United States
Environmental Protection
Agency
Environmental Research
Laboratory
Athens GA 30613
Research and Development
EPA-600/S3-83-054 Jan. 1984
Project Summary
Fluvial Transport and Processing
of Sediments and Nutrients in
Large Agricultural River Basins
David B. Baker
Rivers and streams draining water-
sheds in northwestern Ohio were sam-
pled in a 3- to 5-year study to describe
the transport of nutrients and sedi-
ments in large agricultural basins. Land
use in the watersheds, which ranged in
size from 171 to 16,395 km2, was
dominated by row crop agriculture on
generally fine textured and poorly
drained soils. Automatic samplers were
used to collect at least four samples
per day at 12 U.S. Geological Survey
gauging stations. During storm events
all samples were analyzed, whereas
during non-event periods one sample
per day was analyzed.
For these rivers the concentrations
of suspende jlids, total phosphorus,
nitrate plus nitrite nitrogen and total
Kjeldahl nitrogen increased with in-
creasing stream flow. Storms with
similar peak discharges had widely
varying flux weighted concentrations
of both sediments and nutrients. Large
seasonal and annual variations in flux
weighted concentrations were also ob-
served. The ratio of particulate phos-
phorus to sediments varied greatly
among samples, with lower ratios gen-
erally associated with higher suspended
solids concentrations.
The unit area yields of nutrients were
considered to be in the high range for
agricultural watersheds. Sediment de-
I ivery ratios ranged from 6.2% to 11.9%.
There was no correlation between sed-
iment delivery ratio and basin size.
Furthermore, there was no correlation
between gross erosion rates and unit
area nonpoint phosphorus yields.
Phosphorus entering streams from
point sources was rapidly processed by
the stream system. Subsequent trans-
port of this phosphorus to the lake
depended on resuspension of particu-
late phosphorus during storm events.
The soluble reactive phosphorus ex-
ported during storm events was largely
derived from nonpoint sources. Upon
delivery to Lake Erie, phosphorus from
nonpoint sources had a higher per-
centage availability than phosphorus
derived from point sources and pro-
cessed by the stream system.
The data illustrate many patterns of
nutrient and sediment transport in river
systems. Many of these patterns need
to be taken into account in establish-
ing water quality monitoring programs.
These data should also be useful in
evaluating the effectiveness of con-
servation tillage in controlling agricul-
tural nonpoint source pollution.
This Project Summary was developed
by EPA's Environmental Research Lab-
oratory, Athens GA. to announce key
findings of the research project that is
fully documented in a separate report
(see Project Report ordering informa-
tion at back).
Background
Studies of nonpoint pollution in the
Great Lakes Basin indicate that the rivers
of northwestern Ohio, which drain into the
western and central basins of Lake Erie,
carry the largest loads of agriculturally
derived nutrients and sediments entering
the entire Great Lakes system. These loads
are a product of intensive, row-crop agri-
culture on fine textured soils in the rela-
tively large river basins of this regioa
Because these loads comprise a large
portion of the phosphorus entering Lake
-------
Erie, the transport and processing of sed-
iments and nutrients in the major rivers of
this region have received detailed study.
The objectives of these studies have been
to document and characterize the existing
pollutant loads, to identify critical sub-
watersheds, to compare the transport and
fate of pollutants derived from point and
nonpoint sources, and to develop baseline
data to evaluate water quality benefits of
nonpoint control programs.
Modeling studies of the phosphorus
balance for Lake Erie indicate that reduc-
tions in phosphorus loading from agri-
cultural sources will be necessary to
achieve the desired reductions in eutro
phication. Concurrently with the river
transport studies, the U.S. Army Corps of
Engineers conducted detailed studies to
determine the applicability of various agri-
cultural best management practices for
reducing nutrient and sediment losses from
cropland in these river basins. These
studies concluded that a variety of con-
servation tillage practices, including no-
till, could significantly and economically
reduce phosphorus loading to Lake Erie.
A major tillage demonstration project
was initiated in the Honey Creek Water-
shed of the Sandusky Basin with support
from the U.S. Army Corps of Engineers
and the Agricultural Stabilization and Con-
servation Service. The results of the Honey
Creek Project and other tillage demon-
stration projects in the area, confirm that
many of the area's soils are suitable for
conservation tillage management and that
these methods, when properly imple-
mented, offer economic advantages to
farmers. The Great Lakes National Program
Office of the U.S. EPA has supported a
conservation tillage implementation pro-
gram in 20 agricultural counties within the
Lake Erie Basin.
Studies of nutrient and sediment trans-
port in northwestern Ohio rivers have been
supported by grants and contracts from
several different agencies and organiza-
tions. These include: 1) research grants
from the U.S. EPA for studies of flow
augmentation and river transport; 2) con-
tracts with the Toledo Metropolitan Area
Council of Governments as part of their
208 Planning Study; 3) contracts with the
US. Army Corps of Engineers for data
collection in support of the Lake Erie
Wastewater Management Study; 4) con-
tracts with the Ohio Department of Natu ral
Resources for reservoir management stud-
ies; 5) grants from the Rockefeller Founda-
tion and the Soap and Detergent Associa-
tion for studies of agricultural nonpoint
pollution; 6) grants from the cities of
Tiffin, Bucyrus, and Upper Sandusky for
evaluating instream benefits of phospho-
rus removal programs; and 7) support
from Heidelberg College for matching
funds and data collection during interim
periods between external funding sources.
A selected list of related reports is in-
cluded at the end of this summary.
The results of this study have been used
extensively for the development of recom-
mendations contained in the Final Report
of the Lake Erie Wastewater Management
Study. The data have also been used in
calibrations of several nonpoint source
models. A detailed river transport model
was developed by the U.S. Army Corps of
Engineers using Sandusky Basin data. A
generalized version of the U.S. Army Corps
river transport model is presented in the
Appendix to the complete report
Study Area
The study area, which contains 12 U.S.
Geological Survey stream gauging sta-
tions, is located in northwestern Ohio. The
stations included river mouth stations on
the Maumee, Portage, Sandusky and Huron
rivers where data on tributary loading to
Lake Erie were obtained. In addition, sta-
tions were operated on five major trib-
utaries to the Sandusky River and three
additional Sandusky River sites to identify
critical watersheds and to study the trans-
port and processing of nutrients and sedi-
ments as they move through stream
systems.
Study Methods
Automatic samplers were used to col-
lect a minimum of four samples per day at
each sampling station. During periods of
high flow and/or high turbidity, each sam-
ple was analyzed. Under low flow con-
ditions, only a single sample per day was
analyzed. Analyses routinely included sus-
pended solids (SS), total phosphorus (TP),
soluble reactive phosphorus (SRP), nitrate
plus nitrite nitrogen, ammonia, conduc-
tivity, and chloride. The U.S Geological
Survey provided stage data for the times of
sample collection and stage-discharge
rating tables for each station.
This report focuses on data collected
through the 1979 water year. Atthattime
from three to five years of data were
available at each station. Since 1979,
studies have continued at several of the
transport stations, and pesticide analyses
have been added. All data have been
placed in the STORET system.
Results
Flux-weighted and time-weighted mean
concentrations were developed for major
parameters at each sampling station.
These are shown in Table 1. Parameters
whose concentrations tend to increase
with increasing stream flow SS, TP, and
nitrates have higher flux-weighted than
time-weighted concentrations. The Bucyrus
station is located a short distance down-
stream from the municipal sewage treat-
ment plant At that site, the time-weighted
concentration of TP exceeded the flux-
weighted concentration. Conductivity al-
ways decreased with increasing flow,
whereas the time-weighted concentrations
of SRP were higher than the flux-weighted
concentrations at stations below signifi-
cant point source inputs.
In Figure 1, a typical hydrograph, chemo-
graph and sediment graph pattern for
northwestern Ohio rivers is shown. The
peak concentrations of SS and TP occur in
advance of the peak discharge whereas
peak nitrate concentrations trail the peak
discharge. Concentrations of TP (not
shown) parallel the SS concentrations.
Often SRP concentrations increase in
parallel with the TP concentrations. The
minimum conductivity generally corre-
sponds with the peak discharge.
The trailing nitrate peaks are associated
with the delayed arrival at the stations of
nitrate-laden tile effluents relative to the
arrival of surface runoff with its high SS
and TP concentrations. Sediment resus-
pension during the rising stage of the
hydrograph also helps to account for the
advance sediment and phosphorus peaks.
At each of the sampling stations, the
flux-weighted concentrations of sediments
and nutrients showed large variations be-
tween individual storms, seasons and
years. At the Upper Sandusky station, the
flux-weighted concentrations were calcu-
lated for 52 individual storms. In Figure 2
these concentrations for TP are plotted in
relation to the peak flow for the storm.
Storms with similar peak flows have
widely varying concentrations of both TP
and sediments over the entire range of
peak flows. Although some of the vari-
ability is associated with differences be-
tween snow melt and rainfall induced
events, much of it is related to variations in
rainfall intensity and amount pre-existing
soil moisture and ground cover. This large
variability in phosphorus and sediment
concentrations for storms with similar
peak flows illustrates the complexity as-
sociated with calibrating runoff models for
large watersheds.
There are large annual variations in
discharge, load and flux-weighted mean
concentrations of sediments and nutrients
at the stations. The concentrations depend
-------
Table 1. Comparison of Flux Weighted and Time Weighted Mean Concentrations of Suspended Solids and Nutrients in Northwestern Ohio River
Basins
Location
Maumee
Portage
Huron
Sandusky
Mexico
Weighted
N Concentration
1769 Flux
Time
1842 Flux
Time
2027 Flux
Time
2237 Flux
Time
1578 Flux
Time
Upper Sandusky 2729 Flux
Bucyrus
Tymochtee Cr.
Honey Cr.
Time
2299 Flux
Time
2631 Flux
Time
2115 Flux
Time
Broken Sword Cr. 1766 Flux
Wolf, East
Wolf, West
8 3 "
rx »»
_^-0-
0 i to" 3E. ^ -^
^~ «- t
^,§ -
c *° ^ci.
CTl ^
c§- £ £
++ +ft *c.
<0 U) v,
5 ,-t- O ~^
c- *E *O r/*
*- . **)
<38--I |
£ t c
lo- is !
*
Time
1654 Flux
Time
1699 Flux
Time
§rx
_o I v A
I ^" KrV^
§ - /if \
8 -" \ Ai\ \
* & \HU\ \
?3 yff \ \
s « A/ 1 \
^ ^0 /U/ \
Is /It \ j
£ liA \ /
8 to / \ Jbc
__*_*/
* ; 1 !-** i
10. 12. 14. 16.
Day of April
Suspended
Sediment
mg/L
242
106
164
62
220
69.6
217
83.1
239
85.8
235
105
173
49.6
205
68.7
180
57.8
244
78.3
181
42.4
183
40.8
1
\
\
\ */
\x
. /<
\ / X
\/ ^
y
A
V
i ^^
v ^r^
*^--*~ -,
18
1981
Soluble NO3-NO2
Total P, Reactive P, Nitrogen, Conductivity
mg/L
0.516
0.340
0.402
0.360
0.362
0.343
0.453
0.244
0.428
0.250
0.518
0.482
0.573
1.13
0.419
0.181
0.403
0.195
0.401
0.157
0.416
0.161
0.394
0.232
^ /""
^\^
^
^_
^^
-*^
^sTT-
^~-~-_
20. 22
mg/L mg/L ivnhos
0.116 4.92 517
0.116 3.91 652
0. 1 19 5.89 554
0.191 3.80 854
0.104 3.61 540
0.201 2.50 685
0.093 4.61 487
0.073 3.09 716
0.070 3.50 624
0.069 3.49 750
0.134 3.90 478
0.234 2.60 709
0.219 3.42 460
0.837 3.11 702
0.069 5.12 397
0.040 3.48 751
0.101 4.85 397
0.102 3.74 587
0.061 4.87 428
0.042 3. 1 1 637
0.118 4.71 578
0.063 292 764
0. 100 6. 14 464
0.133 3.45 747
more on the proportion of winter-to-
summer runoff than on the total amount of
runoff for the year.
Calculated paniculate phosphorus (PP)
to sediment ratios show that because
there are substantial variations in annual
flux-weighted concentrations of SS, there
are also large annual variations in phos-
phorus/sediment ratios. Models that esti-
mate PP export based on sediment yields
should take into account the effects of
sediment concentrations on PP/SS ratios.
Mean annual loads of nutrients and
sediments at each station were estimated
using two methods. Either flux-weighted
concentrations were multiplied by mean
annual discharge based on long term
hydrological records for each station or
flow duration tables were used in com-
bination with flux-weighted concentra-
tions for each flow interval. The two
methods gave very similar results. The
resulting mean annual unit area yields are
Figure 1. Typical patterns of changing concentrations of suspended sediments, nitrates and shown in Table 2. These yields are con-
conductivity during a runoff event at the Me/more gaging station on Honey Creek.
sidered to be in the high range for agri-
-------
cultural regions that have been studied,
especially considering the large size of the
drainage areas.
As part of a related study, gross erosion
rates were calculated for each of the study
watersheds. These erosion rates coupled
with the unit area sediment yields allow
direct calculation of sediment delivery
ratios. Both the erosion rates and sed-
iment delivery ratios are also shown in
Table 2.
The sediment delivery ratios ranged
from 6.2% to 11.9%. The general ten-
dency of delivery ratios to decrease with
increasing basin area, however, was not
evident for these streams. These basins
were much larger than those in which
previous measurements of sediment de-
§T
livery have been made. Sediment delivery
ratios were inversely correlated with gross
erosion rates for the study watersheds.
Concentration exceedency calculations
for nitrates, for example, indicate that
Ohio's drinking water standard (10 mg/L
nitrate-N) was exceeded from 2% to 5% of
the time at the transport stations (see
Table 3). Concentrations of 7 mg/L are
exceeded more than 10% of the time at all
of the stations. Widespread adoption of
conservation tillage in this region could
increase the duration of time when nitrate
standards are exceeded by increasing tile
flow in relation to surface runoff.
Flux exceedency calculations confirm
the dominant role of periods of high flux in
total stream transport of sediments and
cs
5
«
+ +
+ ++
/*
0.5
Figure 2.
1.0 1.5
Log (Peak Flow) mfl/s
2.0
2.5
Relationships between flux-weighted mean concentrations of total phosphorus for
individual runoff events and the peak flow for the events. Storm event data were
taken from the Upper Sandusky gaging station on the Sandusky River.
nutrients. For the Maumee station the 5%
of the time with the highest fluxes ac-
counted for the following percentages of
total stream transport: SS, 70%; TP, 57%;
SRP, 39%; dissolved solids, 30.3%; and
water 39.2%. For the Honey Creek station
the corresponding figures were: SS, 79%;
TP, 65%; SRP, 50%; dissolved solids,
30%; and water 44.4%. In general, as the
watersheds become smaller, greater per-
centages of the total transport are ac-
counted for by short periods of high flux.
The nonpoint phosphorus yields from
each basin were calculated by subtracting
the upstream point source inputs. The
resulting unit area nonpoint phosphorus
yields are shown also in Table 2. This
calculation assumes 100% delivery of
point source inputs through the stream
system. It is recognized that for stations
with substantial point source inputs, such
as Bucyrus and Portage, the resulting unit
area nonpoint phosphorus yields are un-
realisticaily low. These same stations have
sediment yields similar to adjacent basins.
One would expect similar nonpoint phos-
phorus-to-sediment ratios regardless of
the presence or absence of point sources
in the basins. These data suggest that the
delivery of point source phosphorus through
stream systems is substantially less than
100%.
Much of the phosphorus that enters
stream systems from point sources is
quickly removed from the flowing water.
In the Sandusky Basin, the point source
phosphorus loading rate upstream from
the Fremont station was 5.3 kg/hr. At this
station the phosphorus flux was less than
5.3 kg/hr 55% of the time while during
this time the average flux was only 1.8
kg/hr. Although these periods of low
phosphorus flux are associated with low
stream flows, most of the flow is derived
from base flow in the watershed rather
than point source inputs. Given the low
flow phosphorus concentrations in water-
Table 2. Unit Area Yields of Sediments and Nutrients for Northwestern Ohio Agricultural Watersheds
Location
Maumee
Portage
Huron
Sandusky
Mexico
Upper Sandusky
Bucyrus
Tymochtee Cr.
Honey Cr.
Broken Sword Cr.
Wolf, East
Wolf, West
Total P,
kgha-'yr-1
1.34
0.97
1.02
1.10
1.17
1.45
1.85
1.06
1.09
1.43
1.40
1.02
Non-PL P
Yield,
kgha'1 yf1
1.14
0.61
0.57
0.96
0.98
1.04
0.68
1.06
1.01
1.43
1.40
1.03
Soluble
Reactive P,
kgha'1 yr1
0.30
0.30
0.29
0.24
0.21
0.39
0.74
0.18
0.27
0.22
0.39
0.26
NOrNO2
Nitrogen
kg ha'' yr'1
13.1
15.1
10.2
12.5
12.2
11.4
11.5
13.4
13.1
17.3
15.8
16.0
Suspended
Sediment
ton ha'1 yr'1
0.63
0.40
0.59
0.52
0.65
0.63
0.54
0.52
0.49
0.87
0.61
0.48
Gross
Erosion
ton ha'' yr'1
6.84
5.00
7.51
8.25
9.37
9.35
7.85
8.41
6.86
9.39
5.11
4.19
Delivery
Ratio,
%
9.2
8.0
7.9
6.3
6.9
6.8
6.9
6.2
7.1
9.2
11.9
11.5
-------
sheds lacking point sources, most of the
low flow phosphorus flux must be derived
from nonpoint sources. Phosphorus con-
centration profiles in stream reaches below
point source inputs confirm the rapid
deposition of point source derived phos-
phorus. Data described above suggest
that some of this point source derived
phosphorus is not resuspended and ex-
ported from the stream system.
Although point source phosphorus
enters streams largely as bioavailable SRP,
this phosphorus is rather quickly assimi-
lated into the stream bed. Its subsequent
export is apparently in the form of PP that
is resuspended during storm events. Com-
parison of the bioavailability of PP, as
measured by traditional extraction tech-
niques, at sites downstream from large
point source inputs with the bioavailability
of PP exported from watersheds lacking
major point sources, shows no differences.
Apparently, point source SRP has under-
gone similar chemical transformations as
fertilizer SRP by the time it is exported
from the watersheds.
During runoff events, much more SRP is
exported from these rivers than can be
accounted for by upstream point source
inputs entering the streams during the
storm events. Watersheds lacking point
source inputs have similar SRP export as
those containing point sources. Conse-
quently, most of the SRP exported during
storm events must be derived from non-
point sources. This SRP exported during
storm events may be an important source
of bioavailable phosphorus entering the
open water of Lake Erie. Its rapid passage
through estuaries and bays would make it
much less subject to estuarine or near-
shore processing than SRP derived from
point sources emptying directly into estu-
aries or the nearshore zone of the lake.
Because of the extensive export of non-
point-derived SRP during runoff events,
the overall bioavailability of nonpoint-
derived phosphorus is greater than the
bioavailability of point source-derived
phosphorus at the time of its export from
the river basins.
Conclusions
1. A combination of both detailed and
long-term studies is necessary to accu-
rately characterize the export of nutri-
ents and sediments from large agri-
cultural watersheds such as those of
northwestern Ohio. Most of the ma-
terial export occurs during runoff eventa
During individual events the concen-
trations of various components change
rapidly and in characteristic patterns.
The flux-weighted mean concentrations
for runoff events of the same size,
however, can show large variations
from storm-to-storm over a wide range
of storm sizes. Only part of this vari-
ability can be explained on the basis of
seasonal effecta Large annual variations
in flux-weighted means are also present
Ratios of phosphorus-to-sediment ex-
port vary from year to year. Attempts to
characterize the export of nutrients and
sediments based on short term studies,
no matter how detailed they might be,
could give very misleading results. An
appreciation of the patterns of vari-
ability observed in these studies could
be useful in designing and interpreting
monitoring studies in other areas.
2. The large variability in nutrient and
sediment loads for storms of equal size
suggests that material export is not
limited by the transport capacity of the
rivers but rather by the movement of
materials from the land surface to the
stream system. Thus, reductions in
river export should accompany non-
point control programs that reduce
material transport from the land sur-
face to the stream systems. Possible
increases in stream bed or stream bank
erosion associated with the sediment
carrying capacity of the stream will not
prevent the effectiveness of land man-
agement control programs.
3. The sediment delivery ratios in the,
study watersheds range from 6.2 to
11.9%, whereas average gross erosion
rates vary from 4.2 ton to 9.4 ton/ha/yr.
The delivery ratios are not correlated
with the size of the watershed but are
inversely correlated with the gross ero-
sion rate. There is no correlation be-
tween gross erosion rates and unit area
nonpoint phosphorus yields. These
observations raise doubts about the
concept of "critical areas" and the use
of gross erosion rates in their identifi-
cation. The sediment and phosphorus
yields from these large watersheds
may be more related to the amount of
clay entrained by rain drop impact and
subsequent surface runoff. Tillage prac-
tices that increase cover and/or de-
crease runoff could reduce sediment
and phosphorus yields from areas of
both high and low gross erosion.
Table 3. Percentage of Time the Indicated Concentrations of Nitrate-Nitrogen fmg/l) were Exceeded at Representative Gaging Stations
Exceedency Maumee
Portage
Tindall
Me/more
Wolf West
Nevada
99%
98%
95%
90%
80%
70%
60%
50%
40%
30%
20%
10%
5%
2%
1%
0.5%
.01 mg/l
.03
.13
.44
1.45
2.01
2.57
3.20
4.30
5.32
6.53
8.02
9.10
10.60
12.30
13.00
.02 mg/l
.06
.21
.47
1.07
1.70
2,21
2.93
3.79
4.80
6.43
8.69
10.00
11.89
13.10
14.10
.02 mg/l .480 mg/l
.03
.09
.240
.510
1.190
1.800
2.560
3.140
4.100
5.410
7.200
8.800
10.900
13.820
16.460
.770
.970
1.260
1.670
2.020
2.340
2.820
3.330
3.990
4.990
7.380
9.360
13.300
16. 1 10
18.520
.040 mg/l
.070
.110
.240
.500
.900
1.55
2.66
3.61
4.54
6.16
8.21
9.70
11.30
14.60
16.39
.059 mg/l
.070
.130
.240
.500
.800
1.620
2.230
2.870
3.800
5.000
7.200
9.440
13.00
16.00
18.80
Watershed
area
Km2
16,395
1,109
3,240
386
171.5
271
Gross Erosion
mt/ha/yr.
6.84
5.00
8.25
6.86
4.19
9.31
-------
4. Both concentration profiles and flux
exceedency data indicate that point
source phosphorus, most of which is in
the form of soluble reactive phosphorus
when it enters streams, is rapidly proc-
essed by the stream sediments. There
is indirect evidence that less than
10096 of the point source derived
phosphorus is subsequently delivered
out of the stream system. Limited data
suggest that the bioavailability of par-
ticulate phosphorus exported during
storms is no greater from watersheds
containing large point source phos-
phorus inputs than from watersheds
lacking point sources. Apparently, the
bioavailability of point source derived
paniculate phosphorus is no greater
than that of nonpoint source derived
paniculate phosphorus. Most of the
soluble reactive phosphorus exported
during storm events is derived from
nonpoint sources. The short retention
time of storm flows in the lower por-
tions of rivers or in estuaries, bays and
the nearshore zone may result in high
deliveries of nonpoint derived soluble
phosphorus through these zones to
the open lake. In contrast, point source
phosphorus entering these zones may
be subject to deposition and trans-
formation to less bioavailable forms.
5. The data sets on nutrient and sediment
transport in northwestern Ohio rivers
provide baseline data upon which the
regional water quality benefits of con-
servation tillage can be assessed. Ac-
celerated conservation tillage imple-
mentation programs are underway
throughout the region.
Related Reports
Cahill, Thomas H., R. W. Pierson, Jr., and
B. R. Cohen. 1979. Nonpoint Source
Model Calibration in Honey Creek Water-
shed. U.S. EPA. Athens, Georgia. EPA-
600/3-79-054. 1 34 pp.
U.S. Army Corps of Engineers, Buffalo
District 1982. Lake Erie Wastewater
Management Study. Final Report 223pp.
Verhoff, Frank H. and David B. Baker.
1981. Moment Methods for Analyzing
River Models With Application to Point
Source Phosphorus. Water Research,
15:493-501.
Verhoff, Frank H., David A. Melf i and David
B. Baker. 1978. Phosphorus Transport
in Rivers. Lake Erie Wastewater Man-
agement Study, U.S. Army Corps of
Engineers, Buffalo District Buffalo, New
York 88 pp.
David B. Baker is with Heidelberg College, Tiffin, OH 44883.
Thomas O. Barnwell, Jr., is the EPA Project Officer (see below).
The complete report, entitled "Fluvial Transport and Processing of Sediments and
Nutrients in Large Agricultural River Basins," (Order No. AD A111894; Cost:
$17.50, subject to change) will be available only from:
National Technical Information Service
5285 Port Royal Road
Springfield, VA 22161
Telephone: 703-487-4650
The EPA Project Officer can be contacted at:
Environmental Research Laboratory
U.S. Environmental Protection Agency
College Station Road
Athens, GA 30613
-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Official Business
Penalty for Private Use $300
US. OFFICIAL MAIL
FPMITY
Fill
IBE SJOO I
US, POSTAGE
s 0.2 ?
SM E T i R
503205
------- |