&EPA
United States
Environmental Protection
Agency
Water
Office of Water and
Waste Management
Water Planning Division
Washington, D.C. 20460
September 1980
Agricultural Land Use
Water Interaction:
Problem Abatement,
Project Monitoring,
and Monitoring Strategies
-------
Review Notice
This report has been reviewed by the Environmental
Protection Agency and approved for publication.
Approval does not signify that the contents neces-
sarily reflect the views and policies of the Envi-
ronmental Protection Agency, nor does mention of
trade names or commercial products constitute
endorsement or recommendation for use.
The apendix referenced in the text is not included
as part of this publication. It is available in
limited number from the Environmental Protection
Agency's Water Planning Division on request.
-------
AGRICULTURAL LAND USE WATER QUALITY INTERACTION:
PROBLEM ABATEMENT, PROJECT MONITORING, AND MONITORING STRATEGIES
by
Jochen Ktthner
Meta Systems Inc
Cambridge, Massachusetts 02138
Purchase Order W-5571-NASX
Project Officer
Robert Moore
U.S. EPA Rural Nonpoint Section
Water Planning Division/Implementation Branch
Washington, D.C.
May 1980
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C.
-------
CONTENTS
Figures ii
Tables iii
Acknowledgment iv
1. Introduction 1
2. Interrelationship of Agricultural Land Use Activities
and Receiving Waters 4
Dominant Variables/Parameters 4
Recognition of Water Quality Impacts 10
Wind Erosion 13
3. Literature Review 14
4. Matrix Development (Agricultural Land Use Activities/
Water Quality Impacts 20
Pollutants Generated by Agricultural Land Use
Activities 20
Pollutant Transport by Runoff/Percolation ... .22
Transport by Sediment 26
Pollutant Impacts on Receiving Waters 27
Review of Individual Pollutants 27
Review of Receiving Water Types 32
Lakes and Reservoirs 35
Agricultural Land Use and Water Quality Impact . . . .37
Considerations in Selecting Abatement Practices/
Measures 40
Examples of Choosing Abatement Practices/Measures. . .52
5. Project Monitoring 54
Background 54
Setting up the Monitoring Network. 57
An Example of a Monitoring Program 60
6. Monitoring Strategy 66
7. Conclusions. 72
-------
FIGURES
Number Page
1 Logic of the report 2
2 Pathways in predicting impoundment water quality. . . 7
3 Control areas of nonpoint source pollutants 8
4 Comparison of practices lowland 11
5 Percent change of highest revenue factor lowland . 12
6 Technical project evaluation 51
7 Control pathways to be studied in the monitoring
program 62
8 Suggested monitoring stations in the watershed. ... 64
ii
-------
TABLES
Number Page
1 Summary of literature review 15
2 Potential generation of pollutants from various
(baseline) land use activities 21
3 Feedlot runoff characteristics 25
4 Potential impact of individual pollutants from
agricultural activities on receiving waters. ... 28
5 Estimated per capita contribution of indicator
microorganisms from human beings and some animals. 33
6 Agricultural land use/water quality impacts ..... 38
7 Principal types of cropland erosion control prac-
tices and their highlights 41
8 Practices for controlling direct runoff and their
highlights 43
9 Practices for the control of nutrient loss from
agricultural applications and their highlights . . 44
10 Practices for the control of pesticide loss from
agricultural applications and their highlights . . 45
11 Some sediment1 control practices for irrigated
agriculture 47
12 Animal holding control practices 47
13 Control practices/measures (orchard/vineyards).... 48
14 Control practices/measures (range and pasture).... 48
15 Control practices (homestead) 49
16 Effectiveness of soil and water conservation
practices in controlling pollutants 50
17 River watershed and reservoir monitoring program. . . 63
18 Project characteristics 68
iii
-------
Acknowledgments
The author wishes to acknowledge the input of Robert Moore and Walt Rittall
of U.S. EPA's Hater Planning Division, and the helpful discussions with Carl
Myers of U.S. EPA and John Peterson of USDA/SCS's Water Quality Section (cur-
rently on detail to EPA).
At Meta Systems, Alfred Leonard and Dianne Wood contributed significantly
to the content, shape, and style of the report. Diane Winkleby, Richard
Cromonic, and Paula Jackson were helpful in the report's production. However,
only the author is responsible for the final outcome of this document.
iv
-------
Section 1
Introduction
This report documents the work performed for the Water Planning Division's
Rural NFS Section (U.S. EPA, Washington, D.C.) under Purchase Order W-5571-
NASX. The project tasks are as follows:
Attempt to generalize agriculturally related NFS pollution problems
in various receiving waters and generally outline their potential
remedies through modification of practices and introduction of new
practices and practice combinations;
discuss the requirements, performance; and limitations of project
monitoring; and
consider the process/strategy of selecting projects for detailed
monitoring/evaluation (M/E) across the United States.
Figure 1 indicates the logic in performing these tasks and how they are inter-
related and covered in each section of-this report.
In our discussion we subscribe to the basic idea that existing water
quality problems must be resolved and that the potential remedial actions
must be geared to these particular problems. Since water quality degradation
is the issue, only those agricultural areas and practices that appear to cause
the problem should be modified, and the degree of modification or the intro-
duction of new practices must be determined by the expected water quality im-
provements. This necessitates identifying the cause-effect relationships in
a preliminary analysis for each project, and gearing the M/E efforts to the
particular problem and its land/water setting.
The following main sources provided the necessary background material:
Studies t
D.A. Haith and R.C. Loehr (eds.), "Effectiveness of Soil and Water
Conservation Practices for Pollution Control," prepared by Cornell
University for U.S. EPA, Athens, GA, October 1979 (EPA-600/3-79-106).
Meta Systems Inc, "Costs and Water Quality Impacts of Reducing Agri-
cultural Nonpoint Source Pollution," U.S. EPA, Athens, GA, August
1979 (EPA-600/5-79-009).
International Joint Commission, Pollution in the Great Lakes Basin
from Land Use Activities, Washington, D.C. 1980.
-------
FIGURE 1. LOGIC OF THE REPORT
I TASKS
I (OF REPORT)
AGRICULTURAL LAND USE/WATER
QUALITY PROBLEMS
AGRICULTURAL LAND USE > TYPE
OF POLLUTANT (AND BEHAVIOR)
POLLUTANTS > RECEIVING WATER
PROBLEM
' AGRICULTURAL LAND USE/WATER
QUALITY PROBLEM (GENERAL)
CONTRIBUTION OF OTHER LOCAL-
IZED FACTORS
ABATEMENT PRACTICES
POTENTIAL GENERALIZATION ABOUT
APPLICATION OF PRACTICES
SECT.
DISCUSS PARAMETERS OF
AGRICULTURAL LAND USE/
WATER QUALITY INTERFACE
(ACTIVITIES; RECEIVING
WATER POLLUTANTS; CATE-
GORIES OF PRACTICES)
SECT, 2
LITERATURE REVIEW
PARAMETERS
RECEIVING WATERS
MONITORING EFFORTS
ANALYSIS
SECT, 3
PROJECT MONITORING
PREMISES
STEPS TO BE FOL-
LOWED
T
SECT, 5
MONITORING STRATEGY
AVAILABLE MATERIAL
PUTTING MATERIAL
GOALS AND NEEDS FOP
SELECTION PROCESS
TOGETHER
SECT, 6
-------
Journals;
Journal of Environmental Quality
Water Resources Research
Journal of Soil and Water Conservation
EPA Documents;
Water Pollution from Cropland (Vols. I and II)
Animal Waste Utilization on Cropland and Pastureland
Guidance Document for NFS Monitoring (NFS Task Force) (Draft of March
1980)
USDA Documents;
RCA Reports/ Resources Conservation Act
Various Meta Systems Documents
-------
Section 2
Interrelationship of Agricultural
Land Use Activities and Receiving Waters
Dominant Variables/Parameters
This section is the basis for the work in the next sections. Here we
select the variables/parameters which are needed for a first-cut assessment
of the water quality impacts due to agricultural land use activities (and
their modifications), and which are needed as well to derive simple rules
for project monitoring and M/E selection strategy.
This scheme entails the following steps:
1) choosing the basic agricultural land uses;
2) choosing the basic receiving waters;
3) describing the basic water quality problems and their indicators;
4) modifying the basic agricultural land uses for water quality improve-
ment according to soil conditions, slope/length, climatic conditions,
and drainage patterns.
We will work through these steps in some detail.
Step 1
There doesn't seem to be a unifying framework of land uses that we can
draw on,* but on the basis of various reviews and discussions, we feel most
comfortable with the following relatively crude classification.
Cropland**
non-irrigated
irrigated
Orchard/Vineyard
Grazing land (including range, improved range, pasture, and improved
pasture)
Animal holding
Homestead
It is obvious that environmental problems created by a particular land use
*Various discussions with John Peterson, SCS (on detail to U.S. EPA), March
1980.
**In contrast to USGS classification, we do not include pastureland in cropland;
we assume, however, that hayland is included, possibly as part of a rotation.
-------
activity need to be evaluated in light of local dominant soil and physio-
graphic conditions. Thus local land form might be more important than the
land use activity itself, as Sonzogni et al. argue.* We, however, deal with
different land use activities by assuming similar local conditions in any
comparisons of environmental impacts.
Step 2
The following receiving waters are considered candidate types (very
much in agreement with EPA's NURB project):
lake/re servoir
small stream
river
estuary
oceanic bay (including Great Lakes Bays)
groundwater
This is basically a sound classification, but it includes receiving water
types that have a low priority under RCWA (as agreed upon by USDA and EPA),
such as estuaries and coastal areas. It is obvious, however, that it is
sometimes impossible hot to trace impacts as far downstream as necessary,
e.g., down to tidal estuaries. It might turn out to be useful in future
refinements of the scheme to differentiate between nonstratified (shallow)
and stratified lakes/reservoirs because the impacts of eutrophication are
quite different, and it might also be helpful to separate reservoir and lakes
in the future since reservoirs permit combinations of land and instream pro-
tection measures/management schemes that are different from those feasible
in lakes.
Thus we select the following receiving water types for our current
review:
lake/reservoir
small stream
river
bays (Great Lakes)
groundwater
Step 3
The following water quality*problems and their indicators are selected:
sedimentation
eutrophication
salinity
pesticides^
pathogens
BOD/organic materials
nitrates
*Sonzogni, W.C. et al., "Pollution from Land Runoff," Environmental Science
and Technology, Vol. 14, No. 2, February 1980.
-------
We recognize that we are lumping together broad water quality impacts
that are caused by a combination of various parameters and that are described
by various parameters (such as eutrophication), and specific one-parameter
problems (such as nitrate). We have chosen this representation because it
is the most common. In any case, laying out the pathway of pollutants indi-
cates very well how different parameters contribute to one "combined" problem
and how, at the same time, they are representative of other "individual"
problems (see Figure 2). Failure to recognize these interactions of various
parameters that can be controlled on the land leads easily to incorrect invest-
ments in control strategies designed to improve water quality.*
Furthermore, we have not included at this point those secondary effects
of the various water quality problems that must be anticipated. These can
be seen either in terms of the anticipated use of the impacted receiving
water or in general terms. For example, if the bacteria count is higher
than permitted, swimming might not be allowed. If there are high incidents
of eutrophication, the fish population changes as a result of the limited
oxygen supply in the hypolimnion. The degree of diversity declines; valuable
fish leave while trash fish survive; and total productivity may increase
due to greater food supply. In the latter case, it can be argued that this
change might provide a benefit to sport fishing, which implies that beneficial
uses could be derived from a generally degraded receiving water if "sport
fishing" is labeled the exclusive water use a highly unlikely prospect.
Step 4
In this step we select those practices/measures that effect water quality.
In Figure 3 we have differentiated among five control strategies that lead
to reduced edge-of-stream loadings and changing water quality:
1) Modification of land use activities without any additional structures
and/or addition of non-structural conservation/control practices;
2} Management strategies;
3) On-site structures attached to/associated with ongoing land use ac-
tivities;
4) Off-site structures capturing and/or modifying runoff and washoff;
5) Streambank and instrearn control measures.
Examples of the first control area are chisel plowing (instead of mold-
board plowing); of the second area, timing and application of fertilizer
and pesticides; of the third area, grassed waterways and terraces; of the
fourth area, sedimentation basins, ponds, and grass strips; and of the fifth
area, fencing against animals and copper sulfate applications against algal
bloom.
*A classic example is the switch from light-limited to phosphorus-limited
eutrophication. Soil conservation methods might not prove very effective
in controlling runoff/washoff of biologically available phosphorus. Thus
reduction in soil loss might increase light penetration in the impoundment
without substantial reduction in available nutrient supply.
-------
LOADINGS
COLOR
SEDIMENT
PHOSPHORUS
NITROGEN
TRAPPING/
DECAY RATES
COLOR
SEDIMENT
> PHOSPHORUS
*- NITROGEN
OUTFLOW/
EPIL1MNETIC
CONCENTRATIONS
IMPOUNDMENT MORPHOMETRIC
AND HYDROLOGIC CHARACTERISTICS
COLOR,
TRANSPARENCY
CHLOROPHYLL-A
CONCENTRATION
HYPOLIMNETIC
OXYGEN DEFICIT
FIGURE 2. PATHWAYS IN PREDICTING IMPOUNDMENT WATER QUALITY
-------
FIGURE 3. CONTROL AREAS OP NONPOINT SOURCE POLLUTANTS
00
LAND USE
ACTIVITIES
RUNOFF
WASHOFF
STRUCTURES
RUNOFF
WASHOFF
STRUCTURES
II
III
IV
Control Areas
modification of
activities with-
out any structures
and/or addition of
conservation/con-
trol activities
management on-site
strategies structures
off-site structures
source control
del livery/trans port
control
RUNOFF
WASHOFF
REC,
WATER
streambank & in-
stream measures
direct problem
control
-------
We feel that this differentiation of control options is useful for control
of erosion as well as other pollutants for all land use activities. It pro-
vides a sense of the geographical separation of control options, a logical
view of what type of practices/measures can be combined, and where and how
their direct effects can be measured. It also indicates at a first glance
which requirements are imposed on any monitoring attempts. Practices listed
in the Conservation Manual* or the USDA/EPA publication on control of water
pollution from cropland** can be put into these categories.
We recognize at this point that control strategies must be planned/de-
signed in accordance with the natural/human-induced conditions of the areas
to be controlled. Therefore, it is desirable to keep, say, a clayey soil,
which has a relatively high enrichment ratio, out of the receiving water
because of the potential inducement of biological activity, but it also be-
comes obvious that clayey soil is difficult to control by off-site structures
that operate solely on physical control principles such as settling basins.
Thus minimizing/preventing detachment might be the only reasonable control
strategy. Based on the composition of the soils predominant in the areas
to be controlled, a first selection of appropriate control mechanisms should
be made.
We prefer to use the type of land use activity as our reference for
control categories (like cropland) and not the general land-based problems
(such as erosion and nutrient wash-off). This means that we have to develop
practices for combinations of each land use activity and its land-based prob-
lem types which in turn are dependent upon the local soil and physiographic
conditions. Clearly, there are overlaps; if erosion occurs from croplands,
erosion control will always interface with fertilizer and pesticides that
are applied.
We also prefer to define baseline conditions for each land use activity
that constitutes (in a simplified sense) the most common practice:
Continuous row crop (non-irrigated cropland), tilled up and down
slope without any particular sophisticated fertilizer and pesticide
application practice/management;
Continuous row crop (irrigated cropland), furrow irrigated without
any particular fertilizer and pesticide application practice/manage-
ment;
Draining from animal holding areas into the drainage of surrounding
land;
On orchards/vineyards, reducing the possible endangerment of the
harvest by intensive chemical spraying with only minimal considera-
tions of other aspects, such as erosion;
Grazing range and pastureland without any advanced plan/management
*USDA, Soil Conservation Service, National Handbook of Conservation Prac-
tices Notice, Washington, D.C., 20013.
**USDA/EPA, Control of Water Pollution from Cropland, Vol. I, November
1975 (see Table 12, p. 63).
-------
of these resources with regard to their carrying capacity;
, Homestead, relying on the surrounding land for absorption of waste/
residuals.
We are aware of the oversimplification of these definitions of "base-
line." But it is essential that we state some baseline conditions in order
to make meaningful estimates of water quality impacts and their modifications.
Recognition of Water Quality Impacts
Having identified the various parameters and components of interest
here, it is appropriate to make a brief statement about previous experiences
in linking land-based activities to receiving waters and in identifying
changes in water quality due to modifications of land-based activities. In
a recent study based on the land data from the Black Creek Project,* Meta
Systems analyzed long-term average impacts of various cropland activities
on the water quality of a synthesized downstream impoundment.
Eleven selected farm practices (ranging from monoculture to rotation
and structural measures) were analytically compared. Figure 4 shows the
results as applied to one soil type, the lowland (silty clay; hydrologic
class D). Various comparisons can be made, such as between practices on
the basis of individual parameters or on an overall basis, or between indivi-
dual parameters of one specific practice. Eigure 5 depicts relative impacts;
the highest net revenue is chosen as base case. The relative results can
be used to identify the impact on individual uses. For example, a switch
from the base case (CB-CH) to a C-B-W-H rotation would result in a decrease
in each of the six water quality parameters, ranging from a decrease in im-
poundment sedimentation of almost 70 percent to a decrease in impoundment
biomass of about 3 percent. This might also indicate that the water quality
problem cannot be solved by control practices of categories 1 and 3, but
only in combination with category 2 techniques such as reduced fertilizer
application.
What is of major interest here is the attenuation effect of the receiv-
ing water.** Even in an impoundment, whose water quality is considered quite
sensitive to input changes, significant edge-of-stream loadings are trans-
ferred to significant water quality changes only in a few instances. Since
in the modeling exercise, a specific practice was implemented in the total
watershed at one time, it becomes obvious what a difficult task the M/E
*Meta Systems, Costs and Water Quality Impacts of Reducing Agricultural
NOnpoint Source Pollution, U-S. EPA, Athens, GA, EPA-600/5-79-009, August, 1979,
**For the overall configuration of watershed area, impoundment surface
area, and impoundment mean depth values of 200 km2, 5 km2, and 4 m, respec-
tively, have been selected as being typical of watershed/impoundment configu-
rations in the midwestern data set used to develop the impoundment models.
With a total flow rate of .25 m/yr from the watershed, the hypothetical im-
poundment has a surface overflow of 10 m/yr and a mean hydraulic residence
time of .4 years.
10
-------
NET REVENUE
tK$)
SOIL LOSS .
(kg/rr.f of
wotershed-yr)
SEDIMENTATION
(kg/m* of <
impoundment-yr)
RIVER NITROGEN
(g/ms) 20
10
RIVER PHOSPHORUS
CoVm') ^
JO
0
RIVER LIGHT EXTINCTION COEFFICIENT
"ML
20
IMPOUNDMENT LIGHT EXTINCTION COEFFICIENT
m-' 2.0
1.0
IMPOUNDMENT BIOMASS
(g Chlorophyll- .03
A/m»)
Key:
CC-CV: Continuous corn,
conventional tillage,
without terracing
CC-CH: Continuous corn,
chisel plowing, with-
out terracing
CC-NT: Continuous corn,
no-till planting,
without terracing
CB-CV: Corn-soybeans,
conventional tillage,
without terracing
CB-CH: Corn-soybeans,
chisel plowing,
without terracing
CB-NT: Corn-soybeans,
no-till planting,
without terracing
CBWH: Corn-soybeans
wheat-hay, conven-
tional tillage for
corn only, without
terracing
CBWH-NT: Corn-soybeans
wheat-hay, no-till
planting, without
terracing
CC-CVT: Continuous corn,
conventional tillage,
with terracing (PTO
terraces)
CC-CHT: Continuous corn,
chisel plowing, with
terracing
CB-NTT; Corn-soybeans,
no-till planting,
with terracing
FIGURE 4. COMPARISON OF PRACTICES LOWLAND
11
-------
REVENUE
(%)
SOIL LOSS
(%)
0
-40
-80.
80
40
0
-40
SEDIMENTATION
Key: See Figure 4.
RIVER PHOSPHORUS
20r
"r
-20r
RIVER LIGHT EXTINCTION COEFFICIENT
(%) 50
0
-20
-60
IMPOUNDMENT LIGHT EXTINCTION COEFFICIENT
1
t^^T^'J
IMPOUNDMENT BICMASS
FIGURE 5. PERCENT CHANGE OF HIGHEST REVENUE FACTOR IXJWLAND
12
-------
projects of RCWP face in delineating impacts from much smaller areas by water
quality monitoring and evaluation techniques.
Wind Erosion
Wind erosion has indirect impacts on the quality of a region's water
bodies. Soil particles (and their attached chemicals) are blown off a field
and fall out onto water (with and without compositional changes in the air).
The fallout has a definite impact, but given our current state of knowledge
it is even more difficult than for the water-based erosion to establish a
relationship between origin, transport (modification), and fallout location.
If wind is the predominant transport mechanism, detailing the transport route
over local and regional distances is extremely difficult. Also, measuring
the composition of the fallout material is a very hard task. For example,
it has to be decided whether and how dust particles should be isolated to
give the quality of precipitation, the main source/force for fallout, which
then implies the dust composition.
Factors similar to those in water-based erosion determine the erosion/
credibility of an area. The combination of these factors for a quantitative
analysis has been attempted in the wind erosion equation, which is analogous
to the Universal Soil Loss Equation**^?he factors considered include wind
credibility, soil-ridge roughness, climate, median unsheltered travel distance
across a field, and equivalent quantity of vegetative cover. Based on the
identification of these factors, essential for wind erosion potential and
controls, respectively, effective control means can be designed to curb wind
erosion. Various practices that control water-based erosion are also effec-
tive in reducing wind erosion; but given the difficulty in relating control
and impact, we can only emphasize the usefulness of these controls in areas
hit by wind erosion without showing their positive impact. Therefore, we
will not discuss wind erosion further in this report.
*See, for example, Skidmore, E.L., F.S. Fischer, and N.P. Woodruff, "Wind
Erosion Equation: Computer Evaluation and Applications," Soil Service Amer.
Proc., Vol. 34, 1970, pp. 331-335.
13
-------
Section 3
Literature Review
The literature survey* covered 36 different articles reporting the find-
ings of monitoring programs designed to answer a variety of questions (Table
1). These studies varied in pollution problems investigated, the aspects
of the pollution phenomena addressed, the type and period of monitoring,
and several other variables. Some general characteristics of the material
surveyed are:
The studies were conducted in 19 different states with the greatest
concentration in states with large agricultural schools.
The pollutants investigated included nutrients, pesticides, path-
ogenic bacteria, plutonium, oxygens and salt.Approximately 75
percent of the studies, however, were concerned primarily wih nutrients.
Host studies listed no specific environmental problem related speci-
fically to the monitoring site. Those water quality problems that
were mentioned were usually theoretical such as a mention of eutro-
phication when nutrients are being investigated. Likewise, receiving
waters were often unspecified, though our interpretation of the arti-
cles gives a distribution of 31 streams, 3 groundwater aquifers, 1
estuary, and 1 retention pond.
The pollution sources addressed were predominantly cropland or crop-
land mixed with pasture, woodland,and/or residential areas.
Abatement practices were investigated (to some degree) in 12 of these
studies. Terraces, contour cultivation, grass-lined waterways, and
various range and crop management techniques were addressed, but
some of the other central techniques (such as sod-based rotation.
and filtration strips) were not.
Monitoring techniques varied widely. Only 11 of the 36 tests used
some automatic equipment; station locations ranged from natural streams
to outlet culverts to runoff collection tanks? and test durations
ranged from a month or 2 to 10 years (14 studies monitored for over
2 years, and an additional 14 monitored for 1 year or more).
Being reasonably representative, these studies suggest that the following
needs should be recognized by the RCWA program. Care must be taken that
the studies selected correlate well with pollution problems' specific data
*The review was not intended to be exhaustive, but was undertaken to docu-
ment some of the existing knowledge in this field.
14
-------
TABLE 1. SUMMARY OF LITERATURE REVIEW
Article
1
2
3
4
5
6
7
8
9
Location
CO
GA
GA
HI<«>
ID
ID
ID
IN
IA
Pollution Problem
salt
N, CL
herbicides.
sediment
herbicide
coliform,
bacteria
coliform
coliform.
BOD
N,P
N, P.
sulfrite,
sediment
Receiving
stream
stream
stream
estuary
stream
stream
Effects
Pollution Source
Land
Class (es)
MX<2)
CR<1}
CR01
Pineapple
CR-I
PAS. MX121
MX
CR(1)
CR/T"'
Area
(acres)
4,734
3.2
3-6.7
2-6.2
203,000
57,600
10OO-1544
.021
7.9-22.7
Soil
Type(s)
grassed
waterways.
terraces.
cover crops
grassed
waterway a.
terraces,
cover crops
silt loant
si It- loam
% Slope
3-12%
0-12%
).2-12.4%
3-G%
Abatement
Practice
various
tillage
systems
drained and
uudr.i i ned
terraces
Time
Period
1 yr.
3 yrs.
3 yrs.
sampler
3 mo.
1 yr.
3 yrs.
2 yrs.
3 yrs.
Monitoring
Frequency
of Sampling
varies (3>
runoff events
runoff events
used was faul
2 -week
Intervals
weekly during
grazing, else
biweekly
2-week
intervals
simulated
storm
events
runoff events
Type ot
Sampler
auto-
matic
y
grab
grab
grab
automati
and grab
grab
Location
of Sampler
streams and
canals
water slitd
outlet '
watershed
outlet
in stream
sub- surface
drains
in stream
(22 stations)
outlet
outlet
at outlet
weir and col-
lection tank
Ul
1. Multiple watersheds. See the abstract for more complete Information.
2. Mixed watershed. See the abstract for more complete information.
3. Multiple sampling stations.
4. Two monitoring studies repeated in one article.
-------
TABLE 1. (CONTINUED)
Article
in
11
12
13
14
15
16
17
ia
f /
location
IA
IA
IA
KY
IA
MI
MN
HB
1
m
Pollution Problem
Parameters
P.N
H.P, alqil
density
N
H.P
N,P,K
N, chloride
N,P, others
H,P
P,DO,H,BOD
Receiving
Mater
stream
stream
stream
stream
groundwater
groundwater
stream
Effects
outroph.
Pollution Source
band
Class (ea)
c*(1)
CR, RS.
MX"'2'
CR(1>
MX(l>
CR, PAS(1>
MX(S)
N PAS
...(2,3)
CR
r*B T *
' (I)
PAS-I11'
Aren
(acres)
74-128
40,000-
3,600,000
74-150
494-
200,000
140,000
.06
31. 3-60.0
Soil
Typ«<«>
loeaa
loesn
silt-loam
varies
silt-loam
loess
% Slope
2-18%
varies
5%
>t
I
Abatement
Practice
various
terrace
types
terracing.
contour
planting
various
tilling
techniques
various
management
techniques
range
management
Monitoring
Time
Period
3,4 yrs
14 mo.
3 yrs.
2 yrs.
1-2 no.
12 mo.
5 yr».
(soil s.
1
"season"
Frequency
of Sampling
7.4 per run-
off event
weekly-summer
monthly
storm event
monthly
storm event
weekly plus
some
scattered
runoff event
moles only)
at 12-hr.
intervals
during irr.
Type of
[Sampler
grab
grab
grab
grab
grab
grab
grab
Location
of Sampler
watershed
outlet
stream (16
stations)
outlet
in streams
tank
in stream
runoff cap-
tured in tank
stream gaging
station
a\
1. Multiple watersheds. See the abtitract for more complete Information.
2. Mixed watershed. See the abstract tor more complete information.
3. Multiple sampling stations.
-------
TABLE 1. (CONTINUED)
Article
19
20
21
22
23
24
25
26
27
Location
NV
NY
NY
OH
OH
OH
OH
Oil
OH
Pollution Problem
Parameters
DO,BOD,N,P
P
N,P
N,P,K
N,P,K
N
P,N, others
Plutonium,
cesium
N.P.sed.,
others
Receiving
Hater
stream
stream
stream
stream
groundwater
stream
retention
ponda
stream
Effects
Pollution Source
Land
Class (es)
CR-I(1)
PAS-I
MX<2>
crop."'
barnlot
CR,HD(1>
RS,MX(2>
PAS(1)
CR.PAS'1'
CR, MXU)
Area
(acres)
21.3-60
81,500
.79
0.4
43.5 and
303.7
304
5.4-8.9
205-262 (1)
Soil
Type(s)
si It- loan
silt-loam
si It- loam
silt-loam
clay, clay
loam and
sandy loam
% Slope
2-4%
13%
12-18%
Abatement
Practice
range
management
various crop
managonent
techniques
range
management
Monitoring
Time
Period
(nee st
20 mo.
1 yr.
21 yrs.
4 yrs.
5 yrs.
2 yrs.
32
Frequency
of Sampling
idy no. 18. T
from several
per day to
2 /month)
varies with
flow
storm event
storm event
every two
weeks and dur
ing runoff
events
usually
weekly
weekly and
runoff events
(sediment uamj
ponds and ne,
samples were
runoff events
Type of
Sampler
ey used
grab and
lutomatic
automatit
and grab
automa-
tic
grab
grab
auto-
matic
les were
rby. No
taken.)
grab
Location
of Sampler
he same data)
stream
outlet and
tank
outlet
stream, at
weirs
springs
watershed
outlet
taken in
water
watershed
outlet
1. Multiple watersheds. See the abstract for more complete information.
2. Mixed watershed. See the abstract for more complete information.
3. Multiple sampling stations.
-------
TABLE 1. (CONTINUED)
irticle
29
29
30
31
32
33
34
35
36
Location
OK
OK
PA
PA
SD
TX
IX
HI
HI
Pollution Problem
Parameters
N.p
, f
N,P,aed.
antracine
pesticide)
P
P,N,SS,COO
arsenic
N
P, Bed.
N,P,K
Receiving
teter
tream
stream
stream
tream
strean
stream
stream
stream
Effects
Pollution Source
Land
Claas(as)
PAfl(1>
CR.RG11'
CRU)
CR.MX
CR1-1'
CR(1>
CR(1)
CR,WD(1>
*X<2) (PAS,
m
Area
(acres)
19-27
01
1,900
8.8-18.7
9.9
9.9
.0003
(20 plots)
(12,000
Soil
Type(s)
silt-loams
silt-loams
ilty clay
oam
loan,
silty loan
loan and
sandy clay
LOdM
loustoir
>lack clay
louston
ilack clay
)raway loan
Loess
% Slope
.7-3.5%
4%
;3%
£3*
4-6%
Abatement
Practice
range
management
various crop
and range
management
techniques
pesticide
application
techniques
various crop
rotations
tillage
techniques
manure
application
practices
Monitoring
Time
Period
10 yrs.
12 mo.
13 mo.
(over
2 yrs.)
33 mo.
2 yrs.
3 yra.
57 mo.
2 yrs.
2 yrs.
Frequency
of Sampling
runoff events
runoff events
storm event
3/wk-14 no.,
wuekly-19 mo.
runoff events
runoff events
runoff events
simulated
runoff events
6-12 per yr.
Type of
Sampler
automatic
and grab
utomatic
grab
grab
grab and
lutomatic
automati
automati
and grab
grab
grab
location
of Sampler
watershed
outlets
watershed
outlet
runoff Lank
strean (1
station at
weir)
watershed
outlet
watershed
outlet
watershed
outlet
vacuum
collection
system
itrean (30
stations
00
1. Multiple watersheds. See the abstract fox more complete information.
2. Mixed watershed. See the abstract for more complete information.
3. Multiple sampling stations.
-------
needs, and similar consideration must be given to see that all of the relevant
pollution control and management practices are addressed. Also, some guide-
lines should be established to ensure that monitoring results from different
monitoring programs are compatible in aggregate statistical analyses.
19
-------
Section 4
Matrix Development
{Agricultural Land Use Activities/Water Quality Impacts)
This section serves primarily to develop a matrix that relates agricul-
tural land use activities to water quality, and to indicate how water quality
is affected by modification of the land use. The discussion is directed
toward those land uses and receiving waters that were outlined in Section
2 and touches on effects of local conditions and controls for agricultural
land use. Examples are presented for some control categories.
Pollutants Generated by Agricultural Land Use Activities
Precipitation and solar radiation are uncontrollable inputs to the agri-
cultural production system, which is characterized by its soils and topo-
graphy. For cropland, these inputs, together with controllable inputs such
as seed, fertilizer, and pesticides, and input activities such as tilling
and disking, result in crop production, but also in the negative impacts
of agricultural runoff and percolation, erosion and sedimentation, and nutri-
ent and pesticide washoff. We are not dealing with agricultural production
here, but with the control of these negative impacts. Precipitation results
in infiltration/percolation and runoff; precipitation, together with runoff,
results in erosion and sediment transport; and runoff and percolation, as
well as sediments, act as carriers for organic matter, chemical compounds,
and pathogens depending upon many variables, such as topography, climate,
and type and sophistication of land use activities. In Table 2 we have brief-
ly summarized the pollutant generation of the various land use activities
(at their baseline conditions) caused by the three transport mechanisms.
:-.At this point we will briefly review the transport mechanisms and their
pollutants, with particular emphasis on cropland.* First, we review runoff
and percolation together, and then sediments' carrying capacity. Runoff
occurs when the rate of precipitation exceeds the infiltration rate of the
particular soil. Precipitation (characterized by its intensity, duration,
and frequency) and, to a certain extent, the resulting runoff control the
detachment process of soil particles together with other natural and man-
induced conditions of land and soil. The detached particles are transported
by the runoff as long as the water's volume and velocity are sufficient to
carry them. They might be deposited on land at one point (with the heavier
particles settling out first), or end up as sediment In the stream. The
*A1 though much of this material can be found in the literature, we present
it here in a condensed form in order to provide a common background for our
discussion.
20
-------
TABLE 2. POTENTIAL GENERATION OF POLLUTANTS FROM VARIOUS (BASELINE) LAND USE ACTIVITIES
Cropland
non- irrigated
irrigated
Pasture/Range
Orchard/Vineyard
Animal Holding
Homestead
Pun-
off
X
X*
X
X
X
X
Perco-
lation
X
X
X
(X)
X
(X)
Sed. N f
X
X
X
X
(X)
(X)
X
(X)
X
X
X
(X)
X
Runoff
Patho-
BOD gens
(X)
X
X
(X)
X
(X)
Pesti-
cides Salt
X
X
(X)
X
X
N
X
X
X
X
Percolation
Patho-
P qens Pest.
(X)
(X)
-------
remaining sediments have higher concentrations of fine particles. In addition
to sediments, sellable fractions of nutrients, pesticides, bacteria, and organ-
ic matter are removed from land. We concentrate here on nutrients and pesti-
cides.
Pollutant Transport Joy Runoff/Percolation
Nutrients from Cropland. During rainfall, small surface depressions
are filled with water prior to the beginning of overland flow (runoff). Sur-
face-applied fertilizers can dissolve into this water. Nitrate, being very
soluble, is leached to a large extent into the groundwater by infiltration
the rate of which depends on the soil type (hydrologic group). When the
water in depressions becomes part of the runoff (after infiltration capacity
is exceeded), it transports a load of nutrients. The longer the time between
application and overland flow, the greater the chance that moist soil has
dissolved fertilizer and that light precipitation has leached the fertilizer
(especially NOg) into the ground, which means that there are fewer soluble ^
components in the runoff. If, however, the infiltrated water moves laterally
and again becomes surface flow (interflow), the nitrate will end up in surface
water.* These facts explain why it is virtually impossible to generalize
the relationship of nitrogen (N) concentration or loading with surface flow.
Phosphorus (P) is quickly immobilized, so soluble inorganic phosphorus
in surface runoff usually represents only a small fraction of the total P
lost to surface waters. Most evidence suggests that subsurface and ground-
water movements of P are not a major source of stream pollution; the chance
of P contributions to groundwater would be greatest on well-drained, coarse-
textured (sandy) soils receiving large amounts of fertilizer and water, as
well as on peat-soil, since these soils have little tendency to react with
P. In general, soil solutions contain less than 0.1 mg of P per milliliter;
thus leaching losses are extremely low even in well-drained soils.
In areas with seasonal snow cover, losses of soluble P may
be appreciable in surface runoff during snowmelt, when soil/water contact
is limited due to frozen soils and P may be leached from surface
crop residues.
Pesticides from Cropland. Chemical and physical characteristics of
pesticides, in combination with environmental factors, determine the potential
losses through runoff and percolation/leaching with individual pesticides
reacting more or less to specific environmental factors. The balance between
dissolved pesticides and solid phase pesticides, often described by the ad-
sorption partition coefficient, is the most important factor in allocating
pesticide losses to the transport phenomena with which we are concerned.**
*It is important to understand soil characteristics in terms of their
hydrology and the geology. Poorly drained soils (hydrologic groups C and D)
should produce larger loads of nitrate in runoff than well-drained soils.
However, this is equalized if the leached nitrate becomes interflow.
**The Cornell study, citing Wauchope (1978) in order to put the usefulness
of pesticide control by water and soil conservation practices (SWCPs) in
22
-------
Various environmental factors influence the site of this equilibrium at vari-
ous times. Potentially important parameters are organic matter content of
soil, soil temperature, acidity, cation exchange, moisture content, and clay
mineral content. Since water in the soil competes with pesticides for adsorp-
tion sites on soil particles, the fraction of the chemical adsorbed increases
as the moisture level in the soil decreases. Precipitation on a dry soil
will therefore desorb a portion of the pesticide, which will then move with
the water in any runoff that occurs.
Thus pesticides might be moved by overland flow or leached down through
the soil (dissolved in water or in soil solution) , perhaps reappearing later
in surface runoff (via interflow) or in groundwater. But, as explained above,
the quantity of a pesticide actually moving in water from a treated area
in any given runoff event depends on factors such as erosion and other runoff-
related factors such as topography, intensity and duration of rainfall, soil
credibility, and land management and cropping practices, as well as on manage-
ment factors, such as the amount of pesticides initially applied (i.e., ante-
cedent soil moisture) and placement of pesticides. These factors are described
in detail elsewhere.* Some of the important factors are:
Characteristically, pesticide losses are highest in the first runoff
occurring after application of the chemical, and the magnitude of
the loss generally decreases as the time between application and
runoff increases. The effect of elapsed time is particularly notice-
able with short-lived pesticides and those that are not incorporated
into the soil. Therefore, time management of pesticides must be
determined by climatological and soil conditions.
Due to the competition between water and pesticides for adsorption
sites on soil particles, some pesticides will have greater losses
in runoff if applied to wet rather than .dry soil; this is particularly
true for runoff events occurring shortly after application.
perspective, states: "The small percentage of toxic material transported
in runoff water and sediment limits the usefulness of SWCPs for reducing
pesticide pollution in comparison with other control procedures. Wauchope
(1978)/ in his extensive review of the literature on pesticide runoff losses,
concluded that losses are generally 0.5 percent or less of the amount applied,
with larger losses indicating the occurrence of a large runoff-producing
rainfall event within one to two weeks after the pesticide application. Wau-
chope estimated losses for organochlorines to be about one percent of the
applied material. He concluded that the major reason for the higher losses
is the persistence of organochlorine insecticides in soil. Wauchope also
noted that losses of herbicides applied as a wettable powder may be as high
as five percent of the applied material, depending on weather conditions
and the slope of the treated field." (p. 210) It is also obvious that the
impacts of even the relatively small losses on runoff and sediment cannot
be ignored and must be investigated on a case-by-case basis.
*Caro, J.H., "Pesticides in Agricultural Runoff," (chapter 5) in USDA/EPA,
op. cit., Vol. II, 1978.
23
-------
Placement of pesticides in soil generally means that there will be
fewer pesticides lost in runoff than if they are left on the surface
or sprayed on foliage.
Persistence of pesticides in soils affects the temporal change of
amounts lost in runoffs occurring at different intervals since appli-
cation. However, many factors influence persistence, such as the
sequence of overlapping loss processes* after application (implying
rapidly changing loss potential in the initial period); weather;
cultural practices; type, temperature, moisture level, and acidity
of the soil; and interactions between chemicals when more than one
pesticide is applied.
Pathogens, Organic Material, and Nutrients from Animal Holding. Runoff
from areas having large animal populations can contribute significantly to
water quality problems. Clearly, the actual impact depends on the waste
production and its characteristics per animal, number of animals, and the
management practices (which implies the magnitude of a slug discharge).
The concentration of nutrients and organic material found in manures
of animal holding areas depends upon the time of year and the age of the
manure. The quality of the runoff will be a function of the physical and
biochemical changes that occur. There is less decomposition in the winter
than in the summer, so that a large concentration of pollutants could accumu-
late. But there are other influences as well. Decomposition depends not
only on temperature, but also on moisture content; also, the longer the manure
remaing wet, the better the chances of biological degradation of the pollu-
tional compounds. In dry climates, when manure dries out rapidly, the pollu-
tional constituents of ihe manure remain essentially constant. When the
manures are wetted again by precipitation, the quality of runoff is essentially
the same as it would have been if the material had been discharged at the
time it'was first deposited on the ground. The concentration of the various
pollutants in the runoff is highest during the initial phase of rainfall;
it decreases as runoff continues.
The results of several studies describing the magnitude and variability
of constituents in runoff have been suannarized in Table 3. The variability of
runoff is illustrated by the BOD range of values, which varies from 500 mg/fc
to 12,000 mg/£. Solids and nitrogen concentration show even wider variations.
The variable nature of the runoff indicates the significant slug effect that
these discharges could have on a receiving water.
Salt from Irrigated Croplands. In areas with irrigated cropland, seep-
age/percolation and return flows from irrigated lands carry soluble salts
into deceiving ditches and thus into receiving water. This can result in
substantially increased salt concentrations over natural levels, so that
a continuous increase of salt concentrations must be expected in the downstream
direction of a river that provides the irrigation water for a region.
*These are volatilization, sorption, leaching, and eventually chemical and
biological degradation (see A.E. Hiltbold, "Persistence of Pesticides in Soil,"
in Pesticides in Soil and Water, (W.D. Guenzi, ed.). Soil Sci. Soc. Amer., Inc.
Madison, Wis., 1974, pp. 205-222.
24
-------
TABLE 3. FEEDLOT RUNOFF CHARACTERISTICS
to
Ul
Range of Values for Constituents, mg/l
Ortho-
Suspended phosphate Organic Ammonia Nitrate
Animal Solids R\ Nitrogen Nitrogen Nitrogen BOD
COD
feference
Cattle 3,400-13,400
Cattle
Cattle 1,000-7,000*
500-3,300
6-800 2-770 0-1,270 1,000-12,000 2,400-38,000
300-6,000
Cattle 1,500-9,000 4,000-15,000
Cattle 1,400-12,000 15-80 1-139 0.1-11 2,500-15,000
Cattle 20-30 600-630 270-410 5,000-11,000 16,000-40,000
Cattle 1,500-12,000 16-140 ~ 3,000-11,000
Cattle 1,400-12,000 66-1,460b 265-3,400 800-7,500
Owens and Griffin,
1968
Hells et al.,1970
Norton and Hanson,
1969
Loehr, 1969b
Miner et al.,1966
Loehr, 1969a
Miner, 1967
Townshend et al.t
1970
^Volatile solids.
Total phosphorus as POt,.
Source: Porcella, D.B. and A. B. Bishop, Comprehensive Management of Phosphorus Water Pollution,
Ann Arbor Science Publishers, Ann Arbor, 1977.
-------
The groundwater is also very much impacted. For example, measurements
at Boise River indicate that the specific conductivity, a measure of the
salt content, is higher during the fall and winter months than during the
irrigation season. This is judged to be due to lower river flows during
this period than during the irrigation season, making groundwater contribu-
tions relatively higher. This implies that the groundwater's salt concentra-
tion must have reached a significant level that impairs beneficial uses such
as drinking water supply.
Transport by Sediment
Nutrients. The second major transport mechanism is sediment. Soil
particles act as carriers of nutrients and pesticides, especially phosphorus
and organic nitrogen and those pesticides with a high adsorption partition
coefficient. The finer soil particles, being transported the longest distance,
have a higher capacity per unit of sediment to adsorb phosphorus and nitrogen;
also, organic matter tends to be associated with the fine particles. This
means that the transported sediment is richer in phosphorus and nitrogen
than the original soil, a phenomenon known as 5enrichment." Predicting
the receiving water impacts of sediment-bound pollutants is complicated by
their relatively low bio-availability, compared with soluble pollutants (see
below).
The ratio of the particle fraction in the sediment to that in the original
soil is called the "enrichment ratio" (ER). The enrichment ratio for clay
(ER ) is usually greater than 1. Although not much is known about organic
matter enrichment (ER ), one can speculate that the transport of organic
matter is much the same as that of clay due to its characteristic low density.
According to the recent Cornell study,* one can generalize that the
enrichment ratio for clay or organic matter increases with an increase of
soil detachment due to raindrop splash relative to that due to overland flow.
These enrichment ratios will also increase as transport energy decreases.
The obvious shortcoming of the enrichment ratio concept is that it does not
indicate the actual amount of clay in the sediment load. A low ER for sedi-
ment originating from a clayey soil may very well mean a much greater amount
of clay in the sediment than a high ER from a sandy soil. A high ER for
a small sediment load may mean less clay than a low ER for a big sediment
load. In general, enrichment ratios of 2 to 6 are cited for first estimates.
Pesticides. Sediment-bound pesticides adsorb preferentially on smaller
soil particles (like nutrients) because of these particles's high surface
area per unit weight. Since the small particles are transported greater
distances than coarser material in runoff, these pesticides might reach the
receiving water. Rill and sheet erosion, which primarily involve surface
soil, tend to favor movement of the most strongly adsorbed pesticides. Higher
pesticide concentrations in eroded material do not necessarily mean that
*Haith, D.A. and R.C. Loehr (eds.), "Effectiveness of Soil and Water Conser-
vation Practices for Pollution Control," prepared by Cornell Oliversity for
U.S. EPA, Athens, GA, October 1979 (EPA-600/3-79-106).
26
-------
gross losses will be greater in the sediment than in runoff water, since
the amounts of water moved are so much greater.
Pollutant Impacts on Receiving Waters
A few general aspects of pollutant impacts on receiving water are:
The impact of type and amount of pollutants on the receiving water
quality is influenced by the type of receiving water.
An individual pollutant's effect depends on its availability.
Primary and secondary effects have to be distinguished e.g., nutri-
ents might cause algal bloom (eutrophication), which might then lead
to a significant depletion of oxygen in the hypolimnion of a lake.
In Table 4, we have summarized those pollutants from agricultural activi-
ties which could have an impact on receiving waters. If agricultural nonpoint
pollutants were combined with point discharges of residential, commercial,
and industrial activities, all entries of the matrix would have to be filled.
Below, we discuss briefly the impacts of the following pollutants; first
in general, and then with reference to the various receiving waters:
sediment (sand, clay, silt)
phosphorus (total soluble and extractable particulate)
nitrogen
organic load (BOD)
pesticides
bacteria
salt
Review of Individual Pollutants
Sediment and Nutrients. Once sediment reaches streams and lakes, it
causes two major types of water quality problems those caused by the sedi-
ment itself and those created by pollutants adsorbed to sediment. The direct
effects of sediment are benthic build-up impacting aquatic life and changing
hydraulic profiles, and high turbidity (fish kills, reduction photosynthesis
by aquatic plants, treatment costs of water supply). As one of the primary
transport mechanisms for pollutants such as pesticides and nutrients, the
interrelationship of pediment-attached nitrogen and phosphorus and dissolved
chemicals in surface wafeers~is"not' well understood. ' GeheralHyV sediment
found in surface waters resembles the soils from which it came, but it normally
has higher silt, clay, and organic matter content, which we have called en-
richment. Thus enrichment means that sediments delivered to surface waters
have higher concentrations of nitrogen and phosphorus than do the original
soils. However, because sediment-bound nutrients are isolated from the water
column due to sedimentation, their potential impacts on receiving water pro-
ductivity are not fully expressed.
Sediment also acts as a "scavenger" in the receiving water. Soluble
phosphorus and other uncharged pollutants are removed from waters by attachment
27
-------
TABLE 4. POTENTIAL IMPACT OP INDIVIDUAL POLLUTANTS FROM AGRICULTURAL ACTIVITIES ON RECEIVING WATERS
to
Impacted Receiving
Water
Lake/Re servoir
Small Stream
River
Great Lakes (Bay)
Grouhdwater
Parameters
Sediment P N BOD
X XXX
X (X) X X
(X) (X)
(X) (X)
X (X)
Pesticides
X
X
(X)
(X)
X
Pathogens
X
X
(X)
(X)
(X)?
Salt Remarks
(X) largely cumulative
impact
X largely transient
impacts
(X) largely transient
impacts
X slow transition; also
cumulative
Note: (X) indicates the impact is not great.
-------
to the sediment particles. Input of nutrients into receiving waters might
cause a significant increase in bioraass (algae). Various factors, such as
temperature, available P and N, C, and light penetration, determine the de-
gree of biomass development (eutrophication), which is generally measured in
grams chlorophyll-a/m^. The degree of biomass development is estimated by
the "Trophic State" Index. While carbon may be limiting in special circum-
stances/ eutrophication is generally limited by P, N, and/or light. Thus
the degree of sedimentation in a receiving water might influence the avail-
ability of nutrients as well as the availability of light.
Preliminary data analysis should indicate whether P is more likely than
N to limit primary production in lakes or reservoirs not subject to point
source discharge. N/P requirements for algal growth generally range from
7 to 10. In many cases, P is more likely to be limiting to algal
growth in impoundments not subjected to point sources (or any sewer influ-
ence) . N becomes increasingly important as a limiting nutrient as portions
of loads from point sources/sewerage increase. This arises because natu-
rally occurring runoff is enriched in N, whereas culturally created pollu-
tion sources tend to be enriched in P relative to algal growth requirements.*
In discussing transport mechanisms/ we have indicated that nitrate is
not significantly adsorbed by sediment. This would tend to reduce the im-
portance of sedimentation as an N removal/trapping mechanism in the receiv-
ing water as compared with its importance to the removal of P. Thus the
amount of P attached to soil particles and suspended in runoff may indi-
cate the potential amount available for receiving water pollution. How-
ever, this amount gives little indication of the amount of material
immediately biologically available for aquatic plant growth. Estimating
the "bio-availability" of particulate P is not yet done according to a
unified theory; jrather,""various arbitrary measures are applied.. Rfimkens
and Nelson** consider the NH^F/HCL extractable portion of the particulate
P (Bray P) as the amount of particulate P that is available, and they
assume that the remaining inorganic and organic particulate forms are un-
available to support algal growth in downstream impoundments. In their
methodological study of the Black Creek area in Indiana, Meta Systems Inc***
concluded that extractable and total particulate P data from soils in the Black
Creek area (collected by Sommers et al.)t generally support Taylor's suggestion
*For example, the N/P ratio for rainfall in the Northeast averages about
62, while the N/P ratio for sanitary sewage ranges from 2 to 8.
**Rflmkens, M.J.M. and D.W. Nelson, "Phosphorus Relationships in Runoff
from Fertilized Soils," Journal of Environmental Quality, Vol. 3, No. 1,
1974, pp. 10-13.
***Meta Systems Inc, op. cit.
tSommers, L.E. et al., "Water Quality Monitoring in Black Creek Watershed,"
Environmental Impact on Land Use and Wate_r__Qual.ity» (J. Lake and T. Morrison,
eds.), 1975.'
29
-------
that about 10 percent of P in soils is available for aquatic plant growth.*
Thus it is clear that "bio-availability" is an important assumption it
is critical in evaluating the effects of erosion controls on eutrophication,
given the current state of knowledge.**
Thus the water quality problems ofsejlimejita^^cjn_jinj3_eut
might be expected~rrom sediments washed jntoreceiving water. A secondary
effect of eutrophication is that with~an~excessive amount'of algal growth,
bottom conditions of receiving waters deteriorate. Dissolved oxygen decreases
(due to the organic loads from algae and sediment sinking to the bottom),
and might cause a decrease in desirable fish species such as trout and an
increase in trash fish, which survive better in marginal conditions (see
discussion below).
Pesticides. When dissolved and sediment-bound pesticides enter a receiv-
ing water, they are distributed within the water in a manner and at a rate
that depend primarily on whether the chemicals are initially dissolved in
the water or adsorbed on particles of eroded soil suspended in the water.
A dissolved pesticide will be diluted in the larger volume of water and will
be subject to processes that dissipate it. In a flowing stream it will
simply be transported away from the point of entry, later to undergo degra-
dation or removal from the water. In a lake or impoundment, it may sorb or
concentrate in algae and aquatic vegetation, or it may attach to suspended
sediment and other particulates in the water such as bacterial floes, diatoms,
and general organic or inorganic fragmentary material. In either case, the
pesticide is eventually deposited on the bottom of the lake unless it is
chemically or biologically degraded before it reaches the bottom or is taken
up by living organisms. Highly soluble pesticides that are only weakly ad-
sorbed may be hydrolyzed or biologically degraded in solution at a rate that
depends on the types and numbers of microorganisms in the water.
Pesticides entering the receiving water adsorbed on sediment will first
distribute with the carrying sediment and then will equilibrate with the
remainder of the aquatic system.*** Sediments entering water bodies will
segregate on a particle-size basis: in a stream the fractionation will depend
on stream velocity? in a relatively stagnant receiving water the particles
will settle on the bed in decreasing order of particle size. The finer parti-
cles containing the highest concentrations of pesticides will be transported
farthest down a stream; in a lake they will settle last and remain at the
water-sediment interface. In large, thermally stratified lakes, density
currents may control the movement and mixing of incoming sediments and settling
may be very slow.
*Taylor, A.W., "Phosphorus and Water Pollution," Journal of Soil and Water
Conservation, Vol. 22, 1967, pp. 228-231.
**It becomes obvious that additional research is required. What happens
to the fractions that are not biologically available? Not much is known;
little more than guessing can be done about the rate at which it might become
available.
***Conditions in the water body such as pH, salinity and temperature may
affect the adsorption and desorption of the pesticide on the sediment.
30
-------
The persistence of the pesticide in the receiving water will depend
to a large extent on the specific type of pesticide. For example, an organo-
phosphorus hydrolyzes rapidly in the aquatic environment, while organochlor-
ines degrade very slowly but decompose at a quicker rate in anaerobic environ-
ments by microorganisms. This shows that it is difficult to generalize the
relationship of pesticide input and receiving water type.
Biomagnification might be anticipated as a problem. There is as yet
no extensive knowledge on this subject, although the Cornell study made
a few statements about it. Biomagnification has generally not been a problem
with most herbicides. Despite persistence and biomagnification occurring
in invertebrates, (for example atrazine residues), in a study of pond fauna
over a two-year period, no significant harmful effects were detected.* An
exception to the general rule of herbicide nonaccumulation is trifluralin,
which has been shown to magnify in aquatic food chains.
There are, however, problems with some herbicides and the organophosphate
and carbamate insecticides, in spite of their generally low levels of bio-
magnification.** Adverse effects have occurred at levels below the established
acutely "safe" level leptophos has been shown to produce delayed neutro-
toxic effects on chickens. Recent evidence that DBCP, a pesticide ingredient
previously thought to be safe, can cause sterility and possibly cancer in
humans at long-term, low-level exposures has brought into question the ade-
quacy of the entire procedure for testing pesticide safety.
Organic Material and Pathogens. Organic putrescible material in receiv-
ing waters is of no great concern in most agricultural land uses. In contrast
to point source pollution of sewage and other putrescible matter, nonpoint
source pollution generates only a relatively small input of putrescible matter,
and this load originates primarily from some confined feedlots (not usually
considered nonpoint source feedlots because of their size) and pasture areas
quite intensively used by animals. The pollutional impact of these sources
is measured in BOD (reflecting the oxygen needed to stabilize the receiving
water) and coliform. In an overloaded receiving water the dissolved oxygen
may become exhausted by the biochemical oxygen demand. In most situations
washoff of organic material from feedlots will not result in exhaustion of
DO, but when algae that feed off the feedlots1 nutrients die, sink down,
and decompose in the hypolimnion of lakes/reservoirs, the DO-household might
be strained to its limit. This situation occurs in highly eutrophic lakes
due to the thermal stratification in the summer and the resulting lack of
import of additional DO into those layers.
The presence of coliform organisms (occurring in the intestinal tracts
of warm-blooded species) is taken as an indication that pathogenic organisms
may also be present; their absence is an indication that the water contains
no disease-producing organisms. The coliform bacteria include the genera
Escherichia and Aerobacter. However, the use of coliforms as indicator organ-
isms is complicated by the fact that Aerobacter (and certain Escherichia) can
*Haith and Loehr, op. cit., p. 225.
**Ibid.
31
-------
grow in soil. Thus the presence of coliforms does not always mean contamina-
tion with human or animal wastes. Apparently Escherichia coli (E. coli)
are entirely of fecal origin. Since it is difficult to determine the presence
of E. coli to the exclusion of the soil coliforms, the entire coliform group
is used as an indicator of fecal pollution. Since tests have been developed
to distinguish among total coliforms, fecal coliforms (FC), and fecal strep-
tococci (FS), the use of the ratio of fecal coliforms to fecal streptococci
has been suggested to show whether the suspected contamination derives from
human or from animal wastes. It had been observed that the quantities of
fecal coliforms and fecal streptococci that are discharged by human beings
are significantly different from the quantities discharged by animals. Typical
data on the ratio of FC to FS counts for human beings and various animals
have been reported. In Table 5 the FC/FS ratio for domestic animals is less
than 1.0, whereas the ratio for human beings is more than 4.0. If ratios
are obtained in the range of 1 to 2, interpretation is uncertain.*
Use of the FC/FS ratio can be very helpful in establishing nonpoint
source pollution in rural and other areas where septic tanks are used. In
many situations where animal pollution is suspected on the basis of coliform
test results, the actual pollution may in fact be caused by malfunctioning
septic systems.
Review of Receiving Water Types
Depending on the type of receiving water into which the pollutant is
discharged, pollutant effects on water chemistry and aquatic ecology will
vary significantly. The water bodies discussed here small streams, rivers,
lakes, impoundments, bays and groundwater each have hydraulic, ecological,
and physical differences which affect the fate of pollutants. This discussion
deals primarily with the mechanisms of pollutant transport and transformation
of sediment and nutrients, and touches upon pesticides.
Small Streams/Rivers. In small streams/rivers** the fate of the entering
*The following constraints are imposed on the interpretations (see D.D.
Mara t Bacteriology for Sanitary Engineers, Churchill, Livingston (Edinburgh),
1974):
1) The sample pH should be between 4 and 9 to exclude any adverse effects
of pH on either group of microorganisms.
2) At least two counts should be made on each sample.
3) To minimize errors due to differential death rates, samples should
not be taken farther downstream than 24 hours of flow time from the
suspected source of pollution.
4) Only the FC count obtained at 44°C is to be used to com-
pute the ratio.
**We talk simultaneously about small streams and rivers under the assumption
that they can be largely differentiated by their water flow velocity and
volume (at average conditions), with small streams having larger velocity
but much smaller volume and generally shallower beds.
32
-------
TABLE 5. ESTIMATED PER CAPITA CONTRIBUTION OF INDICATOR
MICROORGANISMS FROM HUMAN BEINGS AND SOME ANIMALS
Average indicator
density/g
offeces
Animal
Chicken
Cow
Duck
Human
Pig
Sheep
Turkey
Fecal
coliform
10*
13
0.23
33.0
13.0
3.3
16.0
029
Fecal
streptococci
10s
3.4
1.3
54.0
3.0
84.0
38.0
2.8
Average
contribution/
capita-24 h
Fecal
coliform
10'
240
5,400
11,000
2,000
8JOO
18,000
130
Fecal
streptococci
10*
620
31,000
18,000
450
230,000
43,000
1300
Ratio
FC/FS
0.4
0.2
0.6
4.4
0.04
0.4
0.1
Note: gx 0.0022-Ib.
Source: Mara, D.D.f Bacteriology for Sanitary Engineers/ Churchill, Living-
ston (Edinburgh), 1974.
Note: Doran, J.W. , and D.M. Linn* in a recent article, state that the
FC/FS ratio for humans is 4.3; for cattle and other livestock and poultry
it is .104 to .421; while for rabbits, birds, and mice it is 0.0008 to 0.043.
*"Bacteriological Quality of Runoff Water from Pasture Land," Applied
Environmental Microbiology, Vol. 37, No. 5, 1979, pp. 385-391.
33
-------
sediments cannot be determined without knowing the characteristics of the
basin. Slope, water velocity, particle type and size all play an important
role in sediment transport. At best, a few generalities can be made:
As velocity 'increases, the distance of transport increases;
Larger and more dense particles settle out first;
Sedimentation occurs in pool areas;
Turbulence will act to maintain particles in suspension.
The interaction between water velocity and sediment transport is complex.
In general the rate of settling increases as the velocity decreases; if velo-
city increases, bank scour may also increase, and bed load may become sus-
pended, causing an increase in the total solids being transported in the
stream. Sedimentation might affect benthic life directly and impact aquatic
life through turbidity, but sedimentation problems can also occur as a result
of shifting substrates and thereby cause a loss of certain aquatic species.
Sedimentation is thus dynamic, which means that constant shifting of deposit-
ed materials might occur.
A general idea of where sedimentation problems will occur in a stream
can be determined from information on soil types and by locating pool areas
where water velocities decrease.
Much of the transport of nutrients in streams is closely associated
with sediment transport due to adsorption of phosphorus (and of some nitrogen)
onto soil particles. Particulate phosphorus and nitrogen will disperse ac-
cording to the mechanisms of particle dispersion. The dissolved fractions
will be absorbed by aquatic vegetation, adsorbed onto suspended or bank sed-
iments, or carried downstream. Factors affecting these processes are:
water velocity and turbulence
microorganisms and vegetation
turbidity
temperature
particle size and type
channel characteristics
temperature
Assuming no additional inputs of nutrients, concentrations decrease
as materials proceed downstream due to uptake by organisms and dilution or
by dilution from unpolluted interflow.
The extent to which nutrients settling out affect the stream depends
on the particular situation. During storm conditions with accompanying high
flows, sediment material often becomes resuspended, essentially acting as
a new input into the stream.
Very little is known about the impact of pesticide concentrations on
receiving waters. Incidents of surface water contamination by the organo-
chlorine insecticides are still occurring, and apparently several lakes in the
Midwest have been closed to fishing because of accumulations of toxic
34
-------
materials.* It is not known whether this contamination is due to the continu-
ing runoff and erosion of soils carrying these persistent pesticides. Some
researchers have noted, however, that concentrations of organochlorine resi-
dues are high in areas of high sediment losses and turbidity of surface waters.
Thus, for example, the presence of pesticide residues was established in
the Mississippi River at New Orleans and in every major watershed in the
State of Iowa.**
Lakes and Reservoirs
Determining the rate of sedimentation in lakes and reservoirs is diffi-
cult. Sampling inputs from diffuse sources and from output points gives
a clue about the sediment balance, but does not indicate the location of
sedimentation, which can only be found by instream investigations. However,
it has been observed that due to the abrupt decrease in velocity, the initial
point of entry is usually a major sedimentation point. Current pattern,
eddies, and storm events exhibiting flushing effects further determine the
other areas affected.
A few general trends in reservoir sedimentation have been observed:
Reservoirs with greater drainage area/surface area ratios have shorter
lifetimes;
Sediment trapping efficiency increases with hydraulic detention time
(volume/annual flow);
Particles : tend to be deposited in a gradation of particle sizes
along the longitudinal axis of the reservoir. Coarser and heavier
particles are dropped .in the headwater and finer sediments are depos-
ited toward the dam. This is affected by water level; temperature
and dissolved minerals; mineral composition of the sediments, especial-
ly clay-sized fraction; volume relationship of reservoir storage
capacity and influent water; configuration of basin; and amount of
sediment previously deposited.
Eutrophication*** is a problem, particularly exacerbated by the environ-
mental conditions of a lake/impoundment when high nutrient inputs occur.
*Haith and Loehr, op. cit., p. 217.
**Richard, J., G. Junk, et al., "Analysis of Various Iowa Waters for Select-
ed Pesticides: Atrazine, DDT and Dieldrin 1974," Pesticide Monitoring
Journal, Vol. 9, No.3, pp. 117-223, cited in Haith and Loehr, op. cit^
***Traditional strategies for classifying lakes with respect to eutrophication
have relied primarily upon subjective assessments of one or more types of
water quality or biological characteristics. Recently, the increased avail-
ability of data has made it possible to develop more objective criteria for
ranking and classifying lakes at a regional level on the basis of observed
lake conditions or the factors governing them, such as nutrient loading,
hydrology, and morphometry. Lack of an objective basis for specifying stan-
dards or criteria with regard to lake or impoundment water quality has arisen
partially out of the fact that water quality concerns are related to beneficial
35
-------
However, the dynamics of nutrients in lakes and reservoirs are not well under-
stood. In an essentially closed lake system with a long hydraulic detention
time, the nutrients entering remain in the system; they are either utilized
by the vegetation, settle out, or remain in solution. Many reservoirs have
short detention times and cannot be considered closed. In general, soluble
nitrates and phosphates are readily available for plant growth. This is
also true of ammonia, which can be utilized as a nitrogen source by many
types of aquatic vegetation. In most receiving waters influenced by nonpoint
sources, ammonia-N concentrations are at extremely low levels due to high
rates of biological uptake, oxidation to nitrate, and/or adsorption to sedi-
ments. Nutrients associated with particulate matter tend to settle out and
become part of the sediments, but a number of factors may affect this process.
For example, turbulence will tend to keep particles suspended.
To what extent nutrients isolated in the sediment from the water column
are available to aquatic life is not definitely known, especially, the extent
of P-release under anoxic conditions. Phosphates tend to form very stable
complexes with elements such as iron and aluminum. Some evidence suggests
that sediment runoff from particular soils into reservoirs and lakes may
actually reduce dissolved ortho-phosphate levels by forming complexes that
precipitate due to the rapid equilibrium between water and sediments. However,
these sediments could later supply phosphorus to aquatic organisms when the
sediments are stirred up by turnover, turbulence, or even by bottom feeding
fish such as carp and catfish. Bottom sediment in particular are easily resus-
pended due to a flushing effect of storm events in reservoirs. It should
also be recognized that rivers characteristically carry higher concentrations
of nutrients than do lakes. Thus by damming a river, these higher concentra-
tions are retained in the system. The addition of nonpoint sources of nutri-
ents to such a system could accentuate the buildup if flows are never high
enough to flush the system.
Potentially long retention times might lead to long-term effects from
persistent pesticides. The total elimination of aquatic life in those southern
lakes that captured the pesticides from cotton field drainage is well known.
But as in the case of streams/rivers, not much is known about the environmen-
tal impacts at different levels.
use and do not always correspond with traditional trophic state criteria.
While some states may have considered or be considering phosphorus standards,
the nutrient in itself does not hinder water use. It is the indirect effects
of the nutrient on such water quality aspects as transparency, odor, and
dissolved oxygen that are of concern from a water use standpoint. Recent
theoretical and empirical developments indicate that the effects of phos-
phorus supply on primary production and water quality vary with impoundment
morphometric and hydrologic characteristics, and depend upon supplies of
other nutrients. Thus it may not be advisable to establish universal phos-
phorus standards. Standards should be based on those water quality responses
that are of direct concern to water use (See for example, W.W. Walker, "Use
of Hypolimnetic Oxygen Depletion Rate as a Trophic State Index for Lakes,"
Water Resources Research, Vol. 15, No. 6, December 1979).
36
-------
Bay Areas. The bay areas of the Great Lakes are considered, for our
purposes, infinite sinks. Most likely, it will be impossible to detect sig-
nificant differences of water quality parameters and water quality impacts
due to changing agricultural practices in any reasonable length of time.
Groundwater.* Excessive nitrate and salt concentration in some agricul-
tural land use areas are the most reported groundwater impacts due to para-
meters we are here concerned with. They are especially important for ground-
water used as a drinking water source. There has also been some pollution of
groundwater by pesticides due to leaching. The rate at which pesticides
leach through the soil profile is influenced by many of the same factors
influencing losses in runoff namely, adsorption/desorption characteristics
of the pesticide, chemical reactions, solubility, rate of application, ante-
cedent soil moisture, soil structure, and flow velocity. Such pesticides as
2,4-D, atrazine, cyanazine, dicamba, dimethpate, chloramben, dinoseb, monuron,
and methoxychlor all exhibit a greater propensity for movement in the soil.
However, under normal conditions, extensive leaching, and subsequent contami-
nation of groundwater are unlikely** although some pesticides, such as atra-
zine, have been identified*** in groundwater during periods of application
and in wet years. For example, in Iowa the pesticides atrazine, DDE, and
dieldrin contaminated all water originating from shallow wells located in the
alluvial plains of contaminated rivers.
It should be noted here that apparently the state of pesticide analysis
under field conditions is not advanced enough to allow reliable monitoring
of low-level concentrations in runoff and surface receiving waters and ground-
water. This means that there is existing pollution that could produce little
understood chronic and sublethal toxicity effects.
Agricultural Land Use and Water Quality Impact
On the basis of the above discussion. Table 6 shows, in simplified form,
some agricultural land1 use/water quality impacts (as triggered by the "base-
line" performance of these activities).
Other factors that influence the severity of water quality impacts
should be mentioned. They are:
*The quality and quality changes of groundwater are the most difficult
to assess. This is true because, on the one hand data on groundwater quality
are inadequate, and on the other hand, it is most problematic to take ground-
water samples in situ and then extrapolate results at one geographic point
to a larger volume of a system previously defined.
**Haith and Loehr, op. cit., p. 211.
***Leaching will be more of a problem in areas of pesticide disposal, areas
with water tables not far beyond the root zone of crops, and in areas with
sandy soils containing little organic matter or clay to bind pesticides as
they percolate through the soil profile. Groundwater contamination by pesti-
cides may also occur when solid-phase pesticides are washed down deep cracks
which can appear when heavy rains follow periods of drought. Haith and Loehr,
op. cit.
37
-------
TABLE 6. AGRICULTURAL LAND USE/WATER QUALITY IMPACTS
Receiving Water
Land Use Activity
Cropland
nonirrigated
Orchard/Vineyards
Pastureland/Range
Problems
oo
Homestead
Sedimentation
Eutrophication(incl.
hypol. DO-D)
Pesticide
JL trate_
Sedimentation
Salinity
Butrophication(incl.
hypol. DO-D)
Pesticide
itrate
itrophication
'esticide
Titrate
Eutrophication
Pathogens
BOD
Sedimentation
BOD
Butrophication
Pathogens
titrate
BOD
Pathogens
Eutrophication
Note: (X) indicates the impacts are not great.
? indicates the situation is largely unknown.
-------
slope and length/topography
use intensity
climate/precipitation (region)
soil type
hydrologic group
erodibility
drainage density
The most obvious factors are slope/length and use intensity. In general,
the steeper and longer the slopes, the worse the erosion condition that can
be generated; the higher the use intensity, the greater the overall pollution
potential. Several concerns relate to precipitation: intensity, duration,
frequency, and snowfall and melting. If a field has not yet grown any pro-
tective cover and the more intensive a rain is, the higher will be the erosion;
the closer a rain event follows the date of pesticide and fertilizer applica-
tion, the more material applied will be washed off and/or percolated. These
effects will be accentuated if the soil is highly erodible and the materials
applied are largely sediment-bound. Clearly, seasonal and regional differ-
ences play a role here. High intensity rain is more detrimental on uncovered
ground in the early planting season than on covered ground near harvesting
time; rain characteristics vary significantly across the United States. Drain-
age density, i.e., the degree of development of runoff carrying channel capa-
city, affects the transport of material to the mainstream. The higher the
density, the greater the potential of material transport to the receiving
water.
It should be pointed out here that the water quality impact of the events
described above depend at least on the time of year and type of receiving
water. If there is a river or even a small stream under relatively high
flow conditions, the impacts might be marginal. If there is a low flow season
and the final receiving water is a reservoir, the impact could be quite sig-
nificant. Again the land/water interplay is important.
Soils have various characteristics that are of interest: natural nitro-
gen and phosphorus content; organic matter content; distribution of sand,
silt, and clay; structure; and permeability. These factors contribute to
the erodibility of a soil, but they also influence a soil's general control-
lability (See the control categories outlined in Section 2), its overall
eutrophication potential, its carrying/adsorption potential, and the leachate
potential for nonadsorptive parameters. The hydrologic group indicates the
drainage characteristics something closely linked to the soil type. This
might have some influence on the release of pollutants to the surface receiv-
ing water. For example, it could be hypothesized that runoff carries much
of the soluble nitrate fractions from poorly drained soils, while a much
larger fraction infiltrates from a well-drained soil, resulting in pos-
sible nitrate contribution to surface water via interflow. Thus this phased
input of nitrate could have some beneficial impact on the surface water qual-
ity of a stream since no heavy slug of nitrate enters the water at one instant.
But in the case of a lake/reservoir as the ultimate receiving water, the
phased input might play only a minor role if the input is delayed up to the
cold season with its reduced biological activity and higher dilution rates.
39
-------
This brief discussion indicates the inadequacy of a simple impact matrix
(such as Table 6) that neglects the localized conditions, but it also makes
us aware of the lack of knowledge in this field and it cautions us that cer-
tain standard agricultural practices are not simply transferrable across
the nation without a detailed look at the specific problem.*
Considerations in Selecting Abatement Practices/Measures
Tables 7-15 list some of the practices/measures that have been suggested
for the various agricultural land use activities. Host practices belong
to category 1, but we realize (along the lines of our previous discussions)
that certain pollutants, such as pesticides, can only be controlled through
a combination of measures from different categories. The Cornell Study re-
viewed the measures suggested for non-irrigated cropland and classified their
effectiveness as soil and water conservation measures for control of certain
pollutant types (Table 16). Given our emphasis on water quality impacts,
the usefulness of such a classification should be investigated. Our discus-
sion had concluded that the degree of water quality impact caused by agricul-
tural land use depends on the type and intensity of agricultural land use
on localized factors, and, in particular, on the type and size of the receiv-
ing water. That means that every control measure should be tuned to the
pollutant (s) and the resulting water quality problem. Thus the Cornell listing
raises three questions:
1) Does the classification of land-based effectiveness provide us with
an initial idea on which measures to select from?
2) Might a preselection based on land-based effectiveness lead to a
wrong approach due to synergistic control effects for different para-
meters?
3) Do there exist any combinations of measures that are equally effective
and possibly more desirable?
Before these questions can be answered, the development of technical prefer-
ences should be explained.
Technical evaluation (Figure 6) has to be performed in a "2-track" system.
First, after the water quality problem has been described, the sources must
be detected and their potential modifications through practices belonging
to categories 1-5 must be analyzed. Second, the desirable water quality
must be identified and the necessary reduction in pollutant input necessary
to achieve the water quality goal computed (including possible instream
measures). Third, the two tracks must be meshed in order to determine which
practices/measures are capable of meeting the water quality goal. Finally,
the feasibility of technically acceptable measures must be determined in
socio-economic and institutional terms.
This approach to the technical evaluation process helps to answer the
*The discussion of these characteristics also provides a first insight
into fundamentally different monitoring requirements for different land use/
water quality interfaces in different parts of the United States.
40
-------
TABLE 7. PRINCIPAL TYPES OF CROPLAND EROSION CONTROL PRACTICES AND THEIR HIGHLIGHTS*
Category No.
Erosion Control
Practice
Practice Highlights
1
1
1
1
1
1
2
1
1
1
El
E2
E3
E4
ES
E6
E7
E8
E9
E10
No-till plant in prior-
crop residues
Conservation tillage
Sod-based rotations
Meadowless rotations
Winter cover crops
Improved soil fertility
Timing of field
operations
Plow-plant systems
Contouring
Graded rows
Most effective in dormant grass or small grain; highly effective in crop residues; minimizes
spring sediment surges and provides year-round control; reduces man, machine and fuel re-
quirements; delays soil warming and drying; requires more pesticides and nitrogen; limits
fertilizer- and pesticide-placement options; some climatic and soil restrictions.
Includes a variety of no-plow systems that retain some of the residues on the surface; more
widely adaptable but somewhat less effective than E 1 ; advantages and disadvantages generally
same as E 1 but to lesser degree.
Good meadows lose virtually no soil and reduce erosion from succeeding crops; total soil loss
greatly reduced but losses unequally distributed over rotation cycle; aid in control of some
diseases and pests; more fertilizer-placement options; less realized income from hay years;
greater potential transport of water soluble P; some climatic restrictions.
Aid in disease and pest control; may provide more continuous soil protection than one-crop
systems; much less effective than E 3.
Reduce winter erosion where corn stover has been removed and after low-residue crops;
provide good base for slot-planting next crop; usually no advantage over heavy cover of
chopped stalks or straw; may reduce leaching of nitrate; water use by winter cover may reduce
yield of cash crop.
Can substantially reduce erosion hazards as well as increase crop yields.
Fall plowing facilitates more timely planting in wet springs, but it greatly increases winter and
early spring erosion hazards; optimum timing of spring operations can reduce erosion and
increase yields.
Rough, cloddy surface increases infiltration and reduces erosion; much less effective than £ 1
and E 2 when long rain periods occur; seedling stands may be poor when moisture conditions
are less than optimum. Mulch effect is lost by plowing.
Can reduce average soil loss by 50% on moderate slopes, but less on steep slopes; loses
effectiveness if rows break over; must be supported by terraces on long slopes; soil, climatic,
and topographic limitations; not compatible with use of large farming equipment on many
topographies. Does not affect fertilizer and pesticide rates.
Similar to contouring but less susceptible to row breakovers.
*Source: U.S. EPA/USDA, Pollution from Cropland, Vol. I, (Table 12).
-------
TABLE 7. (CONT.)
Category No.
Erosion Control
Practice
Practice Highlights
to
1
3
3
j
1
1
1
3,4
-3
3
1,2
£11
E12
E13
E14
E15
E16
E17
Contour strip cropping
Terraces
Grassed outlets
Ridge planting
Contour listing
Change in land use
Other practices
ii
n
n
n
Rowcrop and hay in alternate 50- to 100-foot strips reduce soil loss to about 50% of that
With the same rotation contoured only; fall seeded grain in lieu of meadow about half as
effective; alternating corn and spring grain noteffective;area must be suitable for across-slope
farming and establishment of rotation meadows; favorable and unfavorable features similar to
E 3 and E 9.
Support contouring and agronomic practices by reducing effective slope length and runoff
concentration; reduce erosion and conserve soil moisture; facilitate more intensive cropping;
conventional gradient terraces often incompatible with use of large equipment, but new
designs have alleviated this problem ; substantial initial cost and some maintenance costs.
Facilitate drainage of graded rows and terrace channels with minimal erosion; involve establish-
ment and maintenance costs and may interfere with use of large implements.
Earlier warming and drying of row zone; reduces erosion by concentrating runoff flow in
mulch-covered furrows; most effective when rows are across slope.
Minimizes row breakover; can reduce annual soil loss by 50%; loses effectiveness with post-
emergence corn cultivation ; disadvantages same as E 9.
Sometimes the only solution. Well managed permanent grass or woodland effective where
other control practices are inadequate; lost acreage can be compensated for by more intensive
use of less erodible land.
Contour furrows,
Diversions
Subsurface drainage
Land forming
Closer row spacing
-------
TABLE 8. PRACTICES FOR CONTROLLING
DIRECT RUNOFF AND THEIR HIGHLIGHTS*
Category No. Runoff Control Practice
Practice Highlights
1
111 ' 1 "
1
1
1
1
2
1
1
3
1
3
3-
1
1
1
3,4
4
Rl
R2
R3
R4
RS
R6
R7
R8
R9
RIO
Rll
R12
R13
R14
R1S
R16
R 17
RIB
No-till plant in prior crop residues
Conservation tillage
Sod-based rotations
Meadowless rotations
Winter cover crop
Improved soil fertility
Timing of field operations
Plow plant systems
Contouring
Graded rows
' Contour strip cropping
Terraces
Grassed outlets
Ridge planting
Contour listing
Change in land use
Other practices
Contour furrows
Diversions
Drainage
Landforming
Construction of ponds
Variable effect on direct runoff from substantial reductions to
increases on soils subject to compaction.
Slight to substantial runoff reduction.
Substantial runoff reduction in sod year; slight to moderate
reduction in rowcrop year.
None to slight runoff reduction.
Slight runoff increase to moderate reduction.
Slight to substantial runoff reduction depending on existing
fertility level.
Slight runoff reduction.
Moderate runoff reduction.
Slight to moderate runoff reduction.
Slight to moderate runoff reduction.
Moderate to substantial runoff reduction.
Slight increase to substantial runoff reduction.
Slight runoff reduction.
Slight to substantial runoff reduction.
Moderate to substantial runoff reduction.
Moderate to substantial runoff reduction.
Moderate to substantial reduction.
No runoff reduction.
Increase to substantial decrease in surface runoff.
Increase to slight runoff reduction.
None to substantial runoff reduction. Relatively expensive.
Good pond sites must be available. May be considered as a
treatment device.
*Erosion control practices with same number are identical.
Source: See Table 7.
43
-------
TABLE 9. PRACTICES FOR THE CONTROL OF NUTRIENT LOSS FROM AGRICULTURAL
APPLICATIONS AND THEIR HIGHLIGHTS
Category
2
2
1
1
1
1
1,2
1
1 ,
1
2
No.
Nl
N2
N3
N4
N5
N6
N7
N8
N9
N10
Nil
Nutrient Control Practice
Eliminating excessive
fertilization
Leaching
Timing nitrogen applica-
tion
Using crop rotations
Using animal wastes for
fertilizer
Plowing-under green
legume crops
Using winter cover crops
Controlling fertilizer
release or transformation
Practice Highlights
May cut nitrate leaching appreciably,
reduces fertilizer costs; has no effect
on yield.
Control
Reduces nitrate leaching; increases
nitrogen use efficiency; ideal timing
may be less convenient.
Substantially reduces nutrient inputs;
not compatible with many farm enter-
prises; reduces erosion and pesticide
use.
Economic gain for some farm enter-
prises; slow release of nutrients;
spreading problems.
Reduces use of nitrogen fertilizer;
not always feasible.
Uses nitrate and reduces percolation;
not applicable in some regions; reduces
winter erosion.
May decrease nitrate leaching; usually
not economically feasible; needs
additional research and development.
Control of Nutrients in Runoff
Incorporating surface
applications
Controlling surface
applications
Using legumes in haylands
and pastures
Decreases nutrients in runoff; no
yield effects; not always possible;
adds costs in some cases.
Useful when incorporation is not
feasible.
Replaces nitrogen fertilizer; limited
applicability; difficult to manage.
Control of Nutrient Loss by Erosion
Timing fertilizer plow- [Reduces erosion and nutrient loss; may
down (be less convenient.
Source: See Table 7.
44
-------
TABLE 10. PRACTICES FOR THE CONTROL OP PESTICIDE LOSS
FROM AGRICULTURAL APPLICATIONS AND THEIR HIGHLIGHTS
Category
2
1,2
1
1,2
2
2
1
2
2
2
1,2
1
No.
PI
P2
P3
P4
P5
P6
P7
P8
P9
P10
Pll
P12
Pesticide Control Practice
Practice Highlights
Broadly Applicable Practices
Using alternative pesti-
cides
Optimizing pesticide
placement with respect to
loss
Using crop rotation
Using resistant crop
varieties
Optimizing crop planting
time
Optimizing pesticide
formulation
Using mechanical control
methods
Reducing excessive treat-
ment
Optimizing time of day
for pesticide application
Applicable to all field crops; can
lower aquatic residue levels; can hin-
der development of target species
resistance .
Applicable where effectiveness is
maintained; may involve moderate cost.
Universally applicable; can reduce
pesticide loss significantly; some in-
direct cost if less profitable crop is
planted.
Applicable to a number of crops; can
sometimes eliminate need for insecti-
cide and fungicide use; only slight
usefulness for weed control.
Applicable to many crops; can reduce
need for pesticides; moderate cost
possibly involved.
Some commercially available alterna-
tives; can reduce necessary rates of
pesticide application.
Applicable to weed control; will re-
duce need for chemicals substantially;
not economically favorable.
Applicable to insect control; refined
predictive techniques required.
Universally applicable; can reduce
necessary rates of pesticide applica-
tion.
Practices Having Limited Applicability
Optimizing date of pesti-
cide application
Using integrated control
programs
Using biological control
methods
Applicable only when pest control is
not adversely affected; little or no
cost involved.
Effective pest control with reduction
in amount of pesticide used; program
development difficult.
Very successful in a few cases; can
reduce insecticide and herbicide use
appreciably.
45
-------
TABLE 10. (CONT.)
Category
2
2
1
1
No.
P13
P14
P15
Pesticide Control Practice
Using lower pesticide
application rates
Managing aerial applica-
tions
Planting between rows in
minimum tillage
Subsurface application
Practice Highlights
Can be used only where authorized;
some monetary savings.
Can reduce contamination of non-target
areas.
Applicable only to row crops in non-
plow based tillage; may reduce amounts
of pesticides necessary.
Reduces concentrations of pesticides
in runoff.
Source: See Table 7.
46
-------
TABLE 11. SOME SEDIMENT CONTROL PRACTICES FOR IRRIGATED AGRICULTURE
(Estimated Sediment Loss Reduction for Selected Control Practices)
Control
Category
2
3
4'
4
1
Control
Practice
Flow Cutback
Vegetative Buffer Strip
Sediment Pond
Mini-Basins
Sprinklers
Sediment Loss
Reduction (%)*
30
50
67
90
100
*Baseline: Conventional Furrow Irrigation
Source: D.W. Fitzsimmons, et. al., Evaluation of Measuring for
Controlling Sediment and Nutrient Losses from Irrigated
Areas, EPA-600/2-78-138, July 1978.
TABLE 12. ANIMAL HOLDING CONTROL PRACTICES
Control Category Control Practices
4 Diversion
4 Retention Points
1 Confinement
2 Proper Location
4 Evaporatiion Points
4 Land Disposal
47
-------
TABLE 13. CONTROL PRACTICES/MEASURES (ORCHARD/VINEYARDS)
Control Category Control Practices
1 Chiseling and Subsoiling
1 Drainage Land Grading
3 Drainage System Structure
3,4 Diversion
4 Ponding
2 Irrigation Water Management
2 Nutrient Management
2 Pesticide Management
4 Access Road Protection
TABLE 14. CONTROL PRACTICES/MEASURES (RANGE AND PASTURE)
Control Category Control Practices
2 Range and Pasture Management and Use Plan
3 Spring Development
3 Ponding (including protection against losses)
4 Water Control Structures
5 Fencing
4 Stock Trails and Walkways
48
-------
TABLE 15. CONTROL PRACTICES (HOMESTEAD)
Control Category Control Practices
3 Septic System
2 Maintenance Schedule (Septic System)
3 Sizing (of Septic System)
3 Dry Wells (runoff from impervious area)
1 Alternative Waste Systems (compost toilets/
double plumbing for waste and grey water)
49
-------
TABLE 16. EFFECTIVENESS OF SOIL AND WATER CONSERVATION PRACTICES IN CONTROLLING POLLUTANTSJ
Control of
SWCP
Strongly Adsorbed
Soil, Organic N, Organic P Availoble Phosphorus
Paraquat
Moderately
Adsorbed
Non-Adsorbed
Effective
no tillage
conservation tillage
ridge planting
sod-based rotations
cover
sod-based rotations
ridge planting
sod-based rota-
tions
Moderately
Hffectivc
Slightly
Effective
Not Effective
or Very
Slightly
Effective
graded rows
contour listing
terraces
contour farming
filter strips
diversions
grassed waterway'
drainage
graded rows
terraces
no tillage
conservation tillage
contour farming
ridge planting
contour listing
cover crops
diversions
filter strips
grassed waterway-*
drainage
contour listing
no tillage
conservation
tillage
contour farming
terraces
graded rows
cover crops
diversions
grassed waterway
filter strips
drainage
sod-based
rotations
contour farming
ridge planting
contour listing
no tillage
conservation
tillage
graded rows
cover crops
diversions
terraces
grassed waterway
filter strips
drainage
These evaluations have been made for general situations. The site specific effectiveness of a
. particular practice as it fits into a total farming system might be different than indicated
-Not effective for pesticides.
' In combination with other practices, grassed waterways are affective.
Source: Haith, D.C. and R.C. Loehr (eds.). op cit.
-------
FIGURE 6. TECHNICAL PROJECT EVALUATION
INFORMATION:
RECEIVING WATER
TYPE
HYDROLOGY
CLIMATE
MORPHOMETRY
IDENTIFY SOURCES
OF WQ PROBLEM
(ANALYSIS/INSPEC-
TION)
LIST APPROPRIATE
CONTROL PRACTICES/
MEASURES (BY CATE-
GORY)/COMBINATIONS
ANALYZE THE REDUC-
TION POTENTIAL FOR
PRIME POLLUTANTS IN
TIME * SPACE (SYN-
ERGISTIC EFFECTS)
NO
IDENTIFY WQ PROBLEM
(PRIMARY & SECONDARY
EFFECTS/IMPAIRMENT
OP WATER USES)
IDENTIFY MAJOR POLLU-
TANTS/PARAMETERS RE-
SPONSIBLE FOR PROBLEM
(STATISTICAL CHARAC-
TERISTICS/SEASONS/
OTHERS)
ANALYZE WQ IMPACT £ RE-
DUCTION NEED OF POLLU-
TANT INPUT TO ACHIEVE
DESIRABLE WQ LEVEL
|(POLLUTANTS/COMBI NATIONS/
IN-STREAM MEASURES)
REDUCTION OF POLLUTANTS
TO NECESSARY LEVEL FEA-
SIBLE WITH PRACTICES?
YES
EVALUATE FEASIBLE REDUC-
TION STRATEGIES WITH RE-
SPECT TO IMPLEMENTATION
CAPABILITY (IN USDA/SC?.
SENSE)
IMPLEMENTATION
M/E
51
-------
above questions about the usefulness of the Cornell listing of practices
for water quality planning. Selection of those practices is only helpful
if soil and water conservation practices (SWCP) are considered as remedial
actions. For example, if erosion and sedimentation cause light-limited eutro-
phication in a lake/ employing the "most effective" SWCP might result in
worse eutrophic conditions because of the potential shift from light- to
nutrient-limited eutrophication. Thus SWCPs must be tuned to the water quality
problem, which means that "highly effective" SWCPs might not be highly effec-
tive water improvement measures. Combining different practices/measures
from different control categories will produce in most cases a few equally
effective load reduction measures; the ease of their implementation and M/E
should determine the final choice.
It becomes obvious that quantifying the land use/water quality relation-
ships is technically a difficult problem. Therefore, only the severest water
quality problems and only those projects in which M/E of certain practices
is desirable warrant this level of detail. This means that at this time
the land use/water quality relationship cannot be quantified for all areas
in all states.
Examples of Choosing Abatement Practices/Measures
For our example, we will assume that an agricultural land use affects
a lake in all four water quality problem areas: sedimentation, eutrophication,
pesticides, NO,. Since we have discussed the interrelationship of sediment,
eutrophication, and NO,, we know that erosion control alone might only reduce
the sedimentation problem without improving the other water quality parameters.
If eutrophication is P-limited, erosion control is helpful; if it is
light-limited, there should be a reduction in the application rate in addition
to some erosion control measures. NO. problems can most likely only be elim-
inated by reduced fertilizer application. Drainage characteristics notwith-
standing, NO. might end up in the reservoir via interflow, a situation not
helped by phasing input since pollutants accumulation is a problem in this
receiving water type. Thus reducing runoff via SWCP does not help. Depending
on the type of pesticide, SWCP might have some impact, but different management
schemes would most likely mitigate the problem; i.e., reducing nutrient and
pesticide application rates and timing their application must be added to
erosion control practices in order to alleviate these combinations of problems.
Parameters such as sand, silt, clay, distribution of soil/sediment, adsorption
capacity, trapping of P and N in the reservoir, excessivity of N over P with
-aspect to algae requirements, turbidity, snowmelt, potential enrichment of
soil, and ratio of particulate P to dissolved P play an important role in
designing the overall abatement strategy. Knowledge of pathways of pollutants
is essential.
If there is a water quality problem of BOD, eutrophication, and bacteria
in a small stream due to some type of animal holding, the remedy should be
less complicated. Phasing the waste input over time into a receiving stream
through a holding pond (and possible waste treatment) reduces its impact.
This is definitely true for bacteria and BOD; if it is also true for nutrients,
52
-------
it must be assessed on a case-by-case basis. Controlled land application
of manure might be the only possible way of controlling nutrient input.
53
-------
Section 5
Project Monitoring
Background
In order to discuss monitoring the water quality impact of an individual
project, several assumptions are made: the project area is larger than a
single watershed;* the project is identified because of its water quality
problems and not because of critical areas in the watershed (in the tradition-
al SCS sense);** there is no uniformity in the river basin with respect to
land use, topography, and soils; and there is no single ownership.
These assumptions lead to the following premises, which are necessary
to the understanding of any project:
The smaller the project area is with respect to the total basin/water-
shed, the more difficult it is to identify clearly any distinct im-
pacts.
The larger the volume of the receiving water is, the more difficult
it is to trace impacts.
If the receiving water is not a lake/reservoir (or possibly an estuary),
it will be most difficult to measure and identify water quality impacts.
The more individual practices are combined together in a so-called
"Best Management Practice" (BMP), the less likely it is that the
effectiveness of individual practices can be identified (since combi-
nations of practices are very unique to a specific problem, the fewer
practices are combined, the more useful are the results).
The more diversified the ownership of the lands under investigation,
the more difficult it is to monitor the agreed-upon practices.
Even though severity of erosion and washoff is very much dependent
upon precipitation events and their intensity, duration, and frequency,
it is wrong to associate immediate agricultural pollution impact
in every case solely with individual storm events; rather, impacts
must be seen in terms of the hydrologic cycle and its interaction
*0ne of the significant characteristics of the watersheds that are addressed
in the literature review turned out to be their small size. There is apparent-
ly little experience in monitoring large area impacts.
**For example, the heavy erosion of an area far away from a stream will
most likely influence the water quality less than the moderate erosion from
a field near the stream.
54
-------
with soil and groundwater (interflow) conditions.*
Since each pollution problem is characterized by a unique setting,
individual practices and combinations of practices must be analyzed
in terms of individual pollutants and their behavior in the specific
setting. (For example, classifying agricultural chemicals according
to their expected dissolved and solid-phase forms can provide a basis
for evaluation of a "baseline" effectiveness of a practice in reducing
the load of agricultural chemicals on the stream, but only detailed
consideration of soils' runoff/infiltration potential will permit
identification of time-phased impacts.)
If the runoff enters a receiving water characterized by relatively
long detention times (e.g.-, impoundment), phased release of pollutants
from the land (e.g., washoff vs. interflow) does not significantly
reduce pollution potential, but phased release is helpful in miti-
gating impacts in a running stream.
In general, all measurements of pollutant loads/concentrations that
are chiefly induced by precipitation show a large variability. Exper-
ience has shown that a good initial working hypothesis is that runoff
varies more between years than between land uses/practices, and that
sediment discharges and particulate P and N vary more between land
uses than years. However, the latter is only obvious for extreme
differences such as cropland vs. pastureland, or row cropping vs.
rio-tillage, but is not necessarily true for marginal land use differ-
ences, as they might be created by the implementation of certain
practices. This also implies that the increase of information due
to extending a monitoring series over long years is only very marginal.
The analytical techniques used up to now to analyze highly variable
phenomena are not well developed. Using simple average values can
be extremely misleading.
In general, a network of monitoring stations must be installed to
develop reasonable data sequences.
In many cases, those parameters that are of concern for an evaluation
of performance are not monitored (for example, monitoring only solu-
ble phosphates instead of total phosphorus and soluble phosphates).
Currently, existing water quality data can generally be used only
to indicate water quality trends; they do not permit establishment
of a historical land use/water quality relationship because avail-
able "characteristic" land use data are inadequate.
There are a few questions that should be addressed before the necessary
steps for a reasonable monitoring program are laid out. First, how can instream
concentrations be characterized? Because of the transient nature of the
processes generating storm loadings and the resultant variabilities in con-
centration and flux, it may not be sufficient simply to consider average
*The surface runoff control by any practices does not necessarily reduce
the total (runoff + percolation) losses of chemicals, so that due to interflow,
there is a time lag between the precipitation event (the cause of this perco-
lation) and the actual input to the stream.
55
-------
concentration levels, for example, in relation to criteria or to the "control"
land use impact. It seems more appropriate to consider the concentration
history over a certain period, which includes both means and extremes of
meteorologic conditions. This would provide a complete picture of water
quality variations in time, which can serve as a basis for contrasting and
evaluating the significance of variations during high-flow periods. Thus
a hypothesis derived from these arguments is that in order to document precip-
itation-based problems effectively, we need to consider a range of flow con-
ditions.
Frequency distribution is one way to characterize concentration levels.
Such a distribution would be derived from monitoring data (or from simulation).
It may be useful to summarize this distribution by fitting appropriate proba-
bility density functions. Time series behavior (serial dependence) and con-
centration/duration statistics should be considered, and the significance
of using alternative time scales for sampling or averaging explored. Appro-
priate time scales might vary with component, water use, and type of water
body, subject to data availability constraints. In general, little work
has been done on this subject.
i
If it is desirable to estimate the impacts of concentration variations
on specific water uses, it is necessary to map the concentration distribution
onto a function that reflects relative degrees of use impairment or damage
for each level of concentration. Such a function could be based upon water
quality criteria or standards and might possibly be specific for each water
quality component/water use combination. This would lead to two questions:
(1) how to express environmental damage; and (2) how to specify water quality
component/water use combination.
Clearly, the purpose of this report is not to answer these questions,
but with regard to M/E, it should be made clear that for the estimation
°f relative degrees of use impairments or harm to aquatic life, for example,
the averaging or sampling/monitoring used to derive the concentration fre-
quency distribution must be consistent or made iteratively consistent with
the characteristic response time of the damage to a change in concentration
(i.e., the time it takes for any damage to be felt). This means that moni-
toring/analysis/evaluation should be performed continuously in order to im-
prove the M/E program.
Since the value for historical water quality data has been constantly
debated, some clarification would be useful at this point. Water quality
data, measured at one location for a given time period, present the "lumped"
impact of upstream land uses. Over time, they might indicate trends for
certain parameters, and they can be analyzed in terms of average, median
variance and range so that degree of variability of concentration and load
can be established. Sometimes they have been related to external parameters,
such as precipitation records and land use distribution. While the first
type of analysis makes sense, the latter is of dubious value. In general,
changing land use conditions and the variety of influencing factors (as ex-
pressed in our previous discussions) do not permit any conclusions about
cause-effect relationships. Sometimes, those data might be useful in calibra-
ting an analytical model designed to simulate land use/water quality
56
-------
interactions. But their value becomes highly questionable if no historical
input data exist. Thus if there is a network of water quality monitoring
stations, the correlation of the data from monitoring stations improves the
state of knowledge and helps to identify critical areas. But again, it is
inconceivable that water quality data alone can be interpreted so as to assess
the effectiveness/performance of certain agricultural practices. The lack
of historical data on the actual "land use/performance" and the high unrelia-
bility of the system do not permit respective conclusions from "lumpy" water
quality measurements.
For example, EPA developed a very extensive data base on eutrophication
of lakes/reservoirs in its National Eutrophication Survey. Using this data
subset on midwestern impoundments to explain phosphorus retention phenomena
was unsuccessful;* however, isolating those impoundments for which sediments
samples taken by USDA and the Corps of Engineers were available made the
phosphorus retention analysis much more conclusive. Thus EPA's results,
that on the average 41 percent of total phosphorus export from 96 agricultural
watersheds (of which 50 were in the cornbelt) is in ortho-phosphorus form,
seem to be incorrect (as was also confirmed by some Black Creek data), which
implies that monthly grab samples of lake tributaries most likely do not
reflect the loadings of particulate P entering during storm events.
In summary, the sampling program and its derived conclusions did not prove
adequate for the analysis and explanation of the actual driving forces of
the system. Additional data on sedimentation indicate the importance of
sediments, but even these data are inadequate to pin down critical areas
and practices in the upstream watersheds. Only a network combining measure-
ments of concentration and flow for the calculation of loads (and ultimately
a receiving water budget) with downstream water quality impact data can permit
determination of cause-effect relationships.
Setting Up the Monitoring Network
In setting up a basinwide monitoring program, five basic questions have
to be answered:
1) What are the types of samples to be taken and measurements to be
made (constituents, flow, etc.);
2) What are the locations of sampling stations;
3) What is the frequency and duration of sampling at the stations;
4) What are the methods to be used in sampling and measurement (i.e.,
equipment, etc.); and
5) How are the data to be processed and stored for later analysis?
Generally, these five characteristics of sampling programs cannot be consi-
dered independently, but it is not necessary to consider the same characteris-
tics for each station; i.e., some stations might be designed to identify
water quality and other trends in order to understand specific relationships.
Thus some might be sampled at a lower or higher frequency for a few or many
*Meta Systems, op. cit.
57
-------
constituents, respectively; high frequency sampling might be by automatic
samplers/ while low frequency sampling might be performed manually.*
The following steps should be followed in order to set up the most effec-
tive monitoring program:
1) Define the critical stretches of the receiving water;
2) Describe the apparent water quality problem in quantitative terms
as much as possible (e.g., to label a problem "eutrophication" is inappro-
priate; rather, define some estimates of nutrient and sediment concentra-
tion, hypolimnetic DO conditions, and fish population);
3) State clearly which are the parameters that have to be controlled
and in which season they are of concern (e.g., what is the limiting factor
of the eutrophication; if P, should total particulate and dissolved P be
monitored);
4) State clearly what the critical land areas are that determine water
quality impact and describe those individual water quality parameters that
are influenced by the critical areas (e.g., if an area is characterized by
heavy gullies, the impacts of erosion and nutrient control might be different
than the impacts of such control measures on a cropland without heavy gully
formation);
5) Assess qualitatively the effects that various BMPs might have on
runoff and edge-of-stream load (and on water quality);
6) Attempt an analysis (possibly with regional data) of the impacts
of various practices and practice combinations on the pollutant loads, the
load distribution, and on the receiving water's quality. On the basis of
this analysis, identify the critical parameters to be monitored***
7) Based on the problem's temporal and geographic characteristics, the
parameters of concern, and the BMP's anticipated effects, lay out the monitor-
ing network (including flow measurements). Since not all the areas are sim-
ilar and not all areas are treated with the same practices, receiving water
loads (i.e., flows and concentrations) must be measured at various points.
(This is particularly important for lakes and reservoirs in order to derive
a pollutant budget.)?
8) Be prepared to monitor water quality and flow in creeks (draining
to critical water quality stretches) during storm events occurring shortly
after applications of fertilizer and pesticides (this generally requires
automatic sampling equipment);
9) Determine the frequency of sampling in creeks/drains to the receiving
water, and in the receiving water's critical stretch. Generally, due to
the higher variability (i.e., more dynamic response to precipitation events),
the frequency of monitoring has to be higher in the upstream parts of the
*What individual(s) or institution(s) should do the sampling is frequently
an important question.
**Haith and Loehr, op. cit., and Meta Systems Inc, op. cit. should provide
some guidance on this type of analysis.
58
-------
basin than in the downstream parts. Further, frequency depends on the infor-
mation needed for characterizing a pollution problem in a receiving water.
Simply stated, when the pollution problem is of a cumulative nature (like
eutrophication in lakes/reservoirs), the critical area needs to be monitored
relatively infrequently because changes do not occur rapidly. However, the
drainage into the major receiving water has to be monitored on a regular
basis in order to establish the input patterns. If a pollution problem is
identified in a running stream, continuous/relatively high frequency monitor-
ing is necessary in order to establish patterns. A nitrate problem (e.g.,
in certain seasons) is a typical example.*
10) Determine the appropriate length of time for a watershed to be moni-
tored. This varies according to the nature of the pollution problem, the
nature of the control or management strategy, and the length of monitoring
chosen for other watersheds providing information to the program. Some pollu-
tion problems occur in a manner that requires long time periods in order
to obtain a statistically significant number of observations. Examples are
pesticides and fertilizer that may only be applied once during the course
of a year and be absent from runoff waters after 3 or 4 runoff events. The
eventual selection between different management and control strategies neces-
sitates some understanding of the relative effectiveness of the different
strategies, in terms of mean or median effectiveness and in deviation from
this average. Longer periods of monitoring reduce the variance of an estimate
at a single site. This leads to a tradeoff between parameter accuracy and
monitoring duration. While long-term monitoring projects may be required
for reduced variance, a greater diversity of projects is desirable so that
the information will provide a more accurate estimate of the possible effec-
tiveness of the management practices at some site that has not been moni-
tored.** Generally, a combination of long- and short-term monitoring projects
is superior to the exclusive use of either one.
11) Given the analytical framework for the analysis, the need for constant
reassessment of the monitoring scheme, and the value of the generated data,
process and store the data in such a way that they are easily accessible
and transformable. All raw data should be, retained. After initial collec-
tion, simple statistical manipulation can be performed, and simple relation-
ships plotted. Thus questions such as the following can be addressed through-
out the monitoring period:
How do assumptions of flow in the model compare with actual flows?
Do we encounter dry, wet, or average years/months?
How do discharge and the concentrations and flux of all components
monitored vary with season?
At each station, what is the relationship of concentration to flow?
(A typical pattern for nonpoint sources is a positive relationship
between flow and the concentration of suspended-phase materials.
Dissolved phase concentrations exhibit a variety of patterns.
*See U.S. EPA NFS Task Force, "Guidance Document on NFS Monitoring," 4th
draft, Washington, D.C., March 1980.
**In an econometric analysis, this tradeoff is reduced bias vs. increased
efficiency of the estimators.
59
-------
Conductivity may decrease with flow due to dilution of base flow.
Dissolved P is usually weakly or totally unrelated to flow. If a
large negative relationship is apparent for any component, an upstream
point source may be indicated.)
Do concentration variations within storm events appear to follow
any pattern with respect to flow or time?
What is the flow-weighted average concentration of each component?
Do there appear to be substantial differences in concentration and
export (mass/watershed area-year) across stations? Could they be
related to differences associated with land use activities?
What is the total P loading?
How does P discharge compare with loading? In case of reser-
voir/lakes, below the reservoir, what was the hydraulic detention
time during the year? What is the measured phosphorus retention
coefficient and how does it compare with that predicted on the basis
of residence time?
The answers to these and similar questions help assess the appropriateness
C;f the monitoring scheme. M/E is so expensive that the scheme has to be
revised according to the information gained in order to make monitoring as
cost-effective as possible.
12) Incorporate into the monitoring programs quality assurance.* Ade-
quate steps must be taken to ensure that the data gathered are reliable.
This includes proper sample collection and preservation methods. Calibration
of field instruments, in particular, the dissolved oxygen meter, is critical.
A quality assurance program should be set up to verify laboratory analyses,
particularly for measurements that have not been made routinely in the past.
Coded replications, spiking of extra samples, and analyses by independent
laboratories are three typical means of verifying analyses. This type of
activity should be emphasized in the early stages of the program and reduced,
but not eliminated, as procedures become established.
An Example of a Monitoring Program
Ideally, an analytical framework of the land use/water quality interface
serves two purposes: guiding the monitoring programs, and continuous analysis
and evaluation. Using such an analytical framework, we present below an
example of a monitoring program in which the analytical framework is oriented
toward "long-term/average" conditions. Thus we explore how the framework's
data needs and the respective monitoring program might be set up for a eutro-
phication problem in a drinking water reservoir that was identified as P-
limited on the basis of analysis with regional data.**
*For example, the NURP program provides some guidance. See U.S. EPA "Data
Collection Quality Assurance for the Nationwide Urban Runoff Program." A
similar document should be prepared for RCWP.
**Assuming a 3-year monitoring program revised periodically as data become
available.
60
-------
The methodology and associated data requirements can be organized in
three general relationships or submodels (Figure 7):
1) land use/stream phosphorus relationships;
2) reservoir phosphorus balance; and
3} trophic state and associated water quality conditions.
By monitoring at selected locations, it should be feasible to identify
the relative impacts of various land uses on stream P levels. Stations should
be located so as to isolate areas within specific land use categories where
possible. Both periodic and storm event sampling should be done. Staff
gages should be installed and calibrated to permit flow measurements at each
site.
Monitoring is also needed to establish the reservoir P balance,
i.e., inflows and outflows expressed on an annual average basis. This will
help to establish relationships among P loadings, nutrients available to
support algal growth in the impoundment, and nutrients discharged to down-
stream stretches. Additional calibration of the phosphorus retention model
is feasible with this type of information. Monitoring of flow and concentra-
tion at .the_inflow and outflow points is required. Since the data require-
ments are similar, stations at reservoir inflow points can be used both for
estimating land use/stream P relationships and for estimating the reservoir
P balance. The latter also requires outflow stations.
Monitoring within the impoundments will provide a means of assessing
existing conditions and further calibrating the trophic index system. The
latter describes relationships among phosphorus, chlorophyll-a, transparency,
and oxygen depletion rate. This represents the linkage between the P mass
balance and the reservoir water quality response. The program should also
cover water quality aspects that may be indirectly related to trophic state,
such as iron, manganese, and trace metals released from sediments during
anaerobic periods. Impoundment monitoring also entails temperature profile
data to determine the extent and period of vertical stratification.
Seven locations for watershed monitoring stations in the Sargent River
system are listed in Table 1-7 (and shown in Figure 8) together with the recommen-
dations for watershed and impoundment monitoring, including variables mea-
sured, spatial frequencies, temporal frequencies, and environmental conditions
recorded. Station A, at the Almar Road drain, should permit some isolation
of urban impact. Three of the stations (B, C, and D) are located on the
lower ends of the three main branches of the river above the lake. Station
£ is below the confluence of three branches and above the lake. To complete
the mass balance on the lake, Stations F and G are located on the eastern
tributary and below the lake, respectively. Because they are critical to
the input/output analysis of the reservoir, we suggest that stations E and
G be given highest priority in terms of flow and quality monitoring activity.
Watershed monitoring is scheduled on a monthly basis, but should be
biweekly during spring runoff. A few storm events should also be monitored
at each station in order to identify the extent of P concentration variations
61
-------
FIGURE 7. CONTROL PATHWAYS TO BE STUDIED IN THE MONITORING PROGRAM
Other Controlling Factors
I 1. I
Land Stream Reservoir Trophic Water
Use ^"Phosphorus*"Phosphorus*~ State Quality
62
-------
TABLE 17. RIVER WATERSHED AND RESERVOIR MONITORING PROGRAM
Watershed Station Locations*
A. Almar Road Drain
B. River, W. Branch
C. River, N. Branch
D. River, E. Branch
E. River, above lake
F. Eastern Tributary
G. Lake Discharge
Environmental Conditions Recorded
- precipitation (including previous
3 days)
- air temperature
- wind velocity direction (quali-
tative)
- sky cover (qualitative)
- odor (qualitative)
- reservoir level **
- reservoir discharge source**
(spillway vs. regulator)
- reservoir flow regulator setting**
Water Variables Monitored
stage (flow)
total phosphorus
dissolved phosphorus
turbidity
dissolved color
conductivity
Reservoir Variables Monitored
temperature
dissolved oxygen
total phosphorus
dissolved phosphorus
ammonia nitrogen
organic nitrogen
nitrate nitrogen
nitrite nitrogen
PH
conductivity
dissolved color
turbidity
transparency
chlorophy11-a
iron
manganese
trace metals
Monitoring Frequencies
a) watershed
spatial: 1 sample, mid-depth,
mid-stream
temporal: monthly (except bi-
weekly during spring runoff);
2 or 3 storm events per year
b) reservoir
Spatial Frequencies
deepwater stas. discharge stas.
profilet
profilef
0-10 ft composite
grab
surface
0-10 ft composite
grab
*Locations are identified in Figure 8.
**Lake discharge station only.
^Profiles at 5-ft intervals except 1-ft near thermocline (15-20 ft)
63
-------
FIGURE 8. SUGGESTED MONITORING STATIONS IN THE WATERSHED
64
-------
during these events and any relationships with flow. Measurements of total
and dissolved P will provide a basis for assessing both the quantity and
quality of nutrient sources in the watersheds. Dissolved P is considered
a more readily available nutrient source for phytoplankton growth than suspen-
ded P. Turbidity, color, and conductivity will provide additional indications
of the origins of the water (i.e., surface runoff vs. base flow) on seasonal
and storm event bases.
Estimates of the annual hydrologic budget should be based on weekly stage
data. Daily measurements would be preferable if convenient. If a continuous
stage recorder and/or automatic sampler is available, installation at Station
E just above the reservoir seems appropriate.
One deepwater and one discharge station should be established to monitor
conditions in the reservoir during the stratified period. Discharge stations
should be located to catch bottom-water releases. These are included to
provide a means of detecting any nutrient, iron, manganese, and trace metal
releases from bottom sediments that may occur if and when hypolimnetic oxygen
levels become critical. Generally, one profile station should be established
at the deepest location of the impoundment, but far enough removed from the
dam to be representative of the open water.
A monthly monitoring frequency between April and October seems adequate.
Sampling should be done under settled weather conditions. The program should
include vertical oxygen and temperature profiles and surface composite samples
of P species, N species, pH, conductivity, color, turbidity, and transparency.
While algal growth in these reservoirs is probably limited, N species are
included to provide a means of assessing any limiting effects of N. Chloro-
phyll-a is the basis of the trophic index and should be monitored or at least
spot-checked if suitable laboratory arrangements can be made. If laboratory
facilities are limited, N and chlorophyll-a analyses can be limited to one
sample in late spring and one in late summer. N could be phased out altogether
if the results indicate that it is insignificant as a limiting factor after
one year of monitoring.
Copper sulfate or other algicide treatments will have influence on assess-
ments of existing conditions using the trophic index system. If such treat-
ments cannot be eliminated altogether, impoundment monitoring should be
scheduled so that surveys do not follow immediately after treatment. In
any event, treatment activities should be monitored and recorded for future
reference in the data analysis.
65
-------
Section 6
Monitoring Strategy
In Section 5 we outlined how to go about setting up monitoring projects
for nonpoint source pollution abatement. In this section we present some
criteria for selection of representative projects across the United States
for M/E. This process should proceed in steps.:
1) Determine the projects that exhibit agriculturally based water quality
problems;
2) Characterize these projects according to the important M/E criteria;
3) Weight the M/E criteria for final choice of project.
For Step 1 the following sequence of actions might occur:
1) Local USDA (SCS; ASCS) and regional EPA sources designate significant
pollution problems in each of the EPA regions with respect to loca-
tion, pollutants, receiving waters, watershed characteristics (land
use/ownership), and proximity to areas previously investigated in
detail. This knowledge can be summarized for each region in a simple
matrix (such as Table 6 in Section 4), where the entries would be
the number of the particular land use/water quality problems in each
region.
2) The land use/water quality combinations that cause the most signifi-
cant deterioration of water quality are marked on the matrix.
3) If there is an obvious tradeoff between a "large number" of specific
problems (land use/water quality combinations) and individual infre-
quent but severely impacting land use/water quality problems, the
choice is made between the "large number" (possibly representing
a typical problem) and the individual, most severe problem. If there
is no obvious tradeoff, additional criteria are used, such as typical
hydrology, ease of problem identification and isolation, and local
capacity (basically the same decision-making process that has been
used up to now).
For Step 2 the characterization of the projects chosen on the basis
of criteria important for M/E project selection should proceed in two ways:
1) A determination should be made based on federal perspective whether
or not a project is typical; and
2) On the basis of technical details, implementation potential, and in-
stitutional capabilities, a ranking should be made of the leftover
projects.
66
-------
For 1 xonder Step 2 above, we suggest the following tools:
Climatic regions (ultimately it is desirable to have at least one project
in each of these regions);
Precipitation/runoff (permits distinction in dry and humid climates);
Livestock population (RCA project) (these indicate the overlap or
Crop intensity (RCA project) isolation of problems of poten-
* * tially great magnitude)
Soil erodibility and drainage maps (adding these maps permits evalu-
Other U.S. maps ation of the seriousness of erosion
^~ problems or fertilizer drainage
in areas under intensive investi-
gation)
USGS, EPA maps on SS, P, N (these generally provide the background
material necessary to put a project in perspective or to select a specfic
project type).
For 2 under Step 2 above, there are some characteristics/criteria that
were tested in our review of the current 13 projects and their suitability
for M/E (Table 18). Not all of these characteristics/criteria carry the
same weight (see below).
For Step 3 the ranking of the projects should be based on the charac-
teristics/criteria of Table 18. All of these characteristics/criteria can
be viewed as purely descriptive, but most of them lead to comparisons in the
review process. Thus the weighting should not be static, but should reflect
the information gained in conducting M/E projects. In this way, statements
can be made about each characteristic/criterion, reflecting, more or less,
the current state of knowledge.
1) In selecting projects across the United States, all areas should
ideally be covered; however, for now it is desirable to focus on
areas that have cold, as well as warm, seasons.
2) A general description of the water quality problem allows a first
indication of the extent to which it can be quantitatively described.
3) It is necessary to isolate as much as possible individual, largely
homogeneous land uses in order to draw any inferences for a poten-
tial project.
4) The choice of irrigated or unirrigated land use is largely a matter
of agency preference.
5) Any point source or septic tank influence should be minimized because
its impact cannot be easily isolated and abstracted from concentra-
tion/load estimates, especially in lake/reservoir receiving water
systems.
6) Overall analysis and evaluation are facilitated if one specific pol-
lution problem can be identified. Different problem types need
different remedies in terms of practices/measures, but the more
individual practices are eventually combined in a so-called "Best
67
-------
TABLE 18. PROJECT CHARACTERISTICS
1. U.S. Location
North/South//East/West (i.e., separation into dry and humid
areas and those impacted by snowfall)
2. Water Quality Problem (general)
3. Major Land Use (acres if available)
cropland (type of crop)
feedlots (covered under RCWP)
animal holdings (except feedlots)
range/pasture
mix (population centers and others)
4. Irrigation/Nonirrigation
5. Point Source Influence
point source
nonsewered/septic tanks
purely nonpoint source agricultural problem
6. Type of Pollution Problem (as defined by review of land)
erosion and associated nutrients
erosion and associated nutrients and pesticides
heavy pesticide use
7. Receiving Water (including hydrologic characteristics)
lake/reservoir
small stream
river
bay (Great Lakes)
groundwater
8. Drainage/Land Characteristics
flat, delta type
unclear drainage to critical water quality areas
clear-cut drainage
slope
68
-------
TABLE 18. (CONTINUED)
9- Project Area (acres)/Watershed
10. Population in Project Area
11. Critical Area (acres
ratio of critical area to project area
12. Number of Farms (in critical land area)
large ( > 200)
medium (100-200)
(these are not absolute limits, but reflect
the 13 projects currently considered)
small ( < 100)
13. Number of Animal Facilities (in critical land area)
14. Water Use
drinking water supply
recreation (contact/noncontact)
fisheries and wildlife
agricultural and industrial water supply
15. Specific Water Quality Problems
coliform
pesticides
eutrophication (P-f N-, light-limited)
nitrate
sedimentation
salinity
16. Parameters Previously Monitored
17. Preliminary Analysis in Application (including^pathway of pollutants)
18. Protective/Preventive Practices (suggested)
19. Suggested M/E plan in Project Application
20. Inclusion in 208 plan?
69
-------
Management Practice" (BMP), the lesser the likelihood that the effec-
tiveness of individual practices can be identified. Thus since com-
binations of practices are very unique to a specific problem, the
fewer problems involved the better and, hence, the fewer practices
combined, the more useful the results.
7) Given the current state-of-the-art of identifying and evaluating re-
ceiving water pollution phenomena, it must be emphasized that if the
receiving water is not a lake/reservoir, it will be difficult to
measure and identify water quality impacts. There is, however, at
least one caveat to this suggestion: if this receiving water's pol-
lution problem has been present for a long time, it cannot be cleared
up quickly, because of the lake/reservoir's "memory" (e.g., nutrients
in sediments). This makes it rather unlikely that any changes in
water quality can be identified as a result of practice changes in
a 5-10 year period.
8) In order to effectively perform M/E, there should be only one drain-
age area and clear-cut drainage patterns.
9 and 11) The smaller the project's critical area with respect to the
total project basin/watershed, the more difficult it is to identify
clearly any distinct impacts; also, the smaller the ratio of receiv-
ing water surface area to watershed area, the lesser the likelihood
that changes due to isolated land practices can be extracted from
water quality data. Furthermore, the type of analysis applied is
somewhat influenced by this configuration; in impoundments with
extremely short hydraulic residence times, seasonal variations of
loadings may become of overriding importance.
10) Since permanent and temporary population may contribute to pollution
(see 5 above), it is advisable to have as small a population as pos-
sible in the project area (including the impacted receiving water
area).
12) The smaller the number of farms the better, since diversified owner-
ship of the lands under investigation makes it more difficult to
monitor the agreed-upon practices. This is especially true if
measures other than structures have to be used.
13) Since a large number of animal holding facilities makes identifica-
tion of the problem area difficult, it is desirable to have a rela-
tively small, observable number of facilities in the area.
14) Definite water use characterization is important, since it implies
standards for respective water quality parameters that would have been
set by the state having jurisdiction over the problem area. Such
standards would provide for clearly identifiable thresholds to be
reached in the receiving water improvement.
15) Some of the specific water quality problems occur mostly in combina-
tion, such as eutrophication, sedimentation, and NO3. This influ-
ences the set of practices to be applied (and monitored) and the
monitoring requirements (space and time) in the receiving water.
Since each pollution problem is characterized by a unique setting,
individual practices and combinations of practices must be analyzed
70
-------
in terms of individual pollutants and their behavior in the specific
setting. The more interplay there is between problems, the more
difficult analysis and monitoring become and thus the more careful
their set-up must be.
16) Historical data can give some clue about past water quality trends,
e.g., the time period a eutrophication problem has prevailed. But
the data base is generally inadequate to draw any conclusions about
the land use activities that caused the problem.
17-20) The more analysis performed prior to any monitoring exercise the
better. The relative level of required analysis depends on the pol-
lutant/practice/receiving water combination for each area (see 15).
On the basis of the above discussion, we feel that the following factors
are most important in selecting M/E projects at this time:
ease of identifying the water quality problem and its cause;
eutrophication problems in lakes/reservoirs or in other relatively
stagnant water bodies generated in a clearly identifiable upstream
area, with focus on those eutrophication problems whose history is
only relatively short;
the potential for reducing the "reason for the water quality problem"
to one land use type (e.g., cropland versus animal holding) and thus
the avoidance of "mixed land uses"?
avoiding areas where uncontrollable septic tank influences are possible;
given the interest in tradeoffs of point source vs. nonpoint source,
isolation of a project in which correction of both problems can be
monitored- (it would be helpful to have historical data on the point
source effluent);
avoiding areas with more than one drainage pattern.
Finally, since the utilization of the data is important, we advise that
their handling and storage be comparable to the EPA (STORET) and USGS systems.
In this way the new data can be fit with other data sources to perform local
and regional analysis of the type needed to initially assess the water quality
problems of certain areas (See our discussion in Section 5).
71
-------
Section 7
Conclusions
In this concluding section, we will reiterate five major points that have
come out of this report.
1) Water quality considerations must be.the driving force behind the
selection of Best Management Practices. The "2-track" planning process (Figure
6) should ensure that the water quality problem is adequately identified and
that appropriate practices/measures of the five control categories (Figure
3) are imposed.
2) Background water quality data are not essential in the selection of
an M/E project. They are valuable in putting a water quality problem in per-
spective (e.g., in identifying the number of years a eutrophication problem
has been present), but they do not alone lead to the identification of the
reasons for past pollution problems.
3) Any design of a monitoring scheme, and the monitoring itself, must
be preceded by analysis to determine the pathways of pollutants and to clarify
which parameters have to be monitored at what frequency and what location.
4) The transferability of results from one project site (intensively
monitored) to another is questionable because of currently limited knowledge.
However, the transferability will improve as additional monitoring projects
are carried out. The complexity of this issue is demonstrated by the length
of time it took to develop regional estimators for hydrologic regions. It
might, however, become possible to uncouple upstream land and drainage systems,
and downstream receiving water systems on the basis of known receiving water
behavior; but this can only be accomplished if it can be determined with cer-
tainty what the receiving water's response will be to changes in input. Thus,
uncoupling might become feasible with lakes/reservoirs.
5) It is necessary to integrate firmly the land/water quality analysis
with monitoring and evaluation of the water quality and the performance of
the practices. This means that the monitoring system must be adjusted to the
results of the monitoring and analysis, but it also means that the practices/
measures must be adjusted as well. It is obvious that, in many cases, conven-
tional measures do not meet established needs; they must be augmented by
management measures in order to yield the desired water quality results.
*Ui GOVERNMENT PfllNTINO OWICti 1»IO 341-012/101 1-1 72
------- |