United States Office of Research and Office of EPA/625/R-97/008
Environmental Protection Development Water December 1997
Agency Washington, DC 20460 Washington, DC 20460
&EPA Proceedings: National
USDA Watershed Water Qua!ity
Project Symposium
-------
FOREWORD
The lessons learned from watershed projects addressing nonpoint source
problems are recorded in these proceedings of the National Watershed Water
Quality Project Symposium, held September 22-26,1997, in Washington, DC.
The Symposium was conducted by the: U.S. Environmental Protection Agency, U.S.
Department of Agriculture; Cooperative State Research Education & Extension Service,
Natural Resources Conservation Service, Maryland Department of Natural Resources, and
Conservation Technology Information Center (CTIC). The Symposium featured
accomplishments of local projects funded under EPA's Section 319 (Clean Water Act) National
Monitoring Program and USDA's Demonstration, Hydrologic Unit Area Programs,
and Management Systems Evaluation Areas. The symposium also featured
lessons learned in the Farm*A*Syst and Home*A*Syst programs.
ACKNOWLEDGMENTS
The success of the symposium and the information transferred from it and
the proceedings is due to the expertise and efforts of many individuals.
The sum of these are greater than the each individual part and will
hopefully be multiplied many times.
Authors
A special thanks goes to the many authors of the papers and abstracts
presented in this document. Their efforts in paper preparation and editing
made this document possible.
Peer Review
Numerous individuals were involved in the peer review process, too numerous
to list, but their invaluable insights and comments improved the scope and
direction of each of the papers that they reviewed.
iii
-------
Program Committee
Appreciation goes out to those individuals on the committee that
represented their organizations in planning the program. Their tireless
efforts provided insight to the agenda and the tours.
Jack Clausen, University of Connecticut, Storrs, Connecticut '
Tom Davenport, U.S. Environmental Protection Agency, Chicago, Illinois
Steve Dressing, U.S. Environmental Protection Agency, Washington, D.C.
Gary Jackson, University of Wisconsin, Madison, Wisconsin <
Lyn Kirschner, Conservation Technology Information Center, West Lafayette,
Indiana !
Pat Lietman, U.S. Geological Survey, Harrisburg, Pennsylvania '
John McCoy, Maryland Department of Natural Resources, Annapolis, Maryland '
Don Meals, University of Vermont, Montpelier, Vermont !
Dan Murray, U.S. Environmental Protection Agency, Cincinnati, Ohio I
Deanna Osmond, North Carolina State University, Raleigh, North Carolina
Lynette Seigley, Iowa Geological Survey, Iowa City, Iowa !
Dan Smith, USDA Natural Resources Conservation Service, Washington, D.C. |
Jean Spooner, North Carolina State University, Raleigh, North Carolina '
Dave Rathke, U.S. Environmental Protection Agency, Denver, Colorado j
Mary Ann Rozum, USDA Cooperative State Research Education & Extension
Service, Washington, D.C.
IV
-------
Contents
Foreword iii
Acknowledgments iii
Monitoring the Effects of Nonpoint-Source Pollution Controls on Sny Magill Creek, Clayton County, Iowa 1
The Missouri MSEA Project: A Model for "The Partnership Approach" to Water Quality Concerns 9
Lake Pittsfield Watershed Project - A Cooperative Effort 15
The Farm/Field*A*Syst Decision Support Systems 21
Utilizing Voluntary Farmstead Assessments to Encourage Best Management Practice Adoption in the
Skaneateles Lake Watershed 29
Multi-year multi-rate demonstrations of manure application for corn fertility in northeast Iowa 35
Adoption of Best Management Practices (BMPs) to Meet Water Quality Goals in the Granger Drain Hydrologic Unit Area 41
Demonstration and Hydrologic Unit Projects in North Carolina: The Team Approach to Improving Water Quality 47
What Have We Learned About Our Nonpoint Source Pollution Education Programs? 53
The Royal River Watershed Eduction Project 61
The Oak Creek 319 (h) National Monitoring Program 67
Communication and Adoption Evaluation of USDA Water Quality Demonstration Projects 73
Gum Creek Water Quality Demonstration Project..-. 79
The Implementation of Innovative Best Management Practices in the Sny Magill Watershed 85
Nitrate Losses Under Various Nitrogen Management Systems ; 91
Results and Lessons Learned from the Beaver Creek Hydrologic Unit Area Project in West Tennessee 97
Subsurface Drainage Outflow Improvement with Constructed Wetland 103
Restoration of the Waukegan River 109
Evaluation of Two Basins Prior to Streambank Fencing in Pastured Areas within the Mill Creek Basin of
Lancaster County, Pennsylvania 117
Lessons Learned in the Long Creek Watershed Project 127
-------
Contents (continued)
The Monocacy River Watershed Water Quality Demonstration Project: A Commitment to Water Quality 133
Environmental Assessments for Real Estate Professionals 139
Sand Mountain-Lake Guntersville Hydrologic Unit Area Progress Report: 1990-1996 145
The Muddy Fork HUA - A Water Quality Success Story 151
Effectiveness of Barnyard Best Management Practices in Wisconsin ; 157
The Nebraska MSEA Project Management of Irrigated Corn and Soybeans to Minimize Ground Water Contamination 169
North Carolina Agricultural Systems for Environmental Quality (ASEQ) Project 175
Assessing the Impacts of Voluntary Pollution Prevention Programs in Agriculture: Lessons Learned from a
Cost-Benefit Evaluation of Farm*A*Syst 181
Analysis of Long Creek Watershed Monitoring Data 195
Jordan Cove Urban Watershed Section 319 National Monitoring Program Project 201
Partnerships in Puget Sound Watershed Remediation: Linking Water Quality to Pollution Controls 207
The National Water-Quality Assessment Program—At 5 Years Old 215
Detecting Fecal Contamination and Enteric Microbes in Watersheds 219
The Role of Consultants in Addressing Water Quality Issues 221
Twenty Years of Change: The Lake Erie Agricultural Systems for Environmental Quality (LEASEQ) Project 223
Elm Creek HUA: Results and Impact 231
HUA Crop Management Service Continues As Farmer-Run Crop Management Association 233
On-Farm Field Trials Demonstrate and Quantify the Environmental and Economic Benefit of BMPs 239
Arizona's Water Quality and Hydrologic Unit Area Projects to Promote BMPs 241
Accomplishments and Challenges of Data Collection and Modeling in North Carolina Demonstration Watershed Project .'243
Phosphate Concentrations in Subsurface Drainage Effluent in East-central Illinois 249
The Success of Farm*A*Syst and Safe H2OM£ Water Quality Programs in Maine 257
VI
-------
Monitoring the Effects of Nonpoint-Source Pollution Controls on Sny
Magill Creek, Clayton County, Iowa
Lynette S. Seigley
Iowa Department of Natural Resources - Geological Survey Bureau, Iowa City, IA
Mike W. Birmingham and Mike D. Schueller
University Hygienic Laboratory, Iowa City, IA
Jayne E. May
U.S. Geological Survey, Iowa City, IA
Gaige Wunder
Iowa Department of Natural Resources - Fisheries Bureau, Decorah, IA
Tom F. Wilton
Iowa Department of Natural Resources - Water Quality Bureau, Des Moines, IA
The Sny Magill Creek watershed in Clayton County, Iowa (Figure 1), is the location of the Sny
Magill Creek Nonpoint Source Pollution Monitoring Project, part of the U.S. EPA's National
Monitoring Program. Sny Magill Creek drains a 35.6 mi2 agricultural watershed, primarily in
forest, forested pasture, rowcrop, cover crop, and pasture. The Sny Magill Project, initiated in
October 1991, is designed to monitor and assess improvements in water quality from the
implementation of best management practices (BMPs) in the watershed. The BMPs are
intended to reduce sediment delivery to Sny Magill Creek and to reduce fertilizer and pesticide
inputs. A paired watershed approach is being used; Bloody Run Creek, the adjacent
watershed to the north, is serving as the control watershed. The objectives of the water-quality
project are to show reductions in sediment, nitrate, and pesticide concentrations in Sny Magill
Creek relative to Bloody Run Creek, and to document improvements in the biological habitat
through a habitat assessment of the stream corridor and monitoring of the benthic
macroinvertebrate and fish populations. At this time, results from the monitoring are mixed.
Improvements in macroinvertebrates and pesticide detections have been measured, while fish,
habitat, and nitrate and sediment loads are unchanged. This paper summarizes the water-
quality monitoring results to date. Information on BMPs implemented in the Sny Magill
watershed can be found in Palas and Tisl (1997, this volume).
Introduction
Sny Magill Creek is a coldwater stream in northeastern Iowa. Identified water-quality
impairments include sediment, nutrients, and pesticides. Sny Magill Creek is managed for "put
and take" trout fishing and is one of the more widely used recreational fishery streams in Iowa.
-------
Iowa's State Nonpoint Source Management Report (IDNR, 1989) identified Sny Magill Creek as
a priority for project action to improve water quality. The stream was further designated as
"high-quality waters" and was to be protected against degradation of water quality. As a result
of water quality concerns, the Sny Magill Hydrologic Unit Area land treatment project was
implemented in the watershed in 1991 to reduce sediment delivery to Sny Magill Creek and to
reduce fertilizer and pesticide inputs (see Palas and Tisl, this volume).
Fish Assessment
An annual fish assessment is conducted at six sites in Sny Magill and Bloody Run watersheds
during the fall of each year (Figure 1). The sample date is selected to minimize stocked trout
numbers and angler interference with fish sampling personnel, and to sample the streams
under baseflow conditions. Two backpack-mounted stream electrofishing units are used to
sample a 300-foot stream reach of mixed pool-riffle habitat at each site.
Results of the fish survey from all years show that the streams' forage fish populations are
typical of Iowa coldwater streams. Twelve fish species have been identified. For all survey
years, the dominant fish is the fantail darter in Sny Magill Creek and the slimy sculpin in Bloody
Run Creek. With the exception of 1995 and 1996, year-to-year fluctuations in fish populations
appear to be a normal response to variations in precipitation, runoff, water clarity, and water
stage. Extremely low numbers were reported for the five Sny Magill sites during 1995 and
1996. The cause of the low numbers is not known. Chemical water-quality data and surveys of
the benthic macroinvertebrate populations showed no negative response during this period.
An herbicide spill at site SN2 on May 19,1995, may have impacted the fish populations.
The index of biotic integrity (IBI) is a widely used tool for evaluating the environmental health of
a stream or river based on its fish population. The IBI of Lyons and others (1996), developed
for coldwater streams in Wisconsin, classified the five Sny Magill sites as "very poor" to "poor"
and the Bloody Run site as "poor" to "fair" for any given year. The IBI scores have shown little
change and low fish numbers from Sny Magill Creek for 1995 and 1996 prevents an IBI from
being calculated for those years. Of the 12 forage fish species identified, only one, slimy
sculpin, is classified by Lyons and others (1996) as intolerant to contamination. Slimy sculpin
dominate the Bloody Run population (43 to 98% any given year), yet are not found in Sny
Magill Creek. It is unlikely that additional fish species, especially intolerant species, will be
found in either Bloody Run or Sny Magill creeks unless the species are physically reintroduced.
Both Sny Magill and Bloody Run are coldwater streams separated from other coldwater
streams by the Mississippi River, a warmwater body. This thermal isolation barrier limits inter-
stream movement of coldwater fish species. The IBI of Lyons and others (1996) has its
limitations (e.g., one metric is based on the number of intolerant species present), but will
continue to be used until a more suitable/appropriate IBI is developed and tested using fish
data from Iowa coldwater streams.
Habitat Assessment
The annual habitat assessment, designed to characterize stream habitat conditions, occurs in
the fall under low-flow, baseflow conditions at eight water-quality sites (Figure 1; site BRSC was
not included). Instream and streamside habitat variables are measured and observed at ten
regularly spaced, cross-sectional stream transects within a 100-foot stream reach. Each
stream reach includes two or three sets of pools and riffles.
-------
The aquatic habitat characteristics were compared using a simple ranking process and habitat
similarity index. The results suggest a relation between the drainage area size and position in
the landscape to the habitat variables. Monitoring sites with similar drainage size showed
greater habitat similarity to each other than to other sites. The apparent interrelatedness of
habitat, drainage area size, and channel slope suggests that physiography and stream
morphological processes such as channel erosion and sediment deposition are important
determinants of monitoring site habitat character.
An original goal of the habitat assessment was to determine if there were any trends related to
implementation of the land treatment changes in the Sny Magill watershed. The annual habitat
assessment is good at characterizing habitat but the frequency of the assessment would need
to be increased to monitor year-to-year trends attributable to land treatment changes.
Benthic Macroinvertebrate Monitoring
Benthic macroinvertebrate monitoring of Sny Magill and Bloody Run creeks occurs each year
on a bi-monthly basis (April, June, August, and October) at eight water-quality sites (Figure 1;
site BRSC was not included). Samples are collected in triplicate at each location for each
month using a Modified Hess bottom sampler. Laboratory processing and data analysis is
performed as described in Plafkin and others (1989).
The benthic community composition for both watersheds is similar. Figure 2 shows the metric
results for the HBI values, taxa richness, EPT Index, and the percent dominant taxa. The metric
values represent the mean, on an annual basis, for the combined sites of SN1 and SN2, and
sites BR1 and BR2. The HBI value is used as a measure of organic pollution. HBI values range
from 0 to 5. Lower values indicate streams of high water quality. Taxa richness, a direct
measurement of the number of distinct taxa present in a sample, generally increases with
increasing water quality, habitat diversity, and/or habitat suitability. The EPT Index measures
the more pollution sensitive insect orders of the mayfly, stonefly, and caddisfly. The EPT Index
generally increases with increasing water quality. The percent dominant taxa, a measure of the
percent contribution of the numerically dominant taxon to the total population sampled, reflects
community evenness and redundancy. A high proportion of dominant taxa (>40%) may
indicate impairment of water quality.
Metrics that may indicate some discernible trends are the EPT Index and the percent dominant
taxa (Figure 2). Both suggest trends of improving water quality in Sny Magill Creek during the
monitoring period. The Bloody Run sampling sites have shown slight decreases in EPT Index
values each year, suggesting steady to worsening water quality. The Sny Magill Creek sites
have shown consistent increases in EPT values since 1992. During the monitoring period, the
percent dominant taxon metric has declined for Sny Magill Creek. The percent dominant taxa
metric values for Bloody Run Creek have fluctuated but shown no substantial improvements.
Until 1995, the HBI values suggested improving water quality at Sny Magill relative to Bloody
Run. In 1995, however, the HBI values increased for both creeks. Based on the HBI values for
all sites and all years, the water quality is rated as "very good."
Physical/Chemical Water-Quality Data
Stream and suspended-sediment discharge, as well as nitrate and pesticide concentrations
and loads, are measured for both Sny Magill and Bloody Run creeks. In spite of just one year
of calibration period data, a significant relationship did exist between Sny Magill and Bloody
-------
Run creeks for discharge, sediment, and nitrate load. The pesticide data, however, exhibited
no correlation during the calibration period.
From the calibration to the treatment periods, there has been no significant decline in the
nitrate loads for Sny Magill Creek relative to Bloody Run Creek. There has been a decline in
the frequency of pesticide detections in Sny Magill Creek, from 60% to 25%; the frequency of
pesticide detections in Bloody Run Creek has remained above 95%.
For both streams, the majority of a year's total sediment load is delivered during two periods:
a spring snowmelt period and a summer storm period. Although BMPs have effectively
reduced the sediment delivered from the uplands to Sny Magill Creek by an estimated 35%
(see Palas and Tisl, this volume), these reductions in sediment have yet to be reflected in the
sediment loads discharged by Sny Magill Creek. It is uncertain whether Sny Magill Creek will
show significant reductions in sediment load as a result of BMPs implemented. In spite of their
close proximity, some intense rainstorms have affected Sny Magill and not Bloody Run. Data
from 1996 illustrate the significance of these rainstorms. In 1996,14 days accounted for 90% of
the year's total sediment load for Sny Magill while 204 days accounted for 90% of Bloody Run's
annual total. There also is the concern over the large volume of historical sediment (post-
settlement alluvium; Bettis, 1994) in the drainage network. Though implementation of BMPs in
the uplands has reduced sediment delivery to Sny Magill Creek, the impact the large quantity
of sediment historically stored in the drainage network may have on the sediment loads
discharged from Sny Magill Creek is poorly understood.
Conclusion
Results from the Sny Magill Creek Project show some indication of improved water quality in
Sny Magill Creek. The frequency of pesticide detections has declined and some of the metrics
calculated for the benthic macroinvertebrates do suggest improving water quality for Sny Magill
Creek relative to Bloody Run Creek. However, the fish and habitat assessments give no
indication of improving water, and nitrate and sediment loads have not declined in Sny Magill.
Acknowledgements
This project is supported, in part, by a grant from the U.S. EPA, Region VII (Nonpoint Source
Program), and is administered by the Iowa Department of Natural Resources.
References
Bettis, E.A., III, 1994, Paleozoic Plateau erosion perspective, in Seigley, LS. (ed.), Sny Magill
watershed project: baseline data: Iowa Department of Natural Resources, Geological
Survey Bureau, Technical Information Series 32, p. 19-27.
Iowa DNR, 1989, State nonpoint source management report - Iowa, 68 p.
Lyons, J., Wang, L, and Simonson, T.D., 1996, Development and validation of an index of
biotic integrity for coldwater streams in Wisconsin: North American Journal of Fisheries
Management, v. 16, p. 241-256.
Palas, EA, and Tisl, J.A., 1997, The need for innovative best management practices in the Sny
Magill watershed, this volume.
-------
Plafkin, J.L, Barbour, M.T., Porter, K.D., Gross, S.K., and Hughes, P.M., 1989, Rapid
bioassessment protocols for use in streams and rivers: benthic macroinvertebrates and fish:
Assessment and Watershed Protection Division, Office of Water, U.S. Environmental
Protection Agency, Washington, D.C., EPA/440/4-89-001.
Figure 1. Location of Sny Magill and Bloody Run watersheds.
Figure 2. Benthic macroinvertebrate metrics for Sny Magill and Bloody Run creeks for HBI (A.),
taxa richness (B.), EPT Index (C.), and percent dominant taxa (D.).
-------
BLOODY RUN WATERSHED
SNY MAGILL WATERSHED
s
<
v
rt
V
1
1 1 I 1
x
i i i r
—
R
Bloody Run Creek
agiII Creek
1000 meters
5000 feet
Clayton
© Monitoring sites for fisheries surveys
I Monitoring sites for water quality
Figure 1
6
-------
2 —
5
1 —
1992 | | 1993 HI 1994
1995
Bloody Run (control) Sny Magill (treatment)
(BR1 + BR2) (SN1 + SN2)
Bloody Run (control)
(BR1 + BR2)
Sny Magill (treatment)
(SN1+SN2)
10
8
1992 | | 1993 ^| 1994
1995
Bloody Run (control) Sny Magill (treatment)
(BR1 + BR2) (SN1 + SN2)
a
£
I
i
1992 1993 ^m 1994 1995
Bloody Run (control) Sny Magill (treatment)
(BR1 + BR2) (SN1 + SN2)
Figure 2
-------
The Missouri MSEA Project:
A Model for "The Partnership Approach" to Water Quality Concerns
Myra Smith, Ph.D., MSEA Education Coordinator
University Extension, University of Missouri, Columbia, Missouri
Introduction
Water is one of the most abundant natural resources in Missouri. About 3.2 million Missouri
citizens obtain their drinking water directly from the Missouri and Mississippi Rivers. Over 80%
of rural Missourians obtain their drinking water from the numerous reservoirs and lakes that dot
the state. Both Missourians and non-Missourians alike enjoy the abundant recreational
activities associated with these water resources. It is vitally important that the quality of
Missouri's water resources be maintained over time.
The objective of the Missouri Management Systems Evaluation Area project (MSEA) is to
identify and/or develop agricultural cropping systems and practices that protect water
resources from contamination by agricultural chemicals and sediment. The objective of the
educational component is to facilitate the transfer of information and technology from all water
quality research programs, including the MSEA project, to the end users; in the form of feasible
practices that maintain and enhance water quality.
The MSEA projects are unique in their implementation approach, incorporating an extension/
educational component in even the earliest project stages. This allows a vast network of
linkages and partnerships to form among research, extension, industry and the ag-community;
ensures good communication and movement of information; and facilitates the voluntary
adoption by producers of environmentally sound practices.
Setting
Missouri's MSEA project activities have extended throughout northern Missouri. The primary
research site, located in a 28 square mile watershed in the north-central part of the state,
contains claypan soils representative of the 10 million acre Midwest claypan soil region. One
hundred forty surface water sites, representing 95 streams and 14 separate river systems
throughout northern Missouri, have also been monitored for herbicide and nutrient
contamination over the last four years. Sampling data from other monitoring programs,
conducted by industry and other agencies, have been evaluated as well. Demonstration sites,
highlighting economically viable low and non-atrazine alternatives and pesticide/nutrient BMPs,
have been initiated in six producer's fields in atrazine-sensitive areas. These demonstration
sites represent a collaborative effort among MSEA personnel, the Integrated Pest
Management Program at the University of Missouri, regional Extension and Natural Resources
-------
Conservation Service (NRCS) personnel, local community watershed committees, industry
representatives, and local dealers and producers. (See Figure 1).
Missouri MSEA
. Research, Demonstration, and Sampling Sites
, Research Site'' .'
A Demonstration Site
* Stream Sampling Site
Figure 1. Site map illustrating the extent of Missouri MSEA project activities.
Education and Technology Transfer
Educational efforts have been targeted at groups and individuals with the greatest potential to
influence others. Information and technology have been transferred to a diverse customer
base through the print media (newsletters, press releases, newspapers, state and national
magazine articles), radio, television, field days, tours, demonstrations, workshops, meetings,
and conferences. Technology transfer and education programs have increased pesticide use
safety, increased awareness of vulnerable water resources (e.g. over 1,300 abandoned wells
have been properly sealed and capped), and improved technical advice and recommendations
given to producers.
Primary emphasis this year was placed on educating communities regarding their local water
quality problems; demonstrating the effectiveness, practicality, and economic viability of
various best management practices (BMPs) as solutions to these problems; and providing
feedback regarding both the economic and environmental impact of these measures on local
surface water quality and agriculture.
Environmental Benefits Measured
No-Till Cropping Practices: The MSEA project was instrumental, and highly
successful, in encouraging the adoption of no-till cropping for control of erosion and
sedimentation. The use of no-till in Audrain County, the primary location of the Missouri
MSEA site, has increased over 8-fold during the 6-year project period; climbing from
10
-------
14,000 acres in 1990 to 139,500 acres in 1996, an increase from 5.4% to 45.2% of the
cropland acreage. For comparison, no-till acres in the state increased from 10% to
30%, and in the Midwest Region from 6% to 18%, over the same time period (CTIC,
1997)B. This success presents an additional water quality challenge, however.
Although no-till systems can reduce soil losses by over 80%, mean annual surface
runoff from no-till on claypan soils is about 18% higher than from other conservation
tillage systems that disturb the entire soil surface ("personal communication", Dr.
Eugene Alberts). Higher surface runoff with no-till, coupled with the surface application
of herbicides, provide increased opportunities for herbicide movement from farm fields.
Surface Water Quality: The most important water quality problem in Missouri is
herbicide contamination of streams, primarily during the critical 45-60 day period
following chemical application in the spring. The restrictive layer in many of our soils
causes nearly 30% of the total precipitation to be lost as runoff. Excessive runoff
usually occurs during the spring when soils become saturated. Of 140 surface water
sites monitored throughout northern Missouri in 1994 and 1995; 55% and 42%,
respectively, exceeded 3 ppb atrazine in the spring. Sampling was less extensive but
much more intensive during 1996 and 1997. During 1996, sampling efforts were
focused on the outlets of 20 major drainage basins which were outfitted with USGS
flow gages. These sites were monitored 3-7 times during May - early July, with all
exceeding 3 ppb until mid May. During 1997, 50 sites were monitored throughout May
and June; half on a weekly basis, and the other half (those without flow gages) on a bi-
weekly basis.
Atrazine concentrations in surface waters still need to come down. Preliminary data
from 1997 show that 85% of these sites exceeded 3 ppb atrazine, 50% exceeded 20
ppb atrazine, over 30% exceeded 30 ppb, and 14 % exceeded 50 ppb. Our sampling
over the last four years also reveals that herbicide concentrations in streamflow are not
related to the land use patterns of a particular cropping region, but are more related to
the hydrologic characteristics of the land resource area (Blanchard, 1995 and 1997).
This finding significantly changes how policy makers and regulators must look at the
surface water contamination problem.
Implementation of Best Management Practices and Alternatives for Atrazine: Label
changes, along with extensive educational efforts, aimed at persuading producers to
voluntarily reduce their reliance on this herbicide, have succeeded in decreasing the
average use rate from 1.48 Ibs/A in 1992 to 1.29 Ibs/A in 1995 (Missouri Farm Facts,
1996). However, atrazine use is still widespread (82% of com acres treated in 1995),
and further reductions are necessary. Efforts to expand the voluntary adoption of
BMPs which protect surface water quality have been hampered by numerous
economic, social, and educational barriers. Identification of these barriers, however, is
an important step in overcoming them.
Economic barriers include the increased costs associated with the use of alternative
herbicides and the costs associated with creating and maintaining filter/buffer strips and
riparian zones. Currently available alternative herbicides require producers and custom
applicators to depart from the convenience of prophylactic, broad spectrum,
preemerge, one-pass applications. Most of the low or non-atrazine alternatives are
11
-------
significantly more expensive; include a combination of herbicides; require additional
applications; and require increased management, due to the narrower spectrum of
weeds they control. Many also have a smaller window of application, decreased crop
safety, and increased potential liability for both producers and custom applicators.
Other BMPs , which encourage applicators to delay applications if fields are saturated
or rains are expected, could limit the number of acres they can handle with their current
equipment and labor (Johnson, 1996).
F
Public perceptions represent additional barriers to adoption. In 1994, 10 of Missouri's (
public water supplies were out of compliance with atrazine levels exceeding the MCL
(maximum contaminant level) of 3 ppb for drinking water. But, with all but one of *
Missouri's public water supplies in compliance throughout 1995, 1996, and 1997, the
level of concern has dropped. Extensive educational efforts will be required to inform
communities regarding herbicide levels in raw water, how these levels fluctuate
throughout the season, the inability of current filtration systems to remove pesticides
above certain levels, and the impact that timing of quarterly public water sampling can
have on whether or not they remain in compliance.
I
Overcoming these barriers will require continued educational and demonstration efforts (
to increase environmental awareness and technical expertise. However, the economic
trade-off between production and environmental goals must be recognized, quantified,
and addressed. Detailed economic analyses of the various options will help determine
the type and amount of incentives required to speed the voluntary adoption of those
BMPs with the greatest potential for improving surface water quality. Efforts are j
underway to conduct these analyses. Identification of areas most in need of |
implementing BMPs will help in the allocation of limited resources. i
i
Documentation of Basin and Regional Herbicide Losses for Northern Missouri and
Southern Iowa Streams: The percentage of applied atrazine lost in a watershed can ;
be determined by monitoring both the atrazine concentration in the river during the j
critical period following application and the volume of streamflow. It is also necessary
to know the number of acres cropped to com and milo, the percentage of acreage
receiving atrazine, and the average atrazine use rate within the basin. Blanchard and !
Lerch (1997) compiled these figures for 1996 to determine an estimate of the total |
atrazine applied in each tributary of the Missouri and Mississippi Rivers for northern ;
Missouri and southern Iowa. Using the atrazine concentrations, daily streamflow, and i
estimated atrazine usage in each river basin, the percent of atrazine lost from each
basin was estimated (Figure 2). Atrazine transport from twelve of the largest northern >
Missouri / southern Iowa river basins comprised 25% of the atrazine discharge in the j
Mississippi River at Thebes, IL; but represented only 2.6% of the drainage area and
11 % of the streamflow. Preliminary data from 1977 show a similar pattern.
Precision Farming: Precision farming technology for automated measurement of soil, l
site, and yield variability within fields has been evaluated. Strategies to tailor nitrogen
fertilizer application to match productivity variations have been developed which show
that nitrogen fertilizer applications within a given field can be reduced by 5 to 15% with
variable-rate field applications (Kitchen, 1995 and 1996). This could translate into an
annual reduction of 14 to 40 million Ibs of nitrogen fertilizer for Missouri's 2-million com
i
12
-------
acres, and reduce leaching of soil nitrate to groundwater. Several research studies to
evaluate the impact of precision farming on herbicide concentrations and losses in
surface runoff are being initiated in 1997.
Estimated % Loss of
Atrazine
May-June 1996
Figure 2. The percent of applied atrazine lost in each drainage basin during May-June 1996.
• Ground Water Quality: Herbicide contamination of ground water is not a major
problem on claypan soils. Of the more than 1,000 ground water samples analyzed
between 1991 and 1994, atrazine was detected in only 66 samples, with concentrations
never exceeding 0.12 ppb. Alachlor was detected in only two samples at concen-
trations near the limit of detection (Blanchard, in press). Nitrate-N contamination,
however, appears to be widespread; occurring primarily during the fall and winter. Of
the more than 1,000 samples collected from 96 wells, about 25% of the samples
exceeded the 10 ppm nitrate-N drinking water standard. In some cases, elevated
ground water nitrates are associated with poor nitrogen management on cropped fields
(Kitchen, 1997). Farming systems and technologies need to be developed to improve
nitrogen crop-use efficiency and to minimize nitrate leaching into ground water.
Lessons Learned
Implementing solutions to water and soil quality problems require clear two-way communication
between researchers and end users. While greater awareness regarding water quality issues
exists today than a few years ago, adoption lags behind knowledge and technology. Adoption
or modification of farming practices generally does not occur as a consequence of any one
specific educational program. Change generally occurs in small increments over a period of
time. Effective change must also include voluntary solutions to these important environmental
issues. It is therefore essential to continue and further develop these ongoing educational
programs that not only provide relevant, up-to-date, research-based information on cropping
practices and systems that will protect our water resources; but that involve research,
extension, government agencies, industry, and the affected communities in a true partnership.
13
-------
References
1. Blanchard, P.E. 1997. Soils, hydrology, and land-use: What controls water quality
degradation? In Journal of Soil and Water Conservation 52(4):283.
2. Blanchard, P.E., and W.W. Donald. Herbicide contamination of ground water beneath
claypan soils in North-Central Missouri. Journal of Environmental Quality, (in press)
3. Blanchard, P.E., and R.N. Lerch. 1995. Regional-scale factors affecting herbicide
contamination of northern Missouri streams. P. 72-76. In Proc. 5th Annual Missouri Water
Quality Conference, Columbia, MO, Feb. 2, 1995. Univ. Of MO, Columbia, MO.
4. Conservation Tillage Information Center (CTIC), 1997, Internet Information:
http//www.ctic.purdue.edu.
5. Johnson, B., F. Fishel, and A. Kendig. 1996. Atrazine: Best management practices and
alternatives for Missouri, Univ. of MO Ag Guide Sheet G4851, Univ. of Missouri, Columbia,
MO.
6. Kitchen, N.R., P.E. Blanchard, D,F. Hughes, and R.N. Lerch. 1997. Farming system
impact on groundwater nitrate underlying claypan soil. J. Soil and Water Conservation.
52(4):272-277.
7. Kitchen, N.R., K.A. Sudduth, S.J. Birrell, and S.C. Borgelt. 1996. Missouri precision
agriculture research and education. P. 1091-1100. In Proc. 3rd Int'l. Conf. on Precision
Agriculture, June 23-26, 1996, Minn., MN. ASA, CSSA, and SSSA, Madison, Wl.
8. Kitchen, N.R., K.A. Sudduth, D.F. Hughes, and S.J. Birrell. 1995. Comparison of variable
rate to single rate nitrogen fertilizer application: corn production and residual soil NO3-N. p.
427-442. In Proc. 2nd International Conf. on Site-Specific Management for Agricultural
Systems, Minn., MN, Mar. 27-30,1994. Am. Soc. of Agronomy, Madison, Wl.
9. Missouri Department of Agriculture. 1996. Missouri Farm Facts: 1996. Agricultural
Statistics Service, Columbia, MO, 76 p.
14
-------
Lake Pittsfield Watershed Project
- A Cooperative Effort
Richard J. Mollahan
Illinois Environmental Protection Agency, Springfield, Illinois
The objective of this project is to create of cooperative partnerships in the Blue Creek watershed
(watershed) to reduce the sediment delivery to Lake Pittsfield and improve water quality through
the implementation of upland cultural, and mechanical best management practices (BMPs).
Lake Pittsfield was constructed in 1961 as a flood control structure and public water supply for the
City of Pittsfield located in west central Illinois. Watershed land uses that drain into Lake Pittsfield
are dominated by row crop agriculture with small livestock operations (hog production lots)
scattered throughout the watershed. Local concern about protecting the watershed's water
resources was the catalyst to start the interest in development of a Lake Pittsfield Watershed
Project (LPWP). A series of studies to determine the nonpoint source pollution impacts in Lake
Pittsfield and cultivation practices within the watershed was the
focus of the project's early years (1979-1992). As the result of
the studies, sedimentation was determined to be the major
water quality problem in Lake Pittsfield. The next phase of the
LPWP (1992) was the construction of 29 water and sediment
control basins (WASCOBs) and a single large sediment
retention basin (SRB) on Blue Creek just prior to the in-flow of
Lake Pittsfield.
In the Section 319 National Monitoring Program Project (NMPP)
land use, land cover and soils determinations generated by the
Geographic Information System (GIS) mapping (Figure 1) were
used to determine NPS pollution impacts. The GIS is a tool to
interpret and then illustrate the data collection, monitoring and
BMPs implemented within the watershed and in Lake Pittsfield
and its direct ravine tributaries.
INTRODUCTION
1979 Land Cover
Ljhc PIllifMJ Wvtcfihcd Monftofina Protect
Figure 1
Since the construction of Lake Pittsfield in 1961, sedimentation originating from the 7,000-acre
watershed has severely reduced the original capacity of the lake. Sediment from farming
operations, direct tributaries, and shoreline erosion has decreased the surface area from 262 acres
to 222 in the last 33 years. By 1979, 25 percent of Lake Pittsfield's original water storage capacity
was lost due to sedimentation. Implementation of agricultural BMPs (no-till conservation
tillage/terraces/grass waterways, etc.) Occurred in the late 1970's, through the Agricultural
Stabilization and Conservation Service (ASCS) utilizing funding under the United States
Department of Agriculture's (USDA) Water Quality Incentive Program (WQIP).
15
-------
The Lake Pittsfield National Monitoring Program (LPNMP) began in 1992 as a joint responsibility
of the Illinois State Water Survey (ISWS), United States Environmental Protection Agency (U.S.
EPA), and the Illinois Environmental Protection Agency (Illinois EPA). The goal of the LPNMP is
to monitor and evaluate the effectiveness of the installed BMPs and implemented land use
cultivation practices (no-till conservation, etc.).
The Pike County Soil and Water Conservation District proposed a series of WASCOBs in the
watershed and a SRB constructed on Blue Creek just prior to the inflow of Lake Pittsfield. The
funding for the construction of the WSCOBs and the SRB, was obtained through Section 319
Nonpoint Source Pollution Control Program of the Clean Water Act in 1992. Twenty-nine WSCOBs
were constructed by the fall of 1995 in the watershed, along with the SRB in the summer of 1996.
DESCRIPTION
Lake Pittsfield sedimentation studies were conducted during the years 1974, 1979, 1985 and in
1992 to access the condition of the lake prior to the proposed WASCOBs and SRB
implementations. The purpose of a lake sedimentation survey was to determine the following
conditions: 1) the present lake volume, 2) the volume and mass of sediment deposited in a lake,
3) the stage-volume relationship within the lake, and, 4) if there have been prior sedimentation
surveys to document the changes in the sedimentation rate over time.
WQIP funding focused on the adoptions of BMPs and cultural practices that will reduce the
transportation of sediment, fertilizer, and pesticides within the watershed. These BMPs and cultural
practices such as conservation tillage, integrated crop management, livestock exclusions, and
wildlife habitat management, are being implemented in the watershed.
1993 Land Cover
Uto PtnsfMd WMwclwd Monhorina Protect
The land uses within the watershed consist of eight
different designations including urban, water, public land,
farmsteads, shrub-brush, deciduous forest, cropland
and pasture/rangeland (Figure 2). The watershed is
roughly divided in half between steep woodlands and
pasture lands to the southern and western regions and
rows crops on flat prairie soils in the northern and
eastern regions.
The primary BMPs implementation in the watershed was
the construction of the 29 WASCOBs. These
WASCOBs have been constructed below some 2,482
acres, or 36 percent of the total watershed size. The
construction of the large SRB (423 acre ft) was in the fall
of 1996. The SRB is predicted to trap 83 percent of
sediment delivery to the watershed prior to the flood
water's discharge into Lake Pittsfield.
Figure 2
16
-------
Locations of the WSCOBs are strategically placed
throughout the watershed (Figure 3). The majority of the
basins are located in the small tributaries of Blue Creek
with other basins located on a western region of the
watershed in the direct tributaries. The WASCOB's sizes
range from 49.75 acre-ft to 5.02 acre-ft. The mean size of
the basins is 24.29 acre-ft. Construction of the SRB (423
acre-ft area) was completed in the summer of 1996.
WATERSHED STREAM MONITORING
The stream monitoring strategy was to establish a series
of stream sampling and designed flow gaging stations
within the watershed. In the fall of 1992, the sampling
network was created to collected data in the determined
five subbasins (Figure 4). The network consists of
automatic samplers and a Doppler flow meter at station B
(Lake Pittsfield) to measure stream flow during lake
backwater episodes. Five automatic samplers were
installed at five separate sampling station locations within
the watershed. These locations included the tributaries of
Blue Creek, the main channel of Blue Creek, and a large
direct tributary to Lake Pittsfield located in the
southwestern area of the watershed.
Retention Basins
Lake PhtafleU Watershed Mooltorino Protect
Figure 3
Monitoring Network
-D
Lake Pittsfiald Watershed Monhorina Project
Figure 4
Watershed yields of a sediment baseline were
formulated by an intensive sampling schedule during
significant flood events during 1993. Normal stream
flows were sampled on a biweekly basis to determine
the base flow. Stream sampling was intensified during
the Spring sampling seasons when the flows'
measurements at chosen stream monitoring stations
were taken every 3.5 days in accordance with the U.S.
EPA National Monitoring Program. All stream samples
follow U.S. EPA methodology and flow measurements
are performed in accordance with USGS procedures.
MODELING/APPLIED TECHNOLOGIES
Geographic Information System (GIS) is an organized
collection of computer hardware, software, geographic
data and personnel designed to efficiently capture,
store, updated, manipulate, analyze and display all
forms of geographically referenced information.
GIS is being used in the watershed to spatially
17
-------
characterize many of the physical and hydrologic features. GIS has made it possible to construct
a high resolution digital model of the land uses/land covers, the hypsography or digital elevation
model, hydrology, soils and other physical features needed to properly define a watershed for
environmental modeling and assessment. The GIS databases are being used in conjunction with
the Agricultural Nonpoint Source Pollution (AGNPS) model to assess the factors affecting the lake's
water quality and to characterize physical and environmental activities and their interaction within
the watershed. A survey of the extensive ravine stream channel network was conducted in 1978
and was scheduled to be repeated this spring. GIS will be used to develop three-dimensional
models of the ravine network which then will be integrated so that physical changes over the two
decades can be assessed.
A digital elevation model (DEM) is the digital form of the hypsography in which the data was used
to determine the aerial extent of the 29 WSCOBs. With the use of the GIS, the WASCOBs
subwatersheds can be characterized for their soil types, slope, slope lengths, land use/land cover
and farming practices. Such a determination will allow for an improved assessment of the
WSCOB's expected life and for a more comprehensive management of the overall watershed
program.
The use of GIS in conjunction with the development of spatial digital databases (SDD), can be
utilized for accurate assessments of the stream profiles, streams slope, and to determine
areas/shapes of an infinite number of watersheds along a stream reach. These various physical
analyses with the integration of the SDD, are making it possible for improvement of the needed
assessments/analyses to support management decisions.
RESULTS
The 1992 study, determined that the lake volume was 2,679 acre-feet, which represented a total
lake volume loss of 24.8 percent or an average annual volume loss of 0.79 percent since 1961.
This compares to the average annual volume loss rate found in the previous studies which varied
from a high of 1.19 percent (1974-1979) to a low of 0.32 percent (1985-1992) (Table 1).
Table 1 . Comparison of Percent Volume Loss
lor Different Survey Periods
Survey Period
1961-1974
1974-1979
1979-1985
1985-1992
1961-1992
Lake Volume Loss
Years (Acre-feet)
13.5 494
4.8 204
6.0 105
7.2 81
31.5 884
Lake Volume Loss
(Percent)
13.9
5.7
2.9
2.3
24.8
Average Annual
Volume Loss
(Percent)
1.03
1.19
0.49
0.32
0.79
18
-------
The study also concluded that
884 acre-ft of sediment has
accumulated since 1961.
This represents a total of
904,800 tons of sediment
deposited from 1961 to 1992,
for an average annual
deposition of 28,700 tons
(Table 2). This represents an
average annual sediment
yield per acre of watershed of
4.2 tons. The average annual
sediment yield per acre of
watershed has varied in each
sedimentation study
conducted. This variation
ranged from 5.9 tons (1961-
1974) to a low of 0.7 tons (1985-1992).
Table 2. Comparison of Sedimentation Rates (in tons) for Different Survey Periods
i Sediment Deposition
Per Acre of Watershed Area (Tons')
Sediment Deposition (Tons')
Survey Period
1961-1974
1974-1979
1979-1985
1985-1992
1961-1992
Total
545,300
190,100
135,600
33,800
904,800
Average Annual
40,400
39,600
22,600
4,700
28,700
Average for
Survey Period
79.3
27.7
19.7
4.9
131.6
Average
Annual
5.9
5.8
3.3
0.7
4.2
Table 3. Spring Flood Sediment Yields
Lake Pittsfield Watershed - 1993-1996
1993 7 storms
Floodwater Discharge (ac-ft)
Average Sediment Cone. (mg/L)
Sediment Yield (tons/acre)
1994 4 storms
Floodwater Discharge (ac-ft)
Average Sediment Cone. (mg/L)
Sediment Yield (tons/acre)
1995 7 storms
Floodwater Discharge (ac-ft)
Average Sediment Cone. (mg/L)
Sediment Yield (tons/acre)
1996 7 storms
Floodwaler Discharge (ac-ft)
Average Sediment Cone. (mg/L)
D Subbasin
1756 acres
509
2,870
1.1
259
4,485
0.9
400
2,878
0.9
298
7,873
C Subbasin
1567 acres
379
7,341
2.4
410
8,299
2.9
368
4,632
1.5
360
6,696
Results of stream monitoring over a four-year
time period (1993-1996) revealed that the
concentrations of sediment for the first three
years were consistently lower (approximately
by half) from the watershed's (eastern,
Subbasin D) flat former prairie uplands then
that of the ( western, Subbasin C) steeper
pasture and woodlands. The sediment
trapping effectiveness of the WASCOBs
appears to be influenced by the locations of
older established ponds upstream. The
WASCOBs in the steeper upland woodland
and pasture lands in the watershed's
southern and western regions appeared to
be more effective in reducing sediment
concentrations. During the recorded 1993-
1996 "Spring Flood Sediment Yields" (SFSY)
revealed an increased sediment in 1996,
which was not consist with the previous
year's reduction after the construction of the
WASCOBs. There is limited evidence that
the increase in the 1996 SFSY may be a
result of channel streambank instability on
Blue Creek in the middle portion of the
stream reach (Table 3).
19
-------
DISCUSSION
The stream monitoring data collected (1993-1996) to fulfil the National Monitoring Program
requirements is still being reviewed to determine BMPs effectiveness. The increase sediment
delivery rate detected in 1996, may just be a temporary adjustment of the Blue Creek's stream
channel to the new hydrologic regime imposted by the WASCOBs. Additional monitoring will
determine the extent of any stream channel adjustments. It is hoped that the landowners with
WASCOBs located on their land will be able to see the effects of their own farming operations on
their basins and give more consideration to conservation farming practices on other areas of their
land. Proposed plans are to implement six loose stone weirs on Blue Creek within Subbasin D.
These structures will serve to help reduce the rate and amount of streambank erosion through
grade control.
20
-------
The Farm / Field * A * Syst Decision Support Systems
Pierre C. Robert and J. L. Anderson
University of Minnesota, St. Paul, Minnesota
The Farm / Field* A*Syst Decision Support Systems (DSS) are user friendly software designed
to help assess farm practice impacts (cropland, pastureland, woodland, farmstead,
wetland/waterbody, and other resources) on water quality. Pollution potential is evaluated by
identifying high risk categories for watersheds or projects. Plan of action to address problems
found in farmsteads and fields while preserving data confidentiality is provided.
Results of field tests indicate that the DSS have a high potential to reduce water pollution
because they provide a precise list of pollution risks for each farm and recommendations to
correct problems. This capability to show and print simple action recommendations for
identified high risk categories accelerates the assessment process, and increases farmer
participation and implementation.
The Farmstead Assessment System (Farm * A * Syst)
The Farmstead Assessment System (Farm * A * Syst) supports voluntary rural water pollution
prevention programs. The program provides unique pollution risk assessment tools for
farmsteads and a flexible program implementation framework that has been successful in
building interagency and private sector partnerships to support rural pollution prevention efforts.
(Jackson, 1995).
The main goals of Farm * A * Syst are to :
• create awareness of farm activities and structures that may cause drinking water
contamination and other environmental problems;
• promote an understanding of pollution prevention and clean-up actions;
• identify sources of technical, educational and financial assistance;
• aid in developing personal, voluntary action plans.
This is accomplished by integrating complex national and state environmental policies and
programs of numerous agencies into an applied decision-making framework for farmers,
ranchers, and rural residents. The framework uses a series of step-by-step worksheets that
evaluate rural activities and structures posing risks to groundwater. Information on available
financial, technical and educational assistance is also provided through fact sheets developed to
coincide with each worksheet. Farmers and rural residents can use Farm * A * Syst on their
own, or in consultation with local experts. The system is designed as a confidential service to
concerned farmers and rural residents.
21
-------
The Farm * A * Syst Decision Support System (FADSS)
Description
The Farmstead Assessment Decision Support System (FADSS ) is a computer software
developed to facilitate the use of the Farm * A Syst worksheets, provide action recommendations
to solve problems associated with current farmstead activities, and help identify important issues
by watershed or project, so educational programs or efforts can be directed to priority areas.
Major components of the software are:
• user-friendly data entry windows to record farmstead rankings for all assessment
worksheets and categories;
• computation of average risk rankings for each worksheet and computation of overall
farmstead risk rankings;
• display all high ratings from individual farmstead assessments and action recommendations
to reduce high risk factors;
• capability to modify action recommendations for high-risk factors by authorized
information managers;
• search capability by sheet categories and risk rankings, computation of risk factor
frequency, and display of summary statistics;
• display of fact sheet information;
• printing of action recommendations and summary statistic reports;
• Microsoft Windows based system.
A more detailed description of FADSS is available in Robert and Anderson, 1995.
Implementation and evaluation
The DSS was used by the Stearns County Soil Water and Conservation District, Minnesota, the
University of Wisconsin Cooperative Extension of Buffalo county, Waupaca county, and UW
-Platteville Extension. It was used on about a total of 100 farms. Visits to each farm were
conducted by a trained conservationist or agent. The DSS was also sent upon request to 31 states
and 3 Canadian provinces for evaluation.
Benefits and lessons learned
Principal benefits of FADSS reported in the four pilot evaluations are:
• a strong capability to reduce water pollution because the system automatically provides a
list of high risks for each farmstead and precise recommendations to correct the problems.
This seems to have increased farmer participation. This is also important for educators or
others assisting in the evaluations because it reduces the follow up time.
• minimizes the time needed to complete an assessment, a factor very important to farmers
and educators. FADDS reduces the time needed for assessment from 2-3 hours to
approximately 1 hour.
• very user friendly.
• favors greater consistency of assessments between farms.
• stores all the data in a database while protecting the confidentiality of assessments.
22
-------
• provides to project managers summary tables of risks and indicates most frequent high risk
categories. For example, in southwestern Wisconsin (UW -Platteville Extension), FADSS
was used to conduct farmstead assessments on 17 farms. FADSS analysis tools
immediately identified that high risks of water pollution were principally in two categories'
Petroleum product storage' (32 cases), followed at a significantly lower frequency by'
Household waste water treatment' (14),' Livestock yard management' (12), and ' milking
center waste water treatment '(11). The ' High Risk Table " indicated that within each of
these types of risks, there were 1 or 2 categories with greater risks. For example, in the
case of petroleum product storage, high risks most often identified were in the categories '
Spill and tank overfill protection ', and' Tank enclosure'.
FADSS was preferred over the printed worksheet and fact sheet system but it does require a
computer notebook with a good color screen. Agents that used it with farmers have
recommended the following improvements:
• capability to display more graphs and print them in color
• more help for the most technical categories
• more explanation or helps indicating why some questions are important and how they relate
to pollution prevention.
• prepare a video tape as a complement to the user's guide to make the learning of the
software easier.
The Field * A * Syst Decision Support System (FIADSS )
The Field * A * Syst Decision support System (FIADSS) is a complement to Farm * A * Syst,
which addresses the management of a whole farm. It is based on a comprehensive and whole
farm planning tool - The Whitewater Whole Farm Planning Manual - developed by USDA-
NRCS with a team of representatives from the farming community, agencies, commodity groups,
and environmental groups, in southeastern Minnesota.
The Basic Conservation Planning - Whole Farm Planning
The goal of the group was to assemble materials that expand environmental awareness, take into
account the interactions between many land uses, and help make decisions on the best possible
management practices for farm profitability and environmental protection.
The objective of the system is to provide a method to develop a farm plan that
covers all farm natural resource needs while meeting financial and personal goals.
The process is voluntary and confidential. It is designed to be flexible to meet a wide variety of
needs and planning decisions. The planning process is driven by the user to meet personal needs
and provides guidance needed to manage all resources. Depending on individual goals, the plan,
or parts of the plan, can be shared with others. There are numbers of agencies, private
individuals, and groups that can assist the producer.
Whole farm planning materials are organized in a logical and methodical way to help in the
planning process and , perhaps, look at farm management in a new way. This may take a while
to complete the first time. But once done, the farm plan is easily maintained and modified as
needed. Worksheets are organized in distinct sections that should be completed in sequence to
get the best benefit from the plan. The sections are:
23
-------
1. Goal setting: statements of farm management goals within a defined time frame to help focus
the planning and the implementation.
2. Inventory: listing of all resources to define current condition, identify basic lacks of certain
resources, and to later assess maintenance, improvement, or loss.
3. Evaluation: help define the current state of resources and management. It attempts to identify
all resource concerns that may occur on farms in accordance to local standards of resource
protection. Worksheets are divided into distinct land uses: cropland, pastureland, woodland,
wetland/waterbody, and farmstead. Resource concerns are divided into a four tier evaluation
process similar to Farm * A * Syst:
• Rank 4: low potential impact on productivity and probably enhances production.
Resource enhancement.
• Rank 3: low-moderate impact on productivity and probably maintain productivity.
Resource maintenance.
• Rank 2: moderate-high impact on productivity and may maintain or reduce productivity.
Resource degradation with some controls.
• Rank 1: high impact on productivity. Productivity will be reduced.
Uncontrolled resource degradation.
4. Alternatives: investigation of alternative actions - what and how - to meet selected needs.
Some actions may be evident and others may require more detailed analysis to ensure that they
meet financial goals. Guide sheets demonstrate the possible combinations of practices and their
interactions.
5. Selection and implementation: selection of steps, actions, and timeline to implement the
selected alternative. Blank forms are designed to help the process. They include:
• Action item
• Land use, resource concern
• Goal for resource concern
• Action items needed to achieve goal, priority, responsibility, deadline to complete, and date
completed.
6. Follow up: study of impacts of actions and selection of some changes when needed
Description of the Field * A* Syst Decision System
The FIADSS software is intentionally very similar to FADSS. The primary goal is to
complement and broaden the objectives of FADSS. The look of the software - dialog, data entry,
and data processing windows - is also intentionally similar to facilitate its use. Only will the key
features be presented, particularly features different from FADSS.
The basic functions of FIADSS are:
• data input
• file and database management
• output of selected action recommendations
• query
• multilevel helps
The principal content additions relate to goal setting, evaluation work sheets, aids in selecting
and implementing actions, and notepad for follow-up.
24
-------
^ Ranking Description 1
Choose a field by Weriuneet 2 : Dopland Re*ouiee
Concern -Page2
cficking a button Category 1 : Nutrients: Saawang Density
®i Oa
Entet you tanking by clicking in one of the buttons:
O2 OlO
O3 Oil
LOW RISK LOW-MOO RISK
O tank 4 Q help O rank 3
O4 O12
O5 O13
OB O14
O7 Ol5
Oa OIB
Fi*UU__-
field SI
•*•
*
Sajaping include*
• nanai IBMIH nl ai«a
fPVCWfC **^*P""*flr
or at least 8 soil
CQfGB eMO COHoClBu
ho* S aaas 01 less
to Ion a coapoiite
soi seiBpkt and at
least 2 coasoowte
soi
lanplesafe
collectod fioai each
field
•ice.
»
—
*
a
MOD-HIGH RISK
help ©Itanir?! Q
At least *oi cores
• •A finllATlMl flflM ^
to 10 acre* lo tone a
and at least 2 soi
sanplesaie
_ _||_j,|— j| -UM f«^f
«•
»
HIBH RISK
help O rank 1
At least 5 soi cote*
10 lo 20 acres to
letei a compn«tte
saBpkL
PI looinola PI loatnoim
Enable Modify
Save
tie*
Olo
•«
jr
—
[7
a >«.,
A tingU toil *«mpto
f-nlljU-jlawl f|-fh.m 3(1 AM
note actos.
otnota
?H«p
^
—
a
Q footnote
Xc«n«l
Figure 1. Dialog window for selecting a field and a risk level, and input of a field local
name
JO »1 : Cropland - Paqe2 (PID: Soil «1. FID: U of M| 1
F
Nutrients
8.1. Sampling Density
C.1. FeitiliMf Rates (N|
C.2.Feirikzef Rates
(Plandflq
C3.Micro Nutrients
D.I. Tia>o«Placement
D.Z (Contl
0.3. (Contl
E.I. Dedit*
E.2. (Contl
F.I.Manuie Application
Avg. Rank lot aach field:
RE
aai
R
H
R
R
H
R
R
••
*
&
•*•
R
it: 1 2 34 5 6 7 8 9 10 11 12 13 14 15 16 1
Rank lot each field
i
i
i
i
i
i
i
5ULT
3
3
3
~
1
3 !
2
1
1
3
2
3
3
=
2
1
3
2
1
1
3
II
2
3
3
2
1
3
2
1
1
3
2
3
3
=
2
1
3
2
1
1
3
2
3
3
2
1
3
2
1
1
3
2
3
3
2
1
3
2
1
1
3
2
3
3
2
1
3
2
1
1
3
2
3
3
2
1
3
2
2
3
3
2
1
3
2
1 1
1 1
3 3
Z
3
3
2~
1
3
2
1
1
3
2
3~
3
2
1
3
2
1
1
3
mono
ZJLZOLU
tt of categories (To"| Total {2521 RiskScote
tt tanked Ral Lavel
tabulate
He*
I/OK
? Hetp
Xc«ncel
2
^^
3
3
2
1
3
2
1
1
3
.
—
=
_
JLJLJLJLJ
—
=
a
Figure 2. Dialog window for risk level input, access to the ranking description dialog
window [ i button ], and to the action recommendation window [ R button]
25
-------
1. Goal setting. A simple form helps define farm goals and select appropriate actions.
2. Evaluation worksheets. New worksheets help evaluate on-farm resources and management
risks:
• cropland
• pastureland '
• woodland i
• farmstead I
• other resources: ground water, urban corridors, recreation I
• wetland / waterbody
Worksheets allow for an evaluation of up to 16 fields or parcels. When a farm has more than 16 j
parcels the farm record can be split into several records identified in the " General Information "
worksheet. Each field can receive, in addition to a code number made of the field ID and a
number 1 through 16, its usual name (e.g., Todd 80 N) from a dialog window in the ' Ranking
Description ' input window (Fig. 1). Each worksheet has several pages grouping into coherent i
dialog windows all types of categories (Fig. 2).
3. Alternatives. Guide sheets designed to assist in evaluating the effect of potential actions and
their impact on agroecosystems.
4. Action Plan. Field * A * Syst provides a list of action recommendations for identified high
risk conditions. They are developed regionally and can be eventually modified locally by
authorized program managers. They can be accessed at the time of evaluation or summarized
and printed by the system after the farm evaluation is finished. In addition, a form is provided to
help implement selected actions.
5. Follow-up. An electronic notepad created to record actions and follow-ups.
Hardware and software requirements
FADSS and FIADSS require a PC computer running Windows 3.1 or Windows 95. The system
should have a least 4 MB RAM and 1 MB of free hard disk free. More disk space is needed to
store farm and field data. A mouse or an equivalent pointer is required.
More information
For more information about the DSS, contact Dr. Pierre Robert. University of Minnesota,
Department of Soil, Water, and Climate, 1991 Upper Buford Circle, St. Paul, MN 55108.
Phone: (612) 625-3125. E-mail: probert@soils.umn.edu
Acknowledgments
The Authors thank M. Kunz and L. Svien, USDA-NRCS, Lewiston and Rochester, MN for their
contribution in the development of the Farm* A*Syst Decision Support System; and D. Fucks
and R. Richner, Soil and Water Conservation District, St. Cloud, MN; G. Blonde, UWEX,
Waupaca, WI; K. Gates, UWEX, Madison, WI; B. Kloster, Buffalo County LCD,WI for
providing field evaluations of FADSS.
26
-------
References
Jackson, G.W. 1996. Accomplishments in Farmstead Management: Building effective rural
pollution prevention partnerships through the Farm * A * Syst and Home * A * Syst programs.
Vol. IV, p. 69-71. In Proceedings of the Conference Clean Water - Clean Environment- 21 st
Century. March 5-8, 1995. Kansas City, MO. ASAE, St. Joseph, MO.
Robert, P.C., and J.L. Anderson. 1996. Farm * A * Syst Decision Support System. Vol. IE, p.
219-222. In Proceedings of the Conference Clean Water - Clean Environment- 21 st Century.
March 5-8, 1995. Kansas City, MO. ASAE, St. Joseph, MO.
27
-------
Utilizing Voluntary Farmstead Assessments to
Encourage Best Management Practice Adoption in the
Skaneateles Lake Watershed
A. Edward Staehr
Cornell Cooperative Extension of Onondaga County, Syracuse, NY
Skaneateles Lake Watershed Agricultural Program
This report provides an illustration of how a voluntary watershed protection
program was established through efforts of local farmers, the City of Syracuse,
Soil and Water Conservation Districts, Cornell Cooperative Extension
associations, the Natural Resources Conservation Service, New York State
Department of Health, and New York State Soil and Water Conservation
Committee. In addition, the report explains how farmstead assessments were
adapted to adequately address priority non point source pollution concerns and
serve as an educational tool for farmers.
Program staff developed materials for farm assessments from existing sources
such as Farm*A*Syst, the New York City Watershed Program, and Ontario
Environmental Plan. Farm assessments are merely one part of a
comprehensive planning process utilized to address water quality concerns in
the Skaneateles Lake Watershed.
A key component of all plans is that they are consistent with operators' business
plans. Moreover, plans cannot negatively impact farm profitability by increasing
Best Management Practices' operating and maintenance costs. The net
economic effect of all proposed practices on a farm business must be zero or
positive.
Introduction
Skaneateles Lake, located in the Finger Lakes region of Central New York, is the
primary source of unfiltered drinking water for the City of Syracuse and
surrounding communities. Watershed area measures 73 square miles, which is
small compared to other Finger Lakes. Approximately 200,000 consumers
purchase Skaneateles Lake water from the City of Syracuse.
The City of Syracuse was faced with either constructing a costly filtration plant,
or implementing a comprehensive watershed protection program covering all
land uses. Agriculture accounts for approximately 48 percent of total watershed
area and was the first land use to be examined by the City of Syracuse.
Watershed protection is not a new concept for the City of Syracuse. Since the
early 1900's, Syracuse has employed watershed inspectors to monitor the public
29
-------
water supply. An example of existing watershed protection efforts is an annual
pesticide survey that all watershed farmers fill out.
Legislation, namely the Surface Water Treatment Rule, was the impetus behind
commencing a watershed protection program. The Onondaga County Soil and
Water Conservation District gathered a group of watershed farmers together to
form an Ad Hoc Task Force and make recommendations to the City of Syracuse
for an agricultural watershed protection program. Task force members were
from three counties comprising the watershed, as well as the City of Syracuse
Watershed Control Coordinator. To guide members in formulating their
recommendations, a member of the New York State Soil and Water
Conservation Committee chaired the group. This chair was also the chairman of
the Onondaga County Soil and Water Conservation District.
In 1994, Syracuse Mayor Roy Bernard! endorsed the Ad Hoc Task Force
recommendations in a public signing ceremony. Watershed farmers were
pleased with the outcome, as they recommended a voluntary program fully
funded by the City of Syracuse.
Program Development
Guidance
Upon formalizing the task force recommendations, the Ad Hoc Task Force
assumed new duties and changed their name to the Watershed Agricultural
Program Review Committee. This group is responsible for providing guidance
to program staff, as well as reviewing Whole Farm Plans. Before program staff
came on board, the committee developed vision and mission statements to guide
the program.
The Skaneateles Lake Watershed vision as described by the group is: The
Skaneateles Lake Watershed will be an environmentally sound region, where a
viable agricultural industry and others benefiting from the lake work together
harmoniously to improve and maintain a high standard of water quality.
The program mission is: To carry out a cost effective, innovative program for the
fanning community that upholds the high drinking water quality standards of
Skaneateles Lake. This mission has served the program well, as external
funding sources, have been used to cover 81 percent of Best Management
Practices implementation costs. The USDA Farm Services Agency is one of
many outside sources used to fund implementation projects.
30
-------
Structure
Three existing agencies either assigned or hired staff to carry out various duties
in the Skaneateles Lake Watershed Agricultural Program. The Onondaga
County Soil and Water Conservation District is responsible for program
administration, nutrient management, plan design, and implementation. Cornell
Cooperative Extension of Onondaga County provides technical assistance for
pathogen management, pesticide risk evaluation, and farm business
management. The Natural Resources Conservation Service is responsible for
soil management, plan design, and implementation.
Implementation and Evaluation Approaches
Initial Meetings
To promote farmer enrollment in the Skaneateles Lake Watershed Agricultural
Program, staff held three informational meetings throughout the watershed.
Approximately 67 percent of eligible farmers attended at least one of the three
meetings and enrolled over 80 percent of all agricultural land in the protection
program. A key factor behind successful program enrollment was the effort of
the Watershed Agricultural Program Review Committee. This group encouraged
their peers to attend informational meetings and enroll their farms in the
program. Another factor was the teamwork between agency directors to
promote the protection program to eligible clientele.
Program staff gathered critical data with farmers regarding general
environmental conditions on their farms through a "Farmer Affirmation
Questionnaire" at each of the three initial meetings. This information indicated
specific assessment worksheets that farmers needed to fill out at a later date.
After staff interpreted questionnaires, they mailed each farm a "Farmer
Affirmation Report" that explained which additional worksheets farmers needed
to fill out and why.
Farm Assessments
There are 15 potential farm assessment worksheets that a farmer may have to
complete, based on needs identified by the initial questionnaire. The worksheet
titles are as follows: 1.) Pathogens, 2.) Manure: Field Application and Storage,
3.) Stream Management, 4.) Milking Center Wash Water, 5.) Silage Storage,
6.) Petroleum Product Storage, 7.) Fertilizer Management, 8.) Pesticide Use,
9.) Barnyards, 10.) Water Well Evaluation, 11.) Waste Disposal, 12.) Soil
Management, 13.) Pasture Management, 14.) Forest Management, and
15.) Neighbor Relations. An interdisciplinary group designed the worksheets to
address concerns in the Skaneateles Lake Watershed. Since the City of
31
-------
Syracuse does not filter Skaneateles Lake water, Giardia and Cryptosporidium
are the primary concern and many worksheets concentrate on these pathogens.
Farmers and program staff completed farm assessments together to ensure
accurate results. Time needed to conduct these assessments ranged from a half
hour to two hours, based on complexity of operation and size of business. A
visual inspection of all areas of concern gave program staff an opportunity to
document positive practices, as well as areas needing improvement.
Following farm visits, team members from Cornell Cooperative Extension, Soil
and Water Conservation Districts, and the Natural Resources Conservation
Service generated a "Worksheet Report". In this report the team outlined
possible Best Management Practices to consider that alleviate potential
concerns, as well as positive actions the cooperator may already have in place to
reduce non point source pollution.
Environmental Benefits Measured
Cropping Changes
Most program participants had soil conservation plans on file, as required by the
Food Security Act, to reduce erosion. However, additional reductions in soil loss
were needed to reduce sedimentation concerns in the watershed. Nine farms
adopted different crop rotations to reduce soil loss. One farm chose to
implement strip cropping, as well as contour farming, while maintaining his
existing rotations. All farms appreciated being presented with alternatives in the
"Worksheet Report". This gave them time to consider other options prior to
working with program staff in developing their Whole Farm Plan.
To measure environmental benefits, the Natural Resources Conservation
Service Resource Conservation Specialist utilized the Food Security Act
Alternatives model to estimate soil loss. Implementing practices such as strip
cropping, contour farming, and alternative crop rotations produced admirable
results. These practices reduced annual erosion by 1,369 tons.
Nutrient Management
Cornell University soil test results are the basis for all nutrient management plan
recommendations. The Skaneateles Lake Watershed Agricultural Program
provides free soil testing for one test recommendation, which is good for three
years. Program participants are responsible for taking their own soil samples
after three years.
32
-------
All nutrient management plans account for livestock manure, as well as crop
rotations. Dairy and livestock farmers reduced the amount of purchased
nutrients the most, while crop farmers altered their commercial nutrients the
least. Savings ranged from $180 to $1,911. Average savings per farm
amounted to $1,180 annually.
To measure fertilizer savings, the author utilized partial budgets. This method is
used to examine specific areas of a farm business and determines the
profitability of making changes in one's operation, measured in net farm income.
A key component in calculating profitability involves utilizing manure nutrient
analysis to place an economic value on manure. By having a quantifiable
economic analysis available to illustrate savings from nutrient management,
program participants are more likely to continue soil testing, owing to a positive
impact on their farms' profitability.
Lessons Learned
Conducting Farm Assessments
Each step in the planning process offers unique educational opportunities.
Utilizing a questionnaire to determine appropriate farm assessment worksheets
provides farmers with an illustration of why certain areas of their business need
closer examination. This is best achieved when a written report accompanies
applicable assessment worksheets.
Another educational opportunity occurs when program staff work with a farmer to
fill out his or her farm assessment worksheets, referred to in the Skaneateles
Lake Watershed Agricultural Program as "Tier II Worksheets". While staff
interact with farmers by visually examining farming practices, a two way
exchange of ideas and suggestions is facilitated. This visual examination helps
reinforce to program participants that no environmental concerns will be reported
to regulatory agencies.
Merely conducting environmental assessments does not ensure that operators
will change their management practices or be receptive to discussing
improvements. However, an environmental performance report can augment
assessments by explaining how farm environmental conditions may impact water
quality. A report should document positive environmental practices that farmers
already have in place, as well as areas needing improvement.
Program Delivery
Agency cooperation is paramount to program success. Staff from the Soil and
Water Conservation Districts, Cornell Cooperative Extension, and the Natural
33
-------
Resources Conservation Service jointly drafted a letter to all watershed farmers
endorsing a voluntary watershed protection program. A farmer member of the
Watershed Agricultural Program Review Committee was also involved in the
program endorsement letter.
All agencies involved should define their roles and describe how they can best
contribute to a program's overall success. Teamwork can build upon the
strengths of each agency. For example, Cornell Cooperative Extension has
research based information on a multitude of areas critical to the Skaneateles
Lake Watershed Agricultural Program including nutrient and pathogen
management, in addition to pesticide data. Soil and Water Conservation
Districts have decades of results in encouraging and assisting producers in
adopting conservation practices, as well as securing resources for them to do so.
The Natural Resources Conservation Service has extensive experience in
providing technical assistance in Best Management Practice design and
implementation.
Conclusion
Results obtained in the Skaneateles Lake Watershed Agricultural Program can
be replicated elsewhere if certain factors are in place. In fact, the program is a
model for New York State's Agricultural Management initiative. Farmer
endorsement and promotion are necessary to facilitate program sign-up. A
voluntary watershed protection program overseen by a farmer review committee
was a key selling point to encourage farmer participation. Moreover, farmer
input is crucial in developing effective Whole Farm Plans to reduce
environmental concerns.
34
-------
Multi-year multi-rate demonstrations of manure application
for corn fertility in northeast Iowa
G. Hanson, E. Palas and J. Rodecap
Agronomy Department, Iowa State University
Northeast Iowa Demonstration Project, 111 W. Greene, Postville, IA 52162
S. Brown and G. A. Miller
Iowa State University Extension, 2104 Agronomy Hall, Ames, IA 50011-1010
Setting
Livestock is raised on nearly 90% of the farms in the four northeast counties of Iowa. In a manure
initiative started in 1987, Iowa State University Extension staff helped area farmers determine the
nutrient content of solid manure from dairy, beef, and swine feedlots, and also from dairy barn
gutters. With the results of over 1,000 manure samples in hand, a project manure management
extension specialist was assigned to assist farmers to optimize use of their manure nutrients. It
was soon apparent that one roadblock preventing many farmers from taking appropriate manure
credits was that they did not know how much manure they were applying to their fields. A set of
wheel scales was purchased with funding by five local banks to calibrate manure spreaders. A
common estimated application rate had been 10 tons of manure per acre. Calibrated rates from
over 100 spreaders showed the average application was 22 tons per acre.
Still, farmers were reluctant to take manure credits - especially for nitrogen. In most cases manure
was surface applied. Farmers were unsure how many nutrients were being lost due to
volatilization or runoff before spring tillage. How much nitrogen would be available for the first
years crop? What about second year credit? There also was the issue of uniformity of manure
application. Without uniformity, manure credits cannot be taken effectively.
Sampling manure for nutrient content so that it can be applied for maximum benefit to crops is
difficult. Although a significant amount of manure is handled as a solid in local dairy and cattle
feeding operations, increased popularity of confinement feeding has increased liquid manure
storage for most swine operations. Liquid handling makes it possible to better manage the time of
application, rates and uniformity; but obtaining a representative manure sample from the storage
structure can be a problem. The sample is usually taken when the storage structure is being
emptied after mixing, the manure has usually been applied before the laboratory results are back.
Considering these management issues and uncertainties about manure crediting, most farmers are
not comfortable relying on manure as a consistent nutrient resource.
35
-------
Introduction
In the karst topography of Northeast Iowa, both ground and surface water resources are
threatened if manure nutrients are not properly managed. The USDA Northeast Iowa Water
Quality Demonstration Project has established a network of demonstrations to thoroughly
document nutrient content of various manures, first-year and carry-over availability, sufficiency of
modest rates for corn production, and the effects of excess application.
On-farm manure demonstrations for com production compare various manure resources, nitrogen
fertilizer (with and without manure) and check plots. Measurements on the replicated plots include
manure nutrient contribution (through manure analysis and spreader calibration), plot yields and
soil and plant nitrate nitrogen analysis. New demonstrations are developed annually throughout
the project area, and long-term demonstrations are also maintained. Both long-term and short-term
manure application demonstrations are designed specifically to address producer concerns about
the concentration of manure nutrients and their seasonal availability, control of spreading rates,
and other management issues related to optimizing the use of this on-farm resource.
Multi-year demonstration - "Burrack-Kregel Site"
The Burrack-Kregel site was a 10 year manure management demonstration initiated in 1987 by
the state-sponsored Big Spring Demonstration Project on continuous com. Demonstrated was
the effect of a fall application of low (1,750 gal.), medium (3,500 gal.), and high (7,000 gal.) rates
of swine finishing manure compared to 180 lb./A. applications of commercial N (urea) applied in
the fall and spring. A check area had no manure or nitrogen applied. The treatments were
replicated three times for eight years (1987-1994). Each treatment area was approximately
30x100 feet. Yields and stalk NO3 tests were hand harvested from the center rows.
Environmental benefits - results and discussion
The "more is not better" observation was borne out by the demonstration. Average continuous
corn yields were identical for 5 years (1989-1994) at 153 bushels per acre on plots receiving
3.500 and 7,000 gallons of manure per acre. The five-year average yield from the 1,750 gallons
manure treatment was 145 bushels per acre. The impact of excessive manure application was
shown at the end of eight years when the end-of-season cornstalk analysis averaged 8,606 ppm
NO3 -N (4x excessive) in the 7,000 gallons manure treatment, 4,046 ppm (2x excessive) in the
3,500 gallon manure treatment, and 605 ppm (marginal) in the 1,750 gallon treatment. Similar
relationships between the manure treatments were present when the soil was analyzed for P and
K, however, soil pH test results were lower following increasing rates of manure application (see
graphs).
Area producers showed considerable interest in the Burrack-Kregel site demonstration results,
which were distributed annually through project newsletters and extension crop meetings. They
began to raise additional questions, "How quickly will a field without manure history respond to
manure application?", and "How much carryover nutrients are available to the next year's crop?"
In fall of 1994 the demonstration design was changed to address their questions. During crop
years 1995 and 1996 neither manure nor commercial fertilizer was applied to the areas which had
previously received manure. Instead, 3,500 gallons of swine finishing manure was applied to the
check areas and to the areas which had previously received 180 lb./A. commercial N in the spring.
36
-------
200
i 150
•1,100
50
280
240
Key to treatments
1987-94
1995-96
7.000 gal./ac. manure _ _ none
3,500 gal./ac. manure none
1.750 gal./ac. manure none
Check 3.500 gal./ac.
manure
\
\
'86 '87 '88 '89 '90 '91 '92 '93 '94'95 '96 '86 '87 '88 '89 '90 '91 '92 '93 '94 '95 '96
Graphs of 1987-1996 Burrack-Kregel manure management corn yield and phosphorus.
The check areas responded immediately to the manure application, producing yields equal to the
fall application of 180 pounds of commercial nitrogen. Likewise the manure application on the
previous spring urea treatments produced similar yields, indicating that a modest application of
manure would provide sufficient N to the crop the first year of application. The areas which had
medium and high rates of manure the previous eight years also produced well the first year with
no additional N fertilizer, indicating there was carry-over N from the previous manure applications.
The late spring soil nitrate and fall cornstalk nitrate tests added information on the environmental
impact of various manure application rates.
The cooperators who provided manure for this demonstration site admitted they were skeptical at
first about manure nutrient credits, and followed results for three years before starting to make
changes in their own crop/manure management practices. Now, however, they have
discontinued starter fertilizer and commercial N following manure application on all of their
continuous corn fields, and they have asked the Demonstration Project staff to assist with field
demonstrations designed to further explore how much commercial fertilizer can be saved on corn
the second year following a manure application. They now hire trucks to move manure to more
distant farms, and have contracted with another livestock producer to purchase additional manure
to apply to a nearby rented farm. They have commented that the work being done by the
Northeast Iowa Demonstration Project provides information they cannot get from any other
source, and that, without having local manure demonstration data; they would not be treating
manure as a resource for their farm.
Multi-site demonstrations
Over the last three years manure demonstrations were established on fifteen additional farms
using a design similar to the long-term demonstration. Time of application varied (fall, winter or
spring) as well as the type, rate and nutrient content of the manure, according to what was
available from the cooperating farmer. Spreading rates for manure applications were calibrated
and manure samples analyzed for each demonstration to determine nitrogen credits.
Results and discussion
Average corn yields and end-of-season cornstalk residual nitrate levels from the fifteen sites are
shown in table 1. Based on these results, project staff advise local producers that applying 50
37
-------
Ib. N per acre in addition to a typical manure application should provide sufficient nitrogen for their
corn crop. Since manure spreading patterns (uniformity) and nutrient concentrations are not
always consistent; the 50 Ib. N application will compensate for this variability. They can also
expect an immediate yield response from moderate manure applications on fields that have not
received manure for several years, while high to excessive rates provide carry-over residual
nutrients for the next crop year.
Table 1. Average corn yields from 15 manure management demonstrations, 1994-1996.
Treatments Com yield (buJA.) Stalk nitrate (ppm)
No manure, no nitrogen 124 542
Manure', no nitrogen 131 1,850
Manure1,50 lb./A. nitrogen 134 2,498
Manured 100 lb7A. nitrogen ; 133 3.936
1The average N credit was 124 Ib/A. from all liquid and solid manure sources.
All producers who have cooperated with field demonstrations, spreader calibrations, and manure
sampling activities have made adjustments in their manure management and reduced purchases
of commercial fertilizer. But the impact of these on-farm demonstrations doesn't stop at the farm
gate.
Technology transfer of demonstration results
Methods used to transfer manure management results from demonstrations to local practice have
included intensive information marketing, one-on-one assistance, and an innovative incentive
education workshop program. Through these efforts, results have reached a large audience.
• Self guided tours of selected demonstrations have allowed farmers to observe manure
and nitrogen management options throughout the growing season. Signage plus a
mailbox containing brochures with previous demonstration details help visitors understand
the manure demonstration.
• A manure management poster has been displayed at more than 25 events attended by !
over 5,000 people, including 12 locations outside the project area. An average of 250 !
results brochures and 150 bumper stickers ("Manure Happens - Take Credit") are i
requested each time the display is used.
• Frequent news releases tailored to the style needed by small community news media, and i
a project newsletter Water Watch with a bimonthly circulation of 1,700 convey
demonstration results to interested local producers.
• A survey conducted by the Demonstration Project in 1995 showed 92 percent of farmers ,
surveyed were aware of the project's Water Watch newsletter and results from N and
manure field demonstrations were cited most often as the most useful information it carried.
One-on-one with local farmers
Farm Services Agency (formerly ASCS-FmHA) offices provided referrals to beginning farmers in
the project area for a manure utilization/crop fertility planning demonstration. The Iowa Leopold
Center for Sustainable Agriculture also provided funding. An extension manure management
i
38
-------
specialist worked individually with producers to develop a manure nutrient inventory, calibrate
spreaders and sample manure. The program had given these young farmers increased
management skills and confidence. Of the 17 who completed the program, only three said they
were making fertilizer decisions for their own farms when the program began. By the end, this
number had increased to 14.
Educational workshops
An innovative incentive education workshop program was designed to reach a large number of
producers making more efficient use of staff time and resources. Demonstration results are used
to reinforce the technical information provided. Participants learn to analyze and sample their soil
resource, set realistic yield goals, develop a manure nutrient inventory, determine manure and
legume credits, and prepare nutrient and pest management plans for their own farms. A second-
year workshop participants' survey indicates that 92 percent have reduced nitrogen use and 82
percent of those did so by taking manure nitrogen credits. Eighty percent of the surveyed
producers indicated they were more involved in soil test interpretation, compared to half before
the program began.
Lessons learned
There are legitimate reasons why farmers hesitate to rely on nutrient credits from their manure
resource, including uncertainties about nutrient content, application rate, and uniformity of spread
patterns as well as the timing of application and cost of moving manure given historically
inexpensive commercial fertilizer sources. For areas where improved manure management is both
an environmental and economic priority, the Northeast Iowa Demonstration Project has shown the
steps needed to effectively cause change in management practices. First, a local database of
manure analytical results and spreader calibrations is generated to quantify the potential nutrient
resource. Second, a long-term series of local, on-farm demonstrations provides credibility for the
economic and environmental benefits of improved manure utilization. Demonstration results
enhance education programs, but farmers want to observe the demonstrations firsthand, and may
still make changes slowly or on limited acres to build confidence.
As farmers become more environmentally aware of manure and try to refine its use they will in turn
expect their fertilizer suppliers, crop consultants and custom manure applicators to recognize
manure as a resource. Demonstration Project staff have provided training on spreader calibration
and manure crediting to crop consultants and ag businesses as a result of farmer-initiated
questions. One farmer relayed that he had requested his custom applicator calibrate his
spreading rate. The custom applicator had never thought of doing this but complied with the
farmer's wishes. A year later the custom applicator returned to the farm and stated calibrating was
the best thing he had ever done for his business and he now lets his other farmer customers
know what is applied per acre so they too can take credit.
Another positive outcome of the manure demonstration effort has been the farmer-to-farmer
dialogue that takes place as a result of the educational outreach efforts. When farmers meet,
either at incentive program meetings or at the local coffee shop, they do talk about what has
happened with the manure demonstrations and how they save money by taking manure credits.
One farmer even wrote a guest editorial for an area newspaper on how developing a manure
management plan reduced his commercial fertilizer needs. This peer reinforcement is ultimately
one of the best ways to get manure best management practices established on the land.
39
-------
Adoption of Best Management Practices (BMPs) to Meet Water
Quality Goals in the Granger Drain Hydrologic Unit Area
R.G. Stevens, T.W. Ley and V.I. Prest
Washington State University, Prosser, Washington
This report provides an overview of the Granger Drain Hydrologic Unit Area which has
been active since 1991. Implementation and evaluation approaches utilized,
environmental benefits measured and lessons learned are reported.
Setting
The Granger Drain Hydrologic Unit Area (Granger HUA) is located in the southern portion
of the Yakima River Valley in central Washington State. The Granger Drain is composed
of a natural and man-made drainage network that drains approximately 17,000 acres of
highly productive irrigated agricultural land. The area within the Granger HUA is part of a
desert climatic zone receiving 7-9 inches of precipitation annually. Crop production is
dependent upon irrigation water from mountain storage reservoirs. Irrigated soils are
predominately silt loams found on rolling topography (2-8%). Irrigation return flows from
surface irrigation systems are collected in a series of sub-drains and are returned to the
Yakima River via the Granger Drain. This highly productive agricultural system supports a
wide variety of crops including: corn, pasture, asparagus, alfalfa, grapes, mint, orchards,
hops, wheat and many specialty crops. The Granger HUA has eighteen dairies within its
boundaries with cow populations ranging from 100 to over 3,000 and averaging over 600
producing cows.
There are approximately 450 agricultural producers in the project area. This number
comprises both commercial operators (275) and noncommercial operators, with outside
employment. Most of the small acreages are utilized as pasture. The area surrounds and
includes two small communities, Granger and Outlook, with a combined population of
2,000.
Suspended sediment, nutrient and pesticide loads from irrigated agricultural areas of the
lower Yakima River basin have long been recognized as serious impairments to water
quality. The effects of soil erosion on farmland and the effects of sediment and
dichlorodiphenyltrichloroethane (DDT) on the aquatic resource have been the focus of
numerous activities by several agencies. Several reaches of the lower Yakima River and
several of its tributaries violate numerous state water quality criteria and federal
41
-------
guidelines (Rinella, et ai. 1992, Ecology, 1994,1995). The Granger Drain (WA-37-1024)
has been cited by the Washington Department of Ecology (Washington DOE) as
exceeding standards in the following parameters: DDT, 4-4-DDE, 4-4'-DDD, Dieldrin,
Endosulfan, fecal coliform, dissolved oxygen, temperature, pH and ammonia. The
Washington DOE estimated that the Granger Drain contributed 60 tons/day of suspended
solids during the 1995 irrigation season (unpublished data Joe Joy, Washington DOE).
Objectives
The overall project goal is to reduce nutrient, biological and sediment loading from the
Granger Drain to the Yakima River mainstream to a level which allows the river to meet
its classification as a "Class A" water according to Washington DOE standards. The
specific water quality objectives are to accomplish the following: 1. Reduce sediment
loading by: a. increasing irrigation use efficiency by improved scheduling;
b. decreasing sediment load in tail water by using Best Management Practices (BMPs);
c. reducing tail water movement off the field by reuse. 2. Reduce nutrient loading to
surface and ground water by: a. proper assessment of yield goals and nutrient needs;
b. reducing nitrogen movement by proper timing and placement; c. reducing excess
nutrient applications through soil testing and crediting all available nutrient sources.
3. Reduce input of E. coli by: a. optimizing waste management and confined feeding
operations; b. optimizing waste application methods and timing; c. renovation and
management of pastures.
The key to all of the above objectives is the implementation of BMPs at the individual field
level as part of a coordinated farm water quality effort.
Implementation and Evaluation Approaches
Project objectives are being met by providing educational materials, demonstrations,
technical assistance and developing working partnerships. Implementations of BMPs has
been directed at individual producers by using a newsletter and CE publications to
provide educational materials, commodity and area meetings and demonstration sites to
share technology and follow-up with individual producers to implement BMPs.
A major focus of the project has been directed at dairy operations and associated nutrient
management concerns. Many of the eighteen operating dairies in the HUA have
increased significantly in cow numbers, with some dairies more than doubling. These
increases have placed an additional strain on waste facilities and nutrient loading. The
Lower Yakima Conservation District (CD) working with NRCS has worked with fifteen of
the eighteen dairies to develop or update dairy waste management plans. This effort has
been mainly directed at improvement in handling facilities to prevent movement of waste
into surface waters. Approximately 44% of the $300,000 of FSA cost share money spent
in the HUA has been spent on dairy waste facilities. Cooperative Extension's role has
been to work with dairymen and other producers receiving manure to implement BMPs
for nutrient management. Nutrient content of dairy waste, estimation of crop yield and
42
-------
nutrient requirement and the use of soil testing have been stressed as part of nutrient
planning. A 1993 survey of dairy storage lagoons in the HUA found that with current
management practices lagoons had significantly lower nutrient levels than other
Northwest production areas (Table 1.). This information allowed dairymen to modify their
application practices and better utilize this resource.
Table 1. A comparison of dairy lagoon nutrient concentration in Pacific Northwest
production areas.
TKN Inorganic N Total P Total K
lbs/1000gal lbs/1000gal lbs/1000gal lbs/1000gal
Granger Drain, WA
Whatcom County1, WA
Willamette Valley1, OR
2.80
13.60
4.88
1.56
7.20
4.46
0.55
3.0
0.37
2.43
14.10
5.10
1 Data collected by Henry Bierlink in Whatcom County CE and by Mike Gangner in the Willamette Valley
Soil sampling to a depth of 4-6 ft in producer fields that have long histories of manure
application have shown significant buildup of residual soil nitrate after harvest. These
levels which often exceed 300 Ibs N/ac have been used to demonstrate that excess
nitrogen is being applied thus increasing the potential risk of significant nitrate being
leached to ground water. Demonstration plots have been utilized to show that manure
applications on these fields can be reduced or eliminated without yield reduction the next
year. Phosphorus (P) soil test values in excess of 200 Ibs P2O5/ac (bicarbonate
extractant) have been found indicating long-term build up of P with its potential for
movement to surface waters. Current efforts are addressing the potential for manure
composting creating a product that can be economically transported greater distances
from the dairies.
Since the major mechanism for the movement of nonpoint pollutants to the Granger Drain
is through runoff from surface "furrow" irrigation, a major effort of the project was limiting
the movement of sediment off the field. Converting surface furrow irrigation to either
sprinkler or drip irrigation is the best long-term solution to this problem, because this
essentially eliminates surface movement of NPS pollutants. However, this conversion is
expensive and, therefore, implementation of this BMP is slow. Approximately 55% of the
FSA cost share monies were used to help producers make this conversion and improve
delivery systems. With proper management this conversion eliminates surface movement
of nonpoint pollutants.
One of the most rapidly adopted BMPs was first introduced by the HUA project in 1994.
Researchers had determined that small amounts of polyacrylamide (PAM) added to
surface irrigation water could effectively reduce soil erosion under furrow irrigation. Some
of Washington's first demonstrations were conducted in the HUA and sediment losses
from the end of furrows were reduced by 90-95%. Producers have continued adopting
43
-------
the use of PAM and CE and NRCS personnel continue providing technical assistance to
producers desiring to start using this practice. The use of PAM is a cost effective way of
improving irrigation infiltration and significantly limiting movement of sediment and
attached chemicals.
In 1992 the HUA was selected for a test site of a new field-level P index used to assess
the potential for P movement. High P index levels were found associated with irrigated
cropping practices where manure applications had been made (Stevens, et.al. 1993).
This information is being used to increase producer's awareness of the long-term effects
of continuous high rates of manure application.
In 1993 the HUA program utilized the Home*A*Syst program educating rural landowners
of potential management practices that may lead to degradation of drinking water
supplies and to introduce management practices that can reduce those risks. This was
the first application of this tool in the state. Participants were solicited by offering free
nitrate testing for domestic wells. Participants reported changes in current practices that
would reduce the potential for drinking water contamination and environmental
degradation.
To date the success of the project has been based on changes in public and producer's
attitudes about water quality and their responsibility as an active part of the problem and
the solution. Success has also been based on the successful implementation and
continued use of BMPs by producers.
Although the Granger Drain HUA is a joint project with Natural Resource Conservation
Service (NRCS), Washington State University Cooperative Extension (CE) and the Farm
Service Agency (FSA), the activities of these groups in the HUA has been a catalyst for
many working partnerships within the HUA and across the greater Yakima River
Watershed. These partnerships are leading to increased efforts towards improving water
quality across the Yakima River Watershed.
Environmental Benefits Measured
Although water quality monitoring has not been a part of this project, the Washington
DOE has monitored portions of the Yakima River. In 1994 and 1995 the Washington DOE
undertook a total maximum daily load (TMDL) evaluation in the lower Yakima River basin
including the Granger Drain to control suspended sediments, turbidity and DDT
contamination. Preliminary results of this study indicate reduced levels of E. coli.
However, sediment levels continue to exceed acceptable levels. Washington DOE has
established TMDL targets for sediment from the Granger Drain and the HUA is working
with producers developing strategies to meet these goals. The TMDL requires return
drains to be at 25 ntu or 56 mg/l for total suspended solids, requiring a 85-95% reduction
in the Granger Drain discharge.
44
-------
Based on the effort of the HUA project and the established TMDL, the local irrigation
districts have initiated a monitoring program that will be used to evaluate the
effectiveness of implemented BMPs and in evaluating future efforts.
Lessons Learned
In 1991 when this project was initiated the general public and producers had not
accepted that a water quality problem existed or that they were part of the solution. The
HUA over the years has served as an example of how water quality problems should be
addressed in other areas in the watershed. During this time a Yakima River Watershed
Council (YRWC) has been formed with an active water quality committee using the HUA
as a focal point. As a part of the YRWC an interagency group has been formed
coordinating efforts and facilitating transfer of technology between agencies and areas of
the watershed.
Rate of adoption of BMPs was found to be directly related to cost of BMP implementation.
Conversion of irrigation systems often costing $800-1,000/ac are much slower to be
implemented than practices such as the use of PAM costing $4-6/ac per application.
However, the implementation of expensive BMPs is often the only long-term solution to
problems. Therefore, improving water quality in these cases should be considered a
long-term effort.
Although the levels of sediment reduction that was initially anticipated have not been
reached, producers and other involved parties are actively working on strategies to make
things happen. One of the major lessons learned here is that it takes time to lay the
groundwork that is often necessary in accomplishing complex goals such as improved
water quality.
References
Ecology, 1994. "1994 Section 303(d) List submitted to EPA" Washington Department of
Ecology, Water Quality Program, Olympia, WA. 63 pgs.
Ecology, 1995. Impaired and threatened water bodies requiring additional pollution
controls-proposed 1996 Section 303 (d) list. Washington Department of Ecology Water
Quality Report: ECY# WQ-R-95-83. Olympia, WA. 25 pgs.
Rinella, J.F., S.W. McKenzie, and G.J. Fuhrer. 1992. Surface-Water-Quality Assessment
of the Yakima River Basin, Washington: Analysis of Available Water-Quality Data
Through 1985 Water Year. USGS Open File Report 91-453, Portland, OR.
Stevens, R.G., T.M. Sobecki, and Thomas L. Spofford. 1993. Using the Phosphorus
Assessment Tool in the Field. J. Prod. Agric. 6:487-492.
45
-------
DEMONSTRATION AND HYDROLOGIC UNIT PROJECTS
IN NORTH CAROLINA:
THE TEAM APPROACH TO IMPROVING WATER QUALITY
Maurice Cook
North Carolina State University, Raleigh, NC
Steve Coffey, Frank Humenik, Greg Jennings, Rich Mclaughlin,
Mark Rice, Casson Stallings
North Carolina State University, Raleigh, NC
George Stem
USDA Natural Resources Conservation Service, Raleigh, NC
Patrick Hunt, Terry Matheny, Jeff Novak, Ken Stone, Ariel Szogi
USDA Agricultural Research Service, Coastal Plains Soil, Plant, and Water Conservation
Research Center, Florence, SC
Introduction
The Goshen Swamp Watershed, in the southern coastal plain of North Carolina, was selected
in 1990 as a site for a Hydrologic Unit Area (HUA) Project. In the same year the Herrings
Marsh Run Watershed, which lies within the Goshen Swamp Watershed, was selected as a
site for a Demonstration (Demo) Project. This pair of watersheds, with their contiguous
relationship on the landscape, provided an excellent opportunity to bring together several
federal, state, and local agencies to focus on agricultural land use and its consequential
effects on water quality.
The overall objective of both projects, to encourage accelerated, voluntary and widespread
adoption of management practices and technologies that cost-effectively reduce impacts on
surface and ground water and result in documented water quality benefits, has been achieved.
The successful accomplishment of this objective required a collaborative and cooperative team
approach to stimulate agricultural producers to adopt best management practices (BMPs).
Setting
The study watersheds are typical of much of the Atlantic Coastal Plain region of the
southeastern United States. Soil parent materials are marine and fluvial sediments containing
mixed sands and clays. Most of the soils in the watershed are sandy and well drained. The
landscape is moderately dissected, consisting of gently undulating uplands and gentle valley
slopes.
47
-------
Two aquifer systems describe the ground water in the project area-the Surficial aquifer and
the Cretaceous aquifer. The Surficial aquifer is the saturated portion of the upper layer of
sediments, typically 7-17 meters thick. The Surficial aquifer is unconfined, i.e., its upper
surface is the water table rather than a confining bed. Thus, it is sometimes called the water
table aquifer. Many shallow wells tap the Surficial aquifer, which is particularly vulnerable to
contamination.
The 5,000-acre Herrings Marsh Run Watershed (Demo) encompasses a broad mix of rural
land uses which includes 120 farms and about 200 residences. Com and soybeans are the
field crops grown on the largest acreage, but there has been a significant shift to cotton
production. Area producers also grow at least 10 other crops of significance, including
tobacco and vegetable crops. Livestock operations include about 25, 000 hogs, 50, 000
turkeys, 130,000 broilers and 110,000 pullets. The number of swine in the watershed has
doubled since 1990, and a continued increase in production is expected.
The Goshen Swamp Watershed (HUA), which covers approximately 130,000 acres, has a crop
and livestock production pattern similar to that in the Herrings Marsh Run Watershed. The
sandy soils, fluctuating water table, and intensive crop and livestock operations throughout the
watershed provide a setting conducive to surface and ground water contamination.
Best Management Practices (BMPs)
Similar BMPs were promoted in both watersheds. Data will be reported for the Goshen
Swamp Watershed since it also includes the Herrings Marsh Run Watershed.
Nutrient management was a major thrust because of the natural setting and the agricultural
enterprises described previously. Recommended nutrient management practices include soil
sampling, plant tissue sampling, waste sampling, crediting for nutrients contained in animal
manures, calibration of application equipment, and split applications of fertilizers, especially
nitrogen. Nutrient management plans have been developed for over 20,000 acres of cropland.
Over TOO animal waste utilization plans have been developed to use more than 280,000 tons
of animal manure on 1,700 acres of cropland.
Pest management plans have been developed for producers on 1,700 acres of cropland. The
plans include scouting to assess the need for pesticide application, reduced rates of pesticide
application, and the use of chemicals that are less persistent in the environment.
Animal waste management practices for water quality protection emphasize good manure
handling, collection and storage, off-site transport of manure, and on-site management of
manure as a plant nutrient or feed source for livestock. Eighty-five animal waste
storage/treatment systems, including three poultry mortality compost facilities, have been
installed on farms throughout the Goshen Swamp Watershed. The compost facilities have
been a factor contributing to the statewide interest in poultry mortality composting. Permits
have been issued for more than 500 mortality composters since the projects began.
The kinds and amounts of BMP implementation were determined through producer surveys
designed to track land use and land treatment activities at field and watershed levels.
Separate surveys for cropping and animal production systems were used. The BMPs
employed in both crop and livestock management were identified and described.
48
-------
Landscape features, which may be viewed as naturally occurring BMPs, were modified to
reduce the amount of nitrogen reaching the streams in the watershed. Beavers constructed a
dam near the Site 2 monitoring station which produced an enhanced in-stream wetland. In the
upper reaches of the subwatershed above Site 2, a field border between a swine wastewater
irrigation field and a small stream was planted to tree seedlings to intercept the nitrogen in the
laterally moving ground water before it enters the stream.
Water Quality Evaluation
Methodology
Water quality of streams in the Herrings Marsh Run Watershed has been monitored using four
continuous sampling stations. The four continuous monitoring stations for stream discharge
and water quality data are located as follows: Site 1 (Red Hill) at the watershed exit; Site 2
(Beaver Dam) along a tributary downstream from intensive swine and poultry operations; Site
3, the background site, along the main stream flowing through woodlands; Site 4 located
upstream from Site 1 to monitor the eastern portion of the watershed.
Sample collection has been continuous from October, 1990. Water samples have been
collected hourly and combined into three-day composite samples. They are analyzed for
nitrate-nitrogen, ammonium-nitrogen, total Kjeldahl nitrogen, ortho-phosphorus, and total
phosphorus. Stream discharge is recorded by the U.S. Geological Survey (USGS).
Monitoring wells were strategically placed to evaluate shallow ground water quality throughout
the watershed. Well screens were placed at depths ranging from about 2 meters (m) to about
13m. The wells were monitored monthly for nitrate-nitrogen and selected pesticides. Current
well sites include a swine waste irrigation field, pasture field receiving turkey mortality compost,
cropped areas for which nutrient and pest management practices are being implemented, and
the turkey mortality composter site.
Biological monitoring has been conducted annually at Site 1 by the North Carolina Division of
Water Quality (DWQ). Aquatic fauna are inventoried, with the primary output consisting of a
species list with indications of relative abundance (rare, common, abundant) for each taxon.
Unstressed streams have a diversity of species, while stressed streams have relatively few
species. Water quality ratings are assigned based on the abundance and characteristics of
the most intolerant invertebrate groups. Streams are classified as Excellent, Good, Good/Fair,
and Fair.
Results
In the first year of the Demonstration Project, mean nitrate-nitrogen concentrations in the
surface water leaving the watershed (Site 1) were twice the background concentrations (Site
3). Over the project period, there has been a continued reduction in nitrate-nitrogen and total
nitrogen concentrations recorded at the outlet of the Herrings Marsh Run Watershed. This
continued reduction indicates the water quality benefits of BMP implementation and landscape
modifications.
Stream water from subwatershed 2 has been consistently higher in nitrate nitrogen compared
to the other subwatersheds and watershed outlet. In the first year, daily nitrate-nitrogen
concentrations at Site 2 sometimes exceeded 10 mg/L. Over-application of animal waste to
fields probably contributed to the elevated nitrate concentrations at this sampling station.
49
-------
Since July, 1991, the maximum nitrate concentration at Site 2 has been 8 mg/L and the mean
has been about 5.5 mg/L.
Stream flow data from the USGS gaging stations were integrated with the stream monitoring
data to calculate the mass loading of nitrate-nitrogen and ammonium-nitrogen. In 1991 and
1992, the mass nitrate-nitrogen leaving the watershed (Site 1) averaged about 30
pounds/acre/day. The tributary (Site 2) received about 20 pounds/acre/day from its sub-
watershed. These levels have decreased slightly with time.
Baseline biological monitoring data indicated a bioclassification of Fair at Site 1. Lack of
quality stream habitat has limited macroinvertibrate diversity, thus there has been little change
in the bioclassification even though nitrate nitrogen concentrations have decreased
Ground water samples were collected monthly from 92 monitoring wells from 1993 to 1995
and quarterly thereafter. Analyses were conducted for three pesticides- aiachlor, atrazine,
and metolachlor, the most widely used pesticides in the watersheds. Although the use of
these pesticides is high, only two wells out of 92 had confirmed detections for aiachlor. Only
one well had a detectable level of atrazine and no wells had a metolachlor detection.
Most of the stream samples collected at the watershed outlet were free of pesticides at the
analytical detection limits. Although there are large applications of herbicides (700-850 kg
annually) in the watershed, current pest management BMPs used by local farmers and
applicators appear to be satisfactory for maintaining acceptable ground water quality.
Lessons Learned
Value of small watershed
The small Herrings Marsh Run Watershed (5,000 acres) was deliberately selected for the
Demonstration Project with the specific aim of showing, during the project life, measurable
water quality changes resulting from land-applied BMPs. Although improvements in water
quality were observed during the project period, we concluded that five years is about the
minimum time period that one should consider in water quality project planning.
Importance of land use data
Complete and accurate land use data are essential for the proper interpretation of results. The
task of obtaining such data was more difficult than we had anticipated. The farmers
themselves are the most knowledgeable source of the information but it became increasingly
difficult to arrange meetings with farmers to obtain it. It became necessary to complement the
farmer surveys with "windshield" surveys and cropping records maintained in county offices.
Modeling constraints
The selection, calibration and verification of models and the development of an interface to link
the model with a geographic information system (GIS) has been a significant activity of the
Demonstration Project. This effort was undertaken to enable one to predict water quality
impacts when a given set of land use characteristics is known. Modeling technology is
dynamic and developing, as a result, the calibration and validation of the models has been a
50
-------
slow and painstaking process. The intense data requirements of the model were challenging
even for the well defined 5000 acre watershed.
A major modeling goal is to extend water quality findings over space and time. Project
modeling has resulted in correlations between modeling and monitoring that range from good
for flow, acceptable for long term transport and poor for short term concentration and transport.
Continued work is needed to develop coordinated modeling and monitoring techniques for
water quality planning and extension of water quality cause and effect relationships over time
and space.
Dynamic reference data
The baseline conditions at the conclusion of the projects are quite different from those at the
beginning. Several events beyond our control have occurred within the last three years which
make quantitative assessment of impacts more difficult.
In the Demonstration Project, the subwatershed for the Site 3 monitoring station was primarily
a natural woodland and, thus, it provided a good reference condition. A large swine operation
is now in place in the subwatershed which has altered the background conditions. Several
new swine operations have been established in other parts of the watershed. It will be
interesting to note the effects, if any, of such operations on water quality.
Heavy rains during the summer, 1995 and the damage caused by two hurricanes in 1996 have
altered stream channels and hindered the nutrient mass yield determinations.
A development of about 40 manufactured homes has occurred in the Demonstration Project
area. On-site waste systems are being employed for the development. There is the potential
for water quality impacts, particularly when a septic system malfunctions.
Necessity for team approach
The excellent results from the project would not have been possible without the outstanding
cooperation and collaboration among the federal, state, and local agencies and organizations.
A single agency could not have carried out the broad range of activities. The Cooperative
Extension Service (CES) provided education and information programs. The USDA Natural
Resources Conservation Service provided technical assistance. The USDA Farm Services
Agency provided cost share assistance to area producers. North Carolina State University
(NCSU) scientists have led the modeling efforts to select, calibrate and validate models and
evaluate correlation with water quality monitoring date. The USDA Agricultural Research
Service (ARS), through its Soil, Plant, and Water Conservation Research Center in Florence,
SC, has conducted detailed monitoring of both surface and ground water, and has assisted
with the nitrate and pesticide modeling efforts. The North Carolina Division of Water Quality
(DWQ) has conducted biological monitoring of selected stream sites in the Herrings Marsh Run
Watershed. The US Geological Survey (USGS) has provided cost share assistance for
installation of automated sampling and flow measuring stations, maintenance for the flow
discharge relationship and flow data management.
Perhaps the most important members of the team are the farmers themselves. The projects
proceeded well due to their willingness to cooperate and assist in many ways. We give them
our thanks and appreciation.
51
-------
What Have We Learned About Our
Nonpoint Source Pollution Education Programs?
Robin Shepard
Water Quality Coordinator, University of Wisconsin-Extension, Madison, Wisconsin
Educational programming is a common element to most watershed protection projects, but the
actual components to an educational strategy vary greatly from project to project and even from
educator to educator. These efforts vary mostly in the level of program intensity and in the way
information is delivered to the public. In Wisconsin, since 1978, nonpoint source pollution
prevention strategies have been targeted to hydrologic units or watersheds in a collaborative
effort by the University of Wisconsin Extension, the Wisconsin Department of Natural Resources,
the Wisconsin Department of Agriculture, Trade and Consumer Protection, and a number of
county agencies. Expanding the watershed-based programs, in 1990 the University of Wisconsin
Extension and Wisconsin's Natural Resources Conservation Service began a USDA
Demonstration Project under the President's Water Quality Initiative.
Educational programming, often referred to as information and education (I & E) strategies,
attempts to provide information to landowners in expectation that it will lead to environmentally
beneficial actions such as the adoption of best management practices on the farm. These I & E
strategies, especially those that seek to reduce nonpoint source pollution from agriculture,
generally rely on a combination of two approaches:
1) Diffuse efforts that involve disseminating, somewhat randomly, information to a wide area -
similar to the way a shotgun disperses lead shot by spraying a target indiscriminately. These
types of information delivery approaches attempt to reach the largest possible number of the
target audience, often through mailings, newsletters and mass media.
2) One-on-one information transfer techniques such as on-farm visits, individual farm trials and
individual farmer consultation.
This analysis will consider the rate of adoption of nutrient management strategies by farmers in
two different Wisconsin watersheds over a five year-period of 1990 to 1995. One watershed, a
USDA Water Quality Initiative Demonstration project, used intensive one-on-one information
transfer processes. The other, a state funded priority watershed project, relied on more diffuse
educational strategies. The educational programming in both watershed-based projects was
coordinated by University of Wisconsin Extension educators. This analysis focuses on
educational programming, and does not consider the potential differences in additional agency
involvement and/or the impacts attributed to other public remediation efforts in the two
watersheds.
The comparison in educational program approaches indicates that more intensive one-one-one
information transfer strategies are more effective in encouraging farmers to lower rates of
nitrogen and phosphorus application on corn ground within the watershed project area.
53
-------
Introduction
In Wisconsin, nonpoint source pollution has been identified as the greatest cause of water quality
problems (Wisconsin Department of Natural Resources, 1992). Over 75 percent of inland lakes,
many of the harbors on coastal waters on the Great Lakes, and substantial groundwater resources
are affected by nonpoint source pollution. The majority of this problem is attributed to
agricultural land uses.
The pervasiveness of these water quality problems are merely symptoms of the more serious
causes stemming from failure to implement existing remedial technologies (Lockertz, 1990;
Nowak, 1983). While many reports have described the physical dimensions of water quality
problems caused by excessive nutrients caused by animal manures, few of these reports provide
reliable indicators of remedial technology adoption.
One of the most serious sources of nonpoint pollution are animal manures. These animal manures
include organic pollutants, chlorides, nitrogen and phosphorus. Wisconsin has spent much time
and money enacting strict and costly limits on municipal and industrial phosphorus dischargers
while animal operations remain largely unregulated. Wisconsin Department of Natural Resources
(WDNR) estimates show the amount of phosphorus generated from all municipal and industrial
sources is approximately six million pounds a year, at least 90 percent of which is treated or
otherwise prevented from reaching surface water. The total phosphorus contribution from these
sources is less than 600,000 pounds. In contrast, manures associated with the state's livestock
industry produces an estimated 143 million pounds of phosphorus per year. WDNR estimates
that at least 10 percent of this amount, approximately 14 million pounds, is lost to surface water
(Wisconsin Department of Natural Resources, 1992). Consequently, mismanagement of livestock I
manures in Wisconsin contribute about 25 times as much phosphorus to the state's surface water i
as all municipal and industrial sources combined. With this in mind, the success of Wisconsin
watershed projects in rural areas should be judged on the extent to which manure management
practices are used and nutrient management plans are developed and followed. I
As stated above, educational programming is a common element to many voluntary watershed I
protection projects. While those responsible for such projects extol the virtue of strong ,
educational programming, the level of staff commitment varies greatly from project to project. If i
a project does assign staff to administer educational programs, the role of educator may be
intermixed with other job responsibilities that are more technical and bureaucratic. Moreover, j
approaches that attempt to educate landowners often focus on randomly selected and
unconnected activities (Geller, Winett and Event, 1982). Carefully designed multi-year strategies i
that reach landowners who need specific assistance most are rare. '
I
Dedicating an individual staff position to a specific watershed for the purpose of implementing '
educational programs is also somewhat unique, especially in watersheds less than 300 square
miles. When such a staffing commitment does occur, there are often differences in opinion as to !
how to best provide educational programs that reach farmers. Even professional educators
disagree over the benefits associated with reaching the large number of farmers with general
information versus a more one-on-one consulting approach with a select farmers in a given area. ,
54
-------
Methods
Two watersheds were selected for a comparison of educational programming approaches. Both
watersheds were selected in 1989 to begin nonpoint source pollution remediation programs. This
designation was due to degraded surface and groundwater quality in conjunction with the impact
of sedimentation on aquatic habitat in the watershed's main river system. Both watersheds
contained numerous dairy farms, making manure runoff from barnyards and fields a major
concern.
In both watersheds, a population of farmers was defined as all farmers who operated at least 40
acres of land and/or have 15 head of dairy cattle. In a northern Wisconsin watershed 101 out of
134 farmers in the project area meeting this criteria responded to an initial questionnaire in 1990
resulting in a 75% response rate. In a southern Wisconsin watershed 208 out of 260 farmers
completed a baseline questionnaire, for a response rate of 80%. These surveys measured salient
nutrient and pest management behaviors. Among these management behaviors, farmers were
asked about their nitrogen and phosphorus application rates in the production of corn.
Specifically, the rates of eight different sources of agricultural nitrogen and phosphorus were
measured. Nitrogen and phosphorus derived from manure application was also measured by
establishing the type of manure applied (dairy, beef, swine, and/or poultry), using estimates of the
capacities of various spreaders used and the number of loads applied to the specific corn fields.
The survey also measured nitrogen added from legumes.
In 1995, both watersheds conducted follow-up surveys using on-farm interviews. Seventy-five
farmers for the detailed "time-two" survey were randomly selected from the each watershed's
original list of baseline survey respondents. This number of follow-up interviews was selected to
allow a direct comparison of each farmer between how they responded in 1990 and 1995. It also
represented an achievable number of on farm visits for a single interviewer during a two month
time period when follow-up interviews were in late winter 1996.
In addition to measuring farmer management, this study also asked the local "watershed-based"
educator to describe their approach to educational programming. Each educator monitored the
amount of time they dedicated to diffuse I & E delivery of information versus the amount of time
they spent on face-to-face or direct information delivery to landowners.
Results
Results from the 1990 baseline assessments indicated that over-application of nitrogen and
phosphorus could be attributed to the availability of nutrients from field applied livestock manure
and prior legume crops in the field rotation. The problem of excess nutrient application can be
attributed to a failure of farmers to reduce commercial nutrient purchases to compensate for the
availability of on-farm nutrients.
During the ensuing years, each project focused on improving nutrient management practices in
their respect watersheds. Each project also relied on a full-time educator to help provide
55
-------
information to farmers. However, an analysis of workload shows how the educators differed
substantially in their approach to targeting landowners and delivering information to target
audiences. While both educators used similar approaches to delivering water quality information,
the degree to which certain techniques and approaches were favored over others distinguished the
two educators. Such as in the southern watershed the educator spent more time working through
advisory committees and assisting other watershed staff in providing information to landowners.
In the northern watershed the educator dedicated nearly three times the number of more working
days in direct consultation to landowners that the southern watershed educator.
The educator in the northern watershed began by specifically targeting 120 of the watershed's
dairy farmers with personal farm visits. The educator also kept records and strived for multiple
farm visits throughout the year. In addition, the northern watershed educator placed more
attention on working with local Co-op agronomists from the watershed's three main farm supply
dealers. Past research has shown Co-op agronomists and independent crop consultants are the
most influential providers of nutrient management information (Shepard, 1993). In this watershed
nutrient planning and integrated crop management workshops were offered to the private sector
information providers - those entities (Le., crop consultants) which farmers pay for information.
The northern watershed educator followed these approaches between 1992 and 1995.
The educator in the southern watershed gave greater attention to working with influential "peer"
farmers in the watershed, often those active in the watershed's citizen advisory committee. The
southern watershed educator also dedicated more time to delivering information through the news
media, project newsletters and local events such as on-farm demonstrations, tours, farm field
days, watershed events, and local schools. In the northern watershed there one local farm
Cooperative which dominated the local fertilizer sales market. This Co-op resisted attempts by
the watershed educator to work collaborative with on-farm demonstrations and field days. The
educator in the southern watershed followed these approaches between 1991 and 1995 (See Table
1).
Other aspects of the projects were very similar in the type of technical and financial assistance
provided beyond purely educational information. Regulatory aspects for both projects were also
similar because they were administered by a separate WDNR program that was available in both
areas if needed.
Changes in critical nutrient management practices that are needed to reduce nonpoint source
pollution in both watersheds did occur, and positive benefits were seen in both watersheds.
However, the extent of change in nitrogen application was greater in the northern watershed
where the educator followed a more targeted information delivery approach (See Figure 1 and
Figure 2).
56
-------
Table 1. The Annual Average Time Dedicated by Watershed Educator on Differing
Techniques of Information Dissemination
Educational Approach
Northern Watershed
Southern Watershed
Number of
Days
10
Percent
Time
46
Number of
Days
31
15
36
16
Percent
Time
12
21
32
8
12
26
4
10
2
14
1. Human relations skills: counseling, 120
interviewing, conflict resolution and
negotiating.
2. Conducting demonstration projects 15
and field research.
3. Conducting tours and field days.
4. Working with small groups
and conducting workshops.
5. Organizing and maintaining
citizen advisory committees.
6. Conducting needs assessments
and evaluations.
7. Making public presentations.
8. Staffing booths, exhibits, fairs
and public events.
9. Writing newsletters and publications.
10. Working with the media.
11. Writing watershed plans.
12. Assisting other watershed
staff with technical issues.
*Both educators worked full time (40 hours per week), an estimated 260 days per year. The
above represent annual estimates of days dedicated to educational approaches.
15
8
20
18
0
0
6
3
7
7
0
0
10
10
34
20
24
34
4
4
13
7
9
13
57
-------
70
Under
University
Recommendations
3- 1990
Within 10% of
University
Recommendations
Over
University
Kromncudntionx
1990 Nitrogen Application Mean = 217 Ibslac*
(Based on 56 randomly selected cases.)
Grossly Over
University
RccouvDcndatjoru
1995
1995 Nitrogen Application Mean = 136 Ibslac*
(Based on 56 randomly selected cases.)
*=agnificant at the .001 leveL
Figure 1. Nitrogen Application Rates of Farmers in The IVorthern Wisconsin Watershed.
58
-------
Under
Univwsily
Wllhin 10% of
University
Over
University
Koconxntudations
HI 1990
1990 Nitrogen Application Mean = 234 Ibs/ac*
(Based on 63 randomly selected cases.)
*=not significant at the .001 level.
Grossly Over
University
Retiouvucmtations
1995
1995 Nitrogen Application Mean = 227 Ibslac*
(Based on 63 randomly selected cases.)
Figure 2. Nitrogen Application Rates of Farmers in The Southern Wisconsin Watershed.
More specifically, farmers in the northern watershed where the educator spent more time
conducting one-on-one farm visits and working directly with Co-op agronomists decreased excess
nitrogen application rates on corn ground by 80 pounds per acre. In the southern watershed
where the educator followed more diffuse information delivery approaches, the rate of decline in
excess nitrogen application was not statistically significant even though the average nitrogen
application rates showed a slight reduction.
The percent of farmers in the northern watershed who took advantage of on-farm nitrogen in
manure and legumes increased from 26 percent to 32 percent during the five year period. In the
southern watershed, only 1 percent of the farmers changed their commercial nitrogen rates due to
manure sources of nitrogen.
Other positive management changes occurred in both watersheds such as: the percent of farmers
reducing commercial nitrogen purchases due to nitrogen from prior legume crops; the percent of
farmers using soil tests; and the percent practicing appropriate manure hauling. Both watersheds
showed increases in environmentally beneficial practices, but the northern watershed experienced
a greater rate of change than the southern watershed.
59
-------
Conclusions
This comparison of educational approaches shows that greater rates of farm management
adoption are found in projects that emphasize direct transfer of information to fanners through
one-on-one contacts when contrasted to more diffused-based efforts that rely more heavily on
secondary transfer of information to farmers through newsletters, mass media and events.
Other findings include:
Superficial program targeting is insufficient. Target audiences should be identified and then
program resources, especially educational programs, should be deployed in ways that ensure that
they actually reach those who need them most.
Emphasizing mass dissemination of information more than one-on-one information transfer
techniques can diminish the impact of educational programs that encourage farmers to make
specific management changes.
Watershed management strategies, annual staff work plans, staff positions and program
approaches should acknowledge a commitment to one-on-one information delivery techniques.
Private sector information providers represent a significant influence on farmer behaviors.
Educational program design and implementation should utilize public-private sector partnerships
in program delivery.
References
Geller, E.S., R.A. Winett and E.B. Evertt. 1982. Preserving the Environment: New Strategies
for Behavior Change. Elmsford, New York: Pergamon Press.
Lockertz, William. 1990. What have we learned about who conserves soil? Journal of Soil and
Water Conservation. 45(5):517-23.
Nowak, Peter J. 1983. Obstacles to adoption of conservation tillage. Journal of Soil and Water
Conservation. (May-June): 162-165.
Shepard, Robin L. 1993. Beyond Superficial Targeting: Designing Educational Strategies for
Water Quality Programs. Ph.D. thesis. University of Wisconsin, Madison, Wisconsin.
Wisconsin Department of Natural Resources. 1992. Wisconsin Water Quality Assessment
Report to Congress. Wisconsin Department of Natural Resources Publication WR254-92-REV.
Madison, Wisconsin.
60
-------
The Royal River Watershed Education Project
John M. Jemison, Jr., and Peter R. Wagner
University of Maine Cooperative Extension
Orono, Maine
David E. Lytle
University of Minnesota
Saint Paul, Minnesota
Given the diffuse nature of non-point source (NPS) pollution, it is often difficult
to identify the source. In some areas, agriculture and natural resource industries
are significant contributors while in others residential land uses may play a more
important role. One solution to this problem is intensive, but broad-based,
educational programming in small watershed areas. In 1992, the University of Maine
Cooperative Extension began a five-year agricultural and residential watershed
education program in the Royal River Watershed area, a sub-watershed of Casco Bay.
Since both agriculture and residential sources were assumed to be contributing to
NPS pollution, we developed several project goals: 1) producers would adopt
practices to reduce NPS pollution; 2) producers would improve pest management
practices within the watershed; 3) rural residents would change activities to reduce
NPS pollution; 4) a volunteer water monitoring group would identify problem areas
within the watershed, and 5) youth awareness of water quality issues.would increase.
With cooperating farms, we have been able to introduce practices that would minimize
water quality impacts such as low input weed control, livestock exclusion from
streams, alternative watering systems, stream-bank stabilization, improved nutrient
management, and many others. Through a benthic macroinvertebrate study, we
documented improved water quality on the dairy farm as a result of implementing
BMPs. The residential program has focused on the Safe H20ME program, newsletters,
watershed stewards program, youth educational programs, and other demonstration
projects. We have seen many successes: growers have fenced out animals from
tributary streams; adopted integrated pest management; and improved manure
management. Our residential clientele have adopted many home BMPs detailed in the
Streamlines newsletter. Over 120 people completed our Safe H2ome Program and
identified specific areas of improvement. Also, 25% of the wells were found to be
polluted with bacteria (15% with E. coli); all were disinfected and retested. From
these results, we have demonstrated that broad-based education in focused areas is
effective to reduce non-point source pollution.
61
-------
Introduction
The Royal River watershed covers approximately 197 square miles, spanning 11 towns
and two counties. Land uses vary in the watershed. Over 55,000 people live in the
watershed area. There are a myriad of business activities including farming,
greenhouses, nurseries, fish hatcheries, a shoe factory, summer camps, and other
small businesses. These land uses have the potential to produce NPS pollution which
can jeopardize fisheries, wildlife, recreation, and the beauty of the river. In
fact, the impact of non-point source pollution has been .well documented in Casco Bay
estuary. Over 40 percent of commercial shellfish beds have been closed due to
bacterial contamination; also there has been a notable decline in the abundant
marine life in the subtidal zone. In 1992, The University of Maine Cooperative
Extension was awarded a grant from USDA-CSREES to begin a non-point source pollution
education project in the Royal River watershed.
Given the mixed land uses in the watershed, we divided our educational activities
into agricultural and residential focuses. Agriculture is a small but significant
interest in the watershed. We also recognized the potentially large NPS pollution
contribution from the residential community. Therefore, we proposed several project
goals: 1) agricultural producers would adopt practices to reduce NPS pollution; 2)
producers would improve pest management practices within the watershed; 3) rural
residents would change practices to reduce NPS pollution; 4) a volunteer water
monitoring group would identify problem areas within the watershed; and 5) youth
awareness of water quality interests would be increased.
Procedure
To reach our agricultural clientele, we identified three demonstration farms (dairy,
beef, and fruit/vegetable) to implement/demonstrate BMPs and potential improvements
to water quality. We identified key problems on each farm, and corrected them over
the period. We used field days, demonstrations, educational programs, video, and
others to demonstrate BMPs to producers.
To meet the goals set for our residential audience, we used a variety of mechanisms
including focused educational programming, newsletters, volunteer monitoring groups,
youth education programs, and other demonstration projects. We have done specific
evaluations on demonstration projects, newsletter, and focused educational programs.
Results and Discussion
Agricultural Education Program Activities and Results
On our dairy demonstration farm, we identified two key areas of improvement:
livestock access to the stream and poor manure management. Animal access had
degraded water quality because of bacteria, nutrients, and stream bank erosion.
62
-------
Secondly, nutrient management on the farm was inefficient due to poor manure
management. As a result, forage quality was poor.
In the fall of 1994, we designed a management plan to address these issues. Before
the producer would agree to fence off the brook, we had to demonstrate that
livestock influenced stream water quality. We decided to test benthic
macroinvertebrate (BMIs) populations in three locations in the stream (above the
farm, in the middle, and below the farm). We found significantly different
populations and differences in species diversity and feeding groups among sites.
These data were convincing to the producer, and he agreed to fence the stream. With
the animals fenced out of the stream, a watering system was required. A Ram pump
was used to push water to the highest point on the landscape; water was delivered by
gravity to each paddock. We also used pasture and solar pumps to show alternative
methods. We intensified the rotational grazing to improve efficiency, reduce weeds,
and improve manure dispersal over the pasture. At field day presentations, producers
showed a great deal of interest in this system. We also planted riparian plant
species to show producers which types of plants grow well in eroded environments.
We have documented improvements in both forage quality, milk production, and stream
bank stabilization. We continued monitoring BMIs in subsequent years and found
increases in pollution intolerant families and species diversity. We have used
these data in other meetings and presentations and have raised awareness of other
producers both inside and outside the watershed.
On the beef demonstration farm, we identified weed management nutrient management,
and rotational grazing as focus points. We conducted demonstrations to identify
ways to eliminate or reduce atrazine use. We compared atrazine to cultivation or
combinations of low-rate herbicides for weed control. We also evaluated narrow-row
corn against conventional wide row planting and found that we could cut rates by one
third. We successfully showed that a spring-tine cultivator was a viable
alternative to pre-emergence herbicides. Cultivation has become the weed control
method of choice for the producer. We also demonstrated our computer nutrient
management program. Nitrogen fertilizer was reduced with use of the pre-sidedress
nitrate test. Rotational grazing system was expanded and intensified as well.
We have worked closely with two fruit and vegetable growers. We focused on
improving nutrient management (PSNT in sweet corn), IPM in apples, and weed
managment in small fruit production. We helped cut herbicide use with cultivation
in strawberries. We improved N efficiency in sweet corn with the PSNT. Apple
producers have cut spray applications by a third.
We held field days with the Soil and Water Conservation District to bring producers
to see these practices and evaluate their effectiveness. We used grower meetings as
well. We seem to have had more success getting producers outside the watershed to
adopt these practices than from within. Many of the beef and dairy producers are
small, part-time operations with limited income, and do not believe their operations
have impact. However, we have used these demonstrations in other educational
programs around the state.
63
-------
Residential Education Program Activities and Results
Our residential education program has been extremely successful. The key features
of the program have been the Safe H2(W£ Program, the Streamlines newsletter,
volunteer monitoring, and the watershed stewards program.
The Safe H2OA/£ Program has been our most effective educational tool. The program
materials consist of five factsheets, each with a worksheet for homeowners to rate
the safety of their activities. Homeowners who completed the materials reported the
activities that were high to moderately high risk (Table 1). Specific changes
resulting from doing the project are reported in Table 2.
Table 1. Results of Targeted Safe H2OWf Evaluations
Self Risk Assessments Percentage with Moderately
High to High Risk
Well construction and maintenance 25.8
Household hazardous waste 23.6
Household wastewater 25.8
Lawn and garden care 23.2
Lead in home 24.1
Table 2. Changes made as a result of completing Safe H2OA/E Program
Areas of intended change Percentage
Change storage practices of hazardous products 38.9
Have water regularly tested 23.4
Have septic system pumped and maintained 13.7
Improve well construction 10.9
Monitor / decrease water usage 5.7
Increase soil organic matter 2J3 j
i
Over 120 residents completed the program materials. When participants returned
their evaluations, they were given a water test. We found that 27% of the wells ,
were contaminated with bacteria and 15% with E. coli. We contacted those
homeowners and discussed disinfection methods. All wells were sanitized and were >
retested. !
Another way we chose to reach the residential community was through our newsletter
Streamlines. During the course of the project we had over 1650 residents subscribe
to the newsletter. Topics have focused on composting, safe gardening, water
testing, wetlands, chlorinating wells, bacteria in water supplies, and many others.
Through surveys, residents told us that their understanding of NPS pollution
increased (81%). they used educational activities presented in the newsletter (24%).
64
-------
made specific changes in home management(55^) including increased recycling efforts,
composting, conserving water, pumped septic systems, used pesticides more safely,
and reduced fertilizer use near the river. The newsletter received two awards. We
placed second in the national Agricultural Communicators in Education contest; and
we won first place in the National Association of County Agricultural Agents/AT&T
contest for team newsletters.
We have worked closely with the Friends of the Royal River (a local volunteer
monitoring group) to develop a long-lasting monitoring program that will identify
problem areas in the river. We have trained volunteers, designed sampling
protocols, and established a process such that the group should be able to continue
well in the future. Similarly, we have trained 15 individual in the Watershed
Stewards Program. Similar to the Master Gardner program, these people received 20
hours of water-based education. In return, they volunteer a minimum of 20 hours of
volunteer community service around a project. Some are working with the monitoring
program, others with schools, towns, and other water-related programs.
Lastly, we have had a very active youth education program. We have worked with
local high school students on the BMI.project. Advanced placement biology students
visited the dairy demonstration farm, evaluated farm activities, and learned about
the dairy industry. On another day, they assisted in sorting the BMIs as well. We
worked with the environmental biology classes with data collected by the Friends of
the Royal River. They studied chemical data collected by the volunteers,
synthesized the data into a report, and presented the material to area residents.
We have also had an active program with elementary students as well. Project staff
have delivered over 20 programs to area 5th and 6th grade classes on environmental
education. We also have assisted with the Southern Maine Children's Water Festival
over the past two years. Over 130 students from the watershed attended the program.
Children have taken these ideas and done class projects from what they have learned.
Conclusions
Our educational programming efforts have been successful in reducing non-point
source pollution within the Royal River watershed. Through our three demonstration
farms, we have worked to present to the agricultural community sound production
practices that will lessen the impact of farming activities on river and estuary
water quality. We have also targeted the residential community and the residents
have responded by implementing many pollution-reducing practices. Taken together.
we feel that these educational programs have and will continue to help improve water
quality in the Royal River, its tributaries, and the Casco Bay estuary.
65
-------
The Oak Creek 319 (h) National Monitoring Program
Ravin Donald, Christine Crabill, Brent Birch, Richard Foust and Gordon Southam
Northern Arizona University, Flagstaff, AZ 86011-5640
Oak Creek, AZ, which has cut a gorge in the southern rim of the Colorado plateau near
Sedona, is a popular recreation area (swimming, hiking, fishing and camping) that is impacted
annually by fecal pollution. The goal of this project is to demonstrate that a set of Best
Management Practices (BMPs) will reduce the amount of non-point source pollutants
(ammonia, nitrate, phosphate and fecal coliforms [FC]) being discharged into the waters of Oak
Creek. BMPs at Slide Rock State Park (SRSP), the site most impacted by planktonic fecal
pollution, include; an attempt to reduce the number of visitors to the park, improved access to
restroom facilities and a public education program. Statistical analysis using pre- and post-
BMP data has shown no improvement to water quality in the park. By expanding our watershed
monitoring program (during 1995 and 1996) to include interstitial (sediment) FC reservoirs in
Oak Creek, we have found that the impact of sediment fecal pollution occurs throughout the
length of the canyon and therefore, the problem is more widespread than at just SRSP. During
summer months, sediment reservoirs of fecal pollution in Oak Creek can exceed 10,000,000 FC
per 100 ml suspended sediment but drop 4 orders of magnitude as winter approaches. This
sediment data also contradicts all of the previous water quality studies performed over the past
25 years which have held recreational users responsible for the fecal-related water quality
impacts. In Oak Creek, the strong correlation between recreational use and fecal pollution does
not prove causation. This study demonstrates the importance of determining whether disrupted
sediment reservoirs of fecal pollution are responsible for observed impacts to water quality.
Introduction
Oak Creek has been plagued with an annual, seasonal deterioration in water quality (fecal
pollution) during the summer monsoon season (Hansen and White, 1992; Jackson, 1981; Obr
et a/., 1978; Rose et a/., 1987). These studies focused on the recreational day-time users as
the primary source of fecal pollution.
During 1995, two major reservoirs of fecal pollution within Oak Creek Canyon were identified:
SRSP and the Switchbacks (located at the upper reach of the canyon). These reservoirs
averaged 2,200 times more FC than the water column. The large distance (10 km) and the
significant reduction in pollution between these two sites is indicative of the existence of more
than one source of fecal pollution. Also, the occurrence of sediment bound fecal pollution at
SRSP prior to the monsoon season suggested that the source of fecal pollution must be close
to the creek because a long-distance transport mechanism i.e., monsoons, is not in place. This
implicates a human (recreational and/or residential) source of fecal pollution at SRSP. Contrary
to this, the correlation between the summer monsoon rains and the FC build-up at the
Switchbacks suggests that fecal material from the abundant elk, deer, and cattle populations on
the surrounding uninhabited plateau impact the creek. This study investigates the fecal
pollution at SRSP.
67
-------
Materials and Methods
Sample sites along Oak Creek included Pine Flats Campground (RFC), Slide Rock State Park
(SRSP), Mazanita Campground (MZC) and Grasshopper Point (GP). Upstream and
downstream samples were taken at each site expect for sediment fecal analysis which for
which only downstream samples were taken. Sediment fecal samples were also taken
upstream at Pump House Wash.
All samples were collected (Protocol 9060 A.), preserved and stored for analysis (Protocol 9060
B.) as outlined in the Standard Methods for the Examination of Water and Wastewater (APHA,
1989). Nitrate analyses were done on field filtered (0.45 |im) samples (EPA, 1983). A
separate sample was collected for ammonia and phosphate analysis by field filtering and
acidifying the samples with sulfuric acid (EPA, 1983). All samples were stored at 4 °C until
analyzed.
Water samples were collected using the grab method. Sediment samples were obtained using
a hand trowel rinsed with stream water. Sediment samples (500 ml volume) were collected
from the upper 10 cm of creek sediment and placed into sterile 1 I plastic bottles. Each
sediment sample was taken immediately following water sampling from a location directly
beneath the water sample.
FC were enumerated using the Fecal Coliform Membrane Filtration Technique (Protocol 9222
D.; APHA, 1989). For sediments, a suspended sediment (SS) fraction was produced by adding
100 ml of 0.85% (wt./vol.) sterile saline to each sample, vigorously shaking for 30 s and then
allowing settlement of the larger stream aggregates. Two 10 ml aliquots of the resultant
supernatant (the SS fraction) were collected for analysis. The first aliquot was placed in a
graduated centrifuge tube and allowed to settle overnight at 25 °C to measure the sediment
load. The second 10 ml SS fraction was added to a Waring® blender containing 90 ml of sterile
saline and mixed for 5 min. Appropriate dilutions were then enumerated according to the
technique used for water samples.
Results and Discussion
Chemical analyses performed on Oak Creek demonstrated that the water was within
acceptable limits and of high quality (Table 1). Nitrate, ammonia and phosphate concentrations
were generally well below Oak Creek Unique Water Standards (Mueller, 1984). They are also
far below the standards set for total nitrogen (2.5 mg/l) and total phosphorous (0.30 mg/l) set
for Oak Creek in the Arizona Administrative Code (Title 18, Ch. 11).
Table 1. Critical Parameters of Oak Creek for Year 1996
Nitrate (mg/l)
Ammonia (mg/l)
Phosphate (mg/l)
Fecals (cfu/100 ml)
Sediment Fecals (cfu/100 ml)
Mean
0.066
0.013
0.020
160
1,236,329
Maximum
0.290
0.043
0.082
3,500
74,400,000
Minimum
0.000
0.001
0.005
0
0
68
-------
Fecal levels in the water column most often exceeded the water quality standard of 800 cfu/100
ml (Mueller, 1984) during summer months, and violations of the standard occurred over 50% of
the time at the SRSP downstream site. No standard has been set for sediment fecal levels.
In 1996, the sediment populations of FC were on average about 8,000 times greater than the
planktonic counts (see Table 1 and Figure 1), suggesting that Oak Creek sediments are
receiving a tremendous amount of fecal pollution. Several studies in Arizona (Brickler and
Morse, 1979; Brickler et a/., 1976; Doyle et a/., 1992; Tunnicliff and Brickler, 1984) have also
demonstrated that sediment FC populations exceeded water FC populations (but by smaller
ratios; 10:1 to 100:1) and impacted water quality.
PFC
SR
MZC
GP
CD (O CD
o> ci> g>
oo *— if)
S S 3
p: p:
CD
en
(A)
(B)
Figure 1. - Sample locations were plotted vs. time to create a category-based spatial and temporal grid.
Log10 planktonic (A) and sediment (B) FC population/100 ml recovered from Oak Creek were each plotted
on this grid to compare water and sediment fecal pollution. In 'B' FC < 1000 cfu/100 ml are included in the
lowest log order.
A second sampling regime for FC analysis of water and sediment was undertaken in 1996
encompassing sites within and around Slide Rock State Park. The same correlation between
sediment reservoirs and fecal pollution that was observed in the creek-wide study (Figure 1)
was also observed within SRSP. While recreational users have been historically held
responsible for fecal pollution in Oak Creek, agitation of the sediment through recreational use
and by monsoons can cause water quality violations.
During the winter of 1995 a series of BMPs were implemented at SRSP in an attempt to limit
the impact of fecal pollution there. BMPs included an attempt to reduce the number of visitors
to the park, improved access to restroom facilities and a public education program.
69
-------
FC data compiled before (1994 and 1995) and after (1996) implementation of BMPs were used
for statistical analysis. In order to apply standard regression modeling techniques,
transformations of the raw (linear) data are required to strengthen the validity of the analysis.
The transformed data for SRSP upstream and SRSP downstream are denoted by LNSRU and
LNSRD respectively. The transformation was performed by applying: In (x + 0.1). Normal
probability plots for LNSRU and LNSRD showed linearity except for heavy left tails for each of
the transformed data set.
The regression model employed in the analysis is:
LNSRD = b0 + b! LNSRU + b2 BMP + b3 BMP*LNSRU + e
where BMP = 0 if the data were collected prior to BMP implementation and BMP = 1 if the data
were collected during BMP implementation. The standard regression assumptions apply in that
the error term e is assumed to be normally distributed and independent. The analysis of
variance results as well as the coefficient estimates are given in Table 2.
Table 2. Analysis of Variance.
Source Sum-of- DF Mean- F-Ratio P
Squares Square
Regression 338.7659 3 112.9220 37.0635 0.0000
Residual 228.5039 75 3.0467
Effect Coefficient Std Error t P-value (2 Tail)
CONSTANT 1.9905 0.3640 5.4677 0.0000
LNSRU 0.8500 0.1115 7.6235 0.0000
BMP -0.3320 0.5912 -0.5615 0.5761
LNSRU*BMP 0.1190 0.1736 0.6857 0.4950
The P-value associated with BMP and LNSRU*BMP indicate that H0: b2 = 0 and H0: b3 = 0
should not be rejected. This indicates that the BMPs are currently not effective or additional
time is needed to further evaluate their effectiveness.
Regression diagnostics showed no blatant violations of the normality distribution except for the
residuals vs. time plot. This plot appeared to show three groups of data points; corresponding
to the seasonal impact of fecal pollution which occurred during the summer of 1994, 1995 and
1996. This pattern may indicate a correlation between response variables over time which is a
violation of one of the regression model assumptions. Future work will focus on selecting the
best time series model that fits the data.
Several overall conclusions can be made at this juncture:
1. Sediment reservoirs of FC are a significant factor in the observed fecal pollution in the water
column.
2. Recreational use (activity in water) and increased stream flow during monsoonal activity
appear to function as a FC distribution system in Oak Creek.
70
I
-------
3. BMPs have been ineffective in controlling fecal pollution at SRSP.
During 1997, we will be using random-primer PCR to genotype Escherichia coli populations
from known human and animal sources as well as those from Oak Creek to determine the
source(s) of fecal pollution in the creek. This will enable us to recommend BMPs that will truly
enhance water quality in Oak Creek Canyon.
References
American Public Health Association. (1989) Standard methods for the examination of water and
wastewater 17th ed. American Public Health Association, Water Works Association and
Water Environmental Federation. Washington, DC.
Brickler, S. K. and D.W. Morse, III. (1979) Baseline water quality analysis of Madera Creek,
Madera Canyon. Report on Cooperative Agreement 03-05-03-75. USDA For. Serv.,
Coronado National Forest, Arizona, University of Arizona, Tucson, Arizona.
Brickler, S. K. and R.A. Phillips and R.M. Motschall. (1976) A water quality analysis of
recreation waters in Sabino Canyon. Report to Colorado National Forest and Cooperative
Agreement 03-04-1-75, University of Arizona, Tucson, Arizona.
Doyle, J.D., B. Tunnicliff, R. Kramer, R. Kuehl and S.K. Brickler. (1992) Instability of fecal
coliform populations in waters and bottom sediments at recreational beaches in Arizona.
Wat. Res. 26:979-988.
EPA. (1983) Methods for the Chemical Analysis of Waters and Wastes, EPA-600/4-79-020,
U.S. EPA (March, 1983).
Field, R. and R.E. Pitt. (1990) Urban storm-induced discharge impacts: US Environmental
Hansen, O. and R. White. (1992) STORET documentation for menu-driven user interface. US
Environ. Protect. Agency, Region IX, Water Quality Branch, San Francisco, Calif.
Jackson, P. D. (1981) Water quality report, Slide Rock and Grasshopper Point swim areas in
Oak Creek Canyon. Summer 1980. U.S. Forest Service, Coconino National Forest, Ariz.
Mueller, B. C. (1984) Unique water nomination for Oak Creek and the West Fork of Oak Creek.
Ambient Water Quality Unit. Arizona Department of Health Services, Phoenix, Ariz.
Obr, J. E., R. H. Follett and J. K. Kracht (1978) Oak Creek water quality report. Arizona
Department of Health Services in NACOG-ADHS, Phoenix, Ariz.
Rose, J.B., R.L Mullinax, S.N. Singh, M.V. Yates and C.P. Gerba. (1987) Occurrence of
rotaviruses and enteroviruses in recreational waters of Oak Creek, Arizona. Wat. Res.
21:1375-1381.
Tunnicliff, B. and S.K. Brickler. (1984) Recreational water quality analyses of the Colorado river
corridor in Grand Canyon. Appl. Envir. Microbiol. 48:909-917.
71
-------
Communication and Adoption Evaluation of
USDA Water Quality Demonstration Projects
Garrett J. O'Keefe, Peter J. Novyak, Susan Anderson
University of Wisc9nsin-Madison
Madison, Wisconsin
Claude F. Bennett
CSREES-USDA
Washington, D.C.
on, Wisconsin
Craig Trumbp
Cornell University
Ithaca, New York
Increasing concern over the Nation's water quality led to a 1988 Presidential Initiative to protect
surface water and groundwater from pollution by fertilizers, pesticides and agricultural wastes.
Responding to this Initiative, in 1989, the U.S. Department of Agriculture (USDA) and its state
and local cooperators launched a Water Quality Program. The Program's goal is to provide
farmers and ranchers with the knowledge and technical means to respond independently and
voluntarily in addressing on-farm environmental concerns and related state water quality
requirements, while maintaining agricultural productivity and profitability.
The Water Quality Program of USDA and state'cooperators brings a new focus to the goal of
protecting the Nation's water resources from pollution. This new focus emphasizes contaminants
from agricultural sources; and it promotes interagency coordination, collaboration and program
integration to achieve program objectives.
To achieve its objectives, USDA's Water Quality Program brings together three interrelated
components: (1) Education, Technical and Financial Assistance, (2) Research and Development,
and (3) Database Development and Evaluation. Water Quality Demonstration Projects in 16
states across the Nation are part of the Education, Technical and Financial Assistance component.
This evaluation focuses on the eight demonstration projects established during fiscal year 1990 in
California, Florida, Maryland, Minnesota, Nebraska, North Carolina, Texas, and Wisconsin.
The Water Quality Demonstration Projects are designed to accelerate voluntary adoption of
agricultural best management practices (BMPs) that protect surface and groundwater, while
maintaining farm and ranch productivity and profitability. Objectives are to:
encourage producers to more quickly adopt cost-effective use of inter-related
pesticide, fertilizer, and irrigation BMPs that can substantially reduce agricultural
pollutants, and
show how quickly and effectively farmers and ranchers can voluntarily modify their
use of BMPs to prevent or reduce the pollution of surface and ground water.
In line with these objectives, this paper summarizes an evaluation of the ability of the 1990
demonstration projects to accelerate producer adoption of BMPs during 1992-1994, the first two
73
-------
years of full operation of the projects. All of the 1990 projects are scheduled for completion by the \
close of FY 1998, with two projects already having been concluded.
Evaluation Components '
The producer adoption evaluation examines producers':
i
changes in awareness of, familiarity with, assessment of, and use of BMPs ,
recommended by the demonstration projects; and
awareness of, participation in, and reactions to the projects.
Three to four of the several BMPs recommended by each of the demonstration projects were ,
designated for adoption process surveys, resulting in examination of a total of 34 BMP cases. ,
The BMPs sampled from the project-recommended BMPs varied in their requirements for extent of >
managerial labor, capital investment and risk. '
Effectiveness of the projects in the demonstration areas ("watersheds") was evaluated through t
use of a quasi-experimental design, comparing producer adoption within the eight demonstration
project areas with producer adoption within nearby matched comparison areas. Surveys of
representative groups of producers were conducted during 1992 and 1994, in both the
demonstration areas and the comparison areas. Comparing the adoption rates of the designated ]
BMPs, between the demonstration areas and their respective comparison areas, allows i
identification of area impacts of the projects. Effectiveness of the projects among project ,
participants was evaluated through use of a simple longitudinal design, examining changes '
among project participants in the adoption variables measured from 1992 to 1994. '
Project effectiveness at the individual producer level was evaluated as well through use of
cross-sectional design, examining correlations between individual producers' characteristics and
their 1994 status with respect to the adoption variables.
Extensive interviews were conducted with demonstration project staff regarding their project
organization, planning and evaluation activities; methods for conducting field demonstrations; and
methods for conveying information and education to producers. These interviews help to interpret
the findings of the adoption surveys.
Program Effort Findings
The eight projects varied considerably in their organization, planning and evaluation, use of
demonstration methods, and in their complementary use of other information and education
methods. Most demonstration projects were unable to assess producers' ad9pter characteristics
prior to their initiation, but most did gather feedback from producers once projects were underway.
Some projects allocated resources to testing the local applicability of BMPs proven effective
elsewhere; and, in some projects, the BMPs initially chosen for evaluation received lower priority
emphasis from 1992 to 1994 than initially planned in 1991.
Some projects were more adequately staffed than others with communication professionals,
and/or made stronger efforts at targeting their audiences with pertinent information and education
strategies. These tended to have more significant increases in BMP awareness, familiarity and
adoption, but not withput exceptions. Again, the need for local BMP testing and changes in BMP
priorities limit the building of inferences about communication impact here. One-on-one
communication with producers was the most emphasized method of information transfer,
supplemented by local media use.
-------
The demonstration projects appear to have operated in a communication environment loaded
perhaps to the point of saturation with agricultural information. It is no small accomplishment for the
projects to have gained the producer recognition and impacts indicated. In some cases, the
complexity of Federal-State relationships among the collaborating USDA and State Cooperator
Agencies made project tasks more difficult to efficiently carry out. USDA funding uncertainties,
lags, and late notification of project extensions to varying extents affected all projects, resulting in
decreased momentum and personnel losses.
Adoption Findings and Conclusions
The project-area analysis examines the extent to which producers in the demonstration
areas—averaging across those who participated directly in the project and those who were
non-participants—made gains with respect to the adoption variables. Demonstration area
producers as a whole:
became more aware of most of the BMPs-with statistically significant increases
among 21% of the 28 BMP cases examined;
became more familiar with most of the BMPs-significantly so for 40% of the 28
BMP cases examined;
did not change their assessments of the BMPs, among the 28 cases examined;
and
increased use of the BMPs in fewer than half of the 26 BMP cases examined-with
only 19% of the increases being statistically significant.
Producer awareness of most of the BMPs was already high at the onset of the demonstrations,
so the bigger gains in familiarity than in awareness are expected. At the onset of the projects in
1992, an average of about 25 percent of producers were already using the designated BMPs.
Thus most of the BMPs examined had already moved through the innovator category of
producers and well into the early adopter category. Among the BMP cases where statistically
significant gains by 1994 were found, BMP awareness, familiarity, and/or use increased among
5% to 23% of producers, with a median increase of 15%.
These percentages suggest that over the first two years of full operation of the demonstrations
projects, overall diffusion spread into the category of early majority adopters. Demonstration area
producers' awareness of, familiarity with, and uses of the BMP? made net gains relative to the
matched comparison areas in only a few instances. One reason is that there were several
occurrences of gains in BMP awareness, familiarity, and use in the comparison area as well. The
low frequency of net gains limits a clear inference that the demonstration projects influenced BMP
adoption variables at the area level. Several factors could account for these gains, including
non-demonstration project coverage of the BMPs and "overflow" of demonstration project
information to the comparison areas. The project participant analysis is underway, with findings
not yet available.
The analysis for individual producers examined whether their management characteristics reflected
their status regarding adoption of the project-recommended BMPs. Producers in the
demonstration and comparison areas who indicated that they had received and attended to
information about the designated BMPs during 1993 were significantly more familiar with them in
95% of the 19 BMP cases examined, and more likely to be users of them in 53% of these cases.
This finding suggests the possibility of demonstration project influences on adoption of
project-recommended BMPs.
-------
The evaluation findings suggest that the demonstration projects were a substantial information
force within the agricultural information system of the demonstration areas. A strong majority of
demonstration area producers were aware of the projects, but adoption of environmentally sound
agricultural practices has been previously shown to be a generally slow process. Previous
research shows that producer adoption of conservation/environmental practices is apt to occur at
a slower rate than for practices that are more clearly economically advantageous.
In addition, two other factors may have attenuated major project impact: (a) the above-noted
need for some projects to delay dissemination efforts until they tested the local applicability of
BMPs; and (b) the lowering during project implementation of priority accorded by some projects to
the BMPs designated for the evaluation. In this context, the increases in BMP awareness,
familiarity, and use indicated here during the first two years of project implementation appear
encouraging. Increased rates of adoption might be detected by subsequent data-gathering. This
evaluation did not examine the separate influences of financial and technical assistance on
adoption.
In any case, it does appear that the adoption processes here are slow and deliberate, too much
so for conclusive inferences to be drawn over a two-year period. Whether the above results are
encouraging enough for subsequent study of these same projects depends greatly upon the
conduct of the projects since 1994, current and projected USDA and state-level policies and
priorities regarding the BMPs being addressed, and the need for project evaluation for USDA
program accountability and improved program management.
Recommendations
Future programs by USDA and its cooperators to accelerate the voluntary adoption of BMPs that
focus on water quality should take more advantage of contemporary research on adoption
processes and on effective information transfer and communication programs. Also recommended
is greater consensus across projects in determining: (1) what constitutes realistic, locally tested
BMPs with respect to agronomic, economic and water resource protection advantages; (2) the
nature, conduct and effectiveness of field result demonstrations; and (3) the nature and conduct of
other information and education strategies. These include use of:
clearly measurable project goals within a national program framework;
comprehensive models of information and education strategy and tactics, based
upon experience and research in related settings;
basic communication planning tools in program design and execution, including
formative evaluation in order to appropriately segment producers by their existing
characteristics and needs;
adaptive research to assure that the practices being recommended are relevant
and economically viable to local producers; and
justification, integration, and focus of project funding to achieve project objectives.
The above elements require staff trained in information transfer and education, as well as in the
technical aspects of practice testing and application. The elements also require project resources
committed to information and education program planning, design, production and evaluation.
Equally important, these elements require consistency of vision across time and circumstance.
This can be quite difficult to achieve in a multi-Agency effort faced with shifting personnel
76
-------
assignments, uncertain budgetary futures, and changing leadership roles. This evaluation also
provides recommendations for USDA's management of its future water quality programs, in such a
way as to increase effectiveness and administrative efficiency of water quality projects funded
by USDA. These recommendations are as follows:
USDA should emphasize and financially support site-specific, adaptive research
as an integral part of projects to accelerate producer adoption of water quality
BMPs;
USDA/CSREES should direct water quality staffs and extramural funding in order
to achieve distribution of funds to continuing projects before the midpoint of the
Federal Fiscal Year; and
USDA and its component agencies, including CSREES, should develop and
adopt a policy on the duration of Federal funding for future state water quality
projects.
This evaluation has been an innovative, complex undertaking. The USDA Water Quality
Program, and other national, multi-Agency initiatives will increasingly need sophisticated
evaluation strategies. Such future evaluations will require streamlined inter-Agency staff and
budget administration, increased interaction between evaluation and project staff, increased
precision of project objectives, and reassessment of techniques for evaluating producer adoption.
77
-------
Gum Creek Water Quality Demonstration Project
Kenneth Lewis, County Extension Coordinator
University of Georgia Extension, Cordele, Georgia
Wayne Adkins, Extension Engineer, Water Quality
University of Georgia Extension, Cordele, Georgia
Joel Wood, District Conservationist
Natural Resources Conservation Service, Cordele, Georgia
The Gum Creek Water Quality Demonstration Project is located in Crisp and Dooly Counties, Georgia.
The watershed is located in the Southern Coastal Plain land resource area. Gum Creek drains into Lake
Blackshear which was constructed on the Flint River in 1930 by Crisp County to provide electricity for
the County. The 8,250 acre lake is heavily utilized for recreational activities. Ninety percent of the land
area in the watershed is utilized in the production of food, fiber or forest products. Most of the city of
Cordele is located within the watershed boundaries, and the city's waste treatment plant discharges into
Gum Creek.
Lake Blackshear was classified as eutrophic in a 1974 EPA eutrophication survey of Georgia lakes.
During the period of 1974 through 1993 the lake was ranked in the top 8 lakes in the state for trophic
state index values. In two of these years Lake Blackshear had the highest readings of any tested lake in
the state. Gum Creek is identified in the December 1989 Georgia Nonpoint Source Assessment Report
and the Georgia Nonpoint Source Management Plan as an agricultural stream likely to be threatened by
agricultural nonpoint sources of pollution.
The watershed is approximately 53,000 acres. Of this, some 25,000 acres are in agricultural production,
23,000 acres are in woodlands and the remainder is in urban/built-up areas including industry, roads and
railroads. Major crops grown include cotton, peanuts, corn, small grains , soybeans, watermelons, pecans
and pasture. Many of these crops were being grown with high fertilizer and pesticide inputs. The area
also has several cattle and hog operations.
The main objective of the project was for farmers to implement cost-shared Best Management Practices
(BMP's) designed to reduce pollution and/or the potential of pollution of surface and ground waters in
the project area while maintaining farmer productivity and profitability. Other objectives included:
increasing landowner knowledge and understanding of agricultural pollution potentials and water quality,
increasing crop production efficiency through better management of natural resources, increasing
awareness of the general public on surface and groundwater contamination and to initiate a state-
administered cost-share program for agricultural BMP's.
The project was designed and supported by some 23 cooperating federal, state, and county governmental
agencies. A technical committee was formed representing most of the cooperating agencies. The
committee had the charge of designing and approving technical aspects of the project including BMP
79
-------
selection and design and monitoring activities. The committee met regularly for the purpose of project
updates and planning. Technical support for BMP's and the education of cooperating farmers was
charged to local NRCS and Extension personnel. BMP's were implemented by individual farmers on
either a 60% cost-share basis or established payments.
Best Management Practices
Cost-sharing for some 25 separate Best Management Practices was offered to farmers in the project area.
These include:
Permanent structures designed to reduce surface water contamination by acting as nutrient or pesticide
sinks and settling areas for sediments.
• Shallow Water Impoundments • Water Impoundment Reservoirs
• Sediment Basins • Water & Sediment Control Basins (Gully Plugs)
Permanent structures designed to act as physical barriers to prevent contamination of soil, groundwater or
surface water from pesticides and/or nutrients.
• Well Head Protection
• Permanent Chemical Mixing/Loading/Storage Facilities
• Portable Chemical Mixing/Loading Systems
Permanent structures designed to reduce sediment, pesticide and nutrient loading of surface water by run-
off water management.
• Terraces • Vegetative Filter Strips
• Grassed Waterways • Conservation Cover Crops
• Riparian Forest Buffer Strips • Forest Road Culverts
• Artificial Wetland Construction
Permanent structures designed to prevent access of cattle to streams thus reducing nutrient loading from
waste products and sedimentation caused by bank erosion.
• Livestock Fencing • Alternative Water Supplies
• Alternative Water Sources
Permanent equipment designed to reduce the number of pesticide applications or increase the efficiency
of such applications and/or irrigation water applications.
• Envirocaster Peanut Leafspot Advisory & Irrigation Scheduling System Hardware
• Watermark Soil Moisture Blocks and Meters
Management activities designed to reduce the application of pesticides, nutrients, irrigation water and/or
increase the efficiency of such applications.
• Integrated Crop Management (ICM) through - soil testing, tissue analysis, nematode sampling,
insect scouting, pesticide record keeping, irrigation water management
Management Activities designed to reduce surface water loading of nutrients, pesticides and sediments
through cropping systems.
80
-------
• Crop Residue Use • Conservation Tillage
• Green Manure Cover Crops
Certain practices were strongly encouraged but did not qualify for cost-share payments or incentives.
• Pesticide Container Recycling • Contour Farming
• Tree Planting
Monitoring Activities
A monitoring plan was devised by the technical committee that was a compromise between what was
desired for good science and what could be afforded under a sparse monitoring budget. Automated
sampling devices were desired but were not within the budget constraints. The plan revolved around
periodic spot sampling of surface water, shallow test wells, pan collectors, and private wells. Surface
water samples were taken primarily after rainfall events. These were taken at 7 established monitoring
stations throughout the watershed. Twelve shallow test wells ranging in depth from 10 feet to 20 feet
were installed down-slope of row-crop fields and in pecan orchards to monitor shallow groundwater.
Samples were withdrawn on a periodic basis. Pan collectors were constructed and placed at the edge of
cotton fields to collect runoff exiting fields from furrows. Collectors were placed in both fields being
farmed by conventional tillage methods and those being farmed with strip till for comparison purposes.
Collectors were emptied after rainfall events. Drinking water and irrigation wells were tested
periodically throughout the project period. An effort was made to sample one-quarter of the wells in the
watershed. Samples were analyzed at the: University of Georgia Ag Services Lab, Athens, Ga; Georgia
Environmental Protection Division, Atlanta, Ga; or the USDA Agricultural Research Lab, Tifton, Ga.
Samples were analyzed for both pesticides and nutrients. In addition to laboratory testing, biological
sampling and assays also were performed periodically on Gum Creek and several tributaries.
BMP Participation
The main objective of the project was to encourage farmers to voluntarily initiate Best Management
Practices. From this standpoint the project was very successful. Cost-share contracts were written with
31 farmers on 12,000 acres in the project area. This represents 98 % of all full time farmers and 48 % of
all cropland in the watershed area. Nineteen of the 23 BMP's were initiated on at least one farm. Most
were initiated on multiple farms. The highest number of acres were enrolled in management activities.
Annual enrollment figures on the most widely adopted practices were as follows: 7980 acres in ICM,
7129 acres in crop residue management, 4168 acres in irrigation water management, 2759 acres in
conservation tillage, 1224 acres in green manure cover crop. Permanent structure installation highlights
are as follows: 3 water holding facilities (ponds), 200 wellheads curbed, repaired or upgraded, 13
portable chemical mixing stations, 1 permanent chemical mixing/loading/storage facility, 7938 feet of
livestock fencing, 1 alternative livestock water source and 1 livestock water supply and 10 miles of
terraces.
Monitoring Results
From 1990 until 1996 approximately 1000 water samples were taken. These were analyzed for nutrients,
14 minerals and 46 pesticides.
A total of 275 water samples were analyzed from field pan collectors during 1993 through 1995. Except
for water collected during one rainfall event in 1993, chemicals were generally below detection levels for
all compounds except fluometuron (Cotoran, etc.) and norflurazon (Zorial). Fluometuron was detected in
41 % of samples. Levels ranged from 0.7 ppb to 78.3 ppb with an average of 14 ppb. Norflurazon was
81
-------
detected in 11 % of samples. Levels ranged from 3 to 114 ppb with an average of 36 ppb. The highest
levels of both norflurazon and fluometuron were detected exiting strip-till fields. Runoff accumulated
more frequently in conventional tillage fields.
Some 273 water samples were collected from Gum Creek and tributaries for 1990-96. Analysis detected
no agricultural chemicals. The fact that fluometuron and norflurazon often were detected at field edges
and not in streams indicates an effectiveness of the filtering ability of vegetation between the edge of
fields and streams.
Using phosphorous and nitrates as indicators, the overall water quality of samples taken from streams
could be classified as fair to good. Phosphorous levels ranged from 0 to 0.38 ppm. Eighty percent of
samples were below 0.2 ppm. Nitrate levels ranged from non-detectable to 3.8 ppm. Sample stations in
the upper and middle portions or the watershed generally had nitrate levels of 1.0 ppm or below. The
highest nitrate and phosphorous levels were consistently detected at two sample stations in the lower
portion of Gum Creek. One was located just downstream of the City of Cordele waste treatment facility.
The other was downstream from a non-participating hog operation sited by EPD for excessive animal
units per acre and subsequently forced to close for economic reasons.
178 water samples were analyzed from shallow test wells located at the edge of row crop fields and in
pecan orchards. Immunoassay analysis of atrazine and alachlor detected low levels of these chemicals
throughout the study. However, these chemicals were applied prior to the beginning of the study. The
most common herbicides in the project area, trifluralin, fluometuron and norflurazon were rarely found.
Nitrate levels in the test wells often exceeded EPA safe drinking water standards. Wells located in pecan
orchards generally had levels of 5 ppm or below. Wells showing highest levels of nitrates (up to 23
ppm) were down slope of large irrigated fields (100 + acres). These fields were in a 2 years of cotton and
1 year of peanuts rotation pattern.
A total of 110 water samples were taken from irrigation, residential and farm wells. None were found to
contain any detectable levels of pesticide. Nitrates exceeding EPA guidelines of 10 ppm were found in
only 3 of the 110 wells. Two of these were traced to improperly constructed wells and home sewage
problems. The third was attributed to an inadequately protected well head and a large fertilizer spill.
Habitat assessment and habitat monitoring was conducted at 5 locations throughout the watershed on 3
different dates from 1992-1994. The conclusions from these studies appear to correlate well with the
results of analytical testing. Habitat quality was lowest near the sampling stations showing consistently
higher nitrate and phosphorous levels.
Water quality monitoring in the Gum Creek basin did not reveal any components that exceed EPA or
Georgia EPD limits. Samples collected at the edge of cotton fields often had low to moderate levels of
herbicides, but they were not detected in Gum Creek and tributaries. Water from shallow groundwater
test wells often contained nitrate nitrogen levels in excess of 10 ppm. However, these elevated levels
were not. detected in streams or (with the exceptions noted previously) in drinking water or irrigation
wells.
Management Practice Modifications
Significant and measurable reductions in pesticide and nutrient applications occurred during the project
period. When the project began cotton farmers averaged spraying for "cotton worms" 6 to 8 times per
season. During the project period farmers began to rely on beneficial natural parasites and predators. As
a result, during the 1995 and 1996 growing seasons, area farmers sprayed cotton on the average of 2.5
times with no decrease in yield or fiber quality. Pecan growers in the project area are spraying less.
Hard lessons have been learned concerning yellow aphid sprays. Improperly timed sprays simply
aggravate the problem. Most growers now wait on natural parasites, predators and diseases to crash
82
-------
damaging levels of aphids. Estimates are that due to ICM activities, average nitrogen use decreased by
some 6.4 Ib/acre and phosphorous applications decreased by some 1.7 Ib/acre during the project period.
Spin-off Projects and Lessons Learned
Well surveying in the project area revealed that most of the well heads were not properly protected by
grouting and concrete pads (curbs), with over half having cavities around the casing. Curbing was
being left to well owners and was not being done. Beginning in July 1996 an educational program was
initiated. Gum Creek personnel curbed 40 wells for farmers to generate interest in the program.
Technical assistance was then offered to rural homeowners and farmers to help them curb their own
wells. Local well drillers were approached to investigate the possibility of incorporating curbing as part
of a well drilling package. On April 22, 1997 the 100th well was curbed with either Gum Creek
personnel labor or technical assistance. During the same time about 75 additional area wells were curbed
and properly protected by local well drillers. Four of the five well drillers in the two county area now
routinely curb all wells they drill.
Disposal of plastic pesticide containers is a major problem for many Georgia farmers and local county
governments. Because of the Gum Creek project, the Georgia Department of Ag and University of
Georgia Extension Service initiated a pilot pesticide container recycling program in Crisp and Dooly
counties. To date, more than 90,000 Ib of pesticide containers have been collected and recycled.
Because of the lessons learned and success of the program in the pilot area, container recycling is now a
state-wide program in Georgia. This program had major influence on the initiation of a Georgia Clean
Day program that has recovered and disposed of tons of unusable pesticides statewide.
The promise of precision agriculture sparked great interest among Gum Creek personnel and farmers. A
Navstar Global Positioning System receiver was purchased with local funds. Soil nutrient mapping was
initiated in the project area and selected farms in Crisp County. One acre grids were utilized. During the
winter of 1997, 2,200 acres were gridded and maps generated detailing distribution patterns of soil pH,
phosphorous, potassium, magnesium, calcium, manganese and zinc. Maps were generated to show the
distribution of recommended nutrients as per University of Georgia recommendations. As a result of soil
mapping efforts, a long term precision agriculture research project to be conducted by UGA scientist has
been initiated.
Perhaps the greatest lesson learned from the project is that profitable South Georgia agricultural
enterprises can co-exist with a healthy environment through the use of agricultural Best Management
Practices. BMP's can be highly effective in reducing pesticide and nutrient contamination of both
surface and groundwater. Farmers are willing to install and implement economically and
environmentally sound BMP's.
Another important lesson learned from the Gum Creek Project is the value of inter-agency cooperation
and teamwork.
83
-------
The Implementation of Innovative Best Management Practices
in the Sny Magill Watershed
Eric A. Palas
Department of Agronomy, Iowa State University, Postville, Iowa
Jeff Tisl
Natural Resources Conservation Service, Elkader, Iowa
The Sny Magill Watershed Project is an interagency effort to improve water quality through
voluntary changes in farm management practices. The project provides technical assistance,
information and education, and cost share assistance to producers within a 22,780 acre agricultural
watershed. The project is designed to reduce sediment, nutrient, and pesticide delivery to Sny
Magill Creek, a coldwater trout stream located in Clayton County, Iowa.
Background
Since 1991, a diverse selection of Best Management Practices (BMPs) such as Integrated Crop
Management (ICM), terraces, water and sediment control basins, and stripcropping have been
successfully applied by the majority of landowners within the watershed (Table 1). Iowa State
University Extension (ISUE) reports that pesticide and nutrient loading has been reduced on
45% of the cropland acres through the delivery of an ICM assistance program. The Natural
Resources Conservation Service (NRCS) estimates that sediment delivery to the stream has
been reduced by over 40%. These achievements are significant, but because of the
watershed's size, many acres were being left untreated.
Table 1: BMPs Applied 1991-1997
Item
Conservation Cropping (Rotations)
Conservation Tillage
Contour Farming
Grade Stabilization Structures
Integrated Crop Mgmt.
Nutrient & Pest Mgmt.
Pasture and Hayland Management
Streambank Protection
Terraces
Timber Management Plans
Water & Sediment Control Basins
Unit
ac.
ac.
ac.
no.
ac.
ac.
ac.
ft.
ft.
ac.
no.
Amount
2,957
1,940
1,447
88
3,095
2,639
828
655
249,540
534
56
In the mid-1990's, the local project coordinators attempted to identify the barriers preventing
additional landowners from adopting the selected BMPs. Through one-on-one inquiries, group
meetings, and questionnaires, it became apparent that most of these barriers centered on
85
-------
economic concerns. Many landowners noted the environmental benefits of the selected BMPs,
but were reluctant to adopt them due to the direct costs involved. '
i
The agencies worked to develop alternative cost-effective BMPs which would overcome many of |
the identified barriers, while meeting the overall project objectives of reducing sediment, nutrient,
and pesticide delivery to the stream. Primary efforts were focused on the areas of streambank
stabilization and nutrient and pest management.
Development of ISUE's Nutrient and Pest Management Incentive Education Program |
It is recognized that improving the environmental sustainablility of agriculture will require producers '
to increase the intensity of their management. For crop nutrient and pest management, there are
well established refinements which publicly supported projects in Iowa have documented to
reduce excess loading of agricultural chemicals, increase the use of on-farm resources, and
maintain or increase profitability.
Existing financial incentives for nutrient and pest management often depend on expert assistance
provided to cooperating producers, and lack an educational component. Producers receiving
incentives hire crop consultants to "deliver" plans, scouting reports, and recommendations. At the
beginning of the Sny Magill project, an integrated crop management (ICM) program was offered.
Involved producers enrolled over 3000 acres, and reduced nitrogen applications by nearly 40000
pounds during the 1994 crop year. While successful from a reduction standpoint, in-progress
project surveys showed little evidence that this program changed producers' long term attitudes
about more sustainable management.
The Nutrient and Pest Management Incentive Education Program is a local initiative developed
by staff of the Sny Magill Creek HUA and Northeast Iowa Demonstration projects. To enhance
long term adoption of practices, it requires participating farmers to learn the basics of managing
their own nutrient and pest management programs. The program is targeted to an area where
private consultants are either not available or are unwilling to serve small crop acreage farms,
including many livestock operations. By moving through the program in a series of workshops
with a group of 8 to 10 participants, the program also establishes a peer support mechanism.
Workshop sessions involve reading soil maps and soil test reports, fertilizing for realistic yield
goals, as well as determining manure inventories and legume and manure fertility credits. The
program allows cooperators to develop, write, and implement their own nutrient, manure utilization,
and pest management plans.
Producers receive incentive payments of $1/acre up to a $250 maximum payment for: 1) all crop
acres in the management program; 2) all acres covered by their manure distribution plan; and 3) all
acres covered by end-of-year field records, according to a record system set up by the project.
Additional payments are made at the end of the second and third years for records and for
completing an annual survey of nutrient and pest management practices. Pest management
planning and basics of field scouting are also introduced. In all, each participant may earn up to
$750 in the first year, and $250 in each of years two and three of the program. The maximum
payment levels were set high enough to compensate for time spent completing program
requirements and to provide some incentive, such as partially offsetting required soil test costs,
but low enough not to compete with private crop consultants' fees.
86
-------
Project staff are available to consult with participants. A biweekly newsletter with useful
scouting, field, and pest management information is distributed to participants during the cropping
season. Ten to twelve issues are normally provided to participants. An economic evaluation of
changes made is part of the end-of-year workshop. This approach is meant to give producers
the knowledge and confidence to control their own nutrient and pest management programs,
whether they choose to do it themselves or to work with suppliers or consultants in the future.
This program pays incentives for performance of specific program components. A flexible
payment source is a critical component of this program, as timing and amounts of payments vary
from one producer to another. Traditional agency programs have been set up to provide one
payment per year to producers.
Since 1994, sixty-six producers in three northeast Iowa water quality project areas have
participated in this program. Ten Sny Magill area producers have been enrolled. Because of the
relatively small numbers of participants in this pilot program, evaluation data (baseline and annual
post-workshop surveys) have been aggregated for all three water quality projects. The three
projects represent different watersheds in the same topography, with a similar mix of producers,
farm resources, and environmental concerns.
Comparison of baseline and annual surveys of program participants show increased confidence
in their own ability to manage fertility programs (rather than relying on suppliers), reduced use of
purchased fertilizers, and improved manure management. Survey responses from 25 producers
completing the first year of the NPMI program are as follows:
• 75 percent indicated they reduced nitrogen use.
• 76 percent indicated it was profitable to reduce N use. The improved net income ranged
up to $5,500 per farm.
• 67 percent changed manure management. Fifty-six percent of those changing manure
management did so by spreading fields that could benefit most from manure. Others
noted that they spread manure more uniformly to gain more nutrient credit, or spread more
acres when they realized that they were overapplying nutrients.
• 70 percent of the NPMI producers plan to change their insect or weed management
programs. Reasons for adjustments include seeking improved weed control with improved
timing of spray application, self spraying to avoid custom applicator skips, and seeking
increased flexibility with products to match weed problems.
Responses to open-ended survey questions indicate these practices are profitable:
• "Taking credit for the manure application saved $1000 in fertilizer".
• "Continuing to back down nitrogen rates on sod to corn saved $480 on my farm".
• "Changing the rate of manure per acre after calculating crop needs and credits increased
the profit on this farm by $1400 to $2000".
87
-------
The Nutrient and Pest Management Incentive Education program is being evaluated on its
effectiveness to: 1) Improve or maintain water quality by refining the use of pesticides and
fertilizers, including manure and legume credits; 2) reach the low acreage, diversified young
farmers that have the greatest potential to effectively use on farm nutrients, thus reduce off farm
fertilizer purchases and as a result, improve farm profit; 3) provide a cost effective incentive
program that can be efficiently delivered and evaluated for performance. The goal is to insure that
publicly funded management incentives result in sustainable crop management practices that
participants will maintain throughout their farming careers.
This program currently represents the primary focus of alternative nutrient and pest management
efforts. In order to further reduce sediment load, additional emphasis was placed on. innovative
streambank stabilization demonstrations.
Alternative Streambank Stabilization Initiative
The Project offers landowners many alternative BMPs for reduction of sedimentation resulting from
rill, sheet, and gully erosion. For example, if landowners cannot afford tile outlet terraces, other
equally effective sediment reducing BMPs such as contour stripcropping could be offered. When
working with streambank erosion problems, however, our available BMP list was limited to
effective, albeit expensive, rock rip-rap based technologies. This problem is not isolated to this
project. Several projects across the country are faced with this situation. Often landowners with
farms along streams simply cannot justify spending their limited income, even with cost share, to
benefit and protect such a small area.
Unfortunately, recent studies in Illinois indicate that as much as 60% of the sediment found in
small-to-mid size Midwestern streams is derived through streambank erosion. Even though
opinions differ on how to resolve streambank erosion problems, few fail to recognize its
significance. Everyone involved with the Sny Magill Watershed Project realized that to make
significant sediment reductions, the issue of streambank stabilization would have to be j
addressed. The use of traditional rock rip-rap based technologies was not the answer since so !
few landowners could afford their use. A more cost effective approach needed to be found. j
In 1994, local project coordinators began exploring the use of soil bioengineering. The basis of
this technology is the use of plant materials in conjunction with limited structural materials.
Together, they provide a sound streambank stabilization installation, but utilize less expensive I
materials. |
Working with NRCS specialists, the local project coordinators invited several federal, state, and |
county agencies to work together, and have subsequently applied a series of installations to
learn more about the uses and limitations of this technology. In the spring of 1995, the first j
installations using willow posts, fascines, and brush mattresses were installed. More were
installed using slightly different combinations in the fall. These installations were completed on a 6
foot high bank and cost approximately $12 per lineal foot, or 1/3 the cost of traditional rock rip-rap !
based methods, and have been inundated by floods at least 4 times with no apparent damage to J
the streambank. !
88
-------
In the summer of 1996, an installation using warm season prairie grasses was attempted. Both
willows and warm season grasses possess deep penetrating root systems, which hold loose
soil in place. Warm season grasses are an alternative for landowners that fear encroachment of
willows into cropland. A major problem with warm season grasses, however, is their slow initial
growth. Several species take two full growing seasons to become established. Therefore, in this
installation, bio-degradable mulch mats were utilized. The mats should hold the loose soil for two
to three years, allowing the warm season grasses the time they need to grow and form a dense
sod layer. The cost of this installation on a 5 foot high bank was $8 per lineal foot, or 1/7 that of
traditional rock rip rap methods.
The use of soil bioengineering technologies does have its limitations. Since the primary
stabilizing measures are living, special care must be taken to ensure their survival. Willows and
grasses must be frequently scouted for insect, beaver, and ice damage. Failure to conduct
periodic inspections and perform required maintenance could lead to failure of the entire installation.
One of the primary instability problems associated with the streambanks is the traffic patterns
created by heavy fishing pressure. One of the willow installations failed to survive because
anglers simply trampled it. For this reason the use of vegetation was abandoned when the most
heavily fished stretch of the stream was stabilized. While surveying, it was noticed that a
significant number of anglers using this stretch of the stream were of limited mobility. As a result,
handicap accessibility was incorporated into the project designs. Traditional installations would
have included concrete and asphalt, which were not used due to high cost and maintenance
considerations. Items such as rock gabion baskets and crushed limestone were used which will
allow proper access, have fewer maintenance concerns, and remain cost effective.
Plans are underway to install a series of low stone weirs across the stream in the fall of 1997.
The weirs will be used to help reestablish a stable pool and riffle sequence, which will not only
help stabilize the channel grade and streambanks, but will significantly increase the in-stream
habitat potential of the stream.
Lessons learned
The Nutrient and Pest Management Incentive Education Program and the Alternative Streambank
Stabilization Initiative represent innovative local efforts to improve water quality.
Providing expert assistance to write crop management plans and recommendations for the
producer does not ensure that they will continue practices on their own. By including an
educational component, the NPMI program allows cooperators to develop, write, and implement
their own nutrient, manure utilization, and pest management plans. This provides for long-term
adoption of practices.
The Alternative Streambank Stabilization Initiative is an effort to develop and demonstrate new,
lower-cost Streambank stabilization technology. The evaluation of these demonstrations is just
beginning, as the installations have been in place less than two years. Efforts focus on
installations that can be completed with minimum cost and equipment requirements. The sediment
load to Sny Magill Creek is directly impacted by these efforts, while indirect results occur from the
technical guidance that is applied to watersheds in other areas.
89
-------
Nitrate Losses Under Various Nitrogen Management Systems
J. Kent Mitchell, Sharyl E. Walker, and Michael C. Hirschi
Department of Agricultural Engineering, University of Illinois at Urbana-Champaign
Nitrate in sub-surface tile flow have been monitored for four years from fields with
various tillage and cropping management practices. The effect of the application of
large amounts of nitrogen fertilizer, particularly as a pre-plant operation, was shown in
the nitrate-N concentrations from tile drains. The pre-plant anhydrous-N application
systems with average nitrogen application of 110 kg/ha/yr. had a mean concentration of
nitrate-N of 9.4 mg/L. The mean concentration of nitrate-N from a permanent meadow
field was 1.0 mg/L.
Introduction
The overall goal of the Little Vermilion River Agricultural Nonpoint Source Hydrologic
Unit Area Project is to reduce the levels of nitrate and pesticides entering Georgetown
Lake. To accomplish this goal, the Cooperative Extension Service (CES), Natural
Resources Conservation Service (NRCS) and Food Security Administration (FSA) are
encouraging the adoption of integrated crop management (ICM) practices throughout
the watershed. Besides helping improve the quality of water in the lake, these practices
should also help prevent degradation in aquifers that serve private wells.
The objective of this research is to determine if selected management practices can
eliminate, reduce, or retard the movement of nitrate to ground water and streams.
Studies are being conducted in the Little Vermilion River Watershed on several fields
that have various practices, including fertilizer and pesticide management systems,
buffer strips and wetlands.
Procedures
The Little Vermilion River Agricultural Nonpoint Source Hydrologic Unit Area is located
in East Central Illinois and includes 48,900 hectares in parts of Vermilion, Champaign,
and Edgar counties. River water quality and sub-surface drain tile flow and water
quality, as well as other practices, have been monitored in the watershed. Eight small
sub-surface drainage systems were selected that have the exact extent of drainage
known. Seven of the sites are in corn-soybean production in various combinations,
while an eighth site is permanent meadow. Tillage practices represented by the seven
sites under production include no-till, reduced tillage and conventional tillage. Nitrogen
applications for the seven sites are listed in Table 1.
91
-------
Table 1. Annual and Mean Nitrogen Applications, kg/ha.
Pre-Plant
Application
Side Dress
or Manure
Site
Reduced-Till
Corn-Beans
Reduced-Till
Beans-Corn
Seed Com
White Corn
Mean
Conventional
Till Beans-Corn
No-Till Com-
Beans
Corn Silage
Mean
Crop Year*
1991-92
0
264
207
190
165
219
0
125
110**
1992-93
220
26
0
0
62
0
155
137
97
1993-94
0
237
186
174
149
206
0
31
79
1994-95
224
24
0
0
62
0
183
125
106
Annual
Rotation
Mean
111
138
98
91
110
106
85
100**
97
*Crop year is defined from harvest to harvest.
"The 1991-92 nitrogen application for Corn Silage is not included in the annual rotation mean
that site has a three year rotation period.
because
The soils are predominantly Flanagan silt loam and Drummer silty clay loam at a
location containing two sub-surface drainage systems of approximately 6.1 and 3.3 ha.
Both were in a reduced-till (R-Till) row-crop management system with one field in com
and the other in soybeans and alternating each year. Both fields were tilled using a
chisel, field cultivator or disc, and all fertilizer was pre-plant applied (entire field
application prior to planting). The R-Till Corn-Beans field received a mean nitrogen
application of 111kg/ha/yr. over the four years of record. The R-Till Beans-Corn field
received a mean nitrogen application of 138 kg/ha/yr. for the same period (Table 1).
A crop year is defined as beginning after harvest.
Three more drainage monitoring sites are located where soils are predominantly
Flanagan silt loam and Drummer silty clay loam soils. One site, designated Seed Com,
has a 6.8 ha sub-surface drainage system where high-nitrogen reduced tillage seed
corn management was alternated with soybeans. The second site is an 8.4 ha sub-
surface drainage system under a corn-soybean reduced tillage management system
where white food grade corn was raised; thus, the name White Com for this site. The
third site is a 20.5 ha sub-surface drainage system under a corn-soybean conventional
tillage management system, and was named C-Till Beans-Corn. The Seed Corn and
White Corn fields were tilled using only discs and field cultivators, and received means
92
-------
of 98 and 91 kg/ha/yr. nitrogen, respectively, in pre-plant applications during the four
years of cropping record. The C-Till Beans-Corn field was moldboard plowed after corn
production with a field cultivator used as the secondary tillage tool; this field received a
mean of 106 kg/ha/yr. nitrogen, with most of that in a side-dress application after corn
planting.
At another location where soils are predominately Sabina and Xenia silt loams, a 7.5 ha
sub-surface drainage system was under no-till row crop management. This field was
named No^Till Corn-Beans because the cropping pattern was alternately corn and
soybeans. This field received no tillage except the planting tool. The mean of 85
kg/ha/yr. nitrogen over the four crop years was predominately a side-dress application.
A second 6.9 ha sub-surface drainage system outlets from a field in permanent grass,
and was named Grass.
One drainage monitoring site where soils are predominately Birkbeck and Sabina silt
loam soils, designated Corn Silage, is a 10.9 ha sub-surface drainage system under
soybean-corn-corn reduced tillage management. The second year corn was harvested
as silage and the only fertilization over the last four years was cattle manure at the rate
of 45 Mg/ha during the winters prior to corn, which was estimated to contain a mean of
100 kg/ha/yr. available nitrogen for the three year crop rotation.
Sub-surface drainage (tile) flow was sampled bi-weekly and additional samples were
taken during increased flow following major rainfall events. These samples were
analyzed for nitrate as well as pesticides. The sub-surface outflow depth was
monitored continuously with a flume and stage recorder. Records of agrichemical
application to and tillage on the monitored fields were collected. Soil sampling was
performed to provide background and periodic concentration of agrichemicals in the
field soil.
Results and Discussion
Results of the nitrate concentrations for the eight tile monitoring stations are presented
in Figure 1. Nitrate-N concentrations were above the US EPA Maximum Contaminant
Level (MCL) of 10 mg/L NO3-N at all but a few sampling times from the Reduced Till
(R-Till), Seed Corn and White Corn locations. Nitrate-N concentrations were generally
slightly above and below the MCL from the Conventional Till (C-Till), No-till, and Corn
Silage locations. However, nitrate-N concentrations from the corn silage location were
quite large in early 1995. Nitrate-N concentrations from the Continuous Meadow
location averaged 1.0 mg/L NO3-N for the period of record. As expected, the seasonal
pattern of nitrate-N concentrations from the tile stations is similar to, but less
pronounced than that found in the river; after all, the tile systems supply the river flow.
93
-------
Nitrate- N Concentrations
At Tile Monitoring Stations
40
.30 -
I
§
o
u
20 •
10
0 •
x.
10/30/91 05/17/92 12/03/92 06/21/93 01/07/94 07/26/94 02/11/95 08/30/95
Date of Sample
R-Till Corn-Beans
-------
The amount of fertilization, as well as the method and timing of nitrogen application
affects the nitrate-N concentrations from field tile. The greatest concentrations were
from R-Till Corn-Beans, R-Till Beans-Corn, Seed Corn and White Corn fields where
pre-plant broadcast fertilization at somewhat greater nitrogen amounts were applied
(Table 1). These four cropping systems received the amount of nitrogen fertilizer for
each crop year of record as shown in Table 1. The average nitrogen applied to these
four systems was 165 kg/ha in 1991-92, 62 kg/ha in 1992-93, 149 kg/ha in 1993-94,
and 62 kg/ha in 1994-95; for an overall average of 110 kg/ha/yr.
The lesser concentrations of NO3-N in the tile discharge come from the C-Till
Beans-Corn and No-Till Corn-Beans fields where side dress application of nitrogen to
corn was used, and the Corn Silage field where cattle manure was the only source of
nitrogen (Table 1). The average nitrogen applied to these three systems was 110 kg/ha
in 1991-92, 97 kg/ha in 1992-93, 79 kg/ha in 1993-94, and 106 kg/ha in 1994-95; for an
overall average of 97 kg/ha/yr.
The average nitrate-N concentrations and standard deviation in the tile flow for the
periods of record are shown in Table 2. The R-Till Corn-Beans, R-Till Beans-Corn,
Seed Corn, and White Corn management systems produced average nitrate-N
concentrations of 15.0 mg/L, 14.8 mg/L, 16.5 mg/L, and 15.6 mg/L, respectively, for the
period of observation (Table 2). The mean concentration for this management group is
15.5 mg/L. The average nitrate-N concentrations for the C-Till Beans-Corn, No-Till
Corn-Beans and the Corn Silage management systems were 9.6 mg/L, 7.3 mg/L, and
11.2 mg/L, respectively. The mean concentration for this management group was 9.4
mg/L which is significantly different than the 15.5 mg/L mean of the other management
Table 2. Mean and Standard Deviation Nitrate-N Concentrations in Tile Outflows.
Management System
R-Till Corn-Beans
R-Till Beans-Corn
Seed Corn
White Corn
Pre-Plant Application Mean
C-Till Beans-Corn
No-Till Corn-Beans
Corn Silage
Side-Dress or Manure Mean
Grass
Period of
Observation
12/91 -12/95
2/92-12/95
9/92-12/95
4/94-12/95
11/92-12/95
10/91 -12/95
8/93-12/95
10/91 -12/95
NOs-N, mg/L
Mean
15.0
14.8
16.5
15.6
15.5
9.6
7.3
11.2
9.4
1.0
Standard
Deviation
6.1
4.5
4.3
5.7
3.6
2.8
4.8
1.1
95
-------
group. The lesser concentrations were from fields where nitrogen was usually applied
to com as a side dress application after planting when the corn can immediately utilize
the nitrogen and where fertilization was animal manure that is a combination of
readily-available and slow-release nitrogen. The pre-plant application treatment means
are significantly different from the side-dress and manure application treatment means
with an overall difference between the treatments of 6.1 mg/L nitrate - N.
Summary and Conclusions
The objective of this study was to evaluate the effectiveness of tillage and cropping
management systems in reducing the movement of nitrate in surface and sub-surface
flow. Nitrate in sub-surface tile flow have been monitored for four years from fields with
various tillage and cropping management practices. Samples have also been obtained
along the mainstream of the watershed for nearly five years.
Concentrations of nitrate differed little among specific sampling locations along the
river, but they definitely followed a seasonal cycle. Nitrate concentrations from tile
drains varied considerably between fields depending upon the cropping management
systems used, with concentrations varying seasonally as in the river.
The effect of the application of broadcast, pre-plant nitrogen fertilizer is clearly shown in
the nitrate-N concentrations from tile drains. The pre-plant anhydrous-N application
systems with average nitrogen application of 110 kg/ha/yr. had a mean concentration of
nitrate-N of 15.5 mg/L while the side-dress and manure application systems with
average nitrogen application of 97 kg/ha/yr. had a mean concentration of nitrate-N of
9.4 mg/L. The mean concentration of nitrate-N from a permanent meadow field was
1.0 mg/L.
96
-------
Results and Lessons Learned from the Beaver Creek
Hydrologic Unit Area Project in West Tennessee
Shannon D. Williams and Tom D. Byl
U.S. Geological Survey, Nashville, Tennessee
George F. Smith
University of Tennessee, Agricultural Extension Service, Knoxville, Tennessee
An overview of the major results and some lessons learned from the Beaver Creek Hydrologic Unit
Area project in West Tennessee is presented in this paper. Agricultural best-management
practices were implemented to improve water quality in the watershed, and an intensive
monitoring program was developed to obtain water-quality data.
Data obtained from the monitoring program were used to identify agricultural nonpoint sources of
pollutants, to examine the effect of agriculture on surface- and ground-water quality in the
watershed, and to evaluate in-field and instream best-management practices. Soil, water, and
biological sampling methods were evaluated and new strategies were developed to characterize
agricultural nonpoint-source pollutants. Awareness of water-quality issues within the agricultural
community was increased through educational and informational activities.
Introduction
Beaver Creek drains about 95,000 acres in Shelby, Tipton, Haywood, and Fayette Counties in
West Tennessee and is a major tributary of the Loosahatchie River (fig. 1). Agricultural activities
are vital to the local economy. Cropland covers about 70 percent of the watershed; cotton,
soybeans, and corn are the principal crops. Many of the streams in the watershed have been
channelized to improve drainage. Soils in the watershed consist of highly erodible, well to poorly
drained silt loams. Conservation-related measures termed "best-management practices" (BMP's)
have been implemented in the Beaver Creek watershed, as in many agricultural areas, to reduce
water-quality impairment.
The Beaver Creek watershed project was initiated in 1989, when efforts by the farming community
in Tennessee facilitated a cooperative agreement between the U.S. Geological Survey and the
Tennessee Department of Agriculture to evaluate agricultural nonpoint-source pollution and
BMP's in the watershed. In 1990, the watershed was added to the U.S. Department of Agriculture
Hydrologic Unit Area (HUA) program.
As part of the HUA program, the primary goal of the Beaver Creek project was to improve surface-
and ground-water quality in the watershed through the implementation of BMP's. An intensive
97
-------
water-quality monitoring program was developed to obtain the information needed by resource
management agencies to implement conservation practices and to improve water quality in the
watershed. Specific objectives of the monitoring program included determining the extent to which
agricultural activities affect water quality in the watershed, evaluating in-field and instream BMP's,
and developing and evaluating methodologies for assessing agricultural nonpoint-source (NPS)
pollution.
Several government agencies, academic and research institutions, and agricultural organizations
were involved in the project (Table 1). A local coordination committee facilitated communication
between the various organizations, prevented duplication of efforts by agencies with similar
functions, and capitalized on the specialities and resources of all organizations involved.
Establishing and maintaining communications, such as those developed during regular meetings
of a local coordinating committee, may contribute to the success of other watershed projects.
TABLE 1. Agencies and institutions involved in the Beaver Creek project
U.S. Geological Survey University of Tennessee Agricultural Extension Service
Tennessee Department of Agriculture U.S. Department of Agriculture,
Clemson University Farm Service Agency
University of Memphis Natural Resources Conservation Service
Tennessee Division of Forestry Tennessee Department of Environment and Conservation
Water Research Federation Shelby, Tipton, Fayette, and Haywood County Soil
Tennessee Soybean Promotion Board Conservation Districts
Technical, financial, and educational assistance was provided to farmers for the implementation
of BMP's. BMP's used included structural practices, such as terraces, diversions, and water and
sediment control basins; and vegetative and tillage practices, such as permanent vegetative
cover, contour strip cropping, conservation tillage, and winter cover crops. Integrated crop
management practices, such as nutrient and pesticide management were also used.
Evaluation of Water-Quality Conditions
An intensive monitoring program was used to obtain water-quality data in the watershed (Fig. 1).
The monitoring program served several functions such as identifying sources and magnitudes of
agricultural NPS pollution, providing information needed to support the selection and
implementation of BMP's, and assessing the effectiveness of BMP's in controlling the transport of
pollutants from agricultural areas (1).
A ground-water quality reconnaissance was conducted to examine the occurrence and
distribution of agricultural chemicals in shallow aquifers in the watershed (2), and a more detailed
study was conducted to obtain additional ground-water quality and land-use data (3). An
investigation of the movement and degradation of aldicarb and its metabolites in the soil profile
was also conducted to examine the potential for the leaching of pesticides into shallow aquifers in
the watershed (4).
Nutrient, sediment, and pesticide data were collected at four small streams with drainage areas
ranging from 28 to 422 acres (5) and at two larger streams with drainage areas of approximately
8,000 acres each (6). Nutrient, sediment, and pesticide data were also collected at the inlet and
outlet of a 0.88-acre constructed wetland cell which received runoff from 47 acres of row crops (7).
98
-------
TENNESSEE
W2T
HAYWOOD
•R~\J*~*
FAYETTE-"
Digital data from the 1:100,000 National Mapping Division Universal Transverse
Mercator projection. Projected in State Plane 1927 North American datum.
0123 KILOMETERS
EXPLANATION
COUNTY LINES
HYDROGRAPHY
WATERSHED BOUNDARY
• SAMPLED WELL
y WATER-QUALITY MONITORING
STATION
Figure 1. Location of the Beaver Creek watershed study area and monitoring sites.
Summary of Water-Quality Results
Ground Water
Results from the ground-water reconnaissance (2) and the soil sampling study (4) indicate that
the potential for the leaching of pesticides such as aldicarb, atrazine, and alachlor into ground
water in the watershed is low. The use of no-till and other conservation tillage practices does not
appear to increase the leaching of pesticides through the soil profile (4).
Elevated nitrate and fecal bacteria concentrations were detected in samples from several wells
indicating the influence of anthropogenic activities (3). Nitrate concentrations were significantly
higher in samples from wells near septic tanks and confined animal areas than from wells near
fertilized fields. Wells deeper than 150 feet did not appear to be affected by surface or near-
surface nitrate sources (8).
99
-------
Surface Water
Suspended sediment resulting from erosion is a major water-quality problem in the watershed.
High sediment concentrations and loads were measured in runoff from agricultural fields (5,6,7).
Most of the nitrogen detected in runoff was in the organic form and was transported during late
winter and early spring storms (9). Herbicides and carbamate insecticides were detected in
surface-water samples collected during storms (5, 7)..Most of the transport of pesticides was
reported during large storms occurring shortly after pesticide application (10). A natural bottom-
land hardwood wetlands and a constructed wetland cell were evaluated as instream resource-
management systems. The natural bottomland wetland had significantly lower sediment and
nutrient loads than a channelized stream draining a basin of the same size and land use (6). The
constructed wetland was effective in reducing nitrogen, phosphorus, sediment, and pesticide
loads by 40 to 90 percent during a 4-month period (7).
Developments in Sampling Methodologies
A major component of the Beaver Creek study was the evaluation of existing sampling methods
and the development of new sampling strategies. The design and implementation of adequate
sampling programs for water, soil, and biological samples is critical in assessing agricultural NFS
pollution and BMP's.
Monitoring programs must be able to address multiple pollution sources. Differentiating sources
of NFS pollution at the watershed level is often difficult. Ground-water studies in the watershed
emphasized the use of detailed land-use inventories to identify possible sources of ground-water
contamination and the selection of appropriate statistical methods to determine the effect of
various land uses on water quality (8).
Surface-water monitoring activities in the watershed indicate that an optimal sampling strategy for
characterizing chemicals and sediment in agricultural runoff includes frequent sampling during
storm flow. Many monitoring programs include the collection of a relatively small number of
samples during storms, when most of the pollutants are transported. A sampling interval equal to
5 percent of the storm flow duration was determined to be adequate in characterizing constituents
during storm flow with an error of less than 5 percent (9).
Soil-sampling strategies are needed to accurately characterize the fate and transport of various
pesticides in the soil profile. Multiple sampling transects at right angles to crop rows with samples
collected at row top, slope, and furrow were used to represent field conditions and describe the '
spatial and temporal distribution of aldicarb and its metabolites (4).
i
Many environmental agencies are endorsing the use of biological monitoring in water-quality '
studies. Evaluations of various biological sampling techniques in the watershed indicate that ,
different sampling methods may produce distinctly different biological assessment results. These
results indicate that a variety of methods may need to be used jointly in biomonitoring projects of
NFS pollution in streams such as in the Beaver Creek watershed (11).
100
-------
Educational Activities
Educational activities were used to improve awareness of water-quality issues within the
agricultural community and to increase support of HUA activities such as implementation of BMP's
and monitoring activities. Pamphlets and short papers (12,13,14,15) were prepared as a
cooperative effort between the U.S. Geological Survey and the University of Tennessee
Agricultural Extension Service (UTAES) to improve the understanding of water-quality issues in
the watershed. A UTAES pamphlet (16) also was prepared to describe ways individuals could
protect wells from contamination and improve ground-water quality in the watershed. These
publications were distributed through the local offices of UTAES and other cooperating agencies
and during annual field days, which also included tours of the watershed, field demonstrations,
and presentations.
References
1. Hankin, H.C., and Smith, G.F., 1994, Research and monitoring needs for the implementation
of the USDA water quality HUA program: the Beaver Creek watershed project, in Pederson,
G.L., ed., National Symposium on Water Quality, Chicago, 1994, Proceedings: American
Water Resources Association, p. 57-64.
2. Fielder, A.M., Roman-Mas, Angel, and Bennett, M.W., 1994, Reconnaissance of ground-
water quality at selected wells in the Beaver Creek watershed, Shelby, Fayette, Tipton, and
Haywood Counties, West Tennessee, July and August 1992: U.S. Geological Survey Open-
File Report 93-366, 28 p.
3. Williams, S.D., 1996, Ground-water-quality data for selected wells in the Beaver Creek
watershed, West Tennessee: U.S. Geological Survey Open-File Report 95-769, 30 p.
4. Olsen, L.D., Roman-Mas, Angel, Weisskopf, C.P., and Klaine, S.J., 1994, Transport and
degradation of aldicarb in the soil profile: a comparison of conventional tillage and nontillage,
in Pederson, G.L., ed., National Symposium on Water Quality, Chicago, 1994, Proceedings:
American Water Resources Association, p. 31-42.
5. Williams, S.D., and Harris, R.M., 1996, Nutrient, sediment, and pesticide data collected at
four small agricultural basins in the Beaver Creek watershed, West Tennessee, 1990-1995:
U.S. Geological Survey Open-File Report 96-366,115 p.
6. Cochrane, H.H., and Williams, S.D., 1996, Nutrient and sediment loads in a channelized
stream and a nonchannelized wetland stream in the Beaver Creek watershed, West
Tennessee, in Byl, T.D., and Carney, K.A., compilers, Instream investigations in the Beaver
Creek watershed in West Tennessee, 1991-95: U.S. Geological Survey Water-Resources
Investigations Report 96-4186, p. 3-8.
7. Smink, J.A., and Byl, T.D., 1996, Evaluation of a constructed wetland to control agricultural
row-crop nonpoint-source pollution, in Byl, T.D., and Carney, K.A., compilers, Instream
investigations in the Beaver Creek watershed in West Tennessee, 1991-95: U.S. Geological
Survey Water-Resources Investigations Report 96-4186, p. 11-17.
8. Williams, S.D., and Roman-Mas, Angel, 1994, Relation between nitrate concentrations and
potential nitrate sources for water-table aquifers in the Beaver Creek watershed, West
Tennessee: a statistical analysis, in Pederson, G.L., ed., National Symposium on Water
Quality, Chicago, 1994, Proceedings: American Water Resources Association, p. 43-50.
9. Roman-Mas, Angel, Cochrane, H.H., Smink, J.A., and Klaine, S.J., 1994, Fate and transport
of nitrogen in an agricultural watershed, in Pederson, G.L., ed., National Symposium on
Water Quality, Chicago, 1994, Proceedings: American Water Resources Association, p. 51.
101
-------
10. Williams, S.D., 1996, Transport of aldicarb and aldicarb metabolites in runoff from agricultural
fields in the Beaver Creek watershed, West Tennessee, in Byl, T.D., and Carney, K.A.,
compilers, Instream investigations in the Beaver Creek watershed in West Tennessee,
1991-95: U.S. Geological Survey Water-Resources Investigations Report 96-4186, p. 29-34.
11. Byl, T.D., and Roman-Mas, Angel, 1994, Evaluation of biomonitoring techniques used in
assessing agricultural nonpoint-source pollution, /nPederson, G.L., ed., National Symposium
on Water Quality, Chicago, 1994, Proceedings: American Water Resources Association,
p. 21-30.
12. Byl, T.D., and Smith, G.F., 1994, Biomonitoring our streams: what's it all about?: U.S.
Geological Survey Open-File Report 94-378,1 sheet.
13. Doyle, W.H., Jr., Whitworth, B.G., Smith, F.G., and Byl, T.D., 1996, The Beaver Creek story:
U.S. Geological Survey Open-File Report 96-398,1 sheet.
14. Olsen, L.D., 1995, Pesticide movement in soils: a comparison of no-tillage and conventional
tillage in the Beaver Creek Watershed in West Tennessee: U.S. Geological Survey Open-File
Report 95-329,1 sheet.
15. U.S. Geological Survey, 1995, An overview of the Beaver Creek Study in West Tennessee:
U.S. Geological Survey Open-File Report 95-312,1 sheet.
16. Smith, G.F., 1995, Protecting wells from contamination: results from the Beaver Creek
watershed: The University of Tennessee, Agricultural Extension Service, SP 392-F, 4 p.
102
-------
Subsurface Drainage Outflow Improvement With Constructed Wetland
P. S. Miller, J. Kent Mitchell, S. E. Walker, M. C. Hirschi
Abstract:
Nonpoint source pollution from agricultural fields is a contributor to the degradation of
water quality. In the Little Vermilion River Agricultural Nonpoint Source Hydrologic Unit Area
Project, several types of cropping systems and management practices are being studied. One of
these practices is the establishment of wetlands which act as possible "sinks" for nitrate nitrogen,
phosphorus, and nine commonly used pesticides.
The objectives of this study are to develop a correlation between agricultural effluent
flow through the wetland and changes in nutrient and pesticide mass and to quantify how
efficiently the wetland can adsorb pollutants.
Subsurface drainaige outflow is the only major hydrologic input to a constructed wetland.
The inflow and outflow have been monitored and sampled for three years.
The data indicate that 19.8% of the nitrate nitrogen entering the wetland is^feduced, or
"sunk" within the wetland. The results of this and the analyses of phosphorus and pesticide
transport through the wetland will be presented.
Introduction:
Often in central Illinois, nutrient and pesticide concentrations found in ground and surface
water exceed the Health Advisory Limit (HAL) set by the US EPA. Most of this contamination
is of a diffuse nature (nonpoint source) and attributable to agriculture. Therefore, soluble
agrochemicals leached into groundwater or shunted to surface waters through drainage tile are a
cause for concern.
A large percentage of land in Illinois, especially Central Illinois, was historically
predominated by wetland ecosystems. Due to the effect of acts such as The Swamp Lands Acts
of 1849, 1850, and 1860, over 40,000 km2 of wetlands, 27% of the total state land area, has been
converted for agricultural purposes. (Mitsch, 1994; Novotny, 1994) The loss of these ecosystems
has delivered a severe blow to the quality of the surface and groundwater of Illinois.
Wetlands occupy transitional niches in the environment. They serve as buffers between
dry land and true aquatic environments. The occupation of this unique niche, usually lowland
tending toward depressional storage of storm water runoff, lends itself toward collection of water
born pollutants. Particularly, agrochemicals used in heavily cultivated areas such as central
Illinois eventually appear in surface or ground water. Due to various natural processes, wetland
ecosystems delay, contain, or transform nutrients and pollutants entering into their systems.
Many studies have shown that wetlands act as "sinks" for nutrients ( Kadlec, 1994; Mitsch, 1993;
Mitsch, 1994), but few studies have been done concerning pesticide degradation and transport
through wetlands (Rodgers et al., 1993). This potential to "sink" nutrients and pesticides could
be a valuable property to improve water quality in Central Illinois.
Created or constructed wetlands for nonpoint source pollution abatement function as a
relatively inexpensive best management practice (BMP). This study attempts to ascertain the
functional value of a small constructed wetland fed by a drainage tile in central Illinois.
103
-------
Test Site:
The 147 m3 wetland, located in Vermilion County, has one major hydrologic input, a
drainage tile draining 27 acres of farmland in corn soybean rotation. The wetland was
constructed by building a dam across the natural outlet of the original depression. Creation of a
dam at the site kept the area fully inundated and facilitated measurement of the outflow. The
inflow and the outflow were measured with factory calibrated flumes-located at the inlet and the
outlet of the wetland.
The depth of water in the inlet flume was measured by a calibrated potentiometer attached
to a float sitting in a still well situated at the side of the flume. As a backup system for both the
inlet and outlet, the depth of water in each flume was also traced by pen on a chart stage gage.
The depth in the flume then allowed calculation of the flow rate. The outlet flow rate was
measured similarly.
The volume of flow into and out of the wetland was calculated and logged by onsite
dataloggers every ten minutes. Rainfall was also logged by an electronic tipping bucket rain
gauge. Water samples were taken at equivalent volume amounts by automatic samplers.
Samplers were checked after major precipitation events or a minimum of every two weeks.
Water Quality Laboratory Analyses:
Samples were then removed and analyzed for concentrations of nitrate, phosphate, and
nine pesticides: atrazine, cyanazine, trifluralin, ethalfluralin, alachlor, metolachlor, butylate,
pendimethalin, and clomazone.
Samples were removed from the field, refrigerated at 4° C, and usually analyzed within
24 hours. If analysis was not done within 24 hours, then 2 mL of sulfuric acid was added to the
sample to halt microbial activity. The sample was then stored at 4° C until analysis could be
performed.
The nutrient concentrations were determined using a Technicon AutoAnalyzer n. The
nitrate in the samples was reduced to nitrite, and the concentration of nitrite was determined
calorimetrically at the ppm level (mg/L). The significance of the original nitrite concentration
was deemed extremely small by previous laboratory analysis. The phosphorus concentration was
measured similarly.
Pesticide concentrations were obtained by using Gas Chromatography (GC) techniques.
A Hewlett-Packard model 438 gas chromatograph detected pesticide concentrations in samples at
the ppb level (ug/L).
Data Analysis:
Rainfall amounts and inflow and outflow volumes were characterized on a daily basis
beginning 11-16-93. Concentrations from the analyzed automatic and grab samples were
correlated with the appropriate day on which they were taken. Chemical concentrations were
related to equal flow volumes. From this relationship, a mass of chemical was calculated on a
daily basis for both the inflow and the outflow.
104
-------
The daily mass values for inflow and outflow were summed to create data sets
representing accumulated mass throughout the time of the study. The inflow cumulative mass
was plotted against the outflow cumulative mass for each constituent represented in Table 1 to
create double mass curves (Figure 1).
Table 1. Linear Regression Values and Statistical Significance of Inflow-Outflow Double
Mass Data
Flow
Nitrate-Nitrogen
Phosphorus
Atrazine
Cyanazine
Alachlor
Trifluralin
Clomazone
Pendimethalin
Butylate
Slope (n)
0.819
0.727
0.954
1.058
0.765
0.646
1.694
0.087
1.863
0.324
R2
0.993
0.996
0.981
0.994
0.972
0.664
0.729
0.521
0.826
0.662
t-test value
67.34
149.2
7.86
7.78
11.8
7.19
7.63
115.7
7.00
34.7
A linear regression was performed using system software to obtain values for the slope
and r-squared listed in Table 1. The slope of the regression line was then tested against a slope
of one using a hypothesis test listed in Hogg et al. (1993) for a student's t random variable. T-test
values are listed in Table 1.
Results:
Flow
The total flow into and out of the wetland was tested for significance. The slope of the
regression was 0.819 (R2 = 0.993). The high t-value (67.34) showed that the outflow was
significantly lower than the inflow.
Nutrients
The slope of the regression of the double mass curve for nitrate nitrogen was 0.727 (R2 =
0.996). The results of the hypothesis test concluded that the difference was significant and had a
high t value (149.2). The slope of the regression line representing phosphorus was 0.954 (R2 =
0.981) and tested as significantly different (t-value = 7.86). Even though phosphorus tested
significantly different from one, the efficiency of the pond at eliminating phosphorus was low
and cyclic. At the end of this study, the cumulative phosphorus mass was greater in the outflow
than in the inflow.
105
-------
Pesticides
The regression slope value for atrazine was 1.058 (R2 = 0.994). The t-value, 7.78,
represents a high level of significance implying release of atrazine from the pond. A reduction in
the net mass of cyanazine entering the pond was noted by the slope of the regression line 0.765
(R2 = 0.972) and corresponding t-value (11.8). The correlation between inflow and outflow,
measured by R2 values, started to worsen with chemical constituents appearing less frequently in
samples. Alachlor, clomazone, and butylate showed significant reductions in mass with
regression slopes of 0.646 (R2 = 0.664 t-value = 7.19), 0.087 (R2 = 0.521 t-value = 115.7), and
0.324 (R2 = 0.662 t-value = 34.7), respectively. Trifiuralin and pendimethalin exhibited similar
responses as to that of atrazine. Regression slopes for trifluralin and pendimethalin were,
respectively, 1.694 (R2 = 0.729 t-value = 7.63) and 1.863 (R2 = 0.826 t-value = 7.00). The
sampled levels of metolachlor and ethalfluralin were not high nor frequent enough to prove
significance; therefore, they have been eliminated from the study.
During April and May of 1994 a series of rainfall events created conditions which flooded
the stream down grade from the wetland. Water flowed over the dam and flooded the area.
High levels of atrazine were then recorded in the outflow. These data were removed from the
test. On May 16th of 1995, the outlet weir became clogged. These data were also removed for
atrazine and cyanazine. The effect of these events on other chemical constituents appeared
negligible.
O)
o>
5
«
1
5
3
d
o
te
O
300
250
200
150
100
50
0
. I I
Inflow = Outflow
50 100 150 200 250 300
Inflow Cumulative Nitrate-N (Kg)
350
Nitrate-N Double Mass Curve
Regression
Figure 1. Nitrate-N Double Mass Curve
106
-------
Removal Efficiency
The results of the efficiency of nutrient and pesticide removal calculated from the
cumulative mass difference between the inflow and outflow over the period of the study are
presented in Table 2.
Table 2. Cumulative Nutrient and Pesticide Removal Efficiency 11-16-93 - 12-30-95
Nitrate-Nitrogen
Phosphorus
Atrazine
Cyanazine
Alachlor
Trifluralin
Clomazone
Pendimethalin
Butylate
Efficiency %
19.8
-3.5
-7.75
20.6
37.0
-28.4
94.6
-22.7
27.2
Conclusions:
Differences in chemical mass may be the result of photodegradation, biological uptake,
chemical adsorption/desorption, and other processes in addition to leaching through the substrate.
Losses due to leaching and evapotranspiration have not been quantified
Though infiltration and evapotranspiration rates are unknown, the slope of the regression
for nitrate nitrogen is lower than that of the flow implying that other processes are at work. The
main pathways of nitrate nitrogen loss are considered to be through denitrification and leaching.
The wetland undergoes seasonal cycles of growth. This seasonal dependence affects the
mass of phosphorus leaving the wetland. The general trend is in reduction of phosphorus, yet,
the cumulative mass leaving the wetland had been greater at the end of this study. Uptake by
algae and duckweed during cycles of growth is re-released into the system when death and
decomposition occurs. The cumulative phosphorus entering the wetland was less than the
phosphorus leaving the wetland when vegetation was blooming during the middle of spring. Any
phosphorus removed is then released during periods of death and decomposition. Also, natural
tree fall litter contributes to the organic pool of the wetland. At its highest, phosphorus retention
was 6.7%, but by the end of this study the wetland served as a source of phosphorus.
From analysis of the data, the wetland seemed to serve as a source of atrazine over the
period of the study. Adjustments to remove the effects of flooding may not have fully taken into
account the true affects on atrazine mass in the wetland.
Results for alachlor, cyanazine, and butylate show similar, approximately 20 to 30
percent, cumulative removal efficiencies. All three chemicals exhibit similar half life decay rates
on the order of 10 - 25 days, moderate levels of water solubility, and moderate levels of
107
-------
adsorption (USDA 1991). !
Conversely, trifluralin and pendimethalin seemed to collect in the wetland. High levels
of adsorption potential and longer half lives tended to work together to keep these chemical !
constituents from degrading. Also, due to the their high affinity for adsorption, sediment j
deposited from flooding may have introduced extra quantities of trifluralin and pendimethalin
into the system. Desorption and equilibrium processes may have resulted in larger quantities '
found in the outflow rather than in the inflow.
Clomazone exhibited the highest level of removal efficiency. Clomazone has a rather |
high half life that is comparable to trifluralin. Of the species tested, though, clomazone had the j
highest level of water solubility, five to ten times greater than other constituents. This level of i
water solubility may account for the increased removal efficiency either through microbial
breakdown or photodegradation.
References:
Hammer, D. A. 1997. Creating Freshwater Wetlands. CRC Press, Inc.
Hogg, R. V., E. A. Tanis. 1993. Probability and Statistical Inference. Prentice-Hall, Inc.
Kadlec, R. H. 1994. Wetlands for Water Polishing: Free Surface Wetlands. In Global Wetlands:
Old World and New, ed. W. J. Mitsch, 335-348. Elsevier, Amsterdam.
Mitsch, J.W. 1993. Landscape Design and the Role of Created, Restored, and Natural Riparian
Wetlands in Controlling Nonpoint Source Pollution. In Created and Natural Wetlands for
Controlling Nonpoint Source Pollution, ed. R. K. Olson, ch. 3, 43-69. EPA.
Mitsch, J.W. 1994. The Nonpoint Source Pollution Control Function of Natural and Constructed
Riparian Wetlands. In Global Wetlands: Old World and New, ed. W. J. Mitsch, 351-360.
Elsevier, Amsterdam.
Novotny, V. H. Olem. 1994. Water Quality: Prevention, Identification, and Management of
Diffuse Pollution. Van Nostard Reinhold. New York.
Rodgers, J. H., A. Dunn. 1993. Developing Design Guidelines for Constructed Wetlands to
Remove Pesticides from Agricultural Runoff. In Created and Natural Wetlands for
Controlling Nonpoint Source Pollution, ed. R. K. Olson, ch. 5, 113-130. EPA.
USDA. 1991. Protecting Water Quality in Illinois: Nutrient and Pesticide Management
Strategies. USDA, Cooperative Extension Service University of Illinois at Urbana-
Champaign.
108
-------
Restoration of the Waukegan River
Richard J. Mollahan
Illinois Environmental Protection Agency, Springfield, Illinois
The purpose of the project was to reduce the sediment load discharge to Lake Michigan from
streambank erosion of the Waukegan River. Erosion was caused by increased urban runoff and
channelization, problems common in many urban streams. To address the Waukegan River
streambank erosion, a partnership was formed between federal, state and local entities. Innovative
streambank restoration techniques were implemented to demonstrate the improvement of water
quality by stabilizing eroding streambanks and creating stable stream habitat.
Best management practices (BMPs), such as biotechnical streambank stabilization techniques were
implemented on the Waukegan River in Washington Park and Powell Park in the City of
Waukegan, Illinois. This project was funded in part with funding from the Section 319 nonpoint
pollution program of the Clean Water Act. Monitoring the effectiveness of the implemented BMPs
is the responsibility of the Illinois Environmental Protection Agency (Illinois EPA).
At severe streambank erosion sites, biotechnical streambank stabilization techniques (BSST)
implementation (structure added to vegetation) was a more cost-effective and environmental
sensitive means to reduce nonpoint source (NPS) pollution then the traditional approaches (i.e. rip
rap, concrete lining).
INTRODUCTION
In 1991, a partnership between the Waukegan Park District,
United States Environmental Protection Agency (U.S. EPA),
and the Illinois EPA, utilizing Section 319 funds, was formed to
address the severely eroded streambanks of the Waukegan
River, which were contributing excessive sediment loading to
Lake Michigan.
The Waukegan River is located approximately 35 miles
northwest of Chicago. The Waukegan River is 12.5 miles long
and drains 7,640 acres. Land uses are residential, agricultural,
commercial and industrial (Figure 1). The watershed is highly
urbanized with most of the urbanized portions of the
Waukegan River represented in the reaches that flow through
Washington Park and Powell Park in the lower portion of the
watershed. Both of these parks are located in the older portion
of the City where little or no stormwater detention was
constructed. Therefore, little mitigation of stormwater quantity
Figure 1
109
-------
or quality occurs, resulting in high runoff rates.
Sources of water quality impairments also include
cross-connections between sanitary and storm
sewers, sanitary sewer overflows during wet
weather events, severe streambank erosion
(Figure 2), and a degraded stream habitat.
In the beginning stages of this project, BSSTs
were implemented to protect the City's sanitary
sewers and restore the environmental and
aesthetics benefits to the park lands. The BSSTs
selected combined riparian revegetation
(grasses, willows, etc.) with structural stabilization.
Figure 2
The structural elements that were tested in this
project included lunkers and interlocking concrete
a-jacks structures (Figure 3). These streambank
restoration techniques were chosen for their ability
to withstand high velocity flow while increasing
riparian habitat and in-stream habitats for the
fisheries' community.
PROJECT DESIGNS
Figure 3
The first installation of the BSSTs occurred on
the North Branch of the Waukegan River in
Powell Park and Washington Park during the
fall of 1991 and 1992. Lunkers and a-jacks
were installed in Powell Park while Lunkers
with stone were installed in Washington Park
(Figure 4). On the two lunker installations,
vegetation (willows, dogwoods, grasses, and
other wetland plants) were placed into the
lower, middle, and upper zones of the lunker
structures. The structures utilized were
chosen to enhance in-stream habitat and also
Figure 4
110
-------
to provide a structural base for riparian revegetation
of the bank.
The next installations of BSSTs were utilized on the
South Branch of the Waukegan River in the fall of
1994 to control severely eroded streambanks in
Washington Park. The BSSTs installed were
lunkers, stone, dogwoods, willows, and grasses
(Figure 5). Specific small streambank erosion sites
on the South Branch were also stabilized with other
BSSTs which included coir coconut fiber rolls,
Figure 5
willows, and grasses (Figure 6).
Figure
In the winter of 1996, seven low stone weirs
(LSWs) formed by granite boulders were installed
to create a series of pool/riffle sequences to
enhance in-stream habitat on the Waukegan
River (Figure 7). These LSWs were constructed
to help resolve the lack of water depths, limited
cobble substrates, and limited stream aeration
needed to enhance the aquatic community in the
Waukegan River at Washington Park.
MONITORING
Figure 7
The U.S. EPA's national monitoring program (NMP) is being used to report the effectiveness of the
BSSTs implemented the Waukegan River. The NMP incorporates three biological elements.
These elements are fisheries, benthos, and in-stream habitats. Stream flow is the other element
ill
-------
that is monitored.
National watershed monitoring stations (S1 & S2) on the South Branch of the Waukegan River
were located in the downstream treatment reach and on the upstream control reach respectively
and followed U.S. EPA's National Watershed Monitoring Protocols (NWMP) (Figure 8). The S2 is
the designed upstream control station and is used as a reference condition. These stations based
on seasonal sampling are monitored three times (Spring, Summer, and Fall cycles) since 1994.
The NWMP requires that a survey of the fisheries and stream habitat be conducted before and after
implementation of BSSTs and the LSWs. Macroinverbrate Biotic Index (MBI), Potential Index of
Biotic Integrity (PIBI), and Index of Biotic Integrity (IBI) are indexes calculated from the data
collected. The monitoring activities are performed by stream Biologists from the Illinois EPA and
Illinois Department of National Resources. All the monitoring and related data are entered into the
U.S. EPA's Nonpoint Source Management System (NPSMS) and STORET database system on
an annual basis.
Additional monitoring sites (N1 & N2)
are utilized for background data
collection on the North Branch of the
Waukegan River. At these two
stations, the chosen BSSTs were
wooden lunker/LSW structures in
Washington Park (N1) and recycled
plastic lunker and a-jack structures in
Powell Park (N2) (Figure 8).
The flow of the Waukegan River is
determined by an automatic flow
logger (AFL) which incorporates an
area velocity sensor which uses the
Doppler effect to directly measure the
average velocity of stream flow.
Springtime stormwater discharges
measured in 1996 were 600 to 700
cfs. The meter on the unit calculates
the discharge using level to area data
points which are inputted by the user.
These data points are determined from
a cross-sectional survey of the
channel at the AFL sensor.
The AFL data collection utilizes three
data elements (level, velocity, and
flow). This data is recorded in ten
minute increments and is downloaded
into a computer for analysis monthly.
Figure 8
112
-------
The AFL is contained in a steel box on a raised
platform to minimize vandalism and is located on the
South Branch of the Waukegan River approximately
100 feet downstream of S1 (National
Monitoring station) and 300 feet upstream of the
confluence with the North Branch of the Waukegan
River (Figure 9).
Figure 9
MODELING/APPLIED TECHNOLOGIES
Geographic Information System (GIS) is an organized collection of computer hardware, software,
geographic data and personnel designed to efficiently capture, store, update, manipulate, analyze
and display all forms of geographically referenced information.
GIS is being used in the Waukegan River to spatially characterize many of the physical and
hydrologic features of the watershed. GIS has made it possible to construct a high resolution digital
model of the land uses/land covers, the hypsography or digital elevation model, hydrology, soils and
other physical features needed to properly define a watershed for environmental modeling and
assessment.
In the watershed, the digital databases along with flood zone boundaries, storm sewer network,
transportation, and zoning areas were used in evaluating the application of BMPs using the AUTO-
Ql water quality model. These high resolution spatial data bases establish a physical benchmark
in time. Since the Waukegan River watershed is now a part of the national monitoring program, the
physical changes occurring, spatial assessment/analysis can be done. These activities correlate
with the water quality and biological changes taking place within the watershed. The continued
updating of the GIS environmental databases, in conjunction with doing change assessments,
makes for better management of the overall quality of the watershed and shows how various
activities on the watershed impact its environmental quality.
The use of GIS in conjunction with the development of spatial digital databases (SDD), can be
utilized for accurate assessments of the stream profiles, streams slope, and determining
areas/shapes of an infinite number of watersheds along a stream reach. These various physical
analyses with the integration of the SDD, is making it possible for improvement of the needed
assessments/analyses to support management decisions.
113
-------
RESULTS
Implementations of the BSSTs have been successful in reducing streambank erosion, creating and
reestablishing a vegetative riparian zone, and protecting the communities public works sanitary
sewer infrastructure. However, no significant improvements (i.e. pool depth, instream aeration) to
the aquatic communities in terms of aquatic species diversities (MIBI, PIBI) were measured after
the installation of the lunkers, a-jack structure and related vegetation practices.
After the construction of the seven LSWs
in the winter of 1996, improved water
quality (i.e. fish species, IBI) was evident in
the monitored fishery community. In the
1994 and 1995 years, the number of
pollution tolerate species remained at
eight. Increased numbers of gamefish
and pollution intolerate species raised the
IBI to 35. This increased changed the
Waukegan River status from degraded
aquatic resource to a moderate aquatic
resource. At the S1 station, the number of
fish species increased from 8 to 16 with
the addition of pollution intolerant species.
Fisheries abundance increased by 400%,
with the lunker/a-jack and LSW habitat
enhancement. The upstream control (S2)
remained a limited aquatic resource with
two to four species with limited population
numbers for the entire period. The IBI was
28 or less (Table 1).
TaWe 1. Comparison of UK Mean Station Values offer Indicts
for SI and SI. tor 1994 to 1996.
1994
SI S2
IBI 25.82 22.18
MBI 6.64 7.26
PIBI 41.51 41.93
FISH SPECIES AND ABUNDANCE
Coho
Btaegffl
Laigemoulh bass 1
Longnosedace
Mottled sculnm
Fatbeadmmnow 4 2
Creek Chub 1
Golden «hmer 1 2
While sucker
Black bullhead
Green sun&h
Mosquilo&h 27 13
Goldfish 1
Brook stickleback
Nirapme stickleback 1
Thrccspinc stickleback 1
No. of species 8 3
AbundaciK withoul 35 17
stickleback
Abundance win 35 17
stickleback
1995
SI S2
Luncker
25.33 26.00
6.26 6.31
41.93 41.79
4
64 4
8
17
24 7
20 4
1
53 54
g 4
138 25
191 69
19%
SI S2
Riffles
34.67 28.00
6.99 8.26
41.34 41.65
2
9
12
44
2
16
1!
2
28
3
8
2 1
1
1
3
84 15
16 2
136 1
224 16
The MBI also reflects improving water
quality in the LSWs reach on the South
Branch. The MBI indicated that water quality did not limit or degrade aquatic resources in 1994 or
1995. In 1996, the MBI score of 8.3 at the S2 station indicated poor water quality. The LSWs at
S1 station appeared to moderate the water quality effects since the MBI remained in the non-limited
classification at 7.0 (Table 1).
The physical habitat evaluations found deeper pools at the S1 station while the S2 station remained
very shallow. The LSWs were designed to transport bedload and scour pools during high flow
events. However, PIBI scores remained constant (41-42) for S1 and S2 for all three years (1993-
1996) (Table 1). The PIBI scores are predicated on the absence of claypan or silt-mud substrates
and the percentage of pools and stream width. The S1 and S2 physical habitat had very little silt
or claypan substrates initially, limiting the expected change in the PIBI.
114
-------
DISCUSSION
This project demonstrated that the BSSTs can be more cost-effective than the traditional
approaches in reducing streambank erosion. BSSTs also provide improvements in water quality
as well as beneficial in-stream habitats as demonstrated by the IB) and MBI. The incorporation of
the LSWs that created the pool/riffles series added to the in-stream physical diversity and to the
increase in the biodiversity. In addition to enhancing habitats, LSWs are also effective in reducing
erosion of the stream bed, improving streambank stability, and increasing water aeration.
Streambank restoration is only one important step in improving the diversity of the fisheries'
community. The other important step is the creation of the in-channel restoration which enhances
the entire fisheries' community. LSWs provide the additional pool depth and in stream stone habitat
necessary for establishing higher quality fish communities in urban streams.
115
-------
EVALUATION OF TWO BASINS PRIOR TO STREAMBANK FENCING
IN PASTURED AREAS WITHIN THE MILL CREEK BASIN OF
LANCASTER COUNTY, PENNSYLVANIA
Daniel G. Galeone
U.S. Geological Survey, Lemoyne, Pennsylvania
Abstract
Streambank fencing within pastured areas is a best-management practice targeted to improve
water quality. A cooperative effort between the U.S. Geological Survey and the Pennsylvania
Department of Environmental Protection is in the fourth year of a 6- to 10-year study designed
to quantify the effect of streambank fencing on surface-water quality in a small basin located
within the Mill Creek Basin of Lancaster County, Pa. The paired-basin study was designed to
compare control and treatment basins during a 3-year period prior to fence installation and
during a 3- to 5-year period following fencing of streambanks located within pasture land in the
treatment basin.
Least-squares regression equations developed between the two basins prior to fencing
indicated that significant changes in most water-quality constituents could be detected after
fence installation. The regressions between basins for nutrients and suspended sediment
during base-flow and stormflow conditions showed statistically significant relations for most
constituents. To detect significant changes in the mean difference between the treatment and
control basins for base flow, a change of 23 percent for suspended-sediment concentrations
and an average change of 31 percent for nutrient-species concentrations would be required.
Significant changes in the mean difference between basins for stormflow would require
changes of 29 and 5 percent for suspended-sediment concentration and yield, respectively,
and averages changes for nutrient species of 43 and 31 percent for concentrations and yields,
respectively.
Biological metrics indicate that benthic-macroinvertebrate communities are relatively healthy
within both basins. Canonical correspondence analysis, used to relate sites, environmental
variables, and genera during the pre-fencing period, will be used after the post-fencing period
to determine if there was a macroinvertebrate response to changes in the stream habitat
caused by streambank fencing.
Introduction
Agriculture is the predominant land use in the Mill Creek Basin of Lancaster County, Pa., and
much of the area along streams is used to pasture dairy cattle. Pastured areas within the basin
117
-------
have been identified as nonpoint sources of suspended-sediment and nutrients to streams.
Streambank fencing to exclude animal access is a best-management practice that is targeted
to reduce suspended-sediment and nutrient inputs to streams by reducing direct nutrient inputs
to streams and stopping streambank trampling. A 10-12 foot vegetative buffer along each side
of the stream will also help to stabilize streambanks and potentially reduce the input of nutrients
to the stream channel by overland or subsurface flow.
A 6- to 10-year study is being conducted in two small paired basins underlain by carbonate
bedrock within the Mill Creek Basin to determine the effectiveness of streambank fencing in
improving stream-water quality within the treatment basin. The paired basins are located in
areas where agriculture accounts for approximately 80 percent of the land use. These basins
were chosen for the study because of their similarities in hydrology and geology and because
of the presence of a stable agricultural community that has historically not deviated significantly
from year-to-year practices for their particular tract of land. This relative constancy is critical to
the study because other changes in agricultural activities could make it difficult to detect
changes in water quality strictly caused by streambank fencing. The majority of the 2.5-3.0
stream miles within both basins is bordered by land used to pasture dairy cattle (fig. 1).
76"I5'30"
EXPLANATION
PASTURES
~ STREAMS
• BASIN BOUNDARY
I TOWNS
A WATER-QUALITY SITES
EQUALS 50
DAIRY COWS
39'59'.
Figure 1. Control and treatment basins.
The study is a cooperative effort between the U.S. Geological Survey and the Pennsylvania
Department of Environmental Protection (PaDEP). Funds from PaDEP are provided through
the National Monitoring Program (NMP) of the U.S. Environmental Protection Agency.
Study Design
The paired basins, hereafter referred to as the treatment basin and control basin, are similar in
size (1.4 and 1.8 square miles, respectively), land use, climate, topography, hydrology, and
geology. Three years of pre-fencing (calibration) water-quality data were collected beginning in
1993 to develop relations between the two adjacent basins prior to fencing of streambanks
within pastured areas of the treatment basin. Nutrient, sediment, and benthic-macroinvertebrate
118
-------
data were collected at the outlets of the treatment (T-1) and control (C-1) basins. Streamflow
was continuously monitored at the outlets of both basins and the physical characteristics of in-
and near-stream habitats were documented. Land-use and agricultural-activity data were
collected for both basins to determine if activities within the basins changed from the pre-
fencing to post-fencing period. For a more detailed description of data being collected in the
basins, see Galeone and Koerkle (1996). Fencing within the treatment basin will be installed by
July 1997, and post-fencing (treatment) data will be collected for 3 to 5 years after fence
installation.
The critical season for this study was defined as the period from April through November.
During this time, dairy cows frequently graze in the pastured areas. The critical season is the
period when effects of streambank fencing on surface-water quality are most likely to be
detected. During the critical season, fixed-time samples were collected by hand every 10 days,
and most storm events were sampled. Fixed-time samples were collected regardless of flow
conditions; however, only samples collected at or below the 90th percentile of streamflow data
for that site were considered base-flow samples. Samples submitted for analysis from storm
events were a flow-weighted composite of discrete samples collected throughout the duration
of the event. Although base-flow and stormflow samples were collected outside of the critical
period, only data collected during critical seasons from April 1994 through November 1996 are
presented in this paper.
Constituent concentrations in base-flow and stormflow samples collected at T-1 and C-1 were
used to develop relations between the paired basins during the calibration period. Least-
squares regression relations developed for water-quality constituents between basins during
the calibration period will be compared to the relations developed during the treatment period to
determine whether streambank fencing in pastured areas along stream corridors had a
significant effect on surface-water quality. Constituent concentrations and yields for T-1 were
regressed against corresponding concentrations and yields for C-1 prior to fencing within the
treatment basin. Regressions generated between basins for the treatment period will be
statistically compared (Clausen and Spooner, 1993) to calibration regressions to determine if
the effect of streambank fencing was significant on the water-quality characteristics of the
treatment basin. Unless land use and agricultural activity within the basins change significantly
from the calibration to treatment period, any significant change in regression relations between
basins can be attributed to streambank fencing in the treatment basin. In a paired-basin
analysis, the control basin is used to account for any climatic or hydrologic variations during the
study. This improves the likelihood of detecting significant changes in water quality caused by
the treatment.
Canonical correspondence analysis (CCA) was conducted using the physical characteristics of
the stream habitat, water-quality concentrations, and benthic-macroinvertebrate data to
determine the relation of physical and chemical characteristics to the biological community
within both basins. Biological metrics were calculated from benthic-macroinvertebrate data to
characterize the health of the stream. Relative deviations between basins in benthic-
macroinvertebrates communities from the calibration to treatment period will be identified using
CCA and biological metrics.
119
-------
Results To Date
Data collected during the calibration period showed elevated nutrient concentrations in base
flow and stormflow and elevated suspended-sediment concentrations in stormflow in both the
treatment and control basins (fig. 2). Agricultural activities in the basins, such as cows pasturing
adjacent to stream channels and nutrient applications to cropland, contribute nutrients and
sediment to the stream.
Nutrient and sediment concentrations are similar between basins; however, streamflow at C-1
is nearly twice that of T-1. Streamflow is greater in the control basin because the basin is larger
(1.8 square miles compared to 1.4 square miles) and contains greater amounts of impervious
surface area associated with residential communities, which increases overland runoff during
storms. Ground-water inflow to the control basin from beyond the surface-water basin
boundaries may also affect the quantity of surface water measured at C-1. Even though
streamflow is different, the streamflow regression relation between basins is highly significant.
UlZ
UjO
U-O
%%
3 =
u£
BASE
6
4
2
0
(79)
(0.84)
•
'$
FLOW STORM
COMPOSITE
STREAMFLOW
(67)
(1»)
•
I
4,
?•
120
80
40
n
(51) (50)
•
. (7.48) (14.58).
j
•i i
SUSPENDED SEDIMENT
\TION, IN MILLIGRAMS PER LITER
I
HORUS
(50) (49)
. . (0.70) .
. (0-76) * .
EXPLANATION
(43) Number ol observations
. Data values outside the
10th and 90th percenliles
90th percentile
75th percentile
Median
25th percentile
10th percentile
(0.70) Median concentration
2
o
8
20 (68)
&U
15
10
5
. ,
• Y
(1l!s2)
T-1
TOTAL NITROGEN
(68) „„ (50)
"f •
(10.3)
e.\i
15
10
5
n
. (6.5)
X
?
C-1 " T-1
(49)
(5.1)
, .
* '
C-1
Figure 2. Distribution of streamflow and concentrations of selected
nutrients and suspended sediment for base flow and storm-composite
samples collected during the critical seasons of 1994 through 1996 at
the outlets of the treatment (T-1) and control (C-1) basins.
120
-------
Least-squares regression equations developed for water-quality constituents and streamflow
indicate significant relations between basins for most constituents measured (table 1). That is,
the control-basin data for most constituents was a significant predictor of the treatment-basin
data based on a 95-percent confidence interval. The significant regressions were used to
determine the percentage change (d) required to detect a significant difference in that
constituent between basins during the treatment period relative to the calibration period.
Techniques for this procedure were derived by Clausen and Spooner (1993). The equation to
calculate d requires input of the mean squares error from the regression model, the F value
based on degrees of freedom of the model, and the number of samples collected during the
calibration and treatment period. The mean square error is directly proportional to the variance,
and as the variance of the model increases, so does the value of d required to detect a
significant change from the calibration to treatment periods. The value for d is the percent
change required to detect a significant difference during the treatment period in the mean for
that constituent for T-1 assuming that the mean for that same constituent for C-1 does not
change significantly. Values for d were calculated for base-flow and storm-composite samples
and for concentration and yield of total and dissolved nutrients, suspended sediment,
streamflow, and field constituents. Data indicate that small changes in field constituents
(2-4 percent) or streamflow during the treatment period are needed to document significant
differences from the pre-fencing period at T-1 (table 1). Large differences in nutrients and
suspended sediment are needed to document statistically significant changes from the pre-
fencing period at T-1. For example, the regressions for the concentration of total phosphorus
indicate a 48-percent change in the base flow and a 32-percent change in stormflow to show a
statistically significant change from the pre- to post-fencing period at T-1, assuming the means
for C-1 data do not change.
The upper and lower 95-percent confidence intervals for the equations give a graphical
indication of the change required in the relation to detect significant effects of fencing.
Generally, a significant change will be indicated if the treatment regression relation is outside of
the range of the confidence intervals of the pre-fencing relation. For the dissolved ammonia
relation for base flow samples, a 20-percent change in the mean concentration for T-1 is
required (fig. 3); for the yield of total phosphorus in storm-composite samples, a 31-percent
change in the mean for T-1 is required (fig. 4).
Benthic-macroinvertebrate sampling of pools and riffles just above the outlets of the basins
indicated that the macroinvertebrate communities are relatively healthy based on biological
metrics (table 2). The taxa richness index, the Hilsenhoff biotic index, and the EPT/
Chironomidae abundance ratio indicate that both basins are fully supportive of designated uses
for these particular water bodies (see Plafkin and others, 1989, for discussion of the different
metrics). Because no reference sites were available for the streams, best professional
judgment was used to determine if the designated uses were fully supported or somewhat
threatened. The typical four-tiered classification system based on the level of impact as
discussed by Bode and others (1993) was used as a guideline for converting the biological data
into terms (such as fully supported) that are in accordance with NMP guidelines. It might be
difficult to detect any improvements in macroinvertebrate communities caused by streambank
fencing in the treatment basin because the metrics for T-1 indicated that the stream is fully
supported before the installation of streambank fencing. However, metrics for benthic-
macroinvertebrate communities will be compared for calibration and treatments periods to see
if significant changes occur in the metrics for the treatment basin relative to the control basin.
121
-------
Table 1. Percentage change in mean values of constituents for the control and treatment basins
that would indicate significant differences In constituents during the critical season (April through
November) for the post-fencing period relative to the pre-fenclng period.
Percentage change required1-2
WUI IOUIUCI 11
Streamflow
Temperature
Specific conductivity
Dissolved oxygen
Suspended sediment - concentration
Suspended sediment • yield
Total phosphorus - concentration
Total phosphorus - yield
Dissolved phosphorus - concentration
Dissolved phosphorus - yield
Dissolved orthophosphorus - concentration
Dissolved orthophosphorus - yield
Total ammonia plus organic nitrogen - concentration
Total ammonia plus organic nitrogen - yield
Dissolved ammonia plus organic nitrogen - concentration
Dissolved ammonia plus organic nitrogen - yield
Dissolved ammonia - concentration
Dissolved ammonia - yield
Dissolved nitrite - concentration
Dissolved nitrite - yield
Dissolved nitrite plus nitrate - concentration
Dissolved nitrite plus nitrate - yield
Base flow
5
3
2
4
23
-
48
-
45
-
42
•
4NS
-
NS
-
20
-
26
-
6
-
Stormflow
7
3.
-
-
29
5
32
31
23
16
22
16
NS
36
35
29
83
84
39
19
22
18
1- Percent changes are only given for significant regression relations between basins.
2- Percent changes are based on a 95-percent confidence interval and the assumption that post-fencing
number of observations will equal pre-fencing number of observations.
3-'-' indicates that data are not available.
4- 'NS' indicates that regression was not significant.
Table 2. Mean biological metrics for benthic-macroinvertebrate sampling from September 1993
through September 1996 (seven samplings events with one in May and September of each year
except for 1993) above the outlets of the control (C-1) and treatment (T-1) basins.
Metric
Taxa richness index
Hilsenhoff biotic index
EPT/Chironomidae abundance ratio
Basin
C-1 T-1
23 26
5.36 6.44
.61 .67
Metric ranges for levels of
Fully
supported
>20
0-6.50
0.61-2.00
Fully supported
but threatened
>10-20
6.51-8.50
0.21-0.60
impairment1
Partially
supported
0-10
8.51-10.00
0.00-0.20
1.
Metric ranges are based on best professional judgment because no references sites were available.
122
-------
0.16
0.14
0.08
0.04
0.02
Y = 0.020 + 0.6SX
ADJ. R2= 0.26
UPPER 95-PERCENT
CONFIDENCE LEVEL
PREDICTED EQUATION
LOWER 95-PERCENT
, CONFIDENCE LEVEL
"
0.05
0.10
0.15
0.20
DISSOLVED AMMONIA IN BASE FLOW AT C-1,
IN MILLIGRAMS PER LITER AS NITROGEN
Figure 3. Concentration of dissolved ammonia at the outlet
of the treatment basin (T-1) as a function of the
concentration of dissolved ammonia at the outlet of the
control basin (C-1) for base-flow samples collected during
the critical seasons (April through November) from 1994
through 1996.
Changes in biological communities after streambank fencing can be identified through
multivariate analysis if the community changes are related to changes in the physical and
chemical characteristics of the stream system. One type of multivariate analysis used with
biological data is CCA. CCA is used to relate biological species data to many environmental
variables such as streamflow, nutrient concentrations, or channel characteristics (Ter Braak,
1987). CCA uses the correlation matrix for species and environmental variables to estimate
eigenvalues. The magnitude of the eigenvalue indicates the amount of variability that the
eigenvalue explains in the data. CCA generates an ordination plot of species, sites, and
environmental variables (fig. 5). To relate species or sites to environmental variables on an
ordination plot, a perpendicular line is drawn from the species or site to the vector that
represents the environmental variable. The intersection of the perpendicular line to the
environmental variable line indicates the relative location of that species or site along the
standardized range of the environmental variable. For example, Gammarus sp. and Baetis sp.
were related to relatively high values of specific conductivity. For each environmental variable,
the mean response is represented by the origin (0,0) of the diagram.
123
-------
-0.02
Y = 0.0024+ 0.79X
ADJ. Ra= 0.78
UPPER 95-PERCENT
CONRDENCE INTERVAL ,,
,-'' LOWER 95-PERCENT
X CONFIDENCE INTERVAL
PREDICTED
' EQUATION
0.05 0.10 0.16 0.20
MEAN STORM TOTAL PHOSPHORUS YIELD AT C-1,
IN KILOGRAMS PER ACRE AS PHOSPHORUS
0.25
Figure 4. Yield of total phosphorus at the outlet of the treatment
basin (T-1) as a function of the yield of total phosphorus at the
outlet of the control basin (C-1) for storm-composite samples
collected during the critical seasons (April through November)
from 1994 through 1996.
The x and y axes of an ordination diagram represent the first and second eigenvalues that
explain the most variation in the data. These first two eigenvalues explain 37 percent of the
variation in the species-environmental variable data (fig. 5). According to biplot scores, which
indicate the correlation of environmental variables to the ordination axes, total phosphorus and
dissolved oxygen (in that order of importance) are the variables most highly correlated to the x
axis. The pH and relative abundance of macrophytes are the environmental variables most
highly correlated to the y axis.The three genera that most influence the ordination diagram are
Gammarus (amphipods), Baetis (mayflies), and Orthocladius (midges). This influence is
determined from the weighted scores that relate the genus to the different ordination axes.
Gammarus sp. is the predominant benthic-macroinvertebrate genus identified in the control
basin; Baetis sp. is the predominant benthic-macroinvertebrate genus identified in the treatment
basin; and Orthocladius sp. is the most common genus found in both basins. The location of
Gammarus sp. on the ordination plot relative to the control sampling events (c1 through c7)
indicates a strong relation between the genus and the site.
124
-------
Total ammonia plus
organic nitrogen
Dissolved
oxygen
Macrophytes
PH
Total phosphorus
Gammarussp. c
Specific
conductivity
Ocl +1.0
Filamentous algae
Bank stability
Baetlssp.
Figure 5. Canonical correspondence analysis ordination plot of 31 species, 13
environmental variables, and 7 sampling events from September 1993 through
September 1996 at the outlets of the treatment (labelled t1 through t7) and
control (labelled c1 through c7) basins for benthic macroinvertebrates. [Rare
taxa were excluded from the analysis.]
Using CCA, changes in species composition at T-1 during the post-fencing period can be
related back to any physical and chemical changes in the stream characteristics caused by the
fence installation. Any changes in environmental variables will be reflected in the eigenvalues
and the correlation of the variables to the ordination axes. These changes along with species
composition and abundance data can be used to infer effects of fencing on benthic-
macroinvertebrate communities.
Conclusions
Significant relations developed between paired basins using least-squares regression analysis
and CCA will be useful in determining the effects of streambank fencing on the chemical,
physical, and biological components of the stream system at the outlet of the treatment basin.
Post-fencing data will be collected for 3-5 years in both basins. Assuming that the variance and
the number of surface-water samples from the calibration to treatment period remain similar
and that agricultural activity within the basins remains relatively similar except for the fencing of
streambanks, changes of 6 to 84 percent in nutrient species and changes of 5 to 29 percent in
suspended sediment between basins from the calibration to treatment period would indicate
that streambank fencing had a significant effect on surface-water quality. CCA will be useful in
125
-------
relating changes in the physical and chemical characteristics of the stream to changes in the
benthic-macroinvertebrate community. Changes in the benthic-macroinvertebrate community
can be attributed to streambank fencing If the biological community responds to changes in the
physical and chemical characteristics of the stream channel in the treatment basin and no such
response is evident in the control basin during the treatment period. CCA ordination plots would
qualitatively identify changes in the relation between benthic-macroinvertebrates and stream
characteristics while the eigenvalues and biplot scores associated with CCA could be used to
quantify the effect of streambank fencing on the biological community.
References
1. Bode, R.W., Novak, M.A., and Abele, L.E., 1993, 20 year trends in water quality of
rivers and streams in New York state based on macroinvertebrate data 1972-1992:
New York Department of Environmental Conservation, 196 p.
2. Clausen, J.C., and Spooner, J., 1993, Paired watershed study design:
U.S. Environmental Protection Agency, 841 -F-93-009, 8 p.
3. Galeone, D.G., and Koerkle, E.H., 1996, Study design and preliminary data analysis
for a streambank fencing project in the Mill Creek basin, Pennsylvania:
U.S. Geological Survey Fact Sheet 193-96, 4 p.
4. Plafkin, J.L., Barbour, M.T., Porter, K.D., Gross, S.K., and Hughes, R.M., 1989,
Rapid bioassessment protocols for use in streams and rivers: U.S. Environmental
Protection Agency, EPA/444/4-89-001,176 p.
5. Ter Braak, C.J.F., 1987, The analysis of vegetation-environment relationships by
canonical correspondence analysis: Vegetatio, v. 69, p. 69-77.
126
-------
Lessons Learned in the Long Creek Watershed Project
Gregory D. Jennings, William A. Harman, Daniel E. Line, Carolyn B. Mojonnier
Biological & Agricultural Engineering Department, North Carolina State University, Raleigh, NC
Martha A. Burris, Richard Farmer
North Carolina Cooperative Extension Service, Gaston County Center, Dallas, NC
The Long Creek Watershed Project in Gaston County, North Carolina, was initiated in 1993 as
a nine-year EPA 319 Nonpoint Source National Monitoring Program Project. Objectives are to
implement and evaluate agricultural and urban nonpoint source pollution control measures for
improving surface water quality. This paper describes lessons learned relative to project
management, pollution control measures, water quality monitoring, and education activities
conducted in the project through 1997.
Pollution control measures include nutrient management, riparian buffers, animal waste
management, soil erosion control, and streambank protection. Water quality monitoring
consists of physical, chemical, and biological monitoring at: (1) a dairy farm to evaluate waste
management and riparian protection practices; (2) cropland to evaluate nutrient and waste
management practices; (3) a municipal water supply intake to evaluate sediment control; (4) a
municipal biosolids management facility to evaluate nutrients, metals, and pathogen runoff
control; and (5) an urban watershed to evaluate urban nonpoint source runoff controls.
Education consists of site visits, newsletters, tours, volunteer monitoring, and demonstrations
of control measures.
Introduction
The Long Creek Watershed drains 28,000 acres in Gaston County, North Carolina, with major
land uses of row crops, dairy, beef, residential, commercial, and recreational facilities. Surface
water quality impairments are caused by streambank erosion and runoff of sediment, bacteria,
metals, and nutrients from agricultural and urban land uses. The upstream portion of the
watershed serves as a municipal water supply for Bessemer City. The intake pool at the water
supply intake, located below an agricultural area, is dredged periodically to remove excess
sediment.
127
-------
Objectives of the Long Creek Watershed Project are to measure the water quality benefits of
nonpoint source pollution control measures implemented on:
• A dairy farm to evaluate waste management and riparian protection practices to reduce
pathogens, sediment, and nutrients;
• Cropland to evaluate nutrient and erosion management practices to reduce sediment and
nutrients;
• Land upstream of a water supply intake to evaluate erosion and sediment control
measures;
• A biosolids management facility to evaluate municipal biosolids handling and application
practices to reduce nutrient and pathogen runoff; and
• An urban watershed to evaluate stormwater and streambank protection measures to
reduce nonpoint source runoff.
Project Management
Because of the complexity of this project and large number of agency cooperators, a formal
project management structure was established. The project is directed by a Steering
Committee that meets bi-monthly. Members of the steering committee include technical
experts from North Carolina State University, North Carolina Cooperative Extension Service,
USDA- Natural Resources Conservation Service, USDA Consolidated Farm Services Agency,
Gaston Soil and Water Conservation District, US Geological Survey, North Carolina Division of
Environmental Management, and North Carolina Division of Soil and Water Conservation. The
steering committee also includes a cooperating landowner. The steering committee directs and
approves all major decisions relating to monitoring objectives and BMP implementation. The
Cooperative Extension Service administers the project by providing budget and local project
management.
A broader project Advisory Committee consists of representatives from more than 15
government and private organizations. The purpose of this committee is to provide diverse
input on project direction and community outreach. Annual workshops provide advisory
committee representatives with project updates, tours of monitoring and BMP
accomplishments and roundtable discussions. The steering committee and the advisory
committee are responsible for setting project goals and objectives, however, the agencies
comprising the steering committee are involved with the day-to- day tasks of implementing
these objectives. Site specific work plans were created to clarify objectives and assign agency
responsibilities and timetables. Also, a "Technician Monitoring Guidebook" was written to
provide training and consistent sampling procedures to a project Water Quality Technician.
Lessons learned relative to Project Management include the following:
1. All project stakeholders must be identified at the beginning of the project and provided with
the opportunity to have input toward project directions. Team building for this group is
essential and requires spending time together learning to understand each other's goals,
needs, and limitations.
128
-------
2. A single lead organization must assume responsibility for project management, including
facilitating communication, decision-making, staffing, budgeting, and reporting. Within the
lead organization, there must be at least one individual who understands all aspects of the
project and coordinates activities.
3. A formal management structure with regular meetings is necessary to ensure that all
project stakeholders' needs are being met.
4. Open communications among major project decision-makers must be facilitated through
meetings, e-mail, telephone conferences, and newsletters to avoid confusion and
misunderstandings.
Pollution Control Measures
Pollution control strategies include educational programs, financial and technical assistance,
and regulatory actions. During the two-year baseline water quality monitoring period,
implementation plans were developed for critical areas and pollution sources using monitoring
and modeling. Project staff designed best management practices (BMPs), including
streambank stabilization, livestock exclusion, urban stormwater controls, and agricultural
management practices to be installed in 1995-1997. Funding for BMPs is provided by EPA 319
grants, the North Carolina Agriculture Cost-Share Program (ACSP), and a USDA Water Quality
Incentive Project (WQIP). Technical and educational assistance is provided through the
Gaston Conservation District, USDA-NRCS, and Cooperative Extension Service.
Streambank stabilization BMPs being implemented throughout the watershed include a
combination of vegetative and structural controls depending on the condition of the bank and
stream flow patterns. Dairy farm BMPs include waste storage and handling, livestock water
supplies, riparian area establishment, heavy use and feeding area improvements, stream
crossings, and pasture management. Cropland BMPs include erosion control and nutrient
management. At the Bessemer City water supply intake, 13 acres of cropland immediately
upstream from the intake were converted to permanent wildlife habitat. Also, to comply with the
NC Water Supply Watershed Protection Act, development restrictions are enforced one half
mile of the area above the intake. At the Gastonia Resource Recovery Farm, municipal
biosolids are applied to cropland using recommended rates based on soil and waste analyses.
Urban stream protection BMPs being implemented in two subwatersheds include stormwater
controls, streambank stabilization, and proper landscaping and lawn maintenance practices.
Lessons learned relative to Pollution Control Measures include the following:
1. All project stakeholders, including landowners and local governments, must be involved in
designing BMPs that will be accepted and meet project needs.
2. Project staff must be flexible and patient in working with landowners and municipal officials
to meet their needs. Compromise with cooperating landowners is essential in order to
meet project needs.
3. BMPs must be designed to directly address identified pollution problems. Systems of
BMPs are often required to effectively reduce nonpoint source pollution. These should
include combinations of practices that reduce pollution at the source, that slow pollutant
transport from the source, and that intercept polluted runoff before entering streams.
129
-------
4. BMPs that require intensive maintenance are most likely to fail. For example, sediment
traps that must be cleaned out following every major storm are likely to be neglected and
become ineffective over time. BMPs such as natural forest riparian zones are most likely to
remain effective because of low maintenance requirements.
5. Project staff must oversee BMP implementation and often must actually assist in
implementation to ensure that BMPs are installed properly and at the correct time in the
project monitoring period.
6. Legal, regulatory, and liability issues must be considered in making land use changes on
private and public property.
Water Quality Monitoring
Baseline water quality monitoring was initiated in 1993. Physical, chemical, biological, and
habitat monitoring are used to determine water quality changes related to implementation of
pollution controls. A combination of upstream-downstream, single-station downstream, and
paired watershed monitoring is being used to evaluate water quality benefits of BMP
implementation. Annual macroinvertebrate and habitat monitoring are conducted by the North
Carolina Division of Water Quality at six locations on Long Creek to determine long-term trends
in stream health. Following completion of baseline monitoring in 1995, BMP implementation
began for a three-year period. Monitoring continues throughout the BMP implementation
period and through 2001 to measure long-term impacts of project efforts.
Lessons learned relative to Water Quality Monitoring include the following:
1. Appropriate water quality monitoring designs must be used which allow project staff to
statistically evaluate project successes.
2. Appropriate quality assurance and quality control (QA/QC) plans must be developed and
followed to ensure valid data.
3. Monitoring must address pollutants of concern that will be controlled through BMP
implementation.
4. Monitoring designs must be flexible to meet project conditions and needs.
5. Adequate baseline monitoring is essential to measure success. The high degree of
variability measured during the baseline period confirms the need for long-term data
records in evaluating water quality changes and also for the need for co-variates such as
rainfall and streamflow.
6. Field observations are essential in order to properly interpret monitoring data. Example
observations include temperature, land use changes, locations of livestock, recent spills in
streams, etc.
7. Adequate training and resources must be provided for sampling technicians to collect
samples, maintain sampling equipment, and follow QA/QC plans.
130
-------
Education
An extensive education and outreach program is being conducted in the watershed. Visuals,
tours, newsletters, volunteer monitoring projects, farm visits, and demonstrations of
implemented BMPs are used for public education and technology transfer throughout the
project. Knowledge gained from BMP implementation and monitoring is transferred throughout
the watershed and to other regions.
Lessons learned relative to Education include the following:
1. Project participants must develop an education plan that addresses targeted audiences.
2. The education plan must be flexible and adaptable to new opportunities to meet
educational needs.
3. Project staff must develop strong working relationships with local media to provide
continuing news coverage of watershed activities. The Long Creek project is covered
regularly by the local newspaper in addition to magazines and other media outlets.
4. A project logo is helpful in providing a visual identity for the project. This logo can be used
on displays, stationary, signs, newsletters, rain gauges, magnets, and other materials to
create awareness of the project.
5. Roadside signs should be placed throughout the watershed to identify project cooperators.
6. Citizen advisory committees provide an outlet for informing local citizens of project
activities and for collecting input on how best to meet local citizen needs. The Long Creek
Watershed Watch citizen advisory committee keeps informed about water quality and
agency activities and participates in citizen monitoring.
7. Regular tours of project sites and hands-on workshops are very effective in transferring
knowledge to project cooperators and others. More than 40 tours have been conducted in
Long Creek for Congressional Staff, local elected officials, students, state and federal
agencies, water quality scientists, media, and citizen groups.
CONCLUSIONS
The Long Creek Watershed Project has been successful in creating awareness of nonpoint
source pollution concerns and in implementing BMPs to address problems. Water quality
monitoring has documented specific problems and will be used to measure project successes.
Project stakeholders have developed a systematic management approach that is being copied
in other watershed projects. Results from this project will be used to more effectively manage
watersheds in other regions.
131
-------
The Monocacy River Watershed Water Quality Demonstration Project:
A Commitment to Water Quality
Theresa Wynn and Patricia Burdette
Cooperative Extension Service
Mark Seibert
Natural Resources Conservation Service
Monocacy River Watershed Water Quality Demonstration Project
Frederick, Maryland
Setting
The Monocacy River watershed covers 899 square miles in the Maryland and Pennsylvania
piedmont (Figure 1). The Monocacy River, a major tributary to the upper Potomac River and,
ultimately, the Chesapeake Bay, was identified as a priority watershed for nonpoint source
(NPS) pollution. Maryland's 208 Water Quality Management Plan for the Middle Potomac River
watershed (1985) noted impairment of surface water beneficial uses due to livestock operations,
lack of compliance with state standards at several wastewater treatment plants, and failing
septic systems. Additionally, the report recommended implementation of an accelerated NPS
control program for nutrients, sediments, and animal waste from agricultural operations in
Frederick and Carroll counties.
In 1990, the University of Maryland System's Cooperative Extension Service (CES), the United
States Department of Agriculture's Natural Resources Conservation Service (NRCS), and the
Farm Services Agency (FSA), in cooperation with the Frederick Soil Conservation District
(FSCD) and the Carroll Soil Conservation District (CSCD), launched the Monocacy River
Watershed Water Quality Demonstration Project (the Monocacy Project). The Monocacy
Project is involved in several large interagency projects combining nutrient management
planning and education by CES, conservation planning by NRCS and the Districts, and water
quality monitoring by the Maryland Department of Natural Resources and the University of
Maryland. While individual agency roles are well-defined, overlap is often necessary to meet
the project goals. Emphasis is placed in individual service and community involvement to gain
and maintain producer trust and support.
Objectives
The primary project goal is to improve surface and ground water resources in the watershed by
accelerating the widespread, voluntary adoption of best management practices (BMPs).
Implementation and Evaluation Approaches
Due to the size of the watershed, three subwatersheds, Linganore Creek, Israel Creek, and
Piney Run-Alloway Creek, were chosen as target areas. These areas have both agricultural
and suburban land uses typical of the watershed. Night meetings were held in each of the
133
-------
PENNSYLVANIA
MARYLAND
FREDERICK
COUNTY
Westminster k CARROLL
COUNTY
Legend
Potomac
River
MONTGOMERY
COUNTY
— — — State Line
Creek/River
County Line
• City
Figure 1. Monocacy River Watershed
original subwatersheds to introduce the Monocacy Project to the community. The Monocacy
Project initially focused on the agricultural community, but as suburban areas grew, the
Monocacy Project included this new audience. In 1995, the Glade Creek subwatershed was
added to the targeted areas, in response to problems with surface contamination of
groundwater that threatens the growing population's drinking water supply. As the demand for
services outside the targeted watersheds increased further, the Monocacy Project ultimately
expanded to include the rest of the watershed.
Information and education activities are of primary importance to the Monocacy Project. The
Monocacy Project promotes BMPs to the agricultural and suburban communities through the
Monocacy Farmer, a quarterly newsletter, presentations to local schools and civic groups,
displays at agricultural field days and county fairs, and tours for both the agricultural community
and non-farm residents of the watershed. The Project often serves as a liaison to the
agricultural community for regulatory officials, foreign visitors, and university researchers.
Several "Understand Agriculture" tours have been conducted to provide an informal forum for
communication, and discussion of agricultural and natural resource policies between farmers
and regulatory agency staff.
The Monocacy Project's Nutrient Management Consultant promotes, develops, and updates
nutrient management plans. This often involves challenging the farmer to a test: A field is split
in half, one half uses nutrient management, and the other uses traditional fertilization practices.
At the end of the growing season, the crop yields are equivalent. The Pre-sidedress Nitrogen-
Nitrate Soil Test (PSNT), a check for soil N during the growing season, is also used to convince
farmers to cut back on initial fertilizer applications. If the PSNT indicates the field needs
additional nitrogen, after following the recommended nutrient management plan, the Monocacy
134
-------
3roject commits to covering the costs of the needed fertilizer. These types of "show-me" tools
are very successful at getting farmers to try new practices.
As BMPs became more recognized, the Project took on the challenge of drinking water
education. A groundwater model, which uses dye to show groundwater and pollutant
movement, provides a highly effective visual tool to teach citizens about groundwater pollution.
A well water testing program was also implemented. Workshops on topics such as well and
septic maintenance, lawn and garden care, water conservation, and water quality around the
home were held.
The success of the Monocacy Project's efforts is traced quantitatively by practices implemented
through the SCDs and by responses to program surveys. Project impact is also evaluated
qualitatively by the increased awareness of water quality, and management options in the
agricultural and suburban communities. Ongoing water quality monitoring efforts will ultimately
provide a scientific evaluation of Project efforts.
Environmental Benefits Measured
Changes in Water Quality
A separate 319 National Monitoring Program Watershed project has collected water chemistry
data since 1993 and is developing a quantitative evaluation of BMP implementation in the
Warner Creek. The most recent quantitative evaluation is found in Shrimohammadi and Felton
(1997). The paired watershed component of this project indicated animal waste management is
still a major concern in the Monocacy Watershed. The upstream-downstream component of the
project indicated nitrate-laden ground water flow was a major vehicle for pollution from animal
waste.
Watershed-scale improvements in water chemistry were not statistically significant between
1993 and 1996. Variations of hydrology were extreme and probably overshadowed any
variations due to BMPs implemented in 1994 or 1995. Also, it is not unreasonable to expect
time lags, such as those classically illustrated in Libra et al. (1987), in water quality response to
BMPs.
BMP Implementation
While water quality monitoring continues, water quality benefits can also be evaluated through
reductions in fertilizer applications and the implementation of management practices. Nutrient
management planning resulted in reduced application of nutrients, usually commercial fertilizer.
Between 1990 and 1996, nutrient applications were reduced by 2.9 million pounds (30.8
Ibs./ac.) of nitrogen and 3.5 million pounds (39.9 Ibs/ac.) of phosphorus. Conservation tillage,
conservation cropping sequence and crop residue management continue to be implemented
throughout the watershed, reducing soil losses. Table 1 shows the practices implemented
within the watershed during the last 7 years.
Qualitative Measures
Much of the success of the Monocacy Project is anecdotal. Recent surveys from a series of
residential environmental workshops indicate these programs provide valuable information and
135
-------
increase participants' awareness of their impact on water quality. The average change in their
knowledge is indicated in Table 2. This environmental awareness often results in more
environmentally conscious decisions on a daily basis.
Table 1. Application of All Erosion & Sediment Control Practices/Activities
Practice / Activity
Soil Conservation and
Water Quality Plans
Conservation Crop Sequence
Conservation Tillage
Cover Crop
Crop Residue Management
Contour Farming
Stripcropping
Critical Area Planting
brassed Waterway
Diversion /Terrace
Pasture / Hayland Management
Pasture / Hayland Planting
Livestock Stream Crossing
[Spring Development /
Watering Trough
Units
Acres
Acres
Acres
Acres
Acres
Acres
Acres
Acres
Acres
Feet
Acres
Acres
Each
Each
1990-1996
Summarv
51,003
28,405
24,815
8,766
17,947
1,497
2,769
149
52
2,188
4,262
572
5
67
The Project always receives a good response at workshops, presentations, and displays.
Requests for repeat performances at other events are common. Additionally, working
relationships with other agencies, as well as local businesses, such as fertilizer dealers and
biosolids applicators, continue to improve.
Table 2. Survey Results
Topics
Household Hazards
Alternative Household Cleaners
Ground and Surface Water Movement
Well Maintenance
Septic System Maintenance
Lawn & Garden Care
Spreader Calibration
Knowledge
Before
average
average
below average
below average
below average
below average
none
Knowledge
After
above average
above average
above average
above average
above average
above average
above average
Lessons Learned
The Monocacy Project has worked closely with watershed citizens for seven years, helping
implement conservation and water quality programs and apply innovative technologies. During
136
-------
this time, many lessons were learned about interagency cooperation and successfully
promoting BMPs.
The first, most important lesson learned was, to make dramatic changes, community support is
a must. This can be gained by working with local citizens and agencies from the beginning.
Starting a new project without local input and initiation can imply current programs are
inadequate, leading to resentment and antagonism from local agencies.
Trust, name recognition and community involvement are necessary for participation in
workshops and field demonstrations. The short-term nature of the project has made these
difficult to achieve. Citizens do not want to rely on an agency that will not exist in a few years.
Also, the temporary nature of the staff positions has caused tremendous staff turnover. This
also can jeopardize community relationships and reduced the effectiveness of the project.
Targeting just one sector of the population can also cause friction. After a few years of targeting
the agricultural community, farmers began to feel they were being singled out. By expanding
the educational programs to suburban and urban communities, the Monocacy Project
emphasized that all citizens play a role in water quality, not just farmers. Additionally,
introducing residents to agriculture through tours and presentations promoted mutual
understanding and cooperation.
Getting BMPs on the ground requires individual assistance, flexibility, creativity, and money.
Many agricultural operations are family farms. It often takes individual attention and
perseverance to change traditions in a family business. Because each farm and farmer are
different, BMPs need to be custom-fit. BMPs must be simple and low maintenance. Achieving
this requires flexibility and creativity on the part of agency personnel and programs. Often,
institutional guidelines and agency restrictions make it difficult to implement innovative,
common-sense solutions.
Lastly, change usually has a price tag attached. Many BMPs do not benefit the farmer directly
or do not yield short-term returns. Because of this, public funding for the protection of public
resources is required. Cost-share programs need continued support and increased flexibility.
New, innovative management practices are not funded by traditional cost-share programs. This
lack of funding often impedes the adoption of new technology.
References Cited
Libra, R. D., G. R. Hall berg, B. E. Hoyer. 1987. Impacts of agricultural chemicals on
ground water quality in Iowa. In: Ground Water Quality and Agricultural
Practices, D. M. Fairchild (ed.), Lewis Publishers: Chelsea, Ml. pp 185-215.
Maryland. 1985. Revised 208 Water Quality Management Plan for the Middle
Potomac River Basin: Draft. Maryland Department of Health and Mental
Hygiene, Office of Environmental Programs: Baltimore, Maryland.
Shirmohammadi, A. and G. K. Felton. 1997. Assessment of watershed water quality
using USEPA National Monitoring Design. ASAE paper no. 97-2006. ASAE:
St. Joseph, Ml. 49085-9659
137
-------
Environmental Assessment For
Real Estate Professionals
by Bob Broz
University of Missouri - Outreach and Extension, Water Quality
ABSTRACT:
Much of the agricultural land surrounding cities and towns has been subdivided for urbanization
and development. Real estate professionals and land appraisers can learn to recognize possible
environmental hazards associated with agricultural practices by using an assessment tool such as
the Farmstead Assessment System (Farm-A-Syst). A course was offered to inform real estate
professionals about potential environmental hazards, present regulations and the importance of an
environmental audit when buying or selling agricultural land.
INTRODUCTION:
Agricultural land surrounding cities and towns are being divided into smaller parcels for individual
home sites, subdivisions and industrial development. Many of the activities that took place on
farmsteads were exempt from certain regulations or were considered as acceptable practices.
Representatives from the real estate profession, land appraisers, lending institutions, Department
of Health, Missouri Department of Natural Resources, University Outreach and Extension, and
the Mark Twain Water Quality Initiative worked cooperatively to develop a course that would
increase awareness of regulations and environmental concerns in Missouri. The Farmstead
Assessment System, Farm-A-Syst, was used as a model to identify specific areas of potential
environmental concern on agricultural property.
The Farm-A-Syst is an assessment tool used to identify potential environmental problem areas
that may occur from practices on a farmstead. The Farm-A-Syst assesses wellhead protection
practices, pesticide storage and handling practices, petroleum storage and handling practices,
fertilizer storage and handling practices, on-site sewage systems, hazardous waste disposal,
livestock manure management practices and farmstead site evaluation.
The course, Environmental Assessment for Real Estate Professionals, is approved through the
Missouri Real Estate commission and the Missouri Land Appraisers commission for six hours of
certification credit. The class provides real-estate professionals and their clientele with the ability
to identify pollution risks on properties and to have some indication of how those risks may affect
land values. The materials included in the course notebook offer cost-effective, voluntary action
to prevent pollution, reduce potential liability and protect personal health. The course also
develops awareness of environmental issues when assessing, buying and selling rural properties.
The course was sponsored by University Outreach and Extension in support of the Mark Twain
Water Quality Initiative (MTWQI). MTWQI is a multi-agency organization that combines the
139
-------
talents and efforts of the Missouri Department of Natural Resources, the Natural Resources
Conservation Service, the Missouri Department of Conservation and University Extension.
METHODS/EDUCATIONAL CURRICULUM:
Several factors influenced the development of the "realtors" course: 1.) the number of rural
landowners selling property for agricultural production, 2.) the selling of rural property in small
parcels for an expanding urban population, 3.) the expansion of urbanization and industrialization
into rural property surrounding cities, and 4.) the change in attitude of where people want to live.
The course curriculum was developed to go from general state regulations to more specific areas
of concern. The class curriculum is as follows:
1. Overview and Pre-Test - 20 minutes
Class participants are given a pre-test containing questions from each of the major areas. This
prepares participants for the types of information the course will offer. An overview of the course
content and notebook allows participants some hands-on activities in using the notebook.
2. Federal and State Regulations - 50 minutes DNR and DoH
The Missouri Department of Natural Resources and Missouri Department of Health officials give
an overview of regulations that affect water and environmental quality. A regulatory directory
has been developed listing areas of concern, action required, agencies of authority, telephone
numbers and environmental and health considerations. The directory is a summary of those
activities associated with agricultural and rural housing development that can be harmful to water
and environmental quality.
3. Site Evaluation - 45 minutes
By using the USDA pamphlet "Making a Solid Investment" and county soil survey reports
developed by the Natural Resources Conservation Service (NRCS), class participants are asked to
determine if a section of agricultural land located on the soil survey map is suited for subdivision
development. By determining the soil characteristics and answering the questions in the
pamphlet, class participants can determine if the land is suited for development.
4. Private Drinking Water Wells - 45 minutes, presented by DNR and DoH
Public drinking water supplies are not always readily accessible to rural locations. The Missouri
Department of Natural Resources and the Department of Health go through the regulations that
are applicable to homeowners with private wells. The Department of Health discusses
recommendations on wellhead protection, location and water testing that should be considered
when buying or selling rural property.
5. On-Site Sewage Systems - 40 minutes, presented by DoH
Missouri's on-site sewage regulations are discussed by local Department of Health officials. Class
participants are shown examples of working and non-working on-site systems. Guidesheets
produced by University Outreach and Extension and technical bulletins by the Missouri
140
-------
Department of Health explain the correct construction, maintenance and care of conventional on-
site sewage systems. By being able to identify potential problems with on-site sewage systems,
real estate professionals can determine if costly repairs or installation costs are needed on the
property.
6. On-Site Sewage System Alternatives - 20 minutes, presented by University Extension
Much of Missouri's soil profile does not meet the minimum percolation standards for conventional
on-site sewage systems. Basic information on alternative systems is reviewed for building
requirements, availability of systems, maintenance and cost. Approved alternatives covered in the
class are: drip irrigation systems, mound systems, sand filter systems, low-pressure pipe systems
and submerged flow wetlands. Many older homes being sold with 1 or 2 acre lots may require
the use of an aeration septic system with an alternative drainage field to meet state guidelines.
Technical bulletins developed by the Missouri Department of Health are included in the
participants' packet of materials.
7. Solid and Hazardous Waste - 40 minutes, presented by University Extension
The two most common means of disposal of solid waste in rural areas has been burning and
dumping. As rural agricultural property has been subdivided, new owners may find themselves
with an unwanted farm dump or burn pile on their property. Property values can be greatly
hindered by the existence of a farm dump or burn pile. The correct procedure for cleaning up or
closing a farm dump is discussed. Information on recycling and hauling refuse to local dumps is
discussed as alternatives to burning or on-site dumping.
Another area associated with rural ownership is illegal dumping by others. The legal
responsibility for cleaning up solid waste that is illegally dumped by others is discussed. Even
though the property owner may not be at fault, they must shoulder the burden of clean up and
proper disposal of unwanted trash.
Many areas now have refuse hauling available, but for those areas where it is not available,
participants will look at ways of correctly disposing of hazardous materials that are generated or
found on rural properties. A pamphlet on household hazardous waste is included in the notebook
to help identify common household products that can be detrimental to the environment.
8. Hot Spots - 40 minutes, presented by University Extension
Practices that were generally considered acceptable on farmsteads may create costly
environmental problems. The "hot spots" section of the class looks at four basic areas that are
common on many rural farmsteads and can be potentially costly to correct
In the past, most farming operations had on-farm fuel storage for convenience. These areas, if not
contained in secondary containment structures, could have an accumulation of petroleum spills
and leaks in the surrounding soil. Petroleum laden soil can be environmentally dangerous and
very costly to correct. Knowing what to look for and where to look for on-farm fuel storage may
save a real estate client from costly environmental clean up.
Livestock lagoons were commonly used on farms for storing animal waste. Once the livestock
141
-------
operation was discontinued, most lagoons were left in place and not properly closed. Participants
are made aware of the regulation affecting new Confined Animal Feedlot Operations and the
procedures for closing a lagoon that is no longer in use.
The storage and handling of pesticides and fertilizers were a common practice on many farms.
Participants can follow the questions in the Farm-A-Syst to determine if there was proper
handling of pesticides and fertilizers at the farmstead or if some environmental damage may have
been done. Improper mixing of pesticides, without using some form of well protection device,
may have contaminated farm deep wells and ponds.
Many rural homesteads had drilled or hand dug wells or cisterns that have been replaced with
deep wells or rural water supplies. Many of these wells have been left abandoned and can create
environmental and safety concerns. Class participants are given information on the environmental
concerns, liability, safety and health concerns of abandoned wells. Information concerning the
proper method of closure and the estimated cost are reviewed.
9. Using the Farm-A-Syst - 20 minutes, presented by University Extension
The purpose of the class is to create awareness of hidden concerns or environmental problems
that can affect the value of rural property. By using an assessment tool, the Farm-A-Syst, real
estate professional can identify areas of concern that may decrease the market value of rural
properties. Class participants are introduced to the Farm-A-Syst packet of information and work
through one section to get an understanding of how to use the assessment form. The Farm-A-
Syst is divided into seven areas with worksheets and accompanying fact sheets to help class
participants ask the right questions when assessing rural properties. The site assessment and
overall assessment make up the complete Farm-A-Syst packet.
10. Summary, Post Test and Evaluation - 25 minutes
RESULTS:
The course has been offered three times with a total of 68 people having attended the class.
Evaluations have ranked very high in determining the usefulness of the program.
Of those attending the course, 45 attended the course for professional certification, 30 attended
for personal/professional development, and 3 attended for other reasons.
The professional roles of those attending were as follows:
Appraisers - 30 Educator - 1
Bankers-5 Broker-17
Commissioner -1 Investor - 2
Health Professional - 3 Regulator - 3
Realtors - 12 Ag Related Profession - 1
A pre and post test was given to participants to determine the level of knowledge on
142
-------
environmental regulations and issues they had before and after the course was conducted. Pre-
test scores ranged from 44 percent to 89 percent with an average of 69.14 percent. Post-test
scores ranged from 56 percent to 100 percent and had an average of 82.92 percent.
When ranking the course, using a scale of 1 being very low satisfaction to 5 being very high, the
participants gave the class an average of 4.35 for overall quality of the course compared to past
and/or comparable experiences.
When asked if the program should be repeated for others, 66% of those responding to the
evaluation strongly agreed, and 34% agreed. Other information indicated that 75% felt a more
in-depth program should be offered on the subject areas, and 30% felt a more basic program
should be developed.
The nature of the course is to heighten awareness of environmental concerns and regulations
about rural properties. Test scores from the class indicate that increased awareness has occurred.
The direct effect the information will have on improving environmental problems on rural
property has not been documented at this time. But by increasing the knowledge level and
showing the correlation between land values, liability issues and environmental concerns, we hope
to see a strong voluntary action to correct some environmental problems when assessing property
for loans or sales.
A follow-up survey is being sent to those attending to see how the subject matter covered in the
program will be used and if the information on the Farm-A-Syst will be helpful in determining the
marketability of rural properties. The survey will also ask for information concerning the most
common areas of concern when doing an environmental assessment on rural property.
SUMMARY:
Agricultural properties are being sold in smaller sections for rural homes, subdivisions and
industrial development. Real estate professionals and appraisers need to be aware of potential
environmental problems that are associated with agricultural properties and farmsteads. These
problem areas may affect land vales. The use of an assessment tool, such as the Farm-A-Syst, to
help identify these areas of concern is very useful.
Though designed to be done by rural land owners, the Farm-A-Syst can be used as an
environmental assessment tool by real estate professionals and land appraisers. The assessment
worksheets walk the user through the different areas of the farmstead that generally have been
known to be at risk of causing environmental problems. The fact sheets for the Farm-A-Syst can
be used to review state and federal regulations and to list contact names and numbers if you have
questions concerning specific situations. The need for public education for environmental and
regulatory concerns of rural property is in great demand. Programs developed and presented by
local agency people give credibility to courses like the Environmental Assessment for Real Estate
Professionals and offer needed information to the public.
143
-------
Sand Mountain-Lake Guntersville Hydrologic Unit Area
Progress Report: 1990-1996
Jesse LaPrade and James Hairston
The Alabama Cooperative Extension System
Auburn University, Auburn, Alabama
Background
The Sand Mountain-Lake Guntersville Watershed covers approximately 400,800 acres in
Northeast Alabama. Major land uses include cropland, pastureland, and forestland. Principal
crops include corn, soybeans, and wheat. Minor crops include hay, grain sorghum, and
commercial vegetables. Approximately 75% of the cropland is subject to excessive
erosion. In addition to row crop production, broilers, eggs, cattle, and hogs are produced in
large numbers.
Surface drainage in the watershed flows north-westward into Guntersville Lake on the
Tennessee River. Guntersville Lake is the major source of water-based recreation in the
area.
Baseline Water Quality Problems
High nutrient concentrations, bacterial pollution, and excessive sediment were the primary
water quality problems at the inception of this project. These problems were affecting Lake
Guntersville, the streams of the watershed, and the area's ground water. Studies by the
U.S. Geological Survey (USGS) revealed that more than ten tons of nitrogen (N) per day
were discharged from Town Creek during high flows. Bacterial pollution of ground water
was a growing problem with more than half of all the well water tested showing pollution
by fecal coliform bacteria,and more than 30% of the wells tested had nitrate levels
exceeding the EPA limit for drinking water. The total sediment load (suspended sediment
plus bedload) of the tributary streams was estimated to be 550,000 tons per year. Studies
showed that 90% of the estimated soil erosion occurred on cropland.
The Sand Mountain-Lake Guntersville Hydrologic Unit Area, a USDA water quality
protection project, began in 1990. The project promoted environmentally and economically
sound agricultural practices by offering a combination of educational, technical, and
financial assistance to producers. Participation is voluntary.
145
-------
Accomplishments
Community Outreach
The Alabama Cooperative Extension System (ACES) county staff have held 123 tours and
or special water quality training sessions to educate 17134 attendees on water quality
issues during the project period (1990-1996). In addition, ACES staff generated 628 news
articles and have held 6082 direct county agent consultations with clients on water quality
issues or conflicts during the project period.
The Alabama Cooperative Extension System State staff developed 148 water quality
publications that were used in this project and provided leadership to maximize the success
of the educational effort.
Nutrient Management
This effort has led to an estimated 90 % of all row crop producers having currently
developed farm level nutrient management plans.
Animal Waste Management and Pollution Prevention
An estimated 25% of all livestock producers in the watershed now have pollution
prevention plans for animal waste management and an estimated 17% currently have
animal waste management systems in place. Approximately 49% of all poultry producers
installed new or upgraded dead bird disposal facilities, representing an estimated 38% of j
the total production in the project area.
i
Erosion and Sediment Control
Erosion and sediment control practices placed on more than 3771 acres include '
conservation cropping sequence, conservation tillage, contour farming, cover and green
manure crop, crop residue use, and pasture and hayland planting. There have been over
9940 feet of terrace installed to control erosion.
Project impacts \
With approximately 79% of total planned nutrient management applied, current
improvements have led to a 40% reduction in nitrogen and phosphorus entering surface j
and ground water.
Technical assistance from the Natural Resource Conservation Service (NRCS), financial
assistance from the Farm Service Agency (FSA) and educational assistance from the •'
Alabama Cooperative Extension System have helped farmers and citizens decrease nutrient
runoff by approximately 202 tons of nitrogen and 38 tons of phosphorus on 21,300 acres
of crop and pastureland, annually.
Overflow from all animal waste lagoons has been significantly reduced. |
146
-------
There has been a significant increase in public recognition of farmers' efforts to protect
water quality.
An Executive Report issued in November 1996 by the Watershed Monitoring and
Evaluation Committee shows that overall water quality of streams is improving.
This report documents improvements in the key water quality parameters such as
increasing pH values, decreasing total organic nitrogen levels, and increasing dissolved
oxygen levels.
The water quality also has improved significantly in Lake Guntersville. State health
standards for body contact recreation have consistently been met at the major swimming
areas of Town, Short, and South Sauty Creek embayments. Overall fish population
dynamics have increased or remained stable, with no fish consumption advisories issued.
While water quality improvements have occurred, there remain some major areas of
concern. Increasing phosphate levels in all monitored surface water streams in the
watershed have been noted. Bacterial levels monitored in surface water streams remain too
variable to establish a trend. Private well water monitoring continues to indicate significant
contamination by fecal coliform bacteria. Sedimentation entering Lake Guntersville
continues to cause concern. Rates of sedimentation have increased for the period 1961 to
1996 compared to the period 1940 to 1961. Currently the total reservoir storage volume
losses are at approximately five percent.
The Sand Mountain-Lake Guntersville Hydrologic Unit Area project has made considerable
progress since it officially began in 1990. Goals to be met include:
• Determine the source of phosphate levels increasing in surface water
streams and reduce levels.
• Delineate the source of fecal coliform bacteria in area well water. If the
bacteria are coming from local septic systems, corrections can be made. If
the bacteria are originating from animal waste, a concentrated effort to
reduce levels must be made.
• Set a goal of reducing sedimentation loading in Lake Guntersville of 5% per
year, compared to 1996 monitoring results, through the year 2000 or 20%
reduction over the period. This goal can be met by continuing to emphasize
reduced tillage in agriculture, utilization of best management practices to
reduce sedimentation on construction sites and practicing best management
practices in timber harvesting.
All Federal and Alabama state agencies participating in this watershed project have helped
make this project successful while the local farmers and citizens have made it happen.
Continued success in reaching goals above can only occur through local participation. All
federal and state personnel are fully committed to providing support for the project. We all
147
-------
look forward to the day that monitoring activity shows that this watershed is pristine
clean!
The 1996 Monitoring and Evaluation Committee Members
SM-LG Conservancy District Alabama Dept of Environmental
Raymond Hamilton Management
Stanley McClendon Steve Foster
Jackson County SWCD Charles Sweatt
John Brown Alabama Division of Game and
Alabama A&M University Fish
David Mays Keith Floyd
Karamat Sistani Auburn University
Marshall County Health Department Cliff Webber
Freeman Smith Tennessee Valley Authority
Natural Resources Conservation Services Doug Murphy
Jim McCullough Charlie Saylor
Jerry Wisener Frank Sagona
Federal and State Organizations participating in the Sand Mountain-Lake Guntersville
Watershed Project
• USDA Agricultural Stabilization and Conservation Service
• Alabama Department of Conservation and Natural Resources
• Geological Survey of Alabama
• The Alabama Department of Public Health
• Alabama Association of Conservation Districts
• Top of Alabama Regional Council of Government
• USDA Soil Conservation Service
• USDA Extension Service
• United States Geological Survey
• The United States Environmental Protection Agency - Region IV
• DeKalb, Etowah, Jackson, and Marshall Counties Soil and Water
Conservation Districts
• The Alabama Department of Environmental Management
148
-------
• Sand Mountain-Lake Guntersville Watershed Conservancy District
• The Alabama Cooperative Extension System
• The Alabama Forestry Commission
• The Tennessee Valley Authority
• Alabama Department of Agriculture and Industries
• Alabama Soil and Water Conservation Committee
This project is based upon work supported by the Extension Service, U.S. Department of
Agriculture (CSREES).
149
-------
The Muddy Fork HUA - A Water Quality Success Story
K.A. league
University of Arkansas Cooperative Extension Service, Fayetteville, Arkansas
B. Moore
Dale Bumpers Small Farms Research Center, Booneville, Arkansas
Introduction
The rolling hills of the Ozarks in Northwest Arkansas are known for their scenic beauty and
abundant ground and surface water resources. Vast groundwater aquifers supply drinking
water to a majority of the rural population while lakes provide municipal drinking water for
urban residents. In addition, networks of streams and lakes provide superb fishing, boating,
and swimming opportunities for local residents as well as tourists. Ground and surface
water in the Illinois River watershed, occupying portions of four counties in Arkansas and
Oklahoma, serve these purposes.
Due to aesthetic values and the demand for drinking water and recreational water activities,
improvement and protection of water quality along the Illinois River has been a priority for
both states in recent decades. In the late-1980's, Lincoln Lake, a reservoir constructed on
one of the Arkansas tributaries to the Illinois River, began demonstrating signs of
eutrophication. Residents of the town of Lincoln complained that the water had a peculiar
taste and odor. Water data collected by the United States Geological Survey on Lincoln
lake during the years of 1988-1989 suggested that the lake was receiving high nutrient
loads. Elevated levels of nitrogen and phosphorus were also found in other nearby lakes
and streams.
Lincoln Lake is located in the Moore's Creek sub-basin of the Illinois River watershed.
Numerous confined poultry, swine, dairy, and beef cattle operations are on the land
surrounding the lake. Because animal waste produced in the basin is land-applied almost
exclusively to pastureland, runoff from these pastures was thought to be a major source of
the nutrient and bacterial loads reaching surface water supplies. In an effort to restore
beneficial uses of water resources in this sub-basin, the USDA funded the Moore's Creek
Water Quality Project in 1990.
The Project
The Moore's Creek Water Quality project began as a joint effort among the University of
Arkansas Cooperative Extension Service (CES), the Natural Resource Conservation Service
(NRCS), and the Farm Services Agency (FSA) in cooperation with other state and federal
151
-------
agencies. The project area initially covered approximately 15,000 acres and included
Lincoln Lake and two of its tributaries, Moore's Creek and Beatty Branch. After the first
year, however, the project area was expanded to encompass a 47,122-acre area known as
the "Muddy Fork of the Illinois River Hydrologic Unit Area (HUA) and included three ,
reservoirs and a network of small tributaries to the Illinois River. J
The primary objectives of the project were to reduce nutrient, bacterial, and sediment I
transport to ground and surface water supplies while maintaining the economic viability of •
local agricultural operations. These objectives were to be accomplished through the !
voluntary adoption of nutrient management, waste utilization, and conservation Best i
Management Practices (BMPs). !
i
Project Methods '
One-on-One Farm Visits
Because the Muddy Fork HUA Water Quality Project has been executed through non-
regulatory agencies such as the CES, there has been strong interest and participation from j
landowners. Most farmers welcome one-on-one visits from NRCS and CES staff who
assess land use and make recommendations on animal production systems. Local CES
personnel relay results of BMP research conducted at the University of Arkansas and State
Specialists provide additional support in answering questions and providing publications on \
proven BMPs. The BMPs that are emphasized involve small management changes that
reduce the impact of agriculture with little cost to the farmer in terms of time and money.
NRCS staff offer technical assistance in the form of nutrient management, waste utilization, >
and conservation plans that are tailored to each farmer's operations. In years when cost- ;
share funding was available, FSA has provided financial assistance to producers who
commit to Long Term Agreements.
On-Farm Demonstrations
Although BMP awareness has been growing, many landowners are often reluctant to try
new management strategies. Some feel that because BMPs were developed on University
research plots, the geology, soil type, and slope may be quite different from those on their
own land in the watershed. In this respect, on-farm demonstrations have been extremely
successful in promoting the effectiveness of BMPs as residents witness the results on land
and operations similar to their own. When a farmer hears his neighbor speak of high yields
or monetary savings due to the adoption of a BMP, he is often ready and willing to
implement the management practice on his own farm.
Educational Programs, Publications, and Media Attention
Many residents of the watershed are willing to take an active role in the protection of their
water supplies, but are not sure what to do. Information has been disseminated to the
general public through fact sheets, videos, newspaper articles, and television spots on how
water quality can be improved and preserved in the karst terrain of Northwest Arkansas.
Educational programs on the water cycle, water conservation, and water quality have been
presented to civic groups, community leaders, school children, and 4-H clubs. For those
directly involved in agriculture, field tours of on-farm demonstrations, annual meetings,
152
-------
newsletters, and presentations at Farm Bureau and Cattleman's Association meetings have
been most effective. Presentations using a groundwater simulator have been well received
in elementary schools as a valuable teaching tool. As "contaminants" move through the
model, students take home the message that activities in one area can have an impact on
other seemingly distant water supplies which are all connected through the water cycle.
To reach more diverse audiences, display boards have been used effectively at county fairs,
local festivals, banks, and libraries to promote the Muddy Fork of the Illinois River HUA
Water Quality Project and general water quality awareness.
Project Accomplishments
Over the past seven years, the Muddy Fork HUA Water Quality Project has been a success.
The impact of this project is evident in the numbers of BMPs installed, the size of the soil
and water database generated from intensive sampling efforts, and the overall increase in
public water quality awareness.
BMP Implementation
Since 1990, nearly 250 farm conservation plans have been completed within the project
area. The voluntary adoption of BMPs which emphasize nutrient management, waste
utilization, and conservation was the primary method to improve and protect ground and
surface water supplies in the HUA. Over the past seven years, whole-farm nutrient
management plans were established on more than 32,000 acres. To better identify the
fertility status of their pastures, farmers have been encouraged to take advantage of free
soil testing services through the University of Arkansas. With this management tool,
fertilizer recommendations were based on forage uptake and were used to supplement
nutrients contributed from land-applied animal manures.
At the start of the project, manure handling may have been the most neglected component
of agricultural production. But, through the cooperative efforts of USDA agencies, 134
waste management systems, including 35 waste storage structures (concrete lagoons and
dry stacking sheds) have been installed. Previously, most waste management decisions
were based on convenience and cost, with little consideration for their impact on water
quality. However, through educational efforts, producers learned that by reducing waste
application rates or splitting applications, utilizing alternative forages, and installing
vegetative grass filter strips, nutrient leaching and runoff from land-applied wastes could be
reduced. Producers were also pleased to realize that BMPs such as these resulted in
greater forage yields, higher feed conversions, and substantial economic savings. In much
the same way, chicken house and litter truck calibrations highlighted the importance of
knowing the quantity and quality of animal manures that were applied to pastures.
The most successful BMP demonstrations may have been the replacement of dead bird pits
with composting units. Although dead bird pits were an accepted method of carcass
disposal since the 1950's, many pits failed to properly decompose the birds. At a study site
located in the Muddy Fork HUA, University of Arkansas researchers found that groundwater
moving through a typical pit transported pollutants such as ammonium, nitrate, phosphate,
organic carbon, and pathogens such as Salmonella to aquifers. At the same time, dead
bird composting units were found to consistently convert the birds into a valuable organic
153
-------
fertilizer in dry, above-ground bins so that groundwater was protected. Today, disposal pits
are no longer being installed in Arkansas and more than 30 composters have been
constructed in the Muddy Fork HUA.
Database Development
One of the primary reasons that the Muddy Fork HUA Water Quality Project was proposed
was that there was insufficient ground truth data to help verify sources for water quality
degradation. Without this type of information, it was difficult to determine the magnitude of
nutrient, sediment, and bacterial contributions from agricultural operations in the HUA. In
order to begin gathering data, the project budget included funding for water, soil, manure,
and forage sampling and analyses.
As the project progressed, a substantial database has accumulated. More than 21,000
acres within the HUA have been soil tested and more than 400 water samples have been
collected from streams, springs, ponds, and wells. Because the results of these analyses
have been stored in DataPerfect, it has been difficult to identify and interpret water quality
trends. However, by using IDRISI, a Geographical Information Systems software, and
GeoExplorer, a hand-held Global Positioning System, we are in the process of converting
the data into digital maps on computer that will depict the impact of BMP implementation on
water quality within the HUA over time.
Water Quality Awareness
Perhaps the greatest benefit of the Muddy Fork HUA Water Quality Project was an increase
in public awareness of water quality, specifically the role of waste utilization and nutrient
management in watershed-scale water quality improvement and protection efforts. Several
thousand individuals have participated in tours, annual meetings, and educational programs
while more than a 100 publications have been developed and distributed to residents in the
HUA. Four videos of BMP demonstrations and one which highlights the Muddy Fork project
have been taped and used in presentations to diverse audiences. Additional outreach
through television news spots and regular news articles has also helped to promote interest
in the project and water quality protection.
Water Quality Monitoring
While consistent water quality monitoring may help define the impact of BMP
implementation, regularly scheduled, site-specific stream, lake, or well monitoring was not a
component of this project. It was decided that the best use of funds for our purposes was
to alleviate the cost of well, pond, stream and spring water analyses as they were requested
by landowners in the watershed.
Coincidentally, an on-going USEPA-funded 319 grant project has evaluated the impact of
BMPs on water quality in Moore's Creek and Beatty Branch. Monitoring results from five
sites along these two streams in the HUA indicate that concentrations of NH3-N and Total N
have decreased significantly over time. Other parameters such as NO3-N, PO4-P, and TSS
have remained fairly stable. Data indicate that concentrations of Total N decreased by 50
to 75% from 1991-1994 and continued to decrease by 50 to 73% during 1995 to 1996.
These results seem to indicate that progress has been made from implementation of BMPs.
154
-------
Lessons Learned
As this project has progressed, it has gained momentum. One of the key reasons that
farmer involvement and participation has been so great is that from the start, the project
was promoted as theirs. Although the ultimate goal may have been to improve ground and
surface water quality, the methodology was to teach and train residents in ways that they
could protect their own resources.
When positive comments and recognition were offered, credit was always given to the
producers. Ruth Parker, the Mayor of Lincoln, wrote a letter stating her appreciation of the
HUA project and credited the project efforts for contributing to the improvements in the city's
drinking water quality. Letters such as these and certificates of appreciation were always
shared with HUA residents so that they realized that their voluntary efforts were noticed and
valued. Several landowners in the watershed, including Hollis and Vera Barker, State
Representative Jerry Hunton, Alan and Robin Reed, and Brian Weaver, have received
Environmental Excellence Awards from the EPA - Region 6. In fact, Brian Weaver went on
to win a National Conservation Award for his role in environmental management.
Initial planning and promotion of the project was designed to include key agricultural leaders
and prominent members of the community as well as local city officials, policy-makers, and
representatives from the three USDA agencies.
Continual input and feedback are essential to effectively addressing dynamic needs of
watershed residents. Consequently, a willingness to modify and adjust objectives as the
project progresses is critical to the continued success of the project.
Conclusions
Now that many of the agricultural non-point sources have been identified and are being
addressed through BMPs, as the project continues, the focus may broaden to include some
of the smaller, less apparent sources of water quality degradation. Currently, a
demonstration is underway that will use GIS and a Global Positioning System (GPS) as
tools for whole-farm nutrient management planning. In addition, future demonstrations may
include the stabilization of stream banks to reduce erosion, especially along areas where
cattle have direct access to the waterways. Other areas of focus will include the impact of
septic systems, gravel roads and urban runoff in the cities of Lincoln and Prairie Grove.
The Muddy Fork of the Illinois River HUA Water Quality Project has been one of the most
successful of its type due to the combined efforts of area residents, state and federal
agencies, local government, and the University of Arkansas. With continued cooperation,
promotion, and information exchange, the project will further demonstrate that individual
efforts can combine to improve water quality within an entire HUA.
155
-------
Effectiveness of Barnyard Best Management Practices in Wisconsin
Todd D. Stuntebeck
U.S. Geological Survey, Madison, Wisconsin
A study has been designed to (1) compute the loads of selected constituents coming from a single critical
barnyard in each of two watersheds in Wisconsin and (2) determine whether implementation of barnyard
Best Management Practices (BMPs) at each site improves water quality in the receiving streams. Using
an upstream-downstream experimental design, investigators are analyzing data for discrete water
samples collected automatically during storm-runoff periods before and after implementation of barnyard
BMPs at Otter Creek and Halfway Prairie Creek. Before BMPs were implemented, downstream loads of
total phosphorus, ammonia nitrogen, biochemical oxygen demand (BOD), and microbial loads of fecal
coliform bacteria were statistically greater than upstream at each site. Downstream loads of suspended
solids were statistically greater than upstream loads only at Otter Creek. Data collected after implementa-
tion of the barnyard BMPs indicate improvements in water quality at both sites. The barnyard BMP at
Otter Creek has reduced loads of suspended solids by 81 percent, total phosphorus by 88 percent,
ammonia nitrogen by 97 percent, BOD by 80 percent, and microbial loads of fecal coliform bacteria by 84
percent; the barnyard BMP at Halfway Prairie Creek has resulted in 67-, 89-, 94-, 91-, and 24-percent
reductions, respectively.
Introduction
The Nonpoint Source Water Pollution Abatement Program was created in 1978 by the Wiscon-
sin Legislature. The program goal is to improve and protect the water quality of lakes, streams,
wetlands, and ground water within selected priority watersheds by controlling sources of
nonpoint pollution. For each selected watershed, the Wisconsin Department of Natural Re-
sources and county Land Conservation Departments draft management plans that guide the
implementation of pollution-control strategies known as Best Management Practices (BMPs).
These plans summarize land-use inventories, describe the results of pollution-source modeling,
and suggest pollution-reduction goals. The U.S. Geological Survey, through a cooperative effort
with the Wisconsin Department of Natural Resources, is studying changes in water quality that
result from the implementation of BMPs. State and county officials are then comparing the
results to the watershed plans to assess progress and determine whether goals are being
realized. Information gained from these studies will help managers make informed decisions
regarding BMP implementation in other priority watersheds.
As part of this monitoring program, an upstream-downstream (above-and-below) experimental
design is being used to (1) compute the loads of selected constituents coming from a single
critical barnyard in each of two watersheds and (2) determine whether implementation of BMPs
at each site improves water quality in the receiving streams. This paper focuses on the methods
used to collect data, the magnitude of the differences in constituent loads between upstream
and downstream sites before implementation of BMPs, and the degree to which the investigated
barnyard BMPs reduced constituent loads to the stream.
157
-------
Study Areas and Data Collection
Two sampling stations were established on each stream (fig. 1). One station is upstream from a
single barnyard-runoff source, and the other station is downstream from that same source.
Station locations were chosen to minimize inflows other than runoff from each barnyard. The
barnyards investigated were identified by each watershed plan as critical nonpoint sources
based on herd size, lot size, proximity to the stream, and downslope overland flow characteris-
tics.
Otter Creek, one of the Section 319 National Monitoring Program projects, is within the
Sheboygan River Priority Watershed, 15 miles west of Lake Michigan, in east-central Wisconsin
(fig. 1). The drainage area of Otter Creek is 9.2 square miles at the downstream sampling
station, and land use in the watershed is 67 percent agricultural (Bachhuber and Foye, 1993).
The stream is typified by reduced aquatic habitat due to excessive sediment and nutrient load-
ing from nonpoint sources—mainly cropland and dairy operations—and recreation is limited by
degraded fisheries and by high fecal coliform counts. The investigated barnyard on Otter Creek
is a dairy operation with approximately 50 cows. Upstream and downstream sampling stations,
each equipped to continuously monitor streamwater levels and to collect discrete water
Otter Creek
Watershed
Halfway Prairie Creek
Watershed
3 KILOMETERS
EXPLANATION
A Sampling station
n
2 KILOMETERS
Figure 1. Location of upstream and downstream sampling stations for Otter Creek (right) and
Halfway Prairie Creek (above).
158
-------
samples, were established on Otter Creek in March 1994. Water samples are collected by
means of a refrigerated water-quality sampler, which is activated by the rise and fall of
streamwater levels.
Halfway Prairie Creek is within the Black Earth Creek Priority Watershed, 20 miles northwest of
Madison, in south-central Wisconsin (fig. 1). The drainage area of Halfway Prairie Creek is 16.1
square miles at the downstream sampling station, and land use in the watershed is 60 percent
agricultural (Eagan and Morton, 1989). Like Otter Creek, this stream is typified by reduced
aquatic habitat due to excessive sediment and nutrient loading and by high fecal coliform
counts. The investigated barnyard on Halfway Prairie Creek also is a dairy operation, with
approximately 100 cows on site. Upstream and downstream sampling stations were established
on Halfway Prairie Creek in April 1995. At the upstream sampling station, streamwater levels
and precipitation are continuously monitored, and discrete water samples are collected by
means of a refrigerated water-quality sampler. The downstream station is equipped to collect
water samples only.
Upstream-downstream sampling designs have the inherent potential for upstream loading
sources to mask the effects of the investigated source, because individual inputs from the
investigated source are often small compared to the cumulative inputs from upstream drainage
areas (Spooner and others, 1985). To reduce the potential for this problem, two enhancements
were added to the sampling design at Halfway Prairie Creek. First, the water-quality samplers
were activated by precipitation and were programmed to collect time-integrated samples for an
initial period. (This enhancement was also added to the Otter Creek sampling design for the
post-BMP monitoring period.) After the initial period, samples were collected in response to the
rise and fall of streamwater levels, as in the pre-BMP setup at the Otter Creek stations. Two
benefits of the enhanced approach are that (1) it allows for sampling of barnyard runoff in the
receiving stream before streamwater-level increases can be sensed, thereby helping to isolate
the barnyard runoff from sources upstream, and (2) it allows sampling during small runoff peri-
ods in which local inputs from the barnyard are apparent, but little runoff from the upstream
areas of the watershed is observed. A second enhancement at Halfway Prairie Creek is that the
upstream and downstream stations were close enough together to allow a direct electronic
connection between automatic samplers and, hence, the collection of concurrent samples from
both water-quality samplers. This design allows for statistical comparisons between concurrent
individual upstream and downstream concentrations in water samples.
The types of barnyard BMPs implemented at Otter Creek and Halfway Prairie Creek are similar.
Clean rainwater is diverted away from the concrete areas of each barnyard to minimize the
amount of water flushing thorough the system. Direct precipitation is conveyed by a sloped
concrete surface and retaining wall to a screened collection box where most of the large solids
are trapped. The remaining liquid is then gravity piped to a concrete pad, which evenly distrib-
utes the liquid on to a grass filter strip. The filter strip at Otter Creek gently slopes downwards
toward the stream, whereas the filter strip at Halfway Prairie Creek is a substantial distance
from the stream. Cows, which were previously allowed to roam the stream and banks at each
site, have been fenced in, and a gravel-lined channel crossing now allows them access to the
stream. Although sampling sites were chosen to minimize inflows other than that from the
barnyard, a field near the investigated barnyard at Otter Creek could have potentially contrib-
uted to the stream loading between the upstream and downstream stations in the pre-BMP
phase, especially during periods of heavy storm runoff. As part of the barnyard BMP, a grassed
swale was installed to help minimize runoff from this field.
159
-------
Water samples from 12 storm-runoff periods were collected for the pre-BMP period of April
1994-October 1995 at Otter Creek. The clean-water diversion and concrete work were com-
pleted in October 1994; however, runoff from the collection box was not conveyed to the filter
strip until October 1995. With the exception of one snowmelt period, all the pre-BMP samples
for Otter Creek were collected for storm-runoff periods between April and October. Water
samples from 11 storm-runoff periods were collected for the pre-BMP period of April-July 1995
at Halfway Prairie Creek! All the barnyard BMP components at Halfway Prairie Creek were
implemented by October 1995. All samples from both streams were analyzed for suspended
solids, total phosphorus, and ammonia nitrogen. Due to holding time and budget constraints,
samples from some storms were not analyzed for biochemical oxygen demand (BOD) and fecal
coliform bacteria.
To date (August 1997), water samples from 12 storm-runoff periods have been collected for the
post-BMP period at Otter Creek, and water samples from 11 storm-runoff periods have been
collected for the post-BMP period at Halfway Prairie Creek. (The post-BMP period began April
1996 at both streams.) The sample collection portion of the study is complete; however, analysis
of water samples and constituent load computations have only been completed for seven storm-
runoff periods at Otter Creek and eight storm-runoff periods at Halfway Prairie Creek.
The continuous streamflow and instantaneous water-quality data were used to estimate loads
for individual storm-runoff periods. Loads were computed—in pounds—for suspended solids,
total phosphorus, ammonia nitrogen, and BOD, by summing the product of instantaneous
concentration and streamflow rate for each storm-runoff period (Porterfield, 1972). Microbial
loads of fecal coliform bacteria were computed similarly, however, the units are in total colony
forming units (in the volume of water that occurred during a storm-runoff period).
Runoff volumes for the downstream sites at Otter Creek and Halfway Prairie Creek were as-
sumed to be equal to the volumes computed for each, upstream site. In reality, this is generally
not the case because some amount of water is contributed by the barnyard during a runoff
period. However, for the pre-BMP period at Otter Creek, meaningful differences in streamflow
between the upstream and downstream stations were difficult to detect, primarily because the
drainage area of the barnyard is small compared to the drainage area of the entire watershed.
Because of the assumption of equal volumes, the loads computed for the downstream stations
for Otter Creek and Halfway Prairie Creek are slightly conservative.
Results and Discussion
Testing of Experimental Design
A critical aspect of validating conclusions for an upstream-downstream experimental design is
determining whether downstream loads are significantly greater than the upstream loads before
BMPs are implemented. At both streams, loads of suspended solids, total phosphorus, ammo-
nia nitrogen, BOD, and microbial loads of fecal coliform bacteria were greater at the down-
stream station than at the upstream station for most periods of pre-BMP runoff (figs. 2 and 3).
Using the Wilcoxon signed ranks test to find differences between paired data sets, study investi-
gators determined that downstream loads were significantly greater than upstream loads (at the
95-percent confidence level) for all constituents except suspended solids at Halfway Prairie
Creek. These significant differences indicate that the investigated barnyard-runoff sources are
important contributors to the loading of total phosphorus, ammonia nitrogen, BOD, and fecal
coliform bacteria for the storms monitored; in addition, the barnyard on Otter Creek is also an
important source of suspended solids.
160
-------
CO
Q
z
o
Q.
O
g
cc
g
CO
P 200,000
100,000
30,000
SUSPENDED SOLIDS
TOTAL PHOSPHORUS
150
100
30
20
10
0
w 60
30
go o
AMMONIA NITROGEN
i I I I r
i l r
iJ liL
BOD
i i i i i i i i i
FECAL COLIFORM BACTERIA* _
RUNOFF VOLUME
i i i i i i i i r
1 2 3 4 5 6 7 8 9 10 11 12
STORM NUMBER
EXPLANATION
UPSTREAM STATION
DOWNSTREAM STATION
Figure 2. Loadings and runoff volumes for pre-BMP storm-runoff periods at
Otter Creek.
161
-------
CO
Q
O
Q.
O
DC
CO
^ eo.ooo r
40,000 jr
8,000
6,000
4,000
2,000
0
SUSPENDED SOLIDS
100 I |
100 J-
TOTAL
i i i i i i
PHOSPHORUS
m ' ' 'i
20
15
10
5
0
AMMONIA NITROGEN
FECAL COLIFORM BACTER A*
4 5 6 7 8
STORM NUMBER
9 10 11
EXPLANATION
UPSTREAM STATION
DOWNSTREAM STATION
Figure 3. Loadings and runoff volumes for pre-BMP storm-runoff
periods at Halfway Prairie Creek.
162
-------
Large variations in meteorological data, such as rainfall, could affect comparisons of pre- and
post-BMP load data. For example, a post-BMP period with much larger rainfalls than the pre-
BMP period could show a reduction in loads simply because the barnyard loads could be
masked by much larger contributions from upstream sources. The median runoff volume, rain-
fall, maximum 30-minute rainfall intensity, and rainfall-runoff ratio for the pre- and post-BMP
runoff periods were compared to see whether meteorologic conditions differed. The Wilcoxon-
Mann-Whitney rank sum test was used to find differences between pre- and post-BMP data
sets. Results show no significant difference between pre- and post-BMP data for either Otter or
Halfway Prairie Creek. Any decrease in load contributed by each barnyard is therefore most
likely due to the implementation of the barnyard BMPs and not to changes in meteorological
variables. Rainfall-runoff relations for the pre- and post-BMP storm-runoff periods at Otter Creek
and Halfway Prairie Creek are shown in figure 4. Boxplots comparing pre- and post-BMP storm-
runoff volumes are shown in figure 5.
250
CO
CJ200
CO
O 150
Li/100
| 5°
> 0 Uj_^
OTTER CREEK
HALFWAY PRAIRIE CREEK
co1^
< 100
o
« 80
O
Z 60
% 40
D
o 20
>
,...,,,.. I .... I .... I ....
'_ • J
.
-
-
— —
- -
A
' ^W , , , . I , . , , I . , , , I , , , , "
0 0.5 1 1.5 2 2.5
RAINFALL, IN INCHES
1234
RAINFALL, IN INCHES
EXPLANATION
A PRE-BMP
• POST-BMP
Figure 4. Rainfall depth versus storm-runoff volume for pre- and post-BMP storms at
Otter Creek and Halfway Prairie Creek.
200
CO
< 150
O
OTTER CREEK
(12) (7)
O
Z
UJ
O
100
50
HALFWAY PRAIRIE
CREEK
(8)
CO
<
0
£2
o
z
UJ
2
1
IUU
80
60
40
20
0
1
— -T- —
_
.
_
-
.
T
Ls^»
EXPLANATION
(12) Number of
observations
90th percentile
75th percentile
Median
25th percentile
10th percentile
PRE-BMP POST-BMP
PRE-BMP POST-BMP
Figure 5. Boxplots of storm-runoff volume, by pre- and post-BMP period, at Otter Creek
and Halfway Prairie Creek.
163
-------
Differences Between Pre- and Post-BMP Barnyard Loads
The difference between upstream and downstream load was computed for each constituent and
each runoff period for pre- and post-BMP conditions. These differences are considered to be the
loads contributed by each barnyard (figs. 6 and 7). The Wilcoxon-Mann-Whitney rank sum test
was used to determine whether the contribution of each barnyard decreased significantly after (
implementation of barnyard BMPs.
i
At Otter Creek, post-BMP loads of suspended solids, total phosphorus, ammonia nitrogen, i
BOD, and microbial loads of fecal coliform bacteria contributed by the barnyard were signifi-
cantly less than the pre-BMP loads (at the 95-percent confidence level). At Halfway Prairie
Creek, post-BMP loads of total phosphorus, ammonia nitrogen and BOD contributed by the
barnyard were also significantly less than pre-BMP loads. Because the pre-BMP data analysis i
at Halfway Prairie Creek showed no significant difference between upstream and downstream i
loads of suspended solids, a statistically significant decrease in post-BMP suspended solids ]
was not detected. Although statistically significant differences were observed between upstream
and downstream microbial loads of fecal coliform bacteria for the pre-BMP period at Halfway
Prairie Creek, high variability in the available data have made it difficult to observe significant
differences between pre- and post-BMP periods. Analysis of samples from the remaining post- I
BMP runoff periods may help reduce variability and allow investigators to detect significant I
differences between pre- and post-BMP periods. |
Effectiveness of Barnyard Best Management Practices j
The Hodges-Lehmann estimator is the median of all possible pairwise differences between two
independent data sets (Helsel and Hirsch, 1992). This estimator was used to determine the
decrease in loads contributed by each barnyard between the pre- and post-BMP periods. This
difference was then divided by the pre-BMP median barnyard contribution, resulting in a per-
centage decrease. At Otter Creek, implementation of the barnyard BMP has reduced the loads
of suspended solids by 81 percent, total phosphorus by 88 percent, ammonia nitrogen by 97
percent, BOD by 80 percent, and microbial loads of fecal coliform bacteria by 84 percent; the
barnyard BMP at Halfway Prairie Creek has resulted in 67-, 89-, 94-, 91-, and 24-percent reduc-
tions, respectively (Table 1). Watershed planners for Otter Creek and Halfway Prairie Creek had
expected that implementation of the designed barnyard BMPs would result in phosphorus load
Table 1. Difference in constituent loads between upstream and downstream sites at Otter Creek
and Halfway Prairie Creek barnyards before and after installation of BMPs, and percentage reduc-
tion in constituent loads achieved by barnyard BMP.
Median load difference between upstream
and downstream stations, in pounds*
Constituent
Suspended solids
Total phosphorus
Ammonia nitrogen
BOD
Fecal coliform bacteria*
Otter
Pre-BMP
2,960
13.0
4.5
205
112.2
Creek
Post-BMP
470
0.8
0.3
11
8.3
Halfway
Pre-BMP
230
7.2
6.4
107
46.2
Prairie Creek
Post-BMP
-19
0.7
0.2
10
20.5
Median percent decrease**
Otter Creek
81
88
97
80
84
Halfway
Prairie Creek
67
89
94
91
24
'Fecal coliform microbial load in 1011 colony forming units.
"Computed by dividing the Hodges-Lehmann estimator for pre- and post-BMP barnyard loads by the pre-BMP median barnyard load.
164
-------
SUSPENDED SOLIDS
«
C/5
Q
0
Q.
z
z"
o
p
m
CC
0
o
Q
g
Q
GC
IT
m
80,000 vfc- •> EXPLANATION
v
15,000
10,000
5,000
0
-5,000
x \
_ A
. A A ®
_ A _
A! «K A
f{A © • -
I I I I I I > I,
*
A PRE-BMP
• POST-BMP
TOTAL PHOSPHORUS
80
60
40
20
0
% %
I I I I I Is!'
A
-
A
-A -
.A A A ® •
&A I «, I % I I I I si,
AMMONIA NITROGEN
10
10
5
I £ I I I I s 1 '
-A
~A A A A
|& 0 0
"»* ^
1 I.I 1 1 1 s k ,
*-* * <
BOD
t>UU
400 v
250 ^
200
1 A 1 A I 1 1 s 1 ' H
•> A "3-
^- A H
A
-A
~ A ~
IS
FECAL COLIFORM BACTERIA*
_> V
HA H s
1,200 J- -J* 2
300 M
200
100
0
-KD
«_ -,
- A
_ A
A
~ A a ~
•• A °
* • A * ~
I I I I I I % k * ,
- .a
E
E
s:
O
"3
o
LL
>•
c
81
0 10 20 30 40 50 60 200 250
STORM-RUNOFF VOLUME, IN CFS-DAYS
Figure 6. Storm-runoff volume versus load contributed by the barnyard for
pre- and post-BMP storm-runoff periods at Otter Creek.
165
-------
SUSPENDED SOLIDS
* *
#
CO
Q
z
o
a.
21
z
o
i-
m
cc
O
o
0
a
rr
L^
>
CC
1,500
1,000
500
0
-500 -
-3,000 "
-6,000
'_ © j
-
© is
- ^ A -
LA e 4
EXPLANATION
A PRE-BMP
« POST-BMP
\
f- A -3T
I 1 1 1 1 1 s k ft , H
< ^
TOTAL PHOSPHORUS
4U
30
20
10
0
-40 v
-50
II II I I v 1
- A J
— —
L A A J
£l£± & m ^
— ^pT © © —
k
1 1 1 1 1 1 » I * , ~
*,
*,
AMMONIA NITROGEN
OQ » s
15
10
5.
0
1 1 1 II 1 < 1
^
-- —
_ A A
A
£& A g.
A
AA (Si
4^ ^ | M, *f 1 |M. I 1 S k 1
?nn T
A J
ffi
V
—
-
-
-
A
A
AA
A Ae •
@® «
1 1 1 1 1 1 s > *
-
—
_
-
-
.
FECAL COLIFORM BACTERIA*
_> ">
ouu 1 1 1 -T 1 1 1 M » ' J •£
400 J*- A -^ 2 c
. o o in
200 1
150
100
50
0
__
_
L AA A :
A A «
A\& Ai w l I 1 I > k
'c o •?
|-° =
= c'|
^f TJ -S
« (0
O O
fl) ~~
u.
0 5 10 15 20 25 30 100 120
STORM-RUNOFF VOLUME, IN CFS-DAYS
Figure 7. Storm-runoff volume versus load contributed by the barnyard for
pre- and post-BMP storm-runoff periods at Halfway Prairie Creek.
166
-------
reductions of approximately 95 percent for each barnyard (Pat Sutler, Dane County Land Con-
servation Department, written commun. 1997). The reductions in phosphorus found in this
study—nearly 90 percent for both barnyards investigated—indicate that this assumption is not
unreasonable.
Summary
The implemented barnyard BMPs at Otter Creek and at Halfway Prairie Creek have significantly
reduced the contributions of total phosphorus, ammonia nitrogen, and BOD to the stream. At
Otter Creek, loads of suspended solids and microbial loads of fecal coliform bacteria have been
significantly reduced as well. Although statistically significant differences were observed be-
tween upstream and downstream microbial loads of fecal coliform bacteria for the pre-BMP
period at Halfway Prairie Creek, high variability in the available data have made it difficult to
observe statistically significant differences between pre- and post-BMP periods. Analysis of
samples from the remaining post-BMP runoff periods may help reduce variability and allow
investigators to detect significant differences between pre- and post-BMP periods.
Watershed planners for Otter Creek and Halfway Prairie Creek had expected that implementa-
tion of the designed barnyard BMPs would result in phosphorus load reductions of approxi-
mately 95 percent for each barnyard. The reductions in phosphorus found in this study—nearly
90 percent for both barnyards investigated—indicate that this assumption is not unreasonable.
Finally, the upstream-downstream experimental design appears to have worked well, not only
for measuring the magnitude of the barnyard BMP sources but also for documenting the magni-
tude of the load reductions due to BMP implementation. This technique will most likely have
merits in studies of other rural nonpoint BMPs, such as streambank erosion, rotational grazing,
and buffer strips.
References
Bachhuber, J., and Foye, K., 1993, Nonpoint source control plan for the Sheboygan River
Priority Watershed Project: Wisconsin Department of Natural Resources Publication WR-
265-93 [variously paginated].
Eagan, L.L., and Morton, A., 1989, A plan for the control of nonpoint sources and related re-
source management in the Black Earth Creek Priority Watershed: Wisconsin Department of
Natural Resources Publication WR-218-89 [variously paginated].
Helsel, D.R., and Hirsch, R.M., 1992, Statistical methods in water resources: New York,
Elsevier, p. 132.
Porterfield, George, 1972, Computation of fluvial-sediment discharge: U.S. Geological Survey
Techniques of Water-Resources Investigations, book 3, chap. C2, 66 p.
Spooner, J., Maas, R.P., Dressing, S.A., Smolen, M.D., and Humenik, F.J., 1985, Appropriate
designs for documenting water quality improvements from agricultural NPS control pro-
grams, in Perspectives on nonpoint source pollution: Washington, D.C., U.S. Environmental
Protection Agency, EPA 440/5-85-001, p. 30-34.
167
-------
The Nebraska MSEA Project
Management of Irrigated Corn and Soybeans
to Minimize Ground Water Contamination
Darrell G. Watts James S. Schepers Roy F. Spalding
University of Nebraska USDA/ARS, University of Nebraska University of Nebraska
Lincoln, Nebraska Lincoln, Nebraska Lincoln, Nebraska
Background
There is a clear need to reduce the contribution of production agriculture to the contamination of
both ground and surface waters by nitrate-nitrogen. The USDA sponsored Management
Systems Evaluation Areas (MSEA) projects have concentrated on two major aspects of the
problem across the cornbelt: 1) gaining a better understanding of how contamination occurs,
including the challenges and limitations faced by farmers in managing production systems, and
2) developing and improving practical technologies that will enable the producer to continue to
operate at a profit, with lower overall environmental impact. The Nebraska MSEA has focussed
on irrigated corn and soybean production in a zone having high levels of ground water nitrate.
Project Setting
The project area in Nebraska's Central Platte Valley (Figure 1) contains over 200,000 ha of
irrigated land underlain by shallow ground water having nitrate-N concentrations above 10 mg/L.
Some areas, such as the one around the MSEA site, have concentrations between 30 and 40
mg/L. While there are a number of contributors to the problem, the greatest single source of
contamination is nitrogen (N) fertilizer applied to irrigated corn, which dominates the production
system in the area.
Factors contributing to ground water contamination include overestimation by corn producers of
N and water requirements, their inability to uniformly distribute water using conventional furrow
irrigation, limitations in timing and form of N application, and the lack of a rapid and reliable
technique for programming supplemental N applications. Furthermore, there are no efficient
means to apply additional nitrogen on furrow irrigated corn after the crop exceeds about 80 cm
in height.
A typical conventional furrow irrigation may apply 2-3 times the amount of water needed.
Irrigation runoff is controlled by diking the lower end of the field from the end of June until mid
September, instead of installing a tailwater reuse system. During the irrigation season all runoff
from both irrigation and precipitation soaks into the ground within the field. The combined
effects of management, terrain and good infiltration rate contribute greatly to the potential for
leaching of agrichemicals, particularly nitrate-nitrogen.
169
-------
NEBRASKA
MSEA Sita.Shatton
Study Area, Central Platte Valley
Figure 1. Principal site and satellite locations for the
Nebraska MSEA project.
Project Objectives
1. Evaluate the impact on water quality of conventional and improved management
systems that employ available technologies.
2. Improve existing "best management practices" and develop new practices and
technologies that are both environmentally friendly and cost effective.
Implementation and Evaluation Approaches
Field research was established in 1990 at a principal site near the town of Shelton and at
satellite locations, both in the Central Platte Valley and at Research and Extension Centers near
Clay Center and North Platte. The field work has dealt primarily with the evaluation and
demonstration of available technology packages and the development of new technologies for
water, nitrogen, and pesticide management on irrigated, monoculture com and corn-soybean
rotations. This work has been complemented by a program of socio-economic research and an
active extension program. This report will focus primarily on results from the field scale
management block research/demonstration area, and some of their implications.
Direct evaluation of the impact of management systems on ground water quality has been made
on four 13.6 ha management blocks (Figure 2). The management blocks include conventional
farmer practice for furrow irrigation and nitrogen management, improved furrow irrigation and
nitrogen management, and center pivot sprinkler irrigation with fertigation, all on corn. In
addition, the fourth block is seeded to sprinkler irrigated alfalfa to evaluate the ability of this
scavenger crop to remove nitrate from the high nitrate ground water which is pumped for
irrigation. The improved technologies are designed to be managed by producers.
170
-------
MSEA Management Systems
Corn
Surge Irrig
Runoff Reuse
Split N App.
*
Irria Scheduling
Corn
Center Pivot Irrig
Fertigation
OW FLOW
Irrig. Scheduling
Corn '
Conventional Irrig
Blocked End Furrows
All Preplant N
Alfalfa
Center Pivot Irrig
Tow Line Sprinkler
= Multilevel
Sampling Wells
Figure 2. Layout for field scale evaluation of
management systems employing available
technology.
Each technology package is established as a single treatment, providing enough land area for a
full-scale demonstration and evaluation while ensuring that treatment associated changes in
ground water quality can be detected. The impact of these management systems on water
quality is being evaluated directly through the use of multi-level samplers placed in wells around
and within the management blocks. Indirect assessment of system impact on water quality is
being made through the use of banks of suction lysimeters to sample the quality of drainage
water leaving the root zone, and by means of tracers applied at the land surface. The amounts
of water and N fertilizer applied to each system, and the resulting crop yields are direct
measures of input changes and their impacts on production.
From the outset we assumed that three steps would be required to have any significant impact
on ground water nitrate concentrations:
1. Reduce N fertilizer. Do this by assuming a reasonable yield goal, while including several
important items in the calculation of N requirement: a) residual mineral N, b) an estimate
of the contribution from mineralization, and c) the amount of N supplied to the crop by the
high-nitrate ground water pumped for irrigation.
2. Time N applications to better coincide with crop uptake. Use a small, handheld
chlorophyll meter on both surge flow and sprinkler irrigated fields to guide decisions to
delay or eliminate N applications by fertigation, thereby decreasing N fertilizer amount
below that computed as "recommended".
3. Improve water management. Schedule irrigations according to soil water content and
weather conditions; control irrigation amounts to minimize leaching of applied N during
the growing season; let the crop deplete more of the soil water at the end of the season,
thereby retaining more of the winter and spring precipitation and reducing off-season
leaching.
171
-------
Environmental Benefits Measured
Within the strictures imposed by weather, soil variability, equipment and our knowledge of the
system, we were reasonably successful in accomplishing what we set out to do. Table 1
presents the mean water and N inputs and resultant crop yields for the three MSEA
management systems during five years of field evaluation. The prime input-output values of
Table 1 are summarized in Table 2 as a percentage of the conventional system.
Table 1. Average Irrigation, Rainfall, Soil Nitrogen, Fertilizer, and Yield Data over
Five Years.
Year/
Irrig.
Conventional
Surge-Flow
Center-Pivot
Irrig.
Applied
(mm)
606
211
175
Rainfall
5/1-9/30
(mm)
445
445
445
Rain*
Irrig.
(mm)
1051
656
620
Residual
N*
(kg/ha)
100
107
75
Starter
N
(kg/ha)
24
24
24
N
Fertilizer
(kg/ha)
196
135
122
Irrig,
N03-N
(ppm)
30.8
28.9
29.4
Irrig.
N
(kg/ha)
228
78
57
Grain
Yield
(Mg/ha)
11.91
11.49
11.18
Total residual N (nitrate-N) to a depth of 0.9 m.
Table 2. Average Management Block Irrigation, Fertilizer,
and Yield as Percent of Conventional Practice.
Conventional
Surge-Flow
Center-Pivot
Irrig.
Applied
(%ofConv)
100
40
33
Total
Fertilizer
(%ofConv)
100
69
63
Grain
Yield
(% of Conv)
100
96
94
The two tables collectively show that improving management systems by using readily available
technologies can substantially reduce the excess applications of irrigation water and nitrogen
fertilizer that contribute to nitrate leaching from the crop root zone in the central Platte Valley.
Suction samplers installed at a depth of 135 cm under the management systems showed
average soil water nitrate concentrations over project life as shown below.
Conventional Management 53.7 mg/L
Surge Flow Irrigation 28.5 mg/L
Center Pivot Irrigation 12.0 mg/L
These numbers by themselves do not indicate the amount of nitrate moving to the water table.
They only represent measurements during the active growing season and provide no •
information on the volume of leachate. However, they clearly indicate the overall quality of the
water leaving the root zone during the indicated period. We do not believe that the
concentration of drainage from the pivot irrigated field can be maintained at 12 mg/L and also
maintain full production.
172
-------
Better management can improve ground water quality. When the project began in 1991, the
nitrate-nitrogen concentration was essentially 30 mg/L through the entire 14 meter depth of
saturated thickness of the aquifer. At present under the center pivot, water quality has improved
in the upper 5 meters of the ground water, with nitrate-nitrogen concentrations in the range of
22-25 mg/L. There is a lesser improvement under the surge system, and none under the
conventional system. The deeper ground water remains at the original high concentration.
Improvement literally comes from the top down. As water of lower nitrate concentration enters
the upper tier of the ground water and the water of higher concentration is pumped from below,
water quality will gradually improve, up to a point.
Lessons Learned
1. The most important lesson learned from this work is that it is not only possible, but quite
feasible to reduce ground water nitrate concentration through careful application of currently
available technologies and BMPs.
Use of surge irrigation, an "intermediate" technology, together with improved nitrogen
management reduced applications of nitrogen and water, resulting in lower in-season losses
of both. Results would have been better if we had not tried supplemental nitrogen
applications by adding N fertilizer to the irrigation water in 1993 and 1994. While "fertigation"
results in a uniform N application under center pivot irrigation, the row to row and point to
point variability of water intake under furrow irrigation resulted in an unsatisfactory N
distribution across the field. The problem was exacerbated both years by adverse weather
phenomena.
Conversion to center pivot irrigation offers the greatest opportunity to improve water quality.
This system permits close control of water quantity applications during the growing season,
and makes it easy to leave the soil profile relatively dry at the end of the season. The latter
practice will allow the soil to store more of the winter and spring precipitation, reducing the
amount of water available to leach residual nitrogen. Use of the chlorophyll meter to
schedule supplemental "spoon feeding" of the crop through fertigation can increase nitrogen
use efficiency and further decrease nitrate loss.
2. Reducing water and nitrogen very close to the minimum levels for full production leaves little
room for error and may entail significant risk or increased cost for the producer when
unexpected weather conditions or other management problems arise. Our experience with
the center pivot is a good example. Errors in placement of "full fertilizer" calibration rows for
chlorophyll application in 1992 resulted in an erroneous decision to withhold N fertilizer.
This resulted in a 12 percent yield loss in that year and had some continuing impact in the
following years.
3. Our MSEA extension program has shown that targeted educational programs can move
producers toward more reasonable levels of fertilizer and water application. However, they
are highly reluctant to voluntarily move to application levels that they perceive to significantly
increase the risk of yield loss. Successful voluntary conversion of large areas to more
environmentally friendly management systems must, at a minimum, be cost neutral and
must address farmers' perceptions of risk resulting from reduced inputs of water and
nitrogen.
173
-------
North Carolina Agricultural Systems For
Environmental Quality (ASEQ) Project
Frank J. Humenik
North Carolina State University, Raleigh, North Carolina
Steve Broome, Steve Coffey, Maurice Cook, Gregory Jennings, Michelle Marra,
Rich McLaughlin, Deanna Osmond, Mark Rice, Casson Stallings
North Carolina State University, Raleigh, North Carolina
Patrick Hunt, Terry Matheny, Jeff Novak, John Sadler, Ken Stone,
Ariel Szogi, Matias Vanotti
USDA-Agricultural Research Service
Coastal Plains Soil, Plant, and Water Conservation Research Center
Florence, South Carolina
George Stem
USDA-Natural Resources Conservation Service
Raleigh, North Carolina
Introduction
The North Carolina/South Carolina Agricultural Systems for Environmental Quality Project is
being conducted in Duplin County, North Carolina, and Florence County, South Carolina, by a
team that has successfully implemented a USDA Water Quality Demonstration Project in the
Herrings Marsh Run Watershed in North Carolina, location of the highest swine producing county
in the country. This area is typical of land utilized for crop and livestock production in the
southeastern U.S., where rapid growth is occurring in the swine, poultry, cotton and truck-crop
industries. Water quality improvements have already been measured in the Water Quality
Demonstration Project area as a result of best management practice (BMP) implementation. The
ASEQ Project offers the additional opportunity to develop, implement and evaluate innovative
BMPs as well as to refine those currently in existence.
Major objectives of the ASEQ project are to:
1) evaluate the capacity of riparian and in-stream wetland systems to protect water quality
2) increase functionality and adoption of site specific farming as a BMP to improve
production cost effectiveness and protect water quality
3) enhance the efficiency of a constructed wetland treatment system to remove nitrogen and
phosphorus from swine wastewater and thus minimize the amount of land required for
terminal treatment and protect soil, air and water quality.
175
-------
The following agencies are providing leadership for the ASEQ effort:
• Cooperative Extension Service in the College of Agriculture and
Life Sciences at North Carolina State University
• USDA Agriculture Research Service - Soil, Water and Plant Research
Center in Florence, South Carolina
• USDA Natural Resources Conservation Service
Other agencies involved include the United States Geological Survey, which has cost shared the
installation and maintenance of automated sampling and stream gaging stations; and the North
Carolina Department of Environment, Health and Natural Resources, Division of Water Quality
which conducts biological sampling to support project water quality evaluations. This cooperative
effort has allowed the Demonstration Project to go beyond the initial charge of voluntary,
accelerated and widespread adoption of BMPs.
An initial goal of the Demonstration Project was to measure water quality changes. A relatively
small 5,000-acre watershed was selected to facilitate surface and ground water monitoring and
tracking of land use activities. ARS was also interested in documenting water quality changes
and agreed to run surface and ground water analyses. A cooperative arrangement was
established with USGS for installation and maintenance of automated sampling stations and the
North Carolina Department of Environmental, Health and Natural Resources for biological
sampling to support water quality evaluations. These provisions for monitoring water quality and
land use set an excellent basis for development and evaluation of models to determine water
quality changes in the study watershed and to extend project results over time and space.
Riparian and In-stream Wetland Systems
A riparian area has been replanted between a spray field for swine lagoon liquid and an adjacent
stream. Lagoon liquid was over applied to the spray field which did not maintain vegetative cover
resulting in ground water nitrate levels averaging 60 mg/l with some wells exceeding 300 mg/l.
Recommended practices are now being implemented for lagoon liquid irrigation, and ground
water levels are decreasing. To minimize the impact of surface and ground water flows from the
spray field, the riparian zone was replanted in 1992 with five species of trees.
Soil samples for denitrification and enzyme assay (DEA) and Phosphorus (P) fractionation will be
collected from four locations along two transects of the experimental area. These locations will
be 1) at the edge of the spray field, 2) at the midpoint of the riparian zone, 3) at the stream edge,
and 4) in the non-treated edge across the stream from the spray field. Sampling sites are
176
-------
Constructed In-stream Wetland
(O.9 ha) with Water Level Control
Restored Riparian Area
Beaver
Enhanced
I n-strea
Wetland
(3.3 ha)
^Surface water sampling stations
Figure 1: Riparian and In-stream Wetland System
located within 2-5 meters of each of eight ground water monitoring wells. Soil samples will be
collected from the upper 15 cm of the soil profile, at a point midway between the soil surface and
the water table, and 15 cm below the top of the water table. DEA analyses will determine the
limiting factors for denitrification (i.e. NO3 or C source).
Stream water nitrate concentrations are still higher downstream of the restored wooded riparian
zone than upstream by about 10 mg/l after two years' growth of the replanted trees. The stream
nitrate concentrations in the restored riparian area have been 12-14 mg/l.
In-stream Wetlands. A beaver enhanced in-stream wetland that was developed during
the Demonstration Project is reported on in the Demonstration Project paper. An in-stream
wetland with water level control was constructed in a dry, one hectare pond (dam had been
breached) that receives flow from the restored riparian area but is above the enhanced in-stream
wetland (Figure 1). This allows the evaluation of three landscape features in series along a
stream reach being 1) the restored riparian area, 2) constructed in-stream wetland, and 3)
beaver-enhanced in-stream wetland. The constructed in-stream wetland will be evaluated for its
capacity to reduce nitrogen at various water levels. Inflow and outflow volume, wetland depth
and concentration of nitrogen and phosphorus will be measured. Vegetation will be surveyed
annually for composition, biomass and nutrient accumulation in three transects. One annual
sediment transect with four sampling points will be taken from the 0-15 cm depth for DEA. One
annual sampling of sediment pore water will be made with diffusion-controlled pore water
samplers placed in duplicate at 0-15 cm depths into the sediment at three points on a transect
that parallels the stream direction.
Constructed Wetland Treatment Systems
Constructed wetland studies funded through an Environmental Protection Agency grant,
"Evaluation of Alternative Constructed Wetland Systems for Swine Wastewater Treatment," will
be continued to determine the most effective use of constructed wetlands in an overall treatment
system to minimize land required for terminal treatment and protect soil, air and water quality.
The research site has a nursery operation with 2600 pigs (average weight equals 13 kg) that
uses a flushing system to recirculate lagoon liquid for house cleaning from a single stage lagoon.
177
-------
Wetlands with two 3.6 x 36 m cells in series were constructed in 1992. Initially, a total Kjeldahl
nitrogen (TKN) loading rate of 3 kg/ha/day was used but a dischargeable effluent was not
achieved, so it was increased to 10 kg N/ha/day during the second year. Mass removal of
nitrogen was 94% at the loading rate of 3 kg TKN/ha/day and decreased to 68% at the higher
rate of 10 kg TKN/ha/day. Total phosphorus (T.P.) mass removal efficiencies ranged from 40 to
100 percent at T.P. loading rates less than 1 kg TP/ha/day and varied from 20-80% when TP
loading rates were 1-4 kg/ha/day.
Pilot overland flow and media filter processes are being evaluated as possible wetland system
components to achieve a treatment train which provides higher removals and results in less land
for terminal treatment. The overland flow treatment unit consists of a 4 x 20 m plot with a 2%
slope. The overland flow system had an hydraulic loading of 2.5 cm/day which resulted in an
ammonia application of 54 kg/ha/day. Total mass nitrogen removal efficiency was 59%, total
phosphorus removal efficiencies varied from 40-80% at loading rates as high as 38 kg
TP/ha/day. Average inflow total Kjeldahl nitrogen was 250 mg/l, outflow total Kjeldahl nitrogen
was about 78 mg/l and nitrate nitrogen was about 30 mg/l. The overland flow plot resulted in
good nitrogen removal and cost effective oxidation to nitrate for subsequent denitrification and
removal of nitrogen as nitrogen gas rather than ammonia which has environmental impacts.
The media filter which was a 1.8 m-diameter x 0.9 m high tank filled with mart gravel received
lagoon wastewater as a fine spray onto surface at hydraulic loading rates of 75 L/m2/hr and total
nitrogen loading rate of 147 g) m2. About 50% of total suspended solids and chemical oxygen
demand were removed from wastewater with just one cycle, and losses of total nitrogen were
11 % with one cycle and 22% with four cycles. With four cycles 24-32% of the influent TKN was
converted to nitrate. Media filters can provide a very cost effective unit process for removal of
carbon and suspended solids as well as nitrification for use as a unit process before
denitrification for additional nitrogen removal.
Precision Agriculture
Current precision agriculture or site specific farming activities directed at more efficient
production, energy savings and reduced chemical use which emphasize and document water
quality benefits will be complemented and expanded. Three commercial combine-mounted yield
monitor systems will be used: two for retrofitting farmer cooperator combines and one for
retrofitting a research combine. Yield mapping will be conducted on as large an area as possible
so that farmers will have multiple case studies to set a basis for evaluating precision agriculture.
Spatial yield research will be conducted with the acquired data by 1) simulating crop growth and
yield for comparison to the mapped yield, 2) identifying correlated parameters that can be used
to determine important cause-and-effect relationships resulting in the variable yield, and 3)
developing and implementing variable nitrogen management using the mapped yield. First,
mechanistic models of crop growth and yield, nutrient removal, water quality and nitrogen
transformations will be used along with corresponding soil and crop parameters to simulate the
effects of variable soils on yield. Second, recent developments in statistical theory, including
innovative techniques to exploit spatial data will be used to extract causal relationships from data
collected at different spatial intensities. Mapped yield, which is taken at high frequency, can be
used to develop strategies for more time-consuming sampling such as soil fertility, soil
characteristics, crop characteristics, etc. The resulting data sets will be examined using state-
space models to narrow the choice to important variables. Third, measurements of nitrogen
availability such as with the chlorophyll meter will be used with mapped yield, mapped soil tests
178
-------
and experience with mineralizeable soil nitrogen factions to guide sidedress nitrogen
recommendations.
Water quality benefits. Work is underway to develop paired fields to evaluate
differences in surface and ground water resulting from contemporary farming and precision
agricultural recommendations. Two 3-acre fields have been obtained at the new Center for
Environmental Farming at the North Carolina Department operated Cherry Farm in Goldsboro,
North Carolina. Six wells ranging from 300 cm to 450 cm have been established in each three-
acre field. Preliminary data shows nitrate concentrations of 20 to 25 mg/l in these fields. These
wells will be sampled every two weeks during the cropping season from about April to October
and about once every month from November to March for nitrate, ammonia and ortho-phosphate.
Surface drainage will be collected from about a one-acre sub-watershed in each field for flow
through a sampling station with an automated sampler and flow meter. Sampling will be initiated
by flow stage and flow proportion samples will be taken at given stage increments so that yield
can be calculated based upon concentration and flow for a given sample. Surface runoff
samples will be analyzed for total suspended solids, nitrates, total Kjeldahl nitrogen and ortho-
phosphate. Ground water concentration differences and surface water yield differences between
the study plots will be determined to evaluate the water quality benefits of precision agriculture at
this site with sandy clay loam soils which are typical of the middle Coastal Plain.
Decision Support System
WATERSHEDSS, is a decision-support system (DSS) that is available electronically has been
developed through an EPA grant, "Understanding the Role of Agricultural Landscape Feature,
Function and Position in Achieving Environmental Endpoints." The purpose of the DSS is to help
watershed managers determine appropriate management and/or BMP systems and their
placement on the landscape for particular water quality problems. Effectiveness data collected
from the riparian and wetland system studies will be included in a BMP database that is part of
WATERSHEDSS. The decision support system is accessible at the following universal resource
locators (URLs):
http://www.bae.ncsu. edu/bae/programs/extension/wqg or
http://h2osparc.wq.ncsu.edu
Technology Transfer
Educational workshops on Precision Agriculture; Landscape Features to Protect Water Quality;
and Constructed Wetlands for Wastewater Treatment are to be conducted in conjunction with this
grant.
A site specific farming workshop was held during the first year because we had the opportunity to
cooperate with North Carolina State University College of Agriculture and Life Sciences, North
Carolina Agricultural Research Service and North Carolina Cooperative Extension Service, and
North Carolina Department of Agriculture for a "Precision Agricultural Field Day at the Center for
Environmental Farming Systems, Cherry Farm, Goldsboro, North Carolina. The field day was
judged a success on the basis of the approximately 200 who participated and the number and
quality of industrial and educational displays. There was about 55 farmers and 15 county
extension agents in attendance. The ARS Coastal Plains Soil, Water and Plant Research Center
had an educational display on their yield monitoring and site specific irrigation work which led to
the CSREES grant objective on site specific farming.
179
-------
Assessing the Impacts of Voluntary Pollution Prevention
Programs in Agriculture:
Lessons Learned from a Cost-Benefit Evaluation of Farm *A* Syst
Robert Moreau, Gary Jackson and Doug Knox
National Farm and Home Assessment Program Office
B 142 Steenbock Library 550 Babcock Dr.
University of Wisconsin-Madison
Madison, Wisconsin 53706-1293
Abstract
With the passage of the Government Performance and Results Act (GPRA) of 1993,
national budget balancing efforts for both voluntary and regulatory programs must
address the costs and benefits of pollution-prevention efforts as well as other traditional
agricultural programs. However, cost-benefit and other similar impact assessment data
have been historically difficult to obtain-and are indeed sparse for voluntary pollution
prevention programs-mainly due to the intrinsic problems associated with measuring
environmental benefits. This paper presents research that provides decision makers
with a model comprehensive impact assessment technique for evaluating voluntary
pollution prevention programs in the agricultural and rural home-owner sector in a
manner that is relatively easy and inexpensive.
A case study is presented in which both economic and non-economic impacts of the
Farm Assessment System (Farm *A* Syst) program in Louisiana were measured.
Results show farmers spent an average of approximately $700~in at least one-time
expenditures-on making changes in their practices and structures. Extrapolation to the
one-third of those farmers who are estimated to participate in programs such as Farm
*A* Syst (approximately 8,551 farmers) produce estimated net economic benefits of at
least $2.4 million, as well as other benefits related to educational and attitudinal impacts
on farmers, all of which are statistically significant. Farmers may very well make
additional changes throughout their lives (as opposed to one-time changes only).
Cash incentives have shown to increase participants' actions to reduce environmental
risks, although the sample size here is very small. Data from the Louisiana study has
been used as a basis for estimating national impacts of the program. It is estimated that
the 42,000 farmers who have undergone farm assessments throughout the US have
spent approximately $30 million in pollution prevention expenditures since late 1989.
This methodology may have direct application to other rural environmental management
programs. That potential is briefly explored.
Introduction
Since Congress is now requesting accountability of government funded programs, such
evaluation-or impact assessment-should become a cornerstone of any new program's
design. The most recent law that addresses this issue is the Government Performance
181
-------
and Results Act (GPRA) of 1993, in which comprehensive impact evaluations of
programs and policies are to be conducted, primarily in the form of cost-benefit
analyses. At the same time, economic data on the impact of programs in agriculture-
especially those programs which are voluntary and pollution-preventive in nature, have
been historically difficult to obtain. The primary reason for this is the fundamental
difficulty in measuring the economic and non-economic benefits (or impacts) that these
programs have on its participants and to society in general.
In meeting this standard, and in addressing the need to develop and conduct such an
evaluation of its own program, the National Farm *A* Syst/Home *A* Syst Office
undertook such an evaluation on Louisiana's version of Farm *A* Syst. This evaluation
was conducted during the course of a two year pilot implementation ending in October
1996 (Moreau, 1996, and Moreau, et. al., 1997). Farm *A* Syst is a voluntary pollution-
prevention program that enables farmers to quantitatively identify~and perhaps reduce--
the risks their farming practices and structures pose to groundwater quality. In many
states this assessment also addresses risks to surface water. A home assessment
program of parallel design is also available in most states, addressing environmental
concerns around the home.
This impact assessment methodology is currently being refined and proposed to Farm
*A* Syst and Home *A* Syst coordinators on the state level as a way of evaluating their
own programs. If Farm *A* Syst is integrated into other programs such as NRCS's
Resource Conservation Planning initiative, or community source water protection
programs, then the methodology can be useful in evaluating those broader programs as
well.The Resource Conservation Planning process has many characteristics that are
similar to Farm *A* Syst, as it places the emphasis of environmental protection activities
in the hands of the farmer or rancher, in the form of flexible, voluntary and confidential
plans that address a broad array of environmental and crop management activities.
This paper details the results from the Louisiana study, and the information gained and
lessons learned. We propose that the methodology be considered for evaluating the
impacts of similar programs. Such analyses should aid many aspects of the decision-
making process, especially as it relates to the long-term funding of programs.
Methods for Measuring Costs and Benefits
Measuring the economic costs and benefits of environmental-related programs such as
Farm *A* Syst has been historically difficult, mainly due to the difficulty in assigning
values to environmental benefits. Fields (1994), Freeman (1993), and Mitchell and
Carson (1989) provide excellent summaries of the different direct and indirect ways in
which environmental benefits can be measured, under both observed and hypothetical
situations. Each of these valuation techniques has pro's and con's related to issues such
as standing (whose benefits and whose costs count?) and parameters of the type of
benefit being measured (e.g., avoided cost of some regulatory action, willingness-to-pay
182
-------
for a desired outcome, etc.).
With regard to agricultural-based programs, many economists believe that the
contingent valuation method that asks people "what they are willing-to-pay" (WTP) for
some environmental benefit is one of the more appropriate measures. This Is due to the
fact that WTP questions take into account a more theoretically acceptable approach to
obtaining values that address the traditional consumer and producer surplus (CS and
PS) measures which are the basis for cost-benefit calculations (Boyle, 1997). However,
contingent valuation techniques can be costly and are highly debated because of their
hypothetical nature and wide variation in response rates (Johnson and Johnson, et. al.,
1990).
Valuation is perhaps even more difficult in pollution prevention programs, where risk and
outcome parameters have a certain degree of subjectivity associated with them. For
example, nitrate passage from the land surface to the underground aquifer, and its
accumulation there to dangerous levels, is a process that can take decades (Ryding,
1992). Not only is it difficult to understand what impact current farming practices are
having on the underlying groundwater, but it is nearly impossible to determine the
effectiveness (and corresponding economic benefit)~in the short term-that a pollution
prevention program has on such practices if a water test is the primary measuring stick.
Such lag-effects would out-live the researchers!
Due to time and budget constraints, and in order to utilize more than one method of
valuation for comparison purposes, the researchers of the cost-benefit study on
Louisiana's Farm *A* Syst focused on two of the more "applied" methods of economic
valuation of environmental benefits noted in the literature. First, the fairly recent concept
of averting expenditures (Abdalla, et. al., 1993, and Laughland, et. al., 1993, among
others) was applied. This approach estimates the lower bound of benefits as those
monies actually spent (or planned to be spent) by farmers who make substantive
changes in their practices due to information gained through the farm assessment
process. These figures can be easily attained from participants through pre- and post-
program surveys. In this study, this amount averaged about $700 per farmer,
conservatively assumed as a one-time expenditure over a 10 year period.
The second method-termed avoided cost of clean-up and/or change in water supply -
looks at the $700 as a sort of "insurance premium" against the future and higher cost of
environmental clean-up. This cost was conservatively estimated at approximately
$80,000 (Shaw, 1995) for a typical petroleum or farmstead hazardous waste type
problem. In order to compare the above two methods to the traditional technique of
contingent valuation (what one is "willing-to-pay" to reduce the risk to groundwater
contamination), figures from an existing and similar WTP study were used. Here, the
lower-bound median monthly willingness-to-pay figure of $6.44--which was documented
by Jordan and Elnagheeb (1993) in a previous study on Georgia households-was
183
-------
utilized (the figure is conservative because the mean was nearly double that amount).
The three methods are summarized mathematically below.
NB=AE-GC (averting expenditures - Method #1)
NB=CU-AE-GC (avoided cost of environmental clean-up - Method #2)
NB=WTP-AE-GC (contingent valuation questions - Method #3)
(where, NBcNet BenelBs, A&rfamws' Avertktg Expenditures, 3O)ederal, stale, and local Governmental Costa ol administering the program, CU»Ctean-Up and alternative water
•uppV costs associated win dMerent contaminant, and WTP=WWngnes»-To-Pay Irom contlnaenl va)u«rlon quasitons m Jordan & Elrtaaheeb study)
Impact Results from the Louisiana Case Study
Participation- Overall, 134 assessments and accompanying pre- and post-assessment
surveys were completed. The one-on-one visits produced a 61% (n=85) response rate
in Phase 1 (the first year of the study), and a 98% (n=49) response rate in Phase 2~the
second year of the study. The "workshop" delivery also utilized in Phase 1 produced a
low 7.5% (approximate n=400) response rate. Here, concern for confidentiality required
anonymity of participants, which prevented follow-up, and which we judge to be the
primary reason for the low response rate under this type of delivery. As one would
expect, one-on-one delivery provides a much more effective delivery (benefit), although
more inputs (costs) are required in terms of resources than other delivery methods.
High Risk Areas Noted- The most common high risk areas noted were in petroleum
product storage, well condition and/or location, and pesticide storage and handling.
Table 1 below lists the percentage of high risk cases noted by the 134 participants.
Table 1:
Areas of High Risk in Louisiana Case Study
Identified during the Farm Assessment Process
Area (Worksheet) High Risk Cases (n=134)
Petroleum product storage 23.3%
Well condition and/or location 17.3
Pesticide storage and handling 12.0
Household waste water treatment 11.3
Farmstead hazardous waste management 9.8
Livestock waste management 7.5
Fertilizer storage and handling 5.3
Poultry litter and carcass management 5. 3
Changes Made and Planned and their Associated Costs - Results show that 43 (32%) of
184
-------
the 134 farmers made, or planned to make, 66 individual changes in their farming practices
after undergoing the assessment. These changes were specified by the participants to have
been made "due to information provided by the Farm *A* Syst materials," and totaled
$91,437 ($682 per farmer based on 134 participants). Farmers' time, which respondents
valued at an average (median) of $12.00 per hour, constituted approximately one-third of
this total.
Changes that farmers either made or planned to reduce their risk to groundwater
contamination as a result of undergoing Farm *A* Syst ranged from strictly "management
type" decisions which required no direct outlay of cash (such as mixing pesticides in the
field away from the well, or properly recycling used oil and antifreeze), to changes that
required low to high out-of-pocket expenditures (such as putting back-flow valves in well
pumps, or pulling a leaking underground storage tank). Figure 1 below shows a summary
of these actions. Please note that the "management type only" changes have a cost that
is in terms of farmers' time only, and do not have any out-of-pocket cost component.
Figure 1:
Types of Changes Made and Planned by Louisiana Farmers due to
Information Gained from Undergoing Farm *A* Svst
See Page 193
Expenditures and Risk - One of the issues addressed in the study was whether or not
farmers were making changes to those areas where they needed to the most~or more
specifically where they noted a high risk. Analysis of the data collected shows that this was
indeed the case. For example, nitrate and/or bacteria problems were noted as a high risk
by 47% of the farmers surveyed, and 44% of the monies spent by farmers was allocated
to changes involving nitrate and/or bacteria issues. A graphical representation of this
analysis is shown in Figure 2 below.
185
-------
Figure 2:
Comparison of High Risks Noted and Expenditures on Changes
See Page 194
Economic (Cost-Benefit) Measurements-The methods of averting expenditures, avoided
cost of clean-up and/or change in water supply, and contingent valuation produced present
value estimates of net benefits (over an assumed ten-year period) of $2.7 million, $2.4
million and $15 million, respectively. Details are presented in Table 2 below. All figures are
in present value terms (PV). As discussed in the previous section, the methods of averting
expenditures and avoided cost are lower bound estimates of net benefits in that benefits
are a function on the observed behavior of expenditures noted. For instance, farmers may
be willing to spend much more to reduce their risk. The method of contingent valuation
takes this into account, and therefore produces a much higher willingness-to-pay. In
addition, the latter method utilizes all farming households in the state as a base for
extrapolation, whereas the first two methods utilize only that portionof the farming
community (approximately one-third of all farming houselhods) that is estimated to
participate in an environmental program such as Farm *A* Syst.
The discount rate used for government costs and farmers costs were the average T-Bill rate
(5.2%) and farm-borrowing rate (7.5625%), respectively, for the time period in which the
study occurred.
Table 2:
Comparison of Net Benefits of Farm *A* Syst Estimated Under Three Methods
Method of Benefit PV Benefits PV Cost 'PV Net Benefits
Estimation (over 10 yrs) (over 10 yrs) (over 10 yrs)
per Year
#1-Avert. Expend's: $3,993,691 $1,262,261 $2,731,430 $273,143
(8,551 farmers)
#2-Avoided Cost 7,615,555* 5,255,951 2,360,604 236,060
(8,551 farmers)
#3-Contingent Value Q's:
-all farm h.hold's 20,409,713 5,255,951 15,153,762 1,515,376
(25,653)*
• In this study, and due to the methodology of the Jordan and Elnagheeb study from which the WTP figures are derived, the contingent
valuation method incorporates both users and non-users of the resource (all farming households), whereas the other two methods
incorporate only users (one-third of all farming households, or only those households expected to be similar to the volunteer sample). It
should be noted that this estimate is a conservative one, as the contingent valuation method could theoretically incorporate other rural
and even non-rural households as well, which would produce estimates in net benefits of over $1 billion.
186
-------
Possible Effects of Economic Incentives to Qualifying Farmers — Providing farmers with a
cash incentive to undergo a voluntary pollution prevention program such as Farm *A* Syst
may help to spur participation as well as increase the amount of changes the farmer will
make. A small portion of participants (5 out of 134) qualified for a $150 cash incentive under
the NRCS's Water Quality Incentive Program (WQIP). Of this amount, $75 was required
to be used towards a water test. Three (60%) of these participants made or planned to
make six changes estimated at a total cost of $16,850, compared to the other 129
participants (termed as the primary sample) which made/planned $74,587 in changes.
Table 3 below details the comparisons between the groups. Note that WQIP has recently
changed to EQIP, or Environmental Quality Incentive Program. It should be stressed that
caution is warranted here with regard to the small sample size. However, the authors felt
the information should be presented so that consideration be given to incentive-based
delivery methods for other state programs.
Table 3:
Comparisons Between WQIP Farmers and Other Farmers
% Farmers Making/Planning Changes
Expenditures per Farmer
Number of Changes Per Farmer
Expenditures per Change
Primary
Sample
(n=129)
30.2%
$578
.46
$1,243
WQIP
Farmers
60.0%
3,370
1.2
$2,808
% Difference
(of %'s)
99%
483%
161%
126%
Non-Economic (Educational) Benefits of Fann *A* Syst — There may be many other benefits that
farmers can gain by undergoing a program like Farm *A* Syst in addition to those that are
measured in economic terms only. For example, it is clear that many farmers in the study had a
187
-------
"positive" experience with Farm *A* Syst. They showed significant gains in knowledge and a
positive change in attitude after having gone through the process.
In order to test for these educational benefits, comparisons of means in response rates of the 21
questions asked both before and after a farmer underwent his/her risk assessment showed, in
most every instance, a positive change in knowledge, attitude, perception or awareness regarding
the potential effects of farming practices on groundwater quality. Many of these changes were
statistically significant. For example 73% of the respondents changed away from a "Don't Know"
response (as opposed to changing to a "Dont Know" response) in the post-survey to the question
"What is the most common source of nitrate found in well water in your area?" Similar statistics
were noted in the question "what extent is pollution of well water as a result of farming practices
a problem in your area?"
Another example is that a significant number of respondents (at 95% confidence level) in the post-
assessment survey said they would "Find source of nitrates and correct the problem" if they found
out their "water supply was polluted with a high level of nitrates," as compared to their answers
given in the pre-assessment survey (their other choices were "install water-treatment system" or
"find a new water supply").
Use or Worth of the Program - It is also very important to have an understanding of what farmers
felt about the program in general. For instance, did they think Farm *A* Syst was worth doing, or
would they recommend it to their neighbors? And how difficult did they think the materials were
and how long did it take to go through the assessment process? These questions were asked in
the surveys and tested for significance, and all had positive results.
Of hose who changed their answers (to either "yes" or "no") after undergoing the assessment, 75%
of those respondents thought the assessment was worth doing ("yes") after they had done it as
compared to 25% who changed in the other direction (to "no")--significant at 99%. Nearly all
respondents thought the worksheets were easy to complete, and that all of the questions were
understandable. Approximately 75% of the respondents were assisted (through a "one-on-one"
delivery) by either an Extension, NRCS, or USDA AmeriCorp member. This sort of delivery system I
is very common for Farm *A* Syst-and programs like it~on a state-by-state basis. The average j
time needed to complete the assessment was about 1.8 hours. I
Implications for Program Coordinators
Impact evaluation is now a key component of any government-funded program, and program j
coordinators should build-in evaluation procedures early on in program development, and ,
especially prior to program implementation. One of the major findings of this research is that the
methodology can be used to quickly evaluate the costs and benefits of similar programs. Likewise,
environmental programs that are designed for non-farm entities (such as the Home *A* Syst
program) can also benefit from the design. ,
188
-------
The cost-benefit information otained in this study is the first of its kind for Farm *A* Syst, and is
the only information currently available to help estimate the economic impact of the program on
a national basis. It is estimated that 42,000 farm assessments have been conducted throughout
the US to date. By applying our best estimate of economic impact~$700 per farming household
due to inforation gained by Farm *A* Syst--the result is nearly $30 million in pollution prevention
expenditures. This type of impact evaluation data is extremely important in terms of maintaining
or increasing support for the program.
Other programs that are based on the Farm *A* Syst approach, such as the new Resource
Conservation Planning initiative-as described in the recent NRCS Materials Development Team
Report-could also benefit from this cost-benefit evaluation methodology. This report recommends
placing the responsibility of these flexible, voluntary and confidential planning exercises in the
hands of the farmer or rancher, including issues such as: crop management, erosion control, water
quality, recreational opportunities, fish and wildlife habitat protection/enhancement, and other
community concerns.
If that approach is taken, cost-benefit information-provided in layman terms relating to monies
saved or costs avoided—could help foster participation. This planning process is much broader
than the Farm *A* Syst program which was evaluated in Louisiana. Thus, additional cost-benefit
indicators would need to be identified.
One of the recommendations by the NRCS Development Team was to "monitor the performance
and outcomes of programs," thus changing the emphasis from individual employee performance
monitoring to program-based-monitoring. The methods utilized in this study could help to reach
this goal, as each of the planning components will undoubtedly be evaluated in terms of economic
and non-economic costs and benefits. Continued or increased funding of these programs could
depend in part on such impact evaluations.
Lessons Learned
The following recommendations should help policy makers and decision makers who wish to
evaluate-in both a social (cost-benefit) and physical (environmental) manner-voluntary pollution
prevention programs in agriculture.
Recommendation #1: The impact evaluation process is one that should be built-in at the fore-front
of the overall planning of new programs so that accurate and timely information can be
documented to evaluate the program once implementation takes place.
Recommendation #2: Decision makers should analyze the economic impact$--in addition to
non-economic impacts-of policies and programs in order to gain a more comprehensive
understanding of the holistic effect of the program. Documentation of such economic impacts
will help to make a case for continued or increased funding of programs that are effective.
189
-------
Recommendation #3: The "model methodology" and set of "model surveys" designed for the
Louisiana Farm *A* Syst study should be utilized as a starting point for evaluators and decision
makers when conducting cost-benefit studies on voluntary pollution prevention programs in the
agricultural sector.
Recommendation #4: Program evaluators should consider measuring the economic costs and
benefits of government programs and policies under more than one method, in order to provide
a better understanding of benefit estimations and limitations, and the ranges of which such
calculations can vary. Three methods that are recommended as a starting point in programs
similar to Farm *A* Syst are: (1) averting expenditures; (2) avoided cost of clean-up and/or
change in water supply, and; (3) contingent valuation.
Recommendation #5: One method that should be considered to statistically test for the non-
economic benefits of voluntary pollution prevention programs is to measure the changes (from
pre- to post-program participation) that participants have regarding inquiries to their knowledge,
perceptions and attitudes about how their farming practices can affect the quality of their water
resources.
Recommendation #6: Decision makers should conduct further testing on the impacts of
providing economic incentives to potential users to increase program participation in voluntary
pollution prevention programs.
Recommendation #7: Program evaluators should, whenever possible, conduct similar
evaluations by using a stratified and randomly selected sample of participants. Additionally,
evaluators should maintain proper control over the survey delivery and follow-up procedures in
accordance with such standards recognized in the literature.
Recommendation #Q: Data collection of impacts of programs such as Farm *A* Syst should
continue and be on-going. Data collection and analysis of delivery methods other than one-on-
one (such as the group workshop method, for instance) is especially encouraged, as
comparisons could then be made with regard to the net-benefits of each. This would help to
determine which type of delivery yields the highest net-benefit and net-benefit ratio.
References
Abdalla, Charles W., Roach, Brian A., and Epp, Donald J., "Valuing Environmental Quality Changes Using Averting
Expenditures: an Application to Groundwater Contamination," Land Economics. 68(2): pp. 163-169, May 1992.
Babbie, Earl, Survey Research Methods, published by Wadsworth Publishing Company, Befmont, CA, 1992.
Bishop, Rich, personal communication, Department of Agricultural Economics, University of Wisconsin-Madison, 1995.
Boyle, Kevin P., NRCS Economist, Department of Rural Sociology, University of Wisconsin-Madison, personal
communication, April 1997. .
Boyle, Kevin J!, "A Comparison of Contingent-Valuation Studies of Groundwater Protection," Department of Resource
Economics and Policy-University of Maine. Staff Paper REP 456, 69 pages, May, 1994.
Branch, Bill, personal communication, Environmental Specialist with Louisiana Cooperative Extension Service,
Louisiana State University, 1993-1995.
190
-------
Cooperative State Research, Education and Extension Service (CSREES), "CSREES Blueprint for Implementing the
Government Performance and Results Act (GPRA) of 1993, Draft, May 1,1995.
Dillman, Don. A., Mail and Telephone Surveys: The Total Design Method. John Wiley & Sons, New York, 1978.
"Environmental Indicators of Water Quality in the United States," United States Environmental Protection Agency, Office
Of Water (4503F), EPA 841-R-96-002, June 1996.
Field, Barry C., Environmental Economics: An Introduction, published by McGraw-Hill, Inc., 1994.
Freeman, A. Myric III, The Measurement of Environmental and Resource Values: Theory and Methods, published by
Resources for the Future, 1993, second paperback printing 1994.
Johnson, Rebecca L, and Johnson, Gary V., Editors, Economic Valuation of Natural Resources: Issues. Theory, and
Applications, published by Westview Press, 1990.
Jordan, Jeffrey L., and Elnagheeb, Abdelmoneim H., "Willingness to Pay for Improvements in Drinking Water Quality,"
Water Resources Research. Volume 29, No. 2, pp. 237-245, February 1993.
Laughland, A.S., Musser, L.M., Musser, W.N., and Shortle, J.S., "The Opportunity cost of Time and Averting
Expenditures for Safe Drinking Water," Water Resources Bulletin. American Water Resources Association, Vol.
29, NO. 2, April 1993.
Mitchell, Robert Cameron, and Carson, Richard T., Using Surveys to Value Public Goods: The Contingent Valuation
Method, published by Resources for the Future, 1989, distributed worldwide by The Johns Hopkins University Press.
Moore, Billy, personal communication, Assistant State Director of Programs in Louisiana, National Resources
Conservation Service, Alexandria, LA, 1995.
Moreau, Robert J., Cost-Benefit Analysis of Voluntary Pollution Prevention Programs in the Agricultural Sector: Case
Study of the Farm Assessment System (Farm *A* Syst). doctoral dissertation, Institute for Environmental
Studies, University of Wisconsin-Madison, under the direction of John Strasma-Professor of Economics.
Moreau, Robert J., Jackson, Gary, and Strasma, John, "Impact Assessment of Voluntary Pollution Prevention
Programs in Agriculture: Case Study of a Cost-Benefit Analysis of Louisiana's Farm *A* Syst," submitted to
Journal of Soil and Water Conservation in April 1997 (awaiting decision on publication).
Nevers, E., Jackson, G.W., Castelnuovo, R., and Knox, D., "Risk Assessments: Site Specific Management Tools for
Preventing Pollution on Farms," Proceedings of the Second Conference on Environmentally Sound Agriculture.
Orlando, FL, April 20-22,1994.
Ryding, Sven-Olof, Environmental Management Handbook. IOS Press, Amsterdam-Oxford, Lewis Publishers, Boca
Raton, Florida, 1990.
Shaw, Leil Hamilton, personal communication, Louisiana Department of Environmental Quality, Office of Solid and
Hazardous Waste, Division of Solid Waste, Groundwater Enforcement, 1995.
U.S. Department of Commerce - Economics and Statistics Administration - Bureau of the Census, "1990 Census of
Population: General Population Characteristics," 1990CP-1-2Q.
"Whole Farm and Ranch Conservation Planning," Materials and Development Team Report, consisting of
representatives from NRCS and other agencies/programs, Draft, June 1996.
191
-------
Acknowledgments
The following organizations and individuals have all played a very important role in the
research and development of this study. Their efforts are sincerely appreciated.
Bill Branch - Environmental Education Specialist,
Louisiana Cooperative Extension Service
Billy Ray Moore - Assistant State Conservationist for Programs in Louisiana,
Steve Nipper - Water Quality Specialist, Monroe District,
and USDA AmeriCorp Representatives involved with the study through the
Natural Resources Conservation Service (NRCS)
Wendell G. Miley - Coordinator of Environmental Affairs in Louisiana,
Farm Bureau Federation
All Participating Farmers in Louisiana
Authors' Names and Addresses
Robert J. Moreau, Ph.D. (primary author)
Impact Evaluations Specialist
National Farm Assessment System (Farm *A* Syst) and
Home Assessment System (Home *A* Syst) Programs
B-142 Steenbock Library 550 Babcock Drive
Madison, Wisconsin 53706-1293
Phone: 608-265-2996 email: moreau@aae.wisc.edu
Gary Jackson - National Director
National Farm Assessment System (Farm *A* Syst) and
Home Assessment System (Home *A* Syst) Programs
B 142 Steenbock Library, 550 Babcock Drive
Madison, Wl 53706-1293
Phone: 608-265-2773 email: gwjackso@facstaff.wisc.edu
Doug Knox-NRCS Representative
National Farm Assessment System (Farm *A* Syst) and
Home Assessment System (Home *A* Syst) Programs
B 142 Steenbock Library, 550 Babcock Drive
Madison, Wl 53706-1293
Phone: 608-265-2772 email: dcknox® facstaff.wisc.edu
192
-------
_
o
c
o
s_
o
<4-
e
H-)
O
4->
O)
3
CU
s-
fO
cu c
T3 -^
to o
co OJ
(U -o
en c
c =>
03
^c E
C_3 O
4- 4-
o
•a
, to
CO
Q
_3
Q.
193
-------
to
CT>
c
ro
-c
o
c:
O
CO
O)
C
OJ
Q.
X
-o
c
CO
OJ
C71
C
o
CO
re
Q.
O
O
cu
S-
-------
Analysis of Long Creek Watershed Monitoring Data
D.E. Line, W.A. Harman, and G.D. Jennings
Biological and Agricultural Engineering Dept., NC State University
Raleigh, NC
Nutrient and solids concentrations in streamflow were monitored during a three-year
period in a 56.6-ha, mostly agricultural, watershed located in southwestern North
Carolina. Nutrient and solids concentrations in weekly grab samples were significantly
greater downstream of an overgrazed cow holding and feeding area compared to upstream
during a two-year period prior to the implementation of best management practices
(BMPs). During the year of monitoring after installing livestock exclusion fencing
nutrient and sediment concentrations decreased significantly, especially downstream of
the cow holding area. Both parametric and nonparametric analysis of nitrogen,
phosphorus, and solids monitoring data indicated statistically significant reductions in the
concentrations of pollutants after BMP implementation
Introduction
The Long Creek watershed is the site of a nine-year comprehensive watershed project
initiated in 1993 to improve stream water quality while documenting the effectiveness of
nonpoint source pollution controls. The project is one of 20 comprehensive watershed
monitoring projects in the U.S. EPA National Monitoring Program (Osmond et al., 1995).
The Long Creek drains a 11,530-ha Piedmont watershed in Gaston County, North
Carolina. It has documented water quality degradation caused by sediment, bacteria, and
nutrients (NC DEM, 1989). Potential pollution sources include agriculture (livestock and
crop production), mining, forestry, urban runoff, septic system outflow, streambank
erosion, and discharges from industries and wastewater treatment plants. Approximately
O T
20,060 m of animal waste and 21,950 m of municipal sludge is applied to agricultural
land in the watershed annually.
The project watershed contains five management or study areas. The most intensively
monitored study area is a dairy farm pasture and heavy use area which is the focus of this
paper. Land use in the monitored watershed (fig. 1) is mostly pasture with areas of
residential houses, small business, woods, and farm buildings along the watershed
boundary. The 42-ha area upstream of site D contains mostly pasture which provides
supplemental forage to between 50 and 100 replacement heifers. Several homes, a small
apartment complex, and a small business are located along the periphery of the upstream
(site D) drainage area. The 14.6-ha area between sites D and E is grazed regularly by
between 50 and 100 adult dry cows and occasionally by another 50 to 100 milking cows.
Due to the increased stocking density, grass in the area between sites D and E is sparse
overall, with relatively large denuded areas in places where the cows are fed.
Additionally, the streambanks between sites D and E are eroding at an accelerated rate
due, at least partially, to unlimited access of the cows to the stream.
195
-------
The average annual rainfall in the monitored area is about 1090 mm. The watershed
geology is typical of the western Piedmont with a saprolite layer of varying thickness
overlaying fractured igneous and metamorphic rock. Soils in the drainage area are
generally well-drained and have a loamy surface layer underlain by a clay subsoil. Both
upstream and downstream drainage areas are hilly with land slopes of 5 to 15 percent.
The small stream draining the watershed flow continuously with discharges ranging from
1.4 L/s to more than 2800 L/s during large storm events. Several normally dry channels
enter the stream between the upstream and downstream monitoring sites; thus, the
discharge at sites D and E are nearly equal during basefiow conditions, but, because of
the heavy use areas and large building roofs, can be quite different during storm events.
Land treatment in the monitored drainage was focused on treating the area between
monitoring sites D and E. The only BMP installed upstream of site D was an alternate
watering system for the 50 plus heifers grazing in the drainage area. BMPs implemented
between sites D and E include an alternate watering system, livestock exclusion fencing,
improved stock trails, heavy use area stabilization, a runoff spreader, and a cattle stream
crossing. The fencing excluded the cows from a corridor ranging from 18 to 31 meters
wide along the stream. Grass has been established on streambank areas with severe
erosion and hardwood trees have been planted along the entire 330 meters of the riparian
corridor. Additionally, volunteer vegetation in the riparian corridor has helped to stabilize
streambanks and channels leading to the stream.
Procedure
Grab samples from the overfall of a v-notch weir at site D and a large culvert at site E
were collected weekly since April of 1993. The grab samples were collected within 20
minutes of each other, iced immediately, and transported to a nearby laboratory within
five hours where they were preserved using approved methods. Samples were analyzed
for nitrite+nitrate nitrogen (NO2+3), total Kjeldahl nitrogen (TKN), total phosphorus (TP),
total suspended solids (TSS), and total solids (TS) concentrations by a U.S. EPA certified
laboratory. Samples were analyzed using methods 353.1, 351.2, 365.4 from U.S. EPA
(1983) for NO2+3, TKN, and TP and 2540D and 2540B from APHA et al. (1989) for TSS
and TS. Split, blank, and spiked samples were prepared and analyzed to verify the quality
and representativeness of the samples.
Results and Discussion
Documenting an effect of a BMP using the upstream-downstream monitoring design can
be accomplished in several ways including regression analysis and parametric or
nonparametric analysis of variance. The regression analysis requires a relationship
between upstream and downstream pollutant concentrations that is different before
compared to after BMP implementation. Due to the large, random influence of cows in
196
-------
the stream, a relationship between pollutant concentrations at D and E before BMP
implementation was weak to nonexistent; therefore, this method was not used. Three
analysis of variance methods were used on the differences between concentrations at sites
DandE.
Figure 2 shows the differences in TSS concentrations at sites D and E with positive
differences indicating concentrations greater at E. For the pre-BMP period (4/3/93 to
2/6/96), observation of bars suggests that TSS concentrations are generally greater at site
E than at D while after BMP implementation differences are nearly zero. Trends were
similar, although the magnitude of the differences in concentrations after BMP
implementation were greater for NO2+3,TKN, TP, and TS or in other words, the BMPs
were not quite as effective at reducing concentrations of the other pollutants.
Univariate analysis (SAS Institute, 1985) of the concentration differences between D and
E indicated that the data were not normally distributed; therefore, the data was log-
transformed before performing an analysis of variance (ANOVA). Additionally,
nonparametric statistical analyses were used to assess the data. The average difference
between D and E pre- and post-BMP implementation along with probabilities of
exceeding the computed test statistic for ANOVA, the Wilcoxin Rank Sum test, and the
Kolmogrov-Smirnov test (SAS Institute, 1985) are shown in Table 1. The tests were used
to compare the differences in concentrations before BMP implementation to after
implementation. Because only one year of monitoring data has been collected since
implementation, fewer (43 versus 143) data exist for the post-BMP period.
Table 1. Probabilities of Exceeding the Test Statistic for Comparisons Between
Differences in Concentrations at Sites D and E Before and After BMPs.
Mean at D Mean of Differences' ANOVA2 Wilcoxin Kolmogorov-
Pollutant Pre-BMP Pre-BMP Post-BMP Rank Sum Smirnov
mg/L mg/L mg/L Pr>F Pr>Z Pr>KSa
NO2+3 0.25 0.59 0.0174 0.0030 0.0082
TKN 7.61 0.50 0.0001 0.0001 0.0001
TP 1.42 0.28 0.0961 0.0001 0.0001
TSS 70.4 0.17 0.0001 0.0001 0.0001
TS 153.0 27.9 0.0055 0.0004 0.0003
1 average of concentrations at site E minus D for 143 pre- and 49 post-BMP samples.
analysis of variance was conducted on the logarithms of the concentrations.
A general indication of the level of pollutants in the stream is given by the pre-BMP
concentrations of pollutants shown in column two of Table 1. These mean levels are also
a rough estimate of post-BMP concentrations considering that the only BMP
implemented upstream of site D was the alternate drinking water supply. Columns three
and four show that the increase in concentrations from upstream to downstream is
reduced by the implementation of BMPs with the exception of NO2+3 The concentration
of NO2+3 increases downstream relative to upstream with the implementation of BMPs.
197
-------
This seemingly negative effect of the BMPs can be explained by an increasing trend in
N02+3 concentrations throughout the duration of the project and the realization that the
riparian area may increase infiltration and shallow ground water movement thereby
increasing NO24-3 movement to the stream.
The ANOVA on the log-transformed data indicates that there is sufficient evidence of a
significant reduction in TKN, TP, TSS, and TS at the 0.10 level of significance and TKN,
TSS, and TS at the 0.05 level of significance. The results from nonparametric Wilcoxin
Rank Sum and Kolmogorov-Smirnov tests indicate relatively strong differences in pre-
and post-implementation data. The nonparametric tests indicate stronger differences than
the parametric ANOVA test, especially for TP. This may be due, at least partly, to the
magnitude of the variability in differences in TP concentrations which ranged from -0.23
to 33.8 mg/L.
Thus, statistical analyses of the pre- and post-BMP implementation concentration data
indicates that the BMPs are having a significant effect on levels of pollutants in the
stream. Observation of a greater diversity and intensity of aquatic life in the stream has
also indicated improved water quality since the implementation of BMPs.
References Cited
APHA, AWWA, WPCF (1989) Standard Methods for the Examination of Water and
Wastewater. 17th edition.
NC DEM. 1989. North Carolina Nonpoint Source Management Program. Report 89-02.
North Carolina Division of Environmental Management, Raleigh, NC.
Osmond, D.L., D.E. Line, J. Spooner. 1995. Section 319 National Monitoring Program:
An Overview. NCSU Water Quality Group, Biological and Agricultural Engineering
Department, North Carolina State University, Raleigh, NC.
SAS Institute Inc. 1985. SAS/STAT Guide for Personal Computers, Version 6 Edition.
SAS Institute Inc. Gary, NC.
U.S. EPA (1983) Methods for Chemical Analysis of Water and Waste. EPA-600/4-79-
020. U.S. Environmental Protection Agency, Cincinnati, OH.
198
-------
N
j Buildings
Waste storage
• Sampling site
Kilometers
0.2
0.4
Figure 1. Map of watershed area.
2,000
=8,1,500
E,
(0
1,000
I 50°
I
Pre-BMP-
^.Kee.
/ ///
Post BMP
Figure 2. Differences in concentrations of suspended solids in grab samples from sites D and E.
199
-------
Jordan Cove Urban Watershed Section 319
National Monitoring Program Project
John C. Clausen, Mary Hull and John Alexopoulos
University of Connecticut, Storrs, CT
Melville P. Cote, Jr.
U. S. Environmental Protection Agency, Boston, MA
Bruce L. Morton
Aqua Solutions, East Hartford, CT
Stan Zaremba
CT Department of Environmental Protection, Hartford, CT
Introduction
Runoff from urban areas is the second leading cause of nonpoint source pollution behind agriculture in the
U. S. (USEPA, 1994). In CT, MA, RI, and TX, urban runoff is the major source of nonpoint pollution.
The major types of pollutants found in urban runoff are more varied than from agricultural areas and
include suspended solids, nutrients, oxygen-demanding substances, pathogens, road salts, hydrocarbons,
heavy metals, and other toxins (USEPA, 1993a). Urban runoff sources also dominate nonpoint loading to
Long Island Sound, an estuary impaired by hypoxia and pathogen contamination (LISS, 1994).
The Nationwide Urban Runoff Program (NURP) evaluated runoff from 28 urban areas around the U. S.
(USEPA, 1983). Among the study's main conclusions were that concentrations of pollutants did not vary
significantly in runoff among differing types of urban land uses, such as residential versus commercial
(USEPA, 1983). However, open/non-urban areas had lower concentrations than urban areas. A few best
management practices (BMPs) were studied during NURP, such as wet detention basins, recharge devices,
and street sweeping. Since that time there have been additional studies of BMPs to reduce nonpoint
sources of pollutants from urban areas. Most of these studies have been related to the effectiveness of
individual BMPs, such as detention basins. Excellent summaries of urban BMP effectiveness can be found
in Schueler (1992) and USEPA (1993b). To date, however, there has not been an attempt to investigate the
effectiveness of an entire suite of BMPs applied to an urban development.
The objectives of this study are to determine the water quantity and water quality benefits from developing
an urban subdivision using pollution prevention BMPs. Site design and landowner education are major
BMPs. Two residential areas will be developed beginning July 1997; one using traditional subdivision
requirements and one using BMPs. Runoff from these two areas is being compared to an existing
residential control watershed. The purpose of this paper is to describe the study design, the BMPs that will
be used, and provide preliminary results from watershed calibration.
201
-------
Study Area
The project is located in Waterford, CT near Long Island Sound. The existing control is a 13.9-acre
residential watershed containing 43 lots ranging from 0.3 to 0.5 acre developed in 1988 (Figure 1). The
traditional subdivision is a 10.6-acre residential/commercial/poultry farm being developed into 18 0.3-acre
lots (Figure 2). The BMP subdivision is a 6.9-acre gravel pit being developed into 12 0.25-acre lots
following cluster guidelines (Figure 3). The control watershed has 30 % imperviousness while the
treatment area is at 9.1% imperviousness.
Methods
The overall study design is based on a paired watershed approach (Clausen and Spooner, 1993) using one
control and two treatment watersheds. This approach uses a calibration period and treatment period.
During a calibration period of 1.5 years, no land use changes in the watersheds occurred. The treatment
period includes two intervals: a 18-month construction period, and a long-term post implementation
monitoring period. During calibration, regressions are developed between paired observations from the
control and treatment watersheds, such as paired concentrations. A new regression is developed following
treatment. Analysis of covariance is used to test the difference between the two regression slopes and
intercepts.
Overland flow is being monitored using H-flumes at the treatment sites and a combination V-notch and
rectangular weir in the stormwater pipe at the control site. During stormflow periods, samples are
automatically taken of overland flow using ISCO samplers using flow proportional sampling and a weekly
composite. Samples are analyzed for total suspended solids, total phosphorus, total Kjeldahl, ammonia,
and nitrate nitrogen, copper, lead, and zinc. Grab samples are analyzed for fecal conform bacteria and
BOD5. Analyses are by EPA approved methods (USEPA, 1983b). Monitoring costs about $50,000/yr.
Results and Discussion
BMP Design
The BMP watershed incorporates several pollution prevention measures as part of its design. A
main feature is the replacement of curb-and-gutter and stormwater collection methods with
bioretention swales (Table 1, Figure 3). The traditional 28 ft wide asphalt road is being replaced
with a 20ft wide concrete paver road that allows 12 % infiltration (UNI Eco-Stone®). A
bioretention cul-de-sac that allows for detention and infiltration of runoff will be constructed in
lieu of a conventional paved area. Individual bioretention gardens are incorporated into each lot
to detain roof and lot runoff (Prince George's County, 1993). Several alternative driveway
surfaces are being installed including traditional asphalt, concrete pavers, porous concrete,
concrete two-track, and gravel. Finally, the lawns will be reduced and replaced with no-mowing
and low-mowing zones including native and other low maintenance vegetation. During
construction additional BMPs will be used, such as phased grading, seeding stockpiles, and post
storm maintenance. Following construction, additional BMPs will be employed over the long-
term (Table 1). Deed convenants will restrict imperviousness, obstructions in the swales and will
require maintenance of vegetation and driveways. Ongoing education programs will be devoted i
to lawn nutrient and pesticide management, pet waste management, and general good
housekeeping practices. Costs of BMPs will be tracked. Planning costs have been about double i
for the BMP lots ($808/lot) as compared to the traditional lots ($407/lot). i
202
-------
Figure 1. Existing Residential Control Watershed with contours. Waterford, CT
* .' K T- r «•*•
STREET /
TREES
Figure 2. Traditional subdivision watershed, Waterford, CT
Figure 3. BMP subdivision watershed, Waterford, CT
203
-------
Table 1. BMPs used in BMP subdivision watershed, Waterford, CT
Pollution Prevention Design BMPs
Long-term Maintenance
Bioretention swales
Narrow road surface
Infiltrating road surface
Bioretention cul-de-sac
Small 2-story house footprint
Roof/lot runoff bioretention gardens
Low-mow and no-mow areas
Reduced lawn area mowed
Driveway treatments
Shared driveway entrances
Cluster subdivision
Deed restrictions
Impervious controls
Swale obstructions
Driveway maintenance
Education efforts
lawn nutrient management
pet waste management
good housekeeping
Table 2. Median Concentrations in stormwater runoff from the Jordan Cove, CT monitoring
sites (11/95-5/97) and the Nationwide Urban Runoff Program (NURP)
Variable (mg/L)
Control
BMP
Traditional NURP
Total suspended solids
Total phosphorus
Total Kjeldahl nitrogen
Ammonia nitrogen
Nitrate-nitrogen
22.0
0.152
1.2
0.30
0.40
Biochemical oxygen demand 2
Fecal Coliform bacteria
Copper
Lead
Zinc
Discharge (ftVwk)
,
18
0.014
0.009
0.062
5001
f 10000 -•
H 8000 •
5 ^*
| 4000 •
1 2^tfijSf\*«
3.5
0.020
0.6
0.10
0.22
2
20
0.009
0.004
0.044
526
• ^
41.0
1.4
4.4
<0.01
0.1
— .
—
0.006
0.009
0.057
0.38
^
/
100.0
0.330
1.5
0.17
0.68
9
14,700
0.034
0.144
0.160
—
0 10000 20000 30000 40000
Control Watershed (cu ft/wfc)
Figure 4. Stormflow calibration regression for existing control and BMP sites, Waterford, CT
204
-------
Calibration Monitoring
Concentrations of pollutants in runoff from the existing residential control are somewhat lower when
compared to event mean concentrations from the NURP studies (Table 2). Runoff from the BMP site
exhibits lower concentrations of most water quality variables than the control site. At times the flow at the
BMP site is dominated by ground water flow from a seep in the old gravel pit.
Calibration for flow has been conducted between the control and BMP watershed. In order to develop the
regression between runoff from the two sites, hydrograph separation (recession baseflow back projection)
of stormflow and baseflow was necessary for the BMP site due to the ground water inputs. The regression
between the two sites (Figure 4) is significant (F=83.0, p=<0.001, R2=0.62). The median runoff from the
existing residential control is about 10 times that from the BMP site. Significant (p=0.05) calibration
regressions have also been establish for total suspended solids, total phosphorus, BOD, fecal coliform
bacteria, copper, and lead concentrations; and the mass export of ammonia, total Kjeldahl - N, total
suspended solids, and total phosphorus. Calibration of the control and traditional site is underway.
Conclusions
Several pollution prevention BMPs have been arranged in a residential neighborhood to reduce runoff of
nonpoint source pollutants from urban areas. Calibration of the BMP site has been achieved for discharge
but only after hydrograph separation of stormflow from baseflow. Calibration has also been achieved for
most water quality variables.
Acknowledgments
Funding to support monitoring and supplemental BMP installation is provided by grants to the CT Dept. of
Environ. Prot. from the USEPA under Section 319 of the federal Clean Water Act. The authors gratefully
acknowledge owners John and Gwen Lombard! and engineering by D. Gerwick .
References
1. Clausen, J. C. and J. Spooner. 1993. Paired Watershed Study Design. U. S. Environmental Protection
Agency. EPA 841-F-93-009. Washington, D. C. 20460
2. Long Island Sound Study. 1994. The Comprehensive Conservation and Management Plan.
3. Schueler, T. 1992. A Current Assessment of Urban Best Management Practices: Techniques for
Reducing Non-Point Source Pollution in the Coastal Zone. Department of Environmental
Programs. Metropolitan Council of Governments. Washington, D. C.
4. Prince George's County. 1993. Design Manual for use of Bioretention in Stormwater Management.
Landover, MD.
5. U. S. Environmental Protection Agency. 1983. Results of the National Urban Runoff Program. NTIS
PB84-185552.
6. U. S. Environmental Protection Agency. 1983b. Methods for Chemical Analysis of Water and Wastes.
EPA 600/4-79-020. Office of Research and Development. Cincinnati, Ohio. 45268.
7. U. S. Environmental Protection Agency. 1993a. Urban Runoff Pollution Prevention and Control
Planning. EPA/625/R-93/004. Office of Research and Development. Washington, D.C.
8. U. S. Environmental Protection Agency. 1993b Guidance Specifying Management Measures for Sources
of Nonpoint Pollution in Coastal Waters. EPA 840-B-92-002. Office of Water. Washington, D. C.
20460.
9. U. S. Environmental Protection Agency. 1994. National Water Quality Inventory. 1992 Report to
Congress. EPA 841-R-94-001. Office of Water. Washington, D.C. 20460.
205
-------
Partnerships in Puget Sound Watershed Remediation:
Linking Water Quality to Pollution Controls
Keith R. Seiders
Department of Ecology, Olympia, Washington
Mid-project water quality data and pollution control information from the Totten and Eld Inlet
Clean Water Projects are examined to determine if the installation of nonpoint source pollution
controls have decreased the concentrations of fecal coliform (PC) bacteria entering shellfish
harvest areas. Sources of bacterial contamination include poor livestock keeping practices at
recreational farms and failing on-site sewage systems. Farm plans were developed on 9 of 22
targeted sites in three study basins and 78 of 88 planned best management practices were
installed. Preliminary analyses of water quality data suggest that PC concentrations: decreased
31% in one creek, did not change in a second creek, and increased 600% in a third creek.
Introduction
Totten and Eld Inlets, in southern Puget Sound, (Figure 1) are highly productive shellfish
growing areas where harvest is threatened or restricted by nonpoint source pollution. More than
30% of the shellfish harvest areas in Puget Sound have been closed or restricted to shellfish
harvest (PSWQA, 1991). Sources of bacterial contamination to these shellfish growing waters
include failing on-site sewage systems and poor livestock keeping practices on small, mostly
recreational farrns. Three streams in the study area fail to meet state water quality standards for
PC (Schneider, Burns, and Pierre) and are included on Washington's (303(d)) list of impaired
waterbodies.
Watershed management plans were developed for the Totten Inlet (44,320 acres) and Eld Inlet
(23,220 acres) watersheds in 1989. Implementation of pollution management actions has
occured only when funding allowed. In 1993 and 1995, nearly $1.8 million in grant funded
projects enabled the Thurston County Environmental Health Division (TCEHD) and the Thurston
Conservation District (TCD) to focus efforts on nqnpoint pollution controls in these two
watersheds through 1999. Much of the effort has focused on working with landowners to repair
failing on-site sewage systems (OSSS) and mitigate pollution from small farm livestock keeping
practices. The goal of these clean water projects has been to to reopen harvest-restricted
shellfish growing areas and protect threatened areas within a relatively short time frame.
Washington's Department of Ecology is studying the effectiveness of nonpoint pollution control
programs in six sub-basins within these watersheds under EPA's National Monitoring Program
(NMP) guidance (EPA, 1991) and funding (approximately $0.5 million over 10 years). The goal
of the monitoring program is to detect trends in water quality and pollution controls and associate
these trends to each other. Water quality is monitored before, during, and after (1992-2001)
pollution controls are installed. Data on the initial installation of pollution controls are collected.
However, data are not available regarding their continued operation and maintenance. Historical
water quality data (1986-1992) generated by TCEHD were also used in project design and
evaluation. The paired watershed and single site over time monitoring strategies are used.
Fecal coliform bacteria is the major water quality parameter of interest, other parameters include:
total suspended solids, turbidity, streamflow, conductivity, temperature, and precipitation. Water
quality is monitored during the wet season on a weekly basis from mid-November through mid-
April (n=23). Information about potential sources of pollution and the management of those
207
-------
sources are collected and examined for the extent of coverage in each basin and possible
effects on water quality.
Implementation and Evaluation Approach
Farms in the Totten and Eld watersheds, and those in the study basins, were inventoried and
prioritized by TCD and TCEHD according to their potential to degrade water quality. TCD then
contacted priority farms and, where landowners were willing, developed farm plans and helped
the landowner implement the recommended Best Management Practices (BMPs). Livestock
commonly found on these farms include horses, beef cattle, llamas, donkeys, goats, sheep, and
chickens. In order to compare animal types and numbers, animal types were converted to
animal units (1 AU=1000 pounds). Farm animal inventory data were then used to estimate the
wet season animal population for each study basin: Schneider - 93.0 AU; Bums - 7.6 AU; and
Pierre - 5.0 AU. TCEHD selected homes in critical areas (stream corridors and marine
shorelines) for sanitary surveys of OSSS using dye-tracing techniques.
Water quality data were analyzed with two approaches to determine if trends in FC exist in
Schneider, Burns and Pierre basins: (1) comparison of pre- and post-BMP period FC
concentrations using notched boxplots which graphically depict the 95% confidence interval
about the median; and (2) comparison of pre- and post-BMP period FC relationships using
paired watershed data.
The paired watershed monitoring approach was applied in Kennedy (13,050 ac.) and Schneider
(4590 ac.) basins. The single site over time approach was used in Pierre (65 ac.) and Bums (82
ac.) basins (Figure 1). Kennedy (control) and Schneider (treatment) basins were cho'sen as
paired basins based on their watershed characteristics and plans for managing nonpoint
pollution. For Kennedy and Schneider, pre- and post-BMP period inter-basin pollutant
relationships were established for flow, FC, and FC loading. McLane (7430 ac.) and Perry (3860
ac.) basins are scheduled to receive agricultural pollution controls into 1999 after which water
quality will be analyzed.
Schneider basin has about 25 small farms that potentially impact water quality (17 were targeted
for pollution controls) whereas Kennedy basin has no farms. About 85% of each basin is
currently forested with the remaining land use classified as agriculture, residential, or other.
While Kennedy may not fit the classic definition of a "control" because agricultural land use is
dissimilar to that of Schneider, bacteria concentrations in Kennedy represent the natural
background condition (i.e. wildlife sources of FC only) and can be viewed as what is achievable
by eliminating human-caused effects through nonpoint pollution controls.
Pollution Control Implementation
Most of the farm planning and BMP installation work in Schneider and Bums basins occurred
from 1993 to 1995 and coincided with grant funding designated for agricultural remediation in
the Totten watershed. In Pierre basin, most of the farm planning and BMP installation effort
occurred between 1990 and 1993. The more recent efforts by TCD represent a significant
achievement in the installation of pollution controls in these basins since the watershed
management plan was completed in 1989. Despite these gains, funding to continue farm
planning and implementation work in these basins has not been made available.
For Schneider basin, 5 of 17 priority farms developed farm plans that included 45 individual
BMPs. The completeness of farm plan implementation for these 5 farms, expressed as the
208
-------
percentage of the number of implemented BMPs divided by the number of planned BMPs, is:
14%, 100%, 100%, 100%, and 100%. For Burns basin, 3 of 3 priority farms developed plans
containing 26 BMPs and completeness of implementation was: 100%, 100%, and 100%. For
Pierre basin, 2 of 2 priority farms developed farm plans containing 17 BMPs and completeness
was 57% and 90%. The most commonly applied BMPs included fencing, prescribed grazing,
filter strips, livestock exclusion, nutrient management, and water troughs (Mead and Konovsky,
personal communication, 1997).
Of the 9 farm plans developed in these basins, 5 of the sites appear to have developed farm
plans solely through voluntary action. The remaining 4 sites were encouraged to develop farm
plans by use of a formal referral process which has regulatory overtones. This process involves
the farm operator, TCEHD, TCD, and Ecology, in the progression of farm planning and BMP
implementation until water quality threats are mitigated to a satisfactory level (Starry, 1990 and
Hofstad, 1993).
TCEHD conducted sanitary surveys at 15 of a targeted 36 on-site sewage systems using dye
trace technology. No repairs were found necessary for systems within the Burns and Pierre
basins. In the Schneider basin, 3 systems were suspected as malfunctioning and confirmation
testing is scheduled (Thoemke, personal communication, 1997).
All homeowners in the Burns and Pierre basins participated in the 1994-95 sanitary survey.
About 36% of homeowners in Schneider basin participated in the 1997 sanitary surveys
(Hofstad, personal communication, 1997). The option for TCEHD staff to obtain an
administrative search warrant for inspecting OSSS was available during the 1994-95 sanitary
surveys. This option was unavailable after 1996 due to changes in the county's Sanitary Code.
This followed a Washington State Supreme Court ruling that administrative search warrants
could not be obtained for such inspection programs (Hofstad, et. al., 1996).
Water Quality Results
Two approaches were used to evaluate water quality: comparison of pre-and post-BMP median
FC concentrations using notched boxplots; and comparison of pre- and post-BMP paired-basin
FC relationships using linear regression. These analyses suggest that fecal coliform
concentrations: decreased 31 % in Schneider Creek, did not change in Burns Creek, and
increased 600% in Pierre Creek. Table 1 summarizes pre- and post-BMP periods that were
defined by examining available farm and BMP implementation data. For the paired-watershed
analysis, Kennedy data were paired according to pre- and post-BMP period data for Schneider.
Table 1. Pre-and post-BMP periods in study basins.
Basin Pre-BMP period Post-BMP period
Kennedy
Schneider
Burns
Pierre
none
1988-1993, 5 seasons
1 989-1 993, 4 seasons
1986-1989, 3 seasons
none
,1995-1997, 2 seasons
1995-1997, 2 seasons
1 993-1 997, 4 seasons
Table 2 summarizes the results of the pre- and post-BMP comparison of the median FC
concentration. Notched boxplots suggest that pre- and post-BMP median FC concentrations did
not change in Schneider or Burns and increased in Pierre.
209
-------
For the paired-watershed analysis with Kennedy and Schneider, pre- and post-BMP period
regression outputs were examined after Zar (1984) and EPA (1993). The slopes of these
regressions were not different while the y-intercepts were different (P<0.001). The difference in
intercepts, rather than slopes, indicates a parallel shift in the regression equation (Figure 2). This
shift in the regression represents a 31% decrease from the pre-BMP period (mean log FC=1.43)
to the post-BMP period (mean log FC=0.99).
Table 2. Median FC concentrations from pre- and post-BMP periods.
Basin Pre-BMP median Post-BMP median significant difference
FC and (n) FC and (n)
Kennedy
Schneider
Burns
Pierre
5 (39)
25 (39)
84 (35)
25 (11)
5 (45)
12 (45)
56 (45)
150 (89)
no
no*
no
yes
* See discussion of paired-watershed results where a difference in the mean log FC
concentration was detected.
Water Quality and Pollution Controls
Linking water quality to pollution controls is difficult due to our incomplete understanding of the
operation and maintenance of pollution controls, changes in farm management, effects of
climate, and sources and fate of FC in the study basins.
In Schneider basin, the decrease in FC appears to be due to the implementation of farm plans
as well as changes in farm ownership and farm management. One farm, just upstream of the
sample site, changed ownership after the original farm plan was developed. Fewer horses have
been observed at this farm during the last two seasons than were observed in previous years.
Twenty-one of a targeted 33 OSSSs were not surveyed in Schneider basin. The performance
and potential impact on water quality from these systems is currently unknown. The potential
effect on water quality from the farm with 14% of the planned BMPs implemented is also not
known.
In Pierre basin, the increase in FC might be attributable to a combination of factors such as
partial implementation of the farm plans, the lack of maintenance of previously installed BMPs,
and/or climate effects. One farm in the basin implemented 57% of the recommended BMPs
between 1990 and 1993; the remaining BMPs have yet to be implemented. The other farm in
Pierre basin implemented 90% of planned BMPs. More time may be needed for the effects of
BMPs to be measurable in Burns basin. The ability to link changes in water quality to pollution
controls could be improved by gaining current, complete, and accurate information about
management of farms and the operational status of BMPs and OSSSs.
The influence of climate on water quality during the study period is poorly understood. Climate
data show that the first two seasons of the NMP monitoring effort coincided with several years of
below-average rainfall while the latter three years (much of the post-BMP period) coincided with
rainfall returning to or exceeding the historical average. Rainfall totals, in inches, for the 6-month
period from November through April are: 1992-93:29.3, 1993-94: 25.7,1994-95: 40.5,1995-96:
210
-------
48.6, and 1996-97: 52.8. Linear regression analyses for each site suggest that FC
concentrations correlate poorly with streamflow and with an antecedent precipitation index which
reflects soil saturation and runoff potential.
The sources and fate of bacteria in the study basins are factors also poorly understood. Farm
animals and OSSS are presumed to be the primary sources of FC in the study area. While farm
animal numbers and their management have been inventoried in the past, resources are
unavailable to monitor the numbers of animals and their management. Animal waste that is
deposited or transported to streams may result in temporary storage, and later release, of FC
bacteria. Sherer et. al. (1992) suggests that for basins with livestock impacts, stream sediments
can be a reservoir and subsequent source of water column FC, particularly during periods of
disturbance such as during increased runoff. Examination of relationships between FC, turbidity,
and total suspended solids in these streams may help explain the role of fine sediments in water
column FC concentrations.
Lessons Learned
Current, complete, and accurate information on the operational status of previously installed
pollution controls is needed to link these controls to improvements in water quality. Much of this
information does not exist since state and local efforts focus on developing conservation plans
and helping landowners with initial installation of BMPs. Programs that actively help landowners
in properly operating and maintaining pollution controls beyond initial installation and procedures
for tracking such activities would contribute to long-term effectiveness and abilities to measure
success.
Accounting for and quantifying the effects of climate with the pre- and post-BMP single-site
monitoring design will be challenging. More thorough examinations of rainfall and hydrograph
data, as well as FC relationships to turbidity and total suspended solids, may help in
understanding the effects of climate.
A better understanding of the sources of FC is needed in order to link control of those sources to
changes in water quality. The amount of animal waste kept out of streams is many times a better
measure of environmental benefit than the enumeration of some BMPs (EPA, 1997). For a
stronger evaluation of pollution controls, information should be collected on the numbers, types,
locations, and management of animals as well as potential in-stream sources of FC.
Expectations for measuring water quality improvement from nonpoint pollution control projects
may need to be reduced because desired levels of voluntary participation may not be reached.
About 45% of the priority farms in the study basins participated in developing farm plans while
55% chose not to participate. However, of those farms with farm plans, the level of
implementation was generally excellent (90-100% implementation for 8 of 10 farm plans). Lower
levels of farm plan implementation could be masking benefits gained from complete
implementation of other farm plans (e.g. Pierre basin).
Regulatory and voluntary factors appear to motivate landowners to participate in OSSS and farm
pollution control efforts. Participation in OSSS survey programs decreased when a regulatory
tool was removed. Of 10 farms where plans were developed, 5 farm plans were voluntarily
developed while 5 plans were developed through the use, or probable use, of a formal referral
process. Farm planning on the remaining 12 priority farms in Schneider basin seems unlikely to
occur due to expiration of grant funds and time.
211 '
-------
References
EPA, 1991. Watershed Monitoring and Reporting for Section 319 National Monitoring Program
Projects. EPA Office of Water, August 1991, Washington D.C.
EPA, 1993. Paired Watershed Study Design. EPA#841-F-93-009. Off ice of Water,
Washington D.C.
EPA, 1997. Techniques for Tracking and Evaluating the Implementation of Nonpoint Source
Control Measures: Agriculture. Final Review Draft, EPA Office of Water, January 1997,
Washington D.C.
Hofstad, L, D. Tipton, and S. Berg, 1996. Shellfish Protection Initiative - Eld Watershed: 1993-
1996. Thurston County Environmental Health Division, Olympia, WA.
Hofstad, L., 1993. Watershed Implementation: Eld, Henderson, and Totten/Little Skookum,
1992-1993. Thurston County Environmental Health Division, Olympia, WA.
Mead, M. and J. Konovsky, 1997. Personal communication, June 1997. Thurston
Conservation District, Olympia, WA.
PSWQA, 1990. 1991 Puget Sound Water Quality Management Plan. Puget Sound Water
Quality Authority, Seattle, WA.
Sherer, 6. M., R.J. Miner, J.A. Moore, and J.C. Backhouse, 1992. Indicator Bacteria Survival
in Stream Sediments. Journal of Environmental Quality. 21:591-595.
Starry, A., 1990. Totten/Little Skookum Inlets and Watershed: 1987-1989 Water Quality and
Remedial Action Report. Thurston County Environmental Health Division, Olympia,
WA.
Thoemke, T. 1997. Personal communication, April 1997. Thurston County Environmental
Health Division, Olympia, WA.
Zar, J. H., 1984. Biostatistical Analysis, 2nd Edition. Prentice-Hall, Inc., Englewood Cliffs, N.J.
212
-------
Study Basins
Figure 1. Study area in southern Puget Sound.
3.0
2.5
2.0
1.5
I 1-0
(O
0.5
0.0
Calibration toQY-OB7-logX+0.76. r'-O.Sl
* *
o°
Treatment togY-0.56«togX+0.53. r'-O.OO
0.0 0.5 1.0 1.5 2.0
Kennedy Log FC (control)
Figure 2. Treatment and calibration
period regressions.
2.5
213
-------
The National Water-Quality Assessment Program —
At 5 Years Old
A Synopsis of the NAWQA Program and its Accomplishments
U.S. Department of the Interior
The primary goal of the National
Water-Quality Assessment
(NAWQA) Program of the U.S.
Geological Survey (USGS) is to
systematically document water-
quality conditions across the
country. This allows comparisons to
be made among river basins and
aquifers and facilitates analysis of
national issues and trends.
NAWQA assesses water-quality
conditions in 59 river basin/aquifer
system nationwide. Selection of
study units reflects significant
population centers, major sources of
drinking water, and diverse environ-
mental settings. Investigative
methods used by the Program are
designed not only to capture
changes in water quality over time
but to explain how human and
natural factors affect water quality
in different parts of the Nation.
Working With Other
Monitoring Programs
The first 2 years of any
NAWQA assessment involve
compiling and analyzing existing
data collected by Federal, State and
university scientists and by commu-
nity groups. Using both this "retro-
spective" information and NAWQA
protocols, study-unit scientists
develop a plan to guide 3 years of
rigorous sampling of physical,
chemical and biological characteris-
tics of a river basin/aquifer system.
U.S. Geological Survey
Study-Unit Investigations
In 1991, NAWQA made the
transition from a pilot program to a
full-scale monitoring program with
the start of the first 20 study-unit
investigations. An additional 16
study units started operation in
1994, and another 17 are slated to
begin in 1997. Six study units have
been selected for assessment but
have not yet been assigned a
starting date.
To make the Program cost-
effective, intensive assessment
activities in each study unit are
conducted in rotation, with about
one-third of the 59 study units
active at any given time.
In full operation now for 5
years, NAWQA is designed to be an
on-going program. Every 10 years,
intensive sampling activities cycle
through each study unit, updating
data and characterizing trends.
National Synthesis Studies
In addition to study-unit inves-
tigations, NAWQA's National
Synthesis Teams carry on analyses
that put local and regional condi-
tions into a larger, national perspec-
tive. Presently, there are three
Synthesis Teams that focus, respec-
tively, on pesticides, nutrients, and
volatile organic compounds
(VOCs). A fourth Synthesis Team
with a focus on trace metals is
planned.
What makes NAWQA unique?
Findings developed by NAWQA help answer questions about the
condition and sustainability of the Nation's water resources: What is
the condition of my drinking water source? Are streams providing good
habitat for fish? Is residential expansion affecting ground water? How
do agricultural effects on water quality compare with urban effects?
Has the money spent cleaning up our waters made any difference?
State and local water-quality monitoring programs, which
complement NA WQA assessments, are generally not designed to
support the range of local, regional and national-scale analyses
accommodated by NA WQA.
215
Elements of a NAWQA
Assessment
Assessments of rivers and
streams focus on:
•pesticides
• nutrients
• polychlorinated
biphenyls (PCBs)
• industrial chemicals
• metals
• stream habitat
• fish communities
Assessments of ground water
focus on:
•pesticides
• nitrate
• dissolved solids
• radon
• industrial chemicals
Use of NAWQA Data and
Assessments
Long-term, consistent, nation-
wide data like these have been
previously unavailable from any
Federal or State monitoring pro-
gram. The lack of such data has
made it impossible to evaluate
many questions of concern to
policymakers, resource managers,
and the public.
Results from the NAWQA
Program form a basis upon which
States and the Nation can identify
existing and emerging water-quality
issues, evaluate the effectiveness of
management strategies, and formu-
late more cost-effective programs.
For example, Washington State
implemented a flexible and cost-
saying monitoring program for
drinking water wells after collabo-
rating with NAWQA on an assess-
ment of pesticides in ground water.
The White House Office of Science
and Technology Policy (OSTP) and
the U.S. Environmental Protection
Agency (EPA) are using NAWQA
data on the occurrence of VOCs and
pesticides to consider policy and
regulatory options.
-------
STUDY-UNIT HIGHLIGHTS
Washington State health officials have
implemented monitoring plans for
drinking water wells based on the
relative risk of pesticide contamination.
The risk survey was designed and
In response to public concerns, bed
sediment and fish tissue in Idahos
Upper Snake River were examined
conducted in partnership with NAWQA. for mercury contamination. While
Savings to the State will exceed $6 mercury was detected, concentrations
__•?!• ___ »_ • • - .... .....
million each year.
Although banned 25 years ago,
DDT persists in the Yakirna and
Quincy-Pasco River Basins, of
Washington State at
concentrations that exceed
wildlife protection guidelines
established by the National
Academy of Sciences (NAS) and
Environment Canada. * The
Royal Lake refuge and Lind
Coulee recreational fishing areas
are of special concern.
Radon, a naturally
occurring radioactive
gas, is commonly
found in household
wells near Lake
Tahoe, Nevada, at
concentrations
above the State
drinking water
standard. State
officials and
NAWQA
collaborated on a
pamphlet that
reported these
results and
provided guidance on
home welt-testing
corrective action.
Evidence of
endocrine
disruption,
linked to the
presence of
numerous ••
chemical compounds in
Las Vegas Bay, has
emerged from joint
investigations between
NAWQA and the National
Biological Service (now
the Biological Resources
Division of USGS).
do not exceed guidelines established
by the U.S. Fish and Wildlife Service
and Environment Canada* for
protection of aquatic organisms or
fish-eating birds and wildlife.
The number of pesticides
detected in streams draining
the suburbs of Washington,
D.C. is greater than the
number found in streams
draining farmland. In urban
streams, some insecticides
typically exceed
Environment Canada's*
guidelines for protecting
aquatic health. State
pesticide management plans
in the Potomac Basin would
be most effective if both
urban ana farming areas
were included.
As a result of NAWQA
findings ofPCBs in
fish of the Mohawk
River, New York State
is establishing health
advisories for
affected parts
of the River
and will
*r further
assess the
scope ofPCB
occurrence.
Ground water supplies drinking water
to one-half the population oflndianas
White River Basin. Under cropland, 17
percent of wells exceed EPA drinking
water standards for nitrate. In 50
percent of urban wells, VOCs are
detectable, though not in excess ofEIA
drinking water standards.
The VOC methyl ten-butyl ether
(MTBE) is detectable in low
concentrations in 80 percent of urban
ground-water samples from the Denver
area. While the monitoring wells
sampled are not used for drinking
water, the presence of MTBE, which is
used to reduce air pollution from car
exhaust, was unexpected in ground
water.
* In the absence of U.S. EPA standards for aquatic/wildlife protection, guide-
lines from the National Academy of Sciences and Envir onment Canada are
frequently used by scientists to provide a context for evaluation.
A 54-percent decline
in phosphorus loads
carried by the
Chattahoochee River
between 1989 and
1991 is explained, in
part, by an 83-
percent reduction in
phosphorus
discharged to the
River from Atlanta's
wastewater treatment
facilities.
Commonly used herbicides occur in
the highest concentrations in
streams that drain row-crop areas.
During periods of high runoff,
nearly all samples exceed Em
drinking water standards for at least
one herbicide.
Sediment cores collected from White
Rock Lake in Dallas reveal upward
trends in lead, DDT, chlordane and
PCBsfrom 1912 to 1994. Increasing
concentrations of these compounds
have followed increases in human
population. Declining concentrations of
these compounds occurred after their
use was banned, but levels have not yet
fallen to those recorded for 1912.
216
-------
EXPLANATION
Begun In 1991 i^B
ACFB Apalachicola-Chattahoochee-Rint River Basin
ALBE Albemarle-Pamlico Drainage
CCPT Central Columbia Plateau
CNBR Central Nebraska Basin
CONN Connecticut, Housatonic, Thames River Basins
GAR. Georgia-Florida Coastal Plain
HDSN Hudson River Basin
LSUS Lower Susquehanna River Basin
NVBR Nevada Basin and Range
OZRK Ozark Plateaus
POTO Potomac River Basin
REDN Red River of the North Basin
RIOG Rio Grande Valley
SANJ San Joaquin-Tulare Basins
SPLT South Platte River Basin
TRIN Trinity River Basin
USNK Upper Snake River Basin
WILL Willamette Basin
WHIT White River Basin
WMIC Western Lake Michigan Drainages
Begun In 1994 t I
ALMN Allegheny and Monongahela Basins
CAZB Central Arizona Basins
EIWA Eastern Iowa Basins
KAMA Kanawha-New River Basin
LERI Lake Erie-Lake Saint Clair Drainage
LINJ Long Island-New Jersey Coastal Drainages
LIRB Lower Illinois River Basin
MISE Mississippi Embayment
PUGT Puget Sound Basin
SACR Sacramento Basin
SANT Santee Basin and Coastal Drainages
SCTX South Central "fexas
SOFL Southern Rorida
UCOL Upper Colorado River Basin
UMIS Upper Mississippi River Basin
UTEN Upper Tennessee River Basin
Beginning In 1997
ACAD Acadian-Pontchartrain
COOK Cook Inlet Basin
DELR Delaware River Basin
DLMV Delmarva Peninsula
GRSL Great Salt Lake Basins
LTEN Lower Tennessee River Basin
MARK Middle Arkansas River Basin
MIAM Great and Little Miami River Basins
MOBL Mobile River and Tributaries
NECB New England Coastal Basins
NROK Northern Rockies Intermontane Basins
OAHU Oahu
SANA Santa Ana Basin
SHPL Southern High Plains
UIRB Upper Illinois River Basin
YAKI Yakima River Basin
YELL Yellowstone Basin
Unscheduled I I
KNTY Kentucky River Basin
CACI Canadian-Cimarron River Basins
l
Fig. 1 Location of Study Units
Fig. 3 Nationwide patterns of atmospheric deposition of nitrogen
Low but detectable concentrations of many pesticides occur in streams and
ground water, as well as in air, fog, snow and rain nationwide. Most com-
pounds found are used locally, although some pesticides are transported
through the atmosphere for hundreds or thousands of miles. Persistent com-
pounds like DDT and toxaphene are still cycling through the atmosphere in the
United States more than a decade after their use was banned. Although most
pesticide concentrations in the atmosphere are very low, adverse effects on
wildlife, caused by continuous deposition and accumulation in the food chain,
have been documented.
Among 37 VOCs detected in ground-water samples from eight cities,
chloroform was detected most often, followed by the automotive fuel oxygen-
ate MTBE. Although no detections of MTBE exceeded EPA health advisory
levels (HALs), its presence in ground water was unexpected and prompted
OSTP to assess the use of oxygenated fuels for air-pollution control. NAWQA
scientists chaired an interagency working group for OSTP to document the
occurrence of fuel oxygenates in the environment.
217
-------
Get More Information on Water Quality From USGS and NAWQA..
... on the internet via the World Wide Web
The NAWQA homepage is located at
(http://wwwrvares.erusgs.gov/nawqa
J
... from NAWQA Study Units
Each Study Unit forms an advisory or "Liaison
Committee" composed of Federal, State and local officials.
For information about Liaison Committee meetings or to
inquire about participating on a Committee, contact the
NAWQA Study Unit Chief through USGS state offices or
from the study-unit information contained in our web site:
http://wwwrvares.er.usgs.gov/nawqa/nawqamap.html
. . . from USGS State Representatives
USGS maintains an office in every state. To find out
how to reach your state's USGS State Representative,
consult the phone book, our internet list at
C http://h2o.er.usgs.gov/public/wrdO 11 .htmT)
or contact :
Office of Water Information
Water Resources Division
U.S. Geological Survey
440 National Center
Reston,VA 20192
(703) 648-5699
... in technical and general interest
publications
Some USGS documents are free. Others are available
for a nominal cost. All publications can be ordered from:
U.S. Geological Survey
Information Services
Box 25286, Federal Center
Denver, CO 80225
(303) 202-4200
For more information about the NAWQA
Program, contact:
Chief, NAWQA Program
U.S. Geological Survey
413 National Center
Reston,VA 20192
Phone:(703)648-5716
fax: (703) 648-6693
email: nawqa_whq(g usgs.gov)
Information from NAWQA Includes:
< Historical water-quality data from 37 river/aquifer basins
1 New water-quality assessments from 20 of those basins, with
an emphasis on identifying human and natural factors that
affect current conditions
1 New investigations in 16 basins, with findings due in 1998-99
1 Evaluation of historical, nationwide occurrence of pesticides,
nutrients, and volatile organic compounds in streams and
ground water
NAWQA Handout #1
March 1997
218
-------
Detecting Fecal Contamination and Enteric
Microbes in Watersheds
Mark D. Sobsey
University of North Carolina, Chapel Hill, NC 27599-7400
Fecal contamination from human and animal sources continues to pose risks to human and animal
health from the disease-causing microorganisms commonly found in such wastes. Fecal
contamination from agricultural animals must be managed effectively because it can contain a
variety of human pathogens. Such pathogens include the enteric protozoans Cryptosporidium
parvum and Giardia lamblia and enteric bacteria such as Salmonella species, Campylobacter
species and pathogenic strains of Escherichia coli. Our current studies indicate that typical animal
waste management practices, such as lagooning, is sometimes inefficient in reducing some of these
pathogens. Therefore, land application of lagooned animal wastes can result in surface water
contamination if runoff is not adequately controlled or ground water contamination if soils are
coarse, fractured or otherwise capable allowing rapid microbial transport. We have found that
surface waters passing though animal agriculture operations often experience dramatic increases in
fecal contamination from animal wastes, as indicated by excessive levels of indicator bacteria and
viruses. Information on the extent of increase in parasitic pathogens such as Cryptosporidium
parvum oocysts from agricultural animal wastes in inadequate, but there is evidence of increased
levels from animal agriculture operations. Studies on the survival and persistence of the more
resistant microbial pathogens in agriculture animal waste treatment practices are now in progress.
We have found that some pathogens are highly persistent at lower temperatures but are rapidly
inactivated at higher temperatures typical of thermophilic waste treatment processes. Because of
the unreliability and high cost of detecting enteric pathogens in water, improved microbial
indicators of these pathogens are needed. We have found that male-specific coliphages and the
spores of the bacteria Clostridium perfringens have considerable value in tracing and identifying
sources of fecal contamination in water. However, further studies are needed to determine the
reliability of these indicators in predicting the risks from some of the more persistent protozoan
pathogens, such as oocysts of Cryptosporidium parvum and cysts of Giardia lamblia.
219
-------
The Role of Consultants in Addressing Water Quality Issues
Richard S. Fawcett
Farming has become much more complicated than in the past.
More information is needed to make decisions on the use of inputs
like pesticides and nutrients, and use of various tillage and crop
production practices. Integrated Pest Management systems employ
crop scouting to determine pest populations and utilize economic
thresholds to determine when pest management strategies will be
initiated. Modern nutrient management often employs intensified
soil or plant sampling regimes. Yield maps generated using GPS
must be analyzed to determine reasons for yield variability.
Layers of data for each field, including yield, soil tests, soil
productivity ratings, and input records can then be integrated to
guide variable rate input applications and variations in crop
varieties and populations to match site specific conditions. More
farmers are looking to consultants to provide advice and analysis.
Many management changes adopted to improve efficiency and
economics will have a positive impact on water quality. Pesticides
can be matched to pest and soil conditions, avoiding inappropriate
or unnecessary treatments. Ground and surface water vulnerability
can be considered in selecting products and rates. Comprehensive
nutrient management plans will take credit for non-fertilizer
sources of nutrients and match fertilizer rates to crop production
potential of soils. Variable rate application can avoid over
fertilizing lower productivity soils, while providing sufficient
nutrients for better soils to produce top economic yields.
Consultants can help select specific Best Management Practices
(BMPs) to protect ground and surface water. For example, no-till
is an effective surface water BMP on most soils, but herbicide
runoff may not be reduced when this practice is used on claypan
soils. The consultant can help select effective alternative BMP's.
Contamination of farm wells by pesticides, nutrients, and
bacteria has often been traced to point sources near wells, such as
pesticide and fertilizer mixing and disposal sites, septic systems,
feedlots, and manure storage. Consultants can play a role in
testing well water, identifying potential causes of contamination,
and changing management practices to protect the well. Well
construction and maintenance deficiencies can also be identified
and corrected.
Farmers are increasingly considering water quality impacts
when making management decisions. Consultants can help farmers
obtain and interpret the data needed to make those decisions.
Richard S. Fawcett is President of Fawcett Consulting, Huxley IA
50124
221
-------
Twenty Years of Change: The Lake Erie Agricultural Systems for
Environmental Quality (LEASEQ) Project
R. Peter Richards and David B. Baker
Water Quality Lab, Heidelberg College, Tiffin, Ohio
Concern over the "impending death" of Lake Erie in the early 1970's led not only to extensive
point-source improvements but to some of the earliest and largest-scale efforts to implement
agricultural best-management practices in the United States. Between 1975 and 1995, a number
of implementation, education, and demonstration projects were carried out. While all were in
response to the general goal of reducing pollutant export into Lake Erie, they were run by different
managers responsible to different agencies. In a sense, this has been an experiment to see if
environmental progress can flow from a mosaic of loosely-integrated programs in which
participation is largely voluntary.
Maumee R.
Sandusky R.
LEASEQ Monitoring Stations
Figure 1. The Lake Erie drainage basin, showing the location of the study watersheds
The Lake Erie basin has also been the site of an ongoing monitoring program of unusual detail,
carried out by the Water Quality Laboratory at Heidelberg College (HCWQL), which has
produced daily and more frequent observations of sediment and nutrient concentrations in the
223
-------
Maumee River at Bowling Green and the Sandusky River at Fremont beginning in 1975. These
datasets now contain in excess of 8500 records each, and represent a valuable resource for
evaluating the water quality benefits of changing agricultural management practices.
LEASEQ is a USDA-funded retrospective evaluation of changes in land management, soil fertility,
farm economics, and water quality in the Maumee and Sandusky River basins of the Lake Erie
watershed in the 21-year period between 1975 and 1995. Due to space constraints, this paper
will report findings for the Maumee River basin only; results for the Sandusky are similar in most
respects.
Environmental Management
Changes in Point Sources
Early efforts to reduce eutrophication in Lake Erie focused on upgrading sewage treatment plants
in the Lake Erie watershed; these improvements were well underway by the beginning of the
study period. As a result, and because,of the relatively small populations served by sewage
treatment plants located upstream of the monitoring stations, point sources are relatively minor
components of annual nutrient loads, as shown for phosphorus in Rgure 2. Point source data for
nitrogen are not available, but our estimates indicate that point sources contribute an even smaller
portion of the total nitrogen loads than they do for phosphorus, even with the nitrification of
ammonia
4000"
-------
20.0"
4 C f\~
15.O
10.0~
5.0"
250.0"
200.0~
150.0"
100.0"
50.0"
4.0"
3.0"
2.0"
1.0"
0.0^
19
Figure 3. Farm
.6
1.2-
2
8
•5 0.8-
w
_g
1
0.4-
On—
.U
19
*•••..
Number of Farms, in thousands
*
Average Farm Size, acres
Total Farm Land, millions of acres
75 1980 1985 1990 1995
land use in the Maumee Basin, 1975-1995
• •* • -r
-£v^^-V-"~^^^
- n ° n D
•~Tr~D~-D ° ^ n corn 0 0
•°*B o— — a.
D D D
0
.
• • • . •
— _ w
wheat . • . « • f • • * *-
•
i . ... | .... | .... |
75 1980 1985 1990 19$
^^
^M
^^
)5
Figure 4. Trends in cropping patterns in the Maumee Basin, 1975-1995
225
-------
co
2>
o
3.CH
T3
t 2.0'
2
O
S 1.0-
I
Conventional
Mulch Till
No Till
I
I
I
I
I
I
I
I
I
•T*
I
!••
••
r
1982
1984
1986
1988
1990 1992 1994
Figure 5. Conservation tillage adoption in northwest Ohio, 1975-1995
Sales of fertilizer phosphorus peaked about 1980 and have since declined by nearly 50%.
Sales of fertilizer nitrogen peaked in the Maumee basin about 1980 and have since decreased by
about 25% (Figure 6). Phosphorus from manure constitutes 15% to 20% of that from commercial
fertilizer, and nitrogen from manure constitutes 30% to 40% of that from commercial fertilizer. Both
have declined about 20% over the 20 year study pattern, primarily since the early 1980's.
80,000
a. 60,000'
CO
CO
I 40,000'
CO
20,000
I
150,000
100,000-
to
o
5 50,000'
CO
I
i
Phosphorus
I
71 73 75 77 79 '81 '83 '85 '87 '89 '91 '93 '95
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
!
I
I
I
I
I
N
litn
I
3ge
I
n
I
71 73 75 77 79 '81
'83
'85 '87 '89 '91 '93 '95
Figure 6. Commercial fertilizer sales in the Maumee Basin, 1971-1995
226
-------
Changes in the Soil Resource
Unit area loads of sediment, phosphorus and nitrogen (nitrate plus Kjeldahl) monitored at the
Maumee River sampling station at Bowling Green, Ohio, average 614,1.2, and 19.9 Ib/acre/yr,
respectively. Over the 21-year study period, these losses amount to more than 26,000,000 tons
of sediment and 50,000 tons of phosphorus, and nearly 850,000 tons of nitrogen for the
watershed as a whole. Since point sources are a minor component of loads, and 88% of the
basin is in agricultural land uses, most of these loads represent losses of resources from
agricultural lands.
Comparisons of soil cores from the 1950's and 1960's, archived at Ohio State University, with
cores from the same locations taken in the past year, indicate that soil fertility has generally
increased during the study period, in spite of losses to erosion. Phosphorus levels are generally
above those required for crop maintenance, and are believed to still be increasing gradually.
However, nutrient use has become much more efficient. Watershed-wide mass balance
calculations indicate that, in the peak three years of the study period, 1979-1981, about half of the
phosphorus coming into the watershed from all sources was not exported, but contributed to
nutrient buildup. In the final three years of the study, the phosphorus surplus was down to five
percent, a very significant improvement in the efficiency of nutrient use.
Change in Crop Yields
Crop yields for the three major crops, measured on a per-acre basis, have increased during the
study period (Figure 7). Corn yields are up about 25%, wheat 30%, and soybeans 17%.
150.0'
% 100.0-
"(8
]=
"I
.Q
1 50.0-
corn
wheat o
soybeans
1975
1980
1985
1990
1995
Figure 7. Crop yields in the Maumee Basin, 1975-1995
Changes in Water Quality
Water quality changes were evaluated by conducting trend studies of concentrations and loads.
Trend analyses were performed on load and concentration data aggregated at the daily, monthly,
and annual levels as well as on storm event loads and concentrations. Most analyses were
done using log-transformed concentrations or loads, using an ANCOVA model which included the
month (categorical variate) and the log of flow (continuous) as well as time (continuous).
227
-------
The results of representative trend analyses are shown in Figure 8 for discharge, suspended
sediment (SS), total phosphorus (TP), soluble reactive phosphorus (SRP), nitrate + nitrite
(NO3), and total Kjeldahl nitrogen (TKN). Within each group of bars, the leftmost is the trend
based on annual loads, the next is annual flow-weighted mean concentrations (FWMCs), the
third is storm event mean concentrations (SEMCs), the fourth is monthly FWMCs, and the fifth is
daily FWMCs. Annual trends were not analyzed for TKN. The percent change over 21 years
relative to the initial values is shown by the length of the bar, and the statistical significance of the
trend analysis is shown by dots within the bar: none, not significant; •, p<.05; ••, p<.01; •••,
p<.0001.
These analyses show that average discharge has increased in the Maumee River. Very major
decreases have occurred in SRP concentrations, and substantial decreases have occurred in SS,
TP, and TKN concentrations. Substantial increases in NO3 concentrations have also occurred.
While most analyses suggested comparable amounts of change for a given parameter, the level
of statistical significance was generally highest for daily data, and lowest for annual and storm
event data. We were somewhat surprised that daily data performed the best, because of its
tremendous variability. However, the use of log transformations and the use of flow and
seasonaliry (months) as covariates substantially reduce this variability, and the large number of
observations lend statistical strength. Apparently, the benefit of reduced variability afforded by
averaging data over monthly or larger time scales is outweighed by the loss of degrees of
freedom which results from the averaging process.
o>
O)
in
N.
OJ
0}
O)
CO
o
Discharge SS
SRP
NO3
TKN
Figure 8. Trends in nutrient and sediment concentrations in the Maumee Basin, 1975-
1995. Within each group of bars, the leftmost is the trend based on annual flow-
weighted mean concentrations (FWMCs), the next is storm event mean concentrations
(SEMCs), the third is monthly FWMCs based on storm data only, the fourth is monthly
FWMCs based on all data, and the fifth is daily FWMCs. The percent change over 21
years relative to the initial values is shown by the length of the bar, and the statistical
significance of the trend analysis is shown by dots within the bar: none, not
significant; •, p<.10; ••, p<.01; •••, p<.0001.
The relatively poor performance of storm event measures was unexpected, since many BMPs
are designed specifically to stop erosion during storm runoff. Analysis of the trends in data
partitioned into three ranges by flow showed that, for parameters for which the overall trend is
downward, the greatest change occurs during the low flow periods, and the smallest during high
flow periods. NO3 shows upward trends in the high and medium flow ranges, but downward
trend in the low flow range. Flows in the highest range themselves show a downward trend,
those in the middle range show no trend, and those in the lowest range show an upward trend.
228
-------
Similarly, flows during the first storm runoff event in a group of closely spaced events have a
downward trend, while flows in the later events in a cluster have an upward trend.
One reasonable interpretation of these results is that infiltration has increased, and overland flow
has decreased. Conservation tillage is at least partly responsible for this change. Trends in
concentrations at different flow regimes suggest that changes in nutrient use patterns and
management practices which target erosion have both contributed to improvements in water
quality. The largest percent changes are typically associated with low flows, when erosion is not
an active process. However, since concentrations and especially loads are usually higher under
high flow conditions, the largest absolute changes are associated with storm runoff.
Lessons Learned
• Management programs sometimes succeed in ways other than planned. No-till was
initially intended as a corn practice, but most of its implementation in the Lake Erie basin
has been with soybeans.
• Substantial water quality improvements can be accomplished at a large watershed scale
by a mosaic of agricultural management programs, even if optimal targeting is not practiced
and a single oversight agency is lacking. However, 10-20 years may be required before
sufficient change occurs to be statistically significant.
• Water quality improvements need not come at the expense of lowered crop yields.
• Water quality improvements can be readily documented given appropriate statistical
approaches and detailed monitoring data over a long enough period of time.
• At least for this study area, storm event concentrations and loads are less efficient metrics
for revealing trends than traditional daily observations. Stratification of the data by flow
showed that decreases, in percentage terms, were generally greater at lower flows, but
were greater in absolute terms at higher flows.
Acknowledgments
We thank our LEASEQ colleagues Frank Calhoun and Don Eckert for providing data on changes
in soil fertility and agricultural practices, respectively. Data on conservation tillage implementation
were provided by Gary Overmyer of the Natural Resources Conservation Service. Point source
phosphorus load data were provided by the International Joint Commission, Windsor Office, and
the Ohio EPA.
229
-------
Elm Creek HUA: Results and Impact
Chuck Burr
Clay County Cooperative Extension, Clay Center, Nebraska
Setting
Elm Creek is a spring fed stream located in south central Nebraska that
supports a put-and-take trout fishery. High peak flows and subsequent
sedimentation caused by high intensity-short duration thunderstorms are
factors which negatively impact instream water quality and habitat.
Objectives and Implementation
Project objectives include reducing peak flows, in-stream sedimentation,
and maximum summer water temperatures, as well as improving aquatic
habitat. Implementation involves public education and cost-share
assistance to encourage adoption of best management practices.
Evaluation includes tracking adoption of BMPs. To determine if
improvements in water quality and aquatic habitat are being achieved,
weekly grab samples, seasonal fish and macro-invertebrate collections,
and substrate and habitat evaluations are being conducted.
Environmental Benefits Measured
The Natural Resources Conservation Service is tracking acres and total
soil savings that have been protected with structural BMPs. Artificial
habitat improvements involved the installation of seven lunkers to
stabilize streambanks while providing habitat and protection from peak
flows. The AGNPS computer model was run at the beginning of the project
and a follow-up analysis will be used at the conclusion of the project.
Lessons Learned
Improvements in water quality or reductions in peak flow are difficult
at best to measure directly since large runoff events may be masking any
improvements that would exist under normal flow conditions. Also, in a
State like Nebraska where the majority of land is privately owned,
having project support from landowners is critical for success.
231
-------
HUA Crop Management Service Continues As
Farmer-Run Crop Management Association
Sarah Gushing, Bill Jokela, Sid Bosworth
Plant and Soil Science Dept, Hills Bldg., University of Vermont
Burlington, VT 05405
Missisquoi Bay is one of the most eutrophic in Lake Champlain with phosphorus
concentrations as high as those in Lake Erie in the 1970's. A significant portion of the
P load in this dairy production area has been attributed to runoff from agricultural
nonpoint sources such as manure, fertilizer and cropland erosion.
As part of the Lower Missisquoi HUA, we established a Crop Management Service to
develop whole-farm nutrient management plans, which included nutrient, pest, and
other crop management services. The goal was to improve management practices,
especially for manure and fertilizer, to reduce the potential for environmental impact
and to increase the farmers' economic return. Between 10 and 14 farmers enrolled a
total of up to 1700 acres of cropland each year of the program (Table 1).
Table 1. Yearly summary of number of farms and acreage enrolled in the
Missisquoi Crop Management Service. 1991-1994.
Crop
Year # Farms Corn Alfalfa Clover Grass Total
1991
1992
1993
1994
10
12
14
10
511
566
462
478
339
434
287
284
Ufl V«O
44
125
169
252
493
459
812
710
1386
1584
1731
1725
The services provided varied with the crop, as follows:
Nutrient Management Package (all crops)
- Fall soil sampling and analysis
- Manure sampling and analysis
- Nutrient management planning
- Fertilizer recommendations
- Pre-sidedress Nitrate Test for corn (sampling and analysis)
- Computerized crop record keeping
233
-------
Corn Pest Management Package
- Spring weed assessment
- Population count
- Scouting through season of insect pests
- Fall weed assessment
- Pest management planning and recommendations
Alfalfa Management Package
- Winter injury assessment
- Spring crown count
- Scouting through season of insect pests
- Fall Crown Count
- Pest management planning and recommendations
Grass, Clover and Pasture Management
- Species assessment
The Vermont computerized crop recordkeeping system was used by the farmers to
monitor crop inputs, manure applications, yields, weeds and pest problems, costs and
returns. Participants received an Input Booklet at the beginning of each cropping
season for recording information and returned it at the end of the season for entry into
a computer database. Computer-generated summary reports were made for each farm
and included comprehensive data by field and by crop. These summaries were then
used by the farmer and the ICM consultant to make management decisions for the next
cropping season.
Nutrient Management Planning
A key element of the Crop Management Services was the development of nutrient
recommendations for individual fields on each farm. The ideal is to have soils that have
adequate nutrient levels to support good crop production but not excessive levels that
have a greater potential for adverse impact on water quality. To these ends, each field
on each of the cooperating farms was soil sampled annually, usually in the fall, for
routine soil testing of pH, available and reserve P, K, Mg and Al. Corn fields were
sampled for nitrate in the summer, as well. Soil test results were combined with crop
and soil information and manure analysis and management information to make
nutrient recommendations using the Vermont Manure Nutrient Manager, a Lotus'!23 !
spreadsheet template. \
i
Soil test results varied greatly on individual fields. For example, in 1993 over half (56%) ,
of the fields were medium to optimum in available phosphorous, about a fourth (24%) !
were low to very low, and the remaining 20% were high to excessive. Almost all farms >
have some fields higher than optimum (high or excessive), which means they need no j
i
!
i
234 . !
-------
additional P for crop production and pose a potential problem for surface water quality.
But most farms also nave some fields with lower than optimum P soil test levels,
indicating a need for additional manure or fertilizer P to maintain crop yields. Results of
the Pre-sidedress Nitrate Soil Test (PSNT) for corn showed a large range in nitrate test
levels among individual fields, as well as differences in overall average values for
different years. Several cooperators saved N fertilizer expense and reduced the
potential for nitrate pollution of surface and groundwater through use of the PSNT.
These results point out the critical importance of soil testing to indicate those fields that
need little or no additional fertilizer to maintain top yields and those that require
additional amounts to maintain top productivity.
Manure sampling was done, as part of the nutrient management package, to assess the
nutrient contribution from manure and to avoid excessive applications. Manure was
sampled at least annually on all farms on the program; either by the farmer or by the
project technician. Considerable variation occurred among farms, as much as a two-
fold difference in nutrient content of manure from different farms, which points out the
importance of basing recommendations on nutrient analysis rather than on standard
estimates. This is particularly important in watersheds such as the Missisquoi where
two-thirds of the phosphorus applied to cropland comes from manure.
Recommended vs. Applied Nutrients: Environmental and Economic Impacts
To evaluate how well farmers were following our manure and fertilizer
recommendations we used information from the crop record keeping system to
compare rates actually applied to those recommended. In 1993 application rates of
phosphorus were within 20 Ibs P2Os/acre of the recommended rate on about 75 percent
of all crop acreage (Fig. 1). Nitrogen fertilizer was applied to corn at rates
recommended by the PSNT (+/- 20 Ib/acre) on 80 percent of the acreage, the average
rate being within 5 Ib/acre of the recommended rate. This appears to indicate fairly
good acceptance of a nitrogen test that, on average, recommends less than
conventional recommendation methods.
Following recommendations gave positive economic and environmental results for
participating farmers. Overall, there was a significant reduction in the use of
commercial fertilizer after the farms enrolled in the crop management service, based on
a study of seven farms in the project (Knight et al., 1997. See these proceedings).
Phosphorus use decreased by an average of 40 percent and potassium by 29 percent
over a 3-year period. Also, farmers changed their timing of nitrogen application in ways
that increase the availability to the crop and decrease losses to the environment. The
farms in the study reduced total expenses by an average of $277 acre, while crop yields
remained constant.
235
-------
Applied vs. Recommended P & N Fertilizer Rates
1993
Corn
Fig. 1. Difference between applied and recommended P and N application rates. 1993.
Transition to a Farmer-owned Crop Management Association
The Crop Management Services Program was designed to be a pilot program that
would eventually become self-sustaining so that it would continue beyond the funding
period of the HUA. In an attempt to accomplish that goal, ICM services were offered
free of charge in the first year to encourage participation without financial risk, but in
successive years farmers were required to pay an increasing share of the cost of the
services (Table 2). The intent was to have a self-supporting program by the end of the
5-year funding period, whether by private consultants, by a crop management
association, or by farmers providing their own crop management services. That goal
became a reality in the 1995 growing season with formation of the Missisquoi Crop
Management Association, a farmer owned and operated crop management association.
The Association offers the same services as were offered by the HUA crop service.
However, the ability to pick and choose services has been greatly increased to allow
members to pay only for services they feel they need. All costs for services are based
on the "actual cost of doing business".
During the 1996 season the CMA had 12 members and over 2000 acres enrolled (800
acres of corn, primarily silage, and over 1200 acres of legume and grass hay). A
complete nutrient management program was done on 152 fields, including most of the
236
-------
crop acreage in the program. Scouting for a variety of pests was carried out on 55 corn
fields (656 acres) and 14 alfalfa fields (170 acres). Corn scouting included fall and
spring weed assessments, plant population counts, corn rootworm counts, and/or a
complete weed management package. Alfalfa scouting options were fall and spring
crown counts and scouting for potato leafhopper and alfalfa weevil. Computerized crop
record keeping was carried out on ten farms (168 fields).
Table 2. Yearly pricing structure for the Missisquoi Crop Management Services.
_^ Crop
Year Com Alfalfa Clover Grass Hay
Cost, $ per acre
1991
1992 2
1993 4
1994
Complete 6
Nutrient only 4
Summary
• Nineteen farms have participat
No Charge
2
4
6
3
ed in the Missisquoi Crop
1
2
3
3
Manage
1
2
3
3
ment Services
since 1991. The Missisquoi Crop Management Association, a farmer owned and
managed independent organization, grew out of the Crop Management Service and
now has twelve members and over 2000 acres enrolled.
• Soil test levels varied greatly among fields (low to excessive) and manure nutrient
content showed two-fold differences among farms, emphasizing the importance of
soil testing and manure analysis for field-specific nutrient management plans.
• Farmers enrolled in the Crop Management Services program generally followed the
nutrient recommendations made by the project technician. This resulted in nutrient
applications that reduced loading of phosphorus to fields and decreased fertilizer
costs to the farmers, contributing to both water quality and economic goals of the
project.
237
-------
On-Farm Field Trials Demonstrate and Quantify the
Environmental and Economic Benefit of BMPs.
Bill Jokela, Sid Bosworth, Jeff Tricou, Sarah Gushing
Plant and Soil Science Dept, Hills Bldg., Univ. of Vermont, Burlington, VT 05405
Missisquoi Bay is one of the most eutrophic in Lake Champlain with phosphorus
concentrations as high as those in Lake Erie in the 1970's. A significant portion
of the P load in this dairy production area has been attributed to runoff from
agricultural nonpoint sources such as manure, fertilizer and cropland erosion.
As part of the Lower Missisquoi HUA, we conducted a number of on-farm trials,
in most cases replicated, to demonstrate best management practices we were
encouraging farmers to adopt and to quantify the economic and potential
environmental effect of the practices. Other demonstration trials were done as
part of an RCWP project in the same county.
All demonstrations were done on farms participating in the crop management
services or association. Most field operations were carried out by the farmers as
part of their normal activities so that conditions would represent those on
Vermont dairy farms. Field sites were included in tours and field days for viewing
by farmers and other ag professionals; the data from the trials has been used at
many educational meetings and included in written materials and mass media
presentations. Selected results include the following:
• Top yields of alfalfa-grass on soils testing "optimum" or higher in phosphorus
can be maintained with application of potash only, thus avoiding economic,
and potentially environmental, costs associated with applying unnecessary
phosphorus fertilizer (e.g. a typical 0-10-40 or 0-10-30 blend). Avoiding P-
containing fertilizer blends on these soils showed reductions in P application
of as much as 60 Ib PaOs per acre and potential cost savings to the farmer of
$10 to $15 per acre.
• Application of either liquid dairy manure or NPK fertilizer doubled yields of
mixed grass-broadleaf hay that had been historically under managed.
Reliance solely on manure to meet crop nitrogen needs, however, provided
excess phosphorus and raised the P soil test into the high category.
Concentrations of P in runoff were highest from the manured plots, reflecting
higher P loading and resultant soil test levels. A combination of lower manure
rates and N fertilizer is probably the best recommendation to optimize use of
resources and reduce adverse water quality impact.
239
-------
• Because of poor nitrogen utilization, semi-solid dairy manure was not able to
meet the nitrogen needs of orchardgrass hay, even when applied in multiple
applications at high rates (about 30 tons/acre/year). However, the
combination of N fertilizer and manure provided excellent yields and forage
quality. Phosphorus removal was primarily a function of crop yield; therefore,
fertilizer N was an important tool in optimizing a P balance (by increasing
yield without adding P). A moderate rate of manure (approximately 10
tons/year) combined with fertilizer N provided adequate P for crop needs
while reducing adverse water quality impact.
• Dairy manure applied in the fall or spring with methods that provide
immediate incorporation (sweep injectors, s-tine/field cultivator) produced
higher corn silage yields than manure surface-applied in the fall. '
Incorporation minimized ammonia N losses and protected manure nutrients
from loss in runoff. Slightly higher yields were obtained with sidedressed
fertilizer N, presumably because the timing avoided the high rainfall periods in
the spring and early summer when N loss was highest.
i
• The pre-sidedress nitrate test (PSNT) reflected the differences in N
availability shown in yields from different manure application methods,
reinforcing its value as a tool to assess N fertilizer need. Several cooperators
saved N fertilizer expense and reduced the potential for nitrate pollution of
surface and groundwater through use of the PSNT.
• Ryegrass or clover covercrops interseeded at last cultivation can be
successfully established in silage corn, providing erosion and runoff
protection at a relatively low cost. >
• Starter fertilizer rates half of those typically used (e.g. 25 vs 50 Ib P2Os/acre) '
consistently produced top corn silage yields, thus saving farmers $10/acre
and reducing P loading and its potential runoff effects. (From St. Albans Bay j
RCWP)
• Well-managed dairy manure applied at a typical rate (20 ton/acre) supplied
the equivalent of 100 Ib N/acre as fertilizer, eliminating the need for all but a
low rate of planter-applied starter fertilizer. The pre-sidedress nitrate test
(PSNT) reflected this increased N availability, supporting its use as '
management tool to avoid excessive N application. (From St. Albans Bay
RCWP)
240
-------
Arizona's Water Quality and Hydrologic Unit Area Projects
to Promote BMPs
Dr. Jack Watson
University of Arizona, Maricopa Agricultural Center, Maricopa, Arizona
This presentation centers on Arizona's experience with projects to minimize losses of fertilizer
nitrogen below the root zone of cotton. Environmental benefits were inferred by the
documentation of changes in production practices that impact groundwater quality. The primary
production practice change measured was use of split applications of nitrogen fertilizer. This
practice provides the single most important practice change to protect groundwater from nitrate
nitrogen leaching losses, for growers who have already minimized irrigation water applications
and total nitrogen applications. This is due to the fact that those who do not split their
applications of nitrogen, usually apply it prior to planting and early season irrigations. For
agricultural production systems using surface application of water, the initial 2 or 3 irrigations
are the least efficient, resulting in the greatest likelihood of water and nitrate leaching losses
below the root zone.
Demonstration plots indicated that the most serious losses of nitrogen occurred when animal
manure was used in the cotton production system. For growers who changed from one time
nitrogen fertilizer applications to split applications, reductions in nitrogen losses ranged from
estimates of approximately 10 pounds per acre, to almost 100 pounds per acre, in one case.
An increase in grower use of split nitrogen fertilizer applications occurred over the first 3 year
period of the projects, as documented by surveys conducted by the Arizona Agricultural
Statistics Service. The percentage of growers applying fertilizer only at the beginning of the
growing season decreased from 12% to 3% from 1991 through 1993.
Additional information obtained from the surveys indicated that the statewide application rates
of nitrogen per acre to cotton decreased during the first 3 years of the project, from 170 to 149
pounds per acre. However, survey data indicated that nitrogen application rates increased to
223 pounds per acre in 1994 and 194 pounds per acre in 1995. A regression analysis indicated
that approximately two-thirds of the variation in N fertilizer applied in any given year could be
explained by the price received for upland cotton in that year. Secondly, yields in 1993 and
1994 were greater than in 1992, likely encouraging growers to plan greater N applications in
subsequent years (1994 and 1995) to meet the crop N needs due to greater expected yields.
Additionally, crop growing conditions in 1995 resulted in a statewide trend to extend the growing
season, requiring the addition of more nitrogen fertilizer late in the growing season, even
though the actual crop yields were lower than they were in 1993 and 1994.
Practice changes related to reductions in nitrogen losses to groundwater can be documented,
particularly the reduction in the percentage of growers applying nitrogen only at the beginning of
the season. Cooperators in the Hydrologic Unit Areas were interested in protecting water
supplies and reducing on-farm costs, and implemented selected Best Management Practices.
241
-------
Accomplishments and Challenges of Data Collection and Modeling in
North Carolina Demonstration Watershed Project
Casson Stallings, Siamak Khorram
Forestry and the Computer Graphics Center, North Carolina State University, Raleigh, NC
Steve Coffey, Mark Rice, Greg Jennings, Frank Humenik, Rodney L Huffman
Biological and Agricultural Engineering, North Carolina State University, Raleigh, NC
Ken Stone
Agricultural Research Service, Florence, SC
Kris Matson, Rich Mclaughlin, Deanna Osmond, Maurice Cook
Soil Science, North Carolina State University, Raleigh, NC
Narayan Rajbhandari
Plant and Soil Science, Alabama A&M University, Normal AL
This report describes some of the water quality modeling accomplishments and the lessons
learned on the North Carolina Demonstration Project. In 1990 the Herrings Marsh Run
Watershed in the Southeastern coastal was chosen as the site of a Demonstration Project. The
overall goal of the demonstration project was to encourage the accelerated voluntary adoption
of management practices and technologies that measurably improve the surface and ground
water quality. The 5000 acre watershed is in Duplin County North Carolina, an area of intense
and diversified cropping with increasing livestock production. There are about 120 farms and
350 fields in the watershed. This is an environmentally sensitive area where some of the
streams are currently suffering water quality impairments. A wide variety of best management
practices (BMPs) have been applied on the watershed as part of the project. See the papers by
Humenik and Cook, both in this volume, for additional project information.
The modeling component of this project is meant to support the overall goal by evaluating
individuals BMPs and specific models at the field and watershed scale. Additionally, we would
like to use watershed scale models to assess the application of BMPs over broader regions.
A variety of data was collected to support the modeling and other components of the project.
Farmer surveys were taken every year to identify the field-specific cropping and management
practices, and the number and management of livestock. Cropping and management
information on the farmer surveys have been augmented by other data sources such and the
US Department of Agriculture (USDA) Farm services Agency (FSA) and Natural Resources
Conservation Service (NRCS) databases. Water quantity and quality data have been collected
by a variety of agencies including the Cooperative Extension Service (CES), the USDA
Agricultural Research Service (ARS), and the US Geological Survey (USGS). Climate data was
collected from an pn-site weather station and augmented with local weather station data.
Spatial data describing soils, topography, hydrology, and non-agricultural land cover was
collected from a variety of sources. The agricultural field boundaries were digitized from FSA
aerial photography. Most of this data has been incorporated into a geographic information
system (GIS) to facilitate data queries and modeling. With respect to modeling, this data has
been used to calibrate or validate GLEAMS, EPIC, and SWAT under a variety of situations, to
classify remotely sensed imagery, and to evaluate BMPs at the field scale.
243
-------
Modeling Accomplishments
Simulating a Swine Waste Field With GLEAMS
The GLEAMS model was used to simulate an overloaded spray field and to investigate the
spray field's recovery as application rates were reduced. Initial monitoring at the spray field
found nitrate-N > 80 mg/L in the shallow ground water and nitrate-N > 8 mg/L in an adjacent
stream. The GLEAMS model was used to simulate the observed high initial ground water
nitrate-N concentration simulating the high loading rates at this site. BMPs were implemented
at this site; the spray field was expanded and the nutrient application rates reduced. Ground
water nitrate-N concentration reduced approximately 40% in three years after the BMPs were
implemented. The GLEAMS model simulated this recovery and reduction of the ground water
nitrate-N concentration with absolute errors of less than 18 mg/L
Simulating Changes in Leaching Due to Reduced Pesticide Application
EPIC was used to simulate the effect of nitrogen (N) and pesticide application rate on pollutant
leaching for several cropping scenarios. For example, one study included a conventional com
production scenario with the same sidedress N application rate for each soil (160 Ibs/acre N)
and an alternative scenario where the sidedress N rate was varied to reflect crop production
and plant uptake capabilities for each soil (55 to 149 Ibs/acre N). Results of a 30-year, daily
time step simulation (Table 1) show that leaching can be reduced by avoiding over application
of N. Conventional and alternative pesticide simulation results show that reducing pesticide
application rate, reduces leaching for the Norfolk soil (Table 2).
Table 1. Shows simulated nitrate leaching for com produced with mineral fertilizers.
Soil Conventional Alternative %Change in Loading
Ibs/acre (ppm) Ibs/acre (ppm) dimensionless
Autryville 28(4) 25(4) 11
Blanton 81(11) 75(10) 7
Foreston 16(6) 12(4) 25
Goldsboro 62(17) 57(15) 8
Norfolk 36(11) 29(9) 19
Rains 43(21) 42(20) 2
Assessing the Leaching Potential of Soil/Pesticide Combinations
EPIC was used to run 30 year simulations for combinations of 17 soils, 12 pesticides, and four
crops typically occurring in the Herrings Marsh Run watershed. The model was quite sensitive
to changes in the pesticide parameters for adsorption and degradation, indicating the need to
provide values that reflect actual field conditions or present results as a range of values.
Different soils also presented a wide range of predicted leaching losses. The results are being
compared to a simple index of soil and pesticide leaching potential as well as actual ground
water detections in order to determine the most efficient and effective approach to predicting
ground water sensitivity to pesticide contamination. Once this evaluation is completed, the
resulting ground water vulnerability assessment will be incorporated into the WATERSHEDSS
online decision support system.
244
-------
Table 2. Simulated pesticide leaching for corn production.
Pesticide Conventional
Alternative
Brand Name
Bleep 6L
Furadan 4F
Lasso 4EC
Aatrex 4L
Lariat 4F
Common
Name
MetolachlorAt
razine
Carbofuran
Alachlor
Atrazine
Alachlor
Atrazine
active
ingredient
apptic. rate
Ibs/acre
1.60
1.20
5.84
3.00
1.50
2.50
1.50
Leachate
oz/acre (ppt]
0.142 (3000)
0.086 (2000)
0.557(11000)
0(24)
0.099 (2000)
0(17)
0.099 (2000)
active
ingredient
applic. rate
Ibs/acre
1.28
0.97
4.49
2.00
1.00
1.56
0.938
Leachate oz/ac
(PPt)
0.114(2000)
0.071 (1000)
0.428 (9000)
0(20)
0.071 (1000)
0(16)
0.057 (1000)
Simulating Changes in Water Quality due to Changes in Riparian Buffer Widths
The Soil and Water Assessment Tool (SWAT) was used to assess the impact of varying
riparian buffer widths on nitrogen load. Digital elevation models were used to divide part of the
watershed into subbasins. Approximate land use was derived from farmer survey data. After
calibration of flow and nutrient loading different subbasin sizes were evaluated. The watershed
was broken up into 3,11 21, and 81 subbasins; 11 subbasins gave the best results.
Hypothetical riparian areas of 0 to 60 meters in width were added around all streams and the
simulations were run again on this new data. The simulation results indicated that increasing
the riparian areas would decrease the nitrogen load. Two caveats should be noted: 1) the
nitrogen loads could not be calibrated well given the data set being used at that time, and 2) the
SWAT model has not been validated for riparian areas. We are working to overcome both of
these limitations.
Deve/opment of Spatial Filters to Incorporate Remotely Sensed Land Cover Imagery
A Landsat Thematic Mapper image was classified using image processing techniques. The
accuracy of the classified land cover data was quantified using overall and class specific
accuracy measures. In addition to the typical rectangular mode (or majority) filter, a polygon
mode filter (also called an object filter) was applied to the classified data. After the application of
the filters the overall land cover accuracy was measured again. Whereas the rectangular filter
used a kernel of fixed size the polygon filter uses a kernel determined by a priori knowledge of
homogenous regions (e.g., fields). Each of these filters reduces the 1salt-and-pepper? or speckle
associated with digitally classified satellite data.
The rectangular mode filters improved the classification accuracy of the land cover data by 4%
to 6%, but decreased the class specific accuracy in the classes containing many small patches.
The polygon mode filter improved the overall accuracy by 14% to 16% and the class specific
accuracy by up to 28%. The polygon mode filter also reduced the heterogeneity of the classified
image so that only one crop was represented for each farm field.
The application of the spatial filters is an important step in incorporating digitally classified
remotely-sensed data into a GIS. The rectangular mode filters while good at reducing image
heterogeneity can destroy small patches. The polygon mode filter improves classification
accuracy better than the rectangular filter and does not destroy small patches, however, it
requires a priori knowledge of the homogenous areas.
245
-------
Lessons Learned
Lack of Response on Farmer Surveys
Initially, it was planned to gather all land use, livestock, and management data via farmer
surveys. Before any surveys were collected, the project and its goals were advertised heavily
within the watershed and the local area. The goal in advertising was to make sure each farmer
heard about the project at least four times before they were contacted personally to complete a
survey. Then, farmers were contacted on the phone or personally by a Cooperative Extension
Service (CES) technician, who explained the project and the reason for the survey, and
distributed an educational brochure.
The number of farmers willing to complete the survey was rather low. The initial survey in 1990
?ot responses covering 40% of the fields. Responses covered only 17% of the fields during
991 to 1993, and 31% of the fields in 1994. The decrease after 1990 could have been due to
the length and time required fill out the first survey tool. The increase in 1994 was due to the
efforts of a new technician trained in animal science.
Recommendations: Keep the surveys as short as possible. Farmer surveys are an excellent
tool to get very specific information for field scale studies. However, plan to use some alternate
sources of information on cropping and management data for watershed studies.
The Reliability of Farmer Survey Information Varied
The reliability of information obtained from the farmers varied considerably. Because some of
the farmers relied on memory to answer questions, their information was not as reliable as
those who kept records. Some farmers were unsure about the crop, tillage, and nutrient or
pesticide applications on a particular field.
Farmers relying on memory were sometimes not able to give basic information for the field.
More often they knew what crop was grown, but could not remember information such as the
planting and harvesting date, and the levels of chemicals applied. This situation was
exacerbated on occasions when the surveys were given to the farmers more than a year after
the end of a growing season.
Recommendations: Identify critical information for each application and try to ensure it is
collected. Use alternate data sources to simplify the farmer surveys (see later discussion).
Survey the farmers soon after harvest.
Additional Sources of Cropping and BMP Implementation Practices are Valuable
As it became clear that the farmer surveys were not going to meet all of the data needs of this
project additional sources of information were utilized. The simplest data to collect were
"windshield surveys". Technicians visually inspected the fields in the watershed twice a year
determining the summer and winter crops on each field. This provided cropping data for every
field in the watershed.
Data from the NRCS was obtained to indicate where BMPs were being implemented. These
data were helpful in determining how some of the farmers were managing their fields. The FSA
also keeps records on summer and winter crops planted on each field. This provided field
specific cropping information for the five previous years. However, when farmers split cropped
the fields this data did not indicate where on the field the each crop was planted.
Recommendations: Use alternate information sources to acquire data or fill in missing data and
246
-------
where possible to simplify the farmer surveys.
Complex Simulation Models are Extremely Data Intensive
Water quality models designed for agricultural areas are fairly complex. The data requirements,
however, can be very intense since site-specific data on cropping and management, climate,
sub-surface variables, and stream reaches must be supplied. Recently, for the purposes of
simulation, this watershed is split into 100 to 500 subwatersheds and contains as many stream
reaches, and about 10 ponds. Detailed descriptions for each of these entities are necessary.
Additionally, multiple years of cropping and management data must be supplied for model
calibration. A calibration run in this situation can require that over 100,000 values be specified.
Several tools are making the job of specifying simulation values easier. Notably windows-based
database interfaces and GIS interfaces can greatly simplify the task. However, it is still nearly
impossible to acquire excellent estimates for all of the site-specific and multi-temporal values
needed.
Recommendation: Identify the most important input variables and concentrate on getting good
estimates for those variables. Estimate the remaining variables as well as possible.
Spatial Generalization of Soils Data Introduces High Variances in Model Results
Physical environmental data; particularly soils, slope, and agricultural fields; describing the
watershed were integrated into a GIS. A GIS-based interface to GLEAMS was used to run
GLEAMS on randomly chosen fields. As a baseline, the original detailed soils and slope data
were used. The results of these simulations were compared to simulations where the slope or
soils data were generalized to the dominant class within each field.
Comparison of these simulation results showed that field scale model results can vary by up to
100% when soils data is generalized within an individual field. This did not account for natural
variation in soil properties. Variation due to the generalization of slope was much less, although
this is a topographically mild area.
Recommendation: Researchers should be very careful about simplifying soils data to be used in
simulation models as it can introduce high variances in some output variables. These variances
should not be ignored. Rather the variances should either be avoided or taken into account
when using modeling results to simulate the effects of BMPs in the landscape.
247
-------
Phosphate Concentrations in Subsurface Drainage Effluent
In East-central Illinois
G. F. Mclsaac, J. Kent Mitchell, and S. E. Walker
University of Illinois, Urbana-Champaign, Illinois
Abstract
The objective of this study was to quantify the movement of dissolved, molybdate-
reactive phosphate (DMRP) to subsurface perforated drainage tubes (tiles) in fine
textured soils with a variety of crop management systems commonly used in the
Midwestern-US. Subsurface tile flow and DMRP concentrations from seven fields with
various crop management practices have been monitored for 22 months and an eighth
field has been monitored for 16 months. DMRP concentrations were also determined at
various locations along the mainstream of the watershed. Concentrations of Bray P-1 in
the top 10 cm of the soil were also measured. Flow weighted mean DMRP
concentrations in the tile drain effluent ranged from 0.05 to 0.13 mg-P/L There was little
apparent relation between Bray P-1 concentration in the surface layers of soil and the
concentrations in the tile drain effluent. Concentrations of DMRP in the river averaged
0.12 mg-P/L.
Introduction
In the 19th century, much of the land in the Midwestern US was converted from prairie to
row-crop agriculture due to favorable soils, climate and social institutions (Bouge, 1963).
Many areas were swampy and water was removed by the construction of drainage
ditches and installation of subsurface drain tubes commonly referred to as tiles (Bouge,
1951). Presently, 90% or more of the land in many Midwestern watersheds is used for row
crop production. Chemical fertilizers and pesticides are used and a significant portion of
fields are underlain with tiles. The movement of agrichemicqls to drain tiles, ditches and
streams may significantly impair down-stream water quality depending on the rate of
chemical movement.
There have been many studies of the movement of nitrate nitrogen to tile drain effluent
and drinking water supplies in this setting (Logan et al., 1980). Studies of phosphorus
transport have been less common. Compared to nitrate, phosphorus is much less
mobile in the soil and does not present a direct risk to human health. Nonetheless,
249
-------
phosphorus is an important water quality constituent because it influences the growth of
algae and the process of eutrophication (Sharpley et al., 1994). Eutrophication can
reduce the quality of water for recreation and wildlife and increase the cost of treating
water for human consumption. In Illinois, more than 93% of the lakes and reservoirs were
classified as either eutrophic or hyper-eutrophic (USEPA, 1992). Total phosphorus
concentrations in Illinois streams average 0.41 mg -P/L and have been increasing in
recent years (Ramamurthy, 1994). Total phosphorus concentrations in excess of 0.05 mg-
P/L are thought to promote eutrophication.
The objective of this study was to quantify the concentration and loads of dissolved
molybdate-reactive phosphate (DMRP) in drain tiles in fine textured soils with a variety of
crop management practices. This is part of a larger study in the Little Vermilion River
Watershed in East-Central Illinois examining nitrate, phosphate and pesticide transport
from several fields where different tillage, fertilizer and pesticide management systems
and conservation practices are employed.
PROCEDURES
The Setting
The Little Vermilion River Watershed is located in East Central Illinois, where average
annual rainfall is approximately 1000 mm. The soils of the watershed are mostly flat, dark
prairie soils with poor internal drainage. Approximately 90% of the watershed is used for
row crop production, primarily com rotated with soybeans. The upper reaches of the
Little Vermilion River are man-made drainage ditches created in the late 19th century
when the area was first drained. The River was impounded at the village of
Georgetown in 1936 to create the 46 acre Georgetown Reservoir which serves as a
drinking water supply to Georgetown and the village of Olivet. The reservoir periodically
has high levels of nitrate, suspended solids, and atrazine, low concentrations of dissolved
oxygen, and an unpleasant taste and odor. The 48,900 ha watershed above the
reservoir is the study area.
In 1991, seven water quality sampling sites were established along the Little Vermilion
River, including the Georgetown Reservoir. Water samples were collected manually at
two week intervals and following major rainfall events. At one of these sites, designated
R05, a stream flow gage and automatic water sampler were also installed. Starting in
early 1994, water samples were analyzed for molybdate-reactive phosphate
concentrations using a Technicon AutoAnalyzer II.
Field Site Monitoring and Management Systems
With the assistance of Natural Resource Conservation Service and Cooperative Extension
Service personnel, eight privately owned, small tile drainage systems were identified for
250
-------
monitoring. The soils and crop management systems employed in these fields are
representative of the range of conditions and cropping systems in the Little Vermilion
River watershed. The areas drained by the monitored systems range in size from 2 to 15
hectares. At each site, a Palmer-Bowlus flow measuring flume was installed in the tile
line. Water depth in each flume was recorded continuously with a stage recorder.
Depth was converted to flow using the flume depth-discharge equation.
Water samples for chemical analysis were collected on the same schedule as the river
stations. Automatic flow monitoring and sampling equipment was installed at 7 field sites
in 1992 and 1993 and at an eight in the summer of 1994. DMRP analysis of samples
began in March, 1994. Samples taken with the automatic samplers were usually
retrieved within 24 hours of the sampling event. Samples were refrigerated at <8C until
chemical analysis could be performed. Dissolved molybdate-reactive phosphate
concentration was determined using a Technicon AutoAnalyzer II. Mass of DMRP load
exiting the tile drain was estimated by assuming that the measured concentrations
represented actual concentrations for variable periods of time before and after
sampling so that a concentration could be assigned to each recorded flow rate. Flow
weighted mean concentrations were also calculated by dividing the cumulative DMRP
load by the cumulative drainage volume.
Crop and fertilizer management decisions in the monitored fields were made by the land
owners. Fertilizer application rates, methods and timing, and crop yields were
communicated to the research team. Seven of the sites are currently used for corn and
soybean production and one site is maintained in permanent grass as part of the
Conservation Reserve Program. Soil types and phosphorus management practices for
these sites are presented in Table 1.
Table 1. Soils, Cropping System, Phosphorus Application Method and Mean Phosphorus
Applications, kg-P/ha-yr.
Soils
Drummer silt loam
& Flanagan silty
clay loam
Site ID
A
B
C
D
E
Cropping System
Reduced-till Corn-
Beans
Reduced-till Beans-
Corn
Seed Corn - Beans
oats-corn-soybeans
Conv.-till Beans-Corn
Phosphorus
Application
Timing
every other fall
each fall - VRP
Fall & winter
Spring
Fall
Mean
Annual P
Application
Rate
(kg P/ha-yr)
22
25
28
28
11
251
-------
Sabina &
Xenia silt loams
Sabina & Birkbeck
silt loams
F
G
H
No-Till Corn-Beans
Continuous Meadow
Soybean-Com-
Com Silage
Spring
None
Winter & Spring
(manure)
19
None
20
1 Variable Rate Technology
Bray P-1 in the top 10 cm of soil was measured after planting and harvest for each
growing season. Samples were collected at four locations at each site. Statistical
differences in mean Bray P-1 concentrations were identified using the test for Least
Significant Difference (SAS).
Five of the sites are located in predominantly Drummer and Flanagan soils. These soils
are among the most highly productive in the state. They were developed under native
prairie vegetation, and typically have 5% organic matter and 70% silt. Two of the
Drummer-Flanagan sites, designated as A and B in Table 1, are chisel plowed and disked
or field cultivated each year. In a given year, one of these sites will be planted to
soybeans and the other planted to corn. At A, 44 kg-P/ha is applied every other year.
At B, P fertilizer is applied each year using variable rate technology (VRT).
At another site on Drummer Flanagan soils, C, seed corn is produced in rotation with
soybeans. At this site, tillage includes chisel plowing and field cultivation. Approximately
56 kg P/ha is applied every other year in the fall after soybean harvest. Site D is similarly
managed. From 1991 to 1994 it had been used to grow oats and corn in rotation. In
1995, soybeans were grown. The tile monitoring equipment at site D was installed during
the 1994 growing season and tile flow was not recorded until August of that year. Hence
the period of record for this site is 6 months shorter than the other sites in this study.
The fifth site with Drummer-Flanagan soils, E, is moldboard plowed and field cultivated
following corn production, and field cultivated following soybeans. Phosphorus fertilizer
is applied every other year in the fall after soybean harvest at an average rate of 22 kg-
P/ha.
Two sites are located in predominantly Sabina and Xenia silt loam soils. These soils are
somewhat less productive than the Drummer-Flannagan soils. They were developed
under deciduous forest, and typically have 2% organic matter in the top 20 cm. One of
these sites, F, is used for corn-soybean production using no-till. Fertilizer is applied as a
side-dress during the growing season. The other site, G, is in permanent grass cover and
no fertilizer is applied.
Finally, one site is located in predominantly Sabina-Birkbeck silt loam soils. These soils
were also developed under forest vegetation, have approximatey 2% organic matter in
the surface horizon and are somewhat less productive than the Drummer-Flanagan soils.
At this site, a three year rotation of soybean-corn-com is produced using chisel plowing
and disking. The second year of corn is harvested for silage. The only phosphorus
applied from 1990 through 1995 has been in the form of cattle manure. Phosphorus
content was estimated to be 1.85 kg-P/Mg of raw manure (Midwest Plan Service, 1975).
252
-------
RESULTS
Mean soil Bray P-l concentrations, tile flow, DMRP loads and concentrations appear in
Table 2. Soil Bray P-1 was significantly greater at sites D and H than any other site. Site H
has received treatments of 20 Mg/ha cattle manure for many years. There is evidence
from historical aerial photos that a portion of site D was a feedlot in the 1940s. The least
soil P-l concentration was observed at site G, which has been in meadow and has not
received any addition of P fertilizer since 1990.
Tile flow ranged from 81 mm/yr for site D to 159 mm/yr for site G. For most sites, flow
tended to be flashy with most of the flow occurring during a few weeks of the year. The
DMRP loads ranged from 0.051 kg-P/ha-yr for site D to 0.202 Kg-P/ha-yr for site F. These
values are generally less than 1 % of the P fertilizer application rates and consequently do
not represent a significant loss from a crop production perspective.
Table 2. Subsurface tile flow, DMRP load, flow weighted concentration and mean concentration
in tile effluent from the tile monitoring stations and the county line river station for the period from
March, 1994 through December, 1995.
Site
A
B
v
D2
:
:
G
H
R05
?iver3
Soil Bray
P-l
(mg-P/kg)
Annual
Flow
Depth
(mm/yr)
66 d'
66 d
74 c
103 a
59 d
64 d
21 e
87 b
139
138
99
81
127
138
159
84
64
DMRP Load
(kg-P/ha-yr
0.126
0.143
0.080
0.051
0.082
0.202
0.122
0.075
0.169
DMRP Concentrations
Flow
Weighted
0.080
0.091
0.072
0.058
0.057
0.130
0.068
0.078
0.230
Mean of Max
samples
mg-P/ L)
0.071 0.48
0.072 0.39
0.065 0.37
0.062 0.31
0.055 0.33
0.134 2.78
0.061 0.18
0.161 1.64
0.205 2.10
0.130 2.10
% of obs. N
>0.05
mg-P/L
53 75
45 29
42 48
40 116
43 94
53 171
50 93
64 103
70 120
69 498
1 values followed by any identical letters are not statistically different at the 0.05 level of significance
2 monitoring at D began in August, 1994
3 river concentrations include station R05
253
-------
Flow weighted average concentrations of DMRP in tile effluent ranged from 0.057 to 0.13
mg P/L The mean concentrations of the individual sample values ranged from 0.161 to
0.055 mg-P/L For most sites, the two measures of average DMRP concentration were
similar. The one exception to this was site H where the flow weighted concentration was
0.078 and the mean of the sample values was 0.161 mg-P/L. This indicates that more
samples were taken when flow was low and concentration was relatively high. All of
these average concentrations are considered to be sufficient to promote
eutrophication.
There was little apparent relation between soil Bray P-l and mean DMRP in the tile drain
effluent for these sites. The greatest and the least concentrations were observed at sites
F and E, respectively, two sites where soil concentrations were intermediate. Similarly,
the sites with the greatest and the least soil Bray P-1 concentrations, sites D and G, both
had low levels of DMRP as compared to the other sites.
The lack of correlation between Bray P-l in the top soil and tile DMRP in the drain effluent
is perhaps because both Bray P-l and DMRP represent only portions of total P, and
perhaps due to the relative immobility of P in the soil. Given this relative immobility there
is likely to be a very long lag time between land management practices and P
concentrations in tile drain effluent.
Maximum concentrations from individual samples of tile effluent range from 0.19 at G to
2.78 at F. Between 40 and 64% of the samples taken from the subsurface drain tiles
exceeded the 0.05 mg-P/L level.
DMRP concentrations at river station R05 were greater than at the tile stations and the
loads at R05 were greater than all tile stations except for F. Seventy percent of the
samples taken from R05 had DMRP concentrations greater than 0.05 mg-P/L.
Concentrations at R05 were also greater than other river stations. This may have been
because an automatic sampler was used at R05, which captured more samples during
storm events and at higher flow rates than the manually collected samples. Due to the
relative immobility of phosphorus in the soils, it tends to accumulate in soil surface layers
and to move more readily in surface runoff than subsurface flow. Consequently,
phosphate concentrations in surface runoff are likely to be greater than in subsurface
flow. Furthermore, river concentrations of DMRP are likely to increase during runoff
events. Although DMRP concentrations observed in tile drain effluent in this study were
less than observed in the river and in other runoff studies (Mclsaac et al., 1995), they are
somewhat greater than observed in tile drainage in pastures in England (Hawkins and
Scholefield, 1996).
LESSONS LEARNED
In summary, this study has shown that the concentration of phosphorus in subsurface
drainage in the study area frequently exceeds the level thought to promote
eutrophication. The fact that P concentrations in drainage from the unfertilized
254
-------
continuous meadow site was greater than those from some of the intensively fertilized
and cropped sites suggests that modifying fertilizer and cropping patterns will not have
an immediate impact on P concentration of the subsurface drainage water and
downstream water bodies.
REFERENCES
Bogue, A. (1963) From Prairie to Corn Belt: Farming on the Illinois and Iowa Prairies in the
Nineteenth Century. University of Chicago Press.
Bogue, Margaret Beattie. (1951) The Swamp Land Act and wet land utilization in Illinois,
1850-1890. Agricultural History 25:169-180.
Hawkins, J.M.B. and D. Scholefield. (1996) Molybdate-reactive Phosphorus in Surface
and Drainage Waters from Permanent Grassland. Journal of Environmental Quality
25:727-732.
Logan, T.J., G.W. Randall, and D. R. Timmons. (1980) Nitrate Content of Tile Drainage
from Cropland in the North Central Region. Ohio Agricultural Research and
Development Center, Research Bulletin 1119, Wooster, Ohio.
Mclsaac, G.F., J.K. Mitchell and M. C. Hirschi. 1995. Dissolved Phosphorus
Concentrations in Runoff from Simulated Rainfall on Corn and Soybean Tillage Systems.
Journal of Soil and Water Conservation. 50(4):383-388.
Midwest Plan Service. (1975) Livestock Waste Facilities Handbook. Midwest Plan Service,
Ames, Iowa.
Ramamurthy, G. S. (1994) Chemical Surface Water Quality: Ambient Surface Water
Qualtiy Trends in Streams and Lakes. In: The Changing Illinois Environment: Critical Trends
Technical Report of the Critical Trends Assessment Project Volume 2: Water Resources,
pp 13-32.
Sharpley, A.N., S.C. Chopra, R. Wedepohl, J.T. Sims, T.C. Daniel, K.R. Reddy. (1994)
Managing Agricultural Phosphorus for Protection of Surface waters: Issues and Options.
Journal of Environmental Quality 23:437-451.
USEPA. (1992) National Water Quality Inventory: 1990 Report to Congress. EPA 503/9-
92/006. USEPA, Washington, DC.
ACKNOWLEDGMENTS
Contribution of the Illinois Agricultural Experiment Station, University of Illinois at Urbana-
Champaign as a part of Project 10-309 and Southern Regional Research Project S-218.
Supported in part with funds from the Little Vermilion River Hydrologic Unit Area Project
255
-------
and by the Illinois Groundwater Consortium (SIUC 92-04). We also gratefully
acknowledge the assistance of the Champaign County Soil and Water Conservation
District that sponsored the installation of the river gaging station. The authors also thank
Steve Maddock for his assistance with installing and maintaining field equipment and
Duane Kimme, Assistant Support Scientist, for laboratory analysis.
256
-------
The Success of Farm *A* Syst and Safe H2OMf
Water Quality Programs in Maine
Peter R. Wagner and John M. Jemison, Jr.
University Of Maine Cooperative Extension. Orono, Maine
The University of Maine Cooperative Extension has begun two programs to
address possible non-point source pollution impacts on water from two sources:
farmlands and residential communities. The Farm *A* Syst program aids
producers to identify how they may be affecting their own drinking water and
nearby water bodies. Eighty farms have participated in the program and where
needed, best management practices have been recommended. Common problem areas
found on farms included livestock yard maintenance, manure storage, pesticide
storage, and nutrient management. Similarly, the Safe H2OA/f Program aids
homeowners to assess the impact of their activities on their well water. One
hundred-twenty families have participated in a program correlating their
activities with actual water analyses. Approximately 25 percent of
respondents indicated that they believed their activities put them at high
risk for potential water quality problems. Twenty-seven percent of home wells
tested positive for the presence of bacteria in water.
Introduction
Maine is a very rural state. Most of its population live in communities with
populations less than 10,000 people. Maine covers approximately 19.8 million
acres. Approximately 89 percent of this area is forested, while only eight
percent is farmland. Maine's 2400 full-time farms produce potatoes, wild
blueberries, milk, brown eggs, chickens, and apples on a large scale. Maine
also has many small scale vegetable farms and their numbers are growing.
Ninety-five percent of Maine's rural communities rely upon well water for
their drinking water supply. With such a high percentage, we are very
concerned about the quality of Maine's water and the health of her people and
environment. Maine's superficial geology is such that Maine has few shortages
of water with its sand and gravel aquifers and abundant lakes and streams. We
are concerned with keeping these waters clean. In the past two years the
University of Maine Cooperative Extension in conjunction with the Maine
Department of Environmental Protection and other state agencies has prepared
the Safe H2OME and Farm *A* Syst programs for the residents of Maine. These
programs are educational tools that help homeowners and farmers to assess how
their activities could negatively impact water quality.
257
-------
Procedure
Both the Safe H2OWf and the Farm *A* Syst programs are organized in a series
of fact sheets and work sheets. The fact sheets are included in the packet to
provide background information to the program participants so that they will
be well versed in program terminology. The work sheets asks the participant
specific information about the given topic areas. Participants then assign a
rating to their activities based on how they assess their impact on water
quality for that topic area.
The Safe H2OWf Program was made available to interested residents of Maine and
to 120 residents of the Royal River watershed in southern Maine where a water
quality education effort was underway. Watershed residents received a free
water test kit in exchange for completing a program evaluation. Individual
assessments could then be compared to actual water test results to determine
the accuracy of a homeowner's impression of their impact on water quality.
The results of that evaluation are given below.
The Farm *A* Syst Program focuses on assessing farm impacts on water quality.
This program was made available in the winter and spring of 1996. This
program was made available to 80 farms in five Maine counties with the help of
Americorps Volunteers associated with the Consolidated Farm Services Agency.
Volunteers met with individual farmers to complete the program evaluation.
Results and Discussion
Safe H£ME Program
Maine's Safe H2OWE Program is largely an educational tool to help Maine
families identify possible pollution risks in or around their homes. The five
fact and work sheets help to evaluate activities around the home that may
present possible health risks to the family by affecting the water supply.
The Safe ^OME Program fact sheets and work sheets cover the following topics:
Table 1. Safe H2OM£ Program Topic Areas
Well construction and maintenance
Household hazardous waste
Household wastewater
Lawn and garden care
Lead in the environment
Through our evaluation, participants living in the Royal River watershed were
asked to assess whether their activities involving the five previous topic
areas put their water supply at low, low-moderate, moderate-high or high risk
for contamination. The results for those participants who felt their water
258
-------
supply was put at moderate-high or high risk of contamination due to their
home activities is presented in Table 2.
Table 2. Evaluation respondents who felt their drinking water supply
could be at moderate-high to high risk for contamination.
Topic Area Percent of Respondents
Well construction and maintenance 25.8t
Household wastewater 25.8
Lead in the home 24.1
Household hazardous waste 23.6
Lawn and garden care 23.2
t Remainder of percentages indicate a response of low to moderate-low risk.
Actual water quality results indicated that the respondents were able to make
a fairly accurate assessment of their impact on their water quality by working
through the Safe H2OA/f Program work sheets. More than 25 percent of the
program respondents had a contaminated water supply (Table 3). For this
program, we were concerned with the following water quality parameters: total
bacteria, fecal coliform, E. coli, nitrate, and arsenic. It is likely that
the wells contaminated with bacteria (27 percent) are related to the
percentage of moderate-high to high risk ratings for the following categories:
well construction and maintenance, household wastewater and lawn and garden
care.
Table 3. Percentage of homeowners with contaminated water supplies.
Contaminant Percent of Water Samples
Total bacteria 27.4
Fecal coliform 15.4
E. coli 15.4
Nitrate (N03) O.Of
mean value 0.68 mg I'1
Arsenic (As) 1.1 §
mean value 0.008 mg T1
~fThose samples that tested above the EPA standard of 10.0 mg I"1 for nitrate.
§ Those samples that tested above the EPA standard of 0.05 mg T1 for arsenic.
The final revealing question on the Safe H2OA/f Program evaluation asked
program participants what activities around the home they would change as a
result of reviewing the Safe H2OA/£ Program fact sheets and work sheets.
Almost 39 percent of the participants felt inclined to reduce the use of
hazardous material around the home (Table 4). Other areas of concern include
regular water testing, septic system maintenance and to improve well
construction.
259
-------
In a follow-up survey, those participants with bacterial contamination were
able to correct the problem through information and education from our
program. Through the Safe H2OMF Program and our evaluation process, we have
been able to assess the actual quality of water supplies in the Royal River
watershed and of the educational needs of citizens so that they may work
towards improving their water supplies.
Table 4. Areas of anticipated change to improve home water quality.
Activity Percent response
Reduce use of hazardous products 38.9
Regular water testing 23.4
Maintain septic system 13.7
Improve well construction 10.9,
Decrease water use 5.7
Install lead free plumbing 3.4
Increase garden soil organic matter 2.3
Increase soil testing h7
Farm *A* Syst
Maine's farming community does not encompass a large part of Maine's
landscape. These farms are located on many of Maine's rivers and near
population centers where farming activities.could have a significant impact on
water quality. With the initiation of Maine's Farm *A* Syst Program. Maine's
farm families will hopefully better understand how their activities can affect
local water quality. By identifying these activities, recommendations for
change will not only benefit water quality but will hopefully improve farm
productivity and quality of life on the farm.
Maine's Farm *A* Syst Program, a series of 11 fact sheets and work sheets.
that help farms to identify activities that may impact water quality. The
Farm *A* Syst Program fact sheets and work sheets cover the following topics:
Table 5. Farm *A* Syst Topic Areas
Drinking water well condition
Pesticide storage and handling
Fertilizer storage and handling
Petroleum product storage
Hazardous waste management
Household wastewater treatment
Livestock waste storage
Livestock yard management
Silage storage
Milking center wastewater storage
Overall farmstead assessment
260
-------
Many producers identified activities that put their home well water at
moderate-high to high risk for contamination (Table 6). Over 20 percent of
the respondents felt the following activities could contaminate their water
supply: livestock yard management, pesticide storage and handling, well
construction and silage storage. As with the Safe H2OMf Program, this
evaluation process provides Cooperative Extension and other agencies important
information about educational needs as identified by farms themselves.
Table 6. Evaluation respondents who felt their drinking water supply
could be at moderate-high to high risk for contamination.
Topic Area Percent of Respondents
Livestock yard management 32.6
Pesticide storage and handling 24.4
Drinking water well condition 21.0
Silage storage 20.5
Petroleum product storage 18.2
Milking center wastewater treatment 15.6
Household hazardous waste 10.5
Fertilizer storage and handling 9.7
Hazardous waste management 3.8
Conclusion
The Safe H2OWf and Farm *A* Syst programs are easy to use sources of
information for homes and farms. Both help the user understand possible
threats to water supplies and to determine if their activities could be a
threat. The evaluation process for these programs is extremely useful in
targeting educational opportunities towards both of these groups. Both groups
of participants identified what areas were of greatest concern to them. With
the Safe H2OM£ Program, we were able to validate these concerns with actual
water test results.
261
AU.S. OOVEnWEOT PRIOTDW OFFICE:1998-6SO-001/80176
------- |