United States
Environmental Protection
Agency
Office of Water (WH-553)
Washington, DC 20460
'©EPA Monitoring Lak
'•{ / i
"*\ '
i iif.tr mi f
in
li
EPA 44 J/4-90-007
August 799
.• 79 tt&u1> *¦¦¦ £> '**&.<' ,
# r-
rf^-KS.-
Printed.on Reeve
-------
Monitoring Lake and
Reservoir Restoration
Technical Supplement to
The Lake and Reservoir
Restoration Guidance Manual
for
U.S. Environmental Protection Agency
Assessment and Watershed Protection Division
Washington, DC
1990
-------
This technical supplement was prepared by the North American
Lake Management Society under EPA Cooperative Agreement No.
CX-814969 from the Assessment and Watershed Protection
Division, Washington, DC.
Cover photograph: McLeodLakeReflections, courtesy of
Patricia Mitchell, Edmonton, Alberta, Canada
Points of view expressed in this technical supplement do not necessarily
reflect the views or policies of the U.S. Environmental Protection Agency
and the North American Lake Management Society nor of any of the con-
tributors to its publication. Mention of trade names and commercial
products does not constitute endorsement of their use.
This document should be cited as:
Wedepohl, R.E., D.R. Knauer, G.B. Wolbert, H. Olem, P.J. Garrison, and K.
Kepford. 1990. Monitoring Lake and Reservoir Restoration. EPA 440/4-90-007.
Prep, by N. Am. Lake Manage. Soc. for U.S. Environ. Prot. Agency, Washington,
DC.
-------
Preface
Lake restoration doesn't work by itself. One cannot begin to restore a
lake without understanding its background and the factors that
comprise its present condition. And to determine whether restoration is
working, one must certainly be able to credibly compare changes with
prior conditions.
There's nothing mysterious about this process: it's a systematic,
scientific collection and analysis of data known as monitoring. The
practice of monitoring is not limited to scientists; many excellent
monitoring programs use citizen volunteers, people without scientific
backgrounds but with an interest in water quality.
Monitoring, then, is simply a tool, an essential tool in restoring a lake.
Monitoring Lake and Reservoir Restoration defines and explains how to
use this tool, including the importance of the often neglected long-term
monitoring necessary to maintain a project's achievements.
In this manual the lake manager will find practical information on how
to design and implement a lake monitoring program during and following
a lake restoration project. In addition to describing monitoring methods
for both the waterbody and the watershed, the manual deals with
monitoring specific in-lake restoration techniques.
Although this manual specifically guides the lake manager who must
meet the Clean Lakes Program Phase II monitoring requirements, readers
will find it helpful as a starting point for more comprehensive studies of
lake ecosystems and useful in designing any lake study. Researchers will
welcome its recommendations for consistent methods and quality
assurance procedures.
Monitoring Lake and Reservoir Restoration is the first technical
supplement to The Lake and Reservoir Restoration Guidance Manual. As
with the parent volume, this manual was prepared by the North American
Lake Management Society for EPA's Clean Lakes Program, which
welcomes comments and suggestions. These should be addressed to the
Clean Lakes Program (WH-583), U.S. Environmental Protection Agency,
401M Street, S.W., Washington, DC 20460.
-------
Contents
Preface iii
Contents v
Acknowledgments x
Chapter 1: Introduction
Overview of the Clean Lakes Program 1-1
Purpose of this Manual 1-1
Intended Audience 1-2
Manual Organization 1-2
Chapter 2; Planning the Monitoring Program
Summary 2-1
Background 2-2
Monitoring Plans . 2-2
Preproject (Phase I) Monitoring 2-2
Monitoring During Phase II Implementation 2-3
Phase II Monitoring Following Treatment 2-3
Long-term and Phase III Monitoring 2-3
Quality Control/Quality Assurance 2-4
Chapter 3: Monitoring Methods
Summary 3-1
Background 3-1
In-lake Sampling Procedures 3-2
Water Chemistry Sampling 3-2
Dissolved Oxygen Measurements 3-4
Chlorophyll a Sampling 3-4
Secchi Disk Measurements 3-4
Sediment Sampling 3-5
Macrophyte Surveys 3-5
-------
Tributary Streams 3-5
Streamflow Measurements 3-5
Instantaneous Flow Measurements 3-6
Continuous Flow Measurements 3-8
Stieamwater Sample Collection 3-9
Sample Handling and Preservation 3-12
Water Chemistry Collection Containers 3-12
Sample Preservation 3-12
Analytical Methods 3-13
Phosphorus 3-13
Nitrogen .3-14
Alkalinity/Acid Neutralizing Capacity 3-15
Dissolved Oxygen Measurements 3-15
Chlorophyll a 3-15
Field QA/QC Samples 3-15
Chapter 4: Watershed Monitoring
Summary 4-1
Background 4-2
Relationship of Phase II to Phase I and III Studies . . 4-3
Watershed Monitoring: AHierarchical Approach 4-3
Level I: Watershed Inventories 4-3
Level II: Limited Stream Monitoring 4-3
Level III: Comprehensive Watershed Monitoring 4-3
The Nature of Nonpoint Source Pollutant Loadings to Lakes 4-4
The Effect of Watershed Size on Runoff Derived Loadings 4-6
Lake Water Residence Times—Implications for Monitoring 4-8
Level I: Watershed Inventories 4-9
Applicability 4-9
Construction Phase 4-10
Post-project Phase 4-10
Level II: Limited Stream Monitoring 4-11
Applicability 4-11
Construction Phase 4-13
Post-project Phase 4-14
Level III: Comprehensive Watershed Monitoring 4-16
Applicability 4-16
Streamflow Monitoring 4-18
Characterization of Constituent Concentrations in Streamflow ........ 4-19
Watershed Inventories . 4-19
Interpretation of Tributary Stream Data 4-19
Loading Calculations 4-19
Source Analyses 4-21
vi
-------
Chapter 5: In-lake Restoration Techniques and Monitoring
Summary 5-1
Background 5-2
OBJECTIVE: Control Nuisance Algae 5-5
CONTROL TECHNIQUE #1: Phosphorus Precipitation/ Inactivation
with Alum 5-5
Technical Considerations 5-5
Monitoring During Treatment 5-6
Considerations for Interrupting Treatment 5-6
Example—Wisconsin's Long Lake 5-7
Monitoring Following Treatment 5-8
CONTROL TECHNIQUE #2: Dilution/Rushing . 5-10
Technical Considerations 5-10
Monitoring During the First Two Weeks of Treatment 5-12
Considerations for Interrupting Treatment 5-13
Monitoring Following the First Two Weeks of Treatment 5-13
CONTROL TECHNIQUE #3: Artificial Circulation 5-15
Technical Considerations 5-15
Monitoring During the First Two Weeks of Treatment 5-16
Consideration for Interrupting Treatment 5-16
Monitoring Following the First Two Weeks of Treatment 5-17
CONTROL TECHNIQUE #4: Hypolimnetic Aeration 5-19
Technical Considerations 5-19
Monitoring During the First Two Weeks of Treatment 5-20
Considerations for Interrupting Treatment 5-20
Monitoring After the First Two Weeks of Treatment 5-20
CONTROL TECHNIQUE #5: Hypolimnetic Withdrawal 5-22
Technical Considerations 5-22
Monitoring During the First Two Weeks of Treatment 5-22
Considerations for Interrupting Treatment 5-22
Monitoring Following Treatment 5-23
CONTROL TECHNIQUE #6: Sediment Oxidation 5-25
Technical Considerations 5-25
Monitoring During Treatment 5-25
Considerations for Interrupting Treatment 5-25
Monitoring Following Treatment 5-26
CONTROL TECHNIQUE #7: Food Web Manipulation 5-27
Technical Considerations 5-27
Monitoring During Treatment 5-27
Considerations for Interrupting Treatment 5-28
Monitoring Following Treatment 5-28
OBJECTIVE: Increase Depth 5-30
CONTROL TECHNIQUE: Dredging 5-31
Technical Considerations 5-31
Monitoring During Treatment 5-31
Considerations for Interrupting Treatment 5-32
Monitoring Following Treatment 5-33
-------
OBJECTIVE: Control Nuisance Plants 5-34
CONTROL TECHNIQUE #1: Water Level Drawdown 5-35
Technical Considerations 5-35
Monitoring During Treatment 5-35
Considerations for Interrupting Treatment 5-35
Monitoring Following Treatment 5-36
CONTROL TECHNIQUE #2: Mechanical or Chemical Control of
Nuisance Plants 5-37
Technical Considerations . 5-37
Monitoring During Treatment 5-38
Considerations for Interrupting Treatment 5-38
Monitoring Following Treatment 5-39
CONTROL TECHNIQUE #3: Biological Control of Nuisance Plants
(Grass Carp) 5-40
Technical Considerations 5-40
Monitoring for the First Year After Fish Stocking 5-41
Considerations for Interrupting Treatment 5-41
Monitoring Following Treatment 5-42
OBJECTIVE: Mitigate Acidic Conditions 5-44
CONTROL TECHNIQUE #1: In-lake Liming 5-45
Technical Considerations 5-45
Monitoring During Treatment 5-46
Considerations for Interrupting Treatment 5-47
Monitoring Following Treatment 5-47
CONTROL TECHNIQUE #2: Watershed Liming 5-49
Technical Considerations . . 5-49
Monitoring During Treatment 5-49
Monitoring Following Treatment 5-50
Chapter 6: A Long-term Monitoring Protocol
Summary 6-1
Background 6-2
Monitoring Water Clarity 6-3
A Basic Lake Water Quality Monitoring Plan 6-3
Elements of the Basic Lake Water Quality Trend Monitoring Program 6-5
A Comprehensive Long-term Lake Monitoring Protocol 6-7
Rationale for Comprehensive Monitoring 6-7
Chapter 7: Case Study: Detection of Trends and Sampling
Strategy Evaluations—Statistical Evaluation of the
Neuse River, North Carolina, Total Phosphorus Data Set
Introduction 7-1
Statistical Model Selection 7-1
Parametric Methods 7-4
Distribution-free Methods (Nonparametric Methods) 7-4
Vlll
-------
Evaluation of the Historical Database 7-4
Data Entry and Preparation 7-4
Summary Statistics .7-5
Trend Analysis 7-8
Determination of Future Sampling Effort 7-8
Chapter 8: References 8-1
Appendix: Cooperative Agreements for Protecting and Restoring
Publicly Owned Freshwater Lakes, U.S. Environmental
Protection Agency 9-1
-------
A cknowledgments
We thank the following individuals who reviewed or assisted in the
preparation of this document:
Thomas Davenport
U.S. Environmental Protection Agency
Joseph Eilers
E&S Environmental Chemistry, Inc.
Warren Gebert
U.S. Geological Survey
Mary Jaynes
North Carolina Division of
Environmental Management
Frank Lapensee
U.S. Environmental Protection Agency
James Longbottom
U.S. Environmental Protection Agency
Kenneth H. Reckhow
Duke University
John Walker
U.S. Geological Survey
William W. Walker
Environmental Consultant
C. Bruce Wilson
Minnesota Pollution Control Agency
X
-------
Chapter 1
Introduction
Overview of the Clean Lakes Program
The Clean Lakes Program, which was initiated in 1972 under section 314 of the
Federal Water Pollution Control Act, was a direct response to widespread public
demands for means to protect and support lakes. Since 1975, the program has
provided more than $102 million to help fund State and local Clean Lakes
projects.
A strong partnership has developed among Federal, State, and local govern-
ments that has greatly aided the planning and implementation of each Clean
Lakes project. Although administration of the program is vested with the U.S. En-
vironmental Protection Agency, each State is encouraged to organize and ad-
minister lake projects that meet its individual needs.
States apply for grants through the EPA regional office for lake projects that
meet EPA and State criteria. After reviewing the grant application, the Agency may
award cost-sharing financial assistance to a State, which may, in turn, fund work
done by a community. Although the State may administer a Clean Lakes project
for a community, local involvement in the monitoring program is necessary to en-
sure complete restoration of the lake and future protection from degradation.
Purpose of this Manual
Clean Lakes regulations require that projects be monitored both during and after
implementation. This manual provides the guidance for both design and im-
plementation of a monitoring program by outlining specific standards for specific
types of lake restoration and protection projects.
This manual uses technical and scientific information to supplement the less
technical discussions on project monitoring found in The Lake and Reservoir Res-
toration Guidance Manual (U.S. Environ. Prot. Agency, 1988). It draws in part on
an initial compilation of lake monitoring techniques (unpublished) completed in
1988 by Science Applications International Corporation of McLean, Virginia. This
manual is intended to guide monitoring carried out in connection with the Phase II
or implementation portion of a lake restoration project. While the information con-
tained in this manual may prove useful during the development of diagnostic and
feasibility studies for lake projects, diagnostic/feasibility (Phase I) monitoring is
not treated directly. Generally, such activities are exploratory in nature and there-
fore more generic in terms of parameters measured. Throughout this manual it is
The Clean Lakes
Program was a direct
response to widespread
public demands for
means to protect and
support lakes.
This manual provides
the guidance for both
design and implement-
ation of a monitoring
program.
1-1
-------
Primary users of this
monitoring manual—
regional EPA Clean
Lakes project officers,
State and local project
managers, and project
sponsors and
considtants.
assumed that a scientifically based characterization of the lake's problem was
developed in a Phase I diagnostic and feasibility study and that the proposed res-
toration or protection measures to be implemented in Phase II are logical and ap-
propriate.
The plans presented in this manual are not intended as substitutes for the more
rigorous, object-specific monitoring that is part of a Phase III or research project.
However, data collected during the monitoring described in this manual may help
to identify potential projects.
Intended Audience
The primary users of this monitoring manual are expected to be
• Regional EPA Clean Lakes project officers,
• State and local project managers, and
• Project sponsors and consultants.
Because Federal Clean Lakes regulations are intentionally flexible regarding
the specifics of Phase II project monitoring, EPA Clean Lakes project officers can
consider the needs of each individual project when approving a monitoring plan.
Primary users should regard this manual as a foundation for the development of
satisfactory and practical monitoring plans that can be federally approved.
Manual Organization
This manual is divided into six major parts, exclusive of this introduction, the refer-
ences, and the appendix: five chapters treat monitoring planning and techniques
for lakes and watersheds, and one chapter is devoted to a case study. Each chap-
ter begins with a summary that highlights the salient points and concludes with a
list of references for readers who wish to consult more in-depth materials on the
individual topics.
Those readers responsible for designing the monitoring component of a lake
restoration or protection project or for evaluating the monitoring plan will find
Chapters 3, 4, and 5 to be the most valuable portions of this manual. Clean Lakes
regulations relating to monitoring are included as an appendix.
The Environmental Protection Agency would appreciate any suggestions on
how this manual could be made more practical and useful. Readers are en-
couraged to send comments and recommendations to:
Chief, Clean Lakes Section
Assessment and Watershed Protection Division (WH-553)
U.S. En vironmental Protection Agency
401MStreet, S.W.
Washington, D.C. 20460
1-2
-------
H Chapter 2
Planning the
Monitoring Program
-
Summary
i.
A monitoring plan is required for all Clean Lakes
projects.
2.
The plan should include provisions for monitoring both
during and after project implementation.
3.
The scope of the monitoring program should be propor-
tional to the cost Of the project and should not unneces-
sarily drain resources needed for implementation. For
projects without a research component, adequate data
can usually be acquired for 10 percent or less of total
project costs.
4.
The Phase II monitoring program's structure should
enable local sponsors to continue monitoring the lake
over the long term without additional Clean Lakes fund-
ing.
5.
Quality assurance must be a foremost consideration in a
monitoring program. A quality assurance project plan is
required for all Clean Lakes projects.
¦¦ ' ' ' ' J
2-1
-------
Background
In most instances, the
costs of a monitoring
program will not exceed
10 percent and may
often be closer to
5 percent of the total
project expenditure.
Every Clean Lakes project requires a monitoring plan. The information gained
from this program serves many purposes over the course of the project. Initially,
monitoring data are used to determine the source of potential or actual lake im-
pairment and to provide a basis for selecting appropriate restoration and protec-
tion techniques. During project implementation, monitoring data assist in
assessing restoration work and help determine the need for adjusting project im-
plementation measures. After the project, monitoring data provide the basis for
evaluating project achievements and the impact of treatment.
Monitoring Plans
The foundation of a successful implementation monitoring program is a monitoring
plan that is tailored to the specific problems and goals of each project. The plan
should cover the periods of time both during and after the treatment phase and,
typically, should call for measurement of both in-lake and watershed charac-
teristics. It will identify the information to be obtained, how and when it will be ob-
tained, and what methods will be used to ensure that the data are reliable.
The following elements should be included in every monitoring plan:
• Sampling frequency and stations
• Number of samples
• Types of samples (grab, composite, etc.)
• Number of field blanks and duplicates needed to meet data quality
assurance requirements
• Field measurement and collection procedures and
• Analytical methods.
The goals, scope, and level of detail vary greatly among Clean Lakes project
monitoring plans. The challenge, therefore, is to focus limited resources to obtain
an appropriate level of information without burdening the project with an excess of
monitoring requirements. It is easy to overdesign a monitoring plan, thereby taking
resources away from project management, data evaluation, or the implementation
of the project itself.
In most instances, the costs of a monitoring program will not exceed 10 percent
and may often be closer to 5 percent of the total project expenditure. Much
depends on the level of planned long-term monitoring and whether extensive
watershed monitoring is required for project evaluation. A project designed and
funded to include more intensive research may greatly exceed this 10 percent rule
of thumb. Note also that section 314 regulations do not adequately reflect the
scope of monitoring work—including long-term, post-restoration studies under
Phase III grants—that is needed to support a research effort.
Preproject (Phase I) Monitoring
Preproject monitoring is generally carried out as part of a lake diagnostic and
feasibility study. In the case of an impaired lake, monitoring can reveal the source
or cause of the impairment and provide information on possible restoration
measures. In addition, preproject monitoring information, usually called "baseline
data," is the foundation against which future monitoring results are compared.
Such comparisons can be used to evaluate implemented projects, whether for
restoration or protection.
2-2
-------
This manual does not treat preproject monitoring directly. For this information,
the reader is referred to Section 8 of the Clean Lakes Program Guidance Manual
(U.S. Environ. Prot. Agency, 1981); regulations governing preproject monitoring
are found in 40 CFR Part 35, Appendix A. These regulations broadly identify lake
monitoring elements and provide generic guidance on how projects might be
monitored. Monitoring results needed for a preproject diagnostic and feasibility
study will, in most cases, produce a sufficiently detailed characterization of the
lake for later comparison and evaluation of the project. The sections of this
manual dealing with monitoring methods and with quality assurance are also ap-
plicable to preproject monitoring.
Monitoring During Phase II Implementation
Because implementation monitoring during treatment is a condition of all EPA
Clean Lakes awards, a monitoring plan must be approved by the EPA project of-
ficer before Phase II work can begin. The regulations for implementation monitor-
ing are found in 40 CFR Part 35, Appendix A. These regulations, while specific,
are not inflexible; they allow the EPA project officer discretion to approve a
monitoring program tailored for a particular lake or project. While the project of-
ficer must at least consider the protocol for developing a monitoring program, the
stated objective of implementation monitoring is to "provide sufficient data that will
allow the State and the EPA project officer to redirect the project, if necessary, to
ensure desired objectives are achieved."
Chapter 5 of this manual presents monitoring considerations and specifications
for the most frequently used in-lake restoration techniques. The plans are consis-
tent with the protocol given in 40 CFR Part 35, Appendix A, of the regulations.
Suggested criteria for interrupting a treatment based on monitoring results are
summarized for each technique.
Phase II Monitoring Following Treatment
EPA regulations require monitoring of Phase II implementation projects for at least
one year after restoration of the lake or installation of pollution control devices.
Also, before other Phase II work can begin, the first year's post-treatment monitor-
ing plan must be approved by the EPA project officer. The purposes of this post-
treatment monitoring are to provide data needed to evaluate the effectiveness of
restoration measures immediately after their implementation and to determine
whether project objectives were achieved. It is important to note that the certainty
of this determination is greatly increased in almost all cases by long-term (three to
five years) monitoring. This is particularly important for watershed monitoring and
certain specific lake restoration techniques (e.g., dredging to increase lake depth).
The discussion of project monitoring in Chapter 5 of this manual identifies in-
lake techniques that require more than the minimum one year of post-treatment
monitoring and outlines key monitoring considerations in addition to in-lake
monitoring specifications for the first year (or years) of post-treatment monitoring.
Long-term and Phase III Monitoring
EPA regulations require only one year of monitoring following,treatment, not be-
cause a longer program is unnecessary or undesirable but because the duration
of cooperative agreements must be limited to a reasonably short time frame. For
many types of projects, a year of post-project monitoring provides the necessary
information while allowing the timely closing of a project grant. If additional
monitoring is necessary, sponsors can be awarded costs beyond the first year as
The regulations for
implementation
monitoring allow the
EPA project officer
discretion to approve a
monitoring program
tailored for a particular
lake or project.
2-3
-------
Locally sponsored
long-term monitoring is
one of the most
cost-effective activities in
all lake management and
it serves as an excellen t
foundation for a
continuing lake
management program.
Quality control, which
ensures that monitoring
data are accurate and
precise, must be a
foremost consideration
in planning and
conducting a monitoring
program.
part of the grant-eligible project budget. However, project periods beyond four
years must receive special approval from the EPA regional administrator.
Locally sponsored long-term monitoring that follows project completion is one
of the most cost-effective activities in all of lake management and should always
be encouraged because it serves as an excellent foundation for a continuing lake
management program. Eventually, however, most long-term monitoring requires a
monetary commitment from the local sponsor in the absence of Federal financial
support, although limited long-term monitoring may often be accomplished
through volunteer efforts and State agency assistance.
To ensure locally financed continuation of the post-treatment monitoring pro-
gram, the project manager must correctly ascertain the level of financial commit-
ment that can be expected from the local sponsor. Key considerations when
developing a long-term monitoring program should be minimization of costs,
education of laypersons for volunteer monitoring, and periodic, professional inter-
pretations of data for the project and local sponsors.
A long-term monitoring program can be as simple as Secchi disk readings that
are collected twice a month during the growing season or as complex as a multi-
faceted program that compiles information on a variety of the lake's ecosystem
components. In Chapter 6, various levels of long-term monitoring after project
completion are outlined. The simplest level is appropriate where the local financial
base is modest and the relative importance of (or threats to) the lake as a resource
are limited. The more ambitious levels of monitoring are appropriate where there
are greater local resources; where there is more likelihood of a lake being
degraded by changing watershed conditions or in-lake biota; and where the lake is
more valuable as a local or regional resource.
After the project is closed, the only monetary support currently available for
long-term monitoring in the Clean Lakes Program is the Phase III post-restoration
evaluation. Phase III funds are used on a limited number of previously completed
and independently selected Clean Lakes implementation projects to verify the lon-
gevity and effectiveness of various restoration techniques. Because these
projects are essentially research-oriented, monitoring requirements are highly
case-specific and therefore not dealt with further in this manual.
Quality Control/Quality Assurance
¦ Quality Control, which ensures that monitoring data are accurate and precise,
must be a foremost consideration in planning and conducting a monitoring pro-
gram. Poor quality monitoring data are worse than none at all since this informa-
tion represents a substantial investment of money and time and serves as the
basis for even larger investments. (One midwestern State spent over $300,000
during the 1970s on poor quality data that jeopardized both the success of several
projects and, ultimately, the State's lake management program as well.) A relative-
ly small additional effort to ensure that reliable data are collected and properly
maintained in a database is indispensable.
¦ Quality Assurance requires that the project sponsor prove that the monitoring
results are accurate and precise. For this reason, EPA specifies minimum require-
ments for quality assurance plans in Clean Lakes projects that it funds. Often,
States and private consultants prepare an umbrella quality assurance program
plan that encompasses EPA Clean Lakes project monitoring requirements. These
requirements are found in Interim Guidelines and Specifications for Implementing
Quality Assurance Requirements for EPA Contracts (U.S. Environ. Prot. Agency,
1980).
2-4
-------
The following are the 16 elements of a quality assurance project plan:
* Title page with provisions for approval signatures
* Table of contents
* Project description
* Project organization and responsibility
* Quality assurance objectives for measuring data in terms of precision,
accuracy, completeness, representativeness, and comparability
* Sampling procedures
* Sample custody
* Calibration procedures and frequency
* Analytical procedures
* Data reduction, validation, and reporting
* Internal quality control checks and frequency
* Performance and system audits and frequency
* Preventative maintenance procedures and schedules
* Specific routine procedures to assess data precision, accuracy, and
completeness of specific measurement parameters involved
* Corrective actions, and
* Quality assurance reports to management.
Chapter 3 of this manual contains details on collecting quality control field
samples.
-------
-------
Chapter 3
Monitoring Methods
Summary
The project officer and project manager must have a
working knowledge of monitoring methods to ensure the
quality of the data.
A confusing variety of analytical techniques and field
collection procedures for lake studies has evolved that
hinders data interpretation and limits comparability. Cer-
tain methods have become somewhat standardized
through repeated use and should be relied upon unless a
deviation can be clearly justified.
Ten percent of all water chemistry samples in addition to
those used within laboratories for quality assurance pur-
poses should be used for field quality control.
Background
Poor quality data that have been collected in a nonstandard fashion are a per-
petual problem when interpreting and comparing lake water quality studies. Mean-
ingful study-to-study comparisons become less precise—if they can be made at
all.
Many of the methods employed in lake and watershed evaluations have
evolved from limnological procedures, others from the hydrological sciences, and
Poor quality data that
have been collected in a
nonstandard fashion are
a perpetual problem.
3-1
-------
To minimize problems
that arise when data from
different projects cannot
be compared, use of the
procedures described in
this chapter are
recommended.
still others from engineering practices. Because of the multidisciplinary nature of
lake studies, less experienced lake scientists often become perplexed or dis-
couraged by the diversity of methods, and even the abilities of the most ex-
perienced are periodically questioned when these scientists deal with data outside
their area of expertise.
Chapter 3 provides brief summaries of the methods and techniques used for
obtaining lake data. The level of detail provided is intended to give project
managers a working knowledge of the more common procedures.
The methods described here are not the only ones used and, in some cases,
may not be the most appropriate for a particular application, such as a research
project or a particular interference problem. However;; because a set ofr stan-
dardized methods can help to minimize problems that arise when data from dif-
ferent projects cannot be compared, use of the procedures described in this chap-
ter are recommended for Clean Lake monitoring programs.
Past problem areas are highlighted throughout Chapter 3, and suggestions are
made on review techniques that a project manager can use to ensure quality data.
The principal methods described here were generally derived from either Environ-
mental Protection Agency recommended methods or from Standard Methods for
the Examination of Water and Wastewater (1989). Other good references on this
tope include publications by the U.S. Geological Survey (1977 a.b); Haveren
(1986); Holtan et al. (1968); U.S. Bureau of Reclamation (1975); and Hiilman et al.
(1986).
In-lake Sampling Procedures
All water samples must
be carefully collected,
properly preserved, and
appropriately analyzed.
Water Chemistry Sampling
All water samples are a subset of the whole lake. To be representative of the lake
component being described, they must be carefully collected, properly preserved,
and appropriately analyzed.
The following sections highlight the more important considerations a field tech-
nician must make when collecting samples to ensure they are representative of
the waterbody and have not been contaminated during the process of collection.
Where needed for a specific purpose, additional sampling requirements will be
identified. When the purpose is not described, the following requirements can
serve as general guidance:
¦ Sample Locations for Shallow Lakes. For the purposes of this manual, a
shallow lake is defined as one that has fairly uniform oxygen concentrations in the
surface-to-bottom profile and does not stratify. For general characterization, a
sample from the one foot depth near the center of the lake will often describe con-
ditions; shallow lakes tend to be well enough mixed so that a single sample is rep-
resentative. Exceptions will be those lakes with complex configurations and the
long, river-run impoundments that often show longitudinal differences.
¦ Sample Locations for Deep Lakes. For the purposes of this document, deep
lakes are defined as those lakes that stratify. Epilimnetic waters are sampled as
shallow lakes, with the presumption that the upper waters are generally mixed.
When sampling the hypolimnion, more careful techniques are required to avoid
vertical chemical gradients that are often present because of the lack of wind
mixing and constituents from lake sediments. The highest concentrations of dis-
solved material are usually observed nearest the lake sediments. The highest con-
3-2
-------
centrations of phosphorus, which are often released under anoxic conditions,
occur immediately above the sediments.
A precise characterization of hypolimnetic conditions would require a vertically
integrated sample that is adjusted for volume. However, hypolimnetic conditions
can almost always be sufficiently characterized for Clean Lake monitoring pur-
poses by collection of two samples, one near the top of the hypolimnion and
another Just above the lake sediments (approximately three feet above the bot-
tom). When collecting the bottom sample, care must be taken to ensure that the
sample is free of bottom sediments.
¦ Water Samplers. The most commonly used containers for collecting water
from deep within a lake are the modified Kemmerer or Van Dorn (Alpha Bottle)
samplers (Fig. 3.1). Water samples can also be collected using peristaltic pumps
and weighted hose. When pumps are used, they are often combined with an in-
line filter (0.45 jim membrane) when sampling for material that is dissolved be-
cause of anoxic conditions, such as dissolved phosphorus in an anoxic
hypolimnion.
Figure 3.1 .—Water samplers (courtesy of Wildlife Supply Co.)—(left) Alpha
Bottle (Van Dorn sampler); (right) Modified Kemmerer sampler.
Samplers should be made of material compatible to the
parameter being analyzed and should always be carefully
cleaned prior to use. For nutrient analyses, the sampling equip-
ment must be rinsed several times with the lake water to be
sampled prior to obtaining the sample. Acid washing of equip-
ment used to obtain chlorophyll samples is not recommended,
as acid quickly destroys the chlorophyll. For most lakes, the
sampler must be rinsed with lake water prior to sampling new
stations. It is good technique to collect the lowest concentration
samples first; e.g., top samples are collected before bottom
samples. Stauffer (1981) is a useful reference on sampling
equipment and methodologies.
Samplers should be made
of material compatible to
the parameter being
analyzed.
To obtain chlorophyll
samples, acid washing
of equipment is not
recommended.
3-3
-------
When dissolved oxygen
meters are used, they
must be calibrated
against Winkler analyses.
... chlorophyll a samples
should be collected on a
depth-integrated basis,
from the top 6 feet of the
water column.
Chlorophyll a samples
should be filtered
through a glass fiber
filter immediately in the
field.
Dissolved Oxygen Measurements
Dissolved oxygen samples for wet chemistry analyses are often collected with a
Kemmerer- or Van Dorn-type sampler. The field sample is immediately fixed by
adding manganous sulfate and alkaline iodide-azide prior to analysis (EPA
Method 360.2).
A dissolved oxygen electrode (EPA Method 360.1) is often used when
numerous determinations are necessary. When dissolved oxygen meters are
used, they must be calibrated against Winkler analyses both prior to and following
each day's use. An additional check during midday is also advisable given the ten-
dency for the calibration of some meters to drift. Many dissolved oxygen
electrodes are susceptible to contamination by hydrogen sulfide or through loss of
temperature compensation capability, with the error not always noticeable when
manufacturer- recommended air calibration procedures are used alone.
Chlorophyll a Sampling
Chlorophyll a is the most common biological parameter measured in lake monitor-
ing programs. To help standardize the data, it is recommended that chlorophyll a
samples be collected, on a depth-integrated basis, from the top 6 feet of the water
column. Although other sampling depths have been suggested—two times the
Secchi disk depth, the entire epilimnion, the photic zone—problems can arise with
each of these approaches. Occasionally, chlorophyll data can be biased by high
concentrations of metalimnetic algae, as can happen when the two times the Sec-
chi disk-based sampling depth method is used. Similarly, high concentrations of
blue-green algae are often found near a lake's surface during periods of calm
weather, which will bias sample results if a surface-based sampling technique is
used. The six foot integrated sample should be a good compromise in almost all
cases.
Integrated chlorophyll samples can be collected with a Kemmerer water
sampler, pump, or tube collector, as described by Kennedy (1985) and Stauffer
(1981).
Chlorophyll a samples must be filtered through a glass fiber filter immediately in
the field and then the filter should be cooled (frozen) and stored in a dark container
until analyzed. Use stainless steel forceps when handling the filters. A good tech-
nique is to place the filter in a 15 mL centrifuge tube painted black or taped and
containing a known volume (i.e., 10 mL) of 90 percent acetone. If raw water
samples are simply cooled and stored for a period longer than a few hours prior to
filtration, pigment can break down. Under no circumstances should raw water
samples be held longer than 24 hours or frozen prior to analysis. Good references
are A Manual on Methods for Measuring Primary Production in Aquatic Environ-
ments (Vollenweider, 1969) and Standard Methods for the Examination of Water
and Wastewater (1989).
Secchi Disk Measurements
Secchi disk readings are obtained with a 20 cm diameter disk. Observations are
made, during midday and without sunglasses, from the shady side of the boat.
The observer, who should be wearing a life vest, makes the reading by looking as
close to the water as possible to minimize glare. Ropes must be made of a non-
stretchable material and periodically checked with a measuring tape. (Rope-
making material will often shrink following several wet-dry cycles.)
3-
4
-------
Sediment Sampling
Chemical characterization of lake sediment is often an important component of
Phase I diagnostic studies; rarely will there be a need to characterize lake sedi-
ments during the Phase II construction or post-monitoring periods. When sedi-
ment chemical characterization is desired, procedures will vary significantly, being
decidedly site- and parameter-specific. Detailed methods for sampling will not be
described here; for more information, an excellent reference on sediment and
sampling techniques is Sedimentation Engineering (Vanoni, 1975).
Macrophyte Surveys
The level of effort required for completion of macrophyte surveys varies greatly.
Generally, the surveys that might be associated with a Phase II project must be
quantitative enough to allow comparisons between surveys and between lakes.
Document species composition, distribution, abundance, and maximum depth of
growth during the growing season by using visual observations as much as pos-
sible. Locate major community types (emergents, floating-leaved, and submer-
gents) and then determine species composition and abundance of each
community (abundant, common, sparse) using methods described by Phillips
(1959). The information is best presented on a hydrographic lake map that il-
lustrates distribution of the communities, with a species list and appropriate abun-
dance symbol for each location. Boundaries of single species stands within the
more general community type should aiso be noted.
Plants should be identified to the species level using a regional identification
manual such as those written by Fassett (1969), Voss (1972), Godfrey and
Wooten (1979), or Muencher (1964). Voucher specimens should be collected,
dried, and pressed for future reference and verification.
Tributary Streams
The following section briefly describes some of the methods recommended for
making discharge measurements in stream channels and for collecting stream
water quality samples. The list of methods is not exhaustive; it only highlights the
level of effort required to measure streamflow and provides guidance on obtaining
representative water samples. Field engineers or technicians must be relied upon
to select methods and equipment best suited for particular situations.
Streamflow Measurements
Discharge or streamflow is defined as the volume rate of flow of water, usually
measured in cubic feet per second, past a specific point in the stream. As
described in Chapter 4 of this manual, most lake responses to watershed loadings
are a function of both water quantity and quality. Therefore, whenever water
samples are collected in a stream, concurrent flow rate must also be known. The
most common techniques for measuring streamflow are described here.
Given the importance of streamflow measurements, it is recommended that
consultants unfamiliar with techniques consult with or enlist the services of a U.S.
Geological Survey Water Resources Division office prior to initiating a streamflow
measurement project. A good reference on flow measurement and computation of
discharge is Measurement and Computation of Streamflow, the U.S. Geological
Survey Water Supply Paper No. 2175, Volumes 1 and 2 (Rantz, 1982). Additional
information may also be found in ERA'S Handbook for Sampling and Sample
Preservation of Water and Wastewater (U.S. Environ. Prot. Agency, 1982).
Macrophyte communities
are best presented on a
hydrographic map of the
lake.
Given the importance of
streamflow
measurements, it is
recommended that
consultants unfamiliar
with techniques consult
with or enlist the services
of the U.S. Geological
Survey.
3-5
-------
Instantaneous Flow Measurements
The velocity-area
method of measuring
discharge is the
principal method for
calculating flow in open
channels.
Instantaneous flow gagings measure the quantity of flow passing a monitoring site
at one point In time. These measurements are commonly made directly in the
stream channel where water quality samples are collected. Methods highlighted
here are used for these open channel measurements. Measurement of flows in
confined conduits, such as storm sewers, is not described but is discussed in the
previously referenced U.S. Geological Survey or EPA publications.
The velocity-area method of measuring discharge is the principal method for
calculating flow in open channels. In the velocity-area method, streamflows are
measured by determining the mean velocity of the water passing through the
cross-sectional area of the channel. This is generally done by taking a series of
velocity, width, and depth measurements across the stream and summing up the
products of the areas and velocities. The formula is expressed as:
where
Q = Sum(AiVj)
Ai is the cross-sectional area and
Vi is the mean velocity.
Stream cross-sectional area is determined using the midsection method where
several depth measurements are obtained across the stream channel with sec-
tions of the stream assigned to each depth. Figure 3.2 shows a typical stream
cross section and the measurement made to calculate cross-sectional area. Flow
Water surface
Initial
point
CN
O
(n—l)
3\
EXPLANATION
1,2,3 n Observation points
b},b2,b3 bn Distance, in feet, from the initial
point to the observation point
dt,d2,d3 dn Depth of water, in feet, at the
observation point
Figure 3.2.—Definition sketch of the midsection method of computing stream cross-sectional
area. (Source: Buchanan and Somers, 1968.)
-------
velocity measurements, usually expressed as feet per second, are also made
within each of the partial sections as shown in Figure 3.2. When the water depth is
less than 2.5 feet, these measurements are made at the 0.6 depth down from the
water's surface to obtain an average velocity in the vertical. Where the stream is
deeper than 2.5 feet, two measurements are made, one at the 0.2 depth and one
at the 0.8 depth, with the measurements then being averaged to define velocity. A
sufficient number of partial sections are needed so that no more than 10 percent
of the total flow is described by any one partial section. Generally, this requires 20
or more partial section measurements. A typical streamflow measurement field
sheet is shown in Figure 3.3.
FLOW STREAM CROSS SECTION & DISCHARGE DATA
A sufficient number of
partial sections are
needed so that no more
than 10 percent of the
total flow is described by
any one partial section.
Location Date:
Flow Meter # Recorders Initials
Start Finish Spin Test
Time Stream Width # Measurements
Distance
from Bank (ft)
Depth
(ft)
Velocity
(ft per sec)
Area
(ft )
Discharge
(cu. ft/sec)
Comment:
(Veeetation- Sludee)
Figure 3.3.—A streamflow measurement field sheet
Streamflow measurements are best made at sites that avoid complications
caused by complex channels, abrupt changes in channel configuration, ponded
conditions, and free-falling water. Careful field notes should be kept, and the
project manager should occasionally review a copy of a streamflow gaging field
data sheet.
¦ Velocity Measurement with Rotating Current Meters. The rotating current
meter, such as the Pygmy or Gurley meter, is the most commonly used device for
measuring velocity. As in making any scientific measurement, care must be exer-
cised to assure that the equipment is properly calibrated and maintained. For ex-
ample, proper field sheets should show that a spin test was conducted on rotating
current meters (if used). These tests are done to ensure that the meters are not
binding, that they spin freely and accurately reflect the velocity of the water.
¦ Velocity Measurement with Floats. Although this velocity measurement tech-
nique is not recommended for day-to-day use, situations occasionally exist where
current meter readings are not possible, and ad hoc methods are needed. In these
circumstances, flow velocities can be estimated simply by using a slightly buoyant
surface float, such as an orange, grapefruit, or wad of tissue paper. To determine
flow velocity, one or more floats are placed in the stream and the time needed to
travel a measured distance is determined. A coefficient of 0.85 is commonly used
to convert surface velocity to mean velocity. Flow is then determined as a product
of the mean velocity and cross-sectional area of the stream.
3-7
-------
Fixed control structures
should be the method of
choice when site
conditions allow their
use.
Since high flow events
are very important... a
special effort should be
made to obtain a direct
measurement.
¦ Other Flow Measurement Techniques. Flow rates are often measured by
using a fixed control structure, such as a dam, sluice gate, weir, or flume. These
control structures can provide excellent flow data and should be the method of
choice when site conditions allow their use.
In some cases, periodic but accurate measurement of small streamflows is im-
portant. In these instances, i.e., where streamflow is less than 2 cubic feet per
second, flow measurement using the velocity-area method can be very difficult.
Small flows are best measured by catching the streamflow into a known volume
(five gallon buckets are often used) and measuring the time taken to fill the known
volume. Precise measurements are also possible by using dyes or other tracer
dilution techniques. Occasionally, it is possible to install a portable weir plate in the
stream channel. Standard tables are then used that relate water level to flow rate.
Continuous Flow Measurements
When at all possible, obtain a continuous record of flow. Water flows are related to
stage (the height of the water in the stream channel): the higher the stage, the
higher the flow. Normally, continuous water flow data are collected in natural
stream channels following development of a stage-discharge relationship.
A rating curve is developed by making several instantaneous streamflow meas-
urements and then plotting them against the stage of the stream at the time of
measurement. To develop a rating curve, a minimum of five direct stream gagings
should be made, with the measurements describing the full range of streamflow
conditions. Often, the higher flows are the most difficult to obtain. Since high flow
events are very important to most lake studies, a special effort may be necessary
to obtain a direct measurement of a high flow. Too often stage relationships are ex-
trapolated beyond the capability of the data; the most important high flows are the
ones having the greatest error.
Once a stage-discharge relationship has been developed, flows can then be in-
directly obtained from knowledge of water levels in the stream channel, alone.
These water level measurements may be made either manually or automatically.
¦ Manual Water Level Measurement. Once a stage-discharge relationship has
been developed, flow rate estimates are simply made by measuring water levels
at periodic intervals. These water levels are usually measured by reading a staff
gage (Fig. 3.4) that has been installed directly in the stream channel. A periodic
resurvey of staff gages is necessary to ensure they have not been disturbed by
debris or ice movements.
Obtaining data from systems exhibiting significant flow variability is a major
problem inherent to manual measurements. Important high flow events can be
easily missed if measurements are done manually. The problem is exacerbated in
the smaller, more flashy streams common to lake studies. Manual stage recording
techniques should be used only in streams where flow is stable, and the prob-
ability of missing high flow events is small.
Manual stage readings commonly are made in lake studies where the dam or
outlet structure serves as a control and a stage-discharge relationship can be es-
tablished. In these cases, a staff gage is often attached to a concrete abutment
near the outflow, but not so close as to be within the drawdown regime of the dis-
charge.
¦ Automatic Water Level Recorders. A variety of continuous water level re-
corders are currently used to obtain a record of streamflow. In the past, the most
common method has been to use a stilling well in which there is afloat attached to
a rotating chart that records stage.
3-:
8
-------
METHOD 1
METHOD 2
METHOD 3
Concrete Headwall
Removable Pipe (used when
measurements during ice
free periods are needed)
Steel Fence Post
/
Staff gage installed
on concrete headwall
with expansion bolts
Q,
V4* L shaped
rod used -»
to pin pipe
7777777
, Staff gage bolted
to board and pinned
to pipe
— Screw clamp
— 1 !£' galvanized pipe
^ Pipe union
Stream or lake bottom
V777777
2" galvanized pipe
driven 2 to 5 feet into
stream or lake bottom. Top
of pipe is below ice depth.
Staff gage bolted
to board and
pinned to post
. Stee! fence post
V
Note: All Staff gages must be referenced to a nearby datum and be resurveyed prior to and after use
Figure 3.4.—Various staff gage Installation techniques.
More recently, bubbler tubes have been used, with stage being related to the
pressure required to force gas from the tube placed on the stream bottom; data
are then recorded automatically on tape. In these newer stations, a continuous
record of flow can be easily calculated by computer once a rating curve has been
developed. A U.S. Geological Survey Water Resources Division office should be
contacted for guidance before a water level recording station is installed. Informa-
tion on monitoring hardware can be obtained from the U.S. Geological Survey's
Hydrologic Instrumentation Facility, a lab that has an excellent ongoing hardware
testing program. Write to USGS's Facility at Bldg. 2101, Stennis Space Center,
MS 39529.
Automatic stations, such as the one illustrated in Figure 3.5, require periodic,
often weekly, maintenance. If water quality samples are also being collected, then
additional servicing is necessary. The cost of a continuous record of streamflow
ranges between $5,000 and $20,000 a year. These stations are more completely
described in the U.S. Geological Survey Water Supply Paper No. 2175 (Rantz,
1982).
Streamwater Sample Collection
Streamwater sampling is a vital part of most lake studies and can easily become
the largest source of error in obtaining water quality information. This fact is not
well enough recognized and needs to be emphasized.
Small samples of water are collected from lake tributary streams to charac-
terize the chemical and physical nature of the water entering a lake, with nutrients
and sediment normally being the constituents of most concern. Concentrations of
these parameters must be combined with streamflow to provide the loading infor-
mation critical for most studies.
Homogeneity of a stream at a cross section is determined by physical factors
such as proximity of inflows and turbulence in the channel. Characteristics are not
necessarily homogeneous across the width and depth of a stream cross section.
Poor lateral or vertical mixing is often observed. Immediately below the con-
Stream sampling... can
easily become the largest
source of error in
obtaining water quality
information.
3-9
-------
r IH) SAM now ~
MMIWWa ttAROM
tu*o«c*oo»CMw«yr*.
Figure 3.5.—An automated stream gaging and sampling station.
Vertical heterogeneity is
also common in streams.
fluence of a stream and a tributary, distinct physical separations may exist be-
tween the water of the tributary and that of the main stream and, particularly in
large rivers, this separation sometimes persists for many miles downstream. Sam-
pling locations where mixing is incomplete should be avoided.
Vertical heterogeneity is also common in streams. Figure 3.6 shows an ideal-
ized description of the velocity, sediment concentration, and sediment discharge
regime present in a stream channel. As can be seen, sediment concentrations are
highest near the stream's bottom. Therefore, for phosphorus and the other con-
stituents often associated with sediments, single grab samples taken from a
stream system are often very poor representations of the whole. Use of careful
techniques will ensure that samples taken from stream cross sections are repre-
sentative.
¦ Manual Sample Collection. Where the stream is well mixed (such as at a site
immediately below an overflow structure), a single grab sample may adequately
represent water quality at that instant. However, where there is some likelihood of
stratification, a vertical composite sampling technique is preferable.
Portable integrating sampling devices that allow water to enter the sample con-
tainer at a rate proportional to the water flow rate at the intake nozzle should be
used. This sampling device, which is described in Porterfield (1972), is raised or
lowered from a selected position in the stream to represent ajl the river flow at the
particular point along the cross section. The process is repeated at other points
along the cross section, and then the individual depth-integrated samples are
combined to reflect average characteristics.
¦ Automatic Samplers. Automatic samplers are commonly installed at
streamflow gaging stations to obtain samples over a routine time period (e.g.,
every 30 minutes, 2 hours, 6 hours) or can be set to collect samples in response to
changing water levels (flows). Automatic samplers are recommended whenever
3-
10
-------
VELOCITY
SEDIMENT CONCENTRATION
SEDIMENT DISCHARGE
Figure 3.6.—Velocity, sediment concentration, and sediment discharge within a stream chan-
nel.
storm events must be characterized. Although water samples may be collected by
hand during storm events, automatic stations generally provide the best informa-
tion. Various types of automatic samplers are described in EPA's Handbook for
Sampling and Sample Preservation of Water and Wastewater (1982). Note that
care should be exercised with older samplers that lack prepurge capabilities;
without a good prepurge, initial samples are often contaminated and do not repre-
sent actual conditions.
Holding time constraints that cannot be met often affect collection of dissolved
phosphorus data when automatic samplers are used. Automatic samplers must
be equipped with a refrigeration unit to keep phosphorus and nitrogen samples
chilled until they can be collected. Samples should be collected from the station
regularly to ensure that holding times prior to analyses are not exceeded.
Care should be
exercised with older
samplers that tack
prepurge capabilities.
3-11
-------
Automatic samplers
require special
maintenance.
Always provide
documentation that
automatic sampler
quality assurance checks
were made.
Prior to sample
collection, bottles and
collectors should always
be double- or
triple-rinsed with the
lake or stream water to
be sampled.
Dissolved phosphorus
samples may not be
acidified; they must be
field-filtered, cooled, and
analyzed within 24 hours.
A correction factor to adjust for the inability to vary sample intake location within
the stream should be developed for all but the smallest tributaries. An individual
correction factor must be obtained at each site by collecting a sample composite
across and vertically within the stream channel as described in the manual collec-
tion techniques section. Then the manually collected sample is compared with the
automatically collected sample to develop a correction factor.
Automatic samplers require special maintenance. For example, because the in-
take hose should be free of obstructions, it must be cleaned routinely to prevent
plugging and growth of periphyton that can cause changes in sample concentra-
tions. Therefore, to assure unbiased data, the following maintenance procedure is
recommended:
• After the sampler has been set up and is ready to use, collect one bottle of
distilled, deionized water as if it were a normal sample running through the
entire hose and sampler.
• Cap the sample bottle and place it in the center of the sampler, where it will
remain chilled.
• Following collection of the actual field samples, retrieve the quality control
check sample and analyze for the same parameters as the actual samples.
If the results of the analyses indicate concentrations above the detection limit,
further study is needed to define the source of the problem. Always provide
documentation that automatic sampler quality assurance checks were made.
Sample Handling and Preservation
This section highlights a few selected procedures of special concern for collecting
and handling water quality samples. More complete information is available in the
Handbook for Sampling and Sample Preservation of Water and Wastewater (U.S.
Environ. Prot. Agency, 1982).
Water Chemistry Collection Containers
All samples must be collected in previously cleaned containers (do not use phos-
phorus detergents) that are appropriate for the parameter being analyzed. Prior to
sample collection, bottles and collectors should always be double- or triple-rinsed
with the lake or stream water to be sampled. For phosphorus and nitrogen collec-
tion, 250 mL acid-washed polyethylene or glass bottles should be used. Do not
use acid-washed containers for chlorophyll a collection.
Sample Preservation
Total phosphorus, ammonium nitrogen, nitrite-nitrate nitrogen, and total Kjeldahl
nitrogen can be analyzed out of a single sample that has been preserved by adding
H2SO4 to acidify to a pH of 2 or less. Samples must be analyzed within seven days
following preservation, according to EPA methods.
Dissolved phosphorus samples may not be acidified; they must be field-filtered,
cooled, and analyzed within 24 hours. Raw samples that have been cooled and fil-
tered at the laboratory will often not reflect concentrations actually present in lake
water. Filtration methods are discussed further in the following section.
3-12
-------
Analytical Methods
Over the past several decades, a major work entitled Standard Methods for Ex-
amination of Water and Wastewater (1989) has evolved to standardize techni-
ques for examining water samples. Both Standard Methods and EPA's equivalent,
Methods for Chemical Analyses of Water and Wastes (U.S. Environ. Prot. Agency,
1983), are intended to ensure accuracy and reproducibility of laboratory results.
Nevertheless, lake scientists who have examined data obtained from samples
split between two or more laboratories sometimes wonder whether that objective
has been attained. Although laboratory analytical techniques are better described
than other elements of lake studies, care is still needed to ensure that proper tech-
niques are followed so that good data are obtained.
The following sections highlight techniques recommended for a set of selected
chemical parameters that are often most critical to lake studies. A complete
description of the analytical methods for these and other parameters can be found
in Methods for Chemical Analysis of Water and Wastes.
Phosphorus
It is strongly recommended that EPA Method 365.1 be used for all phosphorus
analyses completed as part of a Clean Lake Phase II monitoring program. This
automated procedure best fulfills the goals of attaining uniformity between
studies, collecting more reproducible data, and, most importantly, obtaining lower
limits of detection. An alternative method should be used only when well justified
by special circumstances.
The most careful analytical methods are needed when phosphorus concentra-
tions are in the range of 0.001 mg/L to 0.050 mg/L. Occasionally, less sensitive
techniques can be used when phosphorus concentrations are known to be higher
than 0.02 mg/L. Figure 3.7 identifies the various forms of phosphorus that can be
differentiated from a single sample. The two forms that are highlighted, total phos-
phorus (STORET No.00665) and dissolved reactive phosphorus (STORET
No.00671, called dissolved orthophosphate in the EPA methods handbook),
should be reported in Phase II monitoring.
¦ Total phosphorus, as the name implies, is a measure of all the constituent
present in a water sample, including that which may be associated with
suspended solids, colloids, or organic compounds. To allow analysis, the raw
water sample is digested by boiling the sample with strong acids to dissolve ail
material. Digestion techniques occasionally vary, being dependent upon the
amount of suspended sediment associated with the water sample.
¦ Dissolved reactive phosphorus is a representation of that form of phos-
phorus that is most readily available for uptake by algae. By definition, dissolved
phosphorus is that portion that passes through a millipore-type membrane filter
with pore diameters of 0.45 nm. This size opening has come to be generally
adopted as defining "dissolved" as opposed to "particulate" material. Obviously
this definition is arbitrary and may not reflect the real amount of phosphorus avail-
able to algae, but it has evolved as the method of choice. Field filtration of raw
water samples is required for dissolved reactive phosphorus analysis. Dissolved
forms of phosphorus can change quickly following collection. If samples are not
immediately filtered, changes can occur from co-precipitation with metals (this
commonly happens when an anoxic water sample is exposed to air) and from
biological uptake within the sample bottle even if the sample has been cooled.
It is strongly
recommended that EPA
Method 365.1 be used for
all phosphorus analyses
completed as part of a
Clean Lake Phase II
monitoring program.
... total phosphorus
(STORETNo.00665) and
dissolved reactive
phosphorus (STORET
No.00671, called
dissolved orthophos-
phate in the EPA methods
handbook), should be
reported in Phase II
monitoring.
The method
recommended for
collecting dissolved
phosphorus samples
from an anoxic
environment uses an
in-line filter technique
that eliminates any air
contact.
3-
13
-------
f
Residue
SAMPLE
Total Sample (No Filtration)
1
Direct
Colorimelry
Hydrolysis &
Colorlmetry
Persulfatc
Digestion &
Colorimelry
Orthophosphate
Hydrolysable &
Orthophosphate
Total
Phosphorus
Filter (through 0.45 fi membrane filter)
1
Filtrate
I
Direct
Colorimelry
Dissolved
Reactive Phosphorus*
I1
I Hydrolysis & X
f Colorimelry T
Persulfate
Digestion &
Cotorimetry
Diss. Hydrolyzable
& Orthophosphate
Total Dissolved
Phosphorus
* Also referred to as dissolved orthophosphate
or dissolved inorganic phophorus
Figure 3.7.—Analytical scheme for differentiation of phosphorus forms.
The method recommended for collecting dissolved phosphorus samples from an
anoxic environment uses an in-line filter technique that eliminates any air contact
(Stauffer, 1981).
To prevent coitfitsion,
total and dissolved
reactive phosphorus
should be reported as P
and not PO4.
¦ Particulate phosphorus is, by definition, total phosphorus minus total dis-
solved phosphorus. As a quality assurance effort, the project manager should oc-
casionally scan reported data to ensure that dissolved reactive phosphorus
concentrations are not reported to be greater than total phosphorus concentra-
tions for the same water sample.
To prevent confusion, both total and dissolved reactive phosphorus should be
reported as elemental phosphorus, i.e., total phosphorus (P) and not as total
phosphate (PO4). Note that phosphorus reported as PO4 will be 3.133 times
higher than if reported as elemental P.
Nitrogen
Nitrogen, a plant nutrient that limits aquatic plant productivity in some lakes and
river systems, can exist in several different forms. The most common forms of
nitrogen evaluated in lake studies are total nitrogen, ammonium nitrogen, nitrite +
nitrate nitrogen, and total Kjeldahl nitrogen.
¦ Total nitrogen is, by definition, all nitrogen found in a water sample. It is a sum
of total Kjeldahl nitrogen (organic and reduced nitrogen) plus nitrite + nitrate
nitrogen. Total nitrogen to total phosphorus ratios are used to identify which
nutrient is limiting to plant growth in lake waters. A lake is usually defined to be
phosphorus limited if the total nitrogen/total phosphorus ratio is greater than 10:1.
3-14
-------
¦ Ammonia nitrogen (more accurately occurring as ammonium in lakes) is the
form most readily used by lake plants. It is found in highest concentrations below
wastewater discharges or in anoxic lake waters. The common analytical techni-
ques are an automated phenate method (EPA Method 350.1) or an ion selective
electrode method (EPAMethod 350.3).
¦ Nitrite + nitrate nitrogen are usually analyzed together. Although nitrite can be
discriminated from nitrate nitrogen, this more expensive differentiation is usually
not necessary for most lake studies. Rarely will nitrite concentrations be sig-
nificant. The most common analytical method for nitrite-nitrate uses a cadmium
reduction technique (EPA Method 353.3). Similar to phosphorus, all nitrogen
species should be reported as elemental nitrogen (N), not as NO3. For example,
when reported as NO3 instead of as NO3-N, values will be 4.4 times higher.
Nitrite and nitrate
nitrogen are usually
analzyed together.
¦ Total Kjeldahl nitrogen is analyzed by using a digestion technique that con-
verts nitrogen components of biological origin to ammonia. The total Kjeldahl
value will also include any ammonia present in the sample. Organic nitrogen is, by
definition, total Kjeldahl nitrogen minus ammonia nitrogen.
As a quality assurance effort, the project manager should occasionally scan
reported data to ensure that ammonia concentrations are not reported to exceed
total Kjeldahl nitrogen concentrations for the same water sample.
AIkalinity/Acid Neutralizing Capacity
When alkalinities are less than 20 mg/L, the Gran analysis method should be
used. The Gran method for alkalinity provides information that is usually referred
to as acid neutralizing capacity because it includes alkalinity plus additional buf-
fering capability of dissociated organic acids and other compounds.
Dissolved Oxygen Measurements
As discussed under the section on sampling procedures, dissolved oxygen is
usually measured using a modified Winkler titration. Dissolved oxygen meters
must be calibrated against Winkler titrations prior to and following a day's use.
Chlorophyll &
Field filtering is required prior to laboratory analyses for chlorophyll a. Chlorophyll
a concentrations should also be corrected for pheophytin prior to being reported.
The analytical techniques described in Standard Methods are appropriate for
Clean Lake study needs.
Field QA/QC Samples
At least 1 in every 10 water samples should be a field quality control check
sample. The following are various types of field check samples that should be col-
lected during the monitoring project. AH QA/QC sample results should be
reported.
¦ Field Duplicates. A field duplicate is a sample taken to determine variability in
the sampling procedure and the source sampled. It is useful when the concentra-
tion of the parameter being sampled is both close to the detection limit of the
At least 1 in every 10
water samples should be
afield quality control
check sample.
3-15
-------
laboratory and to the level of concern for the parameter. Generally afield duplicate
should be taken after every tenth sample.
¦ Field Blanks. Afield blank is a sample of reagent grade deionized water that is
processed through the sampling equipment in the same manner as the actual
sample. This is done to determine if field cleaning procedures are adequate. A
field blank should be taken with every field duplicate. Ideally, no contaminants will
be detected in the field blank. If contaminants are detected, the validity of the
day's samples must be judged.
¦ Split Samples. A split sample is taken to determine interlaboratory variability.
The sample is collected, preserved, and split into two portions for analysis at two
different laboratories. Split samples are designed to determine analytical
variability between laboratories, not sampling variability.
¦ Spike Samples. Spike samples are used to estimate the accuracy of an
analysis. A known amount of substance is added to the sample, and the amount
recovered is determined. Samples spiked in the field can be used to estimate
sampling efficiency and handling loss.
16
-------
Chapter 4
Watershed
Monitoring
: — :—— .——~——\
Summary
1. Long-term hydrological and chemical data that cover a variety of events
are needed to quantify runoff-derived pollutant loadings. The difficulty
of defining loadings increases with decreasing watershed size.
2. The high costs, long time frames, and high level of effort required to
define long-term, average watershed loadings to lakes will normally
preclude collection of this data during a Phase II implementation
project.
3. At the time of project completion, a post-project watershed inventory
must be conducted to document existing conditions.
4. A short-term, periodic grab sampling of tributary streams can provide
qualitative information on loadings and quantitative information on
pollutant source type.
5. The watershed monitoring protocol should be designed so that the local
sponsor will be encouraged to continue the program on a long-term
basis. Data relevance along with cost and ease of acquisition are impor-
tant considerations.
6. There is a paucity of data on the effectiveness of land management
practices that were installed to meet water quality objectives. Water-
shed evaluations should contain, where feasible, long-term measure-
ments of the effectiveness of the project's watershed improvement
practices.
; ; ; /
-------
Background
Project managers must
be careful not to
over-design a
monitoring program.
The bulk of Phase II
watershed monitoring
information will more
often than not be
comprised of data
obtained from surveys
and inventories.
Lakes are products of theirwatersheds; therefore, a lake's waterquality reflects the
condition and management of the lake's watershed. Many lake restoration and
protection projects include watershed components. In some projects, agricultural
practices are modified; in others, urban runoff is treated; and in still others,
groundwater contamination sources are corrected. All these projects are geared
toward reducing pollutant loadings to lakes.
Like in-iake project components, watershed controls require monitoring to
evaluate effectiveness. However, unlike in-lake techniques, watershed improve-
ments are usually installed in an incremental fashion over long time periods and
generally with less risk of acute damage to a lake during implementation. There-
fore, a watershed monitoring program will rarely dictate discontinuation of water-
shed improvements during the project.
One of the most difficult and problematic decisions a lake manager must make
is deciding how to measure nutrient, sediment, and other pollutant contributions to
a lake. Defining monitoring needs for point source loadings is relatively straightfor-
ward. Measurement of groundwater-carried pollutants is extremely difficult, how-
ever; fortunately, loadings from this source are rarely as significant as those
delivered to a lake from surface runoff. The EPA is also reviewing available
methods for assessing nonpoint source-contaminated groundwater discharges to
surface water. Since the majority of a lake's watershed-derived pollutants are
generated from surface runoff, Chapter 4 will focus primarily on monitoring sur-
face-derived sources.
It is difficult to monitor nonpoint source-derived pollutants. Watershed climatic
conditions are rarely steady state. Tributary loadings of water, nutrients, and sedi-
ments are normally extremely variable, exhibiting significant storm-to-storm,
season-to-season, and year-to-year differences. Documenting long-term average
loadings often requires installation of a comprehensive network of streamflow
gaging and automated sample collection stations that are operated and main-
tained over a long period, often three or more years. In addition, data interpreta-
tion, which can be difficult, requires professional judgment from the analyst.
For these reasons, quantification of watershed-derived sediment and nutrient
loadings to a lake are not normally conducted under a Clean Lakes Phase II study.
The high costs of watershed monitoring and the limited time available for post-
project monitoring usually preclude the long-term, comprehensive tributary
monitoring activities needed to define nonpoint source pollutant loadings and
document effects of improved management practices. Therefore, the project
manager should not attempt to prescribe an intensive, short-term data collection
program with expectations that lake loadings will be quantified. Project managers
must be careful not to over-design a monitoring program by budgeting a dis-
proportionate amount for water sample collection and laboratory analyses at the
expense of project administration, quality assurance, data interpretation, and
general watershed evaluations.
In the context of Phase II studies, the meaning of the word "monitoring" will not
be limited to the collection and analysis of chemical and hydrological data but will
be expanded to include watershed inspections, inventories, and general condition
surveys. These sorts of monitoring activities aid data interpretation and provide
essential information that will help fulfill monitoring objectives at a reasonable
cost. The bulk of Phase II watershed monitoring information will more often than
not be comprised of data obtained from surveys and inventories.
4-2
-------
Relationship of Phase II to Phase I and III
Studies
As explained in Chapter 1, the Phase I study identifies a lake's trophic status, the
magnitude and sources of pollutants, and the lake's expected response to pol-
lutant reductions. Ideally, a Phase II watershed monitoring component would be a
scaled-back version of the Phase I study, designed to measure watershed loading
reductions. Prior knowledge of system variability is an aid in designing an ap-
propriate watershed monitoring component. The Phase III grant program is a
mechanism to obtain the higher level of information usually necessary to quantify
watershed pollutant loadings to lakes and the effectiveness of best management
practices.
Watershed Monitoring: A Hierarchical
Approach
Although the watershed monitoring component of a Phase II study is important, it
is nonetheless only part of a lake restoration or protection project; its design must
be consistent with the overall objectives of each project and will vary in intensity
from one to another.
As a decisionmaking aid, watershed monitoring programs have been divided
into three basic levels. Therefore, when designing a watershed monitoring pro-
gram, the project manager should generally determine the appropriate monitoring
intensity based on the following hierarchy:
Level I: Watershed Inventories
A post-project watershed inventory should always be compiled for any lake im-
plementation project. In some cases, a simple update of the Phase I study infor-
mation will suffice. During the construction phase, inspections are needed if best
management practices are being installed. Watershed inventories can be ex-
tremely helpful when evaluating existing and potential nonpoint source loadings.
Prior knowledge of
system variability is an
aid in designing an
appropriate watershed
monitoring component.
Watershed inventories
can be extremely helpful
when evaluating existing
and potential nonpoint
source loadings.
Level II: Limited Stream Monitoring
Tributary stream sample collection programs should be considered for water-
sheds or sub-watersheds where significant problems have been previously iden-
tified. A common tributary sampling strategy combines flow measurements
(preferably continuous) with collection of water samples that are analyzed for
phosphorus, nitrogen, and suspended sediments. A common sampling interval is
14 to 28 days, with the sample being collected near the point(s) where the most
important tributary stream (s) enter the lake.
A drawback of this protocol is that such time series-based monitoring programs
will almost always underestimate actual loadings if simply combined to estimate
loads. However, this sampling strategy can be especially helpful in identifying sig-
nificant differences between adjacent sub-watersheds.
Level III: Comprehensive Watershed Monitoring
The most accurate information on lake loadings comes from a comprehensive
network that continuously records streamflow, and flow-activated sampling sta-
tions that can characterize storm events. Where the existing database merits,
continuation of a comprehensive network that is already in place should be en-
4-3
-------
Pollutant delivery to
lakes varies from storm
to storm and year to year.
... a description of the
loadings received by a
lake during any one-year
or even two-year period
might fail to reflect the
actual long-term
loadings
Some might even argue
that, between changing
land use and climatic
variability, there may not,
in fact, be a "long-term"
condition that makes any
sense.
couraged. Supplemental funding for acquisition of watershed loading information
may be available and should be sought.
The Nature of Nonpoint Source Pollutant
Loadings to Lakes
Pollutant delivery to lakes varies from storm to storm and year to year. Important
reasons for this variability include climatic factors: storm intensity and duration,
coverage of the storm over the basin, and timing of tributary flows. For the most
important pollutants, the highest concentrations are usually observed during
periods of highest flow and are often associated with short-term but intense storm
events. Smaller watersheds are subject to the greatest variability.
As an example, Figure 4.1 shows data collected by Baker (1988) that illustrate
the typical patterns of concentration changes that occurred during a runoff event
from a 149 square mile watershed near Melmore, Ohio. Total phosphorus,
suspended solids, and nitrate nitrogen were observed to increase with increasing
discharge as is typical of nonpoint source-derived pollutants. During the falling
portion of the hydrograph, total phosphorus concentrations declined but not as
rapidly as suspended solids concentrations. Nitrate concentrations continued to
increase.
This particular storm event produced elevated flows for a relatively long time,
three to four days. However, in urban or smaller rural watersheds, storm events
often produce elevated flows for a time period measured in minutes or hours.
Several samples must be collected before contaminant loadings from even a
single runoff event can be reasonably described.
In addition to within-storm variability, nutrient and sediment loadings also ex-
hibit a great deal of season-to-season and year-to-year variability. Figure 4.2
shows the seasonal and annual variability that was observed over a nine-year
period from the same watershed near Melmore, Ohio. Clearly, a description of the
loadings received by a lake during any one-year or even two-year period might fail
to reflect the actual long-term loadings, no matter how intense the monitoring ef-
fort. Some might even argue that, between changing land use and climatic
variability, there may not, in fact, be a "long-term" condition that makes any sense.
In other long-term studies, similar amounts of year-to-year variability have been
observed. Figure 4.3 shows that yearly phosphorus loadings from a small, 1.27
square mile watershed at White Clay Lake, Wisconsin, ranged from a low of 112
pounds to a high of 646 pounds over the seven-year monitoring period. At Delavan
Lake, Wisconsin (Fig. 4.4), phosphorus loss from a 21.8 square mile watershed
varied from 1,400 pounds to 15,100 pounds over a five- year period. The large
variability observed in these studies is similar to that found by Minns and Johnson
(1979). They concluded from their study of rivers draining into the Bay of Quinte
that year-to-year variation in runoff is the major source of variation in watershed
export of phosphorus.
In summary, although it is certainly desirable to define nutrient and water load-
ings to a lake following installation of improved watershed management practices,
the nature and characteristics of runoff events and complications caused by the in-
cremental implementation of best management practices over the project time
frame often preclude obtaining such information. There is a high level of risk as-
sociated with attempting to draw conclusions about the "average" loadings
received by a lake if data are limited to what can be collected during a one- or two-
year period. Also, where loadings are calculated using data collected only at
regular intervals, true loadings are almost always underestimated because critical
storm events are not well described.
4-4
-------
Figure 4.1.—Typical pattern of concentration changes during a runoff event observed June 1981 at Honey Creek Station near Melmore, Ohio. Solid line represents flow.
The connected squares represent: A. suspended solids; B. total phosphorus; C. nitrite plus nitrate nitrogen; D. atrazine (source: Baker, 1988).
O iysoo
o
5^)00
2^00
%
10 11
JUNE 1381
3,000
O 1.000
> 10 11
JUNE 1981
O >¦«»
¦¦¦¦¦¦¦¦¦
3^00
ta
E,
CO
3
DC
O
X
a
w
0
X
a.
1
2,000
CQ
O 1.000
JUNE 1981
JUNE 1981
-------
LEGEND FOR GRAPHS
~ Summer
Z Spring
iS Winter
¦ Fall
197T 1979 1981 1983
WATER YEAR
| i.OQO
|
1 4,000
s
1 2,000
_ r-i 7>
gp gj gj
1977 1979 1981 1883 1985 AVE
WATER YEAR
1977 1979 1981 1983 198S AVE
WATER YEAR
Peak pollutant
concentrations are
higher in the runoff from
small watersheds than
from large watersheds.
80,000
C. 80,000
S 20,000
197? 1979
19B1 1883
WATER YEAR
1977 1979 1981 1983
WATER YEAR
Figure 4.2.—Annual variability and seasonal distribution of discharge, loading of suspended
solids, total phosphorus, dissolved reactive phosphorus, and nitrite plus nitrate nitrogen at
Honey Creek, near Melmore, Ohio (source: Baker, 1988).
The Effect of Watershed Size on Runoff-derived
Loadings
Baker (1988) studied watersheds In the Lake Erie basin ranging in size from 4.4
square mile to 6,330 square mile and identified significant patterns of stream
delivery of agricultural pollutants. Knowledge of these patterns can help a
manager who is designing a lake watershed monitoring program. Although
Baker's conclusions were based on observations from agricultural watersheds,
extrapolation to urban environments should also be possible. Baker observed that
¦ Peak pollutant concentrations are higher in the runoff from small
watersheds than from large watersheds. This effect is most pronounced
for sediments and sediment-associated pollutants, including particulate
phosphorus.
4-i
6
-------
2,000
tn
3
cc
o
X
Q.
CO
0
X
Q.
1
1,500
1,000
500
P?:;" :.i
r —* "r -1
ii
1974
1975
1976
1977
1978
1979
Figure 4.3.—Phosphorus loadings to White Clay Lake, Wisconsin, from a 1.27 square mile
watershed (source: Persson, 1983).
Summer
Spring
Winter
Fall
20,000
15,000
10,000
5,000
1984 1985 1986 1987 1988
WATER YEAR
Figure 4.4.—Phosphorus loadings to Delavan Lake, Wisconsin, from a 21.8 square mile water-
shed (source: Holmstrom et al. 1988).
The duration of runoff events and associated pollutant loadings in-
creases with the size of the watershed.
The annual variability in pollutant yield is greater in small watersheds
than in large watersheds. This factor complicates the task of evaluating
the effectiveness of watershed abatement practices for lakes that have
small watersheds (i.e., where watershed area to lake area is 10:1 or
less).
The annual variability
in pollutant yield is
greater in small
watersheds than in large
watersheds.
4-7
-------
...it takes more
sampling effort to
accura tely measure
pollutant loadings from a
small watershed ...
... lakes require several
flushings before effects of
loading reductions are
observed.
As watershed size becomes smaller, increasing proportions of the total
annual pollutant loading occur in shorter time windows. Consequently, it
takes more sampling effort to accurately measure pollutant loadings from a
small watershed than it does from a larger one, since it is easier for a sam-
pling program to miss the high loading episodes in the smaller watersheds.
The time periods of peak sediment load (and associated pollutants) dif-
fer between smaller and larger watersheds. In small systems, most sedi-
ment loads occur when there is a combination of bare soils and high inten-
sity rainfall events, generally during spring and early summer. In larger
watersheds, most export occurs earlier, generally during the times of peak
discharge when sediment previously deposited in stream channels is
resuspended and exported.
-Implications for
Lake Water Residence Times-
Monitoring
Phase I lake studies are usually conducted over a single year. A one-year monitor-
ing program will allow a somewhat qualitative description of the seasonality of
watershed loadings; however, it will not allow description of year-to-year
variability. Since most lakes have the capability to "average" the loadings they
receive, consideration must be given to the implications this dampening-out effect
has on the effects of individual event loadings.
Lake water residence times vary greatly, ranging from a few days to tens or
even hundreds of years. For lakes having longer residence times (a year or more),
long-term average pollutant loadings become more important to overall lake water
quality. These lakes, unlike many river-run lakes or large reservoirs, are often
characterized as "completely mixed reactors" that have the potential to retain pol-
lutants from previous years* loadings. Because of this capability, they require
several flushing cycles before the effects of loading reductions might be observed.
As an ideal example, a lake having a water residence time of one year will still
retain 50 percent of its original water after a year of average inflow. Following the
second year (after two flushings), 25 percent of the original water will still remain.
In the third year, the lake will have 12.5 percent of the original water, and so on.
This characteristic requires that the longer the water residence time, the longer
the time frame needed for in-lake observations to detect any response to loading
reduction.
As an example, if an estimated response to loading changes would be observ-
able after 85 percent water replacement, then three years of monitoring would be
necessary for a lake having a one year water residence time. Also, the longer the
residence time, the more likely the lake will show response to average rather than
event loadings. Lakes having very short water residence times (days or weeks)
may show response to seasonal loadings. In those cases, a shorter-term monitor-
ing program could identify a lake's response to loadings by focusing on critical
time periods, such as the summer growing season. Even in these cases, year-to-
year variability might obscure the lake's eventual response.
Moreover, because lakes are not always completely mixed reactors and be-
cause pollutants are not always conservative substances, evaluations may be
even more tenuous. Voilenweider (1976) discusses the concept of phosphorus
residence times and how they differ from water residence times. In a study of
Shagawa Bay, Malueg (1975) observed that the lake responded to phosphorus
reductions more slowly than predicted by water residence time. Similar findings
4-;
8
-------
were also observed in an Irish lake, Lough Ennell (Lennox, 1984), where an im-
portant factor in delaying recovery was the buffering capability of the lake sedi-
ments that acted as both a sink and a source of phosphorus. In other instances,
where the mass of labile sediment-associated phosphorus is low, then a lake
might improve more quickly than predicted by water residence times. Such a
situation could be observed in a lake that received a brief, one-time loading, such
as from a fertilizer spill.
Every completed Phase I study should include an estimate of long-term
average water residence times. Where site-specific, long-term water inflow data
are not available, a first approximation of a lake's water residence time can be ob-
tained by using the following procedure:
By definition, a lake's average water residence time is calculated by dividing
the lake's volume by the average annual water outflow (V/Q), with the volume of
the lake expressed as cubic meters or acre-feet. Volumes are determined from
lake depth surveys. Average annual inflow information can be approximated by
using regional average annual runoff rates and adjusted for groundwater inflows
and evaporation/precipitation if necessary. Surface water runoff will dominate
most lakes' water budgets. The exceptions will be those lakes having very small
watersheds. The map included in the pocket of this manual can be used to obtain
this information. Bent (1971) also described a technique he used in Michigan to
develop estimates of annual flow using multiple regression equations. This
methodology, which relates streamflow to basin and climatic conditions, can pro-
vide an estimate of annual flow along with a standard error of estimate.
As an example, the average annual residence time for a lake located in north-
east Pennsylvania can be calculated as follows: First, it is noted that average an-
nual runoff rates are approximately 25 inches per year in this area (Gebert et al.
1987). It is then determined that, at this location, the difference between evapora-
tion and precipitation is negligible (Natl. Oceanic Atmos. Admin. 1982).
Groundwater inflows are also determined to be insignificant, as this lake has a rela-
tively large watershed area of 20 square miles (12,400 acres). An estimate of the
lake's watershed water loadings would then be 25,833 acre-feet per year (25 in-
ches x 12,400 acres/12 inches per foot). If its volume were 50,000 acre-feet, the
lake's water residence time would then be 1.9 years (50,000 acre-feet/25,833
acre-feet of runoff per year).
This particular Pennsylvania lake would be one that is expected to respond
somewhat slowly to watershed changes. It is also an example of a lake that would
require a period of six or more years before response to watershed reductions of
conservative substances would be observed based on water residence times
alone. Where more precise estimates of water loadings are necessary, Cooke et
al. (1986) provide additional detail on defining water budgets.
Level I: Watershed Inventories
Applicability
Every Clean Lake project monitoring plan should have a component that
describes watershed conditions after construction is completed. The level of detail
that is needed will depend on the complexity of the watershed's land uses and the
actions taken to improve watershed conditions.
As described earlier, the direct measurement of pollutant loadings to a lake is
difficult to accomplish given financial and time constraints inherent to most lake
projects. Fortunately, a relatively simple watershed inventory can provide the cru-
cial information needed to minimize problems during the construction phase of a
project and identify critical areas needing long-term protection.
Every Clean Lake project
monitoring plan should
have a component that
describes watershed
conditions after
construction is
completed.
4-9
-------
Construction Phase
... the watershed should
undergo a periodic
general inspection
during critical time
periods....
In many cases, confined
animal feeding
operations contribute
more than point sources.
Many Clean Lakes projects require the installation of watershed management
practices to minimize sediment and nutrient delivery to a lake. Practices common-
ly used include installation of grassed waterways, treatment ponds, water diver-
sions, construction of animal waste control facilities, changes in cultivation
practices, and septic system maintenance.
A periodic watershed inspection program is warranted if for no other reason
than to document installation of practices as specified in the project design and to
ensure that associated interim control measures, such as short-term erosion con-
trol practices, are being used.
In addition to these specific inspections, the watershed should undergo a peri-
odic general inspection during critical time periods (e.g., late winter and late spring
for small, agriculturally dominated watersheds) to check for new problems that
might be developing, to serve as a verification of earlier surveys, and to identify
possible previously unobserved problems. The inspections required in the
monitoring plan would generally be documented by observer notes and refer-
enced to a map of the watershed. The following is a typical checklist for these in-
spections:
• Are the watershed management practices being installed according to
design?
• Are adequate controls in place to prevent unnecessary erosion or loss of
nutrients at implementation sites?
• Are agricultural practices being improved according to design?
• Are there any new construction sites that were not anticipated and, if so,
are adequate control measures in place?
• Have there been any new building permit applications or zoning changes
that are potentially detrimental to project success?
Post-project Phase
An inventory of watershed conditions immediately following project completion is
essential to establish a baseline for future evaluations and to serve as a model for
the local project sponsor who presumably will be encouraged to conduct routine
inspections in the future.
The level of detail required in documenting watershed conditions may vary, but
the final report must always include a delineation of surface watershed boundaries
and land Uses on a map of appropriate scale. If important (as they might be for
seepage lakes), groundwater contributing areas should also be delineated. In the
majority of lake projects, a topographic map with a scale of 1:24,000 (7.5 minute
quadrangle) is used.
Land uses should be broken into the basic categories identified in Table 4.1 and
characterized as a percentage of the total watershed size. With the exception of
lakes having very large watersheds, Level I information should always be
provided. Level II or higher information should be required to document uses
having high contamination potential such as point sources, confined agricultural
feeding operations, land under development, and strip mines.
As an example, more detailed information is often desired for confined animal-
feeding operations that commonly contribute large nonpoint source nutrient load-
ings to a lake. In many cases, these often difficult-to-quantify loadings exceed
point sources as the most significant contributors of nutrients. Unlike municipal or
industrial waste discharges, however, waste derived from confined animal lots is
4-10
-------
Table 4.1.—Basic land use descriptions
LEVEL I
LEVEL II
1, Urban or built-up land
11. Low density residential
12. Medium density residential
13, High density residential
14. Commercial and Industrial
15. Land under development
16. Other urban or intensively used land
2. Agricultural land
21. Row crops
22. Nonrow crops
23. Pasture
24. Confined feeding areas
25. Mixed agriculture
26. Other agricultural land
3. Forestland
31. Deciduous forestland
32. Evergreen forestland
33. Mixed forestland
4. Water
41. Streams and canals
42. Lakes and reservoirs
43. Forested wetland
44. Nonforested wetland
riot released at a relatively constant rate or normally discharged to a receiving
water from a pipe, nor is the waste monitored under the National Pollution Dis-
charge Elimination System (NPDES).
To evaluate feedlot contributions, variations of an Agricultural Research Ser-
vice (ARS) technique (Young et al. 1981) are commonly used to quantify nutrients
delivered to a receiving water from a confined animal operation. Figure 4.5 and
Tables 4.2 and 4.3 are provided as an example of the type of inventory information
needed to quantify this pollution source when this ARS-type evaluation is used.
The potential magnitude of loadings is quantified from the information identified in
Part A of Table 4.2. The hydrologic and pollutant reduction analyses require the
SCS Group and Curve number information summarized in Table 4.3. Further
guidance on appropriate types and level of information for more detailed evalua-
tion of nonpoint sources can be found in A Conceptual Framework for Assessing
Agricultural Nonpoint Source Projects (N.C. Agric. Ext. Serv. 1981).
In many cases, a fairly detailed description of watershed conditions should
have been compiled as part of the Phase I study data; updates by office reviews
(verified by field inspections) will often be all that is necessary for a Phase II sum-
mary, however. To be useful to the local lake project sponsor, the data compiled
from this effort should not be overly complex and must be presented in an easy-to-
understand format. Existing land use and management should be characterized
as a percentage of the total watershed size and by potential contribution to the
overall nutrient loading. Information on critical areas needing protection can be
identified on a separate map to serve as an ongoing reference for the local or-
ganization.
Level II; Limited Stream Monitoring
Applicability
Although the basic information obtained from the Level I watershed inventories
and evaluations will provide significant insight into watershed conditions, it will not
provide data on the actual nutrient or sediment characteristics of the lake's
tributary streams.
Watershed data obtained from a short-term, time-series-based tributary stream
sampling program have formed the basis from which many Clean Lakes Program
... the data compiled
from this effort
should not be overly
complex...
4-
11
-------
Tributary Area (b)
' Farmstead Tributary Araa (b)
i Barn Roof
] Tributary
Araa (b)
Adjacent Araa (c)
Adjacent Araa (c)
v Buffer Path (d)
Discharge Point
Limits of Animal Lot.
Intermittent or
Perennial Stream
Limits of Adjacent
Tributary and Buffar Areas.
Direction of Surface Water Runoff
(Overland Flow)
Direction of Surface Water Runoff,
(Channelized Flow)
Location of Manure Stack...
Figure 4.6.—"typical sketch of feedlot area showing general hydrologlc Information.
watershed projects have been implemented. In these studies, it has been com-
mon to determine pollutant concentrations from analysis of tributary stream grab
samples collected on a fixed interval basis over one year. In some cases, loadings
have been estimated by using simple averaging techniques applied to both con-
centration and instantaneous flow data. Unfortunately, not all of these monitoring
efforts have provided valuable information.
The tendency to use shortcuts or inappropriate methods is not limited to Clean
Lakes Program projects. Even when researchers have acquired data in attempts
to examine nutrient loadings from various land use practices, the information has
often been of less than desirable quality. Beaulac (1980) cited inappropriate or in-
consistent methods as the primary reason he rejected information on nutrient ex-
port coefficients from many research studies.
However, a limited stream monitoring program can provide insight into water-
shed conditions—in some cases. A program of limited or Level II watershed
monitoring is justifiable where the objectives are limited to verifying conditions
found during the Phase I study and detecting order of magnitude differences be-
12
-------
Table 4.2.—Typical feed lot inventory data sheet with information commonly needed
for detailed evaluations
A. Animal Lot Information
1, Animal data
Animal types:'
Number of animals:
% Time animals in lot:
2, Lot configuration
Animal lot: Size (Acres): Lot Surface: % Paved
% Sod SCS CN
% Bare
Cleaning frequency of lot (days)?
B. Area that drains across the animal lot (Tributary Area b in Figure 4.5)
Landcover:
Size (Acres):
Soil Hydrol. Grp.:
SCS Curve Number:
C. Area that drains to the buffer path (Adjacent Area c in Figure 4.5)
Landcover:
Size (Acres): *
Soil Hydrol. Grp.:2 , '
SCS Curve Number:2
D. Buffer Path (d in Figure 4.5)
Landcover: .
Size (Acres):
Soil Hydrol, Grp.:
SCS Curve Number:
'Animal Types
Slaughter steer Sheep
Young beef Turkey
Dairy cow Chicken
Dairy young stock Duck
Swine Horse
Feeder pig
?See Table 4.3 for SCS Group and Curve Numbers
tween watershed sub-basins. Defining average sediment and nutrient loadings is
an unrealistic objective for most Level II monitoring programs.
A limited stream monitoring program can also be valuable if the local sponsor
will continue the monitoring over a long time period. In these cases, identification
of major changes in runoff quality/quantity and documentation of long-term
average loadings are possible.
Construction Phase
In those lake projects where a limited stream monitoring program is being in-
stituted, there will be little or no difference in protocol between the construction
and post-monitoring phases. And in some cases where major watershed improve-
ment practices are being installed and pollutant delivery conditions are extremely
variable, stream monitoring data acquired during the construction phase, unless
very site-specific, will be of little value to the project manager. By the time enough
data can be obtained to define conditions, the conditions themselves will have
changed.
If short-term, point source discharges are required during the construction
phase, then periodic collection of water quality data from these sources should be
considered. A case example might be where a lake is being hydraulically dredged
and spoil site return carriage water is directed back to the lake or tributary stream.
Monitoring the quality of return carriage water can help ensure that the site is
4
-------
Equally important as the
technical design criteria
are those criteria
necessary to ensure that
the local sponsor will
continue to collect data
following project
completion.
Table 4.3.—Surface condition constant and soil conservation service curve numbers
for various cover conditions (Source: Hydrology Guide for Minnesota, USDA-SCS,
St. Paul, MN)
SCS CURVE NUMBER (CW)
SURFACE SOIL SOIL SOIL SOIL
COVER CONSTANT GROUP A GROUP B GROUP C GROUP D
Fallow
0.22
77
86
91
94
Row crop:
Straight row (up & dn)
0.05
67
78
85
89
Contoured
0.29
65
75
82
86
Small grain
0.29
63
74
82
85
Legumes or rotation meadow
0.29
58
72
81
85
Pasture:1
Poor
0.01
68
79
86
89
Fair
0.15
49
69
79
84
Good
0.22
39
61
74
80
Permanent meadow
0.59
30
58
71
78
Woodland
0.29
- 36
60
73
79
Forest w/beavy litter
0.59
25
55
70
77
Farmsteads
0.01
59
74
82
86
Grass waterways
1.00
49
69
79
84
Animal lot:
Uripaved 91
Paved 94
Roof area 100
'Pasture should be considered "poor" if it is heavily grazed with no mulch. "Fair" pasture has between 50 percent
and 75 percent plant cover, and "good" pasture is lightly grazed and has more than 75 percent plant cover.
being properly operated. Where discharge occurs, the effluent must comply with
standards defined in section 401 of the Clean Water Act. State or local regulatory
agencies will normally require discharge monitoring as a permit condition, which
may preclude the need for developing a separate monitoring plan.
Where watershed conditions are not expected to vary greatly during project im-
plementation, the construction phase monitoring plan can be identical to that of
the post-project plan.
Post-project Phase
A limited stream monitoring program should be instituted if:
• Qualitative information on tributary streams will be of value; or
• Identification of previously unidentified sources of contamination is
possible; or
• The design will serve as an example of a long-term monitoring program that
will be continued by the local sponsor,
¦ Design Considerations for Limited Watershed Monitoring. Much of the
value of a limited monitoring program depends on its long-term continuation. Past
studies have often had problems in attempting to normalize data that were
gathered over a short time period, often over a limited range of runoff conditions.
Most of the following discussion is therefore oriented toward the situation where
the local project sponsor is expected to continue the monitoring program following
completion of the formal Phase II project.
Equally important as the technical design criteria are those criteria necessary to
ensure that the local sponsor will continue to collect data following project comple-
tion. These factors include:
• that the data be relevant to the lake problem;
• that costs be kept low and in perspective with the project as a whole;
4-14
-------
• that data be easily obtained (this is especially important if an unpaid local
volunteer is collecting the data); and
• that a mechanism is in place to provide a periodic, professional
interpretation of the data.
¦ Critical Parameters. The following parameters are those most often needed
for lake projects and most likely to have been obtained during the Phase I study.
These parameters will also be commonly obtained during the post-project phase
under both a limited and a comprehensive monitoring effort. Specifically, they are:
• Suspended solids
• Total phosphorus
• Dissolved reactive phosphorus
• Total Kjeldahl nitrogen
• Ammonia nitrogen, and
• Nitrite + nitrate nitrogen.
Additional parameters obtained during earlier studies should also be
monitored if necessary for specific project concerns. Examples include fecal
coliforms, fluoride (an indicator of municipal wastewater discharges), potassium
(an indicator of feediot pollution), chloride, metais, pH, and pesticides. Again,
careful consideration must be given to the cost of the monitoring program to en-
courage continued local sponsorship. For example, analytical costs will be sig-
nificantly reduced if elimination of the nitrogen series can be justified.
¦ Sampling Frequency. There have been many studies to determine the fre-
quency of sampling necessary to characterize stream conditions. In a review of
several intensive stream monitoring projects, Allum (1977) identified the value of
preliminary data in helping to reduce sampling frequency and of making sampling
decisions on a stream-by-stream basis. On the basis of Allum's work and a similar
analysis by Walker (1977), Reckhow et al. (1980) suggested that a sampling inter-
val of about 14 to 28 days could be used to characterize phosphorus concentra-
tions as a general guideline for larger watersheds. They also proposed that
sample collection not be systematic with respect to time (e.g., every two weeks),
but that it be systematic with respect to flow—that more intensive sampling be
done during high flow periods. In his review, Allum demonstrated that the standard
error of the annual phosphorus flux generally varied between 10 and 20 percent of
the true flux for the 14- to 28-day sampling period.
An analysis of Phase I study information will serve as the starting point for
developing the technical design for a limited monitoring program. To minimize
error, Phase I data should be examined to determine the importance of various
flow events and flow periods so that the sampling strategy will be most intensive
during those periods when highest loadings occur. For example, where there are
important spring runoff events, specifications might state that samples should be
collected once each week between March and May and monthly thereafter. How-
ever, Gaugush (1987) proposes that exact sampling dates should not be specified
and that sampling should be done on a random basis within specified time frames.
When storm events are important to loadings, identifying a sampling design be-
comes more difficult, especially if samples are to be manually collected by a local
sponsor. Storm-generated loadings are usually extremely variable and often do
not exhibit the ideal "first flush" phenomenon because of fluctuating rainfall inten-
Storm-generated
loadings are usually
extremely variable and
often do not exhibit the
ideal "firstflush"
phenomenon.
4-15
-------
... obtaining continuous
flow information is
clearly the preferable
approach.
Few long-term water
quality records exist from
which to judge
effectiveness ofnonpoint
source watershed
controls.
sities. Experience has also shown that the most significant events have the
frustrating tendency to occur during the night or on holidays and weekends.
Although even collection of the minimum 13 to 28 samples per year per
tributary may start stretching the resources of a local sponsor, a lesser sampling
frequency should be avoided as errors may quickly result. Occasionally a careful
analytical evaluation of historical data may show that acceptable information can
be obtained from a lesser sampling frequency. The watershed case study in Chap-
ter 7 illustrates an evaluation of the intensity of future sampling needed to estimate
nutrient concentrations.
The representativeness of the information is also a consideration often as im-
portant as the numbers and timing of sampling. Care must be exercised when
selecting a proper sampling site location, identifying appropriate field sampling
techniques, and acquiring appropriate quality assurance samples. Chapter 3
provides more detail on sampling methods.
¦ Streamflow Measurements. Some method of streamflow estimation is neces-
sary for all post-project monitoring where water quality samples are being ob-
tained, if only to define the representativeness of a particular sample. Too often
volunteers have taken collections from ponded areas, even when no flow is
present, just to obtain a sample. Use of these samples to describe average condi-
tions then introduces significant error to the data set.
Streamflow measurement can be accomplished continuously or instantaneous-
ly as described in Chapter 3. Often the Phase I study will have collected
streamflow data on either a continuous or instantaneous basis. Maintaining the
operation of a continuous record gaging station that is already in place is often jus-
tifiable. In many instances, a local sponsor who is interested in protecting sig-
nificant investments made to improve land use practices will want to install a
long-term station. Nevertheless, obtaining continuous flow information is clearly
the preferable approach.
Often, however, only instantaneous streamflow measurements were made
when samples were collected during the Phase I study. If a stage-discharge
relationship was developed at these sites, flow rate determinations can be made
by simple stage readings. Instantaneous measurements should be made peri-
odically to verify the stability of the stage-discharge relationship. Also needed are
special attempts to improve the rating curve at the station by taking direct stream
gagings during the high flow runoff events.
Although continuous streamflow monitoring is desirable, it can add significantly
to the cost of data acquisition. Even instituting a periodic, instantaneous
streamflow measurement program can be a problem for the local sponsor be-
cause of increased costs (an additional $5,000 to $10,000) and the high level of
expertise needed to obtain this data.
Level III: Comprehensive Watershed Monitoring
Applicability
The purpose of monitoring a watershed where improved practices have been in-
stalled is to determine how much watershed practices have reduced pollutant
loadings to the lake.
Unfortunately, this question has not often been adequately answered. Few
long-term water quality records exist from which to judge effectiveness of nonpoint
source watershed controls, with the problem being especially acute for smaller
watersheds. Without more information that is obtained over a wide range of condi-
tions at different sites, there will continue to be limited understanding of the effec-
tiveness of various watershed practices and the evolution of nonpoint source
4-
16
-------
Typical Plan Element for Intermittent Sampling
The following is an example specification for the intermittent monitoring of
a tributary stream.
1. Instantaneous flow measurements should be made at the tributary
stream locations identified on an enclosed map.1 The flow measure-
ments should be made once a week from March through May and
monthly, thereafter. The techniques used for making these measure-
ments should conform to those recommended in Discharge Meas-
urements at Gaging Stations, Techniques of Water Resources
Investigations, Book 3, Chapter A8 (Buchanan and Somers, 1969).
2. Water samples should be collected at the tributary site(s) where
streamflow measurements are required. Grab samples should be col-
lected where sufficient turbulence exists in the tributary stream to
ensure a representative sample. Where flow is less turbulent, special
collection methods should be used.
3. Water samples should be analyzed for:
* total phosphorus,
* dissolved reactive phosphorus,
* total Kjeldahl nitrogen,
* nitrite + nitrate nitrogen, and
* suspended solids.
Analytical methods used shall be those described in Chapter 3 of this
manual.
control programs. Many attempts to document the reduction of nutrient and sedi-
ment loadings from implementation of watershed controls have been unsuccess-
ful because monitoring programs were too limited and of too short a duration to
provide accurate data.
By their nature, Clean Lakes Phase II post-project watershed monitoring
strategies are rarely comprehensive enough to provide these data, nor should
they be. The best use of funds available during project implementation is to as-
sure proper completion of the project. Nonetheless, where good pre-project data
exist, where a comprehensive monitoring network is in place, and where local in-
terest is high, finding a way to continue data acquisition should be a high priority.
The following brief discussion describes some of the methods used by re-
searchers to document the effectiveness of various watershed control measures
and provides general background information on the level of effort required.
1 Normally, the major tributary to the lake will be sampled. Where significant differences exist be-
tween subwatersheds (e.g., heavily urbanized versus agriculture-dominated) or for lakes having com-
plex pianlmetric configurations, an additional site or sites can be of value. Selection of sampling sites
will always be aided by evaluating Phase I Information.
4-17
-------
¦ Design Considerations for Comprehensive Watershed Monitoring. Several
researchers have described the watershed design criteria necessary to quantify
watershed loading changes. Spooner et ai. (1985) identified three basic ex-
perimental designs for watershed monitoring (before and after, above and below,
and paired watersheds) and described the advantages and disadvantages of each.
Reckhow et al. (1980) identified considerations necessary for designing a sam-
pling program and for analyzing data. They observed that, with the understanding
that water quality varies in time and space as a function of many macroscopic and
microscopic processes, water quality data series may be
* autocorrelated
* censored (due to observations below detection limits)
* non-normally distributed
* irregularly spaced in time (perhaps because of missing values) and
* subject to trends or seasonal patterns.
When considering selection of a statistical method for analysis of stream data,
it is important to consider these characteristics along with the assumptions in-
herent to a particular statistical method. The Neuse River case study in Chapter 7
provides an example of how these factors can be considered during the develop-
ment of a future monitoring strategy.
In some cases, such as where wasteload allocations are being made, the
analytical tools selected to evaluate lake response to nutrient loadings will define
data requirements. In these situations, more detailed guidance can be found in
Chapter 2 of EPA's Technical Guidance Manual for Performing Waste Load Al-
locations, Book IV: Lakes and Impoundments (U.S. Environ. Prot. Agency, 1983).
Associated with most monitoring sites are continuous stage recorders and
automatic water samplers programmed to collect samples on a time-series or
flow-proportioned basis. Data summations that relate stage and rating curve infor-
mation to provide a continuous record of flow are usually computer-generated.
Loadings are developed with integration techniques after water quality data have
been obtained over a wide range of flow events.
It is expensive to obtain this type of quality information. For example, the cost to
monitor a single, easily accessible site can range from $10,000 to $30,000 per
year. Since several years of record and more than one site are normally needed,
the cost to monitor just one project often exceeds $100,000.
Streamflow Monitoring
Sub-watershed,
conditions are often
characterized by
instantaneous
measurements.
Continuous monitoring of flow is almost always a necessity for a comprehensive
watershed monitoring project. The method used to measure flow will depend on
the characteristics of the site where flow is to be measured. Tributary stream flow
measurement sites are most commonly located on open channels. Usually, the
stations installed at these sites incorporate continuous float or pressure gage
stage recorders. Where a natural control is used, a rating curve is developed by
making several instantaneous flow measurements over a variety of flow rates and
noting stage at the time of measurement. Occasionally, control structures with pre-
determined rating curves, such as broad-crested weirs or flumes, are installed in
the streambed. An experienced hydrologist must identify the specific techniques
to measure streamflow for each site. Chapter 3 describes the most important of
these methodologies in more detail.
Additional detail on sub-watershed variability and conditions is often obtained
by supplementing continuous flow record stations with instantaneous flow meas-
4-
18
-------
urements on upstream tributary sites. To account for year-to-year variability, both
watershed and subwatershed sites are normally monitored over a period of
several years.
Characterization of Constituent Concentrations in
Streamflow
In most cases, the use of automatic sample collectors is necessary to ensure col-
lection of storm event flow data; this is especially true for urban runoff sites. Only
larger, more stable river systems can be adequately characterized by manual
sample collection.
Methods for obtaining representative samples are site-specific. In streams ex-
hibiting stratified conditions or longitudinal variability, special compositing techni-
ques must be used. Similarly, when automatic sampling stations are installed,
correction factors must be developed as most samplers will have a fixed sample
intake location that may or may not be representative of overall stream conditions.
Chapter 3 describes appropriate sampling techniques in more detail.
Watershed Inventories
Without supplementary information on watershed conditions, even the best, most
comprehensive tributary stream database will be of little value. Description of
watershed conditions, such as that which might be obtained from a Level I water-
shed inventory, is always a component of a comprehensive watershed monitoring
strategy. It must also be noted that even the most complete monitoring station
defines only an average of the conditions upstream of the monitoring site.
Interpretation of Tributary Stream Data
The following sections describe methodologies often used in both limited and
comprehensive watershed tributary stream data collection programs. Intermittent-
ly collected tributary stream concentration data can be interpreted in several
ways. Tributary concentration data can be used to
• calculate loadings if combined with streamflow data,
• evaluate transport mechanisms, and
• analyze and assess nutrient and sediment sources.
In some instances, loading calculations will be warranted and insights into
transport mechanisms may be possible, although results of such calculations
must be used cautiously. Generally, a further understanding of pollutant sources
will be the most valuable information obtained from a limited watershed sampling
program and, in some cases, trends may be detected. Clearly, the best informa-
tion will be obtained from sampling programs continued over several years.
In most cases, the use of
automatic sample
collectors is necessary to
ensure collection of
storm event flow data.
... even the most
complete monitoring
station defines only an
average of the
conditions upstream of
the monitoring site.
Loading Calculations
Estimation of loadings to lakes is the most common use of streamflow concentra-
tion data. Unfortunately, inappropriate techniques are all too often used to calcu-
lated these loadings. Frequently, loadings are derived by using the product of the
arithmetic average nutrient or sediment concentrations and the arithmetic
average flow rate. This method should be avoided, as it will almost always under-
estimate loadings where storm-generated events are important.
4-
19
-------
Walker (1987) concluded that the flow-weighted concentration combined with
average flow is the best estimator when concentration does not vary greatly with
flow. Verhoff et al. (1980) found that a flow interval method relating phosphorus
flux to streamflow provides the best fit to the tributary data they evaluated.
When continuous flow information is available, an integration technique has
been favored by the U.S. Geological Survey for load calculations. Walker (1987)
developed an excellent software package (called FLUX) that allows easy calcula-
tion of loadings by use of several different techniques.
An example of the mid-interval technique (Porterfield, 1972), which is a method
commonly used where noncontinuous data have been obtained, is described
below.
¦ Mid-Interval Technique. A better method than using arithmetic averages, and
one that can be simply explained to others, is to assign representative sample
data to corresponding flow data. Although this method is still very limited in its
ability to describe loads because of an inability to yield an estimate of precision, it
can be used to provide insight into watershed conditions.
In the simplest calculations, the measurement of concentration is combined
with the streamflow measurement made during sample collection to calculate an
instantaneous loading which is then assumed to characterize the tributary
transport over a certain time interval associated with that sample. Generally, the
time interval used is equivalent to one-half the time interval between that sample
and the preceding sample plus one-half the time interval between that sample and
the following sample. Multiplying the instantaneous loading for each sample by
the time interval for each gives a total load for the time period associated with that
sample. Summing the total loads for all the individual samples yields the total load
for the time period covered by the sampling program. Expressed as a formula this
procedure is:
Total Load = Sum QQiTi
where
Ci = concentration of the ith sample
Qi = instantaneous discharge when sample was collected and
Ti = the time interval associated with the ith sample.
Again, as a caution, loading information calculated from an intermittent sam-
pling program must be used carefully, as nonpoint source generated loadings are
very dependent on storm events.
¦ Time-Weighted Mean Concentrations. When samples are collected over a
uniform, fixed interval, average concentrations can be determined by directly
averaging concentrations since each sample characterizes the stream for the
same length of time. When samples are not collected on a regular basis or are col-
lected over storm events, individual samples do not characterize the stream for
equal lengths of time. Therefore, to estimate the average concentration, each
sample has to be weighted according to the length of time it is used to represent
the stream system. Time weighted mean concentrations (TWMC) are calculated
by:
TWMC = (Sum QTi)/(Sum Ti)
where
Ci is the concentration of the ith sample and
Ti is the time period for which the ith sample is used to
characterize the stream concentration. It is equal to
one-half the time interval between the samples
immediately preceding and following the ith sample.
20
-------
Time-weighted mean concentrations are of most value to analysts interested in
the exposure of biota to particular pollutants; for instance, where exposure to or-
ganisms living in a stream reach is important, and where the corresponding flow
rate is unimportant. Analysts concerned with lake water quality usually do not use
average concentration information, since this technique will not provide good
loading estimates.
¦ Flow-Weighted Mean Concentrations. Where total loadings are of concern
(as they are in most lake studies), average concentrations are estimated by
weighing the individual samples with their associated flows. The resulting average
concentration is referred to as a flow-weighted mean concentration (FWMC) and
is equivalent to the total load divided by the total discharge for the period of inter-
est. The FWMC is calculated by
FWMC = (Sum CiTiQi)./(Sum TjQi)
where
Ci is the concentration of the ith sample and
Ti is the time period for which the ith sample is used to
characterize the stream concentration. It is equal to
one-half the time interval between the samples
immediately preceding and following the ith sample.
, Qi is the instantaneous discharge at the time of the ith sample.
Source Analyses
Where different tributaries or sub-basins are sampled, order of magnitude dif-
ferences can sometimes be observed, leading to identification of unknown
nutrient or sediment sources. This source identification should have occurred
during the Phase I study phase. However, since watershed and hydrologic condi-
tions constantly change, one can never be completely confident that ail potential
sources have been adequately identified and characterized.
The following interpretive procedure was described by Baker (1988) and can
be used to define the relative significance of point source discharges versus non-
point source contributions. In some cases, this procedure can also be used to
define the significance of groundwater contributions.
Where different
tributaries or sub-basins
are sampled, order of
magnitude differences
can sometimes be
observed, leading to
identification of unknown
nutrient or sediment
sources.
¦ Comparisons of Flow-Weighted and Time-Weighted Mean Concentra-
tions. There are often considerable differences between the FWMCs and the
TWMCs when nutrient and sediment information is being analyzed. FWMC to
TWMC ratios greater than 1 indicate that the concentrations are increasing with
increasing discharge—suggesting important nonpoint source pollutants.
Where significant point sources are present in the watershed, the concentra-
tions of pollutants tend to decrease with increasing stream flow as dilution plays a
greater role. FWMC to TWMC ratios less than 1 suggest important point source
contributions.
FWMC and TWMC ratio analyses are an example of a simple comparative
method used to evaluate differences between watersheds and to identify the rela-
tive importance of different pollutant sources.
¦ Regression Analysis. Flow and concentration data can be plotted to deter-
mine if there is a deterministic relationship and, if one is found, a fiow-concentra-
tion regression model can be fitted (Walker, 1987). This type of regression
4-:
21
-------
Phosphorus
concentrations in streams
will decrease with
increasing flows where
point sources
dominate.
analysis provides similar information on the relative importance of point- and non-
point-derived pollutants. In general, phosphorus concentrations in streams will
decrease with increasing flows where point source contributions are important;
where nonpoint sources dominate, phosphorus concentrations will generally in-
crease with increasing flows. Subsequent analyses, if desired, can then be made
on regression model residuals.
¦ A General Approach to Water Quality Monitoring Design. In developing a
water quality monitoring design, the following tasks should usually be considered
(Hirsch et al. 1982; Reckhow et al. 1989):
1. Examine the historical data for patterns that may be attributed to a pre-
vious time trend, a seasonal cycle, or a relationship between streamflow
and concentration. These patterns are called "deterministic" because
their cause is determined, or known.
2. In each case, use the data to describe the deterministic pattern mathe-
matically, with a simple deterministic mathematical equation.
3. Subtract, or remove, the mathematical estimate of each deterministic
pattern from the water quality data, leaving a "residual." The residual is
the observed water quality concentration minus the water quality con-
centration predicted using the deterministic mathematical equation.
4. Examine the residuals to ascertain that they are stationary (e.g., that the
average and the variability do not change over time) and that they lack
persistence (i.e., that the residual at any one sampling date is not corre-
lated with the residual a fixed number of time periods apart).
5. If necessary, transform (e.g., take the logarithm of) the residuals to
achieve stationarity. If appropriate, characterize the persistence using
autocorrelation analyses.
6. Use the residuals to estimate the background variance (noise), correct-
ing for autocorrelation.
7. Finally, use the background variance to define the relationship between
the number of samples and the magnitude of the linear trend, again ac-
counting for autocorrelation.
The watershed case study in Chapter 7 describes how this approach was fol-
lowed in an analysis of a data set obtained from the Neuse River near Smithfield,
North Carolina.
4-:
.22
-------
Chapter 5
In-lake Restoration
Techniques and
Monitoring
. ... . . .
Summary
1. Different lake restoration techniques require different
monitoring considerations.
2. Criteria for interrupting a lake restoration project during
the implementation phase are necessary to prevent en-
vironmental damage.
3. Each in-lake measurement should have a purpose that is
directly related either to project evaluation or to protect-
ing the lake environment from adverse impacts during
the treatment phase.
4. This chapter recommends an in-lake monitoring plan for
each lake restoration technique independently from the
other techniques discussed within this manual, thereby
allowing the user to proceed directly to a monitoring
plan for a specific project.
y
-------
Background
Evaluation should
assess not only the
effectiveness of the
restoration or protection
technique but also, more
broadly, whether the
project as a whole
achieves its objectives.
Chapter 5 presents appropriate in-lake monitoring designs as they pertain to the
most frequently used lake eutrophication and acidification restoration techniques.
These specific monitoring plans are designed to be consistent with U.S. Environ-
mental Protection Agency protocol as specified in 40 CFR Part 35 and to offer a
standard monitoring approach for evaluating each project's success.
In-lake measurements are needed to evaluate most lake treatments. The
parameters that are measured should pertain directly either to evaluating results
of the treatment or to protecting the lake environment from adverse effects of
treatment. Evaluation should assess not only the effectiveness of the restoration
or protection technique but also, more broadly, whether the project as a whole
achieves its objectives; therefore, specifications for monitoring are given for each
restoration/protection technique. However, only rarely do lake restoration projects
consist of only one technique; more often, a combination of two or more techni-
ques are required. Therefore, the suggested monitoring plans may vary according
to the individual project needs.
Chapter 5 concentrates on in-lake restoration techniques and monitoring re-
quirements. More often than not, in-lake restoration measures are implemented in
close conjunction with watershed protection measures. Watershed monitoring Is
treated separately in Chapter 4, and long-term monitoring needs, which are im-
plied for most in-lake restoration projects, are discussed in Chapter 6.
Tables 5.1 and 5.2 summarize the lake restoration techniques discussed in this
chapter and itemize the relative importance of the most common lake water quality
parameters for in-lake measurement during and following treatment. The general
role and importance of each parameter in a lake environment are discussed in
detail in other documents, particularly The Lake and Reservoir Restoration
Guidance Manual (U.S. Environ. Prot. Agency, 1988).
The eutrophication section is further divided into techniques designed primarily
to control nuisance algae, maintain or increase water depth, or control nuisance
plant growth. Discussion of the mode of operation and effects of each technique is
limited to facts needed to understand the monitoring approach. More information
on the operational aspects of each technique is available from numerous in-
dividual publications, but Cooke et al. (1986) have published the most comprehen-
sive review to date.
5-2
-------
Table 5.1.—In-lake monitoring during treatment phase*
RESTORATION
TECHNIQUE
DO TEMP TP DRP NH4-N
NO23-N TKN CHL MACRO pH ALK SD A
TX
Control Algae
Alum
Dilution/flushing
Aeration/circulation
Hypolimnetic aeration
Hypolimnetic withdrawal
Sediment oxidation
Food web manipulation
Increase Depth
Dredging
Control Nuisance
Aquatic Plants
Drawdown
Harvesting and sediment tilling
Chemicals
Grass carp
Mitigate Acidic
Conditions
In-lake liming
Watershed liming
— — — e
e — — —
— — e — — • e
— — e — — e
"This monitoring is independent of the monitoring done during a Phase I diagnostic/feasibility study.
1 "e" represents essential water quality parameters
2"u" represents useful but nonessential water quality parameters
Abbreviations: DO = dissolved oxygen: TEMP = temperature; TP = total phosphorus; DRP = dissolved reactive phosphorus; NH4-N = ammonium nitrogen; NO23-N
nitrogen; CHL = chlorophyll a; MACRO = macrophytes; ALK = alkalinity; SD = Secchi disk; A = algae; Z = zooplankton; TX = see text for specific parameters
nitrite + nitrate nitrogen; TKN = total Kjeldahl
-------
Table 5.2.—In-lake monitoring following treatment phase*
RESTORATION
TECHNIQUE DO TEMP TP DRP NH4-N NQ23-N TKN CHL MACRO pH ALK SD A Z TX
Control Algae
Alum
e1
e
e
e
—
—
—
e
u2
e
e
e
—
—
e
Dilution/flushing
e
e
e
e
e
e
e
e
u
—
—
e
—
—
—
Aeration/circulation
e
e
u
u
u
u
u
e
u
e
— ¦
e
e
—
—
Hypolimnetic aeration
e
e
e
e
e
u
u
e
u
—
—
e
—
—
e
Hypolimnetic withdrawal
e
e
e
e
e
u
u
e
u
e
u
e
—
—
—
Sediment oxidation
e
e
e
e
e
e
—
e
u
e
e
e
—
—
e
Food web manipulation
e
e
u
u
u
u
u
e
u
—
—
e
—
e
e
Increase Depth
Dredging u uuu— — — u e — — u — — e
Control Nuisance
Aquatic Plants
Drawdown
u
u
u
u
—
—
—
e
e
—
—
e
—
—
—
Harvesting and sediment tilling
u
u
u
u
—
—
—
e
e
—
—
e
—
—
—
Chemicals
e
e
u
u
u
u
u
e
e
—
—
e
—
—
—
Grass carp
e
e
u
u
u
u
u
e
e
—
—
e
—
—
e
Mitigate Acidic
Conditions
In-lake liming e e — — — — — e — eee — — e
Watershed liming e e — — — — — e — eee — — e
"This monitoring presumes that, in many cases, comparable data were collected during a Phase I diagnostic/feasibility study.
' "u" represents useful but nonessential water quality parameters
2"e" represents essential water quality parameters
Abbreviations: DO = dissolved oxygen; TEMP = temperature; TP = total phosphorus; DRP = dissolved reactive phosphorus; NH4-N = ammonium nitrogen; NO23-N = nitrite + nitrate nitrogen; TKN = total Kjeldahl
nitrogen; CHL = chlorophyll a; MACRO = macrophytes; ALK = alkalinity; SD = Secchi disk; A = algae; Z = zooplankton; TX = S9e text for specific parameters
-------
OBJECTIVE
Control Nuisance Algae
Summary
1. The following techniques for algal control and suggested monitor-
ing plans are discussed: nutrient precipitation/inactivation, artificial
circulation, hypolimnetic aeration, hypolimnetic withdrawal, dilu-
" " tion/flushing, food web manipulation, and sediment oxidation.
2. The common elements of a sampling design throughout most of the
techniques are total phosphorus, dissolved reactive phosphorus,
chlorophyll a, dissolved oxygen, temperature, and Secchi depth.
An overabundance of nuisance algae is one of the most common symptoms of ac-
celerated eutrophication in lakes. Direct impairments from these infestations in-
clude elevated turbidity in the water, occasional release of toxins, and lastly, taste
and odor problems, which are particularly troublesome for waterbodies that serve
as water supply reservoirs. Indirect impairments include accumulation of decay-
ing biomass that results in lowered dissolved oxygen in the water, muck ac-
cumulation, and adverse changes to the fish community.
Nuisance algal growths in lakes result in large part from excessive supplies of
nutrients, although other factors may contribute to the problem. Techniques to
counter these growths are usually directed toward reducing the supply of
nutrients to the lake. In particular, these techniques target phosphorus, since it is
the nutrient that can be practically controlled to limit the growth of nuisance
algae.
The techniques enumerated in this section are those that have become "stand-
ard" through repeated testing and successful use. A small number of less-tested
techniques developed since the start of the Clean Lakes Program are not included
here.
CONTROL TECHNIQUE #1:
Phosphorus Precipitation/inactivation with Alum
Technical Considerations
The addition of aluminum salts (aluminum sulfate or sodium aluminate) is a
proven lake restoration technique for controlling algal growth by creating a
nutrient-limiting environment. The technique is straightforward and relies upon the
affinity of aluminum complexes for phosphorus.
Alum, as aluminum sulfate is called, is usually applied in liquid form. Once the
alum mixes with lake water, it quickly becomes aluminum hydroxide. Dissolved
phosphorus adsorbs to the aluminum hydroxide, which precipitates toward the
lake sediments and sweeps the water clean of phosphorus. Upon reaching a den-
Aluminum sulfate is
effective on hard water
lakes to control sediment
phosphorus release.
5-5
-------
Sodium aluminate and
aluminum sulfate are
used together to treat
softwater lakes
sity equilibrium in the lake's sediments, the aluminum hydroxide forms a barrier
that sorbs phosphorus, thereby greatly reducing its transport from the sediments
to the overlying waters.
Liquid aluminum sulfate exists as an acidic medium (sulfuric acid); when added
to the water, it will consume a portion of the acid neutralizing capacity (alkalinity)
of a lake. Therefore, the pH and alkalinity of the lake must be measured as the
alum is being applied. Aluminum hydroxide forms best when the lake water has a
pH of 6 to 8. If the pH falls below 6, dissolved elemental aluminum, which is toxic
to lake biota, becomes the dominant form.
Alum application to-hardwater lakes is, depending upon the amount applied,
less likely to significantly lower pH. The application of aluminum salts to softwater
lakes is of much greater concern. The usual technique is to mix aluminum sulfate
with sodium aluminate to buffer the acidity. Sodium aluminate is the preferred buff-
er rather than a carbonate salt because it allows more aluminum hydroxide forma-
tion and, therefore, has more potential to remove phosphorus.
The use of alum should not be considered unless phosphorus loading from the
watershed has been reduced to acceptable levels.
Monitoring During Treatment
The most important parameters to monitor during the addition of alum are pH,
alkalinity, dissolved aluminum, dissolved oxygen, and temperature. The utility of
measuring phosphorus depends upon how long treatment takes. For example, if
the actual application takes from one day to two weeks, it is unlikely that the
results of a phosphorus analysis will be available in time to serve any useful pur-
pose during the treatment period. If the application time is longer than two weeks,
then phosphorus should be included as a measured parameter.
The major variable associated with the sampling design is application techni-
que. There are two basic application techniques for alum: surface and deepwater
applications. Table 5.3 gives the recommended specifications for in-lake monitor-
ing during an alum treatment.
Considerations for Interrupting Treatment
Considerations for interrupting an alum treatment should include the following
criteria:
1. Surface Application:
* If the pH of the treated surface water is 6.0 or less, or
* If the pH of the surface water changes by more than 2 standard
units.
2. Deepwater Application:
* If the pH of the water at 6 feet above or below the application depth
is 6.0 or less, or
• If the pH of the water at 6 feet above or below the application depth
changes by more than 2 standard units.
A variety of other factors can be considered in making a final judgment to stop
or allow an alum treatment. For example, if deepwater alum application is the
method of treatment, the volume of water above the application depth may be
more than sufficient to neutralize any added acidity. The change in acidity and dis-
solved aluminum may exist only until the treated area mixes with the overlying
waters.
5-'
6
-------
Table 5.3.—In-lake sampling design during alum treatment
PHYSICOCHEMICAL
A. Water Chemistry
1. Sampling Location(s)
Water samples should be collected at the site(s) selected by the project manager, usually at
the center of the lake.
2. Depth Distribution
Samples should be collected at 6-foot intervals from just below the surface to the bottom.
Care should be taken not to include suspended bottom sediments in the water samples.
3. Analytical Determinations and Sampling Procedures
(a) If the alum application takes less than two weeks, measure pH and alkalinity.
(b) If the alum application takes longer than two weeks, measure pH, alkalinity, dissolved
aluminum, total phosphorus, and dissolved reactive phosphorus.
4. Frequency and Duration
(a) If the alum application takes less than two weeks, sample every day.
(b) If the alum application takes longer than two weeks, samples should be collected as
follows:
• pH: sample daily
• alkalinity: sample daily
• dissolved aluminum: sample once per week and
• total phosphorus and dissolved reactive phosphorus: once every two weeks,
B. Dissolved Oxygen (DO) and Temperature
1. Sampling Location
Same as for water chemistry.
2. Depth Distribution
Measurements should be made at 3-foot intervals from the surface to the bottom.
3. Analytical Determinations and Sampling Procedures
See Chapter 3 for appropriate analytical and sampling techniques.
4. Frequency and Duration
Measurements should be made at weekly intervals.
The water residence time of the lake is another factor. If the lake volume is ex-
changed rapidly (one to two times a month), then the effect of lower pH and higher
dissolved aluminum may be less on local fauna; the effect on downstream fauna
should be considered, however.
The criteria for interrupting an alum treatment are based upon the toxic effects
of dissolved aluminum. It is generally agreed that concentrations of dissolved
aluminum greater than 50 ng/L can adversely affect trout (Freeman and Everhart,
1971). However, the laboratory turnaround time for aluminum analysis is frequent-
ly too long to allow a decision to be made before an alum application is completed.
The pH of the lake water dictates the chemical form of aluminum, and a pH of less
than 6 drives the formation of dissolved aluminum in the water. Therefore, the
criteria for stopping an alum project are based upon the pH of the lake water.
Nevertheless, dissolved aluminum concentrations should be recorded in the
event that a change is later observed in the biota of the lake. Although this has
never been reported in the literature for an alum-treated lake, lake acidification re-
search suggests that dissolved aluminum is one of the elements associated with
the demise of acidified lakes' game fisheries.
Example—Wisconsin's Long Lake
Long Lake, one of the first softwater lakes treated with a combination of aluminum
sulfate and sodium aluminate, was given a surface application in May 1972. The
project was useful in demonstrating the feasibility of using an aluminum sul-
fate/sodium aluminate mixture to treat softwater lakes without seriously affecting
the pH and alkalinity. Table 5.4 (unpublished data from the Wisconsin Department
of Natural Resources) presents the pH and alkalinity data before and after the
treatment.
5-7
In general, dissolved
aluminum should not
exceed 50 \ig/L.
-------
Table 5.4.—Alkalinity and pH during alum treatment of Long Lake
1 DAY 1 HOUR 3 HOURS 1 DAY
BEFORE TREATMENT AFTER TREATMENT AFTER TREATMENT
DEPTH
ALKALINITY
pH
ALK.
pH
ALK.
PH
ALKALINITY
PH
(«)
(mg/L)
(SU)
(mg/L)
(SU)
(mg/L)
(SU)
(mg/L)
(SU)
0
7.0
6.5
7.5
6.7
7.5
6.7
5.0
7.0
3
7.0
6.1
12.0
7.1
8.5
6.9
6.0
7.2
6
7.0
6.0
12.0
7.2
7.5
6.8
6.0
7.1
9
7.0
5.9
10.0
7.1
7.0
6.8
6.0
7.0
12
7.0
5.9
9.0
7.0
6.5
6.8
6.0
7.0
When the criteria for interrupting treatment are applied to the data collected for
Long Lake, it is evident that application of alum was environmentally acceptable.
The pH of Long Lake surface water was 6.5 at the time of treatment. One hour
after treating a particular segment of the lake, the pH rose to 6.7. The change of
0.2 of a pH unit was well within the limits of acceptability. The treatment was suc-
cessful, and the project proceeded without any problems.
The purpose of an alum
treatment is to achieve a
reduction in algal
abundance by lowering
the phosphorus
concentration in the lake.
Mirror Lake—a
successful lake treatment
with alum.
Monitoring Following Treatment
The purpose of an alum treatment is to achieve a reduction in algal abundance by
lowering the phosphorus concentration in the lake. Success is defined by
decreased algal standing crop (commonly measured by chlorophyll a) and phos-
phorus concentration following the treatment.
The most important phosphorus species to measure is total phosphorus, but a
variety of other forms of phosphorus are present in the lake: total dissolved phos-
phorus, dissolved reactive phosphorus, and particulate phosphorus. Chapter 3
contains a more detailed description of phosphorus forms and analytical techni-
ques.
Total phosphorus must be measured to help demonstrate project success. Total
phosphorus measurements include the sum of the total dissolved organic and in-
organic forms and the total particulate organic and inorganic forms. An alum treat-
ment should affect the dissolved inorganic form of phosphorus (dissolved reactive
phosphorus) by direct sorption on the aluminum hydroxide gel.
Figure 5.1 illustrates the impact of an alum treatment on both total and dis-
solved reactive phosphorus concentrations in Mirror Lake, Wisconsin, which was
treated with alum in May 1978. The project was successful, in part because the
external phosphorus loading to the lake was reduced to acceptable levels before
treatment (Knauerand Garrison, 1980).
An immediate reduction in the particulate phosphorus below the depth of ap-
plication is usually evident after alum addition because the phosphorus is physi-
cally entrapped by the aluminum hydroxide floe settling through the water column.
For example, the phosphorus sedimentation rate in Mirror Lake went from 9
mg/m2/day just prior to an alum treatment to 77 mg/m2/day during the alum treat-
ment as a result of the descending alum floe sweeping the water of both particu-
late phosphorus and the sorbed dissolved reactive phosphorus.
The use of alum will not affect the nitrogen compounds in the lake, however.
Unless there is other interest in nitrogen, there is no need to measure the various
forms present in a lake treated with alum.
Alkalinity and pH are necessary chemical measurements that should be con-
tinued following application of alum. These two measurements are taken to detect
adverse environmental conditions that can occur with the addition of the strong
acid associated with alum and can assist the lake manager in either ruling out or
considering alum as the cause. If adverse environmental lake conditions such as
low dissolved oxygen develop during the post-project monitoring and produce a
5-8
-------
0.16
DISSOLVED REACTIVE PHOSPHORUS
TOTAL PHOSPHORUS
i i i i r i i i i i i i i i r
OJAJOJAJOJAJOJAJOJAJ
1977 1978 1979 1980 1981
Figure 5.1—Total phosphorus and dissolvod reactive phosphorus in Mirror Lake, 1977-81.
summer fish kill, these data will be necessary to determine the cause of the event.
Otherwise, alum additions will automatically be blamed for any adverse lake
problems that occur after treatment.
Dissolved oxygen and temperature profiles should also be measured. These
calculations are needed to determine the occurrence of anoxia in the hypolimnion,
the length of time the lake remains thermally stratified, and the timing of complete
lake mixing.
The seasonal timing of lake mixing was an important consideration in determin- —;
ing why an alum treatment did not work at Pickerel Lake, a shallow waterbody in Pickerel Lake why
Wisconsin that was treated with alum in April 1973 during the spring overturn. The alum treatment did not
lake was thermally stratified by mid-May and remained stratified until late July work
through early August. Between late July and mid-September, Pickerel Lake alter-
natively mixed and stratified (Figure 5.2). During this time period, clumps of blue-
green algae were resuspended from the bottom sediments and, because of
favorable warm weather, a massive bloom occurred during August and Septem-
ber (Knauer and Garrison, 1980).
The addition of alum to Pickerel Lake was very effective in controlling phos-
phorus and algal biomass from the spring addition until the lake mixed in midsum-
mer; at that time, however, the lake experienced algal blooms of a greater
magnitude than the previous year. In the final analysis, the driving force behind
Pickerel Lake productivity was the time of summer mixing; therefore, the alum
treatment was of little value.
Documenting changes to macrophyte communities is useful in that the im-
proved water clarity caused by an alum treatment can stimulate undesirable mac-
rophyte growth in shallow areas of the lake. Table 5.5 outlines a detailed plan for
in-lake monitoring following alum treatment.
5-9
-------
Dilution needs a source
of low nutrient water.
Flushing needs to be
fast enough to wash algal
cells from the lake before
MIX
1972
ALUM
MIX
t
1973
Figure 5-2.—Temperature measurements in Pickerel Lake, 1971-72.
CONTROL TECHNIQUE #2:
Dilution/Flushing
Technical Considerations
This technique is actually two separate lake restoration practices that are often
combined to achieve improved water quality. Dilution involves the addition of low
nutrient waters to a lake or reservoir. To be effective, the low nutrient water addi-
tions should reduce in-lake nutrient concentrations, thereby reducing algal quan-
tities.
Flushing is a physical process related directly to an increase in the flushing rate
of a lake or reservoir. If the increase in flushing rate is sufficient, an increase in
algal cell washout can be expected, thereby decreasing algal biomass within the
system.
Both systems work together to reduce algal densities. The technique was used
effectively in Green Lake, Washington (Oglesby, 1969), where sufficient quantities
5-
10
-------
Table 5.5.—In-lake monitoring design following alum treatment
PHYSICOCHEMICAL
A. Water Chemistry
1. Sampling Location(s)
Water samples should be collected at the site(s) selected by the project manager, usually at
the center of the lake.
2. Depth Distribution
Samples should be collected at 6-foot intervals from just below the surface to the bottom.
Care should be taken not to include suspended bottom sediments in the water samples.
3. Analytical Determinations and Sampling Procedures
Water samples should be analyzed for alkalinity, pH, total phosphorus, dissolved reactive
phosphorus, and dissolved aluminum. Note: dissolved aluminum may be discontinued if two
consecutive samples are below 50 |xg/L. See Chapter 3 for appropriate analytical and
sampling techniques.
4. Frequency and Duration
Samples should be collected at monthly intervals for a minimum period of two years following
treatment.
B. Dissolved Oxygen (DO) and Temperature
1. Sampling Location
Same as for water chemistry.
2. Depth Distribution
Measurements should be made at 3-foot intervals from the surface to the bottom.
3. Analytical Determination and Sampling Procedures
See Chapter 3 for appropriate analytical and sampling techniques.
4. Frequency and Duration
Same as for water chemistry.
C. Secchi Disk Transparency
1. Sampling Location
Same as for water chemistry.
2. Frequency and Duration
Measurements should be made at monthly intervals during the growing season (May through
October) for a minimum of two years following treatment.
BIOLOGICAL
A. Chlorophyll a (corrected for pheophytin)
1. Sampling Location
Same as for Secchi disk.
2. Depth Distribution
A subsample should be obtained from an integrated sample representing a water column
equal to 0-6 feet from the surface.
3. Analytical Determination and Sampling Procedure
See Chapter 3 for appropriate analytical and sampling techniques.
4. Frequency and Duration
Same as for Secchi disk.
of low nutrient water were available from Seattle's potable water supply. Reduced
algal biomass in Green Lake was attributed to a reduction of in-lake phosphorus
concentrations and increased cell washout.
In another Washington project, Moses Lake received dilution water from the
Columbia River and showed a substantial improvement in water quality during the
periods of water additions (Welch, 1979). Algal reduction was attributed to a
lowering of inorganic nitrogen concentrations and cell washout.
While it is possible that other factors may have contributed to the lower
biomass of blue-green algae in these lakes—e.g., iron limitation (Welch and Pat-
mont, 1980)—it should be noted that in both cases when the dilution water was
discontinued, the lake water quality reverted back to the original pretreatment
conditions.
5
-------
Monitoring During the First Two Weeks of Treatment
The success of a dilution/flushing project is dependent upon a source of low
nutrient water in amounts that allow a substantial increase in the lake flushing
rate. In most cases, the input of new water must continue throughout the growing
season. The protocol described in Table 5.6 addresses monitoring only of the ad-
ditional waters during the first two weeks of operation.
Table 5.6.—Monitoring design for the first two weeks of a dilution/flushing project
MONITORING THE ADDITION WATERS
PHYS1COCHEMICAL
A. Flow Measurement
Continuous flow measurements should be made throughout the two-week period at the outlet of
the conduit that delivers the dilution/flushing waters to the lake. Flows should be estimated and
reported on a daily basis. See Chapter 3 for appropriate flow measurement techniques.
B. Water Chemistry
A sample should be collected on a daily basis during the two-week period from the outlet of the
conduit that delivers the dilution/flushing waters to the lake. The samples should be analyzed for
• Total phosphorus
• Dissolved reactive phosphorus
• Total Kjeldahl nitrogen
• Nitrite + nitrate nitrogen and
• Ammonium nitrogen.
IN-LAKE MONITORING
No intake monitoring is necessary during the first two weeks of a dilution flushing project.
The most important parts of the treatment to measure are the inflow volume
and nutrient concentrations within the dilution waters, two factors that will deter-
mine treatment effectiveness. The rationale for these measurements is based
upon the impact of the dilution water on the receiving water. If the algal as-
semblage in the lake is phosphorus or nitrogen limited, the addition of a rate-
limited nutrient may actually stimulate algal growth unless the washout rate is
sufficiently increased.
In addition, the dilution process depends upon a low concentration of incoming
nutrients to effectively dilute the higher concentration in the lake. The flow rate is
needed to calculate the washout process and the mass of incoming nutrients. For
example, using the equation
p _ L I (Vollenweider, 1976)
Zp i + i/Vf
to calculate the steady state phosphorus concentration in a lake, where
P = phosphorus concentration in mg/L,
L=phosphorus load in g/m /yr,
Z = mean depth of lake in meters, and
p = the flushing rate in times per year,
and using a phosphorus loading rate, L, of 0.12 g/m2/yr, a mean depth, Z, of 5 m,
and a flushing rate of 0.25/yr, the calculated phosphorus concentration in the lake
is 0.032 mg/L. If, in this example, the flushing rate is increased 4 times to accom-
modate a dilution/flushing experiment, but the phosphorus load is also increased
4 times, then the expected new concentration in the lake is approximately 0.048
mg/L.
This new concentration is unacceptable. The phosphorus concentrations as-
sociated with the dilution/flushing water in association with the volume of new
12
-------
water must not produce an expected higher phosphorus in-lake concentration.
However, if the phosphorus loading is only doubled and the flushing rate is in-
creased 4 times, the expected in-lake phosphorus concentration would be ap-
proximately 0.024 mg/L. Using this scenario, an improvement in the trophic
condition of the lake is realized.
In the case of Moses Lake, the portion of the lake that was of concern, Parker
Horn, received 10 percent of the lake volume per day during one summer. This
flushing rate produced a lower algal biomass, probably as a result of cell washout.
A simple description of cell washout is given by the following equation:
where
dx/dt = Kx - Dx
K = algal growth rate,
x = biomass, and
D = dilution rate.
K, the algal growth rate, will vary between algal species and within species
depending upon environmental conditions (e.g., herbivore population, sinking
rates).
A reported doubling time for the blue-green alga, Aphanizomenon flos-aqua,
during a growth phase was once every three days (Healey and Hendzel, 1976).
For cell washout to limit biomass, the photic zone of the lake should be flushed at
least once every three days. If consideration is given to a slower growth rate be-
cause of cooler dilution waters, algal sinking velocities, and predation, the 10-day
flushing rate for Parker Horn certainly could include a cell washout process.
The rate of delivery of
the source water will
determine cell washout.
Algal cell washout
needs a rapid exchange
of the lake water, e.g.,
every 3-10 days.
Considerations for Interrupting Treatment
The treatment should not increase the nutrient concentration of the target lake or
reservoir once the project has begun. Other factors such as pH and heavy metal
concentrations of the addition water should not create an unfavorable aquatic en-
vironment.
Dilution/flushing should be discontinued under the following conditions:
1. The phosphorus and/or nitrogen concentration of the addition waters
are higher than the target lake average volumetric growing season con-
centrations in the photic zone.
2, The combination of inflow water volume and phosphorus concentrations
produce an expected higher concentration in the lake based on phos-
phorus loading models.
It is assumed that potential hazardous materials (e.g., pesticides and heavy
metals) in the addition waters were considered in the Phase I portion of the
project.
Monitoring Following the First Two Weeks of Treatment
The monitoring plan described in Table 5.7 is designed to evaluate dilution/flush-
ing water and lake water following the first two weeks of treatment. Documenting
changes to macrophyte communities provides useful secondary information as
they may respond to increased clarity following project implementation.
5-13
-------
Table 5.7.—Monitoring design following the first two weeks of a dilution/flushing
project
MONITORING THE ADDITION WATERS
PHYSICOCHEMICAL
A. Flow Measurement
Continuous flow measurements should be made at the outlet of the conduit that delivers the
dilution/flushing waters to the lake. Flows should be estimated and reported on a daily basis.
See Methods Chapter for appropriate flow measurement techniques.
B. Water Chemistry
A sample should be collected on a weekly basis
dilution/flushing waters to the lake. The samples
• Total phosphorus
• Dissolved reactive phosphorus
• Total Kjeldahl nitrogen
• Nitrite + nitrate nitrogen
• Ammonium nitrogen
• Temperature
• Dissolved Oxygen
from the outlet of the conduit that delivers the
should be analyzed for:
IN-LAKE MONITORING
PHYSICOCHEMICAL
A. Water Chemistry
1. Sampling Location(s):
Samples should be collected at the site(s) selected by the project manager, usually at the
center of the lake.
2. Depth Distribution:
Samples should be collected from just below the surface and at 6 foot intervals to the bottom.
Care should be taken not to include suspended bottom sediments in the water sample.
3. Analytical Determinations and Sampling Procedures:
Total phosphorus, dissolved reactive phosphorus, total Kjeldahl nitrogen, ammonium nitrogen,
and nitrite + nitrate nitrogen. See Methods Chapter for appropriate analytical and sampling
techniques.
4. Frequency and Duration:
Samples should be collected at monthly intervals for the duration of the dilution/flushing
project. Data should be obtained for at least a one year period.
B. Dissolved Oxygen (DO) and Temperature
1. Sampling Location(s):
Same as water chemistry.
2. Depth Distribution:
Measurements should be made at 3 foot intervals from the surface to the bottom.
3. Analytical Determinations and Sampling Procedures:
See Methods Chapter for appropriate analytical and sampling techniques.
4. Frequency and Duration:
Same as water chemistry.
C. Secchi Disk Transparency
1. Sampling Location(s):
Same as for water chemistry.
2. Frequency and Duration:
Secchi disk measurements should be made at monthly intervals for the duration of the
dilution/flushing project during the growing season (May through October). Data should be
obtained for at least a one-year period.
5-14
-------
Table 5.7.—Monitoring design following the first two weeks of a dilution/flushing
project (continued)
BIOLOGICAL
A. Chlorophyll a (Corrected for Pheophytin)
1. Sampling Location:
Same as Secchi disk.
2. Depth Distribution:
A sub-sample should be obtained from an integrated sample representing a water column
equal to 0-6 feet from the surface.
3. Analytical Determination and Sampling Procedure:
See Methods Chapter for appropriate analytical and sampling techniques.
4. Frequency and Duration:
Same as Secchi disk.
CONTROL TECHNIQUE #3
Artificial Circulation
Technical Considerations
Artificial aeration/circulation is a common method of alleviating the problem of dis-
solved oxygen depletion and has a long history of use in lakes and reservoirs.
This method has been successfully used as a lake restoration technique to
prevent fish kills, improve domestic water supplies, and reduce algal biomass, or
cause a major shift from nuisance algae (blue-greens) to other algal types
(greens). Artificial circulation, however, has had mixed reviews as a method to
control algal problems in lakes (Pastorok et al. 1980).
The success of this treatment for algal control is based upon changes in the
lake's physicochemical and biological elements. In most cases, artificially circulat-
ing a lake increases the concentration of dissolved oxygen in the bottom waters,
which influences the redox reactions involving iron and manganese. These ele-
ments complex with phosphorus (Mortimer, 1942) and, in part, determine phos-
phorus inputs from the sediments, thereby reducing nutrient availability to
promote algal growth.
Increases in the temperature of the bottom waters as a result of total lake
mixing may counteract the redox reactions by stimulating decomposition rates
and phosphorus release. Furthermore, an increased zone of oxic water over the
sediments in the deep part of the lake may increase the area of habitat for burrow-
ing macroinvertebrates and rough fish (carp) that contribute to phosphorus cycling
within the lake system. The use of artificial circulation to control algal biomass by
reducing phosphorus cycling from the sediments must consider the impacts of in-
creased habitat for organisms that contribute to increased phosphorus cycling. As
with most in-lake treatments, reduction of external phosphorus loads is also im-
portant.
The physical mixing of deep lakes may dilute the algal biomass throughout a
greater volume of water, thereby increasing water transparency. This occurred
during the first year of aeration at Kezar Lake, New Hampshire, where a com-
pressed air system was installed to artificially mix the lake to alleviate a blue-
green algal problem (N.H. Water Supply Pollut. Control Comm. 1971). In the first
Artificial circulation of a
lake has many uses.
Physical mixing can
distribute algal cells
throughout the entire
lake.
5-
15
-------
Other factors that control
algal biomass during
artificial circulation are
changes in pH...
increases in herbivorous
zooplankton ... and
increases in algal virus
activity.
year of operation, the compressors were started in July. The concentration of
algae at the surface of the lake went from 1 x 106 cells/mL before the compressors
were started to 1 x 104 cells/mL within a month of the start of artificial mixing.
This apparent reduction of algal density was misleading, however. If the cell
densities were represented as a sum of all cells distributed with depth under the
surface of the lake, e.g., cm2 of lake surface area, then total algal biomass ap-
parently changed very little. Before the start of the compressors, the cell count
under a cm2 of lake surface was 1.5 x 108. After three weeks of circulation, the cell
count was 1.0 x 108/cm2.
In the second year of operation, the compressors were started in the spring,
and the lake was mixed throughout the spring and summer. Under these condi-
tions, nuisance algal problems never developed, and the areal cell densities were
much less than the previous year.
Other factors that control algal biomass during artificial circulation are changes
in pH (shift in algal species favoring greens), increases in herbivorous
zooplankton (increased grazing on algae), and increases in algal virus activity that
may reduce blue-greens.
Monitoring During the First Two Weeks of Treatment
The construction phase of an artificial aeration/circulation project begins with
equipment installation in the lake and continues through the first two weeks of
operation. During this phase, it is important to evaluate the effectiveness of the
equipment as well as the lake's responses by carefully monitoring temperature
and dissolved oxygen conditions.
Table 5.8 describes an in-lake monitoring design for the first two weeks after
the aeration/circulation device is operational.
Table 5.8.—In-lake monitoring design for the first two weeks of an aeration/circulation
project
PHYSICOCHEMICAL
A. Dissolved Oxygen (DO) and Temperature
1. Sampling Location(s)
Measurements should be made at a site near the aeration/circulation device and at sites 200
feet and at least 1,000 feet away from the air release point.
2. Depth Distribution
Measurements should be made at 3-foot intervals from the surface to the bottom.
3. Analytical Determinations and Sampling Procedures
See Chapter 3 for appropriate analytical and sampling techniques.
4. Frequency and Duration
Measurements should be made at daily intervals for the first two weeks of operation.
The project must account
for initial oxygen demand
of bottom waters.
Consideration for Interrupting Treatment
The immediate effects of mixing on dissolved oxygen throughout the lake deter-
mine whether aeration/circulation should be interrupted. When a lake is artificially
mixed, bottom waters high in oxygen demand can be distributed throughout the
entire water column, which can result in a lowering of the dissolved oxygen in the
entire lake. If the artificial circulation results in an initial dissolved oxygen depletion
to a level that threatens the support of game fish (<5 mg/L), the project will have
failed to meet its objective.
Figure 5.3 represents an example of an aeration/circulation project that did not
take into account the high oxygen demand of the bottom waters. A compressed air
5-
16
-------
TOTAL
AERATION
AERATION
91 12
OCT
NOV
DEC
JAN
FEB
MAR
APR
MAY
1972
Figure 5-3.—Dissolved oxygen (volume weighted mean) for Mirror Lake, 1971-72. (Note the
decrease to 0.9 mg/L when compressed air unit was first started.)
unit was operated from October 19 through November 21,1972, in Mirror Lake,
Wisconsin. The immediate result of destratification was a marked decline in dis-
solved oxygen to a minimum of 0.9 mg/L. The compressor was operated con-
tinuously for nearly two weeks before the dissolved oxygen concentration
recovered to 5 mg/L {Smith et al. 1975). The very low dissolved oxygen con-
centrations had negative biological consequences for the lake.
Another potential problem associated with complete lake mixing is the possible
increase in algal biomass. This possibility exists if the aeration/circulation device
is underdesigned and so allows a slow intrusion of bottom waters high in nutrients
into the epilimnion, thereby encouraging algal growth.
Consideration should be given to interrupting a whole lake aeration/circulation
project if, within the first two weeks of operation, the dissolved oxygen in the top 6
feet is 5 mg/L or less.
Beware of undersized
lake mixing systems.
Monitoring Following the First Two Weeks of Treatment
An aeration/circulation project is successful if the system design can maintain an
acceptable oxic environment in the lake and either reduce algal densities or shift
species composition to more desirable algal types, e.g., to greens. The task
should be accomplished without adding to the overall problem by increasing algal
biomass.
Both temperature and dissolved oxygen are measurable parameters that can
be used to evaluate project success. Chlorophyll a can be used to determine algal
biomass and Secchi disk measurements to determine water clarity. Algae must be
identified and counted to determine if algal species change. Measurements of pH
are used in combination with algal identification to assist in explaining any chan-
ges. Other useful measurements are those designed to evaluate specific project
objectives such as nutrient control and creation of a zone of refuge for larger-
bodied zooplankton.
A monitoring plan following start-up of the aeration/circulation system,
described in Table 5.9, is designed to measure the success of the aeration/circula-
tion project for one growing season during operation of the device. A growing
season is defined as the period from May through October but may vary depend-
ing upon location within the United States.
5-
17
-------
Table 5.9.—In-lake monitoring following the first two weeks of an aeration/circulation
project
PHYSICOCHEMICAL
A. Dissolved Oxygen (DO) and Temperature
1. Sampling Location(s)
Measurements should be made at a site near the air release point and at sites 200 feet and
at least 1,000 feet away from the air release point.
2. Depth Distribution
Measurements should be made at 3-foot intervals from the surface to the bottom.
3. Analytical Determinations and Sampling Procedures
See Chapter 3 for appropriate analytical and sampling techniques.
4. Frequency and Duration
Measurements should be made at weekly intervals during the first month of operation.
Thereafter, measurements should be made at two-week intervals until DO reaches 80
percent of saturation. After DO has reached 80 percent of saturation, measurements should
be made at monthly intervals for six additional months while the aerator is being operated.
Data should be obtained for at least a one-year period.
B. pH Measurements
1. Sampling Location(s)
Samples should be collected at a site 200 feet from the air release points and at a site at
least 1,000 feet from the air release points.
2. Depth Distribution
Samples should be collected from the surface at the three foot depth and at the six foot
depth".
3. Analytical Determinations and Sampling Procedures
See Chapter 3 for appropriate analytical and sampling techniques.
4. Frequency and Duration
Samples should be collected at weekly intervals for the first month of operation. Thereafter,
they should be collected every two weeks for the duration of the growing season (May
through October). Data should be obtained for at least a one-year period.
C. Secchi Disk Transparency
1. Sampling Location
Secchi disk measurements should be made at the same sites selected for dissolved oxygen
measurements and also near the center of the lake.
2. Frequency and Duration
Same as for dissolved oxygen and temperature.
BIOLOGICAL
A. Chlorophyll a (corrected for pheophytin)
1. Sampling Location
Same as for Secchi disk.
2. Depth Distribution
A subsample should be obtained from an integrated sample representing a water column
equal to 0-6 feet from the surface.
3. Analytical Determination and Sampling Procedure
See Chapter 3 for appropriate analytical and sampling techniques.
4. Frequency and Duration
Samples should be collected every two weeks for the duration of the growing season (May
through October). Data should be obtained for at least a one-year period.
18
-------
Table 5.9,—In-lake monitoring following the first two weeks of an aeration/circulation
project (continued)
B. Algae
1. Sampling Location
Samples should be collected at a site 200 feet away from the air release point,
2. Depth Distribution
A subsample should be obtained from an integrated sample representing a water column
equal to 0-6 feet from the surface.
3. Analytical Determinations and Sampling Procedures
Algal samples should be obtained directly from a water sample; they should never be
collected with a net. Samples should be preserved with Lugol's solution immediately following
collection. Algae should be identified to species by using oil emersion and magnification of
900X or greater. Any algal species that comprise greater than 5 percent of the total should be
enumerated at a magnification of 400X or greater. Use of an inverted microscope is
recommended.
4. Frequency and Duration
Same as for chlorophyll a.
CONTROL TECHNIQUE #4:
Hypolimnetic Aeration
Technical Considerations
Hypolimnetic aeration objectives are usually similar to those of total circulation
with the added objective of providing oxygen in the lake hypolimnion without
destroying the thermal barriers associated with summer stratification. Hypolim-
netic aeration has been employed in European lakes since 1948. The original use
of the technique was to remove dissolved metals from cold hypolimnetic water
before it was used for industrial purposes (Mericer and Perrett, 1949).
In more recent times, hypolimnetic aeration has been used in drinking water
reservoirs to prevent the dissolution of iron and manganese compounds from bot-
tom sediments, thereby averting the need for expensive water treatment facilities
(Ripl, 1980). Another demonstrated use of hypolimnetic aeration has been to con-
trol sediment phosphorus release. The maintenance of an oxic environment over-
lying lake sediments with sufficient available iron can reduce phosphorus cycling,
which translates to a reduction in algal biomass.
In Vadnais Lake, Minnesota, two hypolimnetic aerators were used in conjunc-
tion with addition of iron to reduce the total phosphorus concentrations in the
hypolimnion (Walker et al. 1989). The aerators maintained an average hypolim-
netic oxygen concentration above 0.8 mg/L during the summers of 1987 and
1988. In previous summers, prior to hypolimnetic aeration and iron additions, the
total phosphorus concentrations at fall overturn were 100 to 200 ng/L. In 1988,
after hypolimnetic aeration and iron additions, the total phosphorus concentration
at fall overturn was 35 ng/L.
In another example, liquid oxygen was injected into the hypolimnion of Amisk
Lake, Alberta, to control internal phosphorus release and reduce algal biomass
(Prepas et al. 1989). The total phosphorus rate of accumulation in the hypolimnion
during the summer of oxygen injection was about 40 percent less than previous
years. The chlorophyll a concentrations during the summer of hypolimnetic injec-
The maintenance of an
oxic environment
overlying lake sediments
with sufficient available
iron can reduce
phosphorus cycling.
Vadnais Lake, an
example of a successful
hypolimnetic aeration
project.
5-19
-------
tion ranged from 4 to 16 ng/L. In previous years, the chlorophyll a concentrations
ranged from 10 to 27 ng/L.
An additional consideration is the creation of a zone of refuge for large-bodied
; zoopiankton that allows them to escape predation during the day by remaining in
Creating a zone of refuge poorly lit bottom waters and migrating to surface waters to feed on algae during
for large-bodied the night. Increased grazing should further reduce algal biomass.
zoopiankton....
Monitoring During the First Two Weeks of Treatment
There have been a variety of hypolimnetic aeration designs, both workable and
unworkable. The short-term test of a workable system is its ability to maintain the
thermal layers that divide the lake into the epilimnion, metalimnion, and hypolim-
nion while increasing the dissolved oxygen concentration in the hypolimnion.
Table 5.10 describes a monitoring design for the first two weeks of operation.
Table 5.10.—In-lake monitoring design for the first two weeks of hypolimnetic
aeration
PHYSICOCHEMICAL
A. Dissolved Oxygen (DO) and Temperature
1. Sampling Location
Measurements should be made at a site 50 feet from the discharge of the hypolimnetic
aerator and at a site at least 1,000 feet away from the discharge.
2. Depth Distribution
Measurements should be made at 3-foot intervals from the surface to the bottom.
3. Analytical Determinations and Sampling Procedures
See Chapter 3 for appropriate analytical and sampling procedures.
4. Frequency and Duration
Measurements should be made at daily intervals for the first two weeks of operation.
Stop the project if the
hypolimnetic unit causes
a breakdown of the
lake's natural thermal
barrier.
Considerations for Interrupting Treatment
The rationale for stopping a hypolimnetic aeration project is based upon disruption
of the thermal barriers within the lake. Once the thermal integrity between the
hypolimnion and epilimnion has been eliminated, the original objective of main-
taining oxic waters in the hypolimnion without increasing the water temperature
cannot be achieved.
Hypolimnetic aeration should be stopped under the following conditions:
1. If, during the first two weeks, the operation of the hypolimnetic unit
causes substantial erosion of the thermocline.
2. If, during the first two weeks, the operation of the hypolimnetic unit
causes an increase in temperature of the hypolimnion of 0.5°G per day
or greater.
Monitoring After the First Two Weeks of Treatment
A hypolimnetic aeration project's success is related to the system's ability to in-
crease dissolved oxygen in the hypolimnion while maintaining the thermal integrity
of lake stratification. The primary goals of an oxic hypolimnion are reduced phos-
phorus concentrations in the bottom waters and a reduced algal biomass in the
5-20
-------
epilimnion. To evaluate the success of the treatment, therefore, dissolved oxygen,
temperature, algal biomass, phosphorus, and ammonium concentrations should
be monitored. The duration of the monitoring program should be from start-up until
fall overturn. Table 5.11 describes a recommended monitoring design for hypolim-
netic aeration projects.
Table 5.11.—In-Iake monitoring design for hypolimnetic aeration projects following
the first two weeks of operation
PHYSICOCHEMICAL
A. Water Chemistry
1. Sampling Location
Samples should be collected at a site 50 feet from the discharge of the hypolimnetic aerator
and at a site at least 1,000 feet from the discharge.
2. Depth Distribution
Samples should be collected at 6-foot intervals from just below the surface to the bottom.
Care should be taken not to include suspended bottom sediments in the water sample.
3. Analytical Determinations and Sampling Procedures
Water samples should be analyzed for total phosphorus, dissolved reactive phosphorus, total
dissolved iron, and ammonium nitrogen. See Chapter 3 for appropriate analytical and
sampling techniques.
4. Frequency and Duration
Samples should be collected every two weeks until the dissolved oxygen concentration
reaches 1 mg/L. After it has reached 1 mg/L, samples should be collected at monthly
intervals while the aerator is being operated. Data should be obtained for at least a one-year
period.
B. Dissolved Oxygen (DO) and Temperature
1. Sampling Location
Same as for water chemistry.
2. Depth Distribution
Measurements should be made at 3-foot intervals from the surface to the bottom.
3. Analytical Determinations and Sampling Procedures
See Chapter 3 for appropriate analytical and sampling techniques.
4. Frequency and Duration
Measurements should be made at weekly intervals until the end of the first month of
operation. Thereafter, measurements should be made every two weeks until the dissolved
oxygen concentration reaches 1 mg/L. After it has reached 1 mg/L, measurements should be
made at monthly intervals while the aerator is being operated.
C. Secchi Disk Transparency
1. Sampling Location
Measurements should be made at the sites selected for water chemistry sample collection
and at a site near the center of the lake.
2. Frequency and Duration
Measurements should be made at two-week intervals for the duration of the growing season
(May through October). Data should be obtained for at least a one-year period.
BIOLOGICAL
A. Chlorophyll a (corrected for pheophytin)
1. Sampling Location
Same as for Secchi disk.
2. Depth Distribution
A subsample should be obtained from an integrated sample representing a water column
equal to 0-6 feet from the surface.
3. Analytical Determinations and Sampling Procedures
See Chapter 3 for appropriate analytical and sampling techniques.
4. Frequency and Duration
Measurements should be made at two-week intervals for the duration of the growing season
(May through October). Data should be obtained for at least a one-year period.
5-21
-------
To be effective,
hypoUmnetic withdrawal
must not cause the lake
to destratify early.
CONTROL TECHNIQUE #5:
Hypolimnetic Withdrawal
Technical Considerations
Hypolimnetic withdrawal is a lake and reservoir management technique that has
been used successfully in both Europe and the United States. Employed in lakes
that maintain thermal stratification and develop anoxic hypolimnia with a sig-
nificant hypolimnetic phosphorus mass, this technique maintains a bottom water
withdrawal of high phosphorus waters, thereby allowing a consistent depletion of
the sediment phosphorus pool. However, hypolimnetic withdrawal should not des-
tabilize the thermal structure of stratification to the point of inducing total lake
mixing. An early destratification may result in increased algal biomass in the upper
waters as the result of mixing bottom waters high in phosphorus with the
waterbody's surface layers.
In reviewing the results of 17 lakes where hypolimnetic withdrawal has been
employed, Numberg (1987) reports that, in general, the epilimnetic and hypolim-
netic phosphorus concentrations have decreased. The decline in epilimnetic con-
centrations of phosphorus correlated with the phosphorus exported via
hypolimnetic withdrawal and the relationship improved as a function of years of
operation.
Monitoring During the First Two Weeks of Treatment
The objective of hypolimnetic withdrawal is to reduce internal phosphorus cycling
within a lake or reservoir. If thermal stratification is prematurely destroyed as a
result of a downward displacement of the thermocline, serious algal problems may
soon develop in the surface waters. It is important, therefore, to monitor both the
thermal stability of the water column during the system's initial operation as well
as the outflow from the hypolimnion, as most lakes where this technique will be
used have anoxic conditions with elevated hydrogen sulfide and ammonium con-
centrations. Some States may also require that a permit with specific monitoring
requirements be obtained for this discharge because of potential negative impacts
on downstream water quality. A recommended monitoring design for the first two
weeks of operation is shown in Table 5.12.
Consideration must be
given to adverse impacts
on the downstream
environment
Considerations for Interrupting Treatment
The following possibilities are two major concerns related to hypolimnetic
withdrawal: the destratification of the lake during operation and a negative impact
on the downstream environment from the discharge water. Both are considera-
tions for interrupting a project.
In the absence of specific permit requirements, a hypolimnetic withdrawal
project should be interrupted under the following conditions:
1. If the hypolimnetic discharge flows to a receiving stream, and the final
amount of dissolved oxygen in the discharge water is less than 1 mg/L
2. If the hypolimnetic discharge flows to a receiving stream, and the com-
bination of temperature, pH, and NH4-N produce unionized ammonium
concentrations at levels considered toxic to biota.
3. If the temperature in the hypolimnion increases by more than 0.5°C/day.
5-22
-------
Table 5.12.—Monitoring design for the first two weeks of a hypolimnetic withdrawal
project
QUALITY OF HYPOLIMNETIC DISCHARGE WATER
PHYSICOCHEMICAL
A. Water Chemistry
1. Sampling Location
Samples should be collected from the hypolimnetic discharge.
2. Frequency and Duration
Samples should be collected twice weekly during the first two weeks of operation.
3. Analytical Determinations and Sampling Procedures
Water samples should be analyzed for total phosphorus, dissolved reactive phosphorus,
ammonium nitrogen, and pH. See Chapter 3 for appropriate analytical and sampling
techniques.
B. Dissolved Oxygen (DO) and Temperature
1. Sampling Location
Same as for water chemistry.
2. Analytical Determinations and Sampling Procedures
See Chapter 3 for appropriate analytical and sampling techniques.
3. Frequency and Duration
Measurements should be made daily for the first two weeks of operation.
IN-LAKE MONITORING
PHYSICOCHEMICAL
A. Dissolved Oxygen (DO) and Temperature
1. Sample Location
Measurements should be made at site(s) selected by the project manager, usually at the
deepest part of the lake.
2. Depth Distribution
Measurements should be made at 3-foot intervals from the surface to the bottom.
3. Analytical Determinations and Sampling Procedures
See Chapter 3 for appropriate analytical and sampling procedures.
4. Frequency and Duration
Measurements should be made twice weekly.
Monitoring Following Treatment
To be effective, hypolimnetic withdrawal must be used continuously during
periods of stratification for a number of years. The monitoring plan described in
Table 5.13 is designed to measure the success of the project based upon in-Iake
algal and nutrient responses and to monitor the water quality of the hypolimnetic
discharge to the receiving stream.
To be successful, this
treatment may require
several years of
operation.
5-23
-------
Table 5.13.—Monitoring design for a hypolimnetic withdrawal project following the
first two weeks of operation
QUALITY OF HYPOLIMNETIC DISCHARGE WATER
PHYSICOCHEMICAL
A. Water Chemistry
1. Sampling Location
Water samples should be collected from the hypolimnetic discharge.
2. Analytical Determinations and Sampling Procedures
Water samples should be analyzed for total phosphorus, dissolved reactive phosphorus,
ammonium nitrogen, and pH. See Chapter 3 for appropriate analytical and sampling
techniques.
3. Frequency and Duration
Samples should be collected at monthly intervals while hypolimnetic waters are being
discharged. Data should be collected for at least a one-year period.
B, Dissolved Oxygen (DO) and Temperature
1. Sampling Location
Same as for water chemistry.
2. Analytical Determinations and Sampling Procedures
See Chapter 3 for appropriate analytical and sampling techniques.
3. Frequency and Duration
Same as for water chemistry.
IN-LAKE MONITORING
PHYSICOCHEMICAL
A. Water Chemistry
1. Sampling Location
Samples should be collected at the site(s) selected by the project manager, usually at the
deepest part of the lake.
2. Depth Distribution
Samples should be collected at 6-foot intervals from just below the surface to the bottom.
Care should be taken not to include suspended bottom sediments in the water samples.
3. Analytical Determinations and Sampling Procedures
Water samples should be analyzed for total phosphorus, dissolved reactive phosphorus,
ammonium nitrogen, and pH. See Chapter 3 for appropriate analytical and sampling
techniques.
4. Frequency and Duration
Samples should be collected at monthly intervals during the growing season (May through
October). Data should be obtained for at least a one-year period.
B. Dissolved Oxygen (DO) and Temperature
1. Sampling Location
Same as for water chemistry.
2. Depth Distribution
Measurements should be made at 3-foot intervals from the surface to the bottom.
3. Anaytical Determinations and Sampling Procedures
See Chapter 3 for appropriate analytical and sampling techniques.
4. Frequency and Duration
Same as for water chemistry.
C. Secchi Disk Transparency
1. Sampling Location
Same as for water chemistry.
2. Frequency and Duration
Same as for water chemistry.
5-24
-------
Table 5.13.—Monitoring design for a hypolimnetic withdrawal project following the
first two weeks of operation (continued)
BIOLOGICAL
A. Chlorophyll a (corrected for pheophytin)
1. Sampling Location
Same as for Secchi disk.
2. Depth Distribution
A subsample should be obtained from an integrated sample representing a water column
equal to 0-6 feet from the surface.
3. Analytical Determination and Sampling Procedure
See Chapter 3 for appropriate analytical and sampling procedures.
4. Frequency and Duration
Same as for Secchi disk.
CONTROL TECHNIQUE #6:
Sediment Oxidation
Technical Considerations
Ripl (1976) developed a lake restoration method to oxidize the anaerobic surface
sediments of lakes. The method is dependent upon the ability of iron in the sedi-
ments (either natural amounts or iron added as part of the treatment) to control
phosphorus release. The method involves oxidizing the organic matter in the surfi-
cial sediments through increased denitrification, thereby increasing the binding
capacity of ferric hydroxide complexes with sediment interstitial phosphorus. A
solution of Ga(N03)2 and, in some cases, FeCb and Ca(OH)2 is injected into the
sediments. The technique has been demonstrated in Long Lake in Minnesota and
in several European lakes.
Monitoring During Treatment
The objective of sediment oxidation is the same as for alum, artificial circulation,
and hypolimnetic aeration: the reduction of phosphorus release from lake sedi-
ments. The anticipated reduction in sediment phosphorus release should lower
the available phosphorus for algal growth in the photic zone. Because the treat-
ment involves a one-time injection into the sediments, little monitoring can be
done during the process.
Considerations for Interrupting Treatment
Once the injection of Ca(NC>3)2 begins, there are no easily monitored lake
parameters that will indicate that the process should be stopped. The addition of
Ca(N03)2 is not readily toxic; therefore, the risk of environmental problems is low.
Another method to
immobilize phosphorus
release from sediments.
5-
¦25
-------
Monitoring Following Treatment
The monitoring plan is designed to evaluate the effectiveness of the treatment.
The important parameters to monitor are related to the chemical additions of the
process: Ca, NO3, possible changes in pH, and the target parameters of phos-
phorus and algal biomass. The recommended monitoring pian is shown in Table
5.14.
Table 5.14.—In-lake monitoring after sediment oxidation treatment
PHYSICOCHEMICAL
A. Water Chemistry
1. Sampling Location
Water samples should be collected at the siie(s) selected by the project manager, usually at
the deepest part of the lake.
2. Depth Distribution
Samples should be collected at 6-foot intervals from just below the surface to the bottom.
Care should be taken not to include suspended bottom sediments in the water samples.
3. Analytical Determinations and Sampling Procedures
Water samples should be analyzed for total phosphorus, dissolved reactive phosphorus,
ammonium nitrogen, nitrite + nitrate nitrogen, calcium, alkalinity, and pH. See Chapter 3 for
appropriate analytical and sampling techniques.
4. Frequency and Duration
Measurements should be made at two-week intervals from May through October and monthly
thereafter for a period of one year following treatment.
B. Dissolved Oxygen (DO) and Temperature
1. Sampling Location
Same as for water chemistry.
2. Depth Distribution
Measurements should be made at 3-foot intervals from the surface to the bottom.
3. Analytical Determinations and Sampling Procedures
See Chapter 3 for appropriate analytical and sampling techniques.
4. Frequency and Duration
Same as for water chemistry.
C. Secchi Disk Transparency
1. Sampling Location
Same as for water chemistry.
2. Frequency and Duration
Same as for water chemistry.
BIOLOGICAL
A. Chlorophyll a (corrected for pheophytin)
1. Sampling Location
Same as for Secchi disk.
2. Depth Distribution
A subsample should be obtained from an integrated sample representing a water column
equal to 0-6 feet from the surface.
3. Analytical Determination and Sampling Procedure
See Chapter 3 for appropriate analytical and sampling techniques.
4. Frequency and Duration
Same as for Secchi disk.
26
-------
CONTROL TECHNIQUE #7:
Food Web Manipulation
Technical Considerations
The purpose of food web manipulation is to reduce nuisance algal biomass. Over-
ail, nutrient inputs and dynamics of a lake or reservoir comprise a key control on
the level of production. Benndorf and Miersch (in press) suggest that food web
manipulation in lakes having a phosphorus loading rate less than 0,6 gm P/m2 • yr
have a greater chance for success in reducing algal biomass than lakes with
greater phosphorus loading rates.
The concept of food web manipulation is not new. The early work of Hrbacek et
al. (1961) set the stage in Europe for use of this technique as a lake restoration
tool. Shapiro (1978), Porter (1977), and Carpenter et al. (1985) have extended the
biomanipulation philosophy to North American lakes. In principle, an increase in
the piscivore biomass should bring about a decrease in the planktivore biomass
(the larger predator fish prey upon the smaller fish that consume zooplankton).
Decreases in the planktivore fisheries should increase the biomass of the large-
bodied zooplankton that feed on algae. Because grazing rate increases geometri-
cally with body length, the large-bodied zooplankters graze algae more efficiently.
The ultimate goal of food web manipulation is to maintain sufficient populations of
the large-bodied zooplankton over the summer season to consistently graze down
the excessive amounts of algae.
The actual restoration technique can be applied in a number of different ways:
Manipulating the food
web requires time before
a response may be
noticed in algal biomass.
1. A complete fish kill accomplished with a fish toxicant (such as rotenone)
will eliminate the fisheries and, therefore, predation pressure on the
zooplankton. A dramatic increase in large-bodied zooplankton is fre-
quently observed in these situations, along with a corresponding in-
crease in water clarity.
2. A large increase in the stocking of piscivores could have the desired im-
pact on the planktivores.
3. A zone of refuge that precludes predation can be created for herbivores
by hypolimnetic aeration. The large-bodied zooplankton can avoid
predation during the day by moving into the deep waters where light
limits the efficiency of sight-feeding fish. During the night, zooplankton
can migrate to the surface of the lake to feed upon the algae.
#
Monitoring During Treatment
The in-lake monitoring for a food web manipulation project will depend to some
degree on which method is employed. The results from any of the biomanipulation
methods will take some time, however, to be realized. Once the predation pres-
sure is reduced on the zooplankton, they cannot respond in one day's time. The
in-lake monitoring during treatment will be no different than the monitoring design
following treatment.
5-27
-------
Considerations for Interrupting Treatment
The basis for evaluating a food web manipulation project is its success in increas-
ing the herbivore population and decreasing the algal densities or changing the
species composition. Because evaluation of these projects is possible only after
life cycles are completed, criteria for interrupting a project in the short term do not
exist. It is worth restating, therefore, that serious consideration should be given to
the probability of success before starting a food web manipulation project if the
phosphorus loading rate to the lake exceeds 0.6 gm P/m2 • yr.
Monitoring Following Treatment
The basic goal of a food web manipulation project is to maintain a sufficient
population of the large-bodied zooplankters and thereby decrease the algal
biomass by grazing pressure. The length of the monitoring program depends upon
the technique employed. If fish are stocked in large numbers over several years,
monitoring should be delayed until they reach an effective size.
If a fish toxicant is used, the impact can be noticed within a month, but success
depends upon the fish species selected for the stocking program. For example, it
would be unwise to restock with rainbow trout as they would quickly consume the
large-bodied zooplankters. Once biomanipulation has been implemented,
monitoring should be conducted as outlined in Table 5.15.
The aquatic food chain (notto scale)
PISCIVOROUS
FISH
eat
\
PLANKTIVO ROUS
FISH
eat
HERBIVORES
eat
~
ALGAE
use
t
NUTRIENTS
NUTRIENTS
recycle
BENTHIVOROUS
FISH
(source: Shapiro et al. 1982)
28
-------
Table 5.15.—In-lake monitoring for a food web manipulation project
PHYSICOCHEMICAL
A. Dissolved Oxygen (DO) and Temperature
1. Sample Location
Measurements should be taken at the site(s) selected by the project manager, usually at the
center of the lake.
2. Depth Distribution
Measurements should be made at 3-foot intervals from the surface to the bottom.
3. Analytical Determinations and Sampling Procedures
See Chapter 3 for appropriate analytical and sampling techniques.
4. Frequency and Duration
Measurements should be made at two-week intervals during the growing season (May
through October), and monthly thereafter for at least a one-year period following completion
of the project.
B, Secchi Disk Transparency
1. Sampling Location
Same as for dissolved oxygen.
2. Frequency and Duration
Same as for dissolved oxygen.
BIOLOGICAL
A. Chlorophyll a (corrected for pheophytin)
1. Sampling Location
Same as for Secchi disk.
2, Depth Distribution
A subsample should be obtained from an integrated sample representing a water column
equal to 0-6 feet from the surface.
3. Analytical Determinations and Sampling Procedures
See Chapter 3 for appropriate analytical and sampling techniques.
4, Frequency and Duration
Same as for Secchi disk.
B. Zooplankton
1. Sampling Location .
Same as for chlorophyll a.
2. Depth Distribution
With the exception of a zone of refuge treatment, a Number 10 (156 u mesh) plankton net
should be pulled through a water column equal to the depth of oxygenated waters. The exact
length of the plankton tow must be recorded to calculate the volume of water filtered.
For a zone of refuge treatment, samples should be collected at 6-foot intervals from the
surface to the bottom of the oxygenated water column. Samples should be obtained by use
of a Schindler-Patalas trap, Clarke:Bumpus sampler, or similar apparatus,
3. Analytical Determination
Species identification, representative body length, enumeration, and ratio of eggs to adult
females should be made for the zooplankton in each sample. The individual species density
should be reported as numbers of individuals per liter of lake water at each sample depth.
4. Frequency and Duration
Same as for chlorophyll a.
C. Fish
Fish population should be surveyed one year following completion of the project using methods
appropriate for the species present. To completely characterize both game and nongame
species, most surveys will use gill netting for pelagic species and fyke nets/boom shocking for
others.
5-29
-------
OBJECTIVE:
Increase Depth
Natural lakes may fill
with sediment at a rate
of 0.1 inch/year;
reservoirs fill with
sediment at a faster rate,
e.g., 1.5 inches/year.
Summary
1. The rationale for increasing the depth of a lake or reservoir
is to increase the storage capacity of the reservoir, to in-
crease recreational potential for the waterbody, or to reduce
macrophyte growth.
2. The essential monitoring requirements are to determine the
macrophyte distribution and depth of growth and to re-map
the deepened area of the lake.
Flood control reservoirs built in regions of the country where the uplands
have a high erosion rate are usually designed to have a specific life span—
a period before they are expected to fill in with sediments. Deepening can
prolong the usefulness of these kinds of reservoirs by renewing their
water-holding capacity.
Natural lakes as well as recreational reservoirs that have been built
throughout the United States offer lakeshore property for home sites and
opportunities for public recreation. It is not unusual for the inlet areas of
such lakes and reservoirs to noticeably fill in within a decade after they
are developed or reach designed pool capacity. As the managers of these
recreational waterbodies will often attest, this becomes an unacceptable
environmental development that adversely affects recreational oppor-
tunities.
There are many causes for the rapid sedimentation of certain areas in
lakes' and reservoirs' littoral zone. Those natural lakes that are within the
glaciated part of the United States are over 10,000 years old, and many
have as much as 35 feet of sediment within the original lake basin. These
sediments normally are very organic; their origin is plant production and
decay. The difference in sedimentation rates between manmade reservoirs
and natural lakes reflects the origin of the sediments. Sedimentation rates
measured by using either cesium-137 or lead-210 radioisotope methods
for a series of lakes and reservoirs in Wisconsin ranged from less than 0.1
inch per year in the more remote natural lakes to greater than 1.5 inches
per year in reservoirs (Wedepohl et al. 1983).
Agricultural activities upstream of lakes and reservoirs are a common
cause of accelerated sedimentation. For limited periods during the year,
agriculture disturbs the soil, thereby increasing its erodibility. Other
causes for rapid infilling in the lakes' littoral zones include construction
site erosion from houses and roads built near the lakeshore. Whatever the
cause of rapid infilling to the lake or reservoir, reduction of watershed-
derived sediment is often cost-effective.
5-30
-------
CONTROL TECHNIQUE;
Dredging
Technical Considerations
Dredging is often done as a restoration practice to increase water depth and
thereby reduce nuisance levels of rooted aquatic plants. In a few cases, dredging
has been employed to remove a specific layer of sediments containing a high con-
centration of nutrients. In the majority of cases where dredging is the lake restora-
tion technique of choice, however, it is done to restore water depth lost to
sedimentation.
There are several hydraulic or mechanical techniques that can be used to
dredge sediments from a lake or reservoir (Cooke et al. 1986). Hydraulic dredging
with the use of a cutterhead is probably most often employed; however, small
dredging operations have used front-end loaders, draglines, and backhoes to
remove sediments from reservoirs where the water level was drawn down.
Typically, there is concern about contaminants that may be present in the sedi-
ments to be removed from reservoirs and lakes in agricultural or urban water-
sheds. These chemicals ultimately are transported, usually on the fine sediment
particles, to the receiving waterbody. The chemical composition of the sediments
to be removed dictates the necessary precautions that must be considered for
land disposal. In this manual, it is assumed that the sediment characteristics were
quantified during the Phase I study of the lake.
One symptom of rapid infilling of lakes or reservoirs—overabundant growth of
attached aquatic plants—frequently causes the most use problems. If plant con-
trol is one of the objectives of a dredging project, the depth to be dredged must
reflect the depth of plant colonization. Several equations for estimating the maxi-
mum depth of colonization (MDC) have been suggested by scientists (Canfield et
al. 1985). These relationships between depth of plant growth and depth of light
penetration (as measured by a Secchi disk) will vary for different areas of the
United States. Equations developed from Secchi disk (SD) measurements in
Florida and Wisconsin are as follows:
Florida log MDC = 0.42logSD + 0.41
Wisconsin log MDC = 0.79logSD + 0.25
The reader is referred to Chapter 6 of The Lake and Reservoir Restoration
Guidance Manual (U.S. Environ, Pr6t. Agency, 1988) for further discussion of this
relationship.
Monitoring During Treatment
Dredging is a major disruption of the existing ecology. Habitat for benthic or-
ganisms is drastically altered, and changes in the water column are possible
during treatment. In general, however, these perturbations are transient. The ben-
thos will normally recolonize the dredged area, and the chemical conditions of the
water column above the new sediments will reach an equilibrium with the
chemistry of the newly exposed sediments.
A detailed monitoring plan for a dredging project depends upon the specific
characteristics of the sediments to be removed. If, for example, the sediments
contain materials such as mercury or PCBs, then special provisions must be
made for their removal and disposal. An in-lake monitoring plan would also require
that these parameters be measured in the water column. There is no way to an-
ticipate and design a monitoring plan without prior knowledge, through a Phase I
Various dredging
techniques are hydraulic,
dragline, and backhoes.
Knowledge of sediment
chemistry is very
important for disposal.
Dredge below the
maximum depth of plant
colonization.
Disposal of contaminated
sediments is a serious
problem.
5-;
31
-------
lake restoration study, of potentially dangerous materials in the sediments. A
recommended monitoring plan is described in Table 5.16 for a dredging project
without contaminated sediments.
Table 5.16.—tn-lake monitoring design during dredging
PHYSICOCHEMICAL
A. Water Chemistry
1. Sample Location
Water samples should be collected at the site(s) selected by the project manager, usually at
the center of the lake.
2. Depth Distribution
Samples should be collected at 6-foot intervals from just below the surface to the bottom.
Care should be taken not to include suspended bottom sediments in the water samples.
3. Analytical Determinations and Sampling Procedures
Water samples should be analyzed for total phosphorus, dissolved reactive phosphorus,
ammonium nitrogen, and pH. See Chapter 3 for appropriate analytical and sampling
techniques.
4. Frequency and Duration
Samples should be collected at monthly intervals during the dredging'operation.
B. Dissolved Oxygen (DO) and Temperature
1. Sampling Location
Same as for water chemistry.
2. Depth Distribution
Measurements should be made at 3-foot intervals from the surface to the bottom.
3. Analytical Determinations and Sampling Procedures
See Chapter 3 for appropriate analytical and sampling techniques.
4. Frequency and Duration
Same as for water chemistry.
C. Secchi Disk Transparency
1. Sampling Location
Same as for water chemistry.
2. Frequency and Duration
Same as for water chemistry.
BIOLOGICAL
A. Chlorophyll a (corrected for pheophytin)
1. Sampling Location
Same as for Secchi disk.
2. Depth Distribution
A subsample should be obtained from an integrated sample representing a water column
equal to 0-6 feet from the surface.
3. Analytical Determinations and Sampling Procedures
See Chapter 3 for appropriate analytical and sampling techniques.
4. Frequency and Distribution
Same as for Secchi disk.
Considerations for Interrupting Treatment
The rationale for interrupting a dredging project is based upon an assessment of
the nontargeted dredging area within the lake. Sediment displacement within the
dredged area may create suspended solids problems in other areas, while in-
creasing the nutrient load within the system. In addition to a long list of potential
sediment contaminants, other measurements of concern are dissolved oxygen,
32
-------
nutrients, and unionized ammonium. A considerable increase in the oxygen
demand to the lake as a result of the disturbed sediments could cause increased
stress and mortality to the fishery, as could an increase in the unionized am-
monium concentration, and increases in whole lake nutrient concentrations could
promote algal blooms.
Consideration should be given to interrupting a dredging project under the fol-
lowing conditions:
1, If the dissolved oxygen concentrations in the non-target dredging area of
the lake or reservoir fall below 5 mg/L in the surface waters.
2. If the combination of the NH4-N concentration, pH, and temperature cre-
ate concentrations of unionized ammonium within the lake that are lethal
to fish.
Monitoring Following Treatment
A post-project monitoring plan following a dredging operation should be designed
to evaluate the success of the treatment. If the purpose of the dredging was to
deepen the lake and reduce rooted aquatic plants, then the evaluation should
concentrate on mapping both the new, deeper portion of the lake and the macro-
phyte distribution and density within the project area. Table 5.17 presents a
recommended monitoring plan to be followed after dredging is complete.
Table 5.17.—In-lake monitoring design after dredging
PHYSICOCHEMICAL
A. Mapping the Lake Bottom
Generally, lake bottom contours will be carefully resurveyed following completion of the dredging
to determine quantities for payment to the dredging contractor. In the absence of this survey,
there are several acceptable techniques available to determine the water depth above the
sediments. The most common technique uses a recording sonar unit. Steps to follow during
development of a lake depth map include
,1. Aerial Photographs
Obtain an aerial photograph of the lake. Mark a known straight line distance on the map for
calibration. Mark off and measure transect lines across the portion of the lake that was
dredged. The distance between the transect lines will vary depending upon the size of the
lake dredging area. The closer the transects are to each other, the more accurate the map. A
minimum of 50 feet and a maximum of 100 feet between transect lines is reasonable;
however the configuration of the dredged area will affect distances selected.
2. Benchmark
A benchmark must be established on the lakeshore for use as a reference to record lake
level at the time sonar soundings are made.
3. Sonar Transects
Transect markers should be established on (he shoreline, based upon the aerial photograph.
A boat with a sonar and a strip recorder should traverse between the two established
markers at a given, slow, steady speed.
4. Lake Map
Using the calibrated aerial photograph and the strip chart from the sonar measurements, plot
the depth to sediment surface along each transect. When all the depths are recorded along
the transects, join the identical depths (e.g., all 5-foot depths) to form a lake bathymetric map.
The lake map can be used to determine the amount of sediment removed if it is compared to
the before-dredging hydrographic map using the normalized lake level.
BIOLOGICAL
A. Macrophytes
The aquatic macrophytes should be surveyed during the second year following completion of the
project. They should be surveyed twice during the growing season (usually in late June and
again in August) to determine species composition and distribution, abundance, and maximum
depth of growth and depth from water surface to tops of plants.
5-33
-------
Control Nuisance Plants
Summary
1. Techniques used for controlling nuisance rooted plant growth in-
clude water level drawdown, mechanical or chemical controls, and
biological controls.
2. Common monitoring parameters for all techniques include phos-
phorus, nitrogen, chlorophyll a, macrophytes, and Secchi depth. * .
3. The major long-term problems observed with plant control projects
are damage to a lake's fishery and increased algal growth following
macrophyte control.
A balanced aquatic plant community is essential to the ecological well-being of
all lakes. Aquatic plants benefit lakes by harboring food organisms for fish and
waterfowl, providing spawning areas and protective cover for fish, preventing
shoreline erosion and stabilizing the lake bottom, producing oxygen and organic
material for other life in the lake, and providing food and building materials for a
variety of wildlife species.
Unfortunately, aquatic plants can grow to excess in lakes, particularly those
lakes that have been disturbed by human activities or subjected to the introduc-
tion of non-native plant species. Excessive aquatic plant growth can seriously im-
pair a lake's recreational use by limiting boating, swimming, and fishing. These
nuisance weeds are one of the most common and frustrating problems faced by
lake users and managers in all areas of the United States.
A number of techniques exist for managing excessive aquatic plant growth. In
some cases, nuisance plants can be controlled by limiting excess nutrients and
sediment from watershed point and nonpoint sources. In-lake controls for aquatic
plants (mechanical controls such as harvesting, rototilling, and disturbance of
shallow water sediments; sediment covers; lake drawdowns to expose and com-
pact shallow sediments; and chemical controls) are more or less temporary
management measures that limit the impact of excess plants on desired recrea-
tional uses.
EPA Clean Lakes regulations explicitly state that plant harvesting and her-
bicide treatments are palliative measures that are ineligible for project funding
unless they are proven to be the most cost-effective measures available, and
necessary watershed nutrient controls have been installed. Commercially avail-
able biological controls are currently limited to herbivorous fish such as grass
carp that may provide some degree of longer-term control but also carry the risk
of destroying beneficial, nontarget plant species as well as nuisance plant com-
munities.
Potential risks associated with extensive macrophyte control projects include
the possibility that removal of too much plant growth could damage critical fish
habitat and spawning areas. Also, reductions in macrophyte growth can allow in-
creased algae production if available nutrients (which algae will use) have not
been reduced.
5-34
-------
CONTROL TECHNIQUE #1:
Water Level Drawdown
Technical Considerations
Lowering the water level of the lake exposes littoral sediments, plants, and plant
reproductive parts to drying stresses. The drying is accompanied by freezing
stress in overwinter drawdowns in the northern United States and heat stress in
summer drawdowns. Drawdown can also promote compaction and dessication of
highly organic sediments. Where highly organic sediments are exposed to freez-
ing and dessication, however, high nutrient release rates have been observed fol-
lowing refill.
In addition to aquatic macrophyte control, drawdowns are used to compact
sediments to increase lake depth qnd, by fishery managers, to concentrate fish
either for greater predation or to increase the cost-effectiveness of chemical
eradication treatments. Drawdowns can have both acute and chronic impacts on a
lake's fishery. If they are not conducted properly, dissolved oxygen stress can
result in partial or complete fish kills. Elimination of critical plant habitat can
damage future repopulation, causing long-term changes to a lake's fishery.
The timing of drawdown for aquatic plant control depends upon the regional
climatic characteristics and the lake's recreational uses. Winter drawdowns are
most effective in climates with harsh winter conditions when there would be less
disruption of recreation on the lake; therefore, summer drawdowns are rarely
recommended in such regions.
A drawdown's effectiveness is also highly species-specific. Some macrophyte
species show dramatic decreases after drawdowns, but other species react
variably or even increase in abundance. Drawdown experiences with a variety of
species are summarized by Cooke et al. (1986). The benefits of a lake drawdown
are limited to a few years; therefore, the technique will probably have to be
repeated regularly to maintain the reduced plant population.
Monitoring During Treatment
A distinction should be made between monitoring the first instance of drawdown
on a lake and monitoring during subsequent maintenance drawdowns. If a lake
has not been drawn down for several years, the monitoring needed will be some-
what more intensive than that needed thereafter.
A key parameter to monitor during all lake drawdowns is dissolved oxygen.
Other parameters that may be monitored during lake drawdown are phosphorus
and nitrogen species, chlorophyll a, and Secchi depth.
Table 5.18 describes a recommended plan for monitoring during a lake draw-
down and refilling.
Considerations for Interrupting Treatment
The greatest risk when a lake is being drawn down is loss of dissolved oxygen in
the remaining lake pool. This risk increases as a greater amount of organic matter
enters the remaining pool relative to the volume of the pool. The risk can also be
great if, during a summer drawdown, the level of a thermally stratified lake is
reduced enough to cause mixing of low-oxygen, hypolimnetic waters with higher-
oxygen surface waters. Dissolved oxygen readings become the criterion for deter-
mining if a lake drawdown must be interrupted, or if artificial aeration should be
initiated.
5-35
Some macrophyte species
show dramatic decreases
after drawdowns, but
other species react
variably or even increase
in abundance.
-------
Table 5.18.—Iri-lake monitoring design during complete lake water level drawdown
and refilling
PHYSICOCHEMICAL
A. Dissolved Oxygen (DO) and Temperature
1. Sampling Location
Measurements should be made at the site(s) selected by the project manager, usually at the
deepest part of the lake.
2. Depth Distribution
Measurements should be made at 3-foot intervals from the surface to the bottom,
3. Analytical Determinations and Sampling Procedures
See Chapter 3 for appropriate analytical and sampling techniques.
4. Frequency and Duration
Measurements should be made at 10 percent time intervals from the start of drawdown until
the lake begins to refill, and thereafter at 20 percent increments until the lake has reached full
stage.
B. Downstream Observations •
Periodic observations (at least weekly) of downstream effects of increased discharges (flooding,
erosion, sedimentation) should be made during the period of the drawdown.
A second criterion for interruption is the downstream effects of the drawdown.
Sustained high flows downstream of the lake may result in flooding of lands and/or
destruction of fish and wildlife habitat. These problems can usually be avoided
through careful design and implementation of a drawdown.
The following guidelines should be followed in considering interruption of a lake
drawdown:
1. In-lake:
* If the epilimnetic dissolved oxygen concentration decreases to 5
mg/L or less.
2. Downstream:
* If observations indicate excessive flooding or other damage as a
result of increased discharge volumes.
The purpose of interrupting a drawdown because of in-lake dissolved oxygen
depletion is to weigh the risk of an unanticipated loss of the lake's fish resources
and to allow dissolved oxygen levels to recover. In cases where the present
fishery is not a valuable resource, the decision may be made to continue the draw-
down despite the loss of dissolved oxygen.
Monitoring Following Treatment
The measure of success of a lake drawdown for aquatic plant control is the degree
to which plant growth decreases from pre-project levels. Success will be en-
hanced if the aquatic plants are not replaced by increased algae growth. Macro-
phyte surveys should be carried out annually for the first two years to determine
change in species composition from pre-project conditions. Table 5.19 describes a
recommended monitoring plan following refill of the lake.
-36
-------
Table 5.19.—In-lake monitoring design following lake water level drawdown and re-
filling
PHYSICOCHEMICAL
A. Secchl Disk Transparency
1. Sampling Location
Measurements should be made at the site(s) selected by the project manager, usually at the
center of the lake,
2. Frequency and Duration
Samples should be collected at monthly intervals following refill of the lake. Data should be
obtained for at least a one-year period.
BIOLOGICAL
A, Chlorophyll a (corrected for pheophytin)
1. Sampling Location
Same as for Secchi disk.
2. Depth Distribution
A subsample should be obtained from an integrated sample representing a water column
equal to 0-6 feet from the surface.
3. Analytical Determinations and Sampling Procedures
See Chapter 3 for appropriate analytical and sampling techniques.
4. Frequency and Duration
Same as for Secchi disk.
B. Macrophytes
The aquatic macrophytes should be surveyed during the first and second years following
completion of the project. They should be surveyed twice during the growing season (usually in
late June and again in August) to determine species composition and distribution, abundance,
and maximum depth of growth and depth from water surface to tops of plants.
CONTROL TECHNIQUE #2:
Mechanical or Chemical Control of Nuisance Plants
Technical Considerations
There are a variety of aquatic plant control measures that attack unwanted plants
directly. These techniques are essentially temporary or cosmetic in nature. The
goal of each of these techniques is simply to achieve a short-term reduction in
nuisance plant growth. Because of the similarities in monitoring strategies, they
will be treated as a group for the purposes of this manual.
Mechanical or chemical control of nuisance plants includes those measures
that either physically prevent unwanted plants from growing or remove unwanted
plants from the lake. Specific techniques include:
Survey macrophytes to
determine species
composition, distribution,
abundance, maximum
depth of growth, and
depth from water surface
to tops of plants.
... sediment covers are
spread on an area of
lakebed to retard plant
growth.
¦ Bottom screens and other types of sediment covers, which are spread on
an area of lakebed to retard plant growth. The most effective screen materials
are gas-permeable so they will not be buoyed up off the lakebed by the
gaseous products of plant decomposition. Some covers must be removed and
cleaned annually. Bottom screens are typically used for relatively small areas,
such as around piers, beaches, and in boating lanes.
5-37
-------
Only rarely is it possible
to use harvesting as a
restoration technique by
removing enough plant
nutrients to achieve net
nutrient removal from
the lake.
¦ Mechanical harvesting, which relies on large machinery to cut vegetation
and remove it from the lake. Only rarely is it possible to use harvesting as a
restoration technique by removing enough plant nutrients to achieve net
nutrient removal from the lake. Most often harvesting is used during the grow-
ing season to provide immediate short-term relief from conditions that impair
boating and swimming.
¦ Tilling of lake sediments, which disturbs and dislodges root masses in the
lake sediment. This technique is usually performed in the spring or fall when
there is less vegetative matter in the water. Like harvesting, it is best to
remove as much dislodged vegetative matter as possible after a sediment till-
ing operation.
¦ Chemical controls that involve application of various herbicidal agents to
kill unwanted plants; the specific chemicals used vary depending on the plant
species. Mode of operation, selectivity, and use restrictions that must be
placed on the waterbody also vary from chemical to chemical.
Monitoring During Treatment
The monitoring required during mechanical and chemical treatments depends on
the nature and magnitude of the techniques employed.
Installation of bottom screens, alone, requires no monitoring other than periodic
visual observation of the screen to ensure that it remains firmly in place on the
sediments and does not buckle because of disturbance or entrapment of gasses.
No recommended monitoring specifications are given for bottom screen installa-
tion.
For mechanical harvesting and sediment tilling, the most important parameters
to monitor are changes in macrophyte coverage and subsequent algal response.
Less important parameters include total and dissolved reactive phosphorus, am-
monium, nitrite + nitrate, and total Kjeldahl nitrogen, chlorophyll a, and dissolved
oxygen and temperature profiles. A recommended monitoring plan is described in
Table 5.20.
For large-scale herbicide treatments, changes in macrophyte coverage should
be observed. Also, dissolved oxygen should be measured at regular intervals prior
to and following a herbicide treatment. The length of these intervals will depend on
the waiting period associated with the herbicidal action of each chemical. Over the
longer term, nutrient levels (total and dissolved phosphorus, ammonium, nitrate,
and total Kjeldahl nitrogen) and chlorophyll a should be measured. A recom-
mended monitoring plan is described in Table 5.21.
Considerations for Interrupting Treatment
There are no known serious adverse environmental impacts from mechanical
plant control projects that would warrant immediate interruption of treatment. The
impacts on a lake's fishery or changes in algal densities that might occur from
these techniques can be evaluated over one or more growing seasons and adjust-
ments to the techniques can usually be made to reduce adverse impacts to an ac-
ceptable level.
The most serious potential impacts of a chemical herbicide treatment are direct
toxicity of the herbicide to fish or wildlife and indirect toxicity to fish resulting from
depletion of dissolved oxygen by decomposing plants. Careful adherence to label
directions is essential if the risk of direct and indirect toxicity is to be minimized.
This risk can be managed further by examining the treatment area for indications
of fish and wildlife toxicity after the first 50 percent of treatment is complete and
immediately following treatment.
5-;
¦38
-------
Table 5.20.—In-lake monitoring design for plant harvesting and sediment tilling
PHYSICOCHEMICAL
A. Seechi Disk Transparency
1, Sampling Location
Measurements should be made at site(s) selected by the project manager, usually at the
center of the lake,
2. Frequency and Duration
Measurements should be made at monthly intervals. During sediment tilling operations,
additional measurements should be made at weekly intervals.
BIOLOGICAL
The mass of plants
harvested should be
recorded.
A. Chlorophyll a (corrected for pheophytin)
1. Sampling Location «
Same as for Secchi disk.
2. Depth Distribution
A subsample should be obtained from an integrated sample representing a water column
equal to 0-6 feet from the surface.
3. Analytical Determinations and Sampling Procedures
See Chapter 3 for appropriate analytical and sampling techniques.
4. Frequency and Duration
Samples should be collected at monthly intervals.
B. Macrophytes
Maerophyte regrowth should be observed and documented at monthly intervals during the
harvesting operation and until the end of the growing season. Mass of plants harvested should
be calculated. Generally, several representative loads should be calibrated by weighing with total
mass determined by keeping records of the number of loads removed. Several representative
plant samples should be obtained from the harvested plants and tissue total phosphorus and
percent water determined.
For sediment tilling operations, the aquatic macrophytes should be surveyed during the first
and second years following completion of the project. They should be surveyed twice during the
growing season (usually in late June and again in August) to determine species composition and
distribution, abundance, and maximum depth of growth and depth from water surface to tops of
plants.
Dissolved oxygen depressions can usually be avoided by limiting treatment
areas to a fraction of the lake area and treating only when oxygen levels are 5
mg/L or more. Dissolved oxygen measurements should normally be made as near
to dawn as possible when dissolved oxygen levels will be the lowest because of
nighttime plant respiration. Generally, decreases in dissolved oxygen occur
gradually following the treatment as the herbicide kills the treated plants and they
begin to decompose.
Except for direct herbicide toxicity, any negative effects of mechanical or
chemical plant control techniques would not be observed until well after treatment.
Therefore, no specific criteria for interrupting treatment are given for these techni-
ques.
Monitoring Following Treatment
The measure of success for both mechanical and chemical plant control methods
is the short-term improvement in recreational or other use of the lake resulting
from the reduction in nuisance plant growth. These techniques must generally be
repeated every year to maintain increased use of the lake, although the process
repeated over many years can result In longer-term changes to the lake environ-
ment. Recommended specifications are given in Tables 5.20 and 5.21.
Because mechanical and chemical plant control have the potential to change
the structure of a lake's fish community and can encourage increased algal
production, long-term monitoring of these lake ecosystem components is recom-
mended along with evaluation of longer-term changes in plant growth.
The measure of success
for both mechanical and
chemical plant control
methods is the short-term
improvement in
recreational or other use
of the lake due to the
reduction in nuisance
plant growth.
5-39
-------
Table 5.21.—In-lake monitoring design for large-scale herbicide applications
PHYSICOCHEMICAL
A. Dissolved Oxygen (DO) and Temperature
1. Sampling Location
Measurements should be made at the site(s) selected by the project manager, usually within
the area being treated and near the center of the lake.
2. Depth:Distribution
Measurements should be made at 3-foot increments from the surface to the bottom,
3. Analytical Determinations and Sampling Procedures
Samples should generally be obtained as close to dawn as possible. See Chapter 3 for other
appropriate analytical and sampling techniques.
4. Frequency and Duration
Measurements should be made at two-day intervals, for herbicides with waiting periods
between application and kill of less than two weeks. For herbicides with longer waiting
periods, measurements should be made at weekly intervals.
B. Secchi Disk Transparency
1. Sampling Location
Measurements should be made at site(s) selected by the project manager, usually at the
center of the lake.
2, Frequency and Duration
Measurements should be made at two-week intervals until the end of the growing season,
BIOLOGICAL
A. Chlorophyll a (corrected for pheophytin)
1. Sampling Location
Same as for Secchi disk.
2. Depth Distribution
A subsample should be obtained from an integrated sample representing a water column
equal to 0-6 feet from the surface.
3. Analytical Determinations and Sampling Procedures
See Chapter 3 for appropriate analytical and sampling techniques,
4. Frequency and Duration
Same as for Secchi disk.
B. Macrophytes
Changes in macrophyte coverage and regrowth in the treated areas should be observed and
documented at monthly intervals until the end of the growing season,
C. Toxic Effects on Fish and Wildlife
The treatment area should be examined for indication of fish and wildlife toxicity after the first 50
percent of treatment is complete and after the entire treatment is complete.
CONTROL TECHNIQUE #3:
Biological Control of Nuisance Plants (Grass Carp)
Technical Considerations
The only biological controls currently accepted for use against aquatic plants are
genetically sterile grass carp (also known as white amur) Ctenopharyngodon idel-
la, although research is being conducted on numerous other biological agents, in-
cluding different fish species, insects, and plant pathogens, as well as on the
results of food chain manipulations.
Grass carp are more effective in warmer regions. Formulae used in some
States yield a stocking rate based on climatic region, plant type, and amount of
-40
-------
vegetation in a lake (Wiley et a!. 1987). Use of these fish for plant control is
restricted or prohibited in many States because of their potential for damaging
natural fisheries, cycling nutrients, and destroying beneficial plant communities.
A number of factors combine to reduce grass carp feeding rates in a lake over
the long term. These factors include reduced consumption as fish mature, as well
as mortality and escape. Supplemental stocking of additional fish is often required
to maintain initial vegetation consumption levels. However, where initial consump-
tion levels are too high, this natural reduction in activity can benefit the lake.
Stocking strategies for grass carp take into account this change in feeding rate
over time: in serial stocking, additional fish are placed in the lake at intervals,
while in batch stocking, enough fish are placed in the lake initially to compensate
for decreases in efficiency. Serial stocking reduces the risk of long-term damage
to the biotic systems of the lake and is usually the recommended strategy.
Monitoring for the First Year A fter Fish Stocking
The first-year monitoring program for grass carp projects should focus on macro-
phyte responses as well as in-lake nutrient and algae levels. Experience has
shown that grass carp feed preferentially on certain species of aquatic plants such
as naiads, Chara spp., and most pondweeds. Since it is important to ensure
that stocked grass carp do not destroy all plant species beneficial to fish and
wildlife, leaving large growths of nuisance species such as Eurasian watermilfoil,
periodic macrophyte composition studies should be conducted after the first
stocking.
In some grass carp projects, the water's nutrient levels have increased follow-
ing elimination of large numbers of macrophytes, which suggests the importance
of monitoring for phosphorus and nitrogen after the introduction of grass carp. In
addition, chlorophyll a and Secchi depth measurements should be made regularly
during the growing season.
Table 5.22 describes a recommended plan for monitoring a lake stocked with
grass carp during the first year after stocking.
Considerations for Interrupting Treatment
Grass carp projects are difficult if not impossible to interrupt should serious ad-
verse impacts arise; therefore, serial stocking is recommended. Selective removal
of grass carp from any lake (but particularly from large lakes) is difficult no matter
which method is used. In addition, any criteria for interrupting treatment will
depend upon the initial plant communities that existed in the lake and the objec-
tives of the project in terms of plant removal. Therefore, the guidelines for inter-
rupting a grass carp treatment may prove difficult to follow unless a clear
percentage plant removal target is determined at the beginning of the project. In-
terruption can take the form of eliminating the second stocking of grass carp in a
serial stocking strategy if selective removal of stocked fish is found to be impos-
sible.
Interruption of a grass carp stocking program should be considered under the
following conditions:
1. Removal of significantly greater than target percentage of total plant
cover.
Grass carp projects are
difficult if not impossible
to interrupt should
serious adverse impacts
arise; therefore, serial
stocking is recommended.
2. Evidence of grass carp escape into an adjacent hydrologic system.
(This may require a complete fishery eradication.)
3. Evidence of a significant increase (beyond natural variability) in nutrient
or phytoplankton biomass that adversely affects use of the lake.
5-41
-------
Table 5.22—In-Iake monitoring designed for the first year after herbivorous
fish stocking.
PHYSICOCHEMICAL
A. Dissolved Oxygen (DO) and Temperature
1. Sampling Location
Measurements should be made at the site(s) selected by the project manager, usually at the
center of the lake.
2. Depth Distribution
Measurements should be made at 3-foot intervals from the surface to the bottom.
3. Analytical Determinations and Sampling Procedures
See Chapter 3 for appropriate analytical and sampling techniques.
4. Frequency and Duration
Measurements should be made at monthly intervals during the growing season (May through
October).
B. Secchi Disk Transparency
1. Sampling Location
Same as for dissolved oxygen.
2. Frequency and Duration
Same as for dissolved oxygen.
BIOLOGICAL
A. Chlorophyll a (corrected for pheophytin)
t. Sampling Location
Same as for Secchi disk.
2. Depth Distribution
A subsample should be obtained from an integrated sample representing a water column
equal to 0-6 feet from the surface.
3. Analytical Determinations and Sampling Procedures
See Chapter 3 for appropriate analytical and sampling techniques.
4. Frequency and Duration
Same as for Secchi disk.
B. Macrophytes
The aquatic macrophytes should be surveyed one month following the stocking. They should be
surveyed to determine species composition and distribution, abundance, and maximum depth of
growth and depth from water surface to tops of plants.
C. Fish
Fish populations should be sampled during the first year after initial stocking of herbivorous fish.
Mark-recapture techniques such as those described by Youngs and Robson (1978) should be
used to estimate numbers. Lengths and weights should be obtained from samples of each
species to enable the evaluation of population and size structure.
Monitoring Following Treatment
A grass carp stocking project is a success when the reduction in nuisance plant
growth is consistent with the avoidance of long-term adverse changes to the biotic
integrity of the lake. The reduction in nuisance plant growth can be evaluated by
ground observations or aerial photography, but in a grass carp project it is par-
ticularly important to evaluate changes in species composition by conducting a
macrophyte survey, which should be completed during the first growing season
after stocking is initiated. Similarly, to ensure that adequate comparative data are
available on base year fish population structure, a fishery survey should be com-
pleted during the first spring and fall after stocking.
Over the longer term, macrophyte surveys should be completed once a year
during the first three years to determine if consumption of vegetation is decreasing
5-42
Mark-recapture
techniques should be
used to survey grass
carp populations.
-------
over time. Perhaps more importantly, a fisheries survey should he completed
every two years after grass carp are stocked in any lake in which fish are a valued
resource. Grass carp have the potential to exert strong adverse effects on native
fish populations through consumption of vegetation needed for fish spawning,
cover, and food organisms, and, unless the fish are inspected prior to stocking by
a fish pathologist, through transmission of disease (Cooke et al. 1986). The
fisheries survey should be completed before any decision is made to stock sup-
plemental grass carp to maintain consumption levels.
Table 5.23 describes a recommended monitoring plan for use after the first
year of a herbivorous fish stocking project.
Grass carp have the
potential to exert strong
adverse effects on
native fish populations
through consumption
of vegetation needed for
fish spawning, cover, and
food organisms.
Table 5.23.—In-lake monitoring design after the first year of a herbivorous fish
stocking project
PHYSICOCHEMICAL
A. Dissolved Oxygen (DO) and Temperature
1. Sampling Location
Measurements should be made at the site(s) selected by the project manager, usually at the
center of the lake.
2. Depth Distribution
Measurements should be made at 3-foot intervals from the surface to the bottom.
3. Analytical Determinations and Sampling Procedures
See Chapter 3 for appropriate analytical and sampling procedures.
4. Frequency and Duration
Measurements should be made at monthly intervals for at least two years after the initial
stocking of the herbivorous fish.
B. Secchi Disk Transparency
1. Sampling Location
Same as for dissolved oxygen.
2. Frequency and Duration
Secchi disk measurements should be taken monthly during the growing season (May through
October) for at least two years after the initial stocking of the herbivorous fish.
BIOLOGICAL
A. Chlorophyll a (corrected for pheophytin)
1. Sampling Location
Same as for Secchi disk.
2. Depth Distribution
A subsample should be obtained from an integrated sample representing a water column
equal to 0-6 feet from the surface.
3. Analytical Determinations and Sampling Procedures
See Chapter 3 for appropriate analytical and sampling techniques.
4. Frequency and Duration
Same as for Secchi disk.
B. Macrophytes
The aquatic macrophytes should be surveyed during the first and second years following
stocking. They should be surveyed twice during the growing season (usually in late June and
again in August) to determine species composition and distribution, abundance, and maximum
depth of growth and depth from water surface to tops of plants.
C. Fish
Mark-recapture techniques such as those described by Youngs and Robson (1978) should be
used to estimate numbers. Length and weight should be obtained from samples of each species
to enable the evaluation of population and size structure.
5-43
-------
WiE1isl$lii)k>Jsn
Mitigate Acidic Conditions
Acidic lakes cannot
support healthy fisheries.
Summary
1. Addition of limestone to an acidic lake or its watershed can allow success-
ful stocking of game fish species and help return the lake to a productive
sport fishery.
2. The treatment phase ordinarily lasts approximately one month, beginning
with the initiation of liming and continuing during the time that the lake
chemistry approaches a new equilibrium. Key monitoring parameters are
pH, acid neutralizing capacity, and calcium.
3. The post-treatment phase involves monitoring most of the same parameters
Included in treatment monitoring plus certain parameters likely to change
slowly following base addition, such as the biological parameters.
4. Addition of base materials to watersheds has similar but much longer last-
ing effects than surface water treatment. Monitoring considerations, there-
fore, may be different temporally but will include the same parameters.
"Liming" is a generic term used to connote the addition of any base materials to
neutralize surface water or sediment or to increase alkalinity. The most common
product used to treat acidic lakes is limestone, the same mineral used in agricul-
ture (Olem, 1989). Limestone can be applied to the lake surface, injected into the
sediment, continuously dosed to upland streams, or applied to the watershed.
Fisheries managers have known for years that adding lime to acidic, un-
productive lakes can allow successful stocking of game fish species and help
return the lake to a productive sport fishery. Acidic lakes occur in areas where the
soils have no natural buffering capacity and acid rain and other processes cause
acidification of waterbodies. Many of these lakes are unable to support a healthy
reproducing fishery.
There are other sources of acids to lakes that are not related to pollutants in the
air. Some waters are mildly acidic because of their passage through naturally
acidic soils. Stained lakes, for instance, may have pH levels between 5 and 6.
Acidic deposition to these lakes contributes additional mineral acidity to already
slightly acidic waterbodies.
Acidic drainage from abandoned mines affects thousands of miles of streams
and numerous lakes throughout Appalachia (Olem, 1989). Acid mine drainage
also occurs in the midwestern coal fields of Illinois, Indiana, and Ohio and in coal
and metal mining areas of the western United States, where affected streams and
lakes can have pH levels below 4, In some cases, liming can restore these lakes to
productive use.
All liming projects should include a rigorous monitoring program designed to
characterize changes in key hydrological, physical, chemical, and biological
parameters during and after treatment. The results of a monitoring program will
help determine whether the project meets its water quality and biological objec-
tives. Monitoring may also help determine when it is appropriate to stock a
treated lake with selected fish species.
5-
44
-------
CONTROL TECHNIQUE #1:
In-Iake Liming
The use of base materials to neutralize acidic lakes is a proven technique for res-
toring waterbodies that are acidic for a variety of possible reasons. Addition of
base materials to the lake surface is currently the most common treatment to
mitigate acidic conditions.
A good example of an in-Iake liming project is the Lake Acidification Mitigation
Program (LAMP) funded by the Electric Power Research Institute (Porcella,
1989). Two drainage lakes located in the Adirondack region of New York, Woods
Lake and Cranberry Pond, were treated with limestone in 1985. Both lakes were
characterized as small acidic headwater systems with short residence times and,
although they differ in size and depth, both lakes have similar watershed charac-
teristics.
A fine limestone slurry was distributed in the lakes, which resulted in a high dis-
solution efficiency. Four weeks after liming, dissolution was 86 percent and 79
percent in Woods Lake and Cranberry Pond, respectively. Essentially all of the
limestone was dissolved in both lakes within four months of application with only
minimal accumulation in the bottom sediments.
The short-term changes in the water chemistry of Woods Lake included an im-
mediate increase in pH from less than 5.0 to above 9.0; a stabilization in pH below
8.0 after equilibrium with atmospheric CO2 was reached; increase in calcium,
alkalinity, and dissolved inorganic carbon; and a shift in speciation of aluminum
from a dominance of organic (non-labile) monomeric aluminum to the inorganic
(labile) monomeric form (Fordham and Driscoll, 1989; Driscoll et al. 1989b). After
about one month, marked decreases in aluminum, manganese, and zinc were ob-
served in the water column. These minerals accumulated as mineral precipitates
in lake bottom sediments.
Growth and condition of stocked trout were reported by Gloss et al. (1989) to
be good after liming. Spring fingerling fish survival over the first four months after
stocking in both lakes was nearly identical (66 and 64 percent) to average survival
rates in circumneutral Adirondack lakes.
Overall, no deleterious effects of liming were observed. Maintenance of
suitable water quality conditions allowed the reintroduction and restoration of the
brook trout population (Schofield et al. 1989).
Technical Considerations
A monitoring program for lakes is implemented in three phases: pretreatment,
treatment, and post-treatment. This section discusses monitoring during the tran-
sitional and post-treatment phases.
The treatment phase should normally last about one month. The actual addition
of base will normally take one to five days, depending on the method of applica-
tion. Helicopter application, for example, is usually the fastest method; the entire
surface of the lake can be uniformly covered in several hours. On the other hand,
treatments by other methods may take several days. For instance, application of
slurried limestone to a large lake in New York State took five days when a small
boat was used to apply the material (Brocksen and Emler, 1988).
The actual length of the treatment phase depends on when the water chemistry
has stabilized in terms of its immediate response to the base application. Lime-
stone treatment often causes an immediate pH increase to very high levels until
new equilibrium conditions are reached with respect to carbon dioxide and other
carbonate species. The transition phase measurements will monitor these chan-
ges until water chemistry stabilizes.
Restore acidic lakes by
adding limestone.
In two New York State
lakes that were limed,
trout grew well.
Addition of limestone to a
lake may take 1-5 days.
5-45
-------
Some of the elements described for monitoring lakes treated with base
materials Influence and are influenced by the treatment methodology. For ex-
ample, sediment sampling and analyses are different when the treatment techni-
que involves injection of base materials into sediments rather than the more
common application directly to the water column (Ripl, 1980). Water quality
parameters may also differ depending on the base material applied. Although
limestone Is the most commonly used base material for neutralization, other
chemicals have been used, such as calcium hydroxide, calcium oxide, sodium
carbonate, and sodium bicarbonate (Olem, 1989). Sodium, and not calcium,
would be a monitored parameter when sodium-based neutralizing materials are
used in place of the more common calcium-based chemicals.
Certain key parameters
are likely to change
immediately after
liming: pH, turbidity,
acid neutralizing
capacity, calcium, and
aluminum.
Monitoring During Treatment
The physical and chemical parameters to be monitored during treatment include
those that can indicate the response of the system to the neutralization treatment
and provide information on the base material's distribution throughout the water
column. Certain key parameters are likely to change immediately after liming: pH,
turbidity, acid neutralizing capacity, calcium (when calcium-based materials are
used), and aluminum. Acid neutralizing capacity is a Gran titration method that in-
cludes alkalinity plus additional buffering from dissociated organic acids and other
compounds. All parameters should be monitored during the treatment phase,
preferably on a weekly basis.
Table 5.24 summarizes the recommended in-iake parameters to be monitored
during the treatment phase. Physicochemical parameters should be evaluated
during this period, while characterization of sediment chemistry and biological
parameters should be reserved for the regular post-treatment phase.
Sampling during the treatment phase would be conducted immediately prior to
addition of the base material, during the neutralization process, and weekly for
about one month following treatment. This monitoring characterizes transitional
changes in physical and chemical parameters as the system changes from acidic
conditions to neutral or alkaline conditions. During this transitional period the
Table 5.24.—Monitoring during the treatment phase of an in-lake liming project
PHYSICOCHEMICAL
A. Water Chemistry
1. Sampling Location
Samples should be collected at the site(s) selected by the project manager, usually at the
deepest part of the lake for whole-lake liming.
2. Depth Distribution
Samples should be collected from just below the surface and at 6-foot intervals to the bottom.
Care should be taken not to include suspended bottom sediments in the water samples.
3. Analytical Determinations and Sampling Procedures
Water samples should be analyzed for acid neutralizing capacity (Gran plot), pH, turbidity,
calcium, and dissolved aluminum. See Chapter 3 for appropriate analytical and sampling
techniques.
4. Frequency and Duration
Samples should be collected immediately prior to base addition, once during treatment, and
weekly for one month following treatment.
B. Dissolved Oxygen (DO) and Temperature
1. Sampling Location
Same as for water chemistry.
2. Depth Distribution
Measurements should be made at 3-foot intervals from the surface to the bottom.
3. Frequency and Duration
Same as for water chemistry.
5-
46
-------
water may, until the base materials are mixed, be highly alkaline. There have been
reports of pH values as high as 9.5 immediately after liming, even when slightly
soluble limestone has been used as the neutralizing agent (Fordham and Driscoll,
1989). The treatment monitoring program, therefore, should be more intensive
temporally and spatially and less intensive with respect to monitored parameters
than the post-treatment phase.
The hydrologic parameters, lake level and discharge, should also be con-
sidered. These data will allow more accurate assessment of the water quality
changes that occur. For example, a major storm event immediately following treat-
ment may have effects on water quality that would not occur under normal
hydrologic conditions.
Sampling should be conducted just before treatment begins, during the treat-
ment process, and weekly thereafter until equilibrium conditions are reached. The
minimum water column measurements should include samples collected from the
surface and bottom (three feet from sediment surface). Samples should be col-
lected at the deepest point in the lake and, if possible, in the major embayments. It
is desirable to also collect other samples from the major inlet streams and the lake
outlet.
Considerations for Interrupting Treatment
Because liming the lake surface ordinarily is completed in a few days, no criteria
are presented for interrupting treatment. Addition of excessive amounts of base
material would not generally be noticed until after the planned dosage had been
applied.
Monitoring Following Treatment Reunification may
Post-treatment monitoring may be extended over several annual cycles or for one occur sooner when there
hydrologic retention time. Reacidification may occur sooner because of factors are unusually high storm
such as incorrect dosage calculation or unusually high storm flows. Monitoring fiows
would help determine when retreatment is needed.
AH of the parameters monitored during the treatment period, with the exception
of turbidity, should also be measured during the post-treatment period. Turbidity is
monitored immediately following liming to evaluate how long undissolved lime-
stone remains in the water column after treatment. Significant turbidity has been
shown to last for days and sometimes even weeks following treatment.
Certain parameters (including sediment analyses and all biological
parameters) are likely to change slowly following base addition. They do not need
to be monitored during the transition period but must be checked during the post-
treatment phase to evaluate changes in response to liming. Table 5.25 sum-
marizes the recommended in-lake parameters to be monitored during the
post-treatment phase and the locations for collection.
The success of lake liming depends on maintenance of adequate water quality
to sustain the desired aquatic communities (Bukaveckas, 1989; DePinto et al.
1989; Driscoll etal. 1989a; Roberts and Boylen, 1989; Schaffner, 1989). Monitor-
ing measurements, therefore, should include not only hydrological, physical, and
chemical measurements, but also biological parameters. The exact type of
biological measurements will depend on the management objectives of the
resource. Monitoring measurements for a put-and-take fishery, for instance, will
be different from a lake with a sustained, reproducing fish population.
pH values as high as 9.5
have been observed
immediately after liming.
5-47
-------
Table 5.25.—Monitoring following the treatment phase of an in-lake liming project
PHYSICOCHEMtCAL
A. Water Chemistry
1. Sampling Location
Sampled should be collected at the site(s) selected by the project manager, usually at the
deepest part of the lake for whole-lake liming.
2. Depth Distribution
Samples should be collected just below the surface and at 6-foot intervals to the bottom.
Care should be taken not to include suspended bottom sediments in the water samples.
3. Analytical Determinations and Sampling Procedures
Water samples should be analyzed for acid neutralizing capacity (Gran plot), pH, calcium,
and dissolved aluminum. See Chapter 3 for appropriate analytical and sampling techniques.
4. Frequency and Duration
Samples should be collected at monthly intervals for at least a one-year period following
treatment.
B. Dissolved Oxygen (DO) and Temperature
1. Sampling Location
Same as for water chemistry.
2. Depth Distribution
Measurements should be made at 3-foot intervals from the surface to the bottom.
3. Analytical Determinations and Sampling Procedures
See Chapter 3 for appropriate analytical and sampling techniques.
4. Frequency and Duration
Same as for water chemistry.
C. Secchi Disk Transparency
1. Sampling Location
Same as for water chemistry.
2. Frequency and Duration
Measurements should be made monthly during the growing season (May through October)
for at least a one-year period following treatment.
BIOLOGICAL
A. Chlorophyll a (corrected for pheophytin)
t. Sampling Location
Same as for water chemistry;
2. Depth Distribution
A subsample should be obtained from an integrated sample representing a water column
equal to 0-6 feet from the surface.
3. Analytical Determinations and Sampling Procedures
See Chapter 3 for appropriate analytical and sampling techniques.
4. Frequency and Duration
Same as for Secchi disk.
-48
-------
CONTROL TECHNIQUE #2;
Watershed Liming
Technical Considerations
Addition of base materials to the watershed of a lake is a relatively new technique
for mitigating acidic conditions. The method has been conducted on several
watersheds in Sweden, Norway, and Great Britain (Brown, 1988; Rosseland and
Hindar, 1988; Olem, 1989). In the United States, watershed liming has been prac-
ticed only recently. In 1989, Woods Lake watershed in the Adirondack region of
New York State was treated in the most comprehensive technical evaluation of
watershed liming to date.
The watershed liming projects conducted so far have been designed so that
neutral or alkaline lake water conditions remain for decades. Acidic conditions
may never return in these situations if the source of the acidity is removed, such
as through a reduction in atmospheric emissions of SO2 and NOx.
Suggested monitoring considerations are not dramatically different for water-
shed versus in-lake treatment methods for mitigation of acidic conditions. The
major difference may be the temporal characteristics of the monitoring program.
Watershed treatment often results in much more gradual increases in pH and
changes in other water chemistry parameters compared to in-lake treatment
(Olem, 1989).
The treatment phase should normally last about two to three months. The ac-
tual addition of base materials will normally take 1 to 10 days, depending on the
method of application and area to be limed. Helicopter application, for example, is
usually the fastest method and may allow base materials to be added over the
selected subwatersheds over a period of hours. Some applications, such as ap-
plication of limestone by tractor, may take several days if many acres are to be
treated.
The actual length of the treatment phase depends on when the water chemistry
has stabilized in terms of its immediate response to base treatment. The treatment
phase measurements will monitor these changes until water chemistry stabilizes.
Monitoring During Treatment
Certain key parameters are likely to change during the first precipitation or snow-
melt event following base addition. These include pH, turbidity, acid neutralizing
capacity, calcium (when calcium-based materials are used), and aluminum.
These parameters should be monitored at least twice during the treatment phase,
preferably following a major precipitation or snowmelt event.
Recommended in-lake parameters to be monitored during the treatment phase
and the locations for collection are the same as for in-lake treatment (see Table
5.24).
Sampling during the treatment phase would be conducted immediately before
addition of the base material and later during major precipitation events. This
phase is intended to characterize transitional changes in physical and chemical
parameters as the system changes from acidic to neutral or alkaline conditions.
The minimum water column measurements should include samples collected
from the surface and bottom (three feet from sediment surface). Samples should
be collected at the deepest point in the lake and, if possible, at the major embay-
ments. It is highly desirable to collect samples from the inlet streams of the sub-
watersheds where base materials were applied.
Watershed liming may be
a long-term solution to
lake acidification once
the sources are controlled.
Water quality changes
should become evident
after the first rain or
snowmelt.
5-
49
-------
Monitoring Following Treatment
All of the parameters monitored during the treatment period, with the exception of
turbidity, should also be monitored during the post-treatment period. Turbidity is
monitored immediately following liming to evaluate whether undissolved limestone
is flushed into the lake water column from the watershed. This would normally
occur only when limestone material is distributed at or near the lake shoreline or
directly in the tributary streams. The other parameters should be monitored on a
quarterly basis during the post-treatment period.
Recommended in-lake parameters to be monitored during the post-treatment
phase are the same as for in-lake treatment (see Table 5.25).
The success of aquatic liming depends on maintenance of adequate water
quality to sustain the desired aquatic communities. Monitoring measurements,
therefore, should include not only physicochemical measurements but biological
parameters. The exact type of biological measurements will depend on the
management objectives of the resource. Monitoring measurements for a put-and-
take fishery, for instance, will be different from a lake where a reproducing fish
population will be sustained.
-50
-------
Chapter 6
A Long-term
Monitoring Protocol
Summary
1. A continuing data acquisition program should be estab-
lished to track changing water quality conditions and
guide future lake management actions.
2. The Phase II monitoring protocol should serve as a
model for the local sponsor, who should be encouraged
to continue the program after the project's formal com-
pletion,
3. An ongoing monitoring program should be designed
around common skeletal models to ensure collection of
consistent and comparable information. Additional data
can be obtained to meet specific lake requirements.
4. Long-term monitoring programs can range from Secchi
disk observations that are made every two weeks, to a
more complex program that develops basic water
chemistry and biological information, to a comprehen-
sive effort where all major lake ecosystem components
are tracked.
• ¦¦¦ — - I.', .-i..- ¦¦¦¦¦-
6-1
-------
Lake water quality
trends often cannot be
detected unless a
multi-year monitoring
record is available.
When designing a
monitoring program for
a particular lake, the
project manager's major
challenge is often not
technical but
sociological.
Background
Data on the condition of the major components of a lake ecosystem are a prereq-
uisite to sound management of tie resource. This information, along with an ongo-
ing database on water quality conditions within a lake, provides the basis for
sound decisions on managing lake water quality. If a decline in lake water quality
is noted, then a response must be made to identify and correct the cause. Conver-
sely, if a restored lake exhibits improvement over time, the manager will know the
present management strategy has been effective.
A major difficulty in detecting trends is that lakes, unlike rivers, often respond
slowly to changed external influences. Because of these inherent lags and a
natural background variability, lake water quality trends often cannot be detected
unless a muiti-year monitoring record is available. Wolman (1971), commenting
on the detection of water quality trends, suggested that formal statistical proce-
dures often can be used. He noted that
• Water quality records are often short term;
• Techniques and sensitivities of analytical methods have changed overtime;
• Sampling locations and frequencies have also often changed;
• Numerous interrelated physical, chemical, and biological variables
determine water quality;
• Natural background variability often hides water quality trends; and
• Causal explanation of trends requires a knowledge of human activities,
hydrologic processes, and land use in watersheds.
While these conditions may, in fact, make the use of statistical tools more dif-
ficult, they do not preclude their implementation. However, they do pose complica-
tions that must be considered during the statistical analysis.
Most of the above-noted complications can be addressed by establishing a
consistent, generic monitoring protocol upon which to build individual lake
monitoring programs. The Phase II monitoring effort offers an excellent oppor-
tunity to set up a lake-specific data acquisition program while building a consistent
and comparable nationwide lake database.
A lake monitoring program must be structured so the local sponsor will continue
monitoring to establish a multi-year period of record following completion of the
lake restoration or protection project. Once established, trend detection
methodologies such as those described by Montgomery and Reckhow (1984) can
be used to great advantage.
An ongoing, long-term monitoring program can be as simple as obtaining water
clarity information with a Secchi disk twice each month during the growing season
from a single, centrally located site in the lake, or it can be as complex as an in-
depth multi-faceted program that simultaneously obtains physical, chemical,
biological, and sociological data on a variety of lake ecosystem components.
When designing a monitoring program for a particular lake, the project
manager's major challenge is often not technical but sociological. The project
manager's judgment on the intensity of monitoring needed for a particular lake
must be accurate or the local sponsor may decide to discontinue the program be-
cause of costs, maintenance problems, or hard-to-understand protocol. Usually,
the most important factors are keeping costs low relative to the size of the project
and, most importantly, ensuring that a periodic, professional interpretation of the
data is made and communicated to the local sponsors.
6-2
-------
The following sections outline three levels of long-term monitoring that can
serve as frameworks on which to structure individual, lake-specific programs. The
more intensive efforts will generally be associated with those lakes where resi-
dents have recently completed a costly restoration project; simpler monitoring ef-
forts will often be associated with those lakes that have had minimal management
activity. Unfortunately, lakes that have good water quality and are most sensitive
to management changes are usually not managed. With these waterbodies, local
initiative is all too often stimulated in response to a crisis—when monitoring
programs are finally established after the fact.
Monitoring Water Clarity
Collection of Secchi disk information is nearly always a component of the monitor-
ing program—often it is the only component. Water clarity is an indirect measure
of water quality that is directly related to the public's perception of lake quality. It is
also data that are easy to obtain.
To be of most value, Secchi disk data should be collected once every two
weeks during the growing season. Although more frequent (weekly or daily) meas-
urements can be of some value, Smeltzer et al. (1989) noted that, for estimating
average lake water clarity conditions, weekly and biweekly sampling frequencies
yield almost the same amount of information. In many lakes, a single site located
near the center will provide the least biased information on average conditions
(Stauffer, 1988). Multiple sites may be needed where the lake has a very complex
configuration or when it is a long, river-run reservoir.
Ongoing support, analysis, and feedback to the person or group responsible for
managing the lake is necessary for even these simple monitoring efforts. Many
States have volunteer lake monitoring programs that could be used to provide this
necessary professional assistance; where these programs do not exist, an alter-
nate method of support should be established.
A Basic Lake Water Quality Monitoring Plan
Although water clarity information provided by Secchi disk data provides a basic
indication of lake quality, it offers no insights into causal factors affecting a lake's
condition. An example of an expanded water quality trend-monitoring protocol is
presented in Table 6.1. Although this protocol is very basic and oriented towards
smaller impoundments and natural lake environments, it does begin to provide a
database from which in-lake cause-effect inferences can be drawn and lake-to-
lake comparisons completed. And, most importantly, these data are relatively in-
expensive to obtain.
A similar protocol, presently being followed by the U.S. Geological Survey
(USGS) in some midwestern States, costs approximately $3,500 per year, per
station (1989 dollars). In addition, 50 percent cost-sharing is often available from
the USGS, thereby further reducing expenses.
While this basic monitoring program will often be adequate to describe lake
conditions, additional data are sometimes needed to address specific lake con-
cerns. Although there will always be lake-to-lake or region-to-region exceptions,
the scope of this already low-cost protocol will rarely be reduced. It is more likely
that additional parameters will be sampled or that changes will be made to the
timing of data collection.
Unfortunately, lakes that
have good water quality
and are most sensitive to
management changes are
usually not managed.
Water clarity is an
indirect measure of water
quality that is directly
related to the public's
perception of lake quality.
Ongoing support,
analysis, and feedback to
the person or group
responsible for
managing the lake is
necessary for even these
simple monitoring
efforts.
6-3
-------
Table 6.1 .—A basic water quality trend-monitoring protocol
APPROXIMATE TIMES OF COLLECTION
PARAMETER
MIDWINTER
SPRING DURING
MIXED CONDITIONS
JUNE
JULY
AUGUST
COMMENTS
Water chemistry
X
2 depths—1,5 feet from surface
—3.0 feet above bottom
Parameters; TP, DRP, NH4-N,
N02 + N0a-N, TKN, Ca, CI, Mg,
Na, K, pH, total alkalinity, color,
turbidity, total dissolved solids,
S04, Si02
Total phosphorus,
NOj-N. TKN
X
X
X
X
X
3 depths—1.5 feet from surface
—3.0 feet above bottom
—2.0 feet below the top
of the hypolimnion (if present)
Dissolved oxygen,
temperature, pH,
specific conductance
X
X
X
X
X
Profile—1.5 feet from surface,
proceeding to lake bottom using
3- to 6-foot intervals depending
on conditions and lake depth
Secchi-disk depth
X
X
X
X
X
Every 2 weeks
Chlorophyll a
X
X
X
X
6-foot integrated sample
Lake level
X
X
X
X
X
Every 2 weeks
Note: The sampling site will normally be located at the deepest point of the lake. On large lakes, more than one site may be required to adequately define water quality.
-------
Elements of the Basic Lake Water Quality Trend
Monitoring Program
¦ Comprehensive water chemistry data are obtained once each year to
describe the lake's water quality. To minimize within-lake variability and to
keep costs low, this information is usually collected when the lake is well
mixed, often at the time of spring overturn. Although a single sample might
characterize lake conditions, both top and bottom samples should be col-
lected and analyzed to ensure that vertical differences were not present at
the time of sampling. Ideally, the lake should be sampled more than once
during this well-mixed period. Smeltzer et al. (1989) noted significant in-
creases in the precision of describing spring phosphorus concentrations
when several samples were obtained on different dates during the spring
mix sampling period.
Information on nitrogen and phosphorus is collected because these
nutrients often limit plant production; on silica because it is often limiting to
diatoms; and on chloride and sodium because they are often good indicators
of the degree of watershed urbanization in regions not affected by marine
influences or by natural sodium weathering.
Data are also obtained on potassium, an indicator of animal waste
contamination that is found in high concentrations in cattle manure (Travis,
1988).
¦ Color and turbidity are other indicators of water clarity. Colored lakes,
such as those stained by organic acids, often have naturally low water
clarity. High turbidity in the absence of significant algal production is in-
dicative of suspended sediment that is limiting clarity. Data on alkalinity,
calcium, magnesium, sulfate, pH, and dissolved solids are useful in under-
standing in-lake phosphorus dynamics and sensitivity to acid deposition.
¦ Supplemental total phosphorus, NO3-N, total Kjeldahl nitrogen, and
chlorophyll a analyses should be made during the growing season. Along
with clarity data, these are the most common parameters used to describe
lake trophic status. When these data are compared to each other, insights
can be drawn regarding the importance of phosphorus, nitrogen, and cor-
responding algal production when water clarity is limited. Documenting
lake conditions during periods of strong stratification can help identify the
magnitude and potential of internal loading processes. Where internal
phosphorus loadings are important, it is necessary to obtain additional in-
formation on hypolimnetic iron, manganese, sulfate, and ammonium.
¦ Dissolved oxygen data, which provide the most basic description of
water quality, often serve as an indicator of lake productivity. In stratified
lakes, an anoxic hypolimnion suggests mesotrophic or eutrophic condi-
tions. Anoxia also favors sediment phosphorus release.
¦ Temperature profiles provide information on stratification.
¦ Specific conductance profiles can be used as a quality assurance tool to
evaluate the magnitude of dissolved constituents. If specific conductance
increases, corresponding increases of dissolved substances can be ex-
pected.
Both top and bottom
samples should be
collected and analyzed
to ensure that vertical
differences were not
present at the time of
sampling.
Documenting lake
conditions during
periods of strong
stratification can help
identify the magnitude
and potential of internal
loading processes.
Dissolved oxygen data,
which provide the most
basic description of
water quality; often serve
as an indicator of lake
productivity.
6-5
-------
One centrally located
station near the deepest
part of the lake will
normally provide the
least biased
characterization of
conditions.
Adequate support,
analysis, and feedback
are essential to a local
sponsor who will
continue to acquire
long-term data.
¦ Water level information is necessary to make mass balance calculations
and to determine lake volume.
¦ One centrally located station near the deepest part of the lake will nor-
mally provide the least biased characterization of conditions. As described
by Gaugush (1987) and illustrated in Figure 6.1, exceptions are large
reservoirs that often exhibit longitudinal differences. Gaugush also
describes sampling designs appropriate for reservoirs such as the one
shown.
Data that supplement Secchi disk information will provide a significantly better
documentation of lake condition and may also, in some cases, supply information
from which cause-effect relationships can be identified.
Again, it is necessary to emphasize that adequate support, analysis, and feed-
back are essential to a local sponsor who will continue to acquire long-term data.
Although it is very difficult to develop conclusions based upon one or sometimes
even several years of data, most local sponsors will expect some type of interpre-
tive report.
DISSOLVED OXYGEN (mg/L)
TOTAL PHOSPHORUS O-ig/L)
°1
Q>
Q>
SO
X
h-
100
CL
LU
Q
150
"~4
JULY
.
^50^^.
— 25 ^
MAY
S~S=-
JULY
50
too
NOVEMBER
150
0
5
10
15
20
NOVEMBER
15
20
0
5
10
DISTANCE (miles)
Figure 6.1—Longitudinal and vertical distributions of dissolved oxygen and total phosphorus In
DaGray Lako, Arkansas (source; Gaugush, 1987).
6-
6
-------
A Comprehensive Long-term Lake
Monitoring Protocol
Identifying causal factors of lake water quality changes often requires information
on lake chemistry, macrophytes, phytoplankton, zooplankton, and fish com-
munities, as well as watershed conditions. A sample long-term lake monitoring
protocol designed to describe these elements of the lake ecosystem is presented
in Table 6.2.
Table 6.2.—Typical elements of a comprehensive, long-term monitoring program
Parameter Comments
Water Chemistry Protocol at least as intensive as that in Tabie 6,1,
Macrophytes Survey orice or twice during the growing season. Define species present,
distribution, abundance, frequency of occurrence, percent of lake colonized,
maximum depth of growth, and distance of plant tops from the water's
surface. Repeat survey every 3 years.
Zooplankton Do bottom to top vertical tow with 80 u conical plankton net. Collect 3 times
or more per year during the growing season. Identify dominant species and
size range. Minimum diameter of net opening should be 0.2 meters.
Phytoplankton Do six-foot-deep surface composite. Collect 3 or more times per year during
the growing season. Use Lugol's solution for sample preservation. Identify
dominant species.
Fish Community Do boom shocker transects and gill netting once every 2 years, and fyke
netting every 6 years for lakes with pike or walleye. Identify species, size,
length, and catch per unit of effort.
Watershed Identify major land uses once every 5 years. Track major development and
agricultural changes continually. Establish a continually recording flow gag-
ing station with automatic samplers at the major inlets (see Chapter 4).
Rationale for Comprehensive Monitoring
Lakes are often inappropriately characterized when there are no data on their
major ecosystem components. For example, a scientist might conclude that
general lake conditions had improved because Secchi disk readings increased as
a result of improved watershed conditions. At the same time, local citizens and
lake users might perceive that conditions had worsened because dense growths
of rooted plants, which replaced the algae, were severely limiting lake recreation.
Unfortunately, macrophyte-aigae trade-offs are common in many shallow lakes
where improvement in water clarity, for whatever reason, has resulted in dense
stands of rooted plants.
Knowledge of a lake's macrophyte community can also help prevent errors
when defining reasons for lake water quality changes. For example, water clarity
can be degraded by any of several in-lake "improvement" techniques. If macro-
phytes are controlled by harvesting, chemical treatment, use of grass carp, or
dredging, nutrients formerly used by rooted plants can become available for algae
growth. Without knowledge of changes in the lake's rooted plant community, users
might be led to conclude, wrongly, that the resultant decrease in lake clarity was
caused by increased watershed nutrient loadings.
Similarly inaccurate conclusions can be drawn about the causes of an apparent
water quality improvement that occurred because of changes in a lake's fish com-
munity. For example, a lake's improved water clarity could be attributed to the fact
that a sewer had recently been installed around the lake, when in actuality the im-
Identifying causal factors
of lake water quality
changes often requires
information on lake
chemistry, macrophytes,
phytoplankton,
zooplankton, and fish
communities, as well as
watershed conditions.
Macrophyte-algae
trade-offs are common in
many shallow lakes
where improvement in
water clarity, for
whatever reason, has
resulted in dense stands
of rooted plants.
6-7
-------
Because of (he lack of
zooplankton grazing on
the algae... water
clarity is often much
poorer than would be
predicted based upon
nutrient loadings.
The contributions lakes
receive from watersheds
will largely define
long-term, average
conditions.
provement occurred because the fish population had been restructured following
treatment to eradicate undesirable planktivores. Water clarity improvements are
also observed in lakes that have recently experienced massive winter or summer
fish kills. These improvements, which are often temporary in nature, can occur
even though no changes were made in nutrient loading to the lake.
Massive fish kills are not the only events that affect water clarity. Activities such
as recreational fishing, which create more subtle differences in fish community
structures (Shapiro, 1975), can also play important roles in defining a lake's water
clarity. Knowledge of the zooplankton community will often provide insight into the
type of fishery a lake supports. For example, in lakes where the zooplankton
populations are low and size distribution small, often may be a fishery dominated
by small planktivorous fish. Because of the lack of zooplankton grazing on the
algae in such waterbodies, water clarity is often much poorer than would be
predicted based upon nutrient loadings. If the monitoring program shows an in-
creasing dominance of small zooplankton in a lake, then a management strategy
might focus on increasing the number of higher level predator fish by increased
stocking or by reducing recreational fishing,
A lake's fish community can also have a direct effect on its macrophyte com-
munity. In lakes where the fish population is dominated by bottom-feeding com-
mon carp or plant-eating grass carp, rooted plant communities are usually limited
and water clarity is often poor. Poor water clarity is not always a direct result of fish
activities, however, as high nutrient or sediment loadings to a lake may be provid-
ing desirable habitat for these particular species. Only a long-term monitoring pro-
gram that has tracked these ecosystem components can identify the root cause.
Finally, knowledge of a lake's ever-changing watershed conditions is essential.
Ultimately, lakes will respond to the conditions of their watersheds. Algal, macro-
phyte, and fish populations reach an equilibrium that is dependent in part upon the
nutrient and sediment loads received by the lake. Although lake water quality will
always vary from year-to-year or even decade-to-decade, the contributions lakes
receive from watersheds will largely define long-term, average conditions.
A general description of watershed conditions can often be extrapolated from
the watershed inventory that is part of a monitoring program. As described in
Chapter 4, watershed inventories are often the most cost-effective method of ob-
taining data on the watershed's importance to the lake. However, where extensive
watershed improvements are being implemented and evaluated or where
documentation of problems is necessary before control measures can be imple-
mented, more comprehensive data are necessary. The larger, more important lake
tributary streams are often gaged and sampled to quantify sediment, nutrient, and
pesticide loadings to a lake. In lakes where year-to-year watershed land use chan-
ges are not significant, inventory updates every 5 or 10 years are usually ade-
quate.
Only if information is collected on most of the lake's major ecosystem com-
ponents can definitive judgments be made on long-term trends and their causes.
In addition to community-supported monitoring, State agencies responsible for
making overall judgments on lake water quality should consider instituting a long-
term, comprehensive monitoring program for selected lakes.
6-8
-------
Chapter 7
CASE STUDY:
Detection of Trends
and Sampling
Strategy Evaluations
Case Study of the Statistical Evaluation
of the Neuse River, North Carolina,
Total Phosphorus Data Set
The basis for this case
study was a monitoring
design completed for
the Triangle Area Water
Supply Monitoring
Project, Research
Triangle Park, North
Carolina. The study,
which was conducted by
Ken Reckhow, Craig
Stow, James Mitchell,
and Nicolai Denisov,
was completed, in part,
to statistically evaluate
the effectiveness of
alternative monitoring
strategies for detecting
trends in water quality.
Introduction
The Neuse River near Smithfield, North Carolina (Fig, 7.1), drains a 6,192-
square-mile area that includes runoff from the city of Raleigh, two upstream water
supply reservoirs, and an extensive forested area. Starting in 1981, monthly phos-
phorus data were collected from this site by the U.S. Geological Survey. As shown
in Table 7.1 and Figure 7.2, total phosphorus concentrations ranged from a low of
0.13 mg/Lto a high of 1.8 mg/L over the seven-year period.
A statistical evaluation of these data was completed to provide information on
water quality trends and guidance on the intensity of future sampling efforts
needed to detect phosphorus trends. The project sponsor wanted an evaluation of
the ongoing sampling program—should the monitoring program be continued in
the future and, if so, what level of effort was needed—as well as data on improve-
ment or deterioration of water quality.
Although this case study focuses on evaluating tributary stream data, its ap-
proach and techniques can also be used with data obtained directly from a lake.
7-1
-------
Autocorrelation
indicates that each
observation in a time
series is not independent
of other observations.
This means that some of
the information that is
conveyed in the current
observation has already
been conveyed in the
previous observation.
The result is that the
amount of information
actually available is not
reflective of the number
of samples collected;
that is, completely "new"
information has not been
obtained.
Autocorrelation causes
problems with statistical
analysis because, if it is
present, conclusions
regarding the strength of
the analysis can be
incorrect.
Autocorrelation is often
present in lake water
quality data sets where
sampling frequencies are
high. This problem most
commonly occurs when
conservative substances
(such as chloride) are
sampled in lakes that
have long water
residence tunes. In
essence, the same water
is being sampled again
and again.
LAKE MICHIJE.
' ' FALLS LAKE
AT RALEIGH INTAKE
JORDAN LAKE,
fAT ROUTE 64J
HAW RIVER
ATBYNUM
r. NEUSE RIVER
'atsmithfielo
.CAPE FEAR RIVER
\ atlillinston
REGION U, gggSW
LOCATIONS WHBRB HISTORICAL DATA SETS WBRB ANALYZED "
Figure 7.1—Neuse River study area.
Statistical Model Selection
As described in Chapters 4 and 6, major complicating factors in the analysis of
water quality data are the natural background variations that often obscure cul-
turally induced changes.
Seasonal differences are often noted on a yearly basis because of changes in
solar radiation, temperature, and precipitation. Wind irregularities, rainfall events,
and temperature variations also cause seemingly random water quality variations,
but on a smaller scale.
If some of the components causing natural variability can be distinguished and
eliminated from the data set, time trends can be more easily identified. One of the
first steps in analyzing the Neuse River data was identification and separation of
natural variability in the data set from that induced by cultural impacts. If natural
variability can be mathematically described and removed from the water quality
data set, only background variability, or noise, remains to complicate further
trends analysis.
Two different statistical models were considered. Both parametric methods and
nonparametric methods were evaluated for use with the Neuse River data set
The former methods are ones in which a change can be related to particular physi-
cal parameters, e.g., flow, depth, detention time; the latter, more commonly called
distribution-free methods, do not require the assumption that the data be normally
distributed.
7-2
-------
Table 7.1.—Total phosphorus concentrations (mg/L) observed in the Neuse River, near Smithfield, North Carolina
MONTH (DAY COLLECTED SHOWN IN PARENTHESES)
YEAR
JAN
FEB
MAR
APR
MAY
JUNE
JULY
AUG
SEPT
OCT
NOV
DEC
1981
1.80 (19)
0.62 (10)
0.39 (13)
0.45 (8)
1.10 (11)
1.20(17)
0.68 (29)
1.10(4)
0.50 (15)
1.50 (14)
0.53 (4)
0.82 (9)
1982
0.15 (27)
0.23 (15)
0.23(11)
0.32 (14)
0.51 (11)
0.22 (8)
0.37 (8)
0.76 (12)
0.79 (7)
0.92 (7)
0.52(15)
0.25(15)
1983
0.26(13)
0.21 (10)
0.18 (9)
0.18 (19)
0.27 (12)
0.46 (17)
0.90(19)
1.10(22)
1.10(22)
1.40(6)
1.20(9)
0.71 (2)
1984
0.24 (6)
0.29 (15)
0.13 (23)
0.16(20)
0.18(11)
0.47 (22)
0.32 (18)
0.68 (28)
0.52 (17)
0.79 (18)
0.76 (27)
0.45 (11)
1985
0.22 (10)
0.21 (13)
0.49 (21)
0.68 (22)
0.76 (23)
0.74 (24)
0.53 (22)
0.35 (27)
0.85 (11)
0.57 (30)
0.66(18)
0.14 (11)
1986
0.65 (17)
0.39(18)
0.28 (26)
0.68 (10)
0.91 (15)
1.00 (18)
0.64 (28)
1.10(22)
1.30(6)
0.99 (2)
1987
0.17 (27)
0.20 (24)
0.14 (10)
0.31 (9)
0.58 (27)
0.91 (8)
0.74 (30)
0.88 (23)
0.44 (12)
0.89 (9)
1988
0.29 (21)
0.21 (18)
0.36(17)
0,28 (26)
0.35 (27)
-------
A problem inherent to
parametric models
arises from uncertainty
in the applicability of a
given model.
Parametric models must
be always cautiously
applied since there is
evidence that many
water quality
constituents (including
flow) are log-normally
distributed.
Parametric Methods
Parametric approaches in trend detection involve use of two separate models, one
for detecting the trend itself and another to estimate potential errors. Where a
trend is believed to be continuous rather than abrupt, then ordinary least squares
regression techniques are commonly used. A t-statistic is often used for analysis
of trend where a step trend is expected, e.g., a river system from which a was-
tewater discharge has been reduced or eliminated.
If seasonal patterns exist in the data set or if autocorrelation is present, then
more sophisticated techniques such as ARIMAor Box-Jenkins models may be ap-
propriate (Pankratz, 1983). In addition to seasonal trends, other water quality
changes that can be related to identifiable factors, such as a predictable relation-
ship between flow and concentration, should be removed from the data set. This
removal will further reduce background variability from the trend.
A problem inherent to parametric models arises from uncertainty in the ap-
plicability of a given model to a given data set. in the Neuse River evaluation it was
felt that one of the basic assumptions needed for parametric models—that the
data be normally distributed—did not hold for this data set. This uncertainty
prevented the use of a parametric model for this case study. Parametric models
must be always cautiously applied since there is evidence that many water quality
constituents (including flow) are log-normally distributed.
Distribution-free Methods (Nonparametric Methods)
Although distribution-free methods may not be as powerful as parametric
methods, they do not require the assumption that the data be normally distributed.
However, even with these methods there is still a need for independent (un-
autocorrelated) data.
For the Neuse River Analysis, the seasonal Kendall's Tau Test (Hirsch et al.
1982; Hirsch and Slack, 1984) was the method of choice because the data do not
show a normal distribution: they are skewed significantly, and they display a
seasonal cycle. In addition, the seasonal Kendall's Tau Test is not overly sensitive
to extreme values, a situation commonly observed with water quality data. A more
detailed discussion of the use of this test can be found in Gilbert (1987).
Evaluation of the Historical Database
To facilitate evaluation of the Neuse River data, basic statistical analyses were
performed using the software package WQStat II. To simplify this case study, the
actual equations used in the analysis will not be presented here. However, a copy
of WQStat II can be obtained for a nominal fee from Jim Loftis, Agricultural and
Chemical Engineering Department, Colorado State University, Fort Collins, Co
80523; (303) 491-6172.
Data Entry and Preparation
The data collected on the Neuse River were initially imported from an ASCII file
into WQStat II, and a seasonal interval length was specified. In this case, a month-
ly interval was selected for preliminary evaluation because the data had generally
been obtained on a monthly basis; sampling intervals typically ranged between 25
and 35 days. Had the data been collected much outside of this fairly regular time
frame or on a more frequent basis, consideration would have to have been given
to using a quarterly data input format.
7™
4
-------
Summary Statistics
Once entered and prepared, the following summary statistics were obtained from
the data by using WQStat II;
• General statistics
• Skew and kurtosis statistics
• Time series plot
• Seasonal box and whiskers plot, and
• Correlogram.
General summary statistics are shown in Table 7.2.
The statistics for skew (a measure of the degree of the distribution's asym-
metry) and kurtosis (a measure of the degree of the distribution's flatness), which
are shown in Tabie 7.3, provide information on the normality of the data. If either
the skew or kurtosis tests are significant, the data distribution is probably not nor-
mal. In this case, the skew value of 0.876 (calculated by WQStat il) is significant
at the 0.20, 0.10, and 0.02 (80,90, and 98 percent confidence) levels. This shows
that the data are non-normally distributed.
The statistics for skew
and kurtosis provide
information on the
normality of the data.
Table 7.2.—General summary information on the Neuse River data set
CHARACTERISTIC VALUE
Mean 0.586 mg/L
Median 0.500 mg/l
Standard deviation 0.363
Number of data points 85
Table 7.3.—Skew and Kurtosis normality tests for the Neuse River data set
SKEW TEST (SKEW VALUE = 0.876)
CONFIDENCE LEVEL TEST SIGNIFICANCE
98% 0.876 > 0.613 Significant
90% 0.876 > 0.420 Significant
80% 0.876 > 0.324 Significant
KURTOSIS TEST (KURTOSIS VALUE = 3.40)
CONFIDENCE LEVEL TEST SIGNIFICANCE
98% 2.12 < 3.40 < 4.51 Not Significant
90% 2.30 < 3.40 < 3.83 Not Significant
80% 2.39 < 3.40 < 3.55 Not Significant
The total phosphorus concentration time series plot (Fig. 7.2) indicates that
some seasonal patterns exist.
The seasonal box whiskers plot (Figure 7.3) shows, more specifically, the
data's seasonality. Seasonality is considered significant if any of the boxes shown
in the figure do not overlap. Since many of the boxes fail to overlap, and
seasonality is present, consideration should be given to its removal from the data
set prior to trend analysis.
The correlogram (Fig. 7.4) was produced because it can indicate the presence
of seasonal patterns, trends, and/or autocorrelation. In Figure 7.4, the values
along the horizontal axis represent lag values, which are observations N time
periods earlier. In this case, a lag value of 1 represents values obtained one
month previously, and a lag of six represents observations 6 months apart. The
7-5
-------
a
s
z
o
5
cc
I-
z
LXJ
o
z
o
a
m
3
DC
o
x
a.
CO
0
1
a.
2.00
1.60
1.20
0,80 -
0.40
0.00
1981 1982 1983 1984 1985 1986 1987 1988
Figure 7.2—Neuse River total phosphorus concentration time series plot.
(3
2
Z
o
«£
DC
f—
Z
LU
o
z
o
o
to
3
EC
o
X
a.
CO
o
a.
0.0
i r
2/1-3/1 4/1-5/1 6/1-7/1 8/1-9/1 10/1-11/1 12/1-1/1
1/1-2/1 3/1-4/1 5/1-6/1 7/1-8/1 9/1-10/1 11/1-12/1
SAMPLE DATE
Figure 7.3—Seasonal box and whisker plot of the Neuse River data set
lines extending outward from the center of the correlogram represent the autocor-
relation between values at a particular time and those taken N observation periods
(lags) earlier. The parallel horizontal lines represent the values beyond which cor-
relation is significant at the .05 level (95 percent confidence level). For example, in
Figure 7.4, the line at N=1 shows autocorrelation to be significant for observations
made one month apart.
Seasonality in the data is shown in Figure 7.4 by the high positive autocorrela-
tion values at lags 12 and 24 and large negative values for lags 6 and 18, Trend
and autocorrelation both show up as initially significant correlation values that
7-
6
-------
1
-1
6
i
9
i
12
i
15
i
18
i
21
24
Lag (N)
Figure 7.4—Unadjusted Neuse River data set correlogram.
gradually decay to zero. Although it is difficult to distinguish between trends and
autocorrelation, trends generally cause a slower decay in the correlation values.
The correlogram shown in Figure 7.5 was produced following a detrending and
deseasonalization of the data. It indicates that autocorrelation is no longer sig-
nificant beyond the first lag. In this case, the autocorrelation still present at the first
lag was due to a relationship between concentration and flow that could have
been eliminated by using procedures described by Hirsch et al. (1982).
Lag (N)
Figure 7.5—Deseasonalized/detrended Neuse River data set correlogram.
7
-------
Had the correlogram shown the data to be autocorrelated, the assumption of in-
dependent data required for use of the statistical models would have been vio-
lated. Other techniques such as reentering the data in a quarterly format
(averaged or centered) would have been used and the correlogram recomputed.
Hirsch and Slack (1984) describe a statistical method for removal of autocorrela-
tion within the seasonal Kendall's Tau Test in more detail. This correction must
precede final analysis with WQStat II since their procedure has not yet been incor-
porated into the software.
Trend Analysis
Following data preparation and autocorrelation testing, WQStat II was used to run
the Kendall's Tau trend detection tests. In this case it was not necessary to
deseasonalize the data prior to trend detection, since it was handled within the
seasonal Kendall's Tau Test.
As shown in Table 7.4, the seasonal Kendall's Taut statistic used to test for total
phosphorus trend was -1.050. This value was found not to be significant at the 80,
90, or 95 percent confidence levels, which indicated no significant trend. More
specifically, this test shows that one cannot reject the hypothesis that the phos-
phorus concentration trend was zero over the seven-year monitoring period.
Table 7.4.—Kendall Tau Test for trend detection on the Neuse River data set
SEASONAL KENDALL TEST (TEST STATISTIC = -0.619)
CONFIDENCE LEVEL TEST SIGNIFICANCE
95% -1.960 < -1.050 < 1.960 Hot Significant
90% -1.645 < -1.050 < 1.645 Not Significant
80% -1.282 < -1.050 < 1.282 Not Significant
The Kendall Sen slope
estimate of-0.0100
units/year indicates little
change in phosphorus.
...the database
available precluded
detection of any trends
.., where reductions
were less than 30
percent.
The WQStat II program also calculated the seasonal Kendall Sen slope es-
timate. This value is the seasonal equivalent to Sen's nonparametric estimate of
slope. It is the median of all possible slopes generated between the data points.
The value for slope in the Neuse River data set was -0.0100 units/year, indicating
little change in phosphorus over time. Figure 7.6 is a graph of the data over time,
with the calculated phosphorus trend line.
Determination of Future Sampling Effort
The Neuse River analysis found no significant trend in the data. However, the
database available precluded detection of any trends that might have been
caused by phosphorus reduction in the watershed where reductions were less
than 30 percent. Figures 7.7 and 7.8 were prepared using WQStat II to help deter-
mine the number of samples necessary for detecting future trends. For example,
as shown in Figure 7.7, over 120 monthly samples (a 10-year monitoring program)
would be necessary to detect a 23 percent linear decrease in total phosphorus,
such as might be expected with implementation of an extensive nonpoint source
control program, if an error rate of 10 percent is required. If larger error rates are
acceptable, a less intensive sampling effort would be adequate.
If a step decrease in phosphorus is expected, such as that which could occur
from the upgrading of a sewage treatment plant, then, as can be seen from Figure
7.8, only 64 monthly samples (a seven- to eight-year monitoring program) would
be needed to detect the same 23 percent change in trend.
7-i
8
-------
a
z
o
U!
o
z
o
o
m
3
oc
O
x
Q.
CO
o
2.00
1.60
1,20
0.80
0.40
0.00
1981 1982 1983 1984 1985 1986 1987 1988
Figure 7.6—Neuse River total phosphorus concentration time series plot with calculated trend
line.
120
CO
^ 100
Q.
s
>
—I
I
80
IT
UJ
m
60
40
20
Error=0.10
Error=0.20
Error=0.30
23 %
(30)
28 %
(40)
33 <
(50)
PERCENT DECREASE (INCREASE) OF EXPECTED CONCENTRATION
Figure 7.7—Sample size versus the magnitude of a linear trend for total phosphorus concentra-
tion for the Neuse River.
7-9
-------
120
Error=0.10
Mean = 0.57 mg/i J
Std dev = 0.36 mg/l ]
uj 100
CL
80
60
Error=0.20
u.
40
Error=0.30
20
17%
23%
29%
33%
(20) (30) (40) (50)
PERCENT DECREASE (INCREASE) OF EXPECTED CONCENTRATION
Figure 7.8—Sample size versus the magnitude of a step trend for total phosphorus concentra-
tion of the Neuse River.
-10
-------
WSL
References
Chapter 1
U.S. Environmental Protection Agency. 1988. The Lake and Reservoir Restoration Guidance Manual.
EPA440/5-88-02. Washington, DC.
U.S. Environmental Protection Agency. 1980. Interim Guidelines and Specifications for Implementing
Quality Assurance Requirements for EPA Contracts. EPAQAMS-005/80. Washington, DC.
. 1981. Clean Lakes Program Guidance Manual. EPA440/5-81-003. Washington, DC.
Buchanan, T.J. and W.P. Somers, 1968. Stage measurements at gaging stations. Techniques of Water
Resources Investigations, Book 3, Chap. A7, U.S. Geo. Surv., Washington, DC.
Fassett, N.C. 1969. A Manual of Aquatic Plants. Univ. Wisconsin Press, Madison.
Godfrey, R.K. and J.W. Wooten. 1979. Aquatic and Wetland Plants of Southeastern United States:
Monocotyledons. Univ. Georgia Press, Athens.
Haveren, B.Q. 1986. Water Resource Measurements. A Handbook for Hydrologists and Engineers, Am.
Water Works Ass., Washington, DC.
Hillman, D.C., J.F. Potter, and S.J. Simon. 1986. National Surface Water Survey, Eastern Lake Survey.
Phase I, Analytical Methods Manual. EPA-600/4-86-009. U.S. Environ. Prot, Agency, Las Vegas,
NV.
Holtan, H.,N, Minshali, and L. Harrold. 1968. Field Manual for Research in Agricultural Hydrology, Agric.
Handbook 224. U.S. Dep. Agric., Washington, DC.
Kennedy, R.H., ed. 1985. Limnological Studies at Eau Galle Lake, Wisconsin. Rep. 1: Introduction and
Water Quality Monitoring Studies. Tech. Rep. E-85-2, U.S. Army Corps. Eng. Waterways Exp. Sta.,
Vicksburg, MS.
Muencher, W.C. 1964. Aquatic Plants of the United States. Cornell Univ. Press, Ithaca, NY.
Phillips, E. A, 1959. Methods of Vegetation Study. Henry Holt & Co., New York.
Porterfleld, G. 1972. Computation of fluvial sediment discharge. Techniques of Water Resources Inves-
tigations, Book 3, Chap. C3. U.S. Geo. Surv., Washington, DC.
Rantz, S.E. 1982. Measurement and Computation of Streamflow, Vol. 1: Stage and Discharge; Vol. 2:
Computation of Discharge. Water Supply Pap, No. 2175. U.S. Geo. Surv., Washington, DC,
Standard Methods for the Examination of Water and Wastewater, 1989. Joint Editorial Board, Am.
Public Health Ass., Am. Water Works Ass., and Water Poliut, Control Fed., 17th ed. Washington,
DC.
Stauffer, R.E. 1981. Sampling Strategies for Estimating the Magnitude and Importance of Internal Phos-
phorus Supplies in Lakes. EPA-600/3-81 -015. U.S. Environ. Prot. Agency, Washington, DC.
U.S. Bureau of Reclamation. 1975. Water Measurement Manual. Water Resour. Tech. Pub. U.S. Dep.
Inter., Washington, DC.
Chapter 2
Chapter 3
8-1
-------
U.S. Environmental Protection Agency. 1982. Handbook for Sampling and Sample Preservation of
Water and Wastewater. EPA-600/4-82-029. Washington, DC.
. 1983. Methods for Chemical Analysis of Water and Wastes. EPA-600/4-79-020, Washington,
DC.
U.S. Geological Survey. 1977a. Methods for collection and analyses of aquatic biological and
microbiological samples. Techniques of Water Resources Investigations, Book 5, Chap. A4.
Washington, DC.
. 1977b. National Handbook of Recommended Methods for Water Data Acquisition. Off. Water
Data Coord., U.S. Geo. Surv., Washington, DC.
Vanoni, V.A. ed. 1975. Sedimentation Engineering. Manual and Report on Engineering Practice, No. 54.
New York.
Vollenweider, R.A. 1969. A Manual on Methods for Measuring Primary Production in Aquatic Environ-
ments. Int. Biolog. Program, Handbook No. 12. Bfackwell Scientific Pub!., Oxford/Edinburgh.
Voss, E.G. 1972. Michigan Flora. Part I: Gymnosperms and Monocots. Cranbrook Inst. Sci, Publ. 55.
Bloomfieid Hills, Ml.
Chapter 4
Allum, M.O., R.E. Glessner, and J.H. Gakstatter. 1977. An Evaluation of the National Eutrophication Sur-
vey Data. Work. Pap. No. 900. U.S. Environ. Prot. Agency, Off. Res. Dev., Washington, DC.
Baker, D.B. 1988. Sediment, Nutrient and Pesticide Transport in Selected Lower Great Lakes
Tributaries. EPA-905/4-88-001. Great Lakes Nati. Program Off., U.S. Environ. Prot Agency,
Chicago, iL.
Beaulac, M.N. 1980. Sampling Design and Nutrient Export Coefficients: An Examination of Variability
Within Differing Land Uses. M.S. Thesis. Dep. Resour. Dev. Michigan State Univ., East Lansing.
Bant, P.C. 1971. Influence of surface glacial deposits on streamflow characteristics. Open File Rep. U.S.
Geo. Surv., Michigan Water Resour. Div., Lansing.
Buchanan, T.J. and W.P. Somers. 1969. Discharge measurements at gaging stations. Techniques of
Water Resources Investigations, Book 3, Chap. A8. U.S. Geo. Surv., Washington, DC.
Cooke, G.D., E.B. Welch, S.A. Peterson, and P.R. Newroth. 1986. Lake and Reservoir Restoration. But-
terworth Publ., Boston, MA.
Gaugush, R.F. 1987. Sampling design for reservoir water quality investigations. Instr. Rep. E-87-1, U.S.
Army Eng. Waterw. Exp. Sta„ Vicksburg, MS.
Gebert, W.A., D.J. Graczyk, and W.R. Krug. 1987. Average Annual Runoff in the United States, 1951-
1980. Hydrologic Atlas HA-710. U.S. Geo. Surv., Washington, DC.
Hirsch, R.M., J.R. Slack, and R.A. Smith, 1982. Techniques of trend analyses for monthly water quality
data. Water Resour. Res. 18:107-21.
Holmstrom, B.K., P.A. Kammerer, and R.M. Erickson. 1988. Water Resources Data—Wisconsin Water
Year 1988. Rep. No. USGS/WRD/HD-89/264. U.S. Geo. Surv., Madison, Wl.
Lennox, L.J. 1984. Sediment-water exchange in Lough Ennell with particular reference to phosphorus.
Water Res. 18:1483-85.
Malueg, K.W. 1975. A six-year water, phosphorus, and nitrogen budget for Shagawa Lake, Minnesota.
J, Environ. Qual., 4(2):236-42.
Minns, C.K. and M.G. Johnson. 1979. Temporal variation in the nutrient export of rivers draining in the
Bay of Quinte, Ontario, Canada. Water Res. Bull. 15:1061-72.
National Oceanic and Atmospheric Administration. 1982. Evaporation Atias for the Contiguous 48
United States. Natl, Oceanic Atmos, Admin. Tech, Rep. NWS 33. U.S. Dep. Commerce,
Washington, DC.
North Carolina Agricultural Extension Service. 1981. A Conceptual Framework for Assessing Agricul-
tural Nonpoint Source Projects. N.C. State Univ., Raleigh.
Persson, L.A., J.O. Peterson, and F.W. Madison. 1983. Evaluation of sediment and phosphorus
management practices in the White Clay Lake watershed. Water Res. Bull. 19:753-82.
Porterfield, G. 1972. Computation of fluvial-sediment discharge. Techniques of Water Resources Inves-
tigations, Book3, Chap. C3. U.S. Geo. Surv., Washington, DC.
Reckhow, K.H., M.N. Beaulac, and J.T. Simpson, 1980. Modelling Phosphorus Loading and Lake
Response under Uncertainty: A Manual and Compilation of Export Coefficients. EPA-44Q/5-80-011.
U.S. Environ. Prot. Agency, Washington, DC.
Reckhow, K.H., C. Stow, J. Mitchell, and N. Denisov. 1989. Design of the Triangle Area Water Supply
Monitoring Program. Triangle J. Counc. Gov. Research Park, NC.
Spooner, J„ R.P, Maas, S.A. Dressing, M.D, Smolen, and F.J. Humenik. 1985. Appropriate designs for
documenting water quality improvements from agricultural NPS control programs. Pages 30-34 in
Perspectives on Nonpoint Source Pollution. EPA-440/5-85-001. U.S. Environ. Prot. Agency,
Washington, DC.
U.S. Environmental Protection Agency. 1983. Technical Guidance Manual for Performing Waste Load
Allocations. Book IV: Lakes and Impoundments. Chap. 2; Eutrophication. Off. Water Reg. Stand-
ards, Monitor. Data Support Div., Monitor, Branch, Washington, DC.
2
-------
Verhoff, H., S.M. Yaksich, and D.A, Melfi. 1980. River nutrient and chemical transport estimation. J. Env.
Eng. Div„ ASCE. 196(3):591 -608.
Vollenweider, R.A. 1976. Advances in defining critical loading levels for phosphorus in lake eutrophica-
tlon. Mem. 1st. Itai. Idrobiol, 33:53-83.
Walker, W.W. 1977. Some Analytical Methods Applied to Lake Water Quality Problems. Ph.D. diss., Har-
vard Univ., Cambridge, MA.
. 1987. Empirical methods of predicting eutrophication in impoundments. Applications Manual,
EWQOS Program, U.S. Army Corps of Eng., Vicksburg, MS.
Young, R.A., M.A. Otterby, and A. Roos. 1981. An Evaluation System to Rate Feedlot Pollution Poten-
tial. USDA-ARM-NC-17. Agric. Res. Serv,, U.S. Dep. Agric., Washington, DC.
Chapter 5
Benndorf, J. and U, Miersch, In press. Phosphorus loading and efficiency of biomanipulation. Verh. In-
ternal. Verein. Limnol.
Brocksen, R.W. and P.W. Ernler, Jr. 1988. Living Lakes: an aquatic liming and fish restoration
demonstration program. Water Air Soil Pollut. 41 ;85-94.
Brown, D.J A 1988. The Loch Fleet and other catchment liming programs. Water Air Soil Pollut 41:409-
16.
Bukaveckas, P.A. 1989. Effects of calcite treatment on primary producers in acidified Adirondack lakes.
II. Short-term response by phytoplankton communities. Can. J. Fish. Aquat. Sci. 46:352-59.
Canfield, D.E. Jr., K.A. Langeland, S.B. Unda, and W.T. Haller. 1985. Relations between water
transparency and maximum depth of macrophyte colonization in lakes. J. Aquat. Plant Manage.
23:25-28.
Carpenter, S.R., J.F. Kitchell, arid J.R. Hodgson. 1985. Cascading trophic interactions and lake produc-
tivity. Bioscience 35:634-39.
Cooke, G.D., E.B, Welch, S.A. Peterson, and P.R. Newroth. 1986. Lake and Reservoir Restoration. But-
terworth Publ., Boston, MA.
DePinto, J.V., R.D. Scheffe, W.G. Booty, and T.C. Young. 1989. Predicting reacidificatfon of calcite
treated acid lakes. Can. J. Fish. Aquat, Sci. 46:323-32,
Driscoll, C.T., W.A. Ayling, J.F, Fordham, and L.O. Oliver. 1989a. The chemical response of CaCOs-
treated lakes to reacidification. Can. J. Fish. Aquat Sci. 46:258-67.
. 1989b. Short term changes in the trace metal chemistry of two acidic lakes following calcium
carbonate treatment. Can. J. Fish. Aquat. Sci, 46:249-57.
Fordham, G.F. and C.T. Driscoll. 1989. Short-term changes in the acid/base chemistry of two acidic
lakes following calcium carbonate treatment. Can. J. Fish. Aquat. Sci. 46:308-14.
Freeman, R.A. and W.H. Everbart, 1971. Toxicity of alumnimum hydroxide complexes in neutral and
basic media to rainbow trout. Trans. Am. Fish. Soc. 100:644-58.
Gloss, S.P., C.L. Schofield, R.L Spateholts, and B.A. Plonski. 1989, Survival, growth reproduction and
diet of brook trout (Salvelinus fontinalis) in lakes after liming. Can. J. Fish. Aquat. Sci. 46:277-86.
Healy, F.P. and L.L. Hendzel. 1976. Physiological changes during the course of blooms of
Aphanizomenon flos-aqua. J. Fish. Res. Board Can. 33:36-41.
Hrbacek, J., M. Dvorakova, V. Korinek, and L. Prochazkova. 1961, Demonstration of the effect offish
stock on the species composition and the intensity of metabolism of the whole plankton association.
Verh. Internat. Verein. Limnol. 14:192-95.
Knauer, D.R. and P.J. Garrison. 1980. A comparison of two alum treated lakes in Wisconsin. Pages 412-
16 in Restoration of Lakes and Inland Waters. EPA 440/5-81-010. U.S. Environ. Prot, Agency,
Washington, DC.
Mericer, P. and J. Perrett. 1949. Aeration station of Lake Bret. Monatsbull. Schweiz. Ver. Gas. Wasser-
facht 29:25-30.
Moore, L, and K. Thornton, eds, 1988, The Lake and Reservoir Restoration Guidance Manual. EPA-
440/5-88-002. U.S. Environ. Prot. Agency, Washington, DC.
Mortimer, C.H. 1942. The exchange of dissolved substances between mud and water in lakes. J. Ecol.
30:147-201.
New Hampshire Water Supply and Pollution Control Commission. 1971. Algae Control by Mixing. Con-
cord.
Nurnberg, G.K. 1987. Hypolimnetic withdrawal as a lake restoration technique. J. Environ. Eng.
113:1006-17.
Oglesby, R.T. 1969, Effects of controlled nutrients dilution on a eutrophic lake. Pages 747-57 in Advan-
ces in Water Pollution Research, Pergamon Press, New York.
Olem, H. 1989. Liming Acidic Surface Waters. Pub. Rev. Draft: Rep. 15. Natl. Acid Precip. Assess.
Progr., Washington, DC.
Pastorok, R.A., T.C. Ginn, and M.W. Lorenzen. 1980. Review of aeration/circulation for lake manage-
ment. Pages 124-33 in Restoration of Lakes and Inland Waters. EPA440/5-81-010. U.S. Environ.
Prot. Agency, Washington, DC.
8
-------
Porcella, D.B. 1989. Mitigation of acidic conditions in lakes: an overview of an ecosystem perturbation
experiment. Can. J. Fish. Aquat. Sci, 46:246-48.
Porter, K.G. 1977. The plant-animal interface in freshwater ecosystems. Am. Sci. 65:159-70.
Prepas, E.E., D.J. Webb, and C.L.K. Robinson. 1989. Injection of Oxygen into the North Basin of Amisk
Lake. Final rep. submitted to Alberta Environ., Edmonton, Alberta.
Ripl, W. 1976, Biochemical oxidation of polluted lake sediment with nitrate—a new restoration method.
Amblo 5:132-35
. 1980. Lake restoration methods developed and used in Sweden. Pages 495-500 in Restoration
of Lakes and Inland Waters. EPA440/5-81-010. U.S. Environ. Prot. Agency, Washington, DC.
Roberts, D.A., and C.W. Boylen. 1989. Effects of liming on theepipelic algal community of Woods Lake,
New York. Can, J. Fish. Aquat. Sci. 46:287-94.
Rosseland, B.O. and A. Hindar. 1988. Liming of lakes, rivers, and catchments in Norway. Water Air Soil
Pollut. 41:165-88.
Schaffner, W.R. 1989. The effects of neutralization and addition of brook trout (Salvelinus fontinalis) on
the limnetic zooplankton communities of two acidic lakes. Can. J. Fish. Aquat. Sci. 46:295-305.
Schofield, C.L., S.P. Gloss, B.A. Plonski, and R. Spateholts, 1989. Production and growth efficiency of
brook trout (Salvelinus fontinalis) In relation to food supply and water quality changes in two Adiron-
dack Mountain lakes (New York) following liming. Can, J. Fish. Aquat. Sci. 46:333-41,
Shapiro, J. 1978, The need for more biology in lake restoration. Pages 161-68 in Lake Restoration.
Proceed. Natl. Conf. EPA440/5-79-001. U.S. Environ. Prot. Agency, Washington, DC.
Shapiro, J. et al. 1982. Experiments and Experiences in Biomanipulation. EPA-600/3-82-096. U.S. En-
viron. Prot. Agency, Washington, DC.
Smith, S.A., D.R. Knauer, and T.L. Wirth. 1975. Aeration as a Lake Management Technique. Tech. Bull.
87. Wis. Dep. Nat. Resour, Madison.
Vollenwelder, R.A. 1976. Advances in defining critical loading levels for phosphrous in lake eutrophica-
tion. Mem. 1st. Ital. Ideobiol. 33:53-83.
Walker, W.W., Jr., C.E. Westerberg, D.J. Schuler, and J.A. Bode. 1989. Design and evaluation of
eutrophication control measures for the St. Paul water supply. Lake Reserv. Manage. 5:71 -84.
Wedepohl, R.E., A.T. Hanson, and J.E. Szewczyk, 1983. Lake deepening using in-situ techniques.
Pages 41 -45 in Lake Restoration, Protection, and Management. Proe. Natl. Conf, EPA 440/5-83-
001. U.S. Environ. Prot. Agency, Washington, DC.
Welch, E.B. 1979. Lake restoration by dilution. Pages 133-9 intake Restoration. Proc. Natl. Conf. EPA
440/5-79-001. U.S. Environ. Prot. Agency, Washington, DC.
Welch, E.B. and C.R. Patmont. 1980. Lake restoration by dilution: Moses Lake. Water Res. 14:1317-25.
Wiley, M.J., P.P. Tazik, and S.T. Sobaski. 1987. Controlling aquatic vegetation with triploid grass carp.
Circ. 57, 111. Nat. History Surv., Champaign.
Chapter 6
Gaugush, R. F. 1987. Sampling design for reservoir water quality investigations. Instr. Rep. E-87-1 .U.S.
Army Corps. Eng. Waterw. Exp. Sta„ Vicksburg, MS.
Montgomery, R. H. and K. H. Reckhow 1984. Techniques for detecting trends in lake water quality. Water
Resour. Bull. 20(1 ):43-51.
Shapiro, J., ed. 1975. Biomanipulation: an ecosystem approach to lake restoration. Pages 85-96 in
Symp. on Water Quality Management through Biological Control, U.S. Environ. Prot. Agency and
Univ. Florida, Gainesville.
Smeltzer, E., V. Garrison, and W. W. Walker, Jr. 1989, Eleven years of lake eutrophication monitoring in
Vermont a critical evaluation. In Proc. Natl. Conf. Enhancing States' Lake Management Programs,
N, Am. Lake Manage. Soc., Washington, DC.
Stauffer, R. E. 1988. Sampling strategies and associated errors in estimating epilimnetic chlorophyll In
eutrophic lakes. Water Resour. Res. 24(9):1459-69,
Travis, M. J. 1988. Nitrogen contamination of groundwater from barnyards in the central sand plain
aquifer of Wisconsin. M.S. thesis. College Nat. Resour. Univ. Wisconsin-Stevens Point.
Wolman, G. M. 1971. The nation's rivers. Science 174:905-18.
Chapter 7
Gilbert, R.0.1987. Statistical Methods for Environmental Pollution Monitoring. Van Nostrand Reinhold,
New York.
Hlrseh, R.M. and J.R. Slack. 1984. Anonparametric trend test for seasonal data with serial dependence.
Water Resour. Res. 20:727-32.
Hlrseh, R.M., J.R. Slack, and R.A, Smith. 1982. Techniques of trend analysis for monthly water quality
data. Water Resour. Res. 18:107-21.
Pankratz, A. 1983. Forecasting with Univariate Box-Jenkins Models. Wiley, New York.
-------
endix
Cooperative Agreements for Protecting
and Restoring Publicly Owned
Freshwater Lakes
U.S. Environmental Protection Agency
9-1
-------
Environmental Protection Agency
§35.1603
Subparts F-G—[Reserved]
Subpart H—Cooperative Agreements
for Protecting and Restoring Pub-
licly Owned Freshwater Lakes
Authority: Sees. 314 and 501, Clean
Water Act (86 Stat. 816; 33 U.S.C. 1251 et
seq.).
Source: 45 PR 7792, Feb. 5. 1980, unless
otherwise noted.
§ 35.1600 Purpose.
This subpart supplements the EPA
general grant regulations and proce-
dures (Part 30 of this chapter) and es-
tablishes policies and procedures for
cooperative agreements to assist
States in carrying out approved meth-
ods and procedures for restoration (in-
cluding protection against degrada-
tion) of publicly owned freshwater
lakes.
§ 35.1603 Summary of clean lakes assist-
ance program.
(a) Under section 314 of the Clean
Water Act, EPA may provide financial
assistance to States to implement
methods and procedures to protect
and restore publicly owned freshwater
lakes. Although cooperative agree-
ments may be awarded only to States,
these regulations allow States,
through substate agreements, to dele-
gate some or all of the required work
to substate agencies.
(b) Only projects that deal with pub-
licly owned freshwater lakes are eligi-
ble for assistance. The State must
have assigned a priority to restore the
lake, and the State must certify that
the lake project is consistent with the
State Water Quality Management
Plan (§ 35.1521) developed under the
State/EPA Agreement. The State/
EPA Agreement is a mechanism for
EPA Regional Administrators and
States to coordinate a variety of pro-
grams under the Clean Water Act, the
Resource Conservation and Recovery
Act, the Safe Drinking Water Act and
other laws administered by EPA.
(c) These regulations provide for
Phase 1 and 2 cooperative agreements.
The purpose of a Phase 1 cooperative
agreement is to allow a State to con-
duct a diagnostic-feasibility study to
determine a lake's quality, evaluate
possible solutions to existing pollution
problems, and recommend a feasible
program to restore or preserve the
quality of the lake. A Phase 2 coopera-
tive agreement is to be used for imple-
menting recommended methods and
procedures for controlling pollution
entering the lake and restoring the
lake. EPA award of Phase 1 assistance
does not obligate EPA to award Phase
2 assistance for that project. Addition-
ally, a Phase 1 award is not a prerequi-
site for receiving a Phase 2 award.
However, a Phase 2 application for a
proposed project that was not evaluat-
ed under a Phase 1 project shall con-
tain the information required by Ap-
pendix A.
(d) EPA will evaluate all applications
in accordance with the application
review criteria of § 35.1640-1. The
review criteria include technical feasi-
bility, public benefit, reasonableness
of proposed costs, environmental
impact, and the State's priority rank-
ing of the lake project.
(e) Before awarding funding assist-
ance, the Regional Administrator shall
determine that pollution control meas-
ures in the lake watershed authorized
by section 201, included in an ap-
proved 208 plan, or required by section
402 of the Act are completed or are
being implemented according to a
schedule that is included in an ap-
555
-------
§35.1605
40 CFR Ch. I (7-1.88 Edition)
proved plan or discharge permit. Clean
lakes funds may not be used to control
the discharge of pollutants from a
point source where the cause of pollu-
tion can be alleviated through a mu-
nicipal or industrial permit under sec-
tion 402 of the Act or through the
planning and construction of
wastewater treatment facilities under
section 201 of the Act.
§35.1605 Definitions.
The terms used in this subpart have
the meanings defined in section 502 of
the Act. In addition, the following
terms shall have the meaning set
forth below.
§ 35.1605-1 The Act.
The Clean Water Act, as amended
(33 U.S.C. 1251 et seq.).
§ 35.1605-2 Freshwater lake.
Any inland pond, reservoir, im-
poundment, or other similar body of
water that has recreational value, that
exhibits no oceanic and tidal influ-
ences, and that has a total dissolved
solids concentration of less than 1 per-
cent.
§35.1605-3 Publicly owned freshwater
lake.
A freshwater lake that offers public
access to the lake through publicly
owned contiguous land so that any
person has the same opportunity to
enjoy nonconsumptive privileges and
benefits of the lake 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 this lake or other publicly owned
freshwater lakes in the State, or for
improving the quality of these lakes.
§ 35.1605-4 Nonpoint source.
Pollution sources which generally
are not controlled by establishing ef-
fluent limitations under sections 301,
302, and 402 of the Act. Nonpoint
source pollutants are not traceable to
a discrete identifiable origin, but gen-
erally result from land runoff, precipi-
tation, drainage, or seepage.
§ 35.1605-5 Eutrophic lake.
A lake that exhibits any of the fol-
lowing characteristics: (a) Excessive
biomass accumulations of primary pro-
ducers; (b) rapid organic and/or inor-
ganic sedimentation and shallowing;
or (c) seasonal and/or diurnal dis-
solved oxygen deficiencies that may
cause obnoxious odors, fish kills, or a
shift in the composition of aquatic
fauna to less desirable forms.
§ 35.1605-6 Trophic condition.
A relative description of a lake's bio-
logical productivity based on the avail-
ability of plant nutrients. The range
of trophic conditions is characterized
by the terms of oligotrophic for the
least biologically productive, to eutro-
phic for the most biologically produc-
tive.
§ 35.1605-7 Desalinization.
Any mechanical procedure or proc-
ess where some or all of the salt is re-
moved from lake water and the fresh-
water portion is returned to the lake.
§ 35.1605-8 Diagnostic-feasibility study.
A two-part study to determine a
lake's current condition and to develop
possible methods for lake restoration
and protection.
(a) The diagnostic portion of the
study includes gathering information
and data to determine the limnologi-
cal, morphological, demographic,
socio-economic, and other pertinent
characteristics of the lake and its wa-
tershed. This Information will provide
recipients an understanding of the
quality of the lake, specifying the lo-
cation and loading characteristics of
significant sources polluting the lake.
(b) The feasibility portion of the
study includes: (1) Analyzing the diag-
nostic information to define methods
and procedures for controlling the
sources of pollution; (2) determining
the most energy and cost efficient pro-
cedures to improve the quality of the
lake for maximum public benefit; (3)
developing a technical plan and mile-
stone schedule for implementing pol-
lution control measures and in-lake
restoration procedures; and (4) if nec-
essary, conducting pilot scale evalua-
tions.
556
9-3
-------
Environmental Protection Agency
§ 35.1610 Eligibility.
EPA shall award cooperative agree-
ments for restoring publicly owned
freshwater lakes only to the State
agency designated by the State's Chief
Executive. The award will be for
projects which meet the requirements
of this subchapter.
§ 35.1613 Distribution of funds.
(a) For each fiscal year EPA will
notify each Regional Administrator of
the amount of funds targeted for each
Region through annual clean lakes
program guidance. To assure an equi-
table distribution of funds the target-
ed amounts will be based on the clean
lakes program which States identify in
their State WQM work programs.
Cb) EPA may set aside up to twenty
percent of the annual appropriations
for Phase 1 projects.
§ 35.1615 Substate agreements.
States may make financial assistance
available to substate agencies by
means of a written interagency agree-
ment transferring project funds from
the State to those agencies. The agree-
ment shall be developed, administered
and approved in accordance with the
provisions of 40 CPR 33.240 (Intergov-
ernmental agreements). A State may
enter into an agreement with a sub-
state agency to perform all or a por-
tion of the work under a clean lakes
cooperative agreement. Recipients
shall submit copies of all interagency
agreements to the Regional Adminis-
trator. If the sum involved exceeds
$100,000, the agreement shall be ap-
proved by the Regional Administrator
before funds are released by the State
to the substate agency. The agreement
shall incorporate by reference the pro-
visions of this subchapter. The agree-
ment shall specify outputs, milestone
schedule, and the budget required to
perform the associated work in the
same manner as the cooperative agree-
ment between the State and EPA.
§ 35.1620 Application requirements.
(a) EPA will process applications in
accordance with Subpart B of Part 30
of this subchapter. Applicants for as-
sistance under the clean lakes pro-
gram shall submit EPA form 5700-33
(original with signature and two
§ 35.1620-2
copies) to the appropriate EPA Re-
gional Office (see 40 CFR 30.130).
(b) Before applying for assistance,
applicants should contact the appro-
priate Regional Administrator to de-
termine EPA's current funding capa-
bility.
§ 35.1620-1 Types of assistance.
EPA will provide assistance in two
phases in the clean lakes program.
(a) Phase 1—Diagnostic-feasibility
studies. Phase 1 awards of up to
$100,000 per award (requiring a 30 per-
cent non-Federal share) are available
to support diagnostic-feasibility stud-
ies (see Appendix A).
(b) Phase 2—Implementation. Phase
2 awards (requiring a 50 percent non-
Federal share) are available to support
the implementation of pollution con-
trol and/or in-lake restoration meth-
ods and procedures including final en-
gineering design.
§ 35.1620-2 Contents of applications.
(a) All applications shall contain a
written State certification that the
project is consistent with State Water
Quality Management work program
(see § 35.1513 of this subchapter) and
the State Comprehensive Outdoor
Recreation Plan (if completed). Addi-
tionally, the State shall indicate the
priority ranking for the particular
project (see § 35.1620-5).
(b) Phase 1 applications shall con-
tain: (1)A narrative statement describ-
ing the specific procedures that will be
used by the recipient to conduct the
diagnostic-feasibility study including a
description of the public participation
to be involved (see § 25.11 of this chap-
ter);
(2) A milestone schedule;
(3) An itemized cost estimate includ-
ing a justification for these costs;
(4) A written certification from the
appropriate areawide or State 208
planning agency that the proposed
work will not duplicate work complet-
ed under any 208 planning grant, and
that the applicant is proposing to use
any applicable approved 208 planning
in the clean lakes project design; and
(5) For each lake being investigated,
the information under paragraph
(5)(i) of this paragraph (b) and, when
9-4
557
-------
§35.1620-3
40 CFR Ch. I (7-1-88 Edition)
available, the information under para-
graph <5)(ii) of this paragraph (b).
(i) Mandatory information.
(A) The legal name of the lake, res-
ervoir, or pond.
(B) The location of the lake within
the State, including the latitude and
longitude, in degrees, minutes, and
seconds of the approximate center of
the lake.
(C) A description of the physical
characteristics of the lake, including
its maximum depth (in meters); its
mean depth (in meters); its surface
area (in hectares); its volume (in cubic
meters); the presence or absence of
stratified conditions; and major hydro-
logic inflows and outflows.
(D) A summary of available chemical
and biological data demonstrating the
past trends and current water quality
of the lake.
(E) A description of the type and
amount of public access to the lake,
and the public benefits that would be
derived by implementing pollution
control and lake restoration proce-
dures.
(P) A description of any recreational
uses of the lake that are impaired due
to degraded water quality. Indicate
the cause of the impairment, such as
algae, vascular aquatic plants, sedi-
ments, or other pollutants.
(G) A description of the local inter-
ests and fiscal resources committed to
restoring the lake.
(H) A description of the proposed
monitoring program to provide the in-
formation required in Appendix A
paragraph (a)(10) of this section.
(ii) Discretionary information.
States should submit this information
when available to assist EPA in review-
ing the application.
(A) A description of the lake water-
shed in terms of size, land use (list
each major land use classification as a
percentage of the whole), and the gen-
eral topography, including major soil
types.
(B) An identification of the major
point source pollution discharges in
the watershed. If the sources are cur-
rently controlled under the National
Pollutant Discharge Elimination
System (NPDES), include the permit
numbers.
(C) An estimate of the percent con-
tribution of total nutrient and sedi-
ment loading to the lake by the identi-
fied point sources.
(D) An indication of the major non-
point sources in the watershed. If the
sources are being controlled describe
the control practice(s), including best
land management practices.
(E) An indication of the lake restora-
tion measures anticipated, including
watershed management, and a projec-
tion of the net improvement in water
quality.
(F) A statement of known or antici-
pated adverse environmental impacts
resulting from lake restoration.
(c) Phase 2 applications shall in-
clude: (1) The information specified in
Appendix A in a diagnostic/feasibility
study or its equivalent; (2) certifica-
tion by the appropriate areawide or
State 208 planning agencies that the
proposed Phase 2 lake restoration pro-
posal is consistent with any approved
208 planning; and (3) copies of all
issued permits or permit applications
(including a summary of the status of
applications) that are required for the
discharge of dredged or fill material
under section 404 of the Act.
§ 35.1620-3 Environmental evaluation.
Phase 2 applicants shall submit an
evaluation of the environmental im-
pacts of the proposed project in ac-
cordance with the requirements in Ap-
pendix A of this regulation,
§ 35.1620-4 Public participation.
(a) General. (1) In accordance with
this part and Part 25 of this chapter,
the applicant shall provide for, en-
courage, and assist public participa-
tion in developing a proposed lake res-
toration project.
(2) Public consultation may be co-
ordinated with related activities to en-
hance the economy, the effectiveness,
and the timeliness of the effort, or to
enhance the clarity of the issue. This
procedure shall not discourage the
widest possible participation by the
public.
(b) Phase 1. (1) Phase 1 recipients
shall solicit public comment in devel-
oping, evaluating, and selecting alter-
natives; in assessing potential adverse
558
-------
Environmental Protection Agency
environmental impacts; and in identi-
fying measures to mitigate any ad-
verse impacts that were identified.
The recipient shall provide informa-
tion relevant to these decisions, in fact
sheet or summary form, and distribute
them to the public at least 30 days
before selecting a proposed method of
lake restoration. Recipients shall hold
a formal or informal meeting with the
public after all pertinent information
is distributed, but before a lake resto-
ration method is selected. If there is
significant public interest in the coop-
erative agreement activity, an advisory
group to study the process shall be
formed in accordance with the re-
quirements of § 25.3(d)(4) of this chap-
ter.
(2) A formal public hearing shall be
held if the Phase 1 recipient selects a
lake restoration method that involves
major construction, dredging, or sig-
nificant modifications to the environ-
ment, or if the recipient or the Re-
gional Administrator determines that
a hearing would be beneficial.
(c) Phase 2. (1)A summary of the re-
cipient's response to all public com-
ments, along with copies of any writ-
ten comments, shall be prepared and
submitted to EPA with a Phase 2 ap-
plication.
(2) Where a proposed project has
not been studied under a Phase 1 co-
operative agreement, the applicant for
Phase 2 assistance shall provide an op-
portunity for public consultation with
adequate and timely notices before
submitting an application to EPA. The
public shall be given the opportunity
to discuss the proposed project, the al-
ternatives, and any potentially adverse
environmental impacts. A public hear-
ing shall be held where the proposed
project involves major construction,
dredging or other significant modifica-
tion of the environment. The appli-
cant shall provide a summary of his
responses to all public comments and
submit the summary, along with
copies of any written comments, with
the application.
§ 35.1620-5 State work programs and lake
priority lists.
(a)(1) A State shall submit to the
Regional Administrator as part of its
annual work program (§ 35.1513 of this
§35.1620-6
subchapter) a description of the activi-
ties it will conduct during the Federal
fiscal year to classify its lakes accord-
ing to trophic condition (§ 35.1630)
and to set priorities for implementing
clean lakes projects within the State.
The work plan must list in priority
order the cooperative agreement appli-
cations that will be submitted by the
State for Phase 1 and Phase 2 projects
during the upcoming fiscal year, along
with the rationale used to establish
project priorities. Each State must
also list the cooperative agreement ap-
plications, with necessary funding,
which it expects to submit in the fol-
lowing fiscal year. This information
will assist EPA in targeting resources
under § 35.1613.
(2) A State may petition the Region-
al Administrator by letter to modify
the EPA approved priority list estab-
lished under paragraph (a)(1) of this
section. This may be done at any time
if the State believes there is sufficient
justification to alter the priority list
contained in its annual work program,
e.g., if a community with a lower prior-
ity project has sufficient resources
available to provide the required
matching funding while a higher pri-
ority project does not, or if new data
indicates that a lower priority lake will
have greater public benefit than a
higher priority lake.
(b) Clean lakes restoration priorities
should be consistent with the State-
wide water quality management strat-
egy (see § 35.1511-2 of this subchap-
ter). In establishing priorities on par-
ticular lake restoration projects,
States should use as criteria the appli-
cation review criteria (§ 35.1640-1)
that EPA will use in preparing fund-
ing recommendations for specific
projects. If a State chooses to use dif-
ferent criteria, the State should indi-
cate this to the Regional Administra-
tor as part of the annual work pro-
gram.
§ 35.1620-6 Intergovernmental review.
EPA will not award funds under this
subpart without review and consulta-
tion in accordance with the require-
ments of Executive Order 12372, as im-
plemented in 40 CFR Part 29 of this
chapter.
9
-6
559
-------
§ 35.1630
40 CFR Ch. I (7-1-88 Edition)
[48 PR 29302, June 24, 1983]
§35.1630 State lake classification surveys.
States that wish to participate in the
clean lakes program shall establish
and submit to EPA by January 1, 1982,
a classification, according to trophic
condition, of their publicly owned
freshwater lakes that are in need of
restoration or protection. After De-
cember 31, 1981, States that have not
complied with this requirement will
not be eligible for Federal financial as-
sistance under this subpart until they
complete their survey.
§ 35.1640 Application review and evalua-
tion.
EPA will review applications as they
are received. EPA may request outside
review by appropriate experts to assist
with technical evaluation. Funding de-
cisions will be based on the merit of
each application in accordance with
the application review criteria under
§ 35.1640-1. EPA will consider Phase 1
applications separately from Phase 2
applications.
§ 35.1640-1 Application review criteria.
(a) When evaluating applications,
EPA will consider information sup-
plied by the applicant which address
the following criteria:
(1) The technical feasibility of the
project, and where appropriate, the es-
timated improvement in lake water
quality.
(2) The anticipated positive changes
that the project would produce in the
overall lake ecosystem, including the
watershed, such as the net reduction
in sediment, nutrient, and other pol-
lutant loadings.
(3) The estimated improvement in
fish and wildlife habitat and associat-
ed beneficial effects on specific fish
populations of sport and commercial
species.
(4) The extent of anticipated bene-
fits to the public. EPA will consider
such factors as
(i) The degree, n?.ture and sufficien-
cy of public access to the lake;
(ii) The size and economic structure
of the population residing near the
lake which would use the improved
lake for recreational and other pur-
poses;
(iii) The amount and kind of public
transportation available for transport
of the public to and from the public
access points;
(iv) Whether other relatively clean
publicly owned freshwater lakes
within 80 kilometer radius already
adequately serve the population; and
(v) Whether the restoration would
benefit primarily the owners of pri-
vate land adjacent to the lake.
(5) The degree to which the project
considers the "open space" policies
contained in sections 201(f), 201(g),
and 208(b)(2)(A) of the Act.
(6) The reasonableness of the pro-
posed costs relative to the proposed
work, the likelihood that the project
will succeed, and the potential public
benefits.
(7) The means for controlling ad-
verse environmental impacts which
would result from the proposed resto-
ration of the lake. EPA will give spe-
cific attention to the environmental
concerns listed in section (c) of Appen-
dix A.
(8) The State priority ranking for a
particular project.
(9) The State's operation and main-
tenance program to ensure that the
pollution control measures and/or in-
lake restorative techniques supported
under the project will be continued
after the project is completed.
(b) For Phase 1 applications, the
review criteria presented in paragraph
(a) of this section will be modified in
relation to the smaller amount of
technical information and analysis
that is available in the application.
Specifically, under criterion (a)(1),
EPA will consider a technical assess-
ment of the proposed project ap-
proach to meet the requirements
stated in Appendix A to this regula-
tion. Under criterion (a)(4), EPA will
consider the degree of public access to
the lake and the public benefit. Under
criterion (a)(7), EPA will consider
known or anticipated adverse environ-
mental impacts identified in the appli-
cation or that EPA can presume will
occur. Criterion (a)(9) will not be con-
sidered.
560
9-
7
-------
Environmental Protection Agency
§ 35.1650 Award.
(a) Under 40 CFR 30.345, generally
90 days after EPA has received a com-
plete application, the application will
either be: (1) Approved for funding in
an amount determined to be appropri-
ate for the project; (2) returned to the
applicant due to lack of funding; or (3)
disapproved. The applicant shall be
promptly notified in writing by the
EPA Regional Administrator of any
funding decisions.
(b) Applications that are disap-
proved can be submitted as new appli-
cations to EPA if the State resolves
the issues identified during EPA
review.
§ 35.1650-1 Project period.
(a) The project period for Phase 1
projects shall not exceed three years.
(b) The project period for Phase 2
projects shall not exceed four years.
Implementation of complex projects
and projects incorporating major con-
struction may have longer project pe-
riods if approved by the Regional Ad-
ministrator.
§ 35.1650-2 Limitations on awards.
(a) Before awarding assistance, the
Regional Administrator shall deter-
mine that:
(1) The applicant has met all of the
applicable requirements of § 35.1620
and § 35.1630; and
(2) State programs under section 314
of the Act are part of a State/EPA
Agreement which shall be completed
before the project is awarded.
(b) Before awarding Phase 2
projects, the Regional Administrator
shall further determine that:
(1) When a Phase 1 project was
awarded, the final report prepared
under Phase 1 is used by the applicant
to apply for Phase 2 assistance. The
lake restoration plan selected under
the Phase 1 project must be imple-
mented under a Phase 2 cooperative
agreement.
(2) Pollution control measures in the
lake watershed authorized by section
201, included in an approved 208 plan,
or required by section 402 of the Act
have been completed or are being im-
plemented according to a schedule
that Is included in an approved plan or
discharge permit.
§ 35.1650-2
(3) The project does not include
costs for controlling point source dis-
charges of pollutants where those
sources can be alleviated by permits
issued under section 402 of the Act, or
by the planning and construction of
wastewater treatment facilities under
section 201 of the Act.
(4) The State has appropriately con-
sidered the "open space" policy pre-
sented in sections 201(f), 201(g)(6), and
208(b)(2)(A) of the Act in any
wastewater management activities
being implemented by them in the
lake watershed.
(5)(i) The project does not include
costs for harvesting aquatic vegeta-
tion, or for chemical treatment to alle-
viate temporarily the symptoms of eu-
trophication, or for operating and
maintaining lake aeration devices, or
for providing similar palliative meth-
ods and procedures, unless these pro-
cedures are the most energy efficient
or cost effective lake restorative
method.
(ii) Palliative approaches can be sup-
ported only where pollution in the
lake watershed has been controlled to
the greatest practicable extent, and
where such methods and procedures
are a necessary part of a project
during the project period. EPA will de-
termine the eligibility of such a proj-
ect, based on the applicant's justifica-
tion for the proposed restoration, the
estimated time period for improved
lake water quality, and public benefits
associated with the restoration.
(6) The project does not include
costs for desalinization procedures for
naturally saline lakes.
(7) The project does not include
costs for purchasing or long term leas-
ing of land used solely to provide
public access to a lake.
(8) The project does not include
costs resulting from litigation against
the recipient by EPA.
(9) The project does not include
costs for measures to mitigate adverse
environmental impacts that are not
identified in the approved project
scope of work. (EPA may allow addi-
tional costs for mitigation after it has
reevaluated the cost-effectiveness of
the selected alternative and has ap-
proved a request for an increase from
the recipient.)
9
-8
561
-------
§35.1650-3
40 CFR Ch. I (7-1-88 Edition)
§ 35,1650-3 Conditions on award.
(a) All awards. (1) All assistance
awarded under the Clean Lakes pro-
gram is subject to the EPA General
Grant conditions (Subpart C and Ap-
pendix A of Part 30 of this chapter).
(2) For each clean lakes project the
State agrees to pay or arrange the
payment of the non-Federal share of
the project costs.
(b) Phase 1. Phase 1 projects are sub-
ject to the following conditions:
(1) The recipient must receive EPA
project officer approval on any
changes to satisfy the requirements of
paragraph (a)(lG) of Appendix A
before undertaking any other work
under the grant.
(2)(i) Before selecting the best alter-
native for controlling pollution and
improving the lake, as required in
paragraph (b)(1) of Appendix A of this
regulation, and before undertaking
any other work stated under para-
graph (b) of Appendix A, the recipient
shall submit an interim report to the
project officer. The interim report
must include a discussion of the vari-
ous available alternatives and a techni-
cal justification for the alternative
that the recipient will probably
choose. The report must include a
summary of the public involvement
and the comments that occurred
during the development of the alter-
natives.
(ii) The recipient must obtain EPA
project officer approval of the selected
alternative before conducting addi-
tional work under the project.
(c) Phase 2. Phase 2 projects are sub-
ject to the following conditions:
(l)(i) The State shall monitor the
project to provide data necessary to
evaluate the efiiciency of the project
as jointly agreed to and approved by
the EPA project officer. The monitor-
ing program described in paragraph
(b)(3) of Appendix A of this regulation
as well as any specific measurements
that would be necessary to assess spe-
cific aspects of the project, must be
considered during the development of
a monitoring program and schedule.
The project recipient shall receive the
approval of the EPA project officer
for a monitoring program and sched-
ule to satisfy the requirements of. Ap-
pendix A paragraph (b)(3) before un-
dertaking any other work under the
project.
(ii) Phase 2 projects shall be moni-
tored for at least one year after con-
struction or pollution control practices
are completed.
(2) The State shall manage and
maintain the project so that all pollu-
tion control measures supported under
the project will be continued during
the project period at the same level of
efficiency as when they were imple-
mented. The State will provide reports
regarding project maintenance as re-
quired in the cooperative agreement.
(3) The State shall upgrade its water
quality standards to reflect a higher
water quality use classification if the
higher water quality use was achieved
as a result of the project (see 40 CFR
35.1550(c)(2)).
(4) If an approved project allows
purchases of equipment for lake main-
tenance, such as weed harvesters, aer-
ation equipment, and laboratory
equipment, the State shall maintain
and operate the equipment according
to an approved lake maintenance plan
for a period specified in the coopera-
tive agreement. In no case shall that
period be for less than the time it
takes to completely amortize the
equipment.
(5) If primary adverse environmental
impacts result from implementing ap-
proved lake restoration or protection
procedures, the State shall include
measures to mitigate these adverse im-
pacts at part of the work under the
project.
(6) If adverse impacts could result to
unrecorded archeological sites, the
State shall stop work or modify work
plans to protect these sites in accord-
ance with the National Historic Pres-
ervation Act. (EPA may allow addi-
tional costs for ensuring proper pro-
tection of unrecorded archeological
sites in the project area after reevalu-
ating the cost effectiveness of the pro-
cedures and approving a request for a
cost increase from the recipient.)
(7) If a project involves construction
or dredging that requires a section 404
permit for the discharge of dredged or
fill material, the recipient shall obtain
the necessary section 404 permits
before performing any dredge or fill
work.
562
9-
-------
Environmental Protection Agency
Pt. 35, Subpt. H, App. A
§ 35.1650-4 Payment.
(a) Under § 30.615 of this chapter,
EFA generally will make payments
through letter of credit. However, the
Regional Administrator may place any
recipient on advance payment or on
cost reimbursement, as necessary.
(b) Phase 2 projects involving con-
struction of facilities or dredging and
filling activities shall be paid by reim-
bursement.
§ 35.1650-5 Allowable costs.
(a) The State will be paid under
§ 35.1650-4 for the Federal share of all
necessary costs within the scope of the
approved project and determined to be
allowable under 40 CFR 30.705, the
provisions of this subpart, and the co-
operative agreement.
-------
Pt„ 35, Subpt. H, App. A
40 CFR Ch. I (7-1-88 Edition)
(2) A geological description of the drain-
age basin including soil types and soil loss to
stream courses that are tributary to the
lake.
(3) A description of the public access to
the lake including the amount and type of
public transportation to the access points.
(4) A description of the size and economic
structure of the population residing near
the lake which would use the improved lake
for recreation and other purposes.
(5) A summary of historical lake uses, in-
cluding recreational uses up to the present
time, and how these uses may have changed
because of water quality degradation.
(6) An explanation, if a particular seg-
ment of the lake user population is or will
be more adversely impacted by lake degra-
dation.
(7) A statement regarding the water use of
the lake compared to other lakes within a
80 kilometer radius.
(8) An itemized inventory of known point
source pollution discharges affecting or
which have affected lake water quality over
the past 5 years, and the abatement actions
for these discharges that have been taken,
or are in progress. If corrective action for
the pollution sources is contemplated in the
future, the time period should be specified.
(9) A description of the land uses in the
lake watershed, listing each land use classi-
fication as a percentage of the whole and
discussing the amount of nonpoint pollut-
ant loading produced by each category.
(10) A discussion and analysis of historical
baseline limnological data and one year of
current limnological data. The monitoring
schedule presented in paragraph (b)(3) of
Appendix A must be followed in obtaining
the one year of current limnological data.
This presentation shall include the present
trophic condition of the lake as well as its
surface area (hectares), maximum depth
(meters), average depth (meters), hydraulic
residence time, the area of the watershed
draining to the lake (hectares), and the
physical, chemical, and biological quality of
the lake and important lake tributary
waters. Bathymetric maps should be provid-
ed. If dredging is expected to be included in
the restoration activities, representative
bottom sediment core samples shall be col-
lected and analyzed using methods ap-
proved by the EPA project officer for phos-
phorus, nitrogen, heavy metals, other
chemicals appropriate to State water qual-
ity standards, and persistent synthetic or-
ganic chemicals where appropriate. Further,
the elutriate must be subjected to test pro-
cedures developed by the U.S. Army Corps
of Engineers and analyzed for the same con-
stituents. An assessment of the phosphorus
(and nitrogen when it is the limiting lake
nutrient) inflows and outflows associated
with the lake and a hydraulic budget includ-
ing ground water flow must be included.
Vertical temperature and dissolved oxygen
data must be included for the lake to deter-
mine if the hypolimnion becomes anaerobic
and, if so, for how long and over what
extent of the bottom. Total and soluble re-
active phosphorus (P); and nitrite, nitrate,
ammonia and organic nitrogen (N) concen-
trator must be determined for the lake.
Chlorophyll a values should be measured
for the upper mixing zone. Representative
alkalinities should be determined. Algal
assay bottle test data or total N to total P
ratios should be used to define the growth
limiting nutrient. The extent of algal
blooms, and the predominant algal genera
must be discussed. Algal biomass should be
determined through algal genera identifica-
tion, cell density counts (numbers of cells
per milliliter) and converted to cell volume
based on factors derived from direct meas-
urements; and reported in biomass of each
major genus identified. Secchi disk depth
and suspended solids should be measured
and reported. The portion of the shoreline
and bottom that is impacted by vascular
plants (submersed, floating, or emersed
higher aquatic vegetation) must be estimat-
ed, specifically the lake surface area be-
tween 0 and the 10 meter depth contour or
twice the Secchi disk transparency depth,
whichever is less, and that estimate should
include an identification of the predomi-
nant species. Where a lake is subject to sig-
nificant public contact use or is fished for
consumptive purposes, monitoring for
public health reasons should be part of the
monitoring program. Standard bacteriologi-
cal analyses and fish flesh analyses for or-
ganic and heavy metal contamination
should be included.
(11) An identification and discussion of
the biological resources in the lake, such as
fish population, and a discussion of the
major known ecological relationships,
(b) A feasibility study consisting of:
(1) An identification and discussion of the
alternatives considered for pollution control
or lake restoration and an identification and
justification of the selected alternative.
This should include a discussion of expected
water quality improvement, technical feasi-
bility, and estimated costs of each alterna-
tive. The discussion of each feasible alterna-
tive and the selected lake restoration proce-
dure must include detailed descriptions
specifying exactly what activities would be
undertaken under each, showing how and
where these procedures would be imple-
mented, illustrating the engineering specifi-
cations that would be followed including
preliminary engineering drawings to show
in detail the construction aspects of the
project, and presenting a quantitative analy-
sis of the pollution control effectiveness and
the lake water quality improvement that is
anticipated.
564
9-
-------
Environmental Protection Agency
Pt. 35, Subpt. H, App. A
(2) A discussion of the particular benefits
expected to result from implementing the
project, including new public water uses
that may result from the enhanced water
quality.
(3) A Phase 2 monitoring program indicat-
ing the water quality sampling schedule. A
limited monitoring program must be main-
tained during project implementation, par-
ticularly during construction phases or in-
lake treatment, to provide sufficient data
that will allow the State and the EPA
project officer to redirect the project if nec-
essary, to ensure desired objectives are
achieved. During pre-project, implementa-
tion, and post-project monitoring activities,
a single in-lake site should be sampled
monthly during the months of September
through April and biweekly during May
through August. This site must be located
in an area that best represents the limnolo-
gical properties of the lake, preferably, the
deepest point in the lake. Additional sam-
pling sites may be warranted in cases where
lake basin morphometry creates distinctly
different hydrologie and limnologic sub-
basins; or where major lake tributaries ad-
versely affect lake water quality. The sam-
pling schedule may be shifted according to
seasonal differences at various latitudes.
The biweekly samples must be scheduled to
coincide with the period of elevated biologi-
cal activity. If possible, a set of samples
should be collected immediately following
spring turnover of the lake. Samples must
be collected between 0800 and 1600 hours of
each sampling day unless diel studies are
part of the monitoring program. Samples
must be collected between one-half meter
below the surface and, one-half meter off
the bottom, and must be collected at inter-
vals of every one and one-half meters, or at
six equal depth intervals, whichever number
of samples is less. Collection and analyses of
all samples must be conducted according to
EPA approved methods. All of the samples
collected must be analyzed for total and
soluble reactive phosphorus; nitrite, nitrate,
ammonia, and organic nitrogen: pH; temper-
ature; and dissolved oxygen. Representative
alkalinities should be determined. Samples
collected in the upper mixing zone must be
analyzed for chlorophyll a. Algal biomass in
the upper mixing zone should be deter-
mined through algal genera identification,
cell density counts (number of cells per mil-
liliter) and converted to cell volume based
on factors derived from direct measure-
ments; and reported in terms of biomass of
each major genera identified. Secchi disk
depth and suspended solids must be meas-
ured at each sampling period. The surface
area of the lake covered by macrophytes be-
tween 0 and the 10 meter depth contour or
twice the Secchi disk transparency depth,
whichever is less, must be reported. The
monitoring program for each clean lakes
project must include all the required infor-
mation mentioned above, in addition to any
specific measurements that are found to be
necessary to assess certain aspects of the
project. Based on the information supplied
by the Phase 2 project applicant and the
technical evaluation of the proposal, a de-
tailed monitoring program for Phase 2 will
be established for each approved project
and will be a condition of the cooperative
agreement. Phase 2 projects will be moni-
tored for at least one year after construc-
tion or pollution control practices are com-
pleted to evaluate project effectiveness.
(4) A proposed milestone work schedule
for completing the project with a proposed
budget and a payment schedule that is re-
lated to the milestone.
C5) A detailed description of how non-Fed-
eral funds will be obtained for the proposed
project.
C6) A description of the relationship of the
proposed project to pollution control pro-
grams such as the section 201 construction
grants program, the section 208 areawide
wastewater management program, the De-
partment of Agriculture Soil Conservation
Service and Agriculture Stabilization and
Conservation Service programs, the Depart-
ment of Housing and Urban Development
block grant program, the Department of In-
terior Heritage Conservation and Recrea-
tion Service programs and any other local.
State, regional and Federal programs that
may be related to the proposed project.
Copies of any pertinent correspondence,
contracts, grant applications and permits as-
sociated with these programs should be pro-
vided to the EPA project officer.
(7) A summary of public participation in
developing and assessing the proposed
project which is in compliance with Pare 25
of this chapter. The summary shall describe
the matters brought before the public, the
measures taken by the reporting agency to
meet its responsibilities under Part 25 and
related provisions elsewhere in this chapter,
the public response, and the agency's re-
sponse to significant comments. Section 25.8
responsiveness summaries may be used to
meet appropriate portions of these require-
ments to avoid duplication,
(8) A description of the operation and
maintenance plan that the State will follow,
including the time frame over which this
plan will be operated, to ensure that the
pollution controls implemented during the
project are continued after the project is
completed.
(9) Copies of all permits or pending permit
applications (including the status of such
applications) necessary to satisfy the re-
quirements of section 404 of the Act. If the
approved project includes dredging activi-
ties or other activities requiring permits, the
State must obtain from the U.S. Army
12
565
-------
§ 35.2000
40 CFR Ch. 1 (7-1-88 Edition)
Corps of Engineers or other agencies the
permits required for the discharge of
dredged or fill material under section 404 of
the Act or other Federal, State or local re-
quirements. Should additional information
be required to obtain these permits, the
State shall provide it. Copies of section 404
permit applications and any associated cor-
respondence must be provide to the EPA
project officer at the time they are submit-
ted to the U.S. Army Corps of Engineers.
After reviewing the 404 permit application,
the project officer may provide recommen-
dations for appropriate controls and treat-
ment of supernatant derived from dredged
material disposal sites to ensure the maxi-
mum effectiveness of lake restoration proce-
dures.
(c) States shall complete and submit an
environmental evaluation which considers
the questions listed below. In many cases
the questions cannot be satisfactorily an-
swered with a mere "Yes" or "No". States
are encouraged to address other consider-
ations which they believe apply to their
project.
(1) Will the proposed project displace any
people?
(2) Will the proposed project deface exist-
ing residences or residential areas? What
mitigative actions such as landscaping,
screening, or buffer zones have been consid-
ered? Are they included?
(3) Will the proposed project be likely to
lead to a change in established land use pat-
terns, such as increased development pres-
sure near the lake? To what extent and how
will this change be controlled through land
use planning, zoning, or through other
methods?
(4) Will the proposed project adversly
affect a significant amount of prime agricul-
tural land or agricultural operations on
such land?
(5) Will the proposed project result in a
significant adverse effect on parkland, other
public land, or lands of recognized scenic
value?
(6) Has the State Historical Society or
State Historical Preservation Officer been
contacted? Has he responded, and if so,
what was the nature of that response? Will
the proposed project result in a significant
adversely effect on lands or structures of
historic, architectural, archaeological or cul-
tural value?
(7) Will the proposed project lead to a sig-
nificant long-range increase in energy de-
mands? **
(8) Will the proposed project result in sig-
nificant and long range adverse changes in
ambient air quality or noise levels? Short
term?
(9) If the proposed project involves the
use of in-lake chemical treatment, what
long and short term adverse effects can be
expected from that treatment? How will the
project recipient mitigate these effects?
(10) Does the proposal contain all the in-
formation that EPA requires in order to de-
termine whether the project complies with
Executive Order 11988 on floodplains? Is
the proposed project located in a flood-
plain? If so, will the project involve con-
struction of structures in the floodplain?
What steps will be taken to reduce the pos-
sible effects of flood damage to the project?
(11) If the project involves physically
modifying the lake shore or its bed or its
watershed, by dredging, for example, what
steps will be taken to minimize any immedi-
ate and long term adverse effects of such ac-
tivities? When dredging is employed, where
will the dredged material be deposited, what
can be expected and what measures will the
recipient employ to minimize any signifi-
cant adverse impacts from its deposition?
(12) Does the project proposal contain all
information that EPA requires in order to
determine whether the project complies
with Executive Order 11990 on wetlands?
Will the proposed project have a significant
adverse effect on fish and wildlife, or on
wetlands or any other wildife habitat, espe-
cially those of endangered species? How sig-
nificant is this impact in relation to the
local or regional critical habitat needs? Have
actions to mitigate habitat destruction been
incorporated into the project? Has the re-
cipient properly consulted with appropriate
State and Federal fish, game and wildlife
agencies and with the U.S. Fish and Wildlife
Service? What were their replies?
(13) Describe any feasible alternatives to
the proposed project in terms of environ-
mental impacts, commitment of resources,
public interest and costs and why they were
not proposed.
(14) Describe other measures not dis-
cussed previously that are necessary to miti-
gate adverse environmental impacts result-
ing from the implementation of the pro-
posed project.
566
9
-------
v»EPA
United States
Environmental Protection
Agency
(WH-553)
Washington, DC 20460
Official Business
Penalty for Private Use
$300
------- |