WATER RESOURCES BULLETIN

VOL. 26, NO.3 AMERICAN WATER RESOURCES ASSOCIATION JUNE 1990

TARGETING REMEDIAL MEASURES TO CONTROL

NONPOINT SOURCE POLLUTION'

W. T Dickinson, R. P. Rudra, and G. J. Wall2

ABSTRACT: The watershed model GAMES is used for the evalua- concluded that there would be substantial savings
tion of a targeting approach to control fluvial sedimentation arising from targeting erosion control within watersheds, in
from soil erosion in agricultural areas. The data considered for the
analysis consists of output from the application of the model to terms of soil loss reduction per dollar of program cost;
existing and hypothetical soil and crop management systems in two and Runge et al. (1986), noted that more accurate tar-
small watersheds of southern Ontario, one in the rolling uplands geting could reduce on-site productivity losses and
and the other in a very flat lowland area. The model output minimize the area affected negatively by restrictive
includes estimates of spring sediment yield from field-size cells to
the stream outlet for existing agricultural management conditions, land use practices. In a joint position statement, the
and estimates of sediment yield resulting from the successive Ontario Chapter of the Soil Conservation Society of
implementation of two levels of soil erosion controls under four America (now the Soil and Water Conservation
remedial measures strategies. The results reveal that, for the Society) and the Ontario Institute of Agrologists
rolling upland watershed exhibiting a wide range of soil erosion
and sediment yield rates, targeted control programs can be expect- (1986) recommended that lands with erodible features
ed to provide an extremely effective approach to sediment control. by considered as target lands for subsidization of for-
For flat lowland watersheds, exhibiting relatively uniform soil ero- age crops or new productive woodlots, combined with
sion and sediment yield rates, the strategy of targeting controls cross compliance. Targeting initiatives have in fact
may be somewhat more effective than a random approach to con-
trol, but not as efficient as in the case of watersheds in more rolling become part of soil conservation policy in the U.S.,
terrain. It is evident from the study that a screening model such as with the Congress emphasizing in the '81 farm bill
GAMES provides a very useful tool for the planning and evaluation the need to target conservation expenditures to the
of erosion and sediment control programs. nation's more erosive land; and the Food and
(KEY TERMS: nonpoint source; targeting; erosion control; remedial
strategies.) Agriculture Act of '81 authorizing creation of a special
areas conservation program (Ogg et al., 1982; Ogg and
Pionke, 1986).
Regarding nonpoint source pollution control, it has
INTRODUCTION been strongly recommended that areas within water-
sheds that have a higher potential to deliver pollu-
The targeting of soil conservation practice and poli-
tants be identified and that implementation of
controls be given priority in these areas (PLUA.RG,
cy has been advocated in recent years for effective and 1978; Wall et al., 1978; Nonpoint Source Control Task
efficient control of both soil erosion and associated
Force, 1983; Duda and Johnson, 1985; Maas et al.,
nonpoint source pollution problems. In this context, 1985; Myers, 1986; Ogg and Pionke, 1986; Ontario
"targeting" has meant directing remedial measures Chapter SCSA and OIA, 1986). The Nonpoint Source
and/or assistance to those specific areas most requir- Control Task Force recommendation acknowledged
ing such remediation, as opposed to either a more that only a small number of nonpoint programs had
general areal application of controls or offering of been targeted to those areas that contribute a dispro-
financial assistance.
It has been claimed that targeting assistance for portionately large share of the pollution load, and
that it is necessary with a scarcity of resources to
soil erosion control could more than triple the amount
identify priority management areas and to target con-
of soil saved through the Agricultural Conservation trol expenditures. Without better targeting, solving
Program in the U.S. (Henry, 1981). Lee et al. (1985),

'Paper No.89087 of the Water Resources Bulletin. Discussions are open until February 1, 1991.
2Respectively, Professor and Associate Professor, School of Engineering, University of Guelph, Guelph, Ontario, Canada; and Research
Scientist, Agriculture Canada, Guelph Agricultural Centre, P.O. Box 1030, Guelph, Ontario, Canada.

499 WATER RESOURCES BULLETIN

 Dickinson, Rudra, and Wall

nonpoint source problems is slow and expensive. It and key sediment source areas (Clark et al., 1985; Lee
has been suggested that the key to cost-effective and Goebel, 1986; Ogg and Pionke, 1986).
water quality improvement is the targeting of remedi- Alternatively, soil loss estimates can provide the basis
al programs (So and Singer, 1982; Clark et al., 1985; for selecting problem erosion areas, and sediment
Duda and Johnson, 1985; Lee et al., 1985; Maas et yield estimates (usually determined from the applica-
al., 1985; Ogg et al., 1985); and Spurlock and Clifton tion of a delivery algorithm to soil loss values) can be
(1982) stated that a policy involving differential used for delineating sediment source areas (Beasley
restrictions might be not only more efficient but also et al., 1982; Spurlock and Clifton, 1982; Dickinson et
more equitable, obtaining the most pollution abate- al., 1985; Lee et al., 1985). In some cases, priority ero-
ment per dollar of costs with the costs of control being sion and sediment source areas have been ascertained
borne more equally among the polluters. Duda and qualitatively by the consideration of selected groups
Johnson (1985) categorically stated that interagency of factors relating to soil detachment and transport,
cooperative programs must be targeted to pollution such as erosion rate, distance to nearest watercourse,
hot spots if Clean Water Act goals are to be attained. surface roughness, and land slope (Snell, 1984; Maas
Although there has been strong support for a tar- et al., 1985; Dickinson et al., 1986). A tolerable pro-
geting approach to soil conservation and associated ductivity index reduction (Runge et al., 1986) and the
nonpoint source pollution, there is reason to question SCS land capability class system (Lee and Goebel,
whether it is the most effective and efficient in all sit- 1986) have also been used for discriminating erosion
uations. There is reason to believe that the desirabili- areas, and pollutant reduction per dollar invested
ty and necessity of targeting soil conservation (Nonpoint Source Control Task Force, 1983) has been
programs is primarily a function of the spatial vari- used for targeting nonpoint source areas.
ability of factors relating to soil loss and sediment Yet another approach to identifying target areas
yield rates within and among watersheds. In many has involved a formulation of criteria in terms of a
basins, critical erosion and sediment source areas are framework for soil erosion and sediment problem
very localized, and it is in such basins that targeting identification (Dickinson et al., 1984; Dickinson et al.,
of control measures has been shown to be physically 1985), illustrated in Table 1. Areas in Category I are
effective and economically efficient (Wall et al., 1978; those with estimated soil losses greater than the tol-
Ogg et al., 1982; Duda and Johnson, 1985; Lee et al., erable soil loss chosen for that soil, with sediment
1985). However, such wide spatial variability is not yields greater than the selected sediment yield toler-
present in all rural watersheds, and in situations ance. The soil loss and sediment yield tolerances have
where soil loss and sediment yield rates are relatively been established in light of results such as those pub-
uniformly distributed across the landscape, the notion lished by Pierce et al. (1983) and Schertz (1983).
of targeting may have little or no justification Areas in Category II exhibit a soil loss greater than
(Dickinson and Pall, 1982). Therefore, there is a need the tolerable soil loss value but a sediment yield less
to clarifr the utility of the targeting approach for non- than that deemed tolerable. Areas in Category III
point source sediment pollution control. This paper exhibit soil losses less than the tolerable soil loss
explores the relative importance of targeting conser- value but are estimated to yield more sediment than
vation programs for controlling stream sedimentation is tolerable. Areas in Category IV are considered to
in watersheds with widely different spatial patterns have both soil loss and sediment yield values within
of landform, soil loss, and sediment yield, and demon- tolerance levels. This framework has proven to be a
strates the usefulness of an existing watershed mod- very useful and flexible one, not only for the identifi-
eling technique for such a practical study. cation of apparent problem areas in a watershed but
also as a basis for improved communication amongst
the various parties involved in the planning and
implementation of soil conservation and nonpoint
CRITERIA FOR TARGETING AREAS source pollution programs.
Areas to be targeted in the study watersheds were
Exploring the impact of targeting soil conservation selected in two ways according to two different crite-
measures for controlling nonpoint source pollution ria. On the one hand, target areas were selected to be
associated with soil erosion requires establishing cri- those with estimated sediment yield rates exceeding a
teria for identifying sediment source areas and select- selected tolerable yield rate, whatever the associated
ing a method for delineating and examining such soil loss potential. On the other hand, areas with esti-
areas in study watersheds. Estimated soil loss rates, mated soil loss rates exceeding a selected soil loss tol-
or soil loss rates relative to a selected soil loss toler- erance value were identified as target zones for
ance or T value, have provided the most widely used remedial measures, whatever the associated sediment
single criterion for defining critical soil erosion areas yield rate. Details regarding application of the criteria

WATER RESOURCES BULLETIN 500

 Targeting Remedial Measures to Control Nonpoint Source Pollution

are included in a subsequent section dealing with Smith (1978) to describe sheet and rill erosion, is
remedial measures. written for each season as

A5=R5.C5.K5.LS.P (1)
TABLE 1. Framework for Soil Erosion and
Sediment Problem Identification. where A5 = computed seasonal soil loss per unit area,
R5 = seasonal rainfall factor, C = seasonal land use
Problem Erosion Sediment
Problems Problem5 or cropping factor, K5 = seasonal soil erodibility factor,
Category LS = slope gradient and slope length factor, and P =
I x x erosion control practice factor. The seasonal soil loss
U x - from a given field area, E5, is determined from the
III — x
Iv - - product of A5 and the field area.
The sediment yielded by each field to downstream
Agricultural productivity is adversely affected. fields and streams has been determined from the
Sediment yield to the stream and associated contamination are equation
excessive.

S5=DR5.E5 (2)

SELECTION OF A MODEL where S = seasonal sediment yield delivered from

one cell to another (e.g., one field to another field, or
one field to the stream), and DR5 = seasonal delivery
Soil erosion and sediment source areas have been ratio between the two selected cells. The seasonal
identified in research studies by means of a set of delivery ratio, or percentage of the potential field soil
guidelines (Snell, 1984; Maas et al., 1985) or a dis- loss which can be expected to be delivered to the adja-
tributed parameter model such as ANSWERS cent downslope cell in a given season, is determined
(Beasley et al., 1982), CREAMS (Knisel, 1980), or from the expression
GAMES (Dickinson et al., 1984; Cook et al., 1985).
Possible effects of changes in soil and crop manage- 1/2 H
ment practices on these areas have also been explored 1
(3)
with these models, but the relative impacts of differ- DR5=a[-jj-_sS r8]
ent implementation strategies including targeting
have received only limited attention (Lee et al., 1985). where = seasonal surface roughness as indexed by
GAMES, the Guelph model for evaluating the Manning's n; S = land slope; H5 = seasonal hydrologic
effects of Agricultural Management systems on coefficient, an index regarding the probability and
Erosion and Sedimentation, was selected as the model
depth of overland flow; L5 = length of seasonal over-
for describing the generation and delivery of suspend- land flow path; and a,b = calibrated watershed
ed sediment from field to field, field to stream, and parameters. Delivery from one field cell across several
downstream in the agricultural study watersheds. other cells to a downstream point is computed by
The model's accommodation of field-size cells of vari-
ous shapes and sizes, its orientation to a seasonal
time frame and a watershed scale, and considerable b
experience in the calibration and application of the DR =
model in the study region, prompted selection of
GAMES for this study. The background regarding the a[1//[
Sj
i/2
development of GAMES has been presented in the lit-
erature (Clark, 1981; Rudra et al., 1986), and details where m = number of downstream fields; and n5, 5,
on application of the IBM-PC version of the model are
H5, L5 = seasonal roughness, slope, hydrologic coeffi-
readily available (Cook et al., 1985). cient, and length of the jth downstream cell. For cali-
GAMES has been constructed for the computation bration of the model (i.e., estimation of the a,b
of seasonal potential soil loss (i.e., the soil loss compo- parameters) and for normal operational runs, delivery
nent) and sediment yield (i.e., the delivery compo- is defined for each cell from that cell to the point on
nent) for field-size cells, exhibiting homogeneous soil the main stream at the outlet of the watershed. This
characteristics, slope and land-use. The soil loss com- delivery is determined to include the delivery from
ponent is fundamentally the Universal Soil Loss the field cell across downstream cells to the main
Equation (USLE), modified for seasonal considera- stream, and the delivery down the stream. Delivery
tion. This equation presented by Wischmeier and across a number of cells is computed on the basis of

501 WATER RESOURCES BULLETIN

 Dickinson, Rudra, and Wall

the above equation, and delivery along the stream has Parkhill, Brookston, Harriston, and Perth soils are in
been assumed to be 100 percent. the Stratford Avon Watershed.
For determining representative values of the deliv-
ery ratio parameters (a and b) for each watershed, an
optimization routine was used to calibrate watershed
THE STUDY AREA AND DATA BASE sediment loads calculated by GAMES with measured
sediment loads. The basic output generated on a cell-
GAMES was applied to the Stratford Avon by-cell basis for each watershed included: potential
Demonstration Watershed and the Big Creek soil erosion, soil erosion per hectare, cell delivery ratio
Watershed, two of the PLUARG experimental basins to the adjacent cell, cell delivery ratio to the stream,
located in southern Ontario (Wall et al., 1978; Wall et sediment delivered to the stream, and sediment yield
al., 1979). The Stratford Avon Watershed is comprised per hectare. The range and distribution of the esti-
of 434 hectares (1072 acres) of rolling upland loam to mated soil erosion and sediment yield rates for the
silt loam soils and has been cropped predominantly in two watershed areas are presented in Table 2.
continuous corn. The Big Creek Watershed is charac- It is worth noting Big Creek Watershed is predicted
terized by 3300 hectares (8151 acres) of almost level to yield considerably more sediment, largely due to
lowland clay to clay loam soils (0 to 2 percent slopes). much greater predicted delivery ratios, than the
This basin is almost entirely row-cropped, with a pre- Stratford Avon Watershed. This behavior may at first
dominance of corn, beans, and tomatoes. Fall plough- seem anomolous, Big Creek being the much flatter
ing has been practiced throughout the two study watershed. However, soils in the Big Creek Watershed
watersheds, with few if any conservation practices in are much finer than those in the Stratford Avon
place until quite recently. Watershed, and it has been clearly observed, particu-
Since at least 80 percent of the annual sediment larly in the spring of the year when the flat fields are
load in each basin has been observed to be transport- awash in snowmelt and rain, that large quantities of
ed during the spring months (February-May), the fines remain in suspension in the runoff water,
GAMES was applied for conditions characteristic of even though that runoff is moving slowly over the sur-
that period. The input data base was developed for face. The result is high delivery rates and large sedi-
the 1982-1983 spring season in three stages: ment yields. These results were first observed and
discussed in PLUARG studies (Wall et al., 1978).
(1) Land use, soils and land slope were indepen-
dently ascertained from available maps, aerial pho-
tographs, and field surveys. Watershed maps for each
of the three variables were prepared at a base scale of REMEDIAL STRATEGIES
1:5,000.
(2) A composite overland of land use, soils and The possible effects of sequential changes in the
land slope was developed for each basin to divide the portion of each watershed undergoing soil and land
watershed area into field-size cells, each of which was management alterations have been studied. Four
characterized by a single land use, a single soil type, strategies for implementation of remedial measures
and a single class of slopes. have been explored:
(3) Values of the various variables and factors in
GAMES were determined and assigned to each land • Strategy A: Cells used in the model were select-
cell. ed for remediation at random from the total area of
each watershed in increments of 10 percent of the
Values of the cropping factor C assigned to spring total number of cells.
land use conditions of fall ploughed corn, spring • Strategy B: Cells were selected from targeted
grain, pasture and forages, and woodlot were 0.75, portions of each watershed in increments of 10 per-
0.70, 0.07, and 0.03, respectively. Other C values for cent of the cells in the targeted areas.
various combinations of cropping and tillage condi-
tions during the spring season were obtained from the B-i: The targeted portion included those cells
tabulations of Wischmeier and Smith (1978). Spring with spring sediment yield rates before
soil erodibility K values were set at 0.20 for the treatment of greater than one ton per
Waterloo soil type, 0.30 for the Huron and Parkhill hectare.
soils, and 0.37 for the Brookston, Harriston, and B-2: The targeted portion included those cells
Perth soils. The Waterloo and Brookston soils are with spring soil loss potentials before treat-
found in the Big Creek Watershed; and the Huron, ment of greater than three tons per
hectare.
WATER RESOURCES BULLETIN 502
 Targeting Remedial Measures to Control Nonpoint Source Pollution

TABLE 2. Distribution of Estimated Soil Erosion and Sediment Yield Rates for the Study Watersheds.

Erosion Rate Sediment Yield Rate

Area Exhibiting Area Exhibiting
Potential Estimated Potential Estimated Sediment Estimated Sediment
Erosion Rate Erosion Occurring Yield Rates Yield Contributed
Watershed Category5 Percent ha Percent ha Category55 Percent ha Percent ha
Stratford I 48.2 209.1 1.9 27.7 I 66.3 279.1 3.9 10.0
Avon II 4.1 17.9 2.4 35.6 II 11.9 51.6 7.1 19.6
(Upland) III 21.7 94.2 23.9 347.9 III 8.8 38.2 9.8 27.2
N 18.9 82.0 38.5 562.0 IV 6.0 26.0 12.5 34.6
V 7.1 30.8 33.3 485.8 V 9.0 39.1 66.7 184.8

Big Creek I 5.3 174.9 1.3 108.9 I 8.7 287.8 1.7 37.0
(Lowland) II 38.9 1,283.7 34.8 2,916.6 11 27.6 910.5 15.9 340.4
III 55.8 1,841.4 63.9 5,355.5 III 34.0 1,120.7 35.6 762.2
N 0.0 0.0 0.0 0.0 IV 29.7 981.0 46.8 1,001.8
V 0.0 0.0 0.0 0.0 V 0.0 0.0 0.0 0.0
Categories of Estimated Potential Erosion Rates:
I 0 —1.25tJha
II 1.25 —2.Stlha
III 2.5 —5.Otiha
N 5.0 —10.0 t/ha
V >10.Otiha
55Categories of Estimated Sediment Yield Rates:
I 0 —0.2StJha
II 0.25 —0.5tIha
III 0.5 — 1.0 tfha
N 1.0 —2.0 tiha
V >2.Ot/ha

The selection of cells was on the basis of the esti- substantial increase in n of 200 percent. The lesser
mated sediment yield rates in the case of B-i, and the level of treatment corresponds to the implementation
estimated soil loss rate in B-2, those cells with the of reiedial measures expected to introduce a moder-
greatest yield rates and loss rates being treated first. ate level of control (e.g., cross-slope farming, cessation
of fall moldboard ploughing, non sod-based rotations);
• Strategy C: Cells were selected at random from the greater level, to measures expected to introduce a
the targeted portion of each watershed in increments significant level of control (e.g., residue left over win-
of 10 percent of the number of cells in the targeted ter, conservation tillage, sod-based rotations).
areas. Potential spring soil loss rates and sediment yield
rates were recomputed for all cells of each watershed
C-i: The targeted portion corresponded to that for a number of incremental applications of each
of B-i. strategy and for both levels of control. The effects of
C-2: The targeted portion corresponded to that implementing the two levels of control in the various
of B-2. ways were considered in terms of the percentage
reduction on total watershed sediment yield. These
• Strategy D: Cells were selected from the row results are presented in Figures 1 through 4. For
crop portion of the Stratford Avon Watershed in incre- these watershed runs, for which all cells are similar
ments of 10 percent of the number of row crop cells in size, the percentage of cells can be considered to be
with slopes greater than 5 percent. The cells were equivalent to the percentage of the watershed area.
selected on the basis of the land slope of the cell, those
cells with the greatest slope being treated first.

For each strategy, two levels of remedial treatment DISCUSSION OF RESULTS

were applied to the cells selected: a 30 percent reduc-
tion in the cropping factor C, with an associated 40 Figures 1 to 4 illustrate that the criterion for
percent increase in the surface roughness factor n; selecting target areas on the basis of estimated spring
and a 60 percent reduction in C, with an associated
503 WATER RESOURCES BULLETIN
 Dickinson, Rudra, and Wall

C
C a,
a, E
E Va,
V
a, (0
a) C
C C
C 0
tV0 •0
a,
a,
it
it

% of total cells

Figure 1. Reduction in Sediment Yield From an Upland Basin Figure 2. Reduction in Sediment Yield From an Upland Basin
as a Result of the Application of Various Management as a Result of the Application of Various Management
Strategies Involving a Moderate Reduction in C Values. Strategies Involving a Substantial Reduction in C Values.

C C
a, 0)
E E
V V
a)
a, a)
a)
C C
C
C
0
=
tV0
=
a) a,
it it

60
%of total cells %of total cells

Figure 3. Reduction in Sediment Yield From an Lowland Basin Figure 4. Reduction in Sediment Yield From an Lowland Basin
as a Result of the Application of Various Management as a Result of the Application of Various Management
Strategies Involving a Moderate Reduction in C Values. Strategies Involving a Substantial Reduction in C Values.

soil loss rates has proven to be more useful than the results, the remaining discussion focuses on the out-
criterion based on estimated sediment yield rates. In puts relating to remedial strategies A, B-2, C-2, and
the case of the Stratford Avon Watershed, insufficient D.
cells exhibited estimated sediment yield rates greater It is instructive to consider some of the upper lim-
than one ton per hectare for application of strategies its of the curves presented in Figures 1 to 4 before
B-i and C-i. For the Big Creek Watershed, although exploring their shapes. The upper limits of the curves
upwards of 30 percent of the cells of the basin were developed from application of Strategy A are a func-
found to meet this criterion, the total estimated sedi- tion of the leveVsort of treatment and the watershed
ment load yielded by these cells was significantly less characteristics. The maximum percentage reduction
than the load estimated to emanate from that portion in sediment load is at least as great as the percentage
of the watershed targeted on the basis of estimated reduction in the cropping factor. For basins with mini-
spring soil loss rates. As a consequence of these mal opportunity for deposition of eroded material,
WATER RESOURCES BULLETIN 504
 Targeting Remedial Measures to Control Nonpoint Source Pollution

such as watersheds with few depositional areas for In the very flat Big Creek Watershed, targeting
coarse sediments or basins like the Big Creek remedial measures is more efficient than random
Watershed with fine sediments remaining in suspen- remediation but not as efficient as targeting in the
sion, the maximum percentage reduction in basin sed- upland watershed. Moderate treatment, targeted at
iment load is equal to the percentage reduction in C. the most erodible land constituting 10 percent of the
For a basin such as the Stratford Avon Watershed, Big Creek Watershed cells, can be expected to yield
with rolling terrain, loamy soils and ample opportuni- about a 14 percent reduction in the basin sediment
ty for deposition of the coarser eroded material, the load (Figure 3, Curve B-2); while substantial treat-
maximum percentage reduction is considerably ment, similarly targeted, can yield a 25 percent reduc-
greater than the percentage reduction in C, with a 47 tion (Figure 4, Curve B-2). The random
percent reduction in load for a 30 percent reduction in implementation of conservation practices in this flat
C and an 80 percent reduction in load for a 60 percent basin can be expected to yield less than a 10 percent
reduction in C. The upper limits of the strategy B-2 reduction in the basin sediment load for moderate
and C-2 curves are a function of the distribution of treatment, and about a 15 percent reduction for sig-
soil loss rates and amounts in the watersheds, and of nificant treatment, of 30 percent of the cells.
the soil loss tolerance value selected. A rolling basin Implementing remedial measures at random in the
such as the Stratford Avon Watershed with a wide targeted regions of the watersheds, identified on the
range of soil loss rates going well beyond the toler- basis of erosion rates in excess of a tolerable value
ance value, and in which a relatively small percent- (I.e., strategy C-2), can be expected to be more effi-
age of the area contributes a relatively large cient than nontargeting of the measures, but less effi-
percentage of the erosion, exhibits a relatively large cient than implementing measures in the most
upper limit in the percentage reduction in sediment erodible fields first (i.e., Strategy B-2), until measures
load at a relatively small upper limit in the percent- have been implemented in the entire targeted area.
age of the cells treated. A flat basin such as the Big These results confirm that some targeting remains
Creek Watershed, with a narrow range of soil loss more effective than no targeting, but that treating the
rates, some of which exceed the selected tolerance "worst fields first" is clearly the most efficient way to
value, exhibits a relatively small upper limit in the reduce sediment loads in the study basins.
sediment load reduction percentage. Particularly for For the basin with some relief, the Stratford Avon
basins such as the latter, it is relatively easy to select Watershed, it is interesting to note the targeting of
a soil loss tolerance that is either completely above or soil conservation practices on the basis of slope alone
completely below all estimated soil loss rates, the (i.e., Strategy D) is more efficient for reducing the
range of estimated soil loss rates being so narrow. basin sediment load than nontargeting, is less effi-
Obviously, if the tolerance value were selected to be cient than targeting on the basis of Strategy B-2, and
above the range, the upper limits of the B-2 and C-2 is similar in efficiency to the results from targeting on
curves would be zero; if the tolerance value were the basis of Strategy C-2 (Figures 1 and 2). Strategy
below the range, the upper limits would correspond to D was explored to evaluate the relative effectiveness
those of the strategy A curves. of using a very simple basis for targeting, such as
The results clearly reveal that the targeting of slope gradient, as opposed to the somewhat more
remedial measures is a highly efficient technique for involved application of model algorithms such as
reducing sediment loads in the rolling Stratford those employed in GAMES.
Watershed. For example, moderate treatment (i.e., 30 The relative reductions in sediment load yielded by
percent reduction in C plus an associated change in the two levels of treatment, i.e., the 30 percent and 60
n), targeted at the most erodible land contributing 10 percent reductions in C values plus associated
percent of the watershed cells, can be expected to changes in n, are almost but not necessarily linear.
yield reduction in the basin sediment load of up to 30 That is, the reduction in sediment load due to the
percent (Figure 1, Curve B-2). Substantial treatment greater level of treatment is in the vicinity of double
(i.e., 60 percent reduction in C, etc.), similarly target- the load reduction due to the lesser level of treatment.
ed, can yield a 50 percent reduction in load (Figure 2, The degree of nonlinearity varies from basin to basin.
Curve B-2). In contrast, the nontargeting of remedial Although for the model, selected sediment yield is lin-
measures can be expected to yield only up to 15 per- early related to the cropping factor, C, it is not linear-
cent reduction in basin sediment load for moderate ly related to the surface roughness, n.
treatment of 30 percent of the cells (Figure 1, Curve Another informative way of considering the results
A), and up to a 30 percent reduction for significant is to determine the percentage of cells requiring reme-
treatment of the same percentage of cells in the dial measures, for each level of treatment and for
Stratford Watershed (Figure 2, Curve A). each implementation strategy, to achieve a preselect-
ed percentage reduction in sediment load. Suppose,

505 WATER RESOURCES BULLETIN

 Dickinson, Rudra, and Wall

TABLE 3. Percentage of Watershed Cells Requiring Remedial Measures to Achieve a

Selected Percentage Reduction in Basin Sediment Loads.

Implementation Strategy
A B-2 C-2

FOR A 30 PERCENT REDUCTION IN LOAD

Upland Watershed
30% reduction in C 62 12 25 NA**
60% reduction in C 32 5 15 15

Lowland Watershed
30% reduction in CS 100 NA NA NA
60% reduction in CS 52 13 18 NA

FOR A 10 PERCENT REDUCTION IN LOAD

Upland Watershed
30% reduction in C5 22 3 8 7
60% reduction in C 13 2 4 4

Lowland Watershed
30% reduction in C* 36 6 12 NA
60% reduction in C 17 2 6 NA

5Plus associated changes in surface roughness factor.

Not applicable.

for example, the goal was to reduce sediment loads by source pollution model such as GAMES for exploring
10 to 30 percent. The manners in which such a goal the effectiveness of not only various levels of remedia-
could be expected to be achieved are summarized in tion but also various strategies of implementing soil
Table 3. These results again clearly reveal the sub- and water conservation practices has also become
stantial advantage of targeting remedial measures clear.
where possible. If in addition one considers that the
adoption rate for implementing agricultural extension
recommendations is rarely above 30 percent, these
results further indicate that targeting of remedial LITERATURE CiTED
measures may be not only effective but necessary to Beasley, D. B., L. F. Huggins, and E. J. Monke, 1980. ANSWERS: A
achieve a specified reduction in basin sediment load. Model for Watershed Planning. Trans. ASAE 23:938-944.
For the study basins selected, targeting would appear Beasley, D. B., L. F. Huggins, and E. J. Monke, 1982. A Monitoring'
to be essential to achieve sediment load reductions Modeling Strategy for 208 Implementation. Trans. ASAE 25(3):
654.660.
greater than 20 percent. Beasley, D. B., L. F. Huggins, and E. J. Monke, 1982. Modeling
Sediment Yields From Agricultural Watersheds. Journal Soil
Water Cons. 37(2):114.117.
Clark, D. J., 1981. An Expression for Determining Sediment
CONCLUSIONS Delivery Radio. M.Sc. Thesis, University of Guelph, Guelph,
Ontario.
Clark, E. H., J. A. Haverkamp, and W. Chapman, 1985. Eroding
This study, involving the application of a nonpoint Soils: The Off-Farm Impacts. The Conservation Foundation,
source model to a rolling upland and a flat lowland Washington, D.C.
Cook, D. J., W. T. Dickinson, and R. P. Rudra, 1985. GAMES — The
basin, clearly reveals that the targeting of remedial Guelph Model for Evaluating Effects of Agricultural
soil and water conservation measures can be a highly Management Systems on Erosion and Sedimentation. Tech.
efficient approach, and may be a necessary approach, Report No. 126.71, School of Engineering, University of Guelph,
to reducing basin sediment loads and associated pol- Guelph, Ontario.
Cook, D. J., W. T. Dickinson, and R. P. Rudra, 1985. A
lution. Targeting is more efficient and more likely to Microcomputer Model for the Evaluation of Conservation
be necessary in watersheds in rolling terrain, where Systems. ASAE Paper No. 85-2044, Amer. Soc. Agr. Eng., St.
the ranges of soil loss and sediment yield rates are Joseph, Michigan.
relatively, large, and a small percentage of the area Dickinson, W. T. and R. Pall, 1982. Identification and Control of
Soil Erosion and Fluvial Sedimentation in Agricultural Areas of
generates a relatively large percentage of the soil loss the Canadian Great Lakes Basin. Research Report to Supply
and sediment load. The usefulness of a nonpoint
WATER RESOURCES BULLETIN 506
 Targeting Remedial Measures to Control Nonpoint Source Pollution

and Services Canada, School of Engineering, University of So, S. K. and S. N. Singer, 1982. Rural Nonpoint Source Pollution
Guelph, Guelph, Ontario. and Control. Grand River Basin Water Management Study
Dickinson, W. T., R. Pall, and G. J. Wall, 1984. GAMES: A Method Tech. Report No. 27, Ontario Ministry of the Environment,
for Identifying Sources and Amounts of Soil Erosion and Fluvial Toronto, Ontario.
Sediments. Proc. Canadian Hydrologr Symposium 15:805-824. Spurlock, S. R. and I. D. Clifton, 1982. Efficiency and Equity
Dickinson, W. T., R. P. Rudra, and G. J. Wall, 1986. Discrimination Aspects of Nonpoint Source Pollution Controls. Southern
of Soil Erosion and Fluvial Sediment Areas. Can. Jour. Earth Journal Agr. Econ. 14(2):123-129.
Sci., 22:1112-1117. Wall, G. J., W. T. Dickinson, and L. P. J. van Vliet, 1979.
Dickinson, W. T., R. P. Rudra, and G. J. Wall, 1986. Identification of Agricultural Sources of Fluvial Suspended Sediments. Prog.
Soil Erosion and Fluvial Sediment Problems. Hydrologic Water Tech. 11(6):481-499.
Processes 1:111-124. Wall, G. J., L. P. J. van Vliet, and W. T. Dickinson, 1978.
Duda, A. M. and R. J. Johnson, 1985. Cost-Effective Targeting of Contribution of Sediments to the Great Lakes From
Agricultural Nonpoint-Source Pollution Controls. Journal Soil Agricultural Activities in Ontario. I. J. C. Tech. Rep. Fe:
Water Cons. 40(1):108-111. PLUARG, Windsor, Ontario.
Forster, D. L., T. L. Logan, S. M. Yaksich, and J. R. Adams, 1985. Wischmeier, W. H. and D. D. Smith, 1978. Predicting Rainfall
An Accelerated Implementation Program for Reducing the Erosion Losses — A Guide to Conservation Planning. USDA
Diffuse-Source Phosphorus Load to Lake Erie. Journal Soil Agricultural Handbook No. 537, Washington, D.C.
Water Cons. 40(1):136-141.
Henry, J. R., 1981. Agricultural Conservation Program: An
Evaluation. Water Resources Bulletin 17(3):438-442.
Knisel, W. R., Jr., 1980. CREAMS: A Field Scale Model for
Chemical, Runoff, and Erosion From Agricultural Management
Systems. USDA Soil Cons. Res. Report No. 26:486-492.
Lee, L. K. and J. J. Goebel, 1986. Defining Erosion Potential on
Cropland: A Comparison of the Land Capability Class-Subclass
System With RKLSPF Categories. Journal Soil Water Cons., pp.
41-44.
Lee, J. G., S. B. Lovejoy, and D. B. Beasley, 1985. Soil Loss
Reduction in Finley Creek, Indiana: An Economic Analysis of
Alternative Policies. Journal Soil Water Cons. 40(1):132-135.
Maas, R. P., M. D. Smolen, and S. A. Dressing, 1985. Selecting
Critical Areas for Nonpoint-Source Pollution Control. Journal
Soil Water Cons. 40(1):68-71.
Myers, P. C., 1986. Nonpoint Source Pollution Control: The USDA
Position. Journal Soil Water Cons. 41(3):156-158.
Nonpoint Source Control Task Force, 1983. Nonpoint Source
Pollution Abatement in the Great Lakes Basin — An Overview of
Post-PLUARG Developments. Report to the Great Lakes Water
Quality Board, International Joint Commission, Windsor,
Ontario.
Ogg, C. W., J. D. Johnson, and K. C. Clayton, 1982. A Policy Option
for Targeting Soil Conservation Expenditures. Journal Soil
Water Cons. 37(2):68-72.
Ogg, C. W. and H. B. Pionke, 1986. Water Quality and the New
Farm Policy Initiatives. Journal Soil Water Cons. 41(2):85-88.
Ontario Chapter Soil Conservation Society of America and the
Ontario Institute of Agrologists, 1986. Position Update:
Agricultural Soil and Water Conservation. Ontario Chapter
SCSA, Whitney Block, Queens Park, Toronto, Ontario.
Pierce, F. J., W. E. Larson, R. H. Dowdy, and W. A. P. Graham,
1983. Productivity of Soils: Assessing Long Term Changes Due
to Soil Erosion. Journal Soil Water Cons. 38(1):39-44.
PLUARG, 1978. Environmental Management Strategy for the
Great Lakes System. Final Report of the Pollution from Land
Use Activities Reference Group to the International Joint
Commission, Windsor, Ontario.
Rudra, R. P., W. T. Dickinson, D. J. Clark, and G. J. Wall, 1986.
GAMES — A Screening Model of Soil Erosion and Fluvial
Sedimentation on Agricultural Watersheds. Canadian Water
Resources Journal.
Runge, C. F., W. E. Larson, and G. Roloff, 1986. Using Productivity
Measures to Target Conservation Programs: A Comparative
Analysis. Journal Soil Water Cons. 41(1):45-49.
Schertz, D. L., 1983. The Basis of Soil Tolerance. Journal Soil Water
Cons. 38: 10-14.
Snell, E. A., 1984. A Manual for Regional Targeting of Agricultural
Soil Erosion and Sediment Loading to Streams. Working Paper
No. 36, Lands Directorate, Environmental Conservation Service,
Environment Canada, Ottawa, Ontario.

507 WATER RESOURCES BULLETIN