Content uploaded by David C. Coleman
Author content
All content in this area was uploaded by David C. Coleman on Jun 23, 2015
Content may be subject to copyright.
JOURNAL OF SOIL AND WATER CONSERVATION N
|
D 2005
288
reported that suspended sediment concentra-
tion discharges from construction activities
into streams was as high as 355 mg L-1 (355
ppm), with high concentrations persisting
100 m (328 ft) downstream negatively
impacting macroinvertebrate populations.
When eroded sediment is transported from its
site of origin to nearby surface waters it also
carries fertilizers, pesticides, fuels, and other
contaminants and substances commonly
spilled at construction sites that readily attach
to soil particles (Risse and Faucette, 2001).
For example, total annual loss of nitrogen,
phosphorus, and potassium due to soil ero-
sion in the United States is estimated to be
over 38 million Mg (42 million t). It is esti-
mated that the annual cost to society for on-
site loss of soil, nutrients, water and yield
reduction due to soil erosion is over $27 bil-
lion per year (Brady and Weil, 1996).
In terrestrial ecosystems, surface layers of
organic matter reduce the energy of raindrop
impact and allow water to percolate into the
soil, reducing surface runoff and erosion
(Jordan, 1998). The use of surface applied
organic amendments has been shown to
reduce runoff and erosion (Adams, 1966;
Meyer et al., 1972; Laflen et al., 1978;
Vleeschauwer and Boodt, 1978; Foster et al.,
1985). Because of better soil contact and
reduced susceptibility to movement from
wind or water, wood mulches are superior to
hay and straw mats (Holmberg, 1983; Lyle,
1987). Shredded bark and straw mulches will
intercept and dissipate the energy of raindrops
and prevent soil surface crusting. They also
break up overland flow of runoff and hold
more water at the soil surface allowing more
water to infiltrate the soil (Adams, 1966;
Gorman et al., 2000).
In the last ten years, compost has been used
successfully for slope stabilization, erosion and
sediment control, stormwater filtration, and
vegetative establishment applications. In
The U.S. Environmental Protection
Agency (USEPA) has declared that sedi-
ment contamination of our surface waters
is the greatest threat to our nation’s water
resources. Soil erosion is considered the
largest contributor to nonpoint source pollu-
tion in the United States, according to the
federally mandated National Pollution
Discharge Elimination System (NPDES)
(USEPA, 1997). Soil loss rates from con-
struction sites can be 10 to 20 times that of
agricultural lands (USEPA, 2000). For exam-
ple, forestlands lose an average of 0.36 m t
ha-1 (1 t ac-1) per year, agriculture loses an
average of 5.5 metric tons ha-1 (15 t ac-1) per
year, while construction sites average 73.3
metric t ha-1 (200 t ac-1) per year (GASWCC,
2002). In 2003, the federally mandated
NPDES Phase II program went into effect
extending the stormwater management plan
requirement to any land-disturbing activity
over 0.4 ha (1 ac). The new regulations label
development zones as “point sources” requir-
ing erosion control best management
practices (BMPs), stormwater pollution pre-
vention plans, increased monitoring, and
more site inspections by state and local
officials or certified inspectors.
Areas like construction sites that disturb,
excavate or grade soil are particularly prone
to soil erosion. In many cases, the existing
topsoil has been totally removed reducing soil
quality and fertility which reduces future
plant establishment. In addition, heavy
machinery and traffic compacts the soil
which further increase runoff and poor plant
growth (Risse and Faucette, 2001).
The most serious impacts of soil erosion
occur once the sediment leaves the site and
enters surface waters. Ehrhart et al. (2002)
L. Britt Faucette is research director for Filtrexx
International in Atlanta, Georgia. Carl F. Jordan is
senior ecologist at the Institute of Ecology, Universi-
ty of Georgia in Atlanta, Georgia. L. Mark Risse is
associate professor in the Department of Biological
and Agricultural Engineering at the University of
Georgia in Atlanta, Georgia. Miguel Cabrera is pro-
fessor in the Department of Crop and Soil Sciences
at the University of Georgia in Atlanta, Georgia.
David C. Coleman is a distinguished research pro-
fessor in the Institute of Ecology at the University of
Georgia in Atlanta, Georgia. Larry T. West is a pro-
fessor in the Department of Crop and Soil Science at
the University of Georgia in Atlanta, Georgia.
Evaluation of stormwater from compost and
conventional erosion control practices in
construction activities
L.B. Faucette, C.F. Jordan, L.M. Risse, M. Cabrera, D.C. Coleman, and L.T. West
ABSTRACT: Soil erosion is considered the biggest contributor to nonpoint source pollution in
the United States according to the U.S. Environmental Protection Agency and the federally
mandated National Pollution Discharge Elimination System. Soil loss rates from construction
sites can be 10 to 20 times that of agricultural lands. The use of surface applied organic
amendments has been shown to reduce runoff and erosion, however, with the exception of
animal manure, little research has focused on nutrient loss from these amendments. Four types
of compost blankets, hydroseed, silt fence, and a bare soil (control) were applied in field test
plots. Treatments were seeded with common bermuda grass. A rainfall simulator applied
rainfall at an average rate equivalent to a 50 yr hr-1 storm event (7.75 cm hr-1). Three simulated
rain events were conducted: immediately after treatment application, at three months when
vegetation was established, and at one year when the vegetation was mature. After three
months, the compost generated five times less runoff than hydroseed with silt fence, and after
one year, generated 24 percent less runoff. All treatments proved better than the control at
reducing solids loss. Total solid loads were as much as 3.5 times greater from hydroseed and
silt fence compared to the composts during the first storm, and as much as 16 times greater
during the second storm. Materials high in inorganic nitrogen (N) released greater amounts of
nitrogen in storm runoff; however, these materials showed reduced N loss over time.
Hydroseeding generated significantly higher total phosphorus (P) and dissolved reactive P loads
compared to compost in storm runoff during the first storm event.
Copyright © 2005 Soil and Water Conservation Society. All rights reserved.
www.swcs.org 60(6):288-297 Journal of Soil and Water Conservation
N
|
D 2005 VOLUME 60 N UMBER 6289
Portland, Oregon, yard waste composts used
for erosion control in residential construction
projects exhibited reduced erosion and
improved water quality over conventional
erosion and sediment control measures
(Portland Metro, 1994). Ettlin and Stewart
(1993) found that yard waste compost could
be used for slope stabilization and erosion
control on slopes up to 42 percent. Four
inch compost applications effectively con-
trolled erosion on 45 percent slopes for
three years (Michaud, 1995). A study con-
ducted by the Connecticut Department of
Transportation found composts and mulches
reduced soil erosion ten-fold compared to
bare soil surfaces on a 2:1 slope (Demars and
Long, 1998). Furthermore, Demars and
Long (1998) report that when compared to
silt fences, compost is 99 percent more effec-
tive in keeping sediment out of nearby
surface waters, and 38 percent more effective
than hydroseeding. Glanville et al. (2001)
reported runoff and interr ill erosion rates
were significantly lower on newly constructed
highway embankments when using compost
instead of imported topsoil.
While recent studies have shown the effec-
tiveness of compost and other best manage-
ment practices to reduce runoff and erosion
(Wilson et al., 2001), little attention has been
given to the fate of the fertilizer application
and nutrients commonly used in erosion
con-
trol for vegetation establishment. Nutrient
management plans and budgets are common
in the agricultural industry; however, there is
no equivalent for the construction and ero-
sion control industries where soil erosion can
be much greater (USEPA, 2000). Eghball
and Gilley (1999) reported that unincorpo-
rated compost applied to agricultural fields
with wheat and sorghum residue released
less dissolved phosphorus (P) and bioavail-
able P, but more particulate P than fertilizer
applications during initial rain events.
Additionally, they reported that ammonia
nitrogen (N) concentrations in runoff were
higher from fertilizer applications and nitrate
N concentrations where higher from com-
post applications. However,the inorganic N
contents reported for the composts were
greater than manure of the same feedstock,
which may imply that the compost may not
have been fully composted and should not
be used for erosion control applications
(Alexander, 2003).
Previously, we (Faucette et al., 2004)
evaluated a variety of compost and mulch
products for runoff and solids loss under a
one-time intense rainfall simulation at the
University of Georgia. Results showed high
quality composts performed better than those
of inferior quality; and mulches generally
produced less runoff and erosion than com-
posts. We also reported that compost blan-
kets can have relatively high nitrogen and
phosphorus concentrations and loads in
stormwater runoff, and that some of these
inferior compost treatments would be unsuit-
able for erosion control applications. In addi-
tion, no vegetation or subsequent storm sim-
ulations were evaluated. Vegetation and tem-
poral changes could have a tremendous affect
on short and long-term perfor mance of the
treatments and would better reflect real world
situations. These conclusions led to the treat-
ment selection and experimental design
described in the following section.
The overall objective of this study was to
evaluate the surface water quality impacts
from compost systems and hydroseed with silt
fence (conventional industry best manage-
ment practices) used for erosion and sediment
control applications in construction activities.
Specific objectives were to evaluate each
treatment as a best management practice for
managing stormwater and controlling erosion
by: 1) managing stormwater, 2) controlling
erosion and soil loss, and 3) minimizing risk
related to nutrient application. Effective ero-
sion and sediment control practices should
reduce sedimentation while minimizing
additional risks, such as nutrient loading, to
surface water pollution.
Methods and Materials
Site description. Research test plots were
constructed at Spring Valley Farm in Athens
and Clarke County, Georgia at 33˚ 57' N
latitude and 83˚ 19' W longitude. The soil
was originally classified as an eroded Pacolet
Sandy Clay Loam (USDA, 1968) and has a
high soil erodibility factor (K value) of
approximately 0.36 (Wischmeier and Smith,
1978). The area receives an average annual
rainfall of 1215 mm (48 in), with January
through March as the wettest per iod. The
average annual high temperature for the area
is 22˚ C (72˚ F), the average low is 11˚ C (52˚
F), with a mean annual temperature
of 17˚ C (63˚ F) (Weather Channel, 2004).
The week prior to the first simulated storm
event, the research site received no natural
rainfall. During the three months between
the first and second storm event the site only
received 90.7 mm (3.57 in) of natural rainfall,
with only 16.8 mm (0.66 in) falling in the
third month. The week prior to the second
storm the research site received no natural
rainfall. These extremely dry conditions
likely affected vegetation growth. The week
before the final storm event the research site
received 102.4 mm (4.03 in) of natural rain.
This led to saturated field conditions during
the final simulated storm event.
The testing area was cleared of vegetation
and graded with a grading blade mounted
skid steer, exposing a semi-compacted
(from the skid steer) subsoil (Bt horizon) to
simulate construction site conditions. Test
plot borders were installed to prevent cross
contamination of plots. Fifteen cm (6 in)
wide stainless steel borders were trenched 7.5
cm (3 in) into the soil. The plots were sized
to fit the effective rainfall distribution from
the rainfall simulator, 1.0 m (3.3 ft) wide by
4.8 m (16 ft) long, for a total plot area of
4.8 m2(53 ft2). A removable flume was
installed at the base of each plot prior to each
simulated rainfall event. A removable stain-
less steel border was carefully inserted at the
base of each plot, once the flume was
removed after each storm event, to maintain
the structure and integrity of the soil in the
plot. The soil was carefully compacted
around the removable flume and the remov-
able border after each one was installed for
use. Nine rain gauges were installed in each
plot to measure rainfall quantity. Three each
were placed 1.2 m (4 ft), 2.4 m (8 ft) and
3.6 m (12 ft) from the top of the plot.
Gauges were also spaced evenly across the
width of the plot.
Treatments. Seven treatments, each in
triplicate, were assigned randomly and applied
to twenty-one 1 m by 4.8 m (3.3 ft by 15.7
ft) plots on a cleared field graded to a uniform
10 percent slope on a sandy clay loam surface.
The seven treatments were: 1) a biosolids
compost blanket with filter berm, 2) a yard-
waste compost blanket with filter berm, 3) a
municipal solid waste compost and mulch
blanket with filter berm (2:1 compost to
mulch by volume), 4) a poultry litter com-
post, mulch, and gypsum blanket with filter
berm (2:1 compost to mulch by volume with
five percent gypsum addition by volume),
5) hydroseed with silt fence, 6) hydroseed
with mulch filter berm; and 7) a bare soil
(control) plot. To reduce potential nutrient
losses in runoff, the poultry litter compost
and municipal solid waste compost treatments
Copyright © 2005 Soil and Water Conservation Society. All rights reserved.
www.swcs.org 60(6):288-297 Journal of Soil and Water Conservation
JOURNAL OF SOIL AND WATER CONSERVATION N
|
D 2005
290
warm water bath for one hour after a 16 hour
incubation period using respirometry
methods described by Iannotti et. al (1993).
Chemical characterizations were per-
formed at the University of Georgia Agri-
cultural and Soil, Plant and Water Laboratory
using EPA or Association of Analytical
Communities (AOAC) approved procedures
(University of Georgia Soil, Plant and Water
Analysis Lab, 2004). Total carbon (C) and
total N were analyzed on a Carlo Erba
Analyzer and determined by thermal con-
ductivity from combustion to carbon dioxide
(CO2) and nitrous (N2); organic matter was
determined by weight difference after loss on
ignition at 550˚ C (1022˚ F); nitrate-N and
ammonium-N samples were first extracted
using a 20 ml (0.68 oz) solution of deionized
water and KCl and then filtered with
Whatman 42 filter paper before analysis by
continuous flow colorimetric assay (see water
analysis below for more detail on colorimetric
analysis on Alpkem RFA3000); and soluble
salts were determined by electrical conduc-
tivity. Heavy metals were analyzed and all
of the treatments were below the pollutant
concentration levels as specified in EPA part
503 Table 4 (USEPA, 1993). Treatment char-
acteristics are reported in Table 1.
Sampling and analyses for stormwater
included: rainfall amount, antecedent soil
water, time until start of runoff, time until
steady state of runoff flow rate, total runoff
volume, peak runoff rate, rainfall infiltration,
rainfall infiltration to runoff ratio, total solids
loads, and total solids loss ratios. Rainfall
averages were calculated by averaging the
rainfall depths for nine rain gages placed
equidistant to each other around the perime-
ter and within the test plot area. Antecedent
soil moisture conditions were measured pr ior
to the first and third rainfall simulation using
time domain reflectometry with a Tektronix
Cable Tester (Ferre and Topp, 2002). Each
plot used three time domain reflectometry
probes placed below the soil surface at
were blended with wood mulch on a volu-
metric ratio basis (1:1, compost:mulch). This
ratio was chosen because it was felt this was
the most mulch that could be blended with
these materials without adversely affecting
plant growth from N immobilization. In
addition, the poultry litter compost was
blended with approximately five percent
(volumetric basis) ground gypsum (CaSO4).
This was done to reduce potential P losses in
the runoff from the reaction between P and
CaSO4to form the more stable Ca3(PO4)2.
Compost blankets were manually applied
at 3.75 cm (1.5 in) depths over the entire area
of the plot. Filter ber ms were 60 cm (2 ft)
wide by 30 cm (1 ft) high and situated at the
base of the slope across the width of the plot.
Application depth of the compost blankets
and dimensions of the compost filter berms
followed American Association of State
Highway Transportation Official’s specifica-
tions for erosion and sediment control
(Alexander, 2003). Each treatment, exclud-
ing the control plots, were seeded with a 1:1
mix of hulled and unhulled Common
Bermuda (Cynodon dactylon) grass seed
applied at 3.7 kg ha-1 (20 lbs ac-1), specified by
the Georgia Department of Transportation as
an erosion and sediment control vegetative
measure for slopes 3:1 or less for the Athens,
Georgia region. The compost treatments
were physically, biologically, and chemically
characterized prior to application in the test
plots (Tables 1). Treatments were selected
based on availability and results from previous
research conducted at The University of
Georgia (Faucette et al., 2004).
Rainfall simulator. A Norton Rainfall
Simulator with four variable speed V-jet
oscillating nozzles was used to simulate rain
events. During rain events, nozzle water
pressure was maintained at 0.42 kg/cm2
(6 psi), according to manufacturer’s specifica-
tions, producing an intensity of 7.75 cm
(3.1 in) h-1 for one hour. This is equivalent
to the one-hour storm event for a 50-year
return for the Athens, Georgia region, based
on historical rainfall records (USDOC, 1961).
It was our intention to evaluate these treat-
ments under a “worst-case” scenario, because
most runoff and erosion occurs during these
large events. Municipal tap water was used in
this study containing NO3-N of 0.673 mg
L-1 (0.673 ppm) and P04-P of 0.093 mg L-1
(0.093 ppm).
Three simulated rainstorms were conducted
at the beginning of the experiment, at three
months, and one year. These time intervals
were chosen based on the predicted establish-
ment of the vegetation. The first storm event
was intended to provide information on the
performance of the treatments prior to vege-
tation establishment. The second storm
event was intended to provide information
on how the performance of the treatments
changed when vegetation was newly estab-
lished. The final storm event was to provide
information on how the treatments reacted
once vegetation was fully established. All of
the plots were subjected to natural rainfall
between the simulated rainfall events.
Sampling and laboratory analysis. Physical
and biological analyses of the treatments were
performed at the University of Georgia’s
Bioconversion Research and Education
Center laboratory and followed the proce-
dures outlined in the U.S. Composting
Council’s Test Methods for the Examination
of Composting and Compost (TMECC)
(USCC, 1997). Water content (method
07.09-A) was determined by the difference
between wet and dry weight. Human made
inert analysis (method 07.08) was determined
as the non-compost fraction (e.g. glass, plas-
tic, metal) of the total dry weight of the treat-
ment. Germination rate (method 09.05-A)
was determined by percent cress seed germi-
nation in a water extract of the treatment
(USCC, 1997). Bulk density was determined
as dry weight per known volume of sample
(USDA, 1998), and biological stability was
determined as the oxygen uptake rate in a
Table 1. Physical, chemical, and biological characterization of treatments.
Bulk Stability -
density 02uptake Germination Water SS OM
Treatment (g/cm3) (mg O2/g VM hr -1) index (%) (%) pH (mS/cm) (g kg -1) C:N C N NH4NO3PpH
Biosolids comp 0.51 0.02 96 31.3 7 1.62 202 17 100900 5830 2480 1960 4470 7.0
Yard waste comp 0.5 0.09 100 40.66 7.8 0.645 193 19 97500 5010 40 70 3240 7.8
Poultry litter comp 0.59 0.06 100 32.2 7.2 5.93 212 22 131500 5980 70 240 4290 7.2
MSW*comp 0.32 0.1 100 45.7 8.1 4.96 360 20 175200 8660 140 180 1910 8.1
Soil 2.23 Nd Nd Nd 4.7 Nd Nd 18 250 14 0.74 0.053 348 4.7
All nutrients and metals expressed in mg kg-1.
*MSW = municipal solid waste.
Copyright © 2005 Soil and Water Conservation Society. All rights reserved.
www.swcs.org 60(6):288-297 Journal of Soil and Water Conservation
N
|
D 2005 VOLUME 60 N UMBER 6291
distance intervals of 0.9 m (3 ft) from the top
of the plot. The U.S. Department of Agri-
culture (USDA) Agricultural Research
Service (1980) time domain reflectometry
data acquisition (TACQ) computer program
was used to process and convert wavelengths
to water content.
Runoff sampling procedures and calcula-
tion methods followed procedures used for
the Water Erosion Prediction Project (WEPP)
developed by the USDA National Soil
Erosion Lab which have been used in similar
studies (Glanville et al., 2001; Faucette et al.,
2004). Runoff samples were collected from a
flume placed at the base of each plot. The
first sample was taken once water began to
“trickle” from the flume aperture, the point
determined to be the start of runoff. After the
first sample was collected, samples were taken
every five minutes until the 60-minute storm
was finished. Runoff quantity and total solids
samples used one 500 ml (2.1 c) Nalgene bot-
tle per five minute interval sample, and “sec-
onds-to-fill” bottle times were recorded to
obtain runoff flow rates. The total volume of
each runoff sample and the time over which it
was collected was recorded.
For nutrient samples the first sample was
taken once water began to “trickle” from the
flume aperture and every five minutes until
the 60-minute storm was finished. Separate
nutrient samples were taken using 500 ml
(2.1 c) Nalgene bottles and were filled for five
second durations. For each test plot, each
five-second nutrient sample,taken every five-
minutes, was composited in a 3.8 l (1 gal)
bottle and kept in a cooler and refrigerated
prior to analysis. This created a single
volume weighted sample for each test plot
and each storm event. Nutrient samples
were analyzed for total N, NH4-N, nitrate N
(NO3-N), total P, and dissolved reactive P.
Each 500 ml (2.1 c) runoff/solids sample
was weighed and oven dried at 105˚ C (221˚
F) until constant weight was achieved to
determine the total solids content. The total
solids were measured using methods 2540 B
Total Solids Dried at 103 to 105˚ C (217 to
221˚ F) (USEPA, 1983). Using these data, the
peak runoff rate (once flow reached steady
state conditions) was calculated by averaging
runoff rates observed during the last three five-
minute interval samples, during the simulated
storm, or when runoff rates for three time
adjacent samples were the same. The runoff
rate (known volume per measured time) sam-
pled at five-minute intervals during the simu-
lation were plotted, and the total runoff vol-
ume was calculated by summing the area
under the runoff curve.
Total solid loads were calculated by sum-
ming the average total solids concentration
of two time adjacent concentration samples
multiplied by the average of the same two time
adjacent samples for runoff volume.
Infiltration volumes were calculated by total
runoff volume subtracted from total rainfall
volume, where total rainfall volume was rain-
fall total multiplied by the total area of the plot.
Laboratory analysis of the nutrients in
runoff water was conducted at the University
of Georgia’s Institute of Ecology Analytical
Chemistry Laboratory (2004) and all standard
methods for preparation, analysis, and calcula-
tion can be found on their website. All N
and P forms determined from water samples
were first filtered with a 0.45 micron filter
and were processed on an Alpkem RFA300
continuous flow colorimetric analyzer. After
1000 mg L-1 (1,000 ppm) of colorimetric
reagent was added to each sample, the chem-
ical nutrient concentration in solution was
measured as a function of the amount of light
absorbance at a particular wavelength. Prior
to filtration and colorimetric analysis of total
N and total P, a solution using persulfate,
boric acid, and sodium hydroxide was added
to unfiltered runoff samples at 1:5 for oxi-
dization/digestion pretreatment (Qualls,
1989). Nitrate-N and total N were measured
using EPA standard method 353.2 (colori-
metric, automated, cadmium reduction),
ammonia N using EPA standard method
350.1 (colorimetric, automated phenate), and
total P and dissolved reactive P using EPA
standard method 365.1 (colorimetric, auto-
mated, ascorbic acid) (USEPA, 1983).
Statistical analysis. SAS version 8.2 (SAS,
2001) was used for statistical analysis.
Analysis of variance (PROC ANOVA) used
Duncan’s Multiple Range test for significant
differences between cells to determine any
significant differences between treatments
(p ≤0.05). Correlation analysis (PROC
CORR) was used to determine which of
the independent variables, including: physi-
cal, chemical, and biological treatment char-
acteristics and all vegetation and rainfall
characteristics, were correlated to the
response variables.
Results and Discussion
Runoff and infiltration. There was signifi-
cantly less runoff volume for the compost
blankets than the control for the simulated
rain events at three months and one year, and
less runoff from the compost than the
hydroseed with silt fence at three months
(Table 2). Additionally, compost reduced
runoff more over one year than hydroseed-
ing, by 33 percent and eight percent respec-
tively, and reduced total cumulative runoff
relative to the control by 55 percent and
30 percent, respectively. These findings are
similar to land application of manure in agri-
cultural fields at 62 percent reduction (Gilley
and Risse, 2000).
All but one compost treatment had signif-
icantly greater infiltration than the control for
all storm events, while two were significantly
greater than the hydroseed with silt fence.
During the first storm event, relative to the
control, the municipal solid waste compost
and the yard waste compost treatments
allowed 51 percent more water to infiltrate
the surface, the poultry litter allowed 43 per-
cent more, the biosolids 31 percent, the
hydroseed with silt fence 24 percent, and the
hydroseed with mulch berm 20 percent
more. Similarly, Agassi et al. (1998) found
that under rainfall simulation, compost perco-
Table 2. Total runoff volume (mm) by treatment at day one, three months, and twelve
months, n = 3.
Day one Three months Twelve months
Standard Standard Standard
Treatment Average deviation Average deviation Average deviation
PLC/mulch/gypsum 32.0ab 12.7 5.0c 4.9 15.9c 7.0
Biosolids compost 38.1ab 7.9 9.6c 6.9 21.6bc 17.0
MSW*compost/mulch 22.5b 13.1 1.8c Nd 21.9bc 2.2
Yardwaste compost 33.0ab 5.6 8.1c 4.1 25.0abc 7.0
Hydroseed/mulch berm 36.7ab 5.8 20.2bc 2.4 34.2ab 9.9
Hydroseed/silt fence 30.0ab 11.6 32.3ab 28.3 27.6abc 5.1
Bare soil (not seeded) 42.3a 5.6 45.9a 20.6 40.8a 8.9
Treatments with same letter are not significantly different at α= 0.05 using Duncan’s
Multiple Range test.
*MSW = municipal solid waste.
Copyright © 2005 Soil and Water Conservation Society. All rights reserved.
www.swcs.org 60(6):288-297 Journal of Soil and Water Conservation
JOURNAL OF SOIL AND WATER CONSERVATION N
|
D 2005
292
compost blankets gradually increasing soil
structure and water infiltration at the soil
interface, while the control and hydroseeded
treatments may have experienced soil crust-
ing. Of the compost treatments, the poultry
litter and biosolids compost treatments
reduced runoff rates the most over the one-
year study period, 43 percent and 33 percent
respectively. This was likely due to greater
vegetation cover and biomass exhibited by
these two treatments, which are important
factors in reducing runoff and erosion in the
revised universal soil loss equation (RUSLE)
(Brady and Weil, 1996).
Total solids loss. All treatments signifi-
cantly reduced total solids loss relative to the
control for all storm events (Table 4). The
yard waste compost released nearly 3.5 times
less total solids load, and 16 times less total
solid load than the hydroseed with silt fence
during the first and second stor m events,
respectively. In comparison, Glanville et al.
(2001) reported interrill erosion rates from
bare soil were six times greater than yard
waste compost, while topsoil was nearly
12 times greater. Consistent with this trend,
in Faucette et al. (2004) we found the bare
soil (control) generated six times more total
solids than the yard waste compost blanket.
This difference between the compost system
relative to the hydroseed with silt fence, is
probably a result of the immediate and more
stable soil surface cover provided by the com-
post blankets relative to the hydroseed treat-
ments. In review of 200 studies Doolette and
Smyle (1990) reported, mulching reduces soil
erosion between 78 and 98 percent; each of
the four composts in this study reduced total
solids loss between 97 and 99 percent.
The hydroseed with mulch berm treat-
lated nearly twice as much rain water as a bare
soil. By the final storm event, relative to the
control, the composts allowed 61 to 65 per-
cent more water to infiltrate the surface,
while the hydroseed treatments allowed 43
to 47 percent. The increased infiltration
percentages (compared to the control) that
resulted during the final storm event were
probably due to the increase in vegetation
exhibited by all treatments.
Time to start of runoff and peak runoff
rate. The compost systems appeared to
absorb more rainfall initially, creating a signif-
icantly longer time period for runoff to com-
mence, relative to the control and hydroseed,
immediately after treatment application
(Table 3). Average cumulative time before
runoff commenced was 24 minutes for the
compost system and nine minutes for the
hydroseed treatments. The delay in runoff
(and reduced runoff volumes) is likely from
the high water storage capacity characteristic
to humus rich materials (Brady and Weil,
1996), as well as the porous nature of the
compost blanket created by the heterogeneity
of the particle sizes in the material. This may
provide evidence that compost can be an
effective tool to prevent runoff from occur-
ring during small and medium storm events.
Cumulative peak runoff rates were 60 per-
cent lower for the compost treatments relative
to the control and 34 percent lower relative to
hydroseed. This is consistent with Glanville
et al. (2001), which found compost blankets
reduced runoff rates by 62 percent compared
to a bare soil and by 58 percent compared to
topsoil. This may be further indication that
compost blankets allow greater infiltration
and reduce soil crusting. Additionally, the
heterogeneous particle size distribution and
porosity of compost blankets may slow down
the surface flow of runoff by physical inter-
ruption, and slowing down the runoff may
allow more water to infiltrate, which was reported
in the previous section. Over the one-year
study, all four compost treatments showed a
reduction in peak runoff rate, while the
hydroseed with silt fence runoff rate
remained unchanged and the bare soil runoff
rate increased. This may be the result of the
Table 3. Average time (minutes) until start of runoff and steady state conditions by treat-
ment at day one, three moths, and twelve months, n = 3.
Day one Three months Twelve months
RO steady RO steady RO steady
Treatment RO start state RO start state RO start state
PLC/mulch /gypsum 12.0bc 40.3a 41.0ab 55.3a 21a 44.3ab
Biosolids compost 8.3bcd 26.7ab 32.7b 56.0a 23.7a 40.3a
MSW*compost/mulch 20.0a 40.0a 51.7a >60.0a 14.3a 37.7ab
Yardwaste compost 13.0b 31.3ab 33.3b 54.0a 14.7a 34.7ab
Hydroseed/mulch berm 7.3cde 25.7ab 14.3b 31.0b 9.0a 27.3ab
Hydroseed/silt fence 6.0de 22.7ab 8.0b 19.7b 10.3a 33.7ab
Bare soil (not seeded) 2.7e 9.3c 6.3b 19.7b 3.7a 18.7b
Treatments with same letter are not significantly different at α= 0.05 using Duncan’s
Multiple Range test.
*MSW = municipal solid waste.
Table 4. Average total solids loads (g/m2) and total solids loss ratio (treatment to control) by treatment at day one, three months, and
twelve months, n = 3.
Day one Three months Twelve months
Standard Standard Standard
Treatment Average deviation Ratio Average deviation Ratio Average deviation Ratio
PLC/mulch /gypsum 158.9b 91.3 0.025 14.6b 8.3 0.003 10.8b 4.5 0.010
Biosolids compost 105.8b 13.0 0.016 18.9b 13.2 0.003 8.8b 6.4 0.008
MSW*compost/mulch 191.9b 107.8 0.030 6.0b Nd 0.001 17.8b 6.8 0.016
Yardwaste compost 88.5b 45.3 0.014 13.7b 6.6 0.002 17.1b 6.2 0.015
Hydroseed/mulch berm 265.1b 32.3 0.041 78.1b 21.7 0.014 10.9b 6.1 0.010
Hydroseed/silt fence 307.9b 127.8 0.048 219.6b 72.0 0.039 14.5b 6.7 0.013
Bare soil (not seeded) 6428.1a 2182.7 5464.2a 3290.4 1109.7a 987.7
Treatments with same letter are not significantly different at α= 0.05 using Duncan’s Multiple Range test.
*MSW = municipal solid waste.
Copyright © 2005 Soil and Water Conservation Society. All rights reserved.
www.swcs.org 60(6):288-297 Journal of Soil and Water Conservation
N
|
D 2005 VOLUME 60 N UMBER 6293
ment produced 14 percent and 64 percent
less total solid loading than the hydroseed
with silt fence treatment during the first and
second storms, respectively. This may be the
result of the mulch berm to act as a three
dimensional sediment filtration device rela-
tive to the silt fence, which is a two dimen-
sional filter that appeared to clog easily when
exposed to sedimentation. Additionally, the
heterogeneous particle size distribution of
the filter berm matrix may provide greater
surface area and more pore space diversity
(micro and macro) for trapping sediment.
The decrease in solids loss in the control over
time can likely be attributed to the volunteer
weed growth within the test plots. Visually,
the control and hydroseeded plots had
evidence of rilling, indicating erosion from
flow stress. The composts had no evidence of
rilling but did show some movement of
material, indicating stress from sheet flow.
Correlation analysis for runoff and erosion.
Based on correlation analysis at three months,
particle sizes over 6.3 mm (0.2 in) in the
compost seem to have the greatest affect on
runoff (Table 5). This is likely due to greater
surface porosity, created by a more diverse
particle size distribution, which the larger
particles served to widen. Buchanan et al.
(2000) reported similar findings for soil ero-
sion, where a diverse particle size distribution
of wood chips reduced erosion more (by 86
percent) than either small wood chips
(22 percent), or large wood chips (78 percent)
alone. Additionally, the large particle sizes are
more likely to slow down surface runoff;
therefore reducing the runoff rate and
increasing the potential for water to infiltrate.
General characteristics of high quality com-
post such as high germination rate, a neutral
pH, and sufficient N, P, potassium (K) were
good indicators that there would be greater
infiltration and less runoff. This is likely since
these are prerequisites for good vegetation
establishment and cover, which can lead to
less runoff (Brady and Weil, 1996).
Finally, it appears that the bulk density
and organic matter content of compost is
correlated to solids loss from the plots. This
characteristic is consistent with soil erosion
studies where organic matter influences bulk
density, soil structure, and infiltration, normal-
ly
resulting in reduced erosion (Brady and
Weil, 1996). This provides compelling evi-
dence that compost may be well suited for a
variety of stormwater management applica-
tions, particularly where it can eliminate
runoff, thus preventing most water erosion
from ever occur ring.
Total N loss. Nutrient loading was chosen
to evaluate the treatments rather than con-
centrations. If a treatment exhibits compara-
tively high concentrations of nutrient loss,but
is comparatively more effective at reducing
runoff, then it may be a more desirable
method of erosion control, since it is the
amount of nutrients entering nearby surface
water that are of most concern. The total
amount of N applied by each treatment
was 132 g/m2(5.6 oz/yd2) from the poultry
litter compost, 111 g/m2(4.7 oz/yd2) from
the biosolids compost, 104 g/m2(4.4 oz/yd2)
from the municipal solid waste compost, 94
g/m2(4.0 oz/yd2) from the yard waste com-
post, and 10 g/m2(0.4 oz/yd2) from the
hydroseeded applications.
The first simulated storm event was con-
ducted immediately after treatment applica-
tion; therefore the potential for N loss was
greatest at this time period due to the absence
of vegetation. During the first rainfall simu-
lation total N loads from the biosolids compost
(4060 mg/m
2
, 0.17 oz/yd2
), municipal solid
waste compost (2014 mg/m
2
, 0.08 oz/yd2
),
and both hydroseeded
treatments (1392 and
1008 mg/m2, 0.06 and 0.05 oz/yd2
)
were sig-
nificantly higher than the control (Table 6).
In addition, the biosolids compost was signif-
icantly greater than all other treatments. This
was likely because 76 percent of the original
total N content of the biosolids compost was
inorganic N (ammonium-N and nitrate-N),
which is more mobile and easily lost in
stormwater runoff relative to organic N.
Although this percentage of inorganic N is
not characteristic to quality compost, it is
consistent with previous findings for biosolids
compost (Faucette et al., 2004).
Comparatively, the yard waste compost had
two percent of its total N as inorganic N, the
municipal solid waste compost had four
percent,and the poultry litter had five percent
(Table 1). Generally, in mature (e.g. suffi-
ciently composted) compost the majority of
N is in organic form. Furthermore, total N
loading was highly correlated to the ammo-
nium N and nitrate N content of compost
(Table 7). The higher N loading from the
municipal solid waste compost was likely
because this compost had the greatest total N
content of the four composts used in the
study. Although the hydroseed treatments
had the least amount of N applied, the relatively
high loading was likely because the form of
N applied (from fertilizer) was inorganic, and
therefore more mobile in runoff. Total N
lost in the runoff, combined from all three
storms, as a percent of the total N applied by
the treatments was 15.3 percent from the
hydroseed with mulch,12.2 percent from the
hydroseed with silt fence, 3.9 percent for the
biosolids compost, 2 percent for the munici-
pal solid waste compost, and 0.7 percent for
both the yard waste compost and poultry
litter compost treatments. This gives further
evidence that although the compost blankets
Table 5. Results from correlation analysis for runoff during second storm event and solids loss for all storm events. This table lists all
variables with significant correlation (r > 0.70, α= 0.05, n = 21).
Response variable Independent variable (treatment) with correlation coefficient
Time to runoff start Particle size >25mm (0.71), Particle size >16mm (0.82), Particle size >9.5mm (0.77),
Particle size >6.3mm (0.72), pH (0.73), germination rate (0.80)
Time to runoff steady state Particle size >25mm (0.77), Particle size >16mm (0.89), Particle size >9.5mm (0.82),
Particle size >6.3mm (0.74), pH (0.74), germination rate (0.84)
Rain infiltration volume N (0.74), pH (0.81)
Runoff rate N (0.86), P (0.81), K (0.76)
Total sediment concentration, Storm #1 Bulk density (0.93)
Total sediment load, Storm #1 Bulk density (0.90), organic matter (0.78)
Total sediment concentration, Storm #2 Bulk density (0.87), organic matter (0.71)
Total sediment load, Storm #2 Bulk density (0.73), organic matter (0.77)
Total sediment concentration, Storm #3 Bulk density (0.76)
Total sediment load, Storm #3 Bulk density (0.75)
Copyright © 2005 Soil and Water Conservation Society. All rights reserved.
www.swcs.org 60(6):288-297 Journal of Soil and Water Conservation
JOURNAL OF SOIL AND WATER CONSERVATION N
|
D 2005
294
hydroseed with mulch berm, 1236.6 mg L-1
(0.05 oz/yd2
)
from the hydroseed with silt
fence, 906.3 mg L-1 (0.04 oz/yd2
)
from the
poultry litter compost, 523.2 mg L-1 (0.02
oz/yd2
)
from the yard waste compost, and
271.6 mg L-1 (0.01 oz/yd2
)
from the bare soil.
It should be noted that it is common for
hydroseed to be applied multiple times before
vegetation is completely established; this
could have a major effect on added N load-
ing that was not simulated in this study.
Nitrate N loss. Nitrate nitrogen is a highly
mobile form of N and is easily transferred
to groundwater. High concentrations of
NO3-N in drinking water have been linked
to negative health effects in humans as well as
eutrophication in surface waters. Elevated
nitrate concentrations have been reported
from agricultural fields that apply organic
amendments (Eghball and Gilley, 1999). The
biosolids compost had the greatest nitrate
load during the first storm at 2,568 mg/m2
(0.1 oz/yd2
)
, followed by the hydroseed with
mulch berm at 797 mg/m2 (0.03 oz/yd2
)
, the
hydroseed with silt fence at 644 mg/m2 (0.03
oz/yd2
)
, the poultry litter compost at 527
mg/m2 (0.02 oz/yd2
)
, the yard waste compost
at 88 mg/m2 (0.004 oz/yd2
)
, the control at 53
mg/m2(0.002 oz/yd2
)
and the municipal
solid waste compost at 3 mg/m2 (0.0001
oz/yd2
)
. The biosolids compost and both
hydroseed treatments were significantly
greater than the control. The amount of
nitrate loss by each treatment was reflective of
the amount of nitrate in the treatment at the
time of application, and was positively corre-
lated. This is likely because nitrate N is a
highly mobile form of N (Brady and Weil,
1996). Composts with high nitrate contents
and hydroseed using fertilizer with nitrate
generally apply more total N,hydroseed (with
fertilizer) generally loses more of the N
applied during runoff events, likely because
inorganic N is more mobile than the organic
N in compost.
By the second storm event, total N loads
from experimental treatments were not
significantly different from the control or
between compost and hydroseed treatments.
The biosolids compost was still significantly
different from two of the composts, likely
because the inorganic N content (initially
higher than the other composts) had not been
fully leached from the biosolids compost or
taken up by plants. All treatments exhibited
major load reductions between the two storm
events, as ninety percent of the total nitrogen
loading occurred during the first storm event after
treatment application, signifying the risk of N
loading is greatly diminished for most treat-
ments after the first storm event.
By the final storm event, all total N loads
were significantly less than the control. This
was probably due to assimilation by vegeta-
tion (the control was not seeded) and/or
movement into the soil profile, in addition to
the losses from previous storm events. The
increase in N loading during the final storm
event by the control, is likely from higher
runoff volumes caused by saturated condi-
tions prior to the rain simulations.
Total N mass loads for the entire study
period were 4357 mg L-1 (0.18 oz/yd2
)
from
the biosolids compost, 2083.3 mg L-1 (0.08
oz/yd2
)
from the municipal solid waste com-
post, 1524 mg L-1 (0.06 oz/yd2
)
from the
Table 6. Average load for total nitrogen (N), nitrate N (NO3N), total phosphorus (P), and dissolved reactive P (mg/m2) in runoff by treat-
ment at day one, three months and twelve months, n = 3.
Day one Three months Twelve months
Treatment TN NO3N TP DRP TN NO3N TP DRP TN NO3N TP DRP
Poultry 841.9cde 526.8bc 86.7c 75.3c 24.5b 2.9a 16.2a 13.4a 39.9b 4.7c 16.5ab 13.7b
Biosolids 4060.9a 2568.3a 156.7bc 141.2bc 254.3a 126.1a 53.9a 51.4a 41.8b 9.7bc 46.2a 37.8a
MSW*2014.1b 3.4d 33.2c 2.7c 22.7b 8.5a 7.5a 3.9a 46.5b 5.7c 11.9b 7.4b
Y waste 450.5de 88.2cd 70.1c 56.5c 38.5ab 6.8a 10.3a 7.7a 34.2b 8.4bc 12.5b 9.7b
H/Berm 1391.2cb 796.4b 924.7a 865.6a 89.8ab 64.3a 27.7a 20.3a 43.3b 15.4ab 17.5ab 13.8b
H/fence 1008.3cd 644.3b 483.0b 412.0b 188.2ab 171.6a 41.0a 26.7a 40.1b 13.8abc 20.5ab 12.8b
Bare soil 76.7e 53.4cd 0.6c 0.54c 92.0ab 60.1a 22.0a 0.33a 102.9a 20.1a 26.9ab 19.4ab
Treatments with same letter are not significantly different at α= 0.05 using Duncan’s Multiple range test.
*MSW = municipal solid waste.
TN = total nitrogen.
TP = total phosphorus.
DRP = dissolved reactive P.
Table 7. Results from correlation analysis. This table lists all response variables with
significant correlation (r > 0.70, α= 0.05, n = 21).
Independent variable (treatments) with
Response variable correlation coefficient
Total nitrogen (N) concentration, Storm #1 NH4(0.92), NO3(0.92)
Total N load, Storm #1 NH4(0.93), NO3(0.92)
Nitrate-N concentration, Storm #1 NH4(0.94), NO3(0.94)
Nitrate-N load, Storm #1 NH4(0.94), NO3(0.93)
Ammonium-N concentration, Storm #1 NH4(0.99), NO3(0.98)
Ammonium-N load, Storm #1 NH4(0.97), NO3(0.96)
Total P concentration, Storm #1 C:N ratio (0.88), OM(0.74), C(0.73)
Total P load, Storm #1 C:N ratio (0.89), OM(0.72), C(0.72)
DRP concentration, Storm #1 C:N ratio (0.88), OM(0.70), C(0.69)
DRP load, Storm #1 C:N ratio (0.88), OM(0.71), C(0.769)
Total N concentration, Storm #2 NH4(0.91), NO3(0.91)
Nitrate-N concentration, Storm #2 NH4(0.84), NO3(0.82)
Ammonium-N concentration, Storm #2 NH4(0.96), NO3(0.96)
Ammonium-N load, Storm #2 NH4(0.71), NO3(0.70)
DRP = dissolved reactive P.
Copyright © 2005 Soil and Water Conservation Society. All rights reserved.
www.swcs.org 60(6):288-297 Journal of Soil and Water Conservation
N
|
D 2005 VOLUME 60 N UMBER 6295
may not be desirable for use adjacent to, or
potentially in, surface water.
During the second storm event, major
nitrate load reductions were observed in all
treatments excluding the control and the
municipal solid waste compost, which had
low losses from the first storm. It’s likely that
substantial amounts of nitrate were either
taken up by plants, moved into the soil pro-
file, or was already lost in the runoff from the
first storm for most treatments. Additionally,
it’s possible that minor amounts were lost
to denitrification—particularly because the
treatments were saturated from the first storm
event potentially creating anaerobic condi-
tions favorable to denitrification (Sylvia et al.,
1999), and certainly for all treatments after
the second storm. Three of the compost
treatments had extremely low loads of nitrate,
between 3 to 9 mg/m2 (0.0001 to 0.0002
oz/yd2
)
, mostly because they generated very
little runoff during this stor m event,an attrib-
ute that may make them attractive for other
stormwater management applications.
By the final storm event, the bare soil lost
significantly greater nitrate, followed by both
hydroseeded treatments. This was due to the
higher volume of runoff produced by these
treatments relative to the compost treatments.
All compost treatments lost less than half
the amount of nitrate as the control, ranging
between 5 mg/m2and 10 mg/m2 (0.0002 and
0.0004 oz/yd2
)
. Low nitrate content com-
posts (municipal solid waste compost and yard
waste compost in this study) may be desirable
in other stormwater management projects and
applications near surface water, as the nitrate
loads from these treatments were generally
lower than the bare soil throughout the study.
Ammonium N loads were also determined
(data not shown) as trends were consistent
with nitrate N results for all treatments.
Total P loss. While phosphorus is not
toxic to animals and humans it is one of the
main causes of eutrophication in surface
water,which can lead to impaired water qual-
ity. Total P concentrations generally charac-
teristic to wastewater treatment plant dis-
charges is 5 mg L-1 (5 ppm
)
, while the critical
concentration of total P (particulate P + dis-
solved P) in streams at which eutrophication
is triggered is 0.10 mg L-1 (0.10 ppm
)
, and
0.03 mg L-1 (0.03 ppm
)
for dissolved P
(Brady and Weil, 1996). The total amount of
phosphorus applied by each treatment was 95
g/m2(4 oz/yd2
)
from the poultry litter com-
post, 85 g/m2(3.6 oz/yd2
)
from the biosolids
compost, 23 g/m2(1.0 oz/yd2
)
from the
municipal solid waste compost, 61 g/m2(2.6
oz/yd2
)
from the yard waste compost, and 10
g/m2 (0.4 oz/yd2
)
from the hydroseeding.
During the first storm event the hydroseed
treatments had the highest total P runoff
loads, 925 mg/m2(0.04 oz/yd2
)
from the
hydroseed with mulch berm and 483 mg/m2
(0.02 oz/yd2
)
from the hydroseed with silt
fence treatment, both were significantly dif-
ferent from the control and the composts
(Table 6). This was likely due to the soluble
fertilizer P in the hydroseed. Eghball and
Gilley (1999) evaluated storm runoff from
agricultural fields with sorghum residue and
found higher total P concentrations from fertiliz-
er applications (2.12 kg ha-1; 11.7 lb ac-1) com-
pared to compost (0.93 kg ha-1;5.1 lb ac-1) dur-
ing a second storm event.
During the second storm event, all treat-
ments, with the exception of the control, had
major reductions in total P loads in the
runoff. This was likely because most P was
already lost during the first storm event,as 85
percent of the total P mass loading occurred during
the first storm event, while the vegetation likely
took up lesser amounts and/or it moved into
the soil profile. The hydroseed treatments
had the greatest reductions between the first
and second storm events. This was likely
because the P fertilizer in the hydroseed was
in soluble form and therefore more likely
transported in the initial runoff event.
By the final storm the biosolids compost
(46.2 mg/m2;0.0002 oz/yd2) had greater total P
loads than the yard waste (12.5 mg/m2; 0.0006
oz/yd2
)
and municipal solid waste (11.9 mg/m2;
0.0005 oz/yd2
)
composts, likely because of the
elevated P content of the biosolids compost rel-
ative to the other two composts.
Total P mass load for each treatment over
the one year study period was 970 mg/m2
from the hydroseed with mulch, 545 mg/m2
from the hydroseed with silt fence,
257 mg/m2(0.01 oz/yd2
)
from the biosolids
compost, 119 mg/m2(0.005 oz/yd2
)
from the
poultry litter compost, 93 mg/m2(0.004
oz/yd2
)
from the yard waste compost, 53
mg/m2(0.0022 oz/yd2
)
from the municipal
solid waste compost, and 50 mg/m2(0.002
oz/yd2
)
from the bare soil. The higher the
total P content in the compost, the greater
the P load in the runoff. Additionally, the
compost treatments that lost the most P had
near neutral pH; while the municipal solid
waste compost and yard waste compost pH
levels were near 8.0, potentially meaning that
the P was not as mobile in high pH composts
because it is bound to Ca and/or Mg. The
poultry litter compost with gypsum had 54
percent less P loss than the biosolids compost
although the P content of the treatments was
similar. This may give some evidence that
calcium sulfate (gypsum) can reduce P losses
from compost blankets;however, more testing
is needed to draw any conclusions.
Total P lost in the runoff as a percent of the
total P applied from the treatments for all
three storms combined was 9.7 percent from
the hydroseed with mulch berm, 5.4 percent
from the hydroseed with silt fence, 0.4 per-
cent from the biosolids compost, 0.2 percent
from the municipal solid waste compost, and
0.1 percent for both the yard waste compost
and poultry litter compost. This provides
evidence that soluble P fertilizer in hydroseed
is more likely to be lost in storm runoff than
the organic P supplied from compost, both in
percentage of the total amount applied and in
total loading if runoff reaches surface water.
Dissolved reactive P loss. Reactive P
entering surface water is of particular concern
because it is often available to aquatic plants
for immediate uptake, leading to increased
risk of eutrophication. During the first storm
event dissolved reactive P loads were highest
among the hydroseeded treatments, 866
mg/m2(0.036 oz/yd2
)
and 412 mg/m2 (0.012
oz/yd2
)
, and both were significantly different
from the composts and the control. Eghball
and Gilley (1999) also found elevated dis-
solved and bioavailable P in runoff from fer-
tilizer compared to compost application, in
agricultural fields with wheat and sorghum
residue,after the first rain event. The low dis-
solved reactive P loading exhibited from the
municipal solid waste compost may be the
result of the high pH (8.1) of this compost (in
addition to low runoff volumes), which
would reduce its solubility because of adsorp-
tion to Ca and Mg. These results were inter-
esting for several reasons. First, it is well doc-
umented (Brady and Weil, 1996) that highly
weathered clay soils (like iron (Fe)/aluminum
(Al) oxides represented in this study) have a
relatively high capacity to fix dissolved P (as
PO4-), and it is likely that most of the dis-
solved P in hydroseed comes in contact with
the soil which should create a higher propen-
sity for the soluble P from hydroseeding to
adsorb to clay colloids and become insoluble
and unavailable. Second, organic matter
(much higher in the compost treatments) can
inhibit the adsorption of dissolved P to soil
Copyright © 2005 Soil and Water Conservation Society. All rights reserved.
www.swcs.org 60(6):288-297 Journal of Soil and Water Conservation
JOURNAL OF SOIL AND WATER CONSERVATION N
|
D 2005
296
less easily transported relative to soluble
forms. In light of this, state and federal com-
post specifications for erosion control should
incorporate a nutrient component.
The potential for high losses of P from
hydroseeding applications needs to be
addressed by the policy and regulatory com-
munity, particularly since it is one of the most
ubiquitous erosion control best management
practices in the United States. Erosion con-
trol materials high in nutrients, particularly
nutrients in soluble and inorganic forms,
increase the risk of nutrients entering water
bodies; although, because compost can signif-
icantly reduce runoff, nutrient loads are often
lower from these materials relative to other
best management practices. Future research
should focus on testing and improving exist-
ing state and federal specifications for com-
post use in erosion control and stormwater
management applications. Specific attention
should be given to particle size distribution of
compost materials, flow through rates of filter
berms, slope steepness, potential to remove
other pollutants in runoff such as petroleum
hydrocarbons, and the optimum nutrient
contents and forms in materials used for
stormwater management applications.
References Cited
Adams, J.E. 1966. Influence of mulches on runoff, erosion,
and soil moisture depletion. Soil Science Society of
America Proceedings 30:110-114.
Agassi, M., A. Hadas, Y. Benyamini, G.J. Levy, L. Kautsky, L.
Avrahamov, and H.Zhevelev. 1998.Mulching effects of com-
posted msw on water percolation and compost degradation
rate. Compost Science and Utilization 6(3):34-41.
Alexander, R. 2003. National erosion/sediment control
specifications for composted products. Recycled
Materials Resource Center, University of New
Hampshire and American Association of State Highway
Transportation Officials. Durham, New Hampshire.
Buchanan, J.R., D.C. Yoder, and J.L. Smoot. 2000.
Controlling soil erosion on construction sites with steep
slopes with wood chips. 2000 American Society of
Agricultural Engineer s Annual International Meeting.
Milwaukee,Wisconsin.
Brady, N.C. and R.R.Weil. 1996. The nature and properties
of soils: 11th edition.Prentice Hall, Inc. New Jersey.
Demars, K.R. and R.P. Long. 1998. Field evaluation of
source separated compost and Coneg model procure-
ment specifications for Connecticut DOT projects.
University of Connecticut and Connecticut Depart-
ment of Transportation. December, 1998. Report No.
JHR 98-264.
Doolette, J.B. and J.W. Smyle. 1990. Soil and moisture con-
servation technologies: review of literature. In: J.B.
Doolette and W.B. Magrath (eds.) Watershed develop-
ment in Asia: Strateg ies and technologies. World Book
Technical Paper No. 127. Washington, D.C.
Eghball, B. and J.E. Gilley. 1999. Phosphorus and nitrogen in
runoff following beef cattle manure or compost applica-
tion. Jour nal of Environmental Quality 28:1201-1210.
colloids by physically blocking exchange sites,
chelating Fe and Al, thus preventing reaction
with P ions,and organic acids can displace P ions
by filling up potential exchange sites on clay par-
ticle surfaces (Brady and Weil, 1996). These
conditions make P loss more favorable from
compost than hydroseed; however, the results
indicate just the opposite. This is likely because
the soluble P (more mobile in runoff than the
organic bound P normally characteristic to
compost) from the hydroseed saturated the P
fixing capacity of the soil quite quickly allowing
greater amounts of P to leave the system.
During the second storm event a statistical
difference in dissolved reactive P concentration
(but not loads) was observed between the
biosolids (5.9 mg L-1, 5.9 ppm) and the poul-
try litter composts (1.3 mg L-1, 1.3 ppm).
This may be the result of the soluble P in the
poultry litter compost reacting with the gyp-
sum (calcium sulfate), but further evaluation
would be required to draw conclusions.
During the final storm event, dissolved
reactive P loads had decreased relative to
the first storm event but were similar to the
second storm event, probably due to the
increased runoff volumes experienced during
the final storm. The biosolids compost was
significantly higher than all other treatments
except the control—likely due to the high
runoff volumes generated by the control.
Correlation analysis for nutrient loss.
Results from correlation analysis (Table 7)
were used to evaluate which of the treatment
physical, chemical, and biological characteris-
tics, and rainfall and vegetation growth results
(independent variables) were correlated with
the parameters from nutrient loss results
(response variables).
Total P and dissolved reactive P concentra-
tions and loads from the first storm event were
correlated to organic matter content, C content
and C:N of the compost treatments (r > 0.70).
Generally, the higher the organic matter con-
tent, C content, and C:N of the compost the
lower the P concentration and load in the
resulting storm runoff. This may indicate that
some dissolved inorganic P in the runoff react-
ed with humus colloids in the compost treat-
ments or that microorganisms immobilized
some dissolved inorganic P because of insuffi-
cient P relative to C (Brady and Weil, 1996). It
could also indicate a higher percentage of
organic P relative to soluble P content (although
not directly tested) in compost can lead to less P
loss during the first storm event after applica-
tion, assuming erosion is kept to a minimum.
Summary and Conclusion
Under these experimental conditions com-
post systems generally performed as well or
better than slit fence and hydroseeding in
reducing storm runoff. Specifically, compost
systems produced significantly less runoff
than hydroseed during storms after vegetation
establishment and once vegetation was
mature, with significantly greater infiltration
for all storm events. Compost reduced runoff
more over one year than hydroseeding, by 33
percent and 8 percent respectively, and
reduced total cumulative runoff relative to the
control by 55 percent and 30 percent, respec-
tively. Under intense rainfall, compost
systems significantly delayed the onset of
runoff by a cumulative average of 15 minutes,
compared to hydroseed, and significantly
reduced the elapsed time until peak runoff
rate after vegetation was established.
Cumulative peak runoff rates were 60 percent
lower for the compost treatments relative to
the control and 34 percent lower relative to
hydroseed, thereby reducing erosion poten-
tial. Compost systems reduced total solids by
97 to 99 percent relative to a bare soil, and
had 3.5 times less solids loss during the first
storm after application and 16 times less solids
loss during a second storm, relative to
hydroseed and silt fence.
Materials high in total N and total P are
likely to lose more of each nutrient to storm
runoff; however, N and P loading is greatly
diminished after the first runoff event.
Because hydroseed is applied with inorganic
N and soluble P it is more likely that these
nutrients will be lost to storm runoff and
consequently are in forms more available to
aquatic plants. Mass loading of total P from
hydroseed was five times greater than com-
post, and dissolved reactive P was six times
greater than compost; even though the total
amount of P applied was two to nine times
less from hydroseed relative to compost.
Composts high in inorganic N generated
higher N loads in runoff, therefore compost
with a high percentage of organic N of the
total N is recommended. Additionally, high
concentrations of C, organic matter and Ca
(as added gypsum) in compost may reduce P
loss in runoff. The high C content (and by
relation organic matter content) of composts
is likely very important in minimizing P loss.
This results either from P immobilization
characteristic in high C:P ratio materials or it
is an indicator that the chemical constituents
in compost are organic in form and therefore
Copyright © 2005 Soil and Water Conservation Society. All rights reserved.
www.swcs.org 60(6):288-297 Journal of Soil and Water Conservation
N
|
D 2005 VOLUME 60 N UMBER 6297
Ehrhart, B.J., R.D. Shannon, and A.R. Jarrett. 2002. Effects
of construction site sedimentation basins on receiving stream
ecosystems.Transactions of the ASAE 45(3):675-680.
Ettlin, L. and B. Stewart. 1993. Yard debris compost for
erosion control. BioCycle: Journal of Composting and
Organics Recycling 34(12):46-47.
Faucette, L.B., L.M. Risse, M.A. Nearing, J.W. Gaskin, and
L.T. West. 2004. Runoff, erosion, and nutrient losses
from compost and mulch blankets under simulated
rainfall. Journal of Soil and Water Conservation
59(4):154-160.
Ferre, P.A. and G.C.Toppe. 2002. Time domain reflectome-
try. Pp. 434-445. In: J.H. Dane and G.C. Topp (eds.)
Methods of soil analysis, Par t 4. Soil Science Society of
America, Madison, Wisconsin.
Foster, G.R., R.A. Young, M.J.M. Romkens, and C.A.
Onstad. 1985. Processes of soil erosion by water.
Pp. 137-162. In: Soil erosion and crop productivity.
R.F. Follet and B.A. Stewart (eds.) Agronomy Society of
America, Crop Science Society of America, and Soil
Science Society of America, Madison, Wisconsin.
Georgia Soil and Water Conservation Commission (GaSWCC).
2002. Erosion and sediment control course manual. Georgia
Soil and Water Conservation Commission.
Gilley, J.E. and L.M. Risse. 2000. Runoff and soil loss as
affected by the application of manure. Transactions of
the ASAE 43:6.
Glanville,T.D., R.A. Persyn, and T.L.Richard. 2001. Impacts
of compost application on highway construction sites in
Iowa. 2001 American Society of Agricultural Engineers
Annual International Meeting. Sacramento, California.
Paper No. 01-012076.
Gorman, J.M., J.C. Sencindiver, D.J. Horvath, R.N. Singh,
and R.F. Keefer. 2000. Erodibility of fly ash used as a
topsoil substitute in mineland reclamation. Journal of
Environmental Quality 29:805-811.
Holmberg, G.V. 1983. Landuse, soils, and revegetation.
Pp. 279-350. In: L.V.A. Sendlein (ed.) Surface mining
environmental monitoring and reclamation handbook.
Elsevier Science Publishers, New York, New York.
Iannotti, D.A.,T. Pang, B.L.Toth, D.L. Elwell, H.M. Keener,
and H. Hoitink. 1993. A quantitative respirometric
method for monitoring compost stability. Compost
Science and Utilization 1(3):52-65.
Jordan, C.F. 1998. Working with nature: Resource manage-
ment for sustainability. Harwood Academic Publishers,
The Netherlands.
Laflen, J.M., J.L. Baker, R.O. Har twig, W.A. Buchele, and H.P.
Johnson. 1978. Soil and water loss from conservation tillage
systems.Transactions of the ASAE 21:881-885.
Lyle, E.S. 1987. Surface mine reclamation manual. Elsevier
Science Publishers, New York, New York.
Meyer, L.D., C.B. Johnson, and G.R. Foster. 1972. Stone and
woodchip mulches for erosion control on construction sites.
Journal of Soil and Water Conser vation 27:264-269.
Michaud, M.1995. Recycled materials used as erosion con-
trol mulches. In: Compost Utilization in Horticultural
Cropping Systems. P.J. Stoffella and B.A. Kahn (eds.)
Lewis Publishers.
Portland Metro. 1994. Summary of projects using yard
debris compost for erosion prevention and control.
Final report. June, 1994. Solid waste and planning
Department. Portland, Oregon.
Qualls, R.G. 1989. Determination of total nitrogen and
phosphorus in water using persulfate oxidation: A
modification for small sample volumes. Appendix A.
Pp. 131-138. In: The biogeochemical properties of
dissolved organic matter in a hardwood forest ecosys-
tem: Their influence on the retention of nitrogen,
phosphorus, and carbon. Ph.D. dissertation, University
of Georgia Institute of Ecology, Athens, Georgia,
University Microfilms, Inc., No. DEX9003448.
Risse, L.M. and B. Faucette. 2001. Compost utilization for
erosion control. University of Georgia Cooperative
Extension Service Bulletin No. 1189. CAES-UGA.
Athens, Georgia.
SAS,2001. SAS System 8.2 for Microsoft Windows.The SAS
Institute, Inc. Cary, North Carolina.
Sylvia, D.M, J.J. Fuhrman, P.G. Hartel, and D.A. Zuberer.
1999. Transformations of nitrogen. Pp. 259-293. In:
Principles and applications of soil microbiology.
Prentice Hall, Inc. New Jersey.
U.S. Composting Council. 1997.Test methods for the exam-
ination of composting and Compost. First Edition.The
United States Composting Council, Amherst, Ohio.
U.S. Department of Agriculture Agricultural Research
Service (USDA-ARS). 1980. TACQ time domain
reflectometry (TDR) systems control program.
Bushland,Texas.
U.S. Department of Agriculture Soil Conservation Service
(USDA-SCS). 1968. Soil Survey Clarke and Oconee
Counties, Georgia. Washington, D.C.
U.S. Department of Agriculture (USDA). 1998. Soil quality
test kit guide. U.S. Department of Agriculture.
Washington, D.C, Pp.7-10, 55-58.
U.S. Department of Commerce National Weather Service
(USDOC-NWS). 1961.Technical paper No. 40 of the
National Weather Service. www.erh.noaa.gov/er/hq/
Tp40s.htm.
U.S. Environmental Protection Agency (USEPA). 1983.
Methods for chemical analysis of water and wastes,
EPA-600/4 4-79-020. U.S. Environmental Protection
Agency, Cincinnati, Ohio.
U.S. Environmental Protection Agency (USEPA). 1993. 40
CFR Part 503 - Standards for use and disposal of sewage
sludge: Final rule federal register 58:9248-9415.
U.S. Environmental Protection Agency (USEPA). 1997.
Innovative uses of compost: Erosion control,turf remediation
and landscaping. EPA 530-F-97-043. October,1997.
U.S. Environmental Protection Agency (USEPA). 2000.
Stormwater Phase II final rule: Construction site runoff
control minimum control measure. Office of Water
(4203). EPA 833-F-00-008, Fact Sheet 2.6.
University of Georgia Institute of Ecology. 2004. Institute of
ecology analytical chemistry lab methods for soil chem-
ical analysis. www.uga.edu/sisbl/.
University of Georgia Soil, Plant and Water Analysis Lab.
2004. Soil, plant and water lab methods reference/soil
methods summary - analytical methods for soil chemi-
cal and texture analysis. http://aesl.ces.uga.edu/publica-
tions/methods/stl-soil.html.
Vleeschauwer,D.D.and M.D. Boodt. 1978.The comparative
effects of surface applications of organic mulch versus
chemical soil conditioners on physical and chemical
properties of the soil and on plant growth. Catena
5:337-349.
Weather Channel. 2004. www.weather.com/weather/clima-
tology/monthly/30602
Wilson, B.N., H. Nguyen, and D. Stenlund. 2001. Impact of
erosion control blankets on runoff depth and sediment
yield. Amer ican Society of Agricultural and Biological
Engineer’s International Conference Proceedings,
Sacramento, California. Paper No.012151.
Wischmeier, W.J. and D.D. Smith. 1978. Predicting rainfall
erosion loss: A guide to conservation planning. U.S.
Department of Agriculture Agricultural Handbook No.
537,Washington, D.C.
Copyright © 2005 Soil and Water Conservation Society. All rights reserved.
www.swcs.org 60(6):288-297 Journal of Soil and Water Conservation
