United States
Environmental Protection
Agency
Robert S. Kerr Environmental Research
Laboratory
Ada OK 74820
Research and Development
EPA-600/S2-83-011 Apr. 1983
Project Summary
Effects of Livestock Pasturing on
Nonpoint Surface Runoff
Richard K. White, Robert W. VanKeuren, Lloyd B. Owens, William M. Edwards,
and Robert H. Miller
This study was conducted to evaluate
the effects of livestock pasturing in
humid regions of the United States on
the quality of nonpoint surface runoff.
Specific objectives were to establish
the contribution of pollutants from
nonpoint source pasturing regimes, to
establish what happens to hydrologic
and water quality parameters related to
nonpoint surface runoff from pastures
and to identify pasture management
practices that affect nonpoint source
pollution.
Three pasturing regimes were
monitored and evaluated: summer
rotational grazing and winter feeding of
hay on one pasture; summer rotational
grazing with no livestock on pastures in
the winter, and winter rotational
grazing of autumn regrowth with feed-
ing of field-stored hay.
Winter pasturing and feeding of hay
caused a marked increase in the number
of runoff events and the volume of
runoff as compared to previous
meadow practices, and to summer
pasturing only. Dormant season (OS)
(Nov-Apr) nutrient fluxes were greater
than those in the growing season (GS)
(May-Oct). Analysis of the soluble and
sediment fraction of the runoff samples
indicated that the nutrients were
principally in the soluble phase.
Analysis of discrete sample data
collected throughout a runoff event
indicated that nutrient transport was
essentially matched with runoff, i.e. a
linear relation can be assumed between
runoff volume and nutrient delivery. In
general, soluble nutrient concentra-
tions are a function of hydrologic
variables, e.g. precipitation, rainfall
intensity, and antecedent soil moisture;
whereas, sediment related nutrient
concentrations were affected by
pasture management practices.
Microbiological analyses for total
coliform (TC), fecal coliform (FC), and
fecal streptococci (FS) were
conducted. The number of bacteria
exceeded water quality criteria for
nonpoint source runoff, even from a
non-pastured control watershed.
This Project Summary was developed
by EPA's Robert S. Kerr Environmental
Research Laboratory. Ada. OK, to
announce key findings of the research
project that is fully documented in a
separate report of the same title (See
Project Report ordering information at
back).
Introduction
Livestock production systems utilizing
pasturelandtare one segment of agricul-
ture production for which the extent of
nonpoint source pollution has not yet
been clearly defined, and for which best
management practices (BMPs) have not
yet been designed. Beef cow and call
production is increasing in East Central
United States from Pennsylvania, West
Virginia, and Virginia across Ohio and
Kentucky to Iowa and Missouri.
Currently, this area has about ten million
head of beef cattle, many of which are
raised on small farming operations. In
Ohio, for example, about two-thirds of the
1.4 million cattle raised are on farms with
20-99 head. Most cattle in this region are
pastured during the summer. In winter,
they may be fed in a barn or on an open
lot, fed hay out on pasture, or kept on
winter pasture of autumn-saved
regrowth and field-stored hay. The land
used for such cattle production is
primarily hilly and susceptible to erosion
if cultivated or if the vegetative cover is
-------
lost and the soil exposed. This land, in
permanent pasture, constitutes a large,
important part of the region.
This project was conducted coopera-
tively by the Ohio Agricultural Research
and Development Center, Wooster, Ohio,
and the United States Department of
Agriculture, Science and Education
Administrative-Agricultural Research
Service, Coshocton, Ohio. The pastures
studied were located on the North
Appalachian Experimental Watershed
(NAEW), Coshocton, Ohio. Long-term
hydrologic records from the experimental
watersheds at NAEW provided back-
ground data for this study. The three
pasture regimes monitored were as
follows:
1. Summer rotational grazing was
studied using 25 Charolais beef
cows with calves on 17.2 ha
divided into four pastures seeded
to orchard grass. Hay (square
bales) was brought to one of the
pastures and fed to the herd in
the winter. Within this winter
pasture (3.1 ha) was a gaged
watershed (WS 129) of 1.1 ha. A
moderate level of fertility was
provided with a spring application
of 56 kg nitrogen per hectare
(N/ha) and maintenance of
available phosphorus (P) and
potassium (K) levels at 28 and 168
kg/ha, respectively. Soil pH was
maintained at 6.0 to 6.5. Calves
were weaned and removed from
the study area in October.
2. Summer rotational grazing (no
winter feeding) was studied using
40 Charolais beef cows with calves
on 13.8 ha divided into four pas-
tures of orchard grass. Within one
of the pastures (3.8 ha) was a
gaged watershed (WS 110) of 0.86
ha. A high level of fertility was
provided with an annual applica-
tion of 224 kg N/ha (in three appli-
cations) and maintenance of
available P and K levels at 56 and
336 kg/ha, respectively. Soil pH
was maintained at 6.5 to 7.0.
3. Winter rotational grazing of fall
regrowth and feeding of field-
stored hay (round bales) was
studied using 40 Charolais cows
(those from the second regime) on
10.1 ha divided into four pastures.
The pastures were seeded to tall
fescue. Within one of these
pastures was a gaged watershed
(WS 106) of 0.63 ha. The same
high level of fertility as in the
second pasturing regime was
provided here.
A forested non-pastured watershed
(WS 172) of 17.6 ha was monitored as a
control.
Surface runoff from the watersheds
was measured by precalibrated H-flumes.
Precipitation was measured on each of
the watersheds by corresponding
recording rain gages. Topsoil moisture
was measured biweekly. Intervening
soil moistures were calculated using a
method based upon precipitation meas-
urements and depletion curves developed
from field data on soil moisture.
Proportional runoff samples were
collected by using an offset Coshocton
wheel sampler located at the outfall of the
flume. The collected runoff was delivered
to a temperature-controlled (4°C) storage
shed. The flow was collected in a series of
14 buckets (approximately 5L each) and a
final collecting drum for collecting runoff
from an exceptionally large event. Event
markers on the water stage recorders
noted when the sampling device
advanced to the next container.
Using vacuum filtration or
centrifugation, samples were separated
into a liquid fraction (for analysis of
soluble plant nutrients) and a sediment
fraction (for analysis of attached
nutrients). The sediment fraction was air-
dried and stored for subsequent analysis.
The liquid fraction was stored at4°C until
analyzed. The N, P, and K concentrations
were determined for the soluble fraction
and the sediment. Also, the runoff sample
was analyzed for TOC, COD, and
periodically, for BODs.
Conclusions
Evaluation of three pasture regimes
indicated that winter pasturing increased
the surface runoff. Where winter feeding
of cattle on one watershed (WS 129)
occurred, the runoff volume was
threefold that of a predicted volume
based on 23 years of previous hydrologic
record. Winter rotational grazing and
feeding of field-stored hay (WS 106)
increased the runoff by 75 percent more
than predicted. Summer rotational
grazing only (WS 110) increased the
observed runoff by 17 percent more than
predicted. The increased runoff volumes
were primarily associated with the
dormant season, winter feeding, and
pasturing.
Average DS, November through April,
nutrient fluxes were greater than those in
the GS, May through October, for all
three pasturing regimes. The flux of total
potassium (Tot K), 143 kg/ha, during the
study period was larger than total
nitrogen (Tot N) or phosphorus (Tot P), 73
and 14 kg/ha, respectively. The flux of
sediment nitrogen (Sed N) was larger
than sediment potassium (Sed K) or
sediment phosphorus (Sed P) by a factor
of three to six. The sediment-to-soluble-
f lux ratio was largest for P and least for K,
so that P transport is associated more
with sediment than are N or K. A
comparison of nutrient fluxes from the
control watershed with the pasture
watersheds showed that N and K were
two to ten times larger, and that P flux
was larger by three to ten orders of
magnitude for pasture watersheds.
Periodic samplers were utilized to
partition runoff samples during storms
and the discrete samples were analyzed
separately. Normalized flux scatter plots
indicated that N and K tended toward
early delivery (ED) as related to runoff
volume. Phosphorus tended toward late
delivery (LD). Analysis of the nutrient flux
plots, using 90 percent confidence
intervals, indicated that nutrient delivery
was essentially matched with runoff, i.e.
a linear relation can be assumed between
runoff volume and nutrient delivery.
Linear regression was used to identify
variables that contribute to nutrient flux
in runoff. In general, soluble nutrient
concentrations are a function of
hydrologic variables such as
precipitation, rainfall intensity, and
antecedent soil moisture; whereas,
sediment-related nutrient concentra-
tions were affected by pasture
management, specifically cattle
occupancy and time of occupancy.
The pasture management regimes are
related to water quality in surface runoff
in the following order with the first
practice causing the least flux of
nutrients: (1) summer rotational
pasturing only, (2) winter rotational
pasturing on autumn regrowth and field-
stored hay, and (3) summer rotational
pasturing and winter feeding with hay
brought to the cattle on one watershed.
Microbiological analyses for TC, FC,
and FS confirmed previous studies that
showed water quality criteria were
exceeded for coliform in runoff from
pastures. Ratio of FC to FS for pastured
watersheds and a non-pastured, forested
control varied to the extent that the
source of fecal material, either cattle or
wildlife, was not evident. Analyses
showed little difference in the number of
bacteria in runoff water collected early in
-------
a storm event from water collected near
the end of a storm event. Therefore,
bacteriological water quality criteria may
be inappropriate for characterizing
pasture runoff.
Recommendations
This report established the contribution
of nutrients and pollutants from nonpoint
source pasturing regimes in the cool,
humid region of East Central United
States. Three years of detailed hydrologic
data, water quality data, and livestock-
pasture management data were
collected. The nutrient transport was
principally in the soluble form versus the
sediment form and more transport
occurred during the DS than during the
GS. A marked increase due to livestock
pasturing, in the volume of runoff
occurred during the DS. The results
support the following recommendations:
1. Watershed modeling efforts
should include the impact of live-
stock pasturing on the hydrology,
nutrient flux, and pollutant
transport.
2. Composite, proportional samples
of runoff should be used to
measure nutrient flux from
meadow and pastured watersheds.
3. Alternative winter livestock man-
agement practices should be evalu-
ated to identify BMPs for wintei
pasturing.
4. Bacteriological water quality crite-
ria should not be used as a principal
parameter to identify sources of
nonpoint pollution.
Results and Discussion
The average precipitation for the GS
during this study was 25 percent greater
than the long-term mean. In particular,
the GS in 1978 and 1979 were much
wetter than normal. The average precipi-
tation for the DS was only 4 percent less
than the long-term mean.
Effects on Hydrology
Using a linear regression stepwise
analysis procedure, a predictive equation
for runoff volume based on precipitation,
a precipitation intensity factor, and soil
moisture was developed using prior
meadow year data. A predicted runoff (Qp)
was calculated using this equation and
data from the three-year pasture study.
Table 1 gives a comparison of Qp and the
Season*
WS 106
1977
1 977-78 DS
1978 GS
1978-79 DS
1979 GS
Study period
WS 110
1977 GS
1977-78 DS
1978 GS
1978-79 DS
1979 GS
Study period
WS 129
1977 GS
1977-78 DS
1978 GS
1978-79 DS
1979 GS
Study period
Op
(mm)
14.6
15.5
16.1
22.5
14.9
83.6
30.6
14.9
36.0
24.4
48.8
154.7
33.0
7.2
33.8
8.5
37.7
120.2
Oo
(mm)
6.3
13.9
10.5
63.1
53.8
147.6
14.8
28.3
20.9
24.4
96.0
184.4
57.0
79.1
99.3
49.4
114.4
399.2
No. of
Events
13
21
29
20
16
14
12
17
12
11
38
30
40
27
19
*GS is May-Oct. DS is Nov-Apr.
observed runoff (do) during the pasture
study. A comparison of Qp with Qo on WS
110 (summer rotational grazing only)
does not indicate a marked change in
hydrology, observed flow being only 1.2
times greater than predicted. WS 106
(winter rotational grazing on autumn
regrowth and feeding field stored hay)
had an observed flow 1.8 times that
predicted. WS 129 (summer rotational
grazing and winter feeding of hay) had an
observed flow three times larger than
predicted. It was during the DS that the
marked increase in flow occurred on WS
129 and to a lesser degree on WS 106.
Therefore, winter pasturing of livestock
has a marked effect on increasing the
runoff. A similar conclusion was obtained
by using an analysis of optimized curve
numbers.
Nutrient Concentrations and
Flux
The mean concentration of the
nutrients N, P, and K in the runoff from
the pastures and the control watersheds
is given in Table 2. A comparison of
concentrations for the GS indicates a
higher concentration on those pastures
with a higher fertilization rate, WS 106
and WS 110. During the DS, those
pastures with livestock had a higher
concentration of nutrients in the runoff.
Runoff samples were partitioned into
soluble and sediment fractions for the
analysis of N, P, and K. Figure 1 shows the
nutrient fluxes for the three pastures and
the control. Certain trends are evident
from this flux data:
1. Soluble nutrient flux was largest
for K and least for P.
2. Sediment nutrient flux was largest
for N and least for K as shown by
shading on the bars.
3. The total nutrient flux (both soluble
and sediment) was largest for
year-round cattle occupancy (WS-
129) and least for summer
rotational grazing (WS 110) except
for the very wet 1979 GS.
4. Where winter occupancy of cattle
was involved (WS 129 and WS
-------
Table 2. Summary of Mean (if) Concentration of Nutrients in Runoff from the Pasture and
Control Watersheds
Season and Watershed
Tot N, mg/l Tot P. mg/l
K. mg/l
Growing Seasons
WS 106
WS 110
WS 129
WS 172
•6.3
10.7
6.4
2.3
1.13
1.24
1.87
0.05
20.7
20.7
9.9
2.1
Dormant Seasons
WS 106
WS110
WS 129
WS 172
Water Year
WS 106
WS 110
WS 129
WS 172
13.2
5.4
21.5
1.5
9.8
8.2
12.9
1.9
0.69
0.49
2.02
0.02
0.91
0.88
1.93
0.03
15.2
7.6
45.2
1.6
17.9
14.5
25.1
1.8
106), the trend was toward greater
nutrient fluxes during the OS.
A sediment-nutrient-flux to soluble-
nutrient-flux ratio was calculated for
each runoff event. The sediment-to-
soluble-f lux ratio values f or WS 129 were
significantly larger than for WS 106 or
WS110.
Relation Between Nutrient Flux
and Runoff
Runoff was sampled periodically
throughout each storm event. Normalized
plots of percent cumulative nutrient flux
versus percent cumulative runoff volume
were constructed for runoff events that
had five or more discrete subsamples.
Figure 2 shows a scatter plot for percent
cumulative nitrogen flux in all runoff
events from WS 129. The nitrogen flux
slightly preceded the runoff in the"
majority of the samples. By use of 90
percent confidence intervals of the flux,
the time of delivery of the nutrients in the
runoff volume was evaluated. Table 3
summarizes the nutrient flux delivery
data. Event data were grouped in ten
different ways to determine if ED,
matched delivery (MD), or LD of nutrients
with runoff occurred. In most of the data
groupings, MO was observed. The
confidence intervals that excluded MD
did so by only a small amount (3 percent
or less of the total flux for an event).
Therefore, even when ED is found, it did
not differ greatly from MD and a linear
relation between nutrient flux and runoff
volume could be assumed for the
conditions in this study.
Variables Affecting Nutrient
Transport
Regression analysis was used to
identify statistically significant hydrologic
and pasture management variables.
Analyses were conducted for both the
soluble and sediment N, P, and K.
Independent variables regressed against
the nutrient concentration means were:
runoff volume, amount of precipitation,
precipitation intensity factor, soil
moisture, sediment flux, time after
fertilizer application, percent vegetation,
cattle on or off pasture, and time factor
with cattle on or off pasture. Soluble N, P,
and K showed no consistent pattern of
the independent variables correlated
with the dependent variable for WS 106,
110, or 129. For the sediment nutrients, a
time factor of cattle on and off the pasture
correlated with the nutrient means for all
three watersheds.
A second procedure used to identify
significant independent variables was
stepwise regression with significance set
at the 10 percent level. The soluble
nutrient concentration means were a
function of the hydrological variables of
precipitation amount, precipitation
intensity, and antecedent soil moisture.
The sediment nutrient concentrations
means were a function of cattle
occupancy variables.
Microbiology of Pasture Runoff
Runoff water samples were collected
periodically from the pasture watershed
and a control, forested watershed. When
possible, water samples were collected at
both the beginning and the end of the
runoff events. Each water sample was
analyzed for numbers of TC, FC, and FS.
The numbers of all three indicator
bacteria did not provide a clear relation-
ship to animal occupancy. The numbers
of indicator bacteria in runoff from WS
129 were consistently higher than those
of the other three watersheds.
An evaluation of the FC/FS for the
watershed indicated wildlife sources of
fecal pollution for WS 172 and WS 106.
WS 172 and WS 106 had mean FC/FS
ratios of 0.07 and 0.01, respectively. WS
129 and WS 110 had the greatest
number of days of cattle occupancy and
had mean FC/FS ratios of 0.18 and 0.22,
respectively. FC/FS ratios greater than
0.1 are reported characteristics of cattle
pastures.
These data which are in close
agreement with similar studies
elsewhere, support the conclusion that
recommended water quality criteria may
be inappropriate for characterizing
pasture runoff and other nonpoint sources
of potential water pollution.
4
-------
!
.1
•f
WS172
'no data
nfl,-,
* *
12
a
4
0
16
12
a
4
0
32
28
24
20
16
12
8
4
0
WS 110
i—ii~rr~ir—i
nnfln
n
WSW6
WS129
a = Growing Season 1977
b = Dormant Season 1977-78
c = Growing Season 1978
d = Dormant Season 1978-79
e - Growing Season 1979
f = Dormant Season 1979-80
Soluble phase
Sediment phase
a b c d e f
Nitrogen
abode f
Phosphorus
b c d e f
Potassium
Figure 1. Total nutrient flux for watersheds for May 1977 - March 1980.
-------
100
90
80
I
c
70
60
2 50
40
I
30
20
10
10 20 30 40 50 60 70
% Cumulative Runoff Volume
80
90
Table 3. Apparent Trends in the
Nutrient Flux Delivery as
Related to Time of Delivery*
Nutrient
Watershed
Delivery
TotN
P
K
106
110
129
106
110
129
106
110
129
Mainly MD
Mainly MD
Mainly MD
MD
MD
MD
MD
Mainly ED
Mainly MD
* Based on a 90 percent confidence interval for
MD. MD = matched delivery of nutrients
with runoff.
100
Figure 2. Total nitrogen flux scatter plot for all discrete samples in all selected runoff events
from WS 129 during study period.
Richard K. White, Robert W. VanKeuren, and Robert H. Miller are with the Ohio
Agricultural and Development Center. Wooster, OH 44691 and Ohio State
University, Columbus, OH 43210; Lloyd B. Owens and William M. Edwards are
with the USDA-SEA-ARS, North Appalachian Experimental Watershed,
Coshocton, OH 43812.
R. Douglas Kreis is the EPA Project Officer (see below).
The complete report, entitled "Effects of Livestock Pasturing on Nonpoint Surface
Runoff," (Order No. PB 83-165 456; Cost: $11.50, 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:
Robert S. Kerr Environmental Research Laboratory
U.S. Environmental Protection Agency
P.O. Box 1198
Ada, OK 74820
U. S. GOVERNMENT PRINTING OFFICE: 1982/659-095/1926
-------
United States
Environmental Protection
Agency
Center for Environmental Research
Information
Cincinnati OH 45268
Postage and
Fees Paid
Environmental
Protection
Agency
EPA 335
Official Business
Penalty for Private Use $300
CHICAGO TL
------- |