blue-green algal blooms to occur when N:P
ratios fall below about 29.1 by weight, and for blue-greens
to be rare when the N:P ratio exceeds this value.
MATERIALS AND METHODS
To quantitatively test the hypothesis that low N:P ratios
should promote dominance by blue-green algae, I com-
piled data from 15 natural lakes worldwide. For each lake,
I calculated the relative proportion of blue-green algae, on
a biomass basis, in the epilimnion during the growing season.
I chose cell volume or cellular weight as a biomass measure
in all cases, since I believe that the use of cell numbers in
calculating dominance neglects important variations in cell
volume that exist among phytoplankton species. I also at-
tempted to obtain data for annual mass loadings of nitrogen
and phosphorus, and growing season means of epilimnetic
total nitrogen (TN) and total phosphorus (TP) concentrations
for each lake.
RESULTS
I first examined data from Lake Washington (USA), whose
plankton has changed dramatically in response to the diver-
sion of sewage inputs (Edmondson and Lehman, 1981). A
plot of the proportion of filamentous blue-green algae in the
phytoplankton vers_us loading TN:TP ratios (Fig. 1A) and
loading NO3~N:PO4"P ratios (Fig. 1B) reveals that these blue-
greens are strongly correlated with low N:P ratios. As the
N:P loading ratio of Lake Washington increased following
nutrient diversion, the relative abundance of filamentous blue-
green algae dropped sharply once an apparent "threshold"
ratio was attained. As can be seen in Fig. 1C, the propor-
tion of filamentous blue-greens is related to the epilimnetic
TN:TP ratio in an almost identical manner.
Lake Trummen, Sweden, provided a second example of
the response of phytoplankton to changes in the N:P ratio.
Although sewage and wastewater were diverted from Lake
Trummen in 1958, the lake did not recover until suction
dredging was used to remove the nutrient-rich upper
sediments in 1970-1971 (Bjork, 1979; Cronberg, 1980). As
in the case of Lake Washington, a sharp decline in the pro-
portion of blue-green algae accompanied the increase in the
epilimnetic TN:TP ratio in Lake Trummen (Fig. 2). However,
note that in this case all blue-greens were included in
calculating their proportion in the phytoplankton.
Because nutrient loading data were not available for most
of the remaining lakes, I examined the N:P hypothesis, fur-
ther using the epilimnetic TN:TP ratio. When the data from
Lake Trummen and the 13 remaining lakes were plotted (Fig.
3), an apparent boundary was evident between a region
where blue-greens tended to dominate (TN:TP < 29 by
weight) and a region where blue-greens tended to be rare
(TN:TP>29 by weight). It is interesting to note that two
warmwater lakes (Lake George, Uganda, and Lake Kinneret,
Israel) follow this trend closely, despite the fact that the other
lakes in the data set are from cooler north temperate regions.
In contrast, there was no significant correlation between blue-
green dominance and either TN or TP alone (data not
shown).
DISCUSSION
The data in Figures 1 to 3 provide support for the hypothesis
that N:P ratios and nutrient competition are important fac-
tors shaping the species composition of natural
phytoplankton communities. Other, more qualitative data
from the literature also support this hypothesis. For exam-
ple, bloom-forming species of blue-greens declined sharply
in relative abundance following decreased P inputs and in-
creased TN:TP ratios in Gravenhurst Bay, Ontario (Dillon et
al. 1978). Similarly, blue-green algae became absent or rare
237
-------
Lake Restoration, Protection and Management
in Onondaga Lake, New York following a ban on phosphate-
containing detergents. TN:TP ratios in Onondaga Lake in-
creased steadily from a pre-ban value of 3.2 (1968-69) to
a post-ban value of 20.3 (1974) (Sze, 1975, 1980; Effler et
al. 1981). Smith et al. (1982) also suggest that the recent
decline of nitrogen-fixing species of blue-green algae in
Lough Neagh, Ireland, has resulted from increased inputs
of nitrate from agricultural watersheds.
These studies coupled with field studies cited earlier sug-
gest that blue-green algae generally may be better com-
petitors than other algal groups under conditions of nitrogen
100
et
0
ui
D
GO
i/i
O
Z
LU
s
50-
* POST -DIVERSION
1
10
r~
20
30
INFLUENT TN:TP RATIO
t/> 100-
Z
LU
Ul
ce
0
50-
O
Z
ui
*POST-DIVERSION
20
40
60
INFLUENT NO -N:PO -P RATIO
3 4
^> iuu-
LU
Ul
oc
o
UJ
3
CO
3 50-
0
t^
Z
UJ
_j
U-
c
. ^\
\
. \
*\
\*
\
it
* POST-DIVERSION *
*
*
1 1
10
20
30
EPILIMNETIC TN:TP RATIO
Figure 1.—Relationship between the June-September mean pro-
portion of filamentous bluegreen algae (surface samples) and (A)
loading TN:TP ratios, (B) loading NO3-N:PO4-P ratios, or (C)
epilimnetic TN:TP ratios in Lake Washington (USA). Data are from
W.T. Edmondson (pers. comm.) and Edmondson and Lehman
(1981). It should be noted that this proportion includes both
heterocystous and nonheterocystous filamentous bluegreens.
limitation (low N:P ratios), than under conditions of
phosphorus limitation (high N:P ratios). Recent research in
algal nutrient physiology provides a framework that can be
used to explore the implications of such a difference.
It is now clear that the growth rate ^, d~1) of an alga can
be described by the following empirical relationship (Droop,
1974; Rhee, 1978):
(D
M=min
This equation predicts that the growth rate is a function
of the cellular concentration of an essential nutrient (Qj) pre-
sent in the least quantity relative to the cells' subsistence
quota (kqi) for that nutrient. The optimal nutrient ratio, A at
which a transition from one nutrient limitation to another oc-
curs, can be calculated as the ratio of their subsistence
quotas (Droop, 1974; Tilman, 1977; Rhee, 1980):
A= kq1/kq2
(2)
O
Z
LU 5O
UJ
Q£
O
CO
WINTERKILL*
*POST-DREDGING
10
20
30
EPILIMNETIC TN:TP RATIO
Figure 2.—Relationship between the growing season mean pro-
portion of bluegreen algae and epilimnetic TN:TP ratios in Lake
Trummen (Sweden). It should be noted that in this case (and in
Fig. 3), I included all bluegreens—i.e., unicellular + colonial +
heterocystous and nonheterocystous filamentous species—in the
calculation of their proportion in the phytoplankton. Data are from
Bjork (1979) and Cronberg (1980).
O
UJ 50-
UJ
Q£
O
UJ
3
—J
03
a?
-v:
—I—
10
-T-
30
I—
40
—I—
50
—I—
60
—I—
70
EPILIMNETIC TN:TP RATIO
Figure 3.—Relationship between the growing season mean pro-
portion of bluegreen algae and epilmnetic TN:TP ratios in 14 natural
lakes worldwide. (*), Lake George, Uganda; (•&), Lake Kinneret,
Israel. Sources of the data are available on request.
238
-------
Causes and Control of Blue-Greens: Alternatives to Nutrient Control
100
50-
LABORATORY COMPETITION BETWEEN
MICROCYSTIS AND ASTERIONELL A
(Holm and Armstrong 1981 )
/,= 931 15
ju = 0.11 d~'
0 100 200
Molar Si : P Supply Ratio
Figure 4.—Plot of data from laboratory competition experiments
between the bluegreen alga Microcystis aeruginosa and the diatom
Asterionella formosa (from Holm and Armstrong, 1981). The pro-
portion of Microcystis declines steadily as the silicon:phosphorus
(Si:P) ratio is increased, a trend consistent with competition theory
(see text).
The optimal ratio is one parameter that determines the
coexistence or elimination of species in a mixed
phytoplankton assemblage competing for nutrients. For ex-
ample, recent studies (Titman, 1976; Tilman, 1977) have
shown that nutrient competition between Asterionella formosa
and Cyclotella meneghiniana is mediated by the relationship
of the supply Si:P ratio to the optimal ratios (KqSi:Kqp) for
these diatoms. Similarly, Holm and Armstrong (1981) have
demonstrated the importance of Si:P supply ratios to the out-
come of competition between the blue-green alga Microcystis
aeruginosa and the diatom Asterionella formosa, I have plot-
ted the data of Holm and Armstrong (1981) in Figure 4 to
illustrate the parallels between their study and the data
presented in Figures 1 to 3.
Current competition theory (e.g. Taylor and Williams, 1975)
suggests that two species should coexist only when each
is limited by a different resource. In the case of Asterionella
and Microcystis, this occurs when the diatom is limited by
silicon (low Si:P supply ratios) and when the blue-green is
simultaneously phosphorus-limited (all Si:P supply ratios,
since Microcystis does not require Si for growth). Holm and
Armstrong (1981) estimate that the value of for Si and P
in Asterionella is 93 ± 15 (by moles). In mixed culture these
two species should thus coexist in varying proportions for
Si:P supply ratios which are less than about 93; the theory
predicts that the proportional biomass of the competing blue-
green should decrease as the Si:P supply ratio is increased
to , (cf. Tilman 1982). As can be seen in Figure 4, the
data are consistent with the theory. Microcystis tends to
dominate under conditions of extreme silicon limitation (low
Si:P supply ratios), but is reduced to insignificant proportions
when both algae are P-limited (Si:P >• 93).
If it is true that blue-greens are good nitrogen competitors
but poor phosphorus competitors, this type of physiological
model helps explain the patterns evident in Figures 1 to 3.
Low N:P loading ratios, which impose nitrogen limitation,
should select for blue-greens if the optimum N:P ratios for
this group tend to be lower than those of other algae. This
hypothesis was examined by compiling optimal ratios
(kqN:kqp) for 10 algal species commonly found in lakes (Fig.
5). However, while there is a suggestion of clustering of blue-
green algae with values less than 12:1 by weight, as yet too
few data exist to draw firm conclusions.
This model also requires that the cellular N:P ratios of the
phytoplankton change in response to changes in the N:P
supply ratio. This phenomenon is well documented for batch
cultures (Myklestad, 1977) and continuous cultures of algae
(Rhee, 1978; Goldman et al. 1979) but until recently no data
were available for whole lakes. Data from nine lakes in the
Experimental Lakes Area of Northwestern Ontario (Fig. 6)
show a marked correlation betwen the N:P mass loading ratio
and the summer epilimnetic seston H:P ratio (Healey and
Hendzel, 1980). Data from Northern Quebec lakes and Lake
Norrviken, Sweden, also suggest that seston N:P ratios
covary with the epilimnetic TN:TP ratio (Fig. 7).
It is of interest to note the threshold N:P ratio at which
Lake Washington began to respond (TN:TP = 22-25, Fig.
1C). Assuming that the seston N:P ratio is roughly 0.5 TN:TP
in this region (cf. Fig. 7), this implies an algal N:P ratio of
ca. 11-12.5. Because the Lake Washington phytoplankton
was dominated by Oscillatoria spp. at this time, it is perhaps
of significance that this ratio is almost identical to the op-
timal ratio for Oscillatoria agardhii (12.0, Fig. 5). Furthermore,
u-
1—
I
o
QJ
^
i
O 10-
[0
^
O
^
DC
Q_
z
_i
|
Q.
dn-
^MICROCYSTIS SP. (4.1)R
R
"'SYNEDRA ULNA (4.5)"
^PSEUDANABAENA CATENATA (9.0)H
~ANKISTRODESMUS FALCATUS (9.5)R
^SELENASTHUM C APRIC OHN UTUM (10.8)
^FRAGILLARIA CROTONENSiS (11.3)R
OSCILLATORIA AGAHDHII (12.0)Z
SCENEDESMUS OBLIQUUS (13.1)R
-CRYPTOMONAS EROSA (17.1)H
-SCENEDESMUS QUADRICAUDA (37.9)H
characterized by very dense blooms of Microcystis spp.
30
o
<
cc
CL
9: 20-
^
0_
o
h-
co
LJ IQ-
C/5
Z
Ld
0
r
226 SW
303 239
• c
. 302 N
"304 *302S
226 NE
227
26!
5O IOO
Figure 5.—Values of the optimal N:P ratio, A, for 10 species of
freshwater algae (H, from Healey and Hendzel, 1979; R, from Rhee
and Gotham, 1980; Z, from Zevenboom et al. 1980).
TN: TP LOADING RATIO
Figure 6.—Relationship between the summer mean seston N:P
ratio (PN:PP) and the TN:TP loading ratio for nine Ontario lakes
(from Healey and Hendzel, 1980).
239
-------
Lake Restoration, Protection and Management
<
rr
Q.
Q.
o
o
I—
LJ
(S)
-z.
<
50-
20-
10-
* Lake Norryiken
Schefferville lakes
—T
40
1 1 1—
0 10 HO 30 40 50 60 70
MEAN EPILIMNETIC TN:TP RATIO
Figure 7.—Relationship between the summer mean seston N:P
ratio (PN:PP) and the epilimnetic TN:TP ratio for a group of Nor-
thern Quebec lakes (Diamond and Smith, unpub. data) and Lake
Norrviken, Sweden (from Ahlgren, 1979).
which virtually disappeared after lake restoration (Cronberg,
1980). As is evident in Figure 2, these years were also
characterized by TN:TP ratios < 10. Again assuming seston
N:P = 0.5 TN:TP, the algal N:P ratio should be < 5, a value
very close to the optimal N:P ratio for Microcystis aeruginosa
(4.1, Fig. 5). Thus, although care should be exercised in the
use of N:P ratios to assess nutrient limitation in natural com-
munities (Zevenboom et al. 1982), these field data offer
results consistent with our knowledge of algal nutrient
physiology.
It is clear, however, that the N:P hypothesis alone will be
insufficient to explain the presence or absence of blue-green
algae in all lakes, since it is evident from Figure 3 that many
lakes having TN:TP ratios •< 29 were dominated by non-
blue-greens. Some factor(s) other than N or P must have
limited the growth of Cyanophyta in these lakes. Converse-
ly, one might expect lakes having high N:P ratios to be
dominated by blue-greens if concentrations of N and P are
nonlimiting. For example, Lake Gjersj<(>en (Norway) is
dominated by Oscillatoria spp., despite spring TN:TP ratios
>60:1 by weight (Faafeng and Nilssen, 1981). Van Liere
and Mur (1980) attribute the dominance of blue-greens in
Lake Gjersj<(>en to an overriding light limitation, and I am cur-
rently investigating the interactions between light, nitrogen,
and phosphorus using a larger data set than the one reported
on here. It is also of interest to note that the response of
Lake Trummen phytoplankton to nutrients was altered in the
year immediately following a winter fishkill, suggesting that
the composition of other trophic levels also influences blue-
green dominance (Fig. 2).
However, the data presented here are generally consis-
tent with current ecological theory, and suggest that natural
lakes having TN:TP ratios > 29 by weight will typically ex-
hibit low proportions of blue-green algae. These data are also
of practical significance since modification of N:P ratios can
be achieved in many lakes by sewage diversion, wastewater
P removal, or nutrient precipitation within the lakes
themselves. Epilimnetic TN:TP ratios typically increase as
a result of these lake restoration techniques (Smith and
Shapiro, 1981), and Figures 1 to 3 suggest preliminary
nitrogen:phosphorus ratios towards which agencies con-
cerned with water quality might aim. Alternatively, in small
lakes, nitrogen fertilization may be of practical value (Barica
et al. 1980; Flett et al. 1980; Leonardson and Ripl, 1980).
ACKNOWLEDGEMENTS. I thank V. Bierman, G. G. Ganf, N. P.
Holm, J. Kalff, J. Shapiro, E. B. Swain, and D. Tilman for comments
and advice, and I thank W. T. Edmondson and J. DePinto for the
use of unpublished data. This research was supported by NSF Grant
DEB-7921755 to J. Shapiro and a NATO Post-doctoral Fellowship
to V. Smith. Contrib. No. 260 from the Limnological Research Center,
University of Minnesota.
REFERENCES
Ahlgren, 1.1979. Lake metabolism studies and results at the Institute
of Limnology in Uppsala. Arch. Hydrobiol. Beih. Ergebn. Limnol.
13: 10-30.
Barica, J., H. Kling, and J. H. Gibson. 1980. Experimental manipula-
tion of algal bloom composition by nitrogen addition. Can. J. Fish.
Aquat. Sci. 37: 1175-1183.
Bjork, S. 1979. Lake management studies and results at the Institute
of Limnology in Lund. Arch. Hydrobiol. Beih. Ergebn. Limnol. 13:
31-55.
Cronberg, G. 1980. Phytoplankton changes in Lake Trummen in-
duced by restoration. Long-term whole-lake studies and ex-
perimental biomanipulation. Ph.D. Thesis, Univ. Lund.
Diamond, M., and V. H. Smith. Data from Schefferville Research
Project McGill University, Montreal, Canada (Unpubl.).
Dillon, P. J., and F. H. Rigler. 1974. The chlorophyll-phosphorus
relationship in lakes. Limnol. Oceanogr. 19: 767-773.
1975. A simple method for predicting the capacity of a lake
for development based on the lake trophic status. J. Fish. Res.
Board Can. 32: 1519-1531.
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: 263-271.
Droop, M. R. 1974. The nutrient status of algal cells in continuous
culture. J. Mar. Biol. Assoc. U.K. 54: 825-855.
Edmondson, W. T., and J. T. Lehman. 1981. The effect of changes
in the nutrient income on the condition of Lake Washington. Lim-
nol. Oceanogr. 26: 1-29.
Effler, S. W., S. D. Field, M. A. Meyer, and P. Sze. 1981. Response
of Onondaga Lake to restoration efforts. J. Environ. Eng. Div. Am.
Soc. Civil Eng. 107 (EE1): 191-210.
Faafeng, B. A., and J. P. Nilssen. 1981. A 20-year study of
eutrophication in a deep, soft-water lake. Int. Ver. Theor. Angew.
Limnol. Verh. 21: 412-424.
Flett, R. J., D. W. Schindler, R. D. Hamilton, and N. E. R. Camp-
bell. 1980. Nitrogen fixation in Canadian Precambrian Shield lakes.
Can. J. Fish. Aquat. Sci. 37: 494-505.
Goldman, J. C., J. J. McCarthy, and D. G. Peavey. 1979. Growth
rate influence on the chemical composition of phytoplankton in
oceanic waters. Nature 279: 210-215.
Healey, F. P., and L. L. Hendzel. 1979. Indicators of phosphorus
and nitrogen deficiency in five algae in culture. J. Fish. Res. Board
Can. 36:1364-1369.
1980. Physiological indicators of nutrient deficiency in lake
phytoplankton. Can. J. Fish. Aquat. Sci. 37: 442-453.
Holm, N. P., and D. E. Armstrong. 1981. Role of nutrient limitation
and competition in controlling the populations of Asterionella for-
mosa and Microcystis aeruginosa in semicontinuous culture. Lim-
nol. Oceanogr. 26: 622-634.
Hutchinson, G. E. 1944. Limnological studies in Connecticut. VII.
A critical examination of the supposed relationship between
phytoplankton periodicity and chemical changes in lake waters.
Ecology 25: 3-26.
Jones, J. R., and R. W. Bachmann. 1976. Prediction of phosphorus
and cholorophyll levels in lakes. J. Water Pollut. Control Fed 48'
2176-2182. ' '
240
-------
Causes and Control of Blue-Greens: Alternatives to Nutrient Control
Leonardson, L. and W. Ripl. 1980. Control of undesirable algae and
induction of algal succession in hypertrophic lake ecosystems.
Pages 57-65 in J. Barica and L. R. Mur, eds. Hypertrophic
ecosystems, Junk, The Hague.
Myklestad, S. 1977. Production of carbohydrates by marine plank-
tonic diatoms. II. Influence of the N/P ratio in the growth medium
on the assimilation ratio, growth rate, and production of cellular
and extracellular carbohydrates by Chaetoceros affinis var. willei
(Gran) Hudstedt and Skeletonema costatum (Grev.) Cleve. J. Exp.
Mar. Biol. Ecol. 79: 161-179.
Niemi, A. 1979. Blue-green algal blooms and N:P ratio in the Baltic
Sea. Acta Bot. Fennica 110: 57-61.
Pearsall, W. H. 1932. Phytoplankton in the English lakes. II. The
composition of the phytoplankton in relation to dissolved
substances. J. Ecol. 20: 241-262.
Reckhow, K. H., and J. T. Simpson. 1980. An empirical study of
factors affecting blue-green versus nonblue-green dominance in
lakes. Proj. comple. rep. NTIS PB80-169311.
Rhee, G-Y. 1978. Effects of N:P atomic ratios and nitrate limitation
on algal growth, cell composition, and nitrate uptake. Limnol.
Oceanogr. 23: 10-25.
1980. Continuous culture in phytoplankton ecology. Pages
151-203 in M. R. Droop and H. W. Jannasch, eds. Advances in
Aquatic Microbiology. Vol. 2. Academic, New York.
Rhee, G-Y., and I. J. Gotham. 1980. Optimum N:P ratios and co-
existence in phytoplankton. J. Phycol. 16: 486-489.
Schindler, D. W. 1977. Evolution of phosphorus limitation in lakes.
Science 195: 260-262.
Smith, R. V., R. J. Stevens, R. H. Foy, and C. E. Gibson. 1982.
Upward trend in nitrate concentrations in rivers discharging into
Lough Neagh for the period 1969-1979. Water Res. 16: 183-188.
Smith, V. H., and J. Shapiro. 1981. A retrospective look at the
effects of phosphorus removal in lakes. Pages 73-77 in Restora-
tion of Lakes and Inland Waters. U.S. EPA^40/5-81-010. U.S.
Environ. Prot. Agency, Washington, D.C.
Sze, P. 1975. Possible effect of lower phosphorus concentrations
on the phytoplankton in Onondaga Lake, U.S.A. Phycologia 14:
197-204.
1980. Seasonal succession of phytoplankton in Onondaga
Lake, U.S.A. Phycologia 19: 54-59.
Taylor, P. A., and P. J. Williams. 1975. Theoretical studies on the
coexistence of competing species under continuous-flow condi-
tions. Can. J. Microbiol. 21: 90-98.
Tilman, D. 1977. Resource competition between planktonic algae:
an experimental and theoretical approach. Ecology 58: 338-348.
1982. Resource competition and community structure.
Princeton Monogr. Pop. Biol. 17.
Titman, D. 1976. Ecological competition between algae: Experimental
confirmation of resource-based competition theory. Science 192:
463-465.
Van Liere, L., and L. R. Mur. 1980. Occurrence of Oscillatoria
agardhii and some related species, a survey. Pages 67-77 in J.
Barica and L. R. Mur, eds. Hypertrophic Ecosystems. Junk, The
Hague.
Vollenweider, R. A. 1968. Water management research. Organ.
Econ. Coop. Dev. Paris. DAS/CSI/68.27. Mimeo.
1976. Advances in defining critical loading levels for phos-
phorus in lake eutrophication. Mem. 1st. Ital. Idrobiol. 33: 53-83.
Zevenboom, W., A. B. de Vaate, and L. R. Mur. 1982. Assessment
of factors limiting growth rate of Oscillatoria agardhii in hypertrophic
Lake Wolderwijd, 1978, by use of physiological indicators. Lim-
nol. Oceanogr. 27: 39-52.
Zevenboom, W., K. M. Knip, and L. R. Mur. 1980. Influence of the
nature of the growth-limitation on some physiological properties
of Oscillatoria agardhii grown in continuous cultures. Pages
109-127 in W. Zevenboom. Growth and nutrient uptake kinetics
of Oscillatoria agardhii. Ph.D. Thesis, Univ. Amsterdam.
241
-------
CYANOPHAGE: HISTORY AND LIKELIHOOD AS A CONTROL
PAUL R. DESJARDINS
Department of Plant Pathology
University of California, Riverside
Riverside, California
ABSTRACT
It has been 20 years since the first cyanophage was discovered. Since then additional cyanophages
and strains that infect both unicellular and filamentous cyanobacteria have been found. Cyanophages
are similar to other bacteriophages in many physical, chemical and biological characteristics, but dif-
fer from them in their requirement of light tor absorption to their hosts and their dependence upon
the photosynthetic activity of their hosts for their replication. Light quality and the ratio of red to far-
red light affect virus replication. Cyanophages play a distinctive role in the ecology of their hosts and
probably are effecting some natural control. Certain factors (development of resistant host strains,
specific ion requirements, environmental factors and lysogeny) may affect the potential of the
cyanophages to control their hosts, but these have not been conclusively shown to completely destroy
this potential. There is much need for additional research on the experimental control of nuisance
species In natural water bodies. Preliminary studies suggest that the phages may be more effective
in preventing blooms than in eliminating one already formed. An integrated approach involving several
biological techniques is recommended for control of nuisance populations of cyanobacteria.
INTRODUCTION
Since the North American Lake Management Society has
included a session on the control of blue-green algae in this
Conference, I am sure I do not have to attempt to convince
you of the importance of blue-green algal blooms. Dr. John
Barko of the U.S. Army Engineers Waterways Experiment
Station states that they are one of the most serious water
pollution problems in the United States (pers. eomm.). I was
recently quite surprised, however, to note during the course
of consulting eight different books on limnology, all publish-
ed since 1970, that only four of them had any reference to
blue-green algal (cyanobacterial) blooms. Only one of them
(Willoughby, 1976) discussed cyanophages and their possi-
ble role in the ecology of their hosts and possible control
potential. One possible reason for the dearth of reference
to cyanophages in limnology books is the fact that many of
us who are studying cyanophages do not generally publish
in journals read by limnologists, nor do we participate fre-
quently enough in conferences such as this one.
Although Tiffany in the second edition of his book, Algae—
the Grass of Many Wafers (1958), discounts the seriousness
of accounts that toxins are produced by blue-green algae,
there is ample evidence that some clones of certain species
do Indeed produce very potent toxins (Barton and Johnson,
1978; Carmichael, 1981; Carmichael and Gorham, 1981; Col-
lins, 1978; Gorham and Carmichael, 1979; Holm-Hansen,
1968; Hughes et al. 1958; Porter and Orcutt, 1980). The more
frequent detection of blooms of toxic alga! species in recent
years makes the nuisance aspect of algal blooms even more
serious.
It is obvious, of course, that cyanobacterial populations
in water bodies need to be managed so that bloom concen-
trations do not occur. In this paper we shall consider the
possibility of using cyanophages in control and management
strategies.
HISTORICAL VIEW OF CYANOPHAGES
Before any cyanophage had actually been found, Krauss
(1961) predicted they would be found, based on the obser-
vation that algal blooms sometimes suddenly collapsed when
nutrient levels and environmental conditions were still ideal
for continued algal growth. Shortly after Krauss' prediction
Safferman and Morris (1963) reported the discovery of the
first blue-green algal virus (cyanophage or phycovirus). The
LPP-1 cyanophage they first described 20 years ago infects
cyanobacteria in the genera Lyng/bya, Plectonema, and Phor-
mldium. In subsequent years additional cyanophages and
cyanophage strains, infecting both unicellular and filamen-
tous cyanobacteria have been described (Brown, 1972; Des-
jardins, 1981; Desjardins and Olson, in press; Granhall, 1972;
Hu, N-T. et al. 1981; Khydyakov and Gromov, 1973;
Koz'yakov et al. 1972; Koz'yakov, 1977; Safferman, 1973
a,b; Safferman et al. 1969,1972; Sherman and Haselkorn,
1971). For a comprehensive coverage of the literature on
cyanophages the reader is referred to the Practical Direc-
tory to Phycovirus Literature (with addendums) by Safferman
and Rohr (1979).
SOME CHARACTERISTICS OF
CYANOPHAGES
As has been pointed out earlier (Desjardins, 1981), it is not
too surprising that cyanophages are similar in morphology
and interaction with their hosts to many of the bacteriophages
since the hosts of both groups of phages (I.e., cyanobacteria
and other bacteria) are both prokaryotes and related.
As a virologist, who was initially quite confused by the
system of taxonomy of blue-green algae, I wish to point out
the Importance of definite identification of the cyanobacterial
host before attempts are made to characterize a new
cyanophage. Staneir et al. (1971) have demonstrated the
need for working with pure cultures of cyanobacteria before
attempting classification. Perhaps the use of bacterial rules
of taxonomy and nomenclature, as recommended by Rip-
pka et al. (1979), will result in a taxonomic scheme that will
be less confusing for the nontaxonomist.
The importance of obtaining a cyanophage in a highly
purified state before attempts are made to characterize it
should also be reiterated (Barkley, 1976; Barkley and Des-
jardins, 1977; Kraus, 1980; Safferman, 1973 a,b; Sherman
and Brown, 1978). If one is to avoid a good deal of difficul-
242
-------
Causes and Control of Blue-Groans: Alternatives to Nutrient Control
ty, one should make every effort to whether
a particular has only one or a mix-
ture of and Olson, in Hu, N-T, el
al. 1981),
As the are similar in mor-
to some of the of
from different and
are in the in
1-3. The A-1 that
and which has a long tail and
at the is shown in Figure 1. The
short-tailed LPP-1 which was the first one
by Safferman and Morris (1963), is shown in
Figure 2, Figure 3 the widely AS-1
which the unicellular Anaeysffc
and The has a long
tail.
Rgure 3.
Figure 1.
Figure 2.
adsorb do host to in-
much the same way that many ad-
to 1976; Currier and.Wolk, 1979;
and 1972; 1981;
1973 a, b; Samimi and 1978; and Brown,
1978). The of AS-1 to one of its
S, Is shown in 4, They use the tail
to inject infectious acid into
In the latter respect they resemble many
bacteriophages.
Figure 4,
In another they differ dramatically from
and that is in their to- light for
to and Hutchison, 1976;
and 1979; and Olson, in
Olson and a, b; Sherman and Brown,
1978). In the AS-1 -A.
directly with the intensity of red light (625 rim) (Olson and
Desjardins, 1982 a, b). In the dark, diminishes
to (Alien and Hutchison, 1S76; and
All that to
(Brown, 1972; 1981;
Safferman, 1973 a, b; Sherman and Brown, 1978).
The are over a
pH (pH 4-11) than (pH 5-8)
and 1964; and Brown, 1978).
Their and ability to at pH's
Is to their at pH
of or (Brock, 1973),
Some are lytic, they
lysis and of host The lytic
of a it as a
control This lytic can be II-
by the they on lawns of
ceil. The of the cell
(or lawn) where have by the
243
-------
Lake Restoration, Protection and Management
Figure 5.
cyanophage. Clear plaques induced by the lytic cyanophage
AS-1 on a lawn of A. nictulans are shown in Figure 5.
The phenomenon of lysogeny in cyanophage-host systems
has been established (Desjardins and Olson, in press; Hu.
N-T. et al. 1981; Khydyakov and Gromov, 1973; Koz'yakov
et al. 1972; Rimon and Oppenheim, 1975; Sherman and
Brown, 1978; Shilo, 1972). However, induction of the lytic
state from the lysogenic condition appears to be more dif-
ficult with cyanophages than bacteriophages (Rimon and Op-
penheim, 1975; Sherman and Brown, 1978; Shilo, 1972).
When the lysogenic condition is established the cyanophage
viral genome becomes integrated with the host chromosome
and does not replicate separately nor cause lysis of the host
cell. Many of the viral genes that code for virion structural
proteins, destruction of host DNA, and lysis of the host cell
are not expressed. Lysogeny will be discussed further later
in this paper.
Generally within a range of temperature, the rates of
chemical and biochemical reactions increase with in-
creased temperature. Temperature apparently plays an im-
portant role in the development of natural blooms (Kruger
and Eloff, 1978). As one might expect, temperature also af-
fects the growth cycle of cyanophages (Allen and Hutchison,
1976; Desjardins and Olson, in press; Olson and Desjardins,
1982 a, b; Sherman and Brown, 1978). In the AS-1-A.
nidulans system, the length of the growth cycle varies in-
versely with temperature (Desjardins and Olson, in press;
Olson and Desjardins, 1982 a, b). In the range of 25 to
26° C a 1° increase in temperature shortens the growth cy-
cle by approximately 0.5 hours.
A truly unique feature of cyanophages in which they dif-
fer from their bacteriophage counterparts is their dependence
on the photosynthetic activity of their hosts for their replica-
tion (Adolph and Haselkorn, 1972; Allen and Hutchison,
1976; Desjardins, 1981; Desjardins and Olson, in press;
Olson and Desjardins, 1982 a, b; Padan et al. 1970; Sher-
man and Haselkorn, 1971; Sherman and Brown, 1978).
Although the cyanophages that infect unicellular
cyanobacteria and those that infect filamentous
cyanobacteria react the same in some respects to any
impairment to the photosynthetic activity of their hosts, in
other respects they differ (Allen and Haselkorn, 1972; Sher-
man and Brown, 1978). Both groups are dependent on Photo
System I, and both are inhibited by carbonyl-cyanide m-
chlorophenyl hydrazone (CCCP), an inhibitor of photosyn-
thetic electron transport. In the final analysis both depend
on the availability of adenosine triphosphate (ATP). They dif-
fer in that unicellular cyanophages do not inhibit CO2 fixa-
tion while the filamentous group does. The latter group is
not dependent on Photo System II whereas the unicellular
group is. For further discussion of the differences between
the two groups, the reader is referred to the excellent review
by Sherman and Brown (1978).
In the AS-1-A nidulans system at 30°C an inverse log-
linear correlation was found with far-red light and the length
of the rise period of the growth cycle of the cyanophage.
A linear relationship between the red/far-red ration (ratio of
intensities) and virus yield was also found (Desjardins and
Olson, in press; Olson and Desjardins, 1982 a, b). Thus not
only does the light quality play a role in the replication of
the virus, but the ratio of red to far-red is also effective in
virus yield. The latter has apparently not been recognized
before and is, I believe, quite significant.
POTENTIAL FOR CONTROL OF NUISANCE
CYANOBACTERIA
The features of cyanophages that recommend them as an
"ideal algicide" have been previously described and dis-
cussed (Desjardins et al. 1978; Desjardins, 1981), but will
be briefly reviewed here to serve as a reference for subse-
quent discussions. The features are: (1) Cyanophages are
selective and specific for the nuisance species; (2) they are
nontoxic to other microorganisms in the food chain; (3) they
are harmless to man and animals; (4) they become incor-
porated into the cycling of natural elements; (5) they have
no direct effect on water quality; (6) they are noncorrosive
for mechanical equipment; and (7) they increase rather than
deplete during use.
It is known that cyanobacteria vary in their sensitivity to
the most commonly used algicide, copper sulfate (Palmer,
1977). Safferman and Morris (1964) have stated that treat-
ment of water bodies with copper sulfate and chlorine is
seldom aimed at the specific algae causing the problem, but
rather annihilates the total algal population perhaps disrup-
ting the biologic community and adversely affecting the
aquatic environment. The ability of cyanophages to increase
during use and their specificity should especially recommend
them as algicidal agents.
A number of workers who have worked with and studied
cyanophages believe that these phages play a distinctive
role in the ecology of their hosts and in many instances also
effect some natural control of populations of their hosts. This
belief is based on the fact that in many instances when a
cyanophage is found its host is in either very low cell popula-
tions or cannot be found, while in other instances where host
cell populations are found their fluctuation coincides with in-
creases in cyanophage concentrations (Cannon, 1975; Des-
jardins et al. 1978; Desjardins, 1981; Granhall, 1972; Jackson
and Sladecek, 1970; Safferman, 1968; Safferman and Mor-
ris, 1964, 1967; Shilo, 1969, 1972).
It has also been suggested by a number of individuals that
the lytic cyanophages could act as biological control agents
against nuisance cyanobacterial populations if they were in-
troduced by man (Brown, 1972; Cannon, 1975; Desjardins
et al. 1978; Desjardins, 1981; Desjardins and Olson, in press;
Jackson and Sladecek, 1970; Safferman and Morris, 1964,
1967; Shilo, 1969,1972). This is an area where much addi-
tional research is needed and one additional important con-
sideration before the full control potential of cyanophages
can be realized or assessed is the need to find additional
cyanophages, especially ones attacking nuisance cyano-
bacteria, for which there are presently no cyanophages.
cyanophages.
244
-------
Causes and Control of Blue-Greens: Alternatives to Nutrient Control
PRACTICAL CONSIDERATIONS AND
PROBLEMS
The first obstacle to the use of eyanophages as control
agents that is generally proposed is the problem of the
development of resistant strains of the host in the presence
of the cyanophage. Several workers have carried out
laboratory studies to test this possibility. Cowlishaw and
Mrsa (1975) have studied the interaction of LPP-1
cyanophage and P. boryanum over an extended period
of time. Cannon et at. (1976) have made a similar study
with the same phage-host system but used both lytic and
temperate phages. Jenifer (1977) made a somewhat
similar study with the AS-1 cyanophage and its host, A,
nldulans.
As might be expected under the basic experimental
setup, the eyanophages provided a selective pressure on
the host resulting in the replacement of sensitive host
strains with resistant host strains. Since Cannon et al.
(1976) had included lytic and temperate phages, they con-
cluded that the lytic phage did provide a parasite pressure
on the host, but their studies of lysogenic algae and
temperate eyanophages led them to the conclusion that
a balanced parasitic relationship might be occurring in
nature.
Cowlishaw and Mrsa concluded that although the results
of their research were not encouraging for using
eyanophages for control, the lysis of the nuisance bloom
species might be followed by development of a less
troublesome species of blue-greens,
Jenifer (1977), on the other hand, concluded from his
studies that eyanophages could probably not be used suc-
cessfully as specific control agents.
As pointed out earlier (Desjardins, 1981; Desjardins and
Olson, in press), I feel that it does not seem reasonable
to completely extrapolate results from controlled laboratory
studies with axenic cultures to conditions existing in a large
water body where there is little control of environmental
conditions and a great diversity of microblal species.
Jenifer's conclusion certainly appears to ignore the suc-
cession of dominant species that occurs in natural water
bodies (Lin, 1972). Shilo (1972) has also pointed out that
even resistant hosts might well be expected to be suscep-
tible to suitable host-range phage mutants.
Jenifer (1977) has indicated that the nature of the
resistance of the selected resistant strain of A. nldulans
was a loss of the receptor sites. This may well be the case.
In studies of the plating efficiency of cyanophage N-1 on
different isolates of Anabaena variabilis, Currier and Wolk
(1979) suggested that the variation of plating efficiency
on different isolates might be accounted for by the
presence of a DNA restriction endonuclease in isolates
with low plating efficiency. Apparently heating the
organism increased the plating efficiency. One might
wonder if the apparent development of resistant strains
might not occur by mutation of the host to one containing
a restriction enzyme.
Another aspect to be considered here is the reversion
of the resistant strain back to a susceptible strain. Padhy
and Singh (1978) demonstrated the reversion of N-1 resis-
tant mutant of Nostoc muscorum. Also, in our laboratory
a number of years ago we obtained a strain of A, nidulans
which was resistant to the AS-1 virus (Desjardins and
Barkley, unpubl.) Because of other research interests at
the moment, we set the culture aside temporarily, but con-
tinued to make regular transfers. No attempt was made
to clone the resistant strain, but we simply subcultured
it over a period of months. Sometime later when we decid-
ed to study the culture from the standpoint of phage ad-
sorption to the cells, we found that the culture was once
again susceptible to AS-1 eyanophage. Apparently the
susceptible strain had again become the dominant cell type.
The removal of selection pressure by a temporary absence
of the virus apparently resulted in the re-emergence of the
susceptible strain. Perhaps this might also occur in nature.
Cannon (1975) has pointed out that exhaustive field studies
have failed to reveal a naturally occurring strain of P.
boryanum resistant to the LPP-1 cyanophage in localities
where the LPP-1 eyanophage is found.
I maintain that until field studies prove otherwise, one can-
not state unequivocally that the development of resistant host
strains will preclude the use of eyanophages as biological
control agents.
The phenomenon of lysogeny has been briefly described
earlier in this paper. The phenomenon involves the "intimate
relationship" of a so-called temperate phage with its host.
Also, this intimate relationship confers an immunity to related
lytic phages upon the lysogenized host. Recalling that
lysogenized cyanobacterial hosts are not lysed when en-
vironmental conditions favor the lysogenic state, the ques-
tion arises as to whether lysogeny itself or the development
of temperate cyanophage mutants would present a problem
for the control potential of eyanophages in a natural
ecosystem. This is an area that undoubtedly needs additional
research. It is, however, one that is somewhat difficult to ap-
proach experimentally.
Cannon (1975) has suggested that lysogeny might per-
mit both cyanophage and host ceils to survive when en-
vironmental conditions were not ideal for either one. Later,
when conditions were more ideal, the lytic cycle could be
induced either spontaneously or chemically with the release
of virions that in turn would control blooms of cyanobaeterial
populations. According to Cannon this hypothesis could ac-
count for the apparent nonblooming nature of Plectonema
boryanum which can be lysogenized by cyanophage strains
in the LPP group.
Shilo (1972), discussing the heat sensitive temperate LPP-
phage Splctsl and its host P, boryanum has suggested that
certain factors such as temperature, UV light, or unbalanc-
ed growth conditions influence formation of lysogenized
cultures or the lytic cycle, and thus may play a significant
role in the population balance in nature. If a thermosensitive
lysogenic algal population in nature were exposed to elevated
temperatures, the lytic cycle might be induced, resulting in
a decreased host cell population. Shilo further suggests that
such a phenomenon might be involved in the fluctuations
of blue-green populations sometimes observed in nature.
Although one cannot positively say whether or not
lysogeny would be a limiting factor in the use of eyanophages
as control agents, the author feels that it might be more of
a help than a hindrance.
It has been shown that a number of eyanophages have
certain ion requirements (especially Mg"1"1") for maximum
stability of the virus particle (Desjardins et al. 1978; Desjar-
dins, 1981; Desjardins and Olson, in press; Mendzhul et al.
1975; Sherman and Brown, 1978). The LPP-1 cyanophage
is especially dependent on the Mg^ ion for its stability.
Amla (1981) reported that chelating agents can have a shock
effect on certain strains of AS-1 cyanophage. Presumably
the chelating agents depleted divalent cations from the virion
and infectivity was greatly reduced when the cyanophage
suspensions were diluted. Although depletion of ions required
for virion stability might reduce concentrations of infectious
eyanophages to a certain extent, it does not seem reasonable
to the author that it would completely nullify the control poten-
tial of eyanophages.
In an earlier report (Desjardins et al. 1975) our laboratory
described some effects of freezing on the AS-1 cyanophage
particle. Subsequently, we reinvestigated the effects of freez-
ing on AS-1 particle structure and included parallel studies
245
-------
Lake Restoration, Protection and Management
on the LPP-1 cyanophage in the investigation. The struc-
tural changes in the virions were characterized by electron
microscopy and density gradient centrifugation, and assays
were made for possible changes in infectivity tilers. We found
freezing significantly reduced infectivity of both cyanophages
as wen as somewhat dramatically changing the particle struc-
ture of some virions. Although freezing might reduce liters
of infectious virus in natural waters, it does not mean it would
totally destroy their control potential, especially since in a
large body of water complete freezing probably would not
occur. Obviously large scale field studies in locations where
freezing occurs would be most helpful in assessing freez-
ing effects in nature.
Another factor that could conceivably affect the infectivity
of cyanophage particles is their adsorption on cellular debris,
either of host or other microbial origin; adsorption to or en-
trapment in extracellular polysaccharides; or adsorption to
and inactivation by receptor sites released when their host
cells are lysed (Samimi and Drews, 1978; Sanger and
Dugan, 1972; Schnayer and Jenifer, 1974). It has recently
been shown that the AS-1 cyanophage will adsorb to artificial-
ly prepared liposomes with and without host lipid fractions
(Oliveira et al. 1982). As far as the release of host receptor
sites is concerned, one should keep in mind that the greater
the number of lysed host cells, the higher the population of
cyanophage particles that would be released on lysis. One
should also keep in mind that cyanophages increase rather
than decrease in number with use. Therefore, oven if a por-
tion of the cyanophage population were to be rendered
noninfectious, a certain portion could still infect new host
cells. This factor also seemingly would not totally eliminate
the control potential of the cyanophages.
ATTEMPTS AT CONTROL
Jackson and Sladecek (1970) may have been the first to at-
tempt control of a blue-green alga by using cyanophages
in a natural habitat. To test the LPP-1 cyanophage against
P. boryanum, they tried to establish the host in 5,000-gallon
tanks at a sewage treatment plant. They were unsuccessful
in establishing the alga in the tanks even though over 40
attempts were made to do so. The LPP-1 cyanophage was
already naturally present and was preventing the establish-
ment of the host by controlling it from the start of the
experiment.
Shilo (1972) has briefly described reported attempts by
Russian workers to control Microcystis blooms in a reser-
voir using a cyanophage that attacks this organism. In earlier
reports (Desjardins et al. 1978; Desjardins, 1981) we de-
scribed our success in completely knocking out A. nidulans
with AS-1 cyanophage and P. boryanum with LPP-1
cyanophage in 90-liter plastic tanks in temperature baths in
a greenhouse. The results of individual experiments with
these two virus-host systems are graphically illustrated in
Figure 7, which is taken from our earlier report (Desjardins,
1981). The LPP-1 cyanophage became so well-established
in the temperature bath facility that we were never able to
grow P. boryanum there again, even though we stream-
sterilized the plastic tanks in which the algal hosts were
cultured.
Using the same facility we were also successful in com-
pletely eliminating the Myers strain of Anabaena variabilis
from the tank culture using the A-1 lytic cyanophage re-
ceived from Prof. Gromov in Russia (Desjardins and Olson,
in press).
In earlier studies with outdoor pond facilities, we were not
able to establish bloom concentrations of A. nidulans (Des-
jardins et al. 1978). Later we were successful in establishing
P. boryanum in bloom concentrations in 8,000-liter outdoor
E
O
o
< Ol
A nidulons CONTROL CULTURE
A nidulons CULTURE WITH VIRUS
P boryonum CONTROL CULTURE
P boryonum CULTURE WITH VIRUS
I 234 5 6 7 8 9 10 II 12 13 14 15 16 17 18 19 20 21 22
DAYS AFTER INTRODUCTION OF ALGAL CELLS
Figure 7.—Growth of Plectonema boryanum and Anacystis nidulans
in 90 liter plastic tank with their specific viruses.
ponds (Desjardins, 1981; Desjardins and Olson, in press).
The ponds used are shown in Figure 6. To establish the
cyanobacterial host we had to enrich the ponds with
nutrients. This was done by adding Hughes medium to one
tenth the concentration used in aerated cultures in the
laboratory. In one pond where the host was reasonably well-
established before the virus was added, the host itself was
apparently controlled but was rapidly replaced by other blue-
green and green algal species (Desjardins, 1981). The pond
water did not clear but simply had a bloom involving other
species. This supports our point that a broad spectrum of
cyanophages should be available for control (Desjardins et
al. 1978; Desjardins, 1981).
In another pond where the LPP-1 phage was added before
the alga became established no bloom developed (Desjar-
dins and Olson, in press). Cyanophages probably have a
greater potential for preventing algal blooms than for
eliminating ones that have already formed.
DISCUSSION
Undoubtedly additional research is needed into many
aspects of the potential of cyanophages to serve as biological
control agents. Early research on these phages empha-
sized studies involving their isolation, purification, and
characterization. This emphasis is understandable because
of the initial novelty of this group of viruses.
Figure 6.
246
-------
Causes and Control of Blue-Greens: Alternatives to Nutrient Control
It is unfortunate that more effort has not been directed to
actual control studies. Certainly this is one area that should
be emphasized in future research.
At a recent workshop in Algal Management and Control,
it was recommended that a number of methods of biological
control be combined in an integrated approach (Desjardins,
1981). Such an approach would include bacterial agents
(Burnham, 1981), zooplankton grazing (Porter and Orcutt,
1980), and biomanipulation techniques (Shapiro et al. 1975)
in addition to using cyanophage agents (Desjardins, 1981).
Such an integrated approach will require extensive
cooperation by researchers in many different scientific
disciplines. I think R. M. Klein has appropriately described
this need for cooperation (Klein, 1970). I quote from his 1970
paper: "We are all ecologists, working with different levels
of bioorganization. We are all humanists concerned with the
role of science in human affairs. We are all part of the main."
He goes on to point out that if solutions to problems of the
environment are to be found, we all have the responsibility
to use our knowledge and expertise to contribute to the com-
mon welfare.
One last point that I wish to make with respect to research
on cyanophages is that the level of financial support for
research is going to have to be greatly increased if any real
progress is to be made in finally assessing their biological
control potential. Whether the support comes from govern-
mental sources, from the private sector or from both, the
support itself cannot be narrow in scope nor shortsighted
in its view.
ACKNOWLEDGEMENTS: The research on cyanophage in my
laboratory was supported by the Office of Water Research and
Technology, U.S. Department of Interior, under the Matching Grant
Program of P.L. 88-379, as amended, and by the University of Califor-
nia Water Resources Center. The most current grant was part of
the Office of Water Research and Technology Project No. A-066-CAL
and Water Resources Center Project W-533. Grateful acknowledge-
ment and thanks are extended to Dr. R. S. Safferman and Mrs. M.
E. Rohr for algal cultures, cyanophages, antiserum, and many helpful
consultations. For algal cultures and/or cyanophage strains, I also
thank Drs. L. L. Barton, G. V. Johnson, U. Granhall, M. V. Gromov,
J. Myers, L. A. Sherman and D. L. Suguna. I wish to especially
acknowledge the many contributions made by M. B. Barkley and
G. B. Olson during the course of graduate studies for the Ph.D.
degree carried out in my laboratory. I also thank R. J. Drake, S.
A. Swiecki, and W. M. Young for their contributions to our research
program and M. G. Barglowski, D. Linihan, G. O'Reilly, R. Parmelee
and J. A. Swiecki for their technical assistance.
REFERENCES
Adolph, K.W., and R. Haselkorn. 1972. Photosynthesis and the
development of blue-green algal virus N-1. Virology 47:370-374.
Allen, M.M., and F. Hutchison. 1976. Effect of some environ-
mental factors on cyanophage AS-1 development in Anacystis
nidulans. Arch. Microbiol. 110:55-60.
Amla, D.V. 1981. Chelating agent shock of cyanophage AS-1
infecting unicellular blue-green algae, Anacystis nidulans. Indian
J. Exp. Biol. 19:209-211.
Barkley, M.B. 1976. Ultrastructure of the blue-green algae Anacystis
nidulans infected with AS-1 virus. Ph.D. Thesis, Univ. California,
Riverside.
Barkley, M.B., and P.R. Desjardins. 1977. Simple, effective method
for purifying the AS-1 cyanophage. Appl. Environ. Microbiol.
33:971-974.
Barko, J. 1982. Pers. comm. U.S. Army Eng. Waterways Exp.
Sta., Vicksburg, Miss.
Barton, L.L., and G.V. Johnson. 1978. Characterization of Nutrients
and Algal Blooms at Abiquiu and Cochiti Reservoirs. Tech. Com-
ple. Rep. 3109-069, A-058-NMEX. New Mexico WRRI Rep. No.
094.
Brock, T.D. 1973. Lower pH limit for the existence of blue-green
algae: evolutionary and ecological implications. Science
179:480-482.
Brown, R.M. Jr. 1972. Algal viruses. Adv. Virus Res. 17:243-277.
Burnham, J.C. 1981. The utilization of bacteria in managing cyano-
bacterial populations: a review and update. Pages 230-258 in Proc.
Workshop Algal Manage. Control, Pacific Grove, Calif. March
1980. Tech. Rep. F-81-7.
Cannon, R. 1975. Field and ecological studies on blue-green algal
viruses. Pages 112-117 in Proc. Symp. Water Qual. Manage.
Through Biolog. Control. Rep. No. 07-75-1, Dep. Environ. Eng.
Sci. Univ. Florida.
Cannon, R.E., M.S. Shane, and J.M. Whitaker. 1976, Interaction
of Plectonema boryanum (Cyanophyceae) and the LPP-
cyanophages in continuous culture. J. Phycol. 12:418-421.
Carmichael, W.W. 1981. Freshwater blue-green algae (cyanobacteria)
toxins— a review. Pages 1-13 in W.W. Carmichael, ed. The Water
Environment - Algal Toxins and Health. Plenum Press, New York.
Carmichael, W.W., and P.R. Gorham. 1981. The mosaic nature
of toxic blooms of cyanobacteria. Pages 161-172 in W.W. Car-
michael, ed. The Water Environment-Algal Toxins and Health.
Plenum Press, New York.
Collins, M. 1978. Algal toxins. Microbiol. Rev. 42:725-746.
Cowlishaw, J., and M. Mrsa. 1975. Co-evolution of a virus-alga
system. Appl. Microbiol. 29:234-239.
Cseke, C.S., and G.L. Farkas. 1979. Effect of light on the attach-
ment of cyanophage AS-1 to Anacystis nidulans. J. Bact.
137:667-669.
Currier, T.C., and C.P. Wolk. 1979. Characteristics of Anabaena
variabilis influencing plaque formation by cyanophage N-1. J. Bact.
139:88-92.
Desjardins, P.R. 1981. Cyanophages—are they potential biological
control agents of nuisance blue-green algae? Pages 198-229 In
Proc. Workshop Algal Manage. Control, Pacific Grove, Calif.
March, 1980. Tech. Rep. E-81-7.
Desjardins, P.R., and M.B. Barkley. 1972. AS-1 virus adsorption to
cells and spheroplasts of Synechococcus cedrorum. Pages
332-333 in 30th Ann. Proc. Electron Microscope Soc. Am.
Unpubl. data. Dep. Plant Pathol. Univ. California, Riverside.
Desjardins, P.R., and G.B. Olson. In press, viral Control of Nuisance
Blue-Green Algae. II. Cyanophage Strains, Stability Studies of
Phages and Hosts, and Effects of Environmental Factors on
Phage-Host Interactions. Tech. Comple. Rep. Calif. Water Resour.
Center, Univ. California.
Desjardins, P.R., S.A. Swiecki, M.B. Barkley, and R.J. Drake. 1975.
Effects of certain chemical treatments and freezing on the
Ultrastructure of the AS-1 phycovirus particla. Abstr. No. C469.
3rd Int. Congress for Virology, Madrid, Spain.
Desjardins, P.R., M.B. Barkley, S.A. Swiecki, and S.N. West. 1978.
Viral Control of Nuisance Blue-Green Algae. Contrib. No. 169.
Calif. Water Resour. Center, Univ. California.
Gorham, P.R., and W.W. Carmichael. 1979. Phycotoxins from blue-
green algae. Pure Appl. Chem. 52:165-174.
Granhall, U. 1972. Aphanizomenon flos-aquae: Infection by cyano-
phages. Physiol. Plant. 26:332-337.
Hu, N-T., T. Thiel, T.H. Giddings, Jr., and C.P. Wolk. 1981. New
Anabaena and Nostoc cyanophages from sewage settling ponds
Virology 114:236-246.
Holm-Hansen, O. 1968. Ecology, physiology and biochemistry of
blue-green algae. Ann. Rev. Microbiol. 22:47-70.
Hughes, E.D., P.R. Gorham, and A. Zehnder. 1958. Toxicity of a uni-
algal culture of Microcystis aeruginosa. Can. J. Microbiol
4:225-236.
Jackson, D., and V. Sladecek. 1970. Algal viruses—eutrophication
control potential. Yale Sci. Mag. 44:16-21.
Jenifer, F.G. 1977. Studies on the natural relationships of cyano-
phages and their hosts and the nature of resistance. Comple. Rep.
Water Resour. Res. Inst., New Brunswick, N. J.
247
-------
Lake Restoration, Protection and Management
Khydyakov, I. Ya., and M.V. Gromov. 1973. The temperate cyano-
phage A-4 (L) of the blue-green alga Anabaena variabilis.
Mikrobiologiya 42:904-907.
Klein, R.M. 1970. Future prospects of algae and man. Pages 778-781
in J.F. Fredrick and R.M. Klein, eds. Phylogenesis and Mor-
phogenesis in the Algae. Ann. N.Y. Acad. Sci. 175:413-781.
Koz'yakov, S.Y. 1977. Cyanophages of series A (L), specific for blue-
green algae Anabaena variabilis. Pages 151-171 in B.V. Gromov,
ed. Experimental Algology. Biolog. Sci. Res. Inst. Leningrad State
Univ.
Koz'yakov, S.Y., B.V. Gromov, and I.Y. Khudyakov. 1972. Cyano-
phage A-1 (L) of the blue-green alga Anabaena variabilis.
Microbiology 41:486-489.
Kraus, M.P. 1980. Host-range, plaque morphology studies of cyano-
phage LPP-1. J. Phycol. 16:186-191.
Krauss, R.W. 1961. Fundamental characteristics of algal physiology.
Pages 40-47 in Algae and Metropolitan Wastes. U.S. Dep. Health
Edu. Welfare. SEC TR W 61-3.
Kruger, G.H.J., and J.N. Eloff. 1978. The effect of temperature on
specific growth rate and activation energy of Microcystis and
Synechococcus isolates relevant to the onset of natural blooms.
J. Limnol. Soc. S. Afr. 4:9-20.
Lin, C.K. 1972. Phytoplankton succession in a eutrophic lake with
special reference to blue-green algal blooms. Hydrobiologia
39:321-334.
Mendzhul, M.I., S.P. Bobrovnik, T.G. Lysenko and A.D. Shved. 1975.
Effect of certain physicochemical factors on infectivity of
cyanophages A-1, LPP-1 and LPP-3. Mikrobiol. Zh. 37:73-78.
Oliveira, A.R., J.B. Mudd and P.R. Desjardins. 1982. AS-1 cyano-
phage adsorption to liposomes. J. Gen. Virol. 61:153-156.
Olson, G.B., and P.R. Desjardins. 1982a. Light and temperature
modification of virus-host interactions in cyanophage AS-1. Abstr.
P49:8, XIII Internet. Congress of Microbiology, Boston, Mass.
1982b. The effect of light and temperature on the generation
time, adsorption and yield of cyanophage AS-1. Abstr. No. 63,
Ann. Meet. Am. Phytopath. Soc., Salt Lake City, Utah.
Padan, E., D. Ginsberg, and M. Shilo. 1970. The reproductive cycle
of cyanophage LPP-1-G in Plectonema boryanum and its
dependence on photosynthetic and respiratory systems. Virology
40:514-521.
Padhy, R.N., and P.K. Singh. 1978. Reversion of virus N-1 resistant
mutant of the blue-green alga A/ostoc muscorum. Experientia
34:1565.
Palmer, C.M. 1977. Algae and Water Pollution. EPA-600/9-7-036.
U.S. Environ. Prot. Agency, Cincinnati, Ohio.
Porter, K.G., and J.D. Orcutt, Jr. 1980. Nutritional adequacy, manage-
ability, and toxicity as factors that determine the food quality of
green and blue-green algae for Daphnia. Pages 1-35 in Am. Soc.
Limnol. Oceangr. Spec. Symp. Ill: The Evolution and Ecology of
Zooplankton Communities.
Rimon, A., and A.B. Oppenheim. 1975. Heat induction of the blue-
green alga Plectonema boryanum lysogenic for the cyanophage
SPIctsl. Virology 64:454-463.
Rippka, R., et al. 1979. Generic assignments, strain histories and pro-
perties of pure cultures of cyanobacteria. J. Gen. Microbiol.
111:1-61.
Safferman, R.S. 1968. Virus diseases in blue-green algae. Pages
429-439 in D.F. Jackson, ed. Algae, Man and the Environment.
Syracuse Univ. Press.
. 1973a. Special methods—virus detection in cyanophyceae.
Pages 145-158 in J.R. Stein, ed. Handbook of Phycological
Methods—Culture Methods and Growth Measurements. Cam-
bridge Univ. Press.
. 1973b. Phycoviruses. Pages 214-237 in N.G. Carr and B.A.
Whitton, eds. The Biology of Blue-Green Algae. Blackwell Sci.
Publ.
Safferman, R.S., and M.E. Morris. 1963. Algal virus: isolation. Sci-
ence 140:679-680.
_. 1964. Control of algae with viruses. J. Am. Water Works
Assoc. 56:1217-1224.
. 1967. Observations on the occurrence, distribution, and sea-
sonal incidence of blue-green algal viruses. Appl. Microbiol.
15:1219-1222.
Safferman, R.S., and M.E. Rohr. 1979. The Practical Directory to
Phycovirus Literature. EPA-600/9-79-013. U.S. Environ. Prot. Agen-
cy. Cincinnati, Ohio.
Safferman, R.S., T.O. Diener, P.R. Desjardins and M.E. Morris. 1972.
Isolation and characterization of AS-1, a phycovirus infecting the
blue-green algae, Anacystis nidulans and Syncechococcus
cedrorum. Virology 47:105-113.
Safferman, R.S., et al. 1969. Physovirus SM-1: A virus infecting uni-
cellular blue-green algae. Virology 37:386-395.
Samimi, B., and G. Drews. 1978. Adsorption of Cyanophage AS-1 to
unicellular cyanobacteria and isloation of receptor material from
Anacystis nidulans. J. Virol. 25:164-174.
Sanger, V.K., and P.R. Dugan. 1972. Polysaccharide produced by
Anacystis nidulans: its ecological implications. Appl. Microbiol.
24:732-734.
Schnayer, N., and F.G. Jenifer. 1974. Inactivation of blue-green alga
virus, AS-1, by isolated host lipopolysaccharide. Proc. Am.
Phytopath. Soc. 1:144.
Shapiro, J., V. Lamarra and M. Lynch. 1975. Biomanipulation: An
ecosystem approach to lake restoration. Pages 85-96 in Proc.
Symp. Water Qual. Manage. Through Biolog. Control. Rep. No.
ENV-07-75-1 Dep. Environ. Eng. Sci., Univ. Florida.
Sherman, L.A., and R.M. Brown, Jr. 1978. Cyanophages and viruses
of eucaryotic algae. Compreh. Virol. 12:145-234.
Sherman, LA., and R. Haselkorn. 1971. Growth of the blue-green
algal virus LPP-1 under conditions which impair photosynthesis.
Virology 45:739-746.
Shilo, M. 1969. New approaches to the control of harmful brackish
and fresh water algae of economic importance. Biotech. Bioeng.
Symp. 1:177-184.
1972. The ecology of cyanophages. Bamideh 24:76-82.
Stanier, R.Y., R. Kunisowa, M. Mandel, and G. Cohen-Bazire. 1971.
Purification and properties of unicellular blue-green algae (order
Chrococcales). Bact. Rev. 35:121-205.
Tiffany, L.H. 1958. Algae—The Grass of Many Waters. 2nd ed.
Charles C. Thomas, Publ. Springfield, III.
Willoughby, L.G. 1976. Freshwater Biology. Pica Press, New York.
248
-------
PREDATORY MYXOBACTERIA: LYTIC MECHANISMS AND
PROSPECTS AS BIOLOGICAL CONTROL AGENTS FOR
CYANOBACTERIA (BLUE-GREEN ALGAE)
JEFFREY C. BURNHAM
Department of Microbiology, Medical College of Ohio
Toledo, Ohio
PETER C. FRALEIGH
Department of Biology, University of Toledo
Toledo, Ohio
ABSTRACT
To control problem growths of primary producers in lakes and ponds, especially blooms of blue-green
algae and high densities of macrophytes, a diversity of methods have been proposed and are used.
However, absent in this repertoire are methods of biological control analogous to those that have been
successful in terrestrial ecosystems. Presented here is a discussion of studies that suggest that myx-
obacterial predation may be useful in biological control of blue-green algae in aquatic ecosystems.
INTRODUCTION
Huffaker et al. (1976) indicated that the premise of successful
biological control was based on the fact that organisms have
"natural enemies." These authors indicate that using these
natural enemies may be the best approach to developing
a biological control agent. If they are to be successful, these
control agents must possess several attributes: adaptability
to physical conditions, searching (or trapping) capacity,
multiplication, power of prey consumption, and survival dur-
ing periods of low prey availability. We will present the myx-
obacteria as bacterial organisms that fit these requirements.
The relationships between bacteria and cyanobacteria in-
volve symbiotic, commensal, and antagonistic interactions.
Although the most common relationships in nature are sym-
biotic (Echlin and Morris, 1965; Lange, 1971) naturally oc-
curring antagonism has been described by Fitzgerald (1969).
He demonstrated that bacteria-dominated sewage antagoniz-
ed the cyanobacterium Microcystis aeruginosa but not the
green alga Chlorella. Fallen and Brock (1979) showed that
an antagonistic bacterial population of 103 per ml were pre-
sent in a Michigan lake and that these bacteria depended
on the degraded products of cyanobacteria. Gunnison and
Alexander (1975) described the peptidoglycan cell wall layer
of the cyanobacteria as the weak link in their resistance to
natural antagonistic bacteria. Our paper will amplify this by
examining the development of information about the
predatory myxobacters and describe in detail the mechanism
of the myxococci in lysing captured cyanobacteria. Finally,
we will present our recent research efforts in using (a) co-
predatory colonies of myxococci and actinomycetes against
aqueous cyanobacteria, and (b) stable microcosms as
ecosystem testing systems for myxococcal biological control.
HISTORY
The bacteriolytic ability of myxobacteria has been recog-
nized for a long time (Beebe, 1941) but it has only been since
1967 that their ability to lyse cyanobacteria has been de-
scribed (Shilo, 1967). Earlier reviews are available of the
subsequent papers describing the myxobacterial lysis of
cyanobacteria (Burnham, 1975,1981; Burnham et al. 1981;
Stewart and Daft, 1977) so the details of many of these
papers will be omitted here. Stewart and Brown (1969) show-
ed that a Cytophaga could form plaques on lawns of both
green and blue-green algae. These authors described the
algal lysis as extracellular. Shilo (1970) showed that a myx-
obacter isolate (FP-1) could attach directly to the cell sur-
face of the cyanobacteria and subsequently lyse the host
cell. Although secreted lytic enzymes were not detected, the
fact that shaking the combined cultures prevented algal lysis
suggests that an exoenzyme is involved and that the agita-
tion removed the necessary enzyme concentration from the
microenvironment set up by the attached cells.
A similar problem was encountered by Daft et al. (1971,
1973, 1975) in using several Lysobacter sp isolates to lyse
cyanobacteria. They found agitation interfered with lyses and
that high densities of the lytic bacteria were necessary for
successful aqueous predation. This result conflicted
somewhat with their survey results of Scottish lochs and
reservoirs. These bodies of water all contained lytic myx-
obacteria in concentrations ranging from 1 to more than
100/ml with a mean of 44/ml in seven different habitats.
These authors proved a direct myxobacterial relationship with
prey populations by showing that the population of myx-
obacteria was in direct proportion to the chlorophyll a levels
in these waters. What they did not show and what we do
not yet know is whether these predatory bacteria have any
effect on the ecosystems present in these waters.
Daft and Stewart (1973) demonstrated ultrastructurally that
it was the peptidoglycan layer that was the primary target
of the myxobacterial enzymes providing evidence for the later
theory of Gunnison and Alexander (1975).
In analyzing Daft and Stewart's data, it is possible that
aqueous colonies of myxobacteria did form, as their
enumeration methodology could have underestimated the
predatory population level. The idea that colonies of
predatory bacteria might be involved in aqueous lysis of
cyanobacteria is supported by the results described by Burn-
ham et al. (1981). Myxococcus xanthus PC02 has been
shown to form semispherical colonies that entrapped and
lysed populations of the cyanobacterium Phormidium. The
details of this lytic mechanism will be described in the next
section of this paper; however, it should be mentioned here
that study of this lytic mechanism has provided further
249
-------
Restoration, Protection and Management
evidence that successful lysis in aqueous ecosystems
upon the establishment of a environment.
This then allows the cell wall lytic enzymes to accumulate
to an effective concentration. This mechanism is very
to that by the bdellovibrios, a predatory
group of bacteria that can penetrate into and lyse various
Gram-negative bacteria. Rittenberg and Tbomashow (1979)
the bdellovibrio predator as forming a "cozy en-
vironment" within the confines of the host's cell wall. Here
suitable enzyme concentrations can be maintained for the
orderly transfer of nutrient from the prey cytoplasm to the
predator. The myxoeoeci apparently do essentially the same
thing in providing a environment for enzymatic ac-
tivity, but on a multicellular basis.
Predation
In this section we will review the major characters of this
lytic system. Much of this research has been
previously (Burnham et al. 1979,1980,1983; Burn-
ham, 1980, 1981).
In combined culture, with a sensitive strain of
.cyanobacteria (nine have been found
(Bumham et al. 1983), the myxococci will form semispherical
colonies in which the prey cyanobacteria become concen-
trated in the middle of the colony (Fig. 1) surrounded by a
massive multicellular border of myxococcal vegetative cells
Figure 1 .—Phase contrast micrographs of an M. xanthus PCO2
colony containing the eyanobacterium Phormidium luridum in darker
core. The lighter periphery of the colony is of the swarm-
ing masses of myxococcai vegetative rod-shaped cells.
200 X
(Figs. 1, 2). This ability to form a core a
environment in which the lytic enzymes produced by the
myxococci can the prey. The enzymes for
this lytic to to the of autolysis
and production that characterize this Myxococcus
genus and White, 1974; Wireman and Dworkin, 1977).
Figure 3 the fruiting structures, comprised of swarm-
ing vegetative cells and myxospores, that develop on the
surface of the Myxococcus fuhfus BG02 strain when grown
in cultures of Phormidium luridum. Various en-
zymes including two with amidase activity, a
peptidase, an amidase and a gfucosaminidase have
isolated (Sudo and Dworkin, 1972). It that the for-
mation of myxospores upon a cell-free concen-
tration of cytoplasmie constituents from myxococ-
ci. The cellulytic enzymes are needed for this autolysis.
Because the by individual
Figure 2.—Bright field micrograph of a section from a parafin
embedded axenically grown colony of M. xanthus PC02. The
of myxococcal vegetative rod-shaped cells can be seen envelop-
ing the relatively hollow core. 325 X.
Figure 3.—Light micrograph of the fruiting structures (arrows) that
have formed on the of a predatory colony of M. fulvus
and P. luridum - 150 X.
either diffuse out of the colony or to the core of the col-
ony, it is to propose that this colonial core
becomes an enzyme sink gradually increasing in enzyme
concentration. Enzymatic digestion of the prey occurs
predominantly in this core with subsequent uptake of this
liberated nutrient by the myxococci in the periphery of the
colony.
The entrapment mechanism to the fim-
briae (Dobson and McCurdy, 1979) and the lipopolysac-
charide extrusions (Burnham et al. 1981) on the sur-
face of the vegetative (Fig. 4). These ap-
pear to bind (or the onto the col-
ony surface. The are gradually
moved to the core of the colony via the swarming (due to
gliding motility) of the cyanobacteria on the colony
The entire of entrapment, translocation, lysis, and
nutrient transfer is in Figure 5. It is important
250
-------
Causes and Control of Blue-Greens: Alternatives to Nutrient Control
Rgure 4.—This scanning electron micrograph shows the rod-like
M. PCO2 joined together by small protrusions from
the myxocoocal cell wall. Transmission electron microscopy has con-
firmed that are composed of iipopdysaccharide. The entangle-
ment of the fibers holds the sphere into and suggests a role
in the entrapment of the 7015X. (This figure is
reprinted from Burnham, 1981).
Rgure 5.— model of of filamen-
tous cyanobacteria in eutrophic waters (ew). The eyanobacteria (c)
are to grow autotrophically utilizing light (hv) and in-
organic fi). These filaments are captured by the colony using
the extrusions (f) from the myxococcal cell walls. Trartstocatton from
the surface to the cotony core occurs via swarming activity of the
myxococcal (m). During their life cycle in the col-
ony periphery the myxococci produce cellulytlc enzymes (e) during
the of formation (myxocysts, mc).Iheae enzymes lyse
many of the (Im). The enzymes dif-
fuse either out to the aqueous environment or into the colony core
where they accumulate. When the enter the core they
are by these enzymes releasing nutrient (n) which is subse-
quently by the growing (m) in the colonial
periphery.
6
Figure 6.—Effect on P. luridum density of serial 1% into
mature P. luridum cultures (between 10*-107 cells/ml).The-initial
culture contained approximately IV P. luridumlml and 10» M.
MvustrrA. Vertical lines represent transfer points.
to realize that this under autotrophic conditions.
Therefore, all nutrients for the survival and development of
the myxocoeci must originate with the cyanobacterium. That
this is efficient is demonstrated by the curve in
Figure 6 (from Burnham et al. 1983) which shows the se-
quential lysis of a fresh culture of the cyanobacterium P.
luridum about every 7 for a period of 2 months. The
myxococei were serially transferred at a volume ratio of 1
percent. We found inoculum levels could be very low. If 500
individual myxococci were present per ml, ©f 3 x 107
P. luridumlml population occurred in about 7 days. In other
words, it that the myxococci required a lag period
to the colony structure for lysis
of the cyanobacterial population. It is that early
growth and development of colonies from such low inocula
Is in part by the mpococcl using the secretion of
viable (Daft et al. 1975). Ward and Moyer
(1966) that populations could by
using when they demonstrated that the
number of in an cufture multiplied
relative to the amount of growth.
To the of the
the concept of was using an
actinomycete and the fulvus in
lytic colonies. Much of this work was done with Dr. Melvin
Daft at the Department of Biological University of
Dundee, Dundee, Scotland. Rgure 7 is a micrograph of one
such colony formed by an actinomycete from Dr.
Daft and the M. Mws BG02. These colonies
were formed by placing numbers of in a
dilute medium (Difco Broth plus 0.2 tryp-
tone) and shaking the mixture for 24 to 48 hours. The col-
onies that often as in Figure 7 with the
actinomycete predominantly occupying the of the col-
ony and the predominantly in the
periphery. Light microscopy of (Rg.
8) that the surface protrusion from this colony in-
tertwined rod of bofh predatory
When colonies were in cultures of P. luridum,
lysis of the occurred.
251
-------
Late and
Figure 7.— field of a cotony comprised
of M ftjfvus BGO2 predominantly in lighter appearing
periphery) and a tytic actinomycete (located
predominantly in the darker 225X.
, , ,,1 i ,,, V,, ,,,, -,.,. _, „ . . • •
fc- V -"r'^"i-" "'< -f*'- j-'s* .si."IB'-'.-''";. !_,;*,:,;,''%< •'''•;;',
fr--,,7»;, fc /,-;,>:'"?- V '-'}>a -.1: " : 1??V- -WT ',-:^'
fe?:r^>^""l^rf;;?;lvr^
i« ' P^ ' l ij S!"''";) f'1' ,!' ! * \ ^s t? * i T* " *' "v ,,, *''"""" ^ k l| '''
' '
Rgure 8.—Phase contrast micrograph of the surface of the
colony in Rg. 7. Intertwined of the ao-
tinomycete (a) can be along with the
myxococci (m) forming a pointed colonial protrusion. 1000X.
Rgure 9.—tor graph illustrating the reduction in P. luridum
chlorophyll a in old cultures due to the predatory activities of
an M. Mws alone (P + M. fates within
copredatory colonies with each of two lytic actinornycetes
(P + + C9 and P + + N6); and each actinomycete
alone (P + C9 and P + N6). The P. luridum contained 107 cells/ml
while the actinomycete predatory cultures were inoculated with 1
ml from 5-day cultures in a extract-glucose medium.
Both the actinomycete and M. fulvus colonies were fragmented in
a grinder and mixed in Difco Broth to allow formation
of the combined cultures prior to their being intro-
duced into the cyanobacterial prey cultures.
Figure 9 the a in an
P. culture 9 of interac-
tion with both and colonies. The
of the
C9 or N6 in fytic but not to
the we had for. Part of the
to be due to a less efficient trapping and translocation
mechanism. We this the actinomycete
with the swarming of the
and the of the colonies (Fig. 7),
the for digestion. con-
of lysing P. indicated that in
of the cyanobacterial lysis ap-
to be taking among the of the periphery
of the colony.
When other of were to
or muscorum, the amount of chlorophyll a
per ml in the over time
of as This result confirmed by
following nitrogen-fixing of the i.e., the
systems more productive than the con-
trols in medium). It
that this is partly by the interference by the
actinomycete and partly by the increased nutrient
from cyanobacterial that stimulated the growth of the
remaining —in other words, the growth rate
of the prey in the outstripped
the killing activity of the It must be remembered
that this numbers of both we
it will be to improve on this by chang-
ing the ratio of the predatory and eventually achieve
predation of the cyanobacterial prey.
Predatfort in Algal
Studies of microcosms (Fraleigh and Dibert,
1980; Cooke, 1967) provided an understanding of the
of nitrogen, phosphorus, and inorganic carbon con-
centrations on standing crop. These should pro-
vide an excellent for both experimental and
evaluation of myxococcal effects. We believe these
to be excellent complex in which to
252
-------
and Control of BIu^Qreens: Atttmatlvwi to Nutitent Control
test the lytte aWSfty of the This is with
the (Bumham, 1981) that mpoooo-
cel control be in
introduction into pond or lake
We would like to some of our preliminary
investigations using as they some
for predicting the of biological
control.
microcosms were in a modification
of the medium of Taub and Dollar, 1964, by the
by and Dibert, 1980. These were cultured
from the in the inorganic carbon studies
of FraJelgh and Dibert, 1980. We identified both
and in high concentrations. The domi-
nant genera in our microcosms to be
and
were and an amphiod.
Figure 10 shows a control microcosm with the very
i I
i i
I!
10
Rgure 10.—A microcosm (6-month oW) showing the
of the dominant and algal
text for description}.
community primarily in the of
the column. The test not and
were not during to the
6-month-old were distributed (750 ml)
into 1 liter and were
with high of (HP/ml and 107/mi)
to any on the
While inoculation with the low of did not
to significantly the the high
did. In both the control and microcosms a
maintenance of a of and in
chlorophyll a concentration to to similar (Rg. 11).
In chlorophyll a concentrations in the high-
microcosms to in the
and control microcosms. This was
to an Increase in the of A sam-
ple from one of the high-myxococci microcosms is shown
in Figure 12. This illustrates the high crop tht
as well as one of the major for myxococ-
cal i.e., the colonial form of of
Thus, to is that the
Rgure 11.—Changes in chlorophyll a in 5 ml extracts of
10 ml of the communities of control, lew-
and microcosms. For each treatment
there were 2 microcosms. Urn Is in frncuiar
ten with myxooocd. Chlorophyll a was by the acetone
method and the trichromatic formula of
and
. ,
Rgure 12.—Photomicrograph of the community In a microcosm 10
inoculaton with a high of myxocoocl The smailest
on the micrometer is 2 pm.
of a algal in the high-
microcosms.
in the high-myxocoeeal microcosms ap-
pear to as a of the interaction
of two factors: and nutrient In
other has found to In-
and the of non-
preyed-upon at the prey trophic (Paine,
and Hall,
et al. 1970— the in relative
of in the hlgh-myxocoecal
was in a of
253
-------
Lake Restoration, Protection and Management
tive predation by the myxococci preferred prey. However,
recycling of nutrients as a consequence of myxococseai
predation was probably also important. This factor may have
been responsible for the increase in standing cropsi of
chlorophyll a and of the Microcystls.
Shortly after inoculation of myxococci in the high-
myxococcal microcosms pH decreased, the concentration
of inorganic carbon increased and the percent saturation of
CO2 increased (Rgs. 13,14), These appear to have resulted
pH
10
8
10
Microcosm alone
Microcosm + High BG02
10
8
Microcosm + Low BGO2
i I
I V
r s r
0
s r
2
s r
6
DAYS
s r
9
Figure 13.—Changes in pH in control low-myxococcal, and high-
myxococcal microcosms. Cultures were grown under a 16/8 hour
light/dark cycle. Time is in days after Inoculation with myxococci.
S and R denote the beginning of the dark and light periods, respec-
tively. For each treatment there were two replicate microcosms.
from CC>2 released from myxoeoccal respiration and
respiratory activities of organisms feeding on organic mat-
ter supplied as a result of myxoeoccal predation. These
changes did not occur in the control (Fig. 15) or low-
myxococcal microcosms. However, they did occur in the
cultures of myxococci with Phormidium and of myxococci
INORGANIC CARBON
Microcosm* High BG02
CO2 % SATURATION
<5OO-
14
Figure 14.—Changes in inorganic carbon concentration (from alkalini-
ty titrations) and percent saturation of COz (calculated from bicar-
bonate concentrations and pH) in experimental cultures - microcosms
with high myxococci, microcosms with low myxococci and Phor-
midium cultures with high myxococci. Cultures were grown under
a 16/8 hour light/dark cycle. Time is in days after inoculation with
myxococci. S and R denote the beginning of the dark and light
periods, respectively. For each treatment there were two replicate
microcosms.
INORGANIC CARBON
Microcosm Control
CO, ! SATURATION
Microcosm Control
rsr ar 5r sr
02 69
DAYS
IS
Figure 15.—Changes in inorganic carbon concentrations (from
alkalinity titrations) and percent saturation of CO2 (calculated from
bicarbonate concentrations and pH) in control cultures - microcosm
without myxococci and myxococci alone (high dose).
alone. From the large effect In the cultures of myxococci
alone it seems reasonable to conclude that much of the
change in inorganic carbon and CO2 percent saturation in
high-myxococcal microcosms and in the myxococci and
Phormidium cultures was caused by myxoeoccal respiration.
Furthermore, because Fraleigh and Dibert (1980) have
found that inorganic carbon can be limiting in these
microcosms, the respiratory activity of the myxococci in in-
creasing the inorganic concentration in the environment pro-
254
-------
Causes and Control of Blue-Greens: Alternatives to Nutrient Control
bably stimulated algal growth and contributed to the increase
in the standing crops of chlorophyll a and Microcystis. This
may also have occurred in the co-predation experiment with
Anabaena and Nostoc described earlier and may have con-
tributed to the increase in chlorophyll a in these.experiments.
In addition, recycling of other nutrients, as a consequence
of myxococcal predation and breakdown of organic matter
in prey algal species, and the stimulated metabolism of
saprophage populations, also probably contributed to the in-
crease in chlorophyll a standing crop in the high-myxococcal
microcosms. By day 6, a holotrichous ciliate protozoan had
become abundant in the high-myxococci microcosms. These
were probably feeding on a combination of bacteria and
detritus from myxococcal predation and likely contributed to
nutrient recycling back to primary producers.
Thus, what appears to have occurred in the high-
myxococcal microcosms is the following sequence of events:
Upon inoculation, the myxococci began preying on sensitive
algal species. Myxococcal predation then provided soluble
organic matter for saprophagous bacteria and these bacteria,
myxococci, and paniculate detritus from myxococcal preda-
tion provided a food resource for eukaryotic particulate
saprophages. The metabolism of all of these yielded a high
rate of CO2 production, probably a large release of other
nutrients, and heterotrophic conditions in the microcosms
(as evidenced by the free CC>2 concentration being super-
saturated at both lights-on and lights-off between days 1 and
5). This then apparently stimulated growth on nonsuscepti-
ble species (in this case Microcystis) resulting in an increase
in the standing crop of chlorophyll a and subsequently a
return to autotrophic conditions in the microcosms. That
recovery occurred is indicated by the finding that free CC>2
became undersaturated at both lights-on and lights-off after
day 6, a situation that would occur only if the rate of
photosynthesis exceeded the rate of respiration over a
24-hour period. Consistent with this were conditions in the
myx.ococc\-Phormidium and myxococci alone cultures. In
these, where resistant primary producers were absent, the
CO2 concentration remained supersaturated.
This scenario is probably similar to that which would oc-
cur in a natural ecosystem. Encouraging about these results,
which are preliminary at best, is that the myxococcal predator
can apparently cause a directional change in the species
composition of the primary producer trophic level. Thus, while
this system as used apparently does not have the capacity
for broad spectrum control of algae it may be adaptable to
situations where problems are caused by one or several
species. In addition, by redesigning the predatory system,
for example by (a) incorporation of copredatory colonies, (b)
using preformed myxococcal predatory colonies, and (c) in-
troducing agitation, we believe that more active predation
can be achieved over a larger time period. Not only may con-
trol be possible but conditions may be created favoring more
desirable algal species and the energy resources tied up in
the problem species may become available in the food chain
leading to fish.
SUMMARY
In cultures of Phormidium luridum inoculated with the
predator Myxococcus fulvus BGO2 strain, control of the blue-
green algae was found and mechanisms of the predatory
interaction have been described. Studies of co-predation by
BG02 and two strains of actinomycetes suggest some in-
terference and a variable enhancement of BG02 effec-
tiveness. In fact, we were surprised to find that in some
cultures co-predation resulted in an increase in algal biomass
(chlorophyll a). Preliminary work with microcosms contain-
ing a diversity of algae suggests that selective predation by
Myxococcus occurred and densities of less sensitive algal
species increased. In the microcosms, predation by the Myx-
ococcus also decreased pH, and apparently increased the
density of protozoa resulting in the microcosm becoming
temporarily heterotrophic.
These results are encouraging in suggesting that biological
control of certain lake problems may be possible. Specifically,
where problems exist because of an excessive abundance
of one or several blue-green algal species, introduction of
a myxococcal predator may control these species and create
environmental conditions (esp. a localized lowering of the
pH) favoring more desirable algal species and may also in-
crease energy flow to fish. The latter would occur if myx-
ococcal predation made energy stored in otherwise unusable
blue-green algae more available to heterotrophs in food
chains leading to fish. Needed at this time are further studies
with microcosms as a prelude to field tests. The objective
of these would be to focus on and identify the behavior of
the predator in a community of organisms. These results
would then be useful in designing efficient studies in natural
ecosystems to determine the effect of added predators and
ascertain whether natural populations of predatory bacteria
play a role in shaping ecosystem communities.
ACKNOWLEDGEMENTS: The research reported in this paper was
supported in part by Grant B-086-OHIO from the Ohio Water
Resources Center and the Office of Water Resources and
Technology, U.S. Department of Interior. The technical assistance
of Susan Collart is appreciatively recognized.
REFERENCES
Beebe, J. M. 1941. Studies on the myxobacteria. 2. The role of myx-
obacteria as bacterial parasites. Iowa State J. Sci. 15:319-317.
Brooks, J. L, and S. I. Dodson. 1965. Predation, body size and com-
position of plankton. Science 150:28-35.
Burnham, J. C. 1975. Bacterial control of aquatic algae: Pages
120-125 in PL. Brezonik and J.L. Fox, eds. Water Quality Manage-
ment Theory of Biological Control. Rep. ENV-07-75-1. U.S. En-
viron. Prot. Agency and Univ. Florida, Gainesville.
. 1980. Myxobacterial entrapment and lysis of aquatic
cyanobacteria. Abstr. 2nd Int. Symp. Microb. Ecol. Coventry,
England. 26.3: 163.
. 1981. The utilization of bacteria in managing cyano-
bacterial populations: a review and update. Pages 230-258 in Proc.
Workshop on Algal Management and Control. Tech. Rep. No.
E-81-7. U.S. Environ. Prot. Agency, Las Vegas, Nev.
Burnham, J. C., S. A. Collart, and B. Highison. 1979. Entrapment and
lysis of Phormidium luridum by colonial spherules of a Myxococ-
cus species. Abstr. Ann. Meet. Am. Soc. Microbiol. 120.
_. 1980. Myxobacterial antagonism for Phormidium luridum.
Abstr. Ann. Meet. Am. Soc. Microbiol. 16.
. 1981. Entrapment and lysis of the cyanobacterium Phormi-
dium luridum by aqueous colonies of Myxococcus xanthus PC02.
Arch. Microbiol. 129:285-294.
Burnham, J. C., S. A. Collart, and M. J. Daft. 1983. Myxobacterial
predation of the cyanobacterium Phormidium luridum in aqueous
environments. Submitted to Arch. Microbiol.
Cooke, G.D. 1967. The pattern of autotrophic succession in
laboratory microcosms. Bioscience 17:717-721.
Daft, M. J., and W. D. P. Stewart. 1973. Light and electron micro-
scope observations on algal lysis by bacterium CP-1. New Phytol.
72:799-808.
Daft, M. J., S. B. McCord, and W. D. P. Stewart. 1971. Bacterial
pathogens of freshwater bluegreen algae. New Phytol. 70:819-829.
_. 1975. Ecological studies on alga-lysing bacteria in fresh
waters. Freshw. Biol. 5:577-596.
Dobson, W. J., and H. D. McCurdy. 1979. The function of fimbriae in
Myxococcus xanthus. 1. Purifications and properties of M.xanthus
fimbriae. Can. J. Microbiol. 25:1152-2260.
255
-------
Lake Restoration, Protection and Management
Echlin, P., and I. Morris. 1965. The relationship between blue-green
algae and bacteria. Biol. Rev. 40:143-187.
Fallen, R. D., and T. D. Brock. 1979. Decomposition of bluegreen al-
gal (cyanobacterial) blooms in Lake Mendota, Wis. Appl. Environ.
Microbiol. 37:820-830.
Fitzgerald, G. P. 1969. Some factors in the competition or antagonism
among bacteria, algae and aquatic weeds. J. Phycol. 5:351-359.
Fraleigh, P. C., and P. C. Dibert. 1980. Inorganic carbon limitation
during ecological succession in aquatic microcosms. Pages
369-401 in J. P. Giesy, Jr., ed. Microcosms in Ecological
Research. U.S. Tech. Inf. Center, U.S. Dep. Energy Symp. Ser.
52 (CONF-781101).
Gunnison, D., and M. Alexander. 1975. Basis for the susceptibility of
several algae to microbial decomposition. Can. J. Microbiol.
21:619-628.
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. 15:839-928.
Huffaker, C. B., F. J. Simmonds, and J. E. Laing. 1976. The theoret-
ical and empirical basis of biological control. Pages 41-78 in C.
B. Huffaker and P. S. Messenger, eds. Theory and Practice of
Biological Control. Academic Press, New York.
Kottel, R., and D. White. 1974. Autolytic activity associated with myxo-
spore formation in Myxococcus xanthus. Arch. Microbiol. 95:91-95.
Lange, W. 1971. Enhancement of algal growth in cyanophyta-bac-
teria systems by carbonaceous compounds. Can. J. Microbiol.
17:303-314.
Paine, R. T. 1969. A note on trophic complexity and community sta-
bility. Am. Nat. 103:91-93.
Rittenberg, S. C., and M. F. Thomashow. 1979. Intraperiplasmic
growth-life in a cozy environment. Pages 80-86 in D. Schlessinger,
ed. Microbiology— 1979. Am. Soc. Microbiol. Washington, D.C.
Shilo, M. 1967. Formation and mode of action of algal toxins. Bac-
teriol. Rev. 31:180-193.
_. 1970. Lysis of blue green algae by Myxobacter. J. Bacteriol.
104:453-461.
Stewart, J. R., and R. M. Brown. 1969. Cytophagn that kill or lyses
algae. Science 164:1523-24.
Stewart, W. D. P., and M. J. Daft. 1977. Microbial pathogens of cy-
anophycean blooms. Pages 177-218 in M. R. Droop and H. W.
Kannasch, eds. Advances in Aquatic Microbiology. Vol. 1,
Academic Press, London.
Strickland, J. D. H., and T. R. Parsons. 1965. A manual of sea-water
analysis. Fish. Res. Board Can. Bull No. 125.
Sudo, S., and M. Dworkin. 1972. Bacteriolytic enzymes produced by
Myxococcus xanthus. J. Bacteriol. 110:236-245.
Taub, F. B., and A. M. Dollar. 1964. A CWore//a-Daphnia food chain
study: the design of a compatible chemically defined culture
medium. Limnol. Oceanogr. 9:61-74.
Ward, C. H., and J. E. Moyer. 1966. Ecologic relationships between
bacteria and algae in mass culture. Paper presented at AIBS-
NASA sponsored conference on biogenetics systems. Washington,
D.C. Nov. 15-16.
Wireman, J. W., and M. Dworkin. 1977. Developmentally induced
autolysis during fruiting body formation by Myxococcus xanthus.
J. Bacteriol. 129:796-802.
256
-------
Problems in Lake Restoration
REVIEW OF LAKE RESTORATION TECHNIQUES AND AN EVALUATION
OF HARVESTING AND HERBICIDES
G. DENNIS COOKE
Department of Biological Sciences
Kent State University
Kent, Ohio
ABSTRACT
Procedures with proven effectiveness in providing long-term, ecologically sound control of eutrophication
symptoms include nutrient and silt diversion, dilution/flushing, phosphorus inactivation, sediment removal,
lake level drawdown, and sediment covers. Procedures requiring further demonstration and/or scientific
evaluation are hypolimnetic withdrawal, sediment oxidation, biological controls, artificial circulation,
hypolimnetic aeration, harvesting, and herbicide-algicide applications. Harvesting may benefit lakes because
biomass removal may interrupt nutrient release, tissue decay, and oxygen consumption, and prevent deposi-
tion leading to loss of lake volume. Costs are equal to or less than chemical treatments. Research is needed
on cutting techniques. Data are presented to support the conclusions that herbicide-algicide treatments
are only briefly effective, stimulate nutrient release, increase productivity, and promote oxygen depletion
and invasion of resistant or opportunistic species. Further experiments are needed to evaluate questions
about these chemicals, emphasizing studies of processes at the actual level of biological organization to
which they are applied.
INTRODUCTION
Eutrophic in-lake management and restoration techniques
are procedures whose object is to attempt to bring about
long-term lake improvement through ecologically sound
methods that retard plant growth and/or improve conditions
for lake organisms. Presently, lakes may be improved
through removing selected species, by altering plant
substrate or nutrient supply, by oxygenating or circulating,
or by enhancing the activities of pathogens, predators, or
herbivores. "Long-term" means an attempt to relieve obnox-
ious lake conditions by methods that are effective over
periods of years, not days or weeks. "Ecologically sound
methods" are those that do not include the introduction of
toxic substances whose environmental impact is unknown
and/or unacceptable, nor manipulations for which the scien-
tific evidence regarding possible deleterious effects is scanty
or absent.
The purpose of this report is to briefly examine these ques-
tions: Which lake restoration techniques are effective and
ecologically sound, which are not, and which need more
study? An in-depth examination of these questions is found
in a critical review of the literature by G. D. Cooke, E. B.
Welch, S. A. Peterson, and P. R. Newroth (Cooke et al.
1983).
I have divided this report into sections about techniques
that have been shown to be effective and sections about
those that need more research. I have emphasized har-
vesting and chemical treatments, adding new data regard-
ing their effectiveness.
257
-------
Lake Restoration, Protection and Management
TECHNIQUES WITH PROVEN
EFFECTIVENESS
Procedures to Reduce Nutrient
Concentration
Nutrient Diversion
Diversion of nutrient income is a widely recognized first step
to improving lake trophic state. While it may be possible to
reduce the magnitude of algal blooms by such techniques
as food web manipulation or removal of species such as the
common carp, diversion is highly desirable and in most
cases, essential.
Uttormark and Hutchins (1980) examined the responses
of 23 lakes to diversion and found that while nutrient con-
centration (phosphorus in this case) may have declined, in
only 30 percent of the cases had the trophic state improv-
ed. This may have occurred because diversion was insuffi-
cient to lower concentration to the level that would produce
a mesotrophic lake. Even so, a reduction in algal biomass,
an absence of blue-green scums, and a return of fecal col-
iform colony counts to the acceptable range would make the
diversion a success.
Another reason for a failure to change trophic state is
because of continued sediment and macrophyte release of
nutrients into the water column. This factor, along with mor-
phometric and hydrologic features such as mean depth and
flushing rate, are important in determining how much con-
centration will change after diversion and how long it will take
to reach a new, lower concentration. Ultimately, internal P
release should decline as older, richer sediments are covered
with those of lower nutrient content; lake concentration
should then also decline.
Diversion is thus the first step in stopping eutrophication
and in lowering the concentration of nutrients. Recovery of
the lake to a lower trophic state, however, may be slow, re-
quiring an in-lake treatment such as dilution or sediment
removal and/or watershed controls of nonpoint nutrient
sources such as agricultural runoff.
Silt income control may be very significant in preventing
the rapid development of shallow water and the spread of
nuisance macrophytes. Unless the siltation problem is re-
cent, an extensive littoral area might then require removal
or other procedures before macrophyte growth is controll-
ed. Control of income from the watershed is thus again an
essential first step to improvement in trophic state.
Dilution/Flushing
Dilution is a technique to reduce the concentration of
nutrients in the water column by adding nutrient-poor water.
Flushing is a means of reducing algal biomass by increas-
ing the loss rate of cells. Both effects occur if large amounts
of low-nutrient water are added.
Dilution has been very effective in Moses Lake and Green
Lake, Washington (Welch and Patmont, 1980; Oglesby,
1969). Welch (1981) concluded that the best plan is to have
a continuous low-rate input of low-nutrient water, coupled
with nutrient diversion. Otherwise high-rate inputs would be
required.
While this procedure has been shown to bring about lake
improvement, its general applicability is limited because of
the general absence in most watersheds of nutrient-poor
waters that could be diverted into eutrophic lakes.
Phosphorus Inactivation/Precipitation
The purpose of this technique is to lower the concentration
of P in the water column by either precipitating it or by
preventing its release from sediments (inactivation). Salts of
aluminum (aluminum sulfate or sodium aluminate) are add-
ed to the lake surface or to the hypolimnion and P is sorbed
to the surface of the aluminum hydroxide floe. In the case
of inactivation, sufficient aluminum is added to create a bar-
rier to P release from anaerobic sediments.
Like the use of herbicides and algicides, this technique
can create toxic conditions. This situation could occur through
an increase in soluble aluminum (Al+++) or through a de-
crease in pH. Kennedy and Cooke (1982) have described
a method for determining the maximum dose of aluminum
sulfate for a lake, based on lakewater alkalinity. An alternative
dose procedure for soft-water lakes has been described by
Dominie (1980). These procedures allow for the maximum
in P inactivation without creating toxic conditions.
The P inactivation technique is particularly effective in long-
term control of P concentration in the water column. Cooke
and Kennedy (1981) have reviewed the published case
histories of this technique. The longest period of P control
to date is 6 years (Cooke et al. 1982). Failures have resulted
from insufficient dose, polymixes, and insufficient nutrient
diversion.
Sediment Removal
Sediment removal is one of the most commonly prescribed
techniques for long-term lake improvement. Lake deepen-
ing, and the removal of toxic materials, macrophytes, and
nutrient-rich sediments are its main purposes. Until recent-
ly (Peterson, 1982), little documentation had been publish-
ed about its effectiveness, environmental impact, and costs.
Lake deepening projects are generally successful except
in instances where sediment income remains high. Similar-
ly, nutrient release from sediments can be controlled by
removal, although it is less costly to do so through P inac-
tivation (Cooke et al. 1983). There is little documentation of
macrophyte control by dredging.
There is widespread concern about the negative en-
vironmental impact of this procedure. Most negative effects
(algal blooms, high turbidity) are of short duration. One com-
mon problem is inadequate sizing of the disposal area. Dik-
ing in upland areas is the most frequent method of contain-
ing the dredge spoil, but dikes have failed and wells have
been contaminated.
Dredging can be expensive, with costs ranging from $0.22
to $13.93/m3. Removal of contaminated material may ex-
ceed $25.00/m3. One way of lowering costs is to sell dredge
spoil for topsoil.
Procedures to Control Algal and
Macrophyte Biomass
Drawdown
Lake level drawdown is used successfully to control the
growth of certain macrophyte species, to consolidate floc-
culent lake sediments, to provide an opportunity to repair
docks and dams, to remove sediment or to install sediment
covers, and for fish management. A detailed review has been
provided by Cooke (1980a).
Not all aquatic plants are susceptible to the freezing or
high temperature conditions of a winter or summer
drawdown. Alligatorweed (Altemanthera pn/toxero/des) and
A/a/as ffex/fes (brittle naiad) have been reported to actually
increase in density after a drawdown, whereas water lily
(Nuphar sp.), water hyacinth (Eichhomia crassipes), and musk
grass (Chara vulgaris) have been found to always decrease.
For most macrophytes, the published data are too sparse
to generalize. A danger may be that susceptible species will
be replaced by tolerant ones, leading some investigators to
suggest that the best control will be achieved by 2 to 3 years
258
-------
Problems in Lake Restoration
of drawdown, followed by 2 years of stable water levels (Lantz
et al. 1964; Lantz, 1974). Most drawdowns have been dur-
ing the winter since this does not interfere with most recrea-
tional uses, the lake will probably re-fill during spring melt
and rainfall, and terrestrial plants will not invade at that time.
There are insufficient published data to compare the effec-
tiveness of winter and summer drawdowns.
Drawdown has been effective in fish management. Lantz
et al. (1964) reported that in Louisiana reservoirs, popula-
tions of shad and sunfish are removed by winter drawdown
and their spawning prevented by summer drawdown. In
reservoirs with 5 or more consecutive years of drawdown,
fish standing crop and game fish size and reproduction have
increased.
There are negative effects of water level drawdown. Algal
blooms have occurred upon refilling, possibly from nutrient
release from rewetted, highly organic sediments (Plotkin,
1979). Fish kills in pools of water, and great changes in the
macroinvertebrate community have been reported.
This low cost management technique can be effective in
macrophyte control, and can be beneficial in the sense that
it gives management personnel access to the lake to enact
other techniques.
Sediment Covers
Several sheeting materials have been tested for their effec-
tiveness in preventing macrophyte growth. These studies
have been reviewed by Cooke (1980b). Impermeable
materials such as polyethylene have been found to be ef-
fective but troublesome because of the accumulation of
gases beneath them and the difficulty in applying them. One
way to prevent "ballooning" is to cover them with gravel and
silt, but this has usually provided sufficient substrate for plant
re-growth.
One material, Aquascreen, a PVC-coated fiber glass
screen, has been shown to be completely effective in preven-
ting macrophyte growth, is easy to install, does not balloon,
but is very costly ($140/700 ft2; $2.15/m2). Mayer (1978)
believed that the screen controlled macrophytes through light
limitation, but the study of Perkins et al. (1980) strongly sug-
gested that control was through physical or space limitation.
In most instances, the screens must be re-positioned annual-
ly to remove accumulated silt. Because of its cost,
Aquascreen appears to be best suited for small areas such
as docks, beaches, or waterfronts.
TECHNIQUES REQUIRING MORE
RESEARCH AND DEMONSTRATION
Procedures to Reduce Nutrient
Concentration
Hypolimnetic Withdrawal
The object of this technique is to siphon nutrient-rich
hypolimnetic waters from the lake. This should reduce the
impact of vertical entrainment of nutrients on the epilimnion
and it will remove nutrients from the lake. The siphon is
sometimes called an "Olszewski tube" after its originator
(Olszewski, 1961).
There are apparently no literature reports of the use of
this device in North America. Bjork (1974) listed nine applica-
tions in European lakes. Gachter (1976) and Pechlaner
(1978) describe successful experiences with the siphon in
small lakes. Total P and internal P release decreased,
transparency increased, and fewer blue-greens were present.
This technique deserves further study. It should be de-
pendable and inexpensive. The high P-low dissolved oxygen
content of the hypolimnetic water could pose a threat to
receiving waters. Perhaps the lake discharge could be
used in irrigation. Other possible problems could include the
disruption of the thermocline and excessive water loss.
Sediment Oxidation
Ripl (1976) has described a procedure for oxidizing the
top 15 to 20 cm of anaerobic lake sediment to reduce in-
ternal nutrient release. There are very few published data
regarding the success of this new procedure and its costs.
Chemically, Fe (CI3)is first added. Ferric hydroxide for
P binding is formed and the sulfur in FeS is lost as H2S.
Calcium hydroxide is added next to precipitate iron and
to increase pH to the optimum for denitrification. Calcium
nitrate is finally added to the sediments, oxidizing reduc-
ed carbon and liberating N as a gas. The chemicals are
added by a harrow. A positive result (lower P, NH4 de-
crease, 62 demand lowered) was found in Lake Lillesjon,
Sweden (Ripl and Lindmark, 1978), but costs of chemicals
and equipment were high.
Further laboratory and field evaluations of this tech-
nique are needed.
Procedures to Control Algae and
Macrophyte Biomass
Biological Controls
Biological control methods may be the most promising
because of the possibility of controlling nuisance plant den-
sity without eradication or the use of expensive machinery
or toxic chemicals. Significant progress has been made
in recent years.
Among the most controversial of the proposed biological
control methods is the introduction of the grass carp or
white amur (Ctenopharyngoden idella Vol.), a voracious
consumer of macrophytes. Several apparently successful
introductions have taken place (Mitzner, 1978; Shireman
and Maceina, 1981), and the U. S. Army Corps of Engi-
neers and others are conducting large scale feasibility
studies in Florida. The animal, however, was used as a
macrophyte control agent in the United States long before
adequate scientific studies were conducted on adverse
impacts. While new data are now being accumulated, we
still don't clearly understand the effects of these animals
on other species, their role in nutrient recycling, whether
they transmit fish diseases, and whether or not they will
reproduce and infest non-target waters. Currently the fish
is widespread in waters such as the Mississippi River. We
will be fortunate if this animal proves to be useful and in-
nocuous in North America since it was widely released
before these questions were asked.
A plant pathogen (Cercospora rodmanii) appears to be
effective in controlling water hyacinth (Eichhomia cras-
sipes) and is now being prepared for a large scale test by
the U. S. Army Engineers (Freeman, 1977; Freeman et
al. 1981).
Monophagous insects have been introduced in southern
states for aquatic weed control, following careful screen-
ing under quarantine. Two beetles and a moth have been
released as controls of water hyacinth (Center, 1981). Their
effectiveness is currently under study.
Since the initial report of LaMarra (1975) concerning the
nutrient recycling ability of the common carp, others have
further demonstrated the eutrophying activities of benthi-
vorous fish (Andersson et al. 1978; Keen and Gagliardi,
1981). Shapiro et al. (1975) and Shapiro (1978) have coin-
ed the word "biomanipulation" to include lake improve-
ment activities in which fish such as bullheads are re-
moved, or in which the pelagic food web is altered to
decrease predation on the large herbivorous zooplankton.
These and other procedures, described elsewhere in this
259
-------
Lake Restoration, Protection and Management
volume by Shapiro, deserve considerable further support and
attention. These methods may, for example, be a way to im-
prove lakes where significant nutrient diversion is not
possible.
Artificial Circulation
The purpose of this procedure is to prevent thermal stratifica-
tion or to destratify the lake. The most effective and least
expensive technique is to introduce compressed air at max-
imum depth, creating a curtain of bubbles and intense mix-
ing. Lorenzen and Fast (1977) calculate that an air flow rate
of 9.4 m3/min./km2 should be adequate. Pastorok et al.
(1981) found that this flow rate was achieved in only 20 per-
cent of the cases reviewed, a factor which may account in
large part for some of the reported failures of the technique.
Artificial circulation is included in this section on "tech-
niques requiring more research" because the expected
results often do not occur. The principal improvement is in-
creased aeration of the water column and expanded habitat.
Undersizing the flow rate may prevent this from happening.
Theoretically, aeration and circulation should inhibit P release
from sediments, limit algal productivity by increasing cell
residence time in dark waters, decrease pH and perhaps
shift the competitive balance from blue-green algae to greens
and diatoms, limit the survival value of blue-green pseudo-
vacuoles (buoyancy), and favor the survival of herbivorous
zooplankton. Phytoplankton have decreased in about half
the cases examined by Pastorok et al. (1981). Transparen-
cy has decreased more often than increased. Circulation may
introduce nutrients to the water column, and in nutrient-
limited lakes an increase in biomass may occur. Perhaps
this technique might be most effective in lakes where
nutrients are not limiting. Additional research is needed.
Hypolimnetic Aeration
This is a technique to aerate the hypolimnion without achiev-
ing destratification. It is accomplished by an "air-lift" in which
water is lifted up a cylinder by compressed air to the lake
surface, aerated, and then returned to the hypolimnion. The
procedure is reviewed by Pastorok et al. (1982).
Successful oxygenation is usually achieved, although
metalimnetic oxygen minima may occur with an undersized
aerator. Phosphorus concentration may be reduced follow-
ing the introduction of oxygen. There are few data on bio-
logical effects, including that of phytoplankton, zooplankton,
and fish (Pastorok et al. 1982). More documentation is need-
ed about the responses of lakes to this treatment.
Harvesting
Harvesting is one of the more controversial procedures, in
part because there is so little scientific evidence to demon-
strate any short- or long-term effectiveness and the tech-
nique has therefore been categorized with herbicides as a
cosmetic treatment. Yet unlike the introduction of toxic ma-
terials to lakes, nutrients and oxygen-consuming biomass
are removed during a harvest and there is some evidence
that control of macrophytes can be achieved by using the
correct cutting frequency and cutting technique. The pur-
pose of this section is to describe the advantages and disad-
vantages of harvesting, to present some new data on effec-
tiveness and costs, and to discuss the technique as a poten-
tial lake improvement procedure.
Advantages
1. Nutrients and organic matter are removed from the
lake;
2. There is little or no interference with recreation, such
as the 10 to 14-day waiting periods associated with some
herbicides;
3. A licensed operator is not needed so that lake associa-
tions may themselves control where and when to harvest;
4. Cut plants have been found to be a useful garden
mulch;
5. No potentially deleterious substances are added to the
lake;
6. Non-target species are unaffected, except those in-
advertently removed during harvest; and
7. Costs are generally equal to or less than other
macrophyte control measures.
Disadvantages
1. Operation of the harvester may create turbid waters,
release nutrients from the sediments, and stimulate blooms
of algae;
2. Plant fragments not removed may be dispersed from
the site, thus spreading the plant infestation;
3. Plant removal may be needed in many areas of the
lake simultaneously;
4. The initial costs are high;
5. Small young-of-the-year game fish may be removed;
and
6. Obstacles such as docks, posts, rocks, and logs may
prevent plant removal.
Several recent case histories strongly suggest that effec-
tiveness of harvesting in controlling plant growth, at least
during the season of harvest, is related to harvest frequen-
cy and to the cutting technique.
Figure 1, from Conyers and Cooke (1982; also this volume)
illustrates an effective harvest operation, using the
Aquamarine "Chub" harvester. The regrowth of Cerato-
phyllum, Najas, Chara, and Potamogeton in Twin Lakes was
much less in the harvested area than in the control or the
chemically (Diquat and Cutrine) treated areas. Macrophyte
biomass was persistently and significantly less during the
remainder of the growing season. A directly contrary result
was obtained in LaDue Reservoir (Portage County, Ohio) in
1982, following a July harvest of a Myrop/7y//iym-infested bay
1
A B
pr*
poll i poll 2
IU
» B * BA H^
pr« pail 1 pot 2
> 400
s
I
1
polll
poll 2
C B C B
poll 1 poll 2
Figure 1 .-Pre- and post-treatment biomass of aquatic macrophytes
(gms. dry wt./m2) in control, harvested, and chemically treated plots
in East Twin Lakes, Ohio. A = harvest plot, B = control plot,
C = chemical plot. Depth of water in shallow sampling area ranged
from 0.5 to 1.0 meters, in deep area, 1.0 to 1.6 meters (Convers
and Cooke, 1982).
260
-------
Problems in Lake Restoration
200
150
04
5
a:
O
50
HC
I
6-22
7-8
1982
7-21
8-9
Figure 2.—Pre- and post-harvest biomass of aquatic macrophytes
(gms. dry wt./m2) in harvested and control areas of a bay in LaDue
Reservoir, Ohio. Student t-test shows no significant difference bet-
ween means (n = 6 for each area) except on 7-21-82. The bay was
harvested between 7-11 and 7-19-82.
with an Aquamarine H-650 harvester (Fig. 2). Myriophyllum
biomass was identical in the harvested and control areas
within 19 days of the harvest, although the control area was
far more "obnoxious" than the harvested since plants there
were undergoing senesence and also supported a large
biomass of filamentous algae. An algal bloom composed
mainly of cryptomonads, dinoflagellates, and green algae
followed the harvest.
I believe that the differences between these two harvesting
experiences are explained by differences in harvesting tech-
nique, the size of the area cut, differences in macrophyte
species, and the lowered water level in LaDue Reservoir after
the harvest. In the successful harvest (Fig. 1), the cutter blade
was used directly on the sediment surface and removed
many more root systems than were removed at LaDue. In
addition, the dominant plant in East Twin was the non-rooted
Ceratophyllum demersum. After the harvest in LaDue, cut
shoots of Myriophyllum spicatum 2 to 6 cm. in length were
found throughout the bay; these regrew rapidly. Also, the
harvester failed to cut some plants, or missed them. A scuba
evaluation revealed some intact plants that had been
pushed over by the harvester.
Regrowth in LaDue was enhanced when the water level
was sharply lowered after the harvest to augment a down-
stream reservoir. Depths after this did not exceed 1.0 m
(mainly less than 0.5 m) and the cut shoots were exposed
to much more light. The algal bloom in the bay was probably
caused by the disturbance of nutrient-rich reservoir
sediments by the harvester's paddle wheels since there was
a negligible nutrient income to the bay from allochthonous
sources. In the successful Twin Lakes harvest, a much
smaller harvester was used; many days were required to
harvest an area equal to the area harvested in 5 days (10.4
ha) with the H-650, and therefore the amount of nutrients
released per day was small. This probably explains the
absence of an algal bloom in Twin Lakes.
An earlier example of the effectiveness of cutting the plants
at the sediments was reported by Nichols and Cottam (1972).
As with the Twin Lakes experience, they found that a single
harvest sharply reduced biomass. Three harvests virtually
eliminated plant material and the effect carried over to the
next year. Two harvests per year were nearly as effective
as three. Wile et at. (1977, 1979) also found that regrowth
was inhibited for longer periods when the plants were cut
close to the sediment, and that two to three harvests per
season were most effective in reducing biomass.
At this point in the development of harvesting, it is clear
that additional experimentation is needed on how to use the
machine, on how often to harvest, and on the response of
various species. Machine operators are interested in speed
and in protecting the cutter bar from obstacles. Yet if the
cutter is allowed to just skim the sediment surface, slowing
the machine and exposing the blades to possible breakage,
it seems that the harvest will be more effective because root
systems will be cut or removed. Regrowth may be more rapid
in deeper water where the cutter cannot reach the sediment
surface.
Is harvesting a lake restoration technique? To be so, large
amounts of nutrients must be removed (relative to nutrient
loading), or the control of plant biomass must be better than
just the brief cosmetic effect of a few days, or the removal
of plants must interrupt some process contributing to
eutrophication, such as sediment buildup or internal nutrient
loading. Obviously, the long-term effectiveness of plant
removal in controlling biomass is poorly understood.
Burton et al. (1979) and King and Burton (1980) reviewed
harvesting as a technique to reverse the nutrient enrichment
of lakes by removing nutrients in the harvested biomass.
They concluded that most eutrophic lakes have an external
loading well in excess of the amount that could be removed
in a harvest. The difficulty of removing enough biomass is
compounded by the need to have very high macrophyte den-
sity, a P input less than 1 gm P/rrr/year, large areal
coverage, and annual regrowth. This latter condition may
not be met if proper harvesting procedures (two to three
harvests per year, cutting into root systems) are followed.
Nevertheless, Wile et al. (1979) report a removal of 92 per-
cent of net external loading during the 1975 harvest at
Chemung lake, Ontario. Conyers and Cooke (this volume)
report a range of 13 to 20 percent of gross external loading
and up to 100 percent of net external loading which could
be removed from the Twin Lakes (assuming a single harvest
of half the littoral area), and Welch et al. (1979) calculated
that in Long Lake, Wash, a potential removal of 60 percent
of gross external loading could be accomplished if 41 per-
cent of the lake's surface area were harvested each year.
Thus sufficient nutrient removal to counter a substantial frac-
tion of external loading does appear to be possible.
A more direct beneficial effect of plant removal may be
in the interruption of nutrient and organic matter release to
the water column by sloughing and by decay of plant tissue.
The continual loss of plant tissue appears to be a signifi-
cant source of nutrients to the pelagic zone (Hill, 1979; Pren-
tki, 1979; Barko and Smart, 1980; Carpenter, 1981; Landers,
1982).
Also, sedimentation of detritus contributes to the oxygen
depletion at the sediment-water interface and the release of
nutrients such as phosphorus to the water column under
reducing conditions (Rich and Wetzel, 1978). Also, macro-
phyte remains add to the loss of lake volume. Both the in-
ternal loading from tissue sloughing, plant decay, and from
reduced sediments, and the loss of lake volume, may close
a positive feedback loop that increases the area of sediments
available for macrophyte colonization (Carpenter, 1981).
Harvesting therefore appears to have a value in addition to
any short- or long-term control of macrophyte biomass to im-
prove recreation and lake appearance. Removal of macro-
phyte tissue, whether or not there is a carry-over effect to
the next year in terms of biomass control, should help pro-
tect the lake from filling in and may contribute to control of
pelagic algal blooms. This effect perhaps more than any
other may make harvesting superior to the use of herbicides.
A thorough evaluation of this idea is needed.
The costs of harvesting are no higher than herbicides in
most cases, and can be substantially lower. Costs do vary
widely, however. For example, the harvesting operations in
the Okanagan Valley, British Columbia were very high. This
appears to have been due to the size of the lakes, to the
large distances between cutting sites and disposal areas,
261
-------
Lake Restoration, Protection and Management
Table 1. — Average cost, cost range, and number of observations for aquatic plant harvesting in North America. Values for
Canadian experiences in Canadian dollars.
1
Chemung
Lakes
2
Okanagan
Valley
3
North
American
4
Wisconsin
5
Dane Co.,
Wisconsin
6
Ohio
7
Florida
Number of
observations
Mean cost per
hectare
acre
Range of costs per
hectare
acre
35
$345
$140
$282-390
$114-158
$1088
$440
$976-1258
$395-509
$551
$223
$148-1483
$ 60-600
$282
$115
$186
$ 75
$370-490
$150-198
$168-232
$ 68-94
1. Wile et al. (1979) Canadian dollars
2. P. R. Newroth (pers. comm.)
3. Smith (1979)
4. Dunst and Nichols (1979)
5. Livermore and Koegel (1979)
6. Conyers and Cooke (1982)
7. Sassic (1982)
and to the fact that plants were distributed in a narrow band
around the lake.
Table 1 is a comparison of the cost and cost range for
harvesting in North America, including salaries. If the costs
for the Okanagan Valley operation and for Oneida lake are
omitted from the survey, an omission that is justified because
of the physical problems at Okanagan Valley and the ineffi-
ciency of the Oneida Lake project (Smith, 1979), then the
range of cost (44 observations) for harvesting is $148/ha to
$490/ha ($60 to $198/acre). Fuel-maintenance costs seem
to range between 5.8 percent and 25 percent of total
expenditures.
Harvesting is apparently less costly than herbicide
treatments. Dunst and Nichols (1979) reported a cost range
(n = 300 lakes) of $250 to $1,077/ha ($100 to $436/acre) for
herbicide use in Wisconsin. Sassic (1982) estimated the cost
range for Florida lakes to be $371 to $741/ha ($150 to
$300/acre). If the harvesting costs from Table 1 are represen-
tative ($148 to $490/ha), then harvesting, in addition to hav-
ing several potentially beneficial effects on the lake, can be
less costly than herbicide treatment.
To summarize, much remains to be learned about
harvesting. There is evidence that it can control macrophyte
biomass if there are two to three harvests per season and
if the equipment is used properly in shallow water so that
root systems are cut. There are some serious drawbacks
to harvesting, including the possibilities of stimulating an algal
bloom and of removing substantial numbers of young-of-the-
year sport fish (Haller et al. 1980). However, even without
any long-term control of plant regrowth from year to year,
harvesting appears to have the possibility of improving the
lake through nutrient removal and through removal of littoral
biomass which otherwise would contribute to filling-in of the
lake and to nutrient cycling. Harvesting costs are apparent-
ly less than herbicide costs. Much additional research is
needed with harvesting.
Herbicides and Algicides
Introduction
The purpose of adding herbicides and/or algicides to lakes,
ponds, and reservoirs is to alleviate problems with nuisance
macrophytes or algae by killing the standing biomass. There
are several very effective chemicals for this purpose including
Diquat, endothall, 2,4-D, and copper sulfate. Their use con-
stitutes the most widely accepted form of treating eutrophic
water bodies in North America. The desired results are usual-
ly obtained promptly, and with repeated application over the
growing season can assure reasonable freedom from nui-
sance plants.
All users of these chemicals recognize their great and in-
creasing expense, the need to have a licensed and bonded
applicator who is available at the desired time, and the im-
portant and unanswered questions about their effectiveness
and the short- and long-term consequences of their use. Lake
users and lake managers often ask (1) If these chemicals
are so effective, then why is there such controversy over their
use and why does the lake management community spend
so much time and money trying to develop alternative
methods of controlling macrophyte and algae biomass? (2)
Why do so many professional lake managers rarely advise
the use of these chemicals, or at least recommend that they
be used sparingly or in conjunction with other techniques?
There are concerns about the use of these chemicals, in-
cluding (from Conyers and Cooke, 1982):
1. Plants release nutrients to the water column upon death
and decomposition (Simsiman et al. 1972; Morris and Jar-
man, 1981);
2. Oxygen is depleted at the sediment surface by the
microbes that colonize the decaying plants; this may be fol-
lowed by a release of nutrients from these fertile sediments
(Simsiman et al.1972; Rich and Wetzel, 1978; Carpenter and
Greenlee, 1981);
3. Herbicides can be toxic to non-target species (Ander-
son, 1981);
4. Some plant species may be tolerant to the herbicide
and replace the killed species (Reiser, 1976; McKnight,
1981);
5. Some herbicides in current use are suspected to be
mutagenic and carcinogenic (Shearer, 1980; Valencia, 1981);
6. The waiting period following application (10 or more
days in some cases) interferes with recreational use;
7. Some herbicides are not registered for use in potable
waters; and, perhaps most importantly,
8. Herbicide use does not attack the causes of the
nuisance plant problem, nor does it address conditions for
plant growth (high nutrient concentration, shallow water, ox-
ygen depletions, etc.) They are not lake restoration agents
in any sense.
Some Experiments
My students and I have examined the effectiveness of
several chemicals, the effects of their use on the lake, and
their costs.
262
-------
Problems in Lake Restoration
The first study (Reiser, 1976) was designed to test the
duration of effectiveness of a 5 ppm treatment of Aquathol
K on Najas flexilis, Ceratophyllum demersum, and
Potamogeton crispus in East Twin Lake, Ohio (see Cooke
et al. 1978 for detailed physical, chemical, and biological
description). The chemical, the dose, and the application pro-
cedure were chosen and carried out by an Ohio-licensed
pesticide applicator. Based upon biomass in the 2,000 m2
treatment and control plots, the treatment was effective for
about 25 to 30 days, followed by a rebound in biomass caus-
ed by the invasion of a resistant plant, Chara (Fig. 3). An
additional treatment would have been needed in mid-July,
and by season's end the target species Najas flexilis had
bloomed again.
300,
5 250.
5
§200-
150-
< 100-
9
m
50-
ENDOTHAL
TREATMENT
ENDOTHAL-TREATED
CONTROL
10 20 30J 10 20 | 10 20 3C| 10 20 3(i| 10 20
MAY JUN JUL AUQ SEP
10 20 3c]
OCT
Figure 3.—Mean (n = 15) biomass (gms. dry wt./m2) of endothal-
treated and control plots in East Twin Lake. The only significantly
different (p <0.05) pair of control and treatment points occurred on
20 June 1975. Dose was 5ppm aquathol K. (Reiser, 1976).
co
400
V)
1200
"5.
1100
ci
0
treated
tr * j
Total P
^400
§>300
V)
o200
100
ci
0
A s * o
Soluble React! veP-
control
O
Figure 4.—Changes in soluble reactive and total phosphorus before
and after Cutrine-Plus application (*). Dose was an in situ concen-
tration of 0.2 mg Cu/l. (Myers, 1979).
An identical result with these species in the lake was
reported by Conyers (1983). At the advice of professional
herbicide sales, manufacture, and application personnel, she
used Diquat and Cutrine-Plus. Even with the algicide, Chara
biomass equalled the biomass in her control plots within a
few weeks after treatment (Fig. 1).
Thus against these species our two studies showed that
Aquathol K (dipotassium endothal), Diquat, and Cutrine-Plus
had only short-term effectiveness and lake users would have
had to recall the applicator if abatement past mid-July were
required. These chemicals were no more effective in East
Twin Lake, even with the advice of knowledgeable, experi-
enced applicators, than a harvester would have been, had
the additional disadvantage of a 10-day lake closure after
application, and were more expensive than harvesting (Con-
yers and Cooke, 1982).
The cost range for harvesting (Table 1) is $l48/ha to
$490/ha ($60 to $198/acre), while for herbicides (see sec-
tion on Harvesting for references) the cost range is $250 to
$1,077/ha ($100 to $436/acre). Cost-conscious lake owners
and users therefore must consider the apparent fact that her-
bicides are at least as costly as harvesting, with none of the
advantages of biomass removal.
The planktonic algae of many lakes are limited in their pro-
ductivity by the concentration of phosphorus (P) in the water
column. Does the application of an herbicide or an algicide,
in which killed biomass is left in the lake to decompose, bring
about a release of P into the water column? We have in-
vestigated this question, using both field and laboratory ex-
perimental systems.
The algicide experiment (Myers, 1979) involved two side-
by-side ponds of nearly identical morphometry. Both had
nuisance mats of filamentous green algae (Spirogyra,
Hydrodictyon, Oedogonium) that rendered the ponds nearly
useless for recreation. Liquid Cutrine-Plus was added to one
pond, following the manufacturer's directions regarding dose,
proper water temperature, and pond conditions (sunny, calm)
at a concentration of 0.2 mg C^l, on three dates over the
summer. On each date, half the pond was treated, then a
period of several days was observed before treating the other
half. Figure 4 illustrates the pulses in total and soluble reac-
tive phosphorus which occurred after the applications. Pulses
in concentration were not seen in the control ponds. While
it is unknown whether the ponds were actually phosphorus
limited in their productivity, planktonic (Rg. 5) and community
gross productivity (Fig. 6) were significantly higher in the
treated ponds during plankton blooms which followed treat-
ment. As well, the treated pond frequently had low dawn
dissolved oxygen concentrations (range = 0.18 to 11.17 mg
O2/l). The control pond's dawn oxygen was low on only one
occasion. The low oxygen readings prevented further algicide
1.5
control —
treated —
n=3 ±1 s.e.
M
O
Figure 5.—Mean plankton gross productivity ± one standard error
in treated and control ponds before and after Cutrine-Plus applica-
tions (*). (Myers, 1979).
263
-------
Lake Restoration, Protection and Management
3.0
CM 1.5
O
01
control n-3
treated *1s.e.
M
*J*
O
Figure 6.—Mean gross community productivity ± one standard er-
ror in treated and control ponds before and after Cutrine-Plus ap-
plication (*). (Myers, 1979).
applications to control the plankton bloom. The aigicide was
effective in killing the nuisance algae, but there was a sharp
rebound in pond chlorophyll about 5 days after each
treatment.
Thus Cutrine-Plus can be considered to have been a
satisfactory but short-term aigicide that apparently en-
hanced planktonic productivity of the pond, perhaps through
the release of nutrients. Plankton blooms in this case were
more desirable for the lake user than the mats of filamen-
tous algae. The pond could not have been used for fishing
since the low oxygen following treatment would have
prevented fish survival.
The effects of an endothall (Hydrothal 47) application on
phosphorus concentration, productivity, and species eom-
20 r
5 10
DAYS AFTER TREATMENT
Figure 7.—Dissolved oxygen and phosphorus concentrations in con-
trol and experimental aquaria microcosms following application of
Hydrothol 47 (Granular) at 2.0 ppm active ingredient. (James, 1982).
See text for differences between Fig. 7 and Fig. 8.
position were examined by using laboratory microcosms
(James, 1982), Aquaria containing sediment and water from
West Twin Lake were Innoculated with Potamogeton crispus
turions and a population allowed to develop in each, under
a 16 hr/8 hr light-dark cycle. The experimental systems were
treated with 2 ppm Hydrothal 47 (N,N-dimethyl-alkylamine
derivative of endothall), which killed the target plant. In the
first experiment, with a large standing crop of plants, there
was a prompt release of soluble reactive P and lesser in-
creases in particulate and soluble unreactive P (Fig. 7),
Dissolved oxygen declined slightly, and there was a bloom
of epipelic algae. In a second experiment in which the stand-
ing crop of macrophyte phosphorus was much less than in
the first experiment (95.6 mg P/m2 vs. 287 mg P/m2), no
phosphorus pulse was observed (Fig. 8). However, the
treated microcosms became heterotrophic and gross and
net primary productivity recovered as an algal bloom again
appeared.
As with the field studies of Reiser (1976), Myers (1979),
and Conyers (1983), James' experiments showed that the
herbicide stress stimulated a rebound in community meta-
bolism and biomass through the appearance of resistant
species. Opportunistic species invaded and themselves
became abundant. In the field, these invaders themselves
become nuisances to lake users, necessitating further her-
bicide application. The stress of the chemicals appears to
subsidize the growth of new nuisance plants.
Conclusions
I have intentionally introduced the subject of herbicide and
aigicide use in lake management to prompt debate. We can-
not continue to ignore the fact that many, if not most, pond
and lake users rely solely upon these chemicals for abate-
DAYS AFTER TREATMENT
Figure 8.~-Dissolved oxygen and phosphorus concentrations in con-
trol and experimental microcosms following application of Hydrothol
27 (granular) at 2.0 ppm active ingredient. Points are means, bars
the standard error, and asterisks indicate significant differences (p<
0.05, student's t-test). (James, 1982).
264
-------
Problems in Lake Restoration
ment of nuisance plant growth. In the scientific debate which
I hope will follow, there must be some ground rules. We can-
not be lead by emotional or wishful thinking to urge the use
of or abandonment of any procedure to manage lakes, nor
can we as professionals accept a technique that is expe-
dient but creates additional lake problems. We need data
from well-designed, repeatable experiments that address
questions regarding the effectivenesss of herbicides and
algicides in reducing nutrient concentration and in controll-
ing algal biomass when compared to other techniques.
Which procedures are most cost-effective, efficient, and have
the least negative effects on the system or human health?
Are there ways in which herbicidal chemicals can be used
in conjunction with such techniques as harvesting or the in-
troduction of herbivorous fish in order to increase effec-
tiveness and longevity, and at the same time reduce costs?
And finally, we must remember that lake management in-
volves the manipulation of a complex community. Thus it
is to this level of biological organization that we must ad-
dress experiments with herbicides or other procedures and
reduce our emphasis on responses of species populations.
At present few data support the use of herbicidal chemicals
in lakes, despite some work that indicates that they may have
little or no toxicity to other lake organisms and that under
aerobic conditions they decompose readily without leaving
residues in sediment or fish tissue (e.g. Simsiman and
Chesters, 1975; Serns, 1977). Most experiments have not
dealt with the biological level of organization to which these
chemicals are applied.
The data presented here support the conclusion that en-
dothall, Diquat, and Cutrine-Plus were only briefly effective
in doing what they are supposed to do, cost more than alter-
native procedures, and stimulated nutrient release and the
invasion of more plants. Problems with macrophytes in lakes
and reservoirs can usually be handled with less cost and
equal or greater effectiveness through sediment covers,
drawdown, harvesting, or sediment removal. Problems with
algae can be satisfactorily controlled in most situations
through dilution/flushing, phosphorus inactivation, or sedi-
ment removal. In every instance, the source of the problem
should be controlled through nutrient and silt diversion.
Small ponds present a special problem since they are
often owned and used by one or a few people, thus making
some techniques such as harvesting or sediment removal
too expensive to bear. Until better means of treating pond
problems are developed, relief may have to be obtained
through using herbicides and algicides. One possibility to
avoid the death-decomposition-oxygen release-and-nutrient
pulse following algicide application might be to use the
technique of May (1974), who placed blocks of ferric alum
in ponds and obtained control of blue-green algae.
ACKNOWLEDGEMENTS: I thank Applied Biochemists, Mequon,
Wis. for donating some of the chemicals for our experiments, and
I thank Dredgemasters Int., and the Aquamarine Corp., Waukesha,
Wis., who supported some of my travel expenses. The harvesting
study on LaDue Reservoir was supported by the city of Akron, Ohio.
REFERENCES
Anderson, L.W.J. 1981. Control of aquatic weeds with hexazinone.
J. Aquat. Plant Manage. 19:9-14.
Andersson, G., H. Berggen, G. Cronberg, and C. Gelin. 1978. Effects
of planktivorous and benthivorous fish on organisms and water
chemistry in eutrophic lakes. Hydrobiology 59:9-16
Barko, J.W., and R.M. Smart. 1980. Mobilization of sediment
phosphorus by submerged macrophytes. Freshw. Biol. 10:229-238.
Bjork, S. 1974. European lake rehabilitation activities. Inst. Limnol.
Rep., Univ. of Lund, Sweden.
Burton, T.M., D.L. King, and J.L. Ervm. 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.
Carpenter, S.R. 1981. Submersed vegetation: An internal factor in
lake ecosystem succession. Am. Nat. 118:372-383
Carpenter, S.R., and J.K. Greenlee. 1981. Lake deoxygenation after
herbicide use: A simulation model analysis. Aquat. Bot.
11:173-186.
Center, T.D. 1981. Release and establishment of Sameodes a/bigut-
ta/is for biological control of water hyacinth. Tech. Rep. A-81-3,
Environ. Lab. U.S. Army Corps Eng., Vicksburg, Miss.
Conyers, D.L. 1983. A comparative study of harvesting and
herbicides for the control of aquatic macrophytes. M.S. Thesis.
Kent State Univ., Kent, Ohio.
Conyers, D.L., and G.D. Cooke. 1982. Comparing methods for
macrophyte control. Chemical or mechanical? Lake Line 2 (2):8-10.
Cooke, G.D. 1980a. Lake level drawdown as a macrophyte control
technique. Water Res. Bull. 16:317-322.
1980b. Covering bottom sediments as a lake restoration
technique. Water Res. Bull. 16:921-926.
Cooke, G.D., and R.H. Kennedy. 1981. Precipitation and inactivation
of phosphorus as a lake restoration technique. EPA 600/3-81-012.
U.S. Environ. Prot. Agency, Washington, D.C.
Cooke, G.D., R.T. Heath, R.H. Kennedy, and M.R. McComas. 1978.
Effects of diversion and alum application on two eutrophic lakes.
EPA 600/3-78-033. U.S. Environ. Prot. Agency, Washington, D.C.
1982. Change in lake trophic state and internal phosphorus
release after aluminum sulfate application. Water Res. Bull.
18:699-706.
Cooke, G.D., E.B. Welch, S.A. Peterson, and P.R. Newroth. 1983.
Lake Restoration. CRC Press, Boca Raton, Fla.
Dominie, D.R. II. 1980. Hypolimnetic aluminum treatment of softwater
Annabessocook Lake. Pages 417-423 in Restoration of Lakes and
Inland Waters. EPA 440/5-81-010. U.S. Environ. Prot. Agency,
Washington, D.C.
Dunst, R., and S.A. Nichols. 1979. Macrophyte control in a lake man-
agement program. Pages 411-418 in J.E. Breck, R.T. Prentki, and
O.L. Loucks, eds. Aquatic Plants, Lake Management and
Ecosystem Consequences of Lake Harvesting. Inst. Environ. Stud.,
Univ. Wisconsin, Madison.
Freeman, T.E. 1977. Biological control of aquatic weeds with plant
pathogens. Aquat. Bot. 3:175-184.
Freeman, T.E., et al. 1981. Biological control of aquatic plants with
pathogenic fungi. Tech. Rep. A-81-1, U.S Army Corps Eng.,
Vicksburg, Miss.
Gachter, R. 1976. Die Tiefenuasserableitung, ein Weg zur Sanierung
von Seen. Schweiz. Z. Hydrol. 38:1-28.
Haller, W.T., J.V. Shireman, and D.F. DuRant. 1980. Fish harvest re-
sulting from mechanical control of Hydrilla. Trans. Am. Fish. Sec.
109:517-520.
Hill, B.N. 1979. Uptake and release of nutrients by aquatic macro-
phytes. Aquat. Bot. 7:87-93.
James, W.F. 1982. Effects of endothal treatment on phosphorus con-
centration and metabolism of aquatic communities. M.S. Thesis.
Kent State Univ., Kent, Ohio.
Keen, W.H., and J. Gagliardi. 1981. Effect of brown bullheads on
release of phosphorus in sediment and water systems. Progr.
Fish-Cult. 43:183-185
Kennedy, R.H., and G.D. Cooke. 1982. Control of lake phosphorus
with aluminum sulfate. Dose determination and application tech-
niques. Water Res. Bull. 18:389-395.
King, D.L., and T.M. Burton. 1980. The efficacy of weed harvesting
for lake restoration. Pages 158-161 in Restoration of Lakes and
Inland Waters. EPA 440/5-81-010. U.S. Environ. Prot. Agency,
Washington, D.C.
265
-------
Lake Restoration, Protection and Management
LaMarra, V.J. Jr. 1975. Digestive activities of carp as a major con-
tributor to the nutrient loading of lakes. Verh. Int. Ver. Limnol.
19:2461-2468.
Landers, D.H. 1982. Effects of naturally senescing aquatic
macrophytes on nutrient chemistry and chlorophyll a of surround-
ing waters. Limnol. Oceanogr. 27:428-439.
Lantz, K.E. 1974. Natural and controlled water level fluctuation in a
backwater lake and three Louisiana impoundments. La. Wildl.
Fish. Comm., Baton Rouge.
Lantz, K.E., J.T. Davis, J.S. Hughes, and H.E. Schafer. 1964. Water
level fluctuation — its effects on vegetation control and fish popula-
tion management. Proc. 18th Ann. Conf. SE. Assoc. Game Fish
Comm. 18:483-494.
Livermore, D.F., and R.G. Koegel. 1979. Mechanical harvesting of
aquatic plants: An assessment of the state of the art. Pages
307-327 in J.E. Breck, R.T. Prentki, and O.L. Loucks, eds. Aquatic
Plants, Lake Management, and Ecosystem Consequences of
Lake Harvesting. Inst. Environ. Stud. Univ. Wisconsin, Madison.
Lorenzen, M., and A. Fast. 1977. A guide to aeration/circulation tech-
niques for lake management. EPA-600/3-77-004. U.S. Environ.
Prat. Agency, Washington, D.C.
May, V. 1974. Suppression of blue-green algal blooms in Braidwood
Lagoons with alum. J. Aust. Inst. Agric. Sci. 40:54-57.
Mayer, J.R. 1978. Aquatic weed management by benthic semi-
barriers J. Aquat. Plant Manage. 16:31-33.
McKnignt, D. 1981. Chemical and biological processes controlling the
response of a freshwater ecosystem to copper stress: A field study
of the C/j SO4 treatment of Mill Pond Reservoir, Burlington,
Mass. Limnol. Oceanogr. 26:518-531.
Mitzner, L 1978. Evaluation of biological control of nuisance aquatic
vegetation by grass carp. Trans. Am. Fish. Soc. 107:135-145.
Morris, K., and R. Jarman. 1981. Evaluation of water quality during
herbicide applications to Kerr Lake, Okla. J. Aquat. Plant Manage.
19:15-18.
Myers, D.N. 1979. Changes in the metabolism and interactions in a
pond community after treatments with the algicide Cutrine Plus.
M.S. Thesis, Kent State Univ., Kent, Ohio.
Newroth, P.R. 1982. Personal communication. Minist. Environ. Parlia-
ment Buildings, Vancouver, British Columbia, Canada.
Nichols, S., and G. Cottam. 1972. Harvesting as a control for aquatic
plants. Water Res. Bull. 8:1205-1210.
Oglesby, T.T. 1969. Effects of controlled nutrient dilution on the eutro-
phication of a lake. Pages 483-493 in Eutrophication: Causes,
Consequences, and Correctives. Natl. Acad. Sci., Washington,
D.C.
Olszewski, P. 1961. Versuch einer Ableitung des hypolimnischen
Wassers ans einem See. Ergebnisse des ersten Versuchsjahres.
Verh. Int. Ver. Limnol. 18:1792-1797.
Pastorok, R.A., T.C. Ginn, and M.W. Lorenzen. 1981. Evaluation of
aeration/circulation as a lake restoration technique.
EPA-600/3-81-014. U.S. Environ. Prat. Agency, Washington, D.C.
Pastorok, R.A., M.W. Lorenzen, and T.C. Ginn. 1982. Environmental
aspects of artificial aeration and oxygenation of reservoirs: A
review of theory, techniques, and experiences. Tech. Rep. E-82-3,
U.S. Army Corps Eng., Vicksburg, Miss.
Pechlaner, R. 1978. Erfahrungen mil Restaurierungsmassnahmen
an eutrophierten Badeseen Tirols. Oesterreichische Wasserwert-
schaft 30:112-119.
Perkins, M.A., H.L. Boston, and E.F. Curren. 1980. The use of fiber
glass screens for control of Eurasian watermilfoil. J. Aquat. Plant
Manage. 18:13-19.
Peterson, S.A. 1982. Lake restoration by sediment removal. Water
Res. Bull. 18:423-435.
Plotkin, S. 1979. Changes in selected sediment characteristics due
to drawdown of a shallow eutrophic lake. M. S. Thesis, Univ.
Washington, Seattle.
Prentki, R.T. 1979. Depletion of phosphorus from sediment colo-
nized by Myriophyllum spicatum L. Pages 161-176 in J.E. Breck,
R.T. Prentki, and O.L. Loucks, eds. Aquatic Plants, Lake Manage-
ment, and Ecosystem Consequences of Lake Harvesting. Inst.
Environ. Stud., Univ. Wisconsin, Madison.
Reiser, M.L.E. 1976. Some responses of submerged macrophytes of
a glacial eutrophic lake to the herbicide Endothal. M. S. Thesis,
Kent State Univ., Kent, Ohio.
Rich, P.H. and R.G. Wetzel. 1978. Detritus in the lake ecosystem.
Am. Nat. 112:57-71.
Ripl, W. 1976. Biochemical oxidation of polluted lake sediment with
nitrate — a new restoration method. Ambio 5:132-135.
Ripl, W., and G. Lindmark. 1978. Ecosystem control by nitrogen
metabolism in the sediment. Vatten 34:135-143.
Sassic, N.M. 1982. Harvesting: the future of plant control? Aquatics
4:14-16.
Sems, S.L. 1977. Effects of dipotassium Endothall on rooted aquatics
and adult and first generation bluegills. Water Res. Bull. 13:71-80.
Shapiro, J. 1978. The need for more biology in lake restoration.
Pages 161-167 in Lake Restoration. EPA-440/5-79-001. U.S. En-
viron. Prat. Agency, Washington, D.C.
Shapiro, J., V. LaMarra, and M. Lynch. 1975. Biomanipulation: An
ecosystem approach to lake restoration. Pages 85-96 in P.L.
Brezonik and J.L. Fox, eds. Water Quality Management through
Biological Control. Univ. Florida, Gainesville.
Shearer, R. 1980. Public health effects of the aquatic use of her-
bicides: 2,4-D, Dichlobenil, Endothall, and Diquat. In Literature
Reviews of Four Selected Herbicides: 2,4-D, Dichlobenil, Diquat,
and Endothall. Munic. Metro. Seattle, Wash.
Shireman, J.V., and M.J. Maceina. 1981. The utilization of grass carp,
Ctenopharyngodon idella Val., for hydrilla control in Lake Baldwin,
Fla. J. Fish. Biol. 19:629-636.
Simsiman, G.V., and G. Chesters. 1975. Persistence of endothall in
the aquatic environment. Water Air Soil Pollut. 4:399-413.
Simsiman, G.V., G. Chesters, and T.C. Daniel. 1972. Chemical
control of aquatic weeds and its effect on the nutrient and redox
status of water and sediment. Proc. 15th Conf. Great Lakes Res.
1972:166-180.
Smith, G.N. 1979. Recent case studies of macrophyte harvesting
costs-options by which to lower costs. Pages 345-356 in J.E.
Breck, R.T. Prentki, and O.L. Loucks, eds. Aquatic Plants, Lake
Management, and Ecosystem Consequences of Lake Harvesting.
Inst. Environ. Stud., Univ. Wisconsin, Madison.
Uttormark, P.D., and M.L. Hutchins. 1980. Input-output models as
decision aids for lake restoration. Water Res. Bull. 16:494-500.
Valencia, R. 1981. Mutagenesis screening of pesticides using Droso
phila. EPA-600/51-81-017. U.S. Environ. Prot. Agency,
Washington, D.C.
Welch, E.B. 1981. The dilution/flushing technique in lake restoration.
Water Res. Bull. 17:558-564.
Welch, E.B. and C.R. Patmont. 1980. Lake restoration by dilution:
Moses Lake, Wash. Water Res. 14:1317-1325.
Welch, E.B., M.A. Perkins, D. Lynch, and P. Hufschmidt. 1979. Inter-
nal phosphorus related to rooted macrophytes in a shallow lake.
Pages 81-99 in J.E. Breck, R.T. Prentki, and O.L. Loucks, eds.
Aquatic Plants, Lake Management, and Ecosystem Conse-
quences of Lake Harvesting. Inst. Environ. Stud., Univ. Wiscon-
sin, Madison.
Wile, I., G. Hitchin, and G. Beggs. 1979. Impact of mechanical
harvesting on Chemung Lake. Pages 145-159 in J.E. Breck, R.T.
Prentki, and O.L. Loucks, eds. Aquatic Plants, Lake Management,
and Ecosystem Consequences of Lake Harvesting. Inst. Environ..
Stud., Univ. Wisconsin, Madison.
Wile, I., et al. 1977. An assessment of the practical and environmental1
implications of mechanical harvesting of aquatic vegetation in
Southern Chemung Lake. Ontario Ministry Environ. Rep.
266
-------
LAKE RESTORATION CRITERIA: THE LIMNOLOGIST'S
VIEW VERSUS PUBLIC PERCEPTION
LOWELL L. KLESSIG
University of Wisconsin-Extension
College of Natural Resources
University of Wisconsin
Stevens Point, Wisconsin
NICOLAAS W. BOUWES, SR.
USDA Economic Research Service
Department of Agricultural Economics
University of Wisconsin
Madison, Wisconsin
ABSTRACT
Lake management efforts can be planned more efficiently if limnologists predict the outcome of the various
management options. However, such predictions are only useful if public users concur with the evalua-
tions of the experts. That link between objective and subjective evaluation is tenuous at best. Therefore,
citizens should be directly involved in stating their preferences and, when they cannot be, social scientists
should be used to complement limnologists' judgments.
Any evaluation effort is premised on the ability to measure
and compare one or more dependent variables at a begin-
ning point in time (T^ and an ending point (Tj). These ef-
forts run the gamut from a spelling test in third grade to
measuring the degree of achievement of national racial in-
tegration policies.
Evaluations of lake protection or rehabilitation projects fit
the same general context. Protection projects attempt to
maintain the status quo. The hope is that conditions at T2
equal conditions at TV In a longitudinal study of this type,
if it is possible to document that conditions have stayed the
same—the project was successful. But it is not possible to
certify that the investment of project costs has been justified
unless it is possible to predict how bad conditions would have
become had the project not been undertaken. These situa-
tions require methods that can evaluate projects in a "with
or without" framework.
Evaluation of rehabilitation projects also requires a "with
or without" framework. If no action is taken, conditions at
T2 are not likely to equal conditions at T,. Usually, further
deterioration in water quality would be expected. Thus, it is
not sufficient to simply measure the difference between T2
and T,. It is also necessary to predict what would have hap-
pened if nothing had been done.
Whatever the evaluation framework, someone has to apply
a standard of quality regarding the lake resource. The pur-
pose of this paper is to compare the more objective evalua-
tions of lake managers (experts) with the more subjective
evaluations of the public (users).
WATER QUALITY ACCORDING TO WHOM
Most professional groups assume that only they are qualified
to judge quality in their realm of training and experience.
By definition, an expert has special analytical abilities and
understands the accepted standards in the profession. If one
is measuring the strength of a bridge or the level of
phosphorus, there can be no argument with the prerogatives
of the expert.
However, experts have a natural tendency to inflate the
social significance of their particular expertise. They tend to
push themselves higher up in the means and ends pyramid.
For example, a lake manager who spends his career trying
to control phosphorus input to lakes begins to view
phosphorus control as an end in itself—the ultimate goal.
As shown in Figure 1, it is merely a means to reduce algae
growth, which is but one component of water quality, which
is but one means to a rewarding experience at the lake,
Figure 1.—Means and ends pyramid.
267
-------
Lake Restoration, Protection and Management
which is but part of the recreational opportunity for citizens,
which is—in turn—but part of their overall social wellbeing.
If we examine the means-ends hierarchy, we find the ex-
pert's judgment of diminishing significance and the user's
judgment of increasing significance as we go up the pyramid.
Only the expert can measure phosphorus. The expert can
quantify algae through chlorophyll measurement and species
determination, but the users can also observe or smell
nuisance algae.
By the time we reach overall water quality, the expert's
concept of this difficult-to-define condition is no more valid,
perhaps less, than that of the user. The expert still has
chlorophyll, Secchi disc, dissolved oxygen, and other quan-
titative measurements. However, the user makes important
aesthetic judgments of a more subjective nature that are
equally important. Floating debris, fishing success, shoreline
vegetation, and lakeshore development are among the fac-
tors that contribute to his or her perception of water quality.
The significance of objective measures diminishes as
water quality is plugged into overall satisfaction with lake
recreation. At this point, many other variables intrude and
become part of the experience and hence the valuation of
that experience. Density of users, cleanliness of the bath
house, companionship, weather, and a host of other factors
all combine in the recreation experience.
At the level of recreational opportunity, professional
managers are involved only with adequate public access and
perhaps the competition of other lakes in the region. The
full range of other recreational opportunities provides the con-
text where the users must choose for themselves.
Finally, when we reach social well-being — the pinnacle
of the pyramid — the user is the expert. Only he knows what
makes him happy. And, if we are to assist in maximizing
that happiness, we need to periodically ask him.
THE OBJECTIVE METHODS
Many methods have been used to provide a reproducible
objective measure of water quality. Water quality is typical-
ly defined in terms of its trophic state, with oligotrophic lakes
being high and eutrophic (enriched) lakes being low.
Vollenweider's (1975) system, based on phosphorus loading
and algae response, is most widely referenced. Dillon and
Rigler's (1975) modification has also received widespread
use.
Other efforts have included the use of LANDSAT imagers
from satellites (Scarpace et al. 1978; Scherz et al, 1975) or
the use of a variety of chemical parameters (Shannon and
Brezonik, 1972).
When classifying Wisconsin lakes, Uttormark and Wall
(1975) used a system of penalty points. These points were
given for four parameters as shown in Table 1.
A lake with Lake Condition Index (LCI) of 23 would be very
eutrophic and a lake with 0 penalty points would be an
Table 1. — Penalty points for Lake Condition Index.
Parameters
Points
Hypolimnetic dissolved oxygen 0-6
Transparency 0-4
Fishkills 0-4
Use impairment (extent of macrophyte 0-9
or algal growth)
Total 0-23
oligotrophic extreme. All of Wisconsin's 1,100 lakes over 100
acres were classified.
Uttormark's system is not purely objective, because the
"use impairment" parameter is based on expert judgment
of nuisance vegetation rather than measurement. It is,
however, pure in regard to using information solely from
experts.
THE SUBJECTIVE METHODS
Methods that rely on users as sources of information tend
to be ideosyncratic to individual projects or researchers. Un-
fortunately, there is no public perception analog to
Vollenweider's phosphorus loading equation.
This deficiency results from two factors: (1) Measuring
public perception is not as standardized as measuring
phosphorus levels; and (2) most limnologists are not inclin-
ed to view user public perception as relevant. However, if
a good standard method of obtaining user evaluations were
available, many lake managers might use it.
Some studies, such as Harris' (1975), were designed to
answer specific policy questions. Harris developed a family
of water quality acceptance curves for turbidity, color, and
odor values of water being processed by water treatment
facilities. He was advising municipalities on what kind of
water would be acceptable. He did not use taste, even
though he was measuring drinking water quality.
Harris' parameters, plus taste and more, can be applied
to lake surface water. Kooyoomjian and Clesceri (1974) us-
ed a whole raft of variables as shown in Table 2.
Table 2. — Objections to lake water quality.
too cold
too warm
too choppy and rough
not very clear; muddy
build-up of shoreline growths
strange odors
strange colors
strange taste
growth of algae, plants, or scum
film of gasoline or oil
dead fish
irritating to eyes and skin
floating objects
muddy bottom
steep slope
too shallow slope
too many shallow spots
too many boats
too many people
not enough people
too many water skiers
too many fishermen
While using so many variables inhibits interpretations and
policy recommendations, the investigators did find interesting
differences between the three populations they studied:
fishermen, other recreators, and shoreland owners. The
fishermen were most sensitive to surface effects and
crowding. The other recreationists were least critical of water
quality overall, but most sensitive to temperature, clarity, and
bottom conditions. The cottage and homeowners were most
concerned with shorelines, odors, colors, and taste.
Nicolson and Mace (1975) surveyed campers in three Min-
nesota State parks and found the degradation of water quality
was perceived strictly in visual terms.
ATTEMPTS AT LINKAGE
Given the different parameters that experts and users employ
when evaluating water quality, it is not surprising that they
sometimes reach different conclusions. If we measure
benefits of lake restoration and protection in socioeconomic
terms, we should not be surprised if user perceptions are
better indicators of benefit than expert judgment. In other
words, changes in recreational use or property values are
based on perception of the water quality by users, not on
scientific measurements of it.
268
-------
Problems in Lake Restoration
In his early work for the National Commission on Water
Quality, Dombusch (1975) recognized that perception is reali-
ty in such circumstances. However, he wanted to predict
benefits in order to assist ex ante management decisions.
To achieve that objective, it would be necessary to get a
quick evaluation of present water quality and a prediction
of perceived water quality at some future time under various
management options.
Because of time limitations and the need to ascertain
future perceptions, public surveys are not likely to be feasi-
ble. Expert judgment would be feasible. Therefore, he set
out to link the two: to predict public benefit by predicting
public perception of water quality from expert judgment.
Based on the responses of residents from 17 communities,
a perceived water quality index was defined to include in-
dustrial wastes, debris, clearness, algae, odor, wildlife sup-
port capacity, and recreational opportunities. Water quality
experts were asked the same questions as residents regar-
ding the relative significance of these components. Regres-
sions were run to determine the correspondence between
experts and residents. The results were markedly different,
but when a public access variable was added, the following
relationship was obtained:
PWQIRes = -24.778 + 0.463 (PWQI^) + 15.501 (PA)
(3546) (5.266)
R~2 = .681
Where PWQIRes = Perceived Water Quality Index by
Residents,
PWQIgxp = Perceived Water Quality Index by
Experts, and
R = adjusted R2, and
PA = Public Access: 3 for good access, 2 for
limited access,
1 for restricted access.
The t-values for the estimated coefficients were signifi-
cant at 1 percent level.
Bouwes and Schneider (1979) surveyed recreators to link
public perception to expert evaluation of water quality at
seven Wisconsin lakes. They used Uttormark's Lake Con-
dition Index (LCI) for their expert rating. They asked
recreators to rate the lake water quality on the same 0-23
point scale.
Their results yielded the following working equation:
In Rat = 1.948 + .0364 LCI
(3.37)
.643
Where Rat = average rating of all recreators for each
lake, and
R~2 = adjusted R2.
The t-value on the LCI coefficient was significant at 2
percent level.
We attempted to duplicate these pretest results using a
statewide sample of Wisconsin adults in households to
estimate the impact of water quality improvements on recrea-
tion benefits. We also used the Dornbusch model and Ut-
tormark's LCI to predict perceived water quality at two case
study lakes that had recently undergone major management
projects (Bouwes and Klessig, 1982). The following discus-
sion will focus on the statewide survey.
The sample for the statewide survey was drawn from all
housing units with telephones. Because certain numbers
have a low probability of yielding an interview, a multi-stage,
disproportionate, stratified model was used (Palit, 1979). The
response rate was 72 percent, providing 1,107 interviews.
R2
Some of the nonresponses were ineligible; thus the true
response rate was higher.
The interviewers determined which of the respondents
were lake recreators, then asked them to rate the water quali-
ty of lakes they had visited during the previous year. Roughly
two thirds of the respondents recreated at Wisconsin lakes—
the survey yielded 723 observations of 243 lakes. They
evaluated water quality on the same 0-23 scale devised by
Uttormark, thereby providing a subjective rating of the rele-
vant Wisconsin lakes.
Efforts to duplicate the pretest results were unsuccessful.
The lack of statistical significance of the estimated
parameters suggested that LCI was not very effective in ex-
plaining the public user's perception of water quality.
These disappointing results may be attributable to the dif-
ferent methods used in gathering information. In the previous
research, responses were obtained by on-site interviews,
whereas, the statewide sample was gathered through a
telephone survey. In addition to the telephone survey per
se, the 23 point scale might be inappropriate to use over
the phone, but comprehensible if presented on a card in a
personal interview at the lake as was done for the earlier
pretest.
The survey was conducted in October and November of
1978 and questioned respondents about their recreational
activities between Labor Day 1977 and Labor Day 1978.
Therefore, our telephone survey not only was separated
spatially, but also temporally from the lake experience.
The recall factor bias is aggravated by the nature of lake
recreation. Lakeside or surface water activities tend to be
social events. Over time, the social experience (sick baby,
unpleasant in-laws, or heartwarming reunion with friends) in-
termingles with the natural resource experience. The quali-
ty of the food, the weather, and the trip to the lake all tend
to be added into a single evaluation of the overall experience.
Thus, it may not be surprising that when asked about the
water quality, the subjective responses bore little relation-
ship to the objective LCI.
We are most apt to blame our inability to link the expert's
judgment with the user's perception on the methodology that
depended on recall distant in time and space. However, it
should be noted that the pretest and Dornbusch's earlier
work were based on average perceptions by residents and
users. If the distribution of perceptions is bimodal rather than
a normal bell-shaped curve, the high correlations found
earlier are misleading. The variance in subjective percep-
tions among individual recreators or residents should be
examined.
CONCLUSIONS AND RECOMMENDATIONS
The relationship between experts' judgments of water quality
and users' perceptions of public water resources is not well
understood. The criteria used by limnologists are strikingly
different from the criteria used by the public.
Limnologists use criteria that can be measured and com-
municated with other professionals. Recreators seldom think
about water quality per se in biological or chemical terms.
They respond to what they see in the water and on the
shoreline. They respond to what they smell. They respond
to congestion and to a host of social factors surrounding the
experience at the lake. Thus, it is really not surprising to find
that water quality means different things to the two groups.
Based on the tenuous nature of the link between lim-
nologists' measurements and public perception, we offer the
following recommendations:
1. Using limnologists' criteria to predict public evaluation
is risky and should be done with caution. The public criteria
should be recognized as distinct from the experts' criteria
and both should be recognized as equally legitimate. We
269
-------
Lake Restoration, Protection and Management
are not recommending a public poll before each manage-
ment decision is made, but we are advocating a philosophy
of accepting subjective input. To make lake managers more
comfortable in predicting lay persons' perceptions of water
quality, we recommend more research on the linkage bet-
ween the two sets of criteria.
2. Whenever feasible, ask the public recreator for first-
hand reactions to water quality in terms that lay people can
relate to. When such direct feedback is impractical, use a
reaction panel (advisory council) to supplement limnological
judgments. If a reaction panel is also cumbersome, consult
an expert in public perceptions—a social scientist with ex-
perience in documenting and predicting human reactions.
Again use this information to complement evaluation efforts
by physical scientists. A number of systems, such as the
environmental evaluation system (EES) of the U.S. Bureau
of Reclamation and the judgmental impact matrix (JIM) of
the U.S. Army Corps of Engineers (McAllister, 1982), use
the "Delphi" method of a panel of experts. Such systems
are useful if the panel of experts is appropriately broad and
the parameters reflect both expert and citizen criteria.
3. Finally, we recommend that serious consideration be
given to the variance among lake users. Peterson and
Neumann (1969) found different preferences based on age
and education. Dearinger et al. (1973) found occupation and
life style to change preferences. Some lakes might be
developed for the handicapped, others for young families,
others for sportsmen, etc. Since 1960, resource managers
have been indoctrinated with the multiple-use concept. Like
many such concepts, it has been applied indiscriminately
and real incompatibilities among users, such as canoeists
and water skiers, have too often been ignored. We might
want to tailor individual lake management efforts to different
"clientele." Application of a "dominant use" philosophy
would not eliminate the need to develop an index (or poten-
tially a set of indices) to measure users' perception of lake
quality on criteria that they find significant.
REFERENCES
Bouwes, N. W. and L. L. Klessig. 1982. Socio-economic impact e-
valuation of lake improvement projects and lake management
guidelines. Working pap. No. 17. Center Resour. Policy Stud.
Progr., Univ. Wisconsin, Madison.
Bouwes, N.W., and R. Schneider. 1979. Procedures in estimating
benefits of water quality change. Am. J. Agric. Econ. 61:535-539.
Dearinger, J.A., et al. 1973. Measuring the intangible values of natural
streams. Part II. Rep. No. 66. Water Resour. Inst. Univ. Kentucky.
Dillon, P.J., and F.H. Rigler. 1975. A simple method for predicting
capacity of a lake for development based on lake trophic status.
J. Fish. Res. Board Can. 32:1519.
Dornbusch, D.M. 1975. The Impact of Water Resource Quality Im-
provements on Residential Property Prices. Natl. Comm. Water
Quality, Washington, D.C.
Harris, D.H. 1975. The assessment of water quality. Human Fac-
tors 17:139-148.
Kooyoomjian, K.J., and N.L. Clesceri. 1974. Perceptions of water
qual-
ity by select respondent groupings in inland water-based recrea-
tional environments. Water Resour. Bull. 10:728-744.
McAllister, D.M. 1982. Evaluation in Environmental Planning: Assess-
ing Environmental, Social, Economic, and Political Trade-offs. MIT
Press, Cambridge, Mass.
Nicolson, J.A., and A.C. Mace. 1975. Water quality perception by
users: can it supplement objective water quality measures. Water
Resour. Bull. 11:1197-1207.
Palit, C. 1979. Sampling report and response results for statewide
water quality survey. Proj. 1130-S. Wis. Surv. Res. Lab., Univ.
Wisconsin Extension.
Peterson, G.L., and E.S. Neumann. 1969. Modelling and predicting
human response to visual recreation environment. J. Leisure Res.
3:219-237.
Scarpace, F.L., K. Holmquist, and L.T. Fisher. 1978. LANDSAT
analysis of lake quality for a statewide lake classification program.
Proc. Am. Soc. Photogram. 44th Annu. Meet. Washington, D.C.
Scherz, J.P., D.R. Crane, and R.H. Rogers. 1975. Classifying and
monitoring water quality by use of satellite imagery. Prepared for
Int. Conf. Environment Sensing and Assessment. Las Vegas, Nev.
Shannon, E.E., and P.L. Brezonik. 1972. Relationships between lake
trophic state and nitrogen and phosphorus loading rates. Environ.
Sci. Technol. 6:719-725.
Uttormark, P.O., and J.P. Wall. 1975. Lake Classification for Water
Quality Management. Water Resour. Center, Univ. Wisconsin,
Madison.
Vollenweider, R.A. 1975. Input-Output Models. Can. Centre Inland
Waters. Burlington, Ontario.
270
-------
IMPACT OF WATERSHED ON URBAN LAKE QUALITY
YOUSEF A. YOUSEF
MARTIN P. WANiELISTA
HARVEY H. HARPER
Department of Civil Engineering and
Environmental Sciences
University of Central Florida
Orlando, Florida
ABSTRACT
Lake Eola watershed located in Central Florida consists of approximately 59 hectares of commercial
and residential areas. Quantitative and qualitative watershed impacts and responses were measured
during 1980. The lake is subjected to variations in hydrologic and nutrient loading inputs which
demonstrate the dynamic nature of stormwater runoff impacts. Correlations appeared to exist bet-
ween monthly averages of runoff or rainfall and nutrient concentrations in the lake. The lake may
be considered a large retention pond and most of the nutrients (82.6 to 95.4 percent) were retained
in the sediments. This study illustrates the deficiencies in steady-state models if applied to dynamic
systems.
INTRODUCTION
Eutrophication problems often develop when nutrient
loadings rise above levels which the lake can adequately
assimilate and still maintain the delicate balance of its formed
food chain. Excess nutrients are usually manifested in the
form of excessive algal blooms, followed by death and decay
causing oxygen depletion and sometimes anaerobic condi-
tions. These conditions are responsible for unpleasant odors,
fish kills and impairment of the potential beneficial uses of
the lake. Nitrogen and phosphorus are generally considered
to be the most important limiting nutrients. However, many
investigators indicate that the majority of lakes and reser-
voirs are phosphorus limited alone.
The trophic status of lakes based on areal phosphorus
loading has been predicted by simplified input-output models
(Dillon and Rigler, 1974; Chapra and Tarapchak, 1976;
Vollenweider, 1968,1969,1975,1976; Kratzer, 1979). Also,
phosphorus loading models have been correlated with mean
chlorophyll a concentrations and strong correlations
developed between Secchl disk transparency and chlorophyll
a concentrations (Carlson, 1977; Brezonik, 1978). The
Organization for Economic Cooperation and Development
(OECD) eutrophication modeling study has shown that the
phosphorus load normalized by waterbody mean depth and
hydraulic retention time is directly correlated to mean sum-
mer algal chlorophyll a concentration, Secchi depth, and the
hypolimnetic oxygen depletion rate.
It should be realized that most trophic state studies have
been performed on deep northern lakes. For the most part,
these lakes are subject to receiving an almost instantaneous
loading of nutrients during the spring turnover, and maximum
algal concentrations occur generally in the summer months.
For this reason, it is appropriate to analyze the data under
steady-state or static conditions. However, shallow lakes
typical of the Florida environment are not subject to only in-
stantaneous loadings. They often approach continuous flow
systems receiving intermittent nutrient loadings as a result
of surface water runoff. Also, they are subjected to seasonally
variant precipitation. Models to predict lake quality response
to variations in hydrologic and nutrient loading inputs are
rare or nonexistent. Construction of a model to predict these
impacts will require extensive quality and quantity data to
precisely analyze the overlapping impacts of separate storm
events.
This paper will briefly discuss nutrient loadings from water-
sheds caused by stormwater runoff and the potential impacts
of these nutrients, particularly as they relate to the Lake Eola
watershed.
DETERMINATION OF MASS LOADINGS
Generally the most significant water quality problem that ex-
ists in urban lakes is that of excessive nutrients, and it is
frequently necessary to know loadings of nitrogen and
phosphorus into the lake. Mass loadings can be determin-
ed by obtaining data on water quantity and quality from all
sources including tributary inflows, seepage, point source
discharges, direct precipitation, and internal loading and
recycling. The acquisition of this data is expensive, and it
is desirable to be able to estimate the loadings from reliable
existing sources, by identifying the separate land use areas
and the corresponding export rates. Determination of land
uses can be made using Mark Hurd photoquads, U.S.
Geological Survey 7.5 minute quadrangles (both 1:24,000),
photos from the Agricultural Stabilization and Conservation
Service, and other surveys by interested organizations in the
state such as the Florida Citrus Survey.
Nutrient export rates for various land uses were com-
piled by Reckhow et ai. (1980). Export coefficients from ur-
ban areas ranged from 1.48 to 38.47 kg/ha-yr for total
nitrogen (TN) and from 0.19 to 6.23 kg/ha-yr for total
phosphorus (TP). The average values were 9.97 and 1.91
kg/ha-yr for TN and TP. The wide range in the published
export coefficients for a specific land use resulted from varia-
tions in precipitation, soil type, vegetation, and other factors.
However, many of the export coefficients were based on
short sampling periods and did not include annual variations.
Some investigators calculated mass loadings as the product
of the annual runoff water yield and the average concentra-
tions of limited storm events. Drainage from urban water-
sheds cannot be quantified without the construction of a com-
plete water budget. It is inaccurate to assume that all the
water coming from a watershed results from precipitation fall-
ing on this particular drainage area without consideration of
ground seepage, springs, and other sources originating from
outside the boundaries of the study area. Most of the mass
loading rates are based on the total drainage area; however,
271
-------
Lake Restoration, Protection and Management
Table 1. — Bulk precipitation for various land uses in
Florida.
Land use
Mass Loading, ng/ha/yr
~!NTP ZN PB
Reference
Coastal 5.8 0.31
Urban 7.6 0.50
Agricultural 8.8 0.66
Highway 11.5 0.68 1.2
Brezonik et al. (1981)
1.5 Miller and Mattraw (1982)
Commercial 6.4 0.35 1.8 3.9
it may be more appropriate if they are based on the
hydraulically effective area in the basin (Miller and Mattraw,
1982).
Bulk precipitation itself may be a major source of nutrients
for many lakes, particularly seepage lakes with long deten-
tion times. Bulk precipitation including wetfall plus dryfall was
measured in Florida by Brezonik et al. (1981) and Miller and
Mattraw (1982) as shown in Table 1. Nutrient rates from bulk
precipitation may exceed those from the drainage basin.
STORMWATER RUNOFF IMPACTS
Characterization of urban stormwater discharges in terms
of concentrations and pollutional loads indicates potential
receiving water impacts. Suggested manuals of simplified
methodologies to assess those impacts have been developed
by many investigators (Driscoll et al. 1979; Binder et al. 1981).
These methodologies are appropriate for use at the plann-
ing level where preliminary assessments are made to define
the problem, establish the relative significance of contributing
sources, assess feasibility of control and determine the need
for additional evaluation. Pollutants carried by stormwater
runoff may be characterized as organic compounds,
suspended solids, bacterial contaminants, nutrients, and
heavy metals. However, most of the water quality problems
in urban lakes are related to excess nutrients.
Most Florida lakes are formed as solution or depression
basins with shallow mean depths generally less than 3.0 to
4.0 meters. They range in size of below 1 hectare to more
than 10,000 hectares. Unfortunately, the potential beneficial
use of many of these lakes has been impaired by accelerated
eutrophication, and Vollenweider's phosphorus loading
criteria for classification of these lakes tends to overestimate
their trophic status. In Florida, most lakes do not undergo
stable thermal stratification and seasonal variations in
temperature are less pronounced than in the temperate zone.
It is, therefore, reasonable to expect that seasonal variations
in algal standing crop and the concentration of major
nutrients are minimal. Baker et al. (1981) revised the 1975
Vollenweider phosphorus loading criteria and the 1975 Dillon
criteria to improve their prediction capabilities for Florida
lakes. They also developed a composite trophic status in-
dex (TXI) based on Secchi disk, chlorophyll a, and the limiting
nutrient (P or N) to produce an index that reflects the many
dimensions of the eutrophication concept.
In summary, contaminants in stormwater runoff will even-
tually enter the adjacent environment in significant loadings
depending on many factors, such as a preceding dry period,
land use of the drainage basin, degree of urbanization,
volume and type of traffic, industry, and air pollution fallout.
The magnitude of these loadings and their bioavailability
should be adequately assessed before predicting any im-
pacts. A definite national need exists to study receiving water
impacts. The current studies by the U.S. Environmental Pro-
tection Agency (Myers et al. 1982), Nationwide Urban Runoff
Program (NURP) may clarify some of the uncertainties and
produce defensible and useful results to be transferred to
various geographic areas of the nation.
Lake Eola watershed impacts will be presented here to
demonstrate the complexity of the process.
LAKE EOLA WATERSHED IMPACTS
The Lake Eola watershed is approximately 59 hectares, com-
posed of 33.7 hectares of commercial and 25.3 hectares of
residential area discharging via storm drains into Lake Eola.
The lake itself occupies 11 hectares surrounded by 4.5 hec-
tares of park land that are not considered part of the water-
shed because of infrequent runoff to the lake. The imper-
vious area of the watershed is 49.3 hectares and the per-
vious area is only 9.7 hectares. Extensive studies by
Wanielista et al. (1982) concluded that (1) stormwater is the
major external source of pollution, and (2) phosphorus and
other stormwater pollutants require removal.
Accurate estimates of pollutant loadings should reflect the
dynamic nature of the system; therefore, they require ex-
tensive and continuous recording of the water budget and
pollutant concentrations. Pollutants are added at different
rates during storm events and little, if any, is known about
their fate in receiving waterbodies.
Stormwater runoff estimates were made using a water
mass balance of a simplified steady-state system for a fixed
time period. Monthly estimates of various parameters are
depicted in Figure 1. Changes in Lake Eola water volume
did not exceed 5 percent throughout 1980. The total mass
evaporated appeared to equal or exceed the mass of direct
precipitation on the lake surface. It was noted that 1980 was
a relatively dry year, and the total precipitation of 0.87 m
was below the annual average of 1.30 m. Also, discharge
through drainage wells was directly related to stormwater
runoff. The hydraulic residence time which is the average
o
X
4O
20
20
1O
>oo
2SO
SOO
250
SO
O
so
SOO
2SO
fMCIPITATION
• •••.•
I I
IVAMMMION
llllli.nl
fKMMWMU RUNOFF
MAINAOt Will
INTO IAKI Nil UIMMH
J
OUT Or IAKI
T---T--1
IAKI VO1UMI
Figure 1.—Hydrologic parameters for Lake Eola drainage basin.
272
-------
Problems in Lake Restoration
5
3
"E
O>
Z
o
o
I
O
30
23
10
s
200
100
100
30
10OO
MO
20
W
100
CHLOROPHYLL '«'
1 illlii i III . . 1 i i •
1
±
NITRITE NITROGEN
Illllllllll
|
|
|,
1 1
NITRATE NITROGEN
• AMMONIA
1. ..1 llll.
_L
|
I TOTAL KXLDAHL
Illll .1.
NITROGEN
1 1
NITROOEN
1
1 1 ORTHO -PHOSPHORUS
, mill . . ii
i
TOTAL PHOSPHORUS
llllllllll.lll 1
|
1
,
J J ASONOJ FM
• 1»«0 •}•— mi
Figure 2.—Concentration of water quality parameters for Lake Eola,
1980/1981.
volume of lake water divided by the annual volume of storm-
water during 1980 was calculated to be 0.57 years.
During the course of this study, responses of Lake Eola
water to pollutional loadings, particularly phosphorus and
nitrogen, have been assessed. Routine water samples from
six different locations were collected, analyzed, and the
results presented in Figure 2. This figure shows total
phosphorus and orthophosphorus in ^g P/l, total Kjeldahl,
ammonia, nitrite and nitrate nitrogen in ^g N/l and chlorophyll
a in ^g/l for various analyses throughout 1980. The or-
thophosphorus concentrations were generally less than 20
^g P/l and the total phosphorus concentrations were less
than 60 ^g P/l.
Harper et al. (1981) concluded from bioassay studies that
Lake Eola water is phosphorus limited. The TN to TP ratio
varied between 4.4 and 17.3 and averaged 9.8, and the ratio
of available nitrogen to available phosphorus (N:P) varied
between 8.0 and 52.1 and averaged 21.5 (Yousef et al.
1981). These data suggest that Lake Eola is phosphorus
limited most of the time and higher N:P ratios exist during
wet weather months when compared with dry weather
months. Nitrogen may have a higher mobility than
phosphorus. Also, scour of bottom sediments by stormwater
currents may resuspend the organic debris and increase the
nitrogen content.
Models developed to assess the eutrophic state of lakes
are based on average annual phosphorus loadings, hydraulic
residence time, and sediment retention (Vollenweider, 1969;
Dillon, 1975; Snodgrass and O'Melia, 1975). These models
were developed for steady-state conditions using the mass
balance approach. However, the dynamic nature of Lake
Eola was evident from monthly changes in nutrient loadings,
hydraulic residence time, and sediment retention. During
1980 sediment retention was (in percent) for the following:
total phosphorus, 86.5; orthophosphorus, 95.4; nitrates,
82.6; and total Kjeldahl nitrogen, 83.1 (see Table 2). These
calculations were based on average concentrations for
similar nutrients in stormwater runoff as presented by
Wanielista et al. (1981).
Calculated runoff volumes on a monthly basis correlated
well with monthly precipitation (r = 0.98). Also, floating
averages for lake water quality parameters over a period from
1 to 4 months were correlated with precipitation depth for
the same period. It was shown that correlation coefficients
improved by increasing the time interval with maximum value
at 3- to 4- month periods for total phosphorus, ortho-
phosphorus, nitrate nitrogen, TKN and chlorophyll a (Walsh,
1981). In all cases, as the time interval increased, the slope
of the best fit line increased and the pollutant base line con-
centration at zero precipitation decreased. It appears that
the lake can be assumed hydrodynamically open on both
ends since a continuous source of runoff water is coming
to the lake with no prolonged dry periods available. Change
in concentrations with change in precipitation depth can be
predicted from the slope of the linear relationships shown
in Figures 3 and 4 for total phosphorus and orthophosphorus
concentrations in Lake Eola water. The improved correla-
tion on long-term averages does not necessarily tell the short-
term impact.
TARGET PHOSPHORUS REDUCTION
The major question is to what degree should the bottom sedi-
ment and stormwater be treated to economically reduce
nutrient enrichment, fish and duck kill, and algal activity to
an acceptable level? Using the trophic state models, a target
reduction level of phosphorus loadings into the
oligotrophic/mesotrophic level may reduce algal blooms. In
addition, a chlorophyll a mean concentration of 7 ^g/l may
indicate a mesotrophic state. Table 3 illustrates the target
level and need for an approximate 90 percent reduction in
phosphorus load and concentration.
In that National Eutrophication Study, total phosphorus
concentration of less than 10 ^g/l in the water column was
noted as a target reduction to classify lakes as oligotrophic.
Table 2. — Retention of nutrients released to Lake Eola in stormwater runoff by bottom sediments.
Nutrient
species
Total
Phosphorus-P
Orthophosphorus-P
N03~2-N
TKN-N
Runoff
loading
(kg)
37.9
6.4
65.9
324.5
Runoff
volume
(1000 cubic
meters)
574
574
574
574
Stormwater
average
concentration
(mg/i)
0.48
0.24
0.65
3.30
Estimated
runoff
mass
loading
(kg)
279.8
140.0
379
1924
R
(%)
86.5
95.4
82.6
83.1
273
-------
Lake Restoration, Protection and Management
o
z
o
o
i Y,. 44.2 * 2,ix 12 o.ee
2 Yj« 40.B + 3.4X 12 0,82
3 ¥8« 37,4 •> 4,7 X 12 0.88
4 Y4« a*.B * 4.7X 12 °-*9
4 • » 10
PRECIPITATION DEPTH
-------
Problems in Lake Restoration
Wanielista, M.P., Y.A. Yousef, and J.S. Taylor. 1981. Stormwater Yousef, Y.A. et al. 1981. Impact of stormwater runoff on Lake Eola
management to improve lake water quality. Final Rep. Grant No. water quality. Proc. 2nd Int. Conf. Urban Storm Drainage, Univ.
R-8055800. U.S. Environ. Prot. Agency, Washington, D.C. Illinois at Urbana-Champaign, Vol. II: 236.
Wanielista, M.P., et al. 1982. Stormwater management to improve
lake water quality—project summary. EPA-600/S2-82-048. Munic.
Environ. Res. Lab. U.S. Environ. Prot. Agency, Cincinnati, Ohio.
275
-------
URBAN LAKES: THEIR BENEFITS AND UNIQUE PROBLEMS
D. B. PORCELLA
Tetra Tech, Inc.
Lafayette, California
ABSTRACT
Lake restoration is just one aspect of a larger issue, lake quality management. Management implies that
there are beneficial uses and that we know how to select for appropriate uses and apply lake quality manage-
ment techniques. On a broad scale this is largely true, but the objectives of lake management change
extensively depending on the uses. In urban lakes, lake restoration is being carried forward to provide
uses that appear to be important to the urban community. Based on experience with a variety of urban
lakes, I will illustrate some of their characteristics and uses, and then discuss various techniques for pro-
viding for these uses. This illustration will be based on a recent analysis of urban lakes in Los Angeles,
Calif. (Tetra Tech, Inc., 1982, Pre. limnol. rep. on Lincoln, Echo, and Harbor Lakes, Los Angeles, Calif.
Draft). Finally, I will make some recommendations that apply to future urban lake restoration projects.
A typical urban lake is very shallow, receives storm runoff,
and is located in a park surrounded by roads and buildings
such as offices, industry, or apartments. Usually these lakes
have extreme water quality problems. These include low
dissolved oxygen, odors, litter and other aesthetic problems,
greases and oils, and in some cases, toxicants. Biological
problems include potential public health problems resulting
from direct water transmission and from vectors (e.g. rats,
mosquitos), excessive plant growth (weeds and algal
blooms), and the absence of fish.
Since many urban lakes lie within natural drainages, they
are often used as settling and equalizing basins for storm
drains. However, the resulting sedimentation reduces their
depth and produces water quality problems. Shallowness
in these lakes increases water quality problems largely as
nutrient cycling increases between water and bottom
sediments. Excessive algal blooms that result from the high
nutrient concentrations are common. Recreational uses
(Table 1) affect other uses because of increasing litter and
by contributing to bank erosion as fishermen perch on the
edges of the lake. The products of productivity and other
biological oxygen demand often result in low dissolved ox-
ygen, noxious plant blooms, loss of fish, and related water
quality effects. Intelligent management and design can
minimize these while preserving the multiplicity of uses
desired by the urban community. These are reflected in the
water quality conditions of three soon-to-be-restored lakes
and two that have been restored (Tables 2 and 3).
Before discussing the methods with greatest potential, let
us consider several obvious but impractical choices. I believe
it is impractical to consider diverting urban runoff away from
the lakes, although that would have some benefits to water
quality. Nutrient control is similarly impractical. Within the
magnitude of money available, I feel it is better to accept
the high productivity, minimize the major water quality and
aesthetic problems, and strive to develop a fishery that can
provide recreation to the urban communities. In other words,
efforts should be directed toward creating a "commons" that
can provide a sense of community along with tangible recrea-
tional benefits. This goal requires making a benefit out of
high productivity.
Since most urban lakes are directed by a parks depart-
ment, the restoration approaches that appear to be most
feasible are those that can be operated as part of a con-
tinuous maintenance program, that do not require technical
sophistication, nor frequent investment of capital. They can
be labor-intensive and have substantial operational costs if
the use level is also substantial.
To make a benefit of high productivity, it is necessary to .
maintain oxygen in the water column, have adequate depth
to meet the needs of the fishery, and minimize the litter pro-
blems while maintaining the physical environment of the lake
and lake facilities. The restoration techniques that seem to
offer greatest promise for meeting the operation and
maintenance objectives while serving these purposes include
aeration for mixing, dredging for deepening and circulation,
and flushing BOD waters that result from plant decay, storm
runoff, and other sources with higher quality water.
Whichever combination of these techniques that is selected,
it must be supplemented by an ongoing maintenance pro-
gram to minimize litter and to maintain the physical integri-
ty of the lake.
Table 4 summarizes how these objectives can be
achieved.
Table 1. — Major uses of urban lakes.
Recreational
Other
In-lake
Boating
Fishing
Aesthetic
Riparian
Games and sports
Feeding of animals
Picnicking
Aesthetic
Walking, jogging
Community buildings
Storm drainage
Settling basin
Flow equalization
Wildlife habitat
Flyways
Resident populations
Rare and endangered species^
276
-------
Problems in Lake Restoration
Table 2. — Los Angeles lakes receiving rehabilitation.
Echo Park
Lincoln
Harbor
Dimensions
Area, acres
Volume, acre ft
Mean depth, ft
Bottom
Perimeter
Fresh water
Runoff
15.1
106.
7.
Mud
Concrete
3 days/week
Major retention basin
5.
10.
3.
Mud
Rubblestone
1 day/week
Major retention basin
44.
440.
5.
Mud
Soil
Unspecified
Major retention basin
Recreational activities
Boating
Swimming
Fishing
Stocked
Camping
Picnicking
Food concession
Yes
No
Yes
Yes
No
Yes
Yes
No
No
Yes
Yes
No
Yes
No
Yes
No
Yes
Yes
Yes
Yes
Yes
Park maintenance
CuSo4 or Cutrine
Duck feeding
Litter removal
Trash removal
Aerators
as needed
unspecified
2/week
as needed
proposed
1/week
occasional
7/week
as needed
proposed
none
unspecified
5/week
as needed
proposed
Table 3. — Other Los Angeles urban lakes.
MacArthur*
Reseda*
Hollenbeck
Dimensions
Area, acres
Volume, acre ft
Mean depth, ft
Bottom
Perimeter
Fresh water
Runoff
Recreational activities
Boating
Swimming
Fishing
Stocked
Camping
Picnicking
Food concession
8.2
56.
7.
Mud
Concrete
1 day/week
Street runoff
Yes
No
No
Yes
No
Yes
Yes
1.7
9.5
3.6
Muck
Concrete
7 days/week
None
No
No
Yes
Yes
No
Yes
No
4.5
13.5
5.
Mud
Concrete
4 days/week
Major retention basin
Yes
No
Yes
Yes
No
Yes
No
Park maintenance
CuSo4 or Cutrine
Duck feeding
Litter removal
Trash removal
Aerators
as needed
3 days/week
7/week
as needed
7
1/week
daily
7/week
as needed
3
2/month
2 days/week
14/week
as needed
None
•Received rehabilitation previously.
277
-------
Lake Restoration, Protection and Management
Table 4. — Recommended rehabilitation plans for three Los Angeles lakes.
Restoration techniques
Lake
Name
Lincoln
Aeration
aeration
plus surface
spray
Bank
stabilization
renovate existing
wall
Dredging
deepen to 6 feet
minimum;
line bottom
(asphalt-concrete)
Fish
habitat
broken pipe
Storm
drain
none planned
Echo aeration substantial bank deepen 1.5 acre broken pipe modify drain
plus surface stabilization plus to 4 feet minimum; to reduce
spray renovation of add dam to control debris
existing wall lotus deposition
Harbor aeration bank stabilization tule removal and broken pipe none planned
plus surface plus riprap; dredge 5.23 acres
spray reconstruct to 10 feet
northeastern
shore
278
-------
Contributed Papers
THE DREDGING OF LIBERTY LAKE AND ITS IMMEDIATE
ENVIRONMENTAL IMPACT
STEPHEN A. BREITHAUPT
DAVID S. LAMB
M. A. Kennedy Consulting Engineers
Spokane, Washington
ABSTRACT
Approximately 20 hectares (50 acres) of sediment containing high concentrations of phosphorus were re-
moved from Liberty Lake, Wash. The sediment removal was accomplished using hydraulic dredging tech-
niques and land disposal of the slurry. Water quality impacts were minimal and transitory in nature. Beneficial
impacts were observed on the disposal area cropland. The goal of removing a substantial portion of the
phosphorus reservoir on the lake bottom was accomplished.
INTRODUCTION
Liberty lake, Wash., located 15 miles east of Spokane,
Washington, has historically been subject to nuisance algal
blooms and widespread growth of aquatic macrophytes.
Residents' concern over these cultural eutrophication pro-
blems resulted in the formation of the Ecology Committee
of the Liberty Lake Property Owners Association. This com-
mittee organized the Liberty Lake Sewer District which was
approved by a majority of voters in the spring of 1973.
Preliminary engineering for wastewater collection and treat-
ment began in late 1973, and the draft facility plan was com-
pleted in late 1975 by M. A. Kennedy Consulting Engineers.
William Funk at Washington State University was doing
water quality research concurrent with the activities of the
District's engineer. Detailed research on Liberty Lake's pro-
blems began in spring of 1973, although some work had
been ongoing since 1968. A report of the results of this work
was published in January 1974 (Funk et at. 1974), discuss-
ing several sources of nutrients to the lake. An alum treat-
ment (Al^SO^g 14H20) was performed in fall of 1974 to
precipitate phosphorus present in the lake water.
The Liberty Lake Rehabilitation Program proposed to the
U. S. Environmental Protection Agency In late 1975 was a
multi-phased pollution control and lake restoration program.
The pollution control elements included extensive wastewater
collection and treatment systems to alleviate the leaching
of nutrients into the lake from septic tank drainfields, and
construction of confinement and diversion facilities for the
primary stream feeding the lake to reduce the flushing of
nutrients from the marsh into the lake.
The restoration efforts included dredging, alum treatment,
and a variety of nonstructural activities including a storm-
water management report and public education and involve-
ment. The dredging was designed to reduce the impact of
internal nutrient recycling. Alum treatment inactivated the
phosphorus in the water column by floceulation and precipita-
tion. It also reduced internal nutrient recycling by sealing the
lake bottom. The stormwater management report and public
279
-------
L3Ke nestoration, protection ana Managemeni
involvement reduced the impact of human activity on the land
surface of the lake basin.
This paper will be limited to a discussion of the dredging
aspects of the Liberty Lake Restoration Project: specifical-
ly, how the dredging operation was carried out, problems
confronted during the operation, and the results of en-
vironmental monitoring.
DESCRIPTION OF THE DREDGING
OPERATION
The site map for Liberty Lake as well as the area dredged
is shown in Figure 1.
WICOMICO
BEACH
DRF.AMWOOD L1AY
j_AKE SURFACE ELEVATION = 6Z4.26 METERS
LAKE DEPTHS IN METERS
WKST FORK Y$S^EAST FORK-'
mv CHEEK-^f^LIDERTY CREEK
Figure 1 .—Liberty Lake, Wash, showing the approximate locations
of the dredge area boundary (hatched area) and the dredge area
sampling station (indicated by the "x").
Table 1. —Levels of total phosphorus in sediment
of various physical types from Liberty Lake, Wash.
(Funk et al. 1978).
Sediment physical type1
Total phosphorus content
0-20 cm layer (mg P/kg dry wt.)
Ill Light Organic Deposits 316 ± 69
IV Moderate Organic Deposits 474
V Heavy Organic Deposits 627
VI Heavy Muck 1995 ± 966
See Figure 2.
Studies of the lake sediments revealed a high concentra-
tion of total phosphorus at the southern end of Liberty Lake
(Table 1 and Fig. 2). This heavy muck sediment was also
found to be very flocculent, easily disturbed, and to have
much greater phosphorus release rates than other sediments
found in Liberty Lake. The design for the dredging opera-
tion called for removing the top 0.6 m (2 ft.) of this sediment
from 20 hectares of lake bottom (50 acres) (M. A. Kennedy,
1979).
Removal of sediment from the southern end layer was
anticipated to have several benefits: (1) remove a large phos-
phorus reservoir, (2) expose sediment of lower phosphorus
content, (3) expose sediments of a higher degree of con-
solidation and stabilization, and (4) reduce growth rate and
biomass of aquatic macrophytes in the area dredged (M. A.
Kennedy, 1979).
This sediment was removed by using hydraulic dredging
techniques. It should be noted that two contractors were
used during the dredging operation. The first contractor used
a standard 20 cm cutterhead dredge that removed sediment
in 46m-wide swaths. The second contractor employed a
horizontal auger-type dredge that removed sediment in 2.4m-
wide swaths.
The progress of the dredge across the dredge area was
monitored by surveying techniques using permanent con-
trol points established on the shore adjacent to the dredge
area. The cutterhead dredge was located using two control
points, while the auger-type dredge was located from one
control point with distance obtained from a stadia rod
mounted on the dredge. The monitoring of the depth of cut
used an infrared sediment sensing device as the primary
sensing device. It accurately determined the depth to the
sediment surface. This was used to calibrate a weighted,
inverted cone attached to a calibrated line. The latter device
was used mainly because of the simplicity of operation.
Lake water quality was monitored during the dredging
operation to determine its immediate impacts on Liberty
Lake. Water samples were collected at three lake stations
and at three depths for each of those stations. The samples
were analyzed for alkalinity, acidity, pH, conductivity,
temperature, dissolved oxygen, total suspended solids,
biochemical oxygen demand, chemical oxygen demand,
sulfate, total phosphorus, soluble reactive phosphorus,
nitrate, nitrite, ammonia, and total Kjeldahl nitrogen. This
paper will limit the discussion to the dredge area station (see
Fig. 1). Other results will appear in the forthcoming final
report of the Liberty Lake Restoration Project.
The dredged sediment was pumped across the lake as
a slurry through a 25cm-diameter polyethylene pipe. The
pipeline was marked at about 15 m intervals with colored
buoys, so lake users could see it. When the pipe contained
water it would float, but when it contained the dredged slurry
SCALE
O.L DEPTHS IN FEET
IX 0305 = M)
(MACKENZIE Bfft
DREAMWDOD RAY
I.
II.
III.
IV.
V.
VI.
SAND
ROCK .GRAVEL OR SHINGLE
RELATIVELY LIGHT ORGANIC DEPOSITS WITH Cl.AY AND HACI'I
MODERATE ORGANIC DEPOSITS
MODERATELY HEAVY ORGANIC DEPOSITS, UNCONSOLIDnTfD
HEAVY MUCK AND ORGANIC DEBRIS, HIGHLY UNCONSOI.IMTri)
VI 1. CONSOLIDATED FIBROUS PEAT
WE5T INLET
n ACRES
H ACRES
H78 ACRES
173 ACRES
7 7 ACRES
90 ACRES
7 ACRES
LZZ1
mm
cm
NOTE' BdSED ON DATA REPORTED IN FUNK (M n|,ll!VU)
Rgure 2.—Sediment physical type distribution of Liberty Lake, Wash.
Note that type VI sediment predominates at the south end of the
Lake.
280
-------
Contributed Papers
it sank out of sight. A 20 cm booster pump, powered by a
174 kw (230 hp) diesel engine, jwas used Jo overcome the
head losses in the pipe. A total of 4~040 m of pressure pipe
was used to carry the slurry across the lake, overland, and
to a gravity-flow sewer interceptor, which was 46 to 53 cm
in diameter. The use of this interceptor was at the discretion
of the contractor and if used, it was stipulated that pumping
rates would be maintained between 0.11 to 0.15 rrvVsec (3.9
to 5.6 cfs). This flow rate was required to keep the solids
in suspension as well as to keep from overloading the sewer.
Approximately 2,400 m of the interceptor were used by both
contractors.
The slurry in the sewer interceptor was discharged to a
wet well. This provided storage capacity after dredge shut-
down in the event of an irrigation pump breakdown.
An irrigation pump similar to the booster pump was used
to pump the slurry from the wet well to the disposal area
comprised of approximately 80 hectares (200 acres) of
cropland. The slurry was pumped through a 35m-diameter
steel pipe across the disposal area and distributed from this
"header pipe" with 15cm-diameter pipe to either spray or
flood irrigation sites. Spray irrigation, the disposal method
stipulated for the first contractor, used rotating irrigation
nozzles. The contractor chose to use a 50mm-diameter
orifice that gave a discharge rate of 0.057 m3/sec and a
sprayed circle of approximately 50 m in diameter. It was also
stipulated that ponding and runoff of the spoils should not
occur. However, the contractor was unable to operate the
disposal operation so as to prevent this from occurring. Use
of the 50mm-diameter orifice did not break up the slurry
stream being sprayed, causing gouging of the soil around
the sprayed perimeter. Using a 38mm-diameter orifice did
break up the stream, but the contractor felt its use would
slow their production and therefore it was used only as a
last resort. However, freezing conditions set in at this time,
compounding the runoff problem. As a result the agreement
with this contractor was eventually terminated.
Following discussion with the disposal area property
owner, the second contractor was allowed to use flood ir-
rigation as a disposal method. The disposal method was
changed because of uncertainty about project completion
under spring weather conditions, using spray irrigation, by
the beginning of fishing season (this was a constraint put
on the project by the Washington State Department of
Game). Berms were constructed to contain the total expected
volume of slurry, i.e. 4.3 x 105m3 (352 acre ft.). This much
containment was not needed because slurry makeup water
percolated into the soil relatively fast.
The total volume of slurry produced was 4.05 x
(328 acre ft.), with the first contractor producing
8.26 x I04m3 (67 acre ft.) of slurry and the second contrac-
tor producing 3.22 x I05m3 (261 acre ft.). Production rates
were 2.85 x if^m3 of slurry produced per hectare dredged
(9.5 acre ft./acre) and 1 .77 x 1 04m3/hectare (5.8 acre
ft/acre), respectively. This shows that more slurry was pro-
duced using the standard 20 cm cutterhead dredge than
when using the horizontal auger-type dredge, for approx-
imately the same depth of cut.
Nineteen days after dredging had been completed the
slurry makeup water had percolated into the soil enough so
that the berms containing the slurry could be removed. Three
and a half months after the dredging the residue was dry
enough to begin spreading over the cropland. This was tilled
into the soil and the land eventually planted to barley by the
property owner.
Soil samples had been collected, before application of the
slurry and after tilling of the dried residue into the soil. The
results are discussed later in this paper.
PROBLEMS ENCOUNTERED DURING
DREDGING OPERATIONS
As previously discussed, the first contractor was not able
to manage the disposal area so as to prevent ponding and
runoff of the slurry.
It was suspected that in using the _standard cutterhead
dredge the operators may have been overexcavating to en-
sure that the top 0.6 m of sediment was removed. If such
were the case it is possible that the surface sediment to be
removed may have been undercut and not removed. Inspec-
tion of the cut swath by scuba revealed some areas along
the edge of the cut where sediment and macrophytes had
been pushed into piles, apparently a result of the swiveling
action of this dredge.
There was also a problem with sewer surcharge, where
the slurry flowed out of the manholes. This was a result of
the dredge and booster pumps exceeding the capacity of
the sewer interceptor.
The probable cause for these problems was that this
dredge was oversized for this project and couldn't operate
efficiently within the limits of the slurry disposal system.
No disposal problem or sewer surcharge occurred with
the second contractor. The problems encountered were of
a nature that reflected the different mode of operation of the
smaller dredge.
The horizontal auger dredge was equipped with a boom
that allowed it to cut to a 4.6 m depth. Because of snowmelt
and rise in lakewater level, it soon became evident that the
auger would not remove 0.6 m of the sediment surface in
water deeper than 4.0 m, i.e. along the western and nor-
thern portions of the dredge area (see Fig. 1). An extension
boom was installed on the dredge to extend the cut to 7.0
m, which was adequate for this project's needs.
The dredge made a cut 2.4 meters in width and was pro-
pelled back and forth across the lake by means of a dredge-
mounted winch attached to an anchored cable. The cable
was anchored on east and west shores that were 490 to 600
m apart resulting in slack in the cable that could not be re-
moved. Thus the dredge could move off line, especially near
the center of the cable. As a result it is possible that some
areas may have been missed because of wind action and/or
the force imparted to the dredge by the slurry pipeline con-
nected to it.
Macrophytes, mainly E/ocfea, had a tendency to wrap
around the auger and plug the suction intake of the dredge.
This was a problem mostly at the southern portion of the
dredge area. It slowed production because the operation had
to be shut down while the operator removed the macrophytes
from the auger. As the dredging progressed towards the nor-
thern end of the dredge area, out of the E/octea beds, this
was less of a problem.
RESULTS OF ENVIRONMENTAL
MONITORING
Overall, monitoring showed the dredging operation had no
discernible adverse impact on the lake as a whole. Some
localized, transient effects were noted, however.
Suspended solids and Secchi disk transparency near
the dredge—A plume of solids was created (Fig. 3) on the
cut-side of the auger during the cutting. The plume never
reached the lake surface and did not affect the transparen-
cy of the water column, at least as seen from the surface.
A maximum solids content of 49.0 mg/l was observed in the
cut trough, but decreased to about 10 percent of that within
2 m of the lake surface.
Suspended solids—Overall the suspended solids con-
tent was very low, but an increase was observed in the
281
-------
Lake Restoration, Protection and Management
Figure 4.—Suspended solids concentration (mg/l) as ob-
served at the dredge area station during the spring 1981, Mud Cat
dredging operation.
DISTANCE FROM CUTTERHEAD (FT.)
Figure 3.—Secchi disk transparency (vertical lines with horizontal
bar) and suspended solids concentration (mg/l) near the Mud Cat
dredge on 4/22/81 during cutting operation. The arrow indicates
direction of travel of the dredge.
4/17/81 sampling data (Fig. 4). This increase corresponds
with the closest approach of the horizontal auger dredge to
the dredge area station. Solids produced by the dredging
operation may have caused this increase. However, a diatom
peak was also observed at this time (Gibbons, pers. comm.).
This could also account for the increase in the solids content!
Total phosphorus—An increase in total phosphorus was
observed, with a peak of 15 pg P/l in the bottom waters of
the dredge area on 4/17/81, the time of closest approach
to the station (Fig. 5). This increase might have been caus-
ed either by the passage of the dredge in the vicinity of the
station or the diatom peak observed at this time, as sug-
gested by Gibbons (pers. comm.).
Chemical oxygen demand—There were no readily
discernible trends in chemical oxygen demand (COD) at the
dredge area station associated with the passage of the
dredge near the station (Fig. 6). These sediments have a
high organic matter content, so if the dredging operation was
stirring the sediments into the water column COD would be
expected to increase.
Soil chemistry data comparisons, using a Mann-Whitney
nonparametric test for significance (a = 0.05), of the spoils
disposal site were made for samples collected prior and
subsequent to the disposal operation (Table 2). Most con-
stituents showed no significant changes. However, available
phosphorus (P), nitrate (NO3) and ammonia (NH4) concen-
trations exhibited a significant difference between sample
sets. P and NO3 both increased, while NH4 decreased. This
indicates that the sediment is a good soil amendment.
CONCLUSION
The Liberty Lake Restoration project is a multiphased pro-
ject, with the dredging operation being one of the in-lake
phases. It accomplished the goal of removing a substantial
portion of the nutrient reservoir residing in the sediments of
Liberty Lake. This should help reduce the internal nutrient
recycling found to occur in the lake. In addition, the dredg-
ing operation was accomplished without adverse en-
vironmental impact.
3/4 3/II 3/18 3/25 4/1 4/8 4/15 4/22 4/29 5/6
Figure 5.—Total phosphorus concentration (ng P/l) as ob-
served at the dredge area station during the spring 1981, Mud Cat
dredging operation.
3/4 3/II 3/18 3/25 VI 4/8 4/15 4/22 4/2B S/8
Figure 6.—Cherriical oxygen demand (mg COD/I) as ob-
served at the dredge area station during the spring 1981, Mud Cat
dredging operation.
282
-------
Contributed Papers
Table 2. — Disposal area soils data before (10/30/80) and after (4/27/82) slurry application showing the mean and standard
deviation of sample sets and results of a Mann-Whitney nonparametric test for significance (a = 0.05)
CONSTITUENT
PH
% Org.
P (ppm)
NH4 (ppm)
N03 (ppm)
K (ppm)
Ca (meq/100 gm)
Mg (meq/100 gm)
10/30/80
(n = 4)
6.0 + 0.1
6.4 ± 0.7
1.0 ± 9.2
9.0 + 4.6
6.88 ± 4.44
188+34
13.6 ± 0.8
2.11 + 0.04
4/27/82
(n=10)
5.70 + 0.35
5.30 ± 1.7
2.5 ± 2.1
3.23 ± 4.08
20.5 ± 12.7
191 + 9
13.7 ± 0.9
2.05 ± 0.10
SIGNIFICANCE1
NS
NS
Sig.
Sig.
Sig.
NS
NS
NS
NS: not significant; Sig: significant
REFERENCES
Funk W. H., G. C. Bailey, P. J. Bennet. 1974. Water quality study
of Liberty Lake with special reference to sediment analogies: Rnal
Rep. to M. A. Kennedy Consulting Engineers. Dep. Civil Environ.
Eng. Washington State Univ. Pullman.
Funk, W. H., H. E. Gibbons, G. C. Bailey, and P. J. Bennet. 1978.
Effect of restoration procedures upon Liberty Lake: 1st, 2nd, and
3rd status reports to M. A. Kennedy Consulting Engineers. Dep.
Civil Environ. Eng. Washington State Univ.
Gibbons, H. L. 1981. Personal communication. Dep. Civil Environ.
Eng. Washington State Univ.
Michael A. Kennedy Consulting Engineers. 1979. Liberty Lake resto-
ration plan for Liberty Lake Sewer District. Spokane, Wash.
283
-------
VANCOUVER LAKE: PRE-RESTORATION STATUS AND
RESTORATION PROGRESS REPORT
RICHARD B. RAYMOND
FREDRICK C. COOPER
Cooper Consultants, Inc.
Portland, Oregon
ABSTRACT
Water samples collected weekly or biweekly from Vancouver Lake beginning July 10, 1981, were ana-
lyzed for turbidity, suspended solids, alkalinity, dissolved oxygen, pH, nitrogen, phosphorus, conductivity,
algae, and total chlorophyll. Measurements were made in the lake for water temperature, depth and Sec-
chi disc depth. Results of the analyses indicate that Vancouver Lake is a nutrient-rich, productive, eutrophic
lake. Because it is shallow during most of the summer, the lake does not stratify and does not develop
anoxic conditions at the sediment-water interface. Bottom sediments are frequently mixed into the water
column because the lake is shallow and exposed to frequent winds. The water regime and chemistry of
the lake appear to be controlled by a complex interaction between the varying flow of Columbia River
water into and out of the lake, the chemistry of the major influent stream, Burnt Bridge Creek, and the
effects of very large populations of algae that occur during the summer. Vancouver Lake is unusual because
it does not have large populations of blue-green algae or attached aquatic macrophytes, both of which
would ordinarily be expected in a warm, shallow, .nutrient-enriched lake. In May 1982, dredging began
in the lake as part of the Vancouver Lake Restoration Project. During the summer of 1982, increased
recreational use occurred in those areas already dredged. Suspended solids and algal cell density was
lower in summer 1982 than in 1981, but transparency and turbidity remained at about the same level.
This change in conditions is attributed to the effect of fine material suspended in the lake as a result of
dredging.
INTRODUCTION
Vancouver Lake is located in the Columbia River floodplain,
adjacent to the city of Vancouver in southwestern Clark
County, Wash., within the greater Portland, Ore. metropolitan
area (Fig. 1). The predominant land use adjacent to the lake
is agricultural, although the south and west shorelines are
included in a county park. Industrial activity related to the
Port of Vancouver occurs south of the lake and includes a
large aluminum smelter. The primary residential use close
to the lake is in conjunction with farming. Additional residen-
tial areas are located on lowlands southeast of the lake and
along the top of the east shore bluff.
The low-lying lands to the north, west, and south are sub-
ject to seasonal flooding from the Columbia River which flows
within 1 mile of the southwest shore of the lake. These low-
lands have an elevation of from 3 to 6 meters above mean
sea level (msl). The northeast shore of the lake is formed
by bluffs rising to an average elevation of 60 meters msl.
The climate of the region is maritime Mediterranean with
moderately warm, dry summers and mild, wet winters. Sev-
enty-five percent of the annual precipitation occurs be-
tween October and March. Annual total precipitation is ap-
proximately 100 cm.
Vancouver Lake has a surface area of 1,100 ha, is 4 km
across from east to west, and has a mean shoreline length
of about 12 km. The depth varies seasonally, ranging from
a mean depth of less than 1 meter in September and Oc-
tober (maximum depths as low as .6 m have been record-
ed) to about 4 m in early June. The lake has a virtually flat
bottom except for higher areas at the northern end caused
by sedimentation of materials carried into the lake by Lake
River (Fig. 2). These shallower areas are exposed during
lowest water.
HYDROLOGY
The hydraulic regime of Vancouver Lake is complex and
involves Burnt Bridge Creek, Lake River, Salmon Creek,
and Columbia River (Fig.3).
Burnt Bridge Creek flows east through commercial and
suburban sections of Vancouver and drains an area of ap-
proximately 70 square km. Mean annual flow is about 0.6
m/s, but flows are quite variable. Considerably higher flows
are observed during rainfall. Burnt Bridge Creek flows are
usually between 0.1 and 0.3 m/s (Dames and Moore,
1977.)
Lake River joins Vancouver Lake at its northern extremi-
ty and connects to the Columbia River approximately 14
miles to the north. The volume and direction of flow in Lake
River is directly related to seasonal and tidal changes in
the stage of the Columba River. During April and May while
the Columbia River is rising because of spring runoff, Lake
River flows south into Vancouver Lake with flows that
reach 5.7 cu m/s. In late June the stage of the Columbia
River drops rapidly and Lake River reverses to flow north
out of Vancouver Lake. Flows during this period may reach
4.25 cu m/s.
During much of the year, August through March, the flow
in Lake River is variable. It may reverse daily in response
to tidal changes in the Columbia River, or it may flow north
or south for several days at a time because of longer-term
changes in the Columbia River resulting from weather or
power generation.
Salmon Creek is a tributary of Lake River that drains
a large rural and agricultural area north of Vancouver. It
empties into Lake River about 3 km north of Vancouver
Lake. Water from Salmon Creek can enter Vancouver
Lake during southward flow of Lake River.
The Columbia River has an annual mean flow at Van-
couver of 5,714 cu m/s. Flow is distinctly seasonal, as low
284
-------
Contributed Papers
VlCINlf Y MAP
Figure 1 .—Vicinity map of Vancouver Lake.
Station 1
SAMPLING
STATIONS
Mud Flats
Figure 2.—Sampling stations in Vancouver Lake during restoration.
as 1,398 cu mis in the fall and as high as 18,400 cu m/s
during the spring snow melt.
Because of the great variation in flow, the level of the
Columbia River at Vancouver changes considerably ac-
cording to season. Mean maximum river stage is greater
VANCOUVER LAKE
COMPLEX
Columbia River
Burnt Bridge Creek
Figure 3.— Vancouver Lake and associated waterways,
than 4.6 meters msl, while during August river stage may
fall as low as 0,6 meters msl.
During low water in late summer, the tide fluctuates daily
in the Columbia River stage near Vancouver by approx-
imately 0.6 meters.
The hydrology of Vancouver Lake is directly controlled
by the stage of the Columbia River (Fig. 4). The lake is
at its lowest level in late October. Flow in Lake River can
reverse daily in response to tidal fluctuations in the Co-
lumbia River. As the Columbia River and flow in Burnt
Bridge Creek begin to rise from the winter rains, the lake
level rises accordingly to an intermediate level with
perhaps a mid-winter peak following particularly heavy
rain. Flow in Burnt Bridge Creek increases rapidly as the
rainy season begins. Between November and February
the flow in Lake River is frequently reversed and may flow
into Vancouver Lake for several days at a time. The ef-
fects of tidal fluctuations diminish as the river stage rises.
In late spring (April-May) the Columbia River rises in
•esponse to snowmelt runoff. During this time, until the
MEAN DEPTH
"Tj
DAYS SINCE 7/1/81
Figure 4.—Depth of Vancouver Lake, The reported value is the mean
of depths at stations 2, 3, and 4.
285
-------
Lake Restoration, Protection and Management
highest water in mid-June, flow in Lake River is south in-
to Vancouver Lake. The volume of water in the lake may
increase fourfold from November through early June (8.5
to 35 million cubic m). From mid-June to mid-July the lake
level drops rapidly from its high mark of around 3.7 meters
msl to near the annual low of 1.0 meters msl. During this
time, Lake River flows north, carrying water away from the
lake.
From July to December the lake remains at a low level,
fluctuating around 1 to 1.7 meters msl elevation. While the
flow in Lake River may reverse daily in response to Colum-
bia River tidal fluctuation, it is not clear if water from the Col-
umbia River is actually reaching the lake during this period.
WATER QUALITY
As part of the water quality monitoring program of the Van-
couver Lake Restoration project, water samples have been
taken weekly or biweekly at several sites in the lake and in
Lake River. Sampling began July 10,1981, and will continue
until 1 year after completion of construction on the project.
Table 1 gives summary values for the parameters surveyed.
These results characterize Vancouver Lake as a shallow,
turbid, nutrient-rich, eutrophic lake.
Because of its temperate climate and complex hydrology,
many of the parameters in Vancouver show a seasonal cy-
cle. To more clearly show this seasonal variability, mean
values of data from three stations in the lake (Stations 1,
2, and 3) are represented graphically in Figures 5 through
24. Figure 2 shows station locations.
Vancouver Lake remains completely mixed almost con-
tinuously because it is so shallow and is exposed to frequent
winds. As a result, the lake temperature is very closely cou-
pled to air temperature. Daily temperature variations near
the surface of 8° Celsius were recorded during warm sum-
mer days. Daily temperature variation near the bottom was
as much as 6° C. In all cases, the lake cooled enough dur-
TEMPERATURE
S2 is
8
DAYS-SINCE 7/1/81
Figure 5.—Midday surface water temperature at Vancouver Lake
measured at 0.3 m depth.
ing the night to become isothermal on all but the calmest
days. The greatest difference between surface and bottom
minimum temperature observed to date has been 2°C. The
maximum lake water surface temperature was 28° C in
September 1981. Minimum water temperature recorded was
6° C in December 1981. The lake surface froze completely
on January 8, 1982 and remained frozen for 7 days. The
annual water temperature cycle is presented in Figure 5.
Other parameters whose variability seems to be primarily
related to the seasonal cycle are conductivity and alkalinity.
Conductivity decreases steadily throughout the fall, reaching
a low value in January (Fig. 6). This occurs before there has
been a significant influx of water from the Columbia River,
and the value is lower than the normal mean conductivity
in the Columbia in January of about 100 /^mhos/cm (Portland
G. E. Co., 1981). This decline in conductivity is probably
caused by dilution by low conductivity rainwater from Burnt
Table 1. — Water quality data. Values for Vancouver Lake are means of all weekly or biweekly samples taken between
July 10, 1981 and Oct. 5, 1982. Values for the Columbia River are annual means from Portland G. E. Co. (1981).
Station
Parameters
Dissolved oxygen
Temperature
PH
Conductivity
Suspended solids
Turbidity
Alkalinity
Secchi depth
NO3N
NH3N
Kjeldahl N
Organic N
NO2N
PO4P
Total P
Chlorophyll
Algae
Depth
Units
mg/l
degrees C
units
^mhos/cm
mg/l
NTU
milli eq./l
m
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/l
mg/cu m
cells/ml
m
Vancouver Lake
1
9.6
15.5
7.5
137
49
57
1.07
0.37
0.369
0.233
1.44
1.71
0.008
0.062
0.193
51
31 ,000
2
10.2
17.2
7.8
142
80
72
1.11
0.32
0.174
0.175
1.65
1.46
0.004
0.040
0.218
60
23,000
2.031
3
9.9
15.6
7.8
139
62
66
1.04
0.35
0.250
0.150
1.34
1.07
0.004
0.033
0.219
50
28,000
1.15
4
10.0
16.4
7.8
141
71
71
1.08
0.31
0.225
0.162
1.56
1.31
0.002
0.056
0.225
60
29,000
1.88
5
10.7
16.6
8.0
152
57
44
1.14
0.37
0.345
0.150
1.80
1.56
0.003
0.025
0.228
80
33,000
1.54
Columbia
River
—
10.7
1 1 .6
7.8
135
__
15
0.97
0.84
0.25
0.07
_
0.01
0.066
0.08
12
1,800
286
-------
Contributed Papers
CONDUCTIVITY
ALGAL DENSITY
8 '5
JASOMDJFMAMJJASC
DAYS SINCE 7/1/81
Figure 6.—Conductivity of surface water in Vancouver Lake; mean
of stations 1, 2 and 3.
ALKALINITY
2.00
1.80
1.60
J A S 0
DAYS SINCE 7/1/81
Figure 7.—Alkalinity of surface water in Vancouver Lake; mean of
stations 1, 2 and 3.
Bridge Creek. Evidence to support this comes from
September 1982 when the conductivity at Station 5 declin-
ed sharply following several days of heavy rains. This drop
corresponded with a similar but smaller magnitude drop in
Vancouver Lake.
Alkalinity follows a curve very similar to conductivity. The
alkalinity drops during the winter while the pH remains
relatively constant, indicating a response to dilution (Fig. 7).
PHYTOPLANKTON
Phytoplankton density in Vancouver Lake is high (Fig. 8).
Cell density, averaged for the three stations in the lake pro-
per, reached 200,000 cells per milliliter in August 1981.
Chlorophyll concentration is correspondingly high (Fig. 9).
This high density limits the recreational potential of the lake,
and also influences some aspects of water quality.
Total suspended solids, turbidity, Secchi disc depth,
dissolved oxygen, and pH are all influenced directly by the
number and activity of algal cells in the lake. There is a
significant correlation between algal cell density and total
suspended solids (r2 = 0.8755, p< 0.001), and between
algal cell density and Secchi disc depth (r2 • -0.7118, p<
0.01).
The activity of large numbers of algal cells is also evident
in the pH and dissolved oxygen values (Fig. 10 and 11). Most
samples were taken near midday during maximum photosyn-
thesis. During the period of peak algal densities, pH during
the day exceeded 9.0 and dissolved oxygen content reached
a
NOJFMAKJJ
DAYS SINCE 7/1/81
S 0
Figure 8.—Algal cell density in Vancouver Lake; mean of stations
1, 2, and 3. For solitary species each cell is counted, for typically
colonial forms each colony is counted as one unit.
CHLOROPHYLL
300
3 15C -
N
f
IOO -
DAYS SINCE 7/1/81
Figure 9.—Total chlorophyll concentrations in Vancouver Lake; mean
of stations 1, 2, and 3. Chlorophyll is measured by in vivo
fluorescence and adjusted to chlorophyll using methods of Strickland
and Parsons (1972). The values are not corrected for chlorophyll
degradation products.
DISSOLVED OXYGEN
JASuNDJFHAHJ
DAYS SINCE 7/1/91
A 5 0
Figure 10.—Midday dissolved oxygen in Vancouver Lake; measured
at 0.5 meter depth, or halfway between surface and bottom,
whichever is less. Measurements were made by polarographic
technique.
287
-------
Lake Restoration, Protection and Management
pH
0/IVS SINCE 7/1/81
Figure 11 .—Midday pH in Vancouver Lake; mean of stations 1, 2,
and 3.
154 percent saturation. No data on diurnal variation in dis-
solved oxygen are as yet available.
By far the most common species in the phytoplankton is
a diatom, Syneora fasiculata var. truncate. Next in abundance
are members of the diatom genus Stephanodiscus. Other
commonly occurring algae include members of the genera
Nitzschia, Navicula, and Fragilaria, as well as Cyclotella
glomerta, Cryptomonas erosa and Rhodomonas minuta; all
except the last two are diatoms. Cyclotella glomerata some-
times replaces Synedra fasiculata truncata as dominant at
the mouth of Burnt Bridge Creek (Station 5). This station is
heavily influenced by the flow from the creek and shows
several differences from the stations in the lake. Although
occasionally wind causes local high densities of blue-green
algae, blue-greens were a minor component of the phyto-
plankton, rarely comprising more than 5 percent of the total
count and usually less than 2 percent. Table 2 lists species
comprising at least 1 percent of at least one sample.
Also included in Table 2 is the relative frequency of oc-
currence of each species in all the samples and its mean
rank in the samples in which it appeared. In this tabulation,
a species with a low frequency of occurrence would be ex-
pected to also have a low mean rank. A low frequency com-
bined with a relatively high rank would indicate a species
that was either abundant but present at only one station, or
abundant at several stations, but only for a short period of
time.
The composition of the phytoplankton in Vancouver Lake
is interesting in several respects. The cell density is very high,
up to 250,000 cells/ml. The most abundant alga in the lake,
Synedra fasiculata var. truncata, has not been found in any
of the 100 lakes and rivers in the Northwest previously ob-
served or reported in the literature (Sweet, 1982). Another
feature is the insignificance of blue-green algae in the lake.
Blue-green algae are normally abundant in a shallow, warm,
nutrient-rich lake with relatively high ptH such as Vancouver
Lake.
Synedra fasiculata var. truncata develops best in water of
high conductivity, or in slightly brackish water. The two locally
common Stephanodiscus species, S. ambigua and S. hant-
schia, are typical of eutrophic waters, both in lakes and rivers.
S. ambigua is dominant at times in hypereutrophic Upper
Klamath Lake. S. hantschia is more typical of mesotrophic-
eutrophic rivers, but has been found in a shallow hypereu-
trophic pond at very high densities (Sweet, 1982). Both
Stephanodiscus species are very common, at times domi-
nant, in the Columbia River. The pennate diatoms Nitzschia,
Navicula, and Fragilaria are periphytic algae commonly found
in enriched streams.
Cyclotella glomerata is the only alga in Vancouver Lake
that repeatedly replaces Synedra fasiculata as dominant, and
this occurs only at the mouth of Burnt Bridge Creek (Sta-
tion 5). Cryptomonas erosa, Rhodomonas minuta, and Nitz-
schia palea are at times fairly common. These three algae
are typically associated with organic-rich waters, Nitzschia
palea especially with organic pollution. N. palea is more com-
mon at Station 5 than at the other stations.
Three algal blooms were noted in the lake (Fig. 8). The
earliest occurred in March when Stephanodiscus astrea var.
minutula increased dramatically, becoming the dominant
(most common) species for a short period. This species is
present in the Columbia River, and also in Vancouver Lake
during much of the year. The second bloom, an increase in
the abundance of Aphanazomenos flos-aquae, came later
in the summer, in late June 1982, somewhat later in 1981,
and represents an increase in the abundance of Aphanazo-
menos flos-aquae. This increase in A. flos-aquae occurs just
after the major influx of Columbia River water to the lake.
During this bloom Aphanazomenon becomes the most com-
mon species at some stations because of wind-blown con-
centrations. In the past two summers these have been quite
localized and in general have not reached objectionable pro-
portions. The Aphanazomenon bloom quickly fades to be
replaced by a dramatic increase in the numbers of Synedra
fasiculata var. truncata. This alga is always present in the
flora of Vancouver Lake, is usually the dominant species,
and during its peak growth, commonly comprises 60 to 80
percent of the total cell count. It rarely accounts for less than
30 percent of the cell count.
The algae indicate that Vancouver Lake is nutrient-
enriched (nutrients are not limiting growth) and that the in-
Table 2. — Algal taxa comprising at least 1 percent of at least
one phytoplankton sample. Relative frequency is the percent
of all samples in which the taxon appeared at greater than 1
percent. Mean rank is the arithmetic mean of the taxon's rank in
the samples in which it appeared.
Species
Synedra fasiculata truncata
Stephanodiscus astrea minutula
Staphanodiscus hantzschii
Fragilaria construens
Cyclotella glomerata
Rhodomonas minuta
Melosira ambigua
Nitzschia acicularis
Cryptomonas erosa
Melosira distans
Ankistrodesmus falcatus
Anabena sp.
Navicula minuscula
Nitzschia palea
Nitzschia frustulum
Aphanzomenon flos-aquae
Chodatella Wratiawiensis
CWamydomonas-like
Melosira granulata angustissima
Miscellaneous green algae
Nitzschia holsatica
Scenedesmus quadricauda
Stephanodiscus subsalsus
Melosira italica
Scenedesmus abundans
Nitzschia sp.
Relative
frequency
96
92
96
92
69
55
55
55
49
39
36
34
31
26
24
22
22
22
20
16
14
14
12
12
12
10
Rank
1.27
3.93
4.33
5.75
5.84
6.00
6.64
6.68
8.86
8.59
8.25
6.73
6.00
6.64
10.50
3.56
8.11
10.67
7.88
10.83
7.80
9.80
7.33
8.75
12.23
15.00
288
-------
Table 2. (Continued)
Species
Acnanthes linearis
Selenastrum minutum
Synedra radians
Nltzschia paleacea
Cyclotella Kutzinglana
Navicula muralis
Melosira granulata
Oocystis pusilla
Fragilaria pinnata
Scenedesmus acuminatus
Gomphonema angustatum
Acnathes lanceolata
Miscellaneous blue-green algae
Navicula pupula
Miscellaneous pennate diatoms
Kephyrion spirals
Trachelomonas volvocina
Tetraedron regulars
Navicula cryptocephala
Acnanthes minutissima
Asterionella formosa
Nitzschia amphibia
Tetrastrum staurogeniaeforme
Navicula sp.
Cyclotella pseudostelligera
Cyclotella meneghiniana
Sphaerocystis schroeteri
Stephanodiscus astrea
Mallomanas sp.
Crucigenia quadrata
Nitzschia dissipate
Synedra delicatissima
Synedra ulna
Amphora perpusilla
Navicula viridula
Navicula moumei
Navicula minima
Cymbella minuta
Nitzschia filiformis
Fragilaria vaucheriae
Rhoicosphenia curvata
Nitzschia clausii
Relative
frequency
8
8
8
8
6
6
6
6
6
6
4
4
4
4
4
4
4
4
4
4
4
4
4
4
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
2
Rank
7.00
8.75
9.75
12.25
8.00
9.50
10.00
11.33
12.67
13.33
6.50
8.00
8.50
9.50
10.00
10.50
11.00
11.50
11.50
12.00
12.00
13.00
13.00
14.50
8.00
9.00
9.00
11.00
12.00
13.00
13.00
13.00
14.00
14.00
15.00
16.00
16.00
20.00
21.00
22.00
23.00
25.00
flow from Burnt Bridge Creek may be loading organic pollu-
tion and associated elements into the lake.
An unusual aspect of the biota of Vancouver lake is the
absence of aquatic macrophytes in the lake. A lake as
shallow, warm, and nutrient-rich as Vancouver Lake would
ordinarily be expected to exhibit extensive growth of at-
tached aquatic plants. Vancouver Lake as yet has none.
Three conditions of the lake may provide a plausible ex-
planation. The low transparency of the water during the sum-
mer may limit light penetration to such an extent that plants
on the bottom lack sufficient light for photosynthesis. An alter-
native, or perhaps additional, explanation may lie in the
repeated disturbance of the bottom sediments by wave ac-
tion and fish, which may make it impossible for attached
plants to become established. The third factor that could limit
macrophyte growth is the seasonal variation in depth of the
lake.
A severe problem currently limiting the recreational poten-
tial of Vancouver Lake is the low transparency of the water.
Secchi disc readings as low as 0.05 m have been recorded
during the summer (Fig. 12). The high algal population is
Contributed Papers
a major contributor to this problem, but not the only one.
The shallow water permits bottom sediments to be stirred
easily into the water column by even moderate winds. The
magnitude of this problem is indicated by the sharp rise in
suspended solids in November following 2 days of high wind
(Fig. 13). This reduced the Secchi disc transparency from
near 0.30 m to less than 0.10 m.
Another major influence in keeping sediments suspend-
ed in the lake is the large population of carp. In the shallower
parts of the lake it is easy to follow the movements of fish
by the "smoke trails" of suspended sediments they leave
behind.
SECCHI DEPTH
DAYS SINCE 7/1/81
Figure 12.—Secchi disc depth in Vancouver Lake, mean of stations
1, 2, and 3.
SUSPENDED SOLIDS
A « J J A
DAYS SINCE 7/1/ei
Rgure 13.—Total suspended solids (nonfilterable solids) in Van-
couver Lake; mean of stations 1, 2, and 3.
NUTRIENTS
Nutrient concentrations in Vancouver Lake are high (Table
1) with phosphorus higher than nitrogen in relation to algal
needs. The molar ratio of inorganic nitrogen to total phos-
phorus is approximately 3:1. The ratio of total nitrogen to
total phosphorus on a weight basis is approximately 9:1. Both
these ratios indicate that phosphorus is unlikely to be a
limiting nutrient in this system.
Both ortho and total phosphate are at their highest con-
centrations during the summer when water is shallow,
temperatures warm, algal numbers high and fish very ac-
tive (Rg. 14). Both Burnt Bridge Creek and the Columbia
River are high in phosphorus and could be sources for Van-
289
-------
Lake Restoration, Protection and Management
PHOSPHORUS
JASCNDJFMAW I J A S 0
Figure 14. — Phosphorus concentrations in Vancouver Lake; mean
of stations 1 , 2 and 3. Values are reported as mg P/l.
NITROGEN
J A S C N C
DAYS SINCE 7/1/S1
Figure 15.—Nitrogen concentration in Vancouver Lake; mean of sta-
tions 1, 2 and 3. Values are reported as mg N/l. "ORG" is total
organic nitrogen.
couver Lake. A detailed nutrient budget has not been
prepared for Vancouver Lake, but it appears that much of
the phosphorus load in the lake comes from internal cycl-
ing from the sediments. Bottom feeding fish cause this
cycling.
Organic nitrogen follows a pattern very similar to algal cell
density (Fig. 15). It is likely that most of the organic nitrogen
is found in the algal biomass.
Ammonia and nitrate nitrogen follow a somewhat different
annual pattern, being relatively low in the summer and in-
creasing in the winter months. The increases in ammonia
and nitrate occur when inflow from Burnt Bridge Creek is
most influential on the lake. It seems likely that the increases
in nitrate and perhaps ammonia can be traced to increased
input from Burnt Bridge Creek as flow increases. In
September 1982, a large increase in nitrate concentration
at Station 5 (Burnt Bridge Creek) was observed following high
flows resulting from heavy rain.
RESTORATION
Many plans have been proposed in the past to increase the
use of Vancouver Lake (U.S. Environ. Prot. Agency, 1978).
In the 1920's local farmers wanted the lake drained for
cropland use. The plan was never carried out, and in 1948
it was proposed that Vancouver Lake be dredged to a depth
that would permit the mothballing of Liberty Ships there. The
Port of Vancouver and the Vancouver Chamber of Com-
merce supported this plan which included minor recreational
development. In 1966 the Port of Vancouver prepared
another development plan which included construction of a
barge canal into the lake and barge-loading facilities within
the lake. Significant recreation facilities were included in this
plan.
In 1968 the Port of Vancouver proposed a new develop-
ment plan that emphasized recreation. About this time the
Port contracted with Washington State University to conduct
water quality studies of Vancouver Lake and to propose
methods for rehabilitation. The result of these studies (Baghat
and Funk, 1968; Baghat and Orsborn, 1971; Orsborn, 1972)
was a proposal to dredge the lake to remove nutrient-laden
sediments, provide flushing water from the Columbia River,
and to implement controls on the nutrient inputs to the Lake
from Burnt Bridge Creek.
After several modifications of scope and design, the three-
pronged approach recommended by Washington State
University was proposed in a grant application to the EPA
Clean Lakes program (Dames and Moore, 1977). This docu-
ment proposed dredging between 9 and 11 million cubic
meters of sediment from the lake, connecting the lake to the
Columbia River via a combination of open channel and
culverts to admit up to 17 cu m/s of river water, and im-
plementing nutrient runoff control measures in the Burnt
Bridge Creek basin.
After review by EPA and interested groups, and based
on information from the lake dilution project at Moses Lake,
Wash. (Welch, 1979), a final Operations Plan (Dames and
Moore, 1980) was prepared to guide the design and im-
plementation of the restoration project. This final plan
modified dredge disposal areas to preserve critical wildlife
habitat, reduced dredging to 6.5 million cubic meters to com-
ply with EPA requirements, and reduced the number of
culverts to provide a maximum of 9 cu m/s between 3 per-
cent and 5 percent flushing volume per day, depending on
lake and river levels. Tide gates in the channel prevent
backflow through it from Vancouver Lake to the Columbia
River, and sluice gates permit the channel to be closed en-
tirely as necessary to protect anadromous fish runs or to pre-
vent highly turbid flood waters from entering the lake.
The dredging plan, as finally implemented, is designed
to facilitate the flow of higher quality Columbia River water
to those areas of the lake most heavily used for recreation
and to provide a "short circuit" flow for Burnt Bridge Creek
water to Lake River (Fig. 16).
DREDGING &
DISPOSAL AREAS
Figure 16.—Dredging and disposal areas for the Vancouver Lake
Restoration Project.
290
-------
Contributed Papers
A project of this magnitude was possible because land
suitable for dredge material disposal was available near the
lake. Some land was owned by the Port, some was donated
by other large landowners in the area, and easements were
obtained from landowners for the remaining sites.
Water from the dredge is returned to the lake after set-
tling in the disposal areas and secondary settling ponds, with
silt barrier curtains used around the return flow outfall as
needed to maintain water quality. The contract specifies
water quality standards to be maintained in the return flows
and lake during construction; a water quality monitoring pro-
gram monitors the construction progress.
A major concern throughout the development of the Van-
couver Lake Restoration plan has been the effect that dredg-
ing and its resultant disturbance might have on the fish in
the lake, and the effect that the altered flow, resulting from
the flushing channel, might have on the migration of ana-
dromous fish in the Columbia River. To help answer these
questions, a fisheries monitoring plan was developed. This
plan is being carried out by Envirosphere Inc., of Seattle.
Table 3 presents preliminary results of their initial sampling
efforts.
Samples were taken by beach seine and gill net in April.
Since then samples have been taken by beach seine only. A
surprising result was the large number of juvenile salmon
in the lake in the spring. Most of the salmon were caught
along the steeper northeast shore.
In the early samples, yellow perch were concentrated in
the south of the lake near the entrance to Mulligan Slough,
the low-lying flooded ground to the south of Vancouver Lake.
This area is the spawning ground for many of the fish species
in the lake. In later samples, most of the yellow perch caught
in the seine were young of the current season.
Many carp are in the lake, enough to support a small com-
mercial fishery. Other catches of note include an 1.3 m
sturgeon pumped through the dredge, and several sub-legal
sturgeon (less than 1 m length) found in the possession of
the carp fishermen. Recently, shad have been caught in the
lake.
EFFECTS OF CONSTRUCTION ACTIVITY
Dredging started in the lake in late May 1982. As of Oct. 15,
1982, dredging was 35 percent complete. The Rushing Chan-
nel between the lake and the Columbia River was opened
following ceremonies Oct. 22, 1982. At this time it is not
possible to determine what long-term effects the restoration
will have on the lake.
Some effects noted thus far are increased recreation on
the west side of the lake along the county park, decreased
algal densities throughout the lake, and an apparent qual-
itative change in the relationship between turbidity and
suspended solids.
The most obvious effect of the dredging has been to make
the lake deeper. Although the change has been relatively
small in terms of possible limnological effects, dredging along
the west side of the lake has increased the summertime
water depth from approximately 0.5 m to nearly 2.0 m. This
has permitted increased recreation, and even though the pro-
ject is far from complete, area residents are already taking
advantage of the opportunities.
Summertime algal densities in 1982 did not reach the high
values of 1981. Suspended solids concentrations were also
lower in 1982 than in 1981. Turbidity and Secchi disc depths
were very close to the values of 1981. This contrast of lower
suspended solids with the same Secchi disc values stems
from a change in the nature of the suspended material. The
suspended material in 1982 was more finely divided. This
is apparently the result of fine material from the dredge return
flow remaining in suspension, or being resuspended in the
lake. This source of turbidity kept the water transparency
at levels similar to 1981 while greatly reducing algal cell
densities.
It is not known if the reduced algal counts result from light
limitation caused by the increased turbidity from dredging,
Table 3. — Fish species caught by beach seine in Vancouver Lake. Values are the total number
caught at all stations in the lake.
DATE
Species
Chinook salmon
Yellow perch
White crappie
Black crappie
Pumpkin seed
Peamouth
Carp
Northern squawfish
Prickly sculpin
Mountain whitefish
Largescale sucker
Largemouth bass
Brown bullhead
Other sculpin
Walleye
Stickleback (3 spine)
Bluegill
Goldfish
American shad
Warmouth
TOTAL
4-7-82
52
37
5
1
2
3
4
2
1
0
0
0
0
—
—
—
—
—
—
—
106
4-22-82
54
33
28
8
2
17
12
7
—
—
9
4
1
—
—
—
—
—
—
—
175
5-11-81
13
6
1
3
5
190
10
5
3
—
20
9
1
14
—
—
—
—
—
—
280
5-27-82
12
27*1'
14
7
3
9
5
?
17
—
5
4
1
—
1
5
—
—
—
—
113
6-8-82
8
187<1'
39
19
3
65
27
71
10
—
42
11
12
—
—
23
1
7
—
—
525
9-22-82
18(1>
227
520
5
6
70
—
4
—
1
59
—
—
—
—
24
—
1
1
936
1 Mostly young o( this season
291
-------
Lake Restoration, Protection and Management
or if the algae decreased because of other factors such as
cooler temperatures. In any event, water transparency has
not changed significantly during construction.
Changes in Vancouver Lake as a result of the Restora-
tion Program are anticipated to increase the value of the lake
as a recreational resource in the Portland-Vancouver met-
ropolitan area. Improvements will make the lake more usable
during the summer months because of increased depth from
the dredging, and improve the water quality by introducing
higher quality Columbia River water. The improvements
resulting from the addition of Columbia River water will result
primarily from increased flushing of the system rather than
dilution because the nutrient content of the lake and river
are similar.
Changes in the Burnt Bridge Creek watershed to reduce
the use of septic systems, increase sewers, and reduce other
sources of nonpoint pollution are expected to decrease the
influx of nutrients to the lake.
CONCLUSIONS
The Vancouver Lake restoration is a major structural modi-
fication of a lake aimed at improving its recreational poten-
tial. Construction is well underway and progressing well.
Water quality sampling indicates that construction has af-
fected the lake only in minor ways. Noticeable effects from
the restoration effort have not yet been detected. The in-
creased depth in the areas dredged to date has increased
recreational use of the lake.
REFERENCES
Baghat, S.K., and W.H, Funk. 1968. Hydroclimatto Studies of Van-
couver Lake. College of Engineering. Bull, 301. Washington State
Univ, Tech. Extens. Div., Pullman.
Baghat, S.K., and J.F, Orsborn. 1971, Water Quantity and Quality
Studies of Vancouver Lake, Washington, College of Engineering.
Washington State University, Pullman.
Dames and Moore. 1977, Master Plan: Rehabilitation of Vancouver
Lake, Vancouver, Wash. Reg. Plann. Counc. Clark County.
1980. Operations Plan: Rehabilitation of Vancouver Lake.
Port of Vancouver.
Orsbom, J.F., 1972, Correlated Studies of Vancouver Late: Hydraulic
Model Study. EPA-R2-72-078. U.S. Environ. Prot. Agency,
Washington, D.C.
Portland General Electric Co. 1981. Operational Ecological Monitoring
Program for the Trojan Nuclear Plant. Annu. rep. 1980.
PGE-100&60.
Strickland, J., and T. Parsons, 1972. A Practical Handbook of Sea-
water Analysis. 2nd ed. Bull. Fish. Res. Board Can. 167,
Sweet, J.W. 1982. Consultant, Aquatic Analysts. Portland, Ore. Pets.
cornm.
U.S. Environmental Protection Agency. 1978. Final Environmental
Impact Statement for Vancouver Lake Reclamation Study.
EPA-10-WA-CLARK-POV-CL-77, Seattle, Wash,
Welch, E.B. 1979. Lake restoration by dilution. In Lake Restoration.
EPA 440/5-79-001, U.S. Environ, Prot. Agency, Washington, D..C.
292
-------
AN EVALUATION OF 15 LAKES IN KING COUNTY WITH
PROJECTIONS OF FUTURE QUALITY
JOANNE DAVIS
ROBERT G. SWARTZ
Municipality of Metropolitan Seattle
Water Quality Division
Seattle, Washington
ABSTRACT
Fifteen lakes in the Metro planning area were sampled every other month from May 1979 through March
1980. One purpose of the sampling was to assess changes caused by nonpoint sources of pollution since
some of the lakes had last been evaluated at some time between 1972 and 1976. A major study purpose
was to project, where feasible, future quality by simple modeling methods, based on a range of potential
increases in phosphorus loading resulting from possible increases in urbanization. Three of the lakes were
found to be oligo-mesotrophic, five mesotrophic, two meso-eutrophic, and three eutrophic. Two were not
classified owing to insufficient or inconclusive data. Four seemed to have changed since previously surveyed,
while six apparently had not. None of the lakes was fully resistant to effects of future development in its
basin; but it appears that four of the lakes would show little effect from full-basin development at a low
density (two to five units/acre). Four other lakes would probably be adversely affected by such develop-
ment. These kinds of projections were not possible for seven of the lakes. Some recommendations for
management of the lakes were made.
INTRODUCTION
Background
Metro began studying small lakes in the Cedar and Green
River basins in the early 1970's. As part of its responsibility
as the water quality planning agency for its jurisdictional area
under the then new Federal Clean Water Act, over 30 lakes
were initially surveyed. That was followed by more intensive
surveys, concluded in 1976, of most of those lakes. All of
these surveys were documented in reports by Whitmore et
al. (1973), Uchida et al. (1976), and Brenner et al. (1977).
In those and other Metro studies, the agency saw the ef-
fects of nonpoint pollution and concluded that here—as in
many other geographical areas—increasing urbanization is
the major thteat to the quality of recreational waters. Increas-
ing urbanization generally implies accelerated eutrophication
(increased nutrient concentrations and algal growth) for lakes.
As human activities in a basin increase, natural vegetative
cover is destroyed, soils are disturbed, impervious surfaces
are installed, and nutrient-rich materials such as fertilizers
and sewage are introduced. Increasing proportions of
precipitation tend to run off rather than infiltrate soils, carry-
ing along nutrients that support increasing growth of algae
in downstream receiving waters.
Most problems seen in the local lakes began with in-
creased availability of nutrients for algal growth. Increased
algal growth can lead to poor transparency, scums on the
water surface, unpleasant odors and loss of dissolved ox-
ygen in bottom waters. A slow process of eutrophication—
over hundreds or thousands of years—is natural to the life
cycle of a lake, and ends when the lake has become com-
pletely filled in and ceases to exist. Human activities tend
to greatly speed up the process, thus shortening the life of
the lake. This acceleration of the eutrophication process may
or may not be apparent to individual users of a lake. On a
scale comparable to the human life span, even the ac-
celerated changes in a lake may be slow enough to escape
recognition for a time unless objective records are kept and
consulted. However, if activities in a lake basin are man-
aged to minimize transfer of material into the lake, the useful
life of the lake can be extended for the benefit of current
and future usen
Purpose of the Study
This study proposed to look primarily at lakes that had last
been evaluated at some time between 1972 and 1976, to
further assess the effects of nonpoint pollution, and to see
whether recent deterioration or improvement was apparent.
Several lakes outside the Green or Cedar River basins were
also included — North, Geneva, Fivemile and Green. These
had not previously been surveyed by Metro, but lie within
King County.
The major objectives were to characterize each lake
relative to its water quality, and to show the potential for
future problems.
Criteria for Selection of Study Lakes
Metro's criterion for the selection of small lakes in the original
study (Whitmore et al. 1973) was that the lake be at least
8 ha (20 acres) in area, the State's criterion for inclusion
under the Shorelines Management Act. That criterion was
maintained in this study and another requirement added —
that of public access, defined as a State Game Department
boat launch or a lakeshore public park. This resulted in a
list of about 30 lakes. This report deals with half of the lakes,
studied from May of 1979 to May of 1980.
Description of the Study Area
Land Use
The lakes included in the study and their locations are in-
dicated on Figure 1. Two of the lakes (Union and Green)
were in urban settings, while the rest were in suburban to
rural areas. All were in the Puget Sound lowlands, most in
293
-------
Lake Restoration, Protection and Management
Figure 1,—Location map,
the Green and Cedar River basins within King County. Land
use in the basins typically included single-family residential,
forest, open space, and some nonintensive pasture. The
Green and Union Lake basins included considerable com-
mercial and some multi-family residential areas.
Soils
Alderwood soils predominated over much of the area
(Snyder, 1974) with some Everett soils and smaller amounts
of other types. In general, Alderwood soils are underlain at
shallow depth by relatively impermeable substrata When us-
ed for septic tank drainftelds this can cause surfacing of ef-
fluents, particularly in wet weather. Everett soils, on the other
hand, have very high permeability which can induce
downstream pollution because the soil does not completely
retain pollutants.
Climate
Climate in the area is maritime (Phillips, 1968) with prevail-
ing winds from the southwest, west, and northwest. Annual
precipitation in most of the study area ranges from 87 cm
(35 in.) to 100 cm (40 in.) with about 75 percent falling from
October through March. Summer high temperatures are
typically in the 70's (F) and winter minimums range from the
mid-20's to mid-30's. Snowfall is infrequent. Rainfall is usually
light to moderate over extended periods rather than falling
in intensive episodes.
Description of Study Lakes
The study lakes, with three exceptions, were small lakes
in suburban or semi-rural to rural settings (Table 1). Only
one of these small lakes exceeded 0.4 km2 (100 acres)
in surface area; the smallest was 0.08 km2 (20 acres).
Maximum depths ranged from less than 6 m (20 feet) to
nearly 20m (66 feet) and mean depths were from 2.1 to
8.2 m (7 feet to 27 feet). Drainage basin size was generally
small; only four of the small lakes had basins more than
10 times the area of the lake itself. Channelized inflows
and outflows were for the most part intermittent or absent.
Three of the small lakes had King County parks with
bathing and picnicking facilities, and all but one had State
Game Department boat launch facilities. Four to 75 per-
cent of the individual lake basins were devoted to residen-
tial use — averaging 26 percent. Four of the small subur-
ban lake basins were sewered; the rest were served by
septic tanks.
METHODS
Sample Collection and Analysis
Samples were collected six times — in May, July, Sep-
tember, and November 1979 and in January and March
1980. Generally, three lakes were sampled on a particular
day; the first at 8 or 9 a.m. and the last at 11 a.m. or 12
noon. Each lake was characterized by one station (except
Sawyer, which had two), located over the deepest part of
the lake. Samples were taken at 1 meter below the sur-
face, 1 meter above the bottom, and at either one or two
intermediate depths, depending on the maximum depth
of the lake. An exception was Lake Sammamish where
samples were collected at depths of 1,5,10,15, 20, and
25m.
Characteristics sampled in the water column were
temperature, pH, conductivity, dissolved oxygen (D. O.),
total phosphorus, orthophosphate-phosphorus, chloro-
phyll a, phytoplankton, fecal conform, and Secchi disk
transparency.
With a few exceptions, analyses were completed on the
day of sample collection. Chemical parameters were
analyzed according to Standard Methods. Chlorophyll a
analysis was done in duplicate using a tissue grinder,
acetone extraction in the dark, and centrifugation. Absorb-
ance was read before and after acidification on a Beekman
26 spectrophometer. Phytoplankton counts and identifica-
tions to genus were done with a Zeiss inverted microscope
with Whipple grid. Fecal coliform analysis was done in
triplicate using membrane filter techniques.
Data Analysis
Data used in calculating apparent phosphorus loading
rates are shown in Table 2. The method used was to infer
loading based on measured lake concentration (volume
weighted annual mean) and estimated flushing rate. To
estimate the flushing rates it was first necessary to
estimate the outflows by determining average annual
runoff from a map of that parameter (Gladwell and Mueller,
1967) and multiplying by the basin area. Outflow was then
divided by lake volume to arrive at flushing rate.
The model use'd to relate loading to lake concentration
and flushing was a variant of a basic relationship stated
by Vollenweider (1969): P = M1~R) (Dillon and Riqler,
1974, 1975). ze a
Submodels used in estimating R were: R = L—
(Larsen and Mercier, 1976) and R = 0201e-°-042!qs
+ 0.574e-°-009499s (Ostrofsky, 1978). Loadings are
presented as ranges in the report. Use of the Larsen and
Mercier R model produced the lower number in the range
and the Ostrofsky R produced the higher. Use of the two
R estimates reflects the high uncertainty regarding sedimen-
tation rates which are very difficult to measure or estimate.
294
-------
After a current loading rate was estimated, rates for
hypothetical levels of future basin development were add-
ed on, to estimate the lake's future status. Phosphorus ex-
port rates for the future development were obtained from
Metro's "desk top" urban runoff model (Buffo, 1979). For
purposes of these very generalized projections, the individual
lake basins were assumed to be homogeneous in nature,
and the hypothetical developments were not evaluated
relative to any specific features, either natural or cultural,
within a basin. That, of course, should be done if these
methods were used to evaluate potential effects of a specific
development.
While this modeling approach, in its present state of de-
velopment, does not provide probabilities or statistical con-
Contributed Papers
fidence limits for the results, it still can be useful. What it
does provide is a rational and systematic quantification of
ideas that have usually been stated qualitatively or intuitive-
ly, and it gives some idea of relative risk in various potential
development scenarios. More detailed discussion of these
models and their application in similar studies can be found
in Metro's Early Warning System, Section 2 (Davis, 1980).
RESULTS AND DISCUSSION
Current inferred loading estimates for each lake as well as
loading projections based on hypothetical land use changes
were placed on Vollenweider-type plots (Rg. 2-4) to visualize
current and projected trophic status. Summary information
for several example lakes follows:
Table 1. — Physical and cultural data for lakes studied In 1979 -19801
Lake
Angle
Desire
Dolloff
Fivemile
Geneva
Green
Kathleen
Morton
North
Pipe
Sammamish
Sawyer
Star
Steel
Union
'. Most of the data
Basin area
.8 sq mi
2.07 km2
1.37 sq mi
3.55 km2
.81 sq mi
2.10 km2
1.01 sq mi
2.62 km2
.35 sq mi
.91 km2
.49 sq mi
1.27 km2
.40 sq mi
1.04 km2
.76 sq mi
1.97 km2
.49 sq mi
1.27 km2
97.7 sq mi
253.0 km2
13.0 sq mi
33.7 km2
.59 sq mi
1.53 km2
.38 sq mi
.98 km2
Immediate?
Total-600 sq mi
1,554.0 km2
are from Bortteson et al. (1976).
Land use
Lake - 20%
Residential - 75%
Forest or unproductive - 5%
Lake - 8%
Residential - 10%
Forest or unproductive - 82%
Lake - 4%
Residential - 13%
Forest, unproductive & ag - 83%
Lake - 6%
Residential - 6%
Forest, unproductive & ag - 88%
Lake - 12%
Residential - 42%
Forest, unproductive & ag - 46%
Lake - 16%
Residential - 14%
Forest, unproductive & ag - 70%
Lake - 27%
Residential - 16%
Forest, unproductive & ag - 57%
Lake - 12%
Residential - 12%
Forest, unproductive & ag - 76%
Lake - 18%
Residential 14%
Forest, unproductive & ag - 68%
Lake - 4%
Residential - 4%
Forest, unproductive & ag - 92%
Lake - 9%
Residential - 41%
Forest, unproductive & ag - 50%
Lake - 19%
Residential - 66%
Forest or unproductive - 15%
Industrial, commercial
Lake area
100 acres
.40 x I06m2
71 acres
.29 x I06m2
20 acres
.08xi06m2
38 acres
.15xl06m2
26 acres
.11xio6m2
255 acres
1.03xl06m2
51 acres
.21 x I06m2
68 acres
.28xi06m2
56 acres
.23xi06m2
55 acres
.22x10^2
4,897 acres
19.8xi06m2
300 acres
1.2xi06m2
35 acres
.14xi06m2
46 acres
.igxi^m2
580 acres ex-
cl. port, bay
2.35xi06m2
Mean
depth
25ft
7.6m
13ft
4.0 m
10ft
3.1 m
18ft
5.5 m
19 ft
5.8 m
12.5ft
3.8 m
7ft
2.1 m
15ft
4.6m
14 ft
4.3m
27 ft
8.2 m
58ft
17.7 m
26ft
7.9 m
25ft
7.6 m
13ft
4.0 m
34ft
10.5 m
Inflow
No
Inter-
mittent
None
apparent
No
Inter-
mittent
Springs,
storm
drains
Inter-
mittent
None
apparent
Inter-
mittent
None
apparent
Issaquah,
Tibbetts
Creeks
Peren-
nial
None
apparent
Inter-
mittent
Outflow
from Lk.
Wash.
Outlfow
No
No
Channel
visible
No
Yes
Yes
Yes
Yes
Yes
Yes
Sammam-
ish
River
Yes
Yes
No
Fremont
Cut-Ship
Cnl to PS
295
-------
Lake Restoration, Protection and Management
Table 2. — Data used in calculation of phosphorus loading values for lakes studied May 1979-March 1980
LAKE
Angle
Desire
Dolloff
FIvemile
Geneva
Green
Morton
North
Pipe
Sammamish
Kathleen
Sawyer
Star
Steel
Union
Lk Area
I0«m2 (km*)
1
.405
.29
.08
.15
.11
.28
.23
.22
19.8
.21
1.21
.14
.19
2.35
Mean
depth m
2
7.6
4.0
3.0
5.5
5.8
4.6
4.3
8.2
17.7
2.1
7.9
7.6
4.0
10.4
Volume
lOW
3
3.21
1.15
.25
.86
.63
1.22
.95
1,85
350
.47
9.50
1.07
.74
24.7
Basin
area kmz
4
2.07
3.55
2.1
2.62
.91
1.04
1.97
1.27
253
1.27
33.7
1.53
.98
1544
Average
annual
runoff m
5
.48
.50
.58
.62
.62
.55
.60
.50
.52
.50
.58
.58
Outflow
108ffl'
6
.99
1.78
1.22
1.62
.61
.57
1.18
.64
.66
16.8
.89
.57
Flushing
rate Yr1
7
.31
1.6
4.9
1.9
.97
.47
1.24
.35
1.40
1.8
.84
.77
Areal water
loading
m/Yr
8
2.4
6.1
15.2
10.8
5.5
2.04
5.13
2.91
3.41
13.9
6.36
3.0
Mean P
cone.
glm*
9
15
33
52
21
30
30
15
18
17
12
21
20
26
14
R value
10
.64, .81
.44, .72
.31, .60
.50, .70
.59,.75
.29, .53
,63, .74
.43,,61
.54, .69
.53, .73
Annual P
loading
gnT^r
11
,10-
.40-
1.1 -
.40-
.34-
.08-
.18-
.13-
,50-
.36-
.09-
.20
.72
1.9
.66
.56
.13
.33
.19
.72
.54
.16
Cumnt Condftkxia
1,0 — lew dwuiiy
HD-M»»i
-------
waters, particularly where large developments are involved,
and on smaller scales for smaller operations.
While such construction-related impairments to a lake may
largely dissipate with time, the process may require months
or years. Furthermore, some residual effects that do not dis-
sipate, but bring about a permanent increase in the lake's
productivity, can be expected.
A second aspect of urbanization not directly addressed
in the individual lake analyses is potential proliferation of sep-
tic tank systems for wastewater disposal. Studies by Metro
and others (Harper-Owes, 1981; Jones and Lee, 1979;
Sikora and Corey, 1967) indicate that where septic tanks and
drainfields are functioning properly, algal nutrients or bacteria
are not transmitted to lakes. However, when systems fail,
surfacing effluents can enter nearby streams and lakes, in-
creasing the phosphorus content for algal production and
also presenting potential health hazards.
Much of the soil in the study area is the Alderwood type
which has serious limitations for drainfield use, as briefly
described in the soils section of the Introduction to this report.
Since the area generally could be considered high risk for
septic systems because of inadequate soils, regular inspec-
tions and correction of any failures are recommended.
SUMMARY AND CONCLUSIONS
• The study lakes were mostly in the mesotrophic range,
although the complete range observed was from oligo-
mesotrophic to eutrophic.
• Of the lakes that had previous surveys for comparison,
four (Pipe, Star, Sammamish, and Kathleen) have undergone
possible changes in quality, while six did not seem to sub-
stantially change. Lakes Sammamish and Kathleen had
decreased levels of one or more trophic state indicators, im-
plying an improved condition. Pipe and Star lakes had in-
creased levels, implying a deterioration in quality.
Contributed Papers
• All of the study lakes have limits to their tolerance of
increased urbanization, but sensitivity varies from lake to
lake. Some may be able to tolerate full basin development
^M Current ConAOon
— Protections
Figure 3.—Current trophic status of four lakes (Angle, Dolloff,
Geneva, and Steel) and changes in status projected to result from
hypothetical basin development.
Table 3. — Summary of key data and conclusion for King County lakes studied May 1979-March 1980.
Mean
transparency
Lake m
Angle
Desire
Dolloff
Fivemile
Geneva
Green
Kathleen
Morton
North
Pipe
Sammamish
Sawyer
Star
Steel
Union
5.4
2.2
1.6
1.5
3.1
2.4
2.3
2.9
2.1
4.1
5.3
3.8
4.2
4.5
Mean
chlorophyll
^9/l
1.2
6.3
10.7
2.1
5.5
9.6
2.4
4.8
9.6
1.8
3.0
3.6
5.1
2.5
Mean Annual
phosphorus phosphorus
concentration loading
ng/l gm~2
15
34
52
21
30
29
15
18
17
12
20
26
14
.10-
.40-
1.1 -
.40-
34 -
—
—
.08-
.18-
.13-
—
.50-
.36-
.09-
20
.72
1.9
.66
.56
.13
.33
.19
.72
.54
.16
Trophic
state8
O-M
M-E
E
M
M-E
E
E
M
?
O-M
M
M
M
O-M
Change
from
previous
condition
No
No
No
—
—
—
Yes
No
—
Yes
Yes
No
Yes
No
Existing
wastewater
mgmt.b
1
2,3
1,2
2
2
1
2
2
2
1
1,2
2,3
1
1
4
Projected
response to
land usec
2
6
2
5
2
6
5
6
3
3
6
3
3
2
5
Current
land use
(°/o undeveloped)
5
82
83
88
46
70
57
76
68
92
50
15
1 Odigolrophic; M=mesotrophic; E=eutrophic
k- 1=sewered; no documented sewerage problems; 2=septic systems used, present & potential effects unknown; 3=septic systems with
documented problems: 4=sewered, with some combined sewer overflows.
c' 1-no problem with increased development density; 2=lrttle quality change with full-basin development to low (2-5/acre) density, but
definite limitations needed on high density development; ^limitations needed on both tow and high density development; 4=any increase
in development could be hazardous to lake; 5 - lake or data not amenable to the type of analysis used to make projections; 6 = con-
clusions withheld pending further data collection.
297
-------
Lake Restoration, Protection and Management
Eallmated Annual area! water loading (outflow/lake araa). m*1 yr.
Rgure 4.—Current trophic status of lakes Desire and Rvemile. Rela-
tionship after Vollenweider.
at low density, but none can tolerate full develooment at high
density.
• The impacts projected for each lake from increasing
urbanization are based on effects expected from stable
post-construction development. Short-term augmentation of
these impacts can be expected during construction, to a
degree that can be disastrous if construction is not careful-
ly planned and executed.
• Long-term augmentation of projected impacts from ur-
banization can be expected from septic tank failures in future
developments where septic tanks are used for wastewater
treatment.
SPECIFIC RECOMMENDATIONS
The study report, based on collection and evaluation of data,
was not intended to be a management plan. However, Metro
staff was able to recommend specific actions to agencies
or jurisdictions concerned with addressing the issues
discussed.
• The specific tolerances of the individual lakes to effects
of urbanization should be taken into account in the land use
planning and construction permitting processes.
• Strict enforcement of regulations that apply to plan-
ning and execution of construction needs a high priority.
Developers may need assistance to develop appropriate and
realistic erosion control plans, and frequent inspection of ac-
tual application is recommended.
• A comprehensive program to ensure proper
maintenance of septic tank systems should be undertaken.
• Further investigation should occur to clarify reasons for
conditions or changes seen in several of the lakes.
REFERENCES
Bortleson, G.C. et al. 1976. Reconnaissance data on lakes in Wash-
ington. II. Water Supply Bull. 43, 2. State of Washington, Depart-
ment of Ecology. Olympia.
Brenner, R.N. et al. 1977. An intensive water quality survey of eight
selected lakes in the Lake Washington and Green River drainage
basins. Munic. Metro. Seattle.
Buffo, J. 1979. Water Pollution Control early warning system, sec-
tion 1. Non-point source loading estimates. Munic. Metro. Seat-
tle rep.
Davis, J. 1980. Water Pollution Control early warning system, sec-
tion 2. Assessing current and future status of lakes. Munic. Metro.
Seattle rep.
Dillon, P.J., and F.H. Rigler. 1974. A test of a simple nutrient budget
model predicting the phosphorus concentration in lake water. J.
Fish. Res. Board Can. 31: 1771-1778.
. 1975. A simple method for predicting the capacity of a lake
for development based on lake trophic status. J. Fish. Res. Board
Can. 32: 1519-1531.
Gladwell, J.S., and A.C. Mueller. 1967. Water resources atlas of the
state of Washington. Vol. 2, Part B of An initial study of the water
resources of the State of Washington. State of Washington Res.
Center, Pullman.
Harper-Owes. 1981. Pine Lake restoration analysis: final report.
Munic. Metro. Seattle.
Jones, A.R., and G.F. Lee. 1979. Septic tank wastewater disposal
sys-
tems as phosphorus sources for surface waters. J. Water Pollut.
Control Fed. 51(11): 2764-2775.
Larsen, D.P., and H.T. Mercier. 1976. Phosphorus retention capacity
of lakes. J. Fish. Res. Board Can. 33: 1742-1750.
Ostrofsky, M.L. 1978. Modification of phosphorus retention models
for use with lakes with low areal water loading. J. Rsh. Res. Board
Can. 35: 1532-1536.
Phillips, E.L. 1968. Washington climate. Coop. Exten. Serv., College
of Agriculture, Washington State University. Pullman.
Sikora, L.J., and R.B. Corey. 1976. Fate of nitrogen and phosphorus
in soils under septic tank waste disposal fields. Trans. ASAE. 19(5):
866-875.
Snyder, D.E. et al. 1974. Soil survey, King County area, Washington.
U.S. Dep. Agric., Soil Conserv. Serv. in cooperation with
Washington Agric. Exper. Sta.
Uchida, B.K. et al. 1976. An intensive water quality survey of 16 se-
lected lakes in the Lake Washington and Green River drainage
basins. Munic. Metro. Seattle.
Vollenweider, R.A. 1969. Moglichkeiten and Grenzen elementarer
Modelle der Stoffbilanz von Seen. Arch. Hydrobiol. 65: 1-36.
Whitmore, C.M. et al. 1973. Quality of small lakes and streams in
the Lake Washington and Green River drainage basins. Munic.
Metro. Seattle.
298
-------
LAKE HICKS RESTORATION ANALYSIS
D. R. HEINLE
Environmental Scientist
H. A. ARNOWITZ
Water Resources Engineer
CH2M HILL
Bellevue, Washington
ABSTRACT
Lake Hicks, a small urban lake near Seattle, Wash., was closed to swimming in 1975 because of poor
water clarity and high coliform bacteria levels. The closure prompted the local parks and recreation depart-
ment to undertake a year-long lake restoration analysis in 1981. To determine the lake's water quality
conditions and the composition of its influents, a monitoring program was established, sanitary sewer lines
and storm drainage systems were investigated, and local waterfowl were observed. The results of this
analysis indicated that the fecal coliform pollution was related directly to urban runoff rather than to any
specific point sources. Eighty percent of the runoff originated in an area north of Lake Hicks, which com-
prises commercial and residential development and wetland areas. The poor water clarity was attributed
to direct runoff and the infiltration of waterborne humic substances through a corridor of gravelly sandy
loam. Nearby peat soils were the likely source of the humic material. A number of both nonconstruction
and construction restoration alternatives were evaluated; three were selected. (1) By creating a public
awareness program, the public can become knowledgeable about the problem and help to reduce con-
tamination within the drainage basin. (2) To precipitate the humic substances in the lake, a slurry of hydrated
lime and water will be periodically added to the lake. (3) To prevent runoff from entering the lake during
the summer, a rock berm will be constructed across the upstream end of the main inlet channel of the
lake to facilitate infiltration of the waters into gravelly, sandy loam. Excess runoff not absorbed by the soil
will be diverted to the existing pump station that drains the lake.
INTRODUCTION
Lake Hicks is a small lake located in urban King County near
Seattle, Wash. The lake is supplied with water through
springs in its bed, from surface runoff from the 273-hectare
(677-acre) drainage basin, and directly from rainfall (Fig. 1).
It is part of the 6.5-ha (16-acre) Lakewood Park and both
lake and park are important local recreational resources;
however, swimming has been prohibited in the lake since
1975 because of high fecal coliform counts and poor water
clarity.
By documenting the existing water quality conditions and
the composition of the lake's major influents for a year, the
causes of the water quality problems were identified. Solu-
tions were then developed to alleviate or reverse deteriorating
water quality. The study was conducted in accordance with
Environmental Protection Agency and Department of Energy
guidelines. The study used, in part, results of a previous
preliminary water quality investigation of the lake (CH2M
HILL, 1978) and information about storm drainage im-
provements within the area (CH2M HILL, 1979, 1980).
PHYSICAL ENVIRONMENT OF THE
STUDY AREA
Drainage Basin Description
Lake Hicks covers approximately 1.8 ha (4.5 acres), has a
maximum depth of about 6 meters (20 feet), and a volume
of approximately 49,340 m (40 acre-feet). The average depth
is 2.9 m (9.44 feet). Lake Hicks' drainage basin is exten-
sively developed and largely urban in character. The lake
receives stormwater from a 273-ha (677-acre) drainage basin
comprising vacant land, wetlands, low income housing
developments, parks, single family housing, low- and high-
Rgure 1.—Location map.
299
-------
Lake Restoration, Protection and Management
rise multifamily housing, and a commercial business district.
Immediately adjacent to the lake are single-family homes,
a health center complex, and junior and senior high schools.
Other than storm drains, no joint discharge sources drain
into Lake Hicks.
Precipitation
Mean annual precipitation in the drainage basin is about 87.6
centimeters (34.5 inches). The rainy season extends from
October through March, with November and December the
peak rainfall months.
Geology and Soils
Lake Hicks was formed approximately 14,000 years ago dur-
ing the recession of the Vashon Glacier. Since then, much
of the shallow area surrounding the lake and connecting
drainages has become filled. This natural aging and reduc-
tion in lake area has been accelerated by wetland filling for
development and increased erosion and deposition of
sediments from uplands.
The lands surrounding Lake Hicks are gently rolling, with
average slopes of 6 to 15 percent and a regional relief of
15 m (50 feet). Most of the Lake Hicks drainage basin con-
sists of Alderwood gravelly sandy soils. These soils have
developed from Vashon till, a common substrate of the Puget
lowlands. Alderwood soils are characterized by an imper-
vious, cemented layer of hardpan at a depth of about 71 to
91 cm (28 to 36 inches).
There are two important exceptions to these general soil '
characteristics. First, a corridor of Everett gravelly sandy loam
extends southwest from Lake Hicks; another corridor extends
to the northwest. This soil is well drained, having developed
from recessional sand and gravels. It may provide an un-
derground inlet for water to Lake Hicks from the northwest,
as well as providing an outlet from the lake to the southwest.
Second, the area to the northwest is low, marshy, and char-
acterized by peat soils of 1.5 to 4.6 m (5 to 15 feet) in depth.
Lake Level Control
Lake Hicks receives all the runoff from the 273-ha drainage
basin. At normal water levels (approximate elevation 104 m
(343 feet)), Lake Hicks has a surface area of approximately
1.8 ha (4.5 acres). During times of high water this area may
expand to 5.9 ha (14.5 acres), inundating substantial areas
of parkland.
Lake Hicks is drained by a pump with a normal capacity
of 0.06 m3/s (2 cfs). The estimated peak lake level during
a 20-year storm is 108.5 m (356 feet) (assuming that the
pump ran continuously for the entire storm). At high flood
elevation, flows of 0.01 to 0.02 m3/s (0.4 to 0.8 cfs) leave
the lake as ground water through the permeable Everett soils
along the south side of the lake. Ground water surfaces
through springs at the head of nearby Salmon Creek
Canyon.
INVESTIGATION AND STUDY METHODS
To determine the lake's water quality conditions and the com-
position of its influents, a number of investigations were con-
ducted. A monitoring program was established, sanitary
sewer lines and storm drainage systems were investigated,
and local waterfowl were observed.
Water Quality Monitoring
Thirteen monitoring stations were -established. They were
located in the middle of the lake, a nearby detention pond
that is connected to the lake by a channel, six inlet pipes,
and five locations outside the lake but within the drainage
basin. The lake station, the detention pond station, and three
of the external locations were sampled regularly when flow-
ing, while the other stations were sampled during two storms,
when flowing.
The station in the middle of the lake and the station at
the detention pond were monitored with a Hydrolab Model
6 water quality sampler twice monthly during summer
months and monthly throughout the year-long study (March
1981 through February 1982). Parameters included
temperature, pH, dissolved oxygen, and conductivity. Tur-
bidity was measured within 24 hours of sample collection
on a Turner Designs Model 40 nephelometer.
Also for these stations, nutrients, suspended solids, and
chlorophyll a were analyzed in a laboratory using standard
techniques. Occasional Winkler titrations of dissolved oxygen
were also done to check the accuracy of the Hydrolab water
quality sampler. Duplicate analyses were done on three to
five parameters per date as a quality control check.
Concentrations of total and fecal coliforms and fecal strep-
tococci were measured by the laboratory. Comparison of the
membrane filtration and most probable number (MPN)
techniques during early monitoring trips indicated the MPN
method was the most appropriate for the remainder of the
study (from May on).
At the other monitoring stations, samples were taken and
nutrients, suspended solids, fecal coliforms, and fecal strep-
tococci were measured by the laboratory.
Phytoplankton Community Examination
Composite whole-water surface samples were taken, and two
replicate 25-ml subsamples were settled from each sample
and the first 200 organisms were identified to genera. The
percent composition of these organisms was determined and
averaged for the two replicates, resulting in relative species
composition during the study period.
Storm Event Field Investigation
Characteristics of runoff water to Lake Hicks from the
drainage basin were monitored during two storm events,
once in the summer and once in the winter. Stations were
chosen to obtain sub-basin characteristics as well as waters
with direct access to Lake Hicks. Three samples were taken
periodically during each storm at each site.
Water quality parameters included turbidity, nutrients,
suspended solids, and total fecal coliforms and fecal strep-
tococci. Techniques used to measure these parameters were
the same used in the previously described regular water
quality monitoring.
The limited stormwater sampling was to determine if any
sub-basins were contributing disproportionate amounts of
bacteria and nutrients to the lake and to obtain approximate
field verification of phosphorus loading rates calculated from
a loading model.
Sanitary Sewer Line Investigations
Sanitary trunk sewers adjacent to the lake were suspected
of being sources of human pollution because of their prox-
imity, depth of burial, and heavy groundwater infiltration
characteristics. The assumption was made that exfiltration
could also take place given the proper hydraulic gradient.
Record searches of individual house hookups and associated
dye studies were conducted. Bore holes were dug on the
west side of the lake between the existing 61-cm (24-inch)
diameter sanitary trunk sewer and the shoreline to determine
if exfiltration was taking place. Other investigations includ-
ed locating the drain lines from the sanitary facilities and the
diatomaceous filter backwater from a swimming pool located
to the southeast of the lake.
300
-------
Contributed Papers
Storm Drainage Investigation
A similar effort was carried out to isolate storm sewer in-
lets to the lake that might be sources of pollution. Samples
were taken from manholes and free-outlet storm sewers
during storm and nonstorm events. Other investigations
included possible illegal hookups to storm sewers and
documentation of domestic animals near open water-
courses.
Waterfowl Investigation
Waterfowl were observed and counted during routine
monitoring, storm events, and other visits to the lake. Total
number and location within the lake or shoreline were
recorded.
RESULTS
Temperature
Weak thermal stratification was present in Lake Hicks at
the onset of sampling (March 2, 1981); surface temper-
atures were 9°C and bottom temperatures 4°C. Stratifica-
tion was present until early October, with evidence for
some mixing between the hypolimnion and surface waters.
The thermocline was generally between 3 and 4 meters
deep, but there was often a gradual thermocline over
depths from 2 to 4 meters.
One dye study indicated that the thermocline in Lake
Hicks is modified by the inflow of spring water between
the hypolimnion and epilimnion. A maximum temperature
of 28°C was observed at the surface in early August. The
maximum temperature in bottom waters was 15°C in late
August. The minimum temperature of 5°C occurred in late
January when the lake was well mixed.
Dissolved Oxygen
Concentrations of dissolved oxygen were 12 to 13 mg/l
at all depths on the first sampling date. Shortly after the
onset of thermal stratification, concentrations of dis-
solved oxygen became low (under 2 mg/l) in the hypolim-
nion. That condition persisted until vertical mixing occur-
red in late September through early October. The odor of
hydrogen sulfide was often apparent in water samples from
the hypolimnion, indicating the presence of chemically
reducing conditions.
Concentrations of dissolved oxygen in surface waters
remained at or near saturation throughout the year, ex-
cept following fall mixing when concentrations throughout
the lake were below saturation (e.g. 8 mg/l) in November
at a temperature of 11°C. The reduction in dissolved ox-
ygen following fall mixing probably results from the ac-
cumulated chemical and biological oxygen demand of the
bottom waters.
Turbidity, Suspended Solids, and
Transparency
Turbidity generally ranged between 1 and 5 NTU in sur-
face waters and at mid-depth during summer months,
while turbidity of bottom waters ranged from 4 to 20 NTU.
After fall mixing, turbidity was similar at all depths and in-
creased from 2 NTU during low-inflow periods to 7 to 10
NTU during high-inflow periods (January).
Suspended solids in surface waters ranged from less
than 1 to 6 mg/l during the summer months and from 2
to 13 mg/l during the fall and winter months. Suspended
solids at mid-depth were similar to surface values.
Suspended solids, like turbidity, were higher in bottom
waters ranging up to 22 mg/l, a common feature of lakes
in the Seattle area (Uchida et al. 1976). Highest concentra-
tions of suspended solids in bottom waters occurred in March
and July.
As a result of an oversight, Secchi transparency (clarity),
was measured only in March and from late September
through February. That period, however, included represen-
tative ranges of chlorophyll concentrations, surface turbidi-
ty, and color associated with humic material. It is believed,
therefore, that the observed Secchi depth can be used in
later conclusions about the trophic state of the lake.
Conductivity and pH
Conductivity generally ranged between 50 and 80
miliiohms/cm, with occasional values to 100 milliohms/cm
near bottom. There was no seasonal trend in conductivity.
Generally, pH was between 7.1 and 8.0 during spring and
summer in surface waters with the exception of March 30
when a pH of 6.4 was recorded. From October to January,
pH declined to 6.1 to 6.7 in surface waters, rising to 7.4 in
February.
Phytoplankton
Three phytoplankton population peaks occurred during the
study period. A May 1981 peak was dominated by green
algae, primarily Scenedesmus and Cosmarium. A Stauras-
trium bloom occurred in September 1981. A January 1982
population peak consisted of Euglenids, with Eug/ena and
Trachelomonas being the most abundant organisms.
Relatively few Cyanophyta (bluegreen algae) were identified
during the study period. Bluegreens identified included
Chroococcus and Microcystis.
Nutrient Concentrations
Total phosphate levels measured in Lake Hicks during the
study ranged from 0.009 mg/l to 0.190 mg/l. The mean con-
centration at surface was 0.042 mg/l, at mid depth was 0.057
mg/l and at depth was 0.082 mg/l.
Orthophosphate (soluble phosphate) measurements rang-
ed from below detection levels to 0.071 mg/l. Wide fluctua-
tions are also noted for this measure of phosphate. The
highest levels reported (0.071 mg/l) occurred at depth dur-
ing July.
Concentrations of inorganic nitrogen at the surface were
low throughout the study until winter, when levels reached
0.378 mg/l in February. Low levels of ammonia were also
found at the surface until December when levels reached
0.315 mg/l. Total Kjeldahl nitrogen ranged from less than
0.5 mg/l to 0.9 mg/l with the highest organic nitrogen occur-
ring during the summer.
Water Budget
Direct measurement of inflows to Lake Hicks proved imprac-
tical when vandals repeatedly removed weirs that were in-
stalled on the main inlet. A water budget, expressed as the
number of times the lake volume is replaced in a year (tur-
novers per year), was developed indirectly by two methods.
Runoff from the watershed was calculated from rainfall
records and a loading model (Buffo, 1979; Davis, 1980). An-
nual runoff volume divided by the volume of the lake pro-
vides one estimate of turnovers per year.
During the study year, runoff volume was calculated to
be 826,445 m3(670 acre-feet). When this amount is divided
by the volume of the lake (about 49,340 m3 (40 acre-feet)
at an elevation of 104 m (344 feet), a yield of 17 turnovers
per year is determined. A typical year would have somewhat
less runoff than the study year because of fewer intense
storms. An estimated typical year would have 542,740 m3
(440 acre-feet) allowing calculation of 11 turnovers per year.
301
-------
Lake Restoration, Protection and Management
A second indirect estimate of turnovers was calculated
from the results of two dye studies. Twice, the entire volume
of the lake was dyed with rhodamine WT fluorescent dye.
A final concentration (volume-weighted average for several
depth strata) of 1.2 ppb was achieved in March 1981 and
a concentration of 2.4 ppb was achieved in July of 1981.
Decreasing dye concentrations were observed for 40 to 45
days at regular depth intervals at 22 locations in the lake.
Dye concentrations were plotted as straight lines on a
semi-log scale (Fig. 2), indicating nearly constant dilution of
the dye over time. During the July study, the lake was strong-
ly stratified, but similar dye concentrations were achieved
o
I
1-
0.5-
0.3-
0.2-
DATE
(7/28/82)
0.1-
DATE
(3/20/82)
10
20
30
DAYS
40
50
60
Figure 2.—Dye concentrations.
in the epilimnion and the relatively anoxic hypolimnion. One
week later, a layer of water with relatively little dye was
observed at a depth of 3.63 m (12 feet) throughout the lake,
with higher concentrations above and below, indicating
substantial inflow from springs. Subsequent partial mixing
of the lake caused more uniform, but lower, concentrations
of dye throughout the lake. There was no surface inflow dur-
ing the July through August dye study.
The slopes of the lines in Figure 2, -0.0693 per day (sum-
mer) and -0.033 per day (winter), can be used to calculate
turnover times of 15 to 20 days or 12 to 24 turnovers per year.
The steeper slope (more rapid decrease in dye concen-
tration) during the summer study was probably caused by
photodegradation of the dye during sunny weather. Rhoda-
mine WT was observed to photodegrade at a rate of 0.018
per day during experiments in an estuary (Zankel, unpub.).
The turnovers calculated from modeled runoff, 17 per year,
were in close agreement with those calculated from the dye
study, 12 to 24 per year (or 19 to 24 per year allowing for
photodegradation of the dye).
Lake Hicks, with its relatively small volume of about 49,340
m3 (40 acre-feet) (at its usual summer elevation), has rapid
replacement of its water, primarily from springs in the sum-
mer and from springs and surface runoff in the fall, winter,
and spring.
Trophic Status of Lake Hicks
Algal densities have not been sufficiently high in Lake Hicks
to be perceived as a problem by lake users. Nevertheless,
concentrations of chlorophyll a, an indicator of algal biomass,
were observed to exceed 20 ^g per liter during our studyi
and decomposing algae probably contribute to the oxygen
demand in the hypolimnion during the summer. Algae may
also contribute to decreased transparencies in the fall.
Annual phosphate loadings to the drainage basin were
determined using land use patterns and Buffo's model for
actual rainfall frequencies during the study period.
Given the loading rate of phosphorus per year, the average
depth of the lake (z), and the turnover rate in years (T) (tau),
the trophic state can be expressed (Vollenweider, 1975).
The results of the loading rate model indicate a loading
rate to the lake of 2.8 grams per m2 year. Average depth
of the lake is about 3 meters and T (tau) is between 0.045
and 0.091 year (calculated from the water budget).
Lake Hicks is eutrophic according to Vollenweider's (1975)
model, as shown in Figure 3. Using the water quality data
from other sections (chlorophyll a, transparency, phosphorus
concentration), Lake Hicks scores 65 on Carlson's (1977)
trophic state index. That score indicates a slightly eutrophic
lake and agrees with the classification according to Vol-
lenweider's (1975) model.
10
O
Z
Q
O
0.
w
O
0.
2.8-
1.0
0.1
0.01
EUTROPHIC
OLIGOTROPHIC
0.1
1.0
10
100
z/T (m/yr)
33
SOURCE: Vollenweider's Phosphorus
Loading Criterion (1975)
Figure 3.—Trophic status of Lake Hicks.
Nutrient Limitation
While algal blooms have not interfered with the use of Lake
Hicks in the past, assessment of nutrient limitation might
aid in predicting the impacts of restoration. In addition,
controlling algae might reduce turbidity, a factor in clos-
ing the lake to swimmers. During August through October
a moderate algal bloom occurred in the lake, while the con-
centrations of inorganic nutrients in the surface waters re-
mained relatively low.
It is suspected that the small size of the lake leads to
some vertical mixing of the supply of nutrients from the
relatively rich bottom waters to the surface waters
throughout the late summer and fall. Vertical profiles of
dissolved oxygen and nutrient concentrations support this
hypothesis, particularly in the fall.
Ratios of nitrogen (N) to phosphorus (P) in the lake can
be compared to those expected in algal material (Redfield,
1934) to estimate which of the major nutrients is poten-
tially limiting phytoplankton growth. Comparison of avail-
able nutrients would be desirable (e.g. dissolved inorganic"
N and P); however, sampling protocol did not favor that
comparison. Ratios of total N (nitrate and nitrite and TKN)
to total P suggest nutrient balance during the summer
months.
Phosphorus limitation was suggested only on July 2 and
December 22. Other ratios such as inorganic N to in-
302
-------
Contributed Papers
organic P (given that some contribution from algal cells
was present) and of inorganic N to total P were also cal-
culated. The three ratios allow conclusions about different
aspects of algal growth. Total N to total P predicts the total
algal crop that could develop in a body of water. Roughly,
0.1 mg of chlorophyll a might be expected from each mg
of phosphorus under balanced nutrient conditions (Dillon
and Rigler, 1974). Maximum phosphorus concentrations
of 0.061 mg/l in the spring were followed by maximum
chlorophyll concentrations of just over 0.020 mg/l, sug-
gesting that nitrogen may have been the limiting nutrient.
The ratios of inorganic N to inorganic or total P allow
some speculation that adding nitrogen would be more like-
ly to stimulate algal growth in Lake Hicks than adding
phosphorus. The conclusion assumes that excess P is
stored by algal cells and that N is recycled more rapidly
than P. Algal species composition also indicates that Lake
Hicks may be nitrogen limited. Very few blue-green algae
were noted during this study. The predominant phyto-
plankton division is Chlorophyta. Blooms of Euglenids
were noted in July and September, as well as a gradual
increase in the percent composition of Euglenids during
the wet season with another bloom in January. Given the
rapid turnover of water volume in the lake and the pro-
bable supply of nutrients from the hypolimnion during the
summer, Lake Hicks is probably somewhat insensitive to
additions of nutrients from inflows. Long-term re-
duction in nutrient input might eventually deplete available
nutrients at the sediment surface, however.
Bacterial Concentrations and Ratios
Fecal coliform and fecal streptococcal bacteria were
sampled at the surface and near bottom when the lake
was mixed, and at three depths during stratification. Con-
centrations of fecal coliforms in surface waters exceeded
200 MPN/100 ml on three occasions during the spring and
summer and two occasions during the fall. Instances of
fecal coliform counts over 200 MPN/100 ml in surface
waters appear to be associated with precipitation events
in July, September, and November. High counts in May
were during a period when continuous surface flows to the
lake were still occurring, and when anoxic bottom waters
were observed at a relatively shallow depth.
The observation of 240 fecal coliforms per 100 ml on
August 11 did not coincide with either precipitation or lake
mixing. Generally, concentrations of fecal coliforms were high
in surface waters primarily after inflow-producing rainfall, ex-
cept that the rather steady inflow during December through
April was not associated with high concentrations.
Concentrations of fecal coliforms in bottom waters resem-
bled those in surface waters except that concentrations re-
mained high into early June. High concentrations of fecal
coliforms were observed at mid-depth only on June 17. No
unusual flow or mixing events had been recorded on or short-
ly before. Bacterial concentrations in inflowing surface waters
were high in May and June, after the mid-July storm, dur-
ing late September and October, and during the two storm
events. High concentrations in surface inflows thus coincide
with or precede most high concentrations in the lake.
Concentrations of fecal streptococci (FS) were generally
similar to, or slightly higher than those of fecal coliforms (FC),
resulting in FC and FS ratios of about 1 or less on most
dates. Ratios approaching or exceeding 4, indicating the
possibility of human fecal sources (Geldreich et al. 1968),
occurred in samples from surface waters once in the spring
and twice in the summer. The spring and first summer
samples also contained relatively high counts of fecal col-
iforms, supporting the possibility of contamination by
unknown sources.
The FC and FS ratio at the channel flowing into the lake
was 26 with concentrations of 16,000 MPN fecal coliforms
per 100 ml. The FC and FS ratio in the detention pond
feeding the channel was 67 with 2,400 MPN fecal coliforms
per 100 ml. Ratios of FC and FS at the detention pond chan-
nel into the lake were generally similar to those in the lake,
although bacterial concentrations were considerably higher.
There was no surface flow to the lake coincident with the
high FC and FS ratios in the two summer samples.
Sanitary-Storm Drain Systems and
Nonpoint Source Investigation
The investigation of both the sanitary and storm sewer
systems did not reveal any direct point sources. Verification
of actual sanitary hookups in areas north of Lake Hicks but
within the basin should be carried out; however, budget con-
straints in this study did not permit such additional work. Field
investigations of sanitary-storm drain systems and other
possible pollution sources did provide some information.
A 53-cm (21-inch) diameter outfall comes into Lake Hicks
from a nearby housing development. However, compressed
air and smoke testing did not reveal the actual outlet.
Samples taken from the first manhole upstream from the lake
did not reveal the discharge from this pipe to be a pollution
source.
A swimming pool drain is hooked up directly into a 53-cm
(21-inch) diameter storm sewer. This pool has been drain-
ed only twice since being constructed, and, since both oc-
casions were prior to this study, it was assumed that the pool
had little influence on the water quality of the lake. Additional-
ly, although it was thought that the pool's diatomaceous filter
backwash water drained into a 53-cm diameter storm sewer
that led to the lake, the investigation indicated that the
backwash water drained into a 38-cm sanitary trunk sewer.
Horses graze near the lake, indicating a possible source
of bacteria and nutrients to Lake Hicks.
Waterfowl
Lake Hicks was inhabited throughout the study period by
approximately 40 semi-domesticated waterfowl, predominant-
ly mallard and domestic white ducks. As a group, the ducks
appear to spend about 50 percent of their time in the water
and 50 percent of their time on the grassed shoreline. The
contribution of fecal coliform bacteria to the lake water by
waterfowl is insignificant. Fifty waterfowl contributing
11,000,000 coliform bacteria per day (Millepor Manual, 1973)
would add about 2.75 x 10* coliforms per day, assuming
that only about 50 percent of their time is spent in the water.
Assuming also that no coliform mortality occurs, they would
produce an average density of less than 1 per 100 ml of
water.
RESTORATION ALTERNATIVES
Restoration alternatives were divided into two groups for
evaluation: alternatives requiring no construction and alter-
natives requiring construction.
• Nonconstruction Measures
1. Storm runoff diversion into sanitary sewer
2. Street and catch basin cleaning
3. Animal control
4. Public awareness
5. Waterfowl detraction
6. Addition of hydrated lime to increase
transparency
• Construction Measures
1. Dredging
2. Lake drawdown
303
-------
Lake Restoration, Protection and Management
3. Chlorination
4. Dilution
5. Ground infiltration and excess runoff diversion
Following extensive evaluation and a number of public
hearings, three alternatives were selected to restore the lake:
Public Awareness
Public awareness plays a very important part in any type
of source control and remedial action plan. This alternative
involved making the public aware of the problems and how
they could help reduce contamination within the drainage
basin. Field trips and other outdoor activities were to be ar-
ranged to interest people in the future of this lake.
Addition of Hydrated Lime
Humic substances that have entered Lake Hicks from nearby
peat deposits can be precipitated by adding a slurry of
hydrated lime and water. Increases in transparency of 40
to 60 percent have been achieved by this method (Stress
and Hasler, 1960). The technique has been ineffective only
in lakes that are totally surrounded by peat bogs (Stross and
Hasler, 1960). The important feature is that the lime must
be mixed with lake water before application to achieve max-
imum dispersal.
Another purpose of liming lakes has generally been to in-
crease production of fish. This has usually resulted from the
increased water clarity leading to growth of suitable food
organisms through a greater volume of the lake. Increased
concentrations of oxygen in hypolimnion (bottom) waters can
also occur if transparency is sufficiently improved. Increased
fish production might thus be a secondary benefit of liming.
Ground Infiltration and Excess
Runoff Diversion
This construction restoration alternative was selected for
several reasons. Total diversion of all summer storm runoff
is the only positive way of eliminating the "first flush" pol-
lutants entering the lake from five major input locations. Soil
conditions are also suitable for infiltration of storm runoff.
This was indicated by no-flow conditions into the lake from
the detention pond's inlet channel during two summer storm
events. External coliform loading will be reduced by diverting
all flows up to 0.6 m3/s from the inlet channel between the
detention pond and the lake to the existing pump station.
Flow from three inlet pipes to the south will also be diverted
to the pump station. A small dike is proposed on the inlet
channel, as shown on Figure 4, to permit first flush runoff
to infiltrate into the existing soils. Excess runoff up to 0.06
m3/s will be diverted directly into the wet well of the existing
pump station through 25.4 cm diversion line with its inlet
lower than the top of the rock berm.
During extended flows in excess of 0.06 m3/s, the berm
will be topped and water will enter the lake through the ex-
isting channel. To accept the runoff from the adjacent junior
and senior high schools and homes, an interceptor sewer,
pump station, and force main submerged on the lake bot-
tom will divert runoff up to 0.03 m3/s, also directly into the
wet well of the pump station. During the winter months the
pump station will be shut off and water will enter the lake
through the existing influent lines. The dike on the north side
will be submerged during peak runoffs with little backwater
effect on the existing culverts at the parking lot.
The existing suction line for the outlet pump will be replac-
ed and extended to the deepest part of the lake. During storm
flows of less than 0.06 m3/s, water will be withdrawn from
hypolimnion. This will result in removal of nutrients, turbidi-
ty, anoxic water, and floe that precipitates from the addition
of lime. During the spring months, water higher in fecal col-
iform concentration will be selectively removed from several
inlet pipes. Profiles of the proposed interceptor sewers and
detention pond are shown in Figure 5.
ACKNOWLEDGMENTS: Many people assisted in the Lake Hicks
diagnostic feasibility study by contributing background information,
LAKE HICKS RESTORATION
STORM WATER DIVERSION
PLAN
Figure 4—Lake Hicks restoration stormwater diversion plan.
304
-------
Contributed Papers
MlWnDU
INTERCEPTOR PROFILE
DETENTION POND PROFILE
SCALE'
so 100 iso ago ;so xa FEET
10 20 30 40 SO 60 FEET
LAKE HICKS RESTORATION
STORMWATER DIVERSION
PROFILES
Figure 5.—Lake Hicks restoration stormwater diversion profiles.
working in the field, and actively participating in public meetings.
CH2M HILL also acknowledges the public participation program
directed by Dan P. Cheney from Shapiro and Associates to promote
active input from all members of the Lake Hicks Task Force (Lake
Hicks Advisory Committee). The participation of the task force
members was required to ensure the success of the project, and
consequently is greatly appreciated. Funding for the analysis was
provided through the King County Division of Parks and Recrea-
tion, the Washington State Department of Ecology, and the U.S.
Environmental Protection Agency (Clean Lakes Program).
REFERENCES
Buffo, J. 1979. Water Pollution Control Early Warning System, Sec-
tion 1, Non-Point Source Loading Estimates. Metro, Seattle, Wash.
Carlson, R.E. 1977. A trophic state index for lakes. Limnol. Oceanogr.
22:361-369.
CH2M HILL. 1978. Preliminary Investigation of Water Quality Pro-
blems at Lake Hicks. Bellevue, Wash.
1979. White Center Drainage Improvement Study. Bellevue,
Wash.
1980. Environmental Impact Statement, White Center Drain-
age Improvement Project. Bellevue, Wash.
Davis, S. 1980. Water Pollution Control Early Warning System, Sec-
tion 2, Assessing Current and Future Status of Lakes. Metro, Seat-
tle, Wash.
Dillon, P.J., and F.H. Rigler. 1974. A test of a simple nutrient budget
model predicting the phosphorus concentration in lake water. J.
Fish. Res. Board Can. 31:1771.
Geldreich, E.E., L.C. Best, B.A. Kenner, and D.J. Van Donsel. 1968.
The bacteriological aspects of stormwater pollution. J. Water Pollut.
Control Fed. 40:1861-1871.
Millepor Manual. 1973. Biological analysis of water and wastewater.
AM302.
Redfield, A.C. 1934. On the proportions of organic derivatives in sea-
water and their relation to the composition of plankton. Pages
176-192 in James Johnstone Memorial Volume, Liverpool Univ.
Press.
Stross, R.G., and A.D. Hasler. 1960. Some lime-induced changes
in lake metabolism. Limnol. Oceanogr. 5:265-272.
Uchida, B.K., et al. 1976. An Intensive Water Quality Survey of 16
Selected Lakes in the Lake Washington and Green River Drainage
Basins (1973-1974). Seattle Metro. Mss.
Vollenweider, R.A. 1975. Input-output models with special reference
to the phosphorus loading concept in limnology. Schweiz 2.
Hydrol. 37:53-84.
Zankel, K.L. 1981. A study of optical methods to reduce background
fluorescence in dye studies. Environ. Center, Martin Marietta Corp.,
Baltimore, Md. (Unpubl.)
ADDITIONAL BIBLIOGRAPHY
Entrance. 1978a. Restoration Analysis - Wapato Lake. Submitted to
City of Tacoma, Metro. Park Distr., Wash.
. 1978b. Water Quality and Restoration Analysis - Thurston
County Lakes. Submitted to Thurston County and the City of
Lacey, Wash.
Harper-Owes. 1981. Draft Environmental Impact Statement. Pine
Lake Restor. Anal. Metro, Seattle, Wash.
King County, Division of Planning. 1977. Highline Communities Plan.
Seattle, Wash.
Kramer, Chin & Mayo, Inc. 1977. Lake Ballinger Rehabilitation Study.
Rec. Plan for the City of Mountlake Terrace, Wash. Submitted.
Pitt, R. 1979. Demonstration of Nonpoint Pollution Abatement
Through Improved Street Cleaning Practices. 600/2-79-161. U.S.
Environ. Prot. Agency, Washington, D.C.
Reckhow, K.H. 1979. Quantitative Techniques for the Assessment
of Lake Quality. Michigan State University, East Lansing. U.S.
Environ. Prot. Agency, Washington, D.C.
Smith, SA, J.O. Peterson, S.A. Nichols, and S.M. Bom. 1972. Lake
Deepening by Sediment Consolidation - Jyme Lake. Inland Lake
Demonstration Proj., Upper Great Lakes Comm. Madison, Wis.
Sylvester, R.0.1960. An Engineering and Ecological Study for the
Rehabilitation of Green Lake. Univ. Washington, Seattle.
305
-------
Lake Restoration, Protection and Management
U.S. Environmental Protection Agency. 1976. Proc. Workshop on U.R.S. Company. 1977. Lake Fenwick Restoration Study. Submitted
Microorganisms in Urban Stormwater. Edison, N. J. to the City of Kent Parks and Recreation Dep. Wash.
1977. Microorganisms in Urban Stormwater. Edison, N. J. Washington State University. 1968. Washington Climate. Pullman,
1979. Bacteria, Water Quality Standards Criteria Digest, A Wash-
Compilation of State/Federal Criteria. Washington, D.C. Washington, State of. 1977. Washington State Water Quality Stan-
dards. Dep. Ecology.
306
-------
EFFECTS OF ALUMINUM SULFATE TO MIDGE LARVAE
(DIPTERA; CHIRONOMIDAE) AND RAINBOW TROUT
(SALMO GAIRDNERI)
DAVID S. LAMB
Michael A. Kennedy Consulting Engineers, Spokane, Washington
GARY C. BAILEY
Department of Civil and Environmental Engineering,
Washington State University, Pullman, Washington
ABSTRACT
Two separate research projects have been conducted to assess the effects of lake aluminum sulfate (alum)
treatments on aquatic organisms. The first was a laboratory study that attempted to determine the effects
of alum on benthic invertebrates using larvae of the ehironomid species Tanytarsus dissimilis in both acute
(96 hours) and chronic bioassays. The second project was a field study performed in conjunction with
the Liberty Lake restoration project and is entitled "In situ toxicity of aluminum sulfate to rainbow trout
(Salmo gairdnen)." The studies indicated there was no acute toxicity to either the Tanytarsus larvae or
the rainbow trout. In addition, there were no sub-lethal effects to rainbow trout exposed directly to 0.9
mg Al/l or to trout exposed to the settled alum floe as measured during a 1-month observation period.
A number of effects, including chemical toxicity, physical toxicity, and lengthening of larval development
time, were seen with the Tanytarsus larvae during chronic bioassays.
EFFECTS OF ALUM TO MIDGE LARVAE
The objective of this study was to determine the acute and
chronic effects of alum to Tanytarsus dissimilis larvae. This
species is a member of the Chironomldae, the family of
organisms that occupies a significant portion of the benthic
invertebrate community in lakes and which are important
fishfood organisms.
Methods
The test water for these bioassays was obtained from Liberty
Lake, Wash., and filtered, using 0.45 /^m pore size filters as
soon as possible. The alum stock solutions were prepared
using this water and reagent grade aluminum sulfate
(AI2(SO4)3 H2O) crystals.
After the alum solutions were prepared a pH adjustment
procedure was started, the object of which was to obtain a
pH level of 7.8 in the acute tests and 6.8 in the chronic tests.
The pH of 7.8 was desired because it is indicative of condi-
tions in Liberty Lake during the fall (when an alum treatment
might be performed). The lower pH was used for the chronic
tests because it was observed to remain more stable over
the longer time period. pH's in the solutions, including con-
trol, were adjusted by adding sodium hydroxide (NaOH)
every second or fourth day until the pH stabilized at the
desired level.
The test larvae were collected from stock tanks 15 to 18
hours prior to each test start-up and measured to obtain the
instar desired. Second instar larvae were used for the first
acute test and third instars were chosen for the second and
third acute tests. The larger larvae were used to facilitate
handling and o&servation under a dissecting microscope
while in the test beakers. Second instar larvae were used
for all chronic tests. The chosen larvae were held in petri
dishes containing filtered lake water and algae until the start
of each test.
These bioassays were conducted in 50 ml pyrex beakers
containing 30 ml of solution. Five larvae were transferred
randomly to each beaker as soon as the alum (or control)
solution had been added and a pH reading had been taken.
For the chronic tests a substrate of the alga Selenastrum
capricornutum was established in each beaker prior to ad-
ding solution.
At the beginning of each test-specific conductance, car-
bon dioxide and methyl orange alkalinity were determined
on the stock solutions. At the same time a sample of each
solution was taken and filtered for a series of chemical
analyses. At the conclusion of each test the contents of each
test beaker were filtered, using 0.1 t*m pore size filters, and
acidified for dissolved aluminum analysis.
The acute bioassays (Tests No. 1, 2, and 3) ran for 96
hours, and tested alum doses of 80,160,240, 320,480,560,
720, and 960 mg/l. The chronic tests (Tests No. 4 and 5)
used doses of 10, 80, 240, 480, and 960 mg/l and ran for
55 days. The larvae in the chronic tests were transferred to
beakers containing fresh solution and algae when a yellow-
ing of the substrate was observed.
Results
The results of the acute bioassays indicate no apparent ef-
fect of alum on either second or third instar Tanytarsus lar-
vae. The larvae, including controls, were generally active
throughout the test, exhibiting typical movements and food
searching.
The aluminum hydroxide "floe" layer, which ranged from
a patchy 1 to 2 mm thickness to a solid 3 to 4 mm thick
mat, appeared to have no detrimental effects on the larvae.
This was supported by the fact the larvae used the floe
materials to build the tubes they inhabited.
The results of the chemical analyses performed on the
acute test stock solutions appear in Tables 1 and 2. Of the
parameters listed in Table 1, all except bicarbonate (HCOa)
are fairly indicative of (filtered) Liberty Lake waters. The
HCO3 has been artificially raised by the addition of NaOH
during the pH adjustment procedure. Of the parameters
tested, only specific conductance, sodium (Na) and sulfate
(SO4) were observed to change with alum dose (Table 2).
Other parameters analyzed for included dissolved oxygen
(DO), pH, and dissolved aluminum. The DO measured at
307
-------
Lake Restoration, Protection and Management
Table 1. — Background water quality data for acute
bioassays with alum (mg/l).
Parameter Concentration
Parameter
Concentration
Ca
Mg
K
SiO2
Cl
HCO3
5.0 ±.5
1.2 ±.1
1.1 ±.3
4.3 ± .8
1.4 ±.4
30 ±5
PCyP
t-P
N03
NO2
NH3
C02
.008 ± .007
.014 ±.006
.070 ±.009
.01
.019 ±.008
1.0
Table 2. — Water quality data for acute bioassays with alum.
Alum
Dose
C
80
160
240
320
400
480
560
720
960
Conductivity"
(fjmhos/cm)
73 ± 14
124 + 5
206 ± 28
287 ± 21
360 ± 53
437 ± 32
501 ± 42
570 + 50
710 ± 50
840 + 60
SO4b
(mg/l)
3.0
40
76
113
145
178
217
246
310
410
Nac
(mg/l)
10.5
22.3
38.4
54.0
68.0
81.8
93.6
94.1
109.2
118.8
a Average of Tests #1 and #2 for 560, 720 and 960 mg/l
b Data from Test #1
c Data from Test #2
the end of 96 hours in the Test No. 1 beakers was 5.5 ± .2
mg/l. The mean pH was observed to drop in all acute tests
from 7.71 (±.11) to 6.85 (±.15) over the 96-hour period.
Dissolved aluminum was less than 0.1 ^g/l at all alum doses.
In the chronic bioassays mortalities were observed at all
alum doses tested but with widely different rates of dieoff
(Fig. 1). The time to reach 50 percent mortality was shortest
with 480 mg/l at about 4 days. The mortality time for 80 mg/l
and 240 mg/l was not significantly different (a = .05) and oc-
curred between 8 and 10.5 days. At 960 mg/l it took more
than 23 days for 50 percent of the larvae to die. The 10 mg/l
scy
v/.
Figure 1 .—Cumulative percent mortality with 95% confidence in-
tervals (when > 50%) for chronic bioassays with Tanytarsus dissimilis
in alum solutions (alum doses in mg/l).
dose showed less than 37 percent mortality at the end of
55 days when the test was terminated. Although not shown
on Figure 1, the combined control mortality was 5.4 percent.
Prior to death, the larvae were generally active and were
observed to build tubes and feed on the algal substrate. The
floe layer appeared more dense than in the acute test
beakers, probably because algae were trapped within it, but
it tended to clump more than in the acute beakers. The 10
mg/l beakers contained very little floe.
The results of analyses performed on the chronic test alum
solutions appear in Table 3. Additional testing indicated that
the DO was 4.8 ± .2 mg/l one-half hour before the lights
came on in the test chamber. The pH stayed fairly constant,
with an overall average of 6.63 + .32 obtained from initial,
mid-test, and final readings in all doses. Soluble aluminum
levels were again less than 0.1 ng/l in the test beakers.
There appears to be some chemical toxicity at the lower
levels of alum tested in these chronic bioassays, the effect
being most evident at 480 mg/l. Also, the heavy alum floe
in the 960 mg/l dose appears to cause mortalities after an
extended time, possibly due to its inhibiting feeding and
movement of the larvae. Additionally, some aspect of the
test conditions seemed to cause a physiological stress which
caused an overall lengthening of the larval development time.
The accepted life cycle time for this species is 14 days at
20°C (Nebeker, 1973). In this study no larvae pupated in 55
days.
The results of the chemical analyses performed on the
alum toxicant solutions, including dissolved aluminum, show
no variation to account for the observed chemical toxicity.
As a result a chemical equilibrium computer model, called
REDEQL 2, was used to determine the theoretical distribu-
tion of aluminum and other constituents in these solutions.
This model uses the Newton-Raphson Iterative method to
calculate equilibrium speciation given actual water chemical
characteristics (McDuff and Morel, 1973).
The results of this mathematical analysis indicated that
aluminum is found in the seven forms shown in Figure 2.
3-0-
4-0-
50-
10-0-
150-
AI(OH)
3(s)
AKOH),,
AI(OH)
+ 2
1I I I Ml I I
'% ^&
Alum dose (mg/l)
Figure 2.—Equilibrium distribution of aluminum species predicted
by computer analysis of alum toxicant solutions (negative loo of molar
concentration).
308
-------
Contributed Papers
Table 3. — Water quality data chronic test solutions, (mg/l unless otherwise noted)
(C, 10, 80, and 480, Test #5; 240 and 960, Test #6).
Alum
Dose
C
10
80
240
480
960
Cl
2.7
1.9
2.3
2.1
2.3
HC03
29
28
29
32
—
31
C02
1
1
1
1
1
1
NH3 PO4-P
.01 .010
.065 —
.018 —
.050 —
.020 —
.065 -
t-P NO3
.011 1.15
— .45
— .89
— .13
— .84
- .063
NO2
.01
.01
.01
.01
.01
.01
Conductivity
(pimhos/cm)
60
55
123
235
460
615
This graph shows the concentration of aluminum compounds
(as negative log of the molar concentration) with increasing
alum doses. This graph indicates that after the silicate solid
(AI2Si2O5(OH)4) is precipitated from solution the solid alu-
minum hydroxide controls the distribution of aluminum. This
has in fact been substantiated experimentally (Robertson and
Hem, 1969; Hem et al. 1973). It is unlikely that these solids
cause the chemical toxicity observed.
Because of the polymeric, positively charged nature of the
alum floe (Hem and Robertson, 1967) negatively charged
ions and particles may become absorbed and precipitated
from solution. Some substances removed include phosphate
ion (Hsu, 1975), colloidal solids, and organic compounds
(Clark et al. 1977). This absorption could concentrate poten-
tially toxic materials directly above lake bottom organisms.
The possible toxic or stress effects of DO, pH, SO4 con-
centration, depth of water, crowding, and nutrition were ex-
amined, considering observed chronic test conditions. These
were found from the literature not to be critical. No analyses
were performed for organic compounds in the test solutions;
however, it appears that some interaction between the alum
and the algae or an algal byproduct produced the observed
toxicity. This postulate is supported by the observed discrep-
ancy between the acute and chronic tests. Also algae are
known to release a wide range of organic compounds in-
cluding growth-inhibiting substances that may affect other
organisms (Hellebust, 1974).
IN SITU TOXICITY OF ALUMINUM
SULFATE TO RAINBOW TROUT
The objective of this research was to determine the mortali-
ty or growth retardation of rainbow trout exposed to alum
during the course of a partial alum treatment of Liberty Lake,
Wash. Specifically, effects on fish exposed to a direct dose
of aluminum sulfate and on fish exposed to the settled alum
floe were determined.
Methods
Approximately 3,300 rainbow trout (Salmo gairdneri) approx-
imately 2.5 cm in length were placed in two floating cages
on April 24,1981. The cages were anchored in the northwest
section of the lake (Fig. 3). The cages (2.4 m long, 1.2 m
wide, .9 m deep) were constructed with galvanized conduit
frames and covered with 0.6 cm mesh nylon netting. The
hinged tops of the same construction covered all but a small
area under the feeder. An automatic timed feeder on top of
each cage delivered feed in the amount and frequency
recommended by the feed manufacturer. Dead fish were
removed and additional feed was hand fed daily. The auto-
matic feeders were checked and replenished weekly.
On May 21 the. fish were equally apportioned into three
cages (A = direct dose, B = floe exposure, C = control) and
the fish were sampled for length and weight. On May 25 the
LAKE OUTLET
Figure 3.—Liberty Lake, Wash, showing area of May 1981 alum
treatment.
feeders and hinged tops were removed and nylon netting
was laced over the tops. Cage A was then towed to the treat-
ment zone (Fig. 3), the styrofoam floatation was removed
and the cage was sunk to the bottom in 4 m of water. Cage
C (control) was towed for an equivalent 2 hours and then
sunk in the northwest section of the lake on May 26. The
alum treatment began on May 26 and was completed on
May 27. The alum was mixed with water in flumes on the
boats and applied to the surface of the water as a slurry.
During the course of the alum treatment the boats made
several passes near (within 5 m) of Cage A and one pass
was made directly over the cage. Cage B was towed to the
treatment area and sunk near Cage A on May 27 after com-
pletion of the treatment.
Divers checked the cages for mortality on May 28 and 1
kg of feed was hand fed. On May 30, 400 g of feed were
hand fed to each cage. Cages A and B were raised and tow-
ed to the original northwest location on June 3. Ideally, these
cages would have been left in the treatment area but
previous experience indicated that the probability of van-
dalism was high in this area. On June 4 the control cage
was raised, towed, and anchored near A and B. The fish
were hand fed and the automatic feeders reattached at this
time. The feeders were calibrated to 25 percent excess of
the recommended daily feed rate to assure there was no
food-limiting growth retardation. Thereafter, feeder operation
and fish length and weight were checked weekly and mor-
309
-------
Lake Restoration, Protection and Management
tality checked every other day. On July 8 and 9, 200 fish
were sampled for length and weight; 100 fish were retained
in each cage and the remainder were fin clipped to identify
treatment cage and released. The released fish were to be
recaptured by angling In the fall and the differential survival
was to be used to estimate the treatment effect.
Results
The mortality prior to treatment was generally less than 1
percent per week (Table 4) and showed a decreasing trend
over this acclimation period. This was well below the general-
ly accepted limit of 10 percent for control mortality before
and during toxicant testing.
No mortality was observed in either Cage A or B during
the treatment period while the cages were on the bottom
and only one mortality was observed in Cage C. The fish
fed eagerly during this period indicating they were not greatly
stressed.
During the post-treatment period when sub-lethal effects
should have been seen the number of mortalities did not
increase significantly compared to those observed In ttie con-
trol cage. The experimental fish were to be held as long as
possible for post-treatment observation; however, by July 8
the surface water temperature was 20°C and was not ex-
pected to peak until August, based on previous years' data.
As a result this date was used as the cut-off date for the
post-treatment period.
The post-release mortality increased sharply in Cage A
from July 16 to August 5 (Table 4). This increased mortality
may have been caused directly by disease that occurred ran-
domly in Cage A or to an effect of direct alum exposure.
The indirect probable cause, however, was the increase in
surface water temperature from 22,5 to 25°C during this
period.
The lethal temperature limit (100 percent) for the Kamloops
strain of rainbow trout used in this experiment is 25.7°C when
acclimated at 11°C. Fifty percent mortality is at 24°C (Black,
1953). Subcritical high temperature may also cause mortality
by reducing a fish's resistance to disease or other stress.
The mortality in Cage A was obviously not caused solely by
high temperature because all three cages of fish were ex-
posed to the same temperature. The high mortality of Cage
A may have resulted from disease and its high occurrence
in Cage A simply a random event. The dead fish deteriorated
rapidly at the high temperature but those that had not decom-
posed when removed showed no obvious external symp-
toms of disease.
The increased mortality of Cage A may have been a result
of the alum exposure. If, for example, the alum had caused
gill damage those fish would not be able to survive as well
at increased temperature and reduced dissolved oxygen.
However, a microscopic examination of gill filaments and
lamella of deceased fish showed no difference between A,
B, or C fish. This does not exclude the possibility of some
physiological damage.
If the fish in Cage A were affected by the alum treatment
to the point of increased mortality at high temperature it
should also be expressed as reduced growth during the post-
treatment observation period. We observed no apparent
decrease in weight or length gain of acute treatment fish
when compared to Cage B or Cage C fish (Fig. 4). The mean
weight and/or length of Cage A fish were significantly greater
for the July 8 and August 18 measurement periods than
Cage B or Cage C fish. The actual difference, however, is
slight and is probably well within cage to cage variation. The
major source of this variation is the feeder and although every
effort was made to deliver the same amount of feed to each
cage some discrepancy occurred because of battery failure
and mechanical malfunctions.
Attempts to recapture the treatment fish by angling in Oc-
tober of 1981 were unsuccessful. As a result a creel cen-
sus was conducted on the opening day of fishing season
in 1982. This effort was unsuccessful as well, with only two
recaptures from Cage A, four from Cage B, and one from
Cage C.
The alum treatment exposed the fish to E- i aluminum con-
centration slightly greater than those suggested by Freeman
and Everhart (1971) but the exposure was very brief. The
Table 4. — Mortality for experimental fish
Pre-treatment
April 25 -
May 2 -
May 9 -
May 16 -
May 23 -
Treatment
May 26 -
May 1
May 8
May 15
May 22
May 26
period
June 4
No. mortality
61
27
11
1
0
Cage
ABC
0 0 1
Approximate % mortality
2
1
1
1
0
% Weekly mortality
A B
0 0
C
.1
Post-treatment
June 5-June 11
June 11 -June 18
June 19 -June 25
June 26 - July 2
July 3 - July 8
Post-release
July 9 -July 15
July 18 - July 22
July 23 -July 29
July 30 - August 5
0
45
29
19
8
4
4
16
16
6
0
0
26
31
0
30
28
25
.8
.4
.4
1.6
1.7
0
2
10
1
.6
0
0
2.6
3.2
0
6
3
3
310
-------
Contributed Papers
14-
5 12~
en
S 11-
10-
9-
8.
35-
30-
en
r 25
I 20-
15-
10
5
—
fcl-
_
Aniricl
xi
^D
I/)
O
Q.
X
Ul
I
1-n
n
c|~
A
A
—
P
P
—
—
r
~
r
—
—
A
—
R
—
r
_
_
A
—
R
—
r
—
A
—
—
B
"ay 20 June 11 June 18 July 8 Aug. 18
Figure 4.—Mean lengths and weights with 95% confidence intervals of fish exposed to an aluminum sulfate treatment (A = dosed, B =
exposed to floe, C = control).
actual aluminum concentration measured immediately after
alum treatment was .07 mg/l which is only slightly higher
than the background concentration (Gibbons, 1981).
CONCLUSIONS AND IMPLICATIONS FOR
LAKE TREATMENTS
Lake alum treatments have been performed using a wide
range of alum doses (Dunst et al. 1974). Some of these
treatments used alum in the range of doses causing toxici-
ty in the bioassays with Tanytarsus larvae. However, a
number of factors would tend to mitigate these effects in
natural aquatic environments. The predominant hardness-
causing ions for instance, can compete with Al+3 for both
organic and inorganic ligands and decrease the tendency
to form soluble complexes (Stumm and Morgan, 1970). If
an algal byproduct peculiar to Selenastrum is in fact caus-
ing or contributing to mortality it is unlikely that this would
be found in significant enough amounts in a diverse lake
system to be a problem.
The physical toxicity, however, could be a factor regardless
of dose. A heavy floe layer may not be a problem for already
established larvae with a normal development time, but a
substantial floe layer could inhibit pupae from reaching the
surface and the deposited eggs from reaching the sediments.
Concerning the trout study, no acute toxicity was seen with
rainbow trout exposed directly to 0.9 mg/l of aluminum as
aluminum sulfate or to those fish exposed to the settled alum
floe. There was also no apparent chronic effect of the alum
exposure over a 1-month post-treatment observation period,
as measured by comparative lengths and weights.
Both of these studies point to the importance of thorough
planning before lake alum treatments are performed. Not
only must the limnological character of the water to be treated
be well understood, but the time (season) when a treatment
is to be made may be an equally important factor. Thus a
late fall alum application would find most benthic insects dor-
mant or at low levels of activity, and the resident trout popula-
tion unstressed by high temperatures and more able to avoid
any harmful effects caused by alum exposure.
ACKNOWLEDGEMENTS: The research effort for the study on acute
and chronic effects of alum to midge larvae was performed at the
Environmental Engineering Laboratory at Washington State Univer-
sity with the advice and assistance of Gary Bailey and Dr. William
H. Funk. The study was funded, in part, by a grant from the U.S.
Environmental Protection Agency (Project No. R805604-01-0). The
"In situ toxicity of aluminum sulfate to rainbow trout" study was per-
formed under contract with Michael A. Kennedy Consulting
Engineers. Funding also was provided by EPA (Project No.
CS804487-01-1), the Washington State Department of Ecology (Pro-
ject No. 267601702), and the Liberty Lake Sewer District.
REFERENCES
Black, E.G. 1953. Upper lethal temperatures of some British
Columbia freshwater fishes. J. Fish. Res. Board Can. 10:196-210.
Clark, J.W., W. Viessman, and M.J. Hammer. 1977. Water Supply
and Pollution Control. Harper and Row Inc., New York.
Dunst, R.C., et al. 1974. Survey of Lake Rehabilitation Techniques
and Experiences. Wis. Dep. Nat. Resour. Madison.
Freeman, R.A., and W.H. Everhart. 1971. Toxicity of aluminum hy-
droxide complexes in neutral and basic media to rainbow trout.
Trans. Am. Rsh. Soc, 100:644-658.
Gibbons, H.E. 1981. Personal communication and unpublished file
data. Dep. Civil Environ. Eng., Washington State Univ. Pullman.
Hellebust, J.A. 1974. Algal Physiology and Biochemistry. Univ. of
California Press, Berkeley.
Hem, J.D., and C.E. Robertson. 1967. Form and Stability of
Aluminum Hydroxide Complexes in Dilute Solution. U.S. Geo.
Surv. Water Supply Pap. Washington, D.C.
Hem, J.D., C.E. Robertson, D.J. LJnd, and W.L Pdzer. 1973. Chemi-
cal Interactions of Aluminum with Aqueous Silica at 25°C. U.S.
Geo. Surv. Water Supply Pap. Washington, D.C.
311
-------
Lake Restoration, Protection and Management
Hsu, P.H. 1975. Water Res. 9:1155. Robertson, C.E., and J.D. Hem. 1969. Solubility of Aluminum in the
McDuff, R.E., and F.M.M. Morel. 1973. Description and Use of the J^f"" of, WjjLSSS9 rfr ^"^ U'S' ^ *""'
Chemical Equilibrium Program REDEQL 2. Keck Lab. Environ. Water ***** Pap' Washln9t°n. D-c-
Eng. Sci., Calif. Tech., Pasadena. Stumm, W., and J.J. Morgan. 1970. Aquatic Chemistry. Wiley-
Nebeker, A.V., 1973. J. Kans. Ent. See. 46:160. Interscience, New York.
312
-------
WATER QUALITY MANAGEMENT STRATEGY FOR LAKE
OKEECHOBEE, FLORIDA
FREDERICK E. DAVIS
J. STEVE REEL
South Florida Water Management District
West Palm Beach, Florida
ABSTRACT
Lake Okeechobee, a large (1,732 km2), shallow (Z = 2.5 m) lake in south central Florida, is the major
surface water reservoir for the region and a significant natural resource. The limnological characteristics
of the lake as determined by systematic monitoring since 1973 are hard water, turbidity, and high primary
productivity. The lake is classified as eutrophic based on classic critical water chemistry values for nutrients
and other parameters. Analysis of N:P ratios and primary production limiting factor studies by other resear-
chers suggest that both N and P may be potentially limiting at different times and locations in the lake.
Although the lake is subtropical with potentially high internal loadings, a modified version of the Vollenweider
(1976) model fits the lake well. The model indicates that current N and P loadings are 31 and 37 percent
above the excessive loading rate. Substantial uncontrollable atmospheric loadings will require up to 90
percent load reductions in some basins when the total excessive loads are allocated on an areal basis.
Various structural and best management practices were evaluated for implementation to reduce the Total
N and Total P loads in each basin to its areal allocation in a cost-effective manner. These analyses resulted
in two completely different management strategies in the two major problem watersheds to the lake. Best
management practices are being implemented in one of the watersheds through a Rural Clean Waters
Program administered by Federal agencies in cooperation with the South Florida Water Management District,
State and local agencies, and landowners. In the other major watershed, an approximate 128 km2 former
freshwater marsh will be diked off and restored by detaining runoff generated in the watershed for later
use by agricultural and other water users in the area.
INTRODUCTION
General Overview
This report summarizes the South Florida Water Manage-
ment District's efforts to develop a water quality manage-
ment strategy for Lake Okeechobee, Florida which is cur-
rently classified as eutrophic and stressed by high nonpoint
source nutrient loadings. The strategy calls for reducing
phosphorous and nitrogen inputs to Lake Okeechobee in
two priority watersheds. Both short-term actions (such as a
modified pumping schedule for the Everglades Agricultural
Area (EAA) which would include both drought and flooding
considerations) and long-term solutions (requiring con-
siderable implementation time) are proposed.
Goals
Historically, the primary goals of the District have been to
minimize flooding during periods of excess rainfall and to
maximize water supply capability to alleviate periodic
drought. Now a third major goal of equal importance is pro-
posed: to maintain and improve water quality. Implementing
a water quality management strategy for Lake Okeechobee
would be a major step toward achieving that goal. For Lake
Okeechobee, then, the primary water resource goals are (1)
minimize the impacts of flooding during periods of excess
rainfall, (2) maximize water supply storage, and (3) improve
the water quality of Lake Okeechobee.
FINDINGS AND CONCLUSIONS
General Water Quality and Trophic State
Lake Okeechobee, centrally located in the Rorida peninsula,
ranks (after Lake Michigan) as the second largest freshwater
lake in the United States. The lake is part of an extensive
water resource management system constructed by the U.S.
Army Corps of Engineers in the 1950's and 1960's and cur-
rently operated by the South Florida Water Management
District. The lake is a substantial multi-use resource and is
managed as a drinking water supply reservoir, flood control
reservoir, agricultural irrigation supply, recreational and com-
mercial fishery, and general aquatic wildlife habitat.
The water quality of the lake and its major tributaries has
been monitored extensively by the District since 1973. This
monitoring indicates that the lake is a eutrophic, highly
mineralized hard water lake. Dissolved oxygen concentra-
tions are high because of high gross primary productivity
and good reaeration. However, the average nutrient (nitrogen
and phosphorus), Secchi depth, and chlorophyll a concen-
tration are all well above the accepted critical values for
eutrophic lakes as shown in Table 1. Analysis of the current
Table 1. — Trophic state evaluation.
Parameter
Ortho P (mg/l)1
Total P (mg/l)1' 2
Inorganic N (mg/l)1
Total N (mg/l)2
Secchi depth (m)3
Chlorophyll a (mg/m3)2' "
Trophic state index (TSI)4
TP loading rate (g/m^r)2
TN loading rate (g/m2yr)2
L. Okeechobee
annual means
0.005-0.045
0.049-0.097
0.08-0.26
1.45-2.62
0.5-0.7
19.0-27.0
62.7-65.0
0.127-0.443
1.80-5.05
Critical
values
0.010
0.020-0.040
0.30
0.90
2.0
6.0-10.0
53
0.20
2.75
' Vollenweider, 1968.
'Kratzer, 1979.
3 U.S. Environ. Prot. Agency, 1974.
"Carlson, 1977.
313
-------
Lake Restoration, Protection and Management
nitrogen and phosphorus loadings to the lake, using a
modified version of the 1976 Vollenweider input/output model
recalibrated to Florida lakes, indicated that the current loads
for both nitrogen and phosphorus are more than sufficient
to sustain a eutrophic condition.
Analysis of Tributaries and
Nutrient Sources
Locations of the various inflow points and their tributary areas
are shown in Figure 1. Table 2 provides a summary of water
and nutrient inputs to Lake Okeechobee based on seven
Figure 1 .—Lake Okeechobee study area.
Table 2. — Percent summary of water, phosphorus and
nitrogen inputs to Lake Okeechobee.
Inflow
Rainfall
Kissimmee River
Fisheating Creek
S-2
S-71
Taylor Creek/
Nubbin Slough
(S-191)
S-84
S-3
S-72
S-4
S-133 &S-135
S-127, S-129 &
S-131
Other inflows
Water %
38.8
30.9
5.8
5.6
4.9
4.4
4.0
1.6
1.1
1.0
1.0
0.8
0.1
Total
phosphorus %
16.7
20.3
9.8
5.3
9.0
28.5
1.9
1.1
1.7
2.2
1.7
1.6
0.2
Total
nitrogen %
24.3
24.6
7.0
18.8
6.3
5.8
3.1
4.5
1.6
1.7
1.1
0.8
0.4
years of data. As indicated in the table, the Taylor Creek/Nub-
bin Slough basin (S-191) contributes a disproportionate
amount of total phosphorus compared with its flow input.
Similarly, the S-2 basin shows an analagous situation for total
nitrogen.
Application of the modified Vollenweider model to Lake
Okeechobee indicates that to meet the excessive loading
rates for total phosphorus and total nitrogen, the average
annual loading of total phosphorus must be reduced by 40
percent and that of total nitrogen by 34 percent.
It is clear from Table 2 that the most reasonable approach
to achieve these overall nutrient reductions would be to first
address those watersheds that contribute disproportionate
nutrient loads compared with their flow inputs. This was ac-
complished by ranking the watersheds in Table 2 in terms
of excessive total phosphorus and total nitrogen loadings.
Before the ranking was determined, however, two addi-
tional guidelines were necessary for the evaluation. First,
rainfall was considered a "noncontrollable" nutrient source.
Second, the Upper Kissimmee Chain of Lakes (upstream
of S-65) and Lake Istokpoga (upstream of S-68), receiving
waters themselves, are therefore also considered noncon-
trollable sources until water quality criteria are developed for
their watersheds.
With these guidelines in mind, two different methods were
employed to determine the relative watershed ranking. One
was to rank them according to drainage area (amount of
nutrient contributed per square mile of area drained) and the
other was based on annual inflow to Lake Okeechobee
(amount of nutrient contributed per acre-foot of water
discharged). Both methods ranked the Taylor Creek/Nub-
bin Slough basin (S-191) number one. The S-2 basin was
ranked in the second position with each method. Further-
more, the top seven watersheds were the same for both
methods, although the order differed slightly for positions
three through seven.
Table 3 presents nutrient loading data for the seven
highest ranked (priority) watersheds. Management actions
in these watersheds to achieve the desired load reductions
would meet the total overall target load reductions of 40 per-
cent total phosphorus and 34 percent total nitrogen. Further,
it is significant to note that actions taken in the Taylor
Creek/Nubbin Slough basin (S-191) and the Everglades
Agricultural Area (S-2 and S-3) to achieve load reductions
in those areas would accomplish approximately 70 percent
of the total overall desired load reductions.
After the basin ranking was determined, the next step
toward developing long-term solutions was to determine the
nutrient sources within each watershed. Based on land use
loading rates from previous and ongoing studies and land
use/land cover data, average annual loadings for various land
uses were calculated for each watershed. It was not surpris-
ing to learn that, in the 500 km2 Taylor Creek basin, dairies
and improved pasture are the dominant land uses and con-
tribute most of the total phosphorus and total nitrogen loads
from those watersheds. In the 700 km2 S-2 and S-3 basins,
soil type and land use (of which vegetables and sugarcane
are the majority) are the dominant factors controlling the total
nitrogen and total phosphorus loadings to the lake. The 200
km2 S-4 basin is approximately half improved pasture and
half sugarcane. It is also noteworthy that natural areas con-
stitute a significant percentage (in excess of one third of the
C-38, Fisheating Creek, and S-71 watersheds. Essentially,
these natural areas appear to be assimilating a portion of
the nutrient loads coming from the more intense land uses
such as improved pasture.
Evaluation of Alternatives
A wide range of technical alternatives was considered dur-
ing the evaluation process. These alternatives included:
314
-------
Contributed Papers
Table 3. — Desired load reductions for priority watersheds.
Watershed Current Total P load Desired Total P reduction1 Current Total N load Desired Total N reduction1
Taylor Creek/
Nubbin Slough
(S-191)
S-2
S-3
Kissimmee
River (C-38)
Harney Pond
Canal (S-71)
Rsheating Creek
S-4
Totals
Tons
189
35
7
108
47
65
15
466
(655)2
Tons
168
17
—
33
28
14
8
268
% for
Watershed
89
49
—
31
60
22
53
% of
Total
25.6
2.6
—
5.0
4.3
2.1
1.2
40.8
(40)3
Tons
479
1548
373
997
323
575
142
4437
(8148)2
Tons
302
1392
278
354
158
141
80
2705
% for
Watershed
63
90
75
36
49
25
56
%of
Total
3.7
17.1
3.4
4.3
1.9
1.7
1.0
33.1
(34)3
1 Based on drainage area load allocation.
2 Total load from all sources including rainfall, upper Kissimmee chain of lakes,
Lake Istokpoga, & other minor sources.
3 Overall target reduction levels based on Tech. Pub. 81-2.
1. Regional and sub-regional storage of runoff in each ma-
jor tributary area.
2. Diversions of flow to other basins from selected
tributaries.
3. Conventional and reverse osmosis (R/O) treatment
plants.
4. A number of Best Management Practices (BMPs) in-
cluding on-site runoff storage.
Costs, nutrient reduction potential, and the impact on Lake
Okeechobee's water budget were determined for each alter-
native within the seven priority watersheds except the Kissim-
mee River and most of the BMPs. Several options for reduc-
ing nutrient loads are now being considered by the U.S. Ar-
my Corps of Engineers for the Kissimmee River through their
restudy of that watershed. Since this effort is still underway,
it was deemed inappropriate to completely analyze Kissim-
mee River alternatives. It was determined that quantitative
evaluation of BMPs (except on-site storage of runoff) could
not be performed because of inadequate information regar-
ding their nutrient removal effectiveness. These BMPs are
common sense management techniques that could be us-
ed in conjunction with on-site storage to reduce off-site
nutrient loadings, as more data become available regarding
their nutrient treatment efficiencies.
The various alternatives were ranked according to cost
effectiveness (capital cost/amount of nutrient removed) for
each major watershed, then screened using the guidelines
and goals developed during the study. This process resulted
in a set of preferred alternatives depicted in Table 4. The
alternatives listed are the least-cost options that meet the
guidelines. Essentially, the proposed alternative in the Taylor
Creek basin involves on-site management of runoff by the
use of BMPs listed in Table 5 to achieve the desired load
reductions for individual land uses. This approach was
selected because:
1. It was the least costly alternative that also met all the
study guidelines.
2. Available data demonstrate this option has an excellent
potential for achieving high nutrient removal efficiencies.
3. BMPs can be combined with current drainage practices
with minimal impact on overall farming operations.
4. An institutional framework capable of implementing this
alternative already exists.
In the EAA (S-2 and S-3), regional storage and water
recycling using the state-owned "Holeyland" is the propos-
ed alternative.
There are several reasons for using this option:
1. Regional storage and water recycling is the least-cost
alternative that also meets the guidelines established dur-
ing the study.
2. Regional storage of runoff in the Holeyland provides
an additional water storage area for meeting a portion of the
water supply demands on Lake Okeechobee and WCA-3.
3. Regional storage has a greater probability of achiev-
ing nitrogen load reductions to Lake Okeechobee than on-
site storage since runoff would be physically diverted away
from the lake; otherwise it would be treated to some degree
and released back to the system through on-site storage.
4. Considerable preliminary work has already been ac-
complished regarding the Holeyland storage concept through
both the Special Project to Prevent the Eutrophication of Lake
Okeechobee and current activities of the Army Corps.
Specifically, the Holeyland area is being examined as a
possible additional water storage area in the Corps' Water
Supply Study for South Florida.
RECOMMENDATIONS
General Management Strategy
Implementing management actions in the Lake Okeechobee
region is a very ambitious endeavor; therefore, it is propos-
ed that a phased approach over a number of years be us-
ed. Phase 1 consists of the following elements:
1. Initiation and construction of the Holeyland project in
the EAA.
2. Acceleration of implementation of BMP programs in the
Taylor Creek/Nubbin Slough watershed.
3. Implementation of an expanded regulatory program that
includes water limitations for any new construction of
drainage systems in all areas tributary to Lake Okeechobee.
4. Continuation and completion of the Kissimmee River
Survey Review.
Until the Holeyland project is in place and operational, an
Interim Action Plan will be in effect. The Plan uses the flex-
ibility of the EAA water control system to reduce nutrient
315
-------
Lake Restoration, Protection and Management
Table 4. — Summary of preferred alternatives.
Watershed
Taylor Creek/
Nubbin Slough
(S-191)
S-2 & S-3
Harney Pond
Canal (S-71)
Fisheating
Creek
S-4
Kissimmee
River
(C-38)2
Totals
Capital cost
Alternative $ million
On-site 13.2
management
Regional 14.5
storage on
Holey land
On-site 9.1
management
On-site 12.9
management
On-site 9.4
management
On-site 30.9
management
90.0
Total P reduction
with controls
Tons °/o of total
169.8
38.2
28.8
30.8
4.6-
7.7
40.7
312.9-
316.0
25.9
5.8
4.4
4.7
0.7-
1.2
6.2
47.8
48.2
Total N reduction Net water losses
with controls °/° °f total
Tons % of total Acre-feet Lake inflow
302.7
1724.6
189.4
213.4
44.8-
74.6
493.2
2968.1-
2997.9
3.7 18,000
21 .2 90.6001
2.3 15,000
2.6 20,900
0.5- 5,000
0.9
6.1 67,800
36.4- 217,300
36.8
0.5
2.6
0.4
0.6
0.1
1.9
6.1
11rrigation demands on Lake Okeechobee would be reduced by about 60%; hence, the net loss of water would be
approximately 90,600 AF instead of the total amount of flows diverted away form the lake (approximately 226,500 AF).
2 This is only one of many alternatives currently being considered by the U.S.A.C.E. under the restudy of the
Kissimmee River and has not been selected as the least cost alternative. The figures are presented for
comparative purposes only.
loads to Lake Okeechobee by pumping excess floodwaters
south to the Water Conservation Areas rather than north to
the lake. This project is expected to take 5 years.
In the summer of 1981 the Okeechobee County
Agricultural Stabilization and Conservation Service (ASCS)
was awarded a $1.4 million Rural Clean Waters Program.
This program is designed to help defray the cost of im-
plementing BMPs in the Taylor Creek watershed. Additional-
ly, the State of Florida has appropriated $400,000 for a Taylor
Creek Headwaters Project which essentially parallels the
Rural Clean Waters Program. By directly administering the
Taylor Creek Project and working closely with the ASCS,
the District expects that a substantial portion of the water-
shed will experience water quality benefits from BMPs in the
next 5 years. Also, continued cooperation with the Corps of
Engineers and other agencies involved will assure comple-
tion of the Kissimmee River Restudy at an early date.
Table 5. — List of potential best management practices.
1. Treatment of barn, feedlot, and holding area stormwater runoff
by using oxidation/polishing lagoons.
2. Improved fertilizer management - by using soil testing and plant
analysis to avoid overapplication of fertilizer, timing and placement
of fertilizers to maximize plant uptake.
3. Biological nutrient removal - use of vegetated swales, ditches,
and/or shallow grassed waterways.
4. Dragging pastures and redistribution of barn and feedlot waste
to pasture areas.
5. Improved pasture management by rotating grazing areas and
periodically changing vegetative cover.
6. Fencing of waterways, in conjunction with appropriate place-
ment of salt, minerals, feed supplements, shaded areas, and watering
trough and tank sites away from waterways.
7. Conversion of barn and feedlot waste to methane gas for local
use.
8. Biological nutrient removal - use of water hyacinths in temporary
runoff storage lagoons for nutrient uptake.
Throughout the District, this agency presently regulates
existing and new agricultural and urban surface water
management systems. It is proposed to broaden the
regulatory activity to include water quality requirements for
new agricultural activity in areas tributary to Lake
Okeechobee. This will help prevent an increase in nutrient
loadings to the lake from the surrounding areas. New con-
struction would include modifications to existing systems (due
to more intensive land use or development of raw land) for
agricultural and urban purposes. Finally, Phase 1 includes
continuation of the District's existing water quality monitor-
ing program for Lake Okeechobee and the basins tributary
to it.
The conclusion of Phase 1 will mark a major milestone
and a "fork in the road." At that time, progress toward im-
plementation of management actions will be assessed to
determine what steps will be necessary in Phase 2. Among
the issues to be considered under Phase 2 are the following:
1. Should the District's current regulatory program be ex-
panded to include water quality control requirements for ex-
isting drainage systems in order to achieve compliance with
the load allocation?
2. How much further reduction in nutrient loading is
necessary from the tributaries other than S-191, S-2, and
S-3?
3. How effective have the management actions already
taken been in improving water quality?
REFERENCES
Carlson, R.E. 1977. A trophic state index for lakes. Limnol.
Oceanogr. 22:361-369.
Kratzer, C.R. 1979. Application of input-output models to Florida
lakes. M.E. Thesis, Univ. Florida, Gainesville.
Vollenweider, R.A. 1968. Water management research DAS/CS1/
68.27. Organ. Econ. Coop. Dev. Paris.
U.S. Environmental Protection Agency. 1974. An approach to a
relative trophic state index system for classifying lakes and reser-
voirs. Natl. Eutroph. Surv. Work. Pap. No. 475. Pacific Northw.
Environ. Res. Lab., Corvallis, Ore.
316
-------
A COMPARISON OF THE COSTS OF HARVESTING AND HERBICIDES
AND THEIR EFFECTIVENESS IN NUTRIENT REMOVAL AND CONTROL
OF MACROPHYTE BIOMASS
DIANE L. CONYERS
G. DENNIS COOKE
Department of Biological Sciences
Kent State University
Kent, Ohio
ABSTRACT
Mechanical harvesting and herbicide treatment with Diquat and Cutrine-Plus were compared in side-by-
side test plots in East Twin Lake, Ohio to evaluate their costs and their effectiveness in controlling plant
biomass and in removing lake nutrients. After two treatments the biomass in the harvested plot was
significantly reduced and was significantly less than the biomass in the control plot. The biomass in the
chemically treated plot followed the same seasonal growth pattern as the control plot because treated
plants were replaced by other species. Forty-six to 100 percent of the annual net loading and 13 to 22
percent of the annual gross loading of P to East Twin Lake for 1972-1976 could have been removed by
harvesting 50 percent of the littoral area. Chemical treatments do not remove plant nutrients. The initial
year costs of harvesting and herbicide treatment were identical. Harvesting would be far less expensive
in subsequent years.
INTRODUCTION
Any lake in which there is an accumulation of rich littoral
sediments has the potential of being plagued with a nuisance
growth of macrophytes. Anatomical differences among
submerged macrophytes, specifically the presence or
absence of root systems, determine the physical medium
from which these plants extract their nutrients. Although there
seems to be no morphological or anatomical reason why
shoots or roots should not absorb solutes (Denny, 1980),
plants with roots extract a high percentage of their needed
growth nutrients from the sediments (Carignan and Kalff,
1980; Barko and Smart, 1980). Those without roots fulfill
nutritional needs by absorbing nutrients from the water (Den-
ny, 1972).
Regardless of the source, these nutrients along with
dissolved and particulate organic matter will be released in-
to the water column as plants decompose. Oxidation of the
plant detritus may bring about an oxygen deficit and cause
further release of nutrients from reduced sediments. This
enrichment of lake water with nutrients can stimulate the
growth of algae and more macrophytes. Thick beds of
macrophytes will also accelerate accretion of the littoral zone
as their tissues accumulate at the sediment surface after
senescence. Further discussion of this important positive
feedback system is found in Wetzel (1975) and Rich and
Wetzel (1978).
The result of macrophyte decomposition is not the main
concern of lake users. Their interest lies in the accessibility
of the lake for fishing, swimming, and boating. However, both
of these views emphasize the need for ecologically sound
management of macrophytes.
The most common method of treating nuisance aquatic
plants is to apply herbicides. Although they may temporari-
ly retard a macrophyte nuisance, herbicides have the follow-
ing drawbacks, as listed in Conyers and Cooke (1982):
1 Plants release nutrients upon death and decomposi-
tion (Simsiman et al. 1972; Morris and Jarman, 1981);
2 Oxygen is depleted at the sediment surface by microbes
that colonize decaying plants; this may be followed by a
release of nutrients from these fertile sediments (Simsiman
et al. 1972; Rich and Wetzel, 1978; Carpenter and Greenlee,
1981);
3. Herbicides can be toxic to nontarget species (Ander-
son, 1981);
4. Some plant species may be tolerant to the herbicide
(McKnight, 1981);
5. Some herbicides are suspected to be mutagenic and/or
carcinogenic (Shearer, 1980; Valencia, 1981); and
6. The waiting period (10 or more days in some cases)
following the use of many herbicides interferes with the
recreational use of the lake.
Because herbicide use has these drawbacks, other
methods of macrophyte control are being sought or re-
fined. Harvesting is one of these. A mechanical harvester
is a machine that cuts aquatic plants to a maximum depth
of 1.52 m (5 ft.) and removes them from the water before
nutrient release, and oxygen consumption following plant
death can occur. Drawbacks to harvesting can include:
1. Stimulation of an algal bloom;
2. Physical disruption of the environment and resuspen-
sion of sediments;
3. Removal of young fish;
4. Habitat removal for epibiota and macroinvertebrates;
and
5. High initial cost.
We compared mechanical harvesting with the use of
chemicals. The purposes of our study were (1) to evaluate
and compare the efficacy and cost of these two methods
of macrophyte control; and (2) to determine the optimal time
to harvest the largest quantity of plant nutrients.
The study was done on East Twin Lake in Kent, Ohio (see
Cooke et al. 1978 for a description of these mesotrophic
glacial lakes). East Twin Lake has an area of 26.9 hectares
(66 acres) with an extensive littoral zone of 11.7 hectares
(28.9 acres). The predominant plant species in the study area
were Potamogeton crispus (curly-leaf pondweed) in the
spring, Ceratophyllum demersum (coontail) throughout the
317
-------
Lake Restoration, Protection and Management
summer, and Chara (stonewort) in late summer and early
fall. There were patches of Nymphaea (water lily),
Myriopbyllum (water milfoil), Najas, and Potamogeton fdiosus
throughout the summer.
METHODS
Three 810 m2 (0.2 acres) study plots 20 m apart were
marked with floating buoys.,The depth of the water in the
plots was approximately 0.5 m at the shore end increasing
to 1.6 m at the deep end. The mean water depth was 1.1
m (3.5 ft). Plant species composition was similar in the three
plots, except Nymphaea was absent in the control plot. Most
of the sediment within the plots (approximately 85 percent)
had dense plant coverage and the remaining 15 percent was
sparsely covered or void of vegetation.
Sampling began on June 6,1981, and continued until Oc-
tober 3, 1981. Three random 0.25 m2 samples were taken
weekly from each plot, using scuba. In the laboratory the
plants were washed, dried, weighed, and ashed to deter-
mine the weekly standing organic biomass in each plot. After
the dried plant material had been ground and thoroughly
mixed, three sub-samples were taken to determine the P
content of the plant tissues (Andersen, 1976).
Following four weeks of pre-treatment sampling the
treatments began. On June 28 plants were harvested with
an Aquamarine Chub Harvester (Waukesha, Wls.) and
removed from Plot A. This machine cuts to a maximum depth
of 1.5 m (5 feet). The biomass of the harvested plants from
Plot A amounted to 65 kg. dry weight (143 Ibs.). This was
a large pickup truck load when wet. Seven weeks later
(August 18) this plot was harvested again and 3.8 kg. dry
weight (8.4 Ibs.) of plants were removed (a small wheelbar-
row load).
The middle plot (B) was kept as a control and sampled
weekly to observe the natural seasonal succession of plants
and plant biomass.
The third plot (C) was treated with a tank mix of the
algaecide Cutrine-Plus and the herbicide Diquat. The
chemicals were applied at 0600; surface water temperature
was 28.3°C (83°F) and the lake was calm. We mixed the
chemicals in an 8-liter hand sprayer and slowly rowed back
and forth spraying the entire surface of the 810 m2 (.2 acre)
plot. The chemicals and their dose were chosen and applied
following consultations and recommendations of Applied
Biochemists, Inc., Mequon, Wis. The dose of Cutrine-Plus
was 1,2 gal./acre-foot; the dose of Diquat was 1.4 gaUacre-
foot (Applied Biochemists, Inc., 1976). The total amounts of
chemicals applied were 0.84 gal. (3.6 I) of Cutrine-Plus (.2
surf, acres x 3.5 ft. mean depth = 0.7 acre/ft.; 1.2 gal./acre
ft x 0 7 acre ft. = 0.84 gal.) and 0.98 gal. (3.71) of Diquat
(1.4 gaL/acre ft. x 0.7 acre ft. = 0.98 gal.). Inclement weather
prevented treatment until July 3.
Two weeks after the treatment the biomass samples were
not reduced nor could we observe any other effect of the
treatment. Therefore, the chemicals were reapplied. After
both applications, use of the lake was restricted for 10 days
(following the Diquat manufacturer's warning). Within 2
weeks after the second treatment we observed the sub-
merged species (Ceratophyllum) settling to the sediment and
breaking apart.
The early application of the second chemical treatment
created uneven intervals of time between the chemical and
the harvester treatments (Table 1). As a result, the pre-
treatment, post-treatment 1 and post-treatment 2 biomass
data from the harvest and chemical plots were graphed
separately. The biomass for each time interval (pre, post-1,
and post-2) was compared to the biomass in the control plot
for corresponding periods of time (Fig. 1).
I
A 0 A B A B
prt POll 1 po*> 2
Ill
A B » 8 * B
prc port 1 port 3
I
C 6
P0it1
J
post 1 pt»t~~2 pro pail 1 po«l j
Rgure 1 .-Pre- and pest-treatment biomass of aquatic macrophytes
(gms. dry wUm2) in control, harvested, and chemically treated plots
in East Twin Lake, Ohio. A = harvest plot, B = control plot,
C = chemical plot. Depth of water in shallow sampling area ranged
from 0.5 to 1.0 meters; in deep area, 1.0 to 1.6 meters.
Table 1. — Catalog of sampling Intervals and number of biomass samples for harvested (A) and chemically
treated (C) plots.
Treatment
Plot A
Pre-treatment
Post-treatment 1
Post-treatment 2
PlotC
Pre-treatment
Post-treatment 1
Post-treatment 2
Number
of weeks
4
6
6
4
2
10
Total
number of
biomass samples
12
18
18
12
6
30
Shallow
7
10
6
7
3
13
Deep
5
8
12
5
3
17
318
-------
Contributed Papers
Table 2. — Statistical analysis of pre- and post-treatment biomass samples (gms. dry wt./m2).
A. Results of Student's t-test for pre- and post-treatment analysis within each plot. Pre-treatment values are the mean of samples
taken from weeks 1-4. Post-treatment samples are the mean of samples from weeks 13-16.
Control Chemical
Plot
Harvest
Pre-treatment
Mean
S.E.
Post-treatment
Mean
S.E.
d.f.
Ho: p pre = ^ post
109
30
24
8
22
p<.05
167
24
196
69
22
p>.10
193
40
141
25
22
p>,25
B. Results of one-way ANOVA among plots for pre- and post-treatment biomass (gms-dry wt-/m2).
PRE
POST
Source of
variance
Treatments
Error
Total
Treatments
Error
Total
Sums of
squares
44615.4
410146.9
454762.3
185014.4
757547.2
942561 .6
d.f.
2
33
35
2
33
35
Mean
square
22307.7
12428.7
92507.2
22956.0
F P
1.79 0.18
4.03 0.03
Statistical analysis among plots (Table 2) was performed
on pre-treatment data (weeks 1-4) and post-treatment data
(weeks 13-16). Weeks 13 to 16 were used to represent post-
treatment biomass because by the 13th week both chemical
and harvest plots had been treated twice.
RESULTS AND DISCUSSION
Graphs of pre- and post-treatment data illustrate the effect
of harvesting and chemical treatment on the plant biomass
(Fig. 1). The sampling intervals and number of samples taken
are listed in Table 1. The biomass in the harvested plot (A)
decreased in shallow and deep water after the first and se-
cond harvest, while the biomass in the control plot (B) had
a normal seasonal increase and senescence toward the end
of summer (Fig. 1). Following harvests a slow regrowth of
plants occurred. The growth of plants could have been in-
hibited because plants had been cut at the sediment sur-
face and root crowns were removed or damaged. Another
explanation for slow regrowth is that Ceratophyllum, the
predominant macrophyte in the plots, has no roots. Without
a root stock to propagate new plants, rootless species must
depend on shoot fragments or seeds for re-infestation.
The biomass in the chemically treated plot (C) followed
the same seasonal growth as the control plot (Fig. 1). Even
though the effect of the herbicide treatment could be seen,
the biomass (dry wt.) did not differ significantly from the con-
trol plot (Table 2). This was due to a bloom of the calcareous
alga Chara sp., a serious nuisance plant to swimmers and
other lake users. Chara was present in small amounts prior
to the treatment, but seemed to be resistant to the chemicals.
As expected, Nymphaea was unaffected by these chemicals
and its biomass was included in measurements in the
chemical and harvested plots. Chemical control of this plant
would require an additional herbicide.
An analysis of pre- and post-treatment data within each
plot was performed using Student's t-test (Table 2A). Results
show that after two cuttings the plants in the harvest plot
significantly decreased (p< .05); whereas, no statistically
significant difference was found between pre- and post-
treatment biomass in the control (p>.10) or chemically
treated (p > .25) plots. Biomass data from all plots were
analyzed using a one-way analysis of variance (ANOVA) to
determine if the three plots differed significantly before and
after the treatments (Table 2B). The F value did not indicate
a significant difference among pre-treatment plots (p = .18);
however, the F value for post-treatment shows a significant
difference (p< .05). To determine which of the mean values
were significantly different we used the Student-Newman-
Kuels procedure using the range as the statistic to measure
differences among means (Sokal and Rohlf, 1969). Results
of the Student-Neumann-Kuels test showed that the
statistically significant difference was between the harvest
and control plots (p •< .05), and the harvest and chemical
plots (p< .10), but not between the chemical and control
plots.
The possibility of harvesting having restorative value to
East Twin Lake was explored. The weekly phosphorus con-
tent per gram of plant tissue (Andersen, 1976) was compared
with the standing biomass per m2 (Fig. 2) to determine when
maximal nutrient removal could be accomplished. The mean
quantity of phosphorus that could be removed by harvesting
the plants during July was compared to 4 years of annual
phosphorus loading data (Cooke et al. 1978). Our results fell
well within the range for gms- dry wt-/m2 and gms-P/m2
compiled by Burton et al. (1978) for submerged macrophytes
in the northern United States. We found that harvesting the
entire littoral area of East Twin Lake (11.7 ha.) one time would
remove 26 to 44 percent of the total gross phosphorus
loading and 92 to 100 percent of the annual net phosphorus
loading (Table 3). A more realistic consideration would be
to harvest 50 percent of the littoral zone. In this case, a range
of 13 to 22 percent of the total gross phosphorus loading
and 46 to 100 percent of the net phosphorus loading would
have been removed from the lake by harvesting the plants.
319
-------
Lake Restoration, Protection and Management
Table 3. — The percent of annual gross phosphorus income and net phosphorus loading which can be removed by
harvesting 100 percent and 50 percent of the littoral zone of East Twin Lake. The plants growing in the littoral area con-
tain a total of 56 kg. of phosphorus (.0015 gms P/gm dry wt.x 320 gms. dry wt /m2 x 104 m^/ha. x 11.7 ha. x .001 g/Kg. -
56 kg. P). Phosphorus content of plants, plant density, and areal coverage are based on 1981 data.
1972
1973
1974
1975
1976
Annual gross P income (kg.)
Percent removed by 100% harvest
Percent removed by 50% harvest
Net P loading (kg.)
Percent removed by 100% harvest
Percent removed by 50% harvest
180.8
31
15
38.5
:>100
74
126.8
44
22
17.8
:>100
100
219.4
26
13
62
92
46
218.4
26
13
40.9
100
70
127.4
44
22
-8.1
>100
100
DMT WIIOHT
PHOSPHOflUB
0.* a
O
'•' i
M
>.. 1,.
Rgure 2.—Seasonal biomass (gms. dry wl/m2) and total phosphorus
content (gms./m2) for plants in the control plot in East Twin Lake,
Ohio. Phosphorus per m2 is the product of plant biomass x the total
phosphorus concentration per gram (dry wt.) of plant.
The costs for two chemical and two harvest treatments
are shown for East Twin Lake and the adjacent West Twin
Lake (Table 4). Our results of an additive comparison of the
costs of the harvester plus summer operation expenditures
for harvesting half the littoral area (courtesy of the Twin Lakes
Assoc.), with the cost that would occur with two Cutrine and
one Diquat treatment of half of the littoral areas of both lakes
(recommendation of Applied Biochemists, Inc.) show that in
the second season harvesting would be less expensive than
chemical treatment of both lakes. Over a 5-year period the
cost of macrophyte control with chemicals would be 2.6 times
the cost of controlling the plants with a harvester (Fig. 3 and
Table 4). Chemical control would be even more costly if
plants such as Nymphaea (water lily) were to be controlled
and if we had included the herbicide applicator's fee in our
computation. Harvesting is more expensive in the first year
because of the initial cost of the machine.
CONCLUSIONS
Our conclusions are: (1) Harvesting is much more effective
than the recommended doses of Cutrine-Plus and Diquat
in controlling plant biomass in East Twin Lake; (2) maximum
amounts of nutrients are removed from the lake by harvesting
the plants during peak biomass; (3) harvesting will be less
costly than chemical treatment over a 2-year period; and (4)
harvesting will eliminate plant nuisances which limit recrea-
tional activities.
Harvesting can disrupt aquatic biota since plant removal
constitutes habitat removal. However, this method of
leo.ooo
S 70,000
S 50,000
t 30,000
$10,000
PROJECTED EXPENDITURES
FOR 1981-19B5
CHEMICALS -C
HARVESTING -H
I
C H
1981
C H
1982
C H
1983
C H
1984
C H
1985
Figure 3.—Projected expenditures for chemical and harvesting
treatments of aquatic macrophytes in East and West Twin Lakes,
Ohio, based upon 1981 harvester purchase and operating costs.
Chemical costs do not include expenses for application, equipment,
and liability insurance, only 1981 costs of chemicals. Data are from
Table 4. Projected costs do not include price increases.
nuisance macrophyte control appears to be far more effec-
tive than chemical treatment with Diquat and Cutrine-Plus,
longer lasting, less costly, restorative, and without the well-
known side effects that can occur with toxic chemicals. Ad-
ditionally, plant removal interrupts the release of nutrients
by macrophytes and the buildup of the lake's littoral zone.
ACKNOWLEDGEMENTS: We thank Applied Biochemists, Inc. (Me-
quon, Wis.) for providing the Cultrine-Plus and Diquat for this ex-
periment. The Aquamarine Corp. (Waujcesha, Wis.) prpvided par-
tial support of our travel to Vancouver, B.C. We are grateful to the
Twin Lakes Association for granting us permission to carry out this
study on their lakes and for their generous assistance. We also thank
320
-------
Contributed Papers
David WaJler, assistant professor of Biological Sciences at Kent State
University, for help in data analysis. We are grateful to Ron Raschke
for reviewing the manuscript.
Table 4. — Cost summary for harvesting and chemical treat-
ment of half of the littoral area (13.6 hectares or 33.5 acres)
of East and West Twin Lakes, Ohio. Dose was that recom-
mended by Applied Biochemists (see reference). Following
their recommendations, we assumed the minimum treatment
program of two Cutrine-Plus applications and one Diquat ap-
plication per summer. Cutrine dose was 1.2 gal./acre foot,
Diquat was 1.4 gal./acre-foot, based on 134 acre-feet, and
1981 prices of $22.60/gal. for Cutrine and S56.15/gal. for Di-
quat. Herbicide operator fee is not included.
Chemical treatment $17,802
Harvesting1
Initial investment for harvester and trailer 15,000
1981 Operations
gasoline 350
maintenance 230
auxiliary equipment 80
personnel 3,152
$ 3,812
TOTAL $18,812
' Data provided by the Twin Lakes Association, Kent, Ohio.
Note: The revised 1979 edition of "How to identify and control water weeds and algae"
(Applied Biochemists, Inc., Mequon, Wisconsin) recommends a dose of 2 gallons of
Diquat per acre instead of the 1976 edition's recommendation of 1.4 gallons per acre-
foot for control of Ceratophytlum. We were advised by Applied Biochemists to use the
dose in the 1976 edition. If we had computed the herbicide cost per year with the dose
in the 1979 edition, the per year cost would have been $9,818.85 for both the Diquat
and Cutrine-Plus, excluding herbicide operator fee and assuming no increase in her-
bicide cost in subsequent years. The harvester will pay for itself in the second year
in either case.
REFERENCES
Andersen, J. M. 1976. An ignition method for determination of
total phosphorus in lake sediments. Water Res. 10:329-331.
Anderson, L. W. J. 1981. Control of aquatic weeds with hexazinone.
J. Aquat. Plant Manage. 19:9-14.
Applied Biochemists, Inc. 1976. How to identify and control water
weeds and algae. A guide to water management. Applied
Biochemists, Mequon, Wis.
Barko, J. W., and R. M. Smart. 1980. Mobilization of sediment
phosphorus by submersed freshwater macrophytes. Freshw. Biol.
10:229-238.
Burton, T. M., D. L. King, and J. L. Ervin. 1978. Aquatic plant har-
vesting as a lake restoration technique. Pages 177-185 in Lake
Restoration. Proc. Natl. Conf. EPA 440/5-79-001. U.S. Environ.
Prot. Agency, Washington, D.C.
Carignan, R., and J. Kalff. 1980. Phosphorus sources for aquatic
weeds: Water or sediments? Science 207:987-989.
Carpenter, S. R., and J. K. Greenlee. 1981. Lake deoxygenation
after herbicide use: A simulation model analysis. Aquat. Bot.
11:173-186.
Conyers, D. L., and G. D. Cooke. 1982. Comparing methods for
macrophyte control. Chemical or mechanical? Lake Line 2(2):8-10.
Cooke, G. D., R. T. Heath, R. H. Kennedy, and M. R. McComas.
1978. Effects of diversion and alum application on two eutrophic
lakes. EPA-600/3-78-033. U.S. Environ. Prot. Agency, Washington,
D.C.
Denny, P. 1972. Sites of nutrient absorption in aquatic macrophytes.
J. Ecol. 60:819-829.
1980. Solute movement in submerged angiosperms. Biol.
Rev. 55:65-92.
McKnight, D. 1981. Chemical and biological processes controlling
the response of a freshwater ecosystem to copper stress: A field
study of the CuSO4 treatment of Mill Pond Reservoir, Burlington,
Mass. Limnol. Oceanogr. 26:518-531.
Morris, K., and R. Jarman. 1981. Evaluation of water quality during
herbicide applications to Kerr Lake, Okla. J. Aquat. Plant Manage.
19:15-18.
Rich, P. H., and R. G. Wetzel. 1978. Detritus in the lake ecosystem.
Am. Nat. 112:57-71.
Shearer, R. 1980. Public health effects of the aquatic use of herbi-
cides 2, 4-D, dichlobenil, endothall, and diquat. Pages 2-57 (Sec.
1) in Literature Reviews of Four Selected Herbicides: 2, 4-D,
Dichlobenil, Diquat and Endothall. Munic. Metro. Seattle (METRO),
Seattle, Wash. Unpubl. mss.
Simsiman, G. V., G. Chesters, and T. C. Daniel. 1972. Chemical
control of aquatic weeds and its effect on the nutrient and redox
status of water and sediment. Proc. 15th Conf. Great Lakes Res.
1972:166-180.
Sokal, R. R., and F. J. Rohlf. 1969. Biometry. W. H. Freeman and
Co., San Francisco.
Valencia, R. 1981. Mutagenesis screening of pesticides using
Drosophila. EPA-600/S1-81-017. U.S. Environ. Prot. Agency,
Washington, D.C.
Wetzel, R. G. 1975. Limnology. W. B. Saunders Co., Philadelphia.
321
-------
AN ANNUAL INCREMENT LAKE ACIDIFICATION—FISHERIES
RESPONSE MODEL TO GEOLOGICALLY SENSITIVE AREAS
IN ONTARIO, CANADA
J. E. HANNA
J.E. Hanna Associates Inc.,
Pickering, Ontario
M. F. P MICHALSKI
Hough, Stansbury & Michalski
Rexdale, Ontario
PREFACE
The following paper is a summary from an internal report
entitled: An Approach to Assessing the Effects of Acid Rain
on Ontario's Inland Sports Fisheries. The report was prepared
for the Department of Fisheries and Oceans by Hough,
Stansbury, & Michalski Ltd., and J.E. Hanna Associates Inc.
The Department of Fisheries and Oceans has a respon-
sibility for protecting and conserving Canadian fish stocks
currently threatened by acid rain. Therefore it was deter-
mined that an acid rain/fisheries model could be useful in
predicting future damage to Canadian fisheries.
From the beginning it was recognized that the complexi-
ty of the air/soil/biota interactions and the limited data base
presently available would not permit accurate predictions of
fisheries impacts. However, it was believed that by attemp-
ting such a synthesis, information would be obtained early
in the research program that would identify critical gaps in
fisheries' knowledge of the acidification process. Therefore
at the outset it was anticipated that the important output from
this exercise would be the methodology arising from the ac-
cumulation .and synthesis of available concepts and not the
prediction of impact. The experimental approach or model
conceived was derived from an analysis of the Ontario situa-
tion. This area was selected for a variety of reasons including
the fact that reliable fisheries data were available from the
Ontario Ministry of Natural Resources.
The final report submitted by the contractors was sent out
for critical review and comment to approximately 30 scien-
tists with knowledge of the scientific aspects of acidification.
Their observations and those of DFO's Steering Committee
set up to oversee the contract are summarized as follows:
1. The most common concern expressed by the reviewers
was the need to "calibrate the model" before it could be
used for predictions of change. Observations were made on
the difficulty in verifying many relationships used in the
model, especially between the Morpho-Edaphic Index (MEI),
alkalinity and production of fish, and in the physical/chemical
portions.
2. Some specific concerns expressed by reviewers related
to the:
• lack of treatment of lake-sediment interactions,
• influence of labile aluminum on fish,
• influence of episodic events (heavy rain, snow melt)
• role of organic acids in lake acidity,
• treatment of groundwater influences in the model
• reliability of literature values of "threshold lethal pH"
in fish.
Because of these and other limitations the DFO Steering
Committee that oversaw the work of the contractors does
not endorse the predictions of fisheries losses from pilot runs
of the model. Efforts are proceeding to refine the model and
to eliminate or reduce the uncertainties inherent in its
formulation.
INTRODUCTION
Scientists and resource managers generally agree that many
technical elements relating to the interactions between
deposition of atmospherically-transported mineral acids and
aquatic ecosystems are not yet fully understood. However,
it has been said of research that for every question answered,
five more will arise that require investigation. Many resear-
chers believe it will take 20 to 30 years of experimentation
and data collection before we can significantly refine our
knowledge of the acidification process. On the other hand,
the National Academy of Sciences, after reviewing all the
available information on the subject, concluded that suffi-
cient evidence currently exists to justify significant preven-
tive action (Natl. Acad. Sci. 1981). Regardless of the time
frame, policymakers and society in general must still decide
on the level of precision or, conversely, the amount of risk
or uncertainty that is acceptable for a particular decision; in
fact, inaction is a decision in itself.
A decision to reduce acid-causing emissions may have
profound effects both economically and socially; in this in-
stance, relatively precise cost estimates can be made for
alternative strategies. Conversely, a decision to maintain or
increase current emissions could cause extensive ecological
damage and associated economic and social disruption,
albeit, to a different region. In either case, a comprehensive
quantitative analysis for assessing the potential benefits or
damages is not yet in place.
To provide a means to estimate the potential damages
to inland fisheries in geologically sensitive portions of On-
tario and other parts of Eastern Canada, the Canada Depart-
ment of Fisheries and Oceans began to develop a long-term
lake acidification-fisheries response model. The purpose of
this paper is to provide an overview of the model, and to
discuss some of the key assumptions in the methodology.
A number of attempts have been undertaken to develop
predictive models primarily for the hydrologic-chemical
aspects of the lake acidification phenomenon (Environ. Can.,
1981; Conroy et al. 1974; Hendriksen, 1980; Hesslein, 1979).
While trophic-fisheries responses have been described for
322
-------
Contributed Papers
some cases (Aimer, 1972; Aimer et al. 1978; Schindler et
al. 1980; Schofield, 1976; Beamish and Harvey, 1972; Keller,
1978; Watt et al. 1979; Farmer et al. 1980; Mills, 1983),
generally, less work has been devoted to modeling these
more complex effects. Formulation of most hydrologic-
chemical models has been effected through empirical
analyses with partial theoretical explanations of relationships
(e.g. Henriksen, 1979; Environ Can. 1981). Comprehensive
mass balance studies are underway in many locations such
as Kenora, Sault Ste. Marie, Dorset and Sudbury in Ontario,
Laurentide Park in Quebec, Kejimikujit Park in Nova Scotia,
Hubbard Brook in New Hampshire, Sagamore Lake in New
York, and Coweeta in North Carolina (Ontario Ministry En-
viron., 1981; Dillon, 1982). These will offer potential for con-
structing calibrated cause-effect models with strong
theoretical foundations. However, in some instances (e.g.
the Dorset Lakes Study Area and ILWAS) the lack of cor-
roborative fisheries information will tend to confound exten-
ding the analysis to a primary ultimate point of concern—
fisheries production and survival. The essential questions
to ask about whole lake-watershed system monitoring are:
1. When will there be sufficient information to construct
a rigorous lake-watershed-biological response model
suitable for regional analysis?
2. What is the best course of action in the interim?
3. Do the data bases currently being assembled include
all significant control parameters affecting lake and fishery
responses to acidification?
We are of the opinion that a reasonably full accounting
of key relationships can be achieved when one constructs
an integrated model, and that an appreciation of sensitive
control parameters can best be gained through preliminary
testing of such prototype models. Results can also provide
the basis for immediate and interim decisions that are
necessary until complete data bases have been compiled.
Several attempts at constructing integrated
lake-watershed acidification models have been documented
(Overrein et al. 1980; Andrews et al. 1980; Schreiber et al.
1982). However, specific policy recommendations based on
predicted results are absent from these works; in fact,
Schreiber et al. explicitly state that their model is not suitable
for predictive applications (pg. viii, para. 2). Unfortunately,
no alternative is provided to decisionmakers or the general
public as to an appropriate interim course of analysis and
action and no indication is given as to when to expect a "suf-
ficiently" calibrated model(s) for regional application.
The model described in this paper was designed principal-
ly for regional application and to provide predictive results.
The model may be modified and refined for lake-specific
applications.
A MODEL OVERVIEW
The hydrologic-chemical model used was designed for long-
term predictions of lake alkalinity changes through annual
chemical and water mass balance reactions (Hough Stans.
Mich., Hanna Assoc., 1982). The structure is similar to many
lake nutrient models (e.g. Vollenweider, 1969; Dillon, 1974;
Schindler, 1980; Oglesby, 1977; Reckhow, 1979) in that
sources and annual loadings of alkalinity and acid are iden-
tified in the system (Fig. 1), and complete mixing and an-
nual flushing of the lake are assumed. Four key sources of
alkalinity are: (1) soil cation exchange with incoming
atmospherically-deposited mineral acids; (2) ground water;
(3) internally-generated alkalinity from primary production,
sediment exchange, and sulfur reduction; and (4) residual
lake alkalinity.
These loadings are integrated to produce an alkalinity store
for a lake/watershed system which is available to neutralize
annual hydrogen ion input from the atmosphere. Alkalinity
Figure 1 .—Schematic of alkalinity mass balance budget concept.
and hydrogen ion loads are mixed using equivalent
measures, and the net result is converted to a lake concen-
tration and pH by including a hydrologic budget.
The biological component of the model is based on two
elements: a modified morphoedaphic index (Ryder et al.
1974), and observed lethal acid concentration thresholds (Fig.
2). Regarding the first element, Ryder (1963) pointed out that
alkalinity accounts for most of the total dissolved solids (a
significantly correlated variable with fish productivity) in many
natural unacidified waters. Accordingly, alkalinity was
substituted for total dissolved solids in Ryder's basic MEI
equation. The premise of the approach is that as alkalinity
is depleted by acidification, the long-term sustainable yield
of a water body will also decline since the MEI is reduced.
This response is termed a clinical or sublethal loss, and it
predicts a gradual reduction in productivity.
The second element results from lethal concentrations of
H+ ions and aluminum, and is predicted as a threshold
response above which given species of fish will be lost from
the fish community (Table 1). Nine fish assemblages were
identified for Ontario (Table 2). As is the practice of fisheries
biologists in the Province, a portion of the total annual pro-
duction estimated by the MEI is partitioned to the gross game
fish component within a community (Ontario Ministry Nat.
Resour. 1982); this yield was then distributed proportionately
to each individual game fish species present (Table 3). As
a particular species is lost (i.e., the species disappears from
a community when its threshold pH is reached), a portion
of the yield accounted for by that species is then redistributed
among remaining species (Table 4). This redistribution of pro-
ductivity is repeated until the last game fish species is re-
323
-------
Lake Restoration, Protection and Management
moved. At this point, the lake is termed "extinct." Extinct
lakes are those no longer capable of game fish production,
although overall primary and secondary production may still
be relatively high.
The model reports annual potential fish biomass pro-
duced, and the number and type of fish communities that
have been lost. Total biomass production is reported for each
species within a given assemblage for preset time intervals.
To operate the model for each lake, the following infor-
mation needs to be collected: current acid loading; soil depth,
Clinal {pre-lethal) response
Threshold pH
Threshold (lethal)
response
Lake alkalinity/pH
Figure 2.—Schematic fish species response curve to decreasing lake
alkalinity/pH.
base saturation, cation exchange capacity, bulk density, and
hydraulic velocity; annual precipitation, runoff, and evapora-
tion rates; lake surface area, mean depth, fish community
type, and lake trophic state; and lake watershed area. The
atmospheric, geological, and pedological information can be
interpreted from various published reports and maps (e.g.,
Hoffman et al. 1964; Ontario Ministry Agric. Food, 1975; On-
tario Dep. Lands Forests). Lake surface and drainage
areas can be derived from 1:50,000 or 1:250,000 topo-
graphical maps depending on which are available. Lake
mean depth and fish community are available for approx-
imately 2 percent of the lakes from published reports and
file data. Based on the locations of sampled lakes in a water-
shed, neighboring fish communities, and topography of the
shoreline, mean depth and fish communities can be approx-
imated for the remainder of the lakes. Our experience has
been that these approximations can be reviewed and
modified and/or confirmed by local fisheries managers
based on their personal field experience.
DISCUSSION
The model developed primarily addresses the issue of long-
term acidification. Short-term acid pulses during periods of
high runoff are not explicitly considered, although the
threshold lake pH values were derived from field observa-
tions and are therefore partly representative of fish responses
Table 1. — Threshold of pH cut-off points for the five game fish species of interest in this study.
Fish Species
Bass
Lake trout and
brook trout
Yellow pickerel
Northern pike
pH
5.7
5.5
5.5
4.7
Effects
Below pH of 5.5 - 6.0 reproduction success fails
At pH of 5.5 and below, lake trout reproduction fails and below
a pH of 4.7, populations cannot be maintained. Brook trout
reproduction success is adversely affected at pH 5.5 and
below.
Found absent at a pH of 5.2 - 5.8
Relatively tolerant, with reproduction and hatchability of eggs
seriously impaired at pH 4.5 - 5.0
References
Beamish, 1975, 1976
Lewis and Peters, 1956
Beamish et al. 1976
Menendez, 1976
Beamish, 1974
Keller, 1978
Beamish, 1975, 1976
Beamish, 1975, 1976
Beamish and Harvey, 1972
Vallin, 1953
Table 2. — Fish assemblage categories and the proportion of sustainable yield which can be attributed to the game fish
component.
Game fish assemblage
category
% of total annual sustainable
yield allocated to game fish
1 Northern pike, bass and yellow pickerel or bass only, or, bass and
yellow pickerel, or, northern pike and bass
2. Northern pike only
3. Northern pike and yellow pickerel, or, yellow pickerel only
4. Lake trout, northern pike, bass and yellow pickerel, or, lake trout,
northern pike and bass, or, lake trout and bass
5. Lake trout only
6. Lake trout, northern pike and yellow pickerel, or, lake trout and pike, or,
lake trout and yellow pickerel
7. Brook trout only
8. Brook trout, lake trout and bass, or, brook trout and bass, or, brook trout
and lake trout
9. No game fish
40%
40%
40%
40%
25%
30%
30%
30%
0%
324
-------
to annual variations in acid concentration (i.e., acid pulse
effects). Similarly, the effects of toxic elements such as
aluminum have not been explicitly included; again, however,
the thresholds observed in the field were partly a function
of toxic effects of aluminum and other heavy metals. The
direct inclusion of an acid pulse component would certainly
lead_to larger predicted impacts, both in terms of lost pro-
duction and extinction of fish species; however, the
magnitude of such effects cannot be estimated without con-
structing an appropriate predictive model.
The model does not attempt to consider other pressures
on fish populations such as angling and commercial fishing.
If acidification were to be considered jointly with these other
pressures, for some lakes the effects could be more severe.
This consideration would be most important for lakes receiv-
ing a high level of use; from an economic viewpoint, these
lakes are also the most valuable and therefore a simple pro-
rating of the lost fish production would not be appropriate.
In fact, preliminary results indicate that the most seriously
Contributed Papers
affected areas are those parts of Ontario receiving some of
the highest use, suggesting that the consideration of angl-
ing use is significant.
A primary question which arises concerning the model is
whether the estimates of biological damages from long-term
lake acidification are likely too low or too high. The response
to this question has two parts: the first deals with the
hydrologic-chemical component and the second with the
fisheries response. In terms of the chemical impacts, ground
water and autochthonous alkalinity each are predicted to con-
tribute between 10 percent and 55 percent of the total lake
alkalinity supply, the actual value depending on watershed
and lake characteristics. Both are considered insensitive to
acidification and only runoff alkalinity and residual lake
alkalinity change with time. By altering absolute contributions
from these sources (for example, it has been suggested to
the authors that the autochthonous inputs should be reduc-
ed), the predicted number of lakes going extinct can be
dramatically varied. However, one must reconcile the fact
Table 3. — Proportion of sustainable yield allocated to each game fish species in eight assemblage categories.
Game fish assemblage
1 . Northern pike, bass and
2. Northern pike only
3. Northern pike and yellow
4. Lake trout, northern pike,
yellow pickerel
5. Lake trout only
6. Late trout, northern pike
pickerel
7, Brook trout only
yellow pickerel
pickerel
. bass and
and yellow
8. Brook trout, lake trout and bass
Northern Bass
pike
0.10 0.06
0.40 —
0.20 —
0.05 0.03
— —
0.06 —
— —
— 0.04
Lake
trout
—
—
—
0.20
0.25
0.15
0.30
—
Brook Yellow
trout pickerel
— 0.24
_ _
_ _
— 0.12
_ _
— 0.09
_ _
0.15 0.09
Table 4, — Yield re-distribution following elimination of fish species from each game fish assemblage.
Fish Community
First pH threshold
Second pH threshold
Third pH threshold
Species
lost
% MEI
uptake
Species
lost
% MEI
uptake
Species
lost
% MEI
uptake
1. Northern pike, bass and yellow pickerel
2, Northern pike only
3. Northern pike and yellow pickerel
4. Lake trout, northern pike, bass and
yellow pickerel
5. Lake trout only
6. Lake trout, northern pike and
yellow pickerel
7. Brook trout only
8 Brook trout, lake trout and bass
bass
northern
pike
yellow
pickerel
bass
lake trout
lake trout
and yellow
pickerel
brook trout
bass
70%
0%
50%
80%
0%
40%
0%
70%
yellow
pickerel
northern
pike
lake trout
and yellow
pickerel
northern
pike
brook trout
and lake
trout
50%
0%
30%
0%
0%
northern
pike
0%
northern
pike
0%
325
-------
Lake Restoration, Protection and Management
that the model appears to have a tendency to underestimate
measured lake alkalinity; reducing the contribution of a
significant input such as autochthonous production, would
worsen discrepancies. In addition, preliminary total fish pro-
duction estimates obtained are comparable to published
results (Ontario Ministr. Nat. Resour., 1979) and a substan-
tial reduction in lake alkalinities would lead to inconsisten-
cies between estimates.
The model assumes that runoff alkalinity is contributed
solely through cation exchange in aluminum silicate soils.
In fact, carbonate components are present in some glacial
tills deposited in the Precambrian Shield, and scattered car-
bonate bedrock formations also occur. Where these are pre-
sent, the rate of acidification will be much less than that
predicted by the model and the initial lake alkalinities will
significantly increase. While larger surficial deposits or
geological formations containing carbonate compounds were
excluded from the analysis, we believe that actual acidifica-
tion rates will be slower than predicted by the current ver-
sion of the model because of the influence of carbonate com-
pounds. Restricted local areas devoid of carbonate materials
are more likely to follow the predicted response.
From a biological perspective, most of the losses predicted
can be attributed to clinical reductions of productivity. The
modified morphoedaphic index was developed through
regression analysis; while considerable efforts have been ex-
pended to provide a rigorous theoretical explanation of its
predictive powers, the answer is still not completely satisfac-
tory. In fact, recently Prepas (1983) has demonstrated that
mean depth is the key parameter in the MEI and has sug-
gested that TDS should be deleted. Accordingly, a clinical
type of response in fish production can be questioned at
higher alkalinity and pH levels. The model would overpredict
effects if this relationship is invalid and the only valid impacts
would be threshold lethal effects.
CONCLUSIONS
1. The model appears to be a conservative estimator of
long-term acidification impacts (i.e., is more likely to
overestimate acidification rates than underestimate) but does
not adequately account for short-term acid pulse impacts and
fishing pressure effects which, if included, would result in
estimates of greater impacts on the sports fishery.
2. Any fisheries damages predicted need to be incor-
porated in a socioeconomic assessment to make direct com-
parisons with the costs of alternative abatement strategies.
3. Feasible immediate improvements to the analysis in-
clude developing a short-term acid pulse component and in-
corporating a carbonate solution relationship.
4. Long-term research efforts can help to clarify the con-
tribution to lake alkalinity budgets by groundwater, primary
production, and sediment interactions.
SUMMARY
The preceding discussion is intended to highlight the key
characteristics of a long-term lake/watershed acidification-
fisheries response model. The original report provides more
indepth treatment of the (1) internal relationships; (2) the
strengths and limitations, (3) a procedure for regional ap-
plication, and (4) preliminary results. The information in this
paper is being made available at this time in order to:
1. Help identify knowledge gaps in our understanding of
the whole acid rain problem,
2. Help focus the research effort, and
3. Serve as a "jumping off" point to estimate and advance
more work by other scientists to refine and make more ac-
curate, the models of fisheries losses through acidification.
Copies of the full report are available from: Department
of Fisheries and Oceans, Ontario Regional Office, 3050
Harvester Road, Burlington, Ontario, L7N 3J1.
REFERENCES
Aimer, B. 1972. Effects of acidification on fish populations in lakes
on the Swedish west coast. Inst. Freshw. Res. Drottningholm In-
form. 12:1-47.
Aimer, B., W. Dickson, C. Ekstrom, and E. Hornstrom. 1978. Sulphur
pollution and the aquatic ecosystem. Pages 271-311 in J. O.
Nriagu, ed. Sulfur in the Environment. Part II. Ecological Impacts.
J. Wiley and Sons, Inc., New York.
Andrews, A.K., et al 1980. Results of a modeling workshop con-
cerning acid precipitation. Western Energy and Land Use Team,
Off. Biolog. Serv. U.S. Fish Wildl. Serv.
Beamish, R.J. 1974. The loss of fish populations from unexploited
remote lakes in Ontario, Canada, as a consequence of at-
mospheric fallout of acid. Water Res. 8:85-95.
_. 1975. Long-term acidification of a lake and resulting effects
on fishes. Ambio 4:98-104.
. 1976. Acidification of lakes in Canada by acid precipitation
and the resulting effects on fishes. Soil Pollut. 6:501-514.
Beamish, R.J., and H.H. Harvey. 1972. Acidification of the La Cloche
Mountain lakes, Ontario, and resulting fish mortalities. J. Fish. Res.
Board Can. 29:1131-1143.
Beamish, R.J., C.M. Blouw, and G.A. McFarlane. 1976. A fish and
chemical study of 109 lakes in the Experimental Lakes Area (ELA),
Northwestern Ontario, with appended reports on Lake Whitefish
aging errors and the Northwestern Ontario baitfish industry. En-
viron. Can. Fish. Mar. Serv. Tech. Rep. No. 607.
Conroy, N., D.S. Jeffries, and J.R. Kramer. 1974. Acid Shield lakes in
the Sudbury, Ontario region. Proc. 9th Can. Symp. Water Pollut.
Res. in Canada.
Dillon, P.J. 1974. The prediction of phosphorous and chlorophyll
concentrations in lakes. Ph.D. Thesis, Univ. of Toronto.
. 1982. Mass balance models: An explanation of the ob-
served chemistry of lakes near Sudbury, Ontario. In Sudbury En-
vironmental Study—Studies of Lakes and Watersheds Near Sud-
bury, Ontario: Final Limmol. Rep. Ontario Ministry Environ. Publ.
Environment Canada. 1981. Susceptibility of aquatic ecosystems to
the effects of atmospheric pollutant fallout in Quebec. Sci. Ser.
No. 118. Inland Waters Directorate.
Farmer, G.J., T.R. Goff, D. Ashfield, and H.S. Samant. 1980. Some
effects of the acidification of Atlantic salmon rivers in Nova Scotia.
Can. Tech. Rep. Fish. Aquat. Sci. No. 972.
Henriksen, A. 1979. A simple approach for identifying and measuring
acidification of freshwater. Nature 278:542-545.
1980. Acidification of fresh waters: a large scale titration.
Pages 68-74 in D. Drablos and A. Tollan, ed. Proc. Int. Conf. Ecol.
Impact Acid Precip. Norway, 1980.
Hesslein, R.H. 1979. Lake acidification potential in the Alberta Oil
Sands Environ. Res. Progr. Study Area. AOSERP Rep. No. 71.
Hoffman, O.W., B.C. Mathews, and R. E. Wicklund. 1964. Soil
associations of Southern Ontario. Ontario Soil Surv. Rep. No. 30.
Hough, Stansbury & Michalski Ltd., and J. E. Hanna Associates Inc.
1982. An approach to assessing the effects of acid rain on On-
tario's inland sport fishery. Can. Dep. Fish. Oceans. Unpubl. Rep.
Keller, W. 1978. Limnological observations on the Aurora Trout
Lakes. Ontario Ministry Environ.
Lewis, W.M., and C. Peters. 1956. Coal mine slag drainage. Ind.
Wastes 1:145-147.
Menendez, R. 1976. Chronic effects of reduced pH on brook trout
(Salvelinus fontinalis). J. Fish. Res. Board Can. 33:118-123.
Mills, K. 1983. Experimental acidification of lake 223. Experimental
Lakes Area: responses of fish populations, 1976-1982. Pap.
presented at 36th Annu. Meet. Can. Conf. Fish. Res. Winnipeg,
Manitoba.
326
-------
National Academy of Sciences. 1981. Atmosphere-biosphere inter-
actions: Toward a better understanding of the ecological conse-
quences of fossil fuel combustion. Report prepared by Comm.
Atmos. Biosphere. Washington, D.C.
Oglesby, R.T. 1977. Phytoplankton summer standing crop and an-
nual productivity as functions of phosphorus loading and various
physical factors. J. Fish. Res, Board Can. 34:2255-2270.
Ontario Department of Lands and Forests. Surficial Geology Maps
of Northern Ontario. (5,165, 5,265, 5,365 and 5,465).
Ontario Ministry of Agriculture and Food. 1975 Ontario Soils.
Publ. 492.
Ontario Ministry of Environment. 1981. A submission to the United
States Environmental Protection Agency on Interstate Pollution
Abatement. Docket No. A-81-09.
Ontario Ministry of Natural Resources. 1979. Ontario fish yield esti-
mates. Rep. Working Group No. 4. Strategic Plan for Ontario
Fisheries.
. 1982. Partitioning yields estimated from the morphoedaphic
index into individual species yields. Rep. Working Group No. 12.
Strategic Plan for Ontario Fisheries.
Overrein, L.N., H.M. Seip, and A. Tollan. 1980. Acid precipitation-
effects on forest and fish. Final report of the SNSF Proj. 1972-1980.
Oslo, Norway.
Prepas, E.E. 1983. Total dissolved solids as a predictor of lake bio-
mass and productivity. Can. J. Fish. Aquat. Sci. 40:92-95.
Reckhow, K.H. 1979. Empirical lake models for phosphorus: develop-
ments, applications, limitations, and uncertainty. In 0. Scavia and
Contributed Papers
A. Robertson, eds. Perspectives on Lake Ecosystem Modeling.
Ann Arbor Sci., Ann Arbor.
Ryder, R.A. 1963. Chemical characteristics of Ontario lakes with
reference to a method of estimating fish production. Contrib. No.
637. Ontario Dep. of Lands and Forests, Res. Branch, Maple, Ont.
Ryder, R.A., S.R. Kerr, D.A. Loftus, and H. A. Regier. 1974. The
morphoedaphic index, a fish yield estimator — review and evalua-
tion. J. Fish. Res. Board Can. 31:663-688.
Schindler, D.W. 1980. Experimental acidification of a whole lake:
a test of the oligotrophication hypothesis. Pages 370-374 in D.
Drablos and A. Tollan, eds. Proc. int. Conf. Ecological Impact Acid
Precipitation. Norway, 1980.
Schindler, D.W., R.H. Hesslein. R. Wagemann, and W.S. Broecker.
1980. Effects of acidifcation on mobilization of heavy metals and
radionuclides from the sediments of a freshwater lake. Can. J.
Fish. Aquat. Sci. 37:373-377.
Schofield, C. L. 1976. Acid precipitation: effects on fish. Ambio
5:228-230.
Schreiber, K., O. Loucks, and R. Ellison. 1982. Effects of acid preci-
pitation on aquatic resources: results of modelling workshops. Air
Pollut. Acid Rain Rep. No. 12. Fish Wildl. Sen/., U.S. Dep. Inter.
Vallin, S. 1953. Zwei azidotrophe seen im kustengebiet von
Nordschweden. Rep. Inst. Freshw. Res. Drottningholm. 34:71-84.
Vollenweider, R. A. 1969. Moglichkeiten und grenzen Elementaren
Modelle der Stoffbilanz Von Seen. Arch Hydrobiol. 66:1-36.
Watt, W. D., D. Scott, and S. Ray. 1979. Acidification and other
chemical changes in Halifax County lakes after 21 years. Lim-
nol. Oceanogr. 24:1154-1161.
327
------- |