United States
Environmental Protection
Agency
Office of Research and Development
Office of Water
Washington, DC 20460
EPA/625/R-92/006
August 1992
xvEPA Seminar Publication
The National Rural Clean
Water Program Symposium
10 Years of Controlling Agricultural Nonpoint
Source Pollution: The RCWP Experience
-------
-------
EPA/625/R-92/006
August 1992
PROCEEDINGS
The National RGWP Symposium
10 Years of Controlling Agricultural Nonpoint Source
Pollution: The RCWP Experience
September 13-17, 1992 • Orlando, Florida
Hosted by the
The South Florida Water Management District
in cooperation with
Q.S. Environmental Protection Agency
G.S. Department of Agriculture
Agricultural Stabilization and Conservation Service
Soil Conservation Service
Cooperative Extension Service
Printed on Recycled Paper
-------
The Florida Symposium Planning Committee
Boyd Gunsalus, South Florida Water Management District (Chairperson)
Diane Conway, (JSDA Agricultural Stabilization and Conservation Service (Vice Chairperson)
Greg Sawka, South Florida Water Management District
Kurt Harclerode, South Florida Water Management District
Eugene Badger, USDA Agricultural Stabilization and Conservation Service
Beverly Arrants, USDA Agricultural Stabilization and Conservation Service
Michael Greene, (JSDA Agricultural Stabilization and Conservation Service
Jack Stanley, USDA Agricultural Stabilization and Conservation Service
Vickie Hoge, Florida Cooperative Extension Service
Steve Mozley, USDA Soil Conservation Service
Leroy Crockett, USDA Soil Conservation Service '
Patrick Miller, Florida Cooperative Extension Service
The National Steering Committee
Eugene Badger, USDA Agricultural Stabilization and Conservation Service, Florida
Diane Conway, USDA Agricultural Stabilization and Conservation Service, Florida
Tom Davenport, U.S. Environmental Protection Agency, Region V
Tim Denley, USDA Agricultural Stabilization and Conservation Service
Steve Dressing, U.S. Environmental Protection Agency
Jeanne Goodman, South Dakota Department of Environment and Natural Resources
Boyd Gunsalus, South Florida Water Management District
Richard Magleby, USDA Economic Research Service
Don Martin, U.S Environmental Protection Agency, Region X
Don Meals, University of Vermont
Jim Meek, U.S. Environmental Protection Agency
Dan Murray, U.S. Environmental Protection Agency, Cincinnati
Paul Robillard, Pennsylvania State University
Dan Smith, USDA Soil Conservation Service
Paul E. Smith, USDA Agricultural Stabilization and Conservation Service-CEPD
Jean Spooner, North Carolina State University Water Quality Group
Francis Thicke, USDA Extension Service
NOTICE
The information in this document has been subject to the U.S. Environmental Protection Agency's
peer and administrative review, and it has been approved for publication. The work and opinions
described in these papers are those of the authors and, therefore, do not necessarily reflect the views
of the Agency. No official endorsements should be inferred.
For copies of this publication, contact
CJ.S. Environmental Protection Agency
CERI
Document Distribution (Q-72)
26 Martin Luther King Drive
Cincinatti, Ohio 45268
-------
Foreword
T
I he lessons learned from a 10-year experiment in controlling nonpoint source pollution are
recorded in this proceedings of the 1992 National Rural Clean Water Program (RCWP)
Symposium.
The symposium itself was designed to present the results of the RCWP experience to Federal,
State, and local project managers, landowners, and others interested in solutions to nonpoint
source pollution.
In the interest of providing guidance for State nonpoint source programs and local watershed
projects, the symposium addressed both successes and obstacles experienced by the RCWP. The
results presented in this proceedings have been peer reviewed.
Symposium organizers recognize the enormous contribution of the reviewers of these papers,
among them scientists from the U.S. Environmental Protection Agency's Office of Research and
Development. Their expertise, and the time they gave to this project, assured the technical
quality of the papers presented at the symposium and published in this volume.
Others also deserve our gratitude:
• The National Steering Committee, for its vision and guidance of this program to accurately
present the RCWP experience, and
• The Florida Symposium Planning Committee, for translating the vision into reality and the
RCWP experience into a training ground for the rest of the country.
That, of course, is the mission of the symposium and this proceedings — to build on the foun-
dation of the Rural Clean Water Program a blueprint for controlling agricultural nonpoint source
pollution in this Nation.
The symposium sponsors salute those who worked with the 21 project sites over the past
decade and join with them in presenting the fruits of the Rural Clean Water Program as lessons to
be used by all who strive to meet the challenge of controlling nonpoint source pollution.
iii
-------
The Rural Clean Water Program
• Lake Tholocco, Alabama — Coffee and Dale Counties
• Appoquinimink River, Delaware — Hew Castle County
• Taylor Creek-Nubbin Slough Basin and Kissimmee River, Florida
—Glades, Highlands, Martin, Okeechobee, Counties
• Rock Creek, Idaho — Twin Falls County
• Highland Silver Lake, Illinois — Madison County
• Prairie Rose Lake, Iowa — Shelby County
• Upper Wakarusa River, Kansas — Osage, Shawnee, Wabaunsee Counties
• Bayou Bonne Idee, Louisiana — Morehouse Parish
• Double Pipe Creek, Maryland — Carroll County
• Westport River Watershed, Massachusetts —
Bristol County
• Saline Valley, Michigan — Washtenaw County
• Garvin Brook, Minnesota — Winona County
• Long Pine Creek, Nebraska — Brown, Rock Counties
• Tillamook Bay, Oregon — Tillamook County
• Conestoga Headwaters, Pennsylvania — Lancaster County
• Oakwood Lakes-Poinsett, South Dakota — Brookings, Hamlin,
Kingsbury, Counties
• Reelfoot Lake, Tennessee/Kentucky — Lake, Fulton, Obion Counties
• Snake Creek Project, Utah — Wasatch County
• St. Albans Bay, Vermont — Franklin County
• Nansemond-Chuckatuck, Virginia — City of Suffolk and Isle of Wright
County
m Lower Manitowoc River Watershed, Wisconsin — Brown, Calumet,
Manitowoc, Counties
-------
Contents
Foreword iii
Water Quality and Land Treatment Monitoring
The Evolution of the RCWP Water Quality Monitoring Networks in the Taylor Creek/
Nubbin Slough and Lower Kissimmee River Basins 1
Kathy Osking and Boyd Gunsalus
Nutrient Loadings and Chlorophyll a in the Oakwood Lakes System 15
David R. German
Nitrate and Pesticide Occurrence in Shallow Groundwater During the Oakwood Lakes-Poinsett
RCWP Project 33
Jeanne Goodman, J. Michael Collins, and Keith B. Rapp
Water Quality Trends in the St. Albans Bay, Vermont, Watershed Following RCWP
Land Treatment 47
Donald W. Meals
Understanding the Groundwater System: The Garvin Brook Experience 59
David B, Wall, Mark G. Euenson, Charles P. Regan, Joseph A. Magner, and Wayne P. Anderson
Keeping Bacteria Out of the Bay—The Tillamook Experience 71
J.A. Moore, R. Pederson, and J. Worledge
A Tracking Index for Nonpoint Source Implementation Projects 77
Steven A. Dressing, John C. Clausen, and Jean Spooner
The Effects of Temporal and Spatial Variability on Monitoring Agricultural Nonpoint
Source Pollution 89
Thomas H. Johengen and Alfred M. Beeton
Relating Water Quality to Land Treatment
Effects of Pipe-Outlet Terracing on Runoff Water Quantity and Quality at an Agricultural
Field Site, Conestoga River Headwaters, Pennsylvania 97
Patricia L. Lietman
Effects of Nutrient Management on Nitrogen Flux through a Karst Aquifer, Conestoga River
Headwaters Basin, Pennsylvania 115
David W. Hall and Dennis W. Risser
Relating Land Use and Water Quality in the St. Albans Bay Watershed, Vermont 131
Donald W. Meals
Spatial and Temporal Change in Animal Waste Application in the Jewett Brook Watershed, Vermont:
1983-1990 145
Joel D. Schlagel
Water Quality and Land Treatment in the Rock Creek, Idaho, Rural Clean Water Program 151
Gwynne Chandler and Terry Maret
-------
Effectiveness of Agricultural Best Management Practices Implemented in the Taylor Creek/
Nubbin Slough Watershed and the Lower Kissimmee River Basin 161
Boyd Gunsalus, Eric G. Flaig, and Gary Ritter
Estimation of Lag Time for Water Quality Response to BMPs 173
John C Clausen, Donald W. Meals, and E. Alan Cassell
Water Quality Trends in Big Pipe Creek During the Double Pipe Creek Rural Clean
Water Program 181
John L. McCoy and Robert M. Summers
Effects of Nutrient Management on Surface Water Quality in a Small Watershed
in Pennsylvania 193
Edward H. Koerkle
Land Treatment and Operation and Maintenance of BMPs
Cedar Revetment and Streambank Stabilization 209
Gayle Siefken
Manure Testing and Manure Marketing: Tools for Nutrient Management 217
Leon Ressler
Operation and Maintenance of RCWP BMPs in Idaho to Control Irrigation-induced
Erosion 223
Ron Blake
Project Coordination and Farmer Participation
Coordination is the Project Cornerstone • 235
Michael Kuck and Jeanne Goodman
Taylor Creek-Nubbin Slough RCWP Institutional Arrangement and Program Administration 239
John W. Stanley ' .
Farm Operators'Attitudes About Water Quality and the RCWP 247
Thomas J. Hoban and Ronald C. Wimberley
Factors Leading to Permanent Adoption of Best Management Practices in South Dakota
Rural Clean Water Program Projects 255
Karen Cameron-Howell
Techniques to Obtain Adequate Farmer Participation 261
Richard L Yankey
Farmer Participation in the Double Pipe Creek, Maryland, Rural Clean Water Program Project 265
Elizabeth A. Schaeffer
The Key to Successful Farmer Participation in Florida's Rural Clean Water Program 269
John W. Stanley
Institutional Arrangements, Program Administration, and
Project Spin-Offs
Document It! Procedures for the Documentation of Nonpoint Source Project Data —
LandTreatment , • • • • • • • • • • • • • • • • • •'•.;;.; 273
Betty Hermsmeyer
vi
-------
Problems and Conflicts Associated with the Administration of the Long Pine, Nebraska,
RCWP Project 279
Robert F. Hilske
RCWP—The Federal Perspective 287
James Meek, Carl Myers, Gordon Nebeker, Walter Rittall, and Fred Swader
Thinking About a Postproject Evaluation — Start NOW! 295
Clarence W. Robison and Charles E. Brockway
Utah's Snake Creek RCWP Stimulates Additional Efforts to Improve Water Quality
in Wasatch County /... 301
Ray Loveless, Todd Nielson, and Harry Judd
Information and Education
Information and Education— Lessons Learned from RCWP : 309
Bud Stolzenburg
Diversity of Information and Education Help Obtain Goals for Double Pipe Creek RCWP Project 313
David L. Greene
Nutrient Management Educational Initiative: Using Demonstration and Research Plots
and the Penn State Nitrogen Quick Test in the Upper Conestoga RCWP 321
Robert Anderson ,, -
Involving the Agricultural Chemical Industry in Nutrient Management 333
Jeffrey Stoltzfus, Leon Ressler, and Robert Anderson , •'.-.'•
Socioeconomics, Technology Transfer, Lessons Learned
Economic Evaluation of the Rural Clean Water Program .' 337
Richard Magleby
Technology Application in BMP Planning, Design, and Application 347
Gene Dougherty and Jesse T. Wilson
A Method for Ranking Farms and Tracking Land Treatment Progress in the St. Albans Bay
Watershed RCWP Project, Vermont 351
Richard J. Croft and Jeffrey D. Mahood
Elements of a Model Program for Nonpoint Source Pollution Control 361
Steven W. Coffey, Jean Spooner, Daniel E. Line, Judith A. Gale, Jon A. Arnold,
Deanna L. Osmond, and Frank J. Humenik
Extending the RCWP Knowledge Base to Future Nonpoint Source Control Projects 375
Paul D. Robillard
Research Needs and Future Vision
Research Needs and Future Vision for Nonpoint Source Projects 385
Paul D. Robillard, John C. Clausen, Eric G. Flaig, and Donald M. Martin
Additional Information
Synoptic Survey of Dairy Farms in the Lake Okeechobee Basin: Post-BMP Water
Quality Sampling 393
Gregory J. Sawka, Paul Ritter, Boyd Gunsalus, and Thomas Rompot
vii
-------
-------
The Evolution of the RCWP Water
Quality Monitoring Networks in the
Taylor Creek/Nubbin Slough and
Lower Kissimmee River Basins
Kathy Osking and Bpyd Gunsalus
South Florida Water Management District
Okeechobee, Florida
ABSTRACT
The South Florida Water Management District has supported two Rural Clean Water Program
water quality monitoring projects located in watersheds north of Lake Okeechobee. This region is
primarily agricultural with dairy and beef operations. In 1978, a monitoring network was designed
to evaluate water quality in the headwaters of Taylor Creek. At the inception of the Rural Clean
Water Program in 1980, the monitoring network incorporated the entire Taylor Creek/Nubbin
Slough drainage basin. In 1987, the Lower Kissimmee River basin monitoring network was in-
cluded in the program; data collection in this basin has been ongoing since 1986. The major objec-
tives have been to evaluate the effectiveness of best management practices in reducing nonpoint
source nutrient inputs to Lake Okeechobee and monitor water quality as these changes occur in
land use. Throughout the Rural Clean Water Program, network design has changed for sample fre-
quency, collection for analysis of chemical parameters, and intensity of monitoring on individual
farms.
All water in southern Florida flows through
an intricate network of hydrologically and
ecologically connected marshlands, con-
structed canals, streams, sloughs, rivers, and lakes.
The heart of this ecosystem is Lake Okeechobee,
the wellspring of water from the Kissimmee River
and other northern tributaries to the environmen-
tally unique Florida Everglades.
Lake Okeechobee is a vast body of water cover-
ing an area of approximately 1,890 square kilo-
meters, making it the second largest freshwater lake
within the contiguous United States. Located in
south central Florida, the lake is a vital resourc'e to
all inhabitants of the region. For five surrounding
cities, it is a potable water supply and for
municipalities along Florida's southeast Atlantic
coast, it is a source of aquifer recharge to abate salt
water intrusion. In the agricultural area south of
Lake Okeechobee, farmers use lake water to irrigate
crops, such as sugar cane, rice, and winter
vegetables. A highly productive biological ecosys-
tem, Lake Okeechobee contains large populations of
plants and animals and provides habitat to en-
dangered species such as the Everglade, or snail,
kite (S. Fla. Water Manage. Distr. 1989). The lake is
also a sportsman's paradise for fishing and hunting,
and it supports a lucrative commercial fishing in-
dustry.
Though the eutrophic state of Lake Okeechobee
(Federico et al. 1981) has made it a flourishing,
-------
Proceedings of National RCWP Symposium, 1992
productive resource, the lake is in danger of becom-
ing hypereutrophic because of increased nutrient
input, primarily phosphorus from agricultural ac-
tivities in areas north of the lake (Flaig and Ritter,
1989). Two watersheds identified as making the
most significant nutrient contributions to the lake
are the Taylor Creek/Nubbin Slough and Lower Kis-
simmee River basins. The Taylor Creek/Nubbin
Slough basin contributes 5 percent of the average an-
nual flow to the lake and approximately 28 percent of
the lake's average annual phosphorus load, while the
Lower Kissimmee River contributes 31 percent of
the average annual flow and approximately 20 per-
cent of the average annual phosphorus load to the
lake (Federico et al. 1981). Land uses in this region
are primarily agricultural: beef and dairy operations
with additional improved pastures; sod, vegetable,
and citrus farms. Highly dense populations of
animals present the major source of the phosphorus
input problem.
Several programs aimed at reducing the
elevated levels of phosphorus entering Lake Okee-
chobee have been instituted through the cumulative
efforts of various local, State, and Federal agencies.
The Rural Clean Water Program (RCWP) was
developed in 1980 to assist landowners in the design,
cost-share, and implementation of agricultural best
management practices (BMPs) (Okeechobee RCWP
Local Coor. Comm. 1981). Since the early 1970s, the
South Florida Water Management District has been
monitoring the quality of water in the Taylor
Creek/Nubbin Slough and Lower Kissimmee River
basins.
Network Beginnings—Early
Investigations and Research
(1955-78)
The Taylor Creek/Nubbin Slough basin first existed
in nature as two distinct hydrologic areas. Taylor
Creek originally included a basin of 36,248 hectares
and flowed through the city of Okeechobee into
Lake Okeechobee. Major tributaries discharging
into Taylor Creek include Otter Creek, Northwest
Taylor Creek, Little Bimini, Williamson Ditch, and
Wolf Creek as shown in Figure 1. From 1962
through 1968, improved channels, spillways, and
TAYLOR
E- CREEK
BASIN
DAIRY
TRIBUTARY GRAB SITE
TRIBUTARY AUTOSAMPLER SITE
NUBBIN
SLOUGH
BASIN
01234
SCALE IN MILES
Figure 1.—Taylor Creek/Nubbin Slough basin near Lake Okeechobee, Florida.
-------
K. OSKING & B, OUNSALUS
water control structures were constructed in the
upper Taylor Creek watershed to alleviate flooding
(Knise! et al. 19,85). Discharge from the watershed
was diverted in 1973 into the Lr63 interceptor canal,.
which also collects discharge from Mosquito Creek,
Nubbin Slough, Henry Creek, and Lettuce Creek.
These improvements increased the drainage area to
48,800 hectares. The entire. combined hydrologic
watershed, called the Taylor Creek/Nubbin Slough
basin, drains into the north end of Lake Okeechobee
through control structure S191 (Fig. 1).
A long-term hydrologic study extending from
1955 to 1975 was initiated in the upper Taylor Creek
basin through the cooperative efforts of various
Federal and State agencies (Yates et al. 1982). The
objective of this research effort was to monitor the
hydrologic behavior of the watershed before, during,
and after channelization and other flood control im-
provements in the 1960s and observe resulting ef-
fects upon groundwater elevations, streamflow dura-
tion, water yield, and storm runoff.
The U.S. Department of Agriculture Agricultural
Research Service (ARS) installed seven weight-
recording rainfall gages to measure daily rainfall
(Table 1 lists the sites). In 1959, groundwater wells
were installed at each rain station and equipped with
analog stage recorders.
As environmental concerns for Lake Okee-
chobee increased in the early 1970s, ARS expanded
the study with a water quality survey to identify
sources of nutrient pollution to the upper Taylor
Creek watershed. The objectives of the survey were
to assess the water quality of phreatic and open
channel flows in tributaries of the watershed, es-
timate nutrient loads, and determine relationships of
water quality and nutrient loads to watershed hydrol-
ogy and land use practices—which were primarily
dairy, beef, and citrus operations (Allen et al. 1976).
The survey consisted of 15 sampling locations of
which 12 were collected from open channels in the
Taylor Creek watershed and 3 from groundwater
wells. During part of the 1972 wet season (April
through July), samples were collected biweekly
(Knisel et al. 1985). Dry season sampling occurred
monthly until January 1973, when the survey was
temporarily discontinued. In March 1974, the ARS
reinstated an abbreviated version (which extended
through 1975) of the original open channel
streamflow water quality survey in the Taylor Creek
watershed (Stewart et al. 1978).
,The research found that channel improvements
and water level control structures promoted more
rapid recession of stormwater discharges, but had
only negligible effects on groundwater recharge and
drainage (Yates et al. 1982). Data obtained from both
water quality surveys indicated that overland
transport of nutrients had occurred as nitrate-
nitrogen -and orthophosphorus concentrations were
lower in groundwater wells than in open channels.
The most substantial concentrations of orthophos-
phorus (up to 3.26 mg/L) originated in the Otter
Creek watershed, where intensive dairy operations
drained into the tributary.
Taylor Creek and the upper portion of the S191
basin, Mosquito Creek, became the focus of study in
1975 when the South Florida Water .Management
District initiated a research project to investigate
relationships between water quality and land use
(Federico, 1977). The study's goals were to docu-
ment water quality in four subbasins of the Taylor
Creek/Nubbin Slough basin and determine the tem-
poral variation and land use effects on water quality.
The following four subbasins were selected for
the study based on the diversity of land use prac-
tices: Mosquito Creek, Williamson East Lateral,
Otter Creek, and Northwest Taylor Creek. Water
quality monitoring stations were located at each out-
flow of the four subbasins. The regime consisted of
four sample-collection runs during part of the 1975
wet season Quly through September). For each run,
an ISCO automatic sampler collected surface water
samples every three hours for three days.
Otter Creek and Mosquito Creek showed
elevated concentrations of total nitrogen (up to 6.97
mg/L) and total phosphorus (up to 2.97 mg/L) as a
result of runoff from major dairy operations existing
in the two watersheds. Samples from the Williamson
Table 1.—Agricultural Research Service rainfall and groundwater stations in the Upper Taylor Creek Basin.
STATION NAME
Well Line B
Bassett
Judson
Opal
Williams
Raulerson
Mobley
RAIN STATION
NUMBER
8500
8502
8504
8507
8501
8503
8506
GROUNDWATER
STATION NUMBER
8510
8512
8514
8517
LOCATION
Taylor Creek at R Bar Ranch
Bassett Ranch off Hwy, 68
Wilson Rucks Dairy off Hwy. 441 N
Williamson Cattle Company
McArthur Dairy cajf barn
Raulerson house on Potter Rd.
Lawrence Road off Hwy. 441 N
-------
Proceedings of National RCWP Symposium, 1992
East Lateral Ditch exhibited extremely high chloride
and sodium concentrations resulting from a citrus
operation that was irrigating with water from an ar-
tesian well. The results indicated that land use was a
significant influence on water quality. In particular,
dairy operations appeared to be significant sources
of nitrogen and phosphorus (Federico, 1977).
Taylor Creek Headwaters Project
(1978-81)
Results of early investigations prompted the 1976
Florida Legislature to create the Coordinating Coun-
cil on the Restoration of the Kissimmee River Valley
and Taylor Creek/Nubbin Slough Basin. Repre-
sentatives of various State and Federal agencies
were selected by the coordinating council to make
up a technical advisory committee. Recognizing that
agricultural operations should use management
practices to reduce nutrient runoff, the technical ad-
visory committee established the Taylor Creek
Headwaters Project in 1978 (Allen et al. 1982). The
project prioritized the Otter Creek subwatershed
and allocated approximately $300,000 to provide
landowners 100 percent funding to implement
BMPs, which included
• fencing cows out of streambeds,
• constructing supplemental watering facilities,
shade structures, and cattle crossings over
waterways, and
• establishing vegetative filter strips for
nutrient uptake and detention areas to hold
storm runoff.
The Taylor Creek Headwaters Project estab-
lished a water quality monitoring network designed
to evaluate the effectiveness of BMPs at reducing
nutrient runoff in the basin. The primary objective of
the network was to determine baseline water quality
and nutrient loads in the area. A comprehensive
database was established to provide pre-BMP con-
struction concentration data.
The initial monitoring was conducted by Agricul-
tural Research Service in November 1977 with 16
sampling locations that had been monitored since
1972 as part of the original open channel streamflow
water quality survey (Allen et al. 1976). ARS col-
lected hydrologic data and biweekly water quality
samples, and analyses for chemical and physical
parameters were performed by the South Florida
Water Management District's branch laboratory in
Okeechobee. The technical advisory committee con-
ducted an investigative grab sampling tour on June
7,1979, at 59 locations in the project area to pinpoint
the most significant contributors of nutrient inputs
and provide justification for the design of a long-term
water quality monitoring program. As a result, new
sites were added to the network, increasing the num-
ber of biweekly sampling sites to 28. Two additional
sites were included in 1980. Two automatic samplers
collected samples daily at upstream and downstream
locations on Otter Creek (Sites 207 and 209, respec-
tively, of Fig. 1).
In January 1981, the South Florida Water
Management District accepted responsibility for all
Taylor Creek Headwaters project management (Rit-
ter and Allen, 1982) and relieved ARS of all water
quality and hydrologic data collection in October
1981. Table 2 provides a comprehensive list of all
sample stations monitored throughout the Taylor
Creek/Nubbin Slough basin. Sample sites desig-
nated with the Taylor Creek Headwaters station
code (TCHW) were monitored under the project,
and monitoring continued at the ARS stations lo-
cated in the lower Taylor Creek/Nubbin Slough
basin. These sites represented water quality result-
ing from various land uses, including runoff from
dairy operations, beef cattle pastures, and hayfields.
Samples were collected biweekly at each station
and analyzed within one week. Samples were filtered
through a fiberglass prefilter and a polycarbonate
membrane and analyzed for nitrate, nitrite, am-
monia, and orthophosphorus. Unfiltered samples
were analyzed for total Kjeldahl nitrogen, total phos-
phorus, pH, specific conductivity, turbidity, and
color. Samples of lagoon water were also analyzed
for calcium, manganese, sodium, and potassium.
The period of the Taylor Creek Headwaters
Project (1978 through 1981) represented the pre-
BMP implementation phase. BMP effectiveness is
documented in detail by Gunsalus, Flaig, and Ritter
(1992). Generally, throughout the project, nitrogen
and phosphorus concentrations decreased as a
result of a decrease in annual rainfall in the area,
which helped deplete groundwater levels and dry
out many beef and dairy operations' drainage sys-
tems. The closing of one dairy operation in the Otter
Creek subwatershed resulted in a decrease in total
phosphorus concentration at the downstream dis-
charge station (Ritter and Allen, 1982).
RCWP Inception (1981)
In July 1981, the entire Taylor Creek/Nubbin Slough
basin (48,800 hectares) (Fig. 1) was selected to par-
ticipate in the RCWP as one of 21 federally funded
programs designed to reduce nonpoint source water
pollution problems in predominantly rural water-
sheds. Total funds ($1.3 millon) for the entire RCWP
project were administered by the USDA's Agricul-
-------
K. OSKfNQ & B. GC/NSALGS
Table 2.—Period of record, station code, and location of sampling sites
in the Taylor Creek/Nubbin Slough basin.
PERIOD OF RECORD
01/04/72 to Present
03/1 9/74 to Present
01/04/72 to Present
01/04/72 to 09/11/84
03/19/741009/03/81
03/1 9/74 to Present
01/04/72 to Present .
01/04/72 to Present
01/04/72 to Present
01/04/72 to 09/03/81
03/19/741010/01/91
01/04/72 to 09/1 0/84
03/1 9/74 to Present
03/1 9/74 to Present
03/1 9/74 to Present
11/01/77 to 10/25/83
11/01/77 to Present
09/05/79 to Present
09/05/79 to Present
09/05/79 to Present
09/05/79 to 09/03/81
09/05/79 to 09/24/80
09/05/79 to 10/01/91
09/05/79 to 09/27/83
09/05/791001/01/90
09/05/79 to 08/09/88
09/05/791010/01/91
09/05/79 to 09/03/81
11/19/801010/25/83
11/1 9/80 to 10/25/83
10/01/81 to Present*
10/28/811603/05/86
10/28/81 to 11/29/83
10/28/81 to 02/04/87
06/11/81 to Present
06/11/81 to Present
03/01/76 to 09/31/81
03/01/76 to 09/31/81
STATION
TCHW 01
TCHW 02
TCHW 03
TCHW 04
TCHW 05
TCHW 06
ARS07
ARS 08
ARS 09
ARS10
ARS 11
ARS 12
ARS 13
ARS 14
ARS 15
ARS 16
ARS 17
TCHW 18
TCHW 19
TCHW 20
TCHW 21
TCHW 22
TCHW 23
TCHW 24
TCHW 25
TCHW 26
TCHW 27
TCHW 28
TCHW 29
TCHW 30
TCHW 31
TCHW 32
TCHW 33
, TCHW 34
ARS 39
"'• ARS 40
TCHW 508
TCHW 509
LOCATION
N.W. Taylor Creek at Hwy. 68 .
Little Bimini at Potter Rd.
Otter Creek at S-13B
Otter Creek at Hwy. 68
Otter Creek at Otter Creek Rd.
Otter Creek at Potter Rd.
Williamson Main Ditch
Williamson East Lateral
Williamson Ditch at S-7
Taylor Creek at Hwy. 441
Taylor Creek at Cemetary Rd.
Taylor Creek at Well Line B
Mosquito Creek # Hwy. 71 0
Nubbin Slough at Hwy. 710
Mosquito Creek at Hwy. 70
Nubbin Slough at Hwy. 70
Nubbin Slough at Berman Rd. •
Taylor Creek at S-2
East Otter Creek at Potter Rd.
East Otter Creek at Hwy. 441
Little Bimini at Hwy. 68
F & R Dairy Runoff
Wilson Rucks Dairy Runoff
Remsberg North Runoff
McArthur #1 2nd Stage Lagoon
Otter Creek at McArthur Farms
McArthur Hayfield Runoff
Otter Creek Upstream
Gomez Creek at Hwy. 68 West
Gomez Creek at Hwy. 68 East
McArthur Runoff at Otter Creek
McArthur 2nd Stage Lagoon
T. Rucks 2nd Stage Lagoon
SEZ 2nd Stage Lagoon
Henry Creek at Hwy. 710
Lettuce Creek at Hwy. 71 0
Otter Creek at Potter Rd.
Otter Creek at Hwy. 441
*TCHW 31: No samples collected between 09/11/83 and 08/09/88
tural Stabilization and Conservation Service
(Okeechobee RCWP Local Coor. Comm. 1981). The
major goal of the Taylor Creek/Nubbin Slough
RCWP project was to achieve a 50 percent reduction
in phosphorus concentrations entering Lake
Okeechobee from the Taylor Creek/Nubbin Slough
basin by 1992 through implementation of agricul-
tural BMPs (Okeechobee RCWP Local Coor. Comm.
1982). The RCWP provided participating landowners
with 75 percent of the cost of BMP implementation,
up to $50,000 maximum. Landowners in the Taylor
Creek Headwaters Project area continued to receive
100 percent of the funding for best management
practices implemented under that project in addition
to the RCWP funds.
Each RCWP project had to establish a water
quality monitoring and evaluation element to provide
continuous feedback and documen-
tation of water quality changes
resulting from land treatment prac-
tices (Spooner et al. 1991). As the
Taylor Creek/Nubbin Slough
project was not slated for additional
funding from the RCWP for water
quality monitoring, the South
Florida Water Management District
stepped forward to provide a com-
prehensive monitoring and evalua-
tion program. The South Florida
Water Management District took
responsibility for establishing and
funding the program (as specified in
the approved monitoring and evalua-
tion plan) by enlarging the Taylor
Creek Headwaters Project monitor-
ing network (already in operation) to
include the entire Taylor Creek/
Nubbin Slough basin. A sound data-
base was available to provide back-
ground information for the moni-
toring effort.
The objectives of the RCWP
water quality monitoring project
were to
• evaluate the effectiveness of
agricultural BMPs for reducing
phosphorus concentrations in
tributaries, basin outlets, and
ultimately, Lake Okeechobee,
• identify trends in nutrient
concentrations that occurred as a
result of land use changes or
implementation of BMPs, and
• document the general water
quality throughout the nine
major tributaries to provide justification for
the use of agricultural BMPs to control
phosphorus runoff (Spooner et al 1991).
The sampling method and sample locations
monitored for the Taylor Creek Headwaters Project
continued to be used for the RCWP monitoring net-
work (Okeechobee RCWP Local Coor. Comm.
1982).
The local coordinating committee deleted urban
areas of the Taylor Creek/Nubbin Slough from the
critical area covered by the Rural Clean Water Pro-
gram in 1983 and redefined critical area to include all
dairy farms, all extensively drained beef cattle farms,
and all areas within 0.40 kilometers of streams,
ditches, or channels that continually hold water
(Okeechobee RCWP Local Coor. Comm. 1983).
-------
Proceedings of National RCWP Symposium, 1992
In May 1983, seven analog stage recorders were
installed to provide accurate calculation of nutrient
loads at the outflows of the four major tributaries
along the L-63 canal: Mosquito Creek (upstream and
downstream), Nubbin Slough, Henry Creek, and
Lettuce Creek (upstream and downstream). Water
quality sampling continued biweekly through 1986.
In summary, the data collected can be categorized
into three groups:
• 1978-82: Pre-BMP background
• 1983-87: BMP implementation/transition
• 1986- : Post-BMP Evaluation (completion
dates may vary)
Focus on Lower Kissimmee
River (1986)
Originally, the Kissimmee River meandered like a
snake through a wide, shallow floodplain and
naturally filtered its waters across vast marshlands.
However, in 1971, the river was channelized to
relieve flooding, which created a deep, narrow canal
90 kilometers long (Loftin et al. 1990). Located in
western Okeechobee County and eastern Highlands
and Glades counties, the Lower Kissimmee River
basin covers an area of approximately 90,600 hec-
tares. The basin drains into Lake Okeechobee
through control structure S65E and collects dis-
charge from tributaries including Fish Slough, Gore
Slough, Ash Slough, Cypress Slough, Chandler
Slough, Popash Slough, Yates Marsh, the L-62 canal,
and Maple River, as shown in Figure 2. Land use in
the basin consists of 91 percent beef cattle opera-
tions, 7 percent dairy operations, and 2 percent
woodland or urban areas (Stanley and Gunsalus,
1991). The basin became an area of concern as a
major contributor of phosphorus and nitrogen to
Lake Okeechobee—second only to the Taylor
Creek/Nubbin Slough basin (Federico et al. 1981).
In January 1986, the South Florida Water
Management District began to monitor the water
quality of selected tributaries throughout the Lower
Kissimmee River basin under the Kissimmee River
Eutrophication Abatement program in response to
mandates issued by the Kissimmee River Resource
- DAIRY
- TRIBUTARY GRAB SITE
' - TRIBUTARY AUTOSAMPLER SITE
MILES
I I I I
0 3
Figure 2.—Lower Klsstmmee River basin.
-------
K. OSKJNG & B. GCJNSALLfS
Planning and Management Committee and the Lake
Okeechobee Technical Advisory Committee (Ger-
main and Shaw, 1988). The sample site locations are
listed in Table 3. Sample frequency gradually in-
creased from monthly sampling intervals to biweek-
ly and then weekly intervals. The initial monitoring
efforts in the Lower Kissimmee River basin provided
the groundwork for a more intensive, comprehen-
sive network for both the Lower Kissimmee River
and Taylor Creek/Nubbin Slough basins.
Table 3.—Kissimmee River Eutrophication Abate-
ment (KREA) Project water quality grab sample
stations in the Lower Kissimmee River basin (sam-
pling began 1986).
STATION
CODE
LOCATION
KREA 01 Fish Slough at NW 240th Avenue
KREA 02 Chandler Slough at Chandler Road
KREA 03 Ash/Gore at C-700A bridge
KREA 04 Chandler Slough at N bridge on Hwy. 98
KREA 05 Chandler SLough at S bridge on Hwy. 98
KREA 06 Cypress Slough at Lamb Is. Rd. bridge
KREA07 . Larson #1 West on NW 160th Drive
KREA 08 Larson #1 East on NW 160th Drive
KREA 09 Upstream Larson #1 on Old Peavine Tr.
KREA 10 Butler Dairy #1 on Underbill Road
KREA 11 Butler #1 on Underbill House Road
KREA 12 Larson Dairy Road North side
KREA 13 Larson Dairy Road South side
KREA 14 demons Ranch pump off Larson Dairy Rd.
KREA 15 Yates Marsh off NW 128th Avenue
KREA 16 Brighton Dairy #1 on NW 56th Street
KREA 17 Yates Marsh at Plaits Bluff RR bridge
KREA 18 Yates Marsh at Platts Bluff S RR
KREA 33 Dry Lake Dairy #2 outfall on Hwy. 98
KREA 34 Ferrell Dairy outfall on Hwy. 78
KREA 35 L-59 at C-38 Southwest side
KREA 48 Eagle Bay at Hwy. 78
KREA 50 Pool E Maple River Downstreams
KREA 51 Pool E Maple River Upstream
KREA 52 Pool E East of Maple River
KREA 53 Pool E Yates Marsh E. Brighton Dairy
KREA 54 Pool E Daugherty cutoff
KREA 55 Pool E Larson/Butler 2 runoff
KREA 56 Pool E Yates Marsh
KREA 57 Pool E Butler Dairy runoff
KREA 58 Pool D Butler Dairy runoff
KREA 59 Pool D Chandler Slough Downstream
KREA 60 Pool D Chandler Slough Midpool
KREA 61 Pool D Chandler Slough Upstream
KREA 62 , Pool D Larson Dairy #1 runoff
KREA 63 Pool D Larson Dairy #1 East runoff
KREA 64 Pool D Larson Dairy #1 West runoff
KREA 65 Pool D EEEE Fish Camp Downstream
KREA 66 EEEE Fish Camp U.S. at Hwy. 98 N
KREA 67 Pool E Yates Marsh North
During the summer of 1986, a severe algal
bloom covered over 120 square miles of Lake
Okeechobee. Intense concern for the lake's future
grew as highly concentrated nutrient runoff from
agricultural operations (primarily dairies) north of
the lake was charged with causing the algal bloom.
The effectiveness of the Rural Clean Water Program
was questioned as conservationists demanded im-
mediate cleanup of the lake (Stanley et al. 1986).
In June 1987, the Florida Department of En-
vironmental Regulation adopted the Dairy Rule re-
quiring all dairies to obtain an operating permit and
implement BMPs designed to recycle nutrients on
the farm. Concurrent to institution of the Dairy Rule,
the Florida legislature enacted the Surface Water Im-
provement and Management (SWIM) Act, which re-
quired the South Florida Water Management
District to protect Lake Okeechobee's water quality
and reduce its average annual phosphorus inflow
load to 397 tons (Flaig and Ritter, 1989). The South
Florida Water Management District's SWIM plan es-
tablished total phosphorus concentration standards
for direct inflows to the lake (0.18 mg/L) and for out-
flows from various land uses including dairies (1.2
mg/L) (S. Fla. Water Manage. Distr. 1989). Dairies,
under the regulation of the Dairy Rule, were re-
quired to implement waste management systems to
try to meet the targeted 1.2 mg/L total phosphorus
concentration standard of the SWIM plan.
In 1987, the Lower Kissimmee River basin was
included as an RCWP project aimed at reducing
nutrient loads entering Lake Okeechobee from the
basin (Stanley et al. 1988). The same criteria that had
been used to determine the critical area of the Taylor
Creek/Nubbin Slough RCWP project were
employed for the Lower Kissimmee River project,
which considered all dairy farms to be the most criti-
cal. At this time, the program objectives of the
Taylor Creek/Nubbin Slough RCWP were paralleled
to those of the Lower Kissimmee River RCWP to
coordinate BMP implementation and water quality
monitoring with established Dairy Rule and SWIM
plan regulations (Spooner et al. 1991).
.. A target goal of a 43 percent reduction in phos-
phorus load to the lake was calculated for the Lower
Kissimmee River Rural Clean Water Program based
on the 0.18 mg/L inflow concentration standard set
forth in the South Florida Water Management
District's SWIM plan (Flaig and Ritter, 1989). BMPs
implemented in the Lower Kissimmee River basin
were similar to those of the Taylor Creek/Nubbin
Slough project, with emphasis on intensive waste
management systems designed to capture and
recycle nutrients to achieve a nutrient mass balance
for the farm. .
-------
Proceedings of National RCWP Symposium, 1992
Intensive Dairy Monitoring
(1987-89)
From 1987 through 1989, the South Florida Water
Management District's original water quality
monitoring network in the Taylor Creek/Nubbin
Slough and Lower Kissimmee River basins was
revolutionized and expanded to provide more inten-
sive on-farm, tributary, and basin monitoring to sup-
port the Rural Clean Water Program and respond to
monitoring needs to support State regulation. The
objectives of the monitoring program were as fol-
lows:
1. To provide a detailed assessment of water
quality throughout the basins to determine
the effectiveness of BMPs at reducing
phosphorus concentrations,
2. To predict the movement of phosphorus
throughout the basins, quantify, and estimate
the effectiveness of BMPs at reducing
phosphorus loads entering Lake
Okeechobee, and
3. To evaluate the efficiency of specific BMPs.
To meet these objectives, a multifaceted data col-
lection program was implemented to better quantify
water quality changes that resulted from land use
changes. Table 4 lists all water quality stations. Basin
water quality assessment was accomplished with
data obtained from weekly grab samples from dairy,
beef, and tributary discharge points to provide ade-
quate identification of recent episodic conditions and
data for load calculations (Flaig and Ritter, 1988).
Sample stations were located at points both
upstream and downstream of dairy operations. How-
ever, not all discharge outlets could be monitored at
some operations because of inaccessible or un-
defined discharge outlets. Monitoring of multiple
discharge points is discussed in detail by Sawka, Rit-
ter, Gunsalus, and Rompot (1992).
Surface water samples were collected by field
technicians using established sampling protocol
(South Fla. Water Manage. Distr. 1987). Unfiltered
samples were analyzed for total phosphorus and
total Kjeldahl nitrogen and filtered samples were
analyzed for nitrate, nitrite, ammonia, orthophos-
phorus, chloride, and color. Physical parameters
such as temperature, pH, dissolved oxygen, and con-
ductivity were measured by technicians in the field
using a Hydrolab multiparameter instrument. Flow
conditions were determined by observing the move-
ment of fluorescent dye injected in the stream. The
technicians also recorded the extent of aquatic
vegetation observed at the sample site.
To better quantify nutrient loads as set forth in
the second monitoring objective, American Sigma
automatic samplers were employed to collect real
time continuous water quality samples at major
tributaries, water control structures, and dairies
within the basins. The automatic samplers pumped
21 bottles of sample per week, three bottles per day.
Autosamplers were serviced and samples col-
lected weekly. A routine grab sample was also col-
lected from each site at the time of servicing.
Analysis of the continuous samples was limited to
the number of samples the South Florida Water
Management District laboratory could analyze per
week (250 per project). To reduce the number of
samples submitted to the laboratory, all samples col-
lected by the automatic samplers were screened
each week for soluble reactive phosphate at the
Okeechobee field office using a Hach DR 2000
spectrophotometer. Seven samples were selected out
of the total 21 to best represent the observed trend
in soluble reactive phosphorus concentrations for
the week. Unfiltered samples were analyzed for total
nutrients and filtered samples for dissolved
nutrients. The in-house soluble reactive phosphorus
screening analysis also provided quick information
for evaluating nonroutine grab samples (Sawka et al.
1992).
Several automatic sampler locations were also
equipped with environmental monitoring equipment
for recording stream stage (upstream and down-
stream of primary discharge points), velocity, rain-
fall, pH, conductivity, dissolved oxygen, and
oxidation-reduction potential. Environmental data
measurements were recorded every 15 minutes
through an Easylogger onto a data storage pack,
which was removed from the site monthly and
transported to the South Florida Water Management
District's Data Management Division to be
downloaded to the mainframe computer database.
Specific BMPs—such as irrigated sprayfields,
seepage areas, watering holes for cows, and fencing
of waterways with filter strips—were monitored
through weekly grab sampling and continuous auto-
matic sampler collection, where possible, to better
identify how they affected water quality.
Shallow groundwater monitoring wells (2
meters) were also installed at the majority of dairies
in the Lower Kissimmee River basin to supplement
data obtained from the original Agricultural Re-
search Service rain and groundwater stations in the
Taylor Creek/Nubbin Slough basin and better un-
derstand the rainfall, groundwater, and runoff
relationships existing throughout the basins.
For a project of such magnitude, maintaining a
high level of quality assurance and control was a key
8
-------
K. OSK1NG & B. GUNSALUS
Table 4.— Taylor Creek/Nubbin Slough and Lower Kissimmee River water quality monitoring station locations,
1988.
EASY LOG
DATA CODE
TCN2+
RUC2+
MCA4+
TCLB+
MCA2+
MCA1 +
TCO1 +
TCWR+
TCO2+
RUC3+
RUC1 +
TCEO+
TCS2+
TCWD+
ROFR+
SEZO+
TCWC+
LARB+
LARC+
MWHI+
MOSQ+
DAV2+
NEWP+
REDT+
NUBB+
HENR+
UNDE+
LETT+
MATT+
DAV3+
CHAN+
LARW+
BUT1 +
YATE+
RUCWF+
AUTO SAMPLE
CODE
TCN2
RUC2
MCA4
TCLB
MCA2
MCA1
TCO1
, TCWR
, TCO2
RUC3
RUC1
TCEO
TCS2
TCWD
ROFR
SEZO
TCWC
LARB
LARC
MWHI
MOSQ
DAV2
NEWP
REDT
NUBB
HENR
UNDE
LETT
MATT
DAV3
..•
LKR 04A
LKR 06A
LKR 08
LKR 10
LKR 17A
LKR 21
GRAB SAMPLE
STATION CODE
TONS 201
TCNS 202
TONS 203
TCNS 204
TCNS 205
TCNS 206
TCNS 207
TCNS 208
TCNS 209
TCNS 210
TCNS 211
TCNS 21 2
TCNS 213
TCNS 21 4
TCNS 21 5
TCNS 21 6
TCNS 21 7
TCNS 21 9
TCNS 220
TCNS 221
TCNS 222
TCNS 223
TCNS 225
TCNS 227
TCNS 228
TCNS 230
TCNS 231
TCNS 233
TCNS 235
TCNS 241
TCNS 242
TCNS 243
TCNS 245
TCNS 246
TCNS 248
TCNS 249
TCNS 252
TCNS 253
TCNS 255
TCNS 258
TCNS 259
TCNS 260
TCNS 261
TCNS 262
TCNS 263
TCNS 264
KREA01
KREA04A
KREA06A
KREA07
KREA08
KREA09
KREA10
KREA10A
KREA10B ,
KREA16
KREA 17A
KREA19
KREA 20
KREA 21
HISTORIC SAMPLE
STATION CODE
TCHW 01
TCNS 202
TCNS 104
TCHW 02
TCHW 27
TCNS 206
TCHW 03
TCHW 23
TCHW 06
TCNS 111
TCHW 19
TCNS 212
TCHW 18
ARS09
TCNS 215
OSEZ 01
TCNS 21 7
TCNS 219
ARS15
TCNS 221
ARS13
TCNS 223
TCNS 225
ARS 14B
ARS14
ARS 39
TCNS 231
ARS 40
TCNS 235
TCNS 241
TCHW 25
TCNS 243
ARS 07
ARS 08
ARS 11
ARS 17
TCNS 252
New Station
TCNS 255
New Station
New Station
New Station
New Station
New Station
New Station
New Station
KREA 01
KREA04A
KREA06A
KREA 07
KREA 08
'KREA 09
KREA 10
KREA10A
KREA 1 0B
KREA 16
KREA17A
KREA 19
KREA 20
KREA 21
LOCATION
N.W. Taylor Creek at Hwy. 68 bridge
H&T Rucks #1 at Little Bimini
McArthur #4 & 5 at Little Bimini
Little Bimini at Potter Road
McArthur #1 & 2 hayfield runoff
McArthur #1 & 2 runoff at Remilu
Otter Creek at structure S-1 3b
Wilson Rucks runoff at Otter Creek
Otter Creek at S-1 3 and Potter Road
H&T Rucks #3 runoff to West Otter Creek
H&T Rucks #2 runoff, East Otter Creek
East Otter Creek below Remilu Ranch
Taylor Creek Headwaters at S-2
Williamson ditch below Boys School
Rofra Dairy runoff at Dry Lake
Sloan Ray Dairy runoff at Wolff Creek
Wolff Creek off Cemetary Road
County ditch above Larson #8 at N.E. 80th
Mosquito Creek at Hwy. 70
Murphy White Dairy off Hwy. 71 0
Mosquito Creek below Hwy. 71 0
Davie Dairy #2 runoff on farm
New Palm Dairy runoff
Red Top Dairy runoff
Nubbin Slough below Hwy. 710
Henry Creek below Hwy. 710
Underhill Dairy runoff
Lettuce Creek below Hwy. 71 0
Mattson Dairy runoff off Hwy. 710
Davie Dairy runoff below 1 & 2
McArthur #1 runoff seepage field
West Otter Creek upstream at Hwy. 68
Williamson Ditch on Williamson Cattle
Williamson East Latteral
Taylor Creek at Cemetary Road
Nubbin Slough at Berman Road
Tributary runoff above New Palm Dairy
Runoff above Enrico Dairy at Berman Road
Runoff Newcomer Dairy at Nubbin Slough
Enrico Dairy runoff to Henry Creek
Enrico hayfield runoff at Berman Rd.
H.T. Rucks #3 runoff at Potter Rd.
Rofra Dairy runoff
Larson Dairy #7 runoff
Larson Dairy #8 runoff
Murphy White Dairy runoff at Hwy. 710
Fish Slough at N.W. 240th St.
Chandler Slough at JC Bass East
Cypress Slough at Watford Ranch
Larson #1 W on N.W. 1 60th Dr.
Larson Dairy #1 Tributary E at N.W. 160th Dr.
Larson #1 W UPS on N.W. 203rd Ave.
Butler Dairy #1 Outfall at Underfill] Rd.
Butler #1 runoff from east pastures
Butler #1 retention pond
Brighton #2 N.W. 156th St.
Yates Marsh Outfall at Children's Home
Queen Bee Farms runoff to Maple River
Sandfly Gully at Hwy. 98
WF Rucks Dairy Outfall (flume)
-------
Proceedings of National RCWP Symposium, 1992
Table 4.—Continued.
EASYLOQ
DATA CODE
POPS*
FERR+
WOLF+
RUCR+
WHIT+
LAR2F+
BUT2+
FLYG+
RUCC+
MICC+
EAGL+
BASW+
DL1+
DLI2+
AUTO SAMPLE
CODE
LKR30A
LKR 32B
LKR34
LKR 36
LKR 37
LKR 39
LKR 40
LKR 41
LKR 42
LKR 43A
LKR 44
LKR 45
LKR 47
LKR 68
LKR RI1
LKR IR2
S154
S191
S65C
S65E
S84
GRAB SAMPLE
STATION CODE
KREA23
KREA25
KREA28
KREA30A
KREA31
KREA32B
KREA34
KREA36
KREA37
KREA39
KREA40
KREA41
KREA41A
KREA42
KREA43A
KREA44
KREA44A
KREA44B
KREA45
KREA45A
KREA46A
KREA46B
KREA47
KREA49
KREA66
KREA68
KREAIR1
KREAIR2
S154
S191
S65C
S65E
S84
HISTORIC SAMPLE
STATION CODE
KREA23
KREA25
KREA28
KREA30A
KREA31
LREA32B
KREA34
KREA36
KREA37
KREA39
KREA40
KREA41
KREA41A
KREA42
KREA43A
KREA44
KREA44A
New Station
KREA45
New Station
KREA46A
New Station
KREA47
KREA49
KREA66
KREA68
KREA IR1
KREA IR2
S154
S191
S65C
S65E
New Station
LOCATION
Ash Slough at Viking Property weir
Turkey Slough off U.S. 98
Popash Slough at SCL RR N
Popash Slough at L-62, S-1 54
Newcommer heifer runoff at Hwy. 70
Dry Lake Dairy #1 Outfall to Hwy. 98
Ferrell Dairy Outfall at Hwy. 78
Wolff Dairy Outfall
Ralph Rucks Dairy Outfall
White Dairy Outfall to Smith Farms
Larson Dairy #2 Flume Outfall
Butler Dairy #2 Outfall -
Larson 2/Butler 2 Outfall
Flying G Dairy Outfall to Yates Marsh
C&M Rucks Dairy Outfall
Lamb Island Dairy Outfall
Lamb Island Dairy Outfall east
Lamb Island Dairy outfall at entrance
Micco Dairy Outfall
Micco Dairy at Chandler Ranch
Williamson Dairy Outfall
Williamson Dairy Outfall at Hwy. 98
Eagle Island Dairy Outfall
Dry Lake #1 & #2 outfall at Hwy. 98
Four E's Fish Camp bridge on Hwy. 98
Bass Ranch West Outfall to Ash Slough
Dry Lake #2 East Sprayfield
Dry Lake #2 NW Sprayfield
River Structure S1 54
Lake Structure S1 91
River Structure S65C
River Structure S65E
River Structure S84
element. Established procedures for sampling
protocol were practiced routinely by South Florida
Water Management District technicians to ensure
that the required precision, accuracy, and reliability
of the data were upheld (S. Fla. Water Manage. Distr.
1987). A set of quality assurance samples was sub-
mitted weekly along with the routine grab samples,
which included the following:
• a duplicate sample,
• a duplicate sample with an added spike of
known concentration,
• a replicate sample,
• a blank sample of deionized water, and
• a blank sample spiked with a known
concentration standard solution.
Quality assurance samples were submitted for
the automatic samplers along with routine samples
collected. All samples were transported through a
chain of custody protocol from the time the sample
was relinquished to the chemistry laboratory to the
time of sample disposal following analysis.
Expansion of the monitoring capabilities was
well under way for both RCWP projects by the sum-
mer of 1988. Throughout the project's development,
modifications to the sampling program were made as
necessary. The automatic sampler load of 21 bottles
of sample collected per week at each sample site was
reduced to seven bottles per week at some sites to
lessen the soluble reactive phosphorus screening
load at the Okeechobee field office before sample
submission to the chemistry laboratory. Weekly
screening of continuous samples collected at the
major tributary and structure inflow sites continued.
Other automatic samplers located at dairy sites (par-
ticularly in the Taylor Creek/Nubbin Slough basin)
were programmed to pump 80 mL every two hours
into a single 11-liter bottle to form one weekly com-
posite sample.
To provide greater sample preservation, all
samples to be analyzed for total nutrients were
acidified with 50 percent sulfuric acid to a pH less
than 2.0. All automatic sampler bottles were treated
with sulfuric acid before being placed in the field for
sample collection.
10
-------
K. OSK7/YCJ & B, GUNSALUS
Monitoring Streamlined (1990)
Faced with increasing economic pressure to imple-
ment expensive management options, dairy farmers
requested that the State of Florida institute a buy-out
program, which began in 1989 (Albers et al. 1991).
To date, 18 dairies in the basins have elected to par-
ticipate in the buy-out program, which paid them ap-
proximately $600 per cow and established deed
restrictions prohibiting future dairy activity on the
property. Water quality monitoring of buy-out dairies
continued to document the effects of change in land
use.
Through 1990, the Lower Kissimmee River-
Taylor Creek/Nubbin Slough water quality monitor-
ing program evolved and became more streamlined
to optimize the equipment and human resources
available (Ritter, 1991). Monitoring objectives were
as follows:
1. To monitor water quality in support of the
Florida Department of Environmental
Regulation Dairy Rule in accordance with
the 1.2 mg/L total phosphorus concentration
target standard for dairies,
2. To assess tributary and basin loading and
concentration inputs to Lake Okeechobee to
compare with the 0.18 mg/L total
phosphorus concentration SWIM standard
for inflows, and
3. To develop basin and spatial scale models
designed to predict changes in loads to Lake
Okeechobee as changes in land use occur.
Because of the increased sample load brought
on by the automatic sampler program and the labor-
intensive maintenance and repair work necessary to
keep the continuous sampling equipment in opera-
tion, some automatic samplers were either discon-
tinued or the number of samples collected was
reduced. The resolution of continuous samples col-
lected as a daily composite (seven bottles per week)
was sufficient to observe trends in water quality.
Weekly composite samples were discontinued as
they did not provide an adequate degree of resolu-
tion. Automatic samplers continuously collecting
daily composite samples were maintained at the
major inflows and tributaries in the two project
basins. A higher degree of quality control was estab-
lished for these remaining automatic sampler sites:
intake tubing was changed quarterly, sample
volumes calibrated, and the intake strainer cleaned.
A PVC intake strainer holder was designed and im-
plemented to keep the intake properly positioned to
collect flow in the stream without disturbing bottom
sediments.
Of the seven bottles of samples collected each
week, three were submitted to the South Florida
Water Management District chemistry laboratory in
West Palm Beach to be analyzed for total phos-
phorus and total Kjeldahl nitrogen. Samples were
submitted based on changes in the daily stage read-
ings recorded at each site. Intensified quality control
maintenance procedures for the Easyloggers were
implemented, including cleaning of all probes and
staff gages, removing of debris from the site, probe
calibrating with standard solutions, and replacing
used desiccant.
At times, sample stations were relocated or
added to collect the most representative discharge
point(s) of the dairy operation as completion of large
scale BMP construction altered the hydrology of the
property. Grab samples continued to be analyzed for
total and particulate nutrients, with physical param-
eters measured in the field by the technician using a
Hydrolab. Hydrolab s are calibrated daily with stand-
ard solutions before field use. Each Hydrolab under-
goes a quarterly maintenance/overhauling
procedure to clean the conductivity probe, replace
the dissolved oxygen membrane and reference solu-
tion, and replace the pH reference solution.
Current Monitoring Network
(1991)
To better facilitate the established water quality
monitoring network goals and objectives for the
RCWP projects, the South Florida Water Manage-
ment District modified the network design in Oc-
tober 1991. The network consists of three key
components:
1. Dairy monitoring in support of the Florida
Department of Environmental Regulation
Dairy Rule,
2. Synoptic surveys designed to identify
sources of high phosphorus concentrations
within the dairy operation, and
3. Tributary monitoring to assess water quality
throughout the basins.
Sites were deleted or added to sample all dairy
outflows as the construction of BMPs redirected off-
site discharge. Some buy-out dairy and subtributary
sites were also discontinued.
To provide water quality compliance monitoring
in support of the Dairy Rule, grab samples continue
to be collected on a biweekly basis from 68 sites rep-
resenting all active and selected inactive dairy out-
flows. An unfiltered sample is collected and analyzed
for total phosphorus and total Kjeldahl nitrogen.
11
-------
Proceedings of Hatlonal RCWP Symposium, 1992
Results are reported monthly to landowners, State
regulatory agencies, and other interested organiza-
tions.
A second, innovative component of the network
is the implementation of synoptic water quality sur-
veys on dairies (Sawka et al. 1992). The objective of
the synoptic surveillance program is to identify the
sources of high phosphorus concentrations within
the dairy farm operation. A team of two technicians
surveys an entire dairy operation during high water
conditions and collects grab samples, tracking all
flows within the farm. Samples are returned to the
Okeechobee facility, where they are screened for
soluble reactive phosphorus using the spectro-
photometer. The data collected provide a water
quality assessment of the farm that is reviewed with
the landowner and Florida Department of Environ-
mental Regulation personnel to identify and work
toward eliminating problem areas of high phos-
phorus concentration.
Finally, major tributaries and inflow control
structures to Lake Okeechobee are monitored
through 21 weekly grab samples and collection of
daily composites using automatic samplers at 17
sites (Figs. 1 and 2). Biweekly grab samples are also
collected from nine subtributary sites. Weekly grab
samples are analyzed for total and dissolved nutrient
species; daily composite and biweekly samples are
analyzed only for total nutrients. Continuous
samples collected from the major outflows of the
Lower Kissimmee River-Taylor Creek/Nubbin
Slough basins (S191, S65E, and S154) are all
screened for soluble reactive phosphate each week
at the Okeechobee field office before sample submis-
sion to identify any episodic trends that occurred
during the seven-day sampling period. Environmen-
tal equipment continues in operation at 16 automatic
sampler sites. Original Agricultural Research Ser-
vice rainfall and groundwater stage recorders con-
tinue to be monitored. Constant servicing,
maintenance, and repair of all equipment are essen-
tial to keep the continuous monitoring program
functioning.
The present monitoring program succeeds be-
cause it has experienced personnel specifically dedi-
cated to each component of the network.
of a comprehensive evaluation of the pre- and post-
BMP implementation. Although it was, difficult to
evaluate the effectiveness of individual BMPs, future
monitoring efforts strategically located on dairy
operations should provide valuable information.
The monitoring program has evolved from
routine collection of grab samples into a complex
level of environmental monitoring technology. Many
lessons have been learned throughout the evolution.
Thorough servicing and maintenance of a large
number of continuous monitoring stations is difficult
to sustain because of the considerable investment in
resources and the labor-intensive nature of the work.
By using a moderate number of automatic samplers,
Easyloggers, and other environmental equipment,
technicians have time to perform proper field main-
tenance necessary to provide more reliable data.
The importance of •maintaining field quality as-
surance and quality control protocols for precision
and accuracy are essential to the success of the pro-
gram and ensure quality data.
The monitoring program has been well received
by landowners, whose concerns over regulatory en-
forcement prompted participation in the Rural Clean
Water Program. Landowners are appreciative of the
information produced with the synoptic surveillance
component of the current monitoring program to
help identify and alleviate trouble spots.
The protection of Lake Okeechobee, the heart of
the fragile Kissimmee River-Lake Okeechobee-
Everglades ecosystem, is a primary mission for the
South Florida Water Management District. Efforts
must be ongoing to reduce
-------
K. OSKING & B. GUNSALUS
Allen, L.H. Jr., E.H. Stewart, W.G. Knisel Jr., and R.A. Slack 1976.
Seasonal variation in runoff and water quality from the Taylor
Creek watershed, Okeechobee County, Florida. Proc. Soil
Crop Sci. Soc. Florida 35:126-38.
Allen. L.H. Jr. et al. 1982. Land use effects on Taylor Creek water
quality. Proc. Am. Soc. Chem. Eng. Spec. Conf. Environ.
Sound Water Soil Manage., Orlando, FL.
Davis, F.E. and M.L. Marshall. 1975. Chemical and Biological In-
vestigations of Lake Okeechobee, January 1973-June 1974,
Interim Report Tech. Publ. #75-1. Resour. Plann. Dep.,
Central Southern Florida Flood Control Distr., West Palm
Beach.
Federico, A.C. 1977. Investigations of the Relationship Between
Land Use, Rainfall, and Runoff Quality in the Taylor Creek
Watershed. Tech. Publ. #77-3. Resour. Plann. Dep., S. Florida
Water Manage. Distr., West Palm Beach.
Federico, A.C., KG. Dickson, C.R. Kratzer, and F.E. Davis. 1981.
Lake Okeechobee Water Quality Studies and Eutrophication
Assessment Tech. Publ. #81-2. Resour. Plann. Dep., S.
Florida Water Manage. Distr.^ West Palm Beach.
Flaig, E.G. and GJ. Ritter. 1988. Identification and Quantification
of Phosphorus Loads to Lake Okeechobee, Florida. Pap. No.
88-2115. Am. Soc. Agric. Eng., Rapid City, SD.
. 1989. Water Quality Monitoring of Agricultural Discharge
to Lake Okeechobee. Pap. No. 89-2525. Am. Soc. Agric. Eng.,
St Joseph, MI.
Germain, GJ. and J.E. Shaw. 1988. Surface Water Quality Monitor-
ing Network South Florida Water Management District.
Resour. Plann. Dep., S. Florida Water Manage. Distr., West
Palm Beach.
Gunsalus, B.E., E.G. Flaig, and GJ. Ritter. 1992. Effectiveness of
Agricultural Best Management Practices Implemented in the
Taylor Creek/Nubbin Slough and Lower Kissiramee River
Basins, Okeechobee, Florida. In Proc. National RCWP
Symp., Orlando.
Joyner, B.F. 1971. Appraisal of Chemical and Biological Conditions
of Lake Okeechobee, Florida, 1969-1970. Open File Rep.
71006. Geo. Surv., Water Resour. Div., U.S. Dep. Int., Tal-
lahassee.
Knisel, W.G. Jr. et al. 1985. Hydrology and Hydrogeology of Upper
Taylor Creek Watershed, Okeechobee County, Florida: Data
and Analysis. ARS-25. Agric. Res. Serv., U.S. Dep. Agric.,
Athens and Tifton, GA, and Gainesville, FL.
Loftin, M.K., LA Toth, and J.T.B. Obeysekera. 1990. Kissimmee
River Restoration, Alternative Plan Evaluation and Prelimi-
nary Design Report S. Florida Water Manage. Distr., West
Palm Beach.
Okeechobee Rural Clean Water Program Local Coordinating Com-
mittee. 1981. Project Plan of Work, Taylor Creek/Nubbin
Slough RCWP Project 14. Okeechobee Agric. Stabil. Con-
serv. Serv., Okeechobee, FL.
. 1982. Taylor Creek/Nubbin Slough RCWP 1982 Annual
Report. Okeechobee Agric. StabiL Conserv. Ser., Okeecho-
bee, FL
. 1983. Taylor Creek/Nubbin Slough RCWP 1983 Annual
Report. Okeechobee Agric. Stabil. Conserv. Serv., Okeecho-
bee,FL
Ritter, GJ. and LH. Allen Jr. 1982. Taylor Creek Headwaters
Project, Phase I Report: Water Quality. Tech. Publ. #82-8.
Resour. Plann. Dep., S. Florida Water Manage. Distr., West
Palm Beach.
Ritter, G J. 1991. Taylor Creek/Nubbin Slough and Lower Kissim-
mee River Water Quality Monitoring Network (Proposed
Design). S. Florida Water Manage. Distr., Okeechobee.
Sawka, G.J., PR. Ritter, B. Gunsalus, andT. Rompot 1992. Synoptic
Survey of Dairy Farms in the Lake Okeechobee Basin —
Post-BMP Water Quality,Sampling. In Proc. National RCWP
Symp., Orlando, FL.
South Florida Water Management District 1987. Generic Quality
Assurance Plan. For Florida Dep. Environ. Reg., West Palm
Beach.
. 1989. Interim Surface Water Improvement and Manage-
ment (SWIM) Plan for Lake Okeechobee. West Palm Beach.
Spooner, J. et al. 1991. NWQEP Annual Report Water Quality
Monitoring Report for Agricultural Nonpoint Source
Projects—Methods and Findings from the Rural Clean Water
Program (DRAFT). Nati. Water Qual. Eval. Proj., NCSU
Water Qual. Group, Biolog. Agric. Eng. Dep., N. Carolina
State Univ., Raleigh.
Stanley, J., V. Hoge, and L. Boggs. 1986. Taylor Creek/Nubbin
Slough RCWP 1986 Annual Report. Okeechobee Agric.
Stabil. Conserv. Serv., Okeechobee, FL.
Stanley, J., V. Hoge, L Boggs, and G. Ritter. 1988. Taylor
Creek/Nubbin Slough RCWP 1988 Annual Report.
Okeechobee Agric. Stabil. Conserv. Serv., Okeechobee, FL
Stanley, J.W. and B. Gunsalus. 1991. Ten Year Report Taylor
Creek/Nubbin Slough Project RCWP 1981-1991. Okeecho-
bee Agric. Stabil. Conserv. Serv., Okeechobee, FL.
Stewart, E.H., LH. Allen Jr., and D.V. Calvert 1978. Water Quality
of Streams on the Upper Taylor Creek Watershed,
Okeechobee County, Florida. Proc. Soil Crop Sci. Soc. Fla.
37:117-20.
Yates, P. et al. 1982. Channel Modifications on Taylor Creek Water-
shed. Proc. Am. Soc. Chem. Eng. Spec. Conf. Environ. Sound
Water Soil Manage., Orlando, FL.
13
-------
-------
Nutrient Loadings and Chlorophyll a
in the Oakwood Lakes System
David R. German
Water Resources Institute
South Dakota State University
Brookings, South Dakota
ABSTRACT
The amount of nutrients entering lakes has been shown to be related to productivity when large
groups of freshwater lakes are studied. The assumption is often made that this relationship holds
true for individual lakes. A project in South Dakota provided an opportunity to examine the
relationship between annual nutrient loadings and in-lake water quality. Nutrient, sediment and
hydrologic budgets for the Oakwood Lakes were determined in 1987,1988, and 1989. Chlorophyll
a, total phosphorus, organic nitrogen, and Secchi disk transparency were measured as indicators
of in-lake water quality. Low nutrient input in 1988 (.159 g/m2 total phosphorus in West Oakwood
and .087 g/m2 total phosphorus in East Oakwood) was associated with high chlorophyll a con-
centrations (330 mg/m3 peak in West Oakwood and 339 mg/m3 peak in East Oakwood). High
nutrient input to West Oakwood in 1989 (.464 g/m2 total phosphorus) was associated with much
lower chlorophyll a concentrations (80 mg/m3 peak in West Oakwood). Chlorophyll concentra-
tions were intermediate for both lakes in 1987 as were nutrient loads for West Oakwood (.213 g/m2
total phosphorus). East Oakwood, however, had high phosphorus loadings (.230 mg/m2) in 1987.
The Oakwood Lakes and Lake Poinsett water-
sheds in eastern South Dakota (Fig. 1) were
accepted into the Rural Clean Water Pro-
gram (RCWP) in 1981. The project was designed to
improve water quality through the implementation of
agricultural best management practices (BMPs)
based primarily on conservation tillage. In 1982 a
comprehensive monitoring and evaluation project
was approved to study the water quality impacts of
the Oakwood Lakes-Poinsett RCWP. The main focus
of the comprehensive monitoring and evaluation was
groundwater quality. The Oakwood Lakes System
Study (OLSS) was initiated in 1987 to determine if
the application of BMPs in agricultural watersheds
would affect lake water quality.
Study Area
The Oakwood Lakes are a complex of five small, in-
terconnected, shallow, hypereutrophic lakes in east-
ern South Dakota. The lakes have a mean depth of 2
m, a maximum depth of approximately 3 m, a com-
bined surface area of 971 ha and a total watershed
area of 12,793 ha. The lakes are fed by several inter-
mittent streams that drain a predominantly agricul-
tural watershed estimated to be 50 percent cropland.
The lake system has a single outflow from East Oak-
wood to the Big Sioux River (S.D. Dep. Environ. Nat.
Resour. 1985).
Nonbedrock aquifers in eastern South Dakota
consist of glacial outwash deposits resulting from the
Pleistocene epoch glaciation. Within the project area
surficial deposits are of the Wisconsin stage of
glaciation, specifically the Cary and lowan substage
till, loess, and outwash (Flint, 1955). The Big Sioux
aquifer is the surficial aquifer associated with the
Big Sioux River and its tributaries and is of greatest
importance in the project area. Seepage meters and
monitoring wells indicate that the Oakwood Lakes
are hydraulically connected to shallow unconfined
and confined aquifers located around the lake sys-
tem (U.S. Dep. Agric. et al. 1991).
15
-------
Proceedings of National RCWP Symposium, 1992
Legend
ua Highway
State Highway
Paved County Road
1234
Seal* In Miles
Agricultural Englnaorlng Dept.
South Dakota Stat* Unlvaralty
Unpaved Road
Figure 1.—Oakwood Lakes-Polnsett RCWP project area map
North
drawn by K. Elenkiwlch
South Dakota
16
-------
D.R. GERMAN
Monitoring Strategy
Monitoring before and after BMP implementation to
determine water quality effects was inappropriate
since BMP implementation preceded the OLSS by
several years. A strategy was developed that incor-
porated monitoring of water quality and quantity,
biological surveys, land use modeling, and lake
modeling to evaluate BMP impacts. The approach
relies on the development of nutrient, sediment, and
hydrologic budgets for the Oakwood lakes system
for comparison to reductions in loadings estimated
by watershed models. Predictions of water quality
improvements through BMP implementation were
based on in-lake water quality parameters.
OLSS monitoring included the direct measure-
ment of surface water and groundwater flux through
the lake system. Direct measurement of ground-
water flux through the Oakwood Lakes was con-
sidered to be an important part of the OLSS
monitoring strategy because of the presence of ex-
tensive sand and gravel aquifers adjacent to the
lakes.
In-lake monitoring was designed to determine
the current trophic state and to provide water quality
indicators such as chlorophyll a that could be
regressed against tributary nutrient loadings.
Zooplankton and fish populations, nutrient param-
eters, and weather were monitored to help explain
variability in such characteristics as chlorophyll a,
total phosphorus, and Secchi disk transparency.
Methods
Surface Water
Ten sites were instrumented with 13 stage recorders
and stilling wells to monitor surface water movement
through the Oakwood Lakes system (Fig. 2). At six
tributary sites and the outlet (T-0, 1, 2, 3, 4, 5, 6),
simple rating tables were developed by measuring
velocities with a flow meter, measuring the cross-sec-
tional area, >and calculating discharges for numerous
stage heights (Kennedy, 1984). Two stage recorders
and more complex equations were used at three
sites where water flows between lake basins (IL-1, 2,
3) (Fig. 2). The Manning formula was used when the
culverts were flowing partially full (Olson, 1967).
Q = 1.49/nRh2/3S1/2A
With the exception of the roughness coefficient
(n), the information required to calculate discharge
with the formula can be determined from stage
records. Channel slope S was represented by slope
of the water surface in partially submerged culverts.
Wetted cross-sectional area and hydraulic radius
were based on stage heights and a survey of the cul-
vert The'roughness coefficient was determined by
solving the formula for n, using measured values for
Q. For all sites stage records were digitized and con-
verted to discharge with equations derived based on
flow measurements taken at known stages.
Loadings were calculated by combining the con-
tinuous discharge record for each site with water
quality data. Discharge from each site was divided
into time intervals determined by,the date and time
when a sample had been taken. The water quality
sample was located at the midpoint of the time inter-
val. This method is patterned after the "trapezoidal
rule" proposed by Huber et al. (1979). Frequency of
sample collection varied and was determined by
hydrologic activity. Tributaries were grab sampled
except during storm events when they were sampled
by ISCO flow-actuated automatic samplers.
In-lake water quality samples were collected at
seven sites (L-l, 2,3,4,5, 6,7) every two weeks from
May through October and monthly from November
through April (Fig. 2). Samples were collected with
an integrated sampler designed by the project that
collected a column of water from the lake surface to
within 15 cm of the bottom. The sampler was essen-
tially a 5.08 cm inside-diameter tube that let water in
,as it was lowered into the lake; the tube was sealed
with a rubber stopper set in place by a tripping
mechanism when the sampler, touched bottom. The
sample was poured into a clean plastic bucket to en-
sure complete mixing of the water column from
which subsamples for water quality and chlorophyll
a analysis were taken. Secchi disk transparencies,
dissolved oxygen and temperature profiles, and ver-
tical zooplankton tows were taken at the same time.
Chemical analysis of all samples was conducted
by standard U.S. Environmental Protection Agency-
approved procedures (U.S. Environ. Prot. Agency,
1983). Parameters are listed in Table 1. In-lake
samples were also analyzed for chlorophyll a.
Chlorophyll a samples were filtered .through .45
micron glass fiber filters immediately upon arrival in
the laboratory. An ethanol extraction procedure
(Nusch, 1980) was used with absorbance measured
on a Perkin Elmer Model 35 Jr. spectrophotometer.
Groundwater
There were 18 existing monitoring wells (15 on com-
prehensive monitoring and evaluation field sites) ap-
plicable to OLSS project goals. To complete the
.coverage of the lakes, 26 additional groundwater
monitoring wells were installed (Fig. 3). This
17
-------
Proceedings of national RCWP Symposium, 1992
«5
0)
i
i
*
-------
D.R. GERMAN
Table 1.—Parameters and methods of analysis used for the Oakwood Lakes System Study.
PARAMETERS
Soluble reactive PO4-P
Total PO4-P
Nitrate-N
Nitrite-N
Ammonia-N
TKN
Suspended solids
YEAR APPROVED
1971
1971
1971
1971
1974
1978
1971
METHOD NUMBER
365.2
365.2
352.1
354.1
350.2
351.3
160.2
PROCEDURE
Colorimetric
Colorimetfic
Colorimetric-brucine
Spectrophotometric
Colorimetric-distillation
Colorimetric
Gravimetric 103-1 05° C
brought to 44 the total number of wells available for
analysis of flow direction and groundwater
chemistries.
Land wells were installed through a hollow stem
auger after advancing the auger to the proper depth.
At many well locations more than one well was in-
stalled (referred to as a well nest) to permit sampling
of water quality at different depths. The wells were
constructed of 5.08 cm diameter polyvinyl chloride
(PVC) pipe coupled to #18 slot well screen. Sand and
gravel were allowed to cave in around the well
screen during withdrawal of the auger, which after
development formed a natural gravel pack. When
caving did not occur, pea gravel was installed as a
gravel pack. If the borehole penetrated a confining
layer, bentonite was placed in the confining layer in-
terval to create a seal between the well screen and
upper saturated layers. A bentonite seal was always
placed in the top of the borehole around the well.
Single in-lake monitoring wells were installed to
determine groundwater chemistries within the lake
bottom. In-lake wells were constructed of 2.54 cm
diameter PVC pipe with a 15 cm slotted area at the
bottom, wrapped with fiberglass cloth (epoxied in
place) for a screen. Installation was accomplished by
driving a steel tube, with a steel drive point on the
end, into the lake bottom to the desired depth. The
well was inserted into the tube and the tube was
withdrawn, leaving the well and drive point in place.
Bentonite and a tripod support were placed around
the base of the well, at the lake bottom, to seal the
well and to give it strength, and also to protect the
bentonite. The support was weighted to the bottom
with large cobble-size rocks.
Seepage meters were placed on the lake bottom
(Fig. 3) to measure the volume of water that passes
through, be it in flux or out (Lee, 1977). They were
constructed of the upper or lower portion of a 208 L
(55 gallon) metal drum to which a partially filled pli-
able bag, such as a hospital IV bag, was attached
(Cherkauer and McBride, 1988).
Seepage meter readings were taken every two
weeks during the open water season if wind condi-
tions permitted. Seepage bags were filled with 110
mL of water and the volume was remeasured within
24 hours or less, so that back pressure bag did not
inhibit seepage.
Groundwater monitoring wells were sampled
with submersible bladder pumps or bailers depend-
ing on the site accessibility, well recharge capability,
and weather conditions. Samples were collected
every two months. In-lake wells were sampled every
two months except when the lake was frozen.
Samples were collected from the in-lake monitoring
wells with a peristaltic pump. Groundwater samples
were analyzed for nitrate, nitrite, organic nitrogen,
ammonia nitrogen, total dissolved solids, and total
dissolved phosphorus.
The goal of analysis and evaluation of ground-
water data was to determine the quantity and quality
of groundwater entering and exiting the Oakwood
Lakes system. Seepage rates were calculated at
every available measuring point. Areas of lake bot-
tom with similar seepage rates for selected time in-
tervals were delineated. The area was multiplied by
the seepage rate and the time interval to obtain a
volume. Volume was multiplied by a concentration
determined by available water quality data from in-
lake or land wells to obtain a loading for each area.
The loadings were summed to obtain a total loading
for the entire lake bottom for the year.
Lake Sediment
A study of phosphorus release from Oakwood Lakes
sediment was conducted as part of the OLSS (Price,
1990). As part of the study, 36 sediment cores were
obtained from Oakwood Lakes during March 1989
(Fig. 2). Sampling sites were selected based on a
nested experimental design of 4 cores within a site,
three sites within a basin, and three basins within
the lake system. Each lake basin was sampled along
a transect which extended from the shallow near-
shore area to the deep water area.
Intact sediment cores, approximately 0.91 m
long, were collected with a piston-type sediment
corer equipped with removable polycarbonate tubes
19
-------
Proceedings of National RCWP Symposium, 1992
o
c
1
O)
3
•a
o
Q.
n
o
0)
20
-------
D.R. GERMAN
(4.5 cm o.d. x 1.2 m). Polycarbonate was selected for
use because of its resistance to phosphate adsorp-
tion (Ryden et al. 1972).
Two cores from each site were chemically
analyzed to determine the phosphorus content of the
sediment. The cores were extruded from the
polycarbonate tubes while held in an upright posi-
tion to prevent mixing of the sediment layers. Each
core was divided into 15.2 cm sections for analysis,
except that the sufficial segment was further divided
into two 7.6 cm sections to increase resolution of
phosphorus content near the sediment water inter-
face. A brief description noting core length, sedi-
ment color, sediment texture, and presence of
organic matter was logged for each core. The phos-
phorus content of the sediment was determined by
the phosphorus extraction method of Aspila et al.
(1976). The remaining two cores from each site were
used in an incubation experiment.
To investigate the extent of phosphorus release
from the sediment to the overlying water during
aerobic and anaerobic conditions, two undisturbed
sediment cores from each site were incubated within
their original polycarbonate tubes, one under
aerobic conditions and one under anaerobic condi-
tions. Compressed air and nitrogen were bubbled in
the overlying water to maintain the aerobic and
anaerobic systems, respectively.
The overlying water was replaced with low
nutrient (< .02 ppm total P04-P) groundwater every 2
days during the 32-day incubation period. The re-
placement water was allowed to settle, decanted and
analyzed for total P04-P and ortho P04-P, before and
after incubation. The quantity of phosphorus re-
leased from the sediment was determined by mass
balance. A well adjacent to East Oakwood Lake was
used as a source of replacement water. The in-
cubated cores were sectioned and analyzed to deter-
mine the post-incubation phosphorus content of the
sediment after termination of the incubation experi-
ment.
Methods of statistical analysis used to evaluate
the results of the sediment study included Analysis
of Variance (ANOVA) and Least Significant Dif-
ference (LSD). Analysis was conducted using SAS
Institute, Inc., procedures, including PROC LEAPS
and PROC MEANS (SAS Inst. 1985).
Results
Chlorophyll a
Because of their importance in primary production,
measurements of photosynthetic pigments have
been widely used to quantify phytoplankton-standing
crops and to measure in-lake water quality.
Chlorophyll a is the pigment of choice since it is nor-
mally the most common (abundant) pigment in
living phytoplankton and has been studied extensive-
ly. In this study, chlorophyll a was measured primari-
ly to serve as a dependent variable in regression
analysis to determine the relationship between
tributary loadings and water quality.
Mean chlorophyll a concentrations for West
Oakwood Lake stations are presented in Figure 4.
Data for East Oakwood Lake is presented in Figure
5. Algal biomass as estimated by chlorophyll a con-
centrations were lower in 1987 and 1989 than in
1988. Peak chlorophyll a concentrations were ob-
served in late August or early September of all three
years.
Trophic State
Carlson (1977) proposed a Trophic State Index that
allows lakes to be placed on a scale of 0 to 100 based
on their trophic state, with 0 being the least produc-
tive. Each change of 10 in the scale represents ap-
proximately a doubling of the algal biomass for the
index, based on summer chlorophyll a levels.
Trophic State Indices (TSI) were calculated for total
phosphorus, chlorophyll a, and Secchi disk trans-
parency from measurements taken at seven in-lake
sites in each of ,the years 1987, 1988, and 1989
(Carlson, 1977). The mean TSIs for six sites in West
Oakwood and one site in East Oakwood is presented
in Table 2. TSIs of 80-90, indicating hypertrophy, are
common during the peak recreational season of July
and August. During spring and early summer, much
better water quality is often observed.
Seasonal TSI values for West Oakwood Lake sta-
tions (mean of 1-6) are plotted in Figure 6. TSI values
for East Oakwood Lake (station L-7) are plotted in
Figure 7. The values for total phosphorus tend to
fluctuate less than for the other parameters. Large
fluctuations in chlorophyll a TSIs are observed in
the spring of each year. Apparently enough phos-
phorus is available to support a larger amount of
algal biomass, as estimated by chlorophyll a, than
was realized. Carlson (1977) reported a similar
phenomenon for several Minnesota lakes that
resulted from springtime crashes in the algae
population. Many factors can cause low algae popula-
tions. In the Oakwood Lakes, light limitation is un-
likely because of extended day length and high
transparency during periods with low chlorophyll a
(Figs. 6,and 7). Nutrient limitation is a possibility, but
on dates with low chlorophyll a and high transparen-
cy (May 27, 1987; May 24, 1988; and June 6, 1989)
only a few stations show any indication of nutrient
limitation (Table 3): On May 24,1988, at station L-7,
21
-------
Proceedings of National RCWP Symposium, 1992
0^03/87 ' 11/19/87 ' 66/06/8812/23/88 ' 07/11/89
08/11/87 02/27/88 09/14/88 04/02/89 10/19/89
Date
01/27/90
Figure 4.—Chlorophyll a concentrations In West Oakwood Lake (mean of six stations) from 1987 through 1989.
05)03/87' 11/19/87 ' 06/06/88 12/23/88 07/11/89 01/27/90
08/11/87 02/27/88 09/14/88 04/02/89 10/19/89
Date
Figure 5.—Chlorophyll a concentrations in Eastwood Lake (station L-7) from 1987 through 1989.
during the most transparent period observed,
nutrient limitation was not indicated. Temperature
change cannot be ruled out as a factor, since the
lakes warm rapidly during the spring period.
Zooplankton grazing has been associated with high
transparency in other lakes in the upper Midwest
(Carpenter et al. 1985; Shapiro and Wright, 1984)
and may contribute to low TSI values for chlorophyll
a and transparency in the Oakwood Lakes. Other un-
known factors may also contribute to this
phenomenon.
In August and September of all three years, TSIs
for Secchi disk transparency and chlorophyll a ex-
ceed the TSI value for phosphorus. Late in the sum-
mer, more chlorophyll a is present than would be
indicated by the amount of total phosphorus present.
This indicates that the lake may be phosphorus-
limited during this time of the year. The data is con-
sistent with high nitrogen-phosphorus ratios indi-
cating phosphorus limitation late in the year for 1988
and 1989, and to a lesser extent for 1987 (Table 3).
22
-------
D.R. GERMAN
Table 2.—Trophic state Indices for East and West Oakwood Lakes.
EAST OAKWOOD
WEST OAKWOOD
DATE
April 28, 1987
May 14, 1987
May 27, 1987
June 9, 1987
June 23, 1987
July6, 1987
July 20, 1987
August 4, 1987
August 17, 1987
September 9, 1987
September 23, 1 987
April 19, 1988
May 11, 1988
May 24, 1988
JuneS, 1988
June 20, 1988
Julys, 1988
July 19, 1988
August 1 , 1 988
August 15, 1988
Septembers, 1988
September 27, 1988
May 10, 1989
May 23, 1989
June6, 1989
June 21, 1989
Julys, 1989
July 17, 1989
August 1, 1989
August 15, 1989
Septembers, 1989
September 26, 1 989
TSI (TP)
71
69
74
74
76
74
76
81
77
82
82
—
65
66
61
65
66
70
77
81
77
76
71
71
64
70
67
72
76
71
73
75
TSI (SD)
65
65
64
67
70
72
73
77
82
85
91
—
70
47
62
60
80
83
89
95
85
80
68
62
49
55
59
69
70
76
79
81
TSI (CHL)
—
—
__
—
—
77
81
86
90
90
87
---- -
64
32
70
64
80
89
93
97
88
78
72
69
. 39
58
63
83
77
78
86
84
TSI (TP)
77
79
82
81
78
73
74
80
81
82
82
77
73
73 :,
73
74
70
75
81
84
84
82
70
70
69
71
76
74
75
72
81
68
TSI (SD)
72
72
75
76
72
71
75
78
82
86
'88
69
74
66
73
72
80
82
90
98
99
91
56
52
56
65
73
72
71
76
86
77
TSI (CHL)
74
79
86
89
90
85
79
72
63
74
75
81
86
92
96
96
89
59
53
64
71
75
79
76
78
88
76
100
(855
01/23/87 , 08/11/87 02/27/88 09/14/88 04/02/89 10/19/89
05/03/87 11/19/87 06/06/88 12/23/88 07/11/89
Figure 6.—Trophic state indices for West Oakwood Lake (mean of six stations) from 1987 through 1989.
~~ ~~~~ 23 ' : ~ - ~
-------
Proceedings of National RCWP Symposium, 1992
100
03$>3/87 ' 08/11/87 ' 02/27/88 09/14/88 ' 04/02/89 ' 10/19/89
05/03/87 11/19/87 06/06/88 12/23/88 07/11/89
Figure 7.—Trophic state Indices for East Oakwood Lake (station L-7) from 1987 through 1989.
Table 3.—Inorganic nitrogen to inorganic phosphorus ratios for Oakwood Lake stations for 1987 to 1989.
JOHNSON LAKE JOHNSON LAKE
DATE (L-1) (L-2)
Jan. 22, 1987
Apr. 12, 1987
Apr. 28, 1987
May 14, 1987
May 27, 1987
June 9, 1987
June 23, 1987
July 6, 1987
July 20, 1987
Aug. 4, 1987
Aug. 17, 1987
Sept. 9, 1987
Sept. 23, 1987
Oct. 21, 1987
Nov. 18, 1987
Jan. 18, 1988
Mar. 1, 1988
Apr. 19, 1988
May 11, 1988
May 24, 1988
June 8, 1988
June 20, 1988
Julys, 1988
July 19. 1988
Aug. 1,1988
Aug. 15, 1988
Sept. 6, 1988
Sept. 27, 1988
Oct. 25, 1988
May 10. 1989
May 23, 1989
June 6, 1989
June 21, 1989
July 5, 1989
July 17, 1989
Aug. 1,1989
Aug. 15. 1989
Sept. 5, 1989
Sept. 26, 1989
Oct. 31, 1989
6
57
14
3
2
4
20
2
No data
1
10
13
22
5
18
37
15
48
188
11
99
30
9
4
9
38
162
4
22
13
3
1
3
56
32
61
3
No data
1
5
8
10
9
4
26
206
21
52
179
26
32
15
4
27
26
115
W. OAKWOOD
LAKE (L-3)
2
8
21
2
3
3
14
162
21
4
147
2
No data
1
7
14
13
8
8
16
216
254
225
87
195
90
6
5
19
16
203
W. OAKWOOD
LAKE (L-4)
3
10
5
1
4
21
206
30
255
3
7
1
3
3
12
14
5
166
41
19
373
217
56
35
103
13
4
14
39
105
ROUND LAKE ROUND LAKE
(L-7) (I-6)
2
6
2
4
84
72
3-
-
3
OQ
29
2
10
19
24
26
21
177
249
106
63
•t An
149
139
30
7
4
11
77
59
3
27
8
19
39
31
27
No data
208
44
15
13
19
17
152
198
50
4
4
21
23
116
E. OAKWOOD
LAKE (L-7)
oo
23
37
4
4
5
276
18
135
3
7
i
3
4
10
116
84
387
295
10
No data
2
21
17
299
24
-------
D.R. GERMAN
Nutrient, Sediment, and Hydraulic
Budgets
Data produced by the OLSS monitoring program
permitted calculation of the volume of water and the
mass of nutrients and suspended sediment entering
the Oakwood Lakes system from several tributaries
and groundwater in 1987,1988, and 1989. Tributary
contributions of nutrients and sediments to the Oak-
wood Lakes system are presented in Table 4. Mill
Creek (T-0) represents the amount of material ex-
ported by the system through the outlet. Water flow-
ing through the outlet to Mill Creek exhibited lower
concentrations of nutrients than water flowing into
the lakes from the tributaries.
Hydraulic loadings from most tributaries were
highest in 1987 and were lower in 1988 and 1989.
Loadings for 1987 and 1988 were comprised primari-
ly of snowmelt runoff; in 1989, in addition to snow-
melt, three major storm events in June and July
resulted in significant runoff in several tributaries.
West Creek contributed the most water to the sys-
tem in all three years of the study.
There was less transfer of water between basins
were high at the beginning of 1987 due to several
preceding wet years, and the large flows between
lakes resulted from the lakes draining to normal
levels. Below normal precipitation beginning in 1987
caused lake levels to fall below outlet elevations at
the beginning of 1988 and 1989. Part of the tributary
contribution was needed to fill the basins before out-
flow could begin in 1988. The culverts at Kimball's
Crossing were also partially blocked for most of 1988
and 1989 because of local concern over falling water
levels, thus causing less water to flow into East Oak-
wood Lake from West Oakwood Lake. No water left
the system through the outlet in 1989, so 100 percent
of the hydraulic, nutrient, and sediment load from
surface water was retained, minus the evaporation
that occurred from the water surface.
A summation of seepage through of the lake bot-
tom and estimates of loadings for 1987 through 1989
are presented in Table 5. The data for 1988 and 1989
are superior, since no wells and fewer seepage
meters were installed in 1987. The groundwater con-
tribution to the Oakwood Lakes system was ap-
proximately one-half of the combined tributary and
groundwater loadings. The estimates of ground-
water flow into the lake are probably not as reliable
in each successive year of the study. Lake levels
Table 4.—Total contribution of water, suspended solids, total phosphorus, and total nitrogen for tributary and in-
terlake sites in 1987,1988, and 1989, with flow-weighted mean concentrations in parts per million (ppm).
SITE
Pony Creek (T-3)
Mud Creek (T-4)
Mortimer's Cross-
ing (IL-3)
West Creek (T-1)
Crazy Creek (T-2)
Kimballs Crossing
(IL-1)
Round Crossing
(IL-2)
Loomis Creek
(T-5)
Goose Creek (T-6)
Mill Creek (T-0)
YEAR
1987
1988
1989
1987
1988
1989
1987
1988
1989
1987
1988
1989
1987
1988
1989
1987
1988
1989
1987
1988
1989
1987
1988
1989
1987
1988
1989
1987
1988
1989
DISCHARGE
(m3)
164,195
48,605
22,806
608,421
121
145,814
2,714,332
1,344,522
24
2,714,332
1,344,522
2,389,397
1,064,402
697,858
749,177
4,249,942
1 ,078,553
744,736
1,390,906
90,671
5,798
128,790
23,932
51,812
65,382
19,491
89,561
5,328,618
1,686,111
0
• SUSPENDED SOLIDS
LOADING (kg)
950.4
746.3
234.3
8,026.0
371.9
1,691.0
9,684.1
3,038.4
-366.1
43,301.2
33,165.4
149,400.9
41,675.1
10,840.1
4,575.9
66,887.4
24,994.3
26,985.6
24,798.0
2,524.1
238.5
2,105.7
810.9
6,724.4
1,509.4
327.6
2,771.4
55,581.3
31,041.3
0.0
MEAN (ppm)
5.8
15.4
10.3
13.2
2.5
11.6
15.3
26.9
12.4
16.0
24.7
62.5
39.0
15.5
6.1
15.7
23.2
36.2
17.8
27.8
40.9
16.3
33.9
129.8
23.1
16.8
31.0
10.4
18.4
0.0
TOTAL PHOSPHORUS
LOADING (kg)
96.7
32.6
22.4
263.1
71.1
137.6
106.0
39.5
20.6
553.5
403.1
1 ,625.5
339.4
306.4
532.8
517.1
263.1
133.9
172.3
11.6
0.9
158.6
45.0
182.7
7.8
4.6
27.0
427.9
200.7
0.0
MEAN (ppm)
0.59
0.67
0.98
0.43
0.48
0.94
0.17
0.35
0.70
0.20
0.30
0.68
0.32
0.44
0.71
0.12
0.24
0.18
0.12
0.13
0.15
1.23
1.88
3.53
0.12
0.24
0.30
0.08
0.12
0.00
TOTAL NITROGEN
LOADING (kg)
453.9
110.6
66.9
2,204.8
283.6
823.7
1,197.1
495.8
175.4
6,893.3
3,056.6
7,766.5
3,590.2
1,533.1
2,855.1
6,607.5
3,231.8
2,037.0
2,261.6
230.7
17.9
676.8
174.7
529.0
192.4
66.2
359.2
7,560.9
2,859.0
0.0
MEAN (ppm)
2.76
2.28
2.94
3.62
1.90
5.65
1.89
4.40
5.92
2.54
2.27
3.25
3.36
2.20
3.81
1.55
3.00
2.74
1.63
2.54
3.08
5.26
7.30
10.21
2.94
3.40
4.01
1.42
1.70
0.00
25
-------
Proceedings of National RCWP Symposium, 1992
Table 5.—Groundwater discharge and loadings estimates of total dissolved phosphorus and total dissolved
nitrogen to the Oakwood Lakes in 1987-89. Percent of the combined tributary and groundwater loadings that
YEAR
1987
1988
1989
E. OAKWOOD
(m3}
1,733,944
2.139,096
2,876.800
W. OAKWOOD
(m*)
1,411,260
2,050,275
3,824,200
GROUNDWATER DISCHARGE
(m3)
3,185,204
4,189,371
6,701,020
(40%)
(65%)
(66%)
TOTAL PHOSPHORUS
LOADING (kg)
79.4
102.3
164.4
(05%)
(11%)
(06%)
TOTAL NITROGEN
LOADING (kg)
1 ,853.3
2,252.9
3,717.9
(12%)
(30%)
(23%)
as those calculated for the tributaries, but a substan-
tial portion of the hydraulic budget of the Oakwood
Lakes is contributed by groundwater.
Phosphorus Budget
All basins had a net gain of total PO/i-P during the
three years of the study. The basins were most effi-
cient at trapping total PCU-P in 1989, with ranges
from 87 percent to 100 percent The lake system as a
whole trapped 70 percent, 80 percent and 100 per-
cent of the total P04-P that was contributed by six
monitored tributaries in 1987, 1988, and 1989,
respectively. The Oakwood Lakes system operated
as a phosphorus sink, even though a net loss of
water occurred in West Oakwood Lake during 1987
and in East Oakwood Lake during 1988.
The total P04-P load to the lake system carried
by six monitored tributaries was 1,419 kg in 1987,
993 kg in 1988, and 2,528 kg in 1989. To estimate the
water quality effect on the lake, loading per square
meter of lake surface was calculated. This repre-
sented an annual tributary total PO4-P load to the
Oakwood Lake system of 0.16 g/m2/year in 1987,
0.11 g/m2/year in 1988, and 0.28 g/m2/year in 1989.
West Oakwood Lake received a tributary total PO4-P
load equal to 0.21 g/m2/year in 1987, 0.16
g/m2/year in 1988, and 0.46 g/m2/year in 1989. East
Oakwood Lake received a tributary total P04-P load
equal to 0.230 g/m2/year in 1987, 0.012 g/m2/year
in 1988, and 0.093 g/m2/year in 1989 (Fig. 8). Vollen-
weider (1968) would consider this a dangerous load
(i.e., one that would cause rapid eutrophication) for
lakes much deeper than the Oakwood Lakes (per-
missible load varies directly with a lake's mean
depth). A total P04-P loading of 0.13 g/m /year for a
lake of 5 m mean depth is considered dangerous. In
relation to this standard, West Oakwood Lake (maxi-
mum depth = 3.5 m) received excessive tributary
total P04-P loads in 1987. and 1989, and East Oak-
wood Lake (maximum depth = 3.2 m) received an ex-
cessive load in 1987.
Nitrogen Budget : .
Unlike phosphorus, nitrogen may undergo denitri-
fication and be lost to the system as it is returned to
the atmosphere. Nitrogen may also be added to the
1987 1988 1989
1987 1988 1989
Figure 8.—Tributary phosphorus loadings to the Oakwood Lakes in 1987,1988, and 1989.
26
-------
D.R. GERMAN
system through nitrogen fixation by several species
of blue-green algae. No attempt was made to account
for nitrogen lost to denitrification or .gained by
nitrogen fixation.
Large differences in total nitrogen as N loading
between years was observed for most tributaries
(Table 4). The smallest contributions occurred
during 1988 for all tributaries, except Pony Creek,
which had a smaller contribution in 1989. Highest
contributions of total nitrogen as N for most
tributaries occurred in 1987. The large load of total
nitrogen as N carried by West Creek in 1989 was due
primarily to the July 11 storm.
All basins retained nitrogen in 1988 except
Mortimer's Slough, which had a net loss of 102 kg of
total nitrogen as N. Although the lake system was
operating as a nitrogen sink, nitrogen was trapped
less efficiently than phosphorus by the lake system.
The lake system as a whole trapped 46 percent of the
tributary load in 1987, 44 percent in 1988 and 100
percent in 1989. An accumulation of nitrogen over
time will occur if this condition persists, unless
denitrification exceeds nitrogen fixation or nitrogen
is lost through sedimentation.
The largest tributaries (Mud Creek, West Creek,
and Crazy Creek) contributed the largest amounts of
total nitrogen in all three years. The total nitrogen as
N load to the lake system carried by six monitored
tributaries was 14,016 kg in 1987, 5,095 kg in 1988,
and 12,400 kg in 1989. This represents an annual
tributary total nitrogen as N load to the Oakwood
Lake^ system of 1.57 g/m2/year in 1987, 0.57
g/mVyear in 1988, and 1.39 g/m2/year in 1989.
West Oakwood Lake received a total nitrogen as N
load of 2.8 g/m2/year in 1987, 1.4 g/mVyear in
1988, and 2.8 g/m2/year in 1989. East Oakwood
Lake received a tributary total nitrogen as N load of
3.0 g/m2/year in 1987, 1.3 g/m2/year in 1988, and
1.2 g/m2/year in 1989 (Fig. 9). As with phosphorus,
this represents a load that would lead to rapid
eutrophication for lakes as shallow as the Oakwood
Lakes (Vbllenweider, 1968).
Groundwater contributions of total phosphorus
to the phosphorus budget are insignificant when
compared to tributary contributions. Groundwater
contributed 5 percent, 11 percent, and 6 percent of
the combined tributary and groundwater phos-
phorus loadings in 1987,1988, and 1989, respectively
(Table 5).
Groundwater contributions to the nitrogen
budget are larger relative to phosphorus loads.
Groundwater nitrogen comprised 12 percent, 30 per-
cent, and 23 percent of the combined tributary and
groundwater total nitrogen loads in 1987, 1988, and
1989, respectively (Table 5).
Lake Sediment Study
At the time of the lake sediment sampling, the dis-
solved oxygen concentration in all lakes decreased
with water depth to < 1 ppm, 15 cm from the sedi-
ment surface. The low dissolved oxygen concentra-
tions near the lake bottom may. have allowed a
release of phosphorus from the sediment to the
water column prior to the collection of the sediment
cores. If so, the amount of phosphorus per gram of
1987 1988 1989
1987 1988 1989
Year
Figure 9.—Tributary nitrogen loadings to the Oakwood Lakes in 1987,1988, and 1989.
27
-------
Proceedings of National RCWP Symposium, 1992
sediment, and subsequently, the quantity available
for release to the overlying water under anaerobic
conditions, may have been underestimated by this
study.
The overlying water of all anaerobically in-
cubated cores experienced a net gain of total P04-P,
indicating phosphorus was released from the sedi-
ment (Fig. 10). A smaller net gain of total P04-P oc-
curred in the water of the aerobically incubated
cores. Two cores from West Oakwood Lake ex-
perienced a net loss of total P04-P in the overlying
water, which indicates that the sediment had
removed phosphorus from the water column of
these cores, resulting in the negative aerobic treat-
ment mean.
The LSD analysis of the treatment means shows
that the anaerobically incubated cores released sig-
nificantly more total P04-P (0.305 g/m2) than the
aerobically incubated cores (0.080 g/m2). Based on
a muck-type sediment surface area of 7.12 x 10 m
(85 percent of the lake bottom), the potential sedi-
ment phosphorus load to the Oakwood Lakes sys-
tem is 2,041 kg total P04-P during anaerobic
conditions, and 468 kg total P04-P during aerobic
conditions. These figures are based on the total avail-
able phosphorus that was released during the 32-day
incubation period, although most was released in the
first 3 days (Price, 1990). ANOVA indicated that the
release of total PO4-P varied significantly between
lakes for both aerobic and anaerobic incubation at
the 0.03 significance level.
0.5n
A comparison of the observed tributary phos-
phorus load and the calculated potential sediment
phosphorus load to West Oakwood Lake and East
Oakwood Lake is presented in Figure 11. The sedi-
ment load during anaerobic conditions was equal to
or greater than the tributary loadings of phosphorus
in East Oakwood Lake and, during 1987 and 1988,
for West Oakwood Lake. During aerobic conditions
the sediment load of phosphorus was substantially
less than the tributary loads for both years.
Discussion
When large numbers of lakes worldwide have been
studied, it has been shown that the amount of
nutrients entering a lake is directly related to
productivity within the lake (Schindler, 1978). The
assumption was made at the beginning of this study
that this relationship would hold true for the Oak-
wood Lakes. It was assumed that determining the
mathematical nature of the relationship between
tributary loadings and productivity in the lake would
allow estimates of water quality improvement that
could be attributed to reductions in loadings through
BMP implementations.
To develop a useful mathematical model, it was
desirable that differences in both tributary loadings
and in-lake water quality be observed in the three
years of study. Indeed, this was the case for both
West and East Oakwood Lakes. Tributary phos-
JOHNSON
W OAKWOOD E OAKWOOD
LAKE
ANAEROBIC
AEROBIC
Figure 10.—Flux of phosphorus between sediment cores and overlying water.
28
-------
D.R. GEKMAtt
a.
4
O
Q.
-3
•5
Tributary
Sediment
Tributary Sediment
EAST OAKWOOD
WEST OAKWOOD
Figure 11.—Comparison of observed tributary loadings and calculated potential sediment phosphorus loadings.
phorus loadings in 1989 were nearly three times
higher than in 1988 for West Oakwood Lake (Fig. 8).
East Oakwood Lake received its highest tributary
phosphorus loadings in 1987. Phosphorus loadings
in 1988 and 1989 -were much lower. A similar pattern
was observed for tributary nitrogen loads (Fig. 9).
Simple linear regression was used to determine
the relationship between annual tributary phos-
phorus loadings and mean summer chlorophyll a.
The predictive model that resulted for West Oak-
wood yielded an R2 of 0.71, with a negative slope
(Fig. 12). Taken at face value, this would indicate
that if loadings increased to over 0.65 g/m2 of phos-
phorus over the surface area of West Oakwood,
algae in the lake would be eliminated.
The predictive model for East Oakwood indi-
cated no relationship between phosphorus loadings
and algae populations (Fig. 13). In fact, there ap-
pears to be no relationship between annual phos-
phorus loadings and water quality in either of the
Oakwood Lakes. This holds true when tributary
nitrogen loading is regressed against chlorophyll a.
Adding groundwater loadings does not improve the
predictive equations.
During periods in 1987 and 1988, when tributary
loadings had ended completely, large increases of
phosphorus were observed in the water column.
This was especially true for West Oakwood Lake in
July and August of 1987 and 1988 (Fig. 14). The large
increase in in-lake phosphorus concentrations sug-
gest that internal loading of sediment phosphorus
may be the primary source of phosphorus during
periods when there are no tributary inputs, indicat-
ing a substantial release of phosphorus from the lake
sediments. Although the pulse of phosphorus is
short-lived and phosphorus returns to the sediment
with the fall decline in the algae population, it occurs
during the peak of the recreation season.
The sediments represent the historical nutrient
and sediment loads to the Oakwood Lakes. The
lakes have the ability to support large algae blooms
in years when tributary loadings are quite low. Ap-
parently water quality in the lake is more a function
of .sediment phosphorus release than annual
tributary loadings.
The primary objective of the OLSS was to deter-
mine if BMPs were likely to have an impact on water
quality in the Oakwood Lakes. BMPs have not yet
reduced nutrient loadings to acceptable levels or re-
stored desirable water quality to the Oakwood lakes.
Large reductions in tributary may be needed to have
any impact on water quality in the Oakwood Lakes.
The sediments are laden with nutrients and the trap-
ping efficiency of the lake system is very high.
Without inactivation or removal of the sediment-
stored phosphorus, the lakes are likely to remain
self-sufficient in phosphorus.
Recommendations
Findings of the Oakwood study may prove useful to
nonpoint source projects in the planning stage or just
underway. New projects should recognize that if
their goal is to improve water quality in lakes with
29
-------
Proceedings of National RCWP Symposium, 1992
220
200-
180-
T= 160-|
?
= 140-
<0
JJ, 120-1
• 100-1
60H
40
20
2 _
= .71
3.15
02 025 03 035 0.4 0.45 0.5
Total Phosphorus Loadings g/ma/year
Tributary Loadings Only
Figure 12.—Linear regression of tributary phosphorus loadings versus In-lake chlorophyll a for West Oakwood Lake.
220-
o>
200-
180-
160-
140-
o
§
60
40
20
R2 =.02
3.05
Oil
0.15
0.2
0.25
Total Phosphorus Loadings g/m2/year
Tributary Loadings Only
Figure 13.—Linear regression of tributary phosphorus loadings versus In-lake chlorophyll a for East Oakwood Lake.
algae blooms, internal sources of nutrients may
prevent them from reaching this objective through
watershed treatment alone. This likelihood may
seem obvious to experienced limnologists, but it
does not seem to be widely considered when new
nonpoint source projects are selected. The assump-
tion is often made that reducing nutrient loadings
from the watershed will automatically result in im-
proved in-lake water quality. This may not be true for
lakes similar to the Oakwood Lakes. The danger ,is
that if large numbers of nonpoint source projects fail
to improve water quality through watershed treat-
ment, public support for such projects may decline.
Therefore, internal nutrient loadings should be con-
sidered when new nonpoint source projects are
selected for funding; lakes that are thought to have
30
-------
D.R. GERMAN
U. 0.26-
~ 0.24-
g 0.22-
Ł 0.2-
| 0.18-
z °-16"
8 0.14-
Ł °-12-
^ 0.1-
0.08-
i
/!
n
/:
N :"""' \
1 \ ; . \ A
/ \ / ! /
/ \ I \ i
/ \ I \ /
, / I / \ ' /
k u i ;
» i
\
;\
/ \
; \
: \
i \
•.
\
\ *
\ \
\ f
\ 'A/ ' "
L W
V, rLf
HljJ L_,
-1 5
•0.9 S.
-0.8?
-«^
-0.6 J,
-0.5 U!
-0.4 <Ł
OL
-0.3 >-
QC
-0.2 jS
-0.1 m
Jun-88 Dec-88
Jan-87 Aug-87 Feb-88 Sep-88 Apr-89
LAKE
TRIBUTARY I
Figure 14.—A plot of in-lake phosphorus concentrations and tributary phosphorus loadings for West Oakwood Lake
during 1987 and 1988.
considerable internal loadings should be given a low
priority to receive funding for watershed treatment,
if this is the only treatment planned.
Lake projects that include a component that is
likely to reduce or control the effects of internal
loadings, or that target lakes that have minimal inter-
nal loadings and show a good relationship between
tributary loadings and water quality, should be given
a high priority for funding to control nonpoint source
pollution. Projects of this nature are more likely to
succeed and contribute to the body of evidence that
nonpoint source pollution control is worth the invest-
ment.
References
Aspila, K.I., H. Agemian, and A.S.Y. Chau. 1976. A semi-automated
method for the determination of inorganic, organic and total
phosphate in sediments. Analyst 101:187-97.
Carlson, R E. 1977. A trophic state index for lakes. Limnol.
Oceanogr. 23(2):361-69.
Carpenter, S.R., J.F. Kitchell, and J.R. Hodgson. 1985. Cascading
trophic interactions and lake productivity. BioScience 35:634-
39.
Cherkauer, D. A. and J.M. McBride. 1988. A remotely operated
seepage meter for use in large lakes and rivers. J. Ground
Water Sci. Eng. 26(2):165-71.
Flint, R.E 1955. Pleistocene geology of eastern South Dakota. Prof.
Pap. 262. U.S. Geo. Surv., Reston, VA.
Huber, W.C. et al. 1979. Urban-rainfall-quality database update
with statistical analysis. EPA-600/8-79-004. U.S. Environ.
Prot Agency, Washington, DC.
Kennedy, EJ. 1984. Discharge ratings at gaging stations. Book 3,
Ch. A10 in Techniques of Water-Resource Investigations.
U.S. Geo. Surv., Reston, VA.
Lee, D.R. 1977. A device for measuring flux in lakes and estuaries.
Limnol. Qceanogr. 22:14047.
Nusch, E.A. 1980. Comparison of different methods for
chlorophyll and phaeopigment determination. Ergebn. Lim-
nol. (Suppl. to Arch Hydrobiol.) 14:14-36.
Olson, R.M. 1967. Engineering Fluid Mechanics. 2nd ed. Interna-
tional Textbook Company, Scranton, PA.
Price, M.B. 1990. Sediment as a source of phosphorus in a hyper-
eutrophic prairie lake. M.S. thesis. South Dakota State Univ.,
Brookings.
Ryden, J.C., J.K. Syers, and R.F. Harris. 1972. Sorption of inorganic
phosphate by laboratory ware. Implications in environmental
phosphorus techniques. Analyst 97:903-08.
SAS Institute, Inc. 1985. SAS User's Guide: Statistics, Ver. 5 ed.
SAS Inst., Cary, NC.
Schindler, D.W. 1978. Factors regulating phytoplankton produc-
tion and standing crop in the world's freshwaters. Limnol.
Oceanogr. 23:478-86.
Shapiro, J. and D.I. Wright 1984. Lake restoration by biomanipula-
tion, Round Lake, Minnesota—first two years. Freshw. Biol.
14:371-83.
South Dakota Department of Environment and Natural Resources.
1985. Lake Poinsett/Oakwood Lakes Water Quality Study
Area Report. S.D. Dep. Environ. Nat Resbur., Pierre.
U.S. Department of Agriculture et al. 1991. South Dakota Oak-
wood Lakes-Poinsett Rural Clean Water Program 10-year
Report 1981-1991. Pierre, SD.
U.S. Environmental Protection Agency. 1983. Methods for Chemi-
cal Analysis of Water and Wastes. Off. Water, Washington,
DC.
Vollenweider, RA 1968. Scientific fundamentals of the eutrophica-
tion of lakes and flowing waters with particular reference to
nitrogen and phosphorus eutrophication factors. Tech. Rep.
DAS/CSI/68.27. Off. Econ. Coop. Develop., Paris.
31
-------
-------
Nitrate and Pesticide Occurrence in
Shallow Groundwater During the
Oakwood Lakes-Poinsett RCWP Project
Jeanne Goodman
South Dakota Department of Environment and Natural Resources
Pierre, South Dakota
J. Michael Collins and Keith B. Rapp
Delta Environmental Consultants, Inc.
St. Paul, Minnesota
ABSTRACT
From 1984 through 1990, shallow groundwater was monitored at six farmed fields and one un-
farmed site in eastern South Dakota as part of the Oakwood Lakes-Poinsett Rural Clean Water Pro-
gram comprehensive monitoring and evaluation project. The implementation of best management
practices (BMPs) began concurrently with the initiation of monitoring sites with BMPs, sites
without BMPs, and an unfarmed site. Groundwater samples, collected at varying frequencies from
project monitoring wells, were analyzed for nitrate and pesticides (herbicides and insecticides).
Nitrate concentrations above the U.S. Environmental Protection Agency's maximum contaminant
level of 10 ppm were found in 15 percent of the groundwater samples. Based on crops grown, the
dominant controls on nitrate concentrations in the groundwater appear to be precipitation and in-
filtration, denitrification, and fertilizer applications. Tillage practices did not appear to affect nitrate
concentrations in groundwater. Pesticides were detected in 11 percent of the groundwater samples.
Lasso (alachlor), 2,4-D, and Banvel (dicamba) were the most frequently detected pesticides. Al-
though each field site exhibited detectable pesticide concentrations in the groundwater, con-
centrated leaching of pesticides resulting in sustained groundwater contamination did not appear
to be a problem in the RCWP project area. Several recommendations for future land treatment and
water quality projects in areas of vulnerable groundwater resources were developed based on the
study results.
The Oakwood Lakes-Poinsett Rural Clean
Water Program (RCWP) project was part of
a U.S. Department of Agriculture experimen-
tal program aimed at protecting and improving water
quality in areas with identified water quality
problems. The Oakwood Lakes-Poinsett project area
is located in the glacial lakes region of east-central
South Dakota (Fig. 1). Dryland, row-crop agriculture
is the predominant land use, and corn and soybeans
are the major crops produced, with alfalfa and small
grains less prevalent. Commercial fertilizers and pes-
ticides, primarily herbicides, are commonly used in
the project area with application rates varying from
no application to 100 pounds of fertilizer per acre
(averaging 40 pounds per acre) and pesticides ap-
plied according to label recommendations (U.S. Dep.
Agric.etal. 1991).
The glaciated Oakwood Lakes-Poinsett area is
characterized by numerous lakes, wetlands, and
shallow sand and gravel aquifers. Nearly level to
severely undulating topography occurred in col-
lapsed glacial drift, which consists of deposits of
33
-------
Proceedings of National RCWP Symposium, 1992
1234
Scale in Miles
Figure 1.—Groundwater monitoring field sites and master site location.
sand, gravel, and silly-clay till. The major soil as-
sociations are deep silty, loamy, well-drained soils
(U.S. Dep. Agric. et al. 1991). The average annual
precipitation in the project area is approximately 22
inches; actual annual precipitation levels for the
study years are given in Table 1.
Water quality problems identified in the project
area were (1) excess nutrients in the tributaries and
lakes and (2) evidence of excess nitrate concentra-
tions in the Big Sioux aquifer. The Big Sioux aquifer
is a surficial sand and gravel aquifer ranging from 5
to 80 feet in saturated thickness, with coarse-
34
-------
J. GOODMAN, J.M. COLLfttS, & K.B. RAPP
Table 1 .—Annual precipitation of project area.
YEAR
AMOUNT (IN INCHES)
1982
1983
1984
1985
1986
1987
1988
1989
1990
27.67
22.20
31.35
28.98
27.00
19.92
17.92
16.06
24.47
textured overlying soil (S.D. Dep. Water Nat. Resour.
1984).
The goals of the 10-year Oakwood Lakes-Poin-
sett RCWP project, implemented from 1981 through
1991, were to improve and protect groundwater and
surface water by reducing nonpoint source pollution
from total nitrogen, pesticides, water and sediment-
borne contaminants, and animal waste. Among best
management practices (BMPs), conservation tillage
and fertilizer and pesticide management were em-
phasized, with a secondary emphasis on animal
waste management systems. Fertilizer management
included annual soil testing in the spring so that fer-
tility recommendations could be based on residual
soil nutrients and yield goals. Pesticide management
consisted of weed and pest scouting so that recom-
mendations for chemical use could be based on
treating infested areas only. Conservation tillage,
defined as 30 percent residue at planting before crop
emergence, was typically achieved by not using
moldboard plowing.
The goal of the comprehensive monitoring and
evaluation portion of the project was to monitor and
evaluate the effects on groundwater and surface
water from the implementation of agricultural BMPs.
Specifically, monitoring was conducted to determine
if the BMPs recommended to prevent surface water
pollution would change the amount of nitrogen and
pesticide input to the groundwater.
Monitoring Strategy
The groundwater monitoring approach was site-
specific because the project area (106,000 acres) was
too large to monitor as a single unit and the prob-
ability of detecting land-use-affected-changes in
water quality was increased. The seven monitoring
sites, called field sites, ranged from 10 to 80 acres in
size. Six sites were farmed fields, and one site (OP
site) had not been farmed since the early 1970s. Con-
servation tillage, fertilizer management, and pes-
ticide management BMPs were implemented on five
of the farmed sites; the sixth site (JW site) was a con-
trol site without these BMPs. An additional farmed
site (VC site) reverted to conventional tillage
(moldboard plow) from conservation tillage in 1987.
Locations of the field sites are shown in Figure 1.
Agricultural chemical inputs (in the form of fer-
tilizers and pesticides) and tillage practices were
documented for each field site. Also, field site
operators were interviewed to collect historical land
use patterns. In general, the BL, LK, and VC sites
were planted in corn, small grain, and soybeans;
corn and small grain were grown at the JW site;
soybeans and wheat were rotated at the PH site; and
corn and soybeans were rotated at the RS site (U.S.
Dep.Agric. etal. 1991).
Groundwater monitoring began at field sites in
1984 and continued through 1990. The groundwater
was monitored at 60 locations with 114 monitoring
wells installed with a hollow flight auger drill rig and
constructed of 2-inch polyvinyl chloride (PVC)
casing with 3-foot or 5-foot, 0.018-inch commercial
slot PVC screens. Well depths ranged from 7 to 66
feet and averaged 20 feet. All casing sections were
joined with threaded , couplers. The monitoring
design included well nests and wells placed across
the field sites to identify each site's water quality and
hydrogeology.
Groundwater sampling was conducted at all field
sites at a frequency described in Table 2. Quarterly
and bimonthly samples were analyzed for nitrate,
nitrite, total Kjeldahl nitrogen, ammonia, total dis-
solved solids, pH, electrical conductivity, and dis-
solved oxygen. In addition, monthly and biweekly
sample analyses included a gas chromatograph scan
for the pesticides listed in Table 3.
Table 2.—Sampling frequency for field site ground-
water sampling.
WELLS
All wells
All wells
Wells for pesticide
analysis
Wells for pesticide
analysis
FREQUENCY
Quarterly
Every two
months
Monthly
Every two
weeks
TIME FRAME
1984 to Aug. 1989
Aug. 1989 to Dec. 1990
1984 to Dec. 1990
June/July 1989
June/July 1990
Table 3.—Pesticides tested with scan.
Ambien (chloramben)
Banvel (dicamba)
Counter (terbufos)
Dyfonate (fonofos)
Ramrod (propachlor)
Tordon (picloram)
Treflan (trifluralin)
Tbxaphene
Methoxychlor
Prowl (pendimethalin)
Dursban (chlorpyrifos)
Atrazine
Bladex (cyanazine)
Dual (metolachlor)
Endrin
Lindane
Parathion
Sencor (metribuzine)
Thimet (phorate)
2,4-D
Lasso (alachlor)
35
-------
Proceedings of national RCWP Symposium, 1992
Table 4.—Geozone abbreviations and definitions.
ABBREVIATION DEFINITION
WTLT15 Weathered till (brown color) or silty clay (reworked till) with the screened interval of the well at a depth b.g.s. of
less than or equal to 15 ft. (19 wells)
WTGT15 Weathered till or transition zone (greenish brown zone interpreted as a transition zone between the weathered
and unweathered till) with the screened interval at a depth b.g.s. of greater than 15 ft. (11 wells)
UT
Unweathered till (gray color). (7 wells)
SC
Sllty clay aquitard located between an upper and a lower aquifier system. (2 wells)
SS-A
Alternating layers of thinly bedded fine sand and silt. (4 wells)
SQLT5LT10 Sand and gravel with less than or equal to 5 ft of overlying soil material, with the screened interval less than or
equal to 10 ft below the water table. (30 wells)
SQLT5QT10 Sand and gravel with less than or equal to 5 ft of overlying soil material, with the screened interval less than or
equal to 10 ft below the water table. (15 wells) '
SG5-15LT10 Sand and gravel with less than or equal to 5 ft and less than or equal to 15 ft (4.6 m) of overlying soil material,
with the screened interval less than or equal to 10 ft below the water table. (2 wells)
SG5-15GT10 Sand and gravel with less than or equal to 5 ft and less than or equal to 15 ft (4.6 m) of overlying soil material,
with the screened interval less than or equal to 10 ft below the water table. (3 wells)
SGGT15
Sand and gravel with greater than 15 ft of overlying soil material. (5 wells)
SG-UA
Sand and gravel located under and aquitard as the lower unit of a two-aquifier system. (16 wells)
Note: Overlying soil material refers to all silt- and clay-rich sediments overlying a sand and gravel layer. It includes silt and/or clay loams and, in
some cases, glacial till.
Mots: Whlli It may appear that the SGGT15 and SG-UA may both apply to a monitoring well, SG-UA specifies an aquitard between two sand
and gravel units that both have some portion of their saturated thickness. SGGT15 only implies that there Is 15 feet of burden above the
mentioned sand and gravel unit If there Is an upper sand unit, it was not saturated.
To reduce scatter in the water quality data, a
classification method was developed to identify the
well screen's location in relation to the geology that
was sampled. Table 4 lists the geozones, the number
of monitoring wells in each geozone, and the ab-
breviation used. Depth at which the well was
screened is expressed in two ways: (1) depth below
ground surface (depth b.g.s.) for wells screened in
glacial till; and (2) depth below the water table
(depth b.w.t.) for wells screened in sand and gravel.
Figure 2 illustrates the geozone abbreviations
displayed on a diagrammatic cross section. The
schematic cross section is not specific for a single
site but represents a general composite of the
stratigraphic sequences found throughout the
project area. The water table in the diagram is not in-
tended to suggest flow directions, only the typical
relative position of the water table within the
geozones.
Figure 2.—Schematic cross section illustrating the geozones for Oakwoods Lake-Poinsett RCWP project.
36
-------
J. GOODMAN, J.M. COLLINS, &K.B. RAPP
Well placement and distribution within the
geozones were targeted toward shallower depths.
Most monitoring wells penetrated saturated condi-
tions at 10 feet or less below the ground surface.
Monitoring wells predominantly monitored sand and
gravel outwash at the PH, LK, JW, and VC sites and
till at the RS and BL sites. The OP site was charac-
terized by areas of till and outwash deposits.
To help interpret field site data, an agricultural
chemical leaching study was conducted at the
master site (Fig. 1). Water, nutrient, and pesticide
movement through the soil profile was monitored on
plots of no-till and moldboard plow tillages with corn
and oats in rotation. Fertilizer and pesticide inputs
were strictly controlled.
Nitrate Occurrence
A total of 3,092 groundwater samples were collected
from the monitoring wells between January 1984 and
December 1990 for analysis of nitrate as nitrogen
(NOs-N). All field sites exhibited groundwater
samples with NOs-N concentrations greater than 10
parts per million (ppm), which is the U.S. Environ-
mental Protection Agency's (EPA's) maximum con-
taminant level for NOs-N in drinking water, in at least
one well. Less than 15 percent of all nitrate detec-
tions, however, were in concentrations greater than
10 ppm.
Table 5 lists the statistical NOs-N data for each
geozone. Wells with the highest median NOs-N con-
centrations were monitoring the shallow geozones:
alternating sand and silt layers (SS-A), shallow sand
and gravel (SGLT5LT10), silty clay (SC), and shal-
low till (WTLT15). Median NOs-N concentrations
are reported because testing indicated that the
NOs-N data were not normally distributed (SAS
Inst, 1985; Crawford, 1984). It is important to note
that the silty clay geozone, which had the highest
median NOs-N concentration, appeared hydraulical-
Table 5.—Geozone statistical data.
ly similar to the shallow till and that the median NOs-
N concentration is based on samples collected from
only two wells. The median NOs-N concentrations
from the deep geozones were 2 to 20 times lower
than those from the shallow geozones.
NOs-N concentrations from the 3,092 samples
from 106 monitoring wells were plotted versus depth
of sample point below the water table (Fig. 3). With
few exceptions, NOs-N concentrations above 5 ppm
were not detected at depths greater than 20 feet
below the water table, and NOs-N concentrations
greater than 0.2 ppm were not observed at depths
greater than 30 feet below the water table. The plot
also illustrates that NOs-N concentrations exceeded
50 ppm only twice.
Commercial fertilizer has been used in the
project area for over 20 years; therefore, nitrate has
had sufficient time to migrate to depths greater than
20 feet below the water table (S.D. Dep. Environ.
Nat. Resour. 1989), based on vertical hydraulic
gradients and hydraulic conductivities measured at
the field sites. NOs-N concentrations of up to 75 ppm
in the groundwater also illustrate the amount of
nitrogen that is available to migrate to the
groundwater. It appears that the nitrate has not
traveled to greater depths because of denitrification,
the process by which biological organisms transform
nitrate or nitrite to gaseous nitrogen by using the
oxygen present in the nitrate for metabolism.
Denitrification has been shown to occur in shal-
low sand, groundwater environments (Trudell et al.
1986) similar to the glacial aquifers in eastern South
Dakota. Denitrification has been shown to take
place in low oxygen environments even though it is
an anaerobic process (Focht and Verstraete, 1977).
During the RCWP study period, the dissolved
oxygen concentration in the groundwater decreased
with depth below the water table in a manner similar
to the NOs-N concentrations.
BMPs were implemented on the field sites
before monitoring, so an experiment designed to
GEZONE
SS-A
SG5-15GT10
SGLT5LT10
SGGT15
SG-UA
SGLT5GT10
SC
SG5-15LT10
WTLT15
WTGT15
UT
NUMBER OF
WELLS
4
30
15
2
3
5
2
16
19
11
7
N-SPECIES
NOs-N
NO3-N
NOs-N
NOs-N
NOs-N
NOs-N
NOs-N
NOs-N
NOs-N
NOs-N
NOs-N
MINIMUM
0.02
0.00
0.00
0.00
0.00
0.00
1.60
0.00
0.00
0.08
0.02
MAXIMUM
26.20
4.05
75.25
4.07
34.91
28.54
23.50
0.20
54.75
28.00
15.26
MEDIAN
5.25
0.65
4.32
0.03
0.14
0.29
7.18
0.05
7.03
2.20
0.37
NUMBER OF
SAMPLES
147
57
869
730
433
427
43
64
504
338
185
MEAN
6.78
0.92
6.97
0.24
4.64
3.29
10.08
0.05
9.96
6.32
0.56
Note: Concentrations are in parts per million (ppm) of water, which is equivalent to milligrams per liter (mg/L).
37
-------
Proceedings of National RCWP Symposium, 1992
20 feet below the water table
-50
10
20
30 40 50
NO3-N Concentrations (ppm)
60
70
80
Figure 3.—NO3-N (ppm) versus depth below water table: all sites.
compare before and after BMPs was not possible. To
distinguish effects that could be attributed to BMPs,
evaluation of the differences in water quality result-
ing from BMP implementation was conducted by
making comparisons between fields with BMPs,
fields without BMPs, and the unfarmed field site.
Using the Wilcoxan Rank Sum testing (SAS Inst.
1985), significantly lower median NOs-N concentra-
tions were found at the unfarmed site than at the
farmed sites (S.D. Dep. Environ. Nat. Resour. 1989).
No significant differences in median NOs-N con-
centrations were found between field sites with con-
ventional tillage and conservation tillage (S.D. Dep.
Environ. Nat. Resour. 1989). Differences in tillage
practices did not appear to affect groundwater
quality, either positively or adversely, during the
study period at the field sites. However, the vadose
zone monitoring at the master site measured a faster
rate of movement of rainwater from the soil surface
to the groundwater for the no-till tillage plots than for
the moldboard plow plots. Higher quantities of water
and NOs-N were delivered to the 6-foot depth under
no-till plots (U.S. Dep. Agric. et al. 1991).
Median nitrate concentrations were plotted by
year for each site (Fig. 4). The plots in Figure 4 il-
lustrate several occurrences at the field sites. First,
the plots indicate the median NOs-N concentrations
decreased in 1988 and 1989 and increased slightly in
1990. Second, the median NOs-N concentrations in-
creased between 1984 and 1987 at five sites. Above
normal precipitation from 1984 through 1987, below
normal precipitation during 1988 and 1989, and a
return to above normal precipitation levels in 1990,
correlate with decreases and increases in NOs-N
concentrations during this period with three excep-
tions:
• an increase in NOs-N concentrations occurred
at the OP site in 1988 and 1989 following a
field-controlled burn of the vegetation at the
site;
• a decrease in the concentrations at the JW site
appears biased by the installation of additional
monitoring wells at the site with lower con-
centrations of NOs-N; and
• a high median NOs-N concentration of 7 ppm
at the RS site in 1984 follows corn production
at the site with the highest recorded fertilizer
application rate of 100 pounds per acre in
1983.
Nitrogen applied to the land surface reaches the
groundwater through precipitation and groundwater
recharge events; therefore, it follows that less NOs-N
reaches the groundwater during periods of low
precipitation. It appears that nitrogen applied to the
land surface that is not used in crop uptake during
periods of low precipitation remains in the soil
profile until it is available for transport during in-
creased rainfall periods.
38
-------
J. GOODMAN, J.M. COLLJNS, &K.B. RAPP
85
86 87 88 89 90
YEAR
84 85 86 87 88 89 90
YEAR
84 85 86 87 88 89 90
YEAR
84 85 86 87 88 89 90
YEAR
84 85 86 87 88 89 90
YEAR
84 85 86 87 88 89 90
Figure 4.—Median NO3-N concentrations and nitrogen applications for seven field sites.
39
-------
Proceedings of National RCWP Symposium, 1992
The 1990 precipitation measurements, individual
monitoring well hydrographs, and NOs-N concentra-
tions were evaluated to determine the effect of
precipitation on NOs-N concentrations. Several
major storm events occurred between May 1 and
July 31, 1990, with a total of over 12 inches of rain
received in the area. During these major precipita-
tion events, the NOs-N concentrations in most of the
wells studied decreased initially within days of the
storm event because of the dilution effects of the
recharge. NOs-N concentrations then began to in-
crease gradually and approach the prestorm event
levels. For example, at the PH site, the NOs-N con-
centration decreased from 9 ppm in May to 4 ppm in
July, which was followed by an increase to over 6
ppm by the end of August.
The VC field site was chosen for further analysis
of groundwater NOs-N concentrations because the
site changed from a conservation tillage site to a con-
ventionally tilled site (moldboard plow used every
year) in 1987. Seven wells were sampled at the field
site; three of the shallowest wells were sampled
monthly. All monitoring wells were screened in the
shallow sand and gravel geo2ones between 12 and 30
feet below the land surface.
The site was planted to corn in 1988 and divided
into two areas planted to corn and soybeans in 1989
and 1990. Twenty-eight pounds of nitrogen fertilizer
were applied per acre of corn; soybeans were not fer-
tilized. Two monitoring wells were located in the
area of corn production in 1988 and then planted to
soybeans in 1989 and 1990. In one well, NOs-N con-
centrations decreased from a high of 0.1 ppm in May
1988 to a low of 0.01 ppm in August 1989 to non-
detectable in June 1990. In the second well, NOs-N
concentrations decreased from 15.65 ppm in August
1988 to 0.36 ppm in August 1989 and June 1990.
These changes appeared to be related to the change
in cropping patterns — corn in 1988 and soybeans in
1989 and 1990. The up-gradient monitoring well at
the site indicated no groundwater quality influence
from off-site activities. The lack of fertilizer inputs to
the soybeans during 1989 and 1990 appears to be the
major factor contributing to the decrease in NOs-N
concentrations in the monitoring wells. It appears
that the change in BMPs had not affected the
median NOs-N concentrations at this site, and the
dominant control appears to be the amount of fer-
tilizer applied to the fields.
Pesticide Occurrence
The pesticide groundwater sampling program was
initiated in May 1984 and continued through Decem-
ber 1990. The term pesticide in the study included
both herbicides and insecticides.
The frequency of groundwater sampling for pes-
ticides increased each year in the study from
quarterly sampling in 1984 to a monthly or biweekly
groundwater sampling program thereafter. The num-
ber of groundwater samples collected for pesticides
over the study period are presented in Table 6. The
pesticide sampling program consisted of 1,628
samples collected from 73 wells. Detectable pes-
ticide concentrations occurred in 11.3 percent of the
samples (184 of 1,628 samples had detections; in 26
samples, more than one pesticide was detected). As
the program progressed, the frequency of
groundwater sampling and the number of detections
increased. However, the data suggest that the num-
ber of detections per sample collected remained con-
sistent during the RCWP sampling period.
Table 6.—Number of groundwater samples collected
for pesticide analysis.
YEAR
SAMPLES COLLECTED
1984
1985
1986
1987
1988
1989
1990
Total
55
125
161
219
277
305
486
1,628
Groundwater samples were collected from wells
ranging in depth from 8.5 to 65.5 feet. The average
depth of wells in which pesticides were detected was
17.7 feet. Figure 5 presents the sampling frequency
of the geozones, and Figure 6 depicts the frequency
of geozones when pesticides were detected. The
shallow sand and gravel geozone (SGLT5LT10) was
sampled most frequently, and the geozone with the
most frequent pesticide detections was the shallow
till (WTLT15).
The majority of pesticide detections in the
groundwater sampling program were represented
by a single detection of a single pesticide. During
subsequent sampling, the pesticide was not
detected. This type of "hit-and-miss" detection of pes-
ticides appears to be common for groundwater
monitoring at agricultural sites (S.D. Dep. Environ.
Nat. Resour. 1992). Commonly, the pesticide
detected was applied to the field more than one year
before sampling and was accompanied by the detec-
tion of a pesticide with no application history to the
field.
Because of sporadic, irregular detection of pes-
ticides, long-term or continued trends of pesticides
leaching to groundwater with BMPs were difficult to
ascertain. However, distinct pesticide leaching
events were clearly evident at all sites based on one
40
-------
J. GOODMAN, J.M. COLLINS. &K.B. RAPP
SG-UA (8.2%)
•WTGT15 (9.0%)
GT15 (0.2%)
SGLT5LT10 (41.8%)
LT15 (20.3%)
SGLT5GT10 (6.1%)
UT (4.9%)
SG5-15GT10 & SC (1.0%)
SG5-15LT10 (4.0%)
SS-A (4.5%)
Figure 5.—Percentage of pesticide samples collected by geozone.
UT (5.7%)
SGLT5LT10 (28.6%)
•SS-A (6.7%)
SGLT5GT10 (4.3%)
SG5-15LT10 (3.3%)
SG-UA (8.1%)
WTGT15(11.9%;
Figure 6.—Percentage of pesticide detections oy geozone.
or more wells detecting one or more pesticides
during a single groundwater sampling event. For ex-
ample, at the RS site on July 10,1990, samples from
three shallow wells contained a range of Banvel
(dicamba) from 0.04 to 0.83 parts per billion (ppb),
and samples from two deep wells contained the same
chemical at a range of 0.02 to 0.09 ppb. Banvel
(dicamba) was applied to the field approximately one
month before sampling, and a major precipitation
event occurred between its application and sam-
pling.
The pesticides applied to the field sites are
presented in Table 7. Lasso (alachlor) was the most
commonly detected pesticide, followed by 2,4-D,
Banvel (dicamba), and Sencor (metribuzin) (Fig. 7).
The detection of pesticides was primarily seasonal in
nature, because most detections clustered in the late
spring and summer months (Fig. 8). A variable in the
^-WTLT15(31.4%)
detection of pesticides in groundwater appears to be
the timing of pesticide application in relation to the
occurrence, duration, and intensity of precipitation
events. Undoubtedly, the topsoil and unsaturated
zone play the most important roles in the storage of
pesticide residues. The timing, release, and trans-
port of pesticides to the groundwater occur as the
soil moisture is replenished, and entrained pes-
ticides migrate through unsaturated flow to the
water table.
Table 7.—Pesticides used on field sites.
Banvel (dicamba)
Basagran
Furadan (carbofuran)
Lasso (alachlor)
MCPA
Classic (chlorimuron)
Roundup (glyphosate)
Ramrod (propachlor)
Sencor (metribuzin)
Treflan (trifluralin)
2,4-D
Dual (metolachlor)
41
-------
Proceedings of National RCWP Symposium, 1992
Dyfonate (3.8%)
Sencor(14.8%)
2,4-D (16.2%)
Treflan (3.3%)
Undane(1.0%)
Banvel (15.7%)
Dual (6.2%)
Atrazine (1.4%)
^
Lasso (29.5%)
Rgure 7.—Percent of total pesticide detections.
Jun (14.8%)
Tordon (1.4%)
Parathion (5.2%)
Bladex (1.4%)
May (17.6%)
Jul (26.7%)
Apr (4.8%)
Mar (1.0%)
I—Feb (1.9%)
•Jan (3.3%)
Dec (7.1%)
Aug (10.0%)
Figure 8.—Months when pesticides were detected.
Three pesticides correlated positively with ap-
plication on field sites and the subsequent detection
of the pesticide in groundwater (Table 8). These pes-
ticides were Lasso (alachlor), 2,4-D, and Banvel
(dicamba). These three compounds accounted for
more than 61 percent of the pesticides detected. The
frequent detection of pesticides that have no history
of use at the site on which they were detected
(parathion and Sencor) account for greater than 20
percent of detections. These detections could be a
result of improper reporting of applications by the
operators, atmospheric transport of chemicals, con-
taminated application equipment, pesticides in
groundwater or surface water that flows into or onto
the site; or they may result from the extended time
span during which the chemicals were entrained in
the unsaturated zone.
Overall, concentrations of pesticides detected
during the Oakwood Lakes-Poinsett RCWP study
were extremely low (Fig. 9). Continual leaching of
Sep (7.€
-Nov (1.4%)
-Oct (3.8%)
pesticides resulting in sustained groundwater con-
tamination does not appear to be a problem at any
field sites in the Oakwood Lakes-Poinsett RCWP
project area. Therefore, BMPs for pesticides applied
at label rates did not create a long-term negative ef-
fect on groundwater quality in this study area.
Conclusions and
Recommendations
NOs-N was found in excess of 10 ppm in at least one
monitoring well at each field site, but in total, 15 per-
cent of the samples exceeded EPA's maximum con-
taminant level for NOs-N in drinking water. With few
exceptions, NOs-N concentrations greater than 5
ppm were not found at depths greater than 20 feet
below the water table. Nitrate concentrations in the
groundwater were affected by infiltration events and
crop rotations that included crops requiring higher
42
-------
J. GOODMAN, J.M. COLL1HS, & K.B. RAPP
Table 8.—Pesticide applications and detection.
Ambien
Atrazine
Banvel
Basagran*
Bladex
Classic*
Counter
Dual
Dyfonate
Eradicane
Furadan
Lasso
MCPA*
Ramrod
Roundup*
Sencor
Sutan
Thimet
Tordon
Treflan
endrin
lindane
methoxychlor
parathion
toxaphene
2,4-D
BL
site
A
X
X
X
X
X
X
X
Hi
i«:
D
X
X
X
X
X
X
«i
mm.
%$m
;:::wxx
^i«fci^
m®
lp::!i
•!:;SyĄ;l
x'A;:;:!:;:
A - Pesticide applied at site
D - Pesticide detected at site
* - Pesticides not included in analytical scan
nitrogen application. The three apparent dominant
controls on NOs-N concentrations in the ground-
water were
• the amount of nitrogen applied to the land
surface,
• the volume of infiltration, and
• denitrification in the aquifer.
Projectwide, the pesticide sampling program
resulted in 11.3 percent of the groundwater samples
with detectable concentrations of pesticides. Lasso
(alachlor), 2,4-D, and Banvel (dicamba) were the
most frequently used chemicals on the field sites and
were the most frequently detected chemicals in the
groundwater. Several chemicals were detected in the
groundwater samples from field sites with no history
of application at the sites. Most detections of pes-
ticides were one time occurrences with no detec-
tions of the same chemical in the same well in
subsequent sampling events. The application of a
chemical and the subsequent detection of the chemi-
cal in the groundwater were commonly separated by
up to a year's time. This data clearly indicated that
pesticide residues may remain in the soil profile for
an extended period, tied up in the soil matrix and
released during infiltration events rather than
43
-------
Proceedings of National RCWP Symposium, 1992
Lasso
2,4-D
Banvel
Sencor
Dual
Parathlon
Dyfonate
Treflan
Atrazlno
Bladox
Tordon
Undane
0.1
0.2 0.3
Median Concentration (ug/l)
0.4
0.5
Figure 9.—Median pesticide concentrations: all sites.
moving to the groundwater shortly after application.
Although some detections of Atrazine and Lasso
(alachlor) were in concentrations at or above EPA's
maximum contaminant levels or health advisories,
overall pesticide concentrations were extremely low.
No significant differences in groundwater
quality were found between field sites with conserva-
tion tillage practices and field sites with conventional
tillage; however, at research plots in the study area,
vadose zone monitoring measured differences in the
rate of water movement and quantities of water and
NOa-N to the 6-foot depth between no-till plots and
moldboard plow plots. NOs-N concentrations were
significantly higher at the farmed field sites than at
the unfarmed site.
Water quality monitoring began concurrently
with BMP implementation. Therefore, it was difficult
to establish a positive correlation between BMPs
and water quality impacts. It is evident, however, that
after groundwater monitoring began, the overall
change in groundwater quality attributable to
agricultural chemicals was negligible. It appears that
label rate pesticide application does not have a sus-
tained long-term negative effect on groundwater
quality.
For future land treatment and water quality
projects and in areas defined as vulnerable areas for
groundwater quality, the following are recom-
mended:
1. Fertilizer management practices must ad-
dress timing, rate, and method of application
to minimize the amount of nitrogen available
to leach by maximizing the amount of fer-
tilizer uptake by the crop.
2. Crop rotations that include crops requiring
low to no nitrogen should be used.
3. Pesticide management, especially in areas of
vulnerable water supplies, should be used,
based on the detection of several pesticides in
the groundwater after application to the land
surface.
4. Water supply wells should be constructed so
that the screened portion of the well does not
produce water from the upper portion of the
water-bearing materials where high NOs-N
concentrations can occur.
44
-------
J. GOODMAN, J.M. COLL/NS, & K.B. RAPP
References
Crawford, C.G. 1984. Application of Selected Nonparametric
Statistical Methods to the Analysis of Hydrologic Data. Prep.
for Statistical Analysis of Water Quality Data (G0062). Un-
publ. doc. U.S. Geo. Surv., Washington, DC.
Focht, D.D. and W. Verstraete. 1977. Biochemical ecology of
nitrification and denitrification. Adv. Microbial Ecol. 1:135-
214.
SAS Institute Inc. 1985. SAS User's Guide: Statistics, Version 5.
Gary, NC.
South Dakota Department of Environment and Natural Resources.
1987. Annual RCWP Progress Report — Project 20. Open
File Rep. Pierre, SD.
. 1989. Annual RCWP Progress Report — Project 20. Open
File Rep. Pierre, SD.
. 1992.1991 Pesticide and Nitrate Sampling Program. Open
File Rep. Pierre, SD.
South Dakota Department of Water and Natural Resources. 1984.
The Big Sioux Aquifer Water Quality Study. Pierre, SD.
Trudell, MR., R.W. Gillham, and JA Cherry. 1986. An in-situ
study of the occurrence and rate of denitrification in a shal-
low unconfined sand aquifer. J. Hydrol. 83:251-68.
U.S. Department of Agriculture; South Dakota Department of En-
vironment of Water and Natural Resources; and Brookings,
Kingsbury, and Hamlin County Conservation Districts and
Agricultural Stabilization Conservation Service County Com-
mittees. 1991. Ten-year Report: Oakwood Lakes-Poinsett
Rural Clean Water Program, 1981-1991. Open File Rep.
Huron, SD.
45
-------
-------
Water Quality Trends in the
St. Albans Bay, Vermont, Watershed
Following RCWP Land Treatment
Donald W. Meals
School of Natural Resources
University of Vermont
Burlington, Vermont
ABSTRACT
The St. Albans Bay Rural Clean Water Program project sought to improve water quality in
eutrophic St. Albans Bay and its tributaries by implementing improved management practices to
control nonpoint sources of pollution from dairy agriculture. The goal of comprehensive monitor-
ing and evaluation was to evaluate water quality changes in the bay and its tributaries resulting
from the land treatment program. Despite a significant reduction in point source phosphorus load,
St. Albans Bay water quality did not improve significantly, probably because of internal phosphorus
loading. Sediment concentration and export decreased significantly in most tributaries. Phos-
phorus concentrations increased in most monitored streams, but export did not change significant-
ly. Nitrogen concentrations and export increased across the watershed. Indicator bacteria counts
in all monitored streams, however, declined by 50 to 70 percent, suggesting that the land treatment
program did affect water quality. Because these bacteria are not readily stored in the environment,
their rapid decline may represent the leading edge of greater water quality improvements to come
as current nutrient losses from agricultural land decline and nutrients historically accumulated in
the system are gradually flushed out.
St. Albans Bay on Lake Champlain in Vermont
has been subjected to increasing rates of
eutrophication because of excessive phos-
phorus loads from both point and nonpoint sources.
The primary goal of the St. Albans Bay Rural Clean
Water Program (RCWP) project was to improve
water quality and restore beneficial uses in the bay
and its tributaries by implementing a program of
agricultural best management practices (BMPs) to
control nonpoint sources of pollution. Begun in 1980
and completed in 1991, the St. Albans Bay RCWP
project was one of five nationwide to include com-
prehensive monitoring and evaluation (CM&E) to
assess water quality changes in response to land
treatment.
The effectiveness of watershed land treatment
programs in improving water quality has not been
well documented (Braden and Uchtmann, 1985).
Few of the early watershed agricultural nonpoint
source control programs were able to document
water quality improvements following intensive im-
plementation of BMPs (Harbridge House, Inc.,
1983). This lack of response has been attributed to
the incremental process of land treatment, short
duration of monitoring, and difficulty in controlling
for natural and cultural variability (Morrison and
Lake, 1981; Persson et al. 1983). Recently, however,
improvements in water quality have been reported in
some agricultural nonpoint source projects. Under
the RCWP, projects in Idaho, Florida, and elsewhere
47
-------
Proceedings of National RCWP Symposium, 1992
have begun to show water quality changes following
land treatment (Soil Conserv. Serv. et al. 1991; Flaig
and Ritter, 1989; Natl. Water Qual. Eval. Proj. 1988).
This paper presents the principal findings of the
CM&E effort in the St. Albans Bay Watershed
RCWP project
Study Area
The St. Albans Bay watershed is located in
northwestern Vermont 40 km north of Burlington
and drains 13,000 ha of agricultural, forested, and
urban land into St. Albans Bay of Lake Champlain.
The 700 ha bay is 4.2 km long and 2.2 km wide.
Mean depth is 8 m, maximum depth is 12 m. Four
major tributaries drain the watershed into St Albans
Bay (Fig. 1):
• Jewett Brook (median flow = 0.03 m3/sec);
o
• Stevens Brook (median flow = 0.16 m /sec);
• Rugg Brook (median flow = 0.19 m3/sec); and
• Mill River (median flow - 0.46 m3/sec).
The city of St Albans tertiary wastewater treatment
plant discharges about 91 mYsec to Stevens Brook
wetland at the head of the bay.
The St. Albans Bay watershed is primarily in the
Champlain Lowlands, an area of generally low relief
between Lake Champlain and the foothills of the
Green Mountains. Watershed soils formed on glacial
till or lacustrine deposits and include loams (51 per-
cent) (half of which are poorly drained); silts and
clays (27 percent); rock outcrop (15 percent); and
sands (7 percent). The climate is cool and humid,
with pronounced seasonal variations. Mean annual
temperature is 7.3*C, and the growing season
averages 150 days. Mean annual precipitation is 845
mm, with a minimum average monthly precipitation
in February of 44 mm and a maximum average
monthly rainfall in August of 100 mm. Average an-
nual snowfall is 1,550 mm.
Sixty-five percent of the land in the watershed is
devoted to agriculture, about 20 percent is forested,
and 10 percent is urban and residential. Of the 102
farms in the watershed, dairy farms predominate
and average 134 ha, with a mean herd size in 1991 of
110 animal units, up from 95 animal units in 1980.
Corn for silage is the principal cultivated crop, with
land in corn ranging from 1,336 to 1,821 ha or about
10 to 15 percent of the total watershed. Substantially
more agricultural land is devoted to hay: 3,845 to
4,452 ha or about 30 to 35 percent of the watershed.
Some cropland in the watershed has been in con-
tinuous corn cultivation; a three-year corn/five-year
hay rotation is the prevalent practice. Both manure
and commercial fertilizer are commonly applied to
corn cropland; hayland is generally manured twice a
year, immediately following haying.
Recreation, aesthetics, and other beneficial uses
of St. Albans Bay have been impaired by eutrophica-
tion for many years. The most serious nonpoint
source water quality problems associated with
agriculture in the St. Albans Bay watershed have
been nutrients (principally phosphorus) from runoff
of animal and milkhouse waste and sediment from
cropland erosion. Improper manure management —
year-round 'spreading because of lack of waste
storage — and inadequate milkhouse waste and
barnyard runoff management were thought to be
the principal contributors to water quality problems.
Improper fertilizer management — failure to balance
manure and fertilizer nutrient applications with soil
and crop needs — was also thought to be a sig-
nificant contributor to excessive nutrient loading.
Erosion of cropland soil and streambanks also con-
tributed to the decline in water quality.
Methods
Water Quality Monitoring
Four levels of water quality monitoring were con-
ducted to document water quality changes in St. Al-
bans Bay and its tributaries:
• Level 1, bay sampling,
• Level 2, tributary trend monitoring,
• Level 3, an edge-of-field BMP study, and
• Level 4, randomized grab sampling.
The design and specific methods of these and other
elements of the CM&E program, including biologi-
cal and land use monitoring, have been discussed in
detail elsewhere (Clausen, 1985; Vt. RCWP Coord.
Comm. 1991; Meals, 1992). Only Level 1 and 2
monitoring will be discussed in this paper.
• Level 1. St Albans Bay was monitored by four
Level 1 stations sampled 20 times each year (Fig. 1).
Grab samples were collected at Stations 11 (outer
bay), 12 (inner bay), and 14 (off-bridge) at a 0.5 m
depth and 1.0 m from the bottom using a Kemmerer
sampler. Samples were analyzed for turbidity, total
suspended and volatile suspended solids, total and
soluble reactive phosphorus, total Kjeldahl and am-
monia nitrogen, and chlorophyll a. In situ measure-
ments were made of temperature, dissolved oxygen,
pH, conductivity, and Secchi disk trans- parency. Sur-
face samples collected at Station 13, located just off-
shore of the former public beach, were analyzed for
48
-------
D.W. MEALS
CHAMPLAIN
LEGEND
• level 1
A level 2
A level 3
A level 4
@ precipitation
project boundary
2 miles
SCALE
Figure 1.—Map of St. Albans Bay watershed, showing location of monitoring stations.
fecal coliform and fecal streptococcus bacteria. Addi- • Level 2. Level 2 tributary trend monitoring was
tional biological monitoring in the bay included an- designed to determine long-term patterns of con-
nual macrophyte surveys by aerial photography as centration, streamflow, and load in the four major
well as algal enumerations at Stations 11,12, and 14. tributaries to St. Albans Bay and from the St. Albans
49
-------
Proceedings of National RCWP Symposium, 1992
city wastewater treatment plant. The subwatersheds
monitored by these trend stations comprised 72 per-
cent of the entire St Albans Bay watershed. A fifth
station monitored sediment and nutrient loads from
the wastewater treatment plant, the major point
source in the watershed.
At each of the five Level 2 stations, samples were
collected automatically at eight-hour intervals using
refrigerated ISCO samplers and combined into two
48-hour and one 72-hour composite samples each
week for analysis. During some stormflow periods,
individual samples were analyzed discretely. This
sampling/compositing schedule was dictated by
budgetary restrictions and because the use of pre-
viously ungaged stations precluded flow-proportion-
al sampling. It should be recognized that this
program probably tended to underestimate export
rates by over-sampling low-flow periods and under-
sampling high-flow periods. However, the impor-
tance of consistency in sampling schedules for
long-term trend monitoring outweighed the benefits
of changing to a flow-proportional sampling schedule
after stage-discharge ratings were developed.
All Level 2 samples were analyzed for turbidity,
total suspended and volatile suspended solids, total
and soluble reactive phosphorus, and total Kjeldahl,
ammonia, and nitrite+nitrate nitrogen. Weekly grab
samples were analyzed for pH, conductivity, and
fecal coliform and fecal streptococcus bacteria, while
in sittt measurements were made of temperature and
dissolved oxygen at the time of grab sampling.
Stream stage was recorded continuously at each
Level 2 station using ISCO bubbler-type stage re-
corders and discharge was calculated from site-
specific stage-discharge ratings. Three standard
20-cm weighing bucket gages were used to measure
watershed precipitation.
Data Analysis
The lack of preproject water quality data, the concur-
rent start-up of both monitoring and land treatment,
and the incremental implementation of BMPs made
straightforward before/after comparisons impos-
sible. Since no appropriate control watershed was
available, paired-watershed analysis was also imprac-
tical. Therefore, evaluation of changes in water
quality were based on the detection of trends over
time.
Some adjustments to these data were necessary
to satisfy statistical assumptions. To reduce autocor-
relation problems, concentration, streamflow, and
mass data were typically aggregated to monthly
mean values for trend analysis. Neither original data
nor these monthly means followed a normal distribu-
tion; all parametric statistical analyses were per-
formed on log(base 10) transformed data. All means
reported for parametric tests are antilogs of these
log means (geometric means). While some small
bias was introduced by back-transforming these
means, it should be noted that parametric trend
tests, such as regression and analysis of variance
(ANOVA), were conducted and interpreted in log
space, not on retransformed data. Thus, the in-
fluence of retransformation bias on the conclusions
of trend analysis was probably minimal.
For some analyses, monitoring data were
divided into approximate pre-BMP and post-BMP
periods, based on the progress of implementation.
Although the gradual application of land treatment
did not provide an absolute before/after break, it
was useful to approximate such a division. Project
Years 1982-85 (covering 9/1982 through 8/1986)
were considered pre-BMP because active contract-
ing and implementation took place during this
period. By August 1986, all critical acres were under
contract, over 75 percent of animal waste systems
had been installed, and more than 80 percent of
goals for animal units, nitrogen, and phosphorus
under BMP had been achieved. Thus, Project Years
1986-89 (covering 9/1986 through 8/1990) were
designated post-BMP.
To fully explore the data, a variety of trend
analysis techniques were used: simple linear regres-
sion against time, t-test of pre- and post-BMP
periods, ANOVA, comparison of pre- and post-BMP
flow and concentration relationships, analysis of
covariance (ANCOVA), and nonparametric tests
such as Mann-Whitney and seasonal Kendall. Using
numerous trend tests was somewhat cumbersome
but necessary because of the complexity of the data
set, which made reliance on a single technique un-
wise. Some tests, such as time regression, were used
only as-screening tools because the high number of
observations tended to inflate the apparent sig-
nificance of the regression. The ANCOVA technique
was employed to test for trends in concentration data
while controlling for streamflow variation but proved
inappropriate for export data. Some tests were ap-
plicable to data sets with particular characteristics,
such as seasonality (e.g., seasonal Kendall). Dif-
ferent trend tests were also required to assess
monotonic or step trends. Furthermore, the applica-
tion of several different trend tests provided some
additional confidence in the interpretation of results:
a trend confirmed by all analyses was judged more
meaningful than one indicated by only one test. Dif-
ferent approaches did not disagree on the direction
of change but on magnitude and statistical sig-
nificance. Statistical significance of trend test results
was evaluated at P s 0.10.
50
-------
D.W. MEALS
Results
Land Treatment
Of the 102 farms in the St. Albans Bay watershed, 61
signed RCWP contracts. Land treatment in the
watershed achieved 98 percent of goals for critical
acre treatment and 95 percent of goals for critical
source treatment. The most widely used BMPs were
animal waste management (BMP 2) and cropland
protection (BMP 8), which were included in 93 per-
cent and 92 percent of all contracts, respectively.
Other widely applied BMPs included permanent
vegetative cover (BMP 1), stream protection system
(BMP 10), permanent vegetative cover on critical
areas (BMP 11), and fertilizer management (BMP
15). Overall, 64 manure storage systems, 61 barn-
yard runoff control systems, and 43 milkhouse waste
treatment systems were installed and 3,760 ha of
conservation cropping systems were implemented.
After implementation was complete, 79 percent of
watershed animal units were under BMP, repre-
senting 133 tons of manure phosphorus and 664 tons
of manure nitrogen under proper waste manage-
ment.
St. Albans Wastewater Treatment
Plant
The quality of effluent from the city of St. Albans
wastewater treatment plant improved dramatically
following a 1987 upgrade of the plant to tertiary
treatment. The main result of the upgrade was a
major decrease in sediment and nutrient concentra-
tion and export and a decrease in bacteria counts in
plant effluent. Mean effluent total phosphorus levels,
for example, dropped from over 5 mg/L to 0.50
mg/L after the upgrade, and monthly effluent total
phosphorus load dropped 85 percent from 984
kg/month to 147 kg/month. The distribution of
nitrogen species in effluent shifted from pre-
dominantly total Kjeldahl nitrogen and ammonia
nitrogen to substantial proportions of nitrite-nitrate
nitrogen because of increased nitrification. Despite
significant increases in nitrite+nitrate concentration
and load, effluent total nitrogen concentration and
load decreased nearly 35 percent after the upgrade.
One additional effect of the treatment plant
upgrade was improvement in plant operations, which
eliminated periodic overflows of untreated waste-
water to Stevens Brook. These discharges had
strongly influenced water quality observed at
Stevens Brook Station 22, and improvements in
Stevens Brook water quality were likely a result of
their elimination. ,
Tributary Streams
Year-to-year variation in watershed precipitation and
tributary discharge was significant but without a
consistent pattern. Above-average precipitation and
high streamflows occurred in project years 1983,
1985, and 1989; successive years of below-average
precipitation and low streamflows characterized
most of the second half of the monitoring period
(1986-88).
Sediment and nutrient concentrations, which
varied between watersheds and between years
throughout the project, were influenced by differen-
ces in weather, runoff, and streamflow. Water quality
data over the monitoring period are summarized in
Table 1. The Jewett Brook watershed (Station 21),
with the highest proportion of land in agriculture
and the highest level of agricultural activity in the St.
Albans Bay watershed, consistently exhibited the
highest concentrations of suspended solids, phos-
phorus, and nitrogen and bacteria counts over the
monitoring period (Meals, 1992).
Sediment and nutrient export from the tributary
watersheds varied strongly with season and with
streamflow. Most annual export occurred during
spring snowmelt, midwinter thaws, or summer
storm events. Total annual sediment and nutrient ex-
port across the monitored tributaries to St. Albans
Bay is summarized in Figure 2. During the final year
of monitoring, nonpoint sources contributed an es-
timated 99 percent of the suspended solids, 93 per-
cent of the phosphorus, and 85 percent of the
nitrogen loading to St. Albans Bay.
PROJECT YEAR
lTSS(x10")
!TKN
Figure 2.—Plot of total annual sediment and nutrient ex-
port from monitored tributaries, St. Albans Bay water-
shed, project years 1981 to 1989. Values plotted
represent sum of measured export from the monitored
subwatersheds plus the point source contribution.
Values do not represent total load to St. Albans Bay be-
cause exports from ungaged portions of the watershed
are not Included.
Areal nutrient loading rates from the monitored
tributary watersheds differed among the water-
sheds, and these rates were generally higher than
average values cited for other agricultural water-
51
-------
Table 1.— Range In annual mean1 water quality, Level 2 monitoring stations,
[TSSJ (mg/L)
[VSS] (mg/L)
[TP] (mg/L)
[SRP] (mg/L)
[TKN] (mg/L)
INHa-N] (mg/L)
INO2+NO3-N] (mg/L)
FC(#/100 mL)
FS(W100 mL)
Q(m3/sec)
TSS (kg/mo)
VSS (kg/mo)
TP (kg/mo)
SRP (kg/mo)
TKN (kg/mo)
NHa-N (kg/mo)
NOz+NOa-N (kg/mo)
JEWETT BR.
(STA. 21)
14.6-24.9
4.0-7.0
0.68-1.04
0.33-0.61
2.0-3.7
0.5-0.9
0.2-1.3
230-1,600
50-450
0.01-0.09
790-10.000
260-1,700
30-190
20-120
90-760
30-130
40-250
STEVENS BR.
(STA. 22)
11.1-35.2
2.7-6.4
0.15-0.30
0.07-0.24
0.8-1.0
0.1-0.2
0.4-0.9
180-1,200
50-340
0.11-0.28
5,600-26,000
1,300-3,300
75-340
24-110
250-760
40-180
150-620
RUGG BR.
(STA. 23)
5.6-16.1
1 .6-2.9
0.10-0.15
0.02-0.06
0.6-0.9
0.1-0.2
0.2-0.4
110-330
30-230
0.14-0.41
4,200-64,000
780-2,700
60-240
10-60
230-950
40-130
100-570
St. Albans Bay watershed.
MILL R.
(STA. 24)
5.9-14.6
1.9-2.8
0.08-0.17
0.02-0.06
0.6-0.9
0.1-0.2
0.2-0.5
60-230
20-140
0.26-0.98
9,600-150,000
1 ,800-7,900
110-400
30-160
480-2,000
90-300
210-1,250
WWTP
(STA. 25)
5.6-102.8
3.5-78.9
0.35-5.89
0.08-3.87
2.7-20.0
1.3-10.9
0.1-6.4
1-700,000
<1 -190,000
0.06-0.14
1,800-15,200
970-12,500
100-1,300
20-900
800-4,000
400-2,400
10-2,120
'geometric mean (anti-log of log mean)
TSS «total suspended solids; VSS = volatile suspended solids; TP = total phosphorus; SRP = soluble reactivei phosphorus; TKN= total
SldaW nUrogeniNHa-N = ammonia nitrogen; NO2+NO3-N = nitrite+nitrate nitrogen; FC = fecal conform; FS = fecal streptococcus, Q =
stroamllow.
sheds in the United States (Table 2). The Jewett
Brook watershed tended to exhibit the highest areal
nutrient export rates in the St Albans Bay water-
shed, rates that were higher than those observed in
similar agricultural watersheds in Vermont (Meals,
1990). The Stevens Brook watershed generally
showed the lowest nutrient export rates.
Water quality trends in tributaries to St. Albans
Bay were evaluated using regression, pre- and post-
t-test, ANOVA, ANCOVA, and nonparametric tests
such as the seasonal Kendall and Mann-Whitney.
Monthly mean streamflow, mean concentration of
measured parameters, and monthly sediment and
nutrient export values were used as the basic data
set The results of trend analysis are reported in
detail in the projects final report (V~t. RCWP Coord.
Comm. 1991); tributary trends are summarized in
TableS.
Although tributary stream discharge varied
from year to year, there were no strong, consistent
trends observed that would drive overall trends in
water quality across the St. Albans Bay watershed.
Stream discharge did, however, tend to be slightly
lower in the post-BMP period, compared to pre-BMP.
Turbidity and both concentration and export of
suspended solids declined in all tributaries except
Jewett Brook. According to the seasonal Kendall
test, total suspended solids concentrations declined
by 4.1 mg/L/year in Stevens Brook, 1.1 mg/L/year
in Rugg Brook, and 1.4 mg/L/year in Mill River.
Declines in volatile suspended solids concentration
were estimated at 0.2 to 0.3 mg/L/year. Mean
suspended solids levels in Jewett Brook, however,
showed a significant 10 to 15 percent increase be-
tween pre-BMP (15.3 mg/L) and post-BMP (17.1
mg/L) periods, even when differences in flow were
controlled.
Sediment export from the Stevens Brook water-
shed dropped from a pre-BMP average of 25,000
kg/month to a post-BMP average of 8,274
Bay watersheds, 1982-90.
TSS
TP
SRP
TKN
JEWETT
BR.
7-54
1.1-6.4
0.6-4.0
3.3-24
STEVENS
BR.
104-552
1.1-2.8
0.2-0.9
3.9-7.6
RUGG
BR.
174-1,927
1.3-6.1
0.3-2.4
4.4-16
MILL
R.
kg/ha/yr - - -
156-4,025
0.8-6.2
0.2-1 .2
2.6-15
LaPLATTE R.
WATERSHEDS
(VT)1
11-696
0.2-2.5
0.1-1.2
1.4-20
U.S.
WATERSHEDS2
0.1-0.3
0.09-0.13
5.2-9.8
GREAT LAKES
WATERSHEDS3
3-5,600
0.1-9.1
0.1-0.6
0.6-4.2
r Meals, 1990
20merntk.1976,1977
3 Pollut. from Land Use Act. Ref, Group, 1978
TSS «total suspended solids; TP = total phosphorus; SRP =
soluble reactive phosphorus; TKN = total Kjeldahl nitrogen.
52
-------
D. IV. MEALS
Table 3.—Trends in St. Albans Bay tributary water quality, 1982-90.
Jewett Br.
Stevens Br.
Rugg Br.
Mill R.
WWTP
Jewett Br.
Stevens Br.
Rugg Br.
Mill R.
WWTP
TRB
V
V
V
V
0
9
•
•
A
TSS
A
T
T
TSS
V
T
•
V
T
vss
A
T
V
VSS
V
•
V
T
TP
A
T
A
TP
V
•
•
T
POMPF
SRP TKN
• A
T •
A A
SRP TKN
V V
• A
• •
T T
NTRATI
NH3-N
A
V
A
A
\cc
NH3-N
T
A
•
T
ON
NO, + NO3-N
A
A
. A
A
NO,+NO3-N
V
A
A
A
Fr
V
T
T
T
V
FS
V
V
T
T
T
• = No significant trend
= Increasing or decreasing trend by some but not all statistical tests (P <_ 0.10)
AT = Increasing or decreasing trend by all statistical tests (P <. 0.10)
TRB = turbidity; TSS = total suspended solids; VSS = volatile suspended solids; TP = total
phosphorus; SRP = soluble reactive phosphorus; TKN = total Kjeidahl nitrogen; NH3-N =
ammonia nitrogen; NOa+NOa-N = nitrite+nitrate nitrogen; FC = fecal coliform; FS = fecai
streptococcus; Q = streamflow.
kg/month, a decline of 67 percent. Sediment export
from the Mill River watershed declined 59 percent
from an average of 37,000 kg/month pre-BMP to an
average of 15,000 kg/month post-BMP. Total
suspended solids export from the Jewett Brook
watershed dropped 60 percent from an average
33,000 kg/month to 13,350 kg/month in the post-
BMP project period, according to the seasonal Ken-
dall test.
Phosphorus concentrations increased sig-
nificantly over the monitoring period in Jewett
Brook, Rugg Brook, and Mill River but declined in
Stevens Brook. In Jewett Brook, the mean post-BMP
total phosphorus concentration of 0.67 mg/L was 13
percent higher than the pre-BMP mean of 0.76
mg/L. In Rugg Brook and Mill River, total phos-
phorus concentrations increased by an estimated
0.010 mg/L/year and soluble reactive phosphorus
levels increased by 0.006 mg/L/year. These in-
creases were still significant when concentrations
were adjusted for differences in streamflow using
ANCOVA. In contrast, total phosphorus and soluble
reactive phosphorus concentrations in Stevens
Brook both declined by 0.018 mg/L/year. Few sig-
nificant trends in phosphorus export from the
monitored watersheds occurred
during the study period. In
Stevens Brook, total phosphorus
and soluble reactive phosphorus
export declined by 40 percent
and 55 percent, respectively, in
the second half of the project, but
no other significant trends were
observed.
Concentrations of all forms
of nitrogen generally increased
over the entire St. Albans Bay
watershed during the monitoring
period, although fewer statistical-
ly significant trends were ob-
served than for phosphorus
because of high variability in ob-
served nitrogen levels. Most sig-
nificant changes in nitrogen
concentration seemed to occur in
Rugg Brook and in Mill River.
Concentrations of total Kjeidahl
and nitrite-nitrate nitrogen in Mill
River, for example, increased by
an average of 0.04 mg/L/year
and 0.05 mg/L/year, respective-
ly. In Rugg Brook, ammonia and
nitrite-nitrate nitrogen levels in-
creased by about 0.01 mg/L/
year and 0.03 mg/L/year, respec-
tively. These increases became
most apparent around 1987-88, corresponding to
similar areawide increases in nitrogen levels ob-
served in other Vermont watersheds (Meals, 1990)
and the Great Lakes region (Richards, 1991). The ex-
planation for this phenomenon is unclear. Significant
increases in nitrogen concentrations did not occur in
Stevens Brook, where a 13 percent decline in
average post-BMP NHs-N levels was observed.
Changes in nitrogen export across the St. Al-
bans Bay watershed followed a pattern similar to
changes in nitrogen concentrations. No significant
trends in nitrogen export from the Jewett Brook
watershed were documented. Nitrogen export from
the Stevens Brook watershed tended to decline; in
the latter half of the project, export of total Kjeidahl
nitrogen, ammonia, and nitrite-nitrate nitrogen
declined by 30 percent, 46 percent, and 32 percent,
respectively. As with concentration, nitrogen export
from the Rugg Brook and Mill River watersheds
tended to increase. Monthly export of all forms of
nitrogen from Rugg Brook increased significantly in
the later project years: total Kjeidahl nitrogen +90
percent, ammonia nitrogen +160 percent, and nitrite-
nitrate nitrogen +100 percent. The Mill River water-
shed showed an estimated 110 percent increase in
53
-------
Proceedings of national RCWP Symposium, 1992
monthly nitrite-nitrate nitrogen export; however, no
trends in total Kjeldahl nitrogen or ammonia
nitrogen export were observed.
Fecal coliform (FC) and fecal streptococcus
(FS) bacteria counts declined significantly across all
the monitored St. Albans Bay tributaries. Declines in
FC counts were greatest in Jewett and Stevens
brooks, where mean monthly counts decreased by
an average 130 to 140 per year. Monthly FC counts in
Rugg Brook and Mill River decreased by about 15 to
30 per year. Mean post-BMP FC counts dropped by
50 to 70 percent compared to pre-BMP counts in
each of the monitored streams.
Fecal streptococcus (FS) counts also declined
significantly across all the monitored tributaries.
Again, the declines were greatest in Jewett and
Stevens brooks — drops of 30 to 50 per year — com-
pared to 8 to 16 per year in Rugg Brook and Mill
River. The decrease in post-BMP FS counts was even
greater that the decline in FC: mean post-BMP
counts dropped 60 to 70 percent at all of the tributary
stations. The decline in indicator bacteria counts is il-
lustrated in Figure 3. These declines are comparable
to those observed following land treatment in other
Vermont agricultural watersheds (Meals, 1989). The
decline in FC and FS counts is the strongest and
clearest water quality trend observed in the CM&E
project.
-JEWETT BB.
85 66
PROJECT YEAR
- STEVENS BR. » RUQO BR. ••=- MILL R.
85 86
PROJECT YEAR
-JEWETT BR.
- STEVENS BR.
RUQO BR. •-"•• MILL R.
Figure 3.—Plots of mean annual indicator bacteria
counts In monitored tributaries, St. Albans Bay water-
shed, project years 1982 to 1989. Top: fecal coliform
bacteria; bottom: fecal streptococcus bacteria.
St. Albans Bay
A water quality gradient typically existed from the
north end of the bay (Station 14) to the outer bay
(Station 11) for conductivity, Secchi disk transparen-
cy, turbidity, suspended solids, phosphorus, nitro-
gen, and chlorophyll a levels. Concentrations tended
to be highest at the inner end of the bay, where in-
flow from Jewett and Stevens brooks dominates, and
lowest at the outer bay/Water quality was poorer in
the inner region of the bay, which is close to the in-
fluence of point and nonpoint sources of sediment
and nutrients.
Water quality in St. Albans Bay was variable
throughout the monitoring period. Annual mean
values for total phosphorus, chlorophyll a, and Sec-
chi disk transparency for the three main bay stations
are plotted in Figure 4. In the outer bay, median an-
nual total phosphorus concentrations ranged from
20 to 52 ptg/L, median chlorophyll a levels were 4 to
19 u.g/L, and median Secchi disk transparency was
2.7 to 4.3 m. These values were consistent with
mesotrophic conditions characteristic of the north-
east area of Lake Champlain.
In the inner bay, however, annual median total
phosphorus levels were 54 to 76 u,g/L, median
chlorophyll a ranged from 6 to 20 ng/L, and median
Secchi disk values were just 1.1 to 1.5 m, all sugges-
tive of eutrophic conditions. At Station 14, just off-
shore of the Jewett-Stevens Brook inflow at Black
Bridge, water quality was even further into the
eutrophic range: annual median values were total
phosphorus, 66 to 104 ug/L; chlorophyll a, 7 to 26
ug/L; and Secchi disk transparency, 0.9 to 1.2 m. Al-
though some year-to-year differences in St. Albans
Bay water quality were significant, few" consistent
patterns were observed during the study.
Water quality trends in St. Albans Bay are dis-
cussed in detail in the project's final report (Vt.
RCWP Coord. Comm. 1991); trends are summarized
in Table 4. Turbidity and suspended solids con-
centrations generally decreased in inner St. Albans
Bay but were unchanged in the outer region of the
bay. Phosphorus concentrations increased at both
the outer and inner bay stations but decreased sig-
nificantly at Station 14. Seasonal Kendall tests sug-
gested increases in outer bay total phosphorus and
soluble reactive phosphorus concentrations ^ of 6
ug/L/year and 1 ug/L/year, respectively. Similar
analysis suggested a decline of 3 ug/L/year in total
phosphorus concentrations at the off-bridge station.
Fewer significant changes were observed for
nitrogen. Concentrations of both total Kjeldahl
nitrogen and NHs-N increased at the outer bay sta-
tion but did not change significantly in the inner bay.
54
-------
D.W. MEALS
85 86
PROJECT YEAR
OUTER BAY (Sta. 11)
-*-• OFF-BRIDQE (Sta. 14)
INNER BAY (Sta. 12)
85 86
PROJECT YEAR
' OUTER BAY (Sta. 11)
"*• OFF-BRIDQE (Sta. 14)
INNER BAY (Sta. 12)
85 86
PROJECT YEAR
OUTER BAY (Sta. 11)
•*•' OFF-BRIDQE (Sta. 14)
INNER BAY (Sta. 12)
Figure 4.—Plots of mean annual concentrations of
phosphorus (top), chlorophyll a (middle), and Seech!
disk transparency (bottom) In St. Albans Bay, project
years 1981 to 1989.
Table 4.—Trends in St. Albans Bay water quality, 1981-90.
Station
Off-Bridge (14)
Inner Bay (12)
Outer Bay (11)
TRB
V
V
•
TSS VSS
T •
V A
• •
TP
. T
A
A
SRP
V
•
A
TKN NH3-N
« V
• •
A A
CHLa
A
A
A
S.D.
V
V
• = No significant trend
AV = Increasing or decreasing trend by some but not all statistical tests (P<_ 0.10)
= Increasing or decreasing trend by all statistical tests (P <. 0.10)
TRB = turbidity; TSS = total suspended solids; VSS = volatile suspended solids; TP = total
phosphorus; SRP = soluble reactive phosphorus; TKN = total Kjeldahl nitrogen; NH3-N =
ammonia nitrogen; CHL a = chlorophyll a; S.D. = Secchi disk.
Ammonia levels declined significantly at the off-
bridge station.
Chlorophyll a concentrations increased by about
2 to 3 ug/L/year at St. Albans Bay stations over the
monitoring period. Secchi disk transparency tended
to be lower at the inner and outer bay stations in the
post-BMP period; no trends in transparency were ob-
served at Station 14.
No statistically significant trends were detected
in bacteria counts at the Beach Station (Station 13),
principally because of the high variability in the ob-
served counts. However, frequency of fecal coliform
counts in excess of the Vermont water quality stand-
ard (200 per 100 mL) dropped to near zero during
the last three years of the project and median fecal
streptococcus counts were at or below detection
level for the last two project years. Both patterns sug-
gest declines in indicator bacteria counts near the
beach.
Data on algae populations indicated significant
declines in algal abundance (cells/L) throughout the
bay, but total algal biomass (g/L) tended to increase
significantly over the study period. Blue-green algae
biomass did not change significantly over the
project. Annually, the proportion of blue-green algae
in the bay was less than 10 percent; however, during
the summer, blue-greens dominated algae in the bay.
Annual macrophyte surveys of the entire bay did
not reveal major changes that could be attributed to
land treatments. A decline in overall density and ex-
tent of macrophytes was noted in 1989 and 1990, par-
ticularly for Myriophyllum spicatum, the predominant
exotic species. However, declines could have
resulted from decreases in water clarity caused by
increasing algal populations, elevated lake levels
from above-normal precipitation in 1990, or preda-
tion by an aquatic caterpillar Acentria acentropus and
other organisms recently discovered to feed and
reproduce specifically on M.
spicatum throughout Vermont
(Creed and Sheldon, 1991).
In summary, since the begin-
ning of the RCWP project, St. Al-
bans Bay water quality improved
only very slightly in the inner-
most portion of the bay but
declined elsewhere. Nutrient
concentrations and algal produc-
tion generally increased in most
of the bay, and water transparen-
cy decreased. Only in the inner-
most bay (Station 14) and at the
beach (Station 13) did water
quality remain stable or show
55
-------
yilngs of National RCWP Symposium, 1992
slight improvement in terms of slightly decreased
nutrient levels and lower bacteria counts.
Water quality in St. Albans Bay was tracked
using the Trophic State Index (TSp (Carlson, 1977)
as an aggregate measure of condition. Carlson's TSI
considers values of total phosphorus, chlorophyll a,
and Secchi disk and may range from 0 (oligotrophic)
to 100 (hypereutrophic). Mean annual TSI values for
St. Albans Bay from 1981 to 1990 are plotted in Fig-
ure 5; they suggest a trend from mesotrophic to
eutrophic conditions. TSI values observed during
peak summer algae production were in the eutrophic
range. Individual TSI values recorded during peak
summer algal growth have been as high as 84, 78,
and 63 for Stations 14,12, and 11, respectively. TSI
values at Station 14 have stabilized in recent years,
but most of St. Albans Bay remains in the meso-
eutrophic range.
Conclusions
Despite dramatic reduction in phosphorus load from
the wastewater treatment plant, St. Albans Bay water
quality did not improve significantly over the RCWP
project period. Recent modeling analysis of St. Al-
bans Bay response to phosphorus loading using a
mass-balance model calibrated and verified with
CM&E data predicted that substantial summer phos-
phorus reductions in the bay should have resulted
100
from the treatment plant upgrade (Smeltzer, 1991).
The continued high phosphorus levels observed in
the bay contradict model predictions and suggest
the existence of some other source (s) of phosphorus
for St. Albans Bay.
High phosphorus levels in St. Albans Bay sedi-
ments have been documented (Corliss and Hunt,
1973) and the potential for internal phosphorus load-
ing in the bay has been established by early studies
conducted for the CM&E program (Ackerly, 1983;
Drake and Ackerly, 1983; Fowler, 1984; Fowler et al.
1984). Sediment resuspension and chemical release
under both anaerobic and aerobic conditions are all
possible mechanisms for internal phosphorus load-
ing in the bay. Based on these findings and on model-
ing results, Smeltzer (1991) suggested that a net flux
of phosphorus from the sediments to the water
column is supporting continued high algal produc-
tion in St. Albans Bay. Since phosphorus is being
continually removed from the bay by flushing to the
main lake, this internal load should persist only as
long as phosphorus can be remobilized from the
sediments. Thus, after some lag time, St. Albans Bay
should respond to the reduced point source phos-
phorus load.
Unfortunately, no comparable decreases in non-
point source phosphorus loads from St. Albans Bay
tributaries were observed during the RCWP period;
the dramatic in-stream water quality improvements
1981 1982 1983 1984 1985 1986 1987 1988 1989 1990
YEAR
OUTER BAY
INNER BAY
OFF-BRIDGE
1Carlson's Trophic State Index
Figure 5.—Plot of mean annual Trophic State Index values In St. Albans Bay, 1981 to 1990. The dashed horizontal lines
represent the areas of transition between trophic state categories.
56
-------
D.W. MEALS
anticipated when the St. Albans Bay Watershed
RCWP project began were not realized. While sedi-
ment concentrations and export from most of the
monitored tributaries declined, phosphorus con-
centrations in the streams increased and loads were
essentially unchanged. Increases in phosphorus con-
centrations may have been the result of higher
runoff and streamflows in the last years of the
project.
Nitrogen concentrations and export generally in-
creased across the watershed. Only in Stevens
Brook were reductions in nutrient concentrations
and loads noted and these improvements were
directly attributable to the wastewater treatment
plant upgrade that eliminated discharges of un-
treated wastewater to the stream upstream of the
monitoring station. However, the lack of general im-
provement in tributary water quality is somewhat
puzzling because the effectiveness of many of the
BMPs applied has been documented at the edge-of-
field scale (Meals, 1990; Vt. RCWP Coord. Comm.
1991).
Several reasons can be advanced for the lack of
strong water quality response to the land treatment
program. The nature, timing, or level of land treat-
ments may have been inadequate to change water
quality above background variability. Some prac-
tices, such as riparian zone management or livestock
exclusion, were not applied in the project. Further-
more, despite the relatively high level of farmer par-
ticipation in the RCWP, loads from a few
nonparticipating farms may have overwhelmed
upland treatment effects. For example, winter
manure spreading continued to occur on many non-
contract farms, and evidence shows that manure ap-
"plication rates actually increased in some riparian
areas during the project (Schlagel, 1992).
The lag time in the system between application
of land treatment and improvements in water quality
may exceed the monitoring period (Clausen et al.
1992). Just as St. Albans Bay sediments appear to be
serving as a reservoir for continued phosphorus
supply, so too may watershed soils that have been
farmed and fertilized for decades be acting as a con-
tinuing nutrient reservoir — despite recent changes
in inputs and/or management. Significant supplies of
phosphorus may also remain in stream sediments.
The dramatic 50 to 70 percent decline in in-
dicator bacteria counts in all monitored streams is
encouraging and argues that the land treatment pro-
gram did affect water quality. Relationships between
some land use and agricultural management chang-
es and observed bacteria levels in streams have been
suggested (Meals, 1992). Declines in fecal coliform
and fecal streptococcus counts would be expected to
result from improvements in animal waste manage-
ment. Manure storage alone can reduce bacteria
levels in manure (Patni et al. 1985). Reduction of
daily spreading and increases in incorporation of ap-
plied manure should tend to reduce the quantity of
bacteria that is transported in runoff from the soil
surface. Finally, diversion of barnyard runoff into
manure storage structures or otherwise away from
surface waters may have diminished a potent source
of these bacteria.
It is important to note that, unlike phosphorus,
fecal coliform and fecal streptococcus bacteria are
not readily stored in the environment. Decreases in
their numbers in watershed streams suggest a
decreased rate in supply to surface waters. Their
rapid decline may represent the leading edge of
greater water quality improvements to come as
nutrient losses from agricultural land also decline
and nutrients accumulated historically in the system
are gradually flushed out.
ACKNOWLEDGMENTS:^ St Albans Bay Watershed
RCWP Comprehensive Monitoring and Evaluation Program
received financial support from the U.S. Department of
Agriculture's Agricultural Stabilization and Conservation Ser-
vice and the University of Vermont Assistance from the Soil
Conservation Service was critical to the success of the pro-
gram. The cooperation and assistance of the Franklin County
Natural Resources Conservation District, the Vermont
Cooperative Extension Service, and the Vermont Department
of Environmental Conservation is gratefully acknowledged.
Finally, the hard work and dedication of Dr. John Clausen,
who guided the Comprehensive Monitoring and Evaluation
Program for most of its course, are also acknowledged.
References
Ackerly, S. 1983. Sediment-phosphorus relationships, St Albans
Bay, Lake Champlain. M.S. thesis. Geo. Dep., Univ. Vermont
Burlington.
Braden, J.B. and D.L. Uchtmann. 1985. Agricultural nonpoint pol-
lution control: an assessment J. Soil Water Conserv.
40(l):23-6.
Carlson, RE. 1977. A trophic state index for lakes. limnol.
Oceanogr. 22:361-69.
Clausen, J.C. 1985. The St. Albans Bay watershed RCWP: a case
study of monitoring and assessment Pages 214 in Perspec-
tives on Nonpoint Source Pollution, Proc. Natl. Conf., Kansas
City, MO.
Clausen, J.C.; D.W. Meals, and E.A. Cassell. 1992. Estimation of
lag time for water quality response to BMPs. In Proc. Natl.
RCWP Symp., Orlando, FL.
Corliss, B. and A.S. Hunt. 1973. The distribution of phosphorus in
the sediments of St Albans Bay, Lake Champlain. Lake
Champlain Res. Rep. Issue #4. Geo. Dep., Univ. Vermont,
Burlington.
Creed, R.P. and S.P. Sheldon. 1991. The potential for biological con-
trol of Eurasian watermilfoil: results of the research con-
ducted at Brownington Pond, Vermont and the multi-state
lake survey. Prog. rep. Dep. Biol., Middlebury College, Mid-
dlebury.VT.
57
-------
Proceedings of national RCWP Symposium, 1992
Drake, J. and S. Ackerly. 1983. Preliminary pore water investiga-
tions of Lake Champlain sediments. Abstr. Int Ass. Great
Lakes Res., Annu. Meet
Flalg, EG. and G. Ritter. 1989. Water quality monitoring of agricul-
tural discharge to Lake Okeechobee. Pap. #89-2525, Am. Soc.
Agric. Eng. winter meet, New Orleans, LA.
Fowler, B. 1984. Phosphorus fluxes from sediments of St Albans
Bay. M.S. thesis. Geo. Dep., Univ. Vermont, Burlington.
Fowler, B., J.C. Drake, and D.R. Hemenway. 1984. Phosphorus
fluxes from St. Albans Bay, Lake Champlain sediments.
Abstr. Int Ass. Great Lakes Res., Annu. Meet.
Harbridge House, Inc. 1983. The Model Implementation Program:
an evaluation of the management and water quality aspects of
the Model Implementation Program. Natl. Water Qual. Eval.
Proj., N.C. State Univ., Raleigh.
Meals, D.W. 1989. Bacteriological water quality in Vermont
agricultural watersheds undergoing land treatment Lake
Reserv. Manage. 5(l):53-62.
. 1990. LaPlatte River Watershed Water Quality Monitoring
and Analysis Program Comprehensive Final Report. Prog.
Rep. No. 12. Vt Water Resour. Res. Center, Univ. Vermont,
Burlington.
. 1992. Relating land use and water quality in the St Albans
Bay watershed, Vermont In Proc. Natl. RCWP Symp., Orlan-
do, FL
Morrison, J.B. and J.E. Lake. 1981. Environmental impact of land
use on water quality. Final Rep. Black Creek Proj., Phase II.
EPA-905/9-81-003. U.S. Environ. Prot Agency, Washington,
DC.
National Water Quality Evaluation Project 1988. Status of agricul-
tural nonpoint source projects. N.C. State Univ., Raleigh.
Omernik, J.M. 1976. The influence of land use on stream nutrient
levels. EPA-600/3-76-014. Corvallis Environ. Res. Lab., U.S.
Environ. Prot. Agency, Corvallis, OR.
. 1977. Nonpoint source-stream nutrient level relationships:
a.nationwide study. EPA-600/3-77-105. Corvallis Environ.
Res. Lab, U.S. Environ. Prot Agency, Corvallis, OR.
Patni, N.K, M.R. Toxopeus, and P.Y. Jui. 1985. Bacterial quality of
runoff from manured and non-manured cropland. Trans. Am.
Soc. Agric. Eng. 28(6) =1871-84.
Persson, LA, J.O. Peterson, and F.W. Madison. 1983. Evaluation
of sediment and phosphorus management practices in the
White Clay Lake watershed. Water Resour. Bull. 19(5): 753-
62.
Pollution from Land Use Activities Reference Group. 1978. En-
vironmental management strategy for the Great Lakes sys-
tem. Final Rep., Int Joint Comm., Windsor, Ont, Can.
Richards, RP. 1991. Personal communication. Water Qual. Lab.,
Heidelberg College, Tiffin, OH.
Schlagel, J.D. 1992. Spatial and temporal change in animal waste
utilization in the Jewett Brook, Vermont, watershed, 1983-
1990. In Proc. Natl. RCWP Symp., Orlando, FL.
Smeltzer, E. 1991. The response of St Albans Bay, Lake Champlain
to phosphorus loading reductions. Vermont Dep. Environ.
Conserv., Waterbury.
Soil Conservation Service, Idaho Dep. Environmental Quality,
Agricultural Stabilization and Conservation Service. 1991.
Rock Creek Rural Clean Water Program Ten Year Report.
Rock Creek RCWP, Twin Falls, ID.
Vermont RCWP Coordinating Committee. 1991. St. Albans Bay
Rural Clean Water Program Final Report 1980-1990. Vermont
Water Resour. Res. Center, Univ. Vermont Burlington.
58
-------
Understanding the Groundwater
System: The Garvin Brook Experience
David B. Wall, Mark G. Evenson, Charles P. Regan,
Joseph A. Magner, and Wayne P. Anderson
Water Quality Division
Minnesota Pollution Control Agency
Saint Paul, Minnesota
ABSTRACT
A 15 well monitoring network was established in 1981 to create baseline data and track
groundwater quality changes in the Garvin Brook watershed resulting from Rural Clean Water Pro-
gram (RCWP) best management practices (BMPs). Later, researchers discovered that most of the
15 wells yield water that entered the ground over 30 years ago and will not reflect current land
management practices for decades or centuries. Water monitoring and hydrogeologic investigation
led to additional discoveries: (1) nitrate contamination of groundwater was a significant problem in
the Garvin Brook area and (2) the actual recharge area for major aquifiers in the original project
area extends five miles beyond the Garvin Brook surface watershed boundary. Consequently, mid-
course adjustments were made to the project, including increased emphasis on groundwater
protection BMPs, expansion of the project area to include all of the groundwater recharge area,
and increased monitoring of wells yielding recent water. Trend analysis of data collected between
1983 and 1990 shows decreasing nitrate concentrations in high nitrate domestic wells. Vadose zone
modeling results, however, indicate that it will take several decades before the benefits from
nitrogen management BMPs are fully realized. The value of conducting a diagnostic study before
BMP implementation became increasingly apparent during the Garvin Brook RCWP project
Garvin Brook watershed in Winona County,
Minnesota, is one of 21 Rural Clean Water
Program (RCWP) project sites throughout
the United States established to evaluate the social,
economic, and technical aspects of controlling non-
point source pollution. The Garvin Brook RCWP, one
of three RCWP projects with a major focus on im-
proving groundwater quality, began in 1981 by
providing financial and technical assistance to
farmers implementing best management practices
(BMPs) to prevent nonpoint source pollution of
streams and groundwater.
The original emphasis of the project was to
protect and monitor streams in the Garvin Brook
watershed. However, groundwater protection and
monitoring became increasingly important through-
out the early 1980s. The shift in emphasis from sur-
face water to groundwater was prompted in part by
the discovery that nearly one-fourth of 80 sampled
wells in the Garvin Brook watershed contained
nitrate-nitrogen in excess of 10 mg/L.
This report provides an overview of hydro-
geological and groundwater quality findings in the
Garvin Brook RCWP project and resulting adjust-
ments in project direction and monitoring strategy.
Methods to evaluate the program's effects on
groundwater nitrate concentrations are also dis-
cussed.
59
-------
Proceedings of national RCWP Symposium, 1992
Background
Project Area Description
The Garvin Brook RCWP area was limited to the
30,720-acre Garvin Brook watershed until 1985 when
it was expanded to include 15,800 additional acres, a
major groundwater recharge area for wells in the
Garvin Brook watershed (Fig. 1). Garvin Brook
watershed and the groundwater recharge area,
which, combined, make up the Garvin Brook RCWP
area, are often referred to in this report as two
separate and distinct areas.
Garvin Brook watershed is 58 percent cropland,
with nearly 40 percent of all cropland in corn produc-
tion in any given year; other crops include soybean,
barley, oats, other small grain, and hay. Nearly 25
percent of land in the watershed is forested and the
other 17 percent is mostly pasture and urban. Most
of the farms in the watershed are dairy operations,
which together use over 60 feedlots. The additional
groundwater recharge area is 85 percent cropland,
with 60 percent in corn production. Land uses on the
remaining 15 percent are primarily pasture, wood-
land, urban, roads, and railroads. Average annual
precipitation is 31 inches, with about 75 percent fall-
ing from April to September.
Garvin Brook watershed is characterized by
relatively narrow flat-topped ridges, steeply sloping
valley sides, and deep, relatively broad valleys. The
relief ranges from 400 to 600 feet. The groundwater
recharge area is characterized by broad ridges (one
to two miles wide) with narrow drainageways
throughout, narrow valleys (25 to 75 feet wide), and
total relief less than 200 feet. Ridge soils consist of 3
to 30 feet of silt loam soils underlain by thin localized
deposits of till and outwash. These unconsolidated
deposits overlie carbonate bedrock. Narrow terraces
and floodplains, primarily consisting of silt loam soils
5 to 20 feet thick over sand or cobbly material, char-
acterize the valleys. Below the soil and glacial
deposits is a sequence of Paleozoic units charac-
terized by Mossier and Book (1984) and depicted in
T
GARVIN BROOK
WATERSHED
GROUND WATER
RECHARGEAREA
Figure 1.-—Map showing location of Garvin Brook Watershed and the adjacent groundwater recharge area. Together,
these two areas comprise the Garvin Brook RCWP project area.
60
-------
D. WALL, M. EVENSON, C. REGAN, J. MAGNER, & W. ANDERSON
Figure 2. Five formations supply virtually all drink-
ing water to residents within the project area.
The first bedrock formation encountered below
the glacial deposits in ridge areas is the Prairie du
Chien group, consisting of the Shakopee Formation
and Oneota Dolomite. The carbonate bedrock of the
Shakopee and Oneota is slowly being dissolved by
mildly acidic water, producing solution-enlarged frac-
[SYSTEM!
SERIES
LOWER ORDOVICIAN
UPPER CAMBRIAN
•
GROUP OR
FORMATION
NAME
a.
0 SHAKOPEE
g FORMATION
2
0
0 ONEOTA
u DOLOMITE
CC
Ł
Cu
JORDAN
SANDSTONE
ST. LAWRENCE1
FORMATION
FRANCONIA1
FORMATION
IRONTON &
GALESVILLE
SANDSTONES
EAU CLAIRE2
FORMATION
MT. SIMON2
SANDSTONE
PRECAMBRIAN3
SYM
BOL
Ops
Opo
*
•€«
Cf
*
•Ce
•Cm
p€
LITHOLOGY
^'^///^
^ ^>" ';' ' -..
..•'•^X^. '<:
•'''/'' .' X
-, -t, ,f, ..,.*.
^
>%
ff 4.
/•
o '>-*• ' ^ e ^J
r?" V* '. ":/"" =
p*^l
.^ / ,
— ' /
G • 0??r G •ww'/
G ^ • . G •
•r G / :G ;-
/G .'S-'^G' ^
• f.-.- _|'..: ^d\
*Z?SZ?///Ł'SZŁ.
G. r G
G . G G .
A _ G — G -6
f
"•f
/
(
.^^f^
• ' - '. ' . ' * .
• . • * _ ' . • • '
--•7— 7— T-T— 7;-^---
. . ~ . r* ~* ~ "" *" •"
. ' • • • OO • * * •
^5$?^^%
THICK
NESS
(feet)
90
to
115
160
to
180
100
to
120
50
to
75
140
to
180
90
to
120
90
to
125
290
to
350
DESCRIPTION
Thin-bedded and medium-bedded
dolomite with thin sandstone and shale
beds. Basal 20 to 30 feet is fine-grained
quartzose sandstone. Local red iron
staining. Basal contact minor erosional
"V surface p
Thick-bedded to massive dolomite. Some
sandy dolomite in basal 10 to 20 feet.
Vugs filled with coarse calcite in upper
part. Minor chert nodules. Upper part
near contact with Shakopee commonly
brecciated
and well cemented by calcite. Middle part
is medium- to coarse-grained quartzose
sandstone: generally uncemented'and iron
stained in outcrop Basal 35 to 40 feet is
very fine to fine-grained sandstone
Thin-bedded dolomitic siltstone. Minor
shale partings
Thin-bedded, dolomite-cemented
glauconitic sandstone. Very fine to fine
grained. Contains minor dolomite beds
near base and shale partings throughout
medium-grained quartzose sandstone with
minor glauconite
Galesville: Fine- to medium-grained, well-
sorted quartzose sandstone
Very fine to fine-grained sandstone and
siltstone. Some is glauconitic. Interbedded
shale
Fine- to very coarse grained, poorly
cemented sandstone. Contains pebbles in
basal 20 to 40 feet. Sandstone generally
moderately to well sorted. Greenish-gray
shale mottled with grayish -red in basal
third of formation. Basal contact major
erosional surface
/ D- • • \
— ' tuonnc granite gneiss in eastern part, v
Poorlv known in west
Figure 2.—Stratigraphic sequence of Paleozoic bedrock in the Garvin Brook project area (from Mossier and Book,
1984).
61
-------
flroceed/ngs of National ROMP Symposium, 1992
lures and sinkholes. The resulting karst terrain ren-
ders the groundwater system highly susceptible to
contamination. Many wells in the ridges are
developed in the Prairie du Chien group.
Underlying the Oneota Dolomite is the Jordan
Sandstone, which supplies drinking water to many
domestic wells throughout the project area. Many
wells are open in both the Prairie du Chien group
and Jordan Sandstone area and therefore draw water
from both formations. The Prairie du Chien-Jordan
aquifer is unconfined in the project area and is large-
ly absent along the valleys of Garvin Brook and
Stockton Valley Creek. Depths from land surface to
the water table generally range from 75 to 250 feet in
the ridges to less than 50 feet in the valleys.
Stratigraphically below the confining St
Lawrence Formation, the Franconia Formation is
thin-bedded dolomite-cemented glauconitic sand-
stone that is hydraulically connected to the Ironton
and Galesville sandstones. The Franconia-Ironton-
Galesville aquifer supplies water to a few wells on the
ridges and many wells in the valleys. The Eau Claire
Formation acts as a confining unit between the Fran-
conia-Ironton-Galesville aquifer and the Mount Si-
mon Sandstone. The Mount Simon Sandstone is
confined except near Minnesota City, at the mouth of
Garvin Brook, and serves several wells in that area.
BMPs Implemented
Over 1.7 million dollars of Federal cost-share money
has been contracted with farmers for implementa-
tion of BMPs in the Garvin Brook Project area since
1981. The primary BMPs implemented in the project
area for groundwater protection were pesticide
management, fertilizer management, sinkhole treat-
ment, and animal waste management systems. Pes-
ticide management, which includes the application of
pesticides at the rate, time, and method recom-
mended by Cooperative Extension and the proper
disposal of empty containers, was implemented on
approximately one-third of the cropland receiving
pesticides in the project area.
Fertilizer management was implemented on
about 40 percent of all corn cropland in the project
area. This BMP includes split-application of nitrogen
fertilizer at a rate based on realistic yield goals and
proper crediting for previous crops and manure ap-
plication. A 1989 survey of farmers implementing
fertilizer management under the RCWP indicated an
average 20 percent reduction in nitrogen fertilizer
use. Twenty-eight of 90 sinkholes located in the
project area have been treated since 1985, and 15
animal waste management systems have been con-
structed through the RCWP.
Early Findings
Original Groundwater Monitoring for
RCWP (1981 to 1982)
The Minnesota Pollution Control Agency began sam-
pling groundwater in Garvin Brook watershed in
June 1981. The initial objective was to establish
baseline water quality in the major aquifers so that
groundwater quality changes resulting from RCWP
BMPs could be monitored. Fifteen wells and three
springs were chosen for quarterly sampling. The
Prairie du Chien aquifer (an upper carbonate
aquifer) and the underlying Jordan Sandstone were
each represented by two wells; two of the three
springs were also from the Jordan Sandstone. The
other 11 wells and 1 spring produced water from
older bedrock formations.
The 15 wells and 3 springs were sampled for
nitrate, major ions, metals, and four different pes-
ticides. Nitrate-nitrogen was found at concentrations
exceeding the drinking water standard in both
Prairie du Chien wells (280 and 291 feet deep). At
one of the Prairie du Chien wells, the pesticide,
alachlor, was detected on one occasion. All other
wells had nitrate-nitrogen concentrations less than 2
mg/L and no pesticide detections; many had nitrate-
nitrogen levels less than 0.01 mg/L. These results
are described in more detail in Wall et al. (1989).
During the early 1980s, the RCWP project
focused on surface water and, consequently, few
hydrogeologic studies were conducted. Little effort
was made to understand flow systems, recharge area
boundaries, variability in geologic sensitivity, and
groundwater residence times.
Groundwater Findings (1983 to 1985)
• Groundwater Flow Directions and Recharge
Areas. A geologic atlas for Winona County was
developed during 1983 and 1984 by the Minnesota
Geological Survey (1984). One of the maps
developed for the atlas was a potentiometric contour
map (Kanivetsky, 1984) that showed general
groundwater flow directions (Fig. 3). It was evident
from the potentiometric map that the groundwater
and surface watershed had different boundaries.
The boundary for the area supplying groundwater to
Garvin Brook watershed was found to extend about
five miles west of the surface watershed. The
geologic atlas enabled delineation of the entire
groundwater recharge area providing water for Gar-
vin Brook watershed, and it described and mapped
geologic sensitivity, bedrock and surficial geology,
62
-------
D. WALL, M. EVENSO/Y, C. REGAN, J. MAGNER, & W. ANDERSON
Garvin Brook Watershed
Figure 3.—Potentiometric map for the Prairie du Chien-Jordan aquifer In central WInona County (from Kanivetsky,
1984). Water level monitoring locations are designated by circles. Boundaries for Garvin Brook Watershed and the ad-
ditional groundwater recharge area are depicted.
and sinkhole development, all useful information for of all wells sampled had nitrate-nitrogen concentra-
the Garvin Brook RCWP. tions exceeding 3 mg/L. Nitrate-nitrogen concentra-
_ _ _. . , . ^ tions above 3 mg/L generally reflect human impacts
• Nitrate Analyses m Domestic Wells. In 1983,' in southeastern Minnesota groundwater. These
Wmona County Cooperative Extension, working in results indicated a fairly widespread nitrate problem
conjunction with the Minnesota Department of in Garvin Brook watershed.
Agriculture, began annual sampling of domestic
wells for nitrate. The original purpose of this sam-
pling was to raise the awareness of people in the Gar- jut' i __..__ AJ« f J. j.
vin Brook watershed regarding current and potential Wia-COUrSe /\qJUStmentS IO
water quality problems. Of the 80 wells selected to the GarVlH BfOOk RCWP
uniformly cover Garvin Brook watershed, many
were older, with no construction records. According In response to the geologic and groundwater nitrate
to 1983 to 1985 sampling results, 21 to 24 percent of findings, several programmatic changes were made
the wells exceeded the 10 mg/L drinking water to the RCWP project. The project area was expanded
standard for nitrate-nitrogen, and approximately half during 1985 to include the additional groundwater
63
-------
Proceedings of National RCWP Symposium, 1992
recharge area for Garvin Brook watershed, increas-
ing it from 30,720 to 46,520 acres.
The groundwater recharge area included
numerous sinkholes and more intensive row crop
agriculture. The critical area for BMP implementa-
tion was redefined to include areas where suscep-
tibility to groundwater contamination was classified
as "High" in the geologic atlas, which included 94
percent of the groundwater recharge area.
The Garvin Brook RCWP also switched em-
phasis from surface water protection BMPs to those
that protected groundwater. Sinkhole treatment,
which became more commonly implemented, usual-
ly consisted of cleaning out sinkhole debris, grouting
and sealing the holes with clay and plastic liners, and
filling them with soil. Diversions and berms were
constructed around some sinkholes. To contract in
the RCWP, farmers had to implement fertilizer and
pesticide management BMPs.
Groundwater monitoring also intensified during
the mid-1980s. The annual nitrate sampling program
expanded into the groundwater recharge area and
included an additional 80 wells (160 total). The Min-
nesota Pollution Control Agency and State Depart-
ment of Agriculture began sampling wells for
pesticides, focusing on water quality in the Prairie du
Chien-Jordan aquifer. Results during 1985 and 1986
showed that 54 percent of sampled wells in the
groundwater recharge area and 21 percent in the
Garvin Brook watershed had nitrate-nitrogen levels
that exceeded 10 mg/L. Higher nitrate concentra-
tions observed in the groundwater recharge area as
compared to Garvin Brook watershed were probably
related to a greater number of wells in the
groundwater recharge area drawing water from the
unconfined karst Prairie du Chien group and land
use differences.
Ten of the high nitrate wells in the groundwater
recharge area were initially (1985) tested for pes-
ticides: six had detection levels of atrazine, alachlor,
or both. The range of atrazine concentration in four
wells was 0.79 micrograms per liter (i^g/L) to 5.57
Hg/L, and alachlor concentrations ranged between
0.04 and 1.23 ng/L
By 1988, groundwater monitoring had increased
to include
• 22 wells sampled quarterly for pesticides,
• 12 wells sampled for nitrate every five weeks,
• major ion chemistry performed quarterly,
• 160 wells sampled annually for nitrate, and
• lysimeters installed to better track the water
quality effects from specific land uses and
management.
Evaluating Changes in
Groundwater Quality
In addition to groundwater monitoring, vadose zone
monitoring and modeling helped evaluate changes in
water quality resulting from RCWP BMPs. Ground-
water age dating near the project area proved to be
an important tool for interpreting water quality
results.
Understanding Groundwater
Residence Times
When analyzing groundwater for .changes resulting
from land use alterations, it is important to under-
stand the age or residence time. Groundwater age
dating near the project area reported by Alexander
et al. (1987) and Alexander and Alexander (1987)
determined that the older bedrock aquifiers (such as
Franconia-Ironton-Galesville, Mt. Simon-Hinckley),
which were sampled as part of the original (1981)
RCWP monitoring network, contain mostly water
that predates modern agriculture.
Further age dating work by the Minnesota Pollu-
tion Control Agency in the Garvin Brook Area
showed that several Jordan Formation wells also
draw water that entered the ground before 1953.
These results were based on enriched tritium
analyses in wells of known construction. Tritium is a
radioactive isotope that is released into the atmos-
phere during nuclear testing. Atmospheric tritium
concentrations increased considerably during the
mid-1950s. Well water with nondetectable tritium
concentrations (< 0.8 tritium units [TU]) entered the
ground before 1953; water with > 0.8 TU contains at
least some post-1953 water; water with > 17 TU con-
tains all or mostly post-1953 water. Five of 24 Jordan
Formation wells sampled for tritium had pre-1953
water. All eight Prairie du Chien wells sampled for
tritium contained water with a larger fraction of post-
1953 water (see Fig. 4).
The nitrate-nitrogen concentrations in the pre-
1953 water were < 0.04 mg/L in four wells, and 1.1
mg/L in the fifth well (Table 1). These results are
very consistent with other nitrate/tritium relation-
ships developed in Minnesota (Minn. Pollut. Control
Agency and Minn. Dep. Agric. 1991). Wells with
elevated nitrate in Minnesota (e.g., > 3 mg/L) supply
water that has at least some recent (post-1953) com-
ponent. Many wells in the Garvin Brook area with
elevated nitrate probably reflect land use activities
after 1953 but before 1981, when the RCWP began.
Therefore, in this study, project evaluators had to
sample and analyze trend data from numerous wells
with elevated nitrate to obtain an understanding of
64
-------
D. WALL, M. EVENSON, C. REGAN, J. MAGNER, & W. ANDERSON
1100
1000
900
800~
700
W
Coon Valley Mbr
Jordan Sandstone
one mile
| Open interval of wells containing water with tritium concentrations >17 TU (most or all post-1953 water).
H Open interval of wells containing water with tritium concentrations 0.9-12. JU (mixed water).
Q -Open interval of wells containing water with tritium concentrations <0.8 TU (most or all pre-1953 water).
I Open interval of well containing water with no associated tritium analysis.
Numbers represent well water nitrate-nitrogen concentrations in mg/1.
Figure 4.—Conceptualized cross section showing the position of open boreholes in relation to bedrock straltigraphy.
Nitrate-nitrogen concentrations are shown for each of the 47 wells sampled in west central Winona County during
1990-91. Residence times are Indicated for 31 wells.
Table 1.—Tritium (TU) and nitrate-nitrogen (NOs-N) results in west central Winona County wells. All wells were
constructed after 1960 and have associated well logs indicating the aquifer from which water is withdrawn (JDN
= Jordan, PDC = Prairie du Chien). Median nitrate concentrations are significantly different (p < 0.01) among the
three groupings.
PRE-1953 WATER
AQUIF.
JDN
JDN
JDN
JDN
JDN
TRITIUM N03-N
< 0.8 < 0.01
< 0.8 < 0.01
< 0.8 0.02
< 0.8 0.03
<0.8 1.1
Mean NO3-N = 0.22 mg/L
Median NOa-N = 0.02 mg/L
MIXED WATER
AQUIF.
JDN
JDN
JDN
JDN
JDN
JDN
JDN
JDN
PDC
JDN
JDN
JDN
Mean
Median
TRITIUM
0.9
1.3
1.3
1.6
2.1
2.7
4.2
8.2
8.2
9.7
10.4
11.7
NOs-N =
NOa-N =
N03-N
3.2
2.9
2.7
0.3
2.7
0.3
2.1
2.8
3.2
4.6
2.4
3.0
2.5 mg/L
2.8 mg/U
ALL OR MOSTLY POST-1953 WATER
AQUIF. TRITIUM
PDC 17.3
JDN 18.8
PDC 19.9
JDN 21.3
PDC 25.6
JDN 27.0
JDN 28.0
JDN 28.3
PDC 29.4
PDC 30.0
PDC 33.0
PDC 40.0
JDN 40.0
JDN 42.5
JDN 42.5
Mean NOs-N = 5.9
NO3-N
3.6
6.6
4.7
4.0
1.6
3.5
6.4
10.8
2.8
12.0
8.5
2,0
7.6
7.6
6.9
mg/L
Median NOa-N = 6.4 mg/L
65
-------
Proceedings of national RCWP Symposium, 1992
groundwater quality changes resulting from the
post-1982 implemented RCWP BMPs.
Vadose Zone Monitoring
Complex hydrogeologic conditions of the karst
bedrock made it difficult to use domestic (bedrock)
wells when associating groundwater quality with
specific land uses. To more directly link land use and
water quality and better understand potential
groundwater quality improvements from the RCWP,
soil water samplers were installed at depths between
2 and 22 feet during 1988 and 1989. The primary ob-
jectives for installing the soil water samplers were to
• determine nitrate and pesticide concentra-
tions moving past the rooting zone in
cornfields that had been under RCWP con-
tract for two to four years, and
• evaluate nitrate and pesticide contributions
from agricultural fields, sinkholes, grassland,
woodland, and catchment ponds to better
characterize the major sources and pathways
of groundwater contamination.
Twenty-nine soil water samplers were installed
at 15 sites and sampled two to eight times. Four dif-
ferent types of soil water samplers were used, includ-
ing stainless steel pressure vacuum lysimeters,
BAT"* monitoring systems, glass block percolate
lysimeters, and wick percolate lysimeters. Water
samples were analyzed for pesticides, nitrates, and
major ion water chemistry. For more information on
lysimeter construction, locations, timing of sam-
pling, and specific results, refer to Wall et al. (1989),
Wall et al. (1990), and SCS et al. (1991).
Nitrate-nitrogen concentrations at four- to five-
foot depths below three RCWP cornfield sites
ranged between 20 and 50 mg/L, as compared to 0 to
7 mg/L below nearby grassland (3 lysimeters in
Conservation Reserve Program set-aside land) and
two forested sites. Since no pre-BMP implementa-
tion monitoring was conducted at the cornfield sites,
researchers did not know whether there had been an
increase or decrease in vadose zone nitrate con-
centrations. However, given the relatively high con-
centrations, either more stringent measures were
needed to reduce nitrate leaching or more time was
needed to observe the full benefits from imple-
mented nitrogen management BMPs.
Nitrate concentrations in sinkholes were quite
variable. Sinkholes surrounded by fertilized/
manured fields had high nitrate levels (20 to 100
mg/L) and those surrounded by urban land or
grassland had low levels (< 10 mg/L). Pesticide con-
centrations in one sinkhole (which were much
higher than pesticide leachate concentrations in
cornfields and in other sinkholes) received runoff
from the city of Lewiston. The maximum pesticide
concentrations found in this 2.5 feet deep sinkhole
lysimeter were atrazine 24.7 ppb, metolachlor
7.1 ppb, alachlor 12.3 ppb, and cyanazine 27 ppb.
Subsequent monitoring of runoff water from
Lewiston also showed elevated pesticide concentra-
tions. Comprehensive inspections of four commer-
cial pesticide applicator facilities in that city by the
Minnesota Department of Agriculture showed very
high pesticide residues in the facilities' soil. Eight
pesticides with concentrations over 1,000 ppb were
found in soil samples. The herbicides alachlor,
atrazine, cyanazine, metolachlor, and EPTC in soil
samples were each found at concentrations greater
than 70,000 ppb in at least one sample. EPTC — at
over a million ppb — had the highest concentration
of any pesticide in soil samples at the facilities. Both
on-site leaching at the facilities and runoff into the
sinkhole were likely avenues of pesticide contamina-
tion to groundwater near Lewiston.
The only wells with pesticide concentrations
known to exceed drinking water standards in the
Garvin Brook RCWP were located near Lewiston.
Pesticide point sources were probably the cause of
excessive pesticide concentrations in groundwater
and should be considered in future groundwater
protection efforts.
Modeling Lag Times for Groundwater
Response
The vadose zone component of the subsurface water
and solute transport (SWAST) model (Nieber and
Lopez-Bakovic, 1989) was used to attempt to answer
the following question: once nitrate concentrations
are reduced in the upper soil zone through nitrogen
management BMPs, how long will low nitrate water
take to move through the unsaturated zone into the
carbonate aquifer? An answer to this question helps
define the lag time between nitrogen fertilizer BMP
implementation and decreases in groundwater
nitrate concentrations.
The SWAST model, which was applied at three
representative sites in the groundwater recharge
area, calculates major components of water flow
through the soil matrix but does not incorporate
preferential flow through soil cracks, root channels,
or worm holes. It was assumed for this study that
much of the nitrate within the soil will not move out
of the system by preferential flow but through the
soil matrix. Given this assumption, we would not ex-
pect to achieve maximum water quality benefits from
nitrogen management BMPs until the soil matrix
66
-------
D. WALL, M. EVENSON, C. REGAN, J. MAGNER, & W. ANDERSON
nitrate moves through the soil profile and out of the
groundwater system.
The modeling was performed on a daily step
basis starting April 1, 1985, the beginning of
nitrogen management BMP implementation on
many farms in the Garvin Brook area. Soils in this
region are usually frozen and infiltration is minimal
between November 16 and March 31. The depth of
peak concentration movement through the soil
profile was determined for the periods 1985 to 1991
and 1985 to 2001. The following assumptions were
made:
• Soil moisture levels were at field capacity
each spring.
• Ninety-five-day corn was planted each year
on May 10.
• Each soil textural zone represents a
homogenous medium.
• Denazification did not affect the depth of
maximum nitrate level movement.
• No deep percolation occurred between
• November 15 and April 1.
The primary inputs to the model include soil
properties and climatic conditions. The thicknesses
of soil layers at the three sites are shown on Figure
5. Saturated hydraulic conductivities (Ks) were
determined using the relationships between soil tex-
ture and conductivity devised by McCuen et al.
(1981). Two simulations were conducted for each
profile: one using the mean Ks for each soil texture
determined by McCuen et al. and one using a Ks
value three standard deviations above the mean. Ac-
tual temperature and precipitation data for 1985 to
1991 were input into a climate generator, Cligen
(Arlen Nicks, Agricultural Research Service, Ada,
OK), which then computed other climatic variables
needed for that period. Cligen was used to randomly
generate all climatic data for 1992 to 2001.
Modeling outputs included annual records of
evapotranspiration and deep percolation. Evapo-
transpiration ranged from 40 to 50 cm/year and deep
percolation ranged from 0 to 18 cm/year, reasonable
values for the area modeled. The final depths of peak
concentration movement through the soil profile for
the 7 and 17 year runs are shown in Figure 5. For the
17-year simulations at all three sites and for both
conductivity values, the depth of peak concentration
migration ranged from 5.9 to 8.9 feet. This small
range, despite large differences in Ks, reflects the in-
fluence of evapotranspiration in maintaining soil
moisture levels below field capacity, thereby greatly
reducing the effective hydraulic conductivity. The
predicted depths of peak concentration movement
Sitel
Site 2
7yr. 17 yr. S1L
SiL
siL-sio.
SaCL
SaL
SaCL
SaL
1ft.
12ft.
4ft.
6ft.
4ft.
2ft.
7yr.
IJ
Ave.K(s)
High K(s) (three
std. deviations
above ave.)
Figure 5.—Predicted depth of peak nitrate movement
through the soil matrix at three Garvin Brook RCWP
sites. Four runs of SWAST were conducted for each
profile varying the length of time (1985-91 and 1985-
2001) and the saturated hydraulic conductivity (average
and high values for the soli's textural classes).
for the 1985 to 2001 simulations using the higher
conductivity value were 23, 54, and 74 percent of the
soil profile thickness at the three sites. Thus, for the
portion of nitrate that does not move with preferen-
tial flow, residence times to bedrock are estimated to
range from 15 years to 60 years for most Garvin
Brook area soils.
After percolating water reaches the bedrock,
there will be a lag in the time it takes to move
through 50 to 150 feet of carbonate materials to the
water table. Given the great amount of fracturing and
solution channel development in the karstic bed-
rock, water movement through these materials is
believed to be very fast. However, water movement
into bedrock microfractures may have much longer
residence times.
Groundwater Nitrate Trend Results
Time trend analysis was conducted on data
produced from annual nitrate sampling of numerous
wells in the Garvin Brook watershed and the
groundwater recharge area. Sampling was con-
ducted by Winona County Cooperative Extension
and chemical analyses performed at the Minnesota
Department of Agriculture's laboratory. Well sam-
pling occurred each year between late May and early
67
-------
Proceedings of National RCWP Symposium, 1992
Table 2.—Results of Kendall's tau test of correlation between nitrate and time. Statistics were conducted on data
obtained from annual sampling in Garvin Brook Watershed and the groundwater recharge area between 1983
and 1990.
AREA
NITRATE
RANGE
#OF
WELLS
STATISTIC
#NO
TREND
# INCREAS-
ING
# DECREAS-
ING
Garvin Brook Watershed
Garvin Brook Watershed
Groundwater recharge area
Groundwater recharge area
Garvin Brook Watershed
Garvin Brook Watershed
Groundwater recharge area
Groundwater recharge area
3-10mg/L
3-10 mg/L
3-10 mg/L
3-10mg/L
> 10 mg/L,
> 10 mg/L
> 10 mg/L
> 10 mg/L
35
35
25
25
16
16
35
35
p < 0.05
p<0.10
p < 0.05
p<0.10
p < 0.05
p<0.10
p < 0.05
p<0.10
27 (77%)
24 (69%)
22 (88%)
20 (80%)
10(62.5%)
8 (50%)
30 (86%)
27 (77%)
4(11.5%)
6(17%)
3 (12%)
4 (16%)
0 (0%)
0 (0%)
0 (0%)
0 (0%)
4(11.5%)
5 (14%)
0 (0%)
1 (4%)
6 (37.5%)
8 (50%)
5 (14%)
8 (23%)
July and took about two weeks. During the first three
years of testing, wells chosen for trend analysis had
average nitrate-nitrogen levels that exceeded 3
mg/L and were sampled for a minimum of five con-
tinuous years. Wells with low nitrate (< 3 mg/L)
during the first three years of sampling were not
analyzed for trends because low nitrate water had
been shown to represent water that had entered the
ground before 1953.
Fifty-one wells in Garvin Brook watershed met
the selection criteria: eighty-eight percent were
sampled each year between 1983 and 1990; the
remaining 12 percent were sampled from 1985 to
1990. Sixty wells in the groundwater recharge area
met the data selection criteria. Eighty-three percent
were sampled each year from 1985 to 1990, nine
wells (15 percent) were sampled from 1986 to 1990,
and one well (2 percent) was sampled from 1983 to
1990.
Kendall's tau test, which is a nonparametric test
of correlation, was used to measure the direction
and statistical significance of observed trends. Statis-
tical analysis results are shown in Table 2 and Figure
6. Two measures of statistical significance are noted
in Table 2: p < 0.05 and p < 0.10. Significance levels of
p < 0.05 or p < 0.10 indicate that there is less than a 5
or 10 percent probability, respectively, of obtaining a
slope as different from zero as the observed slope if
in fact there is no relationship between parameter
value and time. Significance levels of p < 0.05 and
p < 0.10 are commonly used in scientific work to
allow reasonable certainty that a relationship be-
tween parameter and time really exists. Trend
results for Garvin Brook watershed and the
groundwater recharge area were similar. In both
areas, trend analysis results differed greatly when
comparing wells that had nitrate-nitrogen between 3
and 10 mg/L and those with greater than 10 mg/L
during the first three years of testing. Seventy-three
percent or 44 wells with nitrate-nitrogen between 3
and 10 mg/L had no significant trend (p < 0.10); 17
percent or 10 wells had statistically significant in-
creasing trends; and 10 percent or six wells had sig-
Nitrate-N 3-10 mg/l
60 wells
Nitrate-N >10 mg/l
51 wells
No significant
trend (p<0.10)
Increasing trend
Decreasing trend
Flgura 6.—Nitrate-nitrogen concentration trends from wells sampled annually (late May to early July) in Garvin Brook
Watershed and the groundwater recharge area for five to eight years (between 1983 and 1990). Kendall's tau test of
correlation was used to determine the significance and direction of trends for the two categories of wells.
68
-------
D. WALL, M. EVENSON, C. REGAN, J. MAGNER, & W. ANDERSON
nificant decreasing trends. However, when consider-
ing only wells that had over 10 mg/L during the first
three years of sampling, the only statistically sig-
nificant trends were decreasing trends (p < 0.10).
Sixty-nine percent of 51 high nitrate wells had no sig-
nificant trend and 31 percent had a decreasing trend.
The major differences in trends between mid-
range and high-nitrate wells could be related to
groundwater residence times. High nitrate-nitrogen
wells (> 10 mg/L) tend to be very sensitive to sur-
face contamination, have short residence times, and
are likely to be more directly affected by local
nitrogen sources. Therefore, the water quality in
these wells may reflect more recent land use
changes and climatic conditions. Conversely, wells
with nitrate-nitrogen between 3 and 10 mg/L tend to
be somewhat less sensitive to surface contamination
and may be less responsive to recent land use and
climatic conditions compared to the high nitrate
wells.
It is unknown whether the observed decreases
in the high-nitrate wells are related to BMPs,
climatic conditions, or a combination of many fac-
tors. There were two extended periods of precipita-
tion extremes during the project duration. From
1981 to 1984, total precipitation was 13.2 inches
greater than normal. Between 1987 and 1989, total
precipitation was 12.0 inches less than normal. It is
possible that during the dry period between 1987
and 1989, nitrate concentrations decreased as a
result of decreased recharge into the aquifer simul-
taneous with denitrification and/or mixing of lower
and higher nitrate waters within the aquifer. Vadose
zone modeling and monitoring results suggested
that the full benefits of nitrogen management BMPs
have not yet been realized. However, it is possible
that nitrate movement into the aquifer through
preferential flow paths was reduced by sinkhole
treatment projects, sealing abandoned wells, and
nitrogen fertilizer management BMPs. Continued
monitoring of these wells over the next 10 to 15
years should help to determine whether the trends
are more likely the result of climatic variability or im-
plemented BMPs.
Conclusions
The Garvin Brook RCWP project maintained enough
flexibility to allow for programmatic adjustments
throughout its duration. Groundwater monitoring
and hydrogeologic analysis prompted the adjust-
ments, which included expansion of the project area
to include all of the groundwater recharge area for
Garvin Brook watershed and a shift in emphasis
from surface water to groundwater protection BMPs
and monitoring.
In future similar studies, a diagnostic study
should be conducted before developing implementa-
tion plans and installing expensive BMPs. If time and
resources allow, the diagnostic study should deter-
mine
• pollutants of greatest concern in surface and
groundwaters,
• specific land uses and areas responsible for
the water quality degradation,
• baseline water quality before land use
changes,
• groundwater recharge area boundaries,
groundwater residence times, and other
hydrogeologic characteristics,
• the flow system and pollutant pathway—
especially as they relate to surface and
groundwater interactions, and
• temporal and spatial variability of water
quality within the aquifer.
A long-term water quality monitoring strategy
and information protocol should be developed at the
very beginning of a project. The strategy and
protocol should consider
• monitoring before, during, and after BMP
implementation,
• how to assess attainment of water quality
goals,
• monitoring overall resource response,
• focused monitoring to assess the effect of
specific BMPs,
• documentation of assumptions and working
hypotheses so others can follow the decision-
making process,
• clearly defining who needs the monitoring
information, and
• how best to communicate monitoring
activities to the target audience.
Vadose zone monitoring and modeling results
suggest that major nitrate loading reductions into
the carbonate aquifer are not yet realized in the Gar-
vin Brook area. Modeling results indicated that the
complete water quality benefits from fertilizer
management BMPs will not be observed until after
the year 2000.
Residence time analyses showed that many
wells in the project area withdraw water that entered
69
-------
Proceedings of National RCWP Symposium, 1992
the ground before 1953. Water in these wells had
very low nitrate; most of the recent water wells had
elevated nitrate. Therefore, to better understand the
data from the RCWP, many high nitrate wells were
analyzed. Trend analysis of high nitrate-nitrogen
(> 10 mg/L) wells indicates generally decreasing
nitrate concentrations between 1983 and 1990.
Trend results were much more variable for wells
with nitrate-nitrogen in the 3 to 10 mg/L range. It is
not known whether the decreasing concentrations in
high nitrate wells are the result of implemented
BMPs, climatic conditions, or a combination of many
factors. Another 10 to 15 years of water quality
monitoring will probably be needed to determine
trends associated with BMP implementation. Vadose
zone monitoring should be considered in future
studies to determine the effect of BMPs within a
shorter time frame.
References
Alexander, E.G. and S.C. Alexander. 1987. A chemical and isotopic
survey of the age or residence times in groundwater in
Rochester and Olmsted County. Rep. City of Rochester and
Olmsted County, MM.
Alexander, B.C., S.C. Alexander, and R.S. Lively. 1987. Recharge of
the Mt Simon Hinckley aquifer: responses to climatic
change and water use. Earth Sci. J. Am. Geophys. 64
(44):1270.
Kanivetsky, R. 1984. Bedrock hydrogeology. /» N.H. Balaban and
B.M. Olsen, eds. Geologic Atlas for Winona County, Atlas C-
2. Minn. Geo. Surv., Univ. Minn., St Paul.
McCuen, R.H., W.J. Rawls, and D.L Brakensick. 1981. Statistical
analysis of the Brooks-Corey and the Green-Ampt
parameters across soil textures. Water Resour. Res.
17(4):1005-13.
Minnesota Geological Survey. 1984. Geologic Atlas for Winona
County, Atlas C-2. N.H. Balaban and B.M. Olsen, eds. Minn.
Geo. Surv., Univ. Minn., St. Paul.
Mossier, J.H. and Book P.R. 1984. Bedrock geology. In N.H.
Balaban and B.M. Olsen, eds. Geologic Atlas for Winona
County, Atlas C-2. Minn. Geo. Surv., Univ. Minn., St. Paul.
Minnesota Pollution Control Agency and Minnesota Department
of Agriculture. 1991. Nitrogen in Minnesota Groundwater.
Rep. Prep. Legislative Water Comm., St Paul.
Nieber, J.L. and I.L. Lopez-Bakovic. 1989. Subsurface Water and
Solute Transport (SWAST) Model. Univ. Minn. Agric. Eng.
Dep., St. Paul.
Soil Conservation Service, Minnesota Pollution Control Agency,
and Winona County Extension. 1991. Special 10-year Report
— Garvin Brook RCWR Lewiston, MN.
Wall, D.B., SA McGuire, and JA Magner. 1989. Water Qualify
Monitoring and Assessment in the Garvin Brook Rural Clean
Water Project Area. Minn. Pollut. Control Agency, Water
Qual. Div., St Paul.
——. 1990. Nitrate and pesticide contamination of ground water
in the Garvin Brook areas of southeastern Minnesota:
sources and trends. Pages 113-27 in Agricultural Impacts on
Ground Water Quality. Natl. Water Well Ass. Conf., Kansas
City, MO.
Young, RA/CA Onstad, D.D. Bosch, and W.P. Anderson. 1987.
AGNPS: Agricultural Nonpoint Source Pollution Model — A
Watershed Analysis Tool. Conserv. Res. Rep. 35. U.S. Dep.
Agric., Washington, DC,
70
-------
Keeping Bacteria Out of the Bay
The Tillamook Experience
J.A. Moore
Department of Bioresource Engineering
Oregon State University
Coruallis, Oregon
R. Pederson
Soil Conservation Service
Tillamook, Oregon
J. Worledge
Agricultural Stabilization and Conservation Service
Tillamook, Oregon
ABSTRACT
Bacterial pollution in shellfishing areas is not uncommon along the coasts of the United States.
Elevated bacterial numbers and the potential for disease have forced several fishing areas to close
and implicated point and nonpoint sources of surface water contamination. Tillamook Bay, along
the Oregon coast, presents an ideal location to study these problems. An important shellfishing in-
dustry is located in the bay, surrounded by a large dairy industry. High bacterial counts have
forced the bay to close for harvesting on numerous occasions. The efforts of the Rural Clean Water
Program (RCWP) have reduced bacterial pollution in the bay, but the challenge continues.
This paper summarizes the Tillamook Bay
Rural Clean Water Program (RCWP) project.
It includes a general literature review on
water quality problems in shellfishing areas; a
description of the location, climate, and hydrology of
the Tillamook basin; some history on the develop-
ment of the water quality problem in the basin; and
the process used to develop farm plans and motivate
dairy farmers in the drainage basin to implement
best management practices (BMPs) for achieving ac-
ceptable water quality. The results and impacts of 10
years' efforts to keep bacteria out of the bay will also
be discussed.
Literature Review
The problem of fecal pollution in estuarine areas can
be defined by two questions: (a) where do the bac-
teria come from; and (b) how can their numbers be
reduced so that natural shellfish resource areas will
not diminish. Much research has been conducted to
determine the sources of fecal contamination in
rivers and estuaries. Runoff from pastures and other
agricultural land has been implicated as a major bac-
teriological source by many researchers (Carney et
al. 1975; Doran and Linn, 1979; Dudley and Karr,
1979; Faust and Goff, 1978; Feachem, 1974). Mc-
Donald and Kay (1981) hypothesized that bacteria
collect on the surface between rainfalls and are
flushed by runoff into rivers and streams. Recrea-
tional areas without sanitary waste facilities also in-
crease the levels of fecal bacteria in adjacent streams
(Varnessetal. 1978).
Bacterial pollution of estuaries and oyster beds
can be caused by urban area flooding (Mackowiak et
al. 1976), marinas and pleasure boats (Mack and
71
-------
Proceedings of National RCWP Symposium, 1992
D'ltri, 1973), upstream activities (Sayler et al. 1975),
and duck production areas (Davis et al. 1966; Gates,
1963). A serious problem was encountered in the
Newport River in North Carolina; all shellfish beds
in the river showed unacceptable levels of bacteria
generated from wildlife, sewage treatment plants,
agricultural runoff, and faulty septic systems (Ga.
Environ. Prot. Agency, 1972).
The Tillamook area provided a good setting to
conduct a water quality study for several reasons:
• it is a small, enclosed river basin;
• it is not predominantly urbanized and
contains within its boundaries many bacterial
sources;
• it has a seasonally distributed, high annual
rainfall (2,337 mm), which increases bacterial
pollution in the bay through runoff;
• its bay is shallow enough to allow interaction
between freshwater inflows and tidal
flushing; and
• its basin contains five different watersheds
with multiple activities and varying influence
in water quality.
Area Description
The Tillamook watershed is located on Oregon's
northwest coast, approximately 116 km west of
Portland (Fig. 1). The city of Tillamook has a popula-
tion of 4,000 and is the largest city in the county. Til-
lamook Bay is approximately 16 km long and 5 km
wide, with a surface area of roughly 46 km at high
tide. Surrounding the bay to the southeast is an area
of tidal floodplains and river terraces. The floodplain
is created by the Tillamook, Wilson, Trask, and Kil-
chis rivers at their entrance to the bay. The fifth river,
the Miami, joins the bay along a narrow floodplain
much nearer the mouth of the bay. Floodplains in the
basin are used for urban development and agricul-
ture, predominantly dairy farming. Steep forested
uplands surround the entire bay, floodplain, and
river terraces. The local cheese plant uses 30 per-
1
o
o
o
c
o
• SEWAGE TREATMENT PLANT LOCATIONS
Figure 1.—Location of Tillamook Bay In Oregon and the river systems that flow Into the bay.
72
-------
JA MOORE, R. PEDERSON, 6 J. WQRLEDGE
cent of the milk produced in Oregon. As a result,
over 210 dairies with 28,000 cow-units are based in
the county; over 16,000 cow-units are located in the
Tillamook drainage basin.
Climate and Hydrology
The Tillamook area climate is dominated by a strong
marine influence from the Pacific Ocean. This in-
fluence leads to wet winters and dry summers, with
a comparatively narrow range of seasonal tempera-
tures. Frequent intense winter storms account for
major rainfall in the area, which averages from 3.8
km on forested uplands to 2.3 m annually in the city
of Tillamook. Freezing temperatures infrequently
occur near the bay, and average temperatures range
from 5°C in January to 15°G in July.
River discharge into the bay varies considerably
by season. Average summer flows of the five rivers
total approximately 0.5 m3/sec. Average winter flows
total 28 m3/sec. The estimated 50-year flood would
produce a combined average flow of 85 m3/sec
Qackson and Glendening, 1982).
Building Solutions
Oregon has been a leader in using interagency
cooperation to achieve a variety of major conserva-
tion goals. The Conservation Resource Management
(CRM) process is the organizational structure that
involves all Federal, State, and local agencies, coor-
dinates their efforts, and provides local landowners
with information and assistance to solve conserva-
tion problems in their communities. We learned in
the Tillamook project that interagency sharing is not
enough; the landowner is the key to a successful
project and cannot be left out of the planning
process.
In the early part of this project, dairy farmers in
the Tillamook Bay drainage basin were not aware
they were the major cause of pollution in the bay.
They had been using advice from the Soil Conserva-
tion Service (SCS) and the Cooperative Extension
Service (CES) in their manure-handling efforts for
years. They had confined the manure to large stacks
outside the barns so that it could be spread in the
spring when the threat of flooding had passed. Some
farmers had even constructed small underground
liquid manure tanks with Agricultural Stabilization
and Conservation Service (ASCS) cost-sharing funds
to hold and then spread their liquid waste on nearby
fields. However, monitoring by the Oregon Depart-
ment of Environmental Quality (DEQ) and the U.S.
Food and Drug Administration (FDA) showed that
coliform levels in the bay were sometimes above safe
levels for commercial shellfish production for inter-
state commerce. Bay closings were frequent, unpre-
dictable, and of great concern among oyster
growers.
Water Quality History
The first indication of bacterial contamination
problems were identified in routine monitoring of
the bay from 1969 to 1971. During heavy rains, water
quality dropped below acceptable levels and resulted
in threats of FDA intervention. Additional studies
from 1972 to 1974 and again in 1976 and 1977 docu-
mented an extensive problem. A threat from FDA to
close the bay led to the formation of several task
forces on the State and local levels to address the
problem.
Funding under a Clean Water Act, section 208
grant directed resources to solve the problems in Til-
lamook Bay. Comprehensive monitoring was ini-
tiated in 1979 and 1980 to identify the sources and
extent of the bacterial problem.
Weather conditions were evaluated to determine
if and how hydrologic extremes in the bay and
drainage basin influence bacterial levels. The results
showed that coliform standards were exceeded in
the bay and tributaries during intense storms. The
Oregon DEQ developed a list of possible sources of
fecal pollution, including failing septic systems,
recreational areas, forested areas, animal (dairy)
wastes, and malfunctioning sewage treatment plants.
By relating the land use with the sampling locations
and results, the major fecal pollution sources were
identified as
• management of dairy manures,
• hydraulically overloaded sewage treatment
plants, and
• failing on-site domestic waste systems.
Agency Activity and
Cooperation
The DEQ and the Tillamook County Health Depart-
ment began to address problems from the failing
septic tanks and overloaded sewage treatment
plants; CES, SCS, and ASCS began work on reducing
the bacteria pollution from the dairy manure
management systems.
CES, SCS, DEQ, and the local Soil and Water
Conservation District (SWCD) cooperated in an
educational effort to convince the dairy community
73
-------
Proceedings of National RCWP Symposium, 1992
it was a major part of the bay's water quality problem
and to demonstrate, discuss, and explain manage-
ment practices that reduce or eliminate bacterial pol-
lution from existing manure management systems.
SCS and ASCS offered options of technical assis-
tance and cost-share programs that would help
finance control measures. In 1981, the area was
selected to participate in the RCWP, which provided
additional financial support for the technical, educa-
tional, and cost-share portions of the facilities im-
provement program for individual dairy farms. The
SCS field office added staff to accelerate the plan-
ning, design, construction, and inspection of agricul-
tural waste facilities.
As the RCWP project began, dairy farmers in
the Tillamook Soil and Water Conservation District
and the county ASCS committee were convinced that
the time had come for dairy farmers to adopt new
practices. Many of these practices were directed to
reduce runoff of manure and contaminated water
into the hundreds of drainage ditches, small
streams, and tributaries flowing into the five major
rivers that empty into Tillamook Bay. At first, the
task appeared impossible, especially when dairy
farmers learned that their original practice of storing
manure for 30 days was no longer adequate.
Through a highly effective information program
by all agencies, many hours of personal contacts by
leaders in the dairy community, and strong support
from the local Tillamook County Creamery Associa-
tion, the RCWP project began. Within the first year,
over 90 percent of the 120 dairy operations in the
basin agreed to participate in the project. A rating
system was developed to rank and identify those
operations that should receive assistance first. A
computer program was written to calculate the fate
and movement of bacteria through the various dairy
operations and manure management practices
(Moore et al. 1981). Input data (site-specific for soil
type, farm size, number of animals, and manure col-
lection type) were used to help the county ASCS
committee and the SWCD board rank operations to
allow treatment of the worst cases first and to
evaluate the BMPs that could best achieve the
largest reduction in bacterial movement and losses.
One task was to develop a list of appropriate
BMPs. In one BMP, for example, large above-ground
tanks (up to 80 feet in diameter and 12 feet high)
were designed by SCS. Special foundations were
often required in boggy lowlands near the bay. A
roofing BMP was developed to divert clean water
from contaminated areas; the roof covered the cattle
walkways and open concrete slabs, where cows wait
to enter the milking parlor. It proved more economi-
cal to divert clean water with roofing than to collect,
store, and handle large volumes of contaminated
water. A BMP for gutters, downspouts, and under-
ground drains was also necessary to move clean
water away from barnyard areas. "Speed bump"
curbing was designed for the edges of concrete hold-
ing areas to contain polluted water, yet allow access
of equipment and livestock. Drain tile systems were
designed and installed to lower the water table so
that animal waste could be applied on pastures
during predicted dry periods.
A challenge faced by the Extension Service and
SCS technicians was how to teach new and improved
manure management methods to dairy farmers. One
major BMP that deals with manure storage manage-
ment takes advantage of "windows" in the weather or
spreading opportunities to apply manure. This ap-
proach allows greater use of manure as a fertilizer
and encourages using less commercial fertilizer.
This project was the first time that BMPs were
developed to reduce the movement and water quality
impact of bacteria from a manure management sys-
tem. While many BMPs relate to physical and chemi-
cal contaminants, few had been applied to address
bacterial movement. The major thrust in research
has been to identify the sources of bacteria, with far
less emphasis on source control. However, any solu-
tion to the problem must consider not only how to
determine when shellfish harvesting in an affected
area has become a public health hazard but also
how we can reduce the magnitude of the bacterial
source.
Project Results
In the past 10 years, 105 operations within the Til-
lamook Basin have participated in the RCWP project
and agricultural conservation programs. Today, 80
percent of the initial construction projects have been
completed. Several key dairies are still completing
major BMPs that will affect runoff quality.
The project's cost figures are important and can
be presented in several ways. The following items
provide some summary statements about initial,
operating, and per ton of manure costs for the total
project: (1) $107.88 was spent per 1,000-pound
animal unit for implementation of BMPs; (2) $13.42
was the average annual operating cost for manure
management systems per animal unit; and (3) it cost
an average of $1.07 to handle a ton of manure. These
average annual implementation costs are equivalent
to the cost of a loan amortized over the life of each
BMP. They do not include the annual operation and
maintenance costs of the manure management sys-
tems. Approximately $8,500,000 will be spent to im-
74
-------
J.A. MOORE, R. PEDERSON, & J. WORLEDGE
prove facilities by the end of the project. Of this
amount, $4,540,278 are RCWP cost-share funds ad-
ministered through ASCS.
In addition to improving the bay's water quality
and creating a positive sense of community pride,
the expenditures have been a boost to the local
economy. Many good things unrelated to water
quality have resulted from efforts to address the
water quality problem. The successes of this project
have caught the attention of dairy farmers
throughout the United States, especially those living
in coastal areas where high rainfall complicates
waste handling. Numerous tours for dairy farmers
and technicians of all agencies have been held in the
Tillamook Bay Drainage Basin, and this activity con-
tinues today. In Oregon, the lessons learned from
this project are now being used in special water
quality projects along the further reaches of the
Oregon coast and in the State's other dairy areas.
Water Quality Today
Although much positive improvement has occurred,
the water quality problem has not been solved.
Monitoring continues in the bay, but the bay is still
closed occasionally during major storms because of
high levels of fecal coliform. Some facilities are yet to
be constructed, and some poor management prac-
tices have been observed. The problems identified in
this watershed were extensive and will require a con-
tinuing effort to reach success.
Unfortunately, the monitoring strategy was
changed during the project and further handicapped
by inadequate background and limited data. A sam-
pling strategy should be established and followed
throughout the project's life. At a minimum, the
parameter set must include data on fecal coliform
bacteria, bay salinity, river flow, and rainfall. Statisti-
cally valid trends have not been successfully estab-
lished using the current data.
All data were entered into EPA's STORET sys-
tem. Data dumps from STORET were imported into
spreadsheets for some analysis and into a statistical
software package for nonparametric analysis. Data
points represent discrete grab samples collected on
a given day. For evaluating trends, data were parsed
to monthly or quarterly values. In some cases, data
from multiple sites were pooled. Frequency his-
tograms were also developed. Trends were es-
timated using the seasonal Kendall test (a
nonparametric trend indicator) and, in some cases,
ordinary least squares regression (a parametric
technique). The ordinary least squares regression
was used for comparison because it is a well-known
test. It is not an optimal test for the data set, however,
because it assumes a normal distribution of optional
data.
Wilson River is the largest river flowing into Til-
lamook Bay (as seen in Fig. 1). One of the sampling
stations is near town, just before it enters the bay.
Figure 2, a plot log of fecal coliform vs. time that
shows the trend but no significant difference, is in-
cluded to give the reader a sense of the variation and
direction of water quality changes. Careful, ongoing
tracking of water quality trends will, however,
• allow better documentation of the water
quality improvements;
« make it possible to take timely enforcement
actions;
• identify systems that require maintenance,
enlargement, or management changes; and
• help refine manure management techniques.
Indications exist that water quality improve-
ments have occurred in Tillamook Bay and its
tributaries. Upstream-downstream sampling can
identify farm locations but, to date, not enough data
have been collected to evaluate post-project BMP in-
stallations. It is important to continue the current
monitoring efforts so that future threats to Til-
lamook Bay's water quality can be quickly identified.
References
Carney, J.S., C.E. Carty, and RR. Colwell. 1975. Seasonal occur-
rence and distribution of microbial indicators and pathogens
in the Rhode River of Chesapeake Bay. Applied Microbiol.
30:771-80.
Davis, R.V., C.E. Cooley, and AW. Hadder. 1966. Treatment of
duck wastes and their effects on water quality. Pages 90-150
in Management of Farm Animal Wastes. Proc. Natl. Symp.
Animal Waste Manage. Mich..State Univ., East Lansing.
Doran, J.W. and D.M. Linn. 1979. Bacteriological quality of runoff
water from pastureland. Applied Environ. Microbiol. 37:985-
91.
Dudley, D.R. and J.R. Karr. 1979. Concentration and sources of
fecal and organic pollution in an agricultural watershed.
Water Resour. Bull. 15:911-23.
Faust, MA and N.M. Goff. 1978. Sources of bacterial pollution in
estuary. Pages 819-39 in Coastal Zone 78 — Symposium on
Technical, Environmental, Socioeconomic and Regulatory
Aspects of Coastal Zone Management Am. Soc. Civil Eng.,
New York.
Feachem, R 1974. Fecal coliforms and fecal streptococci in
streams in the New Guinea highlands. Water Res, 8:367-74.
Gates, C.D. 1963. Treatment of Long Island duck farm wastes. J.
Water Pollut Control Fed. 35:1569-75.
Georgia Environmental Protection Agency. 1972. A Report on Bac-
teriological Pollution Affecting Shellfish Harvesting in New-
port River, North Carolina. Surveillance Anal. Div., Athens.
75
-------
Proceedings of National RCWP Symposium, 1992
WILSON RIVER AT HIGHNAY 101
ALL SEASONS
OLS Slope
Seasonal Sen Slopa
SEASONAL KENDALL
Slope « -0.00790
Nat Slgnlf B0X
P « 0.2371
B3 B4 BS SB B7 BB B9 30
79
Figure 2.—Plot of changes In fecal conform counts against time In Wilson River, the major river entering Tlllamook
Bay.
Jackson J.E. and EA Glendening. 1982. Tillamook Bay Bacteria
Study, Fecal Source Summary Report Oreg. Dep. Environ.
Qual., Portland.
Mack, WN. and R.M. D'ltri. 1973. Pollution of a marine area by
watercraft use. J. Water Poll. Control Fed. 45:97-104
Mackowiak, PA, C.T. Caraway, and B.L Portnoy. 1976.
Oyster- associated hepatitis: lessons from the Louisiana ex-
perience. Am. J. Epidemiol. 103:181-91.
McDonald, A. and D. Kay. 1981. Enteric bacterial concentration in
reservoir feeder streams: base flow characteristics and
responses to hydrographic events. Water Res. 15:961-68.
Moore, JA, M.E. Grismer, S.R. Crane, and J.R. Miner. 1981.
Evaluating dairy waste management systems influences on
, fecal coliform concentration in runoff. Bull. 658. Dep. Agric.
Eng., Oreg. Agric. Exp. Sta., Oreg. Univ., Corvallis.
Sayler, G.S., J.D. Nelson, A Justice, and R.R. Colwell. 1975. Dis-
tribution and significance of fecal indicator organisms in
upper Chesapeake Bay. Applied Microbiol. 30:625-38.
Varness, K.J., R.E. Pacha, and R.F. Lapen. 1978. Effects of dis-
persed recreational activities on the microbiological quality
of forest surface water. Applied Environ. Microbiol. 36:94-
104.
76
-------
A Tracking Index for Nonpoint Source
Implementation Projects
Steven A. Dressing
Office of Water
U.S. Environmental Protection Agency
Washington, D.C.
John C. Clausen
Department of Natural Resources Management and Engineering
University of Connecticut
Storrs, Connecticut
Jean Spooner
NCSU Water Quality Group
North Carolina State University
Raleigh, North Carolina
ABSTRACT
One of the challenges of the Rural Clean Water Program (RCWP) is how to appropriately report
program success and summarize the findings of 21 different projects. Measures of success often
address best management practice implementation, institutional achievements, public involve-
ment, modeling projections, and economics. Water quality and the protection and restoration of
beneficial uses, however, constitute the bottom line for RCWP projects. A nonpoint source index
(NPI) is proposed to help identify water quality status in projects before and after BMP implemen-
tation. Hie NPI ranges from 0 to 100 and is based on measures of use attainment, water quality con-
ditions, and level of land management. The NPI score is calculated as the sum of the scores for
each of four subindices that address these measures. By tracking NPI scores over time, trends in
land management and water quality can be assessed and reported. Analysis of the scores for any of
the subindices can shed some light regarding further nonpoint source treatment needs. Applica-
tions of the NPI to select RCWP projects are presented. This index should be useful to the general
public, managers, and the scientific community.
Simple and clear reporting of progress made
in improving or protecting water quality and
designated beneficial uses of waters for
watershed implementation projects has generally
been difficult to achieve for water quality profes-
sionals. Problems stem from the variety of informa-
tion typically generated by a watershed project to the
variety of ways in which such information can be
evaluated, to lack of consensus regarding the most
meaningful ways to report the information.
Traditionally, projects have reported water
quality monitoring and implementation data to docu-
ment the progress made in meeting water quality ob-
jectives. Recently, more emphasis has been placed
77
-------
•dings of National RCWP Symposium, 1992
on documenting the relationship of pollution control
implementation to water quality (Hopkins and
Clausen, 1985; Meals, 1991), but few have docu-
mented such relationships with statistical confidence
(Spooneretal.1990).
Statistical analyses of water quality data can be
interpreted by the relatively small number of water
quality experts across the Nation, but most program
managers, decisionmakers, and the general public
often find little reward in reviewing such informa-
tion. The problem is not so much that the informa-
tion is abstract or complex, which it often is, but that
the presentation of the information does little to
make it more palpable.
Similar problems have been faced in the past.
For example, a new method of reporting the trophic
status of lakes was developed to provide better com-
munication with the general public about the current
and potential status of lakes (Carlson, 1977). Carlson
developed a trophic state index (TSD with a range of
0 to 100 that could be calculated as a single index
using a range of measured parameters as input.
Other indices have been developed to address the
biological status of streams (Karr, 1981) and the
status of habitat (U.S. Environ. Prot. Agency, 1989).
This paper presents an index for summarizing
water quality status in projects before and after im-
plementation of best management practices (BMPs).
The nonpoint source index (NPI) ranges from 0 to
100 and is based on measures of use attainment,
water quality conditions, and level of land manage-
ment. The NPI score is calculated as the sum of the
scores for each of four subindices that address these
measures. By tracking the NPI scores over time,
trends in land management and water quality can be
assessed and reported. Analysis of the scores for any
subindex can shed some light regarding further non-
point source treatment needs. This index is a tool for
presenting data; it does not reduce data needs.
Development of the Index
The NPI is intended to provide a clear and meaning-
ful measure of a watershed's status with respect to
identified water quality problems. It is intended to
help identify water quality status in watershed
projects before and after BMP implementation. The
NPI is not intended to be used to compare different
watersheds, but users may wish to develop stand-
ardized scoring schemes that allow for such com-
parisons, x
Index parameters were selected that can be
measured at a reasonable cost. Unlike more focused
indices, such as the TSI (Carlson, 1977) and the
index of biological integrity (IBI) (Karr, 1981), the
NPI uses interpreted data rather than raw data. For
example, where the TSI uses Secchi disk, chloro-
phyll a, or total phosphorus data, the NPI uses
measures such as beneficial use support status and
the fraction of sample values meeting certain
numeric criteria. Given the variety of designated
beneficial uses for waters (e.g., drinking, fishing,
wildlife, industrial) and the variety of chemical,
physical, and habitat parameters that can be
monitored, it is impossible to provide a generalized
watershed index using raw data that faithfully repre-
sents the range of information generated by any
given project. It is, however, possible to specify the
process by which projects can analyze raw data to
provide interpreted data as input parameters for the
NPL
The NPI
The NPI ranges from a score of 0 to 100, with higher
scores indicating better water quality and greater
pollution control in the watershed. The scores are
structured to represent extremes that will not be ex-
ceeded by any given watershed. The NPI is a mathe-
matical combination of two separate but related
water quality subindices with two separate but re-
lated land management subindices.
Each subindex is assigned a weight of 25 per-
cent of the NPI score. As with other similar tools
(U.S. Environ. Prot. Agency, 1990b), considerable
subjectivity occurs in the selection and assignment
of weights to subindices and input parameters for
the NPI. We recommend that users establish ap-
propriate weights based on local priorities or condi-
tions; these weights should be set at the beginning of
a watershed project and not changed thereafter.
Equations must be adjusted accordingly if user-
assigned weights are applied.
The overall NPI score is calculated as follows:
NPI = (0.25U) + (0.25W) + (0.25E) + (0.25C) (1)
Where:
U = Beneficial Use Support Status Subindex
W= Water Quality Data Subindex
E - Extent of Critical Area Treatment Subindex
C = Pollution Control Expected from Treatment
Applied Subindex.
NPI scores can be calculated annually, monthly,
daily, or at any other interval supported by the data.
Data used in the NPI must be summarized for the
same time interval for which the NPI is calculated.
For example, to calculate annual NPI scores, month-
ly water quality data would have to be reduced to an-
nual values.
78
-------
S.A. DRESSING, J.C CLAUSEM, & J. SPOONER
Since the NPI indicates the state of a watershed
with respect to specific water quality problems or
threats, users may wish to establish more than one
NPI score for any given watershed. For example,
users may wish to establish NPI indices for
eutrophication (NPlEutrophication) and bacterial con-
tamination (NPlBacterial) in the same watershed to
track progress in preventing both problems. An over-
all NPI score for the watershed could then be calcu-
lated as the arithmetic mean of the eutrophication
and bacterial scores.
Water Quality Subindices
Two subindices are provided for rating water quality
status: (1) an assessment of beneficial use support
status, and (2) an analysis of water quality data.
While not directly related in many cases because of
the general lack of water quality criteria for nonpoint
source pollutants and impacts, the two water quality
subindices are, in theory at least, correlated.
• Beneficial Use Support Status (U). Beneficial
use support status is generally reported in four
categories (U.S. Environ. Prot. Agency, 1991a):
1. fully supporting uses,
2. fully supporting uses but threatened,
3. partially supporting uses, and
4. not supporting uses.
The U.S. Environmental Protection Agency
(EPA) (U.S. Environ. Prot. Agency, 1991a) provides
guidance to States on making use support decisions
based on biological data, evaluative (non-water
quality) data, chemical data, and other indicators
such as fishing restrictions. EPA also provides
guidelines for determining use support status for
multiple-use waterbodies. This paper generally incor-
porates EPA's procedures for use support determina-
tions, with the exception of multiple-use water-
bodies, in which a different approach is taken for the
NPI.
The scoring for use support status (U) ranges
from 0 to 100 and is based on the four categories
developed by EPA. It is assumed that appropriate
criteria and analyses are performed to assess use
support status.
NPI users first select the beneficial uses and
waterbodies within the watershed that are to be
tracked. It is recommended that, users focus on
those uses intended to benefit from the watershed
project. Next, a subscore is determined for each
designated use on each receiving waterbody being
tracked. A subscore for each waterbody is then cal-
culated as the arithmetic mean of the beneficial use
subscores for the waterbody. The overall score for U
is calculated as the arithmetic mean of the subscores
for the waterbodies.
For example, a project may plan to control bac-
terial contamination to restore or protect the follow-
ing uses and water bodies:
• drinking water and swimming uses on the
rivers,
• shellfishing in the estuary, and
• well water supplies.
The U score for this project is derived by cal-
culating the subscore for the river as the arithmetic
mean of support status for the two uses, i.e., URiver =
[Uorink + Uswim3/2, and then calculating U as the
arithmetic mean of support status for the three
waterbodies, i.e., U = [URiver + UEstuary + Ucroundl/3.
For waterbodies with only one designated use,
full use support is assigned a score of 80, full use
support but threatened is assigned a score of 70, par-
tial use support is assigned a score of 50, and no sup-
port is assigned a score of 20 (Table 1). Waters of
pristine conditions are assigned a score of 100. In the
above example, we might assume that swimming
was threatened and drinking water fully supported in
the rivers, shellfishing was not supported, and drink-
ing from wells was threatened. The beneficial use
support status score (U) would then be 55, i.e., U =
[(70+80)72 + 20 + 70J/3. The generalized equation
for use support status is:
Uj
(2)
The score for use support status on waterbody
"j" (Uj) is calculated as the arithmetic mean of the
scores for the beneficial uses on the waterbody. The
overall score for U is calculated as the mean of the
scores for the n water bodies.
• Water Quality Data (W). An assumption made in
the NPI is that all properly collected water quality
data are equal. This assumption is loosely based on
the principles of meta-analysis (Mann, 1990). In
meta-analysis, similar information is combined for
Table 1.—Use support scores for single-use waters.
USE SUPPORT STATUS
SCORE
Pristine waters .• 100
Full support 80
Full support but threatened 70
Partial support so
Not supported 20
Not supported and no chance for support p
Note: Users may wish to assign intermediate scores based on
intermediate conditions.
79
-------
dings of tiattonal RCWPSymposium, 1992
statistical analyses using techniques such as the
Mantzel-HaenszeJ-Peto method. With respect to the
NPI, biological, chemical, and physical data are con-
sidered to provide equivalent information regarding
water quality status in any given waterbody.
It is recommended that NPI users select the key
parameters most related to beneficial uses and most
responsive to pollution control implementation. For
example, in the previous project (see Beneficial Use
Support Status), fecal coliform could have been used
as the parameter for calculating the score of the
water quality data (W) subindex. If streams in the
watershed were affected by sediment, then total
suspended solids, total sediment, or a benthic macro-
invertebrate index would also be appropriate.
For all water quality data, the NPI scores are
determined by comparing observed values versus
benchmark values. The benchmark values are water
quality criteria values for many parameters, but for a
number of biological and habitat parameters, the
benchmark will be the maximum potential attain-
ment level (maximum potential) score. Maximum
potential scores should be determined by calibration
of biological indices through reference site monitor-
ing (U.S. Environ. Prot. Agency, 1991b). If data are
lacking, best professional judgment can be used to
establish these scores. Each monitored parameter
should be scored against the same benchmark value
throughout the course of any given study. For
closed-end indices, the maximum potential score is
the highest value that can be scored (e.g., 60 for the
IBI; Karr, 1981), but for open-end indices, such as
the index of well being (Iwb) (Gammon, 1980), judg-
ment must be made regarding the highest possible
score achievable in the watershed (U.S. Environ.
Prot Agency, 1991b).
When using water quality criteria, W is calcu-
lated as the percentage of samples or observations
that meet the criteria. Since NPI analyses are ex-
ploratory, the minimum number of samples or ob-
servations needed can be based on professional
judgment rather than rigorous statistical analysis.
When using maximum potential scores for biological
or habitat parameters, W is the sample value as a
percentage of the maximum potential score (i.e.,
100*[sample value/maximum potential score]). If
more than one biological or habitat sampling is done,
the arithmetic mean of the sampling scores should
be reported as the score for W. The range of scores
for W is 0 to 100.
Although criteria and maximum potential scores
are not necessarily equivalent, both can be used to
develop W scores because both are used to repre-
sent water quality status with respect to a meaningful
goal for the watershed. The scale (i.e., 0 to 100 per-
cent) for each approach provides similar informa-
tion, yet a 50 percent score against criteria is not like-
ly to be equivalent to a 50 percent score against
maximum potential. This discrepancy results from
several factors, including the fact that criteria are not
typically set to represent the maximum potential but
to establish what is necessary to support designated
uses. In some cases, however, the designated uses
may be established with "maximum potential" in
mind.
Three scenarios generally exist for the use of
benchmark values, and all can be addressed with the
NPI. Most typically, samples with measured values
less than or equal to the criterion are judged as meet-
ing the criterion (e.g., nitrate levels of 10 mg/L or
less). Some criteria are such that measured sample
values within a specified range (e.g., dissolved
oxygen concentrations of 6 to 8 mg/L) are judged to
meet the criterion. Higher scores often indicate bet-
ter conditions for biological and habitat parameters
(e.g., IBI; Karr etal. 1986).
For example, if the water quality criterion for
nitrate is 10 mg/L, the percentage of groundwater
samples or observations not exceeding 10 mg/L
would be reported as the score for that parameter.
NPI users may choose to compare flow-adjusted (or
otherwise adjusted) data values against criteria to ac-
count for some of the natural variability that affects
water quality data. This comparison would, of
course, change the interpretation of scores.
If more than one parameter is tracked per water-
body, W is calculated as the average score for all
parameters for the waterbody. For example, the
monitoring program in a watershed where a lake is
eutrophic and the streams are impacted by sediment
may track total phosphorus (criterion=0.010 mg/L; <
or = meets criterion) in the lake, chlorophyll a (state
criterion=25 mg/L; < or = meets criterion) in the
streams, and the IBI (maximum possible value=60)
in the streams. (Streams are lumped as one water-
body in this example because they are very similar.)
The lake subscore (WLake) is simply the percentage
of samples that are less than or equal to 0.010 mg/L,
and the subscore for the streams is the arithmetic
mean of the chlorophyll a (Wstream-Chla) and IBI (W-
Stream-IBl) subscores. The chlorophyll a score is cal-
culated as the percentage of samples that are less
than or equal to 25 mg/L, and the IBI score is calcu-
lated as 100 times the arithmetic mean of sample
values, divided by 60.
If more than one representative monitoring sta-
tion exists for a waterbody, subscores should be cal-
culated for each monitoring station before calcu-
lating W for the waterbody as the arithmetic mean of
monitoring station subscores. It is best, however, to
80
-------
SA. DRESSING, J.C CLAUSEN, & J. SPOONER
analyze only those Stations that best represent condi-
tions in the waterbody.
Individual station scores for each waterbody are
calculated as the arithmetic mean of parameter
scores for the station. Parameter scores are deter-
mined as either the percentage of parameter values
at the station that meets the criterion or as the arith-
metic mean of the biological or habitat parameter
values expressed as a percentage of the maximum
potential attainment value.
The generalized equation for water quality data
is
(3)
In equation 3, the score of W is the arithmetic
mean of the p waterbody scores, where Wk is the
score for waterbody "k," calculated as the arithmetic
mean of the station scores on the waterbody.
Implementation Subindices
Two implementation subindices are used in the NPI:
(1) a measure of the extent of implementation in the
critical area (E), and (2) a measure of the level of pol-
lution control expected from the practices or
measures implemented (C). The two implementation
subindices are related in theory, but not necessarily
in practice. For example, no predictable relationship
exists between the percentage of critical area treated
and the level of pollution control achieved in the criti-
cal area. However, the overall level of pollution con-
trol obtainable through any given treatment
technology is a function of the extent to which the
technology is applied.
Users are advised to track implementation as it
relates to beneficial uses and pollutants or problems
of concern in the watershed. The intent is to track
implementation information that relates to the infor-
mation tracked for the water quality subindices.
Point sources are factored into each implementa-
tion subindex as necessary. For the purposes of the
NPI, point sources consist of wastewater treatment
plants, direct industrial discharges, permitted animal
operations, and combined sewer overflows. The NPI
does not consider background contributions as pol-
lution sources, and it does not have a capability for
including unknown sources.
• Extent of Critical Area Treatment (E). For non-
point sources, the delineation and treatment of criti-
cal areas is central to the cost-effective control of the
major pollutant sources affecting beneficial uses in a
watershed (Humenik et al. 1987; U.S. Environ. Prot.
Agency, 1990a,b). Targeting procedures (U.S. En-
viron. Prot. Agency, 1990b) and modeling (Daven-
port, 1984) are two tools often used to delineate criti-
cal areas and project the benefits of pollution control
efforts.
The score for the extent of treatment in the criti-
cal area (E) is calculated as the percentage of critical
area under treatment, ranging from 0 to 100. For a
nonpoint source project to be successful, most im-
plementation should occur within the critical area. A
key assumption is that the critical area has been ade-
quately identified based on critical area delineation
criteria (e.g., percentage of pollutant loads from the
critical area) that are linked to specific beneficial
uses, pollutants, and problems of concern in the
watershed. An essential step in establishing critical
areas is to assess all potential pollutant sources or
problems of concern. NPI users should consider for
inclusion within the critical area any well-managed
land uses or activities that would be a source of pol-
lutants or problems if they were not well-managed.
Treatment is defined for the NPI as the applica-
tion of a management or structural measure to con-
trol specific nonpoint source pollutants or problems
of concern identified in the watershed. For example,
stormwater detention basins can be used to treat a
watershed in which sediment is the pollutant of con-
cern. In cases where phosphorus loading is a prob-
lem, .both nutrient management and sediment
control would typically be needed to achieve treat-
ment on cropland. Clearly, judgments must be made
regarding what constitutes treatment versus no
treatment. Generally, if no effort has been made to
control the pollutants or impacts causing the
problems, then treatment has not occurred. The
level of pollution control provided through treatment
is reflected in subscore C.
In many cases, point sources will also contribute
to water quality problems in the watershed. Because
they generally discharge directly to receiving
waters, significant point sources should be con-
sidered when delineating critical areas. Point
sources are generally considered treated when they
have at least primary treatment. Livestock opera-
tions with no-discharge permits are considered
treated.
The percentage of critical units under treatment
can be calculated in several ways, but in all cases
NPI users must be careful to count only treatment
that addresses the pollutants or problems of concern
in the watershed. For cropland watersheds, the per-
centage could be calculated based on acreage
treated. For animal operations, the percentage could
be determined as the percentage of animal units
under treatment. For urban areas, the percentage
treated could be calculated as the percentage of the
81
-------
Proceedings of National RCWP Symposium, 1992
drainage area served by stormwater treatment
devices.
If multiple sources are present, a weighted
average of the percentage treated for each source
should be calculated with the weighting factor based
on the proportion of the potential pollution
generated by or potential problems caused by each
source in the critical area. For example: if 30 percent
of irrigated cropland in the critical area is treated (E-
Irrigate -30) and 20 percent of the problem (e.g.,
suspended solids) from the critical area is potentially
caused by irrigated land; livestock operations have
the potential to cause 15 percent of the problem from
the critical area and 10 percent is treated (Eovestock
-10); point sources potentially contribute 35 percent
of the problem from the critical area and 100 percent
are treated (Epoint =100), and nonirrigated cropland
has the potential to cause 30 percent of the problem
from the critical area and 55 percent is treated (E-
Nonkrigated -55); then, the overall Esuspended solids for
the watershed would be 59 (i.e., E =
[30*0.20M10*0.15]+[100*0.35]+[55*0.30]).
• Pollution Control Expected From Treatment
Applied (C). The level of pollution control expected
from treatment applied (C) in the critical area is
another key factor in achieving an adequate level of
nonpoint source control. Tables 2 and 3 provide ex-
amples of nonpoint source controls and their effec-
tiveness. Table 4 provides examples of the range of
technology used to treat point sources. Effectiveness
ranges from 0 to 100 percent and is measured with
respect to on-site, or source, control. Effectiveness
estimates do not incorporate off-site delivery. Users
will need to assign effectiveness values based on
local data and best professional judgment. For many
control practices, users will have to establish a maxi-
mum storm size (e.g., 24-hour, 25-year storm) for
which effectiveness values apply.
To be meaningful, the C score must reflect the
level of control of the pollutants or problems of con-
cern in the watershed. For example, a technology
that removes 80 percent of suspended solids and 15
percent of total nitrogen would receive a C score of
15 in a watershed in which nitrogen was the major
concern, but it would receive a score of 80 if
suspended solids were the pollutant of concern.
Most land treatment takes the form of systems
of practices rather than individual practices. Overall
system effectiveness, therefore, is the most ap-
propriate parameter to apply in calculating C scores.
For example, a conservation tillage system may in-
clude conservation tillage, grassed waterways, and
filter strips. The combined effectiveness of these
practices (i.e., the system) in keeping pollutants on
Table 2.—Examples of nonpoint source treatment ef-
fectiveness: agriculture.
TREATMENT
Animal waste systems1
Filter strips1 (confined
livestock application)
Conservation tillage1 (for
manure applied to
cropland)
Nutrient management2 (for
manure applied to
cropland)
Filter strips1 (for manure
applied to cropland)
Terrace systems1 (for
manure applied to
cropland)
Buried pipe systems2
EFFECTIVENESS* (POLLUTANT)
0 to 37
11 to 82
24 to 78
31 to 83
37 to 100
Oto35
Oto 15
5 to 52
8 to 29
11 to 71
-40 to 51
-230 to 29
-144 to 89
36 to 82
39 to 97
Oto 73
10 to 41
5 to 73
Oto 55
-21 to 35
Oto 29
15 to 57
Oto 39
7 to 45
-14 to 61
-52 to 24
18 to 75
10 to 61
29 to 64
75 to 95
(Dissolved P)
(Nitrate)
(Total P)
(Total N)
(Fecal coliform)
(Dissolved P)
(Nitrate)
(Total P)
(Total N)
(Sediment)
(Dissolved P)
(Nitrate)
(Total P)
(Total N)
(Sediment)
(Dissolved P)
(Nitrate)
(Total P)
(Total N)
(Dissolved P)
(Nitrate)
(Total P)
(Total N)
(Sediment)
(Dissolved P)
(Nitrate)
(Total P)
(Total N)
(Total sediment)
(Sediment)
(irrigation)
Sediment basins2
(irrigation)
75 to 95 (Sediment)
1RobilIard, 1992
2U.S. Dep.Agric. 1991
*Percent reduction In load vs. control or no treatment
Table 3.—Examples of nonpoint source treatment ef-
fectiveness: urban.
TREATMENT
EFFECTIVENESS* (POLLUTANT)
Wet ponds1
Seeding to establish vegeta-
tion on disturbed
construction site2
Sediment basin for construc-
tion sites2
12 to 91 (Total P)
6 to 85 (Total N)
8 to 95 (Total Pb)
13 to 96 (Total Zn)
50 to 100 (TSS)
55 to 100 (TSS)
Conventional septic system2
Recirculating sand filter2 (on-
slte disposal system)
54 to 83
30 to 60
Oto 58
Oto 95
70 to 98
75 to 98
1 to 94
70 to 90
(TSS)
(BOD)
(Total N)
(Total P)
(TSS)
(BOD)
(Total N)
(Total P)
1Schueleretal. 1991
2U.S. Environ. Prat. Agency, 1992
*Percent reduction in load vs. control or no treatment
82
-------
5,A. DRESSING, J.C. CLAUSEM, & J, SPOOLER
Table 4.—Examples of point source treatment effec-
tiveness.
TREATMENT
No discharge
Breakpoint chlorination
Reverse osmosis
Denitrification
Land application-irrigation
Ammonia stripping
Oxidation ponds
Filtration
Secondary treatment
Primary treatment
Primary sedimentation tanks
Primary sedimentation tanks
with chemical precipitation
Alum addition
Activated sludge process
No treatment
EFFECTIVENESS* (POLLUTANT)
100
80 to 95
80 to 90
70 to 95
60 to 90
50 to 90
20 to 90
20 to 40
10 to 30
5 to 10
50 to 70
25 to 40
25 to 75
80 to 90
70 to 80
80 to 90
Up to 95
85 to 90
0
(N)
(N)
(N)
(N)
(N)
(N)
(N)
(N)
(N)
(N)
(TSS)
(BOD5)
(Bacteria)
(TSS)
(BOD5)
(Bacteria)
(P)
(BOD)
*Percent removal
Source: Metcalf and Eddy, 1979
site should be used to calculate C for the land area
treated by the system.
The C score for a farm or other subarea that has
multiple land uses (e.g., cropland and confined live-
stock) would be calculated as the arithmetic mean of
the weighted land use subscores. Land use sub-
scores would be weighted on the basis of their rela-
tive potential contribution of the pollutant or problem
of concern. Clearly, the relative potential contribu-
tion of any nonpoint source must be determined on
the basis of best professional judgment. We recom-
mend that conventional tillage with excessive
nutrient applications be considered the worst-case
scenario for assessing potential contributions of
nutrients and sediment from cropland. For livestock,
one might assume for the worst-case that all animal
waste would leave the site.
For example, a dairy farm that includes both
cropland and livestock may be treated to keep 80
percent of cropland-generated phosphorus on site
and 90 percent of livestock-generated phosphorus on
site. If cropland has the potential (i.e., conservation
tillage, excessive nutrient applications) to contribute
85 percent of the total phosphorus from the dairy,
and livestock facilities have the potential (i.e., all
animal waste leaves site) to contribute 15 percent of
the total phosphorus from the dairy, the C score for
the dairy would be 81.5 (C = [80*0.85]+[90*.15]). In
summary, 81.5 percent of the phosphorus is kept
from leaving the dairy.
Watersheds with multiple sources in the critical
area can be addressed by calculating a C score for
each source category (e.g., dairies, suburban
residential, and forestry) and then calculating a
weighted average score for the critical area. For ex-
ample, the above dairy could be one of 20 in the criti-
cal area. The C score for dairies (Coairies) would be
the arithmetic mean of the weighted scores (based
on relative, potential contribution of pollutant or
problem of concern) for the 20 dairies. The overall C
score for the critical area would be calculated as the
arithmetic mean of the weighted C scores for all
source categories in the critical area. The weighting
factor for each source category would be based on
the proportion of the critical area pollutants or
problems potentially contributed by that source.
As an example, assume that the above dairy is in
a watershed that has four sources with the following
potential contributions of phosphorus: (1) dairies in
the critical area potentially contribute 30 percent of
the phosphorus from the critical area, (2) a waste-
water treatment plant potentially contributes 55 per-
cent of the phosphorus, (3) cropland may contribute
10 percent of the phosphorus, and (4) urban runoff
may contribute 5 percent. The 20 dairies are treated
to remove an overall average of 85 percent of the
phosphorus (CDairy=85), the wastewater treatment
plant has secondary treatment (Cwwrp=90), the
cropland is generally under conservation tillage and
nutrient management (Ccrop=65), and the urban
runoff is not treated (Curban=0). Then, the overall C
score is
C = (.30)(85) + (.55)(90)
(.05)(0) = 81.5.
If more than one pollutant or problem must be
tracked the C score is calculated as the arithmetic
mean of the pollutant scores. For example, if total
suspended solids were tracked (C score - 50.5) in
addition to phosphorus (C score = 81.5), the overall
C score would be 66 ([50.5+81.5] /2).
Testing
We tested the NPI using data from the Rock Creek
Rural Clean Water Program (RCWP) project in
Idaho and using data from the St. Albans Bay RCWP
project in Vermont.
Rock Creek RCWP, Idaho
The Rock Creek, Idaho, RCWP project, located in
south central Idaho, covered 18,321 hectares (45,000
acres) in an 80,352-hectare (198,400-acre) watershed
(Fig. 1). About 350 farms in the project area produce
dry beans, dry peas, sugar beets, corn, small grain,
alfalfa, and livestock. Annual rainfall is low (21.6
cm/year) and irrigation is required for crop produc-
83
-------
Proceedings of National RCWP Symposium, 1992
SCALE
LEGEND
A monitoring stations
• subbasin monitoring stations
JUH1 streambank erosion study sites
I I town
project boundary
NOTE: Diversions from the High and
Low Line canals are controlled.
High and Low Line canals bypass
Rock Creek.
Figure 1.—Project map for the Rock CreeK, Idaho, RCWP project
tion. Irrigation water is diverted from the Snake
River into main canals and delivered to the farms
through a network of canals, laterals, and drains
(ditches). Irrigation return flows eventually empty
into Rock Creek, which discharges in the Snake
River.
Rock Creek has poor water quality that impairs
the beneficial uses of contact recreation and fishing.
In addition, Rock Creek delivers a disproportionate
load of sediment to the Snake River. Major sources
of nonpoint source pollution in the area are sediment
and associated pollutants (i.e., phosphorus and
nitrogen) from irrigation return flows. Streambank
erosion in the upper reaches of Rock Creek is also a
major problem.
The nonpoint source management strategy is to
reduce the amount of sediment, sediment-related
pollutants, and animal waste discharging into Rock
Creek from agricultural lands. All
irrigated cropland and animal
pro-duction facilities are included
in the critical area, which com-
prises 11,470 hectares (28,159
acres). The primary implementa-
tion objectives are to control field
erosion and prevent sediment
from leaving fields. BMPs in-
clude conservation tillage, sedi-
ment retention structures, irriga-
tion water management, vegeta-
tive filter strips, cover crops, and
animal waste management.
Water quality monitoring ob-
jectives include (1) documenting
beneficial use improvements by
monitoring in-stream habitats,
benthic macroinvertebrate, and
fish populations; and (2) record-
ing changes in concentrations of
sediment and nutrients attrib-
utable to BMP implementation.
Monitoring stations are located
on Rock Creek and at down-
stream and upstream pairs in the
subwatersheds, with the down-
stream stations representing out-
lets from the subwatersheds to
Rock Creek.
Grab samples for chemical
and physical parameters are
taken biweekly during the irriga-
tion season at the Rock Creek
stations; the subwatersheds are
sampled biweekly at the begin-
ning and end of the irrigation
season and weekly during the middle of the season.
Monthly grab samples are taken in Rock Creek
during the nonirrigation season. Fish populations
have been measured in 1981, 1985, 1987, 1988, and
1989. Macroinvertebrate surveys are performed
quarterly. Habitat Evaluation Procedures analyses
were completed in 1981,1984, and 1988.
The suspended sediment concentrations from
the downstream Rock Creek monitoring station (S-2)
were used for the NPI case study (Fig. 1). This sta-
tion should reflect the cumulative nonpoint source
impacts from the project area.
• Beneficial Use Support Status (U). The U
scores were calculated based on Water Quality Index
(WQI) values. WQI values were calculated by the
"WQI" program in EPA's STORET database, which
compares the measured chemical and physical
84
-------
SA DRESSING, J.C. CLACISEN, & J. SPOONER
values with recommended "fishable and swimmable"
criteria. Hie project's 10-year report contained WQI
values and associated use support ratings (U.S. Dep.
Agric. 1991). In contrast to the U scores, a low WQI
indicates good quality waters. The U scores were cal-
culated based on an interpolation between the WQI
values and use support ratings and the status and
scores found in Table 1.
• Water Quality Data (W). The W scores reflect
the percent of samples that met the proposed stand-
ard of 100 mg/L suspended sediment in Rock Creek
(U.S. Dep. Agric. 1991).
• Extent of Critical Area Treatment (E). The ex-
tent of critical area treatment was obtained from the
10-year report and from annual project reports for
1982 to 1990.
• Pollution Control Expected From Treatment
Applied (C). The level of sediment control expected
from treatment applied was estimated with sediment
removal efficiencies determined by the project (U.S.
Dep. Agric. 1991). The project reported a 70 percent
decrease in the sediment load delivered to Rock
Creek from the project area.
' NPI. Scores for each subindex and the NPI for
1982 through 1990 for the downstream Rock Creek
station are given in Table 5 and Figure 2. Although
the project has documented improvements in the
biology and habitat of Rock Creek, the improvement
is far short of results predicted from suspended sedi-
ment measurements. Future analysis of the NPI
using available biological and habitat data should
reflect this discrepancy.
Table 5.—NPI and subindex annual scores for the
downstream Rock Creek water quality monitoring
stations (S-2), Rock Creek, Idaho, RCWP project.
BENEFI- WATER
NPI CIALUSE QUALITY
CRITICAL POLLU-
AREA TION
TREAT- CONTROL
YEAR
1982
1983
1984
1985
1986
1987
1988
1989
1990
SCORE
14.9
33.4
32.2
36.9
45.5
47.1
64.0
58.9
64.1
(U)
30
39
35
43
47
45
53
50
54
(W)
7
53
27
44
59
59
87
79
92
WENT (E)
10
22
30
32
35
42
46
60
75
(C)
12
19
37
28
41
42
70
46
35
1 UU
90
80
70
60
50
40
30
20
10
0
P
tt
»i
J<
J<
s
s
F ^
Xs
Xs
Xs
rfh^
ffl &
4 1 1 X"1
jq
X s
X s
Xs
1 982 1 983
—
—
_ s
(* ;
f X S >
I-1 —
f x s
(*
i*1 X s
^
f Xs
^
(* X S
f
sy
s/
\/
s
s/
S '
s /
S''
\s
s,/
s^
S-"
\^
sy
S''
S''
S'1
" S
xs
xs
xs
x s
xs
X S
xs
xs
xs
xs
— XS
xs
xs
xs
— x s
xs
xs
xs
xs
X S
xs
xs
R
.
« V
X V,
x V
< V
"V
•l
«*
—
x s
X ^
^
i
f
'<
/ x(
/<
/ K
)J
<
X
\
— X S
< s
R
* J
X x ^
x j\
< s
X
X
x
x
x
x
x
X -a
>< s
X s
X s
X s
X s
1 9B4 1 985 1 9B6 1 987 1 9BB 1 989 1 990
NPI
Figure 2.—NPI scores for Rock Creek, Idaho (sediment).
85
-------
Proceedings of National RCWP Symposium, 1992
St. Albans Bay RCWP, Vermont
A complete description of the St. Albans Bay RCWP
is given In Meals (1992). The impaired use was con-
tact recreation in St Albans Bay, which could not be
used for swimming because of elevated fecal
coliform counts.
• Beneficial Use Support Status (U). Before im-
plementing watershed BMPs, St. Albans Bay partial-
ly supported use for contact recreation (URec=50).
Following BMP implementation and upgrading of
the wastewater treatment plant, use for recreation
has been fully supported (URec=80).
• Water Quality Data (W). For this paper, only the
bacteria data from the Vermont RCWP will be dis-
cussed. Table 6 summarizes the frequency in which
the fecal coliform bacteria criterion of 200/100 mL
was exceeded in each stream tributary to St. Albans
Bay during pre- and post-BMP time periods. Overall,
the percentage of bacteria violations dropped from
54 percent during the pre-BMP period (W=46) to 42
percent (W-58) during the post-BMP period. The
study showed an average 58 percent decline in fecal
coliform bacteria concentrations across the water-
shed.
Table 6.—Percentage of samples exceeding the fecal
coliform criterion of 200/100 mL for streams tributary
to St. Albans Bay, Vermont.
PERIOD
Pre-BMP
Post-BMP
JEWETT
BROOK
62
61
STEVENS
BROOK
74
49
RUGGS
BROOK
49
34
MILL
RIVER
32
. 22
• Extent of Critical Area Treatment (E). Critical
areas were defined as those lands receiving animal
wastes. Initially, no critical areas were being treated
(E-0). After the fifth year of the 10-year project and
thereafter, 74 percent of the critical areas were
treated (E-74). The project's goal was to treat 75 per-
cent of the critical areas (79 percent of the animal
units in the watershed were being managed by
BMPs).
• Pollution Control Expected From Treatment
Applied (C). Animal waste management systems
are expected to reduce fecal coliform loads by at
least 37 percent (C-37).
• NPI. Below are the NPI calculations for bacteria
during the pre-BMP condition versus the post-BMP
condition.
NPIp,a=(0.25X50)+(0.25X46)+(0.25XO)+(0.25XO)=24 (4)
NPIpo.i=(0.25X80)+(0.25X58)+(0.25X74)+(0.25X37)=62 (5)
This analysis indicates that the NPI for bacteria
increased as a result of the project (Fig. 3). The NPI
scores for phosphorus are given below. The
threshold used for phosphorus was 0.05 mg/L
NPIpre=(0.25X20)+(0.25X30)+(0.25XO)+(0.25XO)=12.5 (6)
NPIp0st=(0.25X20)+(0.25X25)+(0.25X74)+(0.25X24)=35.6(7)
Although the NPI indicates a higher phosphorus
index following BMP implementation, no improve-
ment in phosphorus concentrations occurred in St
Albans Bay. However, as indicated in Clausen et al.
(1992), the level of implementation should result in
water quality improvements after a new field phos-
phorus equilibrium has been reached.
Conclusions
Results from two case studies indicate that the NPI
can be a useful tool to help identify water quality
status in projects before and after BMP implementa-
tion. In both cases, the NPI provides a convenient
summary of land treatment and water quality
parameters that can be easily interpreted by the
scientific community, managers, and the general
public.
The test results also show that the subindex
scores can be used to explain trends in the overall
NPI scores. For example, the Idaho analysis indi-
cates that suspended sediment data are inadequate
as the sole predictor of use support status in the
watershed. The Vermont analysis illustrates the vast-
ly different response times for bacteria and phos-
phorus.
The NPI does not reduce data collection needs,
but it may help drive project participants toward
more thorough data collection and a greater under-
standing of the relationships among identified
problems, pollutants of concern, and implementation
plans. •
References
Carlson, RE. 1977. A trophic state index for lakes. Limnol.
Oceanog. 22(2):361-69.
Claiisen, J.C., D.W. Meals, and EA Cassell. 1992. Estimation of
lag time for water quality responses to BMPs. In Proc. Natl.
RCWP Symp., 1992. Orlando, FL.
Davenport, T.E. 1984. Field modelling in the Highland Silver Lake
watershed: interim report IEPA/WPC/84-026. Div. Water
Pollut. Control, HI. Environ. Prot, Agency, Springfield.
Gammon, J.R. 1980. The use of community parameters derived
from electrofishing catches of river fish as indicators of en-
vironmental quality. In Seminar on Water Quality Manage-
ment Tradeoffs. EPA-905/9-80-009. U.S. Environ. Prot.
Agency, Washington, DC.
86
-------
S.A. DRESSING, J.C. CLAUSEN, & J. SPOONER
UJ
ac.
o
o
100
90
80
70
60
50
40.
30
20
10
I
I
1
I
Pre-BMP-
• Post-BMP
PROJECT PHASE
Figure 3.—NPI scores for St. Albans Bay, Vermont (bacteria).
Hopkins, R. B. and J. C. Clausen. 1985. Land use monitoring and
assessment for nonpoint source pollution control. Pages 25-
29 in Perspective on Nonpoint Source Pollution. EPA 440/5-
85-001. Off. Water, U.S. Environ. Prot. Agency, Washington,
DC.
Humenik, FJ., M.D. Sraolen, and S.A. Dressing. 1987. Pollution
from nonpoint sources: where we are and where we should
go from here. Environ. Sci. Tech., 21 (8) :73742.
Karr, J.R. 1981. Assessment of biotic integrity using fish com-
munities. Fisheries 6:21-27.
Karr, J.R. et al. 1986. Assessing Biological Integrity in Running
Waters: A Method and Its Rationale. Spec. Publ. 5. HI. Natl.
History Surv., Illinois.
Mann, C. 1990. Meta-analysis in the breech. Science 249:476-80.
Meals, D.W. 1991. Surface water trends and land-use treatment,
Pages 13642 in Seminar Publication: Nonpoint Source
Watershed Workshop. EPA/625/4-91/027; Off. Res. Dev.,
Off. Water, U.S. Environ. Prot. Agency, Washington, DC.
. 1992. Water quality trends in the St Albans Bay, Vermont,
watershed following RCWP land treatment. In Proc. Natl.
RCWP Symp., 1992. Orlando, FL.
Robillard, ED. 1992. Draft materials prep. EPA Grant #X-818240.
Penn. State Univ., University Park, PA
Schueler, T.R., P.A. Kimble, and M.A Heraty. 1991. A current as-
sessment of urban best management practices — techniques
for reducing nonpoint source pollution in the coastal zone.
Rev. Draft. Metro. Washing. Counc. Gov., Washington, DC.
Spooner, J., D.A. Dickey, and J.W. Gilliam. 1990. Determining and
increasing the statistical sensitivity of nonpoint source con-
trol grab sample monitoring programs. Pages 119-35 in Proc.
Internal. Symp. on the Design of Water Quality Information
Systems. Info. Sen No. 61. Colo. Water Resour. Res. Inst,
Colo. State Univ., Fort Collins.
U.S. Department of Agriculture. 1991. Rock Creek Rural Clean
Water Program 10-year Report. Agric. Stabil. Conserv. Serv.,
Twin Falls, ID.
U.S. Environmental Protection Agency. 1989. Rapid bioassessment
. protocols for use in streams and rivers: benthic macroinver-
tebrates and fish. EPA/444/4-89-001. Off. Water,
Washington, DC.
. 1990a. Rural Clean Water Program. EPA 440/4-9(W)12.
Nonpoint Source Control Branch, Oft Water, Washington,
DC.
. 1990b. Urban Targeting and BMP Selection. Water Div.,
Region V, Chicago, IL.
. 1991a. Guidelines for the Preparation of the 1992 State
Water Quality Assessments (305[b] Reports). Off. Wetlands,
Oceans, and Watersheds, Off. Water, Washington, DC.
'-. 1991b. Watershed Monitoring and Reporting for Section
319 National Monitoring Program Projects, Off. Wetlands,
Oceans, and Watersheds, Off. Water, Washington, DC.
. 1992. Draft materials for management measures guidance
for nonpoint source control in coastal zone. Off. Wetlands,
Oceans, and Watersheds, Off. Water, Washington, DC.
87
-------
-------
The Effects of Temporal and Spatial
Variability on Monitoring Agricultural
Nonpoint Source Pollution
Thomas H. Johengen and Alfred M. Beeton
National Oceanic and Atmospheric Administration
Great Lakes Environmental Research Laboratory
Ann Arbor, Michigan
ABSTRACT
The Saline Valley Rural Clean Water Program project was one of 21 projects developed to evaluate
methods of controlling agricultural nonpoint source pollution. Control programs were designed
around voluntary implementation of best management practices) and water quality trends were
monitored at eight stream stations from July 1981 to December 1989, using a fixed, weekly sam-
pling design. An additional monitoring program was established within the Macon Creek subbasin
(Station 9) in June 1988 to quantify temporal and spatial variability in pollutant loads. Macon Creek
was monitored daily for seven days following any storm of more than half an inch of rain. Five sta-
tions were added upstream from existing Station 9 to examine spatial variation in loading rates
hroughout the subbasin. This study describes these storm monitoring results and discusses their
implication to the Saline project's monitoring data.
Saline Valley was one of 21 projects within the
U.S. Department of Agriculture's Rural
Clean Water Program (RCWP) designed to
evaluate methods for controlling agricultural non-
point source pollution. For meaningful results from a
water quality management perspective, it is neces-
sary to quantify the effectiveness of best manage-
ment practices (BMPs) in reducing pollutant
loadings associated with agricultural production.
Most projects found it difficult to establish these
relationships for treatment applied at the watershed
level.
Measuring effectiveness was complicated be-
cause a wide variety of BMPs were adopted gradual-
ly over large areas of land and because the RCWP
relied on voluntary participation, which greatly
limited control over the timing, amounts, and place-
ment of these practices. The lack of treatment con-
trol, the spatial scales involved, and the inherent
variability of meteorological processes posed sig-
nificant problems for the water quality monitoring
programs.
Limited funding and human resources severely
restricted the Saline Valley project's ability to ad-
dress patterns of temporal and spatial variability. The
two most significant problems were the lack of con-
tinuous discharge records and the inability to
monitor individual storm events in detail. The pur-
pose of this project was to quantify patterns of tem-
poral and spatial variability within the watershed by
monitoring individual storms daily and increasing
the number of stations along the stream. These data
were then used to help evaluate and interpret the
monitoring data. Data presented in this discussion
89
-------
Proceedings of national RCWP Symposium, 1992
predominately focus on results from Subbasin 9,
where all of the additional monitoring took place
(Fig. 1.).
Materials and Methods
Temporal Variability
Temporal variability in pollutant loading was as-
sessed on annual and daily time scales. Loading data
collected under the project's weekly sampling
scheme were tabulated and analyzed graphically
through cumulative loading distributions to examine
annual variability. Daily variability was assessed by
monitoring individual storm events; however, only
rainstorms producing greater than 0.5 inches a day
were sampled. Following a storm, Macon Creek
transect stations were monitored daily for seven
days or until discharge returned to baseline condi-
tions. A total of eight storms were monitored be-
tween 1988 and 1989.
Storm monitoring results were used to quantify
potential errors in the project's loading estimates.
Errors in the project's weekly loading estimates
were calculated from the difference between sum-
ming the seven consecutive daily measurements ver-
sus extrapolating a single sampling event over a
week. Errors were calculated for each of the seven
days following a storm to represent all possible inter-
vals between the storm and sampling day of a fixed
schedule.
Results from storm monitoring were also used to
evaluate errors in annual loading estimates. A daily
precipitation record, obtained from the Saline waste-
water treatment plant, was used in conjunction with
average loading values observed for storms to es-
timate annual loads. These computed loading values
were then compared against the project's loading es-
timates.
012 3 4
Scale (km)
MILAN
SALINE VALLEY
RURAL CLEAN WATER PROJECT
Washtenaw and Monroe Counties, Michigan
PROJECT & SUB-BASIN BOUNDARIES
O PROJECT SAMPLING STATIONS
A TRANSECT SAMPLING STATIONS
Figure 1.—Saline Valley Rural Clean Water Project study area located In Washtenaw and Monroe counties, Michigan.
Project stations were located at (3) Saline-Bridgewater Drain, (4) Bauer Drain, (6) Bear Creek, (7) Wanty Drain, and (9)
Macon Creek. Stations 3A, 5, and 8 were located on the Saline River. Transect Stations 9-1 to 9-5 were added along
Macon Creek to examine patterns in spatial variability.
90
-------
T.H. JOHENOEN&A.M. BEETON
Spatial Variability
Spatial variability was examined over both watershed
and subbasin scales. Variability within the project's
watershed was evaluated by comparing patterns in
concentration, loading, and export rates among the
stations. Variability at the subbasin level was ex-
amined by creating a longitudinal transect along
Macon Creek within Subbasin 9 (Fig. 1). Five addi-
tional stations were added upstream of existing Sta-
tion 9, ranging from 0.8 to 2.9 kilometers apart.
Transect stations were established in September
1988 and sampled during routine project sampling
and storm events until the project's termination in
December 1989.
TOTAL PHOSPHORUS
CUMULATIVE LOADING DISTRIBUTIONS
90-
80-
70-
60-
50
40
30
20-
10
Results
Temporal Variability in Pollutant
Loading
The project's annual loading data for Station 9
revealed extreme variability in both magnitude and
timing. Annual loads for suspended solids, total
phosphorus (total-P), soluble phosphorus (soluble-
P), and nitrate (NOs) varied by 14-, 5-, 7-, and 3-fold,
respectively, over the course of the study (Table 1).
Cumulative loading distributions for total phos-
phorus revealed that loading rates were highly vari-
able within a given year as well as among years (Fig.
2). If loading rates were uniform throughout the
year, distributions would plot as straight lines. Dis-
tributions were highly nonlinear and indicated that
annual loads were dominated by a storm events.
Summing the three highest weekly loading es-
timates from the project's data indicated that, on
average, 76, 56, 51, and 50 percent of annual loads
for suspended solids, total-P, soluble-P, and nitrate,
respectively, occurred in only 8 percent of the time
within 28 days of the year.
Table 1.—Annual loads for suspended solids (SS),
total phosphorus (TP), soluble phosphorus (Sol-P),
and nitrate (NOs) at Station 9 for 1983 through 1989.
ANNUAL LOAD (mtons)
YEAR
1983
1984
1985
1986
1987
1988
1989
SS
248
101
79
256
43
604
342
TP
0.49
0.34
0.27
0.85
0.20
1.13
0.71
SOL-P
0.15
0.15
0.09
0.26
0.07
0.49
0.29
N03
15.8
13.6
18.6
19.8
15.3
44.4
34.1
Figure 2.—Total phosphorus cumulative loading dis-
tributions at Station 9 for years 1984 through 1989.
the first 48 hours of a storm and returned to baseline
conditions within three days (Table 2). This
response was fairly consistent for all storms
monitored, which ranged from 0.5 to 2.5 inches per
day of precipitation. The percentage of the week's
load, which occurred on Day One after a storm, was
always greater for particulate parameters than for
dissolved nutrients or discharge levels. On average,
40,85, 71, 65, and 55 percent of the week's total load
for discharge, suspended solids, total-P, soluble-P,
and nitrate, respectively, occurred on this first day
(Table 2). This pattern reflects a "first flush" effect,
where material that has accumulated on the ground
is washed off by overland runoff during the first part
of a storm. Nitrate showed the smallest loading spike
because its concentrations tended to peak one or two
days after the storm when overland runoff ceased
and the majority of input to the stream was derived
from seepage through the soil.
Table 2.—Percent of weekly discharge (DIS),
suspended solids (SS), total phosphorus (TP),
soluble phosphorus (Sol-P), and nitrate (NOs) loads
occurring over seven days following a storm.
#DAYS
AFTER ~
STORM
1
2
3
4
5
6
7
PERCENT OF WEEKLY LOAD
DIS
40
20
13
9
7
6
5
SS
85
9
2
1
1
,1
TP
71
.15
5
3
3
2
1
SOL-P
65
14
10
5
3
2
1
N03
55
15
10
8
7
3
2
Daily monitoring of individual storm events
revealed that the majority of loading occurred within
Results from daily sampling were used to calcu-
late loading errors produced under the project's
fixed weekly sampling schedule. Errors in weekly
91
-------
Proceedings of National RCWP Symposium, 1992
loading estimates varied as a function of the duration
between the storm event and sampling (Table 3). If
sampling occurred during the first 24 hours after the
storm, weekly loads were overestimated by an
average of 505, 405, 365, and 295 percent for
suspended solids, total-P, soluble-P, and nitrate,
respectively. Errors were greater for particulate
species because a larger percentage of their weekly
load occurred on Day One after storms. Conversely,
when sampling occurred five or more days after the
storm, weekly loads were underestimated by ap-
proximately 93 percent for suspended solids, 79 per-
cent for total and soluble phosphorus, and 50
percent for nitrate.
Table 3.—Percent loading error for suspended solids
(SS), total phosphorus (TP), soluble phosphorus
(Sol-P), and nitrate (NOs) based on extrapolating a
single sampling event over seven days versus sam-
pling dally for seven days.
*DAYS
AFTER '
STORM
1
2
3
4
5
6
7
PERCENT ERROR IN WEEKLY LOAD
SS
505
-36
-86
-93
-93
-93
-93
TP
405
+/-
-64
-79
-79
-86
-93
SOL-P
365
+/-
-29
-64
-79
• -86
-93
NO3
295
+/-
-29
-43
-50
-79
-86
The implications of this sampling bias were used
to evaluate the project's annual loading estimates.
Average loading values from storm monitoring were
substituted for existing project estimates when more
than 0.5 inches of rainfall occurred. Average weekly
loading values following storms were 50,000, 60,15,
and 1,700 kilograms for suspended solids, total-P,
soluble-P, and nitrate respectively. On average, rain-
fall of 0.5 or greater occurred 25 days per year in the
project area. Adjusted loads indicated that only 3,13,
20, and 17 percent, respectively, of the annual loads
occurred from outside these storms. More impor-
tantly, adjusted loads indicated the project measured
only 19,34,47, and 46 percent of the annual loads for
suspended solids, total-P, soluble-P, and nitrate
respectively (Table 4).
Table 4.—Project's loading estimate (observed) ver-
sus adjusted loading estimate based on storm
monitoring results (predicted) for suspended solids,
total and soluble phosphorus, and nitrates at Station
MEAN
ANNUAL
LOAD
Observed:
Predicted:
Percent
observed:
SUSP.
SOUDS
(mton) •
240
1,295
19%
TOTAL-P
(kS)
590
1,725
34%
SOL-P
(kg)
220
470
47%
NITRATE
(mton)
24
52
46%
Spatial Variability in Pollutant
Loading
Areal normalized export rates revealed large dif-
ferences in loadings patterns among the project's
subbasins (Table 5). Rates varied the most for
suspended solids (eightfold) and least for soluble-P
(twofold). Differences among basins were specific to
individual parameters. For example, in Subbasin 7,
suspended solid export was nearly sixfold less than
in Subbasin 3 but nitrate export was nearly sixfold
greater (TableS).
Table 5.—Export rates for runoff, suspended solids,
total and soluble phosphorus, and nitrate calculated
as annual mean load divided by watershed size.
EXPORT RATE (kg/ha/yr)
STATION
3
4
5
6
7
8
9
AREA
(ha)
1,610
1,980
11,780
1,000
780
31,200
3,940
RUNOFF '
(cm)
20
19
33
30
61
28
10
SS
193
183
267
33
31
106
61
TP
0.29
0.28
0.38
0.26
0.26
0.50*
0.15
SOL-P
0.07
0.08
0.09
0.11
0.06
0.21
0.06
NO3
8.0
26.7
12.0
19.4
44.5
10.2
6.0
*After subtracting known point source contributions.
Regression of annual mean total-P concentration
versus annual mean discharge also indicated that
subbasins responded differently (Fig. 3). Subbasin
streams exhibited similar phosphorus concentra-
tions during years with lower mean discharge; how-
ever, as mean discharge increased, concentrations
varied greatly, as indicated by the different slopes.
Station 9 showed the greatest increase in phos-
phorus concentration whereas Station 5 showed only
a minimal response. Variations in response could
have resulted from differences in the physical char-
acteristics of the basins and in current land use ac-
tivities that would affect source amounts and
transport efficiencies.
To examine spatial variability at the subbasin
level, a longitudinal transect was created along
Macon Creek (Fig. 1). Concentration patterns were
examined to evaluate whether individual farm sites
were having a disproportionately large effect on pol-
lutant loads and survey changes that occur during
transport. Differences among transect stations were
examined statistically using a nonparametric
Kruskhal-Wallace test (Table 6). Dataware stratified
according to low-flow and stormflow conditions. Con-
centration patterns were quite similar throughout
the transect under both conditions. Surprisingly,
more significant differences occurred for total-P,
soluble-P, and ammonia concentrations under low-
flow versus stormflow conditions. These results
imply that internal cycling during transport affected
92
-------
T.H. JOHENGEN S A.M. BEETON
STATION 9
400-1
g> 300
I
200-
100-
Sta9
Sta6
Sta5
0.0 0.1 0.2 0.3 0.4 0.5
ANNUAL MEAN DISCHARGE (m3/sec)
Figure 3.—Annual mean total phosphorus concentration
plotted against annual mean discharge for each sub-
basin within the Saline Valley watershed. Regressions
were based on means from 1982 through 1989 and were
all significant at the 0.05 level.
concentration patterns more than differences in
input from runoff. Presumably the slower velocities
and less turbid water during low-flow conditions al-
lowed sufficient time for biogeochemical processes
to alter concentration patterns. Under stormflow
conditions, only nitrate and chloride showed consis-
tent differences between stations. These results are
interesting because inputs for both these parameters
tended to be dominated by subsurface flows.
Discussion
Temporal Variability
Temporal variability in loading rates over daily and
annual scales were examined to evaluate the
reliability of the project's monitoring design.
Cumulative load distributions revealed that a sig-
nificant portion of the annual load occurred in only a
few sampling intervals. Specifically, over 75 percent
of the suspended solids and 50 percent of the total
and soluble phosphorus annual loads were received
in 8 percent of the time.
The significance of a few individual storms to an-
nual loads has also been observed in previous
studies (Taylor et al. 1971; Johnson et al. 1976) and
Table 6.—Kruskal-Wallace test for concentration differences among Macon Creek transect stations during
stormflow and baseline conditions (N=58 and 61 respectively) (Alpha < 0.1 reported as NS).
PARAMETER
TOTAL-P
SOLUBLE-P
NITRATE-N
AMMQNIA-N
SILICA
CHLORIDE
STORMFLOW
STA:
9-2
9-3
9-4
9-5
STA:
9-2
9-3
9-4
9-5
STA:
9-2
9-3
9-4
9-5
STA:
9-2
9-3
9-4
9-5
STA:
9-2
9-3
9-4
9-5
STA:
9-2
9-3
9-4
9-5
9-3
NS
9-3
NS
9-3
.00
9-3
NS
9-3
NS
9-3
NS
9-4
NS
NS
9-4
.08
NS
9-4
.00
.07
9-4
.08
NS
9-4
NS
NS
9-4
NS
NS
9-5
NS
NS
NS
9-5
NS
NS
NS
9-5
.00
.01
NS
9-5
NS
NS
NS
9-5
NS
NS
NS
9-5
.05
NS
NS
9
NS
NS
NS
NS
9
NS
NS
NS
NS
9
.00
.00
.04
NS
9
NS
NS
NS
NS
9
NS
NS
NS
NS
9
.00
.00
.02
.07
STA:
9-2
9-3
9-4
9-5
STA:
9-2
9-3
9-4
9-5
STA:
9-2
9-3
9-4
9-5
STA:
9-2
9-3
9-4
9-5
STA:
9-2
9-3
9-4
9-5
STA:
9-2
9-3
9-4
9-5
9-3
.09
9-3
NS
9-3
.00
9-3
.00
9-3
NS
9-3
NS
BASELINE
9-4
.01
NS
9-4
NS
NS
9-4
.00
NS
9-4
.01
NS
9-4
NS
NS
9-4
NS
NS
9-5
.06
.00
.00
9-5
NS
.03
.00
9-5
.00
.00
.00
9-5
NS
.03
.01
9-5
NS
NS
NS
9-5
NS
NS
NS
9
NS
NS
.02
NS
9
NS
NS
.03
NS
9
.00
.01
NS
NS
9
NS
NS
.02
NS
9
NS
NS
NS
NS
9
.02
.03
.02
.01
93
-------
Proceedings of national RCWP Symposium, 1992
appears quite characteristic of nonpoint source pollu-
tion. The magnitude of loads following storms is
great because the effect of increased discharge is
multiplied by that of increased concentration. The
magnitude and duration of increased concentration
appears to be a function of the amount of pollutant
that has accumulated in the watershed and the rate
at which it is washed off. Daily monitoring data of
storm runoff indicated that loading spikes typically
lasted only a few days. Spikes were always greatest
for participate species, reflecting the effect of in-
creased erosion from overland runoff and probably
the increased carrying capacity of the stream during
higher velocities. Dissolved nutrient concentrations
remained elevated for several days after the storm
ended, presumably because of subsurface inputs,
and consequently, daily loading rates decreased
more slowly.
The importance of these few large storms to
total annual loads implies that BMPs must be
designed to handle runoff conditions that develop
under intense storms. If BMPs are not effective
against these major events, they may reduce only an
insignificant portion of the annual loads.
During each study year, annual loads were
dominated by a few events. The timing and frequen-
cy of these loading spikes were, however, highly
variable among years. These results imply that er-
rors in loading estimates using a fixed sampling
schedule would be inconsistent between years. In-
consistent errors will compound the problem of high
variability resulting from meteorological conditions
and could invalidate a monitoring program's con-
clusions about BMP effectiveness.
Monitoring data from individual storm events
revealed that the project's annual loading estimates
were extremely inaccurate and could not be used to
assess the effects of the land treatment program.
The magnitude of variability revealed by this study
strongly suggests a mandate for continuous dis-
charge meas- urements and event-based sampling to
accurately evaluate nonpoint pollution loading. In ad-
dition, the effects of varying amounts of discharge
must be removed from annual loading or concentra-
tion trends to provide a sensitive measure of land
treatment effectiveness. In an earlier analysis of the
project's data, changes in empirical regressions of
concentration versus discharge were examined to
evaluate whether BMPs reduced pollutant export
Qohengen et al. 1989). No significant effects were
found, and it was concluded that there was insuffi-
cient coverage of BMPs. Present results also sug-
gest that sampling biases could have affected
concentration and discharge regressions, greatly
reducing their sensitivity.
Comparisons between storm event sampling and
fixed weekly sampling indicated that the fixed
schedule produced extreme errors in loading es-
timates because the fixed schedule assumed that
loading rates were constant over the entire duration
between sampling intervals. Daily monitoring indi-
cated loading spikes lasted only a few days; there-
fore, loads were overestimated if sampling occurred
within one or two days after the storm and underes-
timated if more than two days passed before sam-
pling. This timing also meant that the weekly
sampling schedule would often miss the loading
from a storm completely. The final result was that
the project's annual loading estimates were only 19
to 47 percent of calculated loads accounting for
storms.
Stevens and Smith (1978) found that even with
load rating curves and continuous discharge meas-
urements, an eight-day fixed sampling schedule un-
derestimated nitrate loading by 18 percent and
overestimated particulate-P loading by 43 percent.
Errors are produced even when using rating curves
because concentration values are predicted from
relationships with discharge, and the fixed schedule
tends to oversample more common low-flow condi-
tions. Johnson (1979) suggested that, in the absence
of automated samplers that can sample on a flow-
proportioned schedule, a varying frequency
schedule should be used that samples all stages of
flow. Sharpley et al. (1976) reported that sampling in-
tervals for some streams should be as short as 60
minutes during peak flow to produce loading es-
timates with less than a 15 percent error. Resources
were not available in this study to examine loading
patterns at such time scales.
Spatial Variability
The RCWP was unique because it attempted to es-
tablish relationships between land treatment prac-
tices applied at the watershed level and resulting
changes in the water quality of receiving water-
bodies. Few projects, however, addressed the issue
of spatial variability within their watersheds. Several
projects concurred that sampling within the sub-
basin helped in examining the relationship between
land treatment and water quality (Nati. Water Qual.
Eval. Proj. 1989). Results were not, however, dis-
cussed in terms of the variability in response among
subbasins or the implications for site-specific effects.
This study revealed extreme variability in pol-
lutant export rates among subbasins; differences
were specific to the individual parameters. Sys-
tematic differences in concentration and loading pat-
terns among subbasins can occur as a result of their
94
-------
T.H. JOrfEffGEN & A.M. BEETOfi
size, topography, soil type, land use, and drainage ef-
ficiency (Baker, 1985). These findings imply that in-
dividual subbasins may respond quite differently to
the applied land treatment, and BMP effectiveness
should be assessed at the subbasin scale. Monitor-
ing at the subbasin level also establishes stations
closer to the sites of BMP implementation and can
reduce the influence from nonparticipating areas or
other sources of pollution.
Transect stations established within Subbasin 9
revealed that internal cycling within the stream can
also affect concentration patterns. These results also
imply the need to monitor water quality trends as
close to the site of land treatment as possible. Al-
though transect stations did,not reveal many sig-
nificant differences in pollutant inputs during
storms, they did catch occasional spikes that were
orders of magnitude different. These site-specific
results could help target critical areas within the sub-
basin.
Variations in loading rates among basins could
be used to focus attention on those areas that may
have the greatest effect on water quality. Targeting
critical areas was expounded as the best way to max-
imize cost-effectiveness (Natl. Water Qual. Eval.
Proj. 1989). One difficulty with this approach is that
extensive mon- itoring must take place before the
land treatment program can be initiated. However,
although this degree of initial monitoring could ul-
timately make BMPs more cost effective, it may be
difficult to convince managers to invest in this ap-
proach.
Conclusions
Temporal and spatial variability in pollutant loadings
were extremely high throughout the study. Cumula-
tive loading distributions indicated that a few storms
can produce over 50 percent of the total annual pol-
lutant load. Loading spikes typically lasted only a few
days; therefore, their impact was often missed be-
cause the project mandated weekly sampling. Load-
ing adjustments based on a daily precipitation record
indicated that the project estimated only 20 of the
suspended splids load and 50 percent of the total and
soluble phosphorus load. Storm monitoring results
suggest that a continuous discharge record and flow-
proportional sampling are necessary to establish ac-
curate loading estimates for nonpoint source
pollution. Without this detailed record, evaluating
long-term trends would be impossible, given the ex-
treme temporal variability.
Differences in concentration and loading pat-
terns aftiong the project's stations revealed a high
degree of spatial variability within the watershed.
For transect stations established along Macon
Creek, variability was greater during low-flow versus
stormflow conditions. Results suggest that monitor-
ing at the subbasin level would help target critical
areas and improve chances of detecting water quality
changes resulting from land treatment application.
ACKNOWLEDGMENT: This paper is Contribution No.
804 of the Great Lakes Environmental Research Laboratory.
References
Baker, D. 1985. Regional water quality impacts of intensive row-
crop agriculture: a Lake-Erie Basin case study. J. Soil Water
Conserv. 40:125-31.
Johengen, T.H., A.M. Beeton, and D.W. Rice. 1989. Evaluating the
effectiveness of best management practices to reduce
agricultural nonpoint source pollution. J. Lake Reserv.
Manage. 5:63-70.
Johnson, A.H. 1979. Estimating solute transport in streams from
grab samples. Water Resour. Res. 15:1224-28.
Johnson, A.H., D.R. Bouldin, E.A. Goyette, and A.M. Hedge. 1976.
Phosphorus loss by stream transport from a rural watershed:
quantities, processes, and sources. J. Environ. Qual. 5:148-57.
National Water Quality Evaluation Project. 1989. 1988 Annual
Report: Status of Agricultural Nonpoint Source Projects. EPA
506/9-89/002. North Carolina State Univ., Raleigh.
Sharpley, A.N., J.K. Syers, and P.W. O'Conner. 1976. Phosphorus
inputs into a stream draining an agricultural watershed. I:
Sampling. Water Air Soil Pollut. 6:39-52
Stevens, RJ. and R.V. Smith. 1978. A comparison of discrete and in-
tensive sampling for measuring the loads of nitrogen and
phosphorus in the River Main, County Antrim. Water Res.
12:823-30.
Taylor, A.W., W.M. Edwards, and B.C. Simpson 1971. Nutrients in
streams draining woodland and farmland near Coshocton,
Ohio. Water Resour. Res. 7:81-9.
95
-------
-------
Effects of Pipe-Outlet Terracing on
Runoff Water Quantity and Quality at
an Agricultural Field Site, Conestoga
River Headwaters, Pennsylvania
Patricia L. Lietman
U.S. Geological Survey
Lemoyne, Pennsylvania
ABSTRACT
Terracing effects on runoff were investigated between 1983 and 1989 at a 22.1-acre agricultural site
in Lancaster County, Pennsylvania, as part of the Rural Clean Water Program. The site, underlain
by carbonate rock, was primarily corn and alfalfa fields with a median slope of 6 percent. Average
annual runoff was 11 percent of 44 inches of precipitation. Runoff quantity, suspended sediment,
nutrient, and precipitation data were collected for 21 months before, and 58 months following, pipe-
outlet terrace construction. Data were analyzed using graphic, regression, covariate, cluster, and
Mann-Whitney techniques.
Terracing changed runoff characteristics. Storm characteristics were similar throughout the
study period; however, after terracing, storms producing less than 0.4 inches of precipitation rarely
produced runoff. Total storm discharge as a function of precipitation did not change significantly
throughout the range of runoff-producing storms after terracing. Multiple discharge peaks that ap-
peared on hydrographs before terracing did not occur after terracing. Then hydrographs reflected
the stepwise draining of each terrace through the pipe outlet
After an initial two-year period of terrace stabilization, suspended sediment yield in runoff
decreased significantly as a function of runoff. This result was expected because terracing
decreased runoff energy and because terrace ponding allowed time for sediment redeposition.
Nitrate plus nitrite yields increased proportionally throughout the range of runoff during the
postterracing period relative to the preterracing period. After terracing, increased nitrate con-
centrations were found in runoff, possibly from a combination of increased soil contact time and in-
creased nitrification caused by wetter soils.
No significant change was found in total nitrogen, ammonium plus organic nitrogen, or total
phosphorus yields relative to runoff before and after terracing. Limited data suggest that fine sedi-
ment particles Qess than 0.62 micrometers in diameter), which continued to be discharged from
the site, transported most of the phosphorus.
The U.S. Geological Survey (USGS), in
cooperation with the Pennsylvania Depart-
ment of Environmental Resources, con-
ducted a study to determine the effects of
agricultural best management practices (BMPs) on
surface and groundwater quality. The study involved
water-quality monitoring in the Conestoga River
headwaters area over a 10-year period (1982 to 1991)
on three scales: regional, small watershed, and field.
A detailed description of the overall study can be
found in Chichester (1988). This project was one of
five comprehensive monitoring and evaluation
projects in the Rural Clean Water Program (RCWP).
This report documents the effects of terracing
on surface water quantity and quality at Field Site 1.
Climatological and agricultural activity data are com-
pared to surface water quantity and quality data col-
lected before and after implementation of terracing.
97
-------
Proceedings of National RCWP Symposium, 1992
The pre-BMP period (January 1983 through Septem-
ber 1984) is Period 1. Data from the study's post-
BMP phase are grouped into three time periods:
October 1984 through September 1986 is Period 2,
October 1986 through September 1988 is Period 3,
and October 1988 through September 1989 is Period
4. Period 3 is the most comparable to Period 1 in
terms of annual precipitation and nutrient applica-
tions to the field. In addition, by October 1986, the
terraces were well established (settled) and crop-
ping patterns were stable. Therefore, data from
Period 3 and Period 1 are compared in this report to
discuss changes related to terracing.
All species of nitrogen or phosphorus are ex-
pressed in their elemental forms. For example,
whether discussing ammonium or nitrate concentra-
tions or loads, all values are expressed as nitrogen.
The term ammonium refers to the ammonium ion
plus free ammonia. A water year is the 12-month
period beginning October 1 and ending September
30; it is designated by the calendar year in which it
ends.
Study Area Description
Location and Description
Field Site 1 is located in the Conestoga headwaters
basin, between Churchtown and Goodville, Lan-
caster County, Pennsylvania (Fig. 1). The 22.1-acre
site, underlain by carbonate rock, is conventionally
tilled cropland, planted primarily in corn and alfalfa.
Soils at the site are classified as Duffield and
Hagerstbwn silt loams. The soils, formed in
residuum of weathered carbonate rocks, are up to 60
inches deep and moderately to well drained. The site
has an overall slope of about 6 percent.
Best Management Practices
The BMPs implemented at Field Site 1 were terrac-
ing and nutrient management. Six pipe-outlet ter-
races, designed to accommodate a 5-inch, 24-hour
storm, were constructed between October 19, 1984,
and November 16, 1984. (The four lower terraces
were rebuilt in May 1985 shortly after runoff from a
thunderstorm on May 21, 1985, overflowed the ter-
races and created severe gully formation.) Terrace
construction changed the site's topography and in-
creased the surface drainage area from 22.1 to 23.1
acres (Fig. 2). The terraces were graded so that each
terrace sloped gently downgradient toward the
field's center. The terraces trapped runoff creating
ponds that drained through a pipe outlet. During the
post-BMP period, surface runoff was water leaving
the upper 90 percent of the site through the pipe-out-
let system plus runoff from the unterraced lower 10
percent of the site. Under extreme storm conditions,
runoff breached the terrace structures. In contrast,
surface runoff during the pre-BMP period was unim-
peded by structural devices and created large feeder
gullies throughout the site and massive receiving
gullies in the lower one-third of the site.
The nutrient-management plan for the site was
developed in November 1984. The plan was based on
nitrogen only and, as a result, recommended an ex-
cess of phosphorus when manure was used to satisfy
the crops' nitrogen needs. The nutrient management
plan recommended an annual nitrogen application
rate of 350 to 375 pounds per acre of nitrogen within
5 to 7 days of cornfield plowing (less than a 15 per-
cent reduction from that applied during the pre-BMP
period). In addition, the plan recommended an
average application rate of 100 pounds per acre of
nitrogen for newly seeded alfalfa fields (with no ap-
plication recommended on established alfalfa).
Nutrient management recommendations were
followed within 25 percent in the 1985, 1986, and
1988 crop years, with adjustment to applications for
rapid incorporation of a substantial portion of the
manure for 1986 and 1988. However, the planned ap-
plications were exceeded by about 2.5 times in the
1987 and 1989 crop years. Nutrient management also
recommended that the timing of nutrient applica-
tions be changed from daily spread to the time the
crops could most readily use these nutrients.
Flexibility in the timing of nutrient applications was
made possible by constructing a facility to store
manure for 6 months.
Sampling Network and Data
Analysis
Precipitation was collected at the site in a 13-inch
funnel mounted above, a plastic receiving pipe;
precipitation quantity in the pipe was recorded every
5 minutes with an Analog Digital Recorder (ADR)
(Fig. 2)
-------
P.L LETMAN
PENNSYLVANIA
40° 15'—
CONESTOGA HEADWATERS
AREA
MORGANTOWN
'
CHUCHTOWN
40'
3 5
1
D 6
10 MILES
1
12 KILOMETERS
Figure 1.—Conestoga headwaters study area and Field Site 1 location.
about 4°C from time of collection until time of
analysis.
Detailed information on methods of data collec-
tion and sample and data analysis is presented in
Chichester (1988). A summary of the data collection
network is given in Table 1, and data collection sites
are shown in Figure 2.
Water quality data collected during this study,
published in U.S. Geological Survey Water Resour-
ces Data Reports PA-83-2 to PA-89-2 (U.S. Geo. Surv.,
1984-90), are stored in the USGS WATSTORE and
U.S. Environmental Protection Agency (EPA)
STORET databases. The data were catalogued by
the USGS using the local identification numbers
used in this report. All agricultural activity and soil
Table 1.—Data collection network at Field Site 1.
data are on file in the USGS office in Lemoyne, Penn-
sylvania.
Statistical procedures were used for summariz-
ing data, making statistically supported inferences
about the data, and defining explanatory relations
between various data. Statistical inferences were
based on results of hypothesis testing at the 95 per-
cent confidence interval except where noted. The
null hypothesis was always that no significant
change occurred from Period 1 to Period 3. Sum-
mary statistics, linear regression, analysis of
covariance, cluster analysis, and Mann-Whitney test-
ing were run on software from P-STAT, Inc. (1989).
To estimate annual load, suspended sediment
and nutrient loads for unsampled runoff events were
estimated by using regression equations derived
MEDIUM
NUMBER OF
LOCATIONS
DATA COLLECTION
FREQUENCY
ADNALYSES PERFORMED
OR DATA COLLECTED
Precipitation
5-minute intervals during storms; 7 times
during study period
Volume
Agricultural activities
Entire site
Biweekly
Nutrient concentration
Manure
Varied
At selected major applications
Nutrient application timing and rates;
planting, plowing, and harvesting
locations and dates
Soil
Varied
Spring, summer, fall
Nutrient concentration
Runoff
All runoff events
Most runoff events
Volume
Suspended sediment and nutrient
concentration
99
-------
Proceedings of National RCWP Symposium, 1992
75°58'56"
75°58'43
40*07'SO"
40*07'38"
40*07'50"
4O*07"38"
BEFORE TERRACING
0 ISO FEET
r-W
0 25 50 METERS
| EXPLANATION
DRAINAGE-BASIN DIVIDE
— 460— TOPOGRAPHIC CONTOUR LINE.
INTERVAL 10 FEET. DATUM IS
SEA LEVEL
— TERRACE CREST
^ PRECIPITATION STATION
01576083 Jk. SURFACE-RUNOFF GAGING STATION
0 INTAKE PIPE FOR TERRACE
DRAINAGE SYSTEM
01578083
Figure 2.—Field Site 1 topography and surface: runoff, groundwater, and precipitation sampling locations before and
after terracing.
from log-transformed loads as a function of log-trans-
formed total runoff from sampled runoff events.
Regression analyses were performed separately for
data from the growing season (May through Oc-
tober) and the nongrowing season (November
through April) to make calculations on the basis of
runoff occurring during similar field conditions. For
the 79 months of study, 19 percent of the suspended
sediment load, 21 percent of the total nitrogen load,
and 22 percent of the total-phosphorus load were es-
100
-------
RL UETMAN
timated. Data for estimated and sampled individual
runoff events were summed to calculate annual
loads.
Climate
Annual precipitation at Field Site 1 ranged from 35.6
inches during the 1986 water year to 59.8 inches
during the 1984 water year (Table 2). Table 2 shows
that Period 3 was the most comparable period to
Period 1 for total rainfall; total precipitation at Field
Site 1 was slightly below normal during the 1983 and
1988 water years, and annual precipitation substan-
tially exceeded the normal average during the 1984
and 1987 water years.
During the pre-BMP Period 1 (1983-84 water
years), 169 storms occurred. The median storm
precipitation was 0.29 inch, the median storm dura-
tion was 2.8 hours, and the median maximum 15- and
30-minute intensities were 0.08 and 0.13 inch,
respectively. Nineteen storms in the period produced
a total precipitation of 1.0 inch or greater, and four
storms produced 2.0 inches or greater. During
Period 3 (1987-88 water years), 148 storms occurred.
The median total storm precipitation was 0.33 inch,
the median storm duration was 3.4 hours, and the
median maximum 15- and 30-minute intensities were
0.07 and 0.10 inch, respectively. Twenty-four storms
produced 1.0 inch or more of precipitation, and six
storms produced 2.0 inches or more of precipitation.
Thus, on the basis of this qualitative assessment,
Periods 1 and 3 were climatologically similar.
Agricultural Activities
Year-to-year crop acreages and summer cropping
patterns for the 1983-89 crop years ("crop year" is
defined as the interval of time beginning immedi-
ately after the harvest of 1-year's corn crop at the site
and ending with the harvest of the next corn crop)
are shown in Figure 3. A substantial shift in cropping
occurred at the site during the study period. Period
3 was the study period most representative of the
terracing BMP because terraces were settled by Oc-
tober 1986, and the planned cropping pattern was es-
tablished.
Annual totals of nitrogen and phosphorus ap-
plications were tabulated by crop year rather than by
water year to determine with the greatest possible
accuracy the amount of applied nutrients available to
a season's crop.
Dairy cattle manure was the primary source of
nutrients applied to the site, comprising about 95 and
85 percent of the total nitrogen and phosphorus ap-
plications, respectively. Commercial fertilizer and
nitrogen from precipitation made up the remainder.
Manure applications were concentrated on the corn
acreage, whereas the alfalfa and soybean acreage
received relatively small applications (Table 3).
Period 3 was the post-BMP period most com-
parable to Period 1 with respect to crops and nutrient
applications to the site (Fig. 3 and Table 3). During
both periods, an average of 65 percent of the site was
planted in corn. The corn planting and harvesting
times were about the same in Periods 3 and 1. A
Table 2.—Annual precipitation at Field Site 1 and the 30-year mean from a nearby raingage at Morgantown, PA.
PRECIPITATION IN INCHES
PERIOD
DATES
ANNUAL PRECIPITATION
FIELD SITE 1
30-YEAR MEAN
MORGANTOWN1 (1951-80)
1
2
3
4
Jan 1. to Sept. 30, 1983
Oct. 1 , 1983 to Sept. 30, 1984
Oct. 1 , 1984 to Sept. 30, 1985
Oct. 1, 1985 to Sept. 30, 1986
Oct. 1 , 1986 to Sept. 30, 1987
Oct. 1, 1987 to Sept. 30, 1988
Oct. 1, 1988 to Sept. 30, 1989
31.4
59.8
41.7
35.6
46.2'
41.3
45.0
31.9
41.5
41.5
41.5
41.5
41.5
41.5
1Data from the National Oceanic and Atmospheric Administration (1985).
Table 3.—Annual nutrient applications to Field Site 1 by crop (all values in pounds per acre).
CORN
ALFALFA
PERIOD
1
1
2
2
3
3
4
CROP YEAR
1983
1984
1985
1986
1987
1988
1989
NITROGEN
150
640
230
290
690
300
540
PHOSPHORUS
33
170
61
72 .-.
130
80
160
NITROGEN
0
31
9
7
24
0
290
PHOSPHORUS
0
8
15
12
0
0
84
101
-------
Proceedings of'National RCWPSymposium, 1992
1983 and 1984 (PERIOD 1)
CORN, 65 PERCENT
ALFALFA, 24 PERCENT
OTHER, 10 PERCENT
1985 (PERIOD 2)
CORN, 77 PERCENT
ALFALFA., 13 PERCENT
OTHER, 5 PERCENT
1986 (PERIOD 2)
CORN, 58 PERCENT
ALFALFA. 38 PERCENT
OTHER, 4 PERCENT
1987 (PERIOD 3)
CORN, 62 PERCENT
ALFALFA, 33 PERCENT
OTHER, 5 PERCENT
1988 (PERIOD 3)
CORN. 67 PERCENT
ALFALFA, 33 PERCENT
OTHER, 0 PERCENT
1989 (PERIOD 4)
CORN, 34 PERCENT
ALFALFA, 59 PERCENT
OTHER, 7 PERCENT
EXPLANATION
CORN
ALFALFA
OTHER
Figure 3.—Cropping patterns at Field Site 1 for the 1983-89 crop years.
102
-------
P.L LIETMAN
winter cover crop was planted during both years of
the pre-BMP period, but little winter cover crop was
planted during the post-BMP period. The average
annual applications for the site were about 320
pounds per acre nitrogen and 82 pounds per acre
phosphorus during Period 3, and 270 pounds per
acre nitrogen and 70 pounds per acre phosphorus
during Period 1.
Effects of Terracing on
Surface Runoff
Quantity
Surface runoff during the pre-BMP period was unim-
peded by structural devices and created small feeder
gullies and larger receiving gullies over much of the
area planted in corn (Fig, 3). A massive gully was
created in the lower cornfield. In contrast, most sur-
face runoff during the post-BMP period, after instal-
lation of the terraces, discharged through the
pipe-outlet system. Twice, under extreme storm con-
ditions, runoff breached the terrace structures.
Measurable runoff from Field Site 1 during Periods
1 (21 months), 2 (24 months), 3 (24 months), and 4
(10 months), was recorded 97, 58, 52, and 29 times,
respectively. As shown in Figure 4, runoff during the
post-BMP periods occurred only when storms were
larger and generally of longer durations and higher
intensities than storms during the pre-BMP period.
These changes were a result of terrace construction,
which changed runoff characteristics.
The change in runoff character was most evi-
dent in the change in the shape of runoff
hydrographs for the different periods (Fig. 5).
During Period 1, before terracing, the hydrograph
had numerous peaks of different sizes caused by
varying intensities of rainfall. Surface runoff from
different parts of the field could also be distin-
guished on the hydrograph. The most downgradient
part of the field peaked first, followed by subsequent
peaks from the upgradient cornfield. After terracing,
water retention in the terraces and steady drainage
through the pipe outlets resulted in a stepwise
decline in stage as the terraces drained, regardless
of precipitation intensity. During summer 1985, the
initial hydrograph peak (Fig. 5) was associated with
runoff from the corn field downgradient from the
first terrace and with initial outflow from the ter-
races. During Period 3, when the field downgradient
from the first terrace was established in alfalfa, the
initial runoff peaks generally did not occur.
For Periods 1, 2, 3, and 4, respectively, 9.8, 15,
13, and 3.9 percent of the total precipitation was dis-
charged from the site as runoff (Table 4). For the
pre-BMP period, the runoff varied from 3 percent of
precipitation for the 9-month study period in 1983
January through September), to 13 percent for the
1984 water year. The distribution of rainfall probably
accounted for the low runoff during 1983. Many
small and moderate rainfalls occurred during the
spring when the winter crop cover was high, and
many small rainfalls occurred in the early summer
on recently plowed, loose soil. Annual variation was
much less during Periods 2 and 3.
Multiple regression analyses of data from the
pre-BMP period (unpublished report on file, USGS
office, Lemoyne, Pennsylvania), which were used to
determine explanatory variables, suggested that
total runoff was primarily controlled by total
precipitation and antecedent soil moisture (based on
total precipitation for 30 days before a runoff event).
Regression analyses also suggested that crop cover
on corn acreage reduced total runoff. Analysis of the
pre-BMP data also showed that nearly all rainfall ran
off when the soil was frozen. Results from multiple
regression analysis of the post-BMP data show that
the total runoff after terracing was controlled by the
same factors — total precipitation and antecedent
soil moisture — as was total runoff before terracing.
After terracing, the measured mean and maximum
storm discharges at the gage were controlled by the
pipe-outlet dynamics in addition to the climatological
Table 4.—Annual runoff, suspended sediment, and nutrient yields in runoff from Field Site 1.
PERIOD
1
1
2
2
3
3
4
WATER
YEAR
19831
1984
1985
1986
1987
1988
19892
TOTAL
RUNOFF
(ft3/acre)
3,734
28,568
24,317
17,859
18,851
21,788
5,692
PERCENT OF
TOTAL
PRECIPITA-
TION IN
RUNOFF
3.3
13.2
16.1
13.8
11.2
14.5
3.9
SUSPENDED
SEDIMENT
YIELD
(tons/acre)
0.70
111.00
3.40
.83
.54
1.00
.28
TOTAL
NITROGEN
YIELD
(Ib/acre)
1.8
12.0
8.0
6.3
4.6
9.3
3.4
AMMONIA +
ORGANIC
NITROGEN
YIELD
(% of total n)
91
92
68
68
61
64
, 78
NITRATE +
ORGANIC
NITROGEN
YIELD
(% of total N)
9
8
32
32
39
36
22
TOTAL
PHOSPHORUS
YIELD
(Ib/acre)
1.0
7.0
3.1
2.9
4.0
3.8
1.3
1January through September 1983.
2Ootober 1988 through July 1989.
103
-------
Proceedings of national RCWP Symposium, 1992
0 0
0 0
O ' O
O O O X
§ ° ° * * *„
|-|M1"I' I' f Ą° ? I'
e
X
In4
1
1
3
G
DETACHED VALUE1
OUTSIDE VALUE2
75 1 H HtHCENTlLE
MEDIAN
LOWER WHISKER3
*78 * 81 * 38
da =b db
o
I •
=5
V)
z
s
25
2
Z>
2
1A value >3 times the interquartile'
range from the box.
2 A value >1 .S and S3 times the
interquartile range from the box.
3 Upper whisker is the largest data
point less than or equal ID the
upper quartile plus 1 Ł times the
interquartile range. Lower whisker
is the smallest data point greater
than minus IS times the
Q .„ i .T interquartile range.
° * n . number of observations in
analysis.
Jj7 J.38
so
40
30
O
f=
oo:
II
a.
Z 20
ur
cc
a.
2
D:
en
10
I I
i 9
,, 45
.-I4?....r^i.....423
M 9 a .a-
ALL
STORMS
1 23 4
PERIOD
STORMS PRODUCING
MEASURABLE RUNOFF
STORMS PRODUCING
NO MEASURABLE RUNOFF
Figure 4.—Total storm precipitation,,maximum 30-minute intensity within each storm, and storm duration for all
storms, storms producing runoff, and storms that did not produce runoff during Periods 1 (January 1983 through Sep-
tembor 1984), 2 (October 1984 through September 1986), 3 (October 1986 through September 1988), and 4 (October
1988 through July 1989).
104
-------
P.L LIETMAN
4.0
3.5
3.0
2.0
1.5
1.0
PERIOD 1
PRE-TERRACING
i r i i i r
PERIOD2
POST-TERRACING
i i i i I i
PERIOD 3
POST-TERRACING
0 24 6 80
TIME FROM START OF STORM, IN HOURS
Figure 5.—Hydrographs of runoff for a storm from Period 1 (June 24,1984), Period 2 (July 31,1985), and Period 3 (July
26,1988) with similar amounts of rainfall (0.65 to 0.80 Inches) and similar soil conditions.
factors such as precipitation quantity and intensity,
and antecedent soil moisture that affected mean and
maximum discharge before terracing.
Analysis of covariance was used to determine
significant differences in regression relations be-
tween total runoff and precipitation for Periods 1 and
3 for all storms that produced runoff (except those
that produced runoff on frozen ground). Runoff as a
function of precipitation was not significantly dif-
ferent in Period 3 and Period i (Fig. 6 and Table 5).
Figure 6 also shows that small amounts of precipita-
tion (Jess than 0.40 inch) rarely produced runoff
during Period 3.
Because small storms rarely produced runoff
during Period 3, regression and covariate analyses
compared pre- and post-BMP storms that produced
different amounts of runoff and precipitation. There-
fore, cluster analysis was used to compare similar
type storms to verify the results of covariate analysis.
Pre-BMP data analysis indicated that total runoff
was affected by total storm precipitation, antecedent
soil-moisture conditions, precipitation duration,
:O
10
0.1
0.01
0.001
0.0001
0.00001
0.01
-I 1—I—I—I I I I
PERIOD 1
Y=1.578x-1.254
i 1 1—i—i—i i i
PERIOD 3
Y=1.960x-1.211
— • PERIOD 1
— D PERIODS
0.1 l 1
TOTAL STORM PRECIPITATION, IN INCHES
10
Figure 6.—Total storm runoff as a function of total precipitation for all storms except storms on frozen ground for
Periods 1 (January 1983 through September 1984) and 3 (October 1986 through September 1988). (Regression statis-
tics are listed In Table 5.)
105
-------
Proceedings of National RCWP Symposium, 1992
Table 5.—Regression statistics for the log of total storm runoff (in inches) as a function of the log of storm
precipitation (in Inches) for all storms in each period, except those occurring on frozen ground. (< denotes less
than.)
PERIOD1
1
3
DEGREES
OF
FREEDOM
84
43
INTER-
CEPT
-1.254
-1.211
COEFFICIENT OF
THE LOG OF
TOTAL
PRECIPITATION
1.578
1.960
T-
STATISTIC
8.47
7.41
P-VALUE
<0.001
<0.001
COEFFICIENT
OF
DETERMINATION
(ADJUSTED R2)2
0.46
0.46
STANDARD ERROR
LOG
UNITS
0.70
.54
PERCENT3
PLUS
401
247
MINUS
80
71
'Period 1, January 1983 through September 1984; Period 3, October 1986 through September 1988.
'Coolfldont of determination (R2) adjusted for degrees of freedom.
'Calculated as described by G.D. Tasker, U.S. Geological Survey (written comm. 1978).
precipitation intensity, and crop cover. Therefore,
these characteristics were used to define eight
clusters for storms occurring during the entire 79-
month study period. Five clusters included all but
seven storms. A characterization of the clusters is
given in Table 6, and data from five clusters are
shown in Figure 7. (For the cluster analysis, a factor
from one to four was used to categorize crop cover: 1
is less than 15 percent, 2 is 15 to 49 percent, 3 is 50 to
85 percent, and 4 is greater than 85 percent.) Data
from storms that produced runoff on frozen ground
were eliminated from the data set, and runoff was set
at zero when no runoff was measured for a storm.
Table 7 summarizes changes that were detected
for four clusters by using the Mann-Whitney test to
determine differences in medians between Periods 1
and 3 for all storms and for storms producing runoff.
Cluster 3 was not included because an insufficient
number of storms occurred for comparison in Period
3. Clusters 2, 4, and 5 were not included because
they included too few storms.
Table 6.—General storm characteristics by cluster and percent of total precipitation by period and cluster (all
storms on frozen ground were excluded from the data set before clustering).
CLUSTER CHARACTERISTICS
1
2
3
4
5
6
7
8
Summer showers on moist soil with crop cover
Three large storms in Dec. 1983, Sept. 1985, and June 1987,
with 3.4 to 5.1 inches of rain
Typical spring and fall all-day storms generally with 0.2 to 0.6
inches of precipitation on soil with little crop coverage
One large Sept. 1 987 storm with 6.7 inches of rain
Three large summer storms, one in May 1 985 and two in July
1988, with 2.8 to 4.5 inches or rain
Thunderstorms occurring predominantly in the summer on
soil with crop cover
Very small storms throughout the year on dry soil; most
storms occurring on soil with little crop cover
Typical spring and fall all-day storms generally with 0.8 to 1.6
inches of precipitation on soil with little crop cover
PERIOD
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
1
2
3
4
NUMBER OF PERCENT OF TOTAL
STORMS PRECIPITATION
31
26
21
16
1
1
1
0
22
2
2
10
0
0
1
0
0
1
2
0
18
20
10
4
67
63
73
25
15
11
12
26
11
10
10
11
3.8
6.6
3.9
0
9.4
.3
1.7
,13
0
0
7.7
0
0
3.6
8.3
0
18
27
14
15
22
21
24
16
21
7
16
9
106
-------
P.L. LfETMAN
Table 7.—Mann-Whitney test results comparing within clusters (Table 7 and Fig. 7) total storm runoff and mean
storm suspended sediment and nutrient concentrations between Period 1 (1983-84) and Period 3 (1987-88);
storms on frozen ground excluded. T = statistically significant increase; | = statistically significant decrease; <->
= no statistically different change; (90) = significant at the 90 percent confidence interval; (95) = significant at the
95 percent interval; n = number of storms; mg/L = milligrams per liter; ft3/s = cubic foot per second; ft3/acre =
cubic foot per acre; and Ib/acre = pound per acre.
CLUSTER 1
CLUSTER 6
CLUSTER 7
CLUSTER 8
PERIOD 1/PERIOD 3 PERIOD 1/PERIOD 3 PERIOD 1/PERIOD 3 PERIOD 1/PERIOD 3
ALL STORMS1
Total storm runoff (ft3/acre)
Change
median
n
1(90)
85/0
31/21
4-»
54/400
18/10
1(95)
0/0
67/73
*-*
205/260
15/12
STORMS THAT PRODUCED RUNOFF
Total storm runoff (ft3/acre)
Mean suspended sediment
concentrations (mg/L)
Mean total phosphorus
concentration (mg/L as P)
Mean total nitrogen
concentration (mg/L as N)
Mean ammonia + organic
nitrogen concentration
(mg/L as N)
Mean nitrate + nitrite
concentration (mg/L as N)
Change
median
n
Change
median
n
Change
median
n
Change
median
n
Change
median
n
Change
median
n
t(90)
120/240
21/7
•«— >
2,870/2,030
19/7
<-*
2.6/2.7
12/7
t(90)
3.4/6.1
12/7
2.7/4.2
12/7
T(95)
.56/1 .7
12/7
102/740
13/9
1(95)
9,040/1,850
9/8
<— >
4.1/3.4
8/7
<-*
5.4/6.2
8/7
*-*
4.6/4.2
8/7
t(95)
.54/1.8
8/7
*->
24/80
26/10
1(95)
3,530/725
22/6
<— >
3.1/3.4
17/3
*->
5.2/7.4
17/3
4-+
4.1/4.2
17/3
T(95)
.59/4.1
17/3
•^->
260/260
13/12
1(95)
1,930/470
7/10
4->
3.1/4.3
6/7
too)
4.1/7.2
6/7
<-»
3.6/4.8
6/7
1(95)
.43/3.0
6/7
'Total and mean discharge set equal to zero if no measurable runoff occurred.
The test results in Table 7 show for clusters 1
and 7 (clusters containing small storms), fewer
storms producing runoff occurred during Period 3
than during Period 1. Terracing (which changed the
surface slopes) decreased runoff velocities, in-
creased water-soil contact time, and increased sur-
face storage. These factors promoted evaporation
and soil wetting and delayed the onset of runoff,
which was particularly apparent during small
storms. Nearly all storms in Clusters 6 and 8, which
produced "about five times more precipitation than
storms in Clusters 1 and 7, produced runoff during
all periods. The total storm discharge for storms in
Clusters 6 and 8 did not change from Period 1 to
Periods.
In summary, data analysis shows that after ter-
racing, storms that produced less than 0.4 inch of
precipitation rarely produced runoff; therefore, in ef-
fect, terracing increased the threshold at which
runoff occurred. The terrace BMP had no sig-
nificant effect on runoff quantity during larger
storms and no overall effect on runoff quantity.
Quality
Annual suspended sediment and nutrient yields
transported by runoff from Field Site 1 are shown in
Table 4. The 1984 annual suspended sediment yield
exceeded the erosion rate (T) of 4 tons per acre per
year recommended for the site by the U.S. Soil Con-
servation Service. The suspended sediment yield
was substantially less than T for all years after ter-
race installation and after one year of stabilization
(Table 4). Annual yields of total nitrogen and phos-
phorus for the post-BMP period were within the
range of yields for the pre-BMP period. However,
during the pre-BMP period (Period 1), less than 10
percent of the annual total nitrogen yield was nitrate
plus nitrite, and more than 90 percent was am-
monium plus organic nitrogen. During post-BMP
Periods 2 and 3, 32 to 37 percent of the annual total
nitrogen yield was nitrate plus nitrite (Table 4).
Pre-BMP data indicate that storm runoff on
frozen ground responded differently from runoff
during other storms. Generally, mean concentrations
107
-------
Proceedings of National RCWP Symposium, 1992
II ll
5 I * v
3* I JJ |j|||
8 I J i! Ifllt!
is iUiii/ll*!*
sr
X^j
X X
5
ss
^^^^^_
"1
K 0
ce
^ 2
"^ 3
Z
QC
UJ
3
O
S3HONI NI 'HBOIS 01 HOIMd SAVQ 02 HOJ NOIlVlldl03ad 1V101
m
ro
0
IN
XXX -
1 III
D
— l_l — 1-
<*»
o.
00 ffl
z
10 Ł
10 <75
0
(2
•o
;nn;r: —
IHCXX ^
i — r— i—
o
1
-) 1 1—
C
"5
!*• 3 (0
i g
Ld *!
«« 1— ™
10 W Jr
-3 -2
0 Ł
CO
S3HONI NI 'A1ISN31NI 3inNW-02
10 *
<
n <
M
m
X
s< H
*
1 '
^
M
o x
o
1
C .».»,•.
m ,—
•- k
" ' (
•oo
r*
i i_i-
^
-i L_J-
n <
3
Cd
00 "^
CD
^ 2
z
3
3
S3HONI NI 'NOiiViidiosad wyois
o
i
a
I
D>
108
-------
P.L LIETMAN
and loads of suspended sediment in runoff were rela-
tively low; total nitrogen concentrations and loads
were relatively high when the ground was frozen
compared to runoff from all other storms. Therefore,
all storms (except those that produced runoff on
frozen ground) were used for data analysis.
Graphic and regression analyses of the total con-
stituent yields as a function of total storm runoff
were used to investigate the relation between yields
and runoff and differences in the relation between
pre-BMP and post-BMP periods over the range of
runoff (Figs. 8 and 9 and Table 8). Suspended sedi-
ment yields as a function of runoff were lower during
Period 3 than during Period 1, that is, after terraces
had stabilized for two years and after the field
downslope from the first terrace was stabilized with
alfalfa (Fig. 8). This result was expected and was the
primary reason for installing the terraces at the site
and establishing the lower alfalfa field. Analysis of
covariance showed statistically significant changes
in the intercept and slope of the regression lines of
suspended sediment yield as a function of runoff
from Period 1 to Period 3: (F= 18.10) > (F« = .025,2,
96 = 3.83). Graphic analysis shows that moderate
storms carried about the same amount of sediment
relative to total storm runoff during Period 1 and
Period 3, but large storms carried much less sedi-
ment relative to runoff during Period 3 than during
Period 1 (Fig. 8, Table 8).
Analysis of covariance indicated that proportion-
ally more nitrate was carried in runoff during Period
3 than during Period 1: (F = 21.99) > (Foe = .025,2, 73
= 3.89). No significant changes were detected be-
tween Period 1 and Period 3 with respect to total
phosphorus, total nitrogen, or ammonium + organic
nitrogen yields as a function of total storm runoff
(Figs. 8 and 9, Table 8).
To better understand the effects of terracing on
runoff quality, mean storm concentration and yield
data for suspended sediment and nutrients were
compared within the clusters (Fig. 7 and Table 6).
a
LU
>-
t—
LJUI
Oz
10
o.
o.oi
o.ooi
0.0001 r
I
0.00001
UU
1
0.1
0.01
0.001
PCQL
Oz 0.0001
0.00001
0.1
~l—I I I I III
"I—I I I I 11
"1 1—I I I I 111
"I 1 I I I I H
n -.
PERIOD 1
Y=1.239x-4.391
PERIOD 3
Y=0.878x-4.143
• PERIOD 1
D PERIOD 3
o.i
10 | r
10
~1 1 1 III!
100
~l—I I I I 11
1.000
1 1—I I I I 111
10,000
i i i 11 a
PERIOD 3
Y=0.925x-3.438
no
PERIOD 1
Y=1.036x-3.799
• PERIOD 1
D PERIODS
10 100
TOTAL RUNOFF, IN CUBIC FEET PER ACRE
1,000
10,000
Figure 8.—Total suspended sediment (top) and total phosphorus (bottom) yield In runoff as a function of total dis-
charge for all storms, except storms on frozen ground, for Periods 1 (January 1983 through September 1984) and 3
(October 1986 through September 1988). (Regression statistics are listed In Table 8.)
109
-------
yN
O
0 "JI
?! "
J? C
*3
CO CO
CO g
•§ g
C •*"
0 ^
Ł• 1
5S' ll
it
*^ CO
c o
.2 S
~ 0)
1"
"5 1
2 w
O (A
!l
*•* Hh
«"§
•S-o
0 g
ll
gs
S|
Ł -
S
!a =
>» CO
c cS
o •»:
TJ *•
CD gg
•0 o
•S °-
11
w o
co 3
«- 3
0 U
o'ss.
0 3=
JC 0
»- c
S 2
8 ~
Ł 0
ja ^.
o 0 c
"« Ł Ł
S! o i
O) c ^
III
*^ 1
l2s&
CO
„ z
! % s
S UJ
S 0
3 g
fe s „
g 1
cc a.
r
Ul
3
a
« §|
'"^ 3
oo 3^
HP °-
ii ^
|1 =
O Hi a
OQ :i
111
5
0
. S
5
CO
2 1;
og
CO 7*
SE
gib
So
tco
a
u
li
—
UJ Q
KgS
UJ cc
a u.
a
m
<
cc
$
I
S:
Q
"o
0
1
Sco ^ in
CD in f
CO 0 05 CM
T- T— t— CO
CM CM •«-
in T- T- 0)
o> o? •* in
*3* ^" CO CM
CJ
CM o in in
co co co co
o
8888
V V V V
ix. CO in ;-
^J- T- CO CM
in r^ CD CM
O> CO CM CD
CM CO O> CO
•«-
§ 1 fe 8
T t co co
in *f o> f^-
CD CO t CM
'c 'c
CO CD
E E
T3 T3
s s g g
"2 H? o> o)
Q) (D o O
"i E ~ *"
o3 o) c c
S- S- 3 -3
Ł w i2 ^
T- CO T- CO
h- r^ co o co in
in M- co in in M-
CO O) O 0 O> CO
co co h- o co co
* T— »•" T— »—
1^ CD i- O O> CM
CO ^~ CO C? ^ CD
CO CM ^ CO CO CM
co co o o in co
8 88888
V V V V V V
1^ O CO O> IO O
h- CO CM LO CO O)
IO i- ^t CD CD CM*
00 CD CO O> CD U)
C7> CO O CO O O
T- T-
O IO CO OO O) 00
CO CO ^J" CO OO CO
? CM ? W ? 01
CD CD
CO « -•Ł.-Ł
_3 ^ _ AS :K
Ł§)Ł§)« | | §
1 g IB t i. 1 1
o'c oc CB [0
5 S
CD
CO CD
CO c
S"l-
•c !9 E
•.- ... o
*J- CD ^
CO S C
f"
J3 — ^^
E'ra fe
S o ^
CD-EI™
jjl
IS!
^ 3 tS
T3 =»•=;
-§?1
S-FSP
110
-------
RL LIETMAN
10
0.1
0.01
0.001
0.0001 r
o:
UI
Q_
0.00001
0.1
o
CL
Q
UJ
10
0.1
0.01
0.001
g 0.0001
o
eg
!i 0.00001
0.1
•s
o
10
1
0.1
0.01
0.001
0.0001 r
0.00001
0.1
TOTAL NITROGEN
PERIOD 3
Y=0.866-3.063
0 :
ai
PERIOD 1
Y=0.942x-3.379
— • PERIOD 1
, D PERIOD 3
-i 1—i—
1
10
100
1.000
10.000
i 1—I MINI 1 1—i i i i i i| 1 r
AMMONIUM PLUS ORGANIC NITROGEN
rn 1 1—i i i i 111 1 1—i i i i i b
PERIOD 3
Y=0.876-3.295
n!
PERIOD 1
Ą=0.948-3.470
—• PERIOD 1
—0 PERIOD 3
10
100
1.000
10.000
II I I I I III 1—I—I
NITRATE PLUS NITRITE
i — i i i i 1 1 1
-1 - 1 — I — I I I III
~I - 1 - 1 I I I I
PERIOD 3
Y=0.869-3.553
PERIOD 1
Y=1.006-4.528
—-• PERIOD 1
O PERIOD 3
10 100 1.000
TOTAL RUNOFF, IN CUBIC FEET PER ACRE
10.000
Figure 9.—Total nitrogen (top), total ammonium plus organic nitrogen (middle), and total nitrate plus nitrite (bottom)
yield In runoff as a function of total discharge for all storms, except storms on frozen ground, for Periods 1 (January
1983 through September 1984) and 3 (October 1986 through September 1988). (Regression statistics are listed In Table
8.)
For storms in clusters 6, 7, and 8, mean storm for deposition of suspended material before the
suspended sediment concentrations in runoff runoff discharged through the pipe outlet. For
decreased from Period 1 to Period 3 (Table 7). The storms in Cluster 1, which contained the smallest
terraces reduced runoff energy and thus its ability to storms, no significant change was detected in
transport sediment. Pooling in terraces allowed time suspended sediment concentrations from Period 1 to
111
-------
Proceedings of National RCWP Symposium, 1992
Period 3 using the Mann-Whitney test. The distribu-
tion of the data may bias the statistical analysis
within cluster 1 because no runoff was produced by
the small storms in this cluster during Period 3.
Graphic examination of the data within Cluster 1
shows that suspended sediment yields per unit of
discharge were generally larger during Period 1
than during Period 3.
No change in mean storm total phosphorus con-
centrations in runoff was detected from Period 1 to
Period 3 (Table 7). Total phosphorus concentrations
did not decrease proportionately with suspended
sediment concentrations throughout most storm
groupings (Table 7). It is believed, on the basis of ob-
servation and other limited particle size analysis,
that most of the fine sediment particles (less than
0.62 micrometers in diameter) continued to be dis-
charged from the site after terracing and that most
phosphorus is sorbed to and transported with fine-
grained particles (Sharpley et al. 1981).
Particle size analysis of instantaneous
suspended sediment samples showed that a sig-
nificantly larger percentage of the sediment in runoff
was silt and clay (sediment finer than 0.62
micrometers in diameter) after terracing than before
terracing. A median of 96 percent of the sediment in
runoff from 44 samples collected during the 1986-88
water years was silt and clay, whereas a median of 86
percent of the sediment in runoff from 175 samples
collected during the 1983-84 water years was silt and
clay.
Limited phosphorus concentration data on in-
stantaneous runoff samples shows that a median of
90 percent of the total phosphorus in runoff was
suspended before terracing (125 samples), and 82
percent of the total phosphorus was suspended after
terracing 0»2 samples). The Mann-Whitney test
detected no change in the total and suspended phos-
phorus concentrations from the pre-BMP to post-
BMP periods, but the dissolved phosphorus
concentrations increased significantly. The small
number of samples weakens any conclusions from
the data analysis, but it is possible that small in-
creases in dissolved phosphorus concentrations in
runoff offset any decreases in suspended phos-
phorus concentrations.
Within each cluster, the mean storm nitrate plus
nitrite concentrations in runoff increased significant-
ly (changes were 3- to 7-fold) from Period 1 to Period
3 (Table 7). Therefore, regardless of storm type,
nitrate transport by runoff increased after terracing.
During storms when no runoff was produced after
terrace construction, overall soil moisture may have
increased, allowing increased nitrification and, there-
fore, increased amounts of nitrate available for trans-
port by runoff. During all storms after terracing, the
soil wetting area probably increased from an in-
crease in sheet runoff and a reduction in gully
runoff. Thus, the increased contact time and possib-
ly increased contact area of the runoff with the
nutrient-rich soils allowed for an increase in the con-
version to and dissolution of nitrate. During many
storms, runoff pooled in the terraces — a process
that also increased nitrification and the solution of
nitrate.
Table 7 shows that the mean values of the storm
total nitrogen concentrations in runoff during Period
3 for each cluster were greater than or equal to the
mean values of the storm total nitrogen concentra-
tions in runoff during Period 1. However, a sig-
nificant increase in mean storm total nitrogen
concentrations in runoff was only detected from
Period 1 to Period 3 for storms in Clusters 1 and 8.
The distribution of data within Cluster 1 may bias
this statistical analysis. The increase in mean storm
total nitrogen concentrations for Cluster 8 storms
resulted primarily from the large increases in nitrate
concentrations.
No changes within clusters were found in mean
storm ammonium plus organic nitrogen concentra-
tions in runoff between Periods 1 and 3. The mean
storm ammonium plus organic nitrogen concentra-
tions in runoff were much more variable than the
mean storm nitrate plus nitrite concentrations in
runoff. Therefore, significant changes occurred in
the nitrate plus nitrite concentrations without cor-
responding significant changes in total nitrogen con-
centrations. Figure 10 shows the change in monthly
mean storm nitrate plus nitrite concentrations in
runoff for storms that were sampled following the in-
stallation of terraces in October 1984.
The monthly mean storm total nitrogen con-
centrations in runoff apparently responds to
nitrogen applications at the site (Fig. 10). However,
because of varying lag times between applications
and runoff and lack of nitrogen data for every runoff
event, no statistical relation between land-surface ap-
plications and monthly mean storm total nitrogen
concentrations in runoff could be established.
Conclusions
Terracing of a 22.1-acre agricultural field site under-
lain by carbonate rock was effective in reducing
suspended sediment losses from the site but ineffec-
tive in reducing nitrogen or phosphorus losses from
the site. Nutrient management, a planned BMP at
the site, did not result in reduced applications of
nutrients to the site.
112
-------
P.L LIETMAN
O
oc.
(ft
<
cc
Ul
Of
Ul
Q.
o:
o
Zi
z
o
Ul
0
z
o
o
z
o
ZE
40
35
25
20
15 -
10 -
5 -
I 1 I I 1 I
55
iii. ii
t.
APPLICATION
O—-TOTAL NITROGEN
• TOTAL NITRATE PLUS NITRITE
4,000
Z
Ul
3,500 o
^
Z
u.
3,000 °
in
z
2.500 CL
O
2,000 p
•^
O
_l
Q.
Q-
1,500 <
a:
ui
M
1,000
500
ONDJFMAMJJASONDJFMAMJJASONDJFMAMJJASONDJFMAMJJASONDJFMAMJJASONDJFMAMJJAS
1982 1983 1984 1985 1986 1987 1988
Figure 10.—Mean monthly water-weighted total nitrogen and total nitrate plus nitrite concentrations in runoff from
sampled storms, and total monthly nitrogen application to Field Site 1.
Terracing did change runoff characteristics.
However, terracing did not cause a change in the
relation of runoff quantity to precipitation, although
small storms (generally less than 0.4 inch) did not
produce runoff after terracing.
Suspended sediment yields as a function of
runoff were significantly reduced after a period of
stabilization following terrace construction and es-
tablishment of crop changes planned as part of the
terracing BMP. Reductions in the suspended sedi-
ment yields were larger during storms with larger
amounts of runoff.
Total phosphorus yields as a function of runoff
did not change significantly after terracing. Limited
data suggest that fine sediment particles, with which
most of the phosphorus is associated, continued to
be transported from the site.
Although nitrate plus nitrite yields as a function
of runoff increased significantly after terracing, total
nitrogen loads did not change significantly. Total
nitrate plus nitrite yields made up 35 percent of the
total nitrogen load after terracing, compared to 10
percent of the load before terracing.
References
Chichester, D.C. 1988. Evaluation of Agricultural Best Manage-
ment Practices in the Conestoga River Headwaters, Pennsyl-
vania: Methods of Data Collection and Analysis and
Description of Sites. Open File Rep. 88-96. U.S. Geo. Surv.,
Denver, CO.
National Oceanic and Atmospheric Administration. 1983-89.
Climatological Data, Annual Summary, v. 88-94, No. 13.
Washington, DC.
P-STAT, Inc. 1989. P-STAT User's Manual. Duxbury Press, Bos-
ton.
Sharpley, AN. et al. 1981. The sorption of soluble phosphorus by
soil material during transport in runoff from cropped and
grassed watersheds. J. Environ. Qual. 10:211-15.
U.S. Geological Survey. 1984-90. Water resources data for Pennsyl-
vania, water years 1983-89. Water-Data Rep. PA-83-2 to PA-89-
2. Lemoyne, PA.
113
-------
-------
Effects of Nutrient Management on
Nitrogen Flux through a Karst Aquifer,
Conestoga River Headwaters Basin,
Pennsylvania
David W. Hall and Dennis W. Risser
U.S. Geological Survey
Lemoyne, Pennsylvania
ABSTRACT
The implementation of nutrient management at a 55-acre farm near Ephrata, Pennsylvania, sub-
stantially reduced applications of nitrogen to cropped fields, nitrate concentrations in groundwater,
and nitrogen loads in groundwater discharge from the site. After implementation of a nutrient-
management plan in October 1986, decreases in applications of nitrogen to farm fields ranged from
39 to 67 percent. By 1990, decreases in nitrate concentrations in measured shallow groundwater at
the site ranged from 8 to 32 percent, based on Wilcoxon median tests. Changes in nitrogen applica-
tions to areas contributing water to five wells were correlated (by Spearman rank-sum correla-
tions) with changes in nitrate concentrations of groundwater samples. The correlation analyses
indicated that the transport time of nitrogen fertilizer from the land surface to the water table was
approximately 4 to 19 months. Annual loads of nitrogen added to and removed from the site were
estimated for 1985 through 1990. Nitrogen was added from manure, commercial fertilizer,
precipitation, and groundwater inflow; these sources averaged 93, 4, 2, and 1 percent of the
nitrogen added to the site, respectively. Nitrogen was removed in harvested crops by groundwater
outflow, volatilization, and surface runoff; these losses averaged 38, 38,24, and less than 1 percent
of the nitrogen removed from the site, respectively. Nitrogen loads in groundwater discharge were
estimated to have been approximately 300 and 200 pounds of nitrogen per million gallons of dis-
charge, respectively, before and after implementation of nutrient management.
From 1985 through 1990, the effects of
managing agricultural nutrients on nitrogen
flux in a karst aquifer were studied by the
U.S. Geological Survey (USGS) in cooperation with
the Pennsylvania Department of Environmental
Resources. The study was part of the U.S. Depart-
ment of Agriculture's Conestoga Headwaters Rural
Clean Water Program (RCWP) comprehensive
monitoring and evaluation project (Chichester,
1988).
Elevated nitrate concentrations in groundwater
within and surrounding the Conestoga Headwaters
River Basin were documented by Hall (1934), Meis-
ler and Becher (1966, 1971), and Poth (1977). In
1979, Pennsylvania developed an agricultural 208
plan (required under section 208 of the 1972 Federal
Water Pollution Control Act) that designated the
Conestoga Headwaters as the top priority watershed
in the State for water quality studies (Schueller,
1983). The objective of the Field Site 2 part of the
115
-------
Proceedings of National RCWP Symposium, 1992
Conestoga Headwaters RCWP project was to deter-
mine the effects of reduced applications of manure
and fertilizer under nutrient management (an
agricultural best management practice [BMP]) on
groundwater quality at an agricultural field site.
This paper presents statistical comparisons of
pre- and postnutrient-management nitrate concentra-
tions in groundwater, correlations of changes in
nitrogen applications to changes in concentrations of
nitrate in groundwater, and estimations of annual
nitrogen additions and removals at the site. Reduced
loads of nitrogen in groundwater discharge that oc-
curred as the result of reduced nitrogen applications
to farm fields are also discussed.
Data on agricultural activity, precipitation, water
table altitude, and groundwater quality were col-
lected at the site during a prenutrient-management
period (water years 1985-86) and a postnutrient-
management period (water years 1987-90). A water
year is the 12-month period that begins October 1
and ends September 30; it is designated by the calen-
dar year in which it ends. Elevations and altitudes
are described in reference to the National Geodetic
Vertical Datum of 1929.
Site Description
Field Site 2, located near Ephrata, Pennsylvania
(Figs. 1 and 2), is used for intensive animal and crop
production. The monitored site was 55 acres, with
approximately 47.5 acres in crops and the remaining
7.5 used for farm buildings, animal confinement
facilities, and a residence. The site, underlain by
karstic limestone and dolomite, is located within the
Conestoga Valley section of the Piedmont
physiographic province (Meisler and Becher, 1971).
Soils at the site are classified as Hagerstown silt
loams and silty clay loams (Typic hapluduljs) (Fox
and Piekielek, 1983; U.S. Dep. Agric. 1985). A clay
hardpan, where present, may retard the downward
movement of solutes. The depth from land surface to
bedrock ranges from 5.5 to 28 ft, and the depth to the
water table has a similar but noncoincident range of
5 to 30 ft below the land surface.
A recording precipitation gage was installed at
Field Site 2 in October 1984. Annual precipitation for
the study period (1985-90) is shown in Table 1. The
30-year average precipitation (1951-80) for Field Site
2 was 43.5 inches, based on records from the Nation-
al Oceanic and Atmospheric Administration (NOAA)
precipitation station at Ephrata, Pennsylvania (Natl
Oceanic Atmos. Admin. 1982). Table 1 shows a com-
parison of annual precipitation measured at Field
Site 2, long-term annual average precipitation from
the Ephrata NOAA station, and percent of measured
precipitation deviation from long-term average an-
nual precipitation. Surface runoff of precipitation oc-
curred briefly during periods of large or unusually
intense rainstorms, periods of rain on frozen ground,
and periods of rapid snowmelt.
Thirteen wells were drilled to depths ranging
from 40 to 350 ft to characterize the hydrogeology of
the site and to serve as monitoring wells for collect-
PENNSYLVANIA
STUDY AREA
CONESTOGA HEADWATERS
RURAL CLEAN WATER
PROGRAM STUDY AREA
1.0 MILES
KILOMETERS
Figure 1.—Location of the Conestoga Headwaters Rural Clean Water Program project study area and Field Site 2 near
Ephrata, Pennsylvania.
116
-------
D.VK WALL & LW. RISSER
4o°irs6"
W1V49"
1M1674
76s 11-15"
0 100 200 FEET
0 30 60 METERS
DATUM IS NATIONAL GEODETIC VERTICAL DATUM OF 1929
EXPLANATION
FARM STRUCTURE
CONTOUR LINE C5 FOOT INTERVAL)
- 360 - ELEVATION ABOVE DATUM
TERRACE CREST
r.-.-.vr;;GRASSED WATERWAY
GEOLOGIC FORMATION BOUNDARY
''' 'CONTINUOUS-RECORD GAGING
STATION WITH ID NUMBER
^ PRECIPITATION STATION
LN SP61
-------
r
Proceedings of National RCWP Symposium, 1992
LN 1680
40* 11'56*-
\
so
\
a'
\\
LN 1677
*
76°IO' 53*
FEET
3060 METERS
DATUM IS NATIONAL GEODETIC VERTICAL DATUM OF 1929
LN 1671
CONTOUR LINE 10 FOOT INTERVAL,
IN FEET ABOVE DATUM
.OGIC
SPRING WITH ID NUMBER
CHARACTERIZATION WELL
WITH ID NUMBER
MONITORING WELL
WITH ID NUMBER
MEDIAN WATER-TABLE ALTITUDE,
IN FEET ABOVE DATUM
Figure 3.—Estimated water table altitude map (cross sections A-A' and B-B' are illustrated in Fig. 4).
420
110
400
3*0
380
jj 310
2 3t°
5 390
i?
S 310
I"
° 3JO
310
300
ISO
980
.LAND SURFACE
VERTICAL EXAGGERATION X 8
DATUM IS NATIONAL QEODETIC VERTCAL DATUM OF 1929
JWU
380
370
360
350
340
330
320
310
300
290
280
270
260
9Cn
WELL
LN 16
LAND SURFACE ^^**~
WELL /X""
" WELL LN 167J/^
LN 1669 j — p
•
-
-
-
-
-
-
Ł• ~^ BEDROCK
„ .„ .ELEVATION
* BEDROCK • •. HFDROrK
< ELEVATION EL!VATION>
_ — UAMHLINU-'
.— — DEPTH
WATER TABLE'
, SAMPLING
' DEPTH
n_
.
-
'
. -
' ,
-
0 100 200 FEET
^ SAMPLING | 1 1
DEPrH 0 30 60 METERS
VERTICAL EXAGGERATION X *
DATUM IS NATIONAL GEODETIC
VERTICAL DATUM OF 1929
Figure 4.—Generalized cross sections A-A' and B-B' from Figure 3.
118
-------
D.W.HALL&D.W.RfSSER
groundwater-flow path was located upgradient from
each well, taking into account the water table con-
figuration. Flow lines, which are perpendicular to the
water table gradient, were then expanded into
wedge-shaped contributing areas by using an ar-
bitrary 2:1 (roughly 25°) ratio of longitudinal flow
distance to distance of lateral spreading. True flow-
to-dispersion ratios of this aquifer are unknown.
Some dispersion undoubtedly occurs as solutes in-
filtrate vertically through the unsaturated zone, and
additional dispersion (anisotropic) occurs during
flow through the aquifer. Therefore, applications of
manure and commercial fertilizer made upgradient
and close to the well could have contributed more
nitrogen to well water than if the applications were
made farther upgradient. Contributing areas were
limited to approximately 1,000 ft upgradient from
each well instead of to the entire area contributing
recharge water to the well, thereby defining a land
area where nitrogen applications had a maximum ef-
fect on groundwater quality.
Agricultural Activities
Agricultural activity data, including times of planting
and harvest, tillage information, and areas and times
of applications of manure and commercial fertilizer,
were collected monthly from the farmer. Ap-
proximately 100 steers, 1,500 swine (three groups —
500 at a time), and 110,000 chickens (five groups —
22,000 at a time) are raised at the site each year.
Crops were fertilized using both manure and com-
mercial fertilizer. The manure was of three types:
hog manure from gestation and finishing operations,
steer manure and bedding mixture from a feedlot,
and poultry manure from a poultry house. The steer
manure mix, and the poultry manure were applied by
surface spreading. The gestating and finishing hog
manure was injected into the soil 8 to 10 inches
below the surface, unless the soil was frozen. In the
case of frozen soil, all manures were surface spread.
The commercial fertilizers applied were ammonium
sulfate, broadcast before planting, and a nitrogen liq-
uid coapplied with preemergent pesticides. Before
implementing nutrient management at the site, all
animal wastes were applied to the 47.5 cropped
acres, which supplied about twice the amount of
manure-nitrogen needed for crop fertilization (Table
:2).
A nutrient-management plan (Graves, 1986a,b)
was implemented at the site in October 1986. Under
nutrient management, quantities of nitrogen applied
76°IO'53*
40*H'49'
LN 1674
76°11'15
Q 100 200 FEET
0 30 60 METERS
DATUM IS NATIONAL GEODETIC VERTICAL DATUM OF 1929
EXPLANATION
FARM STRUCTURE
__ __ _A_) CONTRIBUTING DRAINAGE AREA FOR WELL LN 1677
B_l CONTRIBUTING DRAINAGE AREA FOR WELL LN 1676
C~) CONTRIBUTING. DR'AINAGE AREA FOR WELL LN 1673
D ) CONTRIBUTING DRAINAGE AREA FOR WELL LN 1669
— —»Ł/ CONTRIBUTING DRAINAGE AREA FOR WELL LN 1679
LN SP61 ^'SPRING WITH D NUMBER
LN 1674 0 CHARACTERIZATION WELL WITH ID NUMBER
LN 1669 • MONITORING WELL WITH O NUMBER
Figure 5.—Estimated contributing areas for five wells at the field site.
119
-------
Proceedings of national RCWP Symposium, 1992
Table 2.—Nitrogen in manure and commercial fertilizer applied in the contributing areas of five wells (in pounds
of nitrogen per acre per year).
PRE-OR
POSTNUTRIENT
MANAGEMENT
Pre-
Pre-
Post-
Post-
Post-
Post-
WATER
YEAR
1985
1986
1987
1988
1989
1990
LN 16691
580
390
160
130
230
120
LN 16732
670
490
200
160
360
380
WELL
LN 16763
570
440
180
150
150
180
LN 16774
550
470
290
310
310
340
LN 16796
830
430
290
290
270
160
'Contributing area for LN 1669 = 5.6 acres.
^Contributing area for LN 1673 = 5.6 acres.
'Contributing area for LN 1676 = 5.5 acres.
'Contributing area for LN 1677 = 3.4 acres.
to cropped land are determined by crop-nutrient
needs rather than manure-disposal needs. Nutrient-
management plans for the site were based on crop
yield goals, manure application methods, soil
nitrogen concentrations, nitrogen analyses of ma-
nure samples collected at the site, and past manure-
application practices. Nutrient-management imple-
mentation resulted in the export of about 33 percent
of the animal manure generated at the site. The
reductions in nitrogen applications that occurred
from water years 1986 to 1987 resulted from the im-
plementation of nutrient-management planning
(Table 2). In water years 1988 and 1989, for example,
under nutrient management the farmer was per-
mitted to apply either 5 tons of poultry manure per
acre, supplying approximately 280 pounds of
nitrogen per acre; 9,000 gallons of liquid hog
manure, supplying approximately 410 pounds of
nitrogen per acre; or 20 tons of steer manure, supply-
ing approximately 340 pounds of nitrogen per acre.
Planting and harvesting were scheduled accord-
ing to growth requirements of the crops and weather
Table 3.—Annual crop acreage at Field Site 2.
YEAR
1985
1986
1987
1988
1989
1990
GROWING
SEASON
Summer
Summer
Winter (1984-85)
Summer
Summer
Winter (1985-86)
Summer
Summer
Winter (1986-87)
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Summer
Summer
CROP TYPE
Corn
Tobacco
Rye
Corn
Tobacco
Rye
Corn
Tobacco
Sudan grass
Corn
Tobacco
Fruit and vegetables
Corn
Tobacco
Fruit and vegetables
Corn
Tobacco
Fruit and vegetables
ACREAGE
43.5
4.0
22.5
43.5
4.0
25.0
42.0
5.5
5.5
39.5
5.0
2.5
35.0
2.5
10.0
37.0
4.0
6.5
conditions. Corn, the primary crop, was usually
planted during the last two weeks in April and har-
vested from mid- to late-September. Tobacco, which
requires a shorter, warmer season, was transplanted
from starting beds to the field in mid-June and har-
vested in mid- to late-August. During 1985 and 1986,
a winter cover crop of rye was broadcast seeded
after corn harvesting and covered primarily the pipe-
drained terraces. The rye was not harvested but was
sprayed with herbicide before planting corn. In 1987,
a winter cover crop of sudan grass was planted on 5.5
acres of the site. Beginning in 1988, fruit and
vegetables, in addition to sweet corn, were cul-
tivated. In 1989 and 1990, more acreage was planted
in fruit and vegetables than in tobacco. Table 3 lists
crop acreage for the years 1985 through 1990.
Concentrations of Nitrate in
Groundwater
Groundwater samples, collected monthly near the
water table, were analyzed (Chichester, 1988) to
determine long-term changes in nitrate concentra-
tion caused by implementing nutrient management.
Nitrate concentrations in groundwater appeared to
be unrelated to either depth to the water table or
sampling depth below the water table for four wells.
An exception was well LN 1669, which was sampled
approximately 50 ft deeper below the water table
than the other wells (Table 4). Nitrate concentra-
tions were less in water samples from LN 1669 than
in water samples from the other wells, and water
from this well may have represented groundwater
quality of a deeper zone in the aquifer than the water
from other wells (Fig. 6c).
Neither changes in quantity or timing of
precipitation nor changes in other climatic factors
within the variation that occurred during the study
period had a major effect on nitrate concentrations in
groundwater (Fig. 6a to c). Although the wells at the
120
-------
D.W. HALL 6 D.W. RfSSER
Table 4.—Average depth to water table, sampling depth, lag time, significance of Wilcoxon-Mann-Whitney test,
number of samples, prenutrient management and postnutrient management nitrate concentrations, percent
change in median nitrate concentrations, and percent change in nitrogen applications to contributing areas at
five wells (mg/L = milligrams per liter).
WELL
LN 1669
LN 1973
LN 1676
LN 1677
LN 1679
AVG. DEPTH
TO
WATER
(IN FT)
FROM LAND
SURFACE
18
10
24
30
20
AVG.
SAMPLING
DEPTH
(IN FT)
BELOW
WATER
TABLE
SURFACE
66
23
10
3
13
WILCOXON-
MANN-WHITNEY
TEST,
SIGNIFICANT
TOPRE-TO
POST-
INCREASE (+)
OR
DECREASE (-)?
Yes (+)
Yes (-)
Yes (-)
Yes (-)
Yes (-)
# SAMPLES
(PRE-; POST-
NUTRIENT
MANAGE-
MENT)
15; 35
15; 35
11; 25
10; 33
5; 35
PRENUTRIENT
MANAGEMENT
MEDIAN
NITRATE CON-
CENTRATION
(mg/L)
11
53
82
26
24
POSTNUTRIENT
MANAGEMENT
MEDIAN
NITRATE CON-
CENTRATION
(mg/L)
12
37
56
23
22
CHANGE
FROMPRE-TO
POST- PERIOD
IN MEDIAN
NITRATE CON-
CENTRATION
(%)
+8
-30
-32
-12
-8
CHANGE
FROMPRE-TO
POST- PERIOD
IN APPLICA-
TIONS (%)
-67
-53
-67
-39
-60
site are a maximum of only 1,150 ft apart, and
climatic influences (including precipitation) are vir-
tually identical across this site, nitrate concentra-
tions in groundwater samples from the five wells
demonstrated no climate-induced long- or short
term trend or any consistent seasonal variation. For
example, a period of increased recharge occurred
during the late spring and summer of 1989 (Fig. 6a
and b). During this period, nitrate concentrations in
groundwater samples collected at wells LN 1673 and
LN 1676 were increasing, concentrations in samples
collected at wells LN 1677 and LN 1679 were
decreasing, and concentrations in samples collected
at well LN 1669 were remaining relatively constant
(Fig. 6c). Most variations in nitrate concentrations in
groundwater at each well are explained by amounts
and timing of nitrogen applied in the contributing
area of each well.
Three techniques were used to investigate the
effects of nitrogen applications on water quality.
First, median nitrate concentrations in groundwater
from the five monitoring wells (before and after im-
plementation of nutrient management) were com-
pared. Second, changes in amounts of nitrogen
fertilizer applied to the five well-contributing areas
were statistically correlated with changes in nitrate
concentrations of groundwater. Third, estimates of
the quantities of nitrogen added to and removed
from the site for each year of the study were com-
pared to determine whether nitrogen loading to
groundwater decreased as the result of decreased
applications of nitrogen under nutrient management.
Comparison of Prenutrient and
Postnutrient Management Nitrate
Concentrations in Groundwater
Median nitrate concentrations in groundwater
samples collected during the pre- and postnutrient-
management periods were tested for significant dif-
ferences by use of the Wilcoxon-Mann-Whitney test
(Iman and Conover, 1983). The results of this test
(Table 4) indicate that statistically significant
decreases (at a = 0.05) occurred in nitrate concentra-
tions at four of the five wells. The largest post-
nutrient-management reductions in median nitrate
concentration occurred at wells in which pre-
nutrient-management groundwater contained the
largest nitrate concentrations. The median nitrate
concentration at well LN 1669 increased 1 mg/L
after the implementation of nutrient management.
Test results indicate that nutrient management
reduced nitrate concentrations in groundwater at
four of five monitoring wells.
Correlations of Changes in Nitrogen
Applications to Changes of Nitrate
Concentrations in Groundwater
Cross-correlation functions (Wilkinson, 1987) were
used to select the most significant lag times for
lagged correlations between nitrogen applied to land
areas contributing to the wells and nitrate concentra-
tions in groundwater. For example, the series of
monthly nitrogen applications estimated from an ap-
plication curve on Figure 7 can be correlated with
the series of monthly nitrate concentrations es-
timated from the corresponding concentration curve
with lag times of 1 month, 2 months, and so on up to
20 months. When several adjacent lag times were
significant — for example, lag times for well LN
1676—lags from 2 to 9 months were statistically sig-
nificant— the lag time with the strongest correlation
was reported. Rho and p on Figure 8 are reported
from Spearman-ranked correlations (P-STAT, 1986)
by using the most significant correlation indicated in
the cross-correlation procedure for each well data
set.
121
-------
Proceed/ngs of national RCWP Symposium, 1992
LN 1673 ,«.,
r-\ ..A /""V\
..-•••<• v...-*.. ^.. ...-••**«•**•
POSTNUTRIENT "*
, MANAGEMENT , ,
PRENUTRIENT
, MANAGEMENT
JAN
1985
JAN
1986
JAN JAN JAN JAN
1987 * 1988 1989 1990
Figure 6.—Monthly precipitation (A), depth to water from land surface (B), and nitrate concentrations (C) in
groundwater samples.
—-
-------
DM HALL&D.W, RISSER
01
O
O
CE
CO
<
CO
Q
z
O
D_
LU
CE
<
O
CD
DC
O
O
LU
Q
UJ
CL
Q.
Z
LU
O
O
CE
3,000
2.000
1.000
0
3.000
2.000
1.000;
3.000
2.000
1,000
3.0o8
2.000
1.000
0
3.000
NITROGEN .
APPLICATIONS
O \.
LN 1669 LAG 19 MONTHS
NITRATE CONCENTRATIONS
* $*? NITROGEN
5BHD-OGO APPLICATIONS
LN 1673 LAG 16 MONTHS
NITRATE CONCENTRATIONS ,
>^9 ,
LN 1676 LAG 9 MONTHS
NITRATE CONCENTRATIONS
o-o.
NfTROGEN APPLICATIONS
o ^
Q _r' &.
LN'1677 LAG 18 MONTHS
MITRATE CONCENTRATIONS '
NITROGEN APPLICATIONS
LN 1679 LAG 4 MONTHS
NITRATE CONCENTRATIONS
TIME. IN MONTHS.
20
15
10
5
60
40
20
0
100
80
60
40
20
0
40
30
20
tio
0
40
NITROGEN APPLICATIONS
30
20
10
0
01
O
O
tr
CO
<
cr
01
CE
01
a.
CO
CE
O
of
LU
a
z
O
CE
O
Z
<
CE
LU
O
Z
O
O
LU
CE
Figure 7.—Groundwater nitrate concentration data and applied nitrogen data for wells LN1669, LN 1673, LN 1676, LN
1677, and LN 1679. Note: To Illustrate correlations, nitrogen applications have been moved to the right to match resul-
tant nitrate concentrations; to get real-time relations, move the application curves to the left by the indicated time lag
for each well.
123
-------
Proceedings of National RCWP Symposium, 1992
-
fn§ 1l5°°
OP-
pf**^
|Sg 1000
3S
!!jS
55 500
^CL
2
°c
A °
LNT669
p = .002 «
o
o
o^
o
•ft
^^o
Op o o
o ° Jl
* 88 •
) 5 10" "15 2
1,600
.S1-400
§§1200
O Pj . «nQ
ŁE^- lUUU
j?^ 800
3Q 600
5l5
p = .001 o
0
0
0
8 o
Jo*
o
rfM
O O
o
. _
3 10 20 30 40 50 60 70 80
NITRATE CONCENTRATION 19:MONTHS
AFTER ESTIMATED APPLICATION
NITRATE CONCENTRATION 16 MONTHS
AFTER ESTIMATED APPLICATION
l.'HJV
1200
1>000
800
600
400
200
C j»
LN 1676 « *
. P = .001
o
o
S 0 °
O7O o
o
/* fr o
0 g "
r o ° o
° 0
.» °- °
0 20 40 60; 80 10
f(JU
600
500
400
300
200
100
I
D
LN1677 0° *•
rho = .496 °
p = .001 «| °°
080^
o
0 0
o
O O^B0 O
O
o
8
) 5 10 15 2(3 25 30 35 4(
NITRATE CONCENTRATION 9 MONTHS
AFTER ESTIMATED APPUCATTON
NITRATE CONCENTRATION 18 MONTHS
AFTER ESTMATED APPLICATION
yuu
800
700
600
500
400
300
200
100
°C
_
L LN1679 o°
p = .067
000
o o
§0
O o 00
a o e
§ Oo«8 o0
oo o
o o |o
0 0
O Og O
o
» . -. ** .
). 5 10 15 2ff 25 3<
NITRATE CONCENTRATION 4 MONTHS
AFTER ESTMATED APPLICATION
Figure 8.—Plots of groundwater nitrate concentrations and applied nitrogen and results of Spearman rank correlation
testing using data from wells LN1669, LN 1673, LN 1676, LN 1677, and LN 1679.
124
-------
D.W. HALL&D.W. RISSER
Figures 7 and 8 show the effects of changes in
loads of applied nitrogen on changes in nitrate con-
centrations in groundwater at the five monitoring
wells. Each point on the nitrogen-application graphs
of Figure 7 represents the sum of 4 months of ap-
plications made to the contributing area of each well
The points are plotted on the day of the maximum
application in each 4-month period and are, there-
fore, unequally spaced.
Statistically significant correlations (Spearman
rank correlations, a = 0.10 [P-STAT, 1986]) exist be-
tween the timing and amount of applied nitrogen and
changes in groundwater nitrate sample concentra-
tions adjusted for lag time (Fig. 8). The statistically
significant, cause-effect relations between changes
in the amounts of applied nitrogen and changes in
groundwater nitrate concentrations indicate that a
significant amount of the nitrogen in manure and fer-
tilizer becomes available for leaching and is
transported with recharge to the groundwater sys-
tem within 4 to 19 months following application.
Therefore, any substantial reduction or increase in
quantities of applied nitrogen would be expected to
produce a similar reduction or increase of nitrate
concentrations in groundwater samples collected at
the five monitoring wells.
Lag times required for applied nitrogen to leach
to groundwater at each of the wells range from about
4 months upgradient of well LN 1679 to about 19
months upgradient of well LN 1669, probably be-
cause of differences in the hydraulic conductivity of
the unsaturated zone, soil-moisture content, clay
contents of the-unsaturated zone, and the depth to
the water table upgradient of each well.
Annual Additions and
Removals of Nitrogen
Additions and removals of nitrogen at the site were
calculated for each year of the study (Table 5).
Potential errors caused by assumptions made in the
calculation of loads may greatly affect the numbers
reported in Table 5.
Loads of nitrogen in manure were calculated by
using application data supplied by the farmer and
laboratory analysis of manure samples collected at
the site. Loads of nitrogen entering the site in
precipitation were calculated by using rainfall
volumes estimated on the basis of precipitation data
collected at the site and estimated concentrations of
nitrogen in precipitation samples for the Ephrata,
Pennsylvania, area (Lynch, 1990).
Amounts of commercial fertilizer applied to the
site were reported by the farmer. In spring 1989,
Pennslvania State University announced that the
"quick" nitrogen soil test would be made available on
a limited basis to the public. This test measures the
amount of soluble nitrate-nitrogen available in the
soil and was designed to be used at side-dress time
(early June) to determine if additional nitrogen was
necessary for optimum corn yield. Pennsylvania
State recommended that the test be used on fields
that had recently received manure or following a
legume crop. Recommended nitrogen side-dressing
rates were based on soil nitrate test results and corn
yield goals. For example, a soil test level of 20 ppm of
nitrate nitrogen would result in a recommended
side-dressing rate of 25 pounds/acre for a yield goal
of 100 bushels/acre, but the recommended rate for a
yield goal of 200 bushels/acre should be 100
pounds/acre.
The owner of Field Site 2 requested that the
"quick" nitrogen soil test be used on his farm. The
test showed that some fields had nitrate levels below
25 ppm. These fields were side dressed with urea-
ammonium-nitrate on June 26, 1989, at the recom-
mended rate (either 50 or 75 pounds of actual
nitrogen per acre). Approximately 29 acres of the
field site planted in field corn had additional nitrogen
applied, which accounts for the use of commercial
fertilizer during the nutrient management period
(Table 5).
Quantities of nitrogen removed from the site in
crops were calculated from estimated crop yields,
crop nitrogen-content information (Anderson, 1989)
and discussions presented in Knott (1962). Loads of
nitrogen in surface water were calculated by using
discharge-weighted mean-storm concentrations for
each storm multiplied by gaged discharge. Volatiliza-
tion of nitrogen was estimated as a percentage of the
nitrogen applied: 40 percent for surface-spread
manures and 20 percent for injected manures (on
the basis of information in Graves [1986b]), and a 15
percent value for commercial fertilizers (Pionke and
Urban, 1985). Denitrification losses from soil and
groundwater are unknown; however, large con-
centrations of dissolved oxygen in groundwater may
indicate that losses are small because denitrification
is an anaerobic process. For example, water from
wells LN 1669, LN 1673, LN 1676, LN 1677, and LN
1679 was sampled for dissolved oxygen concentra-
tion during May 1986; dissolved oxygen concentra-
tions ranged from 5.7 to 10.5 mg/L
Approximately 99 percent of the nitrogen leav-
ing the site in water was discharged through the
groundwater system. More than 90 percent of this
nitrogen was in the form of dissolved nitrate.
A groundwater-flow model of the hillslope on
which the site is situated was constructed to help es-
125
-------
Proceedings of National RCWP Symposium, 1992
.
*c"
&
s
"c
CO
CO
CO
c
3
O
a.
=
Ł
M
CO
CD
E
J.
§
O)
s
c
0
CO
1
Ł
•o
CO
co
•a
CO
OJ
"5
E
1
U)
_
ul
K
S
<9
o
|
a
1
i=
!2
Ł
nONSTOSI
5
U
a
§
C3
O
MATED NlTRi
p
ffi
SILZ
pi
131
cc
(3
z
S
S
J
g
5
§
or
5
c .. _
JRFACEYWV
LOADS OF
NITROGEN
_«
CO
z IT) co
UJ Bj OL
o 2 cc
§o
CD
EC
sp1
zg>
sil
2f u -J
O z u.
t 0S
Z E
O
ITROGEN
IN
CIPITATION
2 Ul
Ł
IITROGEN IN
OMMERICAl
FERTILIZER
z O
z tf
2 UJ UJ
UJ CC N
oSP
f iS
z ff
cc
Ul CC
o
o
in
8
CO
h»
S
8
in
CO
g
o
0
o
o
§
CJ
W\l
"s
o>
o
o
CO
o
T-
o
o
CO
1-
s
8
f^.
CO
o
CO
8
CO
o
s
8
u>
co
-g
o
o
in
o
o
o
Tf
s
s
o>
CO
i
§
CO
o
8
•t-
i
80 o o co
pom CO
t~ i** ^" Oi
K t-T CD N
0 O O O «t
O O O CO CM
f- in co i-
"*• in M- in
g § g o ^
8 88 8 §
in *r T- t^
ttf |C CO |C
o o o o -i-
h- O) CO (O
^ '"
O O O O CM
CM CO •* CO
CO CO CO CO
0 8 g g *
CO Is- CO
888 S S
in CM N. -ri
co" rC co rC
o
§J ||f
CO O> O 2O> C'Dm
cocao O'- CD (0 S:
o
s
• —
1
a
"o
as
2
-p
§
c
ra
E'
01
c
^ '
CO
^
8
s- .
'•o co
0)0?
is
JS O)
|e
J aj.
^Ł
II
-ll
•e 2 °
C Ł (0
E 2,^
o>1a~
ssl
Pi
|ll
Q.-— °
0) P? D)
ill
CO 03 (tf
2§8
111
ssl
s^a
126
-------
D.W.HALL&D.W.RJSSER
timate the volumes of groundwater inflow and out-
flow across site boundaries needed for groundwater
nitrogen load calculations. The hillslope was simu-
lated as a two-dimensional, steady-state flow system
in the x-y plane by using the finite-difference model
of McDonald and Harbaugh (1988). Figure 9 shows
the finite-difference grid, aquifer properties, and
boundary conditions used in the model. Transmis-
sivities were based on the site water table gradients
(Fig. 3) in conjunction with Darcy's Law (Driscoll,
1986).
The water budget computed by the model
(Table 6) is based on the long-term average
recharge from precipitation of 19.1 in/yr (inches per
year), estimated to be 44 percent of the long-term
precipitation of 43.5 in/yr recorded at the Ephrata,
Pennsylvania, NOAA station (Natl. Oceanic Atmos.
Admin. 1982). Groundwater inflow during 1985-90
was estimated to be 16 percent of total inflow of
water to the site as indicated by modeling simula-
tions.
Nitrogen in groundwater inflow was calculated
from volumes of flow estimated by groundwater flow-
model simulations multiplied by the average
nitrogen concentrations in two water samples col-
lected at well LN 1674, located on the western
boundary of the site.
The quantity of groundwater discharged annual-
ly across each site boundary was computed by multi-
plying the total annual discharge from the site by the
percentage of annual flow estimated to cross each
site boundary as indicated by output from the
groundwater model. Annual discharge across each
site boundary was then apportioned among months
by using water level hydrograph rises for each
month divided by the total annual water level rise to
obtain .the fraction of annual discharge in each
month. .
Having thus obtained quantities of groundwater
discharge crossing each site boundary during each
month of the study, monthly water samples from
wells located in different parts of the site were
chosen to characterize the nitrogen concentrations
of the groundwater discharges from each part of the
site. Water samples from well LN 1677 were chosen
to characterize the nitrogen concentrations of water
discharged across the northern site boundary, water
samples from wells LN 1676 and LN 1679 were
chosen to characterize the nitrogen concentrations
of water discharged across the eastern site bound-
ary, and water samples from wells LN 1673 and LN
1669 were chosen to characterize the nitrogen con-
centration of water discharged across the southern
site boundary. Estimated groundwater discharges
across site boundaries were multiplied by nitrate
concentrations in groundwater samples from the
northern, eastern, and southern areas to obtain
loads of nitrogen in groundwater discharge.
The estimated loads in groundwater averaged
7,950 Ib of nitrogen per year in the 1985-86
prenutrient-management period and 7,925 Ib per
year in the postnutrient-management period. Be-
cause loads were calculated by multiplying ground-
water discharge by groundwater nitrate concen-
trations, average annual loads appear to be un-
changed because annual groundwater discharge
averaged about 6 inches more during the
postnutrient-management period than in the
prenutrient-management period. However, annual
nitrogen loads normalized for discharge (Table 7) in-
Table 6.—Simulated steady-state groundwater budget for the site.
CUBIC FEET PER SECOND
INCHES PER YEAR
PERCENT OF TOTAL INFLOW
Groundwater Inflow
Recharge from precipitation
Flow across western boundary
Total
Groundwater Outflow
Flow across eastern boundary
Flow across northern boundary
Flow across southern boundary
Total
0.119
0.023
0.142
0.023
0.033
0.084
0.140
19.1
3.7
22.8
3.7
5.4
13.7
22.8
84
16
100
16
24
60
100
Table 7.—Loads of nitrogen per inch and per million gallons of groundwater discharge, 1985 to 1990 (loads of
nitrogen are in pounds).
YEAR
PERIOD
NITROGEN LOAD IN GROUNDWATER
(IN LBS. PER INCH OF GROUNDWATER
DISCHARGE)
NITROGEN LOAD IN GROUNDWATER
(IN LBS. OF NITROGEN PER MILLION
GALLONS OF GROUNDWATER DISCHARGE)
1985
1986
1987
1988
1989
1990
Prenutrient management
Prenutrient management
Transition year
Postnutrient management
Postnutrient management '
Postnutrient management
421
450
408
306
298
304
294
314
285
214
208
212
127
-------
Proceedings of Hatlonal RCWPSymposium, 19'92
EXPLANATION
1,000 FEET
1
100 200 300 METERS
g| CONSTANT HEAD BOUNDARY
1,000 TRANSMISSIVITY, IN FEET SQUARED PER DAY
H, NO-FLOW BOUNDARY
.' TRANSMISSIVITY BOUNDARY
Figure 9.—Finite-difference grid, hydrologic boundaries, and aquifier properties used in the groundwater flow model.
128
-------
D.W. HALL &D.W. RISSER
dicate that nitrogen discharge in groundwater
decreased. Reduced nitrogen applications to the land
surface under nutrient management caused the
decrease in nitrogen that discharged in ground-
water.
The reduced loads of nitrogen in groundwater
that occurred as the result of nutrient management
are further illustrated in Figure 10. For example, a
monthly groundwater discharge of 4 in contained
about 1,750 Ib of nitrogen in the prenutrient-manage-
ment period, but only about 1,200 Ib of nitrogen
during the postnutrient-management period. Dilu-
tion was not responsible for the load reductions. No
consistent relation was found between nitrate con-
centrations in groundwater and the volumes of
recharge. Nitrate concentrations in groundwater in-
creased during some recharge events and decreased
during others at this site (Hall, 1992) and at a similar
site in Lancaster County, Pennsylvania (Gerhart,
1986).
By entering the monthly groundwater discharge
(MGWD) values from the postnutrient-management
period into regression equation (1) in Figure 10,
loads of nitrogen that would have been discharged
from the site if nutrient management had not been
implemented can be estimated for the period 1988-
90. Subtracting the actual 1988-90 loads from the es-
timated 1988-90 loads, it was determined that a
reduction of about 11,000 Ib of nitrogen in ground-
water discharge was achieved as the result of
nutrient management during the period 1988-90.
Conclusions
The implementation of a nutrient management plan
in 1987 was effective in reducing concentrations of
nitrate in groundwater at a 55-acre farm site near
Ephrata, Pennsylvania. Under nutrient management,
applications of nitrogen to the site decreased ap-
proximately 33 percent. Statistically significant
decreases in median nitrate concentrations in
groundwater ranged from 8 to 32 percent at four out
of five monitoring wells. The largest decreases in
nitrate concentration occurred at wells in which
prenutrient-management nitrate concentrations in
groundwater were the largest
Correlations between changes in loading of ap-
plied nitrogen fertilizers (primarily manure) and
changes in nitrate concentrations in groundwater at
five wells were statistically significant. Changes in
nitrate concentrations in groundwater occurred
about 4 to 19 months after changes in applications of
nitrogen fertilizer.
Additions and removals of nitrogen loads from
the site were estimated for the period 1985-90.
Nitrogen entered the site in manure, commercial fer-
tilizers, precipitation, and groundwater inflow; these
sources averaged 93, 4, 2, and 1 percent of the
nitrogen that entered the site, respectively. Nitrogen
was lost from the site in harvested crops,
groundwater outflow, volatilization, and surface
runoff; these losses averaged 38, 38, 24, and less
4.000
3.500
3.000
2.500 -
2,000
1.500 -
1.000 -
500 -
Transition year
(1987)
•-Post-nutrient oanagement
(1988-90)
PERIOD
REGRESSION EQUATION
PRE-NUTRIENT MONTHLY NITROGEN « (457.2) (MONTHLY WATER) - 27.4 0.99 0 001
MANAGEMENT
POST-NUTRIENT MONTHLY NITROGEN =. (303.1) (MONTHLY HATER) - 7.1 0 99 0 001
MANAGEMENT
I I t I
2468
MONTHLY GROUND-WATER DISCHARGE, IN INCHES
10
12
Figure 10.—Relation between monthly groundwater discharge and nitrogen discharge.
-------
Proceedings of National RCWP Symposium, 1992
than 1 percent of the nitrogen removed from the site,
respectively.
Virtually all nitrogen leaving the site that was
not removed in harvested crops or lost to the atmos-
phere by volatilization discharged with the
groundwater. Nutrient management substantially
reduced nitrogen loads in groundwater discharge.
Nitrogen loads in groundwater discharge were es-
timated to have been approximately 300 and 200
pounds of nitrogen per million gallons of discharge,
respectively, before and after implementation of
nutrient management.
References
Anderson, R. 1989. Written communication. Penn. State Univ.
Coop. Exten. Serv., University Park
Bouwcr, H. 1978. Groundwater Hydrology. McGraw-Hill, New
York.
Chtchester, D.C. 1988. Evaluation of Agricultural Best Manage-
ment Practices in the Conestoga River Headwaters: Methods
of Data Collection and Analysis and Description of Study
Areas. Open File Rep. 88-96. U.S. Geo. Surv., Lemoyne, PA
Driscoll, EG. 1986. Groundwater and Wells. Johnson Filtration
Systems, St Paul, MN.
Fox, R.H. and W.P. Piekielek 1983. Response of corn to nitrogen
fertilizer and the prediction of soil nitrogen availability with
chemical tests in Pennsylvania. Bull. 842. College Agric.,
Penn. State Univ., University Park.
Gcrhart, J.M. 1986. Groundwater recharge and its effects on
nitrate concentrations beneath a manured field site in Penn-
sylvania. Ground Water vol/issue?: 483-89.
Graves, R.E., ed. 1986a. Manure management for environmental
protection. Doc. MM1-5. Penn. Dep. Environ. Resoun, Har-
risburg.
. 1986b. Field application of manure. Penn. Dep. Environ.
Resoun, Harrisburg.
Hall, D.W. 1992. Effects of nutrient management on nitrate levels
in groundwater near Ephrata, Pennsylvania. J. Ground Water
30(5):720-30.
Hall, G.M. 1934. Groundwater in southeastern Pennsylvania. Bull.
W-2. Penn. Geo. Surv., Harrisburg.
Hickey, J.J. 1984. Field testing the hypothesis of Darcian flow
through a carbonate aquifer. Ground Water 22(5):544-7.
Iman, R.L. and W.T. Conover. 1983. A Modern Approach to Statis-
tics. John Wiley and Sons, New York.
Knott, J.E. 1962. Handbook for Vegetable Growers. John Wiley
and Sons, New York.
Lynch, J.A 1990. Written communication. Penn. State Univ.,
University Park.
McDonald, M.G. and AW. Harbaugh. 1988. A modular three-
dimensional finite-difference groundwater flow model. Book
6, Ch. Al in Techniques of Water-Resources Investigations.
U.S. Geo. Surv., Washington, DC.
Meisler, H. and AE. Becher. 1966. Hydrology of the carbonate
rocks of the Lancaster 15-minute quadrangle. Progr. Rep.
171,4th sen Penn. Geo. Surv., Harrisburg.
. 1971. Hydrology of the carbonate rocks of the Lancaster
15-minutequadrangle, Southeastern Pennsylvania. Bull. W-
26. Penn. Geo. Surv., Harrisburg.
Memon, B A and E. Prohic. 1989. Movement of contaminants in
karstified carbonate rocks. Environ. Geo. Water Sci. 13(1): 3-
13.
Morrissey, D J. 1987. Estimation of the Recharge Area Contribut-
ing Water to a Pumped Well in a Glacial-drift, River-valley
Aquifer. Open File Rep. 86-543. U.S. Geo. Surv., Washington,
DC.
National Oceanic and Atmospheric Administration. 1982.
Climatological data, annual summary. Washington, DC.
P-STAT, Inc. 1986. P-STAT: Users Manual. Duxbury Press, Bos-
ton.
Pionke, H.B. and J.B. Urban. 1985. Effect of agricultural land use
on groundwater quality in a small Pennsylvania watershed.
Ground Water 23 (1): 68-80.
Poth, C.W. 1977. Summary Groundwater Resources of Lancaster
County, Pennsylvania. Water Resour. Rep. 43. Penn. Geo.
Surv., Harrisburg.
Schueller, J.P. 1983. An Assessment of Agricultural Nonpoint
Source Pollution in Selected High Priority Watersheds in
Pennsylvania. Bur. Soil Water Conserv., Penn. Dep. Environ.
Resour., Harrisburg.
U.S. Department of Agriculture. 1985. Soil Survey of Lancaster
County, Pennsylvania. Washington, DC.
Wilkinson, L 1987. SYSTAT — The System for Statistics.
Evanston, IL. SYSTAT, Inc., Evanston, IL.
130
-------
Relating Land Use and Water Quality
in the St. Albans Bay Watershed,
Vermont
Donald W. Meals
School of Hatural Resources
University of Vermont
Burlington, Vermont
ABSTRACT
The St. Albans Bay Rural Clean Water Program comprehensive monitoring and evaluation program
assessed water quality changes in response to land treatment. Evaluation of response to treatment
must be based not only on detection of water quality trends but also on linking such trends to ob-
served changes in land use or management. To this end, the St. Albans Bay watershed monitoring
and evaluation program included intensive tracking of annual land use and agricultural manage-
ment activities and related them to observed water quality using correlation and linear regression.
Major variables considered included animal management, cropping patterns, and manure manage-
ment. Bacteria counts in streams increased with higher animal density but declined with increasing
percentage of animals under BMP waste management, while lower bacteria counts occurred with
greater use of manure from storage. Although data from another watershed had previously sug-
gested that lower stream nutrient levels would be associated with higher proportions of watershed
animals under BMP, no such relationship was observed in the St. Albans Bay watershed, perhaps
because a sufficiently high level of treatment was not achieved. Future efforts to document land
use/water quality relationships should include improved experimental design, better tracking of
changing land use and agricultural management, and more powerful statistical analysis.
St. Albans Bay on Lake Champlain in Vermont
has been subject to increasing rates of
eutrophication as a result of excessive phos-
phorus loads from both point and nonpoint sources.
The primary goal of the St. Albans Bay Rural Clean
Water Program (RCWP) project was to improve
water quality and restore beneficial uses in St. Al-
bans Bay and its tributaries through a program of
agricultural best management practices (BMPs) to
control nonpoint sources of pollution.
Initiated in 1980 and completed in 1991, the St.
Albans Bay RCWP project was one of only five
projects nationwide to include comprehensive
monitoring and evaluation to evaluate the water
quality impacts of land treatment. The goal of this
program was to assess water quality changes in the
bay and its tributaries in response to land treatment.
The results of the program have been presented in
detail by the Vermont RCWP Coordinating Commit-
tee (1991) and Meals (1992).
Evaluation of water quality response to land
treatment must be based not only on detection of
water quality trends but also on linking such trends
to observed changes in land use or management. To
this end, the St. Albans Bay watershed comprehen-
sive monitoring and evaluation program included in-
tensive tracking of land use and agricultural
management activities. This paper discusses the ap-
131
-------
Proceedings of National RCWP Symposium, 1992
proach taken to relate land use to water quality and
presents some results of the analysis. Annual pat-
terns of agricultural activity were related to ob-
served water quality by using correlation and linear
regression to clarify the effects of treatment.
Study Area
The St Albans Bay watershed is located in
northwestern Vermont 40 km north of Burlington
(Fig. 1) and drains 13,000 ha of agricultural,
forested, and urban land into St. Albans Bay of Lake
Champlain. The 700-ha bay is 4.2 km long and 2.2
km wide and is oriented on a southwest-northeast
axis, opening to the southwest. Mean depth is 8 m;
maximum depth is 12 m.
Four major tributaries drain the watershed into
St Albans Bay: Jewett Brook, Stevens Brook, Rugg
Brook, and Mill River (Fig. 1). The city of St. Albans
tertiary wastewater treatment plant discharges
about 91 mVsec to Stevens Brook wetland at the
head of the bay.
The St. Albans Bay watershed is located primari-
ly in the Champlain Lowlands, an area of generally
low relief between Lake Champlain and the foothills
of the Green Mountains. Watershed soils formed on
glacial till or lacustrine deposits and include loams
(51 percent) (half of which are poorly drained), silts
and clays (27 percent), rock outcrop (15 percent),
and sands C7 percent).
The climate of the St Albans Bay watershed is
cool and humid, with pronounced seasonal varia-
tions. Mean annual temperature is 7.3°C; the grow-
ing season averages 150 days. Mean annual
precipitation is 845 mm, with a minimum average
monthly precipitation in February of 44 mm and a
maximum average monthly rainfall in August of 100
mm. Average annual snowfall is 1,550 mm.
Agriculture is the dominant land use in the
watershed (65 percent); about 20 percent of the
watershed is forested, 10 percent is in urban/
residential use, and 5 percent consists of roads and
surface water. Of the 102 farms in the watershed,
most are dairy operations, with a few fruit and
vegetable or horse farms. Dairy farms average 134
ha in size, with an average herd size in 1991 of 110
animal units, up from 95 in 1980. (One animal unit is
equivalent to a 405 kg dairy animal.) Corn for silage
is the principal cultivated crop: from 1980-90, land
area in corn ranged from 1,336 to 1,821 ha or about
10 to 15 percent of total watershed area. Substantial-
ly larger agricultural land area is devoted to hay,
from 3,845 to 4,452 ha or about 30 to 35 percent of
watershed area.
Some cropland in the watershed has been in
continuous corn, but a three-year corn/five-year hay
rotation is the prevalent practice. Both manure and
commercial fertilizer are commonly applied to corn
cropland; hayland generally receives manure twice a
year immediately following hay cuts, which typically
occur in early June and in mid-July.
Recreation, aesthetics, and other beneficial uses
of St. Albans Bay have been impaired by eutrophica-
tion for many years. The most serious nonpoint
source water quality problems associated with
agriculture in the St. Albans Bay watershed have
been nutrients (principally phosphorus) from runoff
of animal and milkhouse waste and sediment from
cropland erosion. Improper manure management
(e.g., year-round spreading because of inadequate
waste storage) and inadequate milkhouse waste and
barn- yard runoff management were believed to be
the principal contributors to water quality problems.
Improper fertilizer management — failure to balance
manure/fertilizer nutrient applications with soil/
crop needs — was also thought to be a significant
contributor to excessive nutrient loading. Cropland
soil and streambank erosion were also important
contributors to water quality problems.
Methods
Water Quality Monitoring
To document overall water quality changes in St. Al-
bans Bay and its tributaries, water quality monitor-
ing was conducted on several geographic scales:
• Level 1: bay sampling
• Level 2: tributary trend monitoring
• Level 3: edge-of-field BMP study
• Level 4: randomized grab sampling
The design and specific methods of the comprehen-
sive monitoring and evaluation program, including
bay monitoring, edge-of-field studies, and biological
monitoring, have been discussed in detail elsewhere
(Clausen, 1985; Vt. RCWP Coor. Comm. 1991; Meals,
1992). Discussions presented in this paper are based
exclusively on tributary monitoring data (Level 2).
Intensive tributary trend monitoring was
designed to determine long-term patterns of
streamflow as well as pollutant concentrations and
loads in the four major drainages to St. Albans Bay
and from the St. Albans city wastewater treatment
plant (Fig. 1). The subwatersheds monitored by
these trend stations comprised 72 percent of the en-
tire St. Albans Bay watershed. A fifth station
132
-------
D.W. MEALS
LEGEND
• level 1
A Ievel2
A. levels
A level 4
@ precipitation
T — project boundary
• 2 miles
SCALE
Figure 1.—Map of St. Albans Bay watershed showing location of monitoring stations.
monitored suspended solids and nutrient loads from
the wastewater treatment plant, the major point
source in the watershed.
At each of the five tributary stations, samples
were collected automatically at eight-hour intervals
using refrigerated ISCO Model 1680R samplers and
133
-------
Proceedings of National RCWP Symposium, 1992
combined weekly into two 48-hour and one 72-hour
composite sample(s) for analysis. During some
stormflow periods, individual samples were analyzed
discretely. All samples were analyzed for turbidity,
total and volatile suspended solids, total and soluble
reactive phosphorus, and total Kjeldahl, ammonia,
and nitrite+nitrate nitrogen by accepted methods
(U.S. Environ. Prot. Agency, 1983; Standard
Methods, 1985). Weekly grab samples were col-
lected at each station and analyzed for pH, conduc-
tance, and fecal coliform and fecal streptococcus
bacteria, while in situ measurements were made of
temperature and dissolved oxygen at the time of
grab sampling.
Stream stage was recorded continuously at each
station using ISCO Model 1870 bubbler-type stage
recorders and discharge was calculated from site-
specific stage-discharge ratings. Three standard Bel-
fort 20-cm weighing bucket rain gages were used to
measure watershed precipitation.
Land Use Monitoring
Basic land treatment implementation data were col-
lected and maintained by the U.S. Department of
Agriculture (USDA) Agricultural Stabilization and
Conservation Service, including number of par-
ticipating farms, critical acres treated, and number
or area of BMPs completed. However, far more
detailed land use information was needed to comple-
ment the water quality database. The comprehensive
monitoring and evaluation program therefore in-
cluded intensive tracking of land use and agricultural
management activities, which was conducted as a
cooperative venture among land treatment and water
quality monitoring agencies.
As a first step, baseline information was col-
lected on all farms (both contract and noncontract)
in the watershed from 1982 to 1983. These data in-
cluded:
• Livestock: type and number
• Manure management: type, capacity,
schedules
• Land use: location and crop type, by field
• Fertilizer: type, amount, fields, schedule
• Milkhouse: type, effluent disposal
• Barnyard: size, use, paving
• Pesticides: type, amount, fields, schedule
• Drainage: type, location.
For long-term land use monitoring, detailed in-
formation was collected from each watershed farmer
(regardless of contract status) on amount, date, and
location of manure and fertilizer applications and
cropland activity over each calendar year. The goal
was to identify not only what activities took place and
where but also when these activities, such as manure
spreading, occurred.
Information was recorded in a checkbook-style
logbook distributed to each farmer (Fig. 2). Each
logbook included an individual farm map derived
from the farm conservation plan or from aerial
photographs, with each field identified by number.
Farmers were asked to maintain records of activities
by field throughout the year, and this information
was collected during biannual (later annual) inter-
views. USDA Soil Conservation Service personnel
distributed logbooks and collected information from
contract farms, while comprehensive monitoring and
evaluation staff did the same for noncontract farms.
A. MANURE APPLICATION: Name:
DATE
Example
4/23/83
FIELD I.D.
(issrnap)
3b
AMOUNT APPLIED
(luH spreader loirj}
2ft
DATE
INCORPORATED
4/23
COMMENTS 1
•Evenly spread except wal spot on NE corner '
-planted corn 4/28
•elc.
B. OTHER FERTILIZER APPLICATION
(including commercial fertilizer, lime, sewage sludge and whey)
DATE
Example
5/23/83
FIELD ID.
!1
(or all
FORMULATION
- 10-20-10
AMOUNT APPLIED/AC
250lb$/Ac.
HOW
APPLIED
planter
COMMENTS 13
disced In
5/23/33
C. CROPLAND ACTIVITIES
DATE
Example
6/23/83
FIELD I.D.
21
ROTATION
CorH
C '
ACTIVITY
seed or plow or
harvest or
herbicide/pesticide
COMMENTS 25
BlartB* 5ltt/ac
-------
D.W. MEALS
land use, cropping patterns, and manure application
were routinely mapped in the GIS.
Data Analysis
Relationships between land use and water quality on
a fine time scale are typically confounded by weather
and season and their influence on specific farming
activities. Furthermore, the timing of land use and
farm management activities was not determined as
precisely as the timing of streamflow or pollutant
loads. Therefore, analysis focused on annual pat-
terns of land use and water quality.
Although some detail was lost by using only an-
nual values, this simplified approach had the ad-
vantages of paralleling annual cropping cycles and
rotations. The approach was also necessary because,
despite intensive data collection efforts, the pre-
cision of much of the land use data varied widely
among farmers reporting. Some farmers, for ex-
ample, reported manure application data by exact
day and number of spreader loads; others, however,
reported spreading "some manure every week over
the summer." Aggregating data to an annual basis
was the only workable solution to such differences in
data precision. Furthermore, some land treatment
data, such as completion of structures, were typically
reported annually.
The following land use/agricultural activity vari-
ables were hypothesized to affect water quality:
• Animal population (animal units: AU)
• Animal density (AU/ha)
• Animals under BMP (% of total AU)
• Land in corn (ha and % of watershed)
• Land in pasture (ha and % of watershed)
• Land receiving manure (ha and % of
watershed)
• Land receiving manure from storage (% of
manured area)
• Land receiving manure with incorporation
(% of manured area)
• Cornland where manure incorporated (% of
manured cornland)
• Manure applied (tons and tons/ha)
• Manure quantity from storage (% of manure
applied)
• Manure quantity incorporated (% of manure
applied)
• Barnyard runoff control (# of systems)
« Milkhouse waste management (# of systems)
• Erosion control (ha and % of watershed).
Animal population and density, for example,
determine the amount of animal waste available as a
pollutant source. The percent of animals under BMP
is an aggregate measure of the proportion of im-
proved waste management practiced, including
proper storage, application, timing, and incorpora-
tion into the soil. Specific BMPs, such as barnyard
runoff control and milkhouse waste management,
reduce the volume of contaminated runoff and divert
runoff away from surface waters.
Annual water quality values were used for com-
patibility with the annual land use data. As with the
latter data, aggregating water quality data to annual
values helped reduce some of the same confounding
interactions between weather, season, streamflow,
water quality, and agriculture. The water quality and
flow data used in this analysis were annual means
(antilogs of means of log-transformed data). Only
concentrations were considered to avoid the in-
fluence of variations in annual stream discharge on
mass export.
Relationships between flow and pollutant con-
centrations generally were not statistically sig-
nificant at this level of data aggregation (Vt. RCWP
Coor. Comm. 1991). However, hydrologic variables,
such as precipitation and streamflow, were also in-
cluded as independent variables in the correlation
analysis to assess the significance of their effect on
observed annual water quality.
Correlation and simple linear regression were
used to analyze land use/water quality relationships.
Correlations were evaluated separately in each of the
four monitored subwatersheds. Unique charac-
teristics of some of the monitored watersheds (e.g.,
extremely high animal populations in the Jewett
Brook watershed and occasional point source dis-
charges to Stevens Brook) prevented combining all
watershed data into a single analysis. Significant cor-
relations were examined further using simple linear
regression to determine strength and direction of
the relationship.
It was not possible to control rigorously for
hydrologic variation in this analysis because of the
few degrees of freedom (n=8-9). However, correla-
tions between annual water quality values and
hydrologic variables (such as mean annual stream-
flow, total annual discharge, and total annual
precipitation) were tested and generally found to be
weak and inconsistent. In the few cases where cor-
relation between a water quality and a hydrologic
variable was statistically significant, relationships be-
tween land use and that water quality variable were
135
-------
Proceedings of National RCWP Symposium, 1992
considered meaningful only if the strength of the
correlation exceeded the strength of correlation
with the hydrologic variable.
Results
Water Quality
Water quality data and trends over time in the St. Al-
bans Bay watershed have been presented and dis-
cussed in detail elsewhere (Vt. RCWP Coor. Comm.
1991) and in these proceedings (Meals, 1992), and
the reader is directed to these sources for further in-
formation.
Land CIse and Agricultural Activity
The monitoring program was highly effective in
tracking annual patterns of agricultural land use in
the St. Albans Bay watershed. Agricultural land use
in each of the monitored subwatersheds in 1990 is
shown in Table 1. The Jewett Brook watershed had
the highest proportion of agricultural land (86 per-
cent) and the highest proportion of land in corn (32
percent) among the monitored subwatersheds.
Agricultural land comprised 45 to 55 percent of the
other subwatersheds and less than 10 percent of the
land in these subwatersheds was planted in corn.
Overall land use patterns in the St. Albans Bay
Watershed were relatively stable during the RCWP
project. Total agricultural land area in the watershed
declined by just 364 ha (3 percent) from 1983
through 1990. Forest land increased by 200 ha (7
percent) and urban land added 133 ha (10 percent).
The mix of agricultural land uses remained
reasonably constant within the St. Albans Bay water-
shed from 1983 through 1986, with about 12 percent
of agricultural land in corn, 40 percent in hay, and 23
percent in pasture (Fig. 3). After 1987, cornland area
increased, while areas of both hayland and pas-
Table 1.— Agricultural land use in monitored watersheds,
St. Albans Bay
RCWP, 1990.
LAND USE
JEWETT BR.
STEVENS BR.
RUGG BR.
MILL R.
Corn
Hay
Pasture
Farmstead
Ag. use
unknown
Forest
Non-ag.
TOTAL
450 (32%)
557 (40%)
175 (13%)
10 (<0.1%)
1 (<0.1%)
169(12%)
22(2%)
1,384
209 (9%)
532 (22%)
324 (13%)
10 (<0.1%)
15 (1%)
511 (21%)
825 (34%)
2,426
48 (3%)
523 (34%)
227 (15%)
7 (<0.1%)
0 (0%)
339 (22%)
404 (26%)
1,548
380 (6%)
1,712 (29%)
1,020 (18%)
40 (1%)
26 (<.0.5%)
1,804(31%)
828 (14%)
5,810
tureland declined. This trend in land use coincided
with the Dairy Termination Program and a general
downturn in the dairy economy. One result of the
Dairy Termination Program was a trend toward in-
creases in herd size. The increase in cornland may
reflect an effort by farmers to produce additional
feed to support these larger herds.
Total watershed animal populations remained
relatively constant over the study period (9,900 to
10,000 animal units), with a slight change associated
with the Dairy Termination Program (Fig. 4). The ef-
fect of that program and other livestock sales begin-
ning in 1986 can be seen in the downturn in animal
units in Figure 4. Minimum animal population in the
St. Albans Bay watershed occurred in 1987; after
1987, animal numbers rebounded as average herd
size increased. The highest animal population was
recorded in 1990, the final project year.
The monitored subwatersheds generally were
similar with regard to changing animal numbers. In
most parts of the St. Albans Bay watershed, animal
densities were fairly low and comparable to values
elsewhere in Vermont, about 0.5 to 0.7 animal
units/ha. In the Jewett Brook watershed, however,
animal densities were more than double this level,
1.2 to 1.6 animal units/ha.
The estimated quantity of manure applied by
farmers in the St. Albans Bay watershed from 1983
through 1990 is shown in Figure 5. Manure quantity
was probably underestimated in 1983, the first year
of land use monitoring. The quantity of manure
reportedly applied in the watershed appeared to be
reasonably constant from 1984-90 (122,000 to
151,000 tons/year); year-to-year differences re-
flected small changes in animal populations, varying
duration of pasturing, or reduced spreading in wet
years with carryover to the following years.
The proportion of the total mass of manure ap-
plied each year that came from storage increased
through the project (Fig. 5). The percent from
storage may have been overestimated because of
under-reporting by non-BMP daily-
spreaders; however, there is no
doubt that manure application from
storage tended to increase over the
project period. The prevalence of
manure application from storage
and an observed increase in
manure incorporation on cornland
indicated improved manure man-
agement following RCWP land
treatment.
Field-by-field data on land
receiving manure were collected
and compiled into an annual "snap-
136
-------
D.W. MEALS
1983 1984 1985 1986 1987 1988 1989 1990
•B CORN PASTURE I I WOODLAND
Figure 3.—Plot of agricultural land use, St. Albans Bay watershed, 1983-90.
HAYLAND
CO
c 12
03
co ^
=3 10
o
h= 8
{2 6
— i o
< *
n
z u
<
-
-
-
-
i
81
---*---
i -
82
; '
i
83
+
i
84
. . ,; ,
-•-•g----
I
85
+
i
86
";' .
~r
87
+
K
88
v
---«-:-•
i
89
+
-^^S
I
90
OVERALL SABW
RUGG BR.
PROJECT YEAR
+ JEWETT BR.
»*- MILL R.
••*- STEVENS BR.
Animal unit=405 kg dairy equivalent
Figure 4.—Plot of animal populations, St. Albans Bay watershed, 1981-90.
shot" of manure application through use of a CIS; the land receiving manure did not vary dramatically
1990 data are mapped in Figure 6. In general, some across the St. Albans Bay watershed or within most
2,400 to 2,800 hectares of land received manure an- of the monitored subwatersheds. A substantially
nually between 1983 and 1990. Over time, the area of greater proportion of the' Jewett Brook subwater-
137
-------
Proceedings of National RCWP Symposium, 1992
1983 1984 1985
1986 1987 1988 1989 1990
YEAR
Figure 5.—Plot of annual quantity of manure applied In St. Albans Bay watershed, 1983-90.
LEGEND
FROU STORAGE. NOT
INCORPORATED. 1,221 HA
FROU STORAGE,
INCORPORATED. 1.016 HA
DAILY SPREAD. 378 HA
PASTURING. 642 HA
NO MANURE APPLIED.
OR UNKNOWN. 9,759 HA
Figure 6.—Map of land receiving manure In St. Albans Bay watershed, 1990.
138
-------
D.W. MEALS
shed received manure (30 to 50 percent) than did
the other subwatersheds (<20 percent).
Land Use and Water Quality
• Animal Populations. Neither total watershed
animal population nor animal density appeared to
exert a significant or consistent influence on in-
stream sediment or nutrient concentrations. Sedi-
ment and nitrogen concentrations were not
significantly correlated with animal numbers or den-
sity. Positive correlations between phosphorus levels
and animal density were observed in some water-
sheds and negative relationships between these
same variables in other watersheds.
However, in each of the monitored streams, both
fecal coliform (FC) and fecal streptococcus (FS) bac-
teria counts were significantly correlated with
animal density:
CORRELATION COEFFICIENTS (r)
ANIMAL DENSITY (Animal units/ha)
FC
FS
JEWETT BR.
-0.69**
-0.73**
STEVENS BR.
0.60*
0.75**
RUGG BR.
0.56*
0.78**
MILLR.
0.81**
0.77**
**Ps 0.05;* Ps 0.10
In Stevens Brook, Rugg Brook, and Mill River,
higher bacteria counts were associated with high
animal densities, which might be expected since
animal waste is a major source of bacteria. In Jewett
Brook, however, lower bacteria counts appeared to
be associated with higher animal populations.
• Animals Under BMP. The percent of watershed
animals under BMP was expected to be an important
factor in stream nutrient concentrations in the St. Al-
bans Bay watershed as it had been in the nearby La-
Platte River watershed project (Meals, 1990).
However, in the St. Albans Bay watershed, such
relationships were not observed.
CORRELATION COEFFICIENTS (r)
ANIMAL UNITS UNDER BMP (%)
TP
SRP
TKN
NH3-N
NOa+NOs-N
JEWETT
BR.
0.45
0.22
0.06
0.31
0.26
STEVENS
BR.
-0.86**
-0.92**
-0.16
-0.33
-0.27
RUGG
BR.
0.78**
0.66**
0.18
0.41
0.31
BILL
R.
0.74**
0.68**
0.17
0.25
0.51
**Ps 0.05; *P s 0.10; TP = total phosphorus; SRP = soluble reac-
tive phosphorus; TKN = total Kjeldahl nitrogen; NHa-N = ammonia
nitrogen; NOa+NOa-N = nltrlte+nltrate nitrogen.
The negative correlations observed in Stevens
Brook probably reflect changes in water quality
resulting from the wastewater treatment plant
upgrade, rather than changes in nonpoint source
contributions (Meals, 1992). Unlike the pattern ob-
served in the LaPlatte River watershed where phos-
phorus and nitrogen concentrations were negatively
correlated with animal units under BMP and only
cases where animal units under BMP exceeded 85
percent were considered (Meals, 1990), positive cor-
relations were observed in Rugg Brook and Mill
River when all cases were considered (as shown pre-
viously) as well as when early years of low implemen-
tation were excluded (r=+0.45 to +0.75).
The lack of apparent response in nutrient levels
to improvements in animal waste management may
relate to the threshold concept In the LaPlatte River
watershed where 100 percent of animals came under
BMP waste management in some subwatersheds,
response did not seem to occur until more than 85 to
90 percent of watershed animals were under BMP, al-
though exact threshold levels were not established
(Meals, 1990). In the St. Albans Bay RCWP mon-
itored subwatersheds, however, maximum levels of
treatment achieved were considerably lower: 61 to
83 percent of animal units under BMP. Thus, the
level of treatment with regard to animals under BMP
waste management in the St. Albans Bay watershed
may not have been sufficient to generate a response
in stream nutrient levels—at least on an annual
basis.
The level of animals under BMP, however, ap-
peared sufficient to affect bacteria levels (Fig. 7). In
each of the monitored subwatersheds, decreasing
fecal coliform and fecal streptococcus counts were
clearly associated with increasing proportions of
watershed animals under BMP:
CORRELATION COEFFICIENTS (r)
ANIMAL UNITS UNDER BMP (%) (AUBMP)
JEWETT BR. STEVENS BR. RUGG BR.
MILL R.
FC
FS
-0.47
-0.85**
-0.85**
-0.94**
-0.84**
-0.84**
-0.49
-0.78**
**P s0.05; *Ps0.10; FC = fecal coliform; FS = fecal streptococcus
The significant regression relationships were:
• Jewett Brook:
logFS= -1.68(logAUBMP)+4.96 ^=0.72 P<0.01
• Stevens Brook:
logFC= -1.74(logAUBMP)+5.27 r^O.72 P<0.01
logFS= -1.69(logAUBMP)+4.60 r^=0.89 P<0.01
• Rugg Brook:
logFC=-0.96(logAUBMP)+3.70 r^O.70 P<0.01
logFS= -1.44(logAUBMP)+3.92 ^=0.70 P<0.01
• Mill River:
logFS= -2.47(logAUBMP)+5.92 ^=0.61 P=0.01
• Manure Management. Few significant relation-
ships were observed between water quality and
139
-------
Proceedings of National RCWP Symposium, 1992
*-._**
r*-0.24, P-0.18 """•--
20
40 60
% AU UNDER BMP
100
•JEWETTBR. •-+-• STEVENS BR. •-*-• RUGG BR. a MILL R.
"•---.._ + ''-r'-o-sg, P<0.01
r'-0.70. P'0.01 """•-*. D
•
20
40 60
% AU UNDER BMP
100
•JEWETTBR. •-+-• STEVENS BR. .-•*-• RUGQ BR. a . MILL R.
Figure 7.—Plots of indicator bacteria counts vs. percent of watershed animals under BMP waste management in
monitored subwatersheds, St. Albans Bay watershed.
either manure quantity or manured area. Use of
manure storage assessed by either area or quantity
of application did not appear to exert a significant in-
fluence on stream nutrient levels.
Manure storage did appear to have a major ef-
fect on stream bacteria counts. In most of the
monitored subwatersheds, significant negative cor-
relations were observed between percent of
manup«i area from storage and both fecal coliform
and fecal streptococcus counts as well as between
percent of manure quantity applied from storage and
bacteria counts. Significantly lower bacteria counts
in streams occurred with greater use of manure
from storage.
140
-------
D.W. MEALS
CORRELATION COEFFICIENTS (r)
MANURE FROM STORAGE (% of total applied)
JEWETTBR. STEVENS BR. RUGG BR.
MILL R.
FC
FS
-0.73**
-0.64*
-0.78**
-0.64*
-0.89**
-0.70**
-0.59*
-0.43
* P s 0.05; *Pa 0.10; FC = fecal coliform; FS = fecal streptococcus
AREA RECEIVING MANURE FROM STORAGE
(% of manured area)
JEWETTBR. STEVENS BR. RUGG BR.
MILLR.
FC
FS
-0.82**
-0.79**
-0.30
-0.24
-0.97**
-0.84**
-0.74**
-0.70**
** P s 0.05; * P s 0.10; FC = fecal coliform; FS = fecal streptococcus
This pattern may be a result of rapid die-off of
bacteria in stored manure. Reports in the literature
suggest that, by itself, storage of manure may reduce
levels of fecal coliform and fecal streptococcus in the
manure by more than 99 percent (Patni et al. 1985).
Furthermore, improved manure management as-
sociated with storage (e.g., proper timing, control of
application rates) may have further reduced oppor-
tunities for transport and delivery of bacteria to
streams.
No significant relationships were found between
water quality and soil incorporation of applied
manure, either on an areal or quantity basis. The
strong association between the extent of manure in-
corporation and cropping patterns (manure can only
be incorporated on plowed cropland, for example)
may have obscured any meaningful relationships.
• Cropping. Associations between water quality
and cropping patterns, such as land in corn and pas-
ture, were generally weak, inconsistent, and con-
founded by relationships with other variables. In
Jewett Brook, for example, lower bacteria counts
seemed to be associated with greater land in corn.
However, since increases in cornland were also as-
sociated with increases in animals under BMP, the
decline in bacteria counts could have been a
response to greater BMP waste management.
• Specific BMPs. Variables focusing on implemen-
tation of specific BMPs, such as barnyard runoff or
erosion control, were of little use in explaining ob-
served water quality. Implementation of such prac-
tices was progressive and irreversible, remaining
constant after implementation goals were achieved.
Consequently, there was not enough variation in
these measurements to be useful as explanatory
variables. The cumulative progress of implementa-
tion of all of these practices closely paralleled other
changes in agricultural management, such as animal
units under BMP. Thus, even when correlations
were found between higher numbers of barnyard
systems installed and lower bacteria levels on
streams, for example, concurrent increases in
animals under BMP waste management were likely
an important contributor to the observed relation-
ship.
• Riparian Land Use. It was hypothesized that
agricultural activities close to surface waters might
influence water quality more than upland activities.
To test this hypothesis, land use and manure applica-
tion data within a 50 m corridor along either side of
Jewett Brook and its tributaries were explored for
significant correlation with observed water quality.
Use of CIS facilitated the determination of riparian
land use.
Land use for corn, hay, or pasture within 50 m of
Jewett Brook seemed to have a significant influence
on streamwater quality, possibly because of manure
management practices^ In particular, concentrations
of some forms of nitrogen tended to be lower when
more riparian land was in hay or pasture but higher
when more riparian area was in corn:
CORRELATION COEFFICIENTS (r)
JEWETT BROOK RIPARIAN LAND USE (ha)
CORN
PASTURE
HAY
TKN
NH3-N
NOg+NOa-N
0.26
0.85*
0.74*
-0.90**
-0.82**
_ 1
**Ps0.05;*Ps0.10;TKN = total Kjeldahlnitrogen;NH3-N = am-
monia nitrogen; NO2+NO3-N = nitrite+nltrate nitrogen
This relationship makes some sense because
manure applications on pastureland tend to be lower
(i.e., by cows only) and, more importantly, more pas-
tureland implied less cornland where manure
and/or fertilizer would be more likely to be applied
regularly. Positive correlations were, in fact, ob-
served between riparian land in corn and both total
quantity of manure applied and area .receiving
manure. Similarly, there were negative correlations
between both manure variables and land in pasture
or hay. Data from nutrient management studies in
the St. Albans Bay watershed confirmed that manure
nutrient application to corn was nearly twice that to
hay and tended to be in excess of crop need (Vt.
RCWP Coor. Comm. 1991).
Bacteria counts in Jewett Brook also appeared to
be correlated with agricultural activities close to the
stream. Mean fecal coliform and fecal streptococcus ,
counts showed significant negative correlations with
land in corn within the stream corridor:
CORRELATION COEFFICIENTS (r)
JEWETT BROOK RIPARIAN LAND USE (ha)
CORN
PASTURE
HAY
FC
FS
-0.78**
-0.86**
0.30
0.48
0.64*
0.43
** Ps 0.05; * P s 0.10; FC = fecal coliform; FS = fecal streptococcus
141
-------
Proceedings oftiattonal RCWP Symposium, 1992
Lower bacteria counts were associated with
greater cornland area adjacent to the stream. This
apparently anomalous relationship may be explained
by manure management practices. Within the
stream corridor, cornland area was significantly as-
sociated not only with manure quantity but also with
manure storage and incorporation—more manure
was applied to cornland than to hayland or pasture,
but more of that manure was from storage and more
was incorporated. The influence of manure storage
on stream bacteria counts has already been sug-
gested. Thus, the declines in bacteria counts that
were associated with increased cornland near Jewett
Brook may have been the result of improved manure
management. In addition, decreased bacteria counts
may have resulted from the decreased likelihood of
direct deposition of manure by grazing cattle when
more riparian land was in corn.
Discussion
Evaluation of land use/water quality relationships in
the St. Albans Bay watershed was not a simple or
straightforward task, nor were results unequivocal.
Changing land use and agricultural management ac-
tivities did not appear to have significant effects on
stream nutrient concentrations. This may not be
surprising because no significant reductions in phos-
phorus and nitrogen levels in tributaries to St. Al-
bans Bay were observed (Meals, 1992).
The strongest results were noted for levels of in-
dicator bacteria in the streams. Bacteria counts in
streams seemed to be related to animal density —
higher bacteria counts were associated with higher
watershed animal populations. Bacteria counts ap-
peared to decline with increasing percentage of
watershed animals under BMP waste management.
Significantly lower bacteria counts in streams tended
to occur with greater use of manure from storage.
Both 5n-stream nitrogen concentrations and bac-
teria counts appeared to be significantly related to
riparian land use. Lower nitrogen levels seemed to
be associated with more riparian land in pasture or
hay. Greater riparian land in corn tended to yield
higher nitrogen concentrations but lower bacteria
counts.
On a whole watershed basis, few meaningful
relationships were found between cropping patterns
or manure management and water quality. Use of
data on the progress of implementation of specific
BMPs was of little value.
The unique behavior of water quality in the
Jewett Brook watershed deserves note. Compared to
the other monitored subwatersheds, the Jewett
Brook watershed had a higher proportion of its land
in agriculture (86 percent), a higher proportion of its
land cultivated for corn (32 percent), the highest
animal density (1.2 to 1.6 animal units/ha), and the
highest proportion of land receiving manure (30 to
50 percent). Jewett Brook consistently demonstrated
the highest phosphorus and nitrogen concentrations
and areal loads as well as the highest bacteria counts
among the monitored subwatersheds (Meals, 1992).
Some relationships between water quality and
land use in the Jewett Brook watershed were also
distinct from those observed elsewhere in the St. Al-
bans Bay watershed. In Jewett Brook, for example,
lower bacteria counts were associated with higher
animal densities, while the opposite was true in the
other monitored subwatersheds. The explanation for
this contrary behavior may be twofold. First, the
Jewett Brook watershed had the greatest proportion
of land where manure can be incorporated (corn)
and the lowest proportion of area in pasture where
manure cannot be collected and managed, thus pos-
sibly reducing the availability of bacteria for
transport in surface runoff. Second, the proportion
of animals under BMP increased along with animal
numbers in the Jewett Brook watershed, which had
the highest proportion of animals under BMP
among all the monitored subwatersheds. It is pos-
sible that manure was managed better in the Jewett
Brook watershed, despite higher animal numbers.
It must be noted that the simple analysis of an-
nual values reported here was not able to control ef-
fectively for climatic and hydrologic variation nor
was it able to sort out clearly the competing and con-
founding effects of multiple changes in land use and
agricultural management. For example, while the
percent of animal units under BMP waste manage-
ment was proposed as an explanatory variable for
changes in bacteria counts, other aspects of the land
treatment program, such as control of barnyard
runoff, may have contributed to the change in bac-
teria counts. The number of animals under BMP was
advanced as a likely factor because reasonable
mechanisms can be proposed to support such a
relationship; i.e., bacteria die-off in storage and soil
incorporation.
Thus, the land use/water quality relationships
discussed here should be taken as observations and
suggestions, not indications of cause and effect. The
primary lesson learned in attempting to sort out
such relationships in the St. Albans Bay RCWP
project was that there is a long way to go before land
use/land treatment and water quality can be clearly
linked on a watershed scale. Numerous issues must
be addressed along the way.
• Improved experimental design is needed to
document land treatment/water quality relationships
142
-------
D.W. MEALS
at the cause-and-effect level. Improved design would
include
• targeting land treatment so the most serious
problems are treated first;
• accounting for hydrologic variability by using
a control watershed where no treatment
occurs;
• monitoring other critical influences on
in-stream water quality, such as hydrologic
pathways, soil interactions, and pollutant
transport mechanisms; and
• allowing greater manipulation and control of
management activity to influence system
response.
• Better data are needed concerning land use and
changing agricultural management activities. Even
though the monitoring program conducted for this
comprehensive monitoring and evaluation effort was
very ambitious and more successful than an-
ticipated, the accuracy and precision of the land use
management data collected were still far less than
that of the water quality data. Methods must be
developed to obtain more accurate land use and
management data.
• During the course of this long-term monitoring
program, land treatment contracts expired and con-
tinued adherence to implemented practices was not
adequately measured. Furthermore, livestock num-
bers, the composition of livestock populations, and
the distribution of animals among contract and non-
contract farms varied and may have had a significant
impact on the effective level of treatment in different
watersheds. Such changes may have important
water quality implications and must be tracked
throughout the monitoring period, not just during
the active implementation phase of the project.
• Simple counts of BMPs implemented are of ex-
tremely limited usefulness; detailed information
on farm practices and changing patterns of land
use and management is essential. For example,
better information on timing of manure application
might have allowed analysis on a finer time scale
than one year, thereby permitting more sophisti-
cated statistical analysis of land use/water quality
relationships. Future land use monitoring and land
treatment tracking efforts must strive to improve
data accuracy and precision. Without an improved
database, efforts to link land treatment with water
quality will be limited to the simple, suggestive
results discussed here.
• Finally, improvements in experimental design
and data quality would allow more sophisticated
statistical analysis. Use of annual values, even with
the relatively long monitoring period, provided insuf-
ficient degrees of freedom for multivariate models.
Multiple or stepwise regression, cluster analysis, or
other multivariate techniques might be more effec-
tive in sorting out the importance of numerous in-
fluences on water quality.
ACKNOWLEDGMENTS:^ St. Albans Bay Watershed
RCWP comprehensive monitoring and evaluation program
received financial support from the USDA Agricultural
Stabilization and Conservation Service and from the Univer-
sity of Vermont. Assistance from the Soil Conservation Ser-
vice was critical to the success of the program. The
cooperation and assistance of the Franklin County Natural
Resources Conservation District, the Vermont Cooperative
Extension Service, and the Vermont Department of Environ-
mental Conservation is gratefully acknowledged. Finally, the
hard work and dedication of Dr. John Clausen, who guided
the comprehensive monitoring and evaluation program for
most of its course, is also acknowledged.
References
Clausen, J.C. 1985. The St. Albans Bay watershed RCWP: a case
study of monitoring and assessment. Pages 21 to 24 in
Perspectives on Nonpoint Source Pollution, Prdc. Natl. Conf.
Kansas City, MO.
Meals, D.W. 1990. LaPlatte River Watershed Water Quality
Monitoring and Analysis Program Comprehensive Final
Report. Prog. Rep. No. 12. Vt. Water Resour. Res. Center,
Univ. Vermont, Burlington.
. 1992. Water quality trends in the St Albans Bay watershed
following RCWP land treatment. In Proc. Nafl. RCWP Symp.,
Orlando, FL.
Patni, N.K., M.R. Toxopeus, and P.Y. Jui. 1985. Bacterial quality of
runoff from manured and non-manured cropland. Trans. Am.
Soc. Agric. Eng. 28(6) =1871-84.
Standard Methods for the Examination of Water and Wastewater.
1985. 16th ed. Joint Editorial Board, Am. Pub. Health Ass.,
Am. Water Works Ass., and Water Pollut Control Fed.,
Washington, DC.
U.S. Environmental Protection Agency. 1983. Methods for Chemi-
cal Analysis of Water and Wastes. EPA-600/4-79.020. Off.
Res. Dev., Cincinnati, OH.
Vermont RCWP Coordinating Committee. 1991. St. Albans Bay
Rural Clean Water Program Final Report 1980-1990. Vt.
Water Resour. Res. Center, Univ. Vermont, Burlington.
143
-------
-------
Spatial and Temporal Change
in Animal Waste Application in the
Jewett Brook Watershed, Vermont:
1983-1990
Joel D. Schlagel
School of Natural Resources
University of Vermont
Burlington, Vermont
ABSTRACT
The Jewett Brook watershed is part of the St. Albans Bay watershed, located on Lake Champlain in
northwestern Vermont. In 1980, funds were made available under the Rural Clean Water Program
to develop and implement strategies to reduce agricultural nonpoint source pollution in the St Al-
bans Bay by using agricultural best management practices (BMPs). However, implementing BMPs
in the Jewett Brook watershed did not significantly improve water quality over the 10-year monitor-
ing period (Vt. RCWP Coor. Comm. 1991), possibly because significant changes in the spatial pat-
tern of animal waste application also occurred during the monitoring period. If these changes did
occur, they may have obscured the benefits of the agricultural BMPs. A geographic information
system-based nonparametric statistical method was used to investigate this hypothesis. It found
that the annual manure application rate increased on 13 percent of the land in the Jewett Brook
watershed that received manure during the project period, while the application rate decreased on
4 percent of the manured land (PsO. 10). Within 100 meters of Jewett Brook, the annual application
rate increased on 18.2 percent of the manured land and decreased on 3 percent of the manured
land (PsO.10). These changes in manure application rates may help explain the lack of improve-
ment in surface water quality during the project period.
The St. Albans Bay watershed is located 40
km north of Burlington in northern Ver-
mont. The watershed drains 13,000 hectares
of agricultural, forested, and urban land into St. Al-
bans Bay of Lake Champlain. The Jewett Brook
watershed, a largely agricultural subwatershed lo-
cated at the north end of the St Albans Bay water-
shed, represents slightly less than 11 percent of the
total watershed area. St. Albans Bay has been ex-
periencing accelerated eutrophication because of
nutrient inputs from point and nonpoint sources.
Agricultural nonpoint sources of phosphorus
(primarily from dairy farming) account for 93 per-
cent of the total phosphorus reaching St. Albans Bay
(Vt. RCWP Coor. Comm. 1991). Spring runoff from
farm fields where manure had been spread during
the winter was believed to be the most significant
source of nonpoint source pollution in the St. Albans
Bay watershed.
Under the Federal Water Pollution Control Act
of 1980, funds were made available to design and im-
plement the St. Albans Bay Rural Clean Water Pro-
gram (RCWP) project. The St. Albans Bay RCWP
project provided funding to farmers to implement a
series of agricultural best management practices
(BMPs), including installing animal waste manage-
ment systems or containment structures that allow
farmers to store manure until needed, giving the
145
-------
Proceedings of National RCWP Symposium, 1992
farmer an alternative to spreading manure during
winter months. Animal waste management systems
were the most costly BMP implemented in the St. Al-
bans Bay watershed, accounting for 77 percent of
Federal cost-sharing funds (Vt. RCWP Coor. Comm.
1991).
The St. Albans Bay RCWP project also included
comprehensive monitoring and evaluation—an in-
tensive land use and water quality monitoring pro-
gram conducted by the University of Vermont's
Water Resources Research Center from 1981
through 1990. The results of the 10-year comprehen-
sive monitoring and evaluation program suggest that
implementation of recommended agricultural man-
agement practices in the Jewett Brook watershed did
not significantly improve water quality over the
monitoring period (Vt. RCWP Coor. Comm. 1991).
The land use monitoring portion of the com-
prehensive monitoring and evaluation program
gathered data on manure application rate, source,
and application method, as well as time and location
of application. If significant changes in the spatial
pattern of animal waste application occurred over
the monitoring period, these changes would have en-
hanced or offset the benefits of recommended
agricultural BMPs. To investigate this possibility, a
raster geographic information system database con-
taining the location, time, and quantity of manure ap-
plication was generated. A nonparametric test for
trend in manure application rate over time was then
calculated on a per-pixel basis.''The result of this
statistical operation was a geographic information
system coverage representing the spatial distribu-
tion of the statistical significance of trend in manure
application rate over time (Schlagel, 1992).
Study Area
The Jewett Brook watershed is an area of 1,384 hec-
tares at the north end of the St. Albans Bay water-
shed. Twenty-one farms and approximately 2,000
animal units are located within the Jewett Brook
watershed. (One animal unit is equivalent to one
1,000-pound dairy cow for estimation of manure
production [Vt. RCWP Coor. Comm. 1991].) The
Jewett Brook watershed is the most agricultural of
the St. Albans Bay watersheds, with approximately
85 percent of its land in agricultural use. It also has
more land planted in corn and a greater density of
farm animals than other subwatersheds (Vt. RCWP
Coor. Comm. 1991). The Jewett Brook watershed
also had the highest areal phosphorus and nitrogen
loadings to St. Albans Bay (Vt. RCWP Coor. Comm.
1991).
Within the Jewett Brook watershed, 80 percent
of the land was identified as critical for inclusion in
the St. Albans Bay RCWP project. Critical areas were
defined as "all land used for animal waste disposal as
well as all cropland and farmsteads" (Vt. RCWP
Coor. Comm. 1991). Sixteen farms comprising 2,126
critical acres (860 hectares) participated in the
RCWP project, while five farms representing 594
critical acres (240 hectares) did not participate.
During the project period, most agricultural land in
the watershed remained agricultural with hay, corn,
and pasture being the dominant land uses (Table 1).
The intensity of agricultural practice appeared to in-.
crease between 1983 and 1990, planted corn acreage
increased from 21 to 33 percent of the land in the
watershed. Herd size varied from 1,963 animal units
in 1984 to 2,166 animal units in 1990.
Methods
Water Quality
Comprehensive water quality monitoring in ,the
Jewett Brook watershed was conducted .from 1981
through 1990. The monitoring program goals were
to ,'..,.'
• document changes in water quality resulting
from the implementation of BMPs, and
• evaluate trends in water quality in the water-
shed during the St. Albans Bay RCWP.
The water quality monitoring program
methodology has been described in detail elsewhere
(Vt. RCWP Coor. Comm. 1991; Meals, 1992a).
LAND USE
Corn
Hay
Pasture
Woodland
Other*
1983
291 (21)
662 (48)
195 (14)
203 (15)
34(2)
1984
325 (23)
625 (45)
193 (14)
203 (15)
38(3)
1985
378 (27)
581 (42)
194 (14)
195 (14)
35(3)
1986
1987
hectares (percent)
378 (27) 405 (29)
647 (47)
148 (11)
176 (13)
36(3)
558(40)
205 (15)
180 (13)
36(3)
1988
421 (30)
548 (40)
184 (13)
196 (14)
36(3)
1989
407(29)
607 (44)
140 (10)
187 (13)
44 (3)
1990
450 (32)
557(40)
175 (13)
169(12)
33 (2)
*Olhor Includes farmsteads, nonagricultural land, and agricultural land—use unknown
146
-------
J.D. SCHLAGEL
Animal Waste Management
Intensive agricultural land use monitoring of the
Jewett Brook watershed began in 1983. Farmers in
the watershed served as primary source of data for
the monitoring program. Maps of each farm within
the watershed were prepared from 1:5,000 ortho-
photographs and digitized. All farmers within the
watershed, including those who chose not to par-
ticipate in the RCWP, were given checkbook-sized
logbooks containing a map of their farm (with each
field numbered). Here they were asked to record the
date, field, amount of manure and fertilizer applied,
planting and harvesting dates, and other information
about cropping activities. These data were gathered
from farmers semiannually, along with data on the
use of each field for that period (e.g., corn, hay, or
pasture). Information on the type and number of
animals present on the farm as well as subdivision of
land, field boundary, and field ownership changes
were recorded during periodic interviews.
Farmer participation in the land use monitoring
program was far greater than expected, but the use
of the field logs and the quality of the data collected
varied from farm to farm. Roughly one-third of the
farmers kept very accurate records, one-third kept
some information, and one-third did not record data
but were willing to try to reconstruct their activities
when interviewed (Vt. RCWP Coor. Comm. 1991).
Results of each farm interview (including the rela-
tive quality of the data obtained) were coded and
added to a geographic information system database.
This data collection procedure was repeated for each
of the eight years that data were collected. The
ARC/INFO geographic information system (En-
viron. Systems Res. Inst. 1990) was used initially to
store and analyze the land use data.
The manure application date, location, and quan-
tity for each calendar year were aggregated into four
quarterly totals (January to March, April to June,
July to September, October to December) and an an-
nual total. The data were then converted from a vec-
tor to a raster data format and transferred to the
Idrisi geographic information system (Eastman,
1990), with each pixel representing 256 square
meters. The Mann-Kendall trend statistic was then
calculated for each pixel location using the overlay
operations of Idrisi. The Mann-Kendall test is a non-
parametric test for zero slope of the linear regression
of time-ordered data versus time (Gilbert, 1987).
The presence of statistically significant trends
was examined in each quarterly and the one annual
coverage at the PsO.20 and PsO.10 significance
levels. The consideration of the PssO.20 probability
level was deemed important for this data set because
of the small sample size (n=8 years). Figure 1 is a
map that represents results of the Mann-Kendall
trend calculations for annual manure application
data.
Results
Water Quality
Trend analysis of water quality data was conducted
using a variety of nonparametric tests, including the
seasonal Kendall and the seasonal Mann-Whittney.
Based on the results of these analyses, no significant
trends were apparent in either surface water phos-
phorus concentration or the mass of phosphorus ex-
ported from the Jewett Brook watershed. In
addition, strong trends were not present in total or
volatile suspended solids concentration (or export)
or nitrogen export, although nitrogen concentrations
did tend to increase. Fecal coliform and fecal strep-
tococcus bacteria counts decreased more than 50
percent over the project period. Water quality results
are discussed in detail elsewhere (Vt. RCWP Coor.
Comm. 1991; Meals, 1992a; Meals, 1992b).
Animal Waste Management
Over the eight-year monitoring period, the first
quarter manure application rate significantly de-
creased on 9 hectares at the PsO.10 level. No trend
was present on 221 hectares, and no significant in-
crease in application rate occurred on any watershed
land at P=s0.10 (Table 2). The limited area on which
Table 2.—Quarterly and annual trends: manure application rate.
DIRECTION OF TREND & STATISTICAL SIGNIFICANCE
Downward trend PsO. 10
Downward trend PsO.20
No trend PsO.10
No trend PsO.20
Upward trend PsO.1 0
Upward trend PsO.20
Nonmanured area
ANNUAL
41
80
826
621
129
294
389
Q1
9
26
221
200
0
5
1,153
Q2
19
48
720
571
138
258
507
Q3
11
26
64
570
6
68
720
Q4
16
34
566
516
2
34
799
Q1 = January through March; Q2 = April through June; Q3 = July through September; Q4 = October through December
147
-------
Proceedings of National RCWP Symposium, 1992
LEGEND
•elcrs
NO MANURE APPLIED
DOWNWARD TREND: P<=0.10
DOWNWARD TREND: P< = 0.20
NO SIGNIFICANT TREND
UPWARD TREND: P<=0.20
UPWARD TREND: P< = 0.10
Figure L-^Jewett Brook watershed 1983-90: trend In annual manure application rates.
148
-------
J.D. SCHLAOEL
manure was applied during the first quarter and the
limited area with significant decreases in application
rate probably reflects the widespread implementa-
tion of the animal waste management BMP in the
watershed before the beginning of detailed land use
monitoring. Most farmers were already properly
managing manure by 1983, and the first quarter data
indicate that farmers were generally reporting com-
pliance with BMPs over the project period. The 221
hectares that had manure applied (but for which
there was no significant trend) probably reflect
manure spreading by farmers who chose not to par-
ticipate fn the project, the late implementation of
BMPs on some farms, and occasional early spring
spreading of manure.
During the monitoring period, most manure was
spread during the second quarter in the early spring
before planting. The second quarter also saw the
greatest area with increases in manure application
rates during the project period. Statistically sig-
nificant increases in application rate occurred on 138
hectares or 15.8 percent of the manured land at
P=s0.10 (Tables 2 and 3). Significant decreases in ap-
plication rates occurred on 19 hectares or 2.1 per-
cent of the manured land at PsO.lO. The increase in
manure application rates on these fields is probably
associated with an overall increase in reported
manure application and an increase in corn acreage
in the watershed from 1983 to 1990.
Data for the third and fourth quarters reflect
similar patterns. Increases or decreases in manure
application rate statistically significant at PsO.lO
were found on very little watershed land (Tables 2
and 3); however, at the P=sO,20 significance level,
changes were more evident. At the PsO.20 level, a
significant increase occurred on 10.2 percent of the
manured land in the watershed during the third
quarter and 5.9 percent of the manured land during
the fourth quarter.
The pattern of manure application reflected in
the annual totals appears similar to the second
quarter data, largely because the total annual
manure application is dominated by the second
quarter.
The initial trend analysis in manure application
rate deliberately ignored farm and field designations
and boundaries because they varied over the
project's course. Nonetheless, it is useful to consider
the farm as a unit when examining the location of
significant trends in application rate. To do this, the
coverage containing the annual Mann-Kendall result
was intersected with the 1990 land use coverage to
identify those farms responsible for trends in
manure application rates. This analysis revealed that
one farm was responsible for an increase in manure
application rate on 38 hectares (or 29 percent of all
acres) with a significant increase in annual applica-
tion rate at PsO.10. This farm is located in the
watershed's center at the junction of the two
branches of Jewett Brook. A review of the field logs
for this farm revealed that herd size and manure ap-
plication increased consistently over the project
period.
It is also useful to consider the location of trend
with respect to the watershed's physical features that
may be associated with nbnpoint source pollution.
For example, changes in agricultural practice occur-
ring within 50 meters of Jewett Brook were more
highly correlated with changes in surface water
quality than changes in agricultural practice at more
remote locations (Vt. RCWP Coor. Comm. 1991;
Meals, 1992b). It is clear from a visual inspection of
Figure 1 that the largest concentration of fields with
statistically significant increases in annual applica-
tion rate are located adjacent to Jewett Brook.
Geographic information system buffer opera-
tions were used to identify the area in six different
distance classes from the main course of Jewett
Brook (shown as a heavy line in Fig. 1). Table 4 con-
tains a breakdown of the trend in application rate
versus distance to Jewett Brook. Within 50 meters
and 100 meters of the stream course, annual manure
application rate significantly increased on 19.7 and
18.2 percent of the land respectively, as compared to
13 percent of the land throughout the watershed at
P^O.10. Significant decreases (PsO.10) occurred on
2.4 and 3.0 percent of the land within 50 and 100
meters of the stream course as compared to 4.0 per-
cent watershed-wide.
Table 3.—Temporal trend in manure application rate on manured land.
DIRECTION OF TREND & STATISTICAL SIGNIFICANCE
ANNUAL
01
02
Q3
04
Downward trend PsO.1 0
Downward trend PsO.20
No trend PsO.10
No trend PaO.20
Upward trend PsO.10
Upward trend PsO.20
4.0
8.0
83.0
62.4
13.0
29.6
4.1
11.1
95.9
86.8
0.0
2.1
2.1
5.5
82.1
65.1
15.8
29.4
1.7
3.8
97.5
85.9
0.8
10.2
2.8
5.8
96.8
88.3
0.4
5.9
Q1 = January through March; Q2 = April through June; Q3 = July through September; Q4 = October through December
149
-------
Proceedings of National RCWP Symposium, 1992
Table 4.—Temporal trend In manure application rate versus distance to Jewett Brook
DISTANCE (METERS)
DIRECTION OF TREND &
STATISTICAL SIGNIFICANCE
Downward trend PsO.10
Downward trend PsO.20
No trend PsO.10
No trend P*0.20
Upward trend P*0.10
Upward trend PsO.20
so
2.4
3.8
77.9
60.5
19.7
35.7
100
3.0
4.7
78.8
59.5
18.2
35.9
200
3.9
5.7
80.3
58.1
15.8
36.2
300
— percent
4.3
6.0
80.6
58.7
15.0
35.3
400
4.2
6.2
80.2
58.2
15.6
35.6
500
4.8
7.0
79.7
58.9
15.5
34.1
WS*
4.0
8.0
83.0
62.4
13.0
29.6
•WS > watershed total
Discussion
This project investigated the possibility that changes
in the spatial pattern of manure application occur-
ring during the St. Albans Bay RCWP project in
some way offset the benefits of improved agricul-
tural management. For the first, third, and fourth
quarters of the year, relatively little change occurred
in the pattern of manure application from 1983 to
1990. The pattern for the second quarter and the an-
nual aggregate data reflected increases in annual
manure application rate on one farm and several
fields on other farms from 1983 to 1990. Twenty-nine
percent of the land with significant increases in ap-
plication rate was associated with one farm. Despite
a sample size of only eight years, these increases
were statistically significant atP^O.10.
Farm fields with increases in manure application
rate tended to be located adjacent to Jewett Brook.
Statistically significant increases in annual manure
application rates over the study period occurred on
20 percent of the land within 50 meters of Jewett
Brook that received manure during the study period,
compared with decreases on just 2 percent of the
riparian land. Watershed-wide application rates in-
creased on 13 percent of the land receiving manure
and decreased on just 4 percent Increases in
manure application rates, particularly on riparian
land, demonstrate that, despite widespread im-
plementation of BMPs, agricultural practices in the
Jewett Brook watershed are—to some extent—work-
ing against improvements in water quality.
References
Eastman, J.R. 1990. Idrisi Users Guide. The Idrisi Proj., Clark
Univ., Worcester, MA.
Environmental Systems Research Institute. 1990. ARC/INFO
Users Guide. Redlands, CA
Gilbert, R.O. 1987. Statistical Methods for Environmental Pollu-
tion Monitoring. Van Nostrand Reinhold, New York.
Meals, D.W. 1992a. Water quality trends in the St Albans Bay, Ver-
mont, watershed following RCWP land treatment. /« Proc.
Natl. RCWP Symp., Orlando, FL.
. 1992b. Relating land use and water qualify in the St. Al-
bans Bay watershed, Vermont /« Proc. Natl. RCWP Symp.,
Orlando, FL.
Schlagel, J.D. 1992. A GIS-based Statistical Method to Analyze Spa-
tial and Temporal Change. Masters Proj., Univ. Vt., Bur-
lington.
Vermont RCWP Coordinating Committee. 1991. St Albans Bay
Rural Clean Water Program Final Report 1980-1990.'Vt.
Water Resour. Res. Center, Univ. Vt., Burlington.
150
-------
Water Quality and Land Treatment
in the Rock Creek, Idaho,
Rural Clean Water Program
Gwynne Chandler and Terry Maret
Division of Environmental Quality
Idaho Department of Health and Welfare
Boise, Idaho
ABSTRACT
Because of point and nonpoint pollution sources, Rock Creek has long been recognized as one of
the most severely degraded streams in Idaho. By the late 1970s, most point source pollutants had
been eliminated; however, nonpoint sources remained a problem. Rock Creek water quality is af-
fected by sediment and associated pollutants from irrigation return flows and feedlot runoff. Rock
Creek was one of only five national Rural Clean Water Program (RCWP) projects to receive addi-
tional funding for land treatment and water quality monitoring and evaluation. Goals of the Rock
Creek RCWP project included reducing the amount of sediment, sediment-related pollutants, and
animal waste entering Rock Creek by implementing best management practices (BMPs). Land-
owners were contracted to implement BMPs with the length of the contract varying from three to
ten years (depending on the practice contracted). Nonparametric trend analysis has shown sig-
nificant reductions in pollutant loadings and concentrations in lower Rock Creek resulting from
BMP implementation. As part of this evaluation an extensive water quality and land treatment
database for subbasins has been constructed. This data will be used to compare water quality with
various land use/land treatment practices implemented over the 10 years of the project.
The Rural Clean Water Program (RCWP) is
designed to control agricultural nonpoint
source pollution in rural watersheds and ul-
timately improve water quality (Yankey et al. 1991).
Rock Creek, Idaho, is one of 21 RCWP projects in
the United States and one of only five with a com-
prehensive monitoring and evaluation component.
RCWP projects involve both land treatment and
water quality monitoring. Landowners were con-
tracted to implement best management practices
(BMPs), for three to ten years, depending on the
practice being implemented. Most RCWP contracts
began in 1980 or 1981, and ended in 1986, but the im-
plementation of some BMPs continued through
1992.
This paper describes the Rock Creek RCWP's
water quality results and the process of linking land
treatment and water quality data. Changes in water
quality result from changes in the environment.
Therefore, correctly implemented BMPs are ex-
pected to have a positive effect on water quality.
Linking land treatment to water quality is a complex
process. This paper summarizes the techniques
being used to make this connection in the Rock
Creek RCWP's postevaluation phase.
Project Background
Rock Creek, located in Twin Falls County, Idaho,
had long been recognized as one of the most severe-
ly degraded streams in the State because of point
and nonpoint source pollution. For years, the creek
had been used by sugar, meat, and dairy processing
plants, stockyards, and feedlots for waste disposal
and septic tank overflow. In 1960, the Idaho Depart-
151
-------
Proceedings of National RCWP Symposium, 1992
ment of Health and Welfare (IDHW) Division of En-
vironmental Quality (DEQ) identified serious public
health problems in Rock Creek (Idaho Dep. Health,
1960). As a result of the National Pollution Dis-
charge Elimination System program, most direct dis-
charges into the creek had been eliminated by 1980
(Yankey et al. 1991). Nevertheless, Rock Creek was
still carrying heavy loads of sediment and agricul-
tural pollutants, which entered the stream primarily
through irrigation return flows as evidenced by com-
paring water quality tests made during the irrigation
and nonirrigation seasons (Yankey et al. 1991). Ir-
rigation return flows were the largest contributor to
the impairment of Rock Creek's beneficial uses—
namely, cold water biota, salmonid spawning, and
recreation.
Because of continuing water quality problems,
the Idaho Agricultural Pollution Abatement Plan
listed Rock Creek as a top priority stream in 1979
(Idaho Dep. Health Welfare, 1979a). Subsequently,
the Snake River and Twin Falls Soil Conservation
Districts received a Section 208 grant to develop a
detailed water pollution abatement plan for Rock
Creek (Idaho Dep. Health Welfare, 1979a). In 1980,
Rock Creek became one of 21 RCWP projects in the
United States. Goals of the project included im-
plementing BMPs to reduce the amount of sediment,
sediment-related pollutants, and animal waste enter-
ing Rock Creek.
Study Area
Rock Creek is located in Cassia and Twin Falls
Counties in south-central Idaho within the Snake
River Basin/High Desert Ecoregion. Its headwaters
are in the Sawtooth National Forest in western Cas-
sia County. Rock Creek flows northwest ap-
proximately 67.6 km (42 miles) through Twin Falls
County to the Snake River confluence north of the
city of Twin Falls. About 25 miles of Rock Creek
were involved in the project area (Fig. 1).
Soils in the lower watershed are generally thin,
light-colored, medium-textured surface soils and
very strong calcareous, silty subsoils. These highly
productive but highly erosive soils vary in total
depth (from 2.5 to 212 cm) and are underlain by frac-
tured basalt. Soils lie on gently sloping plains, with
slopes averaging between 1 and 2 percent (Yankey et
al. 1991).
The climate is semiarid with moderately cold
winters (mean low of -3'C in January) and warm
summers (mean high of 23°C in July). Annual
precipitation averages about 23 cm (9 inches). Ex-
cept in unusual years, precipitation runoff appears to
be a minor factor in the annual sediment loading to
the stream. The area has a 120-day frost-free grow-
ing season (Yankey et al. 1991).
Rock Creek was historically fed by snowmelt in
the higher watershed and characterized by high
spring flows and low summer and fall streamflows.
Developing the area for irrigated agriculture has
greatly changed this historical flow pattern. Rock
Creek now follows an altered-flow regime. Irrigation
return flows and recent hydroelectric developments
have greatly increased summer streamflows. Maxi-
mum streamflows now occur in spring and fall.
The Rock Creek watershed covers a total of
490,046 ha (198,400 acres) of which 128,211 ha are
irrigated pasture and cropland containing 69,552 ha
that are designated critical areas. Approximately 350
farm units lie within the watershed. Basic crops are
dry beans, dry peas, sugar beets, corn, small grains,
potatoes, alfalfa, and pasture. Because of low annual
precipitation, all crops are irrigated. Irrigation water
is diverted from the Snake River at Milner Dam to in-
dividual farms through a network of canals and
laterals owned by the Twin Falls Canal Company.
Water is available to the farm tract at a constant flow
from about April 15 through October 15.
Methods
Land Treatment and Land Use
Information, including data on participants, was
entered annually into the land treatment database.
The Rock Creek RCWP project's original goal was to
contract 75 percent of the critical acres. A critical
acre was defined as land lying on a steep slope,
rented land, intensively cropped land, or land
without any irrigation improvements.
The Rock Creek RCWP project watershed was
divided into 10 subbasins, and each subbasin varied
in total area as well as percent of critical area (Table
1). In addition, the percent of critical area contracted
in each subbasin ranged from a low of 62 percent to a
high of 100 percent in subbasins 3 and 10, respec-
tively (Table 1). The overall project goal of 75 per-
cent of the critical area contracted was achieved
(74.7 percent, Table 1).
The U.S. Department of Agriculture (USDA)
Soil Conservation Service (SCS) is the lead agency
maintaining the land treatment database. Informa-
tion was collected and stored on all Rock Creek
RCWP project contracts by farm fields. Information
for each contracted field included type of BMP in-
stalled, year installed, last year operated and es-
timated benefited area of the BMP. The benefited
areas and dominance of BMPs will be discussed in
152
-------
G CHANDLER & T. MARET
Rock Creek
Stations
7] Sub basin
!I Boundaries
Subbasin
Stations
Figure 1.—Map of the Rock Creek RCWP project study area.
153
-------
Proceedings of National RCWP Symposium, 1992
Table 1.—Summary of total hectares and critical areas in the Rock Creek RCWP project (Yankey et al. 1991).
SUBBASIN
1
2
3
4
5
6
7
8
9
10
TOTAL
TOTAL AREA
(ha)
6,002.1
8,126.3
5,335.2
10,781.5
7,854.6
12,337.6
16,598.3
15,906.7
11,305.1
17,490.0
111,737.4
TOTAL CRITICAL
AREA (ha)
4,199.0
7,805.2
4,643.6
9,282.2
2,835.5
11,492.9
12,609.3
8,113.9
5,320.4
3,250.5
69,552.4
PERCENT CRITICAL
AREA
70.0
96.0
87.0
. 86.1
36.1
93.2
76.0
51.0
47.1
18.6
62.2
CRITICAL AREA
CONTRACTED (ha)
3,786.5
4,026.1
2,889.9
6,547.9
1,793.2 ,
9,324.2
8,422.7
7,360.6
4,391 .6
3,250.5
51,793.2
PERCENT CRITICAL
AREA CONTRACTED
90.2
51.6
62.2
70.5
63.2
81.1
66.8
90.7
82.5
100.0
74.7
the linkage of water quality to land use/land treat-
ment portion of this paper.
SCS also maintained a land use database by field
on the contracted farms. Information included field
size and crop type by year throughout the length of
the contract. The USDA Agricultural Stabilization
and Conservation Service (ASCS) has provided farm
unit cropping information for the noncontracted
years and nonparticipants. This information is used
in conjunction with crop and field slope to determine
an erosion index for each subbasin (Carter, 1984).
Water Quality Data
The Rock Creek RCWP project's water quality data
were collected and the database maintained on.
STORET by DEQ. Rock Creek had six monitoring
locations and monitoring stations that were placed
on irrigation drains to track changes in sediment
load and associated pollutants as close to their
sources and the BMPs as possible (Fig. 1). Some sta-
tions measured the source of water to the subbasins
(7-1, 5-1, 4-1, 44, 2-1, and 1-1); some stations
measured the input of the subbasins to Rock Creek
(10-1, 7-7, 74, 5-2, 4-2, 4-3, 2-5, 24, 2-3, 2-2, and 1-2);
and other stations were key intermediate sites (7-2,
7-3, and 7-6). Subbasin stations were usually sampled
weekly; Rock Creek station was sampled biweekly
during the irrigation season. Water quality
parameters of interest included
• total suspended sediment,
• volatile suspended sediment,
• total phosphorus,
• orthophosphate,
• total nitrogen,
• fecal coliform bacteria, and
• flow.
Data Analysis
Spooner et al. (1985) described three experimental
designs that can be used to document water quality
improvements from agricultural nonpoint source
control programs. The three experimental designs
are
• time trend analyses,
• above and below analyses, and
• the paired watershed design.
Time Trend Analyses
Water quality information has been collected over
the Rock Creek RCWP project's entire 10 years. The
project is, therefore, perfect for a time trend
analysis, which simply involves monitoring a site or
sites over an extended period to determine any
changes in water quality.
The disadvantage of this analysis is that sen-
sitivity is low unless related variables are measured
so that water quality improvements can be attributed
to land treatment measurements. The Rock Creek
project did measure related variables, including flow
and precipitation. Even though this area has been in
a drought situation for the last six years, water uses
have not changed; therefore, changes in water
quality reflect changes in land treatment of land use.
The null hypotheses that can be tested using
this design for the Rock Creek RCWP project are
• HOI: the concentrations, loads, or levels of
measured variables have not changed
over time with the level of BMP
implementation; and
• Ho2: seasonal mean (or median)
concentrations, loads, or levels of
154
-------
G. CHANDLER & T. MARET
measured variables have not changed
over time.
The first null hypothesis (Hoi) will be tested by
regression analysis of the concentration, load, or
level of the variable versus the BMP implementation
level. Ho2 has been tested using a seasonal Kendall
with WQHYDRO, a software package developed by
Eric Aroner to analyze water quality information
(Aroner, 1991).
Upstream-Downstream Design
This analysis involves sampling a flowing system
over time, above and below a potential impact. This
approach allows one to account for upstream inputs
to the monitoring site of interest. The Rock Creek
RCWP project had many upstream-downstream
paired sites in its original monitoring design (Fig. 1).
Station S-6 in Rock Creek is above most agricultural
nonpoint sources entering Rock Creek; station S-4 is
above the majority of the irrigation return flows; and
station S-2 is just below the highest concentration of
irrigation return flows. Subbasin station pairs set up
for this design include 1-1 vs. 1-2; 2-1 vs. 2-2; 4-4 vs. 4-
3; 4-1 vs. 4-2; 5-1 vs. 5-2; 7-1 vs. 7-2; and 7-2 vs. 7-3 or
7-6.
The appropriate null hypothesis to test with this
design in the Rock Creek RCWP is
• Ho: The difference between upstream and
downstream pollutant concentrations,
loads, or levels have not changed over
time as BMPs were implemented.
The information needed to test this hypothesis in-
cludes paired concentrations and loadings above and
below the potential nonpoint source over time for
both pre- and post-BMP periods. This design also as-
sumes that sampling is timed so that the same parcel
of water is sampled at the above and below sites.
This assumption may have been missing in the Rock
Creek study because of sampling regime.
To test this hypothesis, the differences between
paired samples must first be determined. Then, the
same tests used in the trend analyses can be used
with this data set. In cases in which not all the water
originates within the project area, this experimental
design allows trends to be established with more
certainty than the before and after design because of
the corrections for incoming irrigation water from
the Snake River.
Paired Watershed Design
This design is the most complex, has the most strin-
gent data requirements, and requires much coor-
dination between water quality and land treatment
personnel. The paired watershed design consists of
monitoring downstream from two or more agricul-
tural drainages where at least one drainage has BMP
implementation and at least one does not. This
design entails simultaneous monitoring of two
drainages in close proximity during a calibration and
post-treatment phase.
The major advantage of this design is that it con-
trols for meteorologic variability. Water quality im-
provements related to BMP implementation can be
documented within a much shorter time frame. In
addition, this design provides stronger statistical
evidence of the cause-effect relationship between
agricultural nonpoint source control efforts and
water quality changes.
A disadvantage of this design is that land treat-
ment and water quality personnel must coordinate
closely to match BMP implementation efforts with
monitoring and data analysis needs. This design is
most often used on individual fields or study plots.
As drainage area increases, this approach becomes
less effective.
Another disadvantage is that control basins do
not receive as much land treatment, thus reducing
water quality improvement throughout the project
area.
The hypothesis that can be tested using this
design with the Rock Creek RCWP is
• HO: There will be no difference in pollutant
concentrations, loadings, or levels over
time in subbasins 4 and 5 vs. subbasin 7..
Subbasins 4 and 5 combined have close to the same
amount of acres as subbasin 7. Throughout this
project, subbasin 7 received a greater amount of
BMP implementation (especially conservation til-
lage) than subbasins 4 and 5; therefore, the paired
watershed analysis will be used to determine dif-
ferences relative to BMP implementation levels.
Linear regressions of the water quality variables
of interest for the treatment versus the control water-
shed for the calibration and BMP implementation
periods will be performed. A student's t-test will be
performed to determine if the predicted treatment
watershed values at the mean control watershed con-
centration decrease over time.
A decrease in the predicted treatment watershed
values suggests a positive effect of BMPs on water
quality. This evidence of a cause-effect relationship is
stronger than any derived from the designs pre-
viously discussed because of greater control over the
complex meteorologic, hydrologic, and temporal fac-
tors.
155
-------
Proceedings of National RCWP Symposium, 1992
Water Quality Results
Trend Analysis
Water quality trend analysis for sediment concentra-
tion, sediment loading, phosphorus, and fecal coli-
form bacteria has been completed for all subbasin
stations and Rock Creek stations in the project area.
Because of the data's seasonality, a seasonal Kendall
was run using WQHYDRO (Aroner, 1991). Most
monitoring stations did show significant downward
trends at the 90 percent level (Table 2). Some
noteworthy upward trends, however, also appeared,
especially in the upper watershed with fecal coliform
bacteria. This maybe the result of overwintering cat-
tle in the upper basin during the later years of the
study.
Table 2.—Summary of the seasonal Kendall analyses
performed on the Rock Creek RCWP project data. NS
denotes not significant; (+) denotes a significant up-
ward trend; and (-) denotes a significant downward
trend. Level of significance is 90 percent.
TOTAL
SUSPENDED
STATION SEDIMENT
Rock Creek
S-2
S-3
S-4
S-5
S-6
Subbasins
1-1
1-2
2-1
2-2
2-3
2-4
2-5
4-1
4-2
4-3
4-4
5-1
5-2
7-1
7-3
7-4
7-6
7-7
10-1
H
H
(-)
(-)
NS
(-)
(-)
«
H
NS
(-)
NS
(-)
(-)
(-)
H
H
(-)
NS
NS
H
(-)
H
(+)
TOTAL
PHOS-
PHORUS
H
H
H
H
H
(-)
H
H
H
NS
H
H
H
H
H
H
H
(-)
H
NS
H
H
NS
«
FECAL
COLIFORM
BACTERIA
NS
H
H
«
W
(-)
H
NS
(-)
(-)
NS
(-)
(-)
H
H
H
NS
NS
NS
NS
NS
(-)
NS
(+)
FLOW
NS
NS
NS
NS
H
W
NS
NS
NS
NS
NS
NS
NS
(+)
(+)
NS
NS
NS
NS
NS
(+)
(+)
NS
NS
CJpstream-Downstream Design
The upstream-downstream design was straightfor-
ward for the Rock Creek stations but not for the sub-
basin stations. Station S-4 samples were paired with
samples from S-2 and used as background levels to
determine the reduction of sediment load to Rock
Creek from agricultural irrigation return flows (Fig.
2). Irrigation season loadings adjusted for upstream
inputs significantly decreased (p=0.05) between 1982
and 1990 at S-2.
This design was not so straightforward in the
subbasins because of the water management re-
gime. Lateral splits and runoff types are additional
factors that must be dealt with before any estimate of
background levels.
Linkage of Water Quality to
Land Treatment and Land Use
Water quality improvements result from changes in
the watershed. RCWP projects target reduction of
agricultural nonpoint source loads through BMP im-
plementation. Therefore, land use/land treatment
must be linked with water quality to determine BMP
effectiveness. Making this link is part of the Rock
Creek RCWP project's post-project analysis. The
results will make it possible to discuss overall BMP
effectiveness and may even provide enough informa-
tion to address individual BMPs. Determining in-
dividual BMP effectiveness, given factors such as
soil type, land use, irrigation type, and BMP imple-
mented, should be the ultimate goal of any agricul-
tural pollution abatement program.
The complexities of linking the two databases
are immense. Water quality and land use/land treat-
ment information differ temporally and spatially. In
addition, many external interferences had to be
documented during the project, such as changes in
water management and BMP maintenance.
Temporal Difference
The difference between the land use/land treatment
and water quality databases is simply that land
use/land treatment information is annual (one
record per year) and water quality information has
multiple records per year, depending on the number
of samples collected. The Rock Creek RCWP project
replicated the land use/land treatment information
for each year to equal the number of water quality
samples taken. This does not read into the statistical
procedures as multiple BMPs implemented for that
year; rather, it indicates that when water quality
sample "i" was collected, BMP "j" was implemented,
and crop "k" was on the field. Taking this approach
helped ensure that no data were lost or replicated.
Spatial Differences
Spatial considerations of linking the two databases
are even more complex because water management
156
-------
Q. CHANDLER & T. MARET
x
M
100001
8000
S-2
S-4
(S-2) -211,378 KG/YEAR
(S-4) -27,216 KG/YEAR
5678567878567856785678567856785678
I 82 I 83 I 84 I 85 I 86 I 87 I 88 I 89 I 90 I
8000
6000-
4000
2000
o-
B
(S-2)-(S-4)
SLOPE • -184,162 KG/YEAR
I I 1 II II I I I I
1 I I I I I I I I I I 1 1 i L I
5678567878567856785678567856785678
I 82 I 83 l84l 85 I 86 I 87 I 88 I 89 I 90 1
YEAR
Figure 2.—Seasonal Kendall statistics for (a) Rock Creek stations S-2 and S-4 and (b) Rock Creek station (S-2)-(S-4).
is already complex and fields were often distant from that flows in each direction. Some instances occur in
pertinent monitoring sites. Water management is which a lateral splits below a monitoring site so that
confounded by lateral splits and the amount of water the water from it ends up in two or more monitoring
157
-------
Proceedings of National RCWP Symposium, 1992
sites. This split is the reason that the upstream-
downstream design for water quality analysis is not
straightforward for the subbasins.
The land treatment database is set up such that
each field has a designated nearest monitoring or ir-
rigation control site above and nearest monitoring or
control site below. Irrigation control sites are the
lateral splits. The Rock Creek project has obtained
the flows at the splits for the years of the study;
therefore, each split can be assigned a proportion
denoting the volume of water and the direction it
flows. Ultimately, each monitoring site was mapped
to ensure that the watershed's hydrology and its ef-
fect on each monitoring site was accurate (Fig. 3).
Rock Creek
Low
Line
CAnal
High Line Canal
• Monitoring Site
•" Lateral Split
Figure 3.—Schematic of subbasln 7 showing lateral
splits In relation to monitoring stations.
Another spatial aspect of the data is the distance
a field is from the nearest monitoring site it affects.
This aspect is dealt with in the database as type of
runoff. The four defined categories of runoff include
• direct,
• lateral reapplied,
• reapplied cropland, and
• reapplied pastureland.
Direct runoff denotes runoff that directly affects
the monitoring site; no opportunity exists to re-use
the water. Lateral reapplied runoff is water that is
directed back into a lateral which has opportunity for
reapplication farther downstream. Reapplied crop-
land runoff is water that is directly placed on another
field that is in crop production. Lastly, reapplied pas-
tureland runoff is the same as reapplied cropland
runoff except that the receiving land is permanent
pasture. Each field, depending on its runoff type, is
weighted 1.00, 0.50, 0.35, or 0.10 for direct, lateral
reapplied, reapplied cropland, and reapplied pas-
tureland runoff, respectively. This procedure is an in-
direct way to solve the proximity problem because
the closer a field is to a monitoring site, the less op-
portunity there is for reapplication; therefore, the
more likely it is that the runoff will be direct.
External Interferences
External interferences were documented through-
out the study to help explain discrepancies or
anomalies in the water quality. The water quality data
should not be used as a tool to look back at land
treatment information; it needs to be used the other
way around. Interferences — such as improper
management of BMPs, changes in irrigation water
management, or chaining the canals to remove algal
growth — have profound effects on water quality.
The Rock Creek land use/land treatment database
accounts for this type of information but, in many
cases, the data were only documented in field notes.
This information must be incorporated into the
database before its results are used to explain
variability.
BMP Effectiveness
In the past, improvements in water quality have been
assumed to be the result of land treatment practices.
The postevaluation phase of the Rock Creek RCWP
project will attempt to document water quality im-
provements resulting from implementation of BMPs.
BMPs have been shown to be beneficial on small
plots but little has been done to document BMP ef-
fectiveness on large-scale agricultural projects, such
as the Rock Creek RCWP project. The Rock Creek
project showed significant trends in BMP implemen-
tation that resulted in significant increases in water
quality (Yankey et al. 1991).
BMP effectiveness is also compounded by multi-
ple BMPs benefiting the same field or one BMP
benefiting multiple fields. The latter is particularly
the case with community-sized sediment ponds. Al-
though constructed at one site, the area benefiting
from this type of practice is much greater than the
field on which it is constructed. SCS at Twin Falls
has created a dominant BMP practice list to deal
158
-------
G. CHANDLER & T. MARET
Table 3.—Dominant best management practices incorporated in the Rock Creek RCWP project. The table is read
by reading the vertical entry first (i.e., Filter Strips < Ponds). DOM means practices dominant over all other prac-
tices; < means practices that are less effective than others; and NA denotes not applicable.
Ponds
Mini basins
Slots
Buried pipe
Filter strips
Irrigation
practices
Tillage
Sprinklers
PONDS
—
NA
NA
NA
DOM
—
— •
—
MINI
BASINS
NA
—
NA
NA
NA
—
.—
—
SLOTS
NA
NA
—
NA
NA
—
—
—
BURIED
PIPE
NA
NA
NA .
—
NA
—
—
—
FILTER
STRIPS
- <
NA
NA
NA
—
— .
—
—
IRRIGATION
PRACTICES TILLAGE SPRINKLERS
< < DOM
< < DOM
< < DOM
< < DOM
< DOM DOM
— DOM DOM
— — .DOM
— — —
with this type of problem (Table 3). This list is incor-
porated into the statistical procedure, ensuring that
the more dominant practice is used in the analysis.
Erosion Index
Another way to link land use/land treatment with
water quality variables is by using the erosion index
(Carter, 1984). The erosion index considers crop
type and land slope. Slope information is available as
an average per basin. Cropping information is avail-
able for the participants by field and for nonpar-
ticipants by farm unit. Plots will be developed for
each monitoring site to indicate its percent of acres
treated (x-axis), water quality parameter of interest
(y-axis), and erosion index (z-axis). This approach
may be the best way to incorporate crop type into the
analysis.
Lessons Learned
The Rock Creek RCWP, like every project, offers
many recommendations that can make future
projects smoother/The following are some of these
recommendations.
1. Interagency cooperation is essential for
proper data interpretation. Expertise from
both land use/land treatment and water
quality professionals is a must for a quality
end product.
2. A database and experimental design should
be part of the project proposal; they should
not be done as the project progresses. Both
land treatment and water quality monitoring
designs must be clearly defined, consistent,
and sufficient to meet project goals.
3. The watershed's hydrology should be
mapped before monitoring sites are deter-
mined.
4. Monitoring plans should be clearly written to
define objectives and agency roles.
5. Monitoring schemes and land treatment
tracking should be set up by hydrologic units
to facilitate evaluation of BMP effectiveness.
6. External interferences that may mask
benefits from implemented BMPs must be
documented throughout the project period.
7. Land treatment databases should include
nonparticipant data when water quality and
BMPs are being evaluated at the watershed
level.
Although broad, these recommendations may
be overlooked when trying to start a project. Prelimi-
nary tasks, such as mapping the hydrology, setting
up the relational database, and determining the
proper experimental design at the project's begin-
ning, will save many headaches at its end.
ACKNOWLEDGMENTS: The authors thank Ron
Blake, SCS, for his help throughout the postevaluation phase
with the land use/land treatment database, mapping of the
subbasins, and documentation of external interferences. Also,
we thank Don Zaroban and Bill Clark, DEQ, for their help on
various aspects of this project We thank Jean Spooner, North
Carolina State University Water Quality Group, for her con-
tinued assistance with the land use/land treatment database
and statistical analysis of this project Lastly, we thank three
anonymous reviewers for their comments.
References
Aroner, E. 1991. WQHYDRO (Water Quality/Hydrology
Graphics/Analysis System) Interim Guidance/User's
Manual/Documentation. Unpubl. Olympia, WA.
Carter, D. 1984. Rock Creek Rural Clean Water Program intensive
monitoring project, report of ARS activities for 1984. Agric.
Res. Serv., U.S. Dep. Agric., ffimberly, ID.
Idaho Department of Health. 1960. Report on Pollution Problems
in Rock Creek Cassia and Twin Falls Counties, Idaho, 1959.
Mimeo Report. Boise.
159
-------
Proceedings of National RCWP Symposium, 1992
Idaho Department of Health and Welfare. 1979a. Idaho Agricul-
tural Pollution Abatement Plan. Div. Environ. Qua!., Boise.
. 1979b. Application for Rural Clean Water Program Funds,
Rock Creek Twin Falls County, Idaho. Div. Environ. Qual.,
Boise.
Spooner, J. et al. 1985. Appropriate designs for documenting water
quality improvements from agricultural NFS control
programs. Pages 30-34 i» Perspectives on Nonpoint Source
Pollution, Proc. Natl. Conf. U.S. Environ. Prot Agency,
Washington, DC.
Yankey, R. etal. 1991. Rock Creek Rural Clean Water Program Ten
Year Report Interagency rep. in coop. Agric. Stabil. Conserv.
Serv., U.S. Dep. Agric., Soil Conserv. Serv., Idaho Div. En-
viron. Qual., and Twin Falls and, Snake .River Soil Conserva-
tion Distr., Boise.
160
-------
Effectiveness of Agricultural Best
Management Practices Implemented
in the Taylor Greek/Nubbin Slough
Watershed and the Lower Kissimmee
River Basin
Boyd Gunsalus, Eric G. Flaig, and Gary Ritter
South Florida Water Management District
West Palm Beach, Florida
ABSTRACT
During the past 20 years three generations of best management practices (BMPs), including those
associated with the Rural Clean Water Program (RCWP), have been implemented in the Taylor
Creek/Nubbin Slough and the Lower Kissimmee River basins. Over the last four years intensive
animal waste collection and nutrient recycling systems have been installed at each dairy to reduce
phosphorus loads to Lake Okeechobee from dairy and beef pasture operations. Water quality grab
samples have been collected biweekly at basin outlet structures and tributary sites for the last 10 to
15 years. Grab samples have been collected on dairies since 1986. The samples show a significant
decreasing trend in the median total phosphorus concentration at most tributaries in Taylor
Creek/Nubbin Slough basin where there is a high level of participation in the RCWP program. The
total phosphorus concentration values decreased by an average of 60 percent at the tributaries and
by 50 percent at 75 percent of the dairies. Median total phosphorus concentration values decreased
by 84 percent and phosphorus load by 40 percent for the Taylor Creek/Nubbin Slough basin. How-
ever, the phosphorus load in the Lower Kissimmee River basin increased significantly at the S154
structure.
The South Florida Water Management Dis-
trict participates in two Rural Clean Water
Programs (RCWPs) within the Okeechobee
basin north of Lake Okeechobee (Fig. 1). Lake
Okeechobee is in danger of becoming hyper-
eutrophic because of nutrient inputs, primarily phos-
phorus from agricultural activities (Federico et al.
1981). Historically, this area has contributed nearly
50 percent of the phosphorus load to the lake. The
RCWP was implemented to reduce phosphorus in-
puts to the lake and determine the effectiveness of
agricultural best management practices (BMPs) for
improving water quality. This paper describes three
generations of BMPs implemented over the last
twenty years, and evaluates the changes in water
quality as they relate to land treatment within the
Taylor Creek/Nubbin Slough and Lower Kissimmee
River basins.
The RCWP was initiated in Taylor Creek/Nub-
' bin Slough basin in 1981 and in the Lower Kissim-
mee River basin in 1987. The goal of the Taylor
Creek/Nubbin Slough RCWP was to reduce the
161
-------
Proceedings of National RCWP Symposium, 1992
LOWER KISSIMMEE
RIVER BASIN
TAYLDR CREEK
NUBBIN SLDUGH
BASIN
LAKE
DKEECHDBEE
DAIRY MONITORING SITE
TRIBUTARY MONITORING SITE
DAIRIES
BASIN BDUNDRIES
Figure 1.—Taylor Creek/Nubbin Slough and Lower Kissimee River basins and water quality monitoring stations.
phosphorus concentrations entering Lake
Okeechobee by 50 percent by 1992. The goal of the
Lower Kissimmee River RCWP was to reduce phos-
phorus loads to Lake Okeechobee by 43 percent. In
1987, the Florida Department of Environmental
Regulation promulgated a "dairy rule" requiring that
all dairy operations within the Taylor Creek/Nubbin
Slough and Lower Kissimmee River basins imple-
ment a specific set of BMPs. The dairy rule conser-
vation plans provided for collection and treatment of
surface water runoff from the barn and high inten-
sity areas and for recycling nutrients within the farm.
Concurrent with the dairy rule, the dairy industry
requested that a buy-out program be implemented
for those dairies that either chose not to comply or
could not comply with the dairy rule. This program
162
-------
6. GC/ttSALaS, E.G. FLAK?, 6 G. RITTER
was adopted and dairy owners were offered $602 per
cow. In return, a deed restriction prohibited the
properly from being used for any future dairy farm-
ing practices associated with the milking operation.
The dairy proprietor retained ownership of the cows
and the property but had to remove the cows from
the property after the buy-out contract was signed.
Eighteen dairies participated in the buy-out pro-
gram, removing approximately 14,000 milking cows
from the basins.
In this paper, the changes in surface water phos-
phorus concentrations (total phosphorus concentra-
tion) and total phosphorus loads following
implementation of BMPs will be analyzed. First, the
trends in total phosphorus concentrations for the
tributaries will be presented. Second, total phos-
phorus concentration at the dairy sites will be
presented to determine the changes resulting from
the dairy rule and implementation of the buy-out pro-
gram. Finally, the phosphorus loads and concentra-
tions at the discharge structures to Lake Okee-
chobee will be presented.
Description of Project Area
The Taylor Creek/Nubbin Slough basin consists of
several tributaries connected by canals to drain to
the lake through a gated structure, S191. The Lower
Kissimmee River basin consists of tributaries west of
Okeechobee City that are drained to the lake
through structure S154 and the channelized Kissim-
mee River, C-38, which drains to the Lake through
structure S65E (Fig. 1). The lower two pools of C-38
are included in the Lower Kissimmee River basin
RCWP. This area is characterized by flat landscape,
sandy soils, and poor off-site drainage. The average
annual rainfall for the project area is 127 cm; the
majority of rainfall occurs during the summer wet
season supplemented by occasional tropical storms.
The typical soils of the region are sandy spodosols
that have a low phosphorus holding capacity (Blue,
1970). The combination of rainfall in excess of
evapotranspiration and low phosphorus retention
capacity creates a high potential for phosphorus
transport within the Lake Okeechobee basin (Flaig
and Ritter, 1989).
The largest sources of net phosphorus imports
to the basin are dairy operations and improved beef
pasture operations (Fonyo et al. 1991). Over 95 per-
cent of the project area in the Taylor Creek/Nubbin
Slough and Lower Kissimmee River basins is cur-
rently devoted to agricultural activities. Most of the
improved pasture has been fertilized for increased
forage production and ditched to improve runoff
during the summer wet season. Dairy cows, con-
tained on large dairies of 800 to 1,200 animals each,
tend to congregate in holding areas adjacent to the
milking barn. There were 24 dairy barns in the
Taylor Creek/Nubbin Slough basin and 19 in the
Lower Kissimmee River basin; a total of 23,000 and
25,000 dairy and beef cows, respectively, in the
Taylor Creek/Nubbin Slough basin and 12,700 and
68,000 dairy and beef cows, respectively, in the
Lower Kissimmee River basin. Both of these land
use practices (dairy and beef operations) have great
potential to produce nonpoint sources of poor water
quality (Flaig and Ritter, 1989). Nutrient transport in
the area is governed by intense, localized rainfall and
poorly drained soils.
Best Management Practices
Three generations of BMPs have been implemented
in the Taylor Creek/Nubbin Slough and Lower Kis-
simmee River basins. The following criteria were
used to categorize the BMP generations:
• BMPs implemented before the RCWP;
• BMPs implemented during the RCWP; and
• Dairy rule plan BMPs.
The first generation, 1972 to 1978, consisted of
dairy barn lagoons that primarily trapped solids, t-
These BMPs were the first attempt to eliminate barn
wash from the water in adjacent tributaries. Unfor-
tunately, the lagoons were poorly managed, resulting
in occasional leaks or in some instances continual
lagoon discharge into adjacent tributaries. Phos-
phorus concentrations observed in lagoons ranged
from 20 to 40 mg/L (Goldstein, 1986).
The second generation, 1978 to 1986, included
BMPs implemented in the RCWP project. Second
generation BMPs included several practices. Stream
fencing and stream crossings were used to restrict
dairy and beef cow access to adjacent waterways and
provide a vegetative filter strip. Second-stage
lagoons were altered to provide recycled wastewater
for the dairy barn washwater. Pasture management
was improved by the introduction of portable shade
structures to relieve heat stress and provide better
use of pastures. BMPs implemented during this
generation required an active maintenance program
to ensure their effectiveness for attenuating
nutrients. Problems associated with the first genera-
tion were also experienced in the second generation,
including poor maintenance of lagoons and seepage
fields, which resulted in the degradation of off-site
water quality.
The third generation BMPs, 1987 to the present,
were developed in response to the dairy rule. Three
typical management strategies were used in dairy
163
-------
Proceedings of National RCWP Symposium, 1992
rule conservation plans. The first approach used a
perimeter ditch around the barn and high-intensity
areas to capture surface water runoff and divert it to
a waste storage pond. The wastewater was spray ir-
rigated on silage crops for nutrient recovery.
Animals were allowed to graze and feed in the pas-
tures outside the high-intensity areas. The second
approach was similar, except that the animals were
fed, watered, and shaded in the high-intensity areas.
This technique was known as "dry lot confinement,"
and required a high degree of management to
properly maintain the high-intensity areas. The
manure from the area had to be removed and either
sold or spread on appropriate pasture lands. The
third technique was a total confinement facility that
housed the milk herd in a free stall barn. The
animals were fed, watered, and cooled within the
barn. The barns were flushed with water two to
three times per day, and the water was then routed
to the first stage lagoon. The success of the confine-
ment dairy relied largely on management of the barn
facility.
Analysis of Water Quality Data
Water quality data have been collected in the Taylor
Creek/Nubbin Slough basin since the early 1970s
(Osking and Gunsalus, 1992). In 1972, a biweekly
water sampling program was established to deter-
mine phosphorus loads to the lake at all inflow struc-
tures, including S191, S154, and S65E (Fig. 1). Since
the late 1970s, monitoring in the Taylor Creek/Nub-
bin Slough basin focused on quantifying long-term
trends for the nine major tributaries. Water quality
samples were collected and analyzed for nitrogen
and phosphorus, pH, specific conductivity, turbidity,
and color.
Following implementation of the third genera-
tion BMPs, runoff water quality from each dairy was
monitored. Beginning in 1986, grab samples were
collected weekly at each dairy to provide data on pre-
and post-dairy rule conservation plan water quality.
Monitoring was conducted on each dairy in the both
Taylor Creek/Nubbin Slough and Lower Kissimmee
River basins. At a few dairies, all runoff was
monitored at a single site. At most dairies, the dis-
charge site with the poorest water quality was
monitored; other source areas on the dairy were not
monitored. A synoptic monitoring program was
developed to identify the impacts of those other
source areas not covered in the routine monitoring
program (Sawka et al. 1992). Water quality monitor-
ing also began at selected tributary sites in the
Lower Kissimmee River basin in 1986.
Water Quality Analysis
Three data sets were developed for evaluation of the
effectiveness of BMPs: dairy runoff total phos-
phorus concentration; ambient tributary total phos-
phorus concentration; and total phosphorus
concentration and total phosphorus load data col-
lected at outfall structures draining to the lake. Table
1 lists the complete set of statistical analyses per-
formed. The dairy data were used to evaluate the
site specific effects of the dairy rule conservation
plan BMPs. The tributary data were used to deter-
mine the long-term effects of dairy and beef pasture
BMPs. The total phosphorus values from biweekly
grab samples at the structures were used to evaluate
water quality, particularly loads, for each basin.
Tributary Water Quality
The tributary water quality data were analyzed to
determine the changes in water quality resulting
from the implementation of BMPs in the critical
areas of each subbasin. Nine tributary sites in the
Table 1.—Statistical analysis used for total phosphorus concentrations (TPC) and total phosphorus load data
(TPL) for dairies, tributaries, and structures in the Taylor Creek/Nubbin Slough (TCNS) and Lower Kissimmee
River (LKR) basins.
DATA SET
TCNS dairies
LKR dairies
TCNS tributary
LKR tributary
Structure
S-191
S-65E
S-154
Structure
S-191
S-65E
S-154
PARAMETER
TPC
TPC
TPC
TPC
TPC
TPC
TPC
TPL
TPL
TPL
TEST
Wilcoxon Sum Rank/Satterthwaites T-test
Wilcoxon Sum Rank/Satterthwaites T-test
Seasonal Kendall Tau
Seasonal Kendall Tau
Seasonal Kendall Tau
Seasonal Kendall Tau
Seasonal Kendall Tau
PRE-RCWP
ANOVA 1977-81
ANOVA 1973-80
ANOVA 1973-80
RCWP
1982-85
1980-88
1980-88
TIME PERIOD
1988-92
1086-92
1978-92
1986-92
1978-92
1978-92
1978-92
PRE-DRP POST-DRP
1985-88 1988-92
n/a 1988-92
n/a 1988-92
Note: Pre-RCWP « prior to Implementation of BMPs; RCWP = Implementation of BMPs in TCNS and LRK basins; Pre-DRP (dairy rule plan).
post-RCWP BMPs, coincides with pre-DRP on dairies; Post-DRP • average time frame for post-DRP construction.
— ; -
-------
B. GCJNSALUS, E.Q. FLAIG, & G. RITTER
Taylor Creek/Nubbin Slough basin and three
tributaries in the Lower Kissimmee River basin were
evaluated. The seasonal Kendall Tau test for trend
(Hirsch et al. 1984) was performed on the total phos-
phorus concentration time series from each
tributary. The Kendall Tau test was selected because
it could detect the gradual change in total phos-
phorus concentration values in the presence of
anomalous data.
. The implementation of BMPs occurred over
eight years, distributed spatially throughout the
basin. The change in water chemistry due to adjust-
ments in stream sediments was also gradual. Conse-
quently, the changes in total phosphorus concen-
tration at the tributary scale were gradual. Where
the data are skewed and exhibit seasonality the Ken-
dall Tau test was most powerful (Hirsch et al. 1984).
For the Kendall Tau test the data were grouped into
12 monthly seasons and the monthly median values
were used in the analysis. Taking monthly medians
reduced the effect of serial correlation among the
data and the problems associated with missing data.
However, the exact power of this test when unequal
monthly sample sizes occurred is not .known (Gil-
bert, 1987). Insufficient information? was available at
,each site, to adjust these data for hydrologic
variability such as stream stage, discharge, rainfall,
or groundwater stage. Hirsch et al. (1984) concluded
that the seasonal Kendall Tau is appropriate for both
adjusted and unadjusted time series. With only five
years on record, the significance levels of the test are
less certain. Biweekly sampling historically, and
weekly sampling since 1986, have provided sufficient
data for this analysis.
The Tau test was conducted to determine
homogeneity of trend among months using the Bat-
telle Northwest's Trend program (Gilbert, 1987).
When the trend is heterogeneous among seasons,
the overall trend test is misleading.
Dairy Water Quality Analysis
The dairy water quality data contained weekly grab
sample .values collected from 1987 to 1992. The total
phosphorus concentration data were partitioned into
pre- and post-construction data sets. The frequency
distributions from these data were positively skewed
and the data were log-transformed, which produced
approximately normal distribution values. Bradley
(1968) indicated that determining the degree of
departure from normality is difficult and assessing
the loss of power is tedious and unsatisfactory. The
hypothesis that there were no significant differences
between the two periods was tested using the Wil-
coxon sum rank test and the t-test. Montgomery and
Loftis (1987) showed that both tests are relatively in-
sensitive to the shape of the distribution. However,
they are both sensitive to serial correlation among
observations. Rekhow and Chapra (1983) recom-
mended that both tests be applied when there is
doubt about the assumptions.
The Wilcoxon rank sum test was used to test no
differences in populations. The test is robust to out-
liers but not as powerful as a t-test when the data are
gaussian. This test was used to test the null
hypothesis that the means were not significantly dif-
ferent. Satterthwaite's t-test is adjusted for unequal
sample sizes,and unequal variance (Snedecor and
Coehran, 1980), which was the case in these com-
parisons. These analyses were conducted using SAS
Univariate and t-test procedures (SASInst. 1990).
Basin Water Quality Analysis
Total phosphorus concentration time series and
flow-weighted monthly total phosphorus concentra-
tions were analyzed at structures S154, S65E, and
S191, which control discharge to Lake Okeechobee
from the Taylor Creek/Nubbin Slough basin and
Lower Kissimmee River basins. Two analyses were
performed on the'phosphorus data, the seasonal
Kendall Tau test and analysis of variance. The Ken-
dall Tau was applied to the raw total phosphorus con-
centration data and the analysis of variance
(ANOVA) was used to evaluate significant differen-
ces in loads among the different BMP implementa-
tion periods.
The data used in this analysis were biweekly
total phosphorus concentration values from grab
samples and estimated daily discharge. Monthly
flow-weighted total phosphorus concentration values
were calculated to provide total phosphorus level
normalized for discharge. Although these data are
not sufficiently dense to calculate accurate loads,
they do provide a consistent index for .analysis of
trends and differences between time periods. Com-
pared with daily composite data, these loads may dif-
fer by as much as 40 percent.
The total phosphorus load .data were positively
skewed and effectively normalized by the logarith-
mic transformation. The monthly data were tested
for homogeneity of variance using Bartjett's test
(Reckhow and Chapra, 1983). As expected, a sig-
nificant difference in variance resulted from
seasonality. When the data were partitioned into dry
season (November to April) and wet season (April to
October), the resulting data were homogeneous.
Simple linear regression, log-total phosphorus con-
centration regressed against log-discharge, was per-
formed to account for possible long-term hydrologic
variability simultaneously within ANOVA for compar-
165
-------
Proceedings of National RCWP Symposium, 1992
ing total phosphorus load values for each BMP
period. The residuals were checked for deviations
from normality using the W-test and tested for
autocorrelation using the Durbin-Watson test (Reck-
how and Chapra, 1983). The residuals were ap-
proximately normal and exhibited no auto-
correlation.
The specific test time periods are defined in
Table 1. The resulting mean values for total phos-
phorus load for each BMP period were compared as
multiple treatment effects using Ryan's Q-test (Einot
and Gabriel, 1975). Day and Quinn (1989) recom-
mended Ryan's Q-test for samples of equal variance
and unequal sample size. Ryan's test is powerful and
controls type I error rates for multiple comparisons.
The mean total phosphorus load values were calcu-
lated using the minimum variance unbiased es-
timator (Gilbert, 1987) of total phosphorus load and
the mean flow by season and site. These analyses
were completed using the SAS General Linear
Models procedure (SAS Inst 1990) on a SUN
SPARC1+ work station.
Results of Water Quality
Analysis
Tributary Water Quality
The results of the analyses indicated that significant
decreasing trends in most subbasins (Table 2). The
larger Tau values indicated a steeper, more sig-
nificant trend. The significance level is adjusted for
serial correlation (Hirsch et al. 1984). The monthly
trends were found to be homogeneous at all sites.
The percent change in period of record median total
phosphorus concentration is provided for com-
parison.
Over the course of BMP implementation, the
total phosphorus concentration has been consider-
ably reduced in the Taylor Creek/Nubbin Slough
basin. Overall, capture of animal waste with the dairy
rule conservation plan and improved pasture
management greatly improved total phosphorus con-
centration values. No significant trend was found at
Northwest Taylor Creek (site 01), probably because
of the increase in animal density throughout the
RCWP period and the low percent of critical land im-
pacted by BMPs. The increasing trend at Lettuce
Creek (site 40) probably resulted from a low rate of
land owner participation in the RCWP program and
poor animal waste handling.
The specific effects of BMP implementation on
total phosphorus concentration are illustrated for
Mosquito Creek at site 13 (Fig. 4). An improvement
in pasture management in the early 1980s was fol-
lowed by an increase in animal density in 1984. One
dairy near site 13 participated in the buy-out pro-
gram in 1988, which substantially improved down-
stream water quality. Although total phosphorus
concentrations improved slightly as a result of the
RCWP, the buy-out and dairy rule conservation plan
on adjacent dairies was most effective.
In the Lower Kissimmee River basin, total phos-
phorus concentration values are lower than in the
Taylor Creek/Nubbin Slough basin, reflecting the
lower intensity of animal management. The increas-
ing trend at Chandler Slough (04A) reflects the
presence of one dairy and lack of BMP implementa-
tion for improved pasture. Two dairies were located
in the Cypress Slough (06A) basin trend, one of
which has converted to a beef operation through the
buy-out program. No significant trend was found at
Table 2.—Seasonal Kendall Tau total phosphorus concentration trends for tributary sites in the Lower Kissim-
mee River and Taylor Creek/Nubbin Slough basins.
LOCATION
Chandler SI.
Cypress SI.
Yates Marsh
NW Taylor Ck.
Little Strain!
Otter Ck.
Taylor Ck.
Will. Ditch
Mosquito Ck.
Nubbin SI.
Henry Ck.
Lettuce Ck.
SITE
NUMBER
04A
06A
17A
01
02
06
18
09
13
14
39
40
LENGTH OF
RECORD
(YEARS)
5
5
5
13
13
13
12
14
14
14
10
10
PERCENT
LAND
TREATED
1
8
10
68
100
97
85
99
89
84
57
20
TAU
.52
-.52
-.38
.12
-.23
-.64
-.36
-.34
-.61
-.46
-.45
.47
CC
.05
.07
.17
.24
.13
.001
.007
.008
.001
.007
.015
.006
MEDIAN
SLOPE
.044
-.029
-.024
.009
-.042
-.187
-.046
-.012
-.158
-.092
-.133
.021
PERCENT
CHANGE
+120
-60
-87
-26
-60
-220
-90
-66
-140
-86
-92
+84
166
-------
6. GUHSALUS, E.G. FLAIG, & G. RITTER
Yates Marsh (17A), which drains a substantial wet-
land and one small dairy. The trends for Lower Kis-
simmee River basin sites should be interpreted with
caution because of the short period of record. •
Results of Dairy Analysis
The results of the analysis for total phosphorus con-
centration in dairy runoff are summarized in Figure
2. The results at individual dairies have been com-
bined to illustrate the effectiveness of the overall
dairy rule conservation plan.
Pre-DRP
Post-DRP
#
Dairies
|| TCNS
10
8
6
4
2
n F
m \ [S
«*
v<
V
Fl
j i
;,'•'
-
—
t ~,t 1-1^1 »
|
,-|
r-i
n 11
<1.2 1.2-5 5-10 10 <
<1.2 1.2-5 5-10 10 <
Figure 2.—Number of dairies In Taylor Creek/Nubbin
Slough and Lower Kissimmee River basins with dis-
charge total phosphorus concentration in the range <
1.2 mg/Lto > 10 mg/L.
Results of both the Wilcoxon test and t-test indi-
cated highly significant differences between pre- and
post-dairy rule conservation plan at 25 dairies. The
other three sites either had too few data (N<6) for
comparison, or the total phosphorus concentration
values were small and no significant change was
detectable, or no significant improvement was found.
Implementation of the dairy rule conservation
plans has significantly improved water quality in
monitored discharge from 80 percent of the dairies.
Of the 30 remaining dairies, total phosphorus values
collected at the representative monitoring site on 16
dairies are below the regional 1.2 mg/L average total
phosphorus concentration target (Fig. 2). Prior to
the dairy rule conservation plan only 4 dairies met
the target. Of the dairies that have not met the tar-
get, six have significant site specific problems re-
lated to poorly drained pastures and high animal
densities. Several reasons account for this condition,
including (1) small high-intensity areas that provide
few incentives such as feed and shade to encourage
deposition within the waste management system; (2)
poorly drained swale areas through which manure
phosphorus is readily transported off site; and (3)
old wetland areas that historically received high
phosphorus loads and may be washing out excess
phosphorus. Additional BMPs may be necessary to
resolve those specific water quality problems. On the
other dairies, it appears that poor management is
responsible for poor water quality. As the dairy
managers become more adept at herd rotation,
animal waste redistribution, and application of pas-
ture buffer strips, these dairies should come into
compliance.
The effect of the dairy rule conservation plan on
off-site water quality depends on the degree to which
other BMPs have been installed on the dairy. In the
Taylor Creek/Nubbin Slough basin, the dairies had
installed second generation BMPs that improved
pasture runoff and reused barn washwater. Nearly
70 percent of the Taylor Creek/Nubbin Slough basin
dairies have reached the target total phosphorus
concentration at the representative monitoring site.
Before the dairy rule conservation plan, few BMPs
had been installed on dairies in the Lower Kissim-
mee River basin. Only 40 percent of the.Lower Kis-
simmee River basin dairies have reached the target,
and 40 percent of the dairies have problem areas.
Some of these problems reflect the lack of second
generation BMPs. Where pasture BMPs have been
installed, implementation of the dairy rule conserva-
tion plan produced an immediate improvement in
water quality.
One concern in analyzing these data was the ef-
fect of time lag on dairy rule conservation plan dis-
charge. Review of the data indicates .that those
dairies that effectively installed, the dairy rule con-
servation plan are meeting the appropriate standard.
Dairy rule conservation plan BMPs were completed
recently on three dairies. On these sites, one notes a
significant improvement in total phosphorus con-
centration but the values are still above the target.
The results of dairy rule conservation plan im-
plementation can be illustrated by reviewing one
case study. At one dairy,,Site 40, the milking herds
had free access to farm drainage ditches. After the
construction of fencing, additional cooling ponds,
and shade structures, the cows were restricted from
the ditches. The off-site water quality has significant-
ly improved, with average total phosphorus con-
centration of less than 1.2 mg/L (Fig. 3). The
primary effect of the dairy rule conservation plan
has been to redirect concentrated wastewater to
other land where the nutrients can be recovered.
Dairy Buy-out Program
The buy-out program was expected to produce a
phosphorus load reduction by reducing phos-
phorous imports and concentration of manure. Al-
though the land owners were restricted from dairy
167
-------
Proceedings of national RCWP Symposium, 1992
1MT
1848
1989 1990
Year
1991
Figure 3.—Total phosphorus concentration (TPC) time
aeries for dairy site 40.
and feedlot operations, intensive operations such as
heifer and dry cow operations were not prohibited.
For the 18 dairies participating in the program, most
of the land remains associated with intensive animal
management With the buy-out program 14,040 milk-
ing cows were relocated and replaced by 6,600
heifers or beef cows. Other changes in land use in-
clude citrus and vegetables.
The short-term results on total phosphorus con-
centration in runoff from these operations are;
presented in Figure 4 for each category of current
land use. The average runoff total phosphorus con-
centration results are presented by classes. The
lowest class, average total phosphorus concentration
less than 0.35 mg/L, represents the regulatory tar-
gets for these land uses. The second class, 0.35 to 1.2
mg/L, should include those operations making the
transition from dairy to improved pasture. The other
classes include the number of farms that have
problems meeting regional water quality targets.
Total phosphorus concentration values have sig-
nificantly improved at half the buy-out sites; how-
ever, the total phosphorus concentration values are
still unacceptably high, greater than 0.35 mg/L. At
t
Dairies
Pro-DRp
Rost-DRP
. - . .
I - 1 Heifers
6
5
4
3
2
1
m
p
-]
'.i
-,
r~1 Beef
p
'j
-
1 -ifl-f
-
[S Hay
i':m
*T
OS 3SV2
Figure 4.—Number of dairies by current land use In
Taylor Creek/Nubbin Slough basin and Lower Klsslm-
mee River basins participating In the buyout program
with total phosphorus concentration In the range < 0.35
mg/L to > 3 mg/L total phosphorus.
some farms, the runoff water quality has not im-
proved because the current heifer operations are as
intensive as the previous dairy. This total phos-
phorus concentration data are uncertain because
they reflect the short lapse in time since the change
in operations on some farms. The first cows were
removed from the early buy-out participants in fall
1989. The last participants have recently removed
their milk herds.
Basin Scale Water Quality
In the Taylor Creek/Nubbin Slough basin, which
drains to Lake Okeechobee at S191, there has been a
significant decrease in total phosphorus concentra-
tion and total phosphorus load following implemen-
tation of RCWP and dairy rule conservation plan
BMPs. The median concentration decreased 84 per-
cent between 1979 and 1992 (Table 3). The trend in
total phosphorus concentration at S191 increased
during the 1970s, but has decreased during the
1980s following implementation of BMPs (Fig. 6).
During the 1970s, dairy operations intensified and
additional milk herds were brought into the basin.
The trend in total phosphorus concentration had
decreased through the 1980s as BMPs were in-
stalled. The total phosphorus concentration time
series at each structure was marked by wet season
peaks; extended peaks and troughs indicate the oc-
currence of droughts and hurricanes.
Year
Figure 5.—Total phosphorus concentration (TPC) time
series and trend line (slope = -.0158 mg/L-yr) for
Mosquito Creek (Site 13).
The median total phosphorus concentration in-
creased'at S154 and "decreased slightly at S65E
within the Lower Kissimmee River basin. The trend
at S154 has increased significantly during the 1980s
following the establishment of two dairies that in-
creased the number of milk cows by 40 percent. A
slight decreasing trend at S65E during the project
was significant at the 0.10 level.
168
-------
B. GUNSALUS, E.G. FLAIG, & G. RITTER
2.00
1.75
1.50
O
•0.75
0.50
0.25
0.00
1973 1975 1977 1979 19B1 19S3 1985 1987 1989 1991 1993
Year
Figure 6.—Total phosphorus concentration (TPC) time
series and trend line (slope = -0.049 mg/L-yr) for struc-
ture S191, Taylor Creek/Nubbin Slough basin.
,The total phosphorus concentration data exhibit
strong seasonally due to high runoff during the late
summer wet season. High total phosphorus con-
centration values are usually associated with high
flow, but may also be associated with small intense
storms that produce significant flushes with low total
discharge.
The trend in total phosphorus concentration and
total phosphorus are a function of the area of BMP
implementation in each basin. A much greater por-
tion of the land in the S191 basin was involved in
BMPs. The percent of the land treated in each basin
varied from less than 1 percent in the S65E basin and
8 percent in S154, to 58 percent in the S191. The
BMP coverage in S65E and S65D basins was small
relative to the entire area.
The monthly phosphorus loads, expressed as
average daily loads, have varied greatly during the
period of record. The total phosphorus load values
were strongly influenced by basin discharge. Varia-
tion in discharge accounts .for 30 percent of total
phosphorus load at S154,11 percent at S65E, and 50
percent at S191. Most of the variation in total phos-
phorus load occurred within season and the longer-
term trend was less than 5 percent.
The phosphorus load at S191 decreased 40 per:
cent during the period 1979 to 1992 (Table 4). The
Total phosphorus load decrease from pre-RCWP to
post-dairy rule conservation plan was significant at
the .001 level, with a standard error of 26 kg per day.
A significant increase of 54 percent occurred in Total
phosphorus load at S154 during the RCWP period
(1980-88), coincident with increased dairy activity.
The standard error for total phosphorus load at S154
was 9 kg per day. Although there was no significant
change at S65E over the entire period, the total phos-
phorus load significantly increased during the mid-
1980s and is now decreasing.
General Discussion
The primary effect of the dairy rule conservation
plan has been to redirect concentrated wastewater to
other land where nutrients can be recovered. Ad-
vanced plans have been designed to capture greater
amounts of animal waste by total confinement or by
increasing the areal extent of the high-intensity
areas. The wastewater from the high-intensity areas
is more concentrated and easier to handle. Failures
of the dairy rule conservation plan have occurred so
areas of manure concentration exist outside the
high-intensity areas. Within the high-intensity areas
it has been assumed that the lagoons and waste
storage ponds will seal to prevent groundwater pollu-
tion. Although there is some data indicating the
presence of elevated phosphorus concentrations in
groundwater, nothing indicates that groundwater is a
source of off-site contamination.
Table 3.—Seasonal Kendall Tau test statistics for total phosphorus concentrations at Lake Okeechobee outfall
structures in the Taylor Creek/Nubbin Slough and Lower Kissimmee River basins.
STRUCTURE
S-154
S-191
S-S65E
LENGTH OF
RECORD ~
(YEARS)
13
13
'13
SEASONAL KENDALL TAU TREND STATISTICS (mg/L)
,TAU
.421
-.725
.130
OC
.002
.001
.11
SLOPE
.035
-.049
-.0012
MEDIAN
.72
.76
.083
% CHANGE
+63
-84
-19
Table 4.—Mean total phosphorus loads for outfall structures to Lake Okeechobee in the Taylor Creek/Nubbin
Slough and Lower Kissimmee River basins. -
STRUCTURE
S-154
S-191
S-65E
LENGTH OF
RECORD
(YEARS)
13 •
' 13
13
MEAN TOTAL PHOSPHORUS LOADS (kg per day)
PRE-RCWP
. 46a
374a
370a
RCWP
70b
338a ;
402b
POST-RCWP
n/a
275b
n/a
POST-DRP
- 92b :
••••'• 227c
• ' 33ab
% CHANGE ";
'• '-+54 •••'• «
- I', :-40 /'^
• ',"•<••• j^b-1' ^'*:
169
-------
Proceedings of National RCWP Symposium, 1992
A major effect of the RCWP and the dairy rule
has been to reduce net phosphorus imports into the
basins. Improved recycling of wastewater and forage
production have also reduced the import of feed and
fertilizer. The dairy buy-out program significantly
reduced these imports. Although an exact phos-
phorus budget has not been calculated on these ef-
fects, a previous analysis indicates a strong
relationship between net phosphorus imports and
phosphorus loads to the lake for several Lake
Okeechobee basins (Fonyo etal. 1991).
A source of complication in the analysis of phos-
phorus load reduction is the impact of wetlands. The
phosphorus loads at the structures is greatly in-
fluenced by the presence of wetlands in each basin.
Phosphorus transport through wetlands can be sig-
nificantly attenuated. The S191 basin has few wet-
lands, while large sloughs and marshes occupy a
significant portion of the S154 and S65D basins.
Phosphorus temporarily or permanently stored in
these systems may explain the difference between
the tributary total phosphorus values and the total
phosphorus values measured at the structures.
Many interactions and field observations have
been conducted within the Lower Kissimmee River
and Taylor Creek/Nubbin Slough basins during the
tenure of the RCWP. Interactions have included on-
site discussion with landowners to help identify site
specific water quality problems. Success or failure in
resolving these problems depends on the ability of
the landowners to understand the problems and
respond with appropriate solutions. Unfortunately,
on many occasions the problems were fairly simple
but solutions were complicated and costly. Neverthe-
less, these interactions were important and helped
resolve many problems, thus improving off-site
water quality.
Conclusions
Agricultural best management practices were imple-
mented in the Taylor Creek/Nubbin Slough and
Lower Kissimmee River basins over the last twenty
years to reduce phosphorus loads to Lake
Okeechobee. The BMPs were installed in series of
projects beginning with control of dairy barn wash-
water. Additional projects included improved pasture
management and development of animal waste col-
lection systems and nutrient recycling. Water quality
data were collected at many sites during the projects
and analyzed for changes in total phosphorus con-
centrations. From the beginning of the project, total
phosphorus concentration data were collected from
tributaries and basin discharge structures to
evaluate the effectiveness of the RCWP and related
BMP projects. In 1986 sample collection began at
representative dairy sites to determine the effective- !
ness of animal waste collection systems. These data
were analyzed for trends in median values using the
seasonal Kendall Tau test. In addition significant dif-
ferences in mean values of total phosphorus con-
centration and load were evaluated.
Significant decreasing trends in total phos-
phorus concentration were found for most
tributaries. Total phosphorus concentrations de-
creased approximately 55 percent in the tributaries r
following implementation of BMPs. In a few trib-
utaries where the percent of land treated was below
50 percent, or where the number of animal units has
increased, total phosphorus loads also increased. At
S191 total phosphorous concentrations decreased 84
percent, while total phosphorus concentration at
S154 increased 63 percent. A small total phosphorus
increase of 19 percent at S65E was significant at the
0.11 level.
Total phosphorus concentrations have de-
creased approximately 50 percent at 75 percent of
the dairies. Fifty percent of the dairies have met the
total phosphorus concentration target based on one
to three years of monitoring. The total phosphorus
values have not been adjusted for variation in rainfall
and a few few dairies remain that have exhibited site-
specific difficulties in the management of animal
waste.
The phosphorus load discharged to Lake
Okeechobee from Hie Taylor Creek/Nubbin Slough
basin has decreased 40 percent during the project.
About 75 percent of the reduction occurred during
the RCWP program and the remaining reduction oc-
curred following the dairy rule conservation plan im-
plementation. There was a small load increase in the
Lower Kissimmee River basin. The lack of response
to BMPs in the Lower Kissimmee River basin is
probably due to the low percent of land treated with
BMPs, the presence of dairies with site specific
problems, and large wetland areas that may be
responsible for attenuating phosphorus loads. Over-
all there has been a 20-percent reduction in phos-
phorus load from the Lower Kissimmee River and
Taylor Creek/Nubbin Slough basins during this
project. -
In the future, additional analysis will be con-
ducted to evaluate several factors influencing phos-
phorus loading. The impact of wetlands on
phosphorus assimilation and attenuation will help
quantify the magnitude of total phosphorus buffer-
ing. A basin phosphorus analysis will be completed
to assess the impact of various management
strategies and watershed characteristics on the suc-
cess of BMPs.
170
-------
B. GUNSALUS, E.G. FLAIQ, & G. R1TTER
References
Blue, W.G. 1970. The effect of lime on retention of fertilizer phos-
phorus in Leon fine sand. Soil Crop Sci. Soc. Fl. Proc. 30:141-
50.
Bradley, J.V. 1968. Distribution-Free Statistical Tests. Prentice-
Hall, Englewood Cliffs, NJ.
Day, R.W. and G.P. Quinn. 1989. Comparisons of treatments after
an analysis of variance in ecology. Ecol. Mono. 59:433-63.
Einot, I. and KR Gabriel. 1975. A study of the powers of several
methods of multiple comparisons. J. Am. Stat. Ass. 70:574-83.
Federico, A.C., K.G. Dickson, C.R Kratzer, and F.E. Davis. 1981.
Lake Okeechobee Water Quality Studies and Eutrophication
Assess. Tech. Pub. 81-2. South Florida Water Manage. Distr.,
West Palm Beach, FL.
Flaig, E.G. and GJ. Ritter. 1989. Water quality monitoring of
agricultural discharge to Lake Okeechobee. ASAE Pap. 89-
2525. Pres. Int. Winter Meet ASAE, New Orleans, LA
Fonyo, C. et al. 1991. Final Report Basin Phosphorus Mass Balan-
ces. Biogeochemical behavior and transport of phosphorus
in the Lake Okeechobee basin. Contract C91-2394 South
Florida Water Management District Inst Food and Agric.
Sci. Univ. Florida, Gainesville.
Goldstein, AL. 1986. Upland Detention/Retention Demonstration
Project. Final Rep. Tech. Pub. 86-2. South Florida Water
Manage. Distr. West Palm Beach, FL
Gilbert, RO. 1987. Statistical Methods for Environmental Pollu-
tion Monitoring. Van Nostrand Reinhold, New York.
Hirsh, RM., J.R Slack, and RA Smith. 1984. Techniques of trend
analysis for monthly water quality data. Water Resour. Res.
18:107-21.
Montgomery, RH. and J.C. Loftis. 1987. Applicability of the t-test
for detecting trends in water quality variables. Water Resour.
Bull. 23:653-62.
Osking, M.K. and B,E. Gunsalus, 1992. Evolution of the RCWP
water quality monitoring networks in the Taylor Creek/Nub-
bin Slough and Lower Kissimmee River Basins. In Proc. Natl.
RCWP Symp., 1992, Orlando, FL.
Reckhow, K.H. and S:C. Chapra. 1983. Engineering approaches for
lake management. Vol. ,1 in Data Analysis and Empirical
Modeling. Butterworth Publishers, Woburn, MA
Statistical Analysis Systems Institute. Inc. 1990. SAS Procedures
Manual Version 6th ed. Stat Anal. Systems, Gary, NC.
Sawka, G J. et al. 1992. Post-BMP water quality monitoring: dairy
farm synoptic survey program in the Lake Okeechobee
Basin. In Proc. Nati. RCWP Symp., 1992. Orlando, FL.
Snedecor, G.W., and W.G. Cochran. 1980. Statistical Methods. 7th
ed. Iowa State Univ. Press, Ames, Iowa. -
171
-------
-------
Estimation of Lag Time for
Water Quality Response to BMPs
John C. Clausen
Department of Natural Resources Management and Engineering
Uniuersity of Connecticut
Storrs, Connecticut
Donald W. Meals and E. Alan Cassell
School of Natural Resources
University of Vermont
Burlington, Vermont
ABSTRACT
The St. Albans Bay Watershed Rural Clean Water Program project was one of 21 experimental
projects aimed at improving water quality through use of agricultural best management practices
(BMPs). After 10 years of comprehensive monitoring and extensive BMP implementation covering
74 percent of critical acres, a significant reduction in nutrient concentrations and mass exports was
not observed in either tributary streams or the bay. However, decreases in sediment concentra-
tions and loads as well as bacteria counts did occur. Two reasons offered for the lack of nutrient
response are (1) the BMPs were insufficient to cause change or (2) the study had ended before
changes started to occur. This paper presents a method to estimate the length of time needed to
obtain a given water quality response from a certain level of land treatment. The technique was ap-
plied to several BMP scenarios in the St. Albans Bay RCWP, including pre-BMP, reduced runoff,
reduced inputs, and achievable export cases. The results show that reducing exports may take
many years, depending on the amount of input reduction and the initial field nutrient concentra-
tions. This method can be applied to other land treatment nonpoint source projects interested in
determining lag time to BMP implementation.
When best management practices (BMPs)
are applied in watersheds to control non-
point source pollution, it is often as-
sumed that water quality improvements will occur in
a relatively short period of time. Models such as
CREAMS, GLEAMS, and AGNPS (Knisel, 1980;
Leonard et al. 1987; Young et al. 1989) that are com-
monly used to plan selection of the most effective
BMPs assume watershed conditions are in equi-
librium or steady state—that export values do not
change, beyond normal variation, with respect to
time. However, these models do not provide a means
to predict how long it may take to reach equilibrium.
This uncertainty was exemplified by the St. Al-
bans Bay Watershed Rural Clean Water Program
(RCWP) project, located in northwestern Vermont.
Dairy agriculture dominated land use in this 13,000-
ha watershed with 98 farms, where the major BMP
implemented was animal waste management, includ-
ing waste storage during the winter and waste ap-
plication and incorporation in the spring. However,
other BMPs—such as permanent vegetative cover,
cropland protective system, and fertilizer manage-
ment—were used as well.
Four tributaries to St. Albans Bay were con-
tinuously monitored for nitrogen, phosphorus, sedi-
173
-------
Proceedings of National RCWP Symposium, 1992
ment, and bacteria. Sampling was also conducted in
St. Albans Bay and on individual farm fields. After 10
years of monitoring and the implementation of
BMPs over 74 percent of critical acres, a significant
reduction in nutrient concentration and mass export
was not observed in tributary streams (Vt. RCWP
Coor. Comm. 1991). The question was whether the
lack of nutrient response resulted from an insuffi-
cient use of BMPs or whether insufficient lag time
had been allowed for changes to occur. Lag time is
the time elapsed between initiating a land treatment
and achieving a certain percentage of a new equi-
librium water quality.
This paper describes a method for estimating
the lag time for a quality of runoff to respond to
BMPs. Knowledge of lag times would assist water-
shed planners interested in documenting the
benefits of nonpoint pollution controls. Lag times
may also help evaluate the effectiveness of alterna-
tive BMPs.
Study Area
Farm fields used for the calculation of lag times were
located within the St Albans Bay watershed. The
phosphorus export values were from the 1,384-ha
Jewett Brook watershed, located in the St. Albans
Bay watershed. Over two-thirds of the watershed
soils had formed on lacustrine deposits, the
remainder had formed on glacial till, and over 90 per-
cent of these soils are poorly drained. Flow in Jewett
Brook, which averaged 60 percent of annual
precipitation, occurred during both stormflow and
baseflow periods. The dominant land use in the
watershed was dairy agriculture (86 percent), with
32 percent of the watershed being planted to silage
corn.
Of the 18 farms in the watershed, 16 signed
RCWP contracts. Dairy farms averaged 134 ha, with
an average herd size in 1,991 of 120 animal units. At
the completion of the RCWP project, 74 percent of
the critical acres and 83 percent of the animal units
were controlled by RCWP contracts. The Jewett
Brook watershed has been described in greater
detail by Meals (1992) and the Vermont RCWP Coor-
dinating Committee (1991).
Method
Predicting the time required for lake recovery as a
result of decreasing nutrient inputs is not new to lake
scientists. Equations for this process have been
described by Lorenzen (1974) and reported in
various texts, such as that by Chapra and Reckhow
(1983). These equations may also be suitable for ap-
plication to an agricultural field or watershed receiv-
ing BMPs for several reasons:
• The concentration in the field export is related
to the field surface concentration; the lake
equations assume that the export and lake
concentrations are equal.
• The rate of change of the export concentra-
tion is proportional to the field concentration,
which also is assumed for a lake.
• The assumption of a well-mixed system may
apply equally to a lake or the surface of a soil.
Neither system is perfectly well mixed. For
the simple model presented in this paper, the
soil is mixed once per year to the depth of the
plow layer.
• The law of conservation of mass applies in
both systems.
The phosphorus mass balance for a 1-ha field is
presented in Figure 1. This conceptual diagram
shows phosphorus being added to the field as
manure, fertilizer, and precipitation; outputs from
the field are shown as silage yield, runoff, and
leakage into deeper soil layers. Runoff is composed
of two sources, surface and subsurface (Fig. 1),
which collectively become streamflow. For this
paper, the leakage into subsurface soil layers was as-
sumed to be zero.
Manure Fertilizer Precipitation
Crop
Runoff
Storage
Leakage
Figure 1 .—Phosphorus mass balance diagram for lag
time calculations.
The concentration of phosphorus in the export
from the field (Cg) is assumed to be proportional to
the concentration of phosphorus stored in the sur-
face 1 cm of the field (C) according to
CQ = EC (1)
where E = extraction coefficient (Frere et al. 1980).
These phosphorus concentrations must be con-
sidered in terms of availability for runoff and not
necessarily the availability for plant growth. This
174
-------
J.C. CLAUSEN, D.W. MEALS, & E.A. CASSELL
relationship shows that, to improve the quality of
surface runoff, the field concentration must be
reduced. This emphasis on field concentration
rather than runoff concentration has been widely
overlooked in nonpoint source management
strategies.
The mass balance equation that describes the
rate of change in field concentration of phosphorus
per unit soil volume is:
dC MCm + FCf + PCP _ YCy _ ECQ
dt " V V V
(2)
where dC/dt = change in mass stored in the field/
change in time; C = phosphorus concentration in
upper 1 cm of soil (g/m3); ,t = time (yrs); M, F, P, Y =
amount of manure, fertilizer, precipitation, and crop
yield, respectively (m3/yr); Cm, Cf, Cp, Cy = P con-
centration of manure, fertilizer, precipitation, and
yield, respectively; E = extraction coefficient; Q =
rate of surface runoff (m3/yr); and V = volume of soil
(m ). The term YCy, which represents the loss of
phosphorus in the crop, could be made a function of
the soil concentrations rather than handled as a con-
stant.
The steady-state solution of the equation for the
field concentration of phosphorus is determined
from Equation 2 by setting dC/dt = 0, which gives:
Ce = 19
(3)
where Ce is the equilibrium field concentration of P
(g/m3), I is the net loading (kg/yr) obtained by the
summation of all inputs as shown in Equation 4:
MCm + FCf + PCP - YCy
V
(4)
and 6 is the phosphorus runoff residence time (yrs),
based solely on the total phosphorus export from the
field, given by:
9=E^ <5>
where E = the extraction coefficient and Q is the out-
flow (m3/yr). The quantity EQ is determined by
dividing the mass export in runoff (ECQ) by the con-
centration (C) in that runoff. This runoff residence
time underestimates the total residence time of
phosphorus for the field since it ignores the removal
of phosphorus by the crop or deeper burial of phos-
phorus in the soil. A more complex model could ac-
count for these fluxes.
The general solution of Equation 2 that permits
the calculation of the field phosphorus concentration
at any point in time, given an initial concentration
(C0) at time t - 0 , and solved for t is given by:
where C is the field phosphorus concentration at any
time t, and Co is the initial concentration at time t =
0. Equations (3) and (6) will serve as the basis for
determining the length of the lag time (years) as-
sociated with alternative management strategies.
Two techniques were used to determine lag
times. The first was analytical calculations using
these equations on a hand calculator. The second
technique used a dynamic simulation model that al-
lows observation of the continuous change in both
field and runoff concentrations. The software pack-
age, STELLA II, was used to construct the model
(Bogen, 1989; High Performance Syst. 1990). The
structural diagram for the 1-ha field used in STELLA
to describe Equations 1 and 2 is presented in Figure
2. The model in STELLA uses the same equations as
described previously.
Results
For the purpose of not presenting a highly complex
model to estimate lag times, several simplifications
were made in mass balance. Additional fluxes could
be modeled that would likely improve the estimate of
the lag time. To estimate lag time, the following in-
formation is needed:
• the mass of manure, fertilizer, and
precipitation nutrient inputs,
• the output mass values for crop yield and
runoff export, and
• the field and runoff nutrient concentrations.
The nutrient contents of precipitation, crop
yield, and runoff could be measured or estimated
from literature values, such those provided in Knisel
(1980). Regional values of phosphorus in runoff
would be preferred. However, the nutrient content of
field inputs such as manure and fertilizer should be
measured or estimated more directly. An actual sam-
pling of field nutrient concentration is crucial to
reasonable lag time estimations.
This method to determine the time lag was ap-
plied to several BMP scenarios in the St. Albans Bay
RCWP, including pre-BMP, reduced runoff, reduced
input, and setting an export goal as achievable water
quality. Since actual total phosphorus concentration
for the field surface was unknown, calculations were
made for both a high concentration (1,000 g/m3)
and a low concentration (350 g/m3), as reported in
the literature (Thompson and Troeh, 1978).
Pre-BMP
In the first application, the pre-BMP condition is
given to describe the existing phosphorus mass
175
-------
Proceedings of National RCWP Symposium, 1992
Soil Volume
Field Cone
3,500
Flaure 2.—Structural diagram for dynamic simulation model of agricultural field using STELLA II. Clouds are sources
(arrow out) and sinks (arrow In) of phosphorus. Double-lined arrows are pipelines through which phosphorus flows.
Circles connected to pipelines are controllers that regulate the rate of flow of the mass of phosphorus through the
pipelines. Rectangle represents the storage of phosphorus In the field. Plain circles are converters and convert Inputs
Into outputs. Single-lined arrows are connectors depleting causal linkages In the model.
balance for the watershed. Inputs
of manure and fertilizer were
based on averages from 11 farms
in the St. Albans Bay watershed
as reported by Jokela (1991).
The phosphorus content of the
corn silage yield output was from
Frere et al. (1980). The export of
phosphorus in runoff was
measured in the St Albans Bay
watershed (Vt. RCWP Coor.
Comm. 1990). Based on these
averages, mass inputs of phos-
phorus exceeded outputs (Table
1). These loads would result in
rising field concentrations (Fig.
3) and therefore increasing
runoff export values (Table 1). It
was assumed that only the upper
1.0 cm of soil "actively" produced
phosphorus in runoff (Frere et
al. 1980). Therefore, the volume
(V) of the one hectare storage
unit was 100 m3. If the depth of
this layer were increased, the lag
time would become proportional-
ly longer.
176
••• Achievable export
+ Achievable export
3Ł Reduced Input
* Reduced input
X Pre-BMP
•*• Pre-BMP
101 121
YEARS
181
Figure 3.—Changes In field phosphorus concentrations resulting from alterna-
tive BMP scenarios.
-------
J.C. CLAUSEN, D.W. MEALS, &E.A. CASSELL
Table 1.—Summary of phosphorus mass balance cal-
culation to determine lag time for alternative BMPs.
PARAMETER PRE-BMP
Initial Conditions
MCm (kg/ha) 36.
FCf (kg/ha) 5.
PCP (kg/ha) 0.036
Inputs (kg/ha) 41.036
YCy (kg/ha) 34.
ECQ (kg/ha) 2.16
Outputs (kg/ha) 36.16
Results of BMP Scenarios
ForCo = 1,OOOg/m3
Ce (gm/3) 3,257.
Cgo(g/m3) 3,031.
Cso (g/m3) 2,128.
tgo (yrs) 107.
tso (yrs) 32.
ECeQ (kg/ha) 7.04
For Co = 350 g/m3
Ce(g/m3) 1,140.
C9o(g/m3) 1,061.
Cso (g/m3) 745.
tao (yrs) 37.
tso(yrs) 11.
REDUCED
FERTILIZER
36.
0.
0.036
36.036
34.
2.16
36.16
943.
948.
971.
107.
32.
2.04
330.
333.
340.
18.
11.
ACHIEVABLE
EXPORT
34.
0.
0.036
34.036
34.
0.496
34.496
230.
307.
615.
107.
56.
0.496
80.
107.
215.
37.
11.
V = 100m3
for Co = 1,000 g/m3, 0 = 46.3 yrs; EQom = 2.16 m3/yr
for Co = 350 g/m3, 0 = 16.2 yrs; ECUt = 6.17 m3/yr
Reduced Runoff
This application begins with the pre-BMP condi-
tions, then at 50 years, the runoff export is reduced
by 50 percent. This reduction might occur as a result
of a decrease in runoff as when conservation tillage
replaced conventional tillage or when concentrations
are reduced in runoff as a result of erosion protec-
tion. The results from STELLA are presented in Fig-
ure 4A for the initial field concentration of 1,000
g/m3 and in Figure 4B for the field phosphorus level
of 350 g/m3. For both situations, the reduction of
phosphorus in the runoff export is short-lived. The
field concentration and runoff export continue to in-
crease over time because field inputs exceed out-
puts. Land management practices that are designed
to contain nutrients in the field effectively reduce ex-
ports to a point; a further reduction in exports would
require a reduction in inputs.
Reduced Fertilizer
For this application, phosphorus additions from
manuring and precipitation remained, and no addi-
tional phosphorus was applied to the fields as fer-
tilizer. Assuming a high field concentration of 1,000
g/m (ppm), the new field concentration in equi-
librium with these inputs and outputs would be 943
g/m3 (Table 1). To obtain 90 percent of the change
to this new field concentration would take 107 years
(Fig. 3). The resulting runoff export would not
change appreciably from the original export of 2.16
kg/ha. Only 32 years would be needed to achieve 50
percent of the change to a new concentration.
Assuming a low initial field concentration of 350
g/m3, it would take 18 years to achieve a field con-
centration of 333 g/m3 (90 percent change), and 11
years to reach a concentration of 340 g/m (50 per-
cent change) (Table 1).
Achievable Export
For this application, the runoff export was set to a
final value of 0.496 kg/ha, which represents the
average annual phosphorus export from the control
Subwatershed 3 in the LaPlatte River watershed
(Meals, 1990). We consider this value to be the
achievable runoff export of phosphorus for agricul-
ture in Vermont as described by Clausen and Meals
(1989). Also for this example, the field inputs were
reduced to equal yield outputs. To achieve 90 per-
cent of the change from the original concentration to
a new concentration would require 107 years, assum-
ing a high field concentration (Fig. 4). Using a low
field concentration, 37 years would be needed to
achieve 90 percent of the change and 11 years to ob-
serve 50 percent of the change (Table 1).
Conclusions
The estimation of the lag time for a water quality
response to BMP implementation shows that effects
may take many years to occur, depending upon the
amount of input reduction and initial field nutrient
concentrations. For the St. Albans Bay RCWP, a new
equilibrium in water quality, in response to the
BMPs, was probably not reached during the
monitoring period. This procedure is adaptable to
setting various goals, such as a desired mass export
level or some percent change in export. A significant
reduction might be defined as exceeding two stand-
ard deviations of the average export before BMPs
are implemented.
This method emphasizes the importance of
managing water quality by managing the field con-
centration. Future nonpoint source studies evaluat-
ing the effectiveness of BMPs on nutrients in runoff
should include a soil sampling program of total
nutrient concentrations.
This procedure is applicable to other nutrients.
Nitrogen, for example could be evaluated, although
the mass balance equations would be more compli-
177
-------
Proceedings of National RCWP Symposium, 1992
Field Cone Field Runoff
6000 T8.00 T
. 3000 -
i
• M
1
§
C
o
-------
J.C. CLAUSEN, D.W. MEALS, Ł• EA CASSELL
cated since volatilization, denitrification, and leakage
could be prominent mass fluxes. Also, different soils
and different crops would result in different lag
times; however, the mass balance approach would be
similar. This procedure for estimating the lag time
for a water quality response to BMPs should aid in
selecting appropriate BMPs and documenting the ef-
fectiveness of nonpoint source programs.
References
Bogen, D.K. 1989. Simulation software for the Macintosh. Science
246(4926):13842.
Chapra, S.C. and K.H. Reckhow. 1983. Engineering Approaches
for Lake Management Vol. 2: Mechanistic Modeling. Butter-
worth Publishers, Woburn, MA.
Clausen, J.C. and D.W. Meals. 1989. Water quality achievable with
agricultural best management practices. J. Soil Water Con-
serv. 44(6):593-96.
Frere, M.H., J.D. Ross, and LJ. Lane. 1980. The nutrient sub-
model. Chapter 4 in W.G. Khisel, ed. CREAMS: A Field-scale
Model for Chemicals, Runoff, and Erosion From Agricultural
Management Systems. Conserv. Res. Rep. No. 26. U.S. Dep.
Agric., Washington, DC.
High Performance Systems, Inc. 1990. STELLA II User's Guide.
Hanover, NH.
Jokela, WE. 1991. Field nutrient management project Chapter 10
in St. Albans Bay Rural Clean Water Program, Final Report.
Vt. RCWP Coor. Comm., Vt Water Resour. Res. Center, Univ.
Vt, Burlington.
Knisel, W.G., ed. 1980. CREAMS: A Field-scale Model for Chemi-
cals, Runoff, and Erosion From Agricultural Management
Systems. Conserv. Res. Rep. No. 26. U.S. Dep. Agric.,
Washington, DC.
Leonard, RA, W.G. Knisel, and DA Still. 1987. GLEAMS:
groundwater effects of agricultural management systems.
Trans. Am. Soc. Agric. Eng. 30 (5): 1403-18.
Lorenzen, M.W. 1974. Predicting the effects of nutrient diversion
on lake recovery. Pages 205-10 in EJ. Middlebrooks, D.H.
Falkenborg, and T.E. Maloney, eds. Modeling the
Eutrophication Process. Ann Arbor Science Publishers, Inc.,
MI.
Meals, D.W. 1990. LaPlatte River Watershed Water Quality
Monitoring and Analysis Program. Prog Rep. No. 12. Comp.
Final Rep. 1979-89. Vt. Water Resour. Res. Center, Univ. Vt,
Burlington. • — ••
. 1992. Water quality trends in the St Albans Bay, Vermont,
watershed following RCWP land treatment. In Proc. Natl.
RCWP Symp., Orlando FL
Thompson, L.M. and F.R. Troeh. 1978. Soil and Soil Fertility. 4th
ed. McGraw-Hill Book Co., New York.
Vermont Rural Clean Water Project Coordinating Committee.
1991. St Albans Bay Rural Clean Water Program, Final
Report. Vt Water Resour. Res. Center, Univ. Vt, Burlington.
Young, RA, C.A Onstad, D.D. Bosch, and W.P. Anderson. 1989.
AGNPS: a nonpoint source pollution model for evaluation of
agricultural watersheds. J. Soil Water Conserv. 44(2) :168-73.
179
-------
-------
Water Quality Trends in
Big Pipe Creek During the Double Pipe
Creek Rural Clean Water Program
John L. McCoy and Robert M. Summers
State of Maryland
Department of the Environment
Baltimore, Maryland
ABSTRACT
The Double Pipe Creek watershed in Carroll County, Maryland, was the focus of the Double Pipe
Creek Rural Clean Water Program project initiated in 1980. The watershed encompasses 175
square miles divided into two subwatersheds: Little Pipe Creek and Big Pipe Creek, the latter
being 58 percent of the total area in Double Pipe Creek watershed. Animal waste management
facilities constructed as a result of the project increased animal waste storage capacity in the
Double Pipe Creek basin by 28 percent; soil conservation practices reduced soil erosion by an es-
timated 4 percent Water quality was monitored in the Big Pipe Creek subbasin between 1982 and
1985 and again between 1987 and 1990. Base flow and storm flow water quality were monitored at
the base of the watershed in conjunction with a U.S. Department of the Interior Geological Survey
streamflow gaging station. Results of the trend analysis indicated that ammonium and total organic
carbon concentrations decreased significantly during the project period (44 percent and 51 per-
cent, respectively), while total nitrogen and nitrate-t-nitrite nitrogen concentrations increased sig-
nificantly (25 and 34 percent, respectively). Total Kjeldahl nitrogen, orthophosphate, total
phosphorus, total suspended solids, and fecal coliform showed no significant change. While the in-
crease in manure storage may have decreased nitrogen losses during storage and application, the
amount of nitrogen being applied to cropland and lost through leaching increased, as suggested by
the increasing concentrations of total nitrogen and nitrate+nitrite nitrogen. These results suggest
that in addition to structural practices, more aggressive nutrient management efforts are needed in
the basin.
The Double Pipe Creek watershed, 175
square miles located in Carroll County,
Maryland, is divided into two subwater-
sheds, Big Pipe Creek and Little Pipe Creek (Fig. 1),
which are 58 and 42 percent, respectively, of the total
area in the Double Pipe Creek watershed. Double
Pipe Creek watershed is predominately agricultural;
dairy farming is the main activity. In 1980, ap-
proximately 18,000 dairy cattle produced more than
300,000 tons of manure, most of which was not
stored in adequate facilities.
Water quality problems in the watershed were
twofold: both fecal coliform and turbidity were at
levels consistently in excess of State water quality
standards. These problems were directly related to
the large numbers of dairy cattle found in the basin.
Animal manure is a source of fecal coliform that con-
tributes, when poorly managed, to in-stream coli-
form levels. The basin's turbidity problems were
aggravated by concentrations of animals around the
streams and nutrient enrichment in the streams
from year-round manure application on the fields.
The Double Pipe Creek Rural Clean Water Pro-
gram (RCWP) project was initiated in 1980. Its
primary goal was to improve water quality in the
basin through the application of best management
181
-------
Proceedings of National RCWP Symposium, 1992
Pennsylvania
N
CARROLL COUNTY
LEGEND:
A WatwChJaKy Monitoring Stations
Figure 1.—Carroll County, Maryland, Double Pipe Creek Rural Clean Water Program project area BMP-nonpoInt source
monitoring sites.
practices (BMPs). The project's emphasis was man-
agement of animal waste. Its specific water quality
goals were to reduce the level of fecal coliform bac-
teria in Big Pipe and Little Pipe creeks below the
State standard of 200 MPN/100 mL (MPN is a stand-
ard total multiple-tube test) and to meet the State
standard for turbidity in both streams (150
Nephelometric Turbidity Units [NTUs], Jackson
Turbidity Units LJTUs], or the essentially obsolete
Formazin Turbity Units [FTUs] or a monthly
average of 50 units) at all times. The more general
goal of the project was to improve water quality.
A water quality monitoring program to support
the Double Pipe Creek RCWP was established to
detect long-term trends in water quality, measure the
effectiveness of BMPs, and determine the project's
impacts on turbidity levels and fecal coliform.
Methods
Background
The water quality monitoring program has been con-
ducted in two phases. Versar Inc., under contract
with the Maryland Department of Health and Mental
Hygiene, conducted a water quality monitoring pro-
gram for the Monocacy River basin (Versar, 1982),
with-special emphasis on the Double Pipe Creek sub-
watershed. The work plan developed by Versar for
the first phase of the water quality monitoring pro-
gram had three objectives:
1. To establish a near-term water quality
database for Double Pipe Creek,
2. To evaluate the short-term effectiveness of
animal waste and erosion control practices,
and
3. To plan a long-term monitoring program to
be implemented at the conclusion of the
Versar work.
Data was collected by Versar Inc. at the U.S..
Geological Survey (USGS) gaging station on Big
Pipe Creek at Bruceville and three single land use
sites within the Double Pipe Creek basin from
November 1982 to June 1985. Water quality and flow
data were collected during storm events and peri-
odically during base flow at all four sites.
182
-------
J.L. McCOYS R.M. SUMMERS
The contract with Versar Inc. expired in 1986,
the company provided a final report of its findings
(Versar, 1986), and the monitoring stations were
deactivated. Versar's final report included an am-
bitious design for Phase II of the water quality
monitoring program that was too expensive to imple-
ment.
In 1987, the Bruceville site was reactivated by
the Maryland Department of the Environment to
begin Phase II of the water quality monitoring pro-
gram. The work plan for Phase II (Md. Dep. Environ.
1989) called for monitoring at the Bruceville site on
Big Pipe Creek and installation of a monitoring site
in one small agricultural land use basin. This paper
focuses on data collected at the Bruceville site.
Data Sources
Water quality and flow data during storm events
were collected at the Bruceville site between 1982
and 1985 as part of Phase I and, since 1987, as part of
Phase II of the water quality monitoring program.
The stormwater quality and flow data from the two
phases of the water quality monitoring program have
since been merged into a single data set
Water quality data have also been collected
monthly at the Bruceville site since 1978 as part of
the State's ongoing Core Monitoring Network.
These data have been merged with the stormwater
quality data (collected during Phase I and Phase II)
for purposes of this analysis.
Frequency of Collection
Two to three storms per season (8 to 12 per year)
were sampled at the Bruceville site during both
phases of the water quality monitoring program. To
maintain comparability between the data sets, the
parameters were identical to those collected by Ver-
sar during Phase I.
Sampling Procedures
• Flow. The Big Pipe Creek site is a USGS gaging
station; the monitoring equipment was installed in an
adjoining shed. Equipment was set up to use the
specific streamflow rating tables calculated by
USGS, whose local district office is contacted
regularly for rating curve adjustments and flow
records.
• Water Quality. Base-flow sampling is conducted
monthly at the Big Pipe Creek site. Samples are col-
lected, packed in ice, and sent to the Maryland
Department of Health and Mental Hygiene for
analysis (Md. Dep. Health Human Hygiene, 1981).
Flow proportioned composite stormwater quality
samples are collected at the Big Pipe Creek site,
where the automated water quality sampler is trig-
gered by a calibrated flowmeter. Data on ap-
proximately 10 storm events per year are collected at
the site.
The sampling equipment is stage-activated.
Before a predicted event (any storm occurring
during the work week with a 50 percent or greater
chance of rain), the samplers are iced and serviced.
Following a sampled storm event, field personnel
collect the samples, perform necessary in-field
operations, ice the samples, and return them to the
laboratory. Then staff examine and service the
equipment at the station and set it up to sample the
next storm. Sampling lines are cleaned, flowmeters
are recalibrated, and strip charts are retrieved (Md.
Dep. Environ. 1989).
A portion of each sample is collected in a plastic
bottle, packed in ice, and analyzed for total Kjeldahl
nitrogen, volatile suspended solids, total phos-
phorus, total suspended solids, total organic carbon,
and turbidity. Another portion is filtered through a
0.45 urn membrane filter, packed in ice, and analyzed
in the laboratory for total Kjeldahl nitrogen, am-
monium nitrogen, nitrate+nitrite nitrogen, or-
thophosphate, dissolved organic carbon, and total
dissolved solids. A separate sample is collected in a
sterilized 250 mL plastic bottle pretreated with
thiosulfate and packed in ice until processed in the
laboratory for total and fecal coliform. Total nitrogen
is calculated as the sum of total Kjeldahl nitrogen
plus nitrate+nitrite nitrogen.
Data Handling and Analysis
Water quality data collected by Versar during Phase
I, by the Maryland Department of the Environment
as part of the Core Monitoring Program, and during
Phase II of the RCWP water quality monitoring pro-
gram were compiled in a single database and verified
for analysis in this paper. The data set contains storm
event and base-flow water quality data collected be-
tween 1982 and 1990. The data were then plotted
against time (Fig. 2) and the plots were reviewed.
The Maryland Department of Health and Mental
Hygiene laboratory detection limits for orthophos-
phate, ammonium, arid total phosphorus were 0.1,
0.2, and 0.1 mg/L, respectively, on numerous oc-
casions between January 1983 and November 1985.
Current laboratory detection limits for these
parameters are .004, .008, and .01, respectively.
Changes in detection limits over the time period
of interest complicate trend analysis. When data are
183
-------
AHUONIUU HITROGEH COHCEHIRATIOH UG/l TOFAL NITROGEN CONCENTRATION UG/l
I.I1
O
d
* .
O
4 O
O
O ** O *Oo ° O °
o «oo 0 $ ^ o
U*»**»* J*" **»*•• %J&<^Jkxft^° .
3-
12-
II
10
9'
8'
7
6
5'
!•
y
2
1
0'
o
o
o
o
o
o
0 °
o ° o °^ *
oo. o o*^"o ^ o o^ tfk^ CQ^sP
»0 0 *« 0^%
o °
0 o
f
JIHOWI JJAUSI4 I9UAY87 I2FE89D 08NOV92 26NOV8I 22AUG84 I9UAY87 I2FEB90 08HOV92
DATE DAIE
J01AI KJE18AKI HIIROCEN {WHOLE) CONCENTRATION UC/L HITRATE * KITRITE NITROGEN CONCENTRATION UG/L
16
J-
s
s
1
J
1
1
fl
0
0 «
o
o
* 0Ł 0
oo o
« o oS°oo-°
* °«* * 0° * 00 ° ° °o
* *^0*Jl^ ^jflft.*0*0* °° ^«eww5>cl>^^>O^f
* S^« O*01^ 0 «« 0 O
7
i
5
H
0 .
2 '
u
0 3
2
1
0
o
•>> « . »
',' * « « °0
o do
„ 0 ^.4. o«% °°o?0 «
0 /» ° % a ° « *» •«*
°°* *"V * « ' 0 ^>» *%' «»
^ ° o «> o y ^^»s o?^S^L<4,ti«Ja~«>
. «* o^4«oo ^tftoos»«> ^d w^posgoono^waajo ** 9.08
0.06
0.04
0.02
0.00
ooo , • ' o
» 00
0 • 0 "*>
° 0 °» °00 »
* ^080
\y. *« - ° ° :
o o 0° *' * o o °o
o o* o o * c* V* °AO t« *
» ! o o o A ^^ o *^ o*ow
XMEOOOOO 4»O ^ ^ O v'
JMOVJI 22ABCH I9UAY87 I2FEB90 08NOV92 26NOV8I 22AUC84 I9UAY87 I2FEB90 08NOV92
DATE ' ' DATE
TOIAt SUSPENDED SOLIDS CONCEHTRATIOH UG/l TOTAL ORGANIC CARBON CONCENTRATION UC/L
jrt inn
30!
10!
' o
e
« 0
0 o
0 0
o-
4
0 0
'~,« ^o J!;«
A. ^«&Ł&k>«>i%«aKA>«rifeao «&=S&&teŁ?&»4«*>a»6aag>
40:
30-
20:
10-
n-
o
o
o
O ' o
o
o o
o
•%'• '• . - • o-
° * °«> 0 *
° ° ° 0 ° A% ' * d«° ^
o o * o°B o o o * 0 * 0 0
* 0 <*>* °° ^^ %0°» 0
-------
J.L. McCOY & R.M. SUMMERS
FLOW HEAD DAILY DISCHARGE
FECAL COLIFORU BACTERIA CONCENTRATIOH UPN/IOO III
1600-
1500-
1400'
uoo-
1200-
noo-
1000-
900'
800-
700'
600-
500'
400'
300'
200-
too-
Q-
0
°
0 0
o
0 0
o
" 0
o o
0 0
0 0
° 0° ° o° °
*o *C^ ° &*$Ł o^,*^ ° ^SLw ^xy*a<> o^^a^5^
^jSfc <& °^ ^>4> ^"^ ows' xr»
50000:
40000-
30000'
20000'
10000
o-
oo
^
o
O 00 O
o
0
o o o
o
o o o o o o *
<# o *^"» ot?°° o&°o ^^fifr^-co
-------
Proceedings of National RCWP Symposium, 1992
TIME TREND IK LOG TRANSFORMED
AMMONIUM NITROGEN CONCENTRATIONS
TDTAL NITROGEN CONCENTRATIONS
1912 1983 1984 1985 198$ 1987 1988 1989 1990 1991
YEAR
, LOIESS SMOOTHED ...... OUAORATIC REGRESSION
$0110 STRAIGHT LINE IS I8E SEAN PREDICTED LOG NH4
TIME TREND IN LOG TRANSFORMED
TOTAL KJELDAHL HITROCEN (IHOLE) CONCENTRATIONS
1983 1984 1985 1986 1987 1988 1989 1990 1991
YEAR
LOIESS SMOOTHED --,-,, QUADRATIC REGRESSION
SOLID STRAIGHT LINE IS THE MEAN PREDICTED LOG TN
THE TREND IN LOG TRANSFORMED
NITRATE t NITRITE NITROGEN CONCENTRATIONS
-J
1992 198] 1984 1985 1985 1987 1988 1989 1990 1991 1982
YEAR
__ LCIESS SHOOTHCO „ OUAORATIC REGRESSION
SOLID STRAIGHT LINE IS THE UEAH PREDICTED LOG TKNW
TIKE TREND IN LOG TRAHSFORUED
ORTHOmSPHATE COHCEHTRATIONS
1983 1984 1985 1986 1987 1988 1989 1990 1991
YEAR
LOIESS SMOOTHED QUADRATIC REGRESSION
SOLID STRAIGHT LINE IS THE IIEAH PREDICTED IOC N023
TIME TREND IN LOG TRANSFORMED
TOTAL PHOSPHORUS CONCENTRATIONS
-5
I9J2 198] 1984 1985 1986 1987 1988 1989 I
YEAR
.tS'iSS.SMOOIHED.. .,.gUADRAT | C.REGRESSI ON
Lull rlM
-.,,-,
SOLID STRAIGHT LINE IS IfiE UEAH PREDICTED
TIME TREND IN LOG TRANSFORMED
T01AI SUSPENDED SOLIDS CONCENTRATIONS
1983 1984 1985 1986 1987 1988 1989 1990 1991
YEAR
LOIESS SMOOTHED .„..- OUAORATIC REGRESSION
SOLID STRAIGHT LINE IS IHE UEAN PREDICTED LOG IP
TIME TREND IN LOG TRANSFORMED
TOTAL ORGANIC CARBON CONCENTRATIONS
4
3
L
5 I'
1382 198) 1994
1985 I98S 1987 1988
YEAR
1989 1990 1991
_ LG»ESS SMOOTHED ...... OUAORATIC REGRESSION
SOLID STRAIGHT LINE IS TfiE MEAN PREDICTED LOG TSS
Figure 3.— Time trends at the Big Pipe Creek site at Bruceville.
1983 1984 1985 1986 1987 1988 1989 1990 1991
YEAR-
—— LOIESS SMOOTHED QUADRATIC REGRESSION
SOLID STRAIGHT LINE IS THE UEAH PREDICTED LOG TOC
186
-------
J.L. McCOV & R.M. SUMMERS
TIME TREND IN LOG TRANSFORMED CLOI
TIKE TREND IN LOG IRANSFORUEO
FECAL COLIFORII BACTERIA CONCEHFRAIIOHS
8'
;•
L 6-
G
F 5'
L
0
H 4'
V
2-
+
+ * * ++ {
4. H++ + + ^ t +
. V *+ +++_ *** Vl*...*!.=«f-h
\"^"TJ"+ -^F^T ' "**%. I++ ++;*+j^
++1' + + <• +jt+ + +
+f+ ^* + \
11-
I0:
9-
0 8
G
F 7-
C
0 6'
5-
4:
3-
\ «+ +
4- + + *
+ + + + + t +
-H. -H-+ ++--f+ +
+ + -t- +•
+ + •*•
• + *
1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1982 1983 1984 1985 1986 1987 1988 1989 1990 199
YEAR YEAR
LOWESS SMOOTHED „„,. OUADRATIC REGRESSION
SOLID STRAIGHT LINE ISTHEVEAH PREDICTED LOG FLOW
LO»ESS SMOOTHED .-.— QUADRATIC REGRESSION
. SOLID STRAIGHT LINE 'IS THE MEAN PREDICTED LOG FECAL
Figure 3.—Continued.
ship between concentration and time (trend) can be
determined.
Time trends in flow- and season-adjusted resid-
uals are presented in Figure 4, including a scatter
plot of the flow and season adjusted residuals, a
smoothed line through the data, the regression line
for flow- and season-adjusted residuals against time,
and a reference line at zero. Figure 4 shows a sig-
nificant (90 percent confidence level) decreasing
trend in ammonium concentrations, significant in-
creasing (95 percent confidence level) trends in total
nitrogen and nitrate+nitrite nitrogen concentrations,
and no trend in total Kjeldahl nitrogen concentra-
tions. No trend in orthophosphate, total phosphorus,
and total suspended solids concentrations was evi-
dent; however, a significant (95. percent confidence
level) decreasing trend in total organic carbon con-
centrations was observed. Figure 4 shows no trend
in fecal coliform. When the data were kept at the
upper detection limit for the analysis, there was a sig-
nificant trend (90 percent confidence level) in am-
monium concentrations; however, decreasing trends
(95 percent confidence level) in total phosphorus
and orthophos- phorus were also significant.
Discussion
The project established two specific quality goals.
The first was to meet the State's water quality stand-
ards for fecal coliform at all times. The State water
quality standard for fecal coliform is 200 MPN/100
mL, based on a minimum of five samples taken
within a 30-day period (10 percent of the samples
taken during the 30-day period may not exceed 400
MPN/100 mL [Code Maryland Reg. 1989]).
Figure 5 shows the fecal coliform data collected
during this study and a reference line at the 200
MPN/100 mL level. Most of the data collected
during the study was above 200 MPN/lOOmL.
Seventy-one percent of the samples collected during
this study exceeded 400 MPN/100 mL. These data
indicate that fecal coliform standards are regularly
exceeded in the creek.
The BMPs implemented by this project were
designed to manage manure and reduce soil erosion.
The BMPs used in association with this project did
not appear to reduce fecal coliform densities in Big
Pipe Creek; therefore, additional work is required to
determine what practices would accomplish this ob-
jective.
The second specific water quality goal of the
project was to meet the State's water quality stand-
ard for turbidity at all times. The State standard for
turbidity states that "turbidity in the surface water
resulting from any discharge may not exceed 150
units at any time or 50 units as a monthly average"
(Code Maryland Reg. 1989). The units of measure-
ment may be NTU, JTU, or FTU. Between 1982 and
1990, the turbidity in Big Pipe Creek exceeded the
State standard on several occasions (Fig. 5). Using
the turbidity water quality standard as a measure of
nonpoint source pollution is a problem because the
standard was developed for point sources. (Turbidity
is closely associated with the concentration of
suspended solids in a stream.) As has been shown
(Fig. 2), considerable variation was evident in the
concentration of total suspended solids in Big Pipe
Creek. The variation in in-stream suspended solids
concentrations depends on flow, land use, rain fall in-
tensity, antecedent moisture, and a number of other
factors. The application of this standard during
storm events always produces substandard results.
Since this standard was not designed for nonpoint
source pollution, the selection of the State's turbidity
standard was an inappropriate objective for this
project.
187
-------
r
Proceedings of National RCWP Symposium, 1992
"« ""W?^!^^,^: *OJUSr" T,UE TREND ^^A^O, ^ ^SOH ADJUSTED
4
3
1
4 2
1
i
A 9
\
-I-
,J.
+ *
+* +* +
* * *
* +»*
* + *
V* +" * ++
* * *
\-v + * +*
1992 1983 1984 1985
-%msrat
. 4.
+ ***V * T^?3^
+ + 4-4. + 4- T +*
+ * +•
•**" +•
2-
I
K '•
R
D
U
A
S-
-•)•
+
* ++ +
*> * * . . '*Al^*i)tj^
+V4- JIf »' + l[ \*f* r\ I* ^ +\' ** TjT "•*• +•* •» ^"^
+ 4. + 4.^ + *
T 4-+ * 4-
+
4-
+
1986 1987 1988 1989 1990 1991 1982 1983 1984 1985 1966 1987 1988 1989 1990 1991
TEAR ' YEAR
.,._,_ OUAORATIC REGRESSION LOIESS SUOOTHED QUADRATIC REGRESSION
IS THTTuTATl PREDICTED LOG NH4 SOLID STRAIGHT LINE IS THE BEAN PREDICTED LOG TH
nn wmm w^w"™ TlkŁ TKMAWAWMiANpsrei
i-
j ,.
K *
1
I
1 '
A
^ -1-
-2'
f
>
-•*•.'•" ••*.'
V** *"% **" +^' *+i
+ 4-
1112 lit] 1984 1985
LCICSS SVOOIHED
SOUS STRAIGHT LINE
+
/
+*++*f \ *++
^ ^r^^\^^-
++• +
+
K
3 D-
4
K
0
2 -1-
R
D
U
•-J-
S
+ '- +
4.4i* +-"J$ i * *** -| |+ '1 l/^AJt^JTjfJ4l t 'V^
^+ *t 4f 4- *+ * +4-
+.
- - . : ,
+
1986 1987 1988 1989 1990 1991 1982 1983 1984 1985 I98S 1987 1988 1989 1990 1991
YEAR YEAR . '
,,-„-- QUAORATIC REGRESSION
IS IKE UEAK PREDICTED LOG TKH«
^__ LOIESS SMOOTHED QUADRATIC REGRESSION
SOLID STRAIGHT LIKE IS THE UEAN PREDICTED LOG M02I
me
1 ADJUSI"
TIUE TREND IN STREAK FIO« AHO SEASON ADJUSTED
TOTAL PHOSPHORUS RESIDUALS
«•
* * +
* * * 4. t *
*..4. *+ !++»*+*+.
« +* I *V v 4+ *
"* >T ' +' 4.^' '^"i" "^***SŁL
*" +4- + +* +4- t
* + 4" * 41 *
* + +*
•f- »^t -i^
*
1)2 191] 1984 I98S IS86 1987 1988 1989 1990 199
YEAR
4
3
T
' 2'
I ,.
D n-
5 "
A
L -1-
S
-2-
-3-
1 19
4-
4. "**•
* +* *
•f * " +
p-iJLLj-* V + , + +*S* *+L+ + r*h.— *
"*" . * * + H*1 + .
+ -f * +
* +VH •*••*• ' .
* + +
62 1983 1984 1985 1986 1987 1988 1939 (990 199
YEAR
,. IOICSS SUOOIHEO „,„.- OUAORATIC REGRESSION
55119 S1SAICH1 LINE IS THTTiTAN PREDICTED IOC P04
— LOIESS SUOOTHED -,-,,- QUADRATIC REGRESSION
SOLID STRAIGHT LINE IS THE UTAH PREDICTED LOG IP
5
4
? !^
s
i-l-
-J'
-4-
Tlllt IREN8 l» flVERFLOJ AND SEASON ADJUSTED TIUE TREND IK STREAU FLOI AHO SEASON ADJUSTED
TSIAl SUSPENDED SOLIDS RESIDUALS TOTAL ORGANIC CARBON RESIDUALS
*
•f
+«* Ł 4-4-
4- ^ +
^lli. t + * 4- 4-'*' * 4- *^ a. **
^**«*i.n t Jj __4Jt4H-4- +. |*-J-
** /*t * C^^^ir^^vT'^Tc" "
- * + * ** ** * +
*»*
2-
f
S 0'
0
u
s-'
-2-
+ t t
+^i! \+ + x t + \ *
+ ^7r"""^rr^n+"
•** + + + + v
+
1912 111] 1984 1985 I98S 1987 1988 1989 1990 1991 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991
YEAS YEAR
LCICSS SU901HEI ,,,,,- OUAORATIC REGRESSION
SOUS SISAICHI LIKE IS THE UEAH PREOICIEO LOG TSS
LOIESS SUOOTHED OUAORATIC REGRESSION
SOLID STRAIGHT LINE IS THE UEAH PREDICTED LOG TOC
Figure 4.—Time trends at the Big Pipe Creek site at Brucevllle.
188
-------
J.L McCOY&R-M. SUMMERS
HUE TREND IN RIVER FLOW AND SEASOH ADJUSTED
FECAL COLIFORU BACTERIA RESIDUALS
1982 1983 1984 1985
I98S 1989 1990 1991
IOIESS SUOOIHED
SOLID STRAIGHT LIN
1986 1987
YEAR
QUADRATIC REGRESSION
IS THE »EAN PREDICTED LOG FCOL
Figure 4.—Continued.
The more general goal of this project was to im-
prove water quality. To assess the degree of ex-
pected change in water quality, the magnitude of
change resulting from the land treatment should be
estimated. From 1982 through 1990, the RCWP and
farmers in the basin invested approximately
$3,400,000 in best management practices. The
RCWP's investment in the basin was approximately
$2,700,000. The installation of BMPs in the basin
resulted in an estimated 25,646 ton reduction of in-
field soil erosion from cropland per year and the
storage of 99,919 tons of manure per year by 1989.
Based on an average erosion rate of 9.6 tons of soil
per acre of cropland for the basin, the net reduction
of in-field soil erosion from cropland was ap-
proximately 4 percent. Based on the livestock
population numbers for 1989, the net quantity of
stored manure increased by 28 percent (Table 1).
The reduction in soil erosion and the increase in
stored manure can be converted into pounds of
nitrogen (N) and phosphorus (P) using the conver-
sion factors of 1.1 pounds of P/ton and 5.4 pounds of
N/ton for soil and 1.3 pounds of P/ton and 7.0
pounds of N/ton for manure. (U.S. Environ. Prot.
Agency, 1987). The net reduction of nutrients avail-
able from eroded soil and manure as a result of the
implementation of BMPs was 13 percent for both N
and P.
The water quality results to date indicate that,
during the period from 1982 through 1990, am-
monium and total organic carbon concentrations
decreased with time. Based on the predicted values
from the regression of the flow- and season- adjusted
residual against time, ammonium concentrations
have decreased 44 percent and total organic carbon
concentrations have decreased 51 percent over the
period.
The decreasing trends in ammonium and total
organic carbon concentrations are encouraging.
They indicate that some of the more mobile forms of
nutrients and organic matter are being held up in the
system, possibly the result of land treatment spon-
sored by the Double Pipe Creek RCWP. The project
increased the storage of animal waste in the basin by
approximately 99,919 tons/yr and decreased the
quantity of nitrogen and phosphorus readily avail-
able from manure for transport to the stream system
by 28 percent. The trends may indicate that less
manure is being washed off the land surface (am-
monium and total organic carbon are major con-
stituents in manure runoff).
Estimates show that the project has reduced the
quantity of nutrients available for export from soil
erosion and manure application in the basin by ap-
proximately 13 percent (Table 1). This reduction has
not shown up in orthophosphate and total phos-
phorus constituent trends because a 13 percent
change is too slight to be statistically detected given
the large natural variability in these concentrations
with changes in river flow. Orthophosphate and total
phosphorus are more generally associated with soil
erosion and sediment transport and are thus more
influenced by storm events than ammonium and
total organic carbon, making trends much more dif-
ficult to detect.
Similarly, estimates indicate that the project has
reduced soil erosion by only 4 percent. This reduc-
tion has not shown up in total suspended solids con-
stituent trends because the magnitude of the change
is too small to be statistically detected given the
variability of total suspended solids concentrations
with storm events. '
The increasing trend in nitrate+nitrite nitrogen,
which leads to a corresponding increase in total
nitrogen, indicates that the soluble forms of nitrogen
are leaching through the system. A variety of factors
Table 1.—Manure, soil, and nutrient savings resulting from Double Pipe Creek RCWP implementation.
Manure
Soil
Total
TOTAL —
(TONS/YR)
361 ,228
696,960
NUTRIENT EQUIVALENT
(TONS/YR)
N
1,264
1,882
3,146
P
235
383
618
REDUCTION
(TONS/YR)
99,919
25,646
NUTRIENT EQUIVALENT
(TONS/YR)
N
350
69
419
P
65
14
79
PERCENT
REDUCTION
N
28
4
13
P
28
4
13
189
-------
Proceedings of National RCWP Symposium, 1992
FCOL
50000H
40000-
30000-
20000-
10000-
oH
01JAN81
A
1
01 JUN83
1
01JAN86
DATE
FECAL
01 JUN88
STATE STANDARD 200 MPN/100 ML
BIG PIPE CREEK AT BRUCEVI LLE
TURBIDITY NTU
STATE STANDARD 150 NTU
—T
01 JAN91
TURB
2200-
2000-
1800-
1600-
1400-
1200-
1000-
800-
600-
400-
200-
o-
01J
1
A lfl
J\ /^ ^y\^ Jl
UlAl
- i
i
ll
A 1
L^LjuAjluullL
i i i i i
ANSI 01JUN83 01JAN86 01JUN88 OIJAN91
DATE
TURB
Figure 5.—Big Pipe Creek at Brucevilfe, fecal coliform MPN/100 mL.
5s probably contributing to the observed increase in
nitrate+nitrite nitrogen. The increased use of conser-
vation Ullage in the region over the last 15 years may
have increased infiltration and thus increased the
leaching of soluble chemicals into the groundwater.
The increased storage of animal waste has reduced
nitrogen losses to the atmosphere and to streams
through direct runoff. However, storage of animal
190
-------
J.L. McCOY&R.M. SUMMERS
waste does result in more manure being applied to
fields, which increases the potential for leaching.
The proper timing of manure applications and the in-
corporation of applied manure also reduces atmos-
pheric loses of nitrogen and increases the quantity of
nitrogen being applied. Increased atmospheric
deposition of nitrogen and increasing numbers of
residences with septic systems also contribute to the
nitrogen load increases.
Future Work
The water quality monitoring program is scheduled
to continue indefinitely. Water quality monitoring at
the single land use site will be extend into 1993,
depending on the implementation schedule for
BMPs. The 15-year report for the Double Pipe Creek
RCWP will include a trend analysis for 1982 through
1994 and a before-and-after study of BMP effective-
ness at the single land use site.
References
Cleveland, W.S. 1979. Robust locally weighted regression and
smoothing scatter plots. J. Am. Stat Ass. 7:829-36.
Code of Maryland Regulations. 1989. 26.08.02, Water Quality. Div.
State Doc., Off. Secretary State, Annapolis.
Grant, D.M. 1985. Open Channel Flow Measurement Handbook.
Isco Inc., Lincoln, NE.
Maryland Department of the Environment 1989. Quality As-
surance/Quality Control Han, Nonpoint Source Water
Quality Assessment of the Double Pipe Creek Watershed,
Phase II. Water Manage. Div., Baltimore.
Maryland Department of Health and Mental Hygiene. 1981. Field
Procedures Manual. Water Qual. Monitor. Div., Water
Manage. Admin., Off. Environ. Prog., Annapolis.
U.S. Environmental Protection Agency. 1987. Chesapeake Bay
Nonpoint Source Programs. Region 3, Chesapeake Bay
Liaison Off.,.Annapolis.
Versar Inc. 1982. Quality Assurance Plan for the Non-point Source
Water Quality Assessment of the Monocacy River Basin with
Special Attention to the Double Pipe Creek Watershed.
Springfield, VA.
:—. 1986. Nonpoint Source Assessment of the Monocacy River
with Special Emphasis in Double Pipe Creek Watershed.
Springfield, VA
191
-------
-------
Effects of Nutrient Management on
Surface Water Quality in a Small
Watershed in Pennsylvania
Edward H. Koerkle
U.S. Geological Survey
Lemoyne, Pennsylvania
ABSTRACT
The effects of nutrient-management practices on surface water quality were investigated from 1984
to 1989 in a mostly agricultural watershed underlain by carbonate rock in the Conestoga River
headwaters in Lancaster County, Pennsylvania. Nutrient management applied to approximately
half the cropland in the watershed resulted in no significant change in median base-flow nitrate
nitrogen concentrations after 3.5 years. Changes in surface water quality were evaluated using a
pre- and posttreatment design supplemented by a paired-subbasin experiment Base-flow nutrient
concentration data were collected for 24 months prior to implementation and for 42 months after
the implementation of nutrient management. Dissolved nitrate plus nitrite concentrations in month-
ly base-flow samples from the watershed ranged from 1.7 to 14 mg/L as nitrogen largely as a result
of seasonal- and flow-related dependencies. Dissolved nitrate plus nitrite concentrations exceeded
10 mg/L as nitrogen in 3 percent of the samples collected before and 9 percent of the samples col-
lected after implementation of nutrient-management practices. An average 67 percent of the month-
ly nitrogen load in surface water discharged from the watershed was nitrate nitrogen.
The paired subbasins consisted of two 1.4 square-mile subbasins — one in which nutrient
management was not prescribed and one in which nutrient-management practices were applied to
90 percent of the tillable land. No significant change was detected in median dissolved nitrate plus
nitrite concentrations in the base flow despite an estimated 35 percent decrease in nitrogen applica-
tions to cropland in the nutrient-management subbasin. No significant change was detected in
median dissolved nitrate plus nitrite concentrations in the base flow in the nonnutrient-manage-
ment subbasin. Analysis of covariance detected a significant change in the relation between con-
current dissolved nitrate plus nitrite concentrations in base flow from the subbasins. Dissolved
nitrate plus nitrite concentrations exceeding 10 mg/L as nitrogen were detected in 3 percent of
base-flow samples collected from the nutrient-management subbasin both before and after im-
plementation of nutrient management.
Trends toward expanding farm animal popula-
tions, although of substantial economic
benefit to agricultural producers in the
United States, have left many producers with large
quantities of excess manure. This manure is com-
monly disposed of by application to cropland, and, in
some areas, it greatly exceeds commercial fertilizers
as the largest source of nutrients to agricultural
lands. The result can be excessive fertilization and
nonpoint-source contamination of surface and
groundwater resources. In one small watershed in
the Conestoga River headwaters in Lancaster Coun-
ty, Pennsylvania, excessive fertilization from
repeated and localized land applications of manure,
in addition to applications of commercial fertilizers,
has been implicated as a major source of excessive
nutrient concentrations in the surface and
groundwater resources of the area (U.S. Dep. Agric.
1984). Nitrate concentrations in stream base flow
average about 7 mg/L as nitrogen and have oc-
casionally exceeded the 10 mg/L Maximum Con-
taminant Level Goal for drinking water (U.S.
193
-------
Proceedings of National RCWP Symposium, 1992
Environ. Prot. Agency, 1986). Reducing the effect of
nutrient applications on water quality through
nutrient management, an agricultural best manage-
ment practice (BMP), is a goal of the Rural Clean
Water Program (RCWP) projects. Although im-
plementation of BMPs is an important step toward
the goal of improved water quality, more specific in-
formation about the effectiveness of individual
BMPs is needed. To provide that information, a 5.5-
year investigation of the effects of the nutrient-
management BMP on surface water quality was
completed by the U.S. Geological Survey (USGS)in
cooperation with the Pennsylvania Department of
Environmental Resources (PADER), Bureau of
Water Quality Management, under the aegis of the
RCWP.
Purpose and Scope
This paper summarizes the effects of nutrient
management on surface water quality in a small
watershed (5.82 square miles) and a smaller sub-
basin (1.4 square miles) in the watershed located in
the Conestoga River headwaters (Fig. 1). Hydro-
logic, BMP-implementation, and agricultural-activity
data were collected from April 1,1984, through Sep-
tember 30, 1989 — from 2 years prior to the im-
plementation of the nutrient-management BMP to
3.5 years after implementation. Also presented are
qualitative and statistical inferences regarding the
effects of nutrient management on surface water
quality, principally base-flow nitrate-nitrogen con-
centrations and loads.
Methods
The investigation consisted of a before and after
treatment experiment and a paired-subbasin experi-
ment. The experimental treatment was the im-
plementation of nutrient management, that is, the
use of animal-waste and fertilizer best management
practices as defined by the Agricultural Stabilization
and Conservation Service (ASCS). The paired sub-
basins were added to the study after preliminary
contacts with landowners indicated that contracted
nutrient-management implementation would be
clustered in a specific area of the watershed. This
76'09'
75°52'
PROJECT
AREA
40'OS1
Small Watershed
Study Area ,
EXPLANATION
CARBONATE BOCK
NONCARBONATE ROCK
- . REGIONAL STUDY AREA
SMALL WATERSHED STUDY AREA
• FIELD SITE STUDY AREA
COUNTY LINE
2 4 MILES
I1 I
4 KILOMETERS
Figure 1. — Location of the small watershed.
194
-------
E.H. KOERKLE
area, the nutrient-management subbasin, was lo-
cated in the eastern part of the watershed where
about 90 percent of the agricultural land was under
RCWP contract. The nonnutrient-management sub-
basin, where farm operators chose not to participate
in the RCWP, was located in the northwestern part of
the watershed.
The evaluations presented in this report on the
effects of nutrient management on surface water
quality are qualitative. Initially, a major objective of
the investigation was to quantify the effects of
nutrient management on base-flow water quality.
However, quantitative evaluation requires separation
of all non-BMP influences from BMP influences on
water quality and the development of a statistically
and functionally sound relationship between agricul-
tural activities and water quality. Although a simple
multilinear regression model of the relationship was
constructed, factors, such as undefined variation in
nutrient-application timing and location and ground-
water recharge timing and quantities, prevented the
use of the regression model.
Data Collection
Data including precipitation, agricultural activity, and
water quality were collected according to a collection
schedule summarized in Table 1. Precipitation data
were recorded by an accumulating precipitation
gage located near the southern boundary of the
study site. Nutrient-management plans were
prepared by the Lancaster County, Nutrient-Manage-
ment Office operated by the Pennsylvania State
University Cooperative Extension Service. Agricul-
tural-activity data from the nutrient-management
subbasin were collected by staff from the Lancaster
County ASCS during semiannual interviews with
farm operators. Information collected included
statistics on
• animal populations,
• applications of manure and commercial
fertilizer,
• manure exports, and
• crop acreages.
Agricultural-activity data for remaining areas of
the watershed were limited to crop acreages and
animal populations collected during initial contact by
RCWP personnel. Estimates of manure nutrients
produced in, applied in, and exported from the
nutrient-management subbasin were calculated by
multiplying manure quantities produced, applied, or
exported by nutrient-content values supplied by the
ASCS, for livestock and poultry manure. The
manure quantities produced were calculated from
Table 1.—Data collection schedule for the small
watershed study.
SOURCE
2 continuous-
record stations
3 partial-
record stations
7 soil-sampling
sites
1 precipitation
gage
DATATYPE FREQUENCY
Water quality Monthly during
base flow and
major storms
Water quality Monthly during
base flow
Soil nutrients Spring and fall
Precipitation 5-minute intervals
intensity and total
accumulation
13 farm operators1 Agricultural activity Spring and fall
1Nutrient-management subbasin only.
animal-unit averages for the pre- and post-BMP
periods.
Streamflow was recorded continuously at sta-
tions that drained the nutrient-management sub-
basin and the entire watershed. Instantaneous
streamflow was measured at the time of base-flow
sampling at all other stations. Base-flow water quality
samples were collected by USGS and FADER per-
sonnel and measured or analyzed for total and dis-
solved forms of ammonia, ammonia plus organic
nitrogen, nitrate plus nitrite, nitrite, and total phos-
phorus. All nutrient concentrations are expressed as
elemental nitrogen for the nitrogen species and
elemental phosphorus for the phosphorus species.
All statistical tests were evaluated for significance at
the 95 percent confidence interval (p n 0.05). Addi-
tional detail on data collection and methods for the
Conestoga Headwaters RCWP are published in
Chichester (1988).
Site Description
The 5.82-square mile watershed (Fig. 2), about 50
percent underlain by carbonate rock, is drained by
the Little Conestoga Creek and includes all or part of
43 farms. The nutrient-management subbasin covers
1.42 square miles and the nonnutrient-management
subbasin covers 1.43 square miles. Both subbasins
are underlain by carbonate and noncarbonate rock,
but about 24 percent less of the nonnutrient-manage-
ment subbasin is underlain by carbonate rock. Land
use in the watershed and the subbasins is
predominantly agriculture (Table 2). During the
study, about 50 percent of the total crop acreage in
the nutrient-management subbasin was planted in
corn; 22 percent, in hay (alfalfa or alfalfa and grass
mixes), and about 8 percent was pasture. Limited
land-use data for the nonnutrient-management sub-
basin indicates that about 65 percent of the cropland
was planted in corn.
195
-------
Proceedings of National RCWPSymposlum, 1992
76«57'30"
76" 55'
' 10'
<0'08'30'
EXPLANATION
Iv-.-j'-.' 'I NONCARBONATE ROCK
' ' CARBONATE ROCK
01576035 ^ CONTINUOUS-RECORD STATION AND ID
NUMBER
015760833^ PARTIAL-RECORD STATION AND ID
.
I
/
h.
•; . : '• • '• •x.
.*r'f.'.'.'.'• '. '•'• \
.:-:\-.V/:.;.V..V.vv-;-v:;v:v.
' • J-; STATION 1 I'.''. '.'.'.' '". '.. '.' '.''.'•
• '• '/•. 015760831 V '• '• -'•-'• - • - ,„•„'• -<1
'•.•/&??& —H
,-..-<^"'
015760839
-..STATION 4
/
' 0 0.5 MILES
r [III I I
0 0.5 KILOMETERS
V...-
Figure 2.—Surface water and precipitation station locations.
Table 2.—Land use In the watershed by percent of
total land-use area (computed from U.S. Department
of Agriculture photographs).
Agriculture
Woodland
Urban
WATER-
SHED
76.0
21.4
2.6
NUTRIENT-
MANAGEMENT
SUBBASIN
78.0
20.6
1.4
NONNUTRIENT-
MANAQEMENT
SUBBASIN
68.0
30.6
1.4
• Precipitation. The long-term average annual
precipitation for the watershed is approximately 41.5
inches. Precipitation during each of the study years,
1985, 1987, and 1988 was within 5 percent of the
long-term average. Precipitation averages for the
first year of the pre-BMP period, 1984, and the first
year of the post-BMP period, 1986, were 14 and 25
percent below the long-term average, respectively.
• Streamflow. At Station 3, the average flow for the
study period was 0.9 cubic foot per second per
square mile and, at Station 5, the flow was 1.0 cubic
foot per second per square mile. The base-flow con-
tribution to total monthly streamflow averaged 65
percent.
• Best Management Practices and Agricultural
Activities. Nutrient-management implementation
began in April 1986. In the first year, nutrient-
management plans were prepared for 11 of 14 farms
in the nutrient-management subbasin and 13 of 43
farms in the entire watershed (Fig. 3). By 1989, 11
additional farms in the watershed had implemented
nutrient-management plans. Farm operators in the
nonnutrient-management subbasin did not par-
ticipate in the RCWP or supply agricultural-activity
data; therefore, farming practices there were as-
sumed to remain constant. However, some farms did
receive nutrient-management plans during the
period of the study.
In the nutrient-management subbasin, the plans
included fertilizer management (both manure and
commercial fertilizer) on 10 of the cooperating
farms, and animal-waste management on one farm.
The nutrient-management plan supplied nutrient-ap-
plication recommendations intended to allow desired
crop yields and minimize nutrient availability for
transport to groundwater or in surface runoff. Not
all plans recommended reductions in nutrient ap-
plications. Nutrients deposited by grazing livestock
196
-------
E.H. KOERKLE
NUTIIENT MANAGEMENT FUNS • 1986
"V
NUTIIENT MANAGMENT PLANS • 1)87
NUTIIENT MANAGEMENT PLANS • 191!
NUTIIENT MANAGEMENT PUNS • Hit
EXPLANATION
Farms with plans
•— Watershed boundary
•• Subbasin boundary
Farm boundary
A Gaging station
<^ Precipitation station
Figure 3.—Annual nutrient management plan census from 1986 through 1989 for farms in the watershed.
were not accounted for in the nutrient-management
plans. Because nutrient management plans may re-
quire manure storage for as many as 180 days, one
farm with a large swine population increased its
manure storage capacity to 200 days by the addition
of a storage tank.,
Animal populations in the watershed included
beef and dairy cattle, sheep, swine, poultry, and draft
animals. In the nutrient-management subbasin
animal populations on individual farms varied con-
'siderably from year to year and within a year but, in
the subbasin overall, the proportioning of animal
types was generally stable. The total animal popula-
tion consisted of about 50 percent poultry, 30 percent
dairy, and 15 percent swine by weight. Estimates of
manure and manure-nutrient production indicate
that poultry and dairy manure contributed, by
weight, 40 and 35 percent, respectively, of the total
manure-nutrient production (Table 3). Manure-
nutrients production decreased about 5 percent in
.the post-BMP period.
Agricultural-activity data indicate that manure
applications were the largest nutrient input to
cropland (Table 4). Grazing livestock deposited an
estimated additional 22,000 and' 18,000 Ib of nitrogen,
respectively, and 4,200 and 3,500 Ib of phosphorus,
respectively, annually in the pre- and post-BMP
periods. The contribution of nitrogen and phos-
phorus from commercial fertilizers (Table 4) was
nearly equal to that from manure deposited by graz-
ing livestock. Average annual applications of
nitrogen and phosphorus to cropland in the nutrient-
management subbasin decreased about 35 percent
from the pre- to post-BMP period. Not all of the
decrease was a result of nutrient management;
reductions in the number of animal units and in the
acreage being actively farmed accounted for an es-
timated 5 percent*of the 35 percent decrease. Be-
cause of uncertainties in the agricultural-activity
data, actual reductions in nutrient applications may
differ from estimated reductions by 20 percent or
more.
Comparison of manure production and applica-
tion data.show that between 30 and 50 percent of the
estimated manure production was not accounted for
either as applications or export. A large part of the
"missing" production was probably applied to
cropland outside the subbasin. Because this manure
197
-------
Proceedings of National RCWP Symposium, 1992
Table 3.—Manure and nutrient production in the nutrient-management subbasin.
MANURE
TYPE
Dairy
Beef
Swine
Poultry
Sheep
Horse/mule
Annual
average
PERIOD
Pre-BMP
Post-BMP
Pre-BMP
Post-BMP
Pre-BMP
Post-BMP
Pre-BMP
Post-BMP
Pre-BMP
Post-BMP
Pre-BMP
Post-BMP
Pre-BMP
Post-BMP
AU
476
472
120
107
218
188
254
247
30
0
26
26
1,124
1,040
MANURE
PRODUCTION (tons)
14,760
25,610
2,640
4,120
3,840
5,790
5,590
9,510
396
0
863
1,5-10
14,050
13,300
NITROGEN
PRODUCED (Ib)
147,600
256,100
29,040
45,320
53,760
81,060
167,700
285,300
8,710
0
10,360
18,120
208,600
196,000
PHOSPHORUS
PRODUCED (Ib)
26,570
46,100
9,240
14,420
16,900
25,480
49,190
83,690
1 ,390
0
1,900
3,320
52,600
. 49,430
AU « animal units; BMP =
September 30,1989.
best management practice; pre-BMP = April 1,1984, through March 31,1986; post-BMP = April 1,1986, through
Table 4.—Nutrient applications, In pounds, to cropland in the nutrient-management subbasin.
NITROGEN PHOSPHORUS
DATE
PERIOD MANURE + COMMERICAL FERTILIZER = TOTAL MANURE + COMMERICAL FERTILIZER = TOTAL
April 1984-March 1985
AprH1985-March 1986
April 1986-March 1987
April 1987-March 1988
April 1988-March 1989
April 1989-Sept. 1989
Annual average
Pre-BMP
Pre-BMP
Post-BMP
Post-BMP
Post-BMP
Post-BMP
Pre-BMP
Post-BMP1
83,500
107,000
62,300
53,800
73,000
37,400
95,250
63.000
25,100
24,600
20,200
18,300
18,300
10,900
24,850
18,900
108,600
131,600
, 82,500
72,100
91,300
48,300
120,100
82,000
21,200
28,400
14,300
13,700
19,200
8,700
24,800
15,700
7,100
6,100
5,300
4,200
4,700
2,800
6,600
4,700
28,300
34,500
19,600
17,900
23,900
11,500
31 ,400
20,500
'April 1989 through September 1989 excluded when calculating average.
did not leave the farm on which it was produced, it
was not recorded as export. If the unaccounted
manure production remained in the subbasin, es-
timated decreases in nitrogen input would be sub-
stantially less than stated.
Effects of Nutrient
Management on Surface Water
Quality
Monthly Constituent Loads
Monthly nitrogen, phosphorus, and suspended-sedi-
ment loads for Station 3 (Fig. 4), the nutrient-
management subbasin, and Station 5 (Fig. 5), the
small watershed, were calculated by summing daily
base-flow loads and stormflow loads for the month.
Daily base-flow loads were calculated by multiplying
daily base-flow discharge by the constituent base-
flow concentration. Daily base-flow concentrations
were estimated by linear interpolation from con-
centrations measured monthly in the base flow.
Stormflow loads were calculated by the subdivided
method described by Porterfield (1972) for sampled
stormflows and by a mean concentration to dis-
charge regression for stormflows not sampled.
The monthly nitrogen loads consisted mostly of
nitrate plus nitrite and the proportion of nitrate plus
nitrite in the monthly nitrogen loads did not change
significantly from the pre- to post-BMP period. Maxi-
mum monthly loads of total phosphorus and
suspended sediment at Stations 3 and 5 were lower
during the first year of the post-BMP period than
during the pre-BMP period. Correspondingly,
precipitation was 25 percent below average in 1986.
From April 1987 through the remainder of the post-
BMP period, statistically significant increases in total
phosphorus loads were detected at Station 5 and
statistically significant increases in the sediment
load were detected at Station 3. Increases in
198
-------
E.H. KOERKLE
6.000 |
5.500
5,000
Q
^ +.500
O
^ 4.000
Q 3.500
Q
UJ
3,000
2.500
^ 2.000
_l
j< 1.500
1,000
500
PRE-BMP
POST-BMP
TOTAL AMMONIA+ORGANIC NITROGEN
TOTAL NITRATE+NITRITE NITROGEN
MAMJJASONDJFMAMJJASONDJFMAMJJASONDJFMAMJJASONDJFMAMJJASONDJFMAMJJASO
1984 1985 1986 1987 ,'1988 1989
1.400
- 1.200
a
o
1.00O
o
UJ
00
O
a.
BOO
600
O *°0
O
a.
O
a.
200
ji_ 18 a ll
TOTAL PHOSPHORUS
SUSPENDED SEDIMENT
MAMJJASONDJFMAMJJASONDJFMAMJJASONDJFMAMJJASONDJFMAMJJASONDJFMAMJJASO
1984 1985 1986 1987 1988 1989
Figure 4—Monthly loads of nitrogen as N (above) and total phosphorus as P and suspended sediment (below) at Sta-
tion 3. .
199
-------
Proceedings of HattonatRCWP Symposium, 1992
60.000
so.ooo
in
o
z
=>
O. 40,000
Z
O
-1 30.000
Z
Ul
o
o
Z 20,000
10,000
PRE-BMP
In*
POST-BMP
TOTAL AMMONIA+ORGANIC NITROGEN
TOTAL NITRATE+NITRITE NITROGEN
°MAl!(ijASO^DJFMAMJUs6NDJFMAMJJAs6NbJFMAMJJAS
1984 1985 1986 1987 1988 1989
t/J
Q
9,000
8,000
7.000
6,000
J 5,000
4,000 •
iS 3,000
V)
o
Q.
2,000
0
0
0
0
0
0
0
0
0
I
19
-
•
|
JA
84
I
LIJ
PRE-BMP
nHfl
,J
L
POST-BMP .
^ TOTAL PHOSPHORUS
|
j..lL^
1
\
,
j
ill
• !
1
SUSPENDED SED
1
In
IMENT
Pii
•
la a.
ASONDJ MAMJJASO D J FMAM J J ASOND j FM AM J J ASONO J FM AM J J ASOND J FM AM J J ASC
1985 1986 1987 1988 1989
Figure 5.—Monthly loads of nitrogen as N (above) and total phosphorus as P and suspended sediment (below) at Sta-
tion 5.
.200
-------
E.H. KOERKLE
streamflow accounted for most of the increases in
the monthly loads.
Monthly streamflow was factored out of each
monthly load to produce a discharge-weighted mean
concentration that could be used to determine if
monthly mean concentrations of nutrients Changed
significantly from the pre- to post-BMP period. When
pre-BMP and post-BMP monthly discharge-
weighted mean concentrations were compared, no
significant change was detected in total nitrogen, dis-
solved nitrate plus nitrite, total ammonia plus or-
ganic nitrogen, total phosphorus, or suspended
sediment at either Station 3 or 5.
Base-flow Nitrogen
Monthly base-flow nitrogen concentration data col-
lected at Stations 1 through 5 are summarized in
Figures 6 through 8. The data are separated into the
pre- and post-BMP periods.
Within the nutrient-management subbasin, base-
flow concentrations of dissolved nitrate plus nitrite,
which is primarily in groundwater, increased with
downstream distance (Fig. 6). This increase corre-,
lated with increasing proportions of agricultural land
use and carbonate rock in the contributing drainage
area. In the nonnutrient-management subbasin (Sta-
tion 4), median dissolved nitrate plus nitrite con-
centrations were 3 mg/L lower than in the
nutrient-management subbasin during both the pre-
and post-BMP periods. Although the difference in
nitrate plus nitrite concentrations in base flow be-
tween basins could be a result of the difference in
proportion of agricultural land use (Table 2),
Thomas and others (1992) propose (1) that the
parent geology of a watershed is of greater impor-
tance in explaining average nitrate concentrations in
streamflow than agricultural land use, and (2) that
the largest nitrate concentrations occur in stream-
flow from carbonate rocks. The nonnutrient-manage-
ment subbasin is underlain by about 24 percent less
carbonate rock than the nutrient- management sub-
basin. Fishel and Lietman (1986) found that median
concentrations of nitrate in groundwater, the source
of base flow, were three times greater on average in
areas of the Conestoga River Headwaters underlain
by carbonate rock. The presence of sinkholes and
large fractures in carbonate rock allow direct and
rapid transport of surface applied nutrients to the
groundwater. Nitrate plus nitrite concentrations in
the base flow from the entire watershed (Station 5)
were similar to concentrations at Station 3, though
concentrations at Station 5 had greater variation.
The drainage basins for Stations 3 and 5 are under-
lain by approximately equal percentages of car-
bonate rock.
Unlike nitrate, ammonia nitrogen and total am-
monia plus organic nitrogen are commonly found
CONCENTRATION OF DISSOLVED NITRATE PLUS NITRITE,
IN MILLIGRAMS PER LITER AS NITROGEN
3 M *. 01 to 0 M S 01
STATION 1 STATION 2 STATION 3 STATION 4
X
30 , 4Q\
R ^
X
X , X
T I
X
I x
fl8
n*i
1
'
X
1
42
r
T
"T-r"
STATION 5 ,
30 ;;
7
X
i-
EXPLANATION
o DETACHED VALUE1
X OUTSIDE VALUE2
4 I UPPER WHISKER3
"-1—1 75TH PERCENTILE
T
k/EDlAN
25TH PERCENTILE
LOWER WHISKER3
1A value »3 times the interquartile
range from the box.
2 A value >1 JS and Ł3 times the
interquartile range from the box.
3 Upper whisker i> the largest data
point less than or equal to the
upper quartte plus 1.5 times the
Interquartile range. Lower whisker
is the smallest data point greater
than minus 1.5 times the
. interquartile range.
4 n - number of observations in
analysis.
{ | - PRE-RMP (4-84 to 3-86)
Ey?! - POST-BMP C4-86 to 9-89)
Figure 6.—Concentrations of dissolved nitrate plus nitrite nitrogen in base flow at Stations 1 through 5.
201
-------
Proceedings of National RCWP Symposium, 1992
only in trace concentrations in groundwaters of the
area. Rather, concentrations of these constituents in
surface water have their source in and near the
stream channel. With the exception of Station 1, con-
centrations of each of these constituents were
similar in water samples collected at all stations sug-
gesting similar in- or near-stream conditions (Figs. 7
and 8). Stations 2 through 5 are located in agricul-
tural settings and Station 1 is located in a forested
setting.
Concentrations of nitrate plus nitrite in base flow
exhibited seasonal variation and flow dependency.
2.5
OO
20
§2 '
52
So
m
02
§Z 0.5
O
STATION 1 STATION 2 STATION 3 STATION 4 STATION 5
37
EXPLANATION
o DETACHED VALUE1
X OUTSIDE VALUE2
« I UPPER WHISKER3
'-'—• 7STH PERCENTILE
T
MEDIAN
25TH PEHCEMTILE
LOWER WHISKER3
- POST-BMP, (4-86 IO 9-89)
1A value >3 times iha intan)uartile
range from me box.
2 A value >1 Ł and S3 times the
interquartile range from the box.
3 Upper wtiiiker is tha largest dala
point less than or equal to the
upper quanile plus 13 times the
interquartile ranga. Lower whisker
is the smallest data point greater
than minus 1.5 times the
interquartile range.
4 n - number of observations in
analysis.
| | - PRE-RMP (4-84 to 3-86)
Figuro 7.—Concentrations of dissolved ammonia nitrogen in base flow at Stations 1 through 5.
o
o
Is
to
STATION 1 STATION Z STATION 3 SJATION 4 STATION 5
CO 2
is 6
rfOi 6
— H
zE
0-»
2oi
3K
-J«2 ,
pee
H-o
r
tu
o
o
o
30
x
X
1A value >3 times the interquartile
range from the box.
z A value >1 Ł and 53 times the
interquartile range from the box.
3 Upper whisker is the largest data
point less than or equal ID the
upper quarile plus 1.5 times the
Interquartile range. Lower whisker
is the smallest data point greater
than minus 13 times tfie
interquartile range.
4 n > number of observations in
analysis.
p-^t - POST-BMP (4-86 to 9-89) | | - PRE-RMP i (4-84 to 3-86)
Figure 8.—Concentrations of total ammonia plus organic nitrogen in base flow at Stations 1 through 5..
X
EXPLANATION
o DETACHED VALUE1
X "OUTSIDE VALUE2
30 4 I UPPER WHISKER3
•-*—• 7STH PERCENTILE
T
MEDIAN
2STH PERCENTILE
LOWER WHISKER3
202
-------
E.H. KOERKLE
Examples of the seasonal variation can be seen in
the time-series plot for Station 5 (Fig. 9). The small-
est seasonal variation was found at Station 1 and the
greatest variation was found at Station 5. Seasonal
adjustment was employed in the test for step trends
between the pre- and post-BMP periods. Flow de-
pendency was removed from the observed data by
subtracting observed concentrations from a locally
weighted scatterplot smooth estimate of the con-
centration/discharge function (LOWESS) (Cleve-
land, 1979) and adding the overall mean concen-
tration to the difference. Flow-adjusted concentra-
tions were then tested for trends associated with the
implementation of nutrient management.
Pre-Post Comparison
Before comparing pre- and post-BMP nitrogen con-
centrations in the base flow, the first year of post-
BMP data (April 1986 through March 1987) was
deleted from the data set for the following reasons:
• not all farms in the nutrient-management sub-
basin received their nutrient-management
plans exactly at the start of the post-BMP
period, and
• a lag time was expected between the im-
plementation of nutrient management and any
response in the quality of base flow. At a
hydrogeologically similar site also located in
the Conestoga River headwaters, a lag of
about 6 months to 2 years was observed
before nitrate nitrogen concentrations in
groundwater responded to changes in sur-
face-applied nitrogen (D.W. Hall, pers. comm.
1991).
A seasonally adjusted rank-sum test was used to
determine whether a step trend had occurred in
nitrogen concentrations between the pre- and post-
BMP periods. Few significant changes in median
concentrations of the nitrogen species occurred in
base flow between the pre- and post-BMP period for
observed or flow-adjusted data (Table 5). Median
concentrations of nitrate, the predominant base-flow
nitrogen species, did not change. Significant
decreases of 0.10, 0.03, and 0.28 mg/L, were
detected in dissolved-ammonia concentrations at Sta-
tions 1 and 4 and in total ammonia plus organic
nitrogen concentrations at Station 5, respectively.
These decreases represent a very small percentage
of the median total nitrogen concentrations in base
flow. Moreover, because these constituents have
their source in or near the stream channel, they are
most likely not a result of nutrient management.
Water quality goals established at the start of the
Conestoga Headwaters project (U.S. Dep. Agric.
1984) stated that maximum nitrogen concentrations
should not exceed 10 mg/L nitrate as nitrogen and
1.5 mg/L ammonia as nitrogen. Dissolved-nitrate
plus nitrite concentrations exceeded the 10 mg/L
goal in 3 and 9 percent of the samples collected
before and after nutrient management, respectively,
at Station 5. At Station 3, in the nutrient-management
subbasin, the dissolved nitrate plus nitrite concentra-
tions exceeded 10 mg/L in 3 percent of the base-flow
samples collected both before and after nutrient
management implementation. Dissolved-ammonia
concentrations exceeded the goal in 3 percent of
base-flow samples at Station 5 only in the pre-BMP
period and did not exceed the goal at any of the sta-
tions during the post-BMP period.
Paired-Subbasin Comparison
Large variations in water quality due to seasonal,
climatic, and other non-BMP factors, make it difficult
to conclude that changes detected in water quality
were a result of nutrient management. Many of the
non-BMP factors influencing water quality can be
corrected for using paired-subbasins with nearly
identical characteristics.
A paired-subbasin comparison of nitrate plus
nitrite concentrations in base flow in the nutrient-
management subbasin and the nonnutrient-manage-
ment subbasin was made by use of double-mass
plots and an analysis of covariance. A double-mass
plot will show a regression line of constant slope, if
the relation between fluxes (discharge times and
concentration) from each subbasin remain trend
free during the period of interest. Streamflow meas-
urements indicated an unexpected increase in base-
flow discharge from the nutrient-management
subbasin relative to the nonnutrient-management
subbasin about 12 months into the post-BMP period.
If nitrate plus nitrite concentrations were trend free,
the discharge increase would increase nitrate plus
nitrite flux from the nutrient-management subbasin.
However, a double-mass plot (Fig. 10) shows a small
decrease in nitrate plus nitrite flux from the nutrient-
management subbasin relative to the nonnutrient-
management subbasin, thus suggesting a decrease
in nitrate plus nitrite concentrations.
Analysis of covariance detected significant pre-
to post-BMP changes in the relation between nitrate
plus nitrite concentrations in base flow from the sub-
basins. Although not statistically significant by them-
selves (Table 5), small changes in median nitrate
plus nitrite concentrations in each subbasin com-
bined to result in a significant change in the pre- to
post-BMP relation between concentrations of nitrate
plus nitrite in concurrent base-flow samples from the
subbasins (Fig. 11). A decrease in nitrate plus nitrite
concentrations in the nutrient-management subbasin
accounted for most of the change.
203
-------
Proceedings of National RCWP Symposium, 1992
20-
O
<
QC
11) CC
O UJ
•z. i-
O 13
o a;
CM L1J
o °-
Z CO
co 2
^ <
_J CC
0. O
00 _J
O d
o
CO
CO
o
15-
10--
5- -
1984 1985 1986 1987 1988 1989
CALENDAR YEAR
Figure 9.—Time series of base-flow dissolved nitrate plus nitrite concentrations at Station 5, April 1984 through Oc-
tober 1989.
Table 5—Results of step trend analysis tests between pre-BMP (April 1984 through March 1986) and post-BMP
(April 1987 through September 1989) on observed and flow-adjusted base-flow nitrogen concentrations in the
small watershed detected by use of the seasonal rank-sum test.
OBSERVED DATA
STATION PARAMETER STEP TREND (mg/L)
1
Z
3
4
5
Dissolved nitrate plus nitrite
Dissolved ammonium
Total ammonia plus organic nitrogen
Dissolved nitrate plus nitrite
Dissolved ammonium
Total ammonia plus organic nitrogen
Dissolved nitrate plus nitrite
Dissolved ammonium
Total ammonia plus organic nitrogen
Dissolved nitrate plus nitrite
Dissolved ammonium
Total ammonia pjus organic nitrogen
Dissolved nitrate plus nitrite
Dissolved ammonium
Total ammonia plus organic nitrogen
-.10,
-.10
-.08 '
.00
-.10
-.15
-.78
-.03
-.19
.40
-.03
-.22
.80
-.04
-.28
PROBABILITY (P)
0.21
<.01*
.13
.90
.64
. .19
.20
.09
.30
.36
.01*
.15
.31
, .17
.01*
FLOW-ADJUSTED DATA ,
STEP TREND (mg/L)
-.22
-.25
-.50
.37
.69^
PROBABILITY (P)
0.21
.41
.20 r
.30,
.06
mg/L- milligram per liter; < = less than; — * no flow dependency
•Significant at the 95 percent confidence ievel.
204
-------
EH.KOERKLE
3.500
:* 3.000
. .
H <
t= m
cc m
2.500
to
OL "J^
**• fHf-*
CD
Ul
2.000
s °^-
3
O
E
S
1.000
500
A PRE_BMP_
• POST BMP
0 500 1.000 1.500 2.000 2.500 3,000
CUMULATIVE DISSOLVED NITRATE FLUX.
IN POUNDS OF NITROGEN.
IN THE NUTRIENT-MANAGEMENT SUBBASIN
Figure 10.—Comparison of cumulative base-flow dissolved nitrate flux for the nutrient-management and nonnutrlent-
management subbaslns during the pre-BMP (April 1984 through March 1986) and post-BMP (April 1986 through Sep-
tember 1989) study periods. ,
Conclusions
Median concentrations of dissolved nitrate plus
nitrite in base-flow discharge from the watershed
and from two subbasins within the watershed
showed no statistically significant change 3.5 years
after nutrient management was implemented on 50,
90, and less than 10 percent of the croplands in the
watershed, the nutrient-management subbasin, and
the nonnutrient-management subbasin, respectively.
During the period of nutrient management, dis-
solved nitrate plus nitrite concentrations that ex-
ceeded 10 mg/L as nitrogen occurred more
frequently in base flow from the entire watershed
and the nonnutrient-management subbasin and at
the same frequency in base flow from the nutrient-
management subbasin.
Although these results indicate that implementa-
tion of nutrient management had no significant ef-
fect on median nitrate plus nitrite concentrations,
they should be considered specific to this study for
the following reasons:
1. No correlation was made between the percent
of cropland covered by nutrient-management
plans and the percent of reduction in nutrient
applications. Some plans recommended sub-
stantial reductions in nutrient applications
and others recomrnended none.
2. Real reductions in nitrogen applications were
difficult to determine. Estimates of grazing
deposition and large differences between es-
timates of nutrient production and records of
nutrient application and export in the nutri-
205
-------
Proceedings of national RCWP Symposium, 1992
12
10
rr to
5 <
oe CD
j— CD
ijj tn
8|i
LU—' UJ
H-QS O
""' • <
^ 2
< I
S2
o
O APRIL_1984JHROUGH.MAR_CHJ986_ .
• APRIL 1986 THROUGH SEPTEMBER 1989
• •
y=.633x-.860X
O ^
0 •>
-0 2 4 6 8 10 12
DISSOLVED NITRATE PLUS NITRITE CONCENTRATION,
IN MILLIGRAMS PER LITER,
IN THE NUTRIENT-MANAGEMENT SUBBASIN
Figure 11.—Relation between base-flow nitrate plus nitrite nitrogen concentrations from the nutrient-management and
nonnutrlant-management subbasins.
ent-management subbasin limited the accur-
acy of calculated nitrogen reductions.
3. Large natural variations in nitrate plus nitrite
concentrations could have masked smaller
changes in concentrations expected from the
nitrogen application 'reductions recommend-
ed under nutrient management.
4. Carbonate geology probably limited the
achievable reductions in average nitrate plus
nitrite concentrations in base flow under the
recommendations of nutrient management.
References
Agricultural Stabilization and Conservation Service. 1984. Cones-
toga Headwaters Rural Clean Water Program 1984 Prog. Rep.
U.S. Dep. Agric., Harrisburg, PA.
Chichester, D.C. 1988. Evaluation of Agricultural Best Manage-
ment Practices in the Conestoga River Headwaters, Pennsyl-
vania: Methods of Data Collections and Analysis and
Description of Study Areas. Water Resour. Investig. Rep. 88-
96: U.S. Geo. Surv., Denver, CO.
Cleveland, W.S. 1979. Robust locally-weighted regression and
smoothing scatter plots. J. Am. Stat Ass. 74:829-36.
Fishel, D.K. and P.L. lietman. 1986. Occurrence of Nitrate and
Herbicides in Groundwater in the Upper Conestoga River
206
-------
E.H. KOERKLE
Basin, Pennsylvania. Water Resour. Investig. Rep. 854202.
U.S. Geo. Surv., Denver, CO. ,
Pettyjohn, WA and R. Henning. 1979. Preliminary Estimate of
Groundwater Recharge Rates, Related Streamflow, and
Water Quality in Ohio. Rep. 522. Dep. Geo. Mineral, Ohio
State Univ., Boise.
Porterfield, G. 1972. Computation of fluvial-sediment discharge.
Ch. 3 in Techniques of Water Resources Investigation, Book
3. U.S. Geo. Surv., Denver, CO.
Thomas, G.W., G.R Haszler, and J.D. Crutchfield. 1992. Nitrate-
nitrogen and phosphate-phosphorus in seven Kentucky
streams draining small agricultural watersheds: eighteen
years later. Environ. Qual. 21:147-50.
U.S. Environmental Protection Agency. 1986. Quality criteria for
water, 1986. USEPA-400/5-86-001. Off. Water, Washington,
DC.
207
-------
-------
Cedar Revetment and
Streambank Stabilization
Gayle Siefken
Soil Conservation Service
Long Pine Creek Rural Clean Water Program
Ainsworth, Nebraska
ABSTRACT
Cedar revetments were one of the most innovative and successful practices implemented in the
Long Pine Creek Rural Clean Water Program project. The Long Pine Creek watershed had become
degraded by agricultural runoff that had seriously eroded streambanks. A method was needed to
stabilize and revegetate these areas. Over 19,000 feet of cedar revetments were installed
throughout the Long Pine Creek Watershed under best management practice No. 10—Stream
Protection System. In addition to streambank stabilization, these revetments provided a variety of
benefits to trout and other aquatic life. Before the Rural Clean Water Program, little was known
about revetments in Nebraska. Although the Nebraska Game and Parks Commission had experi-
mented with tree revetments in the watershed area, these structures had not been widely
employed. The Rural Clean Water Program provided 75 percent in cost-share funding for best
management practices implemented through the Long Pine Creek project-
In 1981, the Long Pine Creek Watershed was
selected for the experimental Rural Clean Water
Program (RCWP), established as a 10 to 15-year
experiment in controlling nonpoint source pollution.
The Long Pine Creek project offered cost-sharing
and technical assistance as incentives for voluntary
implementation of best management practices
(BMPs).
Fifteen BMPs were selected to address nonpoint
source problems on 60,242 acres determined as criti-
cal and targeted for treatment in the Long Pine
Creek watershed area. Each practice consisted of
several individual practices that could be imple-
mented to achieve the total BMP..All components
had to meet Soil Conservation Service (SCS) techni-
cal specifications.
The Long Pine Creek RCWP project addressed
both land and stream treatment. Land treatment
practices widely implemented in the Long Pine
Creek RCWP included
• irrigation management structures,
• grazing land protection systems,
• establishing permanent vegetative cover on
critical areas,
• water and erosion control structures, and
• fertilizer and pesticide management.
Stream protection systems implemented during
the project included
• channel vegetation,
• fencing livestock out of the stream, and
• establishing revetments to stabilize the
streambank.
The revetments proved to be one of the most suc-
cessful and innovative practices employed in the
Long Pine Creek project.
209
-------
Proceedings of National RCWP Symposium, 1992
Description
The Long Pine Creek watershed is located in north
central Nebraska. The watershed borders the north-
eastern edge of the Nebraska sandhills, the largest
grass-covered sand dune area in the world. The
sandhills rest upon the Ogallala Aquifer, a 200-mile-
wide corridor of underground water that extends
south through Kansas and Oklahoma into Texas.
Surface elevations throughout the watershed
range from 2,700 to 1,930 feet in the stream channel.
Unique geologic and topographic factors have com-
bined to create a fragile environment. The sandhills
rise to 400 feet and then drop below the groundwater
level, creating many wetland .areas. This unique en-
vironment is home to plants and animals found in
both semiarid and marsh habitats (Conserv. Surv.
Div., Inst. Agric. Nat. Resour. 1989). The stream
channel in the Long Pine Creek watershed was
formed in eolian sands and residuum weathered
from sandstone. These materials are especially sus-
ceptible to streambank erosion when vegetative
cover is destroyed.
Originally settled in the mid-1800s, the area sup-
ported ranching and a small amount of farming. The
native grass cover of sand, big and little bluestem,
switchgrass, Indian grass, and prairie sandreed
grass provided excellent grazing for livestock. In the
mid-1960s, irrigation was developed on 35,000 acres
within the watershed. This sudden change from
grasslands to intensive row crop production had a
significant environmental impact. The natural or
geologic erosion and subsequent sedimentation
process, which was accelerated by increased agricul-
tural activity, field runoff, and severely eroded,
streambanks, had a negative effect on aquatic life
along the creek and its tributaries.
Discussion
A significant amount of pre-BMP implementation
data reflecting deteriorating conditions in the Long
Pine Creek watershed had been collected over the
years. The U.S. Geological Survey (USGS) had been
monitoring surface water quality and flow on Long
Pine Creek since 1948, and the Nebraska Game and
Parks Commission had conducted numerous fish
population studies determining abundance and dis-
tribution of fish species throughout the watershed.
In 1979, the Nebraska Department of Environmental
Control began an extensive surface water quality
monitoring project in the watershed that continued
through 1985. This study (Maret, 1985) was the
primary source of pre-BMP implementation baseline
data used for project evaluation.
Pre-BMP implementation monitoring included
sampling of ambient stream conditions, in-stream
runoff conditions, and rainfall. The stream condi-
tions monitored included physicochemical condi-
tions of the water column, biological conditions (i.e.,
bacteria, periphyton, macroinvertebrates, fish),
aquatic habitat (i.e., flow, substrate, cover), and
riparian conditions along streams.
Maret's report found extremely high sediment.
loads at several monitoring stations, especially in
Subbasins 6,8,10, and 11 (see Fig. 1). The excessive.
sediment was destroying food-producing organisms,
filling in pool habitats required for good trout
production, and creating a wider, shallower, and?
warmer stream. Fisheries at some sites exhibited
low diversity and low populations. Monitoring deter-
mined that aquatic life was being impaired by excess
sedimentation from streambank and gully erosion.
Streambank erosion was occurring mainly along
bends in the creek where the current's velocity was
the greatest A BMP practice had to be applied to ad-
dress the eroding streambanks and improve and;
preserve the coldwater stream.
Although the Nebraska Game and Parks Com-;
mission had experimented with tree revetments,
they were not widely employed in Nebraska. BMP-
10 (Stream Protection System), selected for use in'
the Long Pine Creek RCWP, included component
580, Streambank Protection, the installation of
vegetative filter strips, protective fencing, livestock.
crossings, and mechanical (structural) practices to
protect streams from sediment or chemical pol-
lutants and thereby improve water quality. BMP 10;
was applicable to streams or lakes located on or ad-
jacent to farmland. All practices' under BMP 10:
received 75 percent cost-share on the total cost of im-
plementation and had to be maintained for 10 years
following the installation. , ;
Revetment Method
The revetment method of streambank stabilization
uses trees to trap sediment and promote revegeta:
tion along the streambank. As stabilization techni-
ques, revetments cost less than structures. SCS
technicians believed that tree revetments could be
adapted for effective streambank stabilization
throughout the fragile watershed environment,
.Planning for revetment installation requires
several preliminary steps, which include
• obtaining a permit from the Army Corps of
Engineers before beginning construction,
• selecting the sites for revetment installation,
216
-------
G. SEFKBY
\ -fH • : •
'-• •
•fj /: • i • •• ^-?*!nr?^KJ^^
\ ^-:-
Si : : / i^ j i
J^&S^^-i-
\~S.ff\ . :..!.;.
LEGEND
Watershed Boundary
- RCWP Project Area Boundary
Subbasin Boundary
Surface. Water Drainage Boundary
Nebraska Department of Environmental Control
1992
Figure 1.—Long Pine Creek Watershed: Project area, subbasin, and critical area boundaries. Map from Nebraska
RCWP 10-year report (Hermsmeyer, 1991).
211
-------
Proceedings of National RCWP Symposium, 1992
• determining the type of tree and method for
anchoring trees to the streambank, and
• selecting the type of vegetative cover to be
used for revegetation (when the revetment
traps sediment).
• Permits. The Long Pine Creek RCWP project
needed a permit from the Army Corps of Engineers
before it could begin construction of revetments in
the watershed. A general permit applicable for all
revetment installations in the watershed area was ob-
tained through the Nebraska Game and Parks Com-
mission.
• Site Selection. Selection of sites for revetment
construction was based upon the degree of degrada-
tion of the streambanks, the availability of trees for
revetment construction, and the degree of channel
deflection following installation. The more seriously
degraded streambanks were given priority. To
prevent further erosion, revetment construction had
to begin and end where the bank was not eroding.
The revetment could not deflect the channel to such
an extent that it caused additional streambank
erosion downstream.
• Tree Selection and Method of Anchoring.
Next, trees were selected that had many branches to
better trap sediment. The densely branched eastern
red cedar trees that lined the creeks and streams
throughout the watershed were ideal for revetments.
Live trees approximately 20 to 25cfeet tall and 8 to 10
feet in diameter were used. (Dead trees break apart
before a streambank can be stabilized and should
not be used in revetment construction.) If beavers
are active along streams, trees should be cut before
installation and dried out for several months or even
a year. The first revetments constructed in the Long
Pine watershed suffered damage from beavers be-
cause the trees were green; immediately after con-
struction, the beavers cleaned off branches and
removed the trees. After that, trees were felled in the
fall for use the following spring.
Materials for construction (see Appendix: Con-
struction Notes) and methods of anchoring trees to
the streambank depended on soil types and location
of trees to the streambank. Soils can be sand, clay, or
bedrock and each requires different methods and
materials. The trees must be well anchored. In the
Long Pine Creek watershed, cedar trees grew close
to the stream; therefore, a tractor was not needed to
move them. The streambed consisted of mainly
sandy soil, making tree anchoring fairly easy.
Revetments were constructed with Number 9
wire, 5/16-inch cable, and steel fence posts (see con-
struction notes). Steel posts were driven horizontal-
ly into the streambank about every 10 feet. A dead-
man post was set on the upstream end of the revet-
ment and a 5/16-inch cable was attached to it at the
water line and wired to the fence posts. The trees
were placed horizontally in the stream in the direc-
tion of the flow and wired to the cable. The revet-
ment deflected water away from the eroding
streambank and began trapping sediment. Within a
few days, the revetment was filled.
• Vegetative Cover. After sediment has been
trapped, the site must be revegetated. Tree seed-
lings are often used at revetment sites for this pur-
pose; however, the Long Pine Creek Watershed
needed an immediate cover. The first revetments
were damaged by heavy rains and runoff. In the
search for a way to better stabilize and protect the
newly constructed revetments, reed canary grass, a
deep-rooted perennial plant that spreads rapidly, was
broadcast and sodded on the tops of trapped sedi-
ment. This solution proved to be extremely success-
ful. The canary grass, which spread rapidly across
revetments and throughout the stream, provided
habitat for aquatic life and occasional forage for live-
stock.
Within a year, the reed canary grass reached
maturity and produced seeds, which dropped into
the channel, were carried downstream, and were
eventually deposited along stretches of bare banks.
Thus reed canary grass became established on
many untreated eroding banks and spread through-
out the area.
Within two to three years, reed canary grass had
trapped additional sediment, narrowing the stream
channel. The velocity of the stream increased,
removed the sediment that had been deposited, and
eventually flushed out the channel and exposed
gravel bottom in the,streambed. Within four to five
years, the additional vegetative growth along the
revetment narrowed the stream channel further.
Some revetments installed in the Long Pine Creek
RCWP reduced the stream channel from 15 feet to 3
to 5 feet within five to six years. Although constrict-
ing a stream channel is not always desirable, in the
Long Pine Creek watershed, narrowing the stream
channel helped restore stretches of the stream to its
previous dimensions.
At several points along Long'Pine Creek where
revetments were installed, SCS technicians cross-
sectioned portions of the stream before and after in-
stallation of revetments and measured the depth and
width. Results have indicated a significant decrease
in the stream width and an increase in stream depth.
The narrowing of the channel was also documented
in pictures taken over 10 years and through SCS-con-
ducted annual status reviews. Some revetments have
212
-------
G.SJEFKEN
restored the stream to such .an extent that the
original revetment sites are difficult if not impossible
to find.
Conclusions
Revetments increased the velocity of the channel
and created a deeper, narrower, and cooler stream.
The faster and deeper creek flushed out sand and
sediment and exposed gravel in the streambed, a
condition necessary for trout spawning. Revetments
gave trout a place to hide, feed, and rest. The deeper
and cooler stream was conducive to trout species
and other aquatic life.
Revetments proved to be extremely beneficial to
fish populations, including trout The Nebraska
Game and Parks Commission and Soil Conservation
Service used the Habitat Quality Index Procedures
Manual (Binns, 1982) Model II methodology to
document the amount of change in carrying capacity
after revetment installation. According to this model-
ing system, cedar revetments have increased the
mean carrying capacity of Long Pine Creek from
75.4 pounds/acre to 119.2 pounds/acre, a positive in-
crease of 58.1 percent for the sites analyzed.
Cedar revetments were one of the most innova-
tive and successful practices implemented in the
Long Pine Creek RCWP. To date, 19,287.2 feet of
revetments have been constructed throughout five
subbasins in the watershed through the program
(see Table 1). Forty-eight individual revetments
were installed (7.5 percent of the total practices im-
plemented in the program) on 15 participating con-
tracts (17.6 percent of the total participating
contracts). Installation costs were $55,163. Cost-
share provided through RCWP totaled $41,384 (4.8
percent of the total cost-share expended in the pro-
gram).
Cedar revetments would not have been installed
along Long Pine Creek without funding from the
Rural Clean Water Program. Revetments are not an
approved agricultural conservation practice and
therefore are not eligible for cost-share assistance in
any other annual USDA program. In addition to
streambank stabilization, " the revetments have
provided a variety of benefits to trout and other
aquatic life and have dramatically improved the
health and vitality of the stream and its tributaries.
References
Binns, N.A. 1982. Habitat Qualify Index Procedures Manual.
Wyoming Game Fish Dep., Cheyenne.
Conservation and Survey Division, Institute of Agricultural and
Natural Resources. 1989: An Atlas of the Sand Hills. Univ. of
Nebraska-Lincoln, NE. •
Hermsmeyer, E. 1991. Nebraska Long Pine Creek Rural Clean
Water Program Ten Year Report 1981-1991. Agric, Stabil.
Conserv. Serv., Ainsworth.
Maret, T. 1985. Water quality in the Long Pine Rural Clean Water
Project 1979-1985. Nebraska Dep. Environ. Control, Lincoln.
Table 1.—Nebraska's Long Pine Creek Rural Clean Water Program. Best management practice 10, practice code
580—streambank stabilization—cedar revetment.
YEAR SUBBASIN
1989 6
Subtotal Subbasln 6:
1986 7
Subtotal Subbasin 7:
1985 ' * 8
1987 8
Subtotal Subbasin 8:
1984 10
1985 10
1986 10
1989 10
Subtotal Subbasin 10:
1984 11
1985 11
1986 11
1989 11
Subtotal Subbasin 11:
TOTALS
INSTALLATION COSTS
997,
997 ,
1,350
1,350
10,188
1,261
11,449
1,125
1,593
8,328
,477
11,523
18,251
4,156
5,260
2,177
29,844
55,163
COST-SHARE ($)
748
748
1,013
1,013
7,543
946
8,589
844
1,195
6,248
358
8,645
13,692
3,118
3,946
1,633
22,389
41,384
ACRES SERVED
" - •>"1-'
1
1
1
7
2
9
2
1
7
1
11
21
5
1
2
29
51
UNITS APPLIED
180.0ft.
180.0 ft.
570.0 ft.
570.0 ft
5,436.0 ft.
904.0 ft.
6,340.0 ft.
339.0 ft.
750.0 ft.
2,214.4ft.
225.0 ft.
3,528.4 ft.
5,403.0 ft.
1,133.0ft.
1,102.8ft.
1 ,030.0 ft.
8,668.8 ft.
19,287.2ft.
NOTE: 15 participating contracts (17.6% of total participating contracts); 48 individual practices implemented (7.5% of total practices imple-
mented). Taken from Nebraska RCWP 10-year report (Hermsmeyer, 1991).
213
-------
Proceedings of National RCWP Symposium, 1992
Appendix: Construction Notes
Taken from Nebraska RCWP 10-year report (Hermsmeyer, 1991)
FOR CEDAR REVETMENT
SPACE STEEL "T" POSTS AT APPROXIMATELY 10 FT. INTERVALS. PLACE ANCHQR DEAD-MAN
AT A NON TURBULENT SPOT, SO AS NOT TO "CUT" BEHIND ANCHOR. IF NOT POSSIBLE,
LOCATE ANCHOR APPROXIMATELY. 10 FT. INTO BANK, AWAY FROM CHANNEL. AT ,45 DEGREE
TO 90 DEGREE ANGLE WITH RESPECT TO SHORELINE. DRIVE POSTS INTO GROUND,
EXPOSING 1/2 TO 1 FT. OF "T" POST, NORMALLY IN A STABLE BANK. IN VERTICAL OR
VERY STEEP BANKS DRIVE POST INTO STREAMBANK AT 30 DEGREE TO 45 DEGREES.
STRETCH 3/8 OR 5/18 INCH STEEL (AIRCRAFT TYPE) CABLE FROM ANCHOR DEAD-MAN
(USUALLY CONSISTING OF 2 "T" POSTS CROSSED) TO LAST POST ON REVETMENT, USING A
LOOP OR MULTIPLE LOOPS AND CLAMPS TO SECURE CABLE TO ANCHOR AND LAST POST.
HOLES SHOULD BE DRILLED THROUGH POSTS AND PASS CABLE THROUGH HOLES IN POSTS
THROUGHOUT REVETMENT LINE OR TIE CABLE TO ALL MIDDLE "T" POSTS WITH NO. 9 SOFT
MIRE.
BEGIN REVETMENT AT DOWNSTREAM END AND WORK UPSTREAM.
TIMES WITH WIRE.
TIE TRUNK OF TREE 2 OR 3
TIE APPROXIMATELY 2 TO 4 TREES WITH TRUNK DIA. OF 4 TO 6 INCHES IN A 10 FT.
SPACE OF WHICH 4 FT., MORE OR LESS, REMAINS ON BANK. TREES MAY BE TRIMMED BACK
FROM BASE 3 TO 4 FT., LEAVING STUB BRANCHES 6 TO 8 INCHES TO FACILITATE TIEING
TO CABLE AND FITTING TO STREAMBANK. THE BUSHIER TREE IS BETTER WITH SEVERE
(RAW STEEP) BANKS, A ROW OF CEDARS SHOULD BE PLACED LONGITUDINAL ALONG PATH OF
CABLE AFTER MAIN REVETMENT HAS BEEN PLACED DIAGONALLY.
UNDER NO CIRCUMSTANCES SHOULD REVETMENT INCLUDE MORE THAN i/3 OF STREAM WIDTH
ACCORDING TO RULES OF U.S. CORPS OF ENGINEERS.
PLANT VEGETATION (USUALLY REED CANARY GRASS OR SPRIGS) AS SOON AS SBDIMENT
DEPOSITS ALLOW.
CONSTRUCT I OH' NOTES'
FOR CEDAR WING-DIKE REVETMENT
WING DIKES SHALL CONSIST OF 2 CEDAR TREES PLACED 1ST DIAGONALLY WITH STREAM
CURRENT, AND 2ND LONGITUDINAL ALONG STREAM BANK.
TREES WILL BE ANCHORED WITH 1 "T" POST AND TIED WITH NO. 3 GA. SOFT WIRE A
MINIMUM OF THREE SEPARATE TIES.
WING DIKES WILL BE APPROXIMATELY 20 FT. APART.
TREES WITH TRUNK DIA. OF 4 TO S INCHES AND LENGTHS AVERAGING 15 FT. SHOULD BE
USED. DRIVE POST-BETWEEN BRANCHES AND SECURE TREES TO POST WITH WIRE, OR THE
TREES MAY BE TRIMMED BACK FROM BASE 3 TO 4 FT. LEAVING STUB BRANCHES 2 TO 4
INCHES TO FACILITATE TIEING TO ANCHOR AND FITTING TO STREAM BANK.
UNDER NO CIRCUMSTANCES SHALL REVETMENT DIKE INCLUDE MORE THAN 1/3 OF STREAM
WIDTH ACCORDING TO U.S. CORPS OF ENGINEERS.
PLANT VEGETATION (USUALLY REEDS CANARY GRASS OR SPRIGS) AS SOON AS SEDIMENT
DEPOSITS ALLOW.
214
-------
G. SJEFKEN
ID FT. C. TO C.
TYPICAL CEDAR RE'VETMEWT
LENGTH
.f-J
NEEDFO TO
. CHA NG E STREAM DHECTION
THIS POINT IN LINE wITH
OPPOSITE SHORE"LIME
{ OR. DOWNSTREAM
SfCTIOM 1^3
DISTANCE NEEDED TO KEEP STJfEAW
FJ20M GETTING BEHIND SFCTIOW 2.
fiHGlE OP TKEfS VARY PffOM 30° to 'JO"
IWITH EBSPECTTO S«OR.fLrUŁ
PLAN VIEW
CF D/1R.
215
-------
-------
Manure Testing and Manure Marketing
Tools for Nutrient Management
Leon Ressler
Penn State Cooperative Extension
Lancaster, Pennsylvania
ABSTRACT
Nutrient management is a key part of any livestock operation in the nineties. As a part of the Upper
Conestoga Rural Clean Water Program, Penn State Cooperative Extension staff prepared individual
nutrient management plans for farmers in the watershed. Successful nutrient management plan-
ning requires the use of a number of tools, including manure testing, which enables farmers to bet-
ter determine nutrient content. Over 500 manure samples were tested for moisture, nitrogen,
phosphorus, and potash (some were also tested for micronutrients) and individual nutrient
management plans were developed. During this process, the Upper Conestoga Rural Clean Water
Program identified numerous farms that had surplus animal manure and other farms that needed
to purchase crop nutrients. To encourage redistribution of animal manures, Penn State Coopera-
tive Extension compiled a buy and sell list that proved effective. This concept was then expanded to
southeastern Pennsylvania counties as a part of the Chesapeake Bay Program.
Southeastern Pennsylvania is noted for in-
tense livestock operations on small farms.
Lancaster County farms are home to 99,000
dairy cows, 161,000 beef cattle and dairy heifers,
335,000 hogs, and 8,170,000 laying hens, and they
produce more than 50 million broilers each year.
However, high land values have forced farmers to
seek ways to increase income per acre; increasing
animal units per acre and purchasing additional feed
have been the solutions chosen by many. Farmers
export animal products, such as milk, eggs, beef,
and pork, but they still have to deal with a surplus of
crop nutrients in the form of animal manure. The
Upper Conestoga Rural Clean Water Program
(RCWP) project was located in the northeastern
corner of Lancaster County to address the increas-
ing environmental concerns about agricultural non-
point source pollution from animal manure.
Purpose
In 1986, as a part of the Upper Conestoga RCWP
project, Penn State Cooperative Extension opened a
Nutrient Management Office and assigned two
agents to develop nutrient management plans for
farmers in the watershed. Their planning process in-
cluded
• soil testing to determine soil fertility levels,
• using soil type, farmer management ability,
and past history to determine yield goals,
• using livestock inventories and manure
analysis to estimate the amount of manure
nutrients produced on the farm annually,
• developing a plan to use manure on priority
fields,
217
-------
Proceedings of National ROMP Symposium, 1992
• reducing fertilizer recommendations by the
amount of available nutrients in the applied
manure and the residual nitrogen from
previous manure applications,
• verifying nitrogen availability for corn with
the nitrogen soil test, and
• determining how to market surplus manure.
The agents visited farmers to collect information
on livestock numbers, cropping plans, and acreages
and to take soil and manure samples. To determine
the total amount of crop nutrients produced on the
farm — an essential step in the planning process —
total manure production had to be calculated and
nutrient content determined. One option for estimat-
ing nutrient content is to use values from the litera-
ture for manure, but experience in the Upper
Conestoga RCWP project demonstrates that manure
testing whenever possible is preferable.
Table 1.—Manure sample data of 513 samples, 1986-91.
Study Methods and Results
The agents took more than 500 manure samples
over five years while developing 365 individual
nutrient management plans. A large variation existed
among tests of single species (Table 1) and the
averages of the tests were consistently higher than
shown in the literature on manure values for species
other than poultry. Poultry samples also varied, but
their averages were fairly close to the literature
value. Manure test values vary because of different
feeding programs, bedding sources, and handling
systems. Samples taken by Extension staff were
tested for moisture, nitrogen, phosphorus, and
potash (Table 1), and some were tested for
micronutrient content (Table 2).
Extension staff also investigated techniques to
collect a representative sample for manure nutrient
testing. The staff developed procedures for sampling
MANURE TYPE*
Dairy (solid manure)
Dairy (liquid manure)
Dairy (heifers)
Boof
Swine (fattening)
Swine (farrowing)
Swine (solid manure)
Poultry (layers)
Poultry (broilers)
Turkeys
NO. OF PERCENT
SAMPLES MOISTURE
120 82.2 Mean
Median
Maximum
Minimum
64 Mean
Median
Maximum
Minimum
60 75.8 Mean
Median
Maximum
Minimum
161 75.1 Mean
Median
Maximum
Minimum
27 , Mean
Median
Maximum
Minimum
40 Mean
Median
Maximum
Minimum
10 74.3' Mean
Median ,
Maximum
Minimum
10 57.7 Mean
Median
Maximum
Minimum
14 31.4 Mean
Median
Maximum
Minimum
7 42.4 Mean
Median
Maximum
Minimum
PHOSPHORUS
NITROGEN (P205)
10.8
10.4
19.60
5.20
3.4
3.53
5.52
1.26
13.0
12.8
23.2
7.2
14.20
13.60
27.00
6.80
6.10
6.18
8.91
3.76
3.0
2.80
5.09
0.70
21.6
22.10
33.60
11.80
37.7
36.1
60.20
20.00
59.8
63.5
87.60
27.80
49.3
51 .00
61.80
34.60
5.8
5.13
23.90
2.57
1.6
1.68
3.05
0.47
7.0
6.87
18.20
4.37
7.4
6.90
17.50
3.25
4.40
4.08
9.78
1.30
1.8
1.31
5.42
0.09
18.4
18.30
24.20
10.40
48.20
41.30
88.90
21.20
54.3
52.4
94.90
31.10
45.5
42.4
60.70
38.90
POTASH
(KjO)
9.1
8.23
33.88
2.09
2.9
3.26
5.87
0.85
14.4
14.25
24.70
5.70
12.90
12.30
23.00
1.33
2.50
2.40
4.20
1.03
1.7
1.59
3.05
0.27
15.2
17.25
25.90
5.53
24.80
24.50
40.90
12.80
34.0
35.75
50.00
11.20
30.0
30.30
39.10
19.90
* Manure values for solid manure are based on Ibs/ton. Liquid manure values are expressed in lbs/1 00 gals.
218
-------
L. RESSLER
Table 2.—Micronutrient content of manure samples taken from Pennsylvania farms, 1986-91.
FARM MANURE % MOISTURE
Poultry
1. .Broiler
2. Broiler
3. Broiler
4. Broiler
5. Turkey
6: Layer
7. Broiler
8. Broiler
Livestock
9. Cow
10. Heifer
11. Cow
12. Steer
13. Cow/calf
14. Steer
15. Steer
Swine (liquid manure)
1 6. Sow
17. Fattening
31.6
69.2
38.2
34.7
42.8
21.8
45.0
28.4
79.5
78.6
83.7
80.2
79.8
73.5
80.3
8.54
8.53
S
NA
4.32
5.81
4.96
6.98
11.30
9.57
8.44
0.94
1.62
0.98
4.75
1.25
6.56
2.56
0.69
0.93
Ca
NA
20.60
23.60
25.70
23.20
154.00
41.80
34.00
5.34
6.15
7.26
4.71
5.22
6.14
3.27
1.26
2.55
Mg
NA •
2.59
8.53
3.66
4.01
11.40
2.42
8.15
2.33
3.02 '
3.99
1.94
1.74
2.65
1.38
0.62
1.08
Na
NA
3.21
2.84
5.74
7.67
8.45
6.60
7.72
2.50
1 .79
1.01
1.43
1.54
2.17
1.73'
0.55
0.18
Fe
NA
12.83*
2.47
1.36
0.91
3.19
1.76
5.58
0.50
0.82
0.30
3.76
2.92
5.22
0.26
0.25
0.25
Al
NA
0.80
2.22
0.57
0.86
1.05
2.04
4.82
0.24
0.76
0.23
0.89
0.94
1.92
0.17
0.10
0.11
Mn
NA
0.13
0.28
,0.17
0.28
. 0.44
0.52
Q.37
0.05
0.10
0.06
0.05
0.15
0.10
0.03"
. 0.02
0.02
Cu
NA
0.03
0.04
0.03
0.13
0.10
0.20
0.05
0.01
0.02
0.01
0.01
0.01
0.02
- 0.01
0.03
0.04
Zn
NA
0.25
0.28
0.47
0.33
0.65
0.57
0.52
0.06
0.09
0.09
0.04
0.04
0.09
0.03
0.06
0.03
Cd
13.0
1.5
ND
ND
ND
10.0
ND
1.5
NA
NA
ND
NA
NA
ND
ND
1.0
0.5
Note: All values for poultry and livestock samples are expressed as Ibs/ton exceptfor cadmium, which is expressed as ppm on a dry matter
basis. All values for swine manure are expressed as lbs/100 gallons except for cadmium, which is listed as ppm on a dry matter basis. S = sul-
fur; Ca = calcium; Mg = magnesium; Na = sodium; Fe = iron; Al = aluminum; Mn = manganese; Cu = copper; Zn = zinc; and Cd = cadmium.
NA = not available; ND = cadmium was not detected as 0.5 ppm;* contains rust particles., . ,
animal waste (seethe appendix). Sampling bedded
pack manure that has been trampled by livestock
can be a challenge. To evaluate the difficulty of ob-
taining representative samples under these condi-
tions, the project assistant took three samples of
steer manure from the same pen on the same farm
monthly while the steers were being raised.
The results of this study are listed in Table 3. As
the feeding program increased in grain concentrate
fed, the manure's nutrient content generally in-
creased. Nutrient content in the last two samples
that were taken during high concentrate feeding to
finish the steers for market also increased. The
samples taken on the same day varied as little as 3
percent for a given nutrient to as much as 15 to 20
percent.
Because obtaining representative samples in a
bedded pack situation is difficult, researchers should
take several samples on one farm and use that
average — or averages developed from a number of
farms in a region with similar feeding programs —
instead of sampling once on one farm.
The final step in nutrient management planning
is determining how to deal with surplus manure after
crop needs are met. While many farmers in the
RCWP project area had a surplus, other farmers
needed to purchase plant food. To encourage
redistribution of manure, Extension developed a buy
and sell list for the area that identified farmers with
surplus manure (suppliers) along with a description
of the manure type and test values; a list of farmers
that needed additional crop nutrients (receivers) was.
also developed. Every farmer who had a nutrient
management plan got this list, which was printed in
an RCWP newsletter.
The buy and sell list stimulated phone calls and
activity; obviously, it was an effective way to
redistribute nutrients. Later, when the RCWP Exten-
sion agent was assigned to a regional Extension posi-
tion in nutrient management, the manure marketing
list was expanded to eight counties in southeastern
Pennsylvania.
Extension staff also developed a manure market-
ing survey that was distributed at farmers' meetings
and through farm mailing lists. Farmers completing
this survey indicated whether they were potential
manure suppliers or receivers and included other in-
formation about delivery, custom-spreading, willing-
ness to supply free manure, and availability of
composted manure. Potential receivers indicated
whether they preferred composted manure, were
willing to pay for the manure, or required custom
manure application.
, Supplier and receiver lists were compiled by
county and township. These lists are updated and
sent out twice a year in March and October. Farmers
on the supplier list receive a copy of the receiver list
and vice versa. Individual farmers then contact each
other to make the arrangements.
As of March 1992, almost three times as many
farmers have signed up to receive manure as to
supply it (290 versus 105), indicating that a market-
ing opportunity exists for those with excess manure.
In this dense livestock area, the fact that fewer
219
-------
Proceedings of national RCWP Symposium, 1992
Table 3. — Bedded pack manure study test values.*
%
MOISTURE NITROGEN
9/23
9/23
9/23
Mean
10/19
10/19
10/19
Mean
11/22
11/22
11/22
Mean
12/19
12/19
12/19
Mean
1/10
1/10
1/10
Mean
2/17
2/17
2/17
Moan
3/15
3/15
3/15
Mean
4/12
4/12
4/12
^1 Ifc
Mean
5/26
5/26
5/26
Mean
7/6
7/6
7/6
Mean
Project
Average
Summary
Moisture
Nitrogen
Phosphorus
Potash
78.6 12.80
78.8 10.40
76.3 14.00
77.9 12.40
79.3 9.80
80.5 7.80
79.4 8.00
79.7 8.53
75.5 10.60
75.8 11.40
76.7 11.60
76.0 11.20
75.1 9.00
78.0 11.20
75.6 11.20
76.2 10.47
77.8 10.8
78.6 11.40
80.3 11.00
78.9 11.06
79.2 13.00
79.9 13.00
80.7 14.60
79.9 13.53
80.4 11.80
80.5 12.40
80.3 13.00
80.4 12.40
77.3 12.60
76.7 13.00
77.3 12.80
77.1 12.80
74.2 17.50
76.4 14.00
75.7 13.40
75.4 14.97
73.8 20.60
72.3 19.80
72.8 20.00
73.0 20.13
77.5 12.75
Maximum
Minimum
Maximum
Minimum
Maximum
Minimum
Maximum
Minimum
PHOS-
PHORUS
(P206)
5.57
5.52
6.74
5.94
4.65
4.47
4.05
4.39
6.95
6.20
5.54
6.23
6.15
4.23
5.25
5.21
5.80
5.97
5.23
5.66
6.78
6.98
6.71
6.82
7.70
7.93
8.10
7.91
9.05
8.10
7.48
8.21
7.92
7.67
7.11
7.57
12.70
13.30
12.70
12.90
7.09
80.7
72.3
20.6
7.80
13.30
4.05
16.1
9.13
POTASH
(KzO)
12.80
12.70
13.50
13.00
11.60
9.75
9.88
10.41
11.50
11.60
11.70
11.60
10.80
11.00
9.13
10.31
11.40
11.20
11.00
11.20
10.85
10.07
9.35
10.09
10.50
9.99
10.10
10.20
11.20
10.90
10.90
11.00
11.50
12.50
11.50
11.83
14.80
15.30
16.10
15.40
11.50
• 25 percent can custom apply manure,
• 13 percent can deliver manure but will not
apply it,
• 33 percent will supply the manure free if the
receiver picks it up, and
• 3 percent have a composted product.
Among the receivers,
• 27 percent are especially interested in
compost,
• 49 percent are willing to pay for the manure,
• 39 percent are only interested if manure is
free, and
• 22 percent are only interested if the supplier
can custom apply the manure.
These figures demonstrate that farmers who
offer composted manure or custom application ser-
vice will increase their potential markets. Other sur-
vey answers indicate that 34 percent of the farmers
have tested their manure, 73 percent have never
calibrated their manure spreader, and 13 percent
regularly market manure to other operations. The
survey will be a continuing part of Penn State
Cooperative Extension's program in Lancaster
County and southeastern Pennsylvania.
Manure is being marketed successfully in
southeastern Pennsylvania. Because of its con-
centrated nutrients, poultry manure is shipped the
farthest and in the greatest quantity. Weaver and
Souder (1990) reported that broiler litter can be
economically shipped 100 miles for fertilizer use or
300 miles for feed supplement use in Virginia. Two
Lancaster County firms are each marketing, annual-
ly, over 20,000 tons of broiler litter, a large portion of
which is sold to the mushroom industry in neighbor-
ing Chester County. Some manure is shipped as far
away as 300 miles and sold without subsidy — sales
that are usually in conjunction with a back haul of
another product, such as bark mulch or mushroom
compost With increasing environmental concern
and problems from over-application of crop
nutrients, marketing of surplus manure for agricul-
tural use will continue to grow.
* Manure values are expressed as Ibs/ton.
farmers signed up to supply manure is not an indica-
tion of limited supply; it more likely reflects the fact
that farmers who have excess manure do not want it
known for fear of repercussions.
A summary of the survey indicates that, among
suppliers,
Conclusion
Nutrient management is a key part of any livestock
operation in the 1990s. Successful nutrient manage-
ment planning requires a number of old and new
tools. Traditional tools, such as soil testing and
developing realistic yield goals, are key; new tools in-
220
-------
L. RESSLER
elude manure testing, which enables farmers to bet-
ter determine the nutrient content of their animal
waste.
Developing a plan to deal with surplus nutrients
on a farm is important to sound nutrient manage-
ment planning. Manure marketing provides one
solution to a sometimes difficult problem.
Reference
Weaver, W. D. Jr. and G. H. Souder. 1990 Feasibility and economics
of transporting poultry waste. Proc. Natl. Poultry Waste
Manage. Symp., Auburn Univ., AL
Appendix: Manure Sampling Procedures
Compiled by Jeffrey H. Stoltzfus, project associate, and Leon Ressler, Extension agent
Several precautions should be observed when taking
manure samples.
1. When sampling bedded pack manure, take
samples from 10 to 20 different areas of the pen.
Do not take samples near feed troughs or water
bowls because the manure in these areas may be
abnormally diluted with water, feed, and hay.
2. Do not take samples from freshly bedded manure
packs; it is difficult to obtain the proper proportion
of bedding in these manure samples.
3. Manure pits containing liquid manure should be
agitated before the sample is taken. The best time
to do this is when the pit is being cleaned out. If the
pit cannot be agitated, a sample from several loads
should be taken and mixed in a bucket. However,
the first and last loads should not be used for sam-
pling because they contain either a very high or
very low percentage of solids.
4. If a liquid manure sample cannot be taken during
spreading time, a representative sample should be
taken from the pit. The best way to accomplish this
is by using a section of 2.5 cm or larger PVC pipe
that is long enough to reach the bottom of the pit
and has a screw-on cap on one end. Put the pipe
in the pit, then screw the cap on the top end and
remove the pipe. The suction created should keep
the manure in the pipe. Take 8 to 10 samples this
way and mix them in a bucket to prepare the final
sample.
Averages and published values provide a general
view of the manure's nutrient value. However, manure
values can vary from farm to farm because of differing
feeding programs, types of bedding, amounts of wash
water or bedding added, and manure handling sys-
tems. Nevertheless, manure tests, when done proper-
ly, are the best ways to determine manure values for
individual farms.
221
-------
-------
Operation and Maintenance of RCWP
BMPs in Idaho to Control
Irrigation-induced Erosion
Ron Blake
U.S. Department of Agriculture
Soil Conservation Service
Twin Falls, Idaho
ABSTRACT
The Rock Creek Rural Clean Water Program (RCWP) project in Twin Falls County, Idaho, con-
trolled nonpoint source pollution from irrigated agricultural land on 185 contracted farms. Best
management Practice-12 (BMP-12) — sediment-retention erosion or water-control structures —
was a mandatory contract item planned to be installed annually on fields with a cultivated row crop.
The component practices of BMP-12 included minibasins, I-slots, excavated and embankment sedi-
ment ponds, water control structures, buried pipe runoff control systems, and vegetative filter
strips. Irrigation system improvements were the incentive to encourage voluntary participation.
RCWP contracts were flexible in allowing landowners to exchange one sediment-retention
practice for another. For example, during the Rock Creek RCWP Project, landowner preference for
one component within BMP-12 often changed to another component. A practice that would serve
several fields was frequently selected over a practice that served just one field, thus making com-
pliance easier while achieving equally effective results.
Of the components used, sediment ponds were the most valuable practices for illustrating how
much topsoil was eroding. Community-type sediment ponds often collected irrigation wastewater
from neighbors' fields and farms. Generally, these community ponds were installed on major water-
shed drainages, constructed on nonfarm land, and averaged 87 percent sediment-removal effi-
ciency.
Rock Creek in Twin Falls County, Idaho, had
long been recognized as one of the most
severely degraded streams in the State be-
cause of point and nonpoint pollution sources. While
point source pollutants had been virtually eliminated
by the end of the 1970s, Rock Creek was still carry-
ing high loads of sediment and agricultural pol-
lutants.
The Rock Creek watershed covers a total of
198,400 acres: 51,900 acres are irrigated pasture and
cropland containing 28,159 designated critical acres;
109,000 acres are rangeland; 25,400 acres are wood-
land; and 12,000 acres are in other uses such as
urban land, roads, and farmsteads. Approximately
350 farm units are located within the watershed. Fig-
ure 1 shows the location of the Rock Creek water-
shed in Idaho and subbasin boundaries. The map is
taken from the final interagency report on the Rock
Creek Project. Basic crops include dry beans, dry
peas, sugar beets, corn, small grains, potatoes, alfal-
fa, and pasture. Because of low annual precipitation,
all crops are irrigated. Irrigation water is diverted
from the Snake River to individual farms through a
network of canals and laterals owned by the Twin
Falls Canal Company. Water is available at a constant
flow from about April 15 through October 15.
Idaho's Rock Creek Rural Clean Water Program
(RCWP) project successfully reduced suspended
sediment and total phosphorus in the drainages flow-
ing into Rock Creek by 75 and 68 percent,
respectively. Good operation and maintenance of the
best management practices (BMPs) — those
measures determined to be the most effective and
practical means of preventing or reducing pollution
223
-------
Proceedings of National RCWP Symposium, 1992
CMi
Figure 1.—Map of the Rock Creek, Idaho, RCWP project area and subbasln boundaries.
from nonpoint sources — are the major reason that
water quality improved. BMP-12, which was com-
posed of sediment-retention, erosion, or water con-
trol structures, was particularly effective. Table 1
shows the sediment-removal efficiencies of several
BMP-12 components. Comments on the BMP were
documented by the Agricultural Research Service,
Kimberly, Idaho. This paper will describe how the
components of BMP-12 (minibasins, I-slots, sedi-
ment ponds, buried pipe runoff practice control sys-
tems, vegetative filter strips, and water control
structures) were applied and evolved through the
years. Several techniques for installing the BMPs
will also be described to indicate the problems and
solutions the Soil Conservation Service (SCS) ex-
perienced with voluntary participants.
Minibasins
Minibasins, or small, excavated ponds that retain
waste irrigation water, were an effective practice to
trap the topsoil sediment coming from a single field.
224
-------
R. BLAKE
Table 1.—Evaluation of BMPs (source: Carter, 1987).
SEDIMENT REMOVAL EFFICIENCY (PERCENT)
BMP
AVERAGE
RANGE
COMMENTS
Sediment basins (field, farm,
subbasin) 87
Minibasins 86*
Buried pipe systems 83
Vegetative filters 50*
75-95 Cleaning costly
0-95 Controlled outlets essential; careful management
required
75-95 High installation cost; potential for increased production
to offset costs; eliminates tailwater ditch; good control
of tailwater
35-70 Simple; proper installation and management needed
*Average of those BMPs that did not fail.
Minibasins worked best on flat field bottoms, but
they also required considerable effort to install and
maintain, because their outlets must be protected if
they are to retain their storage capacity. PVC pipe,
sheets of plastic, and grass sod ditch banks were
used to establish and protect the outlet and storage
capacity of each minibasin. Figure 2 shows a
diagram of minibasins.
Early in the Rock Creek RCWP project, several
implements were used to install minibasins. A
bulldozer was used to push out storage capacity,
leaving a large mound of topsoil. When the ex-
cavated topsoil was not spread out to a field, a weed
problem developed on the mounds. A carryall was
used to excavate the minibasin and then deposit soil
to create an earthen berm, which was necessary be-
tween each minibasin to regulate trap efficiency.
Another landowner scraped down an entire field bot-
tom to the design depth, then used a small bulldozer
to blade up the berm between each minibasin. The
minibasin outlet was then surveyed and cut in with a
hand shovel.
Sometimes the dead furrow caused during the
plow's final pass in a field created a problem in the
minibasin component. Normally, this dead furrow
was used as the field's drainage ditch. When a mini-
basin was installed before the field was plowed, how-
ever, the dead furrow created a drainage ditch in
front of the minibasin, preventing it from working
properly. Extensive hand labor was then needed to
cut irrigated furrows into each minibasin.
I-Slots
I-slots are small, excavated sediment ponds dug into
a field's wastewater drainage ditch. Generally, I-slots
are 3 to 4 ft wide, 20 ft long, and 2 to 4 ft deep. They
are used to drain a single field. (Initially, T-slots were
used, but evolved into,the easier to install I-slots.)
Figure 3 shows, a diagram of I-slots. A back hoe was
normally used to install I-slots. Excavated topsoil was
then scraped up and spread out on the immediate or
adjacent field.
I-slots can be used on flat and steeply sloped
field bottoms. On flatter field bottoms, a back hoe
was used to dig a drain ditch lower than the field sur-
face between each I-slot to keep wastewater out of
the field.
Not all field bottoms, however, were good for I-
slots. In some fields, drainage ditches were deep,
resulting in a lack of storage capacity in I-slots. In
such cases, sheets of plastic were used to help raise
and stabilize the outlet of each I-slot to make it func-
tional. Also, I-slots were sometimes dug wider to
achieve the design storage capacity in shallow soils
and to help prevent spreading of caliche and other
rocky debris to the field.
I-slots proved to be an excellent "show-and-tell"
BMP. When they filled up at the end of the first or
second irrigation, landowners and farm operators
soon realized that better irrigation water manage-
ment was needed.
Sediment Ponds
Excavated sediment ponds were installed during the
Rock Creek RCWP project where the pond could col-
lect wastewater from one or more fields. The closer
the pond was to the field (s) it drained, the sooner it
filled with sediment.
Excavated sediment ponds were designed for a
minimum of one- and a maximum of five-years'
storage capacity. The greater-than-one-year storage
capacity was desirable because it helped reduce the
landowner's maintenance expense. To help the sedi-
ment pond last longer, a vegetative filter strip was
planted above the pond along the entire field bottom,
thus reducing maintenance and related expense.
Quite often, landowners who owned excavation
equipment had used sediment ponds even before the
RCWP to collect the valuable topsoil eroding from
225
-------
Proceedings of National RCWP Symposium, 1992
Flgur
Waste Water Disposal Ditch-.
— Protect Outlets with
/^ Plastic, Pipe or Sod-
/ vv^K^ \
's> iS^ T°P °f Berm
/ NSCVN/r
1 ._ W
Irrigation
Direction
TYPICAL PLAN VIEW
"~3®Ss^S^ '"/"
\. Waste Water and A /«v&
^\. Sediment Storage /
TYPICAL CROSS SECTION
NOTE
Construction shall be in accordance with
Construction Specification ID-350-B.
Mini Basins Required
I have reviewed the plans and sperifiratinns
and agree to construct this project to the U.S. DE
best of my ability in arrordance with them. SOIL
Date B™.
NOT TO SCALE IHlTlTI
e 2.— Diagram of mlnlbaslns.
X
/ X\N>\
_^ ^§J
/ \\\\ /
u
9 f1. Waste
~Min.^1 wflU,
\ uitcn-j.
''--Basin Spillway
Crest
MINI-BASINS
PARTMENT OF AGRICULTURE
CONSERVATION SERVICE
""' ***»,.: _...
T«t .
THI*
. ~ Job Class
226
-------
R. BLAKE
1-
J
ss,
NOTE
Waste Ditch Bottom — ^ Flow — »—
I-slot \ I-slot
Sediment X SorHment
Storage Storage
| 1 J
t
IRRIGATION
DIRECTION
TYPICAL PLAN VIEW
Waste Ditch Bottom - — ,
/>-. •-/ -\ .: v-y ; /^ • • -y
•/' ... : --.--. ^ • ^ s? \v . . . . ^
?=•-. Storage ^^ ^, ^, storage ^:-:-3^
, Ditch Flow — *-
IRRIGATION
DIRECTION
TYPICAL ISOMETRIC VIEW
r^-p
Typical I-slot
Cross Section
Construction shall be in accordance
with Construction Specification ID-350-B. I-SLOTS
I-Ont«; Rpnuirprf
I have revi«
and agree tc
best of my e
jwed the plans and specifications
) construct this project to the
minty in accordance with them. u.s. DEPARTMENT OF AGRICULTURE
SOIL CONSERVATION SERVICE
Date -Ktn
... . B^f_^ HtnMU. ...
,___ ™"
TiM ..
NOT TO SCALE ^ ~.*_.. joT> Class
Figure 3.—Diagram of I-slots.
227
-------
Proceedings of National RCWP Symposium, 1992
their own and their neighbors' fields or farms. Fig-
ure 4 is an Idaho standard engineering drawing used
to design excavated sediment ponds.
Community embankment sediment ponds were
often installed where they could trap irrigation waste
water pollutants from the fields of neighbors who
were not project participants. Usually, the embank-
ment pond site was land that was out of production
and not being farmed. For example, at one location,
old sheep corrals next to Rock Creek were con-
verted to an embankment pond with a three-year
sediment storage capacity for an 1,100-acre water-
shed. All embankment sediment ponds resulted in
an immediate improvement of water quality and
eventually developed as good wildlife habitats. Fig-
ure 5 is an Idaho standard engineering drawing used
for embankment sediment ponds and water control
structures.
Water control structures were needed to es-
tablish storage capacity within excavated and em-
bankment sediment ponds. Care had to be taken to
protect these structures from washing out A trash
rack was needed at the inlet, and rock rip-rap was
needed at the outlet to further protect the structure
from failing. Figure 6 is an Idaho standard engineer-
ing design of a pipe inlet or outlet commonly used
for buried pipelines but sometimes used for sedi-
ment ponds.
Buried Pipe Runoff Control
Systems
The buried pipe runoff control system (BPRCS) con-
sisted of a buried plastic mainline, with inlet risers in-
stalled every 20, 40, or 60 feet, depending oh
steepness of the field bottom. Earthen berms were
installed near each inlet riser to control and direct
the irrigation wastewater flowing to each inlet riser.
Hand labor was needed to adjust the berms and to
survey and cut the inlet risers to keep water from
flowing around the berm into the field. In a few in-
stances, the field was irrigated before the inlet risers
were properly adjusted, which caused serious
erosion.
One land operator was particularly fond of
BPRCS. He operated six farms with RCWP contracts
and wanted the BPRCS installed on each farm. Some
fields he chose did not have a convex end (a progres-
sively increasing slope over the lower 20 to 80 ft of
the field into the drained ditch that develops as a
result of farmers keeping drainage ditches lower
than the furrow ends and cleaned so that water will
run off the field bottom quickly). These BPRCS re-
quired a good back berm. If that berm did not exist,
it had to be created, which required minibasins to be
excavated and the topsoil used to create the back
berm and the individual minibasins. The inlet riser
was then surveyed and cut to an elevation that kept
the wastewater in the minibasin and out of the
cropland. Figure 7 is a hand-drawn sketch of the top
and side view of the buried pipe runoff control sys-
tem invented by Dr. David Carter, a supervisory soil
scientist for the Agricultural Research Service. (No
Idaho standard engineering drawing exists for this
system at this time.)
Vegetative Filter Strips
Vegetative filter strips were used annually in Idaho's
RCWP project and planted to spring grain, winter
grain, oats, or an alfalfa and grass mixture. Non-cost-
share filter strips used permanent pastureland or a
strip of alfalfa left along the entire field bottom when
the field was plowed for the next crop.
Soil moisture was critical to get spring-planted
filter strips to grow tall enough before the first irriga-
tion. The corrugates were pulled a third of the way
into the filter strip to get the wastewater into and
across it. Ponding sometimes occurred in front of
the filter strip causing crop damage on flat field bot-
toms with no convex end.
Permanent filter strips were used and were ef-
fective for two to three years. In the third year, sedi-
ment deposition sometimes created the need to dig
berms to force water info and across the filter strip.
The field bottom then had to be reshaped and
replanted following harvest. Additionally, the plow's
dead furrow presented a problem when an alfalfa
strip was left as the filter strip. The dead furrow often
became the wastewater ditch in front of the filter
strip, rendering it ineffective.
Ideally, filter strips should be managed as
grassed waterways. In other words, the field bottom
should be graded down to the depth of the plow's
dead furrow and planted in the fall. A winter-hardy
plant species should then be planted in the graded
area. When the farmer plows the field in the spring,
the plow's depth must blend into the depth of the
graded field bottom. Adequate vegetation will be
present in the spring time to help clean irrigation
wastewater of sediment and fertilizer nutrients.
When sediment bars build up, the filter strip or
grassed waterway should be reshaped and reseeded.
Figure 8 is an Idaho standard engineering drawing
for a grassed waterway. (No Idaho standard en-
gineering drawing for vegetative filter strips exists at
this time.) The author chose to use the grassed
waterway drawing because it depicts the way annual
vegetative filter strips should be managed in the fu-
ture.
228
-------
R. BLAKE
Location Map
*As Shown Above
outlet Pipe
TBM =
*In1et Pipe
Cross Section
Cross Section
Sta.
;ide Slooes I
T
Freeboard =
Notes:
1 . Construction shall be in accordance
. with SCS standards and specifications
2. It is' the landowner's or operator's
responsibility to locate and protect
all public utilities (underground and
overhead) within the work area.
I have reviewed the plans and specifi-
cations and agree to construct this
project to the best of my ability in
accordance with them.
Volume of Storage
Volume, of Excavation
Not To Scale
Date
Excavated Sediment Pit
U. S. DEPARTMENT OF AGRICULTURE
SOIL CONSERVATION SERVICE
... Job Class
Figure 4.—Idaho standard engineering diagram used to design excavated sediment ponds.
229
-------
Proceedings of National RCWP Symposium, 1992
1
I
«#
o
SS
CD #
H
to
UJ CM
W e i
UJa
Q.
Q.
Q
O
Q.
CC.
Ł
i 5 a
•H .d T)
O 4J
O >>
a, w w
W O -H
? "H
-a M T)
co CO • cd
Ł3
ra o
a a
my
Da
I-H o IH
ft 4J O
a)
-------
R. BLAKE
E
1
o.
as
•5
•o
O)
Q.
O
•o
2
•o
o>
i
o
I
(O
"I
O)
231
-------
Proceedings of National RCWP Symposium, 1992
TOP VIEW OF FIELD
major
drainage
ditch
field's
irrigated
furrows
mini-basin
s tor age
a r e a
FIEI.D DRAINAGE DIRECTION
\
mini-basin
berms
mainline
inlet risers
TO
ROCK CREEK
SIDE VIEW OF BPRCS
.irrigated
mini-basin
berms
">/( IJ. s til. o / 7 ~V _^_
' X I / / ^^~~~-\
PIPELINE DRAINAGE DIRECTION storage area
*Protect outlet and inlets from rodents!
BURIED PIPE RUNOFF
CONTROL. SYSTEM (BPRCS)
U. S. DEPARTMENT OF AGRICULTURE
SOIL CONSERVATION SERVICE
TB*, ______
Figure 7.—Hand-drawn sketch of top and side view of a buried pipe runoff control system.
232
-------
R.BLAKE
DESIGN CROSS SECTION
TRAPEZOIDAL
-Seeded Width =
k- •—H'
DESIGN CROSS SECTION
PARABOLIC
Seeded Width =.
I have reviewed the plans and specifications
and agree to construct this project to the
best of my ability in accordance with them.
Date:
Lopperator
CONSTRUCTION SPECIFICATIONS
ID-412 AND ID-14000-274 ATTACHED
GRASSED WATERWAY
COOPERATOR
SCO
CO.,IDAHO
U. & DEPARTMENT OF AGRICULTURE
SOIL CONSERVATION SERVICE
Figure 8.—Idaho standard engineering diagram for a grassed waterway.
233
-------
Proceedings of National RCWP Symposium, 1992
Conclusions
Through the process of installing and evaluating
BMP-12 component practices, the following lessons
have been learned and shared:
• Minibasins work best on flat field bottoms.
Scraping the entire field bottom down to the
design depth and using a small bulldozer to
push up berms is the best way to install this
practice. Be sure to protect the outlet of each
minibasin.
• I-slots can be used on steeper field bottoms by
reducing the length of the I-slots so that all ex-
cavated storage capacity will be used. The I-
slots are excavated in the field's drainage
ditch. A shallow drainage ditch between each
I-slot is desirable to help increase the I-slot's
storage capacity.
• Excavated sediment ponds located in the
corner of a field may eliminate the need for
small ponds along the entire field bottom (i.e.,
minibasins or I-slots).
• A vegetative filter strip can be used along the
entire field bottom above the sediment pond
to keep the pond from filling up with sediment
so quickly.
• Embankment sediment ponds normally drain
more than one field or farm. The best location
to install these larger ponds is in major
drainages and on land not being farmed.
• Structures for water control are a necessary
component of all embankment sediment
ponds. Excavated sediment ponds may need a
structure for water control in the form of a
piped outlet depending on the pond site.
• Buried pipe runoff control systems work best
on steep field bottoms with a severe convex
end. To maintain pipeline gradient, the cor-
rugated plastic pipe mainline needs to be care-
fully hand bedded before being covered.
Protection is needed at the inlet and the outlet
so the system does not get plugged by trash
and rodents.
• Vegetative filter strips became the more
popular BMP when a field needed sediment
retention. Filter strips used on flat field bot-
toms created crop damage by. water scalding.
To solve this problem, field bottoms should be
scraped down to the plow's depth (6 to 8 in-
ches deep) and planted using a winter-grow-
ing plant species. When the farmer plows the
field in the spring, the plow's dead furrow (6
to 8 inches deep) won't adversely affect the fil-
ter strip's effectiveness by creating a was-
tewater ditch in front of the filter strip.
• The Soil Conservation Service has kept the
standards and specifications for installing the
BMP-12 components flexible enough to allow
for any needed modifications in the ex-
perimental RCWP project. .
• Cropland fields with 0 to 1 percent field slopes
have a tolerable soil loss and may not be criti-
cal acres in need of.treatment.
• Cropland fields adjacent to wastewater
drainages in which no opportunity exists for
reusing the wastewater are the most critical ir-
rigated acres to be treated.
Reference
Carter, D. 1987. RCWP Annual Report. Agric. Res. Serv., U.S. Dep.
Agric., Kimberly, ID.
234
-------
Coordination is the Project Cornerstone
Michael Kuck
U.S. Department of Agriculture, Soil Conservation Service
Huron, South Dakota
Jeanne Goodman
South Dakota Department of Environmental and Natural Resources
Pierre, South Dakota ..,...-
ABSTRACT
The success of the Oakwood Lakes-Poinsett Rural Clean Water Program project resulted from the '
coordination that existed between all the agencies involved. At the county level, excellent coopera-i
tion and coordination existed among the Agricultural Stabilization and Conservation Service
(ASCS), the Soil Conservation Service (SCS), county committees, and conservation districts,
which provided an excellent forum for local input and control. This cooperation carried through to
the State level, where ASCS, SCS, Cooperative Extension, and South Dakota Department of En-
vironmental and Natural Resources (DENR) worked closely to ensure that the project would move
ahead smoothly with little interruption. DENR served as the main agency in the comprehensive
monitoring and evaluation portion of the project by contracting with ASCS and the South Dakota
State University Water Resources Institute to implement the project's goals and objectives. SCS
and the Cooperative Extension worked with DENR to support any activities that were needed—
from identifying critical areas and planning project tours to completing annual progress reports. A
team approach, used to accomplish this coordination, gave every agency or group a say in what
was being done and the ability to make changes as needed. This approach also allowed the team to
bring in national-level specialists and experts to assist the project and maintain the national
perspective of the Rural Clean Water Program project.
The goal of the South Dakota Oakwood
Lakes-Poinsett Rural Clean Water Program
(RCWP) project was to reduce the total
nitrogen, pesticides, and animal waste entering the
ground and surface water in the project area by im-
plementing fertilizer and pesticide management and
conservation tillage on 60,000 acres. The com-
prehensive monitoring and evaluation portion of the
project monitored the effects and evaluated the im-
pacts to groundwater and surface water from the im-
plementation of best management practices (BMPs).
The success of the Oakwood Lakes-Poinsett
RCWP project was measured in three areas:
1. Producer participation was high—over 80
percent of the highest priority area was
under RCWP contracts and a majority
maintained BMPs even after the cost-share
period had ended.
2. Cooperation between participating agencies
was outstanding, with all agencies
maintaining an active involvement
throughout the project.
3. Water quality monitoring indicated that the
implementation of BMPs to protect and
improve the surface water quality did not
adversely affect the groundwater.
This paper focuses on the cooperative efforts
and the use of committees, interagency teams, and a
project coordinator to implement successful RCWP
projects.
235
-------
Proceedings of National RCWP Symposium, 1992
Discussion
Committees
At the county level, cooperation and project coor-
dination occurred among county committees, con-
servation districts, the ASCS, and the SCS in all
three counties. Vital local input and control were
possible through a local coordinating committee,
which received support from county, State, and
Federal agencies, lake associations, and local farm
communities. The local coordinating committee was
comprised of individuals from the three county con-
servation districts and county committees with one
member from a county committee serving as chair-
person. Acting as the local project sponsor, local
coordinating committee members used their
knowledge of the project area and information from
the land treatment team to recommend priority
areas, goals, and objectives to the State coordinating
committee for its consideration and action. The local
coordinating committee also served as a mechanism
to distribute information to individuals in the project
area.
The State coordinating committee served as the
project decisionmaking body and was represented.
by the following agencies: ASCS, SCS, Cooperative
Extension Service (CES), Farmers Home Ad-
ministration (FmHA), Statistical Reporting Service
(SRS), Environmental Protection Agency (EPA),
South Dakota Department of Agriculture, Economic
Research Service (ERS), Water Resources Institute
(WRI), and South Dakota Department of Environ-
ment and Natural Resources (DENR). The State
coordinating committee directed most aspects of the
project by acting on recommendations from the local
coordinating committee or the comprehensive
monitoring and evaluation team, depending on the
project objectives. Excluding CES and ERS, all fund-
ing decisions were made by the State coordinating
committee, thus ensuring that adequate funding was
dedicated to meeting project objectives. This ap-
proach allowed the State coordinating committee to
base its decisions on local coordinating committee
recommendations and national coordinating commit-
tee directives; it also encouraged all participating
agencies to share input on how project funds were to
be used. Finally, the State coordinating committee
served as the main contact for the national coor-
dinating committee.
Most agencies represented on these committees
remained active throughout, the 10-year project
period. The State and local coordinating committees
were successful in planning, requesting, and secur-
ing approval from the national coordinating commit-
tee for additional cost-share and monitoring funds to
meet project goals. These committees were success-
ful because everyone serving on the committees
believed in and supported the project from its begin-
ning to end.
Teams
Another concept that worked well in this project was
the use of technical teams. Three teams were
developed to implement planning and contract
development, to monitor the project, and to provide
information and educational activities as needed for
the project. These teams included the land treatment
team, the comprehensive monitoring and evaluation
team, and the report management team. .
• The land treatment team, consisting of Coopera-
tive Extension specialists, three SCS district conser-
vationists, three ASCS county executive directors,
and a conservation district technical planner,: acted
as technical advisors to the local coordinating com-
mittee. The committee guided the team on how to
obtain public acceptance for the project, and in turn,
the team advised the committee on the project's
technical aspects; for example,
• fertilizer management, pesticide manage-
ment, and conservation tillage were identified
as the best BMPs for water quality problems
and the most likely to be accepted for the
area; and
• the critical area and the process for estab-
lishing priority areas was identified ,and
developed.
The land treatment team's main role was to work
directly with producers by supplying information
and education and by developing conservation plans
and water quality contracts for BMP implementation.'
Although all team members worked with producers,
the technical planner provided the one-on-one con-
tact that accelerated producer participation. The
land treatment team also decided which producers
to contact, what BMPs to cost-share, and how to ad-
dress day-to-day problems with contracts and;
operators.
The comprehensive monitoring and evaluation
team's role was defined in contractual agreements
developed by ASCS and DENR (with the concur-.
rence of SCS). This team managed and conducted
the comprehensive monitoring and evaluation
project. DENR, in turn, contracted with several agen-
cies to implement various portions of the project.
This team consisted of five technical people from
DENR and WRI and was directed by the DENR
236
-------
M. KUCK & J. GOODMAN
project manager. It was responsible for monitoring
the cause-and-effect relationships between agricul-
tural BMPs and the water quality in the study area.
Therefore, it monitored the ground and surface
waters and water in the soil profile by using the
strategy and standards established in the contractual
agreements and the project workplan.
An SCS employee coordinated the comprehen-
sive monitoring arid evaluation activities. This coor-
dinator served as a direct link between the State
coordinating committee, DENR, and the monitoring
team and as a passive link between the local coor-
dinating committee and the land treatment team
(see Fig. 1). The coordinator maintained vital com-
munication between the comprehensive monitoring
and evaluation team and the land treatment team,
helped establish schedules for meetings and reports,
and served as an information contact for everyone
involved in the project. Near the project's end, the
coordinator and the comprehensive monitoring and
evaluation team manager served as report coor-
dinators. By consensus of the report management
team, they directed the work and coordinated the
preparation of the final report.
The role of the report management team was to
plan and prepare the 10-year report. The report team
was comprised of the project coordinator, the com-
prehensive monitoring and evaluation team, the in-
formation and education coordinator, the technical
planner, a WRI representative, and several State
coordinating committee members. Using the nation-
al coordinating committee's outline, the team
cooperatively prepared and reviewed each section of
the final report for content and format. As a result,
the final report was concise and correct when com-
pleted.
Lessons Learned
Many lessons were learned from this project, but the
following lessons stand out as strong points in
project coordination:
• The full support of everyone involved—from the
grass roots to the State level—is vital for a project to
succeed. This support was very evident in the Oak-
wood Lakes-Poinsett RCWP project, which had the
full commitment of the local coordinating commit-
tee, State coordinating committee, and the various
groups and agencies they represented.
• Using teams to implement project activities
worked well because each team had an identified
structure, goals, and objectives. Team leaders were
identified who could work with team members to en-
sure that tasks were completed according to the
schedules established in workplans and contractual
agreements. This concept allowed the teams to iden-
tify who would be working with producers on a one-
to-one basis, thus eliminating producer confusion
and helping the project to run much more effectively
and efficiently.
• Using the State water quality agency (DENR) to
conduct comprehensive monitoring and evaluation
also added to the project's efficiency. Again, the team
concept allowed those individuals with monitoring
experience and expertise to control the project's
work in that area and to implement established goals
and objectives.
• The project coordinator (whose role varied some-
what throughout the project) was essential for a suc-
cessful project. For the first five years of the project,
the coordinator was located 200 miles from the
project area, isolating the coordinator from the land
treatment team and most members of the com-
prehensive monitoring and evaluation team. The
coordinator's position was then moved nearer to the
project area, which resulted in greater communica-
tion between the teams and more effective leader-
ship for the report management team. The
coordinator maintained communication between the
comprehensive monitoring and evaluation team
(who were water quality people) and the land treat-
ment team (who were U.S. Department of Agricul-
ture [USDA] personnel). The coordinator also
supplied a direct link between the teams and the
State coordinating committee, where all primary
project decisions were made. This individual, an SCS
employee, was familiar with USDA programs and
workers on the district, area, and State levels, which
made communication among the teams and different
disciplines much easier. The coordinator's location
proved most effective when located in the project
area.
• One weakness in the RCWP project coordination
was that after the sign-up period for BMP contract-
ing had ended, the land treatment team and the local
coordinating committee became less involved in the
project, mainly because its members acquired addi-
tional responsibilities under the 1988 Food Security
Act. More involvement from the local coordinating
committee and the land treatment team throughout
the project would have improved the information ex-
change among teams and producers involved in the
project. This increased involvement would also have
provided a larger group to help develop the land
treatment portions of the 10-year report.
237
-------
Proceedings of National RCWP Symposium, 1992
National Coordinating
Committee
(NCC)
State Coordinating
Committee
(SCC)
Local Coordinating
Committee
(LCC)
Department of Environment
and Natural Resources
(DENR)
Land Treatment
Team
Conser-
vation
District
(CD)
Soil
Conser-
vation
Service
(SCS)
Agri-
cultural
Stabiliza-
tion and
Conser-
vation
Service
(ASCS)
SD
Coopera-
tive
Extension
Service
(CES)
Comprehensive
Monitoring and
Evaluation Team
(CM&E)
DENR
SDSU
Water
Resources
Institute
(WRI)
Figure 1.—Oakwood Lakes-Polnsett RCWP project organizational chart.
Conclusion
Coordination is very important for any project.
Without proper coordination the chances for com-
plete project success are reduced considerably. Ef-
fective project coordination begins before the
project's implementation and continues throughout
the project's life. The coordination of the Oakwood
Lakes-Poinsett RCWP project was exceptional. All
agencies, groups, and individuals worked together
and were highly motivated to make the project suc-
ceed—the primary formula for success. And the
process has continued: the cooperation developed
between agencies during the Oakwood Lakes-Poin-
sett RCWP project has carried over to the Big Sioux
Aquifer Demonstration project and several other
nonpoint source projects now being developed in
South Dakota.
238
-------
Taylor Greek-Nubbin Slough RCWP
Institutional Arrangement and
Program Administration
John W. Stanley
Agricultural Stabilization and Conservation Service
U.S. Department of Agriculture
Okeechobee, Florida
ABSTRACT
Institutional arrangements and program administration were key to many of the successes in the
Taylor Creek-Nubbin Slough Rural Clean Water Program project. The Okeechobee Agricultural
Stabilization and Conservation Service county executive director acted as project administrator
and was assisted on the administrative subcommittee by representatives of three State agencies.
These four people met regularly (and as necessary) to handle the project's day-to-day operations.
Each agency fully understood the other's role; therefore, one agency could assist another with its
problems. This cooperation proved helpful for participants, who could get the whole picture from
any of the four agencies. This arrangement was key to the project's success—89 percent of the
critical area were contracted and 99 percent of the best management practices installed—and
helped to hold the agencies working closest to the project together for 10 years. (This paper is
based on the Taylor Creek-Nubbin Slough 10-year Report) , '..-•
The Taylor Creek-Nubbin Slough Rural Clean
Water Program (RCWP) project is located
primarily in Okeechobee, Florida. Funded in
1981, the only RCWP project in Florida defined
63,109 acres of the Taylor Creek-Nubbin Slough
basin as critical and identified 59 farms that needed
treatment.
The Taylor Creek-Nubbin Slough RCWP project
was coordinated by a project administrator—the
Okeechobee Agricultural Stabilization and Conser-
vation Service county executive director—who
headed an administrative subcommittee. Because
most of project acreage was located in Okeechobee
County, the Okeechobee county executive director
coordinated submission of the project and assumed
the role of administrator responsible for coordinat-
ing all activities, keeping the project on track, and
understanding all project phases and each agency's
roles and responsibilities.
While organizing the project, the local coordinat-
ing committee realized it could not meet often
enough to solve problems as they arose, so it estab-
lished an administrative subcommittee to coordinate
project activities. This subcommittee included repre-
sentatives of agencies that would ,be the most active-
ly involved in the project: the project administrator,
the Soil Conservation Service district conser-
vationist, County Cooperative Extension agent, and a
representative of the local South Florida Water
Management District office.
The administrative subcommittee met monthly
and as needed to keep the project on track. This
project's success resulted from the cooperation and
close working relationship of these four agencies
and their representatives; the group's small size
made it easy to call meetings and keep in contact on
the project's day-to-day operation/Each member un-
derstood the others' roles and therefore could assist
239
-------
Proceedings of National RCWP Symposium, 1992
in solving problems and meeting goals. By working
closely together, the administrative subcommittee
was able to overcome many obstacles early in the
project. The administrative subcommittee and local
coordinating committee met several times a year to
review project progress.
Local Coordinating Committee/State
Coordinating Committee Coordination
The local and State coordinating committees worked
mainly through their respective agencies. Repre-
sentatives from both committees met to review the
annual report and the local coordinating committee
invited State committee members to participate in
field days or tours of the RCWP area. The two com-
mittees gave the administrative subcommittee the
flexibility it needed to run the day-to-day operations
and make recommendations and changes. Any
major changes were considered by the administra-
tive subcommittee, the local coordinating commit-
tee, the State coordinating committee, and finally,
the national coordinating committee.
Tables 1 and 2 list the agencies represented on
the State and local coordinating committees. The
local coordinating committee had an official mem-
bership, but the committee itself was made up of
several members of the different agencies involved.
No formal memoranda of understanding (MOU)
were signed because the local agencies had such
close working relationships. (One MOU was signed
with the South Florida Water Management District
for monitoring, but only because it was mandated
under national procedure.)
Institutional Arrangements
The Agricultural Stabilization and Conservation Ser-
vice handbook 1-RCWP outlined the institutional ar-
rangements and each agency's role. The Taylor
Creek-Nubbin Slough project was not restricted by
the 1-RCWP and expanded these roles as needed.
The four key agencies participated in all areas and
phases of the project, even though they were not
specifically mandated to do so. This coordination and
understanding brought the agencies closer together:
Table 1.—Taylor Creek-Nubbin Slough RCWP project Okeechobee County local coordinating committee mem-
bership. ' _ -
Okeechobee Agricultural Stabilization &
Conservation Service Commitee
609 SW Park Street
Okeechobee, FL 34972
i Okeochobea Sod & Water
Conservation District
419 SE 8th Avenue
Okeochobee, Ft-34974
i USDA-Farmers Home Administration
314 NW 5th Avenue
Okoechobee, FL 34972
i Cooperative Extension Service
501 NW 5th Avenue
Okoschobee, FL 34972
• Game & Fresh Water Fish Commission
3991 SE 27th Court
Okeechobee, FL 34974
• Martin Agricultural Stabilization &
Conservation Service
P.O. Box 385
Palm City, FL 34990
• Dairy Farmers inc.
P.O. Box7854
Orlando, FL 32854
• USDA-Agricultural Research Service
Bldg. 164, University of Florida
Gainesville, FL 32611
• South Rorida Water Management District
1000 NE 40th Avenue
Okeechobee, FL 34972
USDA-Soil Conservation Service
611 SW Park Street
Okeechobee, FL 34972
Florida Division of Forestry
5200 Highway 441 North
Okeechobee, FL 34972
i USDA-Soil Conservation Service
2401 SE Monterey Road
Stuart, FL 34996
i Martin Soil & Water Conservation District
P.O. Box 362
Palm City, FL 34990
i Department of Environmental
Regulations
1900 S. Congress Avenue
West Palm Beach, FL 33406
Table 2.—Taylor Creek-Nubbin Slough RCWP project Florida State coordinating committee membership.
« Florida State Agricultural Stabilization &
Conservation Service
P.O. Box 141030
Gainesville, FL 32614-1030
• Office of the Governor
Tallahassee, FL 32399-0001
• Florida Farmers Home Administration
P.O. Box 1088
Gainesville, FL 32602
* National Agricultural Statistical Service
1222 Woodward Street
Orlando, FL 32803
• U.S. Environmental Protection Agency
345 Courtland Street, NE
Atlanta, GA 30308
U.S. Forest Service
227 N. Bronough Street, #4061
Tallahassee, FL 32301
i South Florida Water Management
District
P.O. Box 24680
West Palm Beach, FL33416
i USDA-Agricultural Research Service
Bldg. 164, University of Florida
Gainesville, FL 32611
i Florida State Soil Conservation Service
P.O. Box1208
Gainesville, FL 32602
• Department of Environmental
Regulations
2600 Blair Stone Road
Tallahassee, FL 32301
• Soil & Water Conservation Bureau
P.O. 60x1269
Gainesville, FL32602
• Cooperative Extension
1038 McCarty Hall'
Gainesville, FL 32611
• Florida Division of Forestry
Collins Building
Tallahassee, FL 32301
240
-------
J.W.STANLEY
agencies did not feel that one was trying to over-
shadow the other.
Agricultural Stabilization and
Conservation Service
The Agricultural Stabilization and Conservation Ser-
vice was responsible for the overall day-to-day opera-
tion of the RCWP project. The Service's Okeechobee
County committee chairman chaired the local coor-
dinating committee. The agency took applications
for participation, and, along with the Soil Conserva-
tion Service and Soil and Water Conservation Dis-
trict, set priorities for each application. After the
plans were written, the Agricultural Stabilization and
Conservation Service county committee reviewed
and approved them and contacted participants for
the formal signing.
The Agricultural Stabilization and Conservation
Service tracked all best management practice
(BMP) implementation, issued cost-share approvals
and payments, and monitored the contracts to en-
sure plans were followed properly. The agency par-
ticipated in the information and education program,
was a member of the administrative subcommittee,
and was responsible for coordinating, writing,
publishing, and distributing the annual and 10-year
reports.
Soil Conservation Service
The Soil Conservation Service was responsible for
planning and implementing the BMPs and took part
in setting priorities with the Agricultural Stabilization
and Conservation Service and the Soil and Water
Conservation District board. Soil Conservation Ser-
vice provided the annual status review on contracts,
worked with the Agricultural Stabilization and Con-
servation Service to make sure BMPs were applied,
and monitored maintenance and operation of the in-
stalled BMPs. The agency participated in the infor-
mation and education program, served on the
administrative subcommittee, and helped write the
annual report.
Cooperative Extension
Cooperative Extension provided news releases and
publications, organized field days and tours for the
project, conducted some field studies, and con-
tributed to the annual report. Cooperative Extension
was also involved in other areas of the project and a
member of the administrative subcommittee. The
University of Florida provided assistance to the
project and participants in conjunction with the local
Cooperative Extension office.
South Florida Water Management
District
Although it did not receive funding, the South
Florida Water Management District provided a com-
prehensive monitoring network and funds for some
research and evaluated water quality for the project.
The agency also participated in the information and
education program, helped write the annual and 10-
year reports, and was a member of the administra-
tive subcommittee, where it worked closely with the
three Federal agencies.
Soil and Water Conservation District
The Soil and Water Conservation District helped set
the priorities for each application and reviewed and
approved the plans written by the Soil Conservation
Service. This State agency participated in field days
and tours and was an active member of the local
coordinating committee.
Other agencies participated as members of the
local coordinating committee but did not have a role
in the day-to-day operation of the project.
Other Representatives
The administrative subcommittee worked closely
with several groups to administer the project. Local
and State dairy organization representatives were
consulted and participated in many phases. The
State Dairy Association's environmental specialist
was a member of the local coordinating committee
and played an active daily role, working to help coor-
dinate implementation of the project with dairymen.
(Local cattlemen also provided similar assistance.)
State and local environmental groups were consulted
and invited to participate in the project. All these rep-
resentatives played a vital role in the project's institu-
tional arrangement and administration.
Successes and Failures
The Taylor Creek-Nubbin Slough RCWP project con-
tracted 54,709 critical acres on 48 farms, which far
exceeded its goal. Cooperation among the four key
agencies helped in this effort: the group coordinated
visits with potential participants individually or as a
group to increase participation. One-on-one contact
with the farmer brought the best results, as il-
lustrated in Table 3, which divides the basin into sub-
watersheds and breaks down the contracted acres,
percent of contacted acres compared to the total
critical acres, total farms and number of farms con-
tracted in each subwatershed by each year of the
241
-------
Proceedings of national RCWP Symposium, 1992
2
Ł
S
_
I?
0 r
f]
w
.2 :
>e
s>
:Ł
• S
N *
»Ł
0 jR
O
13
<
<8 C
is (
m
n m
T> in
o> m
o> m
O tf
CD CO
05 T-
o> o
en o
p
o
o
_ S
S "
1
c
(8
NW Taylor Creek
en «^ co in co vj
§ &> •: co 05
O) > CO °°
^ co in co ^
0) S g
CD go co m ' ^
jl4010 vJi
> ^ co w co ^
li1010 |i
g ^ to in o JR
5 O)
3 i " °
0 g? CD 0 0 ^
ogcoo og
T3 T3
CD CD
T3 o P P c
SJ CO S ^2 CO
O i= O O i:
ra c « 2 c
i= o -5 i= •= i
o ssŁo o ss
CO ^ '
k-. »— Ł-
< UL <
E
in
5
CM in ^ co CM
•8.
co in vŁ co co
88s
- M 3 N§ CO CO
en o>
•CM m ^§ co CM
8»
c CM in ^ co co
f CM CO jo CO i-
f 0 CM ^§ CO T-
C*J JJ
CO «>
ra »
* ° fo ^ W ^
gS
.0 OgCOO
T3
(D
•g -g t3 1
o i i; i :
2 t: c o S i
o o o _ o o
HO O S? H O
w co '
E CD E
J— - Łr Js
LL. < LI-
CD
CD
0
aj : O
CD >>
0 H
CD : -|
0 :':- 5
"Tg
; o
S
! o
O ^
! o
1 CD
! o
3 ^
r °
in ^
4? X|
li
S ^
o
T3
C
O ^
•i= <
CD
<
O N-
O N-
O jj-
N °
) O Ł;
o
> o ^
1
1) CD CM
^^
n co CM
n •* e\
in
CO
P . "5
CD <
o c
S c 1
1
as
LL.
Williamson Ditch
vfl -i— O
> T~ T"
en
vŁ T— O
&''"''"
CD
v^ i— O
3 *~ '"
CD
l>
b *~
CD
s? j: •••
^
CO
o
^ ^ '""
CO
r
T3
0)
,*77 O
m
< Li-
Mosquito Creek
3 v$ CM CO
O ^ CM CO
O>
00 xo CM CM
cn °° *" *~
0 ^ CM CM
3 ^5 cj e\
8I-"
g g? ^ Is-
^ 1^
%$-*
o j$ T- o
O ^~
"O
0}
SI S
O " '
< LL
Nubbin Slough
! c&
CO *"
i«
*•? ^
CM
1 *°
" r
'|ln
!? g
•< [5
||
rt
us
T3
C
"S '
^ 1
o s?
CO
CD
i 2! oo
CM ^
JJ 00
CM *"
T- co g?
'CM *
) i- CO go
S "O
CM ")
} i- CO ^
^ CV
' •*" S <^
' s •*" 5v
i
o^
o o o jg
c
(
•g -g 1
II 1^
F O O ^
CO 0]
UL <
Henry Creek . - ,
.
in *
in o
p
P O
' C
LL
LettucaCreek.
242
-------
J.W. STANLEY
Table 4.—Percent Implementation of best management practices by subwatershed by year.
SUBWATERSHED 1981
NW Taylor Creek
Little Bimini
Otter Creek
Taylor Creek Main
Williamson Ditch
Mosquito Creek '
Nubbin Slough
Henry Creek
Lettuce Creek
1982 1983
2%
31% 36%
10% 35%
1%
8%
60%
1984
43%
48%
50%
6%
25%
14%
60%
1985
56%
87%
75%
57%
80%
20%
39%
94%
5%
1986
69%
91%
77%
90%
100%
50%
81%
98%
10%
1987
100%
95%
97%
99%
100%
77%
97%
100%
77%
1988
100%
99%
99%
100%
100%
99%
100%
100%
100%
1989
100%
99%
99%
100%
100%
100%
100%
100%
100%
1990
100%
100%
99%
100%
100%
100%
100%
100%
100%
project. Note that the contracting period was from
1981 until 1986; therefore, the numbers remained
constant for the rest of the project's lifespan.
The administrative subcommittee targeted sev-
eral areas for contracting. Changes in water quality
in the Taylor Creek headwaters were measured first
because the State had allocated funding for BMPs
and early studies of water quality for this area. The
Taylor Creek headwaters include NW Taylor Creek,
Little Bimini, and Otter Creek subwatersheds. Table
3 shows that most of the contracting in the head-
waters area was completed early in the project.
The second area targeted was Williamson Ditch:
the subcommittee thought it important to have good
participation in an area that was predominately beef
operations—and Table 3 shows their success in
recruiting participants. Dairies were also targeted.
Because of the large herds, dairies were the largest
contributors to water quality problems in the basin.
The project goal—to contract with all dairies in the
project—was accomplished.
Contracting critical acres is important only if
these acres are treated as contracted. In the Taylor
Creek-Nubbin Slough project, 99 percent of the
BMPs had been installed and were being maintained
at the end of the 10 years. Table 4 shows the percent
of implementation in the subwatersheds. Because
many practices were installed early, trend analyses
were possible and indicated that BMPs were having
a positive effect These areas were also targeted for
early implementation. Table 4 shows that in NW
Taylor Creek, Little Bimini, Otter Creek, and Wil-
liamson Ditch most of the implementation was com-
pleted by 1987—another result of the administrative
subcommittee's coordinated effort.
The administrative subcommittee decided to
add more BMPs after they saw that initial plans
lacked the practices needed to mitigate water quality
problems in the Taylor Creek-Nubbin Slough basin.
A request for additional BMPs was submitted to the
local coordinating committee, then to the State coor-
dinating committee, and lastly to the National Coor-
dinating Committee, which gave final approval.
Table 5 lists the BMPs originally selected and those
that were added. The project experienced a wide
variety of problems installing BMPs—but the sub-
committee helped resolve them quickly. (For ex-
ample, most farm plans had to be modified to
accommodate conditions at the time of installation,
but these changes were easily coordinated.)
Table 5.—Approved best management practices for
the Taylor Creek-Nubbin Slough project.
BMP-1 * Permanent vegetative cover
BMP-2* Animal waste management system
BMP-5 Diversion system
BMP-6* Grazing land protection system
BMP-8* Cropland protection system
BMP-10 Stream protection system
BMP-11 * Permanent vegetative cover on critical areas
BMP-12* Sediment retention, erosion, or water control
structures
BMP-13* Improving an irrigation and/or water manage-
ment system
*BMPs added during project
The project weathered several environmental
crises and regulatory programs. Progress was aided
by the project administrator, who understood all
phases of the project and worked closely with the ad-
ministrative subcommittee to make sure it was kept
abreast of project activities. The administrator also
kept other interested parties informed on the
project's progress and kept up-to-date on the scien-
tific and political environment. On several occasions,
the administrator brought farmers and environmen-
tal groups together with the project agencies to dis-
cuss and try to understand differences on issues
affecting the project.
The information and education program was
coordinated by the subcommittee and the local coor-
dinating committee. In retrospect, the subcommittee
recognized the following problems, which were not
identified early enough to be funded in the project:
• Cooperative Extension should have had a
more actively funded program,
243
-------
Proceedings of Hattonal RCWP Symposium, 1992
• Demonstration projects were needed to show
the effectiveness of growing row crops on
dairies, >
• Work on the use of dairy wastewater was
limited by funding, and
• Management guides were needed for
operation and maintenance of installed BMPs.
Farms in the project area were assigned a high,
medium, or low priority as they signed up in the pro-
gram. High priority plans were written first. For ex-
ample, if a farmer came in and signed up, the farm
would be assigned a priority based on the com-
mittee's determination of the water quality problem.
High priority farms would get the next sequentially
high priority number.
Looking back, the local coordinating committee
now thinks each farm identified in this project
should have been preassigned a priority number
based on the degree of its water quality problem. In
this manner, the farm determined to have the worst
problem would be Number 1 and the one with the
least important problem would be Number 59. If the
farm with preassigned Number 1 contracted later in
the sign-up period, its plan would be written immedi-
ately. The method actually used by this project would
have given farm Number 1 the next sequential high
priority number available when the farmer signed an
application, and its plans would have been written
only after the plans for farms with higher priority
numbers. Using this method, plans for farms that
strongly affected water quality might be delayed.
The project worked with State agencies and en-
vironmental and farm groups to get additional fund-
ing for research, cost-share monies, technical
assistance, and water quality monitoring. The State
of Florida put $250,000 into cost-shares in the
project. The South Florida Water Management Dis-
trict provided $12,000 for technical assistance,
funded several research projects with Cooperative
Extension, and provided a comprehensive monitor-
ing program. The District also conducted meetings
with the assistance of other subcommittee members
to show participants the results of BMP implementa-
tion—a tactic that encouraged installation of addi-
tional BMPs.
The overall success of the RCWP project and its
administration should be measured by improve-
ments in water quality. Figures 1 and 2 show the
phosphorous and nitrogen concentration over the
life of the project. The goal of a 50 percent reduction
in phosphorous was achieved; however, the goal of
50 percent reduction in nitrogen fell short at 46 per-
cent. Many factors affect water quality other than
BMP implementation. Because of these factors, such
as cow numbers, rainfall, and surface water levels,
the project has yet to prove statistically that BMPs
caused the water quality improvements. Neverthe-
less, looking at these factors subjectively, the BMPs
seemed to have had a positive effect.
Conclusion
This project was run efficiently and effectively be-
cause it established a subcommittee that was al-
lowed to manage day-to-day activities. The sub-
committee also increased participation and im-
plementation by working closely with key people
from the farmer groups. Because of these factors,
the project's water quality goals were met and par-
ticipation goals exceeded. The subcommittee's
smaller size made decisionmaking easier.
However, these factors only established the
framework for the most important element—people.
This project succeeded because people had the
dedication and desire to see it through to success.
Findings and Recommendations
• One person must be in charge to see that the
project stays on course and that goals are met. This
person, who should be selected by the local coor-
dinating committee, must understand all aspects of
the project so that coordination among the different
agencies can be maintained.
• A close working relationship must be maintained
among the key players of a project. This relationship
among the Taylor Creek-Nubbin Slough project
players made the project work.
• The project players need local autonomy and
flexibility.
• The key agencies must share and understand the
roles of other key agencies so that interactions can
take place in all phases and areas of the project.
• Projects should not limit their search for re-
sources to the Federal sector. This project was
awarded State monies for cost-sharing, technical as-
sistance, research, and water quality monitoring.
244
-------
J.W. STANLEY
1.4
1.2
0.8
I 0.6
0.4
0.2
1.17
0.93
0.85
0.77
0.64
1.12
1 1
1.04
1.04
0.84
0.75
0-79 o.78
0.72
0.59 QSJ 0.59
1973 1975 1977 1979 1981 1983 1985 1987 1989
Pre-RCWP | RCWP
Figure 1.—Total phosphorus annual averages over the life of the project.
3.5
3.05
2.65
2.5
o
§ 1.5
2.05
2.14
2.23
0.5
1.88
1.97
3.15
2.7
2.32
2.2
2.11
2.01
2.06
1.8
1.67
1.46
1973 1975 1977 1979 1981 1983 1985
Pre-RCWP | RCWP
Figure 2.—Total nitrogen annual averages over the life of the project.
1987 1989
245
-------
-------
Farm Operators' Attitudes About
Water Quality and the RCWP
Thomas J. Hoban and Ronald C. Wimberley
Department of Sociology and Anthropology
North Carolina State University •
Raleigh, North Carolina
ABSTRACT
Because participation in the Rural Clean Water Program (RCWP) was voluntary, water quality im-
provements ultimately depended on farm operators' willingness and ability to use best manage-
ment practices (BMPs). Evaluation of the RCWP, therefore, must include a systematic analysis of a
farm operator's knowledge, attitudes, and behavior. For this project, telephone interviews were
conducted with 1,111 farm operators, selected at random from the 21 RCWP project areas. Inter-
views were about equally divided between participants and nonparticipants. This paper describes
the survey results in terms of the following: attitudes about water quality; attitudes and adoption of
BMPs; attitudes about the RCWP; and attitudes about public policies. The interviewers also asked
participants and nonparticipants specifically why they did or did not participate in the RCWP.
The goal of the Rural Clean Water Program
(RCWP) has been to reduce pollution from
agricultural land. Improvements in water
quality depend on changes in farm operators' at-
titudes and adoption of best management practices
(BMPs). Such changes should be facilitated by par-
ticipation in RCWP and increased awareness of
water quality issues. Evaluation of the success of the
RCWP should, therefore, include analysis of farm
operator's knowledge, attitudes, and behavior. To do
this, we developed and conducted a telephone sur-
vey to determine farm operator participation in the
RCWP and adoption of BMPs. This paper describes
the major results of that survey.
This survey builds on a long-standing research
tradition. For the past 50 years, social scientists have
studied the process by which farm operators and
others accept and use new practices (Rogers, 1983).
During the past 15 years, much of this attention has
focused on farm operators' adoption of soil conserva-
tion practices (Nowak, 1984; Buttel et al. 1990). A
subset of this work has specifically considered farm
operators' use of BMPs for controlling nonpoint
source water pollution. Research has shown that
decisions about using BMPs represent an ongoing
process for most farm operators (Korsching and
Nowak, 1983). Water quality improvements from
nonpoint source control depend on changes in farm
operators' attitudes and behavior. Voluntary adoption
of BMPs means that farm operators must initially
gain sufficient awareness of and interest in water
quality problems and related issues (Napier et al.
1986). Next, they must develop favorable attitudes
toward the recommended BMPs (Bultena and
Hoiberg, 1986). Finally, these attitudes must be
translated into adoption of the BMPs.
Research Methods
We established an advisory committee—made up of
agency officials, project personnel, social scientists,
and other people who had experience with the
RCWP—to provide advice on sampling design and
survey content. The survey content was based on
review of previous research as well as comments
from the advisory committee, which had the oppor-
tunity to review the survey instrument.
247
-------
Proceedings of National RCWP Symposium, 1992
One major goal of this project is to compare farm
operators who participated in the RCWP with those
who were eligible but did not participate. Local U.S.
Department of Agriculture Agricultural Stabilization
and Conservation Service (ASCS) offices provided
lists of all farm operators who designated whether or
not they had participated in the RCWP. The sampling
procedure was developed in consultation with a
professor of statistics at North Carolina State Univer-
sity (NCSU). The approach involved a compromise
between drawing a proportional sample from each
RCWP area based on the number of farm operators
in each area and drawing an equal size sample from
each RCWP area regardless of the number of farm
operators. The sample of farm operators drawn from
each area was proportional to the cubic root of farm
operators from that area participating in the RCWP.
Once the target sample size for each RCWP area was
determined, farm operators were selected at random
from the ASCS lists. Careful attention was paid to
completing the target number of participants and
nonparticipants from each RCWP area.
A standard telephone survey was used to collect
data. Telephone interviews were selected because
they provide the most effective and efficient means
to systematically collect comparable information
from a large sample, especially when the populations
are widely dispersed geographically. Data was col-
lected by the Applied Research Group at NCSU, an
organization with over 15 years experience conduct-
ing telephone interviews for all types of public agen-
cies and private organizations. Their telephone
interviewers are well trained and carefully super-
vised.
All telephone interviews, which averaged 20
minutes, were conducted during November and
December of 1991. In many cases, callbacks were
scheduled at times convenient for the farm
operators. The overall response rate for the survey
was almost 85 percent of the selected respondents.
In general, the response rate was even higher for
farm operators who had participated in the RCWP
compared to those who were eligible but did not par-
ticipate. Respondents appear to be representative of
the farm population. Sample demographics and farm
structure characteristics will be published in the
final report.
Results
This research, which will ultimately analyze farm
operators' participation in the RCWP as well as their
adoption of BMPs, will include the development and
testing of a theoretical model that specifies a set of
factors that may influence farmers' actions. Inter-
relationships among the factors will be tested in fu-
ture analysis. This paper describes the general
results and highlights the major findings from the
survey in terms of
• water quality awareness and information,
• attitudes about water quality problems,
• adoption of BMPs,
• participation in the RCWP, and
• attitudes about public policies and programs.
Water Quality Awareness and
Information
Awareness of water quality problems will likely be
related to farm operators' adoption of BMPs and par-
ticipation in the RCWP. In general, farm operators
expressed quite a bit of awareness. When asked
"How much have you heard or read about how
agriculture might affect water quality," 52 percent
reported "a lot"; 32 percent had heard "some"; 14
percent had heard "a little"; and only 2 percent
claimed to have heard "nothing."
Interest in learning more about water quality
was .relatively low, compared to awareness. Farm
operators were asked "Overall, how much more in-
formation do you need about what you can do on
your own farm to help protect water quality." Ten
percent said they wanted "a lot more information";
36 percent said they needed "some more informa-
tion"; 29 percent wanted "a little information"; 24 per-
cent said they did not need any more information
about protecting water quality.
Respondents were also asked about their use of
various information sources: "Farmers can do many
things to protect water quality. How much informa-
tion about protecting water quality have you
received from (source)? Have you gotten a lot, some,
a little, or no information?" Results for the 12 dif-
ferent information sources are shown in Figure 1.
Farm magazines represent the most frequently used
source of information, followed by government
agricultural and conservation agencies, with the U.S.
Department of Agriculture's Soil Conservation Ser-
vice (SCS) being most popular. Newspapers were the
fifth most commonly used source of information.
A substantial drop in .frequency of use was ex-
hibited for the remaining seven sources of informa-
tion. Respondents listed other farm operators next
and reported about the same level of use for
television, meetings, or workshops. Farmers used
pesticide or fertilizer dealers, other farm organiza-
248
-------
T.J. HOBAtt & R.C. WIMBERLEY
Farm Magazines
USDA-SCS
USDA-ASCS
Extension Service
Newspapers
Other Farmers
Television
Meetings/Workshops
Farm Businesses
Farm Organizations
Radio
Tours/Demonstrations
i
0.5 1 1.5 2 2.5 3
Average (Mean) Score
Information Received: 0 = None, 3 = A lot
Figure 1.—Farmers' sources of water quality informa-
tion.
tions, and radio programs infrequently for informa-
tion on water quality. Tours and demonstrations
were the least frequently used sources of informa-
tion.
Attitudes about Water Quality
Problems
Attitudes about the severity and causes of water
quality problems should have important impacts on
adoption of BMPs and participation in the RCWP.
Respondents were first asked "Is water pollution a
serious problem, somewhat of a problem, or. not a
problem in your area?" Only 14 percent said water
pollution was a "serious problem," with the
remainder evenly divided between 43 percent, who
felt pollution was "somewhat of a problem" and 43
percent, who felt it was "not a problem." Problem
perception was even lower when respondents were
asked the same question about their own farm, In
this case, only 2 percent said pollution was "a serious
problem," and over 79 percent claimed no problems
with water pollution on their farms. The rest said
"somewhat of a problem."
Concerns for specific impacts of water pollution
appear moderately high. Respondents were asked
"Are you very concerned, somewhat concerned, or
not concerned about pollution of your own drinking
water?" Responses are evenly split between those
who were very concerned (33 percent), somewhat
concerned (34 percent) and not concerned (34 per-
cent). Two-thirds of all respondents either strongly
agreed (11 percent) or agreed (55 percent) that
"agricultural water pollution is a serious threat to fish
and wildlife."
The survey also determined if respondents felt
that agriculture represented a serious cause of water
pollution. Near the start of the interview, respon-
dents were asked an open-ended question: "What do
you think are the major causes of water pollution in
your area?" After an initial answer, respondents were
asked if there were any other causes and could give
more than one answer to this question. General
results are shown in Figure 2.
Cropland/Farm Chemical
Livestock Waste
Industrial Discharge
Municipal DischargeH|12
Home Septic Tanks
Litter or Garbage | e
Urban Runoff H 5
0 10 20 30 40 50 60 70
Percent Mentioning Cause
Figure 2.—Farmers' perceptions of major causes of
water pollution.
The most frequently mentioned cause of water
pollution (from over a third of respondents) was
"runoff from cropland": almost 25 percent specifical-
ly mentioned pesticides (including herbicides or in-
secticides), while 20 percent mentioned fertilizers or
specific nutrients, such as nitrogen. To account for
respondents who might have reported two or more
of these responses, we combined these three causes
into one indicator of pollution: Cropland/Farm
Chemical. In this case, just over half (59 percent)
saw cropland as a major cause of pollution. Far fewer
respondents (17 percent) recognized livestock or
animal waste as a cause of pollution. Combining crop
and livestock nonpoint source pollution, we found
two-thirds of all respondents believed agriculture to
be a major cause of water pollution.
Respondents reported a variety of other non-
agricultural causes of water pollution in their areas.
Twenty-seven percent said that point sources were
majpr causes of pollution. In this case, 15 percent
said industrial waste or factory discharge was the
major source, while another 12 percent cited
municipal sewage treatment. Other sources of pollu-
tion mentioned included household septic systems
(7 percent), litter or garbage (6 percent), and urban
runoff (5 percent).
Several other questions directly addressed
respondents' attitudes about the potential contribu-
tion of agriculture to water pollution problems. Over
half of all respondents either strongly agreed (8 per-
cent) or agreed (46 percent) that "The farming prac-
tices I use now have no significant effect on water
quality in my area." In fact, over three-quarters
249
-------
Proceedings of national RCWP Symposium, 1992
strongly agreed (14 percent) or agreed (63 percent)
that "agriculture is being unfairly blamed as a cause
of water quality problems."
Adoption of Best Management'
Practices
One major goal of this project was to determine the
extent to which farm operators had adopted recom-
mended BMPs to control water pollution from their
farms. Respondents were read the following: "Some
farming practices are available to help farm
operators protect water quality. These are often
called best management practices or BMPs. Are you
now using any of these BMPs?" Respondents were
then asked specifically about the 17 BMPs being
promoted overall as part of the RCWP (Fig. 3). Fu-
ture analysis will determine in a general way
whether specific BMPs were relevant for a particular
type of farm operation as well as whether specific
BMPs were promoted by a particular RCWP project.
Soil Testing
Paatiddo Management
Conservation Tillage
Grass Waterways
Cover Crops
Hayland/Pasturo Mgmt
Permanent Veg Cover
Grass/Logumo Rotation
Nutrient Management
Diversions
Animal Waste Mgmt
Stream Protection
Filter/Buffer Strips
Sediment Traps
Contour Strip-Crops
Irrigation Improvement
Terraces
• 75
• 75
• 73
I 70
166
I 65
20
40
60
80
100
Percent Using BMP
Figure 3.—Adoption of best management practices
(BMPs) by farmers.
Nine of the BMPs listed were relatively more
popular with the farm operators in our sample.
These BMPs tended to be more management-
oriented than structural. Three-quarters or more of
the respondents were reportedly using soil testing
(82 percent); conservation or reduced tillage (75 per-
cent); and pesticide management or reduction (75
percent). Almost as many were using other manage-
ment practices: grass waterways (73 percent); cover
crops (70 percent); hayland or pasture planting or
management (69 percent); permanent vegetative
cover (66 percent); grasses or legumes in rotation
(65 percent); and nutrient management or reduction
(61 percent).
The remaining eight BMPs do not appear to be
as acceptable or applicable, as evidenced by their
lower adoption rate. Less than half of the respon-
dents reported adoption of diversions (48 percent);
animal waste management or storage (45 percent);
stream protection or fencing (44 percent); and filter
or buffer strips (40 percent). Less than one-third of
all respondents had adopted sediment traps (30 per-
cent); contour strip-cropping (30 percent); irrigation
improvement (26 percent); and terraces (25 per-
cent). Ninety-eight percent of the respondents, when
asked, replied they planned to continue using the
BMPs.
We also wanted to determine the features of a
BMP that affect a farm operator's decision about
using it. Respondents were read the following ques-
tion: "How important would each of the following fac-
tors be in your decision about whether to use a new
BMP to help protect water quality? Would (item) be
very important, somewhat important, or not impor-
tant in your decision to use a new practice?" Results
are shown in Figure 4. It is probably most meaning-
ful to consider respondents who said a particular fac-
tor was 'Very important," since relatively few said
the factors would not be important.
Cost of Practice
Water Quality Improve
Effects on Profits
Ease of Use
Labor/Time Required
Govern Cost Sharing
Others' Experience
Inform from Govern
Inform from Businesses
i
20 40 60 80 100
Percent Response
I Very Important BJ Somewhat H Not Important
Figure 4.—Influences on farmers' adoption of best
management practices (BMPs).
About 64 percent of all respondents said the cost
of the practice would be very important. Almost as
many (61 percent) believed the potential of the prac-
tice to improve water quality would be very impor-
tant, while 56 percent said that the effects of the
practice on profits would be very important. Exactly
half of all respondents thought that ease of use as
well as labor or time required would be very impor-
tant, just under half (47 percent) said availability of
government cost-sharing would be very important,
and 44 percent found the experience of other farm
operators to be very important. Two factors appear
to be relatively unimportant for the farm operators in
this sample: less than a third stated that information
from government agencies (31 percent) and farm
businesses (27 percent) would be very important.
Several additional questions addressed farmers'
attitudes about BMPs. Most respondents either
250
-------
T.J. HOBAN & R.C. WIMBERLEY
strongly agreed (10 percent) or agreed (74 percent)
that "We need less expensive farming practices that
will help protect water quality." Over half either
strongly agreed (4 percent) or agreed (56 percent)
that "Farm practices that protect water quality usual-
ly require more labor." It appears, therefore, that
economic and labor constraints may be important
factors that limit the adoption of BMPs.
Participation in the Rural Clean Water
Program
Another major goal of this project is to determine the
reasons why farmers either did or did not participate
in the RCWP. As discussed earlier, our sampling
design was developed to include both participants
and nonparticipants from each of the RCWP project
areas. We based our sampling design on the written
designation of eligibility and participation provided
by the local ASCS offices and verified the informa-
tion by asking farm operators if they had par-
ticipated. Results, however, showed some significant
discrepancies (about 15 percent) between the official
designation and farmers' own reports. Attempts are
being made to clear up this problem. However, it is
interesting to note that over 10 percent of all farmers
interviewed had never heard of the RCWP despite
the fact that the local ASCS had designated all
respondents as eligible for participation.
Depending on whether or not respondents said
they had participated in the RCWP, they were asked
a somewhat different series of questions. The 630
farm operators who claimed they had participated in
the RCWP were asked the following open-ended
question "What were the reasons you decided to par-
ticipate in the RCWP?" A probe ("Are there any other
reasons you participated?") was then used if respon-
dents gave an initial answer. This means some
respondents gave more than one answer to this
question. Results are shown in Figure 5.
Over half (59 percent) mentioned something re-
lated to concern for water pollution or its effects. Just
over one third (38 percent) said that the availability
of cost-sharing funds was an important reason. Al-
most one in five (19 percent) said they felt it was "the
right thing to do" or in some way expressed a con-
servation ethic. Three other reasons were given by a
sizeable number of respondents: increased farm
productivity (13 percent); concern over future pollu-
tion regulations (12 percent); and assistance or en-
couragement from government agencies (12
percent).
Respondents who participated were also asked
to rate their satisfaction with the RCWP. Overall,
satisfaction was very high. Almost all respondents
Concern for Pollution
Available Cost Sharing
Conservation Ethic
Increased Farm Product
Concern for Regulation
Govern Encouragement
70
10 20 30 40 50 60
Percent Mentioning Reason
Only includes the 630 RCWP participants.
Figure 5.—Farmers' reasons for participating In the
Rural Clean Water Program (RCWP).
were either very satisfied (52 percent) or satisfied
(42 percent) with the technical assistance and infor-
mation they received from the RCWP. Satisfaction
with the financial assistance was just as high with
most being either very satisfied (47 percent) or satis-
fied (46 percent). Another indication of effectiveness
was found by determining the impact of the RCWP
on adoption of BMPs. Respondents were evenly
divided when asked "If the RCWP had not been avail-
able would you have been very likely [13 percent],
likely [35 percent], unlikely [34 percent], or very un-
likely [18 percent] to have used all of the best
management practices you are now using?"
The 481 respondents who had not participated
in the RCWP were asked a somewhat different set of
questions, including the following open-ended ques-
tion: "What were the reasons you decided not to par-
ticipate in the RCWP?" An appropriate probe—"Are
there any other reasons you did not participate?"—
was used for respondents who gave an initial answer.
Some respondents gave more than one answer;
results are shown in Figure 6.
No Pollution Problem
Resistant to Change
Dislike Government
Didn't Hear of RCWP
No Funds or Ineligible
Not Invited to Sign-up
Economic Factors
0
5 10 15 20 25
Percent Mentioning Reason
Only includes the 481 RCWP nonparticipants.
Figure 6.—Farmers' reasons for not participating In the
Rural Clean Water Program (RCWP).
251
-------
Proceedings of National RCWP Symposium, 1992
Almost 25 percent of the nonparticipants said
they did not participate because they did not believe
water pollution was a problem, either on their own
farm or in general. The next major reason for non-
participation (15 percent) involved some form of
resistance to change, which included the idea that/a
farmers' present system works well or that changing
practices is too much trouble. Over 10 percent said
they simply did not like government programs, in-
cluding the red tape or complicated procedures. Al-
most as many (10 percent) said they had never
heard of the RCWP at the time it was available, and 9
percent claimed they had tried to sign up but were
either ineligible or no money was available. In addi-
tion, 8 percent volunteered that they were never
asked to participate. Only 6 percent said anything
about economic factors, including a belief that the
cost-share rates were too low.
Respondents who had not participated in the
RCWP were asked specific follow-up questions about
contacts by government agencies asking them to
participate. Almost two-thirds (59 percent) of the
nonparticipants claimed that no government agen-
cies had ever contacted them about participating in
the RCWP. However, on a more positive note, almost
two-thirds said they would be either very likely (15
percent) or likely (49 percent) to participate if a new
program (like the RCWP) were available today.
The survey also determined how participants
and nonparticipants assessed the overall impacts of
the RCWP on a variety of different outcomes. Both
groups were asked "Would you say the RCWP had a
positive effect, no effect, or a negative effect on
(item)?" (This question was not asked those respon-
dents who had never heard of the RCWP.) Results
are shown in Figure 7. Ninety percent of the respon-
dents thought that the RCWP had a positive effect on
farm operators' knowledge about water quality, and
86 percent also stated the RCWP had a generally
positive effect on surface water quality in the area.
Approximately two-thirds believed the RCWP
had a positive effect on operating costs for participat-
ing farm operators (70 percent) as well as on farm in-
come (66 percent). Only 8 percent of the
respondents felt that the RCWP had negative effects
on these economic conditions. Just over half (56 per-
cent) believed that the RCWP had a positive effect
on drinking water quality in the area, while 41 per-
cent thought the RCWP had no effect on drinking
water quality. These results might be explained by
the fact that few RCWP projects specifically targeted
drinking water problems.
Earlier in the interview all respondents had been
asked "Compared to ten years ago, do you think
water quality in your area is better, about the same,
Farmers' WQ Knowledge
Surface Water Quality
Farm Operating Costs
Farmers' Incomes
Drinking Water Quality
20 40 60 80 100
Percent Response
I Positive m No Effect 11 Negative
Figure 7.—Farmers' perceptions of effects of the Rural
Clean Water Program (RCWP).
or worse?" Respondents were divided when asked to
assess trends in water quality. Over one-third (36
percent) believed that water quality had gotten bet-
ter over the previous decade; almost half (47) felt it
had remained about the same; and nearly one-fifth
(17 percent) felt water pollution was now worse.
Attitudes about Public Policies and
Programs
The RCWP was an experiment to determine whether
agricultural nonpoint source water pollution could be
controlled adequately through voluntary implemen-
tation of BMPs on farms. The key assumption of the
RCWP approach was that targeted high levels of
financial and technical assistance and education
would be adequate to promote widespread coopera-
tion. Some of the questions in this survey have direct
relevance to the issue of voluntary versus mandatory
control of agricultural nonpoint source pollution.
Further analysis of results should provide additional
insights.
Most farm operators appear to support the
voluntary approach taken by the RCWP. Almost all
• respondents either strongly agreed (12 percent) or
agreed (84 percent) that "Water pollution can best be
controlled through educational programs that en-
courage farm operators to use BMPs." As further
evidence of support for the voluntary approach, 9
percent strongly agreed and 60 percent agreed that
"the government should help pay more for water pol-
lution control on farms."
On the other hand, many farm operators recog-
nized that the voluntary approach may not work in
all cases. Most seem resigned to the inevitability of
government regulations. Almost all either strongly
agreed (15 percent) or agreed (74 percent) that "If
farm operators don't do more to protect water quality
on their own, the government will force them to
252
-------
T.J. HOBAN 6 R.C WIMBERLEY
through regulation." In fact, most also either strong-
ly agreed (9 percent) or agreed (76 percent) that
"Farm operators do not have the right to farm in
ways that damage water quality." Almost as many
either strongly agreed (6 percent) or agreed (72 per-
cent) that "land should be farmed in ways that
protect water quality even if this, means lower
profits."
Conclusions and Discussion
This final section will draw several general con-
clusions and discuss some preliminary implications
of this research. More detailed recommendations
and conclusions will be forthcoming once further
analysis is completed. In the next few months, we
will determine the types of farmers who were most
likely to participate in the RCWP as well as adopt
BMPs. Such analysis will assess the relative in-
fluence of demographic variables, farm structural
characteristics, information and awareness, and
various attitudes.
Respondents reported fairly high general aware-
ness of water quality but rather low interest in learn-
ing more about it. The major sources of information
appear to be print media (especially farm magazines)
and government agencies. However, over 10 percent
of all respondents did not recall having heard about
the RCWP. Awareness and concern for the impacts
of agriculture on water quality, on the other hand, ap-
pear rather low. (This is particularly true for
respondents' perceptions of water quality impacts as-
sociated with their farms.) One-third of all farmers
did not see agriculture as a major cause of pollution
in their area, which is contrary to the fact that these
areas were selected as RCWP projects presumably
because of serious nonpoint source pollution
problems from agriculture. Also, respondents were
even less likely to perceive impacts of agriculture on
drinking water quality. However, not all projects
specifically addressed drinking water pollution as a
water quality problem.
Some BMPs appear to have reached a fairly high
level of acceptance. (This is especially true for
management-oriented practices.) Further analysis
will be necessary to account for the extent to which a
specific BMP was promoted as applicable in a par-
ticular area. Likewise, certain practices may be ap-
propriate only for certain types of farm operations.
Respondents did indicate that the costs associated
with certain BMPs would be an important factor in
their adoption decisions.
Participation in the RCWP appears to have been
driven, at least in part, out of concern for the en-
vironment; economic factors do not appear an impor-
tant consideration. This last assumption seems
especially true when one considers the reasons
given for not participating in the RCWP. Many
farmers did not participate because they did not
think water pollution was a problem or did not want
to deal, with changing practices. Furthermore, al-
most one-third of the nonparticipants said they
either had never heard of the RCWP, claimed they
were never asked to participate, or voiced general
dislike for government programs.
An initial reading of these results indicates that
the RCWP should have more actively promoted
awareness of and interest in water pollution
problems associated with agriculture. Further
analysis of these results should help to evaluate the
extent to which the RCWP led to changes in be-
havior, such as BMP adoption. Most farmers seem to
recognize that the completely voluntary approach
will not be viable much longer and seem to accept
the fact that society has a right to demand higher
levels of pollution control from agriculture. Most,
however, would also generally agree that it is "that
other farmer" who has the problem and needs to be
more tightly regulated.
References
Bultena, G.L. and E.G. Hoiberg. 1986. Sources of information and
technical assistance for farmers in controlling soil erosion.
Pages 71-82 in S.B. Lovejoy andT.L. Napier, eds. Conserving
Soil: Insights from Socioeconomic Research. Soil Coiiserv.
Soc. Am., Ankeny, IA.
Buttel, F.H., O.F. Larson, and G.W. Gillespie. 1990. The Sociology
of Agriculture. Greenwood Press, New York, NY.
Korsching, P.P. and PJ. Nowak. 1983. Social and institutional fac-
tors affecting the adoption and maintenance of agricultural
BMPs. Pages 349-73 in F. Schaller and G. Bailey, eds. Agricul-
tural Management and Water Quality. Iowa State Univ. Press,
Ames.
Napier, T.L., S.M. Camboni, and C.S. Thraen. 1986. Environmental
concern and the adoption of farm technologies. J. Soil Water
Conserv. 41:109-13.
Nowak, P.J. 1984. Adoption and diffusion of soil and water conser-
vation practices. Pages 214-37 in B.C. English, J A Maetzold,
B. R. Holding, and E. O Heady, eds. Future Agricultural Tech-
nology and Resource Conservation. Iowa State Univ. Press,
Ames.
Rogers, E.M. 1983. Diffusion of Innovations. 3d ed. The Free
Press, New York.
253
-------
-------
Factors Leading to Permanent
Adoption of Best Management
Practices in South Dakota Rural Clean
Water Program Projects
Karen Cameron-Howell
Soil Conservation Service
U.S. Department of Agriculture
Brookings, South Dakota
ABSTRACT
The Oakwood Lakes/Poinsett Rural Clean Water Program (RCWP) project has had a permanent
influence on the farming practices within and around the project area. Project participation ac-
celerated after key community persons became active and after a local technical planner was hired.
Cash incentives were the primary reason persons signed up for RCWP; however, most continued
applying best management practices (BMPs) even after cash incentives ceased. A post-RCWP
questionnaire was given to all participants to find out how the program influenced them, why they
signed up, and if they changed farming methods as a result of the program. This paper describes
the methodology used to assess these permanent influences.
ander Section 208 of the 1972 Amendments
to the Federal Water Pollution Control Act,
individual states identified nonpoint water
quality problems and began to develop plans to con-
trol these sources (U.S. Congress, 1972). The most
significant problem identified in South Dakota was
sediment and associated chemicals and nutrients
from farmland.
In response to Section 208, a new federal pro-
gram called the Rural Clean Water Program (RCWP)
was created (Federal Register, 1980). This program
offered cost-share and technical assistance to land-
owners who adopted specific erosion control and
water quality practices on a voluntary basis.
These practices were referred to collectively as
best management practices (BMPs). Each par-
ticipant signed up for a minimum of three BMPs.
Participants were also required to adopt conserva-
tion measures that limited soil losses to "T" or less,
that is, to the annual soil loss tolerance for each soil
map unit in each field. Water quality plans were writ-
ten by Soil Conservation Service (SCS) field person-
nel for entire farms.
The Oakwood/Lakes Poinsett RCWP project in-
cluded Brookings, Hamlin, and Kingsbury Counties.
Sign-up was slow the first year until farmers became
familiar with the requirements and benefits of the
project. Eventually, RCWP contracts covered 48,088
acres or 60 percent of the project's critical area. Cost-
share assistance was provided by RCWP funds
through the Agricultural Stabilization and Conserva-
tion Service (ASCS) at a rate of 75 percent. The total
cost-share assistance provided to .producer par-
ticipants was $744,729. To assess the effect of this
project and to better handle other watershed
projects, a follow-up questionnaire was administered
at the project's end to RCWP participants only.
255
-------
Proceedings of National RCWP Symposium, 1992
Methods
The Brookings Conservation District mailed public
input questionnaires to all operators who had an
RCWP water quality plan in the period from 1982 to
1990, This mailing took place as part of the Brook-
ings Conservation District Hydrologic Unit Area
process. A cover letter accompanied each question-
naire explaining the importance of the operators'
responses and how their input could aid future
watershed planning. Ninety-five letters were sent; 49
operators responded and returned their question-
naires to the Conservation District — a 50 percent
response.
The questions were reviewed by SCS personnel
at the field office and area office levels; water quality
specialists at the Water Resources Institute at South
Dakota Sate University (SDSU); and by a social
scientist, also at SDSU. Advice was sought on ques-
tion design and content. The project goal was to
answer questions on farmer acceptance of BMPs,
what motivated them to participate, and whether or
not they had achieved permanent change as a result
of the RCWP participation. Only operators who par-
ticipated in RCWP received a questionnaire. Self-ad-
dressed stamped envelopes were included for their
return to the Brookings Conservation District.
Data analysis includes descriptive statistics, chi-
square, and frequencies. Analysis was performed at
SDSU using the Statistical Analysis System (SAS).
The data are not appropriate for an analysis of
variance. Parts of the questionnaire and individual
questions tended to be interrelated; however,
responses for each part can be individually inter-
preted: First, how much farmer acceptance was
there? Did cost-share influence their decision to par-
ticipate? Or did the program itself influence their
decision to participate? Did the program influence
them enough to change their farming methods after
the cessation of cost-share?
Results
The average age of respondents was 47 years. Two
and three-way comparisons with other responses did
not yield a particular age group that adopted BMPs
more or faster than others.
In general, most operators used the moldboard
plow prior to their participation in the RCWP. Thirty-
two operators (65 percent) used the plow either in
the fall or in the spring as some part of their usual
field preparation. Seventeen (35 percent) used some
type of conservation tillage method (chisel and disk
was the most common system).
Table 1 compares tillage methods (conservation
tillage and moldboard plow) prior to RCWP par-
ticipation to tillage methods after RCWP participa-
tion. In general, 29 (59 percent) changed tillage
methods from plowing to conservation tillage as a
result of RCWP participation. One of the goals of this
study was to assess the degree of permanent change
that RCWP fostered by use of attainable BMPs.
Table 1 shows that only 6 percent continued plowing
after the cessation of cost-share for BMP-9 (conser-
vation tillage).
Table 1.—Comparison of conservation tillage users
before and after RCWP participation.
FREQUENCY (PERCENT)
CONSERVA-
TION TILLAGE
AFTER RCWP
Used conser-
vation tillage
before RCWP 11 (22%)
Plowed
before RCWP 29 (59%)
Total 40 (82%)
PLOWED
AFTER
RCWP
6 (12%)
3 (6%)
9 (18%)
TOTAL
17(35%)
32(65%)
49(100%)
STATISTIC
STATISTICS FOR TABLE 1
DF* VALUE
PROBABILITY
Chi-square
Likelihood
Chi-square
Phi coefficient
Sample size = 49
1 4.974
1 4.751
-0.319
0.026
0.029
*DF = degree of freedom
Question 4 (Table 2) also shows acceptance of
new tillage methods by asking why post-RCWP
changes in tillage were made. Fifty percent of the
respondents have not changed tillage methods since
RCWP participation. Of those that did change, all but
one respondent changed to a more specialized con-
servation tillage method, including ridge-till and no-
till. Forty-eight of the 49 respondents said they
would recommend a program similar to the RCWP
to others.
Table 2.—Question 4: If your current tillage methods
are different than while you were enrolled in RCWP,
explain why you changed (n = 30).
RESPONSE
FREQUENCY PERCENT
1: Stayed the same 16 53
2: Enrolled in CRP or other
program 5 17
3: Changed crop rotation (bases
and farm) 5 ' 17
4: Changed to control more erosion 3 10
5: Conservation tillage did not
work — too wet 1 3
To further pinpoint farmer acceptance among
the 92 percent full-time farmers (the remaining 8
256
-------
K. CAMERON-HOWELL
percent were part-time farmers), respondents were
asked about farm productivity while they were
RCWP participants. Responses were varied depend-
ing on how they defined productivity; however, 55
percent thought their productivity increased, with in-
creased moisture for crop growth the most common
reason for this increase.
Interesting results were achieved from a com-
parison of three questions; two, eight, and nine.
These results are in Table 3. Tallied results of ques-
tions 2 and 9 were compared to the ranked results of
question 8. Table 3 looks at sums of the ranked
responses only; categorical modeling was not ap-
propriate with only a three-way table. In the first
case, persons who used conservation tillage before
RCWP and after the program expired cited
monetary incentives as the most important reason
they signed up for RCWP. The same respondents
cited concern for the environment as their second
most important reason, and fertilizer recommenda-
tions as the third.
The second group was comprised of persons
using conservation tillage before RCWP who
switched to a different system after the program ex-
pired — for example, to ridge-till or no-till. Cost-
share incentives ranked third after the fertilizer
recommendation and concern for the environment.
The third and largest group was composed of
those who plowed before RCWP participation and
used some form of conservation tillage after the pro-
gram. Twenty-nine of the respondents were in this
category, and cash incentives were the primary
reason they got involved in RCWP. The second
reason for their involvement was the "free scouting"
Table 3.—Tillage methods before and after RCWP compared to incentive. Three-way comparison of questions 2,
8, and 9 using ranked results of question 8.
offered by BMP-16 (pesticide management). The
fertilizer recommendations offered by BMP-15 (fer-
tilizer management) ranked as the third most impor-
tant reason for their involvement. These results
indicate that cash incentives are the primary reason
farm operators signed up for the RCWP and that the
RCWP influenced them to change permanently.
In future questionnaires, questions involving
ranking will not be used. Seven responses were not
used (n=42) because the respondents incorrectly
ranked their responses (by simply checking them)
or failed to choose five responses. These seven
responses to question 8 were not used in final tabula-
tion and in comparison with other questions' results.
The last comparison on Table 3 concerns
farmers who plowed before and after RCWP. They
also cited monetary incentives and concern for the
environment; however, the lack of proper ranking
does not allow accurate comparisons for this
category.
Now we know that 32 persons (65 percent)
changed from being nonconservation tillage farmers
to being conservation tillage farmers as a result of
the RCWP experience; and 29 of those 32 persons
went from plow to conservation tillage (91 percent).
The six persons that went from conservation tillage
to a plow method did so because of changes in crop
rotation that required some plowing to break sod, al-
though one person said that his ground was simply
too wet to use conservation tillage methods. Other
farmers continued the same conservation tillage
methods that they had been applying prior to RCWP
participation (see Table 1).
OBSERVATION
Conservation tillage before and after RCWP
Conservation tillage before, plow after
Plowed before RCWP, conservation tillage after
Plowed before and after RCWP
VARIABLE
Cost-share
Pest scouting
Fertilizer management
Concern for environment
Other
Time for change
Project
Cost-share
Pest scouting
Fertilizer management
Concern for environment
Other
Time for change
Cost-share
Pest scouting
Fertilizer management
Concern for environment
Other
Time for change
Project
Cost-share
Concern for environment
SUM
34
12
15
23
11
5
3
7
6
12
13
1
2
79
64
62
57
19
38
3
6
2
257
-------
Proceedings of National RCWP Symposium, 1992
But how did farm operators hear about RCWP?
Fourteen persons heard about RCWP from a neigh-
bor, 13 from the SCS, 10 from ASCS, 6 from news or
radio media, and 1 from a public meeting.
The farm operators' perceptions of their impact
on water quality proved to be positive. Keep in mind
that the project experienced extremely wet condi-
tions in the period from 1984 to 1986 — during the
implementation of BMPs on most acres. Most (76
percent) felt that RCWP increased overall water
quality in Oakwood Lakes or Lake Poinsett. Twenty-
four percent felt that the RCWP did not improve
water quality. Most responses were positive, such as
the 47 percent who said that erosion decreased;
negative responses (16 percent) indicated that more
Table 4.—Question 10: (a) Do you think that RCWP
has increased water quality in Oakwood Lakes
and/or Poinsett? (b) Why?
RESPONSE FREQUENCY PERCENT
(a)
(b)
Yas
No
Decreased erosion
More education needed
Overall, fertilizer and pesticide
management helped
More participation needed to
make Impact
Extreme wetness and flooding
negated effects of RCWP
32
10
15
4
4
4
4
76
24
47
16
13
13
13
operators within the watershed needed to be in
RCWP (Table 4).
Question 11 asked farm operators whether prac-
tices that they specifically applied did or did not help
water quality. Ninety-one percent indicated that the
practices they applied on their farms helped water
quality; 9 percent said the practices did not help.
Interesting responses to questions 12 and 13 are
found in Table 5. Crop scouts provided by SDSU
were generally liked and perceived well by the
respondents. The persons occupying this position
were not the same each year. Some individuals were
not well received and this had an impact on the
benefits of fertilizer and pesticide management, ac-
cording to the respondents.
To better understand how the SCS could im-
prove its technical service to participants, questions
15 and 16 asked farm operators to rank the most and
least beneficial parts of the project. In general, the
SCS did a good job (36 percent), but improvements
could be made by more farm visits, more knowledge
of tillage systems, and better listening skills (See
Table 6). Overall, 95 percent said they were pleased
with the help they received from the SCS; 5 percent
said they were not.
So far, results have shown that farmers "did ac-
cept the program, that they were primarily motivated
by the cost-share incentive, and that 40 of the
respondents continued with conservation tillage
after the program benefits ceased (82 percent).
Table 5.—Question 12: What was the most beneficial part of the program to you? (n = 39)
RESPONSE
The money (cost-share)
Conservation tillage (BMP-9)
Pesticide management (BMP-1 6)
Fertilizer management (BMP-15)
Education
Increased soil quality (less runoff)
Question 3: What was the least beneficial part of the
Pesticide management (BMP-16)
Fertilizer management (BMP-15)
Conservation tillage (BMP-9)
Animal waste system
Cost-share was not enough (participant portion too high)
Restricted crop rotation
FREQUENCY
18
7
6
4
2
1
program to you? (n = 22)
10
5
3
2
1
1
PERCENT
46
18
15
1,0
5
3
45
23
14
9
5
Table 6.—Question 15: How could the Soil Conservation Service have improved technical assistance to you? (n
RESPONSE
FREQUENCY
PERCENTAGE
Soil Conservation Service did a good job
Would like to participate In another program like this
Soil Conservation Service should have helped more on tillage systems
Need more timely visits to farms
Soil Conservation Service needs better listening skills (about values of
crop rotations)
Need more on-farm visits
4
2,
2
1
1
1
36
18
18
9
9
9
258
-------
K. CAMERON-HOWELL
Discussion
Has the project in the Oakwood Lakes and Poinsett
RCWP accomplished its original goals? Farmer ac-
ceptance, as reported, was high. We can conclude
that given agency contacts, media coverage, and
farm visits, the SCS influenced and educated par-
ticipating farmers to change tillage practices per-
manently. Economics was the primary reason why
they chose to participate. Many farmers used the
extra income to help out their farms in general;
others, to purchase specialized conservation tillage
equipment, and in one case, a satellite dish. Never-
theless, the adoption of new conservation practices
are not accomplished in a vacuum.
Concurrent with early (1982-1984) RCWP ef-
forts, the SCS, ASCS, and Cooperative Extension
Services increased their educational efforts by talk-
ing to dealers and universities about alternative til-
lage methods. Ridge-till and one pass soil saver rigs
became popular. Farmers also became increasingly
aware of health and environmental issues. In addi-
tion to these educational sources, a comprehensive
monitoring and evaluation study on the RCWP was
conducted by South Dakota State University Water
Resources Institute, and the Department of Water
and Natural Resources. The study was implemented
on several cooperator plots and on a well-researched
and publicized master site.
Other U.S. Department of Agriculture (USDA)
programs that were concurrent with the RCWP
tended to interfere with practice implementation.
These programs include the Dairy Buy-Out, Conser-
vation Reserve Program (CRP), and Payment in
Kind (PIK). Responses to Question 4 show that of
those who have changed tillage methods since
RCWP, 20 percent enrolled in CRP or a similar pro-
gram. Ten percent changed to better control erosion
and are now using no-till or ridge-till, 20 percent
changed crop rotations to take better advantage of
the USDA Feed Grain program, and 3 percent felt
that the conservation tillage method they had used
did not work well given the excess rainfall in the area
during the mid-1980s. Most of the programs that
USDA sponsored in the 1980s stressed soil erosion
and water quality (for example, the 1985 Food
Security Act). Thus, farmers had a growing aware-
ness of the need for change.
Still, program sign-up patterns in 1982 and 1983
support the adoption diffusion model used by
sociologists (Brandner and Straus, 1959). RCWP was
a unique program because it provided immediate
financial incentives (annual payments). Educational
and cultural efforts made farmers aware of the need
for change, and technology provided the vehicle.
These complementary forces helped farmers to ac-
cept the RCWP and to make permanent changes.
Nowak (1987) points out that all these factors are
necessary if farmers are to adopt new practices. Be-
cause they were in place in the South Dakota RCWP,
the program had a maximum chance for success.
Nowak (1987) also says that "diffusion factors in-
crease in importance as the complexity of the in-
novation increases and decrease in importance as
risk is reduced through institutional support."
Cooperative agency efforts by Cooperative Exten-
sion Service, SCS, and ASCS helped to reduce risk.
In all three counties in the South Dakota RCWP,
sign-ups increased after the project hired a full-time
technical planner. The planner increased one-on-one
farmer contacts. With increased contacts and techni-
cal information, sign ups "snow-balled" as key per-
sons in various townships signed up for the RCWP.
The adoption diffusion model calls these persons
"early adopters." On the other side are persons who
are resistant to change. Once the early innovators
and adopters recognized the problem (erosion and
water quality), they decided whether or not to adopt
conservation practices (Ervin and Ervin, 1982). This
model is complementary to- economic incentives
that led to successful program adoption.
References
Brandner, L. and M. Strauss. 1959. Congruence versus probability
in the diffusion of hybrid sorghum. Rural Sociol. 24(4):381-
83.
Ervin, CA and D.E. Ervin. 1982. Factors affecting the use of soil
conservation practices: hypotheses, evidence and policy im-
plications. Land Econ. 58 (3) :277-92.
Federal Register. 1980. Rural Clean Water Program. (Dec. 21)
7:6202-10.
Hefferman, W.D. 1984. Assumptions of the adoption/diffusion
model and soil conservation. Pages 254-69 in B.C. English et
al. eds. Future Agricultural Technology and Resource Con-
servation. Iowa State Univ. Press. Ames.
Nowak, PJ. and P.P. Korsching. 1983. Social and institutional fac-
tors affecting the adoption and maintenance of agricultural
BMPs. Pages 349-77 in E Schaller and G. Bailey, eds. Agricul-
tural Management and Water Quality. Iowa State Univ. Press.
Ames.
Nowak, PJ. 1987. The adoption of agricultural conservation tech-
nologies: economic and diffusion explanations. Rural Sociol.
52(2):208-20.
U.S. Congress. 1972. Federal Water Pollution Control Act Amend-
ments of 1972. 92d Cong. 2d Sess. PC 92-500, Senate Doc.
52770. Washington, DC.
259
-------
-------
Techniques to Obtain Adequate
Farmer Participation
Richard L. Yankey
Soil Conservation Service
U.S. Department of Agriculture
Twin Falls, Idaho
ABSTRACT
Before 1980, Rock Creek in Twin Falls County, Idaho, had been recognized as one of the most
severly degraded streams in the State. In 1980, the national RCWP coordinating committee
selected Rock Creek as a Rural Clean Water Program (RCWP) project. The projects original goals
were to reduce the amount of sediment entering Rock Creek by 70 percent, phosphorus by 60 per-
cent, total nitrogen by 40 percent, pesticides by 65 percent, and fecal coliform by 70 percent. These
project goals could only be obtained through the adequate participation of Rock Creek farmers. An
Information and Education (I&E) committee was formed to develop an I&E program. The key ele-
ments identified for obtaining farmer participation included actively involving local Soil Conserva-
tion District supervisors and Agricultural Stabilization and Conservation Service county committee
members; identifying and targeting local farm leaders; meeting with farmers to gain information on
best management practice implementation; and holding yearly update meetings with landowners
and the media. By conducting a well-planned I&E program, the Rock Creek RCWP project was able
to meet its project goals.
To obtain adequate farmer participation, a
well-planned information and education
(I&E) program must be developed at the
beginning of a water quality project with input from
all participating agencies and organizations. But the
I&E program should also be continued during the
project's life span to demonstrate progress and
publicize success. Above all, a good program will re-
quire input from local farmers. Once developed, the
I&E program should be directed by someone with a
thorough knowledge of the water quality project.
The Rock Creek Rural Clean Water Program
(RCWP) project was able to meet its contracting goal
because these premises were followed.
Project Description
Rock Creek in Twin Falls County, Idaho, had long
been recognized as one of the most severely
degraded streams in the State. Point source pol-
lutants had been virtually eliminated by the end of
the 1970s, yet Rock Creek still carried high levels of
sediment. In 1980, the national RCWP coordinating
committee selected Rock Creek as an RCWP project.
It was also selected for comprehensive water quality
monitoring and economic evaluation.
The Rock Creek watershed covers a total of
198,400 acres (Fig. 1), of which 51,900 acres are ir-
rigated cropland and pasture. Approximately 28,000
acres of irrigated cropland were defined as critical
acres because they contributed excessive amounts
of sediment to Rock Creek. The original goals were
to reduce the amount of sediment entering Rock
Creek by 70 percent, phosphorus by 60 percent, total
nitrogen by 40 percent, pesticides by 65 percent, and
fecal coliform by 70 percent. Best management prac-
tices (BMPs) used on the Rock Creek Project in-
cluded
• conservation tillage,
• sediment-retention structures,
261
-------
Proceedings of National RCWP Symposium, 1992
Figure 1.—Map of the Rock Creek RCWP project In Twin Falls County, Idaho.
• irrigation improvements,
• irrigation water management, and
• animal waste control systems.
Information and Education
When the project began, an I&E committee was
formed (see Fig. 2). Representatives from the
Agricultural Stabilization and Conservation Service
(ASCS), Soil Conservation Service (SCS), Coopera-
tive Extension Service (CES), Idaho Division of En-
vironmental Quality, and the Snake River and Twin
Falls Soil and Water Conservation Districts par-
ticipated on the committee. The Snake River Soil and
Water Conservation District became the lead I&E
group. In 1985, it was able to hire a full-time I&E
specialist, who coordinated the I&E program and im-
plemented many I&E activities.
The I&E program consisted of a mixture of news
releases, newsletters, flyers, brochures, slide presen-
tations, informational meetings, demonstrations,
262
-------
R.L. YANKEY
NATIONAL RCWP
COORDINATING COMMITTEE
ROCK CREEK RCWP
STATE COORDINATING COMMITTEE
ROCK CREEK RCWP
LOCAL COORDINATING COMMITTEE
ROCK CREEK RCWP
TECHNICAL COORDINATION
COMMITTEE
ROCK CREEK RCWP
INFORMATION AND
EDUCATION COMMITTEE
MEMBERS
Twin Falls County Extension Service
Agricultural Stabilization and
Conservation Service
Soil Conservation Service
Snake River Soil Conservation District
Twin Falls Soil Conservation District
Division of Environmental Quality
Figure 2.—Flow chart of the Rock Creek RCWP project committees.
tours, posters, displays, public service an-
nouncements on radio and television, and awards
programs. The Rock Creek RCWP Final Report con-
tains a complete bibliography. (Copies of the report
may be requested by contacting the author; a section
of it is appended in this report.) Six hundred and
eighteen separate reports, newsletters, and publica-
tion are cited. Evaluation of the I&E program after
the end of the contracting period identified the fol-
lowing elements as the most valuable in obtaining
farmer participation:
• actively involving local Soil Conservation Dis-
trict (SCD) supervisors and ASCS county
committee members; ' , -
• identifying and contacting local farm leaders;
• meeting with farmers to gain input on how to
implement and manage BMPs; and
• meeting with landowners, and the media to
provide a yearly update on water quality
monitoring and BMP implementation.
The active involvement and leadership of SCD
supervisors and ASCS county committee members
was effective in gaining the trust of local farmers.
SCD supervisors spent many hours visiting farmers,
to explain the RCWP project. The supervisors' par-
ticipation made farmers receptive to the suggestions
and ideas presented by soil conservationists. After
water quality plans were jointly developed by, the
farmer and soil conservationists, they were reviewed
and approved by SCD supervisors. Tlie ASCS county
263
-------
Proceedings of National RCWP Symposium, 1992
committee members met with each farmer before
the contract was signed to provide cost-share funds
for implementing the water quality plan. This
process helped ensure that the farmers thoroughly
understood the project rules and procedures.
To help get the RCWP project off to a good start,
farm leaders were identified and contacted first.
When the program had been thoroughly explained
to them, they then helped promote it among other
farmers. This technique was very successful. During
the project's contracting period, the ASCS office al-
ways had a waiting list of farmers requesting con-
tracts.
Many BMPs approved for use in the Rock Creek
RCWP project were new practices for Rock Creek
farmers. The BMPs had been developed, re-
searched, and tested by the University of Idaho and
U.S. Department of Agriculture's Agricultural Re-
search Service. They had not, however, been pre-
viously implemented by area farmers. Many farmers
initially believed that BMPs would not be compatible
with their farming operations. Small group meetings
were held with farmers after they initially imple-
mented BMPs. At these meetings, farmers shared
information on how BMPs could best be installed
and managed. The information benefited the farmers
in attendance, and soil conservationists later shared
it with new applicants. These meetings were suc-
cessful because they consisted of farmers telling
other farmers how to implement BMPs.
Keeping farmers and the public informed on the
progress and achievements of the RCWP project was
very important. During the contracting period, a
meeting was held every winter. All Rock Creek
farmers and the local media were invited. These
meetings were well attended. They served not only
to provide water quality and land treatment data but
also to build enthusiasm and pride for the Rock
Creek Project.
It is also important to note that not only was the
contracting goal of 75 percent of the critical area
achieved, but very few farmers cancelled contracts
after they were signed. The key elements noted pre-
viously were effective in obtaining trust, participa-
tion, understanding, commitment, and pride for the
Rock Creek Rural Clean Water Program. This, in
turn, resulted in the achievement of the water quality
goals established for the project
Reference
Yankey R. et al. 1991. Rock Creek RCWP Final Report. Soil Con-
serv. Serv., U.S. Dep. Agric. and Idaho Div. Environ. Qual.,
Boise.
Appendix
from Rock Creek RCWP Final Report (Yankey et al. 1991)
4.1 FINDINGS AND RECOMMENDATIONS
• Information and education activities for the Rock Creek Project were invaluable aids in
explaining the Project to landowners, operators, and the general public and in enlisting
their involvement.
• Through the use of news releases, newsletters, flyers, brochures, slide presentations,
informational meetings, demonstrations and tours, posters, public service announcements
on radio and television, awards programs for participating farmers, cooperation and
support was gained from Project participants and non-participants alike.
• Information and education activities should be initiated at the project's inception to
explain its concept, continue throughout the project to demonstrate its progress, and
proceed to the project's completion to emphasize its success.
• An information and education specialist should be hired at the outset of the project in
order to facilitate information and education activities.
• Personal contact between SCD supervisors and landowners is an extremely valuable
information and education activity.
• A mixture of media is essential in information and education activities since different
types of communications reach different people.
• The need for information and education exists at two levels: the public's level and the
farmer's level. Information and education for the public should be provided by a
communication expert, while the farmers should receive information and education from
those who are more technically oriented.
• Someone with a deep knowledge of the project should organize communication efforts
so that a coordinated I & E message is disseminated.
• Major emphasis should'be placed on educating the public about water quality benefits
derived from the Project. The public was interested in and eager to learn about those
benefits of the Rock Creek Project.
• Monitoring results are helpful in an information and education program. Such data
reassures the public and farmers alike that BMPs are producing the desired results.
264
-------
Farmer Participation in the
Double Pipe Creek, Maryland,
Rural Clean Water Program Project
Elizabeth A. Schaeffer
Agricultural Stabilization and Conservation Service
U.S. Department of Agriculture
Westminster, Maryland ' '
ABSTRACT
Initiated in 1980, the Double Pipe Creek Rural Clean Water Program project was established to pro-
vide farmers with cost-share assistance as an incentive to voluntarily implement best management
practices. The primary objective of the project was to improve water quality and reduce sedimenta-
tion in the Double Pipe Creek Watershed. Farmer participation was affected by economic condi-
tions, current government program requirements, and farmer-agency relationships. By the
project's conclusion, farmers will have been awarded over $3 million in cost-share funds.
The Double Pipe Creek Rural Clean Water
Program (RCWP) project is located in Car-
roll County, Maryland, within the multi-
county Monocacy River basin that runs from
Pennsylvania south to the Potomac River above
Washington, D.C. Historically, dairy and cash grain
farming have predominated in Carroll County. Other
crops in this piedmont area include corn, soybeans,
small grain, and hay. Farm size averages 136 acres.
At the time of the project, livestock on participants'
farms ranged from 400 dairy cows to 5 beef cows.
Water quality plans covering 20,273 acres had
been written and approved for 149 contracts by the
end of the contracting period in 1986. Two priority
critical areas were established:
1. farms where livestock and animal waste
management and severe gully erosion
presented water quality problems and
2. farms where sheet and rill erosion produced
control problems.
The local coordinating committee coordinated
activities of the participating agencies — the Agricul-
tural Stabilization and Conservation Service (ASCS),
which provided administrative assistance; the Soil
Conservation Service (SCS), which gave technical
assistance; and the county Cooperative Extension
Service, which acted as the primary agency in lead-
ing the information and education efforts. Water
quality monitoring was conducted by the Maryland
Department of the Environment.
Farmer Participation
Approach Taken
Program participation was encouraged through
project tours, government literature, and articles in
local newspapers. An SCS field office soil conser-
vationist provided general information about project
benefits and, following a request for a contract,
worked with the applicant to develop a water quality
265
-------
Proceedings of National RCWP Symposium, 1992
plan outlining the best management practices
(BMPs) to be installed or carried out and the es-
timated cost-share funds needed. The ASCS office
maintained contract information and issued pay-
ments.
The BMPs installed included:
• permanent vegetative cover,
• animal waste control systems,
• strip-cropping systems,
• diversions,
• grazing land protections,
• grassed waterways,
• cropland protective cover,
• conservation tillage systems,
• stream protection systems,
• permanent water control structures, and
• fertilizer and pesticide management
The topography of the watershed and the need
for improved animal waste containment made animal
waste control systems (BMP-2) and grassed water-
ways (BMP-7) the most heavily used practices, ac-
counting for 85 percent of the cost-share funds
earned — over $2 million.
The Double Pipe Creek Watershed now serves
as a demonstration of how water quality goals can be
accomplished through voluntary adoption. Many of
the practices would not have been installed without
the incentive provided by the cost-share program.
The cost-share limitation of $50,000 gave the SCS the
flexibility to design innovative practices large
enough to meet participants' needs. These BMPs
now serve as examples of installations that meet
water quality goals. Farmers from within and outside
the targeted area have visited participant farms to
view projects completed with RCWP funds.
Since 1988, Carroll County has been the focus of
a gamut of water quality projects that have resulted
from successful farmer participation in the Double
Pipe Creek Watershed project. The Piney/Alloway
Creek Watershed, which borders Double Pipe Creek
to the north, has been targeted for an ASCS Special
Water Quality Project, State of Maryland-Targeted
Watershed Project, and a Federal Demonstration
Project. The justification for these activities focused
on local agencies' ability to successfully implement
these programs and farmers' willingness to par-
ticipate in demonstrations, as illustrated by the suc-
cess achieved through the RCWP project.
Factors Affecting Participation
The farmers in the project area were very receptive
to the Rural Clean Water Program. The ASCS, SCS,
and the county Cooperative Extension Service had
developed an excellent working rapport with local
farmers before the start of the project, which made
farmer recruitment an easy task. County farmers
were enthusiastic about the opportunity to install in-
novative BMPs with a more liberal payment limita-
tion than traditional Federal programs had afforded.
However, during the 1980s, many outside factors
affected the RCWP project area, including a three-
year drought that placed a severe financial hardship
on the project participants and problems within the
dairy industry, which is the most prevalent type of
farming in Carroll County. During the 1980s, 33 per-
cent of the dairy producers in the county went out of
business. Milk prices fluctuated from $12.90/cwt in
1986 to $14.90/cwt in 1990, which produced uncer-
tainties about market conditions that negatively af-
fected dairy operators' willingness to spend large
amounts of money to construct conservation BMPs.
State and Federal Programs
Several State and Federal programs have been im-
plemented during the RCWP project. Those pro-
grams that affected the project area are described in
the following paragraphs.
• Maryland Agricultural Land Preservation
Program. For a farm to be accepted in the
Maryland Agricultural Land Preservation Founda-
tion, a conservation plan must be developed when
easement rights are sold. Since 1980, 21,473 acres
have been accepted into this program in the Double
Pipe Creek Project area.' By enrolling in the Agricul-
tural Land Program and making a financial commit-
ment to water quality through RCWP, many farmers
have become committed to remaining in farming.
• Food Security Act of 1985. Approximately 75
percent of the land in Carroll County is highly
erodible, according to 1985 Food Security Act
criteria. The SCS's district conservationist has es-
timated that 25 percent of the cropland acreage will
need cropping system modification or additional con-
servation measures to bring the soil loss each year
within the established acceptable level. These
changes should affect the project area's water
quality positively by reducing soil particles, animal
waste, commercial fertilizer nutrients, and pesticides
in the water.
• Wheat and Feed Grain Program. Since 1986,
the 1985 Food and Security Act Wheat and Feed
266
-------
E.A. SCHAEFFER
Grain Program has become a major influence in the
project area. Each year, enrollment has fluctuated
because of changing markets, prices, and producer
needs; however, the average county enrollment has
been 200 farms, with 140 in the project area. Ap-
proximately 4,200 acres have been designated as an
acreage conservation reserve (ACR). Farm oper-
ators must use control measures when needed for
erosion, insects, and weeds on their ACR land. In
general, ACR land is on hillsides or other areas
where erosion control is important. The Wheat and
Feed Grain Program through ACR is decreasing
erosion and preventing animal waste and pollutants
from reaching streams in the RCWP area.
• Conservation Reserve Program. Carroll Coun-
ty has limited participation in the Conservation
Reserve Program because of high land values; how-
ever, 75 percent of the land enrolled (approximately
700 acres) is in the project area.
Time Line of Events
This summary time line of the Double -Pipe Creek
Project is not all-inclusive but highlights major
events. Information for this summary was extracted
from local coordinating committee minutes for the
last 10 years.
• 1979 — Proposal for Double Pipe Creek
Project submitted
• 1980 — Proposal approved for Double Pipe
Creek Project
— Local coordinating committee
determined that any farm in the proj-
ect area that critically affected water
quality, regardless of the distance to
streams, would be considered a high
priority request
— Cost-share earned: $8,953
—•. Contracts approved: 5
• 1981 — Cost-share earned: $176,518
— Contracts approved: 25
• 1982 — Cost-share earned: $283,642
— Contracts approved: 25
• 1983 — Versar, Inc., completed baseline data,
Stage 1 of monitoring
— Cost-share earned: $316,515
— Contracts approved: 18
1984 — Law passed requiring that a
conservation plan accompany ease-
ment rights bid to enter Maryland
Agricultural Land Preservation Pro-
gram
— State agreed to continue monitoring
when Versar contract expired
— Demonstration Day held (85 people
attended)
— Cost-share earned: $187,815
— Contracts approved: 21
1985 — Fertilizer and pesticide management
added to the project
— Contracting period extended through
December 31,1986
— Cost-share earned: $502,836
— Contracts approved: 19
1986 — Cooperative Extension agent hired to
write nutrient management plans and
carry out BMP-15 and BMP-16
— Cost-share earned: $382,125
1987 — Cost-share:earned: $489,773
1988 — Best Management Progress Tour to
demonstrate progress and practices
completed (150 attended)
— Cost-share earned: $236,014
— Contracts approved: 1
1989 — Cost-share earned: $168,033
1990 — Cost-share earned: $136,464
Results
By the end of the sign-up period in 1986, contracts
had been completed for 149 farms, all located in the
critical area as defined by the local coordinating
committee. Ninety-four percent of the cost-share
monies used for BMP application was spent for
animal waste control systems (BMP-2), grazing land
protection (BMP-6), and grassed waterways (BMP-
7), the primary practices installed. These BMPs
were c6mpleted in 80 percent of the contracts. Of the
60 percent of landowners with active 'contracts who
completed animal waste control systems, 74 percent
installed one or more grassed waterways and 44 per-
cent installed grazing land protection BMPs. A total
267
-------
Proceedings of National RCWP Symposium, 1992
of 203 animal waste control systems (BMP-2) were
recorded, including repairs, modifications, and addi-
tions to existing practices. Over the 10 years,'221,616
feet of waterway were cost-shared and 110 grazing
land protection systems installed.
In 1990, the local coordinating committee con-
ducted a survey to evaluate farmers' perceptions.of
the project. Thirty percent of the questionnaires
were returned, and many responses reflected
producers' positive feelings toward the project.
"Tlie BMPs installed on the farm were very
influential in preventing erosion, runoff, and
other problems we were having."
"TJiere were no weaknesses. Strengths — it
stopped the erosion, which is what it was sup-
posed to do; the ASCS people we worked with
are very good."
"I think the field manpower was good, getting
through the practices smoothly was a .real
plus. Having plenty of time (extensions) also
is a real plus. I never felt pushed at any time."
"The major strengths were adequate cost-shar-
ing and flexibility of design, with my wishes '
and ideas being incorporated."
"Tlte only way to improve would be to con-
tinue the project. Tliere are probably still a lot
of problem areas around the county."
Lessons Learned
The Double Pipe Creek RCWP truly served as a
demonstration project. The local coordinating com-
mittee compiled a list of conclusions, some of which
follow:
• Implementation is correlated to farm income
and farm income is difficult to predict over 10
to 15 years.
• Contractors Who are not only available but
als^o have equipment and experience are a
necessity. (Contractors were not a problem for
this project.)
• Because of the experimental nature of the
. "project, a wide range of innovative options
.were planned and installed, particularly
through BMP-2.
• The flexibility to try new ideas during practice
application and the $50,000 payment limitation
was a great asset.
• .ASCS's administration of the project was a
" great advantage because the agency already
had experience in administering cost-sharing
programs.
• Landowners were willing to' try new ideas
after ASCS staff explained that, through the
. project, their contract could be modified to
alter the BMP if the original facility did not
, provide expected results.
' • The water quality improvements placed on the
' land and subsequent publicity greatly im-
*''"" •""' proved the agricultural community's image in
the eyes of the public and county government.
By offering voluntary incentives to achieve
farmer participation and exhibiting a positive,
cooperative approach, the agencies involved at the
local level of the Double Pipe Creek RCWP project
successfully administered $2,680,366 in cost-share
funds through 1990. •••'•<••
268
-------
The Key to Successful Farmer
Participation in Florida's
Rural Clean Water Program
John W. Stanley
Agricultural Stabilization and Conservation Service
U.S. Department of Agriculture
Okeechobee, Florida
ABSTRACT
The high percentage of farmer participation in Florida's Rural Clean "Water Program resulted from
several factors, many not controlled by direct contact with the participants. Different strategies
were used to arouse interest. A farmers' meeting held to inform potential participants of the pro-
gram and encourage them to contract had limited attendance. The media was used as well as local
agency newsletters and small group meetings—again with limited results. The most successful
strategy was one-on-one contact with potential participants. An external factor in the process was
what the local regulatory agency called the "carrot and stick approach," the carrot being the Rural
Clean Water Program and the stick, a regulatory program. In addition, in the 1960s, most of the
dairies in the project area had sold land elsewhere at a profit and moved to Okeechobee. In the
1990s, these dairies would not gain the same financial rewards by moving away from regulation,
The Taylor Creek-Nubbin Slough Rural Clean
Water Program (RCWP) project exceeded
its goals for farmer participation and im-
plementation of best management practices (BMPs).
Two factors influenced this success: things the
project could control and those it could not. This
paper, based on the Taylor Creek-Nubbin Slough 10-
year Report (Stanley and Gunsalus, 1991), reviews
both factors and draws some conclusions about how
each influenced participation and how the lessons
learned here could aid other projects.
Background
The Taylor Creek-Nubbin Slough RCWP defined
63,109 acres as critical and 59 farms that needed
treatment. The project used the following definitions
for determining the critical areas:
• All dairy farms in the project are considered
critical areas.
• All beef cattle farms that have been
extensively drained are considered
• critical areas.
• All areas within one quarter mile on each
side of a stream, ditch, or channel that hold
water year-round are considered critical.
After five years of contracting, 48 farms (Fig. 1)
and 54,709 critical acres (Fig. 2) were under con-
tract, representing 89 percent of the total critical
acres contracted. An additional 5,267 acres on eight
farms, originally considered critical acres, were
determined not to have a water quality problem
when surveyed by the Soil Conservation Service.
This left 3,133 acres (5 percent) on three farms that
needed treatment but were not contracted. The total
acres contracted far exceeded the goal set by the
RCWP project (75 percent of the critical acres). Fig-
ure 1 shows the progression of contracts signed at
the end of each year. Projects had five years from
their inception to contract farms and 10 years to
complete implementation of BMPs.
2619
-------
Proceedings of National RCWP Symposium, 1992
60
50 -
40 -
1
B
•3
t-4
JJ
30 -
20 -
10 -
1982 1983 1984 1985 1986 1987 1988 1989 1990
Years 1982 -1990
Figure 1.—Active RCWP contracts In the Taylor Creek-Nubbin Slough basin.
60
50
40
30
5 20
10
0
I
54709 54709 54709 54709 54709
49532
1982 1983 1984
Figure 2.—Contracted acres for the Taylor Creek-Nubbin Slough basin
1985 1986 1987
Years 1982 - 1990
1988 1989 1990
270
-------
J.W.STANLEY
Ninety-nine percent of the practices scheduled
were completed (Table 1), which is important be-
cause it shows that the farmers were not only willing
to sign a contract, but they were committed to help-
ing solve an environmental problem. Many of the
subwatersheds had a substantial number of signed
contracts and acres treated early in the project.
Table 1 shows that Little Bimini, Taylor Creek Main,
and Henry Creek subwatersheds (all predominantly
dairy) had most implementations completed by
1986. Williamson Ditch, a predominately beef cattle
subwatershed, had 100 percent implementation by
1986. Table 1 and Figure 2 for 1987 show that most
of the critical acres in all subwatersheds were under
contract and most of the project's BMPs had been
implemented by this time. This progress enabled
early trend tests to detect BMP effectiveness.
Farmer Participation
Controlled Factors
Defining the area should have been the first step in
getting participation. Initially in the Taylor Creek-
Nubbin Slough RCWP proposal, the critical area was
defined as the total Taylor Creek-Nubbin Slough
basin, which was 110,000 acres and approximately
150 farms. In 1982, the local coordinating committee
reduced the critical acres after a 1-RCWP handbook
amendment clarified the definition of critical area to
be "that area that needs treatment"—not the total
basin, as originally defined. The committee reviewed
aerial photographs and reduced critical land area by
15 percent to 93,500 acres.
In 1983, the committee established a final critical
area definition after initial contract planning showed
that not all acres were critical. Using the three
criteria mentioned earlier and deleting urban areas
that fell within the project, the committee deter-
mined 63,109 acres with 59 farms were critical areas
that heeded treatment. These important steps
defined the potential participants and kept the
project focused on the most critical farms. Neverthe-
less, time should have been taken in the initial stages
to better define critical areas. However, as it became
clear that changes were needed, the project made
them.
Farmer meetings were held to encourage par-
ticipation but with limited success. In these meet-
ings, the purpose and the goals of the project were
stated, and each agency involved explained its role.
Articles about RCWP were placed in news media and
agency newsletters, but they were of limited success
in getting participants to sign up for the program.
However, these communications reinforced the ini-
tial meetings.
One-on-one contact with potential participants,
was the most successful, overall. Meeting people
face-to-face and listening to their individual problems
and concerns made the process more personal. This
success was augmented by the structure and opera-
tion of the administrative subcommittee, whose
members visited individual farms to explain the pro-
gram and encourage participation.
The administrative subcommittee consisted of
the Agricultural Stabilization and Conservation Ser-
vice county executive director, Soil Conservation
Service district conservationist, County Cooperative
Extension water quality specialist, and a repre-
sentative from the South Florida Water Management
District. This group ran the day-to-day operation of
the project.
Each subcommittee member understood the
role and goals of the project and each other; there-
fore, a potential participant could talk to any one of
these four agency representatives to get information
about the farmer's role and that of each agency. Sub-
committee members either individually or as groups
visited the farms that had not signed initially. The
project's goal was to identify and visit all potential
participants.
Dairies in this project were an easier sell be-
cause the high animal densities on their farms ap-
peared to be the most obvious problem (and because
of uncontrollable factors that will be discussed later).
Beef operators, with fewer livestock, did not think
they had a problem. The subcommittee used two ap-
Table 1.—Percent implementation of best management practices by subwatershed by year.
SUBWATERSHED 1981
NW Taylor Creek
Little Bimini
Otter Creek
Taylor Creek Main
Williamson Ditch
Mosquito Creek
Nubbin Slough
Henry Creek
Lettuce Creek
1992 1983
2%
31% 36%
10% 35%
1%
8%
. 60%
1984
43%
48%
50%
6%
25%
14%
60%
1985
56%
87%
75%
57%
80%
20%
39%
94%
5%
1986
69%
91%
77%
90%
100%
50%
81%
98%
10%
1987
100%
95%
97%
99%
100%
77%
97%
100%
77%
1988
100%
99%
99%
100%
100%
99%
100%
100%
100%
1989
100%
99%
99%
100%
100%
100%
100%
100%
100%
1990
100%
100%
99%
100%
100%
100%
100%
100%
100%
271
-------
Proceedings of National RCWP Symposium, 1992
preaches to convince beef operators to sign: it
pointed out that, collectively, beef operators had as
many animals as dairies and that cattle also sought
(and polluted) water. The subcommittee noted that,
if beef operators participated, they could help prove
their animals were not a problem.
To try to obtain maximum participation, the sub-
committee allowed farmers to select the order in
which the BMPs would be installed. Most par-
ticipants first chose to install practices that they
thought might bring financial or other benefits and
not necessarily those that would have the greatest
effect on water quality. In retrospect, this flexibility
on BMP installation was a poor decision because
BMPs were not implemented first on the most criti-
cal acres.
Many factors entered into the project's success,
but the fact is that most of the farmers understood
the need to do their part in preserving the environ-
ment and therefore completed 99 percent of their
scheduled work.
Uncontrollable Factors
During the contracting period, Florida was entering
into a regulatory phase that defined agriculture's
role in protecting State resources. Part of the reason
these farms signed RCWP contracts was the threat
of regulation. The local regulator agency called this
the "carrot and stick" approach, the carrot being the
RCWP program and the stick, regulation.
The South Florida Water Management District
began to assess water quality along with water quan-
tity. Specifically, the District was looking at ways to
regulate the quality of surface water runoff. The
Taylor Creek-Nubbin Slough basin was the area of
greatest concern and, therefore, the District's first
priority. The District implied that people who did not
participate in the RCWP project would have to get a
surface water management permit Because of
public pressure, this threat became more of a factor
toward the end of the sign-up period. Now, after the
end of the contracting period (1981 to 1986), surface
water management permits are required on most
farms in the area, including RCWP farms.
The farm economy also played a role in par-
ticipation. In 1983, milk prices dropped considerably,
keeping some dairies from signing and causing
some to cancel their request for contracts. In 1985,
prices improved and all the dairies were signed by
the end of the contracting period. Depressed beef
prices also slowed beef operators' participation.
In the 1960s, many dairies had moved to
Okeechobee from south Florida (primarily from
Dade and Broward counties) because, as the popula-
tion grew up around these dairies, their holdings had
become valuable development land. In south Florida,
dairies sold out at inflated prices, moved to
Okeechobee, and paid cash for land. However, the
opportunity to sell this land for development at in-
flated prices was not available in Okeechobee in the
1980s, so dairies could not as easily move away from
regulation. This situation made the RCWP project
more attractive to dairy farmers.
Summary
Individual contact was the most successful method
in getting farmers to participate. When agency rep-
resentatives sat down with a potential participant to
explain how RCWP would affect an operation and
enumerate benefits that could be realized, that per-
son was more likely to sign on to the project. Farmer
meetings, news media, and other resources were im-
portant in laying the foundation for individual con-
tacts and keeping the project visible. Also important
were the people making contacts: they had to be able
to explain the program and the roles of participants
and agencies.
As much as this project's personnel would like to
claim credit, external factors played a large part in
successful participation. Threat of regulation, the
economy, and a changing real estate climate had an
effect. However, project personnel used these uncon-
trollable factors to achieve the project's goals.
Findings and Recommendations
• Farmer meetings were good places to explain the
project but few contracts were signed at meetings.
• One-on-one contact had the most success in get-
ting participants to sign contracts.
• The individual key players must understand the
role of each agency, so that participants can talk to
one person to get initial information.
• The project area and potential participants must
be defined or redefined to keep the project focused
on the most critical areas.
• To get participation, the subcommittee allowed
flexibility in selecting the order in which practices
would be installed. In retrospect, this had little effect
on participation.Therefore, BMPs giving the great-
est benefit to water quality should be addressed first.
Reference
Stanley, Ł and B. Gunsalus. 1991. Taylor Creek-Nubbin Slough
RCWP 10-year Report. Okeechobee Agric. Stabil. Conserv.
County Off., Okeechobee, FL.
272
-------
Document It! Procedures for the
Documentation of Nonpoint Source
Project Data — Land Treatment
Betty Hermsmeyer
U.S. Department of Agriculture
Agricultural Stabilization and Conservation Service
Long Pine Creek RCWP Project
Ainsworth, Nebraska
ABSTRACT
The Rural Clean Water Program (RCWP) was established as a 10- to 15-year experimental program
offering cost-sharing and technical assistance to farm operators as incentives for implementing
best management practices (BMPs) to control agricultural nonpoint source pollution. The RCWP
emphasized interagency cooperation, and as many as 15 agencies were involved in some capacity
at various stages of project development. To provide project continuity and establish a base of con-
sistent information, procedures for documenting economic and technical data must be established
and assigned. This paper defines procedures for identifying and documenting critical acres, acres
under contract, project development, and subbasins on a master aerial photograph of the water-
shed and a master set of. aerial section map photographs. It also provides methods to record
economic and technical project information on a standard nonpoint source pollution data log for-
mat, addresses documentation of nonproject land treatment data, and discusses problems with the
use of acres served data from CRES forms. It recommends that agencies adopt a standard data log
format for all nonpoint source projects involving BMPs, determine standard methods for reporting
acres served or benefited for each individual practice, and adopt procedures for documenting other
project and nonproject land treatment data.
The Long Pink Creek watershed is located in
north central Nebraska on the northeastern
edge of the Nebraska Sandhills, the largest
grass-covered sand dune area in the world. In the
mid-1960s, irrigation was developed on 35,000 acres
within the watershed. This sudden change from
grasslands to intensive row crop production had a
significant environmental effect on the area. Exces-
sive sediment from agricultural runoff and erosion
filled creeks and streams and negatively affected
aquatic life. Agricultural chemicals were found to be
leaching into the groundwater. In 1981, the Long
Pine Creek watershed was one of 21 watersheds in
the United States selected for the experimental Rural
Clean Water Program (RCWP). The program em-
phasized interagency cooperation. A total of 15 State
and Federal agencies made significant contributions
to the project.
The Nebraska RCWP project addressed both
land and stream treatment. Fifteen best manage-
ment practices (BMPs) were selected to address
nonpoint source pollution problems on 60,242 acres
identified as critical. A wide variety of land treatment
practices were employed throughout the watershed,
including
• water management systems,
• water and erosion control structures,
• grazing land protection systems,
273
-------
Proceedings of national RCWP Symposium, 1992
• permanent vegetative cover,
• stream protection systems, and
• fertilizer and pesticide management.
Although plans for water quality monitoring
were extensive, the documentation of land treatment
data to be linked to water quality monitoring data
was not fully realized. Land treatment data is com-
plex and includes many variables. To link water
quality monitoring to BMP effectiveness, significant
land treatment documentation by hydrologic unit
must take place. The Nebraska Department of En-
vironmental Control (NDEC) began a three-year
postmonitoring project in the Long Pine Creek
watershed in spring 1992. A significant amount of re-
search is necessary to provide the NDEC with the
land treatment documentation it will need to link
BMP installation to any changes in water quality.
The following recommendations for documentation
of project and nonproject land treatment data have
evolved out of the Nebraska RCWP experience. This
paper addresses procedures for identifying and
documenting critical areas, subbasins, and contract
acres and for collecting land treatment data.
Identification and
Documentation Procedures
At the beginning of each project, the project area,
critical area, and hydrologic units (subbasins) must
be identified and documented. Data log formats
must be created to record all project data, including
cost-share and non-cost-share practices, technical
and economic data. Technical personnel must decide
what other land treatment data will be needed in ac-
cordance with each project area's agricultural and
urban activities. Formats to document procedures to
collect this land treatment data must then be estab-
lished and data log formats distributed to facilitate
the process.
Step 1
Order a large (four or eight inches to a mile) aerial
photograph of the watershed and surrounding area
from the Cartographic Unit of the Soil Conservation
Service (SCS).This map is an invaluable visual aide
in determining critical areas. The SCS and the
State's Department of Environmental Control, along
with other agencies with technical expertise, should
work together to determine and delineate the follow-
ing:
• The Project Area. The general watershed
area.
• The Hydrologic Units — (Subbasin Lines).
Use the large aerial photograph in conjunction
with topography maps to determine hydro-
logic units that are unlikely to change because
they are based on drainage. In the future,
post-monitoring may be a part of all projects;
therefore, the agencies involved in monitoring
should agree with other technical agencies on
the locations of drainage lines. Delineate the
subbasins on the large aerial photograph.
• The Critical Area Targeted for Treatment
For surface water projects, use topography
maps to determine the acres contributing to
surface runoff and pollutant delivery within
each subbasin. These lines will be delineated
. on the large aerial photograph of the water-
shed and can be used to determine critical
areas. Critical areas will be adjusted according
to each project's variables, which can include
amount of cropland, cropping practices,
' animal unit density, erosion rates, proximity to
• streams and lakes, and percentages of'cover.
Delineate the critical area(s)-on the large
aerial photograph.
Critical areas are likely to change 'as the project
evolves. If changes occur, document them. Include a
detailed description and rationale for each change.
Step 2
Assign responsibility for documentation. Someone
such as the project coordinator ,or Agricultural
Stabilization and Conservation Service (ASCS) pro-
gram clerk must establish and maintain permanent
documentation of critical areas, subbasins, acres
under contract, and the like. ASCS, which is involved
in RCWP projects as the administrative body and as;
head of the local coordinating committee, should
provide a full-time person to establish and maintain
RCWP maps during the contracting period.
After the project area, subbasins, and critical
areas have been determined and delineated on the
large aerial photograph, ther ASCS RCWP clerk will.
• pull small aerial photographs (section, legal,
or subbasin maps) of all sections within the
critical area to form the permanent RCWP file
(these photographs are identified .by legal
descriptions); ' ••••••.•
• stamp "RCWP MAP" on each photograph in
the upper left-hand corner; , ^ .;
• delineate the critical area on the small aerial
section photograph and write the number of
274
-------
B. HERMSMEYER
acres in the critical area in the upper left- hand
corner;
• delineate (if necessary) the subbasin areas
and indicate subbasin identification and acres
in the upper left-hand corner of the map or
subbasin area; and
• assign each field a "Farm Field ID" number to
link photographic information to the geo-
graphical information system and other
databases.
The large aerial photograph of the watershed
and the small aerial section map photographs are
now the permanent record of the project. As acres
are put under contract, the ASCS clerk will
• delineate the acres under contract and write
the contract number, year, and name (ex-
ample: RCWP-25-1992, Bob Smith) on the
large aerial photograph, and
• pull the small aerial section map photo-
graph (s) and delineate the acres under con-
tract with the same notation as the large aerial
photograph.
If acres are added to or taken out of the critical
area, changes should first be drawn on the large
aerial photograph with a notation (example: added
[or deleted] 1993) and then documented on the
small aerial photographs.
Note any activity within the critical area first on
the large aerial photograph and then on the small
aerial photograph. This method is consistent with
ASCS procedures for making changes. When agen-
cies need photographs, make a copy of the small
aerial photographs with current information. At the
end of the project, all of the activity regarding critical
acres, acres under contract, years, producers, and
subbasins will be documented on the maps.
As practices are completed and documented in
the proposed standard nonpoint source data log, the
ASCS program assistant should note the designation
of each practice and the site on both the large and
small aerial photographs. This procedure is consis-
tent with ASCS procedures for documenting Agricul-
tural Conservation Program (ACP) practices on
large and small aerial photographs. These notations
can be extremely valuable as reference points and
visual aides in determining implementation progress
through project development.
Establishing Data Log Formats
The data log records for future evaluation all per-
tinent data for each practice that is installed or imple-
mented. Although much attention has been given to
water quality monitoring data, in many instances,
land treatment data have not been sufficiently ad-
dressed. Land treatment data are complex: in addi-
tion to data on an immense array of agricultural
activities, including dairies, feedlots, and livestock,
feed grain, hay, food, and fiber production, these
data include information about participants and non-
participants, cost-share and non-cost-share, manage-
ment, and cropping practices. Data logs and
procedures must be established to provide consis-
tent and uniform reporting of this technical and
economic information, which must cover two basic
areas:
1. documentation of project data and
2. documentation of other data deemed
essential for evaluation or future use.
In all nonpoint source, projects, project data
should be reported using a standard format. The
BMPs' effectiveness will vary according to location,
and these differences cannot be properly evaluated if
nonstandard formats are used. (An example of a
standard format for reporting data appears at the
end of this paper.) Documentation of other pertinent
land treatment data must be tailored to each project.
In establishing procedures for land treatment
documentation, several things must be kept in mind.
• The information must be recorded at the
easiest point in project implementation.
• The data must be consistently reported at all
stages of data collection.
• The data log should be easily accessible
throughout each project period so that
interested agencies can immediately view or
analyze specific data.
• Annual summaries of data for all practices
completed should be made by subbasin.
Documentation of Project Data
• Technical and Cost-share Data. The technical
and economic data for all cost-share practices are
present whenever a payment is issued. According to
the RCWP manual, upon certification by SCS that
BMPs have been properly installed, the producer
can receive cost-share payment for each practice. In
the Long Pine Creek (Nebraska) RCWP, SCS cer-
tified that the practices were completed by returning
a copy of the Conservation Reporting and Evaluation
System Data Sheet (AD-862 CRES Form) with the /
ACP-245 to ASCS.
ACP-245s are issued at the beginning of each fis-
cal year for each planned cost-share BMP to serve as
275
-------
Proceedings of National RCWP Symposium, 1992
a reminder of scheduled BMPs, provide claims for
cost-share payment, and account for disbursement of
cost-share funds. The CRES form filled out by SCS
technicians records all technical data pertaining to
each practice. At this point, all of the information
pertaining to BMP costs and technical data are in
place. The ASCS program assistant should record
pertinent information in the standard nonpoint
source data log for every practice completed, regard-
less of overlapping detail (i.e., two or more practices
completed in the same field).
• Documentation of Non-cost-share and Man-
agement Practices. Each year, SCS technicians
perform annual status reviews on all contracts im-
plementing BMPs. At this point, they can use the
standard nonpoint source data log forms to docu-
ment technical and economic information for non-
cost-share and management practices that were
implemented.
Documentation of Other Land
Treatment Data
At the beginning of each project, technical personnel
from all agencies must decide, in addition to project
data,
• what other data will be needed for future use,
• what information is obtainable,
• what procedures are there for obtaining data,
and
• who will record the information.
Data logs should be created to record the
selected data. Design formats should be used in con-
junction with the standard nonpoint source project
data log format For example, if annual data is col-
lected on the standard nonpoint source project by
subbasin, then annual data by subbasin should be
collected in other formats.
Each project will have a separate set of goals and
data needs based on the area's agricultural and
urban activities. Because information other than
project data must be specifically tailored for future
use, technical personnel must decide how extensive
ly to collect data. Some data, such as cropping his-
tories, management practices, fertilizer and pesticide
use, and waste management, may be documented
annually.
Cropping histories are available in the ASCS of-
fice on the ASCS-578, which is the certification of a
producer's annual cropping history. Both par-
ticipants and nonparticipants certify their activities
yearly; farm program participants are required to
certify, but nonparticipants generally also certify to
retain crop base records.
Several approaches exist for recording these
data. Copies could be made of maps accompanying a
producer's certification and given to the RCWP
clerk. These maps identify the fields, acres, and
crops for each specific year and are the same section
maps used in the master aerial photograph section
map file of the critical area. The RCWP clerk could
file these maps in the master small aerial photograph
map file. Also plan to obtain cropping histories on
farms that do not certify each year. Mail maps to
producers and ask them to identify crops planted or
visit the producers. You may have to send data logs
that require detailed reporting of specific informa-
tion by field or farm (for example, fertilizer use) to
producers so they can document this information.
Uniform Reporting of Technical Data
"Acres served" and "units applied" data are used to
evaluate cost and water quality monitoring informa-
tion to determine
• the cost-effectiveness of specific BMPs,
• the general effectiveness of specific BMPs,
and
• the effectiveness of each project in general.
These data are recorded by SCS technicians on
CRES forms whenever practices have been com-
pleted. SCS employs over 140 practices, and SCS
technical standards explicitly define reporting "units
applied" for each BMP.
However, guidelines for determining acres
served are less explicit. Acres served is defined for
10 practices. A general guideline given for other
BMPs states that acres served is "land in whole
acres served, benefited, or protected by application
of a conservation practice or system. When installa-
tion benefits adjacent land, report the acres benefit-
ing from the practice, plus the area taken up by the
installation. Acres affected little or not at all should
not be reported as acres served."
In the Long Pine RCWP project, acres served
was thought of in different ways by different tech-
nicians. One technician considered a diversion as
serving only the acre on which it was applied,
another believed the practice served a whole field,
while another thought the BMP served acres both
above and below or adjacent to the diversion site.
Sometimes this definition of acres served even in-
volved land not in a contract. Therefore, acres
served were inconsistently reported.
276
-------
B.HERMSMEYER
In addition, acres served data often do not
reflect the amount of land treated annually within the
contract, within a subbasin, or within the project
area. In the Long Pine RCWP project, if the acres
served data are added by practice for each subbasin
(taking into account practices serving the same
acres, such as BMP 6-516-pipeline, 614-tanks, and
642-wells), the acres served will not reflect the num-
ber of acres treated.
Often, the outline of the critical area was drawn
around an entire quarter section (160 acres) that was
contributing to surface runoff. This section may
have contained a 125-acre center pivot where BMP
15-fertilizer management and BMP 16-pesticide
management were implemented. That year, a 20 per-
cent Acreage Conservation Reserve (ACR) (set-
aside requirement) was in effect. Therefore, the
acres-served figure for BMP implementation
documentation will not include the corners around
the pivot or the ACR, which was not considered
treated by the practice because it was not physically
treated. Cost-share was issued at $1 per acre or acres
treated. Acres treated was reported as acres served.
Many similar situations occurred in the Long
Pine Creek RCWP project. Acres within the critical
area were actually considered treated but not docu-
mented as treated. Therefore, data are not available
on the amount of acreage under contract that was
treated annually. ASCS only knows that, when all of
the practices in the contract are complete, all of the
acres in the contract will have been treated.
Conclusions
If agencies throughout the United State use acres
served data to determine a practice's cost-effective-
ness or its effectiveness in general, then a standard
method of reporting acres served for each practice
must be established. Because this information is be-
coming increasingly important in determining a
project's effectiveness and in planning and applying
future BMPs, SCS should clearly define standard
methods of reporting acres served for every practice
code and various situations (specifically, for acres
under contract in nonpoint source projects). The
standard methods for reporting acres served or
benefited should be included in the SCS technical
manuals under each practice code description.
A workshop for technical personnel from all
agencies should be held to adopt a standard data log
format for all nonpoint source projects involving
BMPs and to discuss and determine standard
methods for reporting acres served (or benefited)
for each BMP.
Standard procedures and a documentation plan
for all nonpoint source projects are vital to future
evaluation. Many areas can also be expanded, includ-
ing alternative data collection formats, use of com-
puterized database programs, computer mapping,
and documenting of BMP implementation. However,
the data log format serves as the base for document-
ing all nonpoint source project data and is the source
of all data used in any computer program.
Standard Nonpoint Source Data Log Format: Technical and Cost-share Data Recorded by BMP Item.
DATE
PRACTICE
INSTALLED
PRACTICE
RCWP
CONTRACT
SUBBASIN
NUMBER
FARM
FIELD ID #
BMP
NUMBER
PRACTICE CODE
DESCRIPTION
COST-
SHARE
NON-COST-
OR SHARE
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
NOTE:
According to the RCWP manual, individual ACP-245S are issued at the beginning of each fiscal year for each
planned cost-shared BMP. These serve as a reminder of scheduled BMPs, provide claim for cost-share pay-
ment, and account for disbursement of cost-share. Upon certification by SCS that BMPs have been properly
installed, the producer can receive cost-share payment. SCS certifies that the practice has been completed by
returning a copy of the Conservation Reporting and Evaluation System Data Sheet (AD-862 CRES Form) with
the ACP-245 to ASCS. The cost-share payment can then be made to the producer.
At this point all information pertaining to BMP costs and technical data are in one place. The CRES form has all
of the technical data pertaining to each BMP. The installation costs and cost-share information are present. This
is when the data log should be addressed. Whenever a cost-share payment is made, the RCWP program assis-
tant can use the data log to document cost-share practices. When annual status reviews are performed, SCS
technicians can use the data log to record noncost-share practices. This data log would be easily accessible at
any point during the project. At the end of the project, all of the information regarding all of the BMPs is in one
place. The data can be entered into a computer at any time and sorted accordingly.
277
-------
Proceedings of National RCWP Symposium, 1992
Data Log Format: Technical and Cost-share Data Recorded by BMP Item.
INSTALLATION
COSTS
COST-
SHARE
ACRES
SERVED
UNITS
APPLIED
*MISC. UNITS
REPORTED
SOIL LOSS SAVINGS
(T/ACRE)
WATER SAVED
(AC.-FT.)
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
NOTE:
All columns are not applicable at every practice. The purpose of the data log is to record ALL of the relevant
data for each practice installed or implemented. Because it is difficult to know what data will be pertinent a
decade from now, it will be necessary to record all available information regarding each practice. If data other
than that noted in the format are available, applicable columns should be added to this two-page format and
the data recorded. All practices must be entered in the data log as they are completed. This includes overlap-
ping data (i.e., two or more practices completed in the same field). Many situations will occur: (1) acres
served may be less than acres in the field; (2) a practice may serve several fields; and (3) several BMPs may
serve the same field. Record all data pertaining to every BMP.
* Misc. Units Reported are units of interest that are reported, such as the number of trees for tree planting,
cubic yards in regard to structures, and so on. Misc. Units Reported are not the "Units Applied" data that are
officially reported by the Soil Conservation Service technicians on ORES forms.
278
-------
Problems and Conflicts Associated
with the Administration of the
Long Pine, Nebraska, RCWP Project
Robert F. Hilske
Middle Niobrara Natural Resources District
Valentine, Nebraska
ABSTRACT
This presentation addresses problems associated with the administration of the Long Pine Rural
Clean Water Program project in Nebraska. On the local level, administration was hampered by staff
turnover, burnout, leadership problems, and personal conflicts. A review of those problems discus-
ses how they were overcome to result in a successful project and provides suggestions for improv-
ing local administration of projects. Finally, this presentation focuses on problems between the
State and local coordinating committees that led to three years of little or no communication, ex-
amines the possible causes, and gives ideas for avoiding these problems.
The Long Pine Creek watershed located in
north central Nebraska includes Long Pine
Creek and three major tributaries, Bone
Creek, Willow Creek, and Sand Draw. According to
the Nebraska Game and Parks Commission, Long
Pine Creek is the longest (26 miles) self-sustaining
trout stream in Nebraska. Anglers who travel from
all parts of the State in search of the brown and rain-
bow trout in this coldwater stream are having a posi-
tive impact on the area's economy.
The focus of the Long Pine Rural Clean Water
Program (RCWP) project was to reduce the amount
of agricultural contaminants and sediment being in-
troduced into Long Pine Creek because both these
nonpoint sources affected the stream's ability to sus-
tain trout fisheries. In addition to its recreational
benefits, Long Pine Creek is one of the few rapid-
flowing, coldwater streams in Nebraska, which
makes its watershed an even more important
resource for the entire State.
Like other RCWP projects throughout the
United States, the Long Pine RCWP project was ad-
ministered locally by a coordinating committee that
was responsible to the State coordinating committee
(see Fig. 2). This presentation focuses on the local
administration of the project, the agencies involved,
and how they worked together, and it outlines suc-
cesses and failures of this arrangement. Although all
RCWP projects are unique, the lessons we learned in
Nebraska may be of value to future water quality
projects.
As manager of the Middle Niobrara Natural
Resources District, I was a member of the local coor-
dinating committee. Because I became involved in
the project halfway through the 10-year process, I
could observe the administration of the Long Pine
project from a different perspective than members
who had initiated the effort. Conversations with
committee members as well as reviews of meeting
minutes and letters have led me to some strong con-
clusions about the successes and failures associated
with the project (other members who worked on the
project during the same five years have similar con-
clusions). However, problems or failures associated
279
-------
Proceedings of National RCWP Symposium, 1992
Figure 1.—Location of Long Pine Creek RCWP project In Nebraska.
National Coordinating Committee
State Coordinating Committee
Executive Subcommittee
(Chaired by Agricultural
Stabilization and
Conservation Service)
Local Coordinating Committee
Technical Subcommittee
(Chaired by
Soil Conservation
Service)
Information and Education
Subcommittee
(Chaired by Cooperative
Extension Service)
Figure 2.—Long Pine Creek RCWP management structure.
with the Long Pine project should not overshadow
the program's success and the efforts of the mem-
bers of the local coordinating committee. In general,
conditions in the Long Pine Creek watershed have
improved over the last 10 years, and knowledge
developed through the RCWP should have a positive
effect on this resource in the future.
It would be impossible to put together a perfect
project. Considering the number of people involved,
the Long Pine Creek project probably fared quite
well. For the Long Pine project, the problems as-
sociated with local level administration fell into four
major areas:
1. Declining interest and attendance at local
coordinating committee meetings,
2. Inadequate planning and goal setting,
3. Communication problems with the State
coordinating committee, and
4. Frequent Soil Conservation Service staff
turnover.
Declining Interest and
Attendance at Local
Coordinating Committee
Meetings
The Long Pine Creek RCWP project involved a total
of 16 agencies and groups at the local level. Key
agencies included the Agricultural Stabilization and
Conservation Service, Soil Conservation Service,
280
-------
R.F. HJLSKE
Cooperative Extension Service, Nebraska Game and
Parks Commission, Nebraska Department of En-
vironmental Control, and the Middle Niobrara
Natural Resources District. Several other agencies
and groups were also part of the local committee,
among them the Brown County Board, Ainsworth Ir-
rigation District, the Long Pine Landowners Associa-
tion, and North Central Research Conservation and
Development (see Fig. 3). Generally, communication
between local agencies was acceptable, and
problems could usually be worked out over a dough-
nut and cup of coffee. However, it was difficult to
keep all those agencies and groups focused on the
administration of the project. Reviews of the minutes
indicate that, after five or six years, several par-
ticipants began to attend infrequently or no longer
took part in the meetings.
AGENCY
LEVEL
Agricultural Stabilization and Federal
Conservation Services
Soil Conservation Service Federal
Middle Niobrara Natural Resources Local
District
Cooperative Extension Service State/Local
North Central Research Conservation Federal
and Development
Game and Parks Commission State
Ainsworth Irrigation District Local
Brown County Board Local
Nebraska Department of State
Environmental Control
Nebraska Forest Service State
Long Pine Landowners Association Local non-
government
Fertilizer representative Local non-
government
Farmers Home Administration Federal
U.S. Environmental Protection Agency Federal
Agricultural Research Service Federal
Economic Research Service Federal
Figure 3.—Members of the Long Pine RCWP local coor-
dinating committee.
From 1984 through 1986, meeting attendance
averaged 15 persons and 39 local coordinating com-
mittee meetings were held; 19 meetings (49 percent)
were attended by 15 or more people, and only one
meeting was attended by less than 10. Compare
these figures to 1989 through 1991, when 25 meet-
ings took place and the average attendance dropped
to 10 participants. Even more dramatic, during those
three years, only three meetings drew more than 12
people and 13 involved less than 10 participants. The
minutes show that attendees represented the key
program agencies: Soil Conservation Service,
Agricultural Stabilization and Conservation Service,
Cooperative Extension Service, Middle Niobrara
Natural Resources District, and the Game and Parks
Commission. Other members became uninterested
and frequently failed to attend. Several factors led to
this drop in attendance.
• Loss of enthusiasm for the project. It is not un-
common for any effort to generate initial interest
that fades, leaving a small group of devoted mem-
bers who are burdened with the task of seeing the
project through. In the early 1980s, enthusiasm ran
high. As the years passed, interest declined and at-
tendance at meetings suffered.
• Lack of accomplishment at local meetings.
Many people (particularly those who value their
time) tend to get very frustrated at meetings that ac-
complish nothing. As this project moved into its mid-
dle stages, it became obvious that several goals
would not be met. Some members had pet projects
that did not progress at what they considered an ap-
propriate pace. Many local meetings turned into ax-
grinding sessions, accomplishing little. Some
members believed that the process had failed and
lost interest in the meetings; others who were tired
of hearing the same gripes hashed and rehashed be-
came infrequent participants.
• Not enough direct involvement with the
project administration. Several of the original
members (Nebraska Forest Service, fertilizer repre-
sentative, Farmers Home Administration, and the
Agricultural Research and Economic Research Ser-
vice) were essentially on the local committee to pro-
vide expertise. Because the need for their advice was
infrequent, these representatives may not have felt
part of the project administration. At any rate, they
soon took an inactive role in the project.
Proposed Solutions
Limit the Size of the Local
Coordinating Committee
Public boards and committees generally work best if
their size is limited. Much (or more) work can be ac-
complished if the members are strong, dedicated in-
dividuals. For RCWP projects, the local coordinating
committee should be limited to 7 to 11 members to
provide better local administration. Members should
include representatives of key agencies, such as the
Agricultural Stabilization and Conservation Service,
Soil Conservation Service, local Research Conserva-
tion and Development, Cooperative Extension Ser-
vice, the conservation district, and Game and Parks
— plus a landowner. Two or three additional mem-
bers could be considered if appropriate (see Fig. 4).
281
-------
Proceedings of National RCWP Symposium, 1992
Other interested parties should be incorporated into
the subcommittee structure rather than the local
coordinating committee. Each subcommittee (as-
suming a limited number) might be allowed to have
a representative on the local coordinating commit-
tee.
• Agricultural Stabilization and Conservation
Service
• Soil Conservation Service
• Research Conservation and Development
• Conservation District
• Cooperative Extension Service
• State Fish and Game Representative
• Landowner Representative
• Optional Seat
• Optional Seat
Figure 4.—Proposed membership of a local coordinat-
ing committee.
Encourage Strong Local Coordinating
Committee Leadership
The success of any committee is dependent on its
leadership. A strong chairperson should be desig-
nated to ensure that meetings accomplish their in-
tended goal. The chairperson should be responsible
not only for chairing the meeting but also for estab-
lishing the meeting agenda, properly notifying mem-
bers, providing copies of appropriate documents,
and distributing meeting minutes. For the purpose of
the Rural Clean Water Program, it might prove ad-
vantageous to use the same chairperson for a num-
ber of years.
Hold Occasional "Fun" Events to
Maintain Interest
Nothing will discourage a committee more than
month after month of long, dry meetings. Our ex-
perience with the Long Pine Project showed that
meetings followed by tours tended to attract the
largest audience. Nothing will invigorate a commit-
tee like seeing the fruits of its labor. Many times,
tours were preceded by before and after slides of
problems tackled by the project. The tours not only
gave an overview of the project, they helped bond
members of the committee. Other like events might
include visiting with project landowners, assisting
with BMP installation, or scheduling a dinner or bar-
becue in conjunction with a meeting.
Inadequate Planning and Goal
Setting
Many of the local coordinating committee members
clearly did not understand the purpose of the Rural
Clean Water Program. Some members were simply
uneducated about RCWP; others were unwilling to
accept the purpose of the program. Some members
viewed acceptance of the Long Pine Creek water-
shed into the nationwide Rural Clean Water Program
as a savior for the watershed's ills. These people
visualized RCWP's access to funding as a way to con-
struct projects that had been talked about for years
but never built. A few of these projects failed to fit
into the scope of the Rural Clean Water Program,
and when the State coordinating committee turned
them down, supporters became defensive and
blamed the State for not understanding the situation.
As the Long Pine Creek project winds down, some
members still harbor hard feelings over issues that
were never settled, and others view the program as a
failure even though numerous accomplishments
were realized.
The importance of following a well-thought-out
plan was overlooked by the Long Pine local coor-
dinating committee, and the plan developed at the
beginning of the project was not used effectively. Old
ideas were rehashed, stalling meeting productivity.
Except for minor attention given to the plan during
the writing of the annual report, the local coordinat-
ing committee as a whole never formally reviewed it
to discuss progress and make revisions. Members
who joined the local coordinating committee in mid-
project had little idea as to the project's direction and
were forced to set their own path, leading some
agencies in different directions.
Proposed Solutions
Make Sure Local Members
Understand the Project's Purpose
The ground rules for what the project can and.can-
not accomplish should be established at the start.
Time should be spent explaining why specific
projects may not fit into the overall program. All
those involved should be allowed to express their
opinion on the purpose of the project so that
everyone understands the interests of others. If
questions occur regarding the purpose of the
project, answers should be sought from appropriate
agencies. The purpose and goals of the project
should be periodically reviewed not only to benefit
282
-------
R.F. H/LSKE
new members but to serve as a reminder for veteran
members.
Develop Comprehensive Project Plans
and Hold Annual Review Sessions
At the outset of the project, all agencies and groups
should meet and develop a plan that includes all the
perceived needs in the project area. Ideas for estab-
lishing management practices as well as larger
projects should be considered. Needs should be
prioritized and timetables developed for their com-
pletion. At that time, projects and programs that
would benefit the overall project and meet the estab-
lished guidelines should be included as part of the
plan. If a specific need does not fit within the scope of
the overall project, evaluate other methods outside
the RCWP to address it, including agencies that
could be responsible as well as potential sources of
funding.
Once a plan is developed and agreed to, all ef-
forts should focus on completing elements of the
plan as outlined. A comprehensive plan review
should be held annually. Elements of the plan that
have been completed should be highlighted. If a
specific project is no longer a priority or will not be
completed, it should be dropped from the overall
plan. If an item is dropped from the plan, it should
not be discussed at meetings.
Communication Problems with
the State Coordinating
Committee
During the Long Pine Creek RCWP project, com-
munication broke down between the State and local
coordinating committees. Early in the 10-year effort,
a number of what the local coordinating committee
felt were key projects were not supported by the
State coordinating committee. This disagreement
became a catalyst for communication problems that
lasted throughout the project as members of the
local coordinating committee blamed the State coor-
dinating committee and the agencies it represented
for the demise of these projects. Communication to-
tally broke down for several years, when no meet-
ings were held between the two groups. However,
better communication was established between the
two groups during the final years of the project.
Much of the problem could be blamed on simple
geography. The project area was located 280 miles
from Lincoln, Nebraska, where the State coordinat-
ing committee resided. Most communication had to
be done by mail or telephone because State commit-
tee members could not attend local meetings
regularly. As a result, the members of the State coor-
dinating committee seemed to have a hard time get-
ting a feel for what was occurring at the project level.
However, the reverse was also true. Decisions were
made by the State coordinating committee that mem-
bers of the local committee did not understand or
agree with, which made accomplishing the project's
goals very difficult.
Proposed Solutions
Form a Strong Bond Between the
State Coordinating and Local
Coordinating Committees
When the project is established, the State and local
coordinating committees should cooperate to set its
scope and goals. By working together, each entity
will get a better understanding for the other's con-
cerns. A strong effort should be made to include the
State coordinating committee in the initial planning
process. The State coordinating committee should
be included in as many project tours, BMP installa-
tions, and social functions as possible.
!
Encourage Greater Participation By
the State Coordinating Committee
The State coordinating committee should actively
participate at local committee meetings. In situations
where distance is a major factor (as in Nebraska),
one State committee member should be appointed to
attend local committee meetings on a rotating basis.
Regular attendance would give local members the
assurance that interest exists on the State level and a
feeling that committee concerns are being heard. In
turn, State committee members would have a better
feel for the attitudes on the local level. State coor-
dinating committee members could indicate which
BMPs, projects, or ideas they think inappropriate,
potentially avoiding conflicts between the two com-
mittees.
Include More Technical Staff
Many members of the Long Pine Creek State coor-
dinating committee were heads of State and Federal
agencies (see Fig. 5) and generally did not have the
time to work with local members administering the
program. To form a stronger bond between the State
and local coordinating committees, include technical
staff from State and Federal agencies who may have
283
-------
Proceedings of National RCWP Symposium, 1992
more time to work directly with the project on the
State coordinating committee, where their technical
expertise could be an asset. Or have the State coor-
dinating committee establish appropriate subcom-
mittees that would include technical staff from the
State and Federal agencies who could advise and
make recommendations about the project. Members
of the subcommittee would then serve as a direct
link to the local coordinating committee, passing
ideas and advice from the State to the local level (and
vice versa).
AGENCY
LEVEL
Agricultural Stabilization and Federal
Conservation Service
Soil Conservation Service Federal
Game and Parks Commission State
Nebraska Department of Environmental State
Control
Cooperative Extension State
Nebraska Natural Resources Commission Local
Fanners Home Administration Federal
U.S. Environmental Protection Agency Federal
Figure 5.—Present membership of the Long Pine State
Coordinating Committee.
to transfer. When putting together the 10-year report
for the project, members found that this turnover
rate not only affected relationships with cooperators
but project administration as well.
Proposed Solutions
Require a Full-time Project
Coordinator
Any project that involves over $1 million and re-
quires 10 years to complete should employ a full-
time project coordinator. The position should be
established at a GS-11 level to maintain uniformity
over the life of the project and reduce turnover. It
would also make one person responsible for oversee-
ing the entire project rather than many people with
other, outside responsibilities. The project coor-
dinator would help coordinate activities between
agencies and assist cooperators. In addition, the
project coordinator could serve as the local coor-
dinating committee chairperson (or assist the chair-
person).
Frequent Soil Conservation
Staff Turnover
Two agencies were involved with most of the one-on-
one contact with the individual cooperators: the
Agricultural Stabilization and Conservation Service
and the Soil Conservation Service. The Soil Conser-
vation Service played an important role in the project
as it provided almost all of the technical assistance, a
role that required a high level of one-on-one contact
with landowners, who needed to trust the technician
they were working with. Each Soil Conservation Ser-
vice technician also had to understand a cooperator's
operation before suggesting ideas and developing
contracts. Unfortunately, turnover among soil con-
servationists working on the RCWP project was
high. (It was not uncommon for soil conservationists
to transfer after one or two years of working on the
project.) The continuity of the RCWP program suf-
fered greatly, and many of the cooperators never be-
came comfortable with the program or developed
trust in Soil C jnservation Service staff.
Because the Soil Conservation Service staff as-
signed to work on RCWP were usually at GS-7 or GS-
9 levels, project membership became a stepping-
stone position rather than one that was (at least)
semipermanent. Even though soil conservationists
enjoyed their work with the program, if they wanted
to improve their status and salary, they were forced
Designate Agency Staff to the RCWP
All agencies or groups that provide staff support to
the effort should designate the person (s) that will be
working on the project (particularly the Agricultural
Stabilization and Conservation and Soil Conserva-
tion services because they provide the bulk of the
staff time). Too many people provided technical as-
sistance over the life of the Long Pine Creek RCWP
project. By the end of the project, local soil conser-
vationists did not know how much time (if any) they
should be spending on the program. Agencies
providing staff assistance need to complete an an-
nual workload analysis to assign staff members to
the project and determine how much time they
should devote to the effort.
Use More State and Local Resources
Generally, State and local agencies have greater
flexibility than Federal agencies. By using the con-
servation districts, Cooperative Extension Service,
Game and Parks, and other local groups, some turn-
over problems could be reduced. As an example,
Federal funds could be passed to the local agencies
to supply technical assistance and project ad-
ministration as is commonly and successfully done
with other Federal programs (319, Clean Lakes,
Dingle-Johnson). This option would further address
personnel turnover because agencies, such as the
284
-------
R.F. H1LSKE
local conservation district or Cooperative Extension,
tend to have fewer personnel changes.
To support this option, I reviewed a 1989 survey
that documented statistics showing that 141 full-time
employees averaged 8.4 years with Nebraska's 23 lo-
cally governed natural resources districts. In some
cases, local groups or agencies may be able to match
Federal dollars with local funds or in-kind services
that could further improve the program and perhaps
allow incorporation of projects outside the scope of
RCWP into the program.
Conclusions
The Rural Clean Water Program provides a unique
experience for Federal, State, and local agencies to
work together for a common goal. However, as with
any program involving multiple agencies, problems
can occur. One of the best aspects of the RCWP con-
cept is that it was designed to be a learning exper-
ience. A final judgment on the contributions of an
RCWP project cannot be made if problems, mis-
takes, and weaknesses are not recognized along with
the successes.
We had several recognizable problems when ad-
ministering the Long Pine Creek project, including
size of the local coordinating committee, failure to
communicate the purpose of the project to local coor-
dinating committee members, inadequate use of an
overall plan, communication problems with the State
coordinating committee, and excessive staff turn-
over. Closer evaluation of these problems reveals
one commonality — a failure to communicate. Im-
proved communication would have had the strong-
est positive impact on the Long Pine project.
In addition to the need for better communication
what else was learned? If the clock could be turned
back to 1980 and we were to start the process over, I
would offer the following key suggestions for better
RCWP project administration:
1. Keep the local coordinating committee
membership manageable. Seven to 11
members is an appropriate size for this type
of project.
2. Make sure all committee members fully
understand the purpose and goals of the
project.
3. Develop a solid, workable plan and review it
annually.
4. Communicate with the State coordinating
committee and keep it an active part of the
project.
5. Employ a full-time project coordinator.
Had we followed these suggestions, the Long
Pine Creek RCWP project would have achieved even
greater success. The RCWP experience provided a
great opportunity to put into practice what many al-
ready knew — that cooperation and good com-
munication is the best formula for success. Let us
hope that future water quality programs will use
what was learned from the RCWP and incorporate it
into their efforts. If they fail to do so, then much of
what was learned during the past 10 years will be
wasted.
285
-------
-------
RCWP—The Federal Perspective
James Meek and Carl Myers
U.S. Environmental Protection Agency
Washington, D.C.
Gordon Nebeker
Agricultural Stabilization and Conservation Service
U.S. Department of Agriculture
Lander, Wyoming
Walter Rittall
Land Treatment Program Division, Soil Conservation Service
U.S. Department of Agriculture
Washington, D.C.
Fred Swader
Working Group on Water Quality
U.S. Department of Agriculture
Washington, D.C.
ABSTRACT
The Rural Clean Water Program, like any experimental program, has had some false starts and
marginal results, but it has been a success from a number of viewpoints. It influenced policymakers
within the U.S. Department of Agriculture (USDA) to make water quality the second most impor-
tant priority in policy set through the Resource Conservation Act process, exceeded only by
erosion control, the USDA's initial mission. The continued debate over how to resolve water quality
problems brought about policy changes in other USDA programs and led to the development of
tools to address nutrient and pesticide management The resulting partnerships provided a consis-
tent framework for developing proposals for congressional consideration. The program's uninter-
rupted 10-year funding ensured results that have helped develop new nonpoint source programs.
The Presidents Water Quality Initiative (which came on-line in 1990) benefited particularly from
the interagency coordination that developed during the program. This paper examines how the
program came into being, the original expectations, how it has worked to meet those expectations,
and what has been learned at the Federal level.
Section 208 of the 1972 Clean Water Act,
which called for areawide wastewater plan-
ning and management, included language
that, for the first time, addressed agricultural
sources of pollution. Section 208 (j)-—added by the
1977 amendments—authorized $800 million over
four years for implementation of agricultural ele-
ments in State and local nonpoint source programs.
(No implementation funding was included in the
original section 208.)
Initially, implementation was assigned to the Soil
Conservation Service (SCS), which completed an en-
vironmental impact statement and regulations in
1978. However, the USDA's Agricultural Stabilization
and Conservation Service (ASCS) had traditionally
administered money or grant programs. The 1980
287
-------
Proceedings of National RCWP Symposium, 1992
USDA appropriation bill included a provision for an
experimental Rural Clean Water Program (RCWP) to
be administered by ASCS in lieu of funding section
208(j). The first year, $50 million was provided; $20
million was added in 1981, making a total of $70 mil-
lion for the experimental RCWP. This amount was
later reduced to $64 million as part of a Federal
budget reduction measure.
Since SCS had developed regulations and held
four national meetings to explain its program under
208Q, there was initial confusion in the field with the
shift to the program funded under the Appropriation
Bill and administered by ASCS. However, when fund-
ing was made available, the agencies quickly or-
ganized to set up a committee structure to manage
the experimental program, develop new guidance for
the field, and select projects. Memoranda of under-
standing were developed among the agencies to
define specific roles and transfer funds.
These initial agreements within USDA and with
the U.S. Environmental Protection Agency (EPA)
have continued to this day. ASCS has been the ad-
ministering agency for new cost-share programs,
with funds transferred directly to SCS and the Forest
Service for technical assistance, to the Cooperative
Extension Service for educational assistance, and to
the Economic Research Service for monitoring and
evaluations.
The first challenge was to design a cooperative
project selection process and select appropriate
projects from some 75 applications. The projects
were rated and ranked by each of the participating
agencies; competition was stiff and congressional in-
terest high. The National Coordination Committee,
made up of heads of the participating agencies, took
its task seriously and debated every project recom-
mended for funding. This agency interaction trans-
lated into positive support for the program and made
coordination easier. The second challenge was to
develop an evaluation program including selecting
projects that would receive additional resources for
formal evaluations. Initially, 10 percent of the re-
sources were reserved for five projects chosen to
represent a range of agricultural water quality
problems. One project in Kansas was terminated
after a review confirmed that a water quality problem
did not exist. However, contracts existing at the time
were honored. Table 1 lists the projects and funding.
Program Expectations
An earlier effort, the Model Implementation Pro-
gram, taught the agencies several important les-
sons—but especially that the imperial "we" from
Washington, D.C., does not sell at the local level.
Consequently, RCWP, which was designed to work
directly with the project managers at the field level.
The RCWP was a natural fit for EPA's nonpoint
source program as an expansion of the existing sec-
tion 208 nonpoint source emphasis. EPA expected to
initiate needed implementation and testing of non-
point source controls, while USDA hoped to gain
more experience with water quality projects through
increasing responsibility and to show that conserva-
tion practices on agricultural lands produce positive
water quality benefits.
In general, USDA expected that the RCWP
would play a major role in water quality as it related
to impacts from agricultural nonpoint sources. Be-
coming an active manager in protecting water
quality from agricultural nonpoint source pollution
was logical and necessary for effective use of avail-
able resources. Thus, water quality managers in
EPA, USDA, and the States used the RCWP to in-
stitutionalize cooperation for implementation pro-
jects focused on priority watersheds. An increasing
role for USDA in water quality was initiated by the
following objectives, which were spelled out in the
1980 regulations for RCWP:
The program provides long-term financial
and technical assistance to owners and
operators having control of agricultural land.
The purpose of this assistance is to install and
maintain best management practices to con-
trol agricultural nonpoint source pollution for
improved water quality.
Congress intended appropriated funds to be
used in geographic areas with critical water quality
problems. Only the highest priority projects would
be funded from the applications submitted to the
Secretary of Agriculture and approved in consult-
ation with the EPA's Administrator.
This was the first specific inclusion of protection
and/or maintenance of water quality as a USDA ob-
jective. Early efforts to coordinate USDA activities
under the section 208 water quality management
program were frustrated; the department was un-
able to fund participation because water quality was
not a department mission. Funds are now ap-
propriated directly, and USDA places more emphasis
on water quality.
The stated objectives of the RCWP were to
• achieve improved water quality in the project
area in the most cost-effective manner pos-
sible in keeping with the provision of adequate
supplies of food, fiber, and a quality environ-
ment,
288
-------
J. MEEK, C. MYERS, G. HEBEKER, W. R1TTALL, &RSWADEK
Table 1. — Rural Clean Water Program projects.
PROJECT
Lake Tholocco
Appoqulnimink
Taylor Creek-
Nubbin Slough/
Lower Kissimmee
River
Rock Creek2
Highland Silver
Lake2
Prairie Rose Lake
Upper Wakarusa
Reelfoot Lake
Bonne Idee
Double Pipe Creek
Westport River
Saline Valley
Garvin Brook
Long Pine Creek
Tillamook Bay
Conestoga2
Oakwood-Lake ,
Poinsett2
Reelfoot Lake
Snake Creek
St. Albans Bay2
Nansemond-
Chuckatuck
Lower Manitowoc
TOTAL:
LOCATION
Alabama
Delaware
Florida
Idaho
Illinois
Iowa
Kansas
Kentucky
Louisiana
Maryland
Massachusetts
Michigan
Minnesota
Nebraska
Oregon
Pennsylvania
South Dakota
Tennessee
Utah
Vermont
Virginia
Wisconsin
ESTIMATED
COST1
$1,850,827
972,319
2,982,111
5,328,108
3,974,849
732,607
2,230,728
932,186
4,354,188
4,993,057
733,924
2,727,997
2,525,100
2,384,711
5,500,428
4,165,038
3,646,134
3,934,774
463,223
5,179,120
2,116,131
2,270,619
$63,998,179
STATUS
Implemented in 1980; 80% of its critical acres and 95% of its best
management practice (BMP) funds are under contract.
Implemented in 1980; 87% of its critical acres and 99% of its BMP are
under contract.
Nubbin Slough project was implemented in 1981 ; 87% of its critical
acres and 98% of its BMP funds are under contract. Lower Kissimmee
River was added in 1988; 58% of its critical acres and 62% of its BMP
funds are under contract.
Implemented in 1980; 73% of its critical acres and 90% of its BMP
funds are under contract.
Implemented in 1980; 82% of its critical acres and 94% of its BMP
funds are under contract.
Implemented in 1980; 97% of its critical acres and 84% of its BMP
funds are under contract.
Implemented in 1980; terminated 1982.
Implemented in 1980; 64% of its critical acres and 90% of its BMP
funds are under contract.
Implemented in 1980; 75% of its critical acres and 88% of its BMP
funds are under contract.
Implemented in 1980; 100% of its critical acres and 88% of its BMP
funds are under contract.
Implemented in 1981 ; 67% of its critical acres and 88% of its BMP
funds are under contract.
Implemented in 1980; 32% of its critical acres and 98% of its BMP
funds are under contract.
Implemented in 1981; 42% of its critical acres and 100% of its BMP
funds are under contract.
Implemented in 1981 ; 79% of its critical acres and 77% of its BMP
funds are under contract.
Implemented in 1981 ; 98% of its critical acres and 99% of its BMP
funds are under contract.
Implemented in 1981 ; 40% of its Headwaters critical acres and 84% of
its BMP funds are under contract.
Implemented in 1981; 58% of its critical acres and 60% of its BMP
funds are under contract.
Implemented in 1980; 58% of its critical acres and 99% of its BMP
funds are under contract.
Implemented in 1980; 100% of its critical acres and 100% of its BMP
funds are under contract.
Implemented in 1980; 75% of its critical acres and 99% of its BMP
funds are under contract.
Implemented in 1981 ; 75% of its critical acres and 92% of its BMP
funds are under contract.
Implemented in 1980; 57% of its critical acres and 88% of its BMP
funds are under contract.
11ncludes funds for best management practices, information,
monitoring and evaluation
2 Comprehensive monitoring projects
and education, Federal and non-Federal technical assistance, and comprehensive
289
-------
Proceedings of,Hatlonal RCWP Symposium, 1992
• assist agricultural landowners and operators
in reducing agricultural nonpoint source
water pollutants and in improving water
quality in rural areas to meet water quality
standards or goals, and
• develop and test programs, policies, and pro-
cedures for the control of agricultural non-
point source pollution.
Program Organization
To administer the program, a series of committees
was established at the Federal, State, and local level.
A National Rural Clean Water Coordination Commit-
tee, whose members were the administrators of the
participating agencies, was set up to manage the
overall program. Responsibility for administration of
the program was delegated to ASCS and coordina-
tion of technical assistance was assigned to SCS.
State Rural Clean Water Coordinating commit-
tees were set up to assist the State ASCS commit-
tees, and local Rural Clean Water Coordinating
committees were established at the project level to
assist the county ASCS committees. All these com-
mittees were made up of agencies that have respon-
sibility or interest in water quality. This structure
encouraged and facilitated leadership from those
managing the program and resulted in effective
coordination among the three levels. Table 2 il-
lustrates RCWFs administrative structure.
Table 2.—RCWP's administrative structure.
Previous experiences with water quality projects
demonstrated the absolute need for water quality
monitoring. A concerted effort was made to ensure
water quality data would be collected, analyzed, and
used to evaluate the activities' effectiveness. Five
projects were designated for comprehensive
monitoring, and State water quality agencies were
generally funded through the Clean Water Act, sec-
tion 208, to monitor the other projects.
To ensure that data evaluation would be a con-
tinuing effort, a project was established with North
Carolina State University to provide guidance, tech-
nical assistance to projects, develop workshops and
annual reports, and evaluate data to determine best
management practice (BMP) effectiveness. The key
elements for effective monitoring and evaluation are
identified in Table 3.
This administrative/evaluation structure pro-
vided an opportunity to look at USDA's traditional
conservation programs in an environmental/water
quality context. Transitions are never easy, and the
agencies were concerned about the change. SCS
employees tended to think only in terms of
erosion—not water quality; Federal cost-share dol-
lars were viewed as entitlements to be used on a
first-come, first-served basis, rather than targeted at
the highest water quality priorities; and least-cost al-
ternatives were not always developed, nor was in-
novation pursued at all levels.
At the national level, a major component of
managing RCWP was establishing a process (a
team) to review projects regularly to determine how
The Rural Clean Water Program Is managed through
the following committees:
National Rural Clean Water Coordinating Committee
• Policy Committee
• Technical Committee
• Program Review Teams
Secretary of Agriculture administers
in consultation with EPA
• Secretary delegates responsibility for administration to ASCS.
•. Approvals made through the ASCS system
• Secretary delegates the coordination of technical assistance to
SCS
—» Assists the administrator of ASCS in managing the program
Advises the Coordinating Committee
Advises the Policy, Technical Committees
Assists the State RCWP Coordinating Committee
State Rural Clean Water Coordinating Committee
Assists the State ASC Committee
Local Rural Clean Water Coordinating Committee
Assists the County ASCS Committee
Nolo: Committees made up of representatives from USDA agencies with programs or interest in water quality to include ASCS, SCS, FmHA,
ES, FS, ARS, EPA, and (as appropriate), water quality agencies at State and local levels as well as farmers.
290
-------
J. MEEK, C. MYERS, G. NEBEKER, W. RJTTALL, & E SWADER
Table 3.—RCWP's monitoring and evaluation struc-
ture.
RCWP projects will be monitored in sufficient detail to deter-
mine BMP application progress and document water quality
improvement trends using the following elements:
• description of WQ monitoring
• schedule
• parameters
• analytical methods
• existing data and trends
• funding from State/local (no RCWP funds allowed)
Selected projects for comprehensive monitoring Q'oint
USDA/EPA) use the following elements:
• objectives — purpose and scope
• strategy to identify changes in WQ
• socioeconomic impacts and cost/benefit on landowners
• institutional aspects for participating agencies
• educational aspects for individual landowners,
• quality assurance
• datastorage
• reporting to NCC
• funding
Evaluation of the RCWP projects will be carried out
annually.
well objectives were being met and whether mid-
course corrections were needed—and to report suc-
cesses or problems. Farmer participation in many of
these project reviews was enthusiastic and direct.
Problems were discussed at 'these open meetings
(often to the chagrin of USDA and EPA field person-
nel) and resolved: new practices were authorized,
policies were developed or revised, and partnerships
were formed where adversarial relationships had ex-
isted.
The review team concept focused decisionmak-
ing and provided a constancy that had not existed in
earlier projects. Even in cases where national ap-
peals were made, review team decisions were
upheld. The review teams developed a "can-do" at-
titude and made changes to ensure that each water
quality problem could be fully addressed.
New practices were instituted as a result of the
review team visits. For example,
• portable pumps were cost-shared to allow
recycling of wastewater for prewashing cows
before milking, reducing the need to pump
groundwater,
• animal waste spreading equipment was cost-
shared where multiple-use agreements were
in place,
• portable shade structures for cows were ap-
proved to allow animal exclusion from water-
courses,
• a hog operation was relocated from a lake
shore,
• new furrow irrigation techniques developed in
Idaho are now cost-shared in other approved
States, and
• nutrient and pest management practices were
developed and implemented. (This last effort
became the basis for national standards and
specifications for BMPs put in place in 1990.)
National workshops were effective ways to ex-
change information among projects and manage
their direction. Generally, the workshops resulted in
• a tightly knit program,
• advances in statistical approaches to
analyzing nonpoint source data,
• advances in nonpoint source modeling
approaches,
• improvements in reporting land treatment
activities,
• increased understanding of the need for
groundwater monitoring, and
• improved nonpoint source monitoring
protocols.
Annual reports have provided standard sum-
maries of RCWP and other agricultural nonpoint
source projects, statistical analyses of data from
selected projects, and analyses of lessons learned
from the RCWP.
As USDA conservation programs are broadened
to address water qualify, the RCWP experience of in-
novatively changing traditional approaches to keep
the focus on water quality must be emphasized. The
pendulum must not swing back to the old traditional
soil erosion conservation programs—such as the
conservation compliance provision of the Food
Security Act of 1985—because these programs do
not fully address water quality.
Project Selection
The agencies made a concerted effort to ensure that
the RCWP would not repeat past mistakes. First,
they decided that RCWP should not reinvent
programs but build on existing ones. To avoid a new
round of planning, a proposed RCWP project had to
be in an area where a critical water quality problem
was identified during the section 208 planning
process and be consistent with the section 208 water
quality management plan. RCWP would implement a
select group of projects to demonstrate BMP effec-
tiveness in controlling agricultural nonpoint source
pollution—not an easy task.
291
-------
Proceedings of National RCWP Symposium, 1992
Projects had to
• have strong institutional and local support,
• protect identified beneficial uses,
• deal with a critical problem,
• correct point sources,
• have benefits proportional to the investment,
and
• be of manageable size.
Table 4 specifies the criteria for project eligibility
and selection. A series of regional workshops were
held to develop applications. Additional workshops
were held for selected projects to explain the work
plan requirements. About 800 agency professionals
from USDA and EPA attended these nationwide
workshops.
Table 4.—RCWP project selection.
ELIGIBILITY
Priority water quality areas Identified under 208 WQM; i.e.,
hydrotoflte boundaries with water quality problems related to
agricultural nonpolnt source pollutants (only critical areas
and sources of pollutants significantly contributing to water
quality problems).
CRITERIA
• Severity of the problems
- Designated uses affected
- Miles or acres of waterbody affected
• Public benefits
- Population affected
- Human health
• Feasibility
-Size
- Cost per participant
- Cost-effectiveness of BMPs
• Suitability
- As an experiment/test of programs
- Potential for monitoring (baseline data/evaluation)
• State/local
- Funding
- Commitment of leadership
• Contribution to national water quality goals
• Participant's water quality plan
- Developed with assistance from SCS
- BMPs to reduce agricultural chemicals
- Treatment of all critical areas
- Compliance with alt applicable laws
- Time schedule
National Level Results
The successes and failures of the RCWP program
revealed several factors that differed from the
national perspective.
Resources
At the outset, the ASCS set aside full funding for
each project plus contingencies—technical assis-
tance, other agency participation, and funding for an
outside group to evaluate the effectiveness of
selected BMPs. Such assured funding for the life of
the project allowed projects to plan and carry out ac-
tivities without the uncertainty of dependence oh an-
nual budgets or appropriations. More importantly, it
allowed the projects to operate for 10 years — long
enough to look at the water quality impact of chang-
ing farm management practices — and also allowed
a consistent level of program support throughout
this period while other programs were being
reduced.
Structure for Managing
While considerable guidance was given at the nation-
al level to get the process started, the projects were
completely developed on the grass roots level, with
organizational support through a series of local com-
mittees. Local institutional support came principally
from agricultural interest groups, which created a
strong institutional base and ensured a consistency
in approach and staffing. Where this did not happen,
projects did not meet all their objectives.
Staffing
At the national level over the 10 years of this project,
a fairly consistent group of people have been directly
involved or in the background. ASCS experienced
the most changes but it has retained some staff who
were part of the initial process. This consistency
adds stability instead of unsettling changes of direc-
tion to fit new personalities. RCWP has a relatively
low profile because top management are not con-
tinuously involved. Delegation of responsibility has
been a major factor in the success of the program.
Communication
The participants have done an excellent job in
moving information from the national to the field
level—and back. National workshops have allowed
excellent exchange of information. Newsletters and
directives facilitated communications among the
projects and provided both support and interaction.
Evaluation
From the beginning, agencies recognized a need for
an independent group to carry out continuing
evaluation of projects and to document what did and
did not work. The North Carolina Agricultural Ex-
tension Service was selected to carry out this func-
tion because it could devote full time and attention to
RCWP. The ensuing series of workshops, annual
reports, on-site evaluations, and technical assistance
292
-------
J. MEEK, C. MYERS, G. NEBEKER, W. RTTTALL, & F. SWADER
have been summarized in an EPA document, "Les-
sons Learned." EPA and USDA have also tapped this
group's expertise for help with other nonpoint
source issues.
In summary, the program benefited from several
factors: adequate resources, stability of staff, better
than usual communication, and a strong evaluation
component.
How RCWP Changed Some
State Programs
• In Alabama, the RCWP provided the impetus for
one county to deal with severe erosion on road
banks and implement conservation measures on
land owned by the Department of Defense. In addi-
tion, RCWP was the model for the new State cost-
share program.
• In Oregon, a computer model was developed to
explore management options for animal waste
management and coliform bacteria in Tillamook Bay.
The model approach was expanded to other es-
tuarine drainage areas, and RCWP project staff
provided training.
• In Vermont, local staff developed new bacterial
tracing methods and computer models to determine
management options for controlling barnyard
runoff.
• In Iowa, the RCWP triggered complete restora-
tion of Prairie Rose Lake. Without the program, the
State might never have undertaken the fisheries res-
toration project.
• In South Dakota, the RCWP was responsible for
developing both groundwater and lake management
expertise in the State.
• In Florida, the success of the RCWP project
prompted additional Federal and State funding to ad-
, dress the problems of Lake Okeechobee.
What RCWP Learned from
States
• The Pennsylvania RCWP project showed that
some farmers are very reluctant to sign contracts
with the Federal Government and that other
mechanisms may be needed to accomplish program
goals.
• The Idaho RCWP project demonstrated the im-
portance of identifying the actual problem. Agricul-
tural BMPs applied to farmland are (not
surprisingly) ineffective in dealing with unrelated
strearhbank erosion problems.
• The Oregon RCWP project discovered that the
possibility of a lawsuit or the loss of a market for
farm products stimulated interest in applying BMPs.
• The Kansas RCWP project illustrated the neces-
sity of field reviews and the importance of coopera-
tive planning in developing project applications and
operating plans.
• Several projects demonstrated the importance of
face-to-face interactions with the project review
team, the local coordinating committee, and the
State coordination committee as well as the impor-
tance of a "can-do" attitude and the need for imagina-
tion and program flexibility.
• USDA has also learned where the Food and Drug
Administration bans shellfish harvesting and what
should be done—beyond land treatment—to get
these areas reopened, knowledge that has increased
USDA's confidence when operating in estuarine
areas.
The Future
The program name may change but the concept will
live on. As this paper was written section 208(j) (the
original Rural Clean Water Program) has become
part of the 1992 reauthorizing bill for the Clean
Water Act (S-1081). The initial funding level would
provide $200 million to SCS through the section
304 (k) transfer mechanism and $400 million to EPA
for grants to States. S-1081 leaves the program
specifics up to USDA and EPA
EPA and the States are stressing a watershed ap-
proach to comprehensively abate point and nonpoint
pollution sources. The lessons learned from RCWP
about implementing targeted watershed solutions
will continue to influence overall control and
management of agricultural and other nonpoint
sources.
293
-------
-------
Thinking About a Postprqject
Evaluation — Start NOW!
Clarence W. Robison and Charles E. Brockway
University of Idaho Research and Extension Center
Kimberly, Idaho
ABSTRACT
Evaluating water quality improvement projects on irrigated areas requires a complex and complete
database that includes information on land use, best management practice (BMP) implementation,
cropping patterns, irrigation system operations, and water quality information for the project area,
including the nonparticipating lands. In surface irrigated areas, runoff and associated pollutants
are returned to irrigation laterals and drains, which supply irrigation water to other users. This
canal network is often highly complex and regulated differently by watermasters each year. The
quality and quantity of runoff entering a natural stream from an irrigated watershed are very de-
pendent on the operation of the canal network and irrigation systems, in addition to the land use.
Under the Rock Creek Rural Clean Water Program project, the U.S. Department of Agricul-
ture (USDA) Soil Conservation Service tracked BMP implementation, and the Idaho Division of
Environmental Quality monitored and tracked water quality. BMP cost data was collected by the
USDA Agricultural Stabilization and Conservation Service. Planted crop data was collected by the
local soil conservation districts. After 10 years, a great deal of information was available on plan-
ning, BMP implementation and installation, and water quality and cropping patterns. However, cru-
cial data elements for a complete and rigorous postproject evaluation were lacking, including
nonparticipant implementation, implementation and operational status after contract completion,
cropping information by field and year, and canal network operation. This paper covers data ele-
ments that must be addressed at the inception and initial implementation of water quality and
erosion control projects on irrigated areas for which a postproject evaluation is planned or even
considered.
Because of point and nonpoint source pollu-
tion, Rock Creek in Twin Falls County,
Idaho, has long been recognized as one of
the most affected streams in Idaho. While most point
source loads had been addressed and minimized by
end of the 1970s, Rock Creek was still carrying high
loads of sediment and nutrients from irrigated
agricultural activities. -In 1980, the stream was
selected as one of the Rural Clean Water Program
(RCWP) projects. Additionally, the project was
selected to be one of five funded for comprehensive
water quality monitoring and economic evaluation.
At the end of the RCWP, the project was also chosen
for a postproject evaluation.
A tributary to Snake River, Rock Creek is located
almost entirely within Twin Falls County in south-
central Idaho. The headwaters are located in the
Cassia District of the Sawtooth National Forest at
elevations of 2,133 meters (7,000 ft). Shortly after
leaving the national forest, Rock Creek flows
through the Twin Falls tract, an intensively farmed,
irrigated area of 82,500 ha (204,000 acres) that has
been farmed since 1907. The Rock Creek RCWP
project watershed covers approximately 18,200 ha
(45,000 acres) of irrigated pasture and cropland,
rangeland, woodland, and urban lands (Sterling,
1983). The soils are highly productive but extremely
erosive in the project area and lie on gently sloping
plains. The climate is semiarid with annual precipita-
tion averaging about 228 mm (9 inches) and a grow-
ing season of 120 days. Most precipitation occurs
during the winter months. Because of low annual
precipitation, all crops in the project area are ir-
rigated (Yankey et al. 1991).
The predominant irrigation method is furrow ir-
rigation. Irrigation water for the project area is
295
-------
Proceedings of national RCWP Symposium, 1992
diverted from the Snake River near Milner, Idaho.
Twin Falls Canal Company delivers irrigation water
to and removes surface and subsurface runoff from
the individual farms through a network of canals,
laterals, and drains. When the irrigation tract was
developed in the early 1900s, natural streams and
drainage patterns were used to deliver irrigation
water and carry away surface runoff. Recharge to
the underlying groundwater caused the water table
to rise over 80 feet in a 20- to 30-year period. The
canal company and farmers installed gravity
drainage systems to lower the water table. These
systems discharge the subsurface drainage water to
the delivery system's surface drains. The canal com-
pany uses runoff from higher farms to supply irriga-
tion water to lower farms. From early April through
mid-October, water is continuously supplied to
farms.
Problem Statement
In performing the postproject evaluation, three
major problem areas were encountered with BMP
implementation, land use, and water quality data. A
key element associated with each problem was the
network of canals, laterals, and drains used to deliver
the irrigation water and remove surface runoff and
subsurface return flows. Another element was the
agencies' definition of the hydrologic or treatment
basins (subbasins). The subbasins delineated by the
agencies were not based on hydrological features
and should not have been referenced with the word
"basin." Subbasiij boundaries were established along
property lines, not by topography. The water quality
monitoring network was designed to document in-
stream changes in water quality of Rock Creek and
overall changes in water quality from subbasins and
not within subbasins. The last element was the
agencies' focus on data supporting their activities by
subbasins.
Because land and water use data were not col-
lected for fields in farms not under a current RCWP
contract, evaluation of water quality improvements
resulting from levels or types of BMP implementa-
tion was difficult for the postproject evaluation. The
postproject evaluation question — what mixture of
BMPs and implementation levels in a subbasin ex-
hibits the most improvement in the quality of water
leaving the basin — could not be adequately
answered.
Irrigation System Network
The potential loads of nutrients and sediments being
transported past a water quality monitoring station
depend on their origin and the canal company's
operation of the various laterals and drains. In the in-
itial planning stages of the Rock Creek RCWP
project, subbasins were delineated and prioritized
for planning activities. These subbasins were iden-
tified with irrigation return flow drains entering
Rock Creek, and monitoring stations were estab-
lished along the drain. The monitoring stations were
usually located near the drain's mouth, with one near
the primary canal on a turnout or diversion. How-
ever, subbasins were not hydrologically formulated
— property lines were the prime delineating features
of the subbasin boundaries. Often, laterals or drains
enter a subbasin, traverse across the subbasin to a
primary drain, and mix with the drain water. Water is
then rediverted and traverses the subbasin to the op-
posite boundary and leaves the subbasin. A monitor-
ing station located at the end of the primary drain
will never reflect the entire sediment and nutrient
load of surface returns from fields upland of the
traversing lateral or drain. It may reflect a portion,
depending on how the canal company operates the
diversion where the lateral and drain mix. Fields in a
farm unit along a subbasin typically never drain to
the same primary drain. To illustrate this monitoring
problem, this paper will examine the delivery sys-
tem, followed by an examination of a typical farm
unit.
Overall Irrigation System
To deliver and remove irrigation water, canals,
laterals, and drains traverse the Rock Creek RCWP
project watershed. As water is delivered to upstream
farms, runoff water is collected by various drains
and laterals and delivered again to downstream
farms. Over the entire Twin Falls Canal Company
system, 36 percent of the diverted irrigation water
(plus precipitation) is used by crops consumptively,
50 percent is lost to subsurface drainage, and 14 per-
cent returns to Snake River through surface streams
such as Rock Creek (Carter et al. 1971). However,
approximately 50 percent of the water delivered to
an individual farm unit leaves as surface runoff
(Yankey et al. 1991). Thus, eroded soil and nutrients
leaving upland fields with surface runoff enter the ir-
rigation distribution system and have high potential
to be diverted to another field before leaving the
watershed or passing a water quality monitoring sta-
tion.
Lateral or Drain Operation
Figure 1 is a simplified schematic diagram showing
the complexity of water flow within a typical farm
drain system. The canal company delivers water to
Farm A at Point 1 and to Farms B and C at Point 3.
The surface runoff leaves Farm A and reenters the
296
-------
C.W. ROB1SON & C.E. BROCKWAY
lateral at Point 2. The canal company has to spill suf-
ficient water at Diversion 1 to supply Farms B and C,
with allowance for runoff from Farm A. At Diversion
3, the company will deliver this mixture as irrigation
water to Farms B and C and spill any excess water.
This excess water flows down the drain and is
monitored by the water quality station at Point 5.
Thus, the surface runoff from Farm A is partially ap-
plied to fields in Farms B and C and a portion of the
runoff passes the water quality monitoring station.
The nutrients and sediments associated with the
runoff from Farm A applied on Farms B and C will
be deposited. "New" sediments and nutrients will be
transported off fields from these farms in their sur-
face runoff. The runoff from Farm B enters the drain
at Point 4, above the water quality station (Fig. 1).
The runoff from Farm C enters a lateral that supplies
other farms at Point 6. Thus, the water passing the
monitoring station is a mixture of the lateral supply,
runoff from Farm A, and runoff from Farm B. If the
canal company minimizes the spill at Diversion 3, the
mixture will be primarily composed of runoff from
Farm B. Thus, water quality improvements
measured at the monitoring station are not entirely
associated with BMP implementation on Farm A, but
are instead masked by operation of Diversion 3 and
the land use and BMP implementation on Farm B.
Canal
Within Farm Operations
Field irrigations are based on some schedule — the
calendar, soil moisture, or crop use. Figure 2 depicts
the fields of Farm C. Each field will be irrigated in-
dividually in sets that may or may not cover the en-
tire field. Sometimes irrigation will occur on multiple
fields. At other times, no fields will be irrigated, and
the irrigator will bypass the water being delivered to
the farm to Point 6 without using it. In the earlier ex-
ample, runoff from this farm was identified as leav-
ing in one location. That was true with the exception
of the field identified as C4 (Fig. 2). The surface
runoff from this field enters the drain receiving the
runoff from Farm B at ppint 7..The runoff from Field
Cl can be routed for use on Field C3 or to the usual
runoff location at corner of Field C3. The runoff
from Field C4 will affect the water quality at the
monitoring station only when it is being irrigated.
Because of crop rotations, the amount of sediment
and nutrients leaving this field will change annually.
Runoff from alfalfa will have the least effect at the
monitoring station. However, runoff from beans or
sugar beets will have a high effect on water quality at
the monitoring station. The measured effect of ir-
rigation of this field with or without treatment also
depends on whether it was being irrigated at the
Figure 1.—Portion of a lateral and drain network within surrounding farms.
297
-------
Proceedings of National RCWP Symposium, 1992
Figure 2.—Farm C and Its associated fields.
time water quality samples were taken at the
monitoring station.
Rock Creek Aspects
Participation in the RCWP was voluntary. In the
Rock Creek project, many farm units along the lower
portion of drains (like Farm B) did not participate. A
cooperator in the program did not have to participate
for the project's entire 10 years. Several farms were
only involved with the project for three to five years.
Because cropping and BMP implementation data
were not collected on these farms during the nonpar-
ticipation period, the cause of water quality improve-
ments associated with BMP implementation on
participating farms could not be evaluated.
Changes in the operation of the Twin Falls Canal
Company distribution system network occurred
during the 10-year project period. During that time,
different watermasters operated the lateral system.
Some watermasters control the system for minimal
waste (spill); others work better with a larger waste
amount. Because of the multiple interconnections
between laterals and drains, watermasters may
change their primary spill or drain location. During
the project, no agency collected information on how
the watermasters operated the various diversions
along a drain. Because watermasters serve an entire
irrigation season, this lack of data confuses year-to-
year comparisons.
Agency Data Collection Efforts
The objectives of the Rock Creek RCWP project
were originally to reduce sediment and sediment-
related pollutants entering Rock Creek by imple-
menting BMPs and to control animal waste entering
Rock Creek by applying animal waste management
systems (Yankeyetal. 1991).
The primary agencies involved in the project
were the U. S. Department of Agriculture (USDA)
Agriculture Stabilization and Conservation Service
(ASCS), USDA Soil Conservation Service (SCS),
Twin Falls and Snake River Soil Conservation Dis-
tricts (SCO), USDA Agricultural Research Service
(ARS), and the Idaho Department of Health and Wel-
fare Division of Environmental Quality (DEQ). The
ASCS administered the project. The SCS provided
technical planning and, with the assistance of the
SCDs, attempted to track BMP implementation and
associated cropping information. DEQ was respon-
sible for implementing a water quality monitoring
program for irrigation return flows and in-stream
298
-------
C.W. ROB/SON 6 C.E. BROCKWAY
conditions of Rock Creek. The Snake River and Twin
Falls SCDs collected cropping information and
helped track BMP implementation. The ARS was
responsible for evaluating individual BMPs.
Agency performance was in part judged with
these project objectives in mind; thus, an agency's
data collection and tracking information were
focused on attaining its goals. For example, the SCS
was primarily involved with planning and implement-
ing BMPs. Therefore, its land treatment database
only contained information on planned and installed
practices on a farm unit while it was under contract.
The DEQ was responsible for collecting and main-
taining water quality data for Rock Creek and its
various tributaries and irrigation return flows. Thus,
its database only included water quality for Rock
Creek at several locations, and irrigation return flow
entering Rock Creek and at the head of the drains
(where they leave the main canal). The individual ef-
fect of a BMP on runoff associated with the practice
was documented by the ARS. This effect was ob-
tained by monitoring the inflow and outflow of a
practice. Man-agement does affect various BMPs
and was well documented, along with BMP efficien-
cies and erosion rates for various crops (Carter,
1984). No agencies collected information about
operating the lateral and drain network within sub-
basins.
Recommendations
At the beginning of a water quality improvement
project, clear definitions of project objectives and
procedures for measuring goals should be estab-
lished. Potential questions that would be asked in a
postproject analysis should be formulated or hy-
pothesized. Data requirements to answer the poten-
tial postproject questions and for documenting at-
tainment of project goals should be clearly identified.
A single agency should be responsible for data
management. This agency should determine early in
the project whether the proper type, amounts, and
quality of data are being collected. The irrigation
delivery and drainage system should be defined on a
clear hydrological basis.
Effects on water quality leaving a subbasin be-
cause of implementation of a mixture of practices
within the subbasin are difficult to determine
without intensive water quality monitoring of each
inflow and outflow along a drain and land and water
use information. One subbasin in the Rock Creek
RCWP project includes over four canal company ir-
rigation diversions from the drain, not including
other diversions along laterals in the subbasin and
associated inflows from the various farms. To pro-
vide adequate data for hypothetical postproject ques-
tions, the water quality monitoring network should
be designed to account for the intricate irrigation
delivery and drainage system configuration and
management.
Land use, management, and BMP implementa-
tion for all lands in the project area should be col-
lected during the entire project period. Collecting
land use, treatment, and water management data for
every farm in the subbasin would help explain chan-
ges in water quality that cannot be explained by
BMP mixtures and implementation levels 'of par-
ticipants in the subbasin. Clearly, some items will be
financially unattainable, which can affect a post-
project evaluation outcome.
Conclusions
If surface runoff from irrigated fields may be reused,
water quality improvements associated with BMPs
before reuse will be masked. Therefore, it is ex-
tremely important to adequately describe the irriga-
tion and drainage channel network. The last fields
using the water before a monitoring station will have
the largest effect on the measured water quality. The
operation of the lateral system by the watermaster
controls surface runoff effects from upland fields on
water quality at a monitoring station. Even with suffi-
cient knowledge about the operation of the canal sys-
tem and drainage patterns, land and water use
information (practice implementation and crops) is
required for all lands in the watershed for the entire
project period. The monitoring of the drain end al-
lows assessment of the effect of general BMP im-
plementation. However, without land and water use
information for all watershed lands, it is not possible
to determine which BMP mixture and level caused
the improvement in the subbasin runoff water
quality.
References
Carter, D.L, JA Bondurant, and C.W. Robbins. 1971. Water-
soluble NOa-Nitrogen, PO4-Phosphorus, and total salt balan-
ces on a large irrigation tract Pages 331-35 in Proc. Soil Sci.
Soc. Am., Madison, WI.
Carter, D.L. 1984. Rock Creek Rural Clean Water Project Intensive
Monitoring Project: Report of ARS Activities for 1984. Agric.
Res. Serv., U.S. Dep. Agric., Mmberly, ID.
Sterling, R.P. 1983. Stream Channel Response to Reduced Irriga-
tion Return Flow Loads. Thesis. Dep. Civil Eng., Univ. Idaho,
Moscow.
Yankey, R. et al. 1991. Rock Creek Rural Clean Water Program
Final Report, 1981-1991. Meragency Rep. Agric. Stabil. Con-
serv. Serv.; Soil Conserv. Serv.; Agric. Res. Serv.; U.S. Dep.
Agric.; Idaho Div. Environ. Qual.; and Twin Falls and Snake
River Soil Conserv. Distr., Kimberly, ID.
299
-------
-------
Utah's Snake Creek RCWP Stimulates
Additional Efforts to Improve
Water Quality in Wasatch County
Ray Loveless
Mountain/and Association of Governments
Provo, Utah
Todd Nielson
U.S. Soil Conservation Service
Provo, Utah
Harry Judd
Utah Department of Environmental Quality
Salt Lake City, Utah
ABSTRACT
Deer Creek Reservoir, located in central Utah on the Provo River, holds 152,000 acre-feet of
municipal, industrial, agricultural, and recreational water. In 1980, environmental concerns
prompted Utah's governor to organize a committee to prepare a water quality management plan for
the Provo River drainage, where agriculture is the largest contributor of nonpoint source pollution.
Also, in 1980, a Rural Clean Water Program grant was awarded to help reduce the impact of local
agriculture on water quality. The Soil Conservation Service, Mountainland Association of Govern-
ments, Agricultural Stabilization and Conservation Service, Wasatch Soil Conservation District,
Utah State University Extension Service, and local dairymen formed a partnership to implement
conservation plans on individual dairies located on the Snake Creek drainage (a tributary to Provo
River). The Rural Clean Water Program paved the way for additional planning and implementation.
A Clean Lakes project funded by the U.S. Environmental Protection Agency in 1984 finished what
was started by the program. Wasatch County adopted a no-phosphorus discharge policy and now
requires county developers to produce water quality plans before construction. Sediment basins
have been installed for new road construction and newly developed ski resorts. Water quality
within the watershed is showing significant improvement, noxious species of blue-green algae have
been reduced, and the reservoir is responding to the reduced phosphorus load.
The Snake Creek Rural Clean Water Program
(RCWP) project provided implementation
grants to six dairy and livestock operators
on 700 acres of irrigated farmland. The project's
specific objectives were to reduce pollution entering
Deer Creek Reservoir from agricultural nonpoint
sources and determine best management practices
(BMPs) that would reduce pollution.
Several animal waste control BMPs were applied
to lands within the project area. The main measures
were
• establishing waste management systems for
livestock operations,
• limiting livestock access to streams,
• improving irrigation efficiencies, and
301
-------
Proceedings of National RCWP Symposium, 1992
establishing vegetative cover on streambanks. Comprehensive Planning
The project included BMPs on six sites along
Snake Creek and Huffaker Ditch. About 90 percent
of significant animal waste pollution problems within
the project area were treated from 1980 to 1982,
resulting in an annual reduction of over 1,000 kg of
total phosphorus.
Project Area Description
The Snake Creek RCWP project area is located in
Heber Valley ofWasatch County, Utah. Heber Valley,
a high mountain valley surrounded by higher moun-
tains, experiences a wide variation in climate. The
floor of the valley is approximately a mile above sea
level; 15 miles to the west, the Wasatch Mountains
tower to more than 10,000 feet
Soils range from being well-drained and deep to
moderately well-drained, with some poorly drained,
deep soils on floodplains, low stream terraces, and
the valley bottom. Slopes range from 1 to 6 percent;
therefore, snowmelt, heavy rainwater, and tailwater
from flood irrigation systems readily flow across the
fields and transport manure, rich in nutrients, to
Snake Creek.
An estimated 600 acres of alfalfa grown on the
valley bottom are fertilized with 29 metric tons of
nitrogen (including 1.4 metric tons of commercial
fertilizer), 22 metric tons of phosphorus (including 4
metric tons of commercial fertilizer), and 61 metric
tons of potassium (no commercial fertilizer) each
year. Most of the nutrients applied to the cropland
are animal waste products; only 4 percent of nitrogen
and 18 percent of phosphorus are applied as com-
mercial fertilizer.
livestock enterprises within the project area in-
clude four dairies, four small beef operations (less
than 50 head), and two horse farms. Most of the
operations are located on a waterway. When the
project began in 1980, livestock totaled approximate-
ly 650 dairy cattle, 100 beef cattle, and 35 horses. As
a result of the Federal dairy buy-out program, dairy
cattle have been reduced to approximately 400.
Deer Creek Reservoir, which is approximately
seven miles long and up to three-quarters of a mile
wide, contains 152,000 acre-feet of water at capacity.
Completed by the Bureau of Reclamation in 1941,
Deer Creek dam stores water primarily for
municipal, industrial, and irrigation purposes in Utah
Valley and Salt Lake Valley. Development of the Deer
Creek State Park has increased the already heavy
recreational demand on the reservoir, where fishing,
boating, and waterskiing are increasingly popular.
As plans were being formulated for the Snake Creek
RCWP, Sowby and Berg (1984) were developing ad-
ditional plans for the remainder of the watershed.
The Deer Creek Reservoir and Proposed Jordanelle
Reservoir Water Quality Management Plan was com-
pleted and approved by Wasatch County, the State of
Utah, and the U.S. Environmental Protection Agency
(EPA) as an amendment to the Mountainland As-
sociation of Governments' Areawide Water Quality
Management Plan. Local government was assisted
by the Deer Creek-Jordanelle Reservoir technical
advisory committee, which was composed of repre-
sentatives from local, State, and Federal agencies
and other water quality consultants. The committee's
role was to develop a technically sound plan that
could be implemented through the legal authority of
the county and State.
The Snake Creek RCWP, the first attempt in
Wasatch County to address nonpoint source water
quality problems, provided a demonstration project
where farmers could see the results of a combined
effort between government agencies and local
people. Other farmers in Wasatch County and across
the State benefited from this knowledge. Assistance
through the Clean Lakes Program, Agricultural
Stabilization and Conservation Service cost-share
programs, and the technical advisory committee
provided funding for water quality problems derived
from animal waste on additional farms.
The Clean Lakes Program in Heber Valley is ad-
ministered by the Utah Department of Environmen-
tal Quality. A Phase I study (diagnostic/feasibility)
kicked off the program in 1981; in 1986, Phase II im-
plementation funds were awarded. The 13 participat-
ing dairy and livestock operators have installed
BMPs patterned after those implemented under the
Snake Creek RCWP. The Agricultural Stabilization
and Conservation Service provided additional assis-
tance through the Agriculture Conservation Pro-
gram and its special water quality grants.
During the RCWP project, the U.S. Department
of Interior (1979) developed an environmental im-
pact statement (EIS) for the proposed Jordanelle
Dam, which is to be located on the Provo River
upstream of Deer Creek Reservoir. In a review of
that EIS, the EPA found that no management plan
had been developed for Jordanelle Reservoir. As a
result, in 1979, then-Governor Scott Matheson com-
mitted the State to develop a management plan for
the proposed municipal and industrial system of the
Central Utah Project's Bonneville Unit. This commit-
ment was followed by Bureau of Reclamation action
to include a reservoir management plan in the list of
302
-------
R. LOVELESS, T. NIELSON, & H. JUDD
mitigating measures for construction of Jordanelle
Reservoir. The Mountainland Association of Govern-
ments (1979) supported this concept after reviewing
a section 208 water quality study that recommended
comprehensive reservoir management plans to
protect Deer Creek and Jordanelle reservoirs.
Local county government took the lead role to
determine what, if any, land use or nonpoint source
control regulations should be considered in im-
plementing the plan. The county actively enforced
planning and zoning regulations to require that
water quality consideration be given to all proposed
construction projects. This comprehensive local,
State, and Federal planning effort included the
Snake Creek RCWP as one component.
The Jordanelle Reservoir Technical Advisory
Committee has continued to meet regularly to
review results of the Snake Creek RCWP, Clean
Lakes Program, and other projects. The Soil Conser-
vation Service (as a member of that committee) has
provided technical assistance to each RCWP, Clean
Lakes and Agriculture Conservation Program par-
ticipant. Soil Conservation staff developed detailed
conservation plans that involved inventory of re-
sources, development of alternatives with costs, as-
sistance for landowners, final design of all facilities
and management components, and implementation
of all BMPs in each contract.
The Wasatch County Soil Conservation District
Board of Supervisors reviewed each RCWP and
Clean Lakes contract with participants to ensure that
the developed plan would to meet the program's
overall objectives. It also held joint meetings with the
Wasateh Agricultural Stabilization and Conservation
Service county committee to review requests for
contracts and establish priorities for contract
development.
Technical expertise in planning animal waste
facilities was provided by an animal waste specialist
from Utah State University Cooperative Extension
Service. This important link to state-of-the-art tech-
nology in handling these wastes was extremely valu-
able in planning effective waste management
facilities. In addition, several private companies that
dealt with agricultural waste handling equipment
acted as consultants in planning and designing waste
storage facilities and handling equipment. Mountain-
land Association of Governments and the Utah
Department of Environmental Quality took the lead
in implementing water quality management plans
and evaluating project effectiveness through water
quality monitoring and reporting.
As a result of an interagency, interdisciplinary
approach to planning animal waste systems, the
Snake Creek RCWP and other programs have pulled
together the best technical knowledge to develop
sound alternatives and solve resource problems
found in the critical areas. Wann (1986) quoted an
EPA official as saying, 'The work at Snake Creek
seems to have been successful from several
standpoints: water quality has significantly im-
proved, the farms are now much more efficient
operations, and a model for other similar projects
has been created." In addition, the charge given by
Lawrence Jensen (1986), then EPA's assistant ad-
ministrator for water, for upstream users to manage
the resource properly so that downstream uses are
protected, has been achieved. In 1990, Wasatch
County was recognized with an award by Utah's
governor for the State's most outstanding water
quality program.
The total installation cost of BMPs in the RCWP
was $210,508. Of this amount, 68.3 percent
($143,684) was paid by project participants. Al-
though the program authorized cost-sharing at a
level of 75 percent, a $50,000 payment limitation
reduced the overall percentage. • ', .
Implementation of a water quality plan in Heber
Valley has significantly benefited the reservoir. Point
and nonpoint sources have been abated through the
Snake Creek RCWP and Clean Lakes Program. After
studying the phytoplankton in the reservoir for a
number of years, Rushforth (1988) noted,
It seems to be very likely that the phosphorus
limitation program that has been instigated
for Deer Creek Reservoir has improved the
biological water quality of the water and will
continue to lead to further water quality im-
provements. It is significant that the water
quality seems to have improved each year
during the past four years. Long term
monitoring is providing an excellent oppor-
tunity to chronicle what appears to be the sig-
nificant recovery of Deer Creek Reservoir.
Water Quality Investigations
Water quality studies conducted on Deer Creek
Reservoir during the 1970s (Mountainland Ass. Gov.
1979) and early 1980s (Utah Dep. Health, 1984;
Sowby and Berg, 1984) concluded that the reservoir
was eutrophic and phosphorus was the nutrient of
concern. Excessive nutrients were entering the
reservoir from human-caused pollution as well as
natural sources. Studies showed that the average an-
nual phosphorus load reaching Deer Creek Reser-
voir before implementation of water quality manage-
ment practices was approximately 25,000 kg/yr.
Water quality models predicted that the phosphorus
303
-------
Proceedings of National RCWP Symposium, 1992
load had to be reduced by 11,000 kg/yr to protect
the beneficial uses of the reservoir.
To achieve the desired reduction level of
nutrient loading to the reservoir, the Jordanelle
Technical Advisory Committee interdisciplinary task
force identified phosphorus sources. Although
dairies and feedlots represented only 12.1 percent of
the total phosphorus load, they were a nutrient
source that could be controlled cost effectively.
Water quality goals and project objectives were
planned to reduce agricultural nonpoint sources of
pollution by 40 percent to Deer Creek Reservoir and
determine if selected BMPs could improve water
quality effectively.
Water quality studies determined that title Snake
Creek drainage was contributing excessive amounts
of nutrients to the reservoir. Data indicated that the
lower Snake Creek drainage (which covers 7,302
acres_ 1.4 percent of the Deer Creek drainage) was
contributing 12.73 percent of the total phosphorus
loads; in addition, these studies identified and quan-
tified other sources of phosphorus pollution. Other
programs, including the Clean Lakes Program, tar-
geted those sources for improvement.
The RCWP project goal was to reduce total phos-
phorus in Snake Creek by 650 kg (50 percent); in
Huffaker Ditch by 600 kg (75 percent); and in Bun-
nel Ditch by 120 kg (75 percent). The following three
animal waste control measures were identified for
this project:
1. Control transport of animal waste to Deer
Creek Reservoir through Snake Creek by ap-
plying waste management systems to live-
stock operations within the project area.
2. Protect stream water quality by limiting live-
stock access to streams in the project area,
thereby reducing the amount of animal
wastes directly entering the stream system.
3. Reduce overland transport of nutrients by im-
proving irrigation efficiency on all irrigated
lands in the project area.
Responsible for an approximately 1,000 kg/yr
reduction of the total load, the Snake Creek RCWP
was a major stimulating force for additional nutrient
decreases occurring under other programs. This
RCWP project developed the model that is now
being used by all other livestock waste management
systems designed in the county as well as the State.
Results
Beginning in November 1979, 20 stations were
selected on the basis of their location within the
Snake Creek drainage system and monitored
regularly to determine which agricultural activities
actually contributed the highest level of nutrients to
Snake Creek. Sites included eight on Snake Creek,
two on Epperson Ditch, three on Huffaker Ditch,
four on Bunnel Ditch, two on the Provo River, and
one on an unnamed ditch.
Monitoring on sites immediately above and
below dairies participating in the project provided
background data on water entering Heber Valley.
Other sites provided estimates of runoff from irriga-
tion practices and data that compared impacts from
point versus nonpoint sources. Deer Creek Reser-
voir and other tributaries in the watershed were also
monitored.
Water samples from all monitoring sites were
analyzed for
• total and fecal coliform bacteria,
• organic nitrogen,
• total Kjeldahl nitrogen,
• ammonia,
• nitrate,
• nitrite,
• total phosphorus,
• orthophosphorus,
• biochemical oxygen demand,
• total suspended and dissolved solids,
• conductivity, and
• pH.
Water samples were collected monthly in the Snake
Creek drainage. During spring runoff when most of
the snowmelt water flows into Deer Creek Reservoir,
water samples were collected biweekly and analyzed
by a certified analytical laboratory. Monitoring
began in 1979 and has continued to the present, with
modifications.
Tables 1, 2, and 3 show the average concentra-
tion of total phosphorus as well as the total phos-
phorus load from sites on Snake Creek between
1980 and 1990. During 1980 and 1981, before im-
plementation (Table 1), the total cumulative phos-
phorus load to Snake Creek at Sites 03 and 04 (Pride
Lane Dairy) resulted in an increase of 802 kg/yr and
300 kg/yr, respectively; in 1982, during implementa-
tion (Table 2), the increase dropped to 282 kg/yr.
During 1983, the load measured higher above the
dairy than below it; from 1984 to 1990, total phos-
phorus loads were 196, 516, 223, 301, 69, 35, and 0
kg/yr, respectively.
304
-------
R. LOVELESS, T. NELSON, & H. JUDD
Table 1.—Average of total phosphorus .concentration
(mg/L) and total load (kg/yr) from sites on Snake
Creek before RCWP implementation (1980-81).
1980
1981
STATION NO.
01
02
03
04
mg/L
.018
.026
.039
.082
kg/yr
163
739
737
1,539
mg/L
.019
.023
.027
.200
kg/yr
107
335
486
786
Table 2.—Average of total phosphorus concentration
(mg/L) and total load (kg/yr) from sites on Snake
Creek during RCWP implementation (1982).
1982
STATION NO.
01
02
03
04
mg/L
.034
.053
.043
.075
kg/yr
308
1,270
1,233
1,515
Even though some water quality impacts oc-
curred during 1985 and 1987, data shows that the
average total phosphorus load before implementa-
tion was 551 kg/yr, whereas following implementa-
tion, the average phosphorus load was 217 kg/yr.
The following observations were made by com-
paring average phosphorus concentrations above
and below the dairy:
1. Before implementation (1980-81), the average
total phosphorus concentration above the
dairy was 0.033 mg/L; the average concentra-
tion of total phosphorus below the dairy was
0.141 mg/L. The difference between those
average concentrations is 0.108 mg/L
2. Following implementation (1983-90), the
average total phosphorus concentration
above the dairy was 0.044 mg/L and the
average concentration below the dairy was
0.057 mg/L. The difference between those
average concentrations is 0.013 mg/L.
Tables 4, 5, and 6 show the average concentra-
tion of total phosphorus as well as the total phos-
phorus load on Huffaker Ditch between 1980 and
1990. During 1980, the average phosphorus load
from Vincent's Dairy reached a high of 553 kg/yr
(Fig. 1). During implementation, that load decreased
to 138 kg/yr in 1982 (Fig. 1). Following implementa-
tion (1983 through 1990; see Table 6 for average con-
centrations), the phosphorus yearly load to Huffaker
Ditch from Vincent's Dairy was 67 kg, 74 kg, 16 kg,
63 kg, 14 kg, 0 kg, 11 kg, and 4 kg, respectively (Fig.
1). In 1988, Vincent's Dairy participated in the
Federal dairy buy-out program and dairy cattle were
removed from that site.
Table 4.—Average of total phosphorus concentration
(mg/L) and load (kg/yr) from sites on Huffaker Ditch
before RCWP implementation (1980-81).
1980
1981
STATION NO.
mg/L
kg/yr
mg/L
kg/yr
05
06
.029
.200
99
652
.050
.202
184
600
Table 5.—Average of total phosphours concentration
(mg/L) and total load (kg/yr) from sites on Huffaker
Ditch during RCWP implementation (1982).
1982
STATION NO.
mg/L
kg/yr
05
06
.200
.256
703
841
Similar water quality benefits are being seen at
sites near dairies and feedlots participating in the
Clean Lakes Program. Wasatch County's aggressive
enforcement of planning and zoning regulations has
also been a factor in improving water quality.
Groundwater monitoring was not an original
component of the Snake Creek RCWP water quality
monitoring program. Beginning in 1986, concerns
were expressed that BMPs used to improve surface
water quality might only be rerouting, storing, and
delaying transport of nutrients to the reservoir.
These opinions were based on a growing under-
standing of the makeup of the unconsolidated Heber
Valley fill and the quality of groundwater found
there.
Jeppson et al. (1991) from Utah State University
installed vadose zone samplers near the Pride Lane
Table 3.—Average concentration of total phosphorus (mg/L) and total load (kg/yr) from sites on Snake Creek fol-
lowing RCWP implementation (1983-90).
1983 1984 1985 1986 1987 1988 1989 1990
STATION .
NO. mg/L kg/yr mg/L kg/yr mg/L kg/yr mg/L kg/yr mg/L kg/yr mg/L kg/yr mg/L kg/yr mg/L kfl/yr
01
02
03
04
.027 442
.047 2,360
.059 2,275
.053 1,675
.038 754
.041 1 ,546
.050 1,665
.058 1 ,885
.026
.029
.029
.048
484
924
869
1,397
.063 1,051
.041 1 ,279
.058 1,593
.069 1,816
.033
.027
.036
.061
166 .022
376 .039
523 .032
824 .052
80
599
275
344
.015
.021
.038
.039
41
225
238
273
.023 65
.023 219
.028 156
.023 95
305
-------
Proceedings of National RCWP Symposium, 1992
Table 6.—Average concentration of total phosphorus (mg/L) and total load (kg/yr) from sites on Huffaker Ditch
following RCWP Implementation (1983-90). •' - : ' • '-
1983
1984
1985
1986
1987
1988
1989
1990
STATION
NO.
05
06
ng/L kg/yr mg/L kg/yr mg/L kg/yr mg/L kg/yr mg/L kg/yr mg/L kg/yr mg/L kg/yr mg/L kg/yr
.057 120 .048
.088 187 .076
110 .030
184 .043
69 .094 172 .047
85 .146 235 .056
74
88
.042
.039
45 .028
39 .032
21
32
.029 22
.024 26
600
550
500
450
4OO
35O
3OO
250
200
15O
10O
5O
O
ieSO 1981 19SS 1983 1984 1985 1 9SB 1 987 1988 1989 1 99O
YERR
Figure 1.—Total phosphorus load from Vincent's Dairy.
Dairy lagoon at depths of 54, 70, and 90 inches to
determine if this waste pond was contaminating un-
derlying groundwater. Only small amounts of water
were extracted even though the samplers were
within the free surface seepage zone that would
theoretically exist if the porus media below and sur-
rounding the lagoon were homogenous. These dry
soil conditions suggested that seepage from the
lagoon into the groundwater was minimal. In addi-
tion, water from the lagoon was observed flowing
into the field immediately south of that pond.
Vadose samplers were also installed in a similar
manner near a liquid waste lagoon at a second dairy
that was not part of the RCWP project. This study
determined that most of the nutrients that flowed
into that lagoon probably moved into the
groundwater system over time. This could be a con-
cern for future lagoons, which may have to incor-
porate an impervious liner.
Conclusion
The Snake Creek RCWP is a dramatic success. Phos-
phorus concentrations have steadily decreased in
the Snake Creek drainage since the project was im-
plemented in 1982. The goals of the project were ac-
complished because management practices were
assertive and dealt with the problem directly. The
various BMPs have included manure bunkers, liquid
waste lagoons, off-stream watering systems, fencing
of streams, piping of ditches, and fertilizer manage-
ment practices.
Participants are satisfied with the results. Others
who did not participate in the original RCWP ob-
served the success of the project and decided to take
part in similar projects through the Clean Lakes Pro-
gram.
Deer Creek Reservoir is responding to the
project's broad-based approach to improving and
protecting the water quality in the watershed. As a
result of the RCWP project, the beneficial uses of the
water are better protected.
References
Jensen, LJ. 1986. How people matter in nonpoint cleanup. EPA J.
12(4): 3-4.
Jeppson, R.W., J. Mclean, C.G. Clyde, and S. Korom. 1991. Studies
related to nutrients entering groundwater from the Heber
Valley sewer farm and dairies. Civil Eng. Dep., Utah State
Univ., Logan.
306
-------
R LOVELESS, T. NIELSON, & H. JUDD
Mountainland Association of Governments. 1979. Areawide water
quality management plan for Summit, Utah, and Wasatch
Counties. Mountainland Tech. Rep. No. 16. Provo, UT.
Rushforth, S.R. 1988. Algal floras from Deer Creek Reservoir,
Wasatch County, Utah 1987. Brigham Young University,
Provo.
Sowby and Berg, Consultants. 1984. Deer Creek Reservoir and
Proposed Jordanelle Reservoir Water Quality Management
Plan. Wasatch County, UT.
U.S. Department of Interior. 1979. Central Utah Project, Bon-
neville Unit, Municipal and Industrial System, Final Environ-
mental Statement. FES 79-55. Upper Colo. Reg. Off., Salt
Lake City.
Utah Department of Health. 1984. State of Utah Deer Creek Reser-
voir Phase I Clean Lakes Study. Bur. Water Pollut. Control,
Salt Lake City.
Wann, D. 1986. A "Fitting Solution" at Snake Creek, Utah. EPA J.
307
-------
-------
Information and Education —
Lessons Learned from RCWP
Bud Stolzenburg
University of Nebraska
Cherry County Extension Service
Valentine, Nebraska
ABSTRACT
The value of information and education (I&E) to the administration of Rural Clean Water Program
(RCWP) projects was recognized at a very early stage. I&E was identified as a distinct component
in planning and was to be an integral part of each project. In actual practice, its role and function
varied considerably among the various RCWP projects, possibly because the RCWP concept was in
its infancy. Many other factors also influenced the evolution and scope of I&E; still, after 10 years of
practical experience with the projects, I&E committees have learned much and can make definite
recommendations based on their experience. A significant lesson is the proper distinction between
information and education. Most I&E committees did an excellent job with information but may
need to re-examine the role of education in future water quality projects. This paper will consider
how information and education are incorporated into RCWP projects, with an emphasis on the ap-
propriate separation and coordination of these components. The author contends that the ultimate
goal of water quality demonstration projects is behavioral modification, which places great em-
phasis on education.
The value of information and education (I&E)
was recognized early by RCWP project
managers and others even though no one
could predict the outcome of I&E efforts, and no one
had any prior experience with water quality projects
to give us direction. We didn't know how the fully
developed, mature water quality project should or
would look—it is difficult to evaluate something that
hasn't happened yet
Many factors shaped and influenced the way
that I&E efforts developed in the various projects.
These factors relate to the widely differing environ-
ments of the RCWP projects. Environmental dif-
ferences affecting each project were physical and
cultural, economic and social.
Some projects were designed to address surface
water concerns; others to address groundwater
problems. Project size varied greatly. The Long Pine
Creek RCWP project involved a watershed of
325,000 acres; others dealt with problems confined
to a much smaller physical area. The number of par-
ticipants, the size and nature of the individual farm-
ing operations, the proximity of industry, and local
cultures and traditions influenced the character and
variety of the RCWP projects and their goals and ob-
jectives.
We also need to remember that some best
management practices (BMPs) may be relatively in-
expensive and quite easy to implement, while others
may be costly and unable to yield an immediate, evi-
dent economic return. Some RCWP projects as-
signed full-time I&E responsibility to their staffs; in
other cases, the workload was shared among several
people or handled as an extended responsibility by
one person. Some projects had an I&E component
309
-------
Proceedings of National RCWP Symposium, 1992
throughout their duration; others had initial support
only, while still others had I&E on an as-needed
basis. A significant difference also existed between
projects with a comprehensive monitoring and
evaluation component and those without it.
The definition of I&E also varied from project to
project. In some cases, a vague position description
made it difficult to have a clear understanding of
responsibility. But now, as I&E staff view the RCWP
projects in retrospect, they can see that they have
learned a great deal about I&E.
The I&E involvement in the RCWP has been a
learning process—an exercise in education. It offers
10 years of valuable hands-on experience that can en-
hance the prospect of success in future water quality
projects and help us avoid pitfalls. The lessons
learned about the role of information and education
can be applied to many parallel situations in which
the end result is behavior modification. However, our
concern here and now is to look at the lessons we
have learned in terms of water quality.
The following ideas were shared at an RCWP
workshop in South Dakota in September 1990. The
meeting included a brainstorming session for the
I&E component in future water quality projects.
• Appropriate personnel are needed to contact
producers. Ongoing I&E responsibility should
be assigned to one agency that directly con-
tacts producers, with other agencies providing
input and support.
• Initial I&E efforts are very important, but I&E
needs funding throughout a project's life.
• I&E funding needs to be less restricted so that
demonstration projects and field days receive
sufficient funding. Both are helpful ways to
disseminate information.
• I&E needs to be the RCWP project's primary
component. I&E is important; it sells a project
to farmers, informs the public, and provides
interagency training.
• I&E expertise is needed to coordinate project
efforts, assess efficiency of methods, evaluate
progress, and recommend changes.
• Extension Service offices can be the lead
agencies for I&E, but support is also needed
from other agencies involved in the project
• I&E includes communicating and transferring
information among agencies at each level—
Federal, State and local—and across levels.
• The local coordinating committee is the most
important component for ensuring that an
RCWP I&E effort succeeds.
Many other suggestions were expressed at this
meeting but these few suffice to illustrate I&E value.
Of course, these observations take nothing away
from other aspects of a water quality project. Rather,
they attach appropriate significance to the role of
I&E.
We have traditionally included "information" and
"education" in the same phrase, knowing that they
were closely related. Perhaps we did not give ap-
propriate credit to their differences. It is one thing to
know the facts and be aware of the information; it is
quite another to absorb this information usefully.
I&E Observations and Ideas
In the last 10 years, the importance of I&E was
recognized but not its modus operandi. We didn't
know how to handle the education part, or perhaps
we just didn't plan to handle the education part. Any
teaching must be based on sound information that
has been evaluated through solid research.
We did an excellent job with information in
terms of collecting, organizing, and publishing it. We
learned a great deal about handling water quality in-
formation.
At the same time, one of the most important
things we learned from RCWP is the difficulty of
taking the step from knowledge to adoption of best
management practices. Taking this step mandates
that we enter an "educational" mode.
When we initially staffed the RCWP projects, we
hadn't had the opportunity and experience to ap-
preciate how important it is to select staff carefully
with this particular concern in mind. Consequently,
we selected people who were good information
people, knowledgeable about their subject, and good
at man-aging data. And, the people we selected did a
good job at what they were asked to do.
One important result of the RCWP is the amount
and kind of information that we collected, organized,
evaluated, categorized, and summarized. Many fu-
ture decisions can be based on the significant data
that RCWP I&E staffers gained through observation
and research.
A second important result is the lesson we
learned about "teaching," that is, about effecting be-
havioral modification. It would seem to me that we
should have some people who are trained to teach.
310
-------
B. STOLZENBCfRG
RCWP from an Educational
Perspective
By comparing RCWP to a school system, we can
draw several parallels between the lessons we have
learned and the classroom. The following table illu-
strates the parallels I have in mind.
SCHOOL SYSTEMS
RCWP PROJECTS
Superintendent needed to
work with the whole district.
Project coordinator needed
for overall supervision.
Big school systems are dif-
ferent from one-room schools.
Project area size determines
the program approach.
Not all students have the
same needs.
Different projects in different
areas have different needs.
Classrooms should have a
maximum of 20 students.
The number of project
participants governs staff
numbers (or vice versa).
A one-on-one relationship
with a student is important in
the learning process.
One-on-one contact with
producers is important in the
adoption of change.
Students who understand the
benefits of education learn
more readily.
People who appreciate the
long-term benefits are more
ready to adopt change.
Diverse cultural backgrounds
may challenge the teacher.
Local culture and tradition
may affect the project
approach.
"That's the superintendent's
job."
Responsibilities must be
defined and understood.
Society has a common goal
to educate children.
Agency cooperation and
partnerships contribute to an
effective program.
Learning is greatly enhanced
for children who are actively
involved with hands-on
participation.
On-farm demonstrations can
increase project participation.
Good teachers use long- and
short-term lesson plans.
Each project needs a definite
plan of action—with modifica-
tions if needed.
School boards need to
budget for the whole year.
I&E needs to be funded for
the duration of a project.
Report cards are distributed
quarterly.
I&E includes systematic and
regular project evaluation.
Good public support helps a
school system.
Good public support helps a
water quality project.
Teachers have opportunities
for continuing education.
Project leaders enhance their
skills through appropriate
trainingand workshops.
Teachers use a variety of
aids and techniques to
improve the learning atmos-
phere.
Good project I&E requires a
variety of methods and
approaches.
"I can't afford to go to
college."
The economics of particular
BMPs may demand unique
methods for their adoption.
These comparisons illustrate many of the educa-
tional aspects of water quality protection and en-
hancement. They are features that need to be in-
cluded and coordinated for project success—par-
ticularly for long-term success.
Conclusion
A definite relationship exists between information
and education. They belong together and comple-
ment and supplement each other, but they are dis-
tinct components of RCWP projects and each is
important. They cannot be merged to the point that
the identity of either is lost.
As it relates to information and education, the
RCWP's biggest present challenge is the appropriate
transfer of "lessons 'learned""'to present and future
projects. I believe we are doing well with informa-
tion. How will we handle the education? The chal-
lenge in many endeavors and not least in the future
of RCWP is to move from theory to reality, from plan
to practice, from information to education.
311
-------
-------
Diversity of Information and Education
Help Obtain Goajs for
Double Pipe Creek RCWP Project
David L. Greene
University of Maryland Cooperative Extension Service, Carroll County
Westminster, Maryland
ABSTRACT
A wide diversity of communication methods were used to promote the Double Pipe Creek Rural
Clean Water Program (RCWP) project .in Carroll County, Maryland. These methods were
developed at the project's start by the Information and Education (I&E) subcommittee and at-
tached to a five-year timetable. The subcommittee met semiannually in the early years, then an-
nually the final two years. As a result of the I&E efforts, the following findings were made relative
to the promotion of the project (1) an initial sharing of ideas among the 21 RCWP projects would
have provided additional mechanisms to reach potential participants; (2) the excellent cooperation
and support developed between sister agencies, farm organizations, and public officials helped
reduce duplication of efforts; (3) having I&E funds handled through the State Cooperative Exten-
sion Service enabled I&E projects to be approved and funded on a more timely basis; (4) a greater
effort in developing one-on-one contacts at the project's .end would have resulted in a higher num-
ber of participating contracts; and (5) guidelines should have been developed at the project's begin-
ning to better track and evaluate the results of I&E initiatives.
The primary objective of the Double Pipe
Creek Rural Clean Water Program (RCWP)
project was to improve water quality in the
drainage basin within Carroll County and down-
stream. The major project emphasis was to manage
livestock waste rather than concentrate on sediment
reduction. To accomplish this objective, the goal was
to have an acreage equal to 50 percent of the critical
area under RCWP contract by the end of the third
year.
The Information and Education (I&E) subcom-
mittee goals were to accomplish the project goals
and inform the general public and government offi-
cials about the progress of the RCWP effort. This
paper addresses the following components of the
I&E effort in the Double Pipe Creek RCWP project:
• Plan of Action to accomplish I&E goals,
• I&E subcommittee activities conducted to
implement the Plan of Action,
• adjustments to the initial I&E plan of action,
• factors influencing the rate of program
participation,
• evaluating the success of the I&E effort, and
• recommendations to improve future water
quality demonstration projects.
Plan of Action
The long range objectives of the I&E program were
divided into two stages. The first was the awareness
stage: the I&E subcommittee wanted to inform the
general public and the farming community about the
project. The nature of the project and its goal to im-
prove the water quality in farm runoff needed to be
communicated along with information about how the
project would be administered and funded. The ob-
jective in the second stage of the I&E subcom-
313
-------
Proceedings of national RCWP Symposium, 1992
mittee's plan was to encourage farmer and land-
owner participation in the RCWP project.
Even though most members had little prior ex-
perience in creating a large educational effort, the
I&E subcommittee developed a diverse number of
promotional ideas to accomplish the designated ob-
jectives, among them:
• holding a press conference to announce the
approval of the Double Pipe Creek RCWP
project,
• distributing an informational flyer to describe
the general outline of the project,
• sponsoring three community information
meetings at different locations within the
Double Pipe Creek area to discuss the project
and answer questions from the farm com-
munity and the general public,
• preparing newsletters and news releases to
encourage farmers to sign up for the project
and to describe agricultural best management
practices (BMPs) that could be implemented
to help farmers meet water and soil conserva-
tion goals, and
• creating informational and media material to
localize the project in Carroll County and
Maryland, promote the project among area
farmers, and inform the farming community
and the general public of the importance of
improving water quality.
Media and informational materials in the last
named category needed to be carefully planned and
diverse to appeal to a variety of age and interest
groups. Some were promotional; others, more tech-
nical, and some were intended to appeal to school
children. Thus, for example, the committee's plans
included a slide and tape series on the Carroll Coun-
ty RCWP and the BMP components of the project; a
booklet called WATER "The Basics of Life" for dis-
tribution with other RCWP literature to county
school children; and an album of sample BMPs to be
developed by Soil Conservation Service (SCS) tech-
nicians to show participants what these practices en-
tailed. Other projects planned by the I&E
subcommittee are detailed in the implementation
section of this paper.
Activities Conducted to
Implement the Plan of Action
A wide array of I&E activities were conducted during
the project's first six years to encourage participation
in the RCWP project, including the following:
Press Conference. This meeting was held
when the Double Pipe Creek project first
received approval.
Tours. Two waste management tours were ar-
ranged so that farmers could observe com-
pleted BMPs. Another tour was held for
county, State, and Federal administrators to
observe the progress of the Double Pipe
Creek project.
Television. Two telecasts on the Double Pipe
Creek Project were aired on public TV (Chan-
nel 67).
Radio. Several RCWP announcements were
made on the local radio station by the county
extension agent during his weekly radio pro-
gram and by a soil conservationist in special
; public service announcements.
i News Releases. Announcements were writ-
ten and printed in local county newspapers
detailing the Double Pipe Creek meetings,
tours, and field days.
i News Articles. News and feature articles
were printed in local county newspapers and
the DelMarVa Farmer.
i Newsletters. Articles on the Double Pipe
Creek Project appeared many times in county
Agricultural and Stabilization Service (ASCS)
newsletters, Soil Conservation District (SCO)
newsletters, and Cooperative Extension Ser-
vice (CES) "Farm Notes."
i Informational Displays. Models were built
for the Carroll County Fair, CES mid-winter
meetings, and the Maryland State Fair.
i Slide Tape Show. A general information
series on RCWP was created and has been
used numerous times at meetings, civic clubs,
and local feed and machinery stores.
i Picture Album. Photographs of sample
BMPs were assembled and made available to
all agencies working in the project area and
used as a review with prospective applicants.
i Metal Signs. The local coordinating commit-
tee designed and distributed signs to par-
ticipants after they had completed the first
BMP in their contract. These signs read
"RCWP Cooperating Farm."
i Communication. Throughout the project
communication has been encouraged. In par-
ticular, vocational agriculture teachers in the
county high schools were encouraged to bring
students to field days to see completed BMPs
314
-------
D.L. GREENE
and to discuss the Double Pipe Creek project
in their classes.
• Individual Farm Visits. Landowners in criti-
cal areas were visited and encouraged to par-
ticipate in the Double Pipe Creek project. Soil
Conservation Service supervisors mounted a
campaign to visit various farms in the project
with known soil conservation problems.
• Meetings. A meeting was held with contrac-
tors who build soil conservation BMPs to in-
form them of the project, and topics
describing the benefits of participating in the
Double Pipe Creek project were included on
the agenda of Cooperative Extension Service
Mid-winter Meetings.
• Plaques. Plaques were presented to RCWP
cooperators who had completed all of their
BMPs.
• Mechanical Groundwater Model. A display
was used to show the interaction between
ground and surface water and the effects of
human activity on that water. Demonstrations
were given to school classes, service clubs,
and farm meetings.
• Manure Spreader Calibration Demonstra-
tions. Technical demonstrations were in-
cluded during field days and tours.
• The RCWP Newsletter. This publication was
initiated and mailed to all farms under con-
tract on a quarterly basis.
The use of published and printed information
augmented other types of I&E activities and adver-
tised and promoted special events. Published pieces
included
• Booklets
• Best Management Practices Tour—
December 8,1981
• Animal Waste Management Tour—
November 16,1982
• Annual Report and Conservation
News—1984
• Rural Clean Water Program: Double Pipe
Creek Project Tour—March 11 and 12,1985
• Best Management Progress Tour—July 28,
1988
• Flyers
• Double Pipe Creek Rural Clean Water
Project
• Questions and Answers About Soil and
Water Conservation
• Rural Clean Water Program Double Pipe
Creek Water Quality Management
• RCWP Update Session— October 1982
• Rural Clean Water Demonstration Day —
November 7, 1984
• Best Management Progress Tour— July 28,
1988
• Invitation
• Best Management Progress Tour — July
•28, 1988
• Newspaper Articles
• Contracts Available for Pipe Creek Area —
Farm News: Clean Water Contracts
Open— July 25, 1980, Carroll County Times
Double Pipe Creek: Water Cleanup Project
Funded— August 9, 1980, The Evening Sun
Ceremony to Mark First Rural Clean Water
Contract— September 16, 1980, DelMarVa
Farmer >
1st in Nation— September 18, 1980, Carroll
County Times •',-•'
Project Reduces Animal Wastes in Little
Pipe— September 18, 1980, Carroll County
Times
Father-Son Team Get First Check in
RCWP— September 23, 1980, DelMarVa
Farmer
Carroll Gets First Facility— October 1980,
Maryland Farmer
Water Meetings Begin — December 10,
1980, The Evening Sun
Pure Water Pilot Program Launched —
December 10, 1980, Carroll County Times
Farm Briefs: Clean Water — December 30,
1980, DelMarVa Farmer
Rural Clean Water Program: It's Helping to
Manage Waste in Carroll County —
December 15, 1981, DelMarVa Farmer
Double Pipe Creek: A New Outlook on
Handling Waste— April 28, 1981, DelMarVa
Farmer
"Clean Water" the Byword as Double Pipe
Creek Farmers Spruce Up Their
Operations — November 2, 1982, DelMarVa
Farmer
315
-------
Proceedings of National RCWP Symposium, 1992
• Farming: Rural Clean Water Project Right
On Schedule—November 2,1982, Carroll
County Times
• Waste Management Tour Tuesday—
November 15,1981, Carroll County Times
• Pits Help End Runoff—April 25,1983,
Carroll County Times
• Rural Clean Water Program—Summer
1983, Carroll County Times
• Ag Briefs: Rural Clean Water Program—
October 26,1984, Carroll County Times
• Carroll County Farmers Face Tough
Decision—October 30,1984, DelMarVa
Farmer
• Maryland to Hold BMP Open House
Wednesday—October 30,1984, Lancaster
Fanning
• RCWP Day Set in Carroll—November 6,
1984, Frederick Post
• Carroll Farmers Join Clean Water
Program—November 6,1984, Carroll
County Times
• Farm/Business: The Clean
Way—November 19,1984, Carroll County
Times
• Farmers Urged to Apply for RCWP Plan-
November 20,1984, Frederick Post
• Extension Service to Expand Offices—
Carroll County Times
• Folders
• Rural Clean Water Program: Double Pipe
Creek—Water Quality Management
• Fact Sheet
• University of Maryland developed a
Grassed Waterway Maintenance fact sheet.
Adjustments to the Initial I&E
Plan of Action
The I&E subcommittee's plan, like any long-range
plan, needed certain midcourse adjustments during
the implementation process to compensate for pro-
gram changes and variations in program participa-
tion. The I&E subcommittee was able to make these
adjustments without dramatic changes to the initial
plan of action.
The I&E subcommittee's initial goal was to meet
semiannually to plan various I&E projects and
evaluate work accomplished to date. This schedule
was adhered to the first six years. After the sign-up
period had lapsed, the committee met annually but
usually only the agency people attended.
Additional educational activities and promotions
were added over the course of the project's first six
years. These additions included recognition of farms
that had completed at least one BMP: farmers were
presented metal signs signifying that they were
"RCWP Cooperating Farms." During a particularly
slack period in sign-ups, a meeting was arranged
with contractors who construct BMPs to inform
them of the project and to see if they could help
stimulate interest.
Near the end of the sign-up period, a list was
developed of farms that had been targeted as
priorities but had not shown any interest. These tar-
geted individual farmers were visited by SCS, ASCS,
County Extension, or SCD personnel. In addition,
vocational agriculture teachers in the high schools
were kept informed about the RCWP project and en-
couraged to discuss the project with their classes. All
adjustments to the original I&E goals were made
near the end of the sign-up period to help increase
participation.
Factors Influencing the Rate
of Program Participation
Obviously, economic and political factors play a sig-
nificant role in the willingness of farmers to par-
ticipate in any government cost-sharing program.
During the 1980s, the project area experienced three
major droughts that placed a severe economic
hardship on farmers and, to some extent, limited
their willingness to participate in this cost-sharing
program. But also during the 1980s, the environmen-
tal concern for cleaning up the Chesapeake Bay
created an awareness in the farm community that a
problem existed with nonpoint source pollution com-
ing from farms and that farm owners would be ex-
pected to help solve the problem.
Other State and Federal government programs
affected the rate of farmer participation in the RCWP
effort either positively or negatively; some programs
are still in existence, and others have gone. The fol-
lowing programs are large enough to have had some
effect on the project area.
Maryland Agricultural Land
Preservation Program
For a farm to be accepted in the Maryland Agricul-
tural Land Preservation Foundation, a conservation
316
-------
D.L. OREENE
plan must be developed before easement rights are
sold. Since 1980, 21,473 acres have been accepted
into the Maryland Agriculture Land Preservation
Program in the Double Pipe Creek project area. By
enrolling in this program, many farmers have com-
mitted to remain in farming. Therefore, these land-
owners are willing to make a long-term commitment
to water quality through RCWP. The Maryland
Agriculture Land Preservation program has had a
very positive effect on RCWP participation.
Food Security Act of 1985
Preliminary information indicates that approximate-
ly 75 percent of the land in Carroll County is highly
credible based on the 1985 Food Security Act
criteria. The SCS district conservationist estimated
that 25 percent of the cropland acreage will need
cropping system modification or the establishment
of additional conservation measures to bring the soil
loss each year within the established acceptable
level.
The emphasis in this act on the need to establish
conservation plans if farmers want to participate in
government programs had a positive effect on the
willingness of farmers to participate in the RCWP
program and eventually on water quality in the
project area.
Conservation Reserve Program
Carrpll County has limited participation in the Con-
servation Reserve Program because of high land
values and investment opportunities. Although par-
ticipation in this program is not widespread, 75 per-
cent of the land that is enrolled in the county
(approximately 700 acres) is in the RCWP project • 1979
area. By taking highly credible cropland out of
Table 1.—Carroll County, Maryland, dairy summary: 1980 to 1990.1
production, the Conservation Reserve Program had
an effect on improving water quality, but a slight one
because of the limited acreage enrolled in the pro-
gram.
Dairy Situation
Although the dairy industry is the most prevalent
type of farming in the county, the 1980s saw a 33 per-
cent reduction in the number of dairy producers. As
indicated in Table 1, not only did the number of
dairy producers decline but the wholesale price of
milk remained flat or declined during the sign-up
period and did not keep up with the increase in the
inflation rate. Although these are countywide statis-
tics, a vast majority of the dairy farms are in the
project area.
The economics of the dairy situation during the
1980s tended to have a negative effect on the willing-
ness of dairymen to participate in the RCWP project.
Although most dairy farms were located in critical or
high priority areas, many dairymen just did not have
the money to match RCWPs cost-sharing assistance.
Therefore, a large percentage of I&E efforts focused
on encouraging these dairy farmers to participate in
the program because of its high potential to affect
water quality.
Rate of Program Participation
The following summarizes several significant events
that occurred during the past decade within the
Double Pipe Creek project. The summary also
shows that the rate of participation was fairly consis-
tent between 1981 through 1986.
- Proposal for Double Pipe Creek
project submitted.
YEAR
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
COW NUMBERS
17,700
18,000
18,100
18,800
19,500
19,800
20,000
17,500
13,500
12,900
12.9004
MILK PRODUCED
(1.000LBS.)
252,200
259,800
262,000
262,800
252,200
271,000
262,900
249,300
240,000
235,000
241,900s
NUMBER OF
PRODUCERS2
293
289
284
266
247
238
222
215
215
197
192
WHOLESALE MILK DOLLARS
PER HUNDRED-WEIGHT (CWT)3
13.60
14.30
14.00
14.00
14.00
13.10
12.90
13.10
13.00
14.50
14.90
1 Source: Maryland Agricultural Statistics Summaries from 1979-89.
2Souroe: Federal Milk Market Administrator, Middle Atlantic Milk Marketing Area. All figures are for the December reporting period.
3Source: Average price, F.O.B. or receiving station.
4Source: Estimates.
5Source: Estimates.
317
-------
Proceedings of National RCWP Symposium, 1992
• 1980 - Proposal approved for Double Pipe
Creek Project
- The Local coordinating committee
determined that "any farm in the
project area having a critical impact on
water quality regardless of the
distance to streams, will be
considered a high priority request."
- Cost-share earned: $8,953.00
- Contracts approved: 5
• 1981 - Cost-share earned: $176,518
- Contracts approved: 25
• 1982 - Cost-share earned: $283,642
- Contracts approved: 25
• 1983 - Versar, Inc., completed baseline data
Stage I of monitoring.
- Cost-share earned: $316,515
- Contracts approved: 18
• 1984 - Signs given to participants completing
BMPs.
- Critical area definition was questioned
by the national coordinating
committee.
- Law was passed requiring that a
conservation plan accompany
easement rights bids on land entering
the Maryland Agriculture Land
Preservation Program.
- State agreed to continue monitoring
when Versar contract expired.
- Demonstration Day held (85 people in
attendance).
- Cost-share earned: $187,815
- Contracts approved: 21
• 1985 - BMP-15 and BMP-16 added to the
project.
- Contracting period extended through
December 31,1986.
- Cost-share earned: $502,836
- Contracts approved: 19
• 1986 - Extension agent hired to write
nutrient management plans and to
implement BMP-15 and BMP-16.
- Cost-share earned: $382,125
- Contracts approved: 15
• 1987 - Cost-share earned: $489,773
• 1988 - Lease Brothers received special
approval for a contract in the
expectation that monitoring on their
farm would benefit the project.
- Best Management Progress Tour was
held to demonstrate progress and
practices completed (150 participants).
- Cost-share earned: $236,014
- Contracts approved: 1
• 1989 - Cost-share earned: $168,033
• 1990 - Cost-share earned: $136,464
Evaluating the Success of the
I&E Efforts
The Double Pipe Creek RCWP project's overall goal
of having 50 percent of the designated critical areas
in the watershed under contract was not quite met,
However, the individual BMP goals for the project
were met. According to the results of the monitoring
program conducted so far, water quality improve-
ment goals have been achieved for phosphorus but
not for nitrogen. A continuing effort to reduce
nitrogen levels in streams coming from farm land
will be made through the efforts of Maryland's
Nutrient Management Program conducted by the
Cooperative Extension Service.
The goals and objectives of the I&E effort were
completed as originally outlined. The I&E subcom-
mittee was very active throughout the sign-up period
and received excellent cooperation from all govern-
ment agencies. Extension provided the leadership in
organizing and supervising the I&E effort. Funding
for the educational effort was administered through
the Maryland Extension Service.
The plan developed to inform farmers was fol-
lowed completely with only minor changes and addi-
tions. Just about every type of creative method of
disseminating information and motivating farmers to
participate was tried in this program.
Of all the promotion efforts used, obviously
some were more effective than others. The tours had
excellent participation and always resulted in an in-
creased rate of contract sign-up. Feature and news
articles in the local and county papers had the most
positive effect by constantly keeping the project
before potential participants. Individual farm visits
318
-------
D.L GREENE
near the end of the program helped recruit farmers
from highly critical areas. All other educational ef-
forts had less impact but were useful in encouraging
participation to some degree.
Many of these same innovative methods of draw-
ing attention to the need for improving water quality
and promoting conservation have been implemented
in other water quality projects such as the Piney-Al-
loway Creek Watershed project (also located in Car-
roll County, Maryland).
Recommendations to Improve
Future Water Quality
Demonstration Projects
The I&E effort for the Double Pipe Creek RCWP
project was implemented primarily during the
program's first six years. Our efforts centered on
securing participation in the program. Now that the
project is complete, it is easy to see that some things
should have been done differently, not at all, or more
extensively.
Recommendations to improve future water
quality demonstration projects include the following:
1. Although a wide array of information and
• education activities were used to promote the
Double Pipe Creek project in Carroll County,
a sharing of ideas with some of the 21 other
' RCWP projects could have provided addition-
al ideas during the first four years. RCWP
project sharing should be included in future
I&E planning.
2. In the. Double Pipe Creek project, excellent
cooperation existed between all the agencies.
The local coordinating committee and I&E
subcommittee had good participation from
local farm organizations and public officials.
Having I&E funding handled separately
through the Cooperative Extension Service
worked extremely well and enabled projects
to be funded quickly. This funding method
should be continued.
3. As the sign-up period was nearing the end,
more time should have been devoted to
making one-on-one contacts with potential
participants. A greater effort during-this time
could have resulted in a higher number of
contracts. One-on-one methods should be
more vigorously pursued in the initial stages
of future I&E efforts.
4. A separate budget line should be developed to
provide for the printing of the annual and 10-
year reports.
5. The original guidelines for the project should
indicate the expected contents of the yearly,
10-year, and close-out reports.
6. A tracking and record system should be
developed at the beginning of the project to
facilitate easier and more pertinent data col-
lection. In addition, the forms used to report
the data should be designed early in the
project.
Conclusion
All government agencies worked closely with the
local coordinating committee and the I&E subcom-
mittee under the CES leadership. It is impossible to
separate activities performed by individual agencies
because all projects were coordinated under the
aegis of the I&E subcommittee; however, the bulk of
work was done by CES, ASCS, and SCS. The local
coordinating committee and,the State coordinating
committee approved and helped promote various
projects. The agencies responsible for water quality
monitoring had a very limited role in the I&E work.
The I&E subcommittee made a substantial contribu-
tion to the promotion and eventual success of the
Double Pipe Creek RCWP project.
=319
-------
-------
Nutrient Management Educational
Initiative: Using Demonstration and
Research Plots and the Penn State
Nitrogen Quick Test in
the Upper Conestoga RCWP
Robert Anderson
Penn State Cooperative Extension
Lancaster, Pennsylvania
ABSTRACT
The Upper Conestoga Rural Clean Water Program project was started in the late 1970s to address
water quality problems, including excessive runoff and soil erosion. By 1985, U.S. Geological Sur-
vey monitoring identified nutrient over application and leaching as the major contributors to high
nitrates in surface and groundwater. In January 1986, Penn State Cooperative Extension began
developing nutrient management plans for farmers. Between 1986 and 1991, 365 farms developed
nutrient management plans. An average of 10 demonstration and research plots were conducted
each year to demonstrate the validity of the nutrient management plans, and a soil nitrogen quick
test was administered (both developed by Penn State Extension). Using a randomized block design
with 3 to 4 replications, plots were scattered throughout the project to help farmers see the effect
of the proposed recommendations on crop yields. Annual reports showing the amount of fertilizer
recommended and its results as measured by yield increases were sent to farmers in the project
area. By allowing farmers to experience these demonstrations, the recommendations made by the
nutrient management plan and soil nitrogen quick test became believable to the farm community.
The Conestoga River drains approximately
477 square miles of land in northeastern and
central Lancaster County and portions of
Lebanon, Berks, and Chester counties (see Fig. 1).
Major streams within the Conestoga watershed in-
clude the Cocalico, Muddy, Mill, and Little Cones-
toga creeks. About 35,000 people live on the
predominantly rural farmland, mostly in rural homes
and small villages. The project area contains 132
square miles (110,000 acres), half of which is used
for farming, while much of the remaining is open
woodland. Approximately 1,250 farms are located in
the project area, each with an average of 52 acres of
cropland. About 55 percent of the cropland (36,000
acres) is planted to corn; 11 percent (7,500 acres), to
various small grain crops (barley, wheat, and oats);
28 percent (18,000 acres), to alfalfa and other hay
crops; and 2 percent (1,500 acres), to tobacco and
other cash crops. Most farms in the project area are
intensive livestock farms.
Area residents depend on water that originates
within the watershed. With the exception of the city
of Lancaster and the borough of New Holland, resi-
dents living within the watershed rely solely on
groundwater.
The topography is rolling with 0 to 40 percent
slopes; slopes on farmed land ranges from 0 to 10
percent. The soils, mostly well-drained, derive from
limestone, with smaller areas of shale and sand-
321
-------
Proceedings of National RCWP Symposium, 1992
Figure 1.—Maps of the Conestoga Headwaters RCWP project area.
stone. Soil depth ranges from 2 to 8 feet. The lime-
stone soils, located near rivers, are bounded on the
north by upland soils formed from Triassic rocks
and on the south by soils derived from ingenues and
metamorphic rock. Average yearly rainfall in the
area is 42 inches, with 22 inches occurring during
the growing season. About 20 inches of the total rain
in the area will evaporate or transpire, 9 inches will
run off, and the remaining 13 inches will percolate
into the soil to help recharge the groundwater sys-
tem.
The farming population, comprised almost en-
tirely of people with German and Swiss heritage, is
often referred to as the Pennsylvania Dutch or Plain
322
-------
R. ANDERSON
People. More than 70 percent of all farm families are
members of the Old Order Mennonite or Amish
churches, which inculcate a close tie to land and
farming. Education is limited to one-room parochial
schools. Students are not educated beyond the age
of 14, and teachers are usually young women with no
formal education. The most conservative believers
use horses and mules for power on the farm and
transportation; some of the more liberal use tractors
on steel wheels, while a few are so liberal as to drive
tractors with rubber tires and cars painted entirely
black (including the chrome). Very few farm families
have radio or television.
The Nutrient Management
Problem
The area's access to major markets (Philadelphia,
Baltimore, Washington, D.C., and New York), the
population's desire or church mandate to farm, and
the limited land available for farming have con-
tributed to dwindling farm size and an increased de-
pendence on livestock as an income source for area
farmers. Families have divided farms several times
to help other family members earn their living on the
land. Livestock numbers have increased by import-
ing feed. Data from the Pennsylvania "Crop and Live-
stock Summary" for 1960,1970, 1980 and 1990 show
this dramatic change on farms in Lancaster County
(seeTables land2).
In 1960, Lancaster County had 482,579 acres of
farm land; however, by 1989, that area had declined
to 304,212 acres—a decrease of 37 percent. In 1960,
33 percent of the total farm land was planted to corn,
but by 1989 that percentage had increased to 52 per-
cent. Increased animal numbers (and manure), less
land to spread the manure on, and an increasing
proportion of land in row crops contribute to in-
creased soil erosion and transport of nutrients to
surface water. Compounding these problems is the
fact that Lancaster County has the fastest growing
population in Pennsylvania. The population is
projected to exceed 440,000 by the year 2000.
Several studies in the area indicate problems of
bacterial contamination and high nitrate levels in
surface and subsurface waters. Agriculture is not the
only nonpqint source of water pollution; other con-
tributing factors include
• residential septic systems,
• land application of sewage sludge, and
• lake eutrophication resulting from large
numbers of wild ducks and geese.
In 1977, as part of a statewide plan for agricul-
ture and earth-moving activities, the Conestoga
River was designated as the watershed with the
greatest potential to be polluted by agricultural non-
point pollution sources. This conclusion was based
on (1) the number of farms, (2) the animal density
on the farms, (3) the percentage of the land used in
growing row crops, and (4) previously identified
water quality problems in the area. Detailed informa-
tion about using commercial fertilizer, manure, and
pesticides was gathered and tabulated for the project
area.
Table 1.—Lancaster County farms: land use changes 1960 to 1989.
CROPS
Corn
acreage:
Oat
acreage:
Barley
acreage:
Wheat
acreage:
Tobacco
acreage:
All hay
acreage:
Total all
acreage:
1960
113,500
9,100
20,800
58,400
27,700
97,000
345,000
1970
142,600
3,600
23,500
36,300
17,270
89,200
330,000
1980
200,700
3,000
8,600
28,500
12,000
81,800
336,150
1989
157,000
2,800
9,700
15,400
8,640
82,000
304,217
%CH.
+38
-69
-53
-74
-69
-15
-12
Table 2.—Lancaster County farms: livestock number changes 1960 to 1989,
ANIMALS
Dairy cows
All cattle
Hogs
Chickens:
Layers
Broilers
1960
64,000
N/A
44,000
3,100,000
7,468,000
1970
67,100
N/A
109,000
3,109,000
18,028,000
1980
94,000
273,700
337,000
7,579,400
42,106,000
1989
97,000
259,000
358,000
8,979,700
55,268,640
%CH.
+52
+714 .
+190
+640
323
-------
Proceedings of National RCWP Symposium, 1992
In June 1982, a 208 assessment survey of 409
farms, called "Assessment of Nonpoint Sources of
Pollution from Agriculture in The Conestoga River
Watershed," showed that an average of over 400
pounds of nitrogen per acre was being applied to
land used to grow corn. In addition to the nitrogen,
the survey showed that farmers were applying 250
pounds of phosphorus and 200 pounds of potassium i
per acre each year to the same land. These nutrients
came from manure or manure plus commercial fer-
tilizer.
Critical sources of pollution identified in the
Rural Clean Water Program (RCWP) funding
proposal included
• animal waste,
• high rates of commercial fertilizer and
pesticides applied to cropland, and
• inadequate erosion and sediment control on
intensively farmed cropland.
The initial project emphasized soil erosion con-
trol as the primary means to improve water quality.
The 208 assessment made the following recommen-
dations to improve water quality:
1. Better management of manure and
commercial fertilizer applications is needed.
2. Fertilizer recommendations should consider
the nutrient value of manure.
3. Since only 29 percent of the farmers
reported having a conservation plan, more
effort is needed to encourage farmers to
develop and implement conservation plans
that include practices, such as terraces,
waterways, diversions, and manure storage,
to reduce soil erosion and to handle manure
properly.
4. Minimum and no-till farming systems should
be promoted to reduce soil erosion.
5. Encourage farmers to incorporate manure as
soon after application as possible.
6. Animal waste storage structures should be
monitored to evaluate groundwater pollution.
Nutrient management was identified as a priority
in the RCWP project application (on page 8 under a
discussion of soil loss):
Soil loss in the critical area has been es-
timated to be approximately 9 tons per acre
per year. One of the goals of this project is to
reduce soil loss to 4 tons per acre. This reduc-
tion should result in a pollution reduction of
approximately 40 pounds of nitrogen (land
Resource Area 148) (ST soil loss reduction x
8 pounds nitrogen/ton = 40 pounds) and 10
pounds of phosphorus (ST soil x 2 pounds/T =
10 pounds) per acre of cropland. This is a cal-
culated annual reduction of 480,000 pounds
of nitrogen and 120,000 pounds of phos-
phorus entering the water in the project area
from the 12,000 acres targeted for treatment.
Although the project application identified a very
close relationship between erosion or runoff and
nutrient management, the major problems as-
sociated with nitrates on page 9 of the 1984 Progress
Report states, "The goal of this project is to reduce
the levels of agricultural pollutants in both ground
and surface waters." When the project was funded,
money was also provided for the U.S. Geologic Sur-
vey (USGS) to monitor several small watersheds
(which were confined to a single farm) and several
streams for water quality improvement that resulted
from conservation practices being installed on the
land. A review of their findings in 1985 showed that
conservation practices used to control erosion would
not to be enough to deal with all water quality
problems in the project area.
Between 1980 and 1985, several research
projects were conducted in the area by Penn State's
Department of Agronomy staff, including Richard
Fox's effort to develop a soil or tissue test to deter-
mine how much nitrogen (if any) should be added to
corn as a sidedress application to optimize yields;
Dale Baker's work on corn-growth and soil nitrate-
leaching models to help determine how excessive
nitrates enter the soil profile; and Les Lanyon's
development of a farm monitoring system to follow
inputs and outputs of nutrients. In addition, Douglas
Beegle developed a computer program to help
farmers determine how much manure their livestock
produced and where it should be spread to meet fer-
tilizer recommendations based on soil test results.
Dale Baker, a professor of soil chemistry,
studied 23 fields during the 1983 growing season.
His study indicated that nitrogen was not the limiting
factor as measured by yield. Table 3 shows his find-
ings. Research since that date by Richard Fox sug-
gests little response of corn to additional nitrogen if
the spring nitrate-nitrogen level is above 20 to 25
ppm (76 to 95 pounds of nitrogen) per acre.
By 1984, USGS had established a network of 42
wells and one spring in the area for monitoring. It
found that 67 percent of the wells in the carbonated
rock portions of the project area had nitrate levels
above 10 mg/L of nitrogen, while only 27 percent of
the wells in the noncarbonate areas exceeded 10
mg/L.
324
-------
R. ANDERSON
Table 3.—Summary of selected field data collected in 1983 on 22 farms located in the Conestoga RCWP project
area.
NITROGEN APPLIED:
NOs-N LBS/ACRE
YIELD WHOLE PLANT
SITE NO.
201
501
1002
1201
1401
1801
1901
2101
FERTILIZER
N
50
165
7
112
12
8
90
118
MANURE
N
400
100
650
0
200
720
180
400
BY SOIL ANALYSIS
SPRING* FALL*
227.6
267.9
695.8
64.6
39.9
122.7
315.4
315.4
47.5
121.6
106.4
19.0
19.0
62.7
49.4
39.9
DRY MATTER
TON/A
3.62
5.44
7.52
5.71
5.63
4.83
7.64
8.95
* Caoulated by multiplying ppm values of NO3-N by 3.8, which is based on the number of pounds of soil in an acre-foot.
Surface water monitoring indicated that the base
flow in streams accounted for about 75 percent of the
total nitrates, and median concentrations of dis-
solved nitrate increased steadily in the watershed's
eastern end where the Little Conestoga River leaves
the noncarbonate forested area and enters the inten-
sively farmed carbonate valley. The maximum of 8.0
mg/L occurred at two sites in an area already
marked by land use data (collected by the Lancaster
County Conservation District) as having soils rich in
nutrients available for transport to the stream (see
Table 4).
Table 4.—Pre-BMP sampling results for selected
sites.
2A
3A
9
CHURCHTOWN
Streamflow 0.59 1.05 1.93 2.38 7.11
(ft3/second)
Nitrate 2.40 3.40 8.00 5.00 7.10
(mg/L)
Nutrient Management Begins
Based on these Penn State research projects and the
USGS monitoring efforts, the following major state-
ments were made in the interim evaluation of the
1984 Progress Report for the Conestoga Head-
waters:
1. Soil conservation practices are effective for
controlling surface losses of chemicals (dis-
solved in runoff or sediment-associated) but
not necessarily for controlling nitrate losses
out of the root zone that ultimately reach
groundwater. Sound management of nutrients
from commercial fertilizer, animal manures,
and legumes is necessary to prevent undue
contamination of groundwater drinking sup-
plies by nitrates, organic nitrogen, and fecal
coliform bacteria.
2. Manure storage that inhibits volatilization and
maximizes the plant nutrients available for
field application is a potential problem. Ani-
mal waste systems on farms with high animal-
to-land ratios should be designed so that plant
nutrients in manure do not exceed crop fer-
tilization requirements. Consideration should
be given in the future to storage structures
that minimize nitrogen availability on those
farms with high animal-to-land ratios.
In the economic evaluation section of the 1984
report, Dr. Edwin Young, U.S. Department of
Agriculture, Economic Research Service, Pennsyl-
vania State University, used the CREAMS model (the
Agricultural Research Survey's field-scale model for
chemicals, runoff, and erosion from agricultural
management systems) to estimate the surface losses
of soil, nitrogen, and phosphorus that were delivered
to the edge of fields. He also used the model to es-
timate the percolate losses of nitrate-nitrogen
leached out of the root zone:
While structural and other [best management
practices or] BMPs are effective means of con-
trolling nutrient losses, the most effective
method for controlling them is to reduce the
nutrients applied to the field via fertilizer and
manure management. A reduction in manure
applications from 40 to 20 tons per acre
results in a 40 percent reduction in nitrogen
and phosphorus losses. Storage of manure to
improve timing of applications leads to reduc-
tions in nutrient losses at recommended ap-
plication rates. However, if excess manure is
being applied to the field, nutrient losses may
actually increase.
The only BMP that reduced the amount of
nitrogen ^getting into the groundwater using the
CREAMS model was no-till planting, which leaves
most of the nutrient on the surface; however, no-till
planting oh soils with heavy manure applications is
very difficult.
325
-------
Proceedings of National RCWP Symposium, 1992
Young predicted that BMPs that reduce runoff
would also affect the amount of nitrogen leached into
groundwater supplies. This prediction was later con-
firmed by monitoring. Young concludes:
... nutrient management has the potential to
be the single most effective measure for con-
trolling nutrients losses. The CREAMS model-
ing showed no practice nearly as effective as
reducing the application of nutrients to levels
compatible with crop needs. Education of
farmers is needed to assist them to better
manage nutrients, particularly manure
nutrients. Timing, rates of application, and
methods of application should be part of an
overall farm plan to provide adequate fer-
tilization for crops while minimizing losses of
nutrients in surface runoff and deep perco-
late, .. . Further efforts to contract or simply
provide technical assistance for nutrient
management for farms not in the program are
expected to significantly reduce the loss of
nutrients from fields and barnyard areas....
Reliable tests for soil nitrogen and manure
nutrients are necessary if farmers are to make
informed choices in the fertilization of their
crops. Further research and educational ef-
forts are needed to provide local SCS or other
technicians with the continuing information
they need to establish state-of-the-art nutrient
management plans for farmers. Factors such
as animal density, crops grown, tillage prac-
tices, type of animal operations, and length of
manure storage will require that technicians
be able to tailor individual plans for each
farm.
A proposal to begin a Nutrient Management
Education Initiative was submitted to the national
RCWP committee in 1985. This initiative outlined a
plan to educate farmers in the project area about
nutrient management and develop nutrient manage-
ment plans using the approved technologies of the
time-soil and manure testing. The portion of the
nutrients available in the manure was to be deducted
from the soil test fertilizer recommendations for the
yields obtained on the farm. Goals of the Nutrient
Management Plans were to reduce
• nitrogen applications by 750,000 pounds on
20,000 acres by 1988 (about 37.5 pounds per
acre) and
* phosphorus applications by 375,000 pounds
on 20,000 acres by 1988 (about 18.75 pounds
per acre).
The National RCWP Committee approved the
transfer of funds set aside for the construction of
conservation structures to this initiative. On January
2, 1986, Penn State Extension opened its Nutrient
Management. Office in New Holland, Pennsylvania,
with two staff members. (Approval for the project
had come earlier in 1985, and the Eastern Lancaster
County School District helped by granting me an 18-
month leave to work on this project.)
During October, November, and December
1985, soil samples were taken in this area on 11
farms in a small watershed monitored by USGS. The
Lancaster Conservation District had also been
gathering information about manure spreading and
fertilizer applications in this area. Therefore, this
area was viewed as a good place to introduce
nutrient management.
Nutrient Management Planning
Nutrient management plans were developed using
the following procedure:
1. Soil test each field to determine the soil
reserves of P2Os and K20.
2. Obtain an official Penn State
recommendation for fertilizer needed to
yield the projected goal for each field.
3. Determine livestock numbers, types, and
average weights; the number of days
livestock were present on the farm; and the
number of days manure was collected from
the livestock for field application.
4. Perform a nutrient analysis of each manure
type used on the farm.
This information was computer-analyzed, and
each farmer received a plan containing a manure ap-
plication rate for each field and a recommendation
for any additional fertilizer nutrients needed to ob-
tain optimum yields. Because of time limitations,
plans for the first 11 farms were developed without
using a manure analysis. Tables 5 and 6 show the
manure values used for these farms derived from the
1985 to 1986 Penn State Agronomy Guide.
Initially, the first 11 plans were to be written
using book values for manure nutrients so that we
could begin the first growing season with some
nutrient management plans in place. Manure tests
would be taken after the spring rush, and plans
changed, if necessary, for the 1987 growing season.
However, a large pool of nitrogen was identified that
had not been addressed in the planning process.
For example, a dairy farm is applying 25 tons of
dairy manure per acre each year; the manure is
spread on a daily basis. The amount of nitrogen per
326
-------
R. ANDERSON
Table 5.—Total plant nutrient values from manure.
KIND
DAILY MANURE PRODUCTION
(lb/1000 Ib LIVE WEIGHT)
APPROXIMATE NUTRIENTS (Ib/ton)
KzO
Cattle (fresh, 85% water)
Poultry
(liquid, 95% water) .
(fresh, sticky, 75% water)
(moist, crumbly, 50% water)
(crumbly, 30% water)
(dry, 15% water)
Swine
Sheep
Horses
85
300
61
32
22
18
100
40
50
10
f
10
30
40
60
90
10
12
6
3
7
20
40
55
70
6
7
3
5
3
10
20
30
40
9
15
10
Important: Only a portion of the total nitrogen is available to crop in the year in which it is applied. The actual amount depends on handling
(see the Penn State Agronomy Guide).
Table 6.—Nitrogen available from manure.
N AVAILABILITY FACTOR*
% POTENTIAL LOSS OF AVAILABLE N
APPLICATION METHOD
POULTRY
OTHER
POULTRY
OTHER
Immediate incorporation
Incorporation after 2 days
Incorporation after 4 days
Incorporation after 7 days
or no incorporation
0.75
0.45
0.30
0.15
0.50
0.35
0.30
0.20
0
40
60
80
0
30
40
60
"Use these factors and total nitrogen values or a manure analysis to calculate available nitrogen as follows: Available N « (total N) X (N
availability factor). •••.-.-
ton using the book value in Table 5 is 10 pounds. The
Penn State Agronomy Guide (1985 to 1986) proce-
dure for calculating nitrogen is Total nitrogen = (tons
of manure applied) x (nitrogen per ton) x (nitrogen
availability factor [Table 6]) In this example, Total
nitrogen = (25 tons) x (10 pounds per ton) x (0.20
availability factor), which calculates to a total of 50
pounds of nitrogen available from the manure ap-
plication. The total nitrogen in the manure is 25 tons
times 10 pounds per ton or 250 pounds of nitrogen.
Approximately 75 pounds of nitrogen (30 per-
cent) was lost to the atmosphere because it was not
incorporated. Therefore, after subtracting the 50
pounds of nitrogen for the crop plus the 75 pounds
lost to the atmosphere from the 250 total pounds of
nitrogen in the manure, 125 pounds of nitrogen
remains that has not been accounted for.
Major Change in Nutrient
Management Planning
On October 21, 1987, a letter was sent to Dr. Myron
Johnsrud, administrator of the Cooperative Exten-
sion Service, stating:
Manure nitrogen management using the
traditional approach can result in applying as
much as 875 pounds of manure nitrogen to
. supply 175 pounds of nitrogen to a corn crop
(20 percent of 875 = 175 IbsJ. This excess
nitrogen (700 Ibs/acre in this example) will
escape into the environment. With the lack of
research addressing the residual nitrogen in
manure, the nutrient Management -Office
made the following assumption: if the amount
of manure applied per acre was constant over
a period of years, 35 percent of the residual
nitrogen will become available in any given
year. Using this approach, it is only necessary
to apply 318 pounds of manure nitrogen to
supply the 175 pounds of nitrogen needed by a
corn crop ([20 percent + 35 percent] of 318 =
175 IbsJ. This leaves only 143 pounds of
nitrogen to escape into the environment, a
savings of 557 pounds of nitrogen. During the
1986 and 1987 growing season many test
plots were conducted to validate this assump-
tion. Farmers were asked to apply livestock
manure in their usual manner. After corn
was planted, the nutrient management staff
applied nitrogen at the following rates: zero
pounds, 50 pounds, 75 pounds, and 100
pounds per acre. Plots were randomly repli-
cated in blocks of four. An attempt was made
to have plots on farms that had different
amounts of nitrogen recommended using this
assumption. Plots were hand harvested and
yields calculated on the per acre basis.
The results of this data are shown in Table 7.
327
-------
Proceedings of National RCWP Symposium, 1992
Table 7.— 1986 and 1987 yield data (assume: residual
N from manure will contribute an additional 35 per-
cent nitrogen).
NITROGEN NUMBER OF OPTIMUM YIELDS
RECOMMENDATION PLOTS (no. at rate of N)
0 12 10atO
2 at 50
50 6 5 at 0
1 at 50
75 1 1 at 0
100 1 1 at 50
180 1 1 at 200
Over the two-year period, a total of 21 plots was
used to check the validity of thjs' assumption. Only
three fields responded to more nitrogen than
projected, while eight fields needed less nitrogen
than was estimated using the assumption. Ten fields
had optimum yields at the projected amount of
nitrogen. Eighty-six percent of the fields tested had
adequate nitrogen using this assumption, and none •
of the fields that had yield increases with additional
nitrogen above the recommended level had yields
that were significantly different.
If this assumption were followed on the 3,736
acres of corn grown in the project area, an additional
186,800 pounds of nitrogen could be reduced from
the fertilizer recommendations to bring the total
reduction to approximately 67 pounds of nitrogen
per acre per year. Penn State University's College ,of
Agriculture adopted a change in its, 1989-1990
Agronomy Guide to reflect this residual nitrogen
from manure. Because the limited data was created
in only a small portion of the State, the values
adopted in the Guide are less than the 35 percent
used in the RCWP project area. The residual
nitrogen credit ranged from 7 percent to 25 percent
Table 8.— Average total nutrient content of manure.
DAILY PRODUCTION MANURE
ANIMAL TYPE (lb/1000 Ib live Wt.) % DRY MATTER
Dairy cattle:
Solid 82 13
Liquid
Veal 63 2
Beef cattle 60 12
Swine:
Pifls 65 9
Gestating sows 32 9
Sow and pigs 88 9
Boar 31 9
Liquid
Sheep 40 25
Horses 45 20
Poultry:
Fresh 61 • 25
Moist 32 50
Crumbly 22 ' '70
Dry 18 • 85
in the Agronomy Guide. These credits are shown in
Tables 8 and 9, which also reflect increased informa-
tion on manure nutrients. Starting in 1986, crop
management meetings sponsored by the Coopera-
tive Extension Service have stressed the need for
manure analysis. (A program to test manure, which
was also administered through county offices,
generated the additional information included in
Tables 8 and 9.)
The use of starter fertilizer was also analyzed in
the project area. If a significant reduction in phos-
phorus was to be achieved, only two areas of soil fer-
tility management had possibilities (other than
reducing manure application rates):
• phosphorus applications would have to be
removed from the starter fertilizer (if this
could be done without affecting yields) , and
• potassium would have to be taken out of the
hay fertility programs.
Removing potassium from the hay program was
not a problem because soil test recommendations for
hay management on soils high to excessive in potas-
sium do not recommend additional application. Infor-
mation was not available, however, to support the
elimination of phosphorus from corn starter fer-
tilizer. Between 1986 and 1987, 17 test plots were
conducted on farms in the project area to assess if
the starter fertilizer applications could be eliminated.
Participating farmers were asked to disconnect the
fertilizer applicator drive on half their corn planter.
With a four-row planter, this process created a plot of
four rows with, and four rows without, starter fer-
tilizer, replicated a minimum of four times. An at-
tempt was made to have plots on farms with different
(Ib/tonor100gal)
N PjOs K20
10 4 8
2.8 1.3 2.5
8 2 11
11 7 10
14 11 11
14 11 11
14 11 11
14 11 11
3.5 2.0 1 .5
23 8 20
12 5 9
30 20 10
40 ,40 20
60 55 30
100 70 , 40
Important note: Have manure analyzed when possible. Actual values may vary over 100 percent from averages in table.
328
-------
R. ANDERSON
Table 9.—Percentage of total manure nitrogen remaining available to crops after storage and handling as af-
fected by application method and field history.
N AVAILABILITY FACTOR*
A. CURRENT YEAR, TIME OF APPLICATION,
AND INCORPORATION
POULTRY MANURE
OTHER MANURE
Manure applied for com or summer annual following year:
Applied in spring
Immediate incorporation
Incorporation within 2 days
Incorporation within 3-4 days
Incoporation within 5-6 days
Incorporation within 7 days
No incorporation
Applied previous fall or winter with no cover crop
Applied previous fall or winter with cover crop
Harvested as silage*
Used as green manure
Manure applied for small grains
Applied previous fall or winter
0.75
0.50
0.45
0.30
0.15
0.15
0.15
0.50
0.50
0.50
0.40
0.35
0.30
0.20
0.20
0.20
0.40
0.40
B. HISTORY FREQUENCY OF MANURE APPLICATION ON
THE FIELD
Rarely received manure in the past
Frequently received manure (4-8 out of 10 yrs.)
Continually received manure (>8 out of 10 yrs.)
0
0.07
0.12
0
0.15
0.25
* Low availability factors do not Indicate a net loss of nitrogen (N). A large amount of N is removed in the cover crop silage. This N will be
recycled in the manure when silage is fed.
Note: To calculate total amount of nitrogen available from manure, add the appropriate value from Table 9, section A based on application
method and livestock type to the appropriate value from Table 9, section B based on the frequency of manure applications and livestock type,
then multiply this answer by the manure analysis value or the nitrogen value from Table 8.
levels of phosphorus. Results of these plots can be
seen in Table 10.
Although it is difficult to draw any conclusions
from the data, it appears that a large percentage of
the farms in the project area could eliminate starter
fertilizer applications without reducing yields. Data
for the 10 replicated plots conducted in 1986 showed
that only three locations had increases in yield, three
had the same yield, and four locations had decreased
yields when starter fertilizer was used (Table 10).
Statistical differences were not calculated for the
plots. However, the three plots that had increased
yields with starter fertilizer averaged an increase of
9 bushes per acre, while, the four plots that had
decreased yields with starter fertilizer averaged a
decrease of 21.5 bushels per acres. Less dramatic
differences were found in 1987 (Table 10). Over the
two years, 17.6 percent of the fields (3 of 17) showed
a response to starter, 47.1 percent (8 of 17) had in-
creased yields by not applying starter fertilizer, and
Table 10.—Starter fertilizer vs. no starter fertilizer.
YEAR
1986
1987
NO.
PLOTS
10
7
STARTER FERTILIZER
NO. WITH INCREASE
IN YIELDS
BUSHELS/ACRE
INCREASE
3
9bu
3
7.5 bu
NO STARTER FER-
TILIZER NO. WITH
INCREASE IN YIELDS
BUSHELS/ACRE
INCREASE
. 4 , .
21 .5 bu
4
8.5 bu
35.3 percent (6 of 17) had the same yield with or
without starter fertilizer. The nonstarter plots
showed an average increase of 16.8 bushels per acre,
while those that received starter fertilizer showed an
increase of 8.25 bushels per acre.
Using Demonstration and
Research Plots
The educational initiative portion of the RCWP
project was successful primarily because participat-
ing farmers had personal contact with Extension
agents that resulted in one-on-one instruction related.
to nutrient management and water quality issues
(leading to the development of a site-specific plan
based on soil and manure tests) and the use of on-
farm demonstration and research plots. During each
year of the program, each agent taught approximate-
ly 40 farmers how to develop a nutrient management
plan, and test plots were a valuable teaching aid.
Over 50 demonstration and research plots were con-
ducted between 1986 and 1991. Although 28 plots
compared the recommended rate of fertilizer in the
nutrient management plan with a higher rate of fer-
tilizer, only one showed a significant yield increase
by adding more nitrogen than the nutrient manage-
ment plan recommended (perhaps because of a wet
growing season and identification). Twenty-three
329
-------
Proceedings of National RCWP Symposium, 1992
plots compared starter fertilizer with no starter fer-
tilizer; only three showed a significant increase in
yield resulting from starter fertilizer.
The design of the nitrogen demonstration and
research plots was a simple 4x4 random block ar-
rangement. Each replication contained four or eight
rows of corn 25 to 35 feet long. Under most cir-
cumstances the nitrogen application rate was (1) 50
pounds per acre less than the plan called for, (2)
what the plan called for, (3) 50 pounds more than the
plan called for, and (4) 100 pounds per acre more
than the plan called for. Harvesting was done by
hand.
The starter demonstration and research plots
used a simple strip plot design. Depending on the
number of rows planted with one planter pass, the
plot would vary from two to four rows in width. On
some occasions, the starter strips overlaid the ran-
dom block nitrogen plots. When this design was
used, nitrogen plots were always eight rows wide—
four with and four without starter fertilizer. Plots
were located where they could be seen easily by the
cooperating farmer and several neighbors. When-
ever possible, the farmer was asked to help harvest
and evaluate the plots.
The Penn State Nitrogen
Quick Test
Nitrogen fertilization rates are the most difficult to
formulate. Many factors (including the weather,
nitrogen source, time of application, and crop) in-
fluence how much nitrogen is needed to produce op-
timum crop yields. Research continues to seek a
chemical analysis method for nitrogen either in the
soil or in the plant at early stages of growth to
calibrate the addition of nitrogen to reach optimum
yields.
The total nitrogen requirement of plants can be
met in several ways. First, soils supply nitrogen
through organic matter decay. Second, legumes and
manure can supply significant amounts of nitrogen.
Third, nitrogen fertilizer can be used. However,
nitrogen behavior in the soil is very complex (Fig.
2). Over 98 percent of the nitrogen in a typical soil
profile is unavailable to plants, according to Fox. The
availability of nitrogen as nitrate is based on microor-
ganisms in the soil that are very sensitive to weather
and soil conditions. Soil nitrates are most available in
late spring when soils are warm and in late fall when
crop demand drops. Since corn has its greatest
demand for nitrogen 30 to 40 days after emergence,
a soil test used at this time is ideal for determining if
additional nitrogen is needed to optimize crop yields.
Figure 2.—The nitrogen cycle.
The graph in Figure 3 shows the results of 87
field research experiments by Fox and associates at
Penn State before the test was made available in
Pennsylvania. The graph indicated that the nitrate-
nitrogen level from this test was very good for iden-
tifying soils in which a yield increase will occur after
fertilizing with nitrogen. In the graph, relative corn
yield is the yield with no nitrogen applied relative to
the maximum yield achieved by adding nitrogen fer-
tilizer. A relative yield near 1 indicates that the yield
without nitrogen is near or equal to the maximum
yield, which means little or no yield increase can be
expected from adding nitrogen fertilizer. The test is
primarily recommended for use on fields with sig-
nificant organic nitrogen contributions from manure
applications or forage legumes in a rotation. How-
ever, the test has not been able to predict nitrogen
from alfalfa in a no-till farming system. The Nitrogen
Quick Test became the final educational tool and in-
itiative that the Extension staff had to help farmers
deal with balancing crop nutrient needs.
0 10 20 30 40 50 60 70 80 90
Soil Test Nitrate-Nitrogen (ppm)
Figure 3.—Yield comparisons.
The procedure to use the Nitrogen Quick Test is
relatively easy:
330
-------
R. AttDERSOtt
1. Apply manure based on nutrient
management plan recommendations.
2. If nitrogen is needed based on the plan,
apply a minimum at planting. If nitrogen is
used as a herbicide carrier, use a maximum
of 50 pounds per acre.
3. When the corn is 12 inches tall, take 10 to 20
core samples to a depth of 12 inches if
possible.
4. Combine, crumble, and dry the samples as
soon a possible.
5. Using the Penn State Nitrogen Quick Test,
analyze the soil for nitrate-nitrogen.
6. Determine the sidedress nitrogen
recommendation from Table 11.
Accomplishments
The potential value of fertilizer savings is estimated
to be $165,415.75 for nitrogen at $0.25 per unit;
$100,973.10 for phosphorus at $0.30 per unit; and
$42,223.20 for potassium at $0.15 per unit for a total
potential fertilizer savings on the 365 farms of
$308,612.05 annually. Actual reductions were slightly
less. Two surveys were performed to see what
farmers had actually done with the recommenda-
tions. The first survey conducted in 1987 found that
farmers had reduced their nitrogen application rates
by 55 percent of recommended levels and their phos-
phorus application rates by 33 percent. A second sur-
vey in 1989 found that farmers had reduced their
nitrogen application rates by 79 percent of the
recommended levels and their phosphorus applica-
tion rates by 45 percent. Actual accomplishments in
the nutrient management project are shown in Table
12.
Table 12.—Project accomplishments.
GOALS
ACCOMPLISHMENTS
Plans 20,000 acres
Reduce N 750,000 Ibs.
Reduce PzQs 375,000 ibs.
Reduce KzO No goal
24, 134 acres
661, 663 Ibs.
336,577 Ibs.
281 ,488 Ibs.
crops cannot be grown without allowing some
nitrogen to leach beyond the root zone. Elimination
of nitrate leaching will be accomplished only when
plants are able to glean 100 percent of the nitrogen in
the soil and when application rates do not exceed
plant needs. Currently, however, to meet the
nitrogen needs of a plant, nitrogen must be applied
at rates beyond this level.E
References
Lancaster County Conservation District. 1982. Assessment of
Nonpoint Sources of Pollution from Agriculture in the Cones*
toga River Watershed. Farm and Home Center, Lancaster,
PA.
Pennsylvania Agricultural Statistics Service. 1990. Statistical Sum-
mary 1989-90 and Pennsylvania Department of Agriculture
Annual Report. Harrisburg.
Pennsylvania Crop Reporting Service. 1960. Pennsylvania Crop
and Livestock Annual Summary. Harrisburg.
-. 1970. Pennsylvania Crop and Livestock Annual Summary.
Conclusion
Nutrient management and soil nitrate testing are
great tools to help reduce excessive applications of _
nitrogen. However, even with the use of these tools,
Table 11.—Sidedress N fertilizer recommendations for corn.
Harrisburg.
. 1980. Pennsylvania Crop and Livestock Annual Summary.
Harrisburg.
Pennsylvana State Agricultural Stabilization and Conservation
Service. 1982. Conestoga Headwaters Rural Clean Water Pro-
gram Progress Report. Agric. Stabil. Conserv. Serv., U.S.
Dep. Agric., Harrisburg.
. 1982. Conestoga Headwaters Rural Clean Water Program
Comprehensive Monitoring Plan. Agric. Stabil. Conserv.
Serv., U.S. Dep. Agric., Harrisburg.
. 1984. Conestoga Headwaters Rural Clean Water Program
1984 Progress Report. Agric. Stabil. Conserv. Serv., U.S.
Dep. Agric., Harrisburg.
. 1985. Conestoga Headwaters Rural Clean Water Program
1985 Progress Report. Agric. Stabil. Conserv. Serv., U.S.
Dep. Agric., Harrisburg.
Pennsylvania State University. 1986. The Penn State Agronomy
Guide, 1985-86. Coll. Agric. Ext. Serv., University Park.
. 1989. The Penn State Agronomy Guide, 1988-89. Coll.
Agric. Ext Serv., University Park.
CORN YIELD GOAL: GRAIN (bu/a) OR SILAGE (T/A)
SOIL TEST LEVEL
(ppm NOa-N)
0-10
11-15
16-20
21-25
25 +
100/17
100
75
50
.25
0
125/21 150/25 175/29
SIDEDRESS N RECOMMENDATIONS (Ibs. N/A)
130 160 190
100 125 150
75 100 125
50 75 100
000
200/33
220
150
125 .
100
0
331
-------
-------
Involving the Agricultural Chemical
Industry in Nutrient Management
Jeffrey Stoltzfus, Leon Ressler, and Robert Anderson
Penn State Cooperative Extension Service
Upper Conestoga RCWP
Smoketown, Pennsylvania
ABSTRACT
Nutrient management planning was a major component of the Upper Conestoga Rural Clean Water
Program project. Nutrient management plans called for the total reduction of 661,663 pounds of
nitrogen, 336,577 pounds of phosphorus, and 281,488 pounds of potash, all of which amounted to
$308,612/year in fertilizer savings for the 365 farmers. A financial bonus for farmers, these reduc-
tions cut into the profits of the local fertilizer companies that served them. From the opening of the
nutrient management office in 1986, a yearly fertilizer dealer meeting was held to keep fertilizer
and pesticide dealers abreast of the Upper Conestoga's goals and accomplishments. These meet-
ings also surveyed business opportunities related to nutrient management, such as crop consulting
and manure marketing. Fertilizer dealers participated in field testing the Pre-Sidedress Nitrogen
Soil Test. Communication channels opened through these meetings were crucial to the success of
nutrient management plans and the support of fertilizer dealers in the farm community was critical
to the credibility of the RCWP nutrient management program.
Nutrient management planning was a major
component of the Upper Conestoga Rural
Clean Water Program (RCWP) project.
Nutrient management plans that were developed
called for the total reduction of 661,663 pounds of
nitrogen, 336,577 pounds of phosphorus, and
281,488 pounds of potash, reductions that amounted
to $308,612/year in fertilizer savings for the 365
farmer-participants. Although this was a financial
bonus for farmers, it cut into the profits of the local
fertilizer companies.
Support from the agricultural chemical industry
was important to the success of an educational
nutrient management program. Fertilizer salesman
were respected and valued crop consultants in the
Upper Conestoga watershed. The local fertilizer ven-
dors had a long history of working with the Coopera-
tive Extension Service and local agriculture teachers
on educational programs. Most fertilizer vendors
were following then-current Penn State recommen-
dations when calculating manure nutrients and fer-
tilizer needs. However, as a result of agronomic re-
search done in the Upper Conestoga RCWP, in 1987
Penn State changed the way manure nitrogen
availability was calculated — a change that ac-
counted for most of the fertilizer reduction gained in
the RCWP nutrient management program.
In 1986, Penn State Cooperative Extension
opened a nutrient management office for the RCWP
project and hired two staff to develop nutrient
management plans for local farmers. In January
1987, the first fertilizer dealer meeting was held, and
22 participants from 13 dealerships attended. These
dealerships represented most of the fertilizer sales
in the county. At the meeting, dealers learned about
the nutrient management component of the project,
heard some preliminary results from water quality
monitoring, and discussed soil fertility as well as
Penn State's nutrient management computer pro-
gram.
In 1988, the fertilizer vendor meeting included
an explanation of the new figures recommended by
333
-------
Plroceedfngs of National RCWP Symposium, 1992
Penn State's Agronomy Department for calculating
residual nitrogen from manure applications. These
figures gave credit to the breakdown of organic
nitrogen from previous years' manure applications
and led to a decrease in fertilizer recommendations
for fields receiving manure. Results of local test plots
conducted by RCWP staff and Penn State re-
searchers were shared each year. Information on
pesticide safety was also incorporated into the
programs because most fertilizer vendors also sell
pesticides.
In 1989 and 1990, fertilizer vendor meetings
began to focus on business opportunities created by
nutrient management, among them custom manure
hauling, crop consulting, manure marketing, and
processing and packaging of poultry litter. Atten-
dance at these meetings increased to 45 or 50 people
representing approximately 25 fertilizer companies;
many of these visitors, although located in other
parts of Pennsylvania, were interested in nutrient
management and its impact on their businesses.
Increased fertilizer vendor attendance could be
attributed to their excitement about the opportunity
to be a part of the solution. The agricultural chemical
industry had historically felt unfairly blamed be-
cause the increase in livestock numbers — not
pounds of commercial fertilizer — had exacerbated
nutrient problems. However, of all the agricultural
service people (including veterinarians, feed sales-
men, and milk inspectors) who educate farmers
regularly, the agricultural chemical salesmen had
the most expertise in nutrient management and
agronomic crop production. Although nutrient
management in many cases reduced fertilizer sales,
vendors recognized farmers' need to use crop con-
sultants. Through fertilizer sales, agricultural chemi-
cal dealers had traditionally profited by increasing
farmers' profitability. Now that fertilizer sales were
not increasing profits, fertilizer dealers were
anxiously looking for new opportunities to serve
farmers.
The meetings were the most visible aspect of the
fertilizer vendor involvement. In addition to provid-
ing information about RCWP, these meetings en-
couraged an open dialogue that enabled RCWP
personnel to better understand how to help the fer-
tilizer industry meet nutrient management goals and
set the stage for intensive fertilizer industry involve-
ment in the Pequea Mill Creek Hydrologic Unit
Project, which serves as the sequel to the Upper
ConestogaRCWP.
One nutrient management tool that fertilizer
vendors used was the Pre-Sidedress Nitrogen Soil
Test. An effective addition to a nutrient management
plan, this test determines which cornfields need ad-
ditional nitrogen fertilizer. Field-tested in the RCWP
nutrient management office in 1989, this test was
released to the public in 1990. It is taken in early
June when the corn is 6 to 18 inches tall and some of
the organic nitrogen has broken down into nitrate.
(vendors often find it hard to apply nitrogen at this
time of year because of bad weather conditions.) For
the test, fertilizer vendors pulled soil samples and
analyzed them for their clients; samples taken from
heavily manured fields rarely called for additional
nitrogen. This test took some of the pressure off the
spreading schedule and helped further reduce
nitrogen applications. Fertilizer vendors received ap-
proximately $5 per sample for testing, which helped
them profit from their investment in testing while
still allowing farmers to save money on fertilizer.
More fertilizer vendors have begun to use this test,
and farmers are relying on it more heavily to plan
fertilizer applications.
The dialogue created by the agricultural chemi-
cal dealer meetings also allowed an opportunity to in-
clude them in Cooperative Extension test plots. In
1990, one fertilizer company allowed Extension to
test a micronutrient enhancer with the starter fer-
tilizer on a high phosphorus soil. This product
produced a 20-bushel yield increase. Making sure
farmers achieve their maximum economic yield is
essential for effective nutrient management. Higher
yielding crops take up more nutrients; therefore,
farmers must use the fertilizer vendors' agronomic
expertise to promote nutrient management.
In the future, nutrient management in Pennsyl-
vania (particularly within the Chesapeake Bay water-
shed) and agricultural chemical industry involve-
ment will continue to evolve. The cooperation and
communication developed in the Upper Conestoga
RCWP project is now the basis for a new level of in-
dustry involvement in water quality projects. The Pe-
quea Mill Creek Hydrologic Unit Project funded in
1991 in an adjoining watershed is certifying agricul-
tural chemical dealers and private consultants to
write nutrient management and integrated crop
management plans. If this program is successful, it
may spread statewide and create a positive impact
long after the Pequea Mill Creek project has ended.
The agricultural chemical industry is excited about
this project; interested companies had to be turned
away from the first training session to keep a
manageable group size.
Conclusions
When the focus of water quality programs shifts to
cultural practices such as nutrient management, in-
tegrated pest management, and farmstead assess-
334
-------
J. STOLTZFUS, L. RESSLER, &R. ANDERSON
ment, private consultants, vendors, and other
agricultural service industries should be brought
into the process. Whether the program is national,
State, or watershed-specific, the process is the same.
The first step is to identify which agricultural service
sector has expertise in the practice we want to
promote. (For example, milk company fieldmen are
experienced in performing farmstead assessments
and wellhead protection.) The second is to identify
benefits the market can provide the agricultural ser-
vice industry for promoting this BMP — e.g., adver-
tising benefits or a new service for farmers.
Most importantly, an uncomplicated method for
implementing the practice should be developed to
decrease the amount of paperwork required while
still allowing a minimum level of tracking. Although
the agricultural service industry is eager to be a part
of the water qualify solution, they still need to sur-
vive in the marketplace. An overly bureaucratic
process can cut well-intentioned businesses out of
the process. Finally, the Federal cost-share and in-
centive programs should include private industry.
Currently, most Agricultural Stabilization and Con-
servation Service programs specifically exclude
private industry.
As we attempt to address environmental issues
in the information age, we need to become partners
with the agribusiness industry. Feed and fertilizer
salespeople, veterinarians, and the host of other
professionals in service industries whom farmers
rely on for information must become a part of the en-
vironmental solution. Government has too often
taken the "us versus them" approach in dealing with
private industry. As bureaucratic budgets tighten at
all levels of government, private industry will have to
be involved in any successful regulatory or educa-
tional environmental effort The Upper Conestoga
RCWP project laid the groundwork for a new level of
industry cooperation that will be essential for ad-
dressing the water qualify issues of the 1990s.
335
-------
-------
Economic Evaluation of the
Rural Clean Water Program
Richard Magleby
Economic Research Service
U.S. Department of Agriculture
Washington, D.C.
ABSTRACT
.The key objective of the Rural Clean Water Program (RCWP) was to "achieve improved water
quality in the approved project area in the most cost-effective manner possible." The program also
called for socioeconomic evaluation to "identify the positive and negative impacts on the land-
owners in the project area and estimate the community or off-site benefits expected of the project if
completed as planned."
This paper discusses some insights gained and lessons learned from the economic evaluation
conducted by the U.S. Department of Agriculture's Economic Research Service. Strong points of
the economic evaluation included successful linking of physical and economic models and some
state-of-the-art analyses of the cost-effectiveness of existing and alternative practices and ap-
proaches. The evaluation could have been improved by greater consistency across projects. En-
ding the evaluation in 1987 also created problems for projects needing updated information for
subsequent annual and 10-year reports.
Economic models used in the RCWP ranged from simple budgeting to multiactivity linear
programming. Links were made to physical models, ranging from simple soil loss and delivery
models to complex field and watershed models. Results indicated that the RCWP projects generally
increased farmers' net returns and reduced pollutant loadings. However, water quality benefits and
program efficiency could have been significantly improved by greater targeting of efforts to the
most critical areas and greater emphasis on management practices rather than structural
measures. The economic efficiency of future programs may be increased if economic assessments
are initiated much earlier to provide information for project selection and best management prac-
tice planning.
The Rural Clean Water Program (RCWP) was
envisioned and authorized as a large-scale,
$600 million-a-year effort to reduce agricul-
tural nonpoint pollution sources. Appropriations
were only provided, however, for a much smaller "ex-
perimental" program. Although the reduced scale
was disappointing at the time, the experimental
RCWP has permitted the gathering of valuable ex-
perience in planning, coordinating, implementing,
monitoring, and evaluating water quality projects
before launching a larger-scale effort. This paper as-
sesses the RCWP and its projects from an economic
perspective.
Assessments were based on model predictions
and the use of alternative scenarios. Monitoring data
were not used because several alternatives needed
to be evaluated. The U.S. Department of Agri-
culture's (USDA) Economic Research Service (ERS)
was responsible for economic evaluation of the
RCWP, with concentration on five comprehensive
monitoring and evaluation projects in Idaho, Illinois,
Pennsylvania, South Dakota, and Vermont. Less-
detailed evaluations of the 15 other projects and an
overall program evaluation were proposed but not
funded.
Economic Aspects of a Project
Economic aspects of a project include its
• costs,
• cost-benefits to producers,
« cost-effectiveness, and
• the off-site economic benefits generated in
relation to the costs.
337
-------
Proceedings of national RCWP Symposium, 1992
Federal regulations for RCWP included these
aspects as expectations for the program:
Tlie objectives of RCWP are to: (a) Improve
impaired water use and quality in the ap-
proved project areas in the most cost-effective
manner possible in keeping with the provision
of adequate supplies of food and fiber ..."
(Federal Regulations, Section 700.2 as
amended).
(3) Socioeconomic Impacts. Identify the posi-
tive and negative impacts on the landowners
in the project area and estimate the com-
munity or off-site benefits expected of the
project if completed as planned. (Federal
Regulations, Section 700.41). -
Figure 1 helps put these economic aspects into
perspective with other aspects of a rural nonpoint
water quality project. Beginning at the top of the
chart, the project extends direct educational, techni-
cal, or financial assistance to farmers who implement
practices to reduce pollutants leaving the site or
leaching downward. The expenditures and time re-
quired to extend this assistance are project costs. Be-
cause some best management practices (BMPs)
require greater assistance to the producer than
others, it is important to determine the differences :
among BMPs relative to time and assistance.
The adoption of BMPs affects the farm operation
by altering the farmer's costs and returns. The ex-
tent to which farmers implement BMPs and con-
tinue to use them after direct assistance winds down
depends on these effects, both on the average and
from year-to-year.
Farmers in the project area not receiving direct
assistance to implement water quality practices may
be influenced indirectly through project publicity or
contacts with program participants. Some may im-
plement water quality practices on their own, par-
ticularly if such practices are more profitable. Others
may continue or engage in practices that increase
pollutant generation. What takes place on these
"other farms" is often not assessed, yet it can be cru-
cial to achieving water quality results. Other farmers
and point sources will — it can be hoped — also con-
tribute to the reductions in pollutant loadings
achieved among program participants so that an
overall reduction will occur in pollutant loadings to
the impaired waterbodies.
To generate water quality economic benefits, the
reductions in pollutant loadings and the resulting
preservation or improvement in water, quality have
to be large enough to increase water use or enjoy-
ment or to reduce user costs. This point is crucial. It
is possible to implement practices that reduce site-
level pollutants and therefore reduce pollutant load-
ings to a waterbody to improve water quality slightly
without significant changes in use or generation of
benefits. It is also possible that the preservation or
improvement in water quality may occur over such a
long time that the current value ;of those benefits is
small. Both possibilities occurred in,the RCWP.
Two measures of economic efficiency are cost-
effectiveness and benefits. Cost-effectiveness is the
cost of an action divided by some measure of the ef-
fectiveness of that action in reducing a critical pol-
lutant or improving a key parameter of water quality.
Costs may include government and net producer
costs. To be useful, cost-effectiveness must be as-
sessed comparatively. One action is more cost effec-
tive than another if either the cost is lower or the
effectiveness is greater, other things equal.
Comparing economic benefits to costs can occur
either as a ratio of one to the other, or by subtracting
costs from benefits to estimate net economic
benefits. A project — or one configuration of a pro-
ject — is more economically efficient than another if
it has a higher benefit cost ratio,, i.e., more economic
benefits per dollar of cost. Ideally, economic benefits
of a project or program .should exceed costs by as
high a margin as possible.
Which of the following do you monitor and
evaluate in a water quality project and program?
• BMP implementation; •
• reduced pollutant loadings at the site level;
;• reduced loadings to the impaired water
bodies;
« water quality preservation or improvement
sufficient to generate economic benefits; or
• generation of benefits that exceed costs.
Some people believe that BMPs are effective
and see no need for evaluation beyond monitoring
BMP implementation. Others focus on public desire
to have water quality problems resolved and believe
that voluntary programs must be made as effective
and .beneficial as possible if regulations are to be
avoided.
Costs of RCWP and Its
Projects
The experimental RCWP was funded in fiscal years
1980 and 1981 with Federal appropriations totalling
$70 million. This amountwas latercutto $64 million
as a Federal budget-reducing measure. Fortunately,
the funds were "no-year," allowing their disbursal,as
338
-------
R, MAGLEBY
Education and technical
assistance
Financial
assistance
Other
farmers
T
Farmers receiving
direct assistance
( BMP
Vadoption/use?
i
Change in
chemical use,
other inputs,
and outputs?
• BMP
adoption/use
Change in
chemical use,
other inputs,
and outputs
•f Reduction in
( loadings to bottom of root
Vzoneand edge of field?
f Reduction in \
f loadings to bottom of root
\^ zone and edge of field? J
Reduction in
loadings to ground
and surface waters?
Water quality
preserved or
improved?
Changes in water
uses and values?
Economic benefits? /
Cost effectiveness?
Benef its > costs?
Most benefits per $?
Figure 1 .—Flow chart for rural water quality projects.
needed over time. Of the $64 million in Federal
funds, about 53 percent provided financial assistance
to farmers implementing BMPs in 21 projects; 30
percent was applied to technical and educational as-
sistance, and 17 percent to monitoring and evalua-
tion. In addition, State and local governments con-
tributed over $4 million to the program.
The funds allocated to and spent by RCWP
projects varied greatly, depending on their size, the
nature of the water quality problems, the BMPs im-
339
-------
Proceedings of National RCWP Symposium, 1992
plemented, and the degree of monitoring and evalua-
tion required (Table 1). Projects ranged in total cost
from $364,000 for the Utah project to $5.7 million for
the Vermont project, including State and local con-
tributions. State and local assistance was substantial
in some projects; in the Florida project, for example,
local contributions comprised nearly one-third of the
total expenditures.
Nearly two-thirds of RCWP financial assistance
supported structural BMPs, with about three-
fourths of this put into animal waste systems (Table
2). Several projects allocated nearly all financial as-
sistance to structural BMPs. Management BMPs
received about one-fourth of total RCWP financial as-
sistance, with the balance to permanent-cover
BMPs.
Accounting procedures in the projects kept
track of expenditures for specific BMP implementa-
tion but seldom for technical assistance or informa-
tion and educational expenditures related to specific
BMPs. The latter are significant components of
some BMPs, and their accounting in future
programs would facilitate economic evaluation and
the determination of the most cost-effective BMPs.
Effects on Farm Costs and
Returns
Effects on farm costs and returns were evaluated for
the five comprehensive monitoring and evaluation
projects by using BMP and enterprise budgeting
(Table 3). In general, farmers in these projects who
implemented BMPs with financial and technical as-
sistance increased their net returns, and thus had
economic incentives to participate. In the case of
conservation tillage properly implemented and
managed, net returns would have been higher even
without cost-sharing. In the Illinois and South
Dakota projects, upward trends in conservation til-
lage use existed before the projects, but the projects
further increased the rate of adoption in both cases.
In the Pennsylvania and Vermont projects,
farmers implementing manure storage and manage-
ment practices with project assistance increased
their net returns. They would, however, have
lowered their net returns had they not received as-
sistance. The Vermont project achieved high par-
ticipation. An interesting issue in this project's
economic evaluation was whether a lower cost-share
rate, one that provided an increase in net returns but
not as much, would have achieved as much or nearly
as much participation. Some project personnel felt
strongly that the full incentive was needed. In Penn-
sylvania, cultural factors discouraged farmers' accep-
tance of financial assistance and hence their par-
ticipation in manure storage.
In the Idaho project, farmers implemented sedi-
ment basins and other sediment-retention devices
with project assistance. However, analysis indicated
that there would be no incentive of increased returns
to maintain these basins and devices after the con-
Table 1. — Expenditures by
RCWP projects and area treated.
GOVERNMENT EXPENDITURES (In thousands)
PROJECT
Alabama
Delaware
Florida
Idaho,
Illinois3
Iowa
Kansas3
Louisiana
Maryland
Massachusetts
Michigan3
Minnesota
Nebraska
Oregon
Pennsylvania
South Dakota
Tenn-/Kentucky
Utah
Vermont
Virginia
Wisconsin
FINANCIAL
ASSIST.
1,215.1
873.2
1,267.5
1,988.8
331.2
2,300.0
3,576.1
518.4
1,747.0
978.2
3.574.6
794.0
744.7
2
143.7
1,686.3
1,721.0
817.1
TECHNICAL
ASSIST.
490.0
304.1
432.9
1,446.7
19.7
830.3
946.0
215.1
623.8
422.7
680.5
1,268.4
614.6
2
88.8
3,227.8
496.6
125.0
I&E
14.4
0.0
79.0
135.5
2
2
511.8
10.8
119.7
230.6
7.5
88.5
163.9
2
85.8
18.7
65.9
1.9
M&E,
OTHER
315.0
400.0
1,389.0
2
2
2
55.0
2
198.6
2,614.1
2,581.6
2
46.2
2,414.1
120.4
5.0
TOTAL
1,719.5
1,492.3
2,179.5
5,242.5
421.0
3,130.4
5,033.9
917.2
2,545.5
1,631.5
4,461.2
4,765.0
4,043.1
2
364.5
5,779.5
2,403.9
949.0
STATE
AND LOCAL
EXPENDI-
TURES1
105.9
2
722.1
280.7
.7
64.4
270.1
1.0
303.5
3
127.8
2
644.6
2
2
640.8
98.4
5.0
CRITICAL
AREA
(ACRES)
: 6,300
13,000
93,500
28,159
3,920
44,800
2
24
20,255
60,242
5,300
2
41 ,830
2
15,431
23,908
26,000
PROPOR-
AREA TION
TREATED TREATED
(ACRES) (PERCENT)
3,811
9,750
54,709
21,147
3,359
27,103
,000
24
10,793
42,879
2,544
17,700
33,465
2
11,277
16,655
23,598
75
59
75
90
60
100
53
71
48
80
80
73
70
91
Source: Project 10-year reports.
I&E «Information and Education; M&E = Monitoring and Evaluation
'Included In "Government expenditures." Excludes farmers' costs.
sNot reported.
310-year report missing.
340
-------
R. MAGLEBV
E
n
D)
o
Q.
1
1
i
o
"a
tc
CD
Ł
•D
C
3
HT
c
to
CO
3
O
C
ss.
!
a
0)
a.
X
CD
CD
O
I
_to
'to
CO
CO
75
u
n
s=
a.
!
CM
0)
Z
|2
*»
»
m
i-
S
i
i
pi g
w l-
co
^
in
0
a>
CO
(O
n
Ł
go
rr
3
S
3
t—
CO
Ł*"
ca
c
1
i-
i
Ł
a.
_j
CO O
52
CM
T-
in
9
CM
_,
CO 2
5
T-
-
,-
PROJECT
•4- if) in oo o t-- CM
in CM co i- CM CM
7) h* O Q> &)
T- CM in
03 T-
CM CO
o •*-
^5 cS
0>
i- CO 00 ^-
oo ffi i- T-
co ^*
O} O3 CM CM
i- CO CM
i- CO
CM i- 00
*f & co
CM
oo co CM in t- CM r«
co 03 co r*. - co r^ 03
in CM CM o_ co i-
t— T—
CO
1—
o CM in co in CM
o in. CM co "- co co
co co CM
CO CM
00 O CM CM T-
^ ^ 0> CO 0)
•^
CM in oo en o ^
CM in co a> O)
CO CM t~-
co
S? m fO
r^ CO CO
CO O CO T- *f i-
CM CM CO CD in
J: * CM g CD CO
"•"
s
s
g 2 «M = 1 1
i 1 •§ ~o N§ „ I 3 1 1
-9J5~^ogc'5^<2
CoQ)OwJ.t>COo™™
oo
O) f— CO
CO CM
!§ W CM
03
CM
fe - ro
CD T-
in en co co
12 g § *
co
o
?
S O3 ^
^- CD
CM *
K §3 S S
co co
co
in co i- •<-
M- CO CO
CM
s
1- 0 T-
TJ* T-
cn en in
o> f- T- in
l-~ CM CM
CM ^
CO 5
«u « « 11
3 ° •« c "5. Q
111 HI
5 S Z O tL en
in in co co
T- CO CO
*-
CO
CM
CO
-• 2 Ł 1
,_-
t s
*-
cS
^"
• ^--co-- . en
in
CM CO CM
CO CO Ol 00
CO CM O CO
T- co_ co co
^~ ^
Si '18
'-"
i- TO CM
CO CD
CO C3
if~
' CM i- CO
*~
CM
co
co o in oo
^— CM O OO
T- CM oo co
r CO
CO t~ CD
^ 0
CM"
S3
CD T- «fr
5
Is** IO O
T~~ O) ^"
T"
IO T- O>
CM CM O
"*
I S -g f 1
IIIIII
E
til
2!
.9 J5
2 o
CD Q
=6 §
i4
tl
(0
—^ i
ill
III
111
.9-o.c
w E w
wl|
tf"g)
S o ro
w E E
111
^" (0 *^
ffS)^
if?
ill
i §•§.
•- R ">
§ " 2
TO [| -^^
II CO H
is S 71
fp
S
> § i"
§ tj g
Q> -t-« *3
O) O CO
>|"E
flf
f^~
7". E~2
8,f a
•— 03 .=
s|I .
Elf -i
C "crt C "CO
'I i ra co
0 II C g-
»!i!i
Q S w m *te
S" ^ C Ł (D
*~ Ł •— ^ Ł
eg o 13 g •*=
Ł !••§ s *
2 i^-« 5
"•§111
tu o nj ^ ta
•o1"™^ §
Q_ Q- c: ±3 4_>
- - 2 'N w o
1 »-lI
Ifi§j?
341
-------
Proceedings of national RCWP Symposium, 1992
Table 3.—Major components of the economic evaluations in five RCWP comprehensive monitoring and evalua-
tlon projects.1
Idaho Rock Creek Project:
—Best management practices (BMPs) and enterprise budgets of costs and returns
—Farm level simulation of economic effects of BMPs
—BMP cost-effectiveness (CE) analysis on representative sites
—Project CE analysis using Universal Soil Loss Equation, sediment delivery ratios, and linear programming (LP)
—Off-site benefit estimates based on existing studies
Hlnots Highland Silver Lake Project: ,
—BMP and enterprise budgets of costs and returns '
—Representative farm analysis using LP
—Project CE analysis using AGNPS (Agricultural Nonpoint Source model) and LP
—Recreational fishing estimate based on special survey.
—Other off-site benefit estimates based on special studies
Pennsylvania Conestoga Headwaters Project:
—BMP and enterprise budgets of costs and returns
—Site-level CE analysts of BMPs using CREAMS (Chemicals, Runoff, and Erosion from Agricultural Systems) and budgets
—Feasibility evaluations of alternative uses of manure
—Drinking water benefit estimates based on cost of alternatives
South Dakota Oakwood Lakes-Poinsett Project:
—BMP/enterprise budgets of costs and returns
—Project CE analysis using AGNPS and CREAMS
—Recreation benefit estimates based on special survey
—Drinking water benefit estimates based on cost of alternatives
Vermont St. Albans Bay Project:
—BMP/enterprise budgets of costs and returns
—BMP analysis using AGNPS
—Representative farm analysis using LP
—Recreation benefit estimates based on special survey ,
—Property value benefit estimates based on special survey
Source: compiled by author.
1The level of funding allocated to the separate economic evaluations varied widely and, in some cases, limited the scope of the evaluation.
tracts expire, which appears to be the case in an ad-
joining watershed, where some farmers who had im-
plemented these practices under another program
have not maintained them after the contractual
period.
Insights on Cost-Effectiveness
In keeping with the RCWP's objective of improving
impaired water use and quality "... in the most cost-
effective manner possible," ERS evaluated cost-effec-
tiveness in the five comprehensive monitoring and
evaluation projects. Methods and procedures dif-
fered among the projects (Table 3). Budgeting and
linear programming were used to generate the cost
measures. Physical models were used to predict the
pollutant reduction effectiveness for the project
when the BMPs were fully implemented as planned
— and for alternative scenarios. Monitoring results
were not used because they did not cover all alterna-
tives being considered and could not be easily con-
trolled for weather variations.
The first step in evaluating cost-effectiveness
was to identify significant trends that would have
continued in the project area in the absence of
RCWP, so as to not overestimate the effects of the
program. In the Illinois and South Dakota projects,
significant upward trends occurred in the use of con-
servation tillage. These trends were projected for-
ward and incorporated into "future without RCWP"
runs of the models. Subtracting the results of "future
without" from "future with RCWP" provided an es-
timate of the net effect of RCWP. In the two projects,
the predicted net effectiveness of RCWP was one-
third to one-half lower than biased estimates that did
not consider the conservation tillage trend.
Model predictions showed the net effectiveness
of the RCWP in the Idaho, Illinois, South Dakota,
and Vermont projects as follows:
• Idaho: a reduction of 13 to 56 percent in
agricultural sediment entering Rock Creek
would be achieved, depending on the level of
adoption of conservation tillage and the main-
tenance of structural BMPs over time.
• Illinois: a net reduction of 14 percent in
agricultural sediment entering Highland Sil-
ver Lake would result from the RCWP, com-
pared to a 24 percent reduction if the
conservation tillage trend was not factored
• out.
• South Dakota: the RCWP would provide a
net reduction of 13 percent' in phosphorus
from agricultural sources entering the
project's three impaired lakes, compared to a
342
-------
R. MAGLEBY
21 percent reduction if the conservation til-
lage trend was not excluded.
• Vermont the RCWP would provide a reduc-
tion of 17 percent in phosphorus from agricul-
tural pollutants entering St. Albans Bay.
To gain insights into how these projects might
have achieved greater cost-effectiveness, alternative
scenarios were developed that included different
mixes of BMPs or greater targeting of treatment to
the most critical areas. The alternatives were
developed in consultation with project staff. For the
Idaho and Illinois projects, the model results indi-
cated that larger acreage in conservation tillage and
less use of structural measures would have in-
creased effectiveness for the same or less cost.
For the South Dakota project, the results indi-
cated that adding a more ideal form of fertilizer
management (split application and injection, in addi-
tion to soil testing and nutrient budgeting) to exist-
ing contracts could greatly improve the project's
effectiveness in reducing nutrient loss at minor
added cost. In contrast, urging farmers with existing
contracts to change from mulch to ridge- or no-till
was shown not to be cost-effective because the effort
would have added to project cost with no significant
reduction in nutrient loadings.
Model results for the Idaho, Illinois, and South
Dakota projects indicate that greater targeting of
conservation tillage or ideal nutrient management to
the most critical acres would have added more to ef-
fectiveness than to cost. The results also indicate
that greater financial incentives could have been of-
fered to farmers in critical areas to speed adoption of
conservation tillage and split application of fertilizer,
and these practices would still have been the most
cost-effective alternative.
Is putting lands into the Conservation Reserve a
cost-effective way of achieving pollutant reductions?
For the South Dakota project, model results indicate
that ideal fertilizer management would be a much
less costly and more cost-effective to achieve
nutrient reductions. However, taking the land out of
production and placing it in permanent cover may
contribute to other societal objectives.
In considering the cost-effectiveness of alterna-
tives and deciding on practices to implement, it is
helpful to know how much reduction is needed in
critical pollutants to achieve water quality objectives
in the impaired resources. Some less cost-effective
measures may be needed in addition to cost-effective
ones to achieve- the needed reductions. For the
Idaho, Illinois; and South Dakota projects, reduc-
tions over 60 percent in agricultural pollutant load-
ings would require incentives sufficient to induce
farmers to.place most cropland under improved
management practices, or to put some of the critical
land into structural measures or permanent cover
(such as in the Conservation Reserve Program). For
the Vermont project area, the reduction of loadings
sufficient to eventually restore St. Albans Bay re-
quired adding agricultural practices to the more
cost-effective sewage treatment program.
In the Pennsylvania project, only the cost-effec-
tiveness of practices at the field- or farm-level was
evaluated. The analysis shows that conservation til-
lage and nutrient management, contouring, and sod
waterways were much more cost-effective than ter-
races and permanent vegetative cover in reducing
sediment and nutrients in runoff. The CREAMS
model also indicated that terraces may significantly
increase the infiltration of nitrogen into ground-
water. Monitoring results have since confirmed this.
Alternative uses or disposal of manure to reduce
field application were also found to have low eco-
nomic feasibility.
Insights on Economic Benefit
Generation
The RCWP projects have reduced pollutants at the
site level, and most projects will reduce loadings to
some waterbody. However, only 10 to 13 projects
(one-half to two-thirds) have or may be able to
measure significant improvements -to water quality
in the impaired water bodies (Natl. Water Qual. Eval.
Proj. 1988; Coffey, pers. comm.) Possibly another 3
to 6 projects, while not making a significant impact
on water quality, could contribute to water quality
preservation or improvement if they become part of
larger regional programs that do make a significant
difference.
Clearly, projects that do not significantly con-
tribute to water quality improvement will not
generate significant economic benefits. Economic
benefit estimates for other RCWP projects were
developed by estimating the range of potential
benefits that might be generated from preserving
water quality from further deterioration and from im-
proving water- quality over preproject conditions.
These two sources of benefits were projected for 50
years, adjusted -to reflect likely, changes in water
quality, and discounted to current value.
The estimates of the potential benefits for the
Idaho, Illinois, South Dakota, and Vermont projects
involved special surveys or studies. (A bibliography
including these .studies is appended to this paper.)
Travel costs and contingent valuation techniques
were used in some .cases. Estimates for other
343
-------
Proceedings of National RCWP Symposium, 1992
projects were based on available information on
water uses, potential users, and the values of dif-
ferent uses. For all projects with drinking water im-
pairments, potential benefits for this use were based
on costs of obtaining water from alternative sources,
such as bottled water.
Results of this benefit analysis suggest that two
to four of the projects may generate relatively high
economic benefits; five to six, moderate benefits; and
four to five, low benefits. For six to twelve of these
projects, generation of economic benefits may be
sufficient to exceed project costs. Six other projects
received an "uncertain" rating because of the poten-
tial for benefit generation outside the project area if
they become part of larger regional programs that
successfully preserve or improve water quality in
major downstream resources. These results indicate
that not all water quality projects will generate high
benefits and that a high economic return on water
quality dollars would require a development of
economic information for use in project selection.
In two projects (South Dakota and Oregon),
available information indicates that the RCWP will
probably generate significant economic benefits
from preserving water quality that would otherwise
deteriorate. Nine projects may generate moderate-to-
high economic benefits by improving water quality
from baseline conditions.
The four projects with high potential economic
benefits involve lakes, bays, or estuaries in close
proximity to, and highly affected by, the project area.
The waterbodies are also in high demand for recrea-
tion and, in some cases, commercial fishing. Four
projects with moderate potential economic benefits
also involve lakes or bays highly used for recreation.
Preserving or improving water quality in a lake or
bay does not automatically generate sizeable
benefits, especially if there are few users, or if
restrictions exist on use. For example, in the Illinois
project, recreational use of the lake is restricted to
fishing in low-powered boats and only one ramp is
permitted. A low level of economic benefit genera-
tion does not necessarily mean that economic
benefits won't exceed costs; costs can also be kept
low. Two projects that produced relatively low
economic benefits could have had positive net
benefits if more cost-effective BMPs had been imple-
mented.
Steps for Increasing Economic
Efficiency of Future Programs
ERS staff learned from the economic evaluation of
the RCWR Strong points of the evaluation included
the successful linking of physical and economic
models and the application of state-of-the-art
analyses of cost-effectiveness to existing and alterna-
tive practices and approaches. Some early results
also provided insights for modifying project im-
plementation. The evaluation could have been im-
proved by greater consistency across projects.
Ending the evaluation in 1987 created problems for
projects needing updated economic information for
subsequent annual and 10-year reports.
For the most part, the economic evaluation of
the RCWP occurred midway during project im-
plementation. Results and experience indicate that
initiating economic evaluation much earlier in future
programs would provide program and project
decisionmakers more timely information for increas-
ing economic efficiency. In particular:
• Before project selection. Conduct feasibility
assessments of proposed or potential projects
to provide information for comparing projects
on at least these aspects:
• Extent Of water use impairments;
• Potential economic benefits if impairments
are reduced;
• Reductions in, pollutant loadings from
various sources needed to reduce
impairments, with attention to the possible
need for a regional program, and the levels
of expected improvement in both surface
and groundwater quality;
• Treatment levels needed to achieve
desired reductions in loadings, considering
local trends in agriculture;
• Likelihood and cost of achieving needed
treatment levels in comparison to potential
economic benefits; and
• Other environmental or nonuser
considerations.
• After project selection but before im-
plementation. Identify and appraise the cost-
effectiveness of feasible alternative targeting
and BMP strategies for achieving the needed
reductions in pollutant loadings. To what ex-
tent can management BMPs be used to
reduce project costs? What incentives will
achieve the needed BMP adoption? Will the
BMPs be maintained?
• During project implementation. Peri-
odically reappraise implementation strategies
to see if changes would contribute to greater
cost-effectiveness. Can costs be reduced or
344
-------
R. MAGLEBV
greater reductions in pollutant loadings be
achieved? Are higher incentives needed for
the more cost-effective BMPs?
After project completion. Review , the
achieved and expected results to see what in-
sights on cost-effectiveness and benefit
generation can be gleaned of value for the fu-
ture.
Economic Evaluation of the
Rural Clean Water Program —
A Bibliography
Inquiries about these reports should be directed to
Richard Magleby, ERS Room 534, 1301 New York
Avenue, N.W., Washington, DC 20005-4788.
Telephone (202) 219-0436.
General or Summary Reports
Magleby, R. 1988. Cost-effectiveness of BMP implementation.
Pages 143-45 in Rural Clean Water Program Workshop Proc.
Natl. Water Qual. Eval. Proj., N.C. State Univ., Raleigh.
Magleby, R. and C.E. Young. 1985. Controlling agricultural runoff:
government's perspective. Pages 234-36 in Proc. Natl. Conf.
Perspectives on Nonpoint Source Pollution. EPA 440/5-85-
001. U.S. Environ. Prot Agency, Washington, DC.
Magleby, R., S. Piper, and C.E. Young. 1989. Economic insights on
nonpoint pollution control from the Rural Clean Water Pro-
gram. Pages 63-69 in National Nonpoint Source Conf. Proc.
Natl. Ass. Conserv. Distr., League City, TX.
Piper, S., C.E. Young, and R. Magleby. 1989. Benefit and cost in-
sights from the Rural Clean Water Program. J. Soil Water
Conserv. 44(3) =203-08.
Piper, S., R. Magleby, and C.E. Young. 1989. Economic benefit con-
siderations in selecting water quality projects — insights
from the Rural Clean Water Program. Staff Rep. No. 89-18.
Econ. Res. Serv., U.S. Dep. Agric., Washington, DC.
Young, C.E. and R. Magleby. 1987.'Agricultural pollution control:
implications from the Rural Clean Water Program. Water
Resour. Bull. 23(4):701-07.
. 1985. Economic benefits of three rural clean water
projects. In Proc. Symp. Off-site Costs of Soil Erosion. Soil
Conserv. Serv., U.S. Dep. Agric., Washington, DC.
Reports on the Illinois Highland Silver
Lake Project
Carvey, D. et al. 1986. Economic Evaluation chapters in Highland
Silver Lake, Illinois, RCWP Annual Reports for 1981-86.
Agric. Stabil. Conserv. Serv., U.S. Dep. Agric., Washington,
DC.
Setia, P. and R. Magleby. An economic analysis of agricultural non-
point pollution control alternatives. J. Soil Water Conserv.
42(6)427-31.
. 1988. Economic efficiency of agricultural nonpoint pollu-
. tion controls. Unpubl. Pap. Annual Meet. Northe. Agric.
Resour. Econ. Ass., Orono, Maine.
. 1988. Measuring physical and economic impacts of con-
trolling water pollution in a watershed. J. Lake Reservoir
Manage. 4 (1):63-71.
Setia, P., R. Magleby, and D. Carvey. 1988. Illinois Rural Clean
Water Project: an economic analysis. Staff Rep. AGES880617.
Econ. Res. Serv., U.S. Dep. Agric., Washington, DC.
Southwestern Illinois Metropolitan and Regional Planning Com-
mission. 1983. Assessment of offsite socioeconomic impacts
— Highland Silver Lake project Rep. No. SWIL-MAPC83-06.
Collinsville.IL.
1982. Economic baseline for the Highland Silver Lake
project Unpubl. Rep. Collinsville, IL.
Starr, V.B. 1983. An Economic Evaluation of Rural Clean Water
Program Policies on Representative Farms in the Highland
Silver Lake Watershed. M.S. thesis. Univ. Illinois. Urbana.
Reports on the Idaho, Rock Creek
Project
Gum, R., R. Magleby, arid J. Kasal. 1986. Economic Evaluation
chapters in Rock Creek, Idaho, RCWP Project Annual Project
Reports: 1980-86. Agric. Stabil. Conserv. Serv., U.S. Dep.
Agric., Washington, DC.
LaPlant, D., D. Martin, L Wear, and R. Gum. 1984. Wildlife habitat
Impacts. Unpubl. Pap. Econ. Res. Serv., U.S. Dep. Agric.,
Washington, DC.
. Economics of Controlling Sediment from Irrigation: An
Idaho Example. Staff Rep. No AGES89-33. Econ. Res. Serv.,
U.S. Dep. Agric., Washington, DC.
Walker, D.J., D.T. Noble, and R.S. Magleby. 1980. Effective cost-
share rates and the distribution of social costs in the Rock
Creek, Idaho, Rural Clean Water Program project J. Soil
Water Conserv. 45(4):477-79.
Walker, D.J., P.E. Patterson, and J.R. Hamilton. Costs and benefits
of improving irrigation return flow water quality in the Rock
Creek, Idaho, Rural Clean Water Project. Res. Bull. No. 139.
Univ. Idaho, Moscow.
Reports on the Pennsylvania
Conestoga Headwaters Project
Crowder, B.M. and C.E. Young. 1985. Evaluating BMPs in
Pennsylvania's Conestoga Headwaters Rural Clean Water
Program. Unpubl. Pap. Nonpoint Pollution Abatement
Symp., Milwaukee, WI.
-^-. 1985. An economic analysis of the Conestoga Headwaters
RCWP Project. Unpubl. Rep. Econ. Res. Serv., U.S. Dep.
Agric., Washington, DC.
. 1985. Modeling Agricultural Nonpoint Source Pollution
for Economic Evaluation of the Conestoga Headwaters
RCWP Project Staff Rep. No. AGES850614. Econ. Res. Serv.,
U.S. Dep. Agric., Washington, DC.
. 1987. Bridging the gap between private incentives and
public goals for agricultural nonpoint pollution control. Pages
177-86 in Optimum Erosion Control at Least Cost. Am. Soc.
Agric. Eng., St. Joseph, MI.
. 1988. Managing farm nutrients: tradeoffs for surface and
groundwater quality. Agric. Econ. Rep. No. 583. Econ. Res.
Serv., U.S. Dep. Agric., Washington, DC.
. 1987. Soil conservation practices and water quality: is
erosion control the answer? Water Resour. Bull. 23(5):897-
902.
345
-------
Proceedings of National RCWP Symposium, 1992
Young, C.E., J.R. Alwang, and B.M. Crowder. 1986. Alternatives for
Dairy Manure Management. Staff Rep. No. AGES860422.
Econ. Res. Serv., U.S. Dep. Agric., Washington, DC.
Young, C.E. and B.M. Crowder. 1986. Managing nutrient losses:
some empirical results on the potential water quality effects.
NatlJ. Agric. Resour. Econ. (October) :130-36. '•
Young, C.E., G. Lengerich, and J. Beierlein. 1984. The feasibility of
using a centralized collection and digestion system for
manure: the case of Lancaster County, Pennsylvania. Proc.
Conf. Poultry Waste. Univ. Park, PA
Young, C.E. et al. 1986. Economic Evaluation chapters in Annual
Reports for Conestoga Headquarters, Pennsylvania, RCWP
Project: 1981-86. Agric. Stabil. Conserv. Serv., U.S. Dep.
Agric., Washington, DC.
Reports on the South Dakota
Oakwood Lakes-Poinsett Project
Erickson, M. and J. McMartin. 1984. Crop Budgets by Alternative
Tillage Systems and Crop Rotations for Selected Soils, Oak-
wood Lakes-Poinsett Rural Clean Water Project Area South
Dakota. Unpubl. Working Mat. Econ. Res. Serv., U.S. Dep.
Agric., Washington, DC.
Erickson, M., R. Magleby, and B. Crowder. 1984. Economic
Evaluation chapters in 1984 Annual RCWP Progress Report:
Oakwood Lakes-Poinsett, South Dakota. Agric. Stabil. Con-
serv. Serv., U.S. Dep. Agric., Washington, D C.
Hoover, H. and M.W. Erickson. 1985. Soil erosion and water
quality farmer workshop: Oakwood Lakes-Poinsett Rural
Clean Water Program, South Dakota. Unpubl. Staff Pap.
Econ. Res. Serv. U.S. Dep. Agric., Washington, DC. •
Magleby, R., M. Erickson, and D. Carvey. Economic Evaluation
chapters in Annual Reports for 1981-88. South Dakota, Oak-
wood Lakes-Poinsett RCWP Project. Agric. Stabil. Conserv.
Serv., U.S. Dep. Agric., Washington, DC.
Piper, S., M.O. Ribaudo, and A. Lundeen. 1987. The recreational
benefits from an improvement in water quality at Oakwood
Lakes and Lake Poinsett, South Dakota.' North Central J.
Agric. Econ. 9(2):279-87.
Reports on the Vermont, St. Albans
Bay Project
Bouwes, N.W. Sr. 1983 Baseline economic conditions in the St Al-
bans Bay watershed. Unpubl. Pap. Econ. Res. Serv., U.S. Dep.
Agric., Washington, DC.
Bouwes, N.W. Sr. and C.E. Young. 1983. Procedures in estimating
on-site impacts of BMP adoption. Pages 66-77 in Proc. Conf.
on Agricultural Nonpoint Source Pollution Management in
Viriginia. VAPolytech. Inst. and State Univ., Blacksburg.
Frevert, K. and B.M. Crowder. 1987. Analysis of agricultural non-
point pollution control options in the St Albans Bay water-
shed. Staff Rep. No. AGES870423. Econ. Res. Serv., U.S. Dep.
Agric., Washington, DC.
Ribaudo, M.O. and D. Epp; 1984. Importance of sample discrimina-
tion in using the travel costmethod to estimate the benefits of
improved water quality. Land Econ. 60(4):397-403.
Ribaudo, M.O., C.E. Young, and J.S. Shortle. 1986. Impacts of
water quality improvement on site visitation: a probabilistic
modeling approach. Water Resour. Bull. 22(4):559-63.
Ribaudo, M.O., C.E. Young, and D. Epp. 1984. Recreation benefits
from an improvement in water quality at St. Albans Bay, Ver-
mont Staff Rep. No. AGES840127. Econ. Res. Serv., U.S.
Dep. Agric., Washington, DC.
Young, C.E. and J.S. Shortle. 1989 Benefits and costs of agricul-
tural nonpoint source pollution controls: the case of St Al-
bans Bay. J. Soil Water Conserv. 44(l):64-67.
Young, C. E. and EA Teti, The influence of water quality on the
value !of recreational properties adjacent to St Albans Bay,
Vermont Staff Rep. No. AGES831116. Econ. Res. Serv., U.S.
Dep. Agric., Washington, DC.
Young, C.E. 1984. Perceived water quality and the value of
seasonal homes. Water Resour. Bull. 20(2) :163:66.
Young, C.E. et al. 1986. Economic Evaluation chapters in Annual
Reports for the St. Albans Bay, Vermont, RCWP Project,
1981-86. Agric. Stabil. Conserv. Serv., U.S. Dep. Agric.,
Washington, DC.
346
-------
Technology Application in
BMP Planning, Design, and Application
Gene Dougherty and Jesse T. Wilson
Soil Conservation Service .
U.S. Department of Agriculture
Gainesville, Florida
ABSTRACT
Initially, traditional survey, design, and drafting methods were used in the Okeechobee Rural Clean
Water Program project. However, evolution and expansion of the original project required develop-
ment of detailed waste management plans, irrigation systems for wastewater, and best manage-
ment practices (BMPs), such as waste storage ponds, floodwater-retarding structures, and
methods to treat high intensity areas. Because of compliance deadlines imposed by State
regulatory agencies, the Soil Conservation Service had to revise its traditional survey, design, and
layout methods, which had proven to be too time-consuming. New survey, design, and drafting
technologies were implemented to prepare engineering designs and provide technical assistance in
their installation within the time frame allowed by State regulatory agencies. These new tech-
nologies ultimately improved the quality of these designs and staff productivity, which was, stimu-
lated by the need to complete topographic and planimetric surveys and design waste management
systems under a time constraint Increased quality stemming from accurate field data and waste
management system designs resulted in more efficient use of land resources.
The original Rural Clean Water Program
(RCWP) focused on the Taylor Creek-Nub-
bin Slough drainage basin, which contained
24 dairies on approximately 33,000 acres. The
average dairy milked 1,000 cows. A 1978 study by
the South Florida Water Management District deter-
mined that this basin contributed 4 percent of the
wastewater flow to Lake Okeechobee and 27 percent
of the total phosphorus load; moreover, it attributed
most of the lake's phosphorus and nitrogen loading
to the dairy operations. Approximately 64,000 acres
(which included all dairies in the basin) were re-
quired to have conservation plans addressing water
quality.
Implementation of the original RCWP project
began in 1981, with goals of a 50 percent reduction of
phosphorus discharge and conservation plans on 75
percent of the critical acres and 100 percent of the
dairies. Typical BMPs installed under the original
RCWP project included
• fences to keep dairy cattle out of waterways,
• waterway crossings,
• portable livestock shade structures, and
• wastewater recycling.
As a result, project goals were met or exceeded in
1986. Among the accomplishments were a 50 per-
cent reduction of phosphorus discharge and conser-
vation plans for 96 percent of the critical acres and all
dairies.
347
-------
Proceedings of National RCWP Symposium, 1992
In 1986, the occurrence of large algae blooms in
Lake Okeechobee prompted the Florida Department
of Environmental Regulation (FDER) to enact a rule
specifically for the dairies in the lake's drainage
basin. Commonly called the "Okeechobee Dairy
Rule," the FDER rule, which became effective June
3, 1987, as part of Rule 17-6.300 of the Florida Ad-
ministrative Code, imposed requirements on 25 addi-
tional dairies covering approximately 52,000 acres in
the Lower Kissimmee River basin and specified that
"discharge of dairy farm wastewater and runoff to
the waters of the State shall not cause or contribute
to a violation of water quality standards." This expan-
sion of the RCWP project led to the development of
new, nontraditional BMPs for the dairies in the Lake
Okeechobee drainage basin. BMPs included those
previously used in the Taylor Creek-Nubbin Slough
drainage basin and
• waste utilization,
• irrigation systems,
• subsurface drainage,
• waste storage ponds,
• floodwater-retarding structures,
• surface drainage field ditches, and
• water table control.
With the need to design and install waste
management systems on. 49 dairies with ap-
proximately 50,000 dairy cows within three years,
the South Florida Water Management District con-
tracted with the Soil Conservation Service (SCS) to
design waste management systems for 30 of the
dairies. Waste management system designs comply-
ing with these requirements were more detailed
than those developed for the original RCWP project;
therefore, SCS had to develop water budgets that
would account for the large amount of wash water
used as well as evapotranspiration, rainfall, runoff,
and drainage.
SCS also developed cropping sequences to uti-
lize nutrients from dairy wastes. Since phosphorus
was the limiting nutrient, sequences were planned to
increase crop uptake of phosphorus. These cropping
sequences required landowners to adopt practices
new to the south Florida dairy industry.
To maximize phosphorus uptake by crops, SCS
provided water table control systems by installing
drainage ditches or subsurface drainage tile with
structures for water control. In many cases, existing
natural outlets did not have the capacity for the re-
quired drainage rate; therefore, floodwater-retarding
structures were constructed to provide an adequate
outlet for the irrigation field through pumping or
gravity flow.
Traditional Methods for
Installing BMPs
Planning
For the initial RCWP project, planning was based on
aerial photographs, U.S. Geological Survey quad-
rangle maps, and site visits, which were used to
determine the areas that needed detailed surveys.
The initial planning phase also established a priority
system for surveying and designing individual
dairies involved in the project.
Surveying
A team of three or four persons was needed to per-
form the detailed topographic and planimetric sur-
veys for planning and design. The design required a
detailed topographical survey of the high intensity
areas, which are nonvegetated areas around the
milking barn where dairy cows are concentrated for
feeding and watering before milking. Typically, high
intensity areas ranged between 10 and 30 acres and
were usually covered by two to three feet of manure,
making them difficult to survey. In addition, they
were often overgrown with brush and vegetation
along the numerous fence lines, a condition that re-
quired multiple instrument setups for standard
levels and transits. The waste storage ponds were
also difficult to survey because of their location,
vegetative cover, and size — an average of 40 to 50
acres, with the largest being 300 acres.
The size of the irrigation sprayfield, which
ranged from 150 to 674 acres, was based on crop
nutrient uptake or the volume of wastewater to be ap-
plied. A complete topographical survey was needed
on these fields to properly design the drainage and
irrigation system.
A survey for the high intensity areas, waste
storage pond, and irrigation sprayfields required ap-
proximately three to four weeks for the average
dairy, followed by up to four days to reduce survey
notes and plot, by hand, the topographic information
for use in the design phase.
Design
Waste management systems were designed based
on crop use of nutrients, with phosphorus being the
limiting nutrient. The result was a waste manage-
ment system that collected wash water from the
barn and rainfall and runoff and drainage water from
high intensity areas for temporary storage in a waste
storage pond until it could be used for irrigation.
348
-------
G. DOUGHERTY & J.T. WJLSON
Development and analysis of alternative systems re-
quired six to eight weeks.
Traditional methods for designing waste storage
ponds were based on an average monthly water
budget that included dairy wash water, rainfall,
runoff and drainage water from the high intensity
area, and evapotranspiration. The irrigation spray-
field size was determined based on crop nutrient use
and volume of wastewater applied.
To achieve high crop uptake rates (up to 75
pounds of phosphorus per acre), it was necessary to
improve the drainage in the irrigation sprayfieldi If
the available outlet could not provide adequate
drainage for optimal plant growth, the system was
designed to include a floodwater retarding structure,
the size of which was based on the required crop
drainage coefficient and the discharge rate allowed
by the South Florida Water Management District.
These alternatives were presented to the land-
owners for their review and selection of the best al-
ternative for their particular operation. However, the
limited availability of computers and the time re-
quired to perform these evaluations reduced the
number of alternatives that could be analyzed and
presented to the landowners.
Implementation
Soil Conservation Service staff performed periodic
and final construction checks on all practices in-
stalled. Construction layout was performed by SCS
technicians for the first several dairies using stand-
ard levels, transits, and electronic distance measur-
ing equipment. Cross sections were taken before
and after construction to compute earthwork vol-
umes.
Current Technology for
Installing BMPs
Because of the size and complexity of the dairies, the
traditional technology used to design waste manage-
ment systems was inadequate to meet the time
frame imposed by State regulatory agencies; there-
fore, SCS decided to implement newer technology to
provide waste management plans to landowners.
Surveying
Introduction of the new technology proved to be
beneficial, particularly in the areas of survey and
design. Surveying quality and accuracy noticeably
improved with the use of total station surveying in-
struments and data collectors. A crew of three could
complete the survey of a typical dairy in one to two
weeks (instead of four). The data collectors allowed
electronic coding and storage of survey measure-
ments and physical features, such as existing fences,
drainage ditches, structures, and buildings. The use
of total station surveying instruments allowed for
more accurate measurements of the physical fea-
tures of sites that were hard to access.
SCS downloaded survey data from the data col-
lector to the computer and used commercial
software to produce topographic maps for design
and planning. These maps were available within
three hours of survey completion, as opposed to
several days when survey notes were reduced and
plotted by hand.
Design
The introduction of computer-aided drafting (CAD)
helped develop standard construction details that
could be used on several designs. Many of the waste
management system plans were drawn with the CAD
system, which allowed computerized changes in the
drawings, saved time, and resulted in quality draw-
ings.
Paralleling the use of CAD was the development
and implementation of several computer models that
help to evaluate existing or proposed drainage sys-
tems and determine required capacity of the waste
storage ponds and floodwater-retarding structures.
SCS used DRAINMOD, a drainage simulation
computer model developed by R.W. Skaggs of North
Carolina State University, to evaluate and design
drainage and irrigation systems. DRAINMOD incor-
porated 20 years of historic rainfall and temperature
data for these evaluations. SCS estimated potential
evapotranspiration and appropriate factors for south
Florida conditions for the DRAINMOD model.
The output from DRAINMOD was then used in
an SCS-developed water budget model known as
WASTESTO. WASTESTO took the daily output files
from DRAINMOD to size waste storage ponds by
balancing all sources of water, including rainfall,
runoff, drainage, evapotranspiration, and wash
water. This water budget model allowed SCS en-
gineers to simulate field scale installations with vary-
ing drain spacings and depths, size of irrigation
sprayfield, size and depth of waste storage ponds,
and amount of dairy wash water to optimize land use
and provide alternatives for the landowners.
Floodwater-retarding structures, where needed,
were sized using the computer model, RESROUT,
developed by SCS. The required drainage rate
(gravity or pumped), South Florida Water Manage-
349
-------
flnoceedfngs of National RCWP Symposium, 1992
ment District allowable discharge rate or outlet
capacity, drainage area, and structure stage-dis-
charge curves were input into RESROUT, which
quickly evaluated the adequacy of the floodwater
retarding structure to retain runoff and drainage
water. SCS could also vary these inputs to evaluate
alternative sizes of floodwater-retarding structures
to help the landowner make decisions on the
drainage rate and size of floodwater-retarding struc-
ture based on the economics of the system.
Use of these computer models resulted in in-
creased productivity of field office staff and im-
proved quality of overall waste management system
plans.
Implementation
As the project evolved and construction occurred
concurrently with design surveys, SCS technicians
found it necessary to reduce time required to per-
form construction layout and checkout. SCS ac-
complished this by providing benchmark elevations
and setting initial alignments and control points for
the contractor. The use of total station surveying
equipment allowed for improved accuracy and
reduced time in setting the initial alignments by ref-
erencing existing points set during the planning
stage.
The construction check of installed practices
was improved by adopting Ohio Engineering
Software, developed by SCS in Ohio. This software
can compute earthwork volumes, complete as-built
drawings, and allows quick computer plotting of
planned versus as-built cross sections and quantity
computations.
Future Technology
With the ongoing development and introduction of
geographic information systems and global position-
ing systems and improvements in total station sur-
veying equipment and computer modeling software,
the quality and accuracy of BMP planning, design,
and implementation will continually improve.
Implementing these technologies will allow for
evaluations of several alternatives and their impacts
on water quality and quantity. New technologies will
provide the necessary alternatives for landowners to
make environmentally and economically sound land
use decisions.
AS programs and services of the U.S. Department of Agriculture, Soil Conservation Service, are offered on a nondlscriminatory basis, without
regard to race, color, national origin, sex, age, marital status, or handicap.
350
-------
A Method for. Ranking Farms and
Tracking Land Treatment Progress
in the St. Albans Bay Watershed
RGWP Project, Vermont
Richard J. Croft
CI,S. Department of Agriculture, Soil Conservation Service
.Winooski, Vermont
Jeffrey D. Mahood
U.S. Department of Agriculture, Soil Conservation Service
Harrisburg, Pennsylvania
ABSTRACT
The St. Albans Bay Watershed Rural Clean Water Program project strived to reduce agricultural
nonpoint source phosphorus loadings to the bay. In 1980 when the project was formulated, few
tools existed to evaluate phosphorus runoff from agricultural sources. The concept of critical acres
was used in the project; however, critical acres were broadly defined and did not adequately ad-
dress sources of nutrients on farms and implementation of BMPs that might affect them. The
project's comprehensive monitoring and evaluation program provided in-stream loading data, but
these values could not be readily related to individual farm operations. Soil Conservation Service in
Vermont developed several desktop computer models to estimate phosphorus loads from various
on-farm sources and employed them to evaluate the pre- and post-treatment phosphorus loading
conditions from each of the watershed's farms and at the subwatershed and watershed level. The
models were used to estimate pre- and post-BMP phosphorus loadings from various non-
point sources on each farm, array farms by phosphorus loads for priority implementation con-
siderations, and track implementation progress toward phosphorus reduction goals.
Challenges in nonpoint source management
projects include tracking progress and
determining when a satisfactory level of
treatment has been achieved — and both are
hindered when the project is formulated on per-
ceived land-based' problems without ideal charac-
terization of pollutant sources, their control, and
their relationship to the impaired waterbody.
In 1980, when the Rural. Clean Water Program
(RCWP) projects were created, the ability to evaluate
the land and receiving water relationship was
limited. Consequently,* most projects featured some
level (often 75 percent) of application of best
management practices (BMPs) to critical areas in
the watershed. Similarly, projects frequently used a
subjective rating procedure to rank applications for
assistance. Quantification of the nonpoint source
problem on each farm would have helped to better
align this ranking with the farm's pollution potential.
Progress tracking was often described in terms
such as critical areas treated, BMPs applied, farms
treated, dollars obligated, and erosion controlled.
These factors were important to track but gave no
clear indication of the pollutant load reduction level.
351
-------
Proceedings of National RCWP Symposium, 1992
Through experience in the RCWP and other
projects, Vermont has determined that a set of
simple planning models is most useful in planning,
implementing, and tracking nonpoint source water
quality projects.
Background
An explanation of the Vermont RCWP project plan of
work development and implementation management
is necessary to appreciate the need for an improved
tracking system.
In the St. Albans Bay watershed project, im-
paired water quality resulting in eutrophication had
affected uses of the bay. Phosphorus concentrations
in the bay were high and a major source was thought
to be farms in the watershed.
After considering numerous studies by the Ver-
mont Agency of Environmental Conservation, the St.
Albans Bay RCWP local coordinating committee
(1980) acknowledged phosphorus as the target
nutrient for nonpoint source management and
developed a phosphorus runoff control strategy "to
encourage the use of those management practices
on the farm which will have the largest effect in
phosphorus reduction." The committee commis-
sioned an assessment of agricultural runoff prob-
lems in the watershed to identify critical acres and
sources that needed treatment and to develop a plan
of work. This assessment included a 30 percent
sample of the farms to characterize their waste
management and soil loss status; later, this sample
was expanded to represent the total watershed.
Critical acres were defined as that "portion of
the farm on which animal wastes were spread or ex-
cessive cropland erosion was occurring" (Vt. RCWP
Coor. Comm. 1980). Critical sources were generally
considered to be the barn and adjoining areas where
the wastes were generated. The assessment did not
include phosphorus load estimations because Ver-
mont had not adopted a measurement technique.
The project set treatment goals as 75 percent of
the total critical acres and sources. Treatment needs
were based on an expansion of BMPs needed to ad-
dress these areas on sample farms. Table 1 provides
the baseline treatment needs and target goals con-
tained in the original plan.
Implementation of the
Vermont RCWP Project
At the start of the project, the committee recom-
mended an application rating system (adopted by the
U.S. Department of Agriculture's Agricultural Stabi-
lization and Conservation Service county commit-
tee) that was based on scores from questionnaires
completed by applicants (Fig. 1). The committee
designed these questionnaires to determine the
highest priority farms with critical acres and sources
needing treatment and the proximity of the farms to
streams and the bay, the presence of nonpoint
Table 1.—Treatment needs and target goals, St. Albans Bay RCWP original plan of work.
Information/Education
Treatment needs
Acres needing treatment
Sources needing treatment
Dairies (no.)
Truck farms
RCWP contracts No.
Ac.
TOTAL
NEEDS
100%
15,355
84
1
85
15,355
CUM.
TARGET
75%
11,584
63
1
64
11,584
1980 1981
TARGET TARGET
8 20
1,448 3,620
1982
TARGET
15
2,715
1983
TARGET
15
2,715
1984
TARGET
6
1,086
BEST MANAGEMENT PRACTICES (BMPs)
1 . Perm. veg. cover
2. Ag. waste mgt.
3. Strip-cropping
4. Terrace systems
5. Diversions
6. Grazing land prot.
7. Waterways
8, Cropland prot. cover
9. Conservation tillage
10. Stream protection
1 1 . Perm. veg. on crit. areas
12. Sediment retention struc.
14. Tree planting
15. Fertilizer management
Cons, cropping system
Pasture/hay management
Ac.
No.
Ac.
Ac.
Lf.
No.
Ac.
Ac.
Ac.
Lf.
Ac.
No.
Ac.
Ac.
Ac.
Ac.
93
75
2,520
40
73,733
21
32
1,162
116
5,733
88
6
20
815
7,200
7,800
70
60
1,890
30
55,300
16
24
872
87
4,300
66
4
15
611
5,400
5,850
5
3 5
1
50
100
3
11
350 450
400 450
10
12
100
2,000
1
2
50
200
6
50
500
500
10
10
150
4,000
1
2
50
5
300
8
100
500
550
10
10
150
4,000
2
2
50
10
400
8
1
2
100
500
600
352
-------
R.J. CROFT &J.D. MAHOOD
Lagoon
Other
QUESTIONNAIRE:
1. Do you spread manure daily?
2. Do you have a disposal system for milkhouse waste?
Check one:
•Filter field Holding pipe
Dry well Open pipe
3. Do you plant corn on your land or land rented by you?
4. Do you plant corn continuously oh the same pieces?
5. Do you plant corn on sloping pieces?
6. Do you feel you have an erosion problem on planting sloping pieces?
7. Do you plant a winter cover crop after corn harvest?
8. How often do you rotate corn pieces to grass cover?
9. Does your livestock have access to a stream, brook or river?
10. Distance your farm is from St. Albans Bay?
11. Does a major water course run through your land?
12. Do you have an erosion problem due to a water course running through yourland?
13. Do you have a water storage facility in grazing areas?
14. Year in which you will have the financial capability to carry out the
needed practices in your Water Quality Flan
SIGNED
YES
ALL CATTLE
NO. OF HEAD
ACRES
ACRES
YEARS
MILES
SPECIFY
YEAR
SCORING:
SCORE
1. If yes
2.
Filter field
Dry well
Lagoon
3. If yes
4. If yes
5. If yes
6. If yes
7. If no
8. Number of years
9. If yes
10. Number of miles
11. If yes
12. If yes
13. If no
10
2
4
6
Holding pipe 8
Open pipe 10
Other 12
5 for 10 acres or less
8 for 31 to 40 A.
10
10
10
10
10
10
10
5
6 for 11 to 20 A.
9 for 41 to 50 A.
7 for 21 to 30 A.
10 for 51 to 60 A., ect.
14. Enter year
Figure 1.—Application rating questionnaire for St. Albans Bay RCWP.
source phosphorus loads that could reach streams,
the potential for efficiently controlling more than one
major phosphorus source on the farms, and the
applicants' financial capability to follow an implemen-
tation schedule.
The committee kept records of the critical acres
treated and BMPs installed as the project pro
gressed (Fig. 2). The success of the project in terms
of physical effects was judged on these parameters,
as measured against the original project goals. Com-
prehensive monitoring and evaluation proceeded
concurrently; however, there was a lag between im-
plementation of BMPs and the period required to
process and evaluate monitoring data.
353
-------
Proceedings of Hatlonal RCWP Symposium, 1992
Cent, fam
«», »g>. N*ne
t 32
2
3
I
5
6
7
a
9
10
11
12
15
u
IS
14
17
IB
19
20
21
22
21
24
2*
26
27
24
29
30
It
32
33
34
JS
34
J7
M
39
40
41
42
43
44
45
44
47
4*
49
10
SI
52
S3
54
55
54
57
58
59
48
95
17
72
13
42
123
44
83
74
44
54
44
14 -,
53 Z
84
31
82
38
70
42
1
49
34
4
77
7t
S ,
27 *
2«
110
47
18
114
55
44
Cancelled FY SB
45
43
2
123
Cancelled 2/BS
89
91
5
41
197
150
114
292
35
99
30
104
185
120
142
154
74
40 C 63
41 r 79
42 Cancelled FT 08
U Cmdied 7/86 (DTP)
44 CM
4i 0 71
41 lolel
•• Average
' Ml contract! expire December 31st
Contract1
Period
9/80-86
9/UO-90
10/80-84
10/80-83
10/80-85
10/80-84
10/80-84
11/80-86
11/80-84
3/81-87
3/81-86
4/81-89
5/81-88
5/81-88
8/81-86
8/81-85
8/81-84
8/81-85
8/81-84
8/81-87
8/81-88
9/81 -88
11/81-85
11/81-89
11/81-85
11/81-85
11/81-85
11/81-84
11/81-85
4/82-86
5/82-85
5/82-88
2/82-87
5/82-85
5/82-86
6/82-85
6/82-87
6/82-88
9/82-87
9/82-85
12/82-87
2/83-86
5/83-86
5/83-87
5/83-90
5/83-89
5/83-86
6/83-86
10/63-86
1/84-87
3/84-87
5/84-88
7/84-88
7/84-88
9/84-87
4/85-88
7/85-88
7/85-88
8/85-88
8/85-89
8/87-90
of the year
Ho.
Tears
6.3
10.3
4.2
3.2
5.2
4.2
4.2
6.2
4.2
6.8
5.8
0.7
7.7
5.7
5.4
4.4
3.4
4.4
3.4
6.4
7.4
7.3
4.2
8.2
4.2
4.2
4.2
3.2
4.2
4.7
3.7
6.7
5.9
3.7
4.7
3.6
5.6
6.6
5.3
3.3
5.1
3.9
3.7
4.7
7.7
5.7
3.7
3.6
3.2
4.0
3.8
4.7
4.5
4.5
3.3
3.7
3.5
3.5
3.5
4.4
1=4
..
4.9
indicated.
Critical
Acres
150
260
207
162
90
94
199
126
52
110
164
81
425
87
370
262
212
155
184
198
309
146
154
191
162
161
217
142
166
124
310
159
7
358
U5
236
200
18J
154
144
227
41
315
156
455
251
168
176
226
175
188
87
147
160
105
74
189
87
295
280
219
11,277
185
Federal
Contract 3
Amount BHPs
$50.000 1 2 6
24,293 125 78 10
44,564 2 8 10
48,521 1 2 5
25,830 2
41,603 125
6,153 1 2 5
47,889 2
37,487 2
18,527 2
21,910 2
24,601
23.283
35,581
33,668
33.164
19,574
18.344
1,166
13,100
24,745
28.392
12,249
50.000
15,583
18,270
25.839
2
2
2
2 7
2 6
2
2 5
2
2
2
2
2
2
2
2
27,586 12
47,361 12 7
18,344 1 2
18,494 1 2
27,064 12
33,882 1 2
16,143 1 2
35,198 1 2
17,321 1 2 5
50,000 12
27,886 1 2
23,929
50.000
6,119
27.520
20.416
25,357
25,412
35,330
31,235
8,633
30,004
2,165
35,628
34,915
21,699
21,446
14,132
36,454
11,257
19.428
8,975
2
2 7
2 6
2
2 6
2
2
2
2
2
2
2
2
2
2
2
2
2
16,747 2 7
41,581 1 2
44,506 2
29.846 1 2
10
9
10
10
10
10
10
10
10
10
10
10
10
10
to
10
10
10
10
10
10
10
10
10
10
10
10
10
10
1,686,349
27,645
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
11
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
12
15
12
12
12
15
15
15
15
15
15
15
15
15
15
15
15
!5
15
15
15
15
15
15
•15
15
15
15
- 15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
15
16,
15
. Paid
10
Date
150,000
24,293
44,564
48,521
25,830
6,153
6,153
47,889
37,487
18,527
21,910
24,601
23,213
35,581
33,668
.33,164
19,574
18,344
1,166
13,100
24,745
28,392
12,249
50,000
15,583
18,270
25,839
27,586
47.361
18,344
18,494
27,064
33,882
16,143
35,198
17,321
50,000
27,886
23,929
50,000
6,119
27.520
20.416
25.357
25,412
35,330
31,235
8,633
30,004
2,165
35,628
34,915
21,699
21,446
14,132
36,454
11,257
19,428
8,975
747
41,581
44,506
29.846
1,686,349
27,645
* 01» • Dairy Temlnatlon Program carticipent.
3 KtPt are displayed In Table 1
Figure 2.—St. Albans Bay RCWP signed contract status as of September 30,1990.
Vermont Evaluation Models
At the same time, the Soil Conservation Service was
studying agricultural runoff and developing manage-
ment plans for other agricultural watersheds in Ver-
mont. To develop strategies for nonpoint source
phosphorus management, the Soil Conservation
Service developed simple desktop computer models
— now known as BARNY, FIELDS, MILK, and
STACK (Keeler, 1989) — to provide estimates for ex-
isting and treated condition phosphorus loading
from barnyards, fields where manure was spread or
where there was significant erosion, milkhouse ef-
fluent discharges, and runoff from stacked manure.
354
-------
R.J. CROFT &J.D. MAHOOD
The Soil Conservation Service also applied the
models to the St. Albans Bay watershed to provide
estimates of agricultural nonpoint source phos-
phorus loads. The model estimates afforded a com-
parison with observations from the comprehensive
monitoring and evaluation activities and provided
another means of tracking progress. These models
are particularly useful for tracking nonpoint source
project progress as well as planning treatment for
farms and project areas (Mahood, 1988).
Procedures for Using the Models
The models use data from on-farm inventories made
from a baseline data inventory worksheet, Form "VT-
PDM-1 (Fig. 3 shows the first page of this
worksheet). The inventory, which includes informa-
tion on crops grown; types, numbers, and manage-
ment of livestock; handling of milkhouse wastes;
manure management practices; and cropland
management practices, characterizes a dairy farm's
SECTION I: OPERATIONAL UNIT
Landuser
Farm Number
Town
Subwatershed Stream
Interest in project participation
(High, Med., low, None)
ACRES OWNED/RENTED OUTSIDE WATERSHED
Cropland Avg. Am. Row Crop
Continuous Hayland
All Other Land
TOTAL OUTSIDE WATERSHED
ACRES WITHIN WATERSHED
Cropland Owned
' Cropland Rented
Awg. Ann. Row Crop
Cont. Hayland Owned
Cont. Hayland Rented
Pasture
Woodland
Other
TOTAL IN WATERSHED
SECTION II; LIVESTOCK
t
Nunber
Milk Cows
Heifers
Beef Cattle
Pasture Season
(Manths/Yr)
CONFINEMENT (Hrs/Day)1
Pasture Season Non-Pasture Season
I/ Confinement - time livestock are in the barn and/or a scraped LCA resulting in
accunulated manure which must be removed.
SECTION III; MILKHOUSE WASTE SYSTEM; See Page 3 for Buffer info.
MILKING SYSTEM
( ) 1.BUCKET/DUMP STA.
( ) 2. PIPELINE
( ) 3. PARLOR
( ) 4.NON-DAIRY
Sheet Flow
BUFFER
DATA
Cover Factor
Length (ft)
Slope (%)
*Conplete Buffer Data
Notes:
NOW
Plan
or/
and
Channel
Now
1.0
Plan
1.0
WASTE DISPOSAL SYSTEM
Now
( )OUTSIDE WATERSHED
( ) DIRECT TO STREAM
( )THRU NON-SCS BUFFER
( ) LAGOON
(___)INTO MANURE STORAGE
( )THRO SCS FILTER*
___ SEPTIC FIELD/DRY WELL
(__) WORKS PROPERLY
( ) WITH OVERFLOW*
Plan S.O. use
0
0
1
1
3
5
6
Figure 3.—Baseline data Inventory worksheet.
355
-------
Proceedings of National RCWP Symposium, 1992
activities for both input to the planning models and
use in detailed planning and design of the BMPs.
Model Results
The models process data from the worksheets to
provide a farm-by-farm analysis (Table 2). The
analysis estimates phosphorus runoff values for ex-
isting conditions, the magnitude of which can be
evaluated for management at the source, farm, sub-
watershed, and watershed levels (Table 3) to enable
project managers to target certain subwatersheds
and farms for priority assistance.
The models provide estimates of phosphorus
loads from each of the various source types so the
significance of each source can be evaluated. Figure
4 shows that manure spreading was a principal
treatable phosphorus source, while cropland erosion
was least significant in the St. Albans Bay watershed.
Table 2 ranks farms in terms of critical or
treatable nonpoint source phosphorus. The critical
portion of nonpoint source phosphorus is con-
sidered to be that portion of total phosphorus that
can be reasonably controlled on the farm. This por-
tion includes 85 percent of the animal waste phos-
phorus (an average value found from model runs
Table 2.—St. Albans Bay watershed: farms ranked by pre-RCWP average annual critical total phosphorus (TP)
(Ibs/yr).
FARM
NUMBER
48
24
61
49
27
4
65
25
131
1
16
121
64
115
102
6
50
28
140
26
43
13
124
141
59
32
29
122
2
113
11
15
120
53
39
36
3
144
127
8
139
33
45
18
20
106
30
SHEET/RILL -
EROSION TP
(EROS. >T)
475
0
0
0
0
0
0
17
0
0
0
7
0
0
0
0
124
0
41
0
0
11
0
23
0
0
7
0
9
0
0
3
0
56
0
0
0
8
0
0
0
0
0
48
0
0
6
AGRICULTURAL WASTE (TP)
SPREAD
MANURE
139
384
234
243
147
162
144
125
70
182
125
8
159
3
78
156
52
122
51
144
90
14
79
168
165
69
102
43
137
31
89
148
59
105
107
57
79
142
142
72
77
20
63
49
114
63
75
STACKED
MANURE
14
0
0
0
0
0
0
0
172
0
0
179
0
198
45
0
0
0
73
52
0
0
17
0,
0
46
0
106
2
71
0
0
49
0
37
0
0
0
0
0
0
0
51
22
0
0
53
BARN-
YARDS
30
17
42
105
112
79
90
107
1
35
111
35
59
42
104
56
26
93
20
22
7
174
105
3
20
7
62
26
0
82
42
9
50
2
19
61
86
8
0
0
51
109
16
16
23
82
0
MILK-
HOUSES
0
35
132
34
38
41
37
19
27
46
16
22
25
0
11
25
11
20
27
0
105
0
0
0
11
71
19
10
34
0
51
20
21
0
9
54
0
0
18
84
28
24
22
6
9
0
7
85% OF
TOTAL
156
371
347
325
252
240
230
213
230
224
214
207
207
207
202
" 201
76
200
145
185
172
160
171
145
167 •
164
156
157
147
156
155
150
152
91
146
146
140
128
136
133
133
130
129
79
124
123
115
- CRITICAL
NPS*
PHOSPHORUS
631
371
347
325
252
240
230
230
230
224
214
214
207
207 ,
202
201
200
200
186
185
172
171
171
168
167
164
163
157
156
156
155 .
153
152
147
146
146
140
136
136
133
133
130
129
127
124
123
121
RCWP
CONTRACT?
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
No
Yes
No
No
Yes
Yes
Yes
No
Yes
Yes
Yes
No
No
Yes
Yes
Yes
No
Yes
No
Yes
Yes
No
Yes
Yes
Yes
Yes
No
No
Yes
No
Yes
Yes
Yes
Yes
No
Yes
356
-------
R. J. CROFT 6 J.D. MAHOOD
Table 2. — Continued.
FARM
NUMBER
14
148
116
12
38
23
34
21
19
128
22
155
51
129
135
118
57
35
152
62
37
56
40
7
44
31
10
143
41
9
147
47
149
108
46
17
138
58
134
126
5
133
150
60
114
54
52
142
132
55
137
Totals:
98
SHEET/RILL -
EROSION TP
(EROS. >T)
0
0
0
0
0
0
0
0
0
0
0
0
0
0
21
39
0
0
21
0
13
3
0
0
14
0
0
8
5
0
0
0
0
0
0
0
0
0
0
0
0
14
0
0
0
0
0
0
0
0
0
973
AGRICULTURAL WASTE (TP)
SPREAD
MANURE
141
72
33
19
123
128
54
82
50
114
77
113
79
0
4
0
52
37
79
74
68
27
49
48
32
31
48
58
0
17
73
31
9
45
39
45
7
16
0
7
29
11
22
20
3
14
12
0
6
0
0
7,185
STACKED
MANURE
0
0
73
0
0
0
15
0
0
0
0
0
0
0
4
28
0
17
0
0
0
0
24
0
0
36
0
14
13
0
0
13
37
1
7
0
27
0
13
14
0
0
0
0
6
0
0
0
0
0
0
1,529
BARN-
YARDS
0
55
28
113
8
0
0
13
45
0
6
0
8
15
11
31
36
27
0
0
13
9
7
3
25
18
15
1
26
39
0
22
12
18
9
2
12
0
17
9
0
0
0
0
8
0
0
9
0
0
0
2,916
MILK-
HOUSES
0
11
0
0
0
0
54
26
25
6
33
0
24
95
64
0
15 ,-,
22
0
23
0
53
12
37
13
0
22
0
33
18
0
0
6
0
7
0
0
18
0
0
0
0
0
0
0
0
0
0
0
0
0
1,786
85% OF
TOTAL
120
117
114
112
111
109
105
103
102
102
99
96
94
94
71
50
88
88
67
82
69
76
78
75
60
72
72
62
61
63
62
56
54
54
53
40
39
29
26
26
25
9
19
17
14
12
10 •
8
5
0
0
11,404
- CRITICAL
NPS*
PHOSPHORUS
120
117
114
112
111
109
105
103
102
102
99
96
94
94
92
89
88
88
88
82
82
79
78
75
74
72
72
70
66
63
62
56
54
54
53
40
39
29
26
26
25
23
19
17
14
12
10
8
5
0
0
12,377
RCWP
CONTRACT?
Yes
No
No
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
No
Yes
No
No
No
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
No
Yes
No
No
Yes
Yes
No
Yes
No
No
Yes
No
No
Yes
No
Yes
Yes
No
No
Yes
No
63: Y; 35: N
*NPS = nonpoint source
Table 3.—A comparison of agricultural nonpoint source phosphorus
loads by subwatershed — St. Albans Bay Watershed RCWP project.
ITEM
Total P (Ibs/yr)
Percent of total
Critical P (Ibs/yr)
Percent of total
SUB-WATERSHED
MILL
7,350
35.1
5,190
38.7
WETLAND
10,975
52.5
6,900
51.5
GUAYLAND
1,025
4.9
210
1.6
DIRECT TO BAY &
1,570
7.5
1,100
8.2
WATER-
HED TOTAL
20,920
100
13,400
100
357
-------
Proceedings of National RCWP Symposium, 1992
Spread Manure
49.4%
Barnyards
20.0%
Other Erosion
0.1%
Cropland Erosion
7.7%
Stacked Manure
10.5%
Figure 4.—St. Albans Bay watershed: agricultural
sources of treatable phosphorus, pre-project.
and monitoring) and total phosphorus associated
with enriched sediment yield at the watershed outlet
from accelerated sheet and rill erosion. Only sedi-
ment derived from erosion in excess of "T" (or the
soil's regeneration level) is considered. Erosion con-
trol below "T" is generally cost-prohibitive on Ver-
mont farms.
Use of Targeting and Tracking
Table 2 would have been very useful in ranking
farms for project assistance. Several of the higher
priority farms did not enter into contracts to treat
problems, while some of the lower priority ones did.
This information would have provided an oppor-
tunity to direct more attention to bringing the higher
load farms under contract. However, other farmer
variables — such as attitude, financial considera-
tions, and expected tenure — may have been
governing this process.
Aside from these variables — and assuming all
farmers were willing to participate — the project
could have achieved phosphorus load reductions
more quickly. Figure 5 compares the phosphorus
load reductions that would have been achieved each
year if the project had been based on treating for
critical phosphorus rather than critical acres. The
project might possibly have achieved the estimated
phosphorus reductions by 1982 rather than 1987, a
full five years earlier.
Also, assuming all farmers were willing to par-
ticipate, the project might have been more efficient
in achieving phosphorus load reductions. Figure 6
shows that 61 farms entered into contracts to
achieve the goal of treating 75 percent of the critical
acres. The same level of phosphorus reduction could
have been achieved by entering into contracts on
only 45 farms, if the highest priority farms had been
treated first At an average contract cost of $27,645,
this might have saved over $442,000 in cost-share
funds.
80 81 82 83 84 85 86 87 88 89 90
Figure 5.—Comparison of estimated phosphorus reduc-
tions by year: model priority (treat 75 percent of critical
phosphorus) vs. actual Implementation (treat 75 percent
of critical areas).
0 10 20 30 40 50 60 70 80 90 100
Farms with Contracts to Implement BMPa
Figure 6.—Comparison of estimated phosphorus reduc-
tions by farm: model priority vs. actual Implementation.
Conclusion
Managers have often evaluated progress in agricul-
tural nonpoint source projects (such as the St. Al-
bans Bay watershed) in terms of treatment
measures applied rather than units of loading reduc-
tion. The Vermont Soil Conservation Service has
developed planning models to estimate loads at the
edge of field, farm, and watershed level for various
treatment conditions, and the results can be used to
better plan nonpoint source projects. During im-
plementation, managers can track project progress
directly in terms of pollutant load reduction rather
than indirectly in terms of treatment measures ap-
plied.
References
Keeler, EM. 1989. Planning agricultural nonpoint source treat-
ment projects. Pres. at Natl. Nonpoint Source Conf . Nati. Ass.
Conserv. Distr., St. Louis, MO.
358
-------
R.J. CROFTS J.D.MAHOOD
Mahood, J.D. 1988. Tracking progress in agricultural nonpoint
source pollution control activities. Pres. at 43rd Annu. Meet.
Soil Water Conserv. Soc., Columbus, OH.
Vermont Rural Clean Water Program Coordinating Committee.
1980. Rural Clean Water Program St Albans Bay Project Plan
of Work. Agric. Stabil. Conserv. Serv., Burlington.
. 1983. St. Albans Bay Rural Clean Water Program 1983 An-
nual Report. Vt. Water Resour. Res. Center, Univ. Vt., Bur-
lington.
—. 1987. St. Albans Bay Rural Clean Water Program 1987 An-
nual Report. Vt. Water Resour. Res. Center, Univ. Vt., Bur-
lington.
—. 1988. St. Albans Bay Rural Clean Water Program 1988 An-
nual Report. Vt. Water Resour. Res/Center, Univ. Vt., Bur-
lington. , '
—. 1991. St Albans Bay Rural Clean Water Program 1991
Final Report. Vt Water Resour. Res. Center, Univ. Vt., Bur-
lington. ' ,
359
-------
-------
Elements of a Model Program for
Nonpoint Source Pollution Control
Steven W. Coffey
Jean Spooner
Daniel E. Line
Judith A. Gale
Jon A. Arnold
Deanna L. Osmond
Frank J. Humenik
National Water Quality Evaluation Project
NCSU Water Quality Croup
Department of Biological and Agricultural Engineering
North Carolina State University
Raleigh, North Carolina
ABSTRACT
Rural Clean Water Program (RCWP) projects have contributed significantly to the knowledge
necessary for reducing nonpoint source pollution and achieving water quality goals. An RCWP
evaluation, conducted during 1991 and 1992 by the National Water Quality Evaluation Project
(NWQEP) at North Carolina State University, shows that many of the 21 projects were highly effec-
tive and others had some effective elements. When expected results were not achieved, the
NWQEP attempted to analyze program and project deficiencies that may have affected the out-
come. Despite difficulties, the RCWP is the best program to date for evaluating agricultural non-
point source pollution control methods, and it should serve as a model for developing future
programs. RCWP was effective because it had good overall program management and institutional
arrangements that encouraged consultation between the U.S. Department of Agriculture (USDA)
and the U.S. Environmental Protection Agency, excellent program guidance, and effective techni-
cal support for reviewing reports and providing ongoing evaluation.
NWQEP has developed a model program and a model project based on the RCWP and added
refinements to strengthen weaker elements identified during the evaluation. The model program
includes a technical support group with access to resources for visiting project sites to assist in
project selection, monitoring, evaluation, and developing the plan of work. The model program
needs technical support from the USDA Agricultural Research Service (ARS) for developing and
evaluating best management practices. Increased assistance for monitoring is also needed and
should be provided by the U.S. Geological Survey.
NWQEP suggests three project levels, based on complexity and level of monitoring detail.
This paper discusses the NWQEP model and lists monitoring protocols that should be incor-
porated in program guidance to improve the chances of detecting water quality trends. All projects
need a preimplementation plan of work development to strengthen the problem definitions/select
the critical area, model the watershed to set treatment goals, and establish a means for land treat-
ment tracking. Projects also need a manager, technical support, and core project staff to improve
efficiency and encourage accountability. . . -•
A growing awareness of agriculture's con- have influenced Federal agencies to respond to
tribution to the nonpoint source pollution these concerns by developing a demonstration of
problem, increasing concern about water nonpoint source pollution control capabilities. In
quality, and pressure from special interest groups response to the 1989 President's Water Quality Initia-
__
-------
Pmceedtngs of National RCWP Symposium, 1992
(ive, the U.S. Department of Agriculture (USDA) has
developed programs that accelerate soil conserva-
tion and best management practice (BMP) im-
plementation on farms, ranches, highly erodable
lands, and watershed projects. Implementing these
programs produces many important benefits, includ-
ing increased adoption of soil conservation practices
and BMPs that improve water quality. Many
programs, however, do not target specific critical
area pollutant sources. With only limited targeting of
pollution sources (and even less water quality
monitoring to document the linkage between land
treatment and water quality), our knowledge of the
water quality benefits in these measures will not ex-
pand appreciably.
New nonpoint source control programs must
build on current knowledge to be effective. The
evaluation of the section 108a Great Lakes
Demonstration projects showed that nonpoint
source pollution is more persistent and more dif-
ficult to treat than previously thought (Newell et al.
1986). It also showed that using a pollutant runoff
model to determine critical areas is an efficient way
to use project funds. In addition, the Model Im-
plementation Program (MIP) evaluation demon-
strated that a project should target critical areas for
treatment to improve the likelihood of success, and
that BMPs should be selected and applied to pro-
mote water quality results (Nati. Water Qual. Eval.
Proj. and Harbridge House, 1983a,b).
Lessons learned from the Rural Clean Water
Program (RCWP) provide critical information about
nonpoint source pollution control technologies and
approaches for the U.S. Environmental Protection
Agency (EPA), U.S. Department of Agriculture
(USDA), and other Federal, State, and local nonpoint
source pollution control agencies and programs. The
RCWP is significant among nonpoint source control
programs because it combines land treatment with
water quality monitoring to document the effective-
ness of nonpoint source controls.
The RCWP has 21 projects located in nearly
every region of the United States that address a wide
range of water quality problems. The program is uni-
que in that it received a higher level of up-front fund-
ing for a longer period (10 to 15 years) than other
federally sponsored nonpoint source programs. The
longevity and dependability of RCWP funds en-
hanced efforts to establish a clear link between
water quality and land treatment, and several RCWP
projects have been able to demonstrate such a link.
The publication of RCWP rules and regulations in
the Federal Register (1980a) provided clear
guidelines for RCWP projects, facilitating the overall
program evaluation by standardizing many of the
projects' administrative and technical aspects. Final-
ly, the approach taken to address water quality
problems — providing Federal cost-share funds to
producers willing to implement BMPs — makes the
RCWP experiment important as a way to evaluate
voluntary versus regulatory approaches to the
problems of agricultural nonpoint source pollution.
Because of its unique characteristics as an ex-
periment in nonpoint source control, the RCWP is an
important source of insights and technology transfer
for the many ongoing and future nonpoint source
programs, including the 319 National Monitoring
Projects, other shorter-term 319 projects, the USDA
Demonstration and Hydrologic Unit Projects, the
Clean Lakes Program, and State nonpoint source
programs, among others. Because so many other
nonpoint source programs are being planned and
conducted, the need for clear articulation and dis-
semination of the lessons learned from the RCWP is
even more important. To share these valuable les-
sons in the most effective way possible, the National
Water Quality Evaluation Project (NWQEP) has re-
stated them as a set of recommendations for a model
nonpoint source pollution control program and
project.
NWQEP's evaluation of the RCWP has been con-
ducted to establish a set of recommendations for
developing Federal nonpoint source pollution con-
trol and water quality programs — programs whose
primary goal is to evaluate the water quality improve-
ments from nonpoint source controls. The objectives
of the evaluation were to assess
• cooperation among project team members,
committees, and agencies;
• agreement between the water quality
problem and the choice of solutions;
• project achievements;
• results of monitoring and assessment of
project impacts; and
• project findings to compile lessons learned.
Methods
For the program analysis, we reviewed the MIP
evaluation (Nati. Water Qual. Eval. Proj. and
Harbridge House, 1983a,b) and the section 108a
Great Lakes Demonstration Programs (Newell et al.
1986). We also reviewed literature for the USDA
President's Water Quality Initiative and the EPA's
section 319 Nonpoint Source Program (U.S. Environ.
Prot Agency, 1991). From these reviews we gained
valuable insights on methods that could be used to
evaluate the RCWP.
362
-------
S.W. COFFEVETAL
Including our own past experience, we used five
sources of information to evaluate RCWP projects:
1. an in-person interview questionnaire for
project personnel during site visits,
2. a short answer questionnaire administered
to project personnel,
3. a telephone survey of producers who did not
participate in the 21 projects,
4. 10-year reports from the RCWP projects, and
5. NWQEP's own 10 years' experience in
offering technical assistance to the projects
and performing program evaluations.
For site-visit evaluations, an interagency evalua-
tion team (led by a NWQEP member) visited each
project. In-person interviews of local and state
project staffs using a standardized questionnaire
were conducted during site visits (Coffey and
Smolen, 1991). Questions were designed to gather
specific information on project elements, including
State and local coordination, local program ad-
ministration, information and education, land treat-
ment, and water quality monitoring and evaluation.
Project staff responses to a short'answer ques-
tionnaire (Coffey and Hoban, 1992) were used to
gather information on project coordination, advisory
committees, project effectiveness, Information and
Education (I&E), farm operator participation, and
BMP implementation. A companion telephone sur-
vey of farm operators was used to determine factors
that influenced participation and BMP implementa-
tion (Hoban and Wimberley, 1992). RCWP projects
also produced detailed 10-year reports that provided
important insights, findings, and recommendations
for each project element.
For each RCWP project, the NWQEP wrote a
comprehensive analysis, including
• a project synopsis;
• a section on findings, successes, and
recommendations for each of the project
elements; and
• a detailed project description.
At the foundation of the analysis were the RCWP
regulations and the findings from individual RCWP
project evaluations. The results of the RCWP
analysis are presented here as a set of recommenda-
tions for a model program and a model project, in-
cluding selected examples from RCWP projects that
support the results.
Results
Based on NWQEP's review of agricultural nonpoint
source pollution control programs, the RCWP is, to
date, the best program available for achieving water
quality goals. For example, the RCWP had a set of
rules and regulations (Federal Register, 1980), tech-
nical oversight, and secure, long-term funding. Some
projects have documented water quality improve-
ments, and all projects have contributed to a greater
understanding of water quality problems and to
cooperation among agencies charged with address-
ing nonpoint source pollution.
The overall RCWP assessment has shown that it
was not possible to document water quality benefits
for RCWP projects in'which
,• agricultural activities were not the primary
pollution source, ,
• the areal extent and magnitude of land
treatment was inadequate,, or
• the monitoring designs were inadequate to •
document water quality improvements.
However, each project did have one or more
nonpoint source pollution control benefits, including
• development of cooperative relationships
among Federal, State, and local agencies
necessary to achieve an effective nonpoint
source pollution control program;
• achievement of widespread adoption of
BMPs to improve water quality under this
assistance program;
• visual improvements in water quality
associated with the use of BMPs; or
• water quality improvements documented by
water quality monitoring.
Therefore, the model program and project described
herein builds on the RCWP's structure and essential
features, while adding refinements to strengthen
weaker components identified during the RCWP
evaluation.
Elements of a Model Program for
Evaluating Nonpoint Source Pollution
Controls
Guidance written in the form of regulations must be
available to help implement the program (Federal
Register, 1980a). The model program's major fea-
tures (as outlined in the RCWP regulations) will in-
clude
363
-------
Proceedings of National RCWP Symposium, 1992
• clearly defined responsibilities of Federal,
State, and local agencies and landowners or
operators;
• criteria for project selection, approval, and
implementation;
• contracting requirements for technical and
financial assistance to farm operators;
• provisions for project funding and
termination;
• requirements for making cost-share
payments to participants; and
• plans for program and project monitoring and
evaluation.
The model program guidance will include these
important features and strengthen water quality and
land treatment monitoring, evaluation, and report-
ing.
Program guidance would also list the roles of
project staff at the Federal, State, and local levels,
and would help staff understand the responsibilities
of interagency counterparts.
The RCWP objectives were to
• achieve improved water quality in the most
cost-effective manner possible in keeping with
the provision of adequate supplies of food,
fiber, and a quality environment;
• help agricultural landowners and operators
reduce agricultural nonpoint source pollutants
and improve water quality in rural areas to
meet water quality standards or goals; and
• develop and test programs, policies, and pro-
cedures for the control of agricultural non-
point source pollution.
These objectives can be restated as model pro-
gram objectives that are relevant, comprehensive,
and nonoverlapping. Thus, the model pfbgram is to
• achieve improved water quality to restore and
protect the designated use of surface or
groundwater resources,
• help agricultural landowners and operators
reduce agricultural nonpoint source pollutants
and habitat perturbations, and
• develop, test, and evaluate policies and pro-
grams to control agricultural nonpoint source
pollution.
Program Administration and Management
The model nonpoint source pollution control pro-
gram should be administered by a single depart-
ment The USDA would administer the model
program in consultation with the EPA Administrator
and the Director of the U.S. Geological Survey
(USGS). Or, the Secretary of Agriculture could
delegate the responsibility of program administra-
tion to the ASCS, which has a long history of pro-
gram administration and leadership that contributed
to the RCWP's success. Local project funding would
have to be received on time through the State ASCS
office. Technical assistance for identifying and
documenting the water quality problem through
monitoring and evaluation could be provided by
EPA. Technical assistance for land treatment and
land treatment monitoring would become a joint
responsibility of USDA Soil Conservation Service
(SCS) and Extension Services. The Extension Ser-
vice (ES) should be responsible for information,
education, and BMP recommendations. The SCS
should be responsible for the development of farm
plans and structural BMPs.
The Agricultural Research Service (ARS) would
also provide technical assistance for developing and
evaluating BMPs. Technical assistance for water
quality monitoring and linking land treatment data to
water quality data would be coordinated by EPA,
with additional technical assistance on sampling, in-
strumentation, and data management from USGS
and ARS.
The model program will also need a national
coordinating committee to oversee functions cur-
rently defined for this committee by the RCWP, in-
cluding
• developing program regulations and
cost-share rates,
• reviewing technical aspects,
• selecting projects to fund based on a
technical assessment of likelihood of success;
• developing annual project reviews, and
• reviewing project progress.
The national coordinating committee should
have the ability to assign provisional status to
projects if State or local program staffs are not meet-
ing minimum performance standards. In addition,
the committee should have the authority to ter-
minate projects that fail to meet minimum require-
ments after two complete years on provisional status.
• Program Planning. Program planning is neces-
sary to ensure adequate attention to all project ele-
ments and stages of development. Problem iden-
tification, the selection of critical areas, and the
development of project proposals precede funding
(Fig. 1). The first two elements may extend into the
first year to allow refinement. Assistance from a tech-
nical support group is also needed before funding
and throughout the project.
364
-------
S.W.COFFEYETAL
All Prnirrts
PROPOSAL
Problem Identifica
Critical Area Selec
High-Level
Projects
Medium-Level
Projects
Low-Level
Projects
1 1
INITIAL FUNDING FIRM BUDGET PERIOD
ion ^.
ion ^
PlanofWor^
Development"^
Land Treatment Planning _ Land Treatment Implementation
Technical Assistance & Ongoing Project Evaluation from National Group
Information and Education
Intensive Water Oualitv Monitoring
Land Treatment / Use Monitoring
Water Oualitv Monitoring
Land Treatment / Use Monitoring
Water Oualitv Evaluation
Land Treatment / Use Monitoring
i i
01 2
Time Period (Years)
Form Establish Establish Evaluate Evaluate
Federal Guidelines National Plans Of Projects '
Program Technical Work
Support
Group
8-15
Complete
Project
Figure 1.—Model program and project timetable.
Project funding occurs in stages (Fig. 1). The
first two years are designated for initial funding only,
whereby projects may be terminated at any time.
Successful advancement to the firm budget period
(after which the project will be funded continuously
throughout its life) requires
• detailed and accurate problem identification,
• adequate'selection of critical areas based on
problem pollutants,
• a detailed plan of work for land treatment and
water quality monitoring, and
• demonstrated progress toward key agency
cooperation and the development of
institutional arrangements.
Land treatment planning before BMP implemen-
tation is critical to ensure adequate targeting of
resources. We suggest allocating two years for this
activity so that technical assistance may be sought if
needed. A baseline should also be established for
water quality monitoring before BMP implementa-
tion. Succeeding land treatment periods are for in-
stalling structural and management BMPs. Water
quality monitoring must be consistent before,
during, and after the implementation periods.
Program Technical Support
The model program must be structured so that
projects are carefully selected to improve chances
for meeting program objectives and obtaining water
quality improvements. A national technical support
group (outside the administrative organizations and
the national coordinating committee) should be in
place at program initiation to help develop program
and project guidelines. This group should also pro-
vide technical assistance during (and after) the plan-
ning period for project selection, critical area and
BMP designation, watershed modeling, land treat-
ment and water quality monitoring, and project
evaluation.
This technical support group must be provided
adequate resources for site visits to projects to gain
information and develop cooperative relationships
before project selection, during BMP and monitor-
ing implementation, and during final project evalua-
365
-------
Proceedings of National RCWP Symposium, 1992
tion. The support group may respond to program
and project technical requests from administrative
agencies and individual projects and be responsible
for verifying the accuracy and completeness of water
quality analyses. Finally, the support group may take
responsibility for the final evaluation by emphasizing
lessons learned, identifying water quality improve-
ments, and making recommendations for future
programs.
Project Selection
Project selection is a key factor for program success.
Selection criteria are needed to ensure that all pos-
sible projects are evaluated for their potential to con-
tribute to program objectives. Projects should be
selected because they have
• high priority water resources with docu-
mented water quality problems, or
• highly valued water resources threatened by
documented agricultural nonpoint source pol-
lution (because prevention of severe degrada-
tion is often more cost-effective than restor-
ation).
In addition, projects should have the following char-
acteristics:
• water resources having the highest public use
value (e.g., recreation or water supply) be-
cause these projects can show a significant
economic benefit;
• smaller watersheds of less than 30,000 acres
because problems in these areas can be more
readily identified, are easier to treat, and
respond more rapidly to treatment;
• the potential for effective control of nonpoint
source pollutants;
• the capability to use water quality models and
monitoring to determine if significant pollu-
tion reductions are likely with BMP im-
plementation;
• clearly stated objectives and goals related to
water quality impairments or conditions
threatening designated use;
• the ability to establish and maintain strong in-
teragency cooperation and institutional
project coordination;
• well-defined critical areas in which implemen-
tation of BMPs targeted to a specific pollutant
(or group of pollutants) can be emphasized;
• the potential for a high level of landowner par-
ticipation in the critical area;
• the potential that landowners will accept and
implement the necessary BMPs and, perhaps,
adopt alternative agricultural systems (e.g.,
changing from row crops to hayland or pas-
ture), which are integrally tied to water quality
improvements and project goals;
• a plan of work development process to obtain
baseline monitoring data, determine prob-
lems, refine critical areas and develop BMP
systems, conduct I&E programs, and docu-
ment effective project administration staffing
and cooperative relationships;
• the ability to conduct an effective I&E pro-
gram in advance to determine if key BMPs
(e.g., fencing or dairy waste use) will be ac-
ceptable to farm operators;
• the characterization of the hydrology arid pol-
lutant transport system to allow adequate
development of water quality goals and
monitoring systems; and ' .
• the ability to monitor explanatory variables,
such as season, stream discharge, water table
depth, precipitation, other hydrologic and
meteorologic variables, and land use changes.
The most successful RCWP projects were those
that met most or all of these criteria. The Florida,,
Idaho, Utah, Vermont, and Oregon RCWP projects
contained most of these elements and were among
the most successful projects in implementing land
treatment and documenting water quality improve-
ments as a result of RCWP treatment. For example,
the Utah RCWP project was relatively small (700
acres) with a well-defined critical area in which
BMPs were targeted to the major source of pol-
lutants (i.e., the dairies). Also, the project had a high
level of landowner participation in the critical area.
The Utah project's commitment to a two-year
preproject monitoring program proved to be the key
monitoring element that helped document substan-
tial water quality improvements.
Several other effective projects contained many
of the stated criteria but could have been
strengthened if the missing elements had been
present. The Nebraska RCWP project, for example,
suffered in its early years from the lack of clearly
defined water quality and land treatment goals. How-
ever, this project developed quantitative water
quality and land treatment goals, a critical area
definition that included BMPs targeted to sediment
and erosion control, and a strong I&E program —
resulting in a high level of landowner participation in
the critical area.
366
-------
S.W. COFFEYETAL.
The Delaware and Maryland RCWP projects
were successful but lacked preproject water quality
monitoring baselines, which impeded the ability to
make quantitative statements regarding water
quality improvements. The Iowa project contained
most of the suggested components but had only a
one-year pre-BMP monitoring database and initially
did not understand that the turbidity problem in
Prairie Rose Lake resulted not only from incoming
sediment but also from resuspended sediment and
algal growth.
Although the Massachusetts RCWP project met
several key project selection criteria in that the
Westport River estuary was a high priority resource
with significant economic value (shellfish beds), the
source of the water quality problem was not well
documented. This lack of clarity was one of several
factors that contributed to a lack of consensus within
the community and, therefore, to poor producer par-
ticipation. The Kansas RCWP project also lacked a
clearly documented water quality problem that could
be linked to a critical area pollutant source. Careful
application of project selection criteria could have
prevented the selection of this project and its sub-
sequent termination three years later.
The Michigan RCWP project had only vague in-
formation indicating that the Saline River was a large
contributor of nutrients (mainly phosphorus) to
Lake Erie. The project had not clearly identified the
critical pollutant source or critical area, and the
project did not document any water use impair-
ments. On the other hand, the Pennsylvania RCWP
project presented a documented water quality im-
pairment of agricultural origins and had the high
visibility of a project that could reduce pollutants
entering Chesapeake Bay. However, careful evalua-
tion of project potential would have shown that the
large number of small farms and the conservative
nature of the farmers would impede BMP accep-
tance and implementation, thereby limiting the
project's potential.
Program Funding
In the RCWP, all funds were identified and made
available at each project's initiation so that long-term
project planning and budgeting were possible. In
contrast, budgets for the current USDA Demonstra-
tion and Hydrologic Unit projects must be approved
each year. The associated delays have caused work
plan uncertainties, budgetary burdens on State and
local agencies, and incompatibility with fiscal budget
requirements. In the model program, funds should
be provided for preproject planning periods, which
may last from six months to two years (as defined
under "Project Proposal and Plan of Work Develop-
ment").
Elements of a Model Project for the
Evaluation of Nonpoint Source
Pollution Controls
The model program, which is based on RCWP
guidelines, carefully selects the individual projects
that will be undertaken. The model project is based
on the outline provided by RCWP regulations. The
following discussion of the model project is sup-
ported by examples from the RCWP projects. The
model project would operate under the primary
authority of USDA with consultation and concur-
rence from EPA. ASCS would be the administrative
lead agency. SCS should be responsible for the
development of structural BMPs, and ES should be
responsible for I&E and management BMPs. While
agencies supervise project activities, committees
would be responsible for setting priorities and coor-
dination. All agencies, committees, and program par-
ticipants would be guided by model program
regulations published in the Federal Register.
Project Administration and Management
Because of its management abilities, administration
for the model project at the State and local levels
should remain with ASCS. To implement the project
at the State level, each successful project must have
strong administrative and technical support from a
State coordinating committee, which also provides a
link to the national coordinating committee and the
local coordinating committee. The local coordinating
committee needs to have strong and continual sup-
port from the State coordinating committee, which
must establish and maintain open communication
lines and a willingness to allow the local coordinating
committee to implement the project.
The fundamental project administration and
management elements are a local coordinating com-
mittee, a county Agricultural Stabilization and Con-
servation (ASC) committee, a project manager, and
project advisory committees. The local coordinating
committee should provide guidance for the agencies,
community leaders, and citizens to oversee the ad-
ministrative and technical tasks of a local project.
The committee serves many functions, including
• assuring an adequate level of public
participation,
• developing a plan of work,
• enlisting the help of needed agencies,
• overseeing information and educational
activities, ,
• determining priorities for water quality plans,
367
-------
Proceedings of National RCWP Symposium, 1992
• enlisting the help of one or more agencies for
land treatment and water quality monitoring
and evaluation,
• developing a plan for critical area selection,
• creating a plan for implementing targeted
recommendations,
• establishing a plan for linking land treatment
and water quality data and analysis, and
• developing a plan for project reporting.
The Florida, Vermont, Idaho, South Dakota,
Pennsylvania, Oregon, Delaware, Utah, Maryland,
and Iowa RCWP projects all had strong local commit-
tees that contributed profoundly to the success of
their projects.
The county ASC committee, elected by county
farm operators, is responsible for encouraging
project participation and compliance. It can also play
a major role in promoting the project. The involve-
ment of the county ASC committee for the Appo-
quinimink River RCWP project in Delaware was a
significant factor affecting participation: BMPs were
implemented in 87 percent of the project's critical
area.
A project manager is also essential (Brichford
and Smolen, 1991). The manager should have a
water quality and management background, ideally
should work with the project from its inception, and
hold the designation for the length of the project.
The manager coordinates and monitors all project
activities, including project reports, and has the
authority to exert pressure on agencies or in-
dividuals not performing adequately. The project
manager is responsible to the local and State coor-
dinating committees and can report problems and
successes directly to the national coordinating com-
mittee.
Examples of RCWP projects that used managers
with positive results are South Dakota, Kansas, Vir-
ginia, and Minnesota. The South Dakota project
hired a temporary, full-time manager during its ini-
tial phase to conduct individual visits with farmers to
lay groundwork for their participation. The manager
also organized project activities and compiled infor-
mation so that the local coordinating committee
could operate quickly and efficiently. The position
continued until the last few years of the project.
Dkewise, Minnesota RCWP project recommenda-
tions suggested that a manager should be hired at
the program's start who is familiar with all govern-
ment agencies involved in the project but
autonomous. A half-time manager was hired in Min-
nesota after the project had begun. As a respected
area farmer, the project assistant was able to en-
courage the participation of his neighbors through
one-on-one visits, well testing, and newsletter
preparation.
Project advisory committees (e.g., administra-
tive, technical, I&E, land treatment and water quality
monitoring and evaluation, and modeling) are useful
for gaining progress in areas where input from a
smaller, more focused group improves decisionmak-
ing. Advisory committees should be formed,
disbanded, or regrouped as needed. For example, an
advisory committee comprised of land treatment and
water quality monitoring and modeling personnel
can help coordinate efforts to link land treatment
and water quality information. In the Vermont
RCWP, an advisory committee proved to be highly
effective; it ensured cooperation among agencies
and kept work activities on schedule. Similarly, the
key to success in the Florida RCWP project was the
.implementation of an administrative subcommittee.
The subcommittee (comprised of major agencies)
met regularly to coordinate project activities.
Project Proposal and
Plan of Work Development
Activities before project start-up influence the opera-
tion and success of each project and the total pro-
gram. Pre-project programs and periods are
specified in Figure 1 for three different levels of
projects based on problem magnitude, monitoring
intensity, and project complexity.
Initially, the Federal program administration is
formed to develop and publish program and project
guidelines. Thereafter, a proposal development
period without funding is specified for all three
project levels. The national technical support group
provides leadership for proposal evaluation and
determines which projects will be funded for plan of
work development.
The high-level, or most complex, projects are re-
quired to have baseline water quality monitoring
data or to initiate water quality monitoring during
the proposal development period. Monitoring of
water quality explanatory variables and land treat-
ment are to continue throughout the total project.
Medium-level projects may begin water quality
monitoring during the plan of work development and
continue for the total project. Land treatment
monitoring will be conducted throughout the project
period. Sampling design for water quality and ex-
planatory variables would be less comprehensive at
this level than in a high-level project.
Projects at the lower level may require periodic
water quality evaluation, such as visual examinations
or simple measurements of an unambiguous water
368
-------
S.W. COFFEYETAL.
quality problem, or a citizens' group may provide
monitoring.
After a successful two-year initial funding period,
a firm budget can be allocated and guaranteed for
the duration of the project as long as satisfactory
progress continues on the project.
Project Technical Support
To provide technical support for the project's first
two or three years, a minimum core project staff
must be created using individuals from the cooperat-
ing agencies. Core project staff will be responsible
for project activities and required to work cooper-
atively with the project manager. Core project staff
and the lead administrative agencies will have pri-
mary authority over project technical activities but
will also seek input from other agency staff, farm
operators, and local groups. Final technical decisions
need not require a consensus of local coordinating
committee members as long as decisions are consis-
tent with program guidance and recommendations
from the national technical support group.
Because they will be accountable for project
progress, the core project staff will have a great in-
vestment in the project. Agencies must establish a
mechanism for accountability and credit for good
performance. The minimum core project staff
should consist of a land treatment planner, and an
I&E specialist. In the Alabama RCWP project where
over 100 percent of the critical area was treated with
BMPs, an extension agent was instrumental in en-
couraging producer participation.
A full-time planner will be needed to help
develop farm plans, assist in BMP installation, help
farm operators maintain practices, and track land
treatment. Other core project staff positions beyond
the minimum (e.g., an engineer, a water quality
monitoring specialist, and an agronomist) may be
needed.
When an adequate level of technical capability is
not available at the project level, outside help should
be employed to assist the project. Core project staff
at the local level will enjoy greater freedom of com-
munication and have a larger team of experts for
technical support, compared to the limited com-
munication that happens when technical assistance
must be sought through line agency procedures. In
the Idaho RCWP project, ARS provided valuable re-
search and recommendations regarding the develop-
ment and evaluation of conventional and new BMPs,
particularly conservation tillage and no-tillage.
Because staff turnover can be problematic, in-
centives should be provided to encourage core
project staff to make a minimum commitment of
three years to the project. In the Louisiana RCWP
project, annual turnover of the SCS soil scientist
hired specifically to help implement the RCWP made
it difficult to track BMP implementation and main-
tain consistency.
Problem Definition
Water quality monitoring cannot be left as an after-
thought in an effective nonpoint source project.
Monitoring must be used to identify specific pol-
lutants (and their variability) responsible for the im-
pairment or threat to designated use. Initital problem
identification monitoring serves to help the project
team understand sources and response charac-
teristics of the affected water resource. The RCWP
projects have vividly illustrated that clear identifica-
tion of the source of the water quality problem and
acceptance of this information by the public and
producers are crucial to project success.
In Iowa, heavy sediment and a blanket of corn
stalks covering a recreational lake surrounded by
farmland helped make the problem and its source
especially clear. RCWP projects in Utah, Vermont,
Florida, Idaho, Nebraska, Oregon, and Pennsylvania
also had ample visual and analytical evidence of
problems in the receiving waters. In Massachusetts,
however, where both intensive dairy farming on
small acreages and booming residential devel-
opment were taking place adjacent to an estuary con-
taining important shellfish resources, the source of
the problem needed to be more clearly documented
to generate community support for project activities.
South Dakota's project required several intense
monitoring programs to gain a thorough under-
standing of the water quality problem and its causes
because of complex interactions between the surface
and groundwater sources feeding the target lakes.
Refinement of problem definition may occur as
the result of new information obtained from water
quality monitoring or modeling. Monitoring provides
a way to track BMP effectiveness and progress
toward water quality goals. Feedback on project ef-
fectiveness provided by monitoring is important to
land treatment personnel and farm operators. For ex-
ample, Vermont's RCWP project was able to reduce
bacterial contamination enough to reopen public
beaches for swimming. This accomplishment was
heavily promoted in the news media, which gave the
participating farmers pride and an investment in
nonpoint source control and their project.
Project Plan of Work and Time Frame
The plan of work is a written strategy used to or-
ganize agencies, project staff, and interested parties
for project implementation. An effective plan is dif-
369
-------
Proceedings of National RCWP Symposium, 1992
ficult to write, primarily because the linkage between
land treatment and water quality is not known with
certainty. A national technical support group is
needed to help the project address key obstacles,
define the water quality problems, and develop effec-
tive land treatment and water quality monitoring
strategies.
Project objectives and goals as stated in the plan
of work must be measurable, quantitative, and (for
the most part) attainable, given best available infor-
mation. Project objectives and goals must be critical-
ly reviewed to ensure consistency with overall
program objectives and goals.
• Time Frame. A model project should last from 6
to 15 years, depending on size and the ability to im-
plement land treatment. The median project length
should be 8 to 10 years, but some projects may need
12 to 15 years to implement enough practices and
document results. Larger areas could require long
periods to show improvement Examples of projects
that successfully made use of longer time frames are
the Idaho, Florida, Oregon, and Utah RCWP
projects. The long pre- and post-BMP water quality
and land treatment monitoring time frames for these
projects, along with high levels of BMP implementa-
tion, made it possible to track irrigation water
management, sediment control structures, and con-
servation tillage in the Idaho project, and animal
waste management in the Florida, Oregon, and Utah
projects. On the other hand, the Pennsylvania RCWP
project found that more time was needed than
originally expected to establish firmly the reduction
in nutrient levels from BMP implementation on
experimental sites.
• Critical Area Definition. Critical areas are pol-
lutant source areas in which the greatest improve-
ment in the water resource can be obtained for the
least investment in BMPs (Maas et al. 1987). The ef-
fectiveness of a nonpoint source pollution control
program is likely to be a function of where, when,
and how many BMPs are installed. Therefore, cost-
share funding should only be available for the treat-
ment of critical areas. Smolen (1988) reports that in
critical areas cause and effect are clear, hydrology is
simple, and response time to treatment is short. The
Utah, Oregon, and Vermont RCWP projects docu-
mented major reductions in bacterial concentrations
resulting from land treatment efforts in animal waste
management The project areas exhibit simple sur-
face water hydrology, and treatment occurred in the
critical areas. Bacterial populations, especially in sur-
face waters, respond to BMP implementation, thus
making bacteria in water a prime candidate to
demonstrate project effectiveness.
• Targeting BMP Systems. BMP systems directed
at water quality improvements are far more effective
than the installation and maintenance of individual
BMPs. In Oregon, for example, the development and
use of BMP systems to store and use manure were
essential in reducing fecal coliform levels in Til-
lamook Bay. However, whether a BMP system or an
individual BMP is to be used, each should be tar-
geted to control specific pollutants identified in the
water quality problem definition and project plan of
work.
For example, BMP systems used to control lake
sedimentation may be different from and target a dif-
ferent soil particle size than systems used to control
lake turbidity. The South Dakota RCWP project tar-
geted its BMPs to a specific problem; consequently,
nutrient management was found to be the most ef-
fective BMP for reducing nutrient contamination in
an area dominated by cropland with only a few scat-
tered animal operations. On the other hand, the Utah
RCWP project saw marked improvements in phos-
phates through animal waste management systems
in a watershed totally composed of animal opera-
tions.
Implementing the Plan of Work
Federal agencies and committees provide direction
and funding to support local administration and coor-
dination of project activities such as I&E and land
treatment Local committees, however, are respon-
sible for carefully defining project objectives and im-
plementing project activities to meet goals. In
addition, local committees receive guidance and sup-
port from the State coordinating committee and the
national coordinating committee.
• Information and Education. Extension Service
should provide leadership for the development, im-
plementation, and coordination of I&E programs for
agricultural nonpoint source water pollution control.
The local coordinating committee, the county ASC
committee, the soil and water conservation district,
and SCS should help with I&E efforts to ensure that
the I&E message is being received by participants.
During the proposal development, the com-
munity and relevant agencies must be informed
about problems in the project area, objectives, and
design. Local people also need to take part in
decisions from the start. An advance I&E effort
should be used to ensure that the majority of the
population and project staff agree about the problem,
its causes, and the treatment approach. The effect of
general and farm community support (or lack of sup-
port) was clearly demonstrated in several RCWP
projects. In the Iowa RCWP project, three public
370
-------
S.W.COFFEVErAL,
meetings were held to inform the community about
the RCWP before the Prairie Rose Lake project ap-
plication was submitted. This strategy of early com-
munity involvement helped the project to a strong
start. Delaware producers also participated in the
selection of and planning for the Appoquinimink
River RCWP project, again contributing to a success-
ful effort with strong producer participation. The
Westport River RCWP project in Massachusetts, on
the other hand, would have benefited from advanced
information and education programming (as well as
water quality monitoring for baseline data collection)
to address and resolve conflicting views on the
source of the water quality problem and the validity
of the approach being recommended in the RCWP
project.
Informational and educational efforts are take
part in stages that change over time. Initially, the
I&E team seeks to develop general awareness of the
water quality problem and support for the project
through mass media and public educational
programs. Then, I&E seeks to increase farm
operators' knowledge about nonpoint source control
and improve their agricultural management skills
through educational programs and one-on-one con-
tact. Ultimately, I&E works to modify behavior by
promoting the adoption of BMPs for improved
management of agricultural chemicals, conservation
of irrigation water, use of animal wastes, and conser-
vation of soil.
The I&E message was received and imple-
mented differently by the RCWP projects. For ex-
ample, in Vermont, the efforts of the local Extension
Service office were essential in informing producers
and convincing them to participate in the RCWP. In
Tennessee, every farmer received at least one (and
sometimes three) personal visit from an I&E team
member to encourage participation. In Florida, field
days, demonstration sites, and tours were the most
effective methods for promoting land treatment and
presenting accomplishments in the RCWP project.
Where fertilizer management and pesticide
management are important parts of the BMP pro-
gram, the I&E staff assists with soil sampling or pest
scouting and provides tailored recommendations to
project participants. The I&E program develops or
strengthens existing commodity associations to sup-
port integrated pest management and other special-
ized programs.
Extension Service can also initiate other
programs to improve water quality. A good example
is the Pennsylvania RCWP project. There the Exten-
sion office set up an animal waste trading exchange
to enable farmers who wanted animal manure to find
farmers who had excess manure. The Nebraska
RCWP developed a strong fertilizer testing and
management program, along with pest scouting.
Both components resulted in a significant decrease
in the use of fertilizers and pesticides.
• Producer Participation. Water quality improve-
ments depend on changes in farm operators' atti-
tudes, knowledge, and BMP implementation. Hoban
and Wimberley (1992) surveyed eligible participants
and nonparticipants from the 21 RCWP project
areas. Their findings on the farm operators' water
quality awareness, need for more information, at-
titudes about water quality problems, adoption of
BMPs, and participation in RCWP and other
programs provide significant information on ways to
improve education and participation in water quality
programs. In addition, results from the short answer
questionnaire (Coffey and Hoban, 1992) show that
cost-share funding was a key incentive to participa-
tion.
Other important factors affecting producer par-
ticipation in RCWP projects included:
• strong leadership within the farm community
(as demonstrated in Iowa and Oregon),
• consensus within the farm community and the
general public on the source of water quality
problems and the importance of water resour-
ces (for example, the high value placed on
local recreational lakes by the Iowa and
Delaware farmers in their projects' critical
areas),
• the threat of regulation if the sources of pollu-
tion were not voluntarily reduced (as in the
Taylor Creek-Nubbin Slough project in
Florida),
• economic penalties for producers who did not
participate (as in the Oregon RCWP project
where producers received lower milk prices
from the local cheese cooperative if they were
not implementing BMPs), and
• producer perception that BMPs implemented
to reach the project goals would also benefit
the farming operation (as in Alabama).
Producer participation also came about through
other means. Concern for stewardship of the land
encouraged many Pennsylvania farmers to par-
ticipate (many implemented BMPs but refused cost-
share funding). In Vermont, a long-standing
commitment to keep the community clean was the
impetus for participation.
• Land Treatment The Soil and Water Conserva-
tion District (SWCD) participates on the local coor-
dinating committee, prepares applications, and
371
-------
Proceedings of National RCWP Symposium, 1992
promotes the project. The SWCD, together with the
county ASC committee, determines the priority of
technical assistance among applicants for water
quality plans based on criteria developed by the local
coordinating committee. The SWCD also approves
water quality plans and revisions.
SCS coordinates technical assistance for BMPs
and recommends the appropriate agency for assist-
ance. SCS provides technical assistance for setting
priorities among applicants and developing and cer-
tifying their water quality plans. The role of SCS as
the lead technical agency for land treatment should
be retained; however, the contribution that can be
made by other agencies and opportunities for inter-
agency cooperation in achieving land treatment
goals should be recognized. As a result of the Mas-
sachusetts RCWP project, a new approach to farm
visits was developed by the local USDA agencies;
ASCS and SCS staff members now routinely visit
farms together to perform their duties under several
USDA programs.
The role of Extension Services should be ex-
panded to emphasize management practices to com-
plement structural practices. For example, during
the latter phases of the Pennsylvania RCWP project,
most of the land treatment effort was facilitated by
the ES through individual contacts and nutrient
management plans. For this project, the high num-
ber of farms needing animal waste storage facilities
and the resistance to installing such facilities made
the use of the ES and nutrient management plans the
only effective way to reduce nutrients in the area
streams.
• Water Quality and Land Treatment Monitor-
ing. The State water quality agency should par-
ticipate on the State and local coordinating
committees and monitor and evaluate the project's
effectiveness. Because Federal assistance is re-
quired to encourage consistent and continuous
water quality and land treatment monitoring
throughout the project period, Federal funding for
water quality monitoring must be authorized as a
part of the model program. Funding for monitoring
is required to document progress, the need for con-
tinued treatment, and water quality changes. Fund-
ing would be provided to all projects to meet
minimum monitoring requirements for both land
treatment and water quality.
Greater accountability by the State water quality
or other monitoring agency is needed to ensure ade-
quate water quality monitoring. Where applicable,
USGS, ARS, local universities, SCS, and Extension
Services should provide technical assistance for
monitoring program design and implementation.
Minimum monitoring protocols for high- and
medium-level projects should be reported in the
Federal Register, Projects would risk cancellation if
monitoring efforts fail to meet minimum require-
ments.
All approved projects should have monitoring to
determine BMP application progress and to docu-
ment trends in one or more variables related to the
water quality problem. Stream water quality monitor-
ing requirements for high-level projects should be
consistent with the EPA 319 National Monitoring
Protocol (U.S. Environ. Prot. Agency, 1991; Spooner,
1992). The protocol requires 20 samples per season
at a weekly or biweekly frequency for physical and
chemical variables and measurements of ex-
planatory variables (e.g., flow and.precipitation) for
each sample. If biological monitoring is desired,
biological and habitat variables should be monitored
one to three times per year. Land use and land treat-
ment data must be reported on a drainage basin rela-
tive to the water quality monitoring station. In
addition, paired watershed studies are strongly en-
couraged.
The protocol's main objective for high- and
medium-level projects is to monitor water quality and
land treatment simultaneously to determine if water
quality changes can be documented and associated
with changes in land treatment. Two features of this
objective must be met: (1) detecting significant or
real trends in both water quality and land treatment
implementation, and (2) associating water quality
trends with land treatment trends.
Guidance for minimum monitoring of land treat-
ment and associated water quality changes for the
model program and its projects should be main-
tained and enhanced by EPA and USGS in consult-
ation with other Federal, State, and local agencies.
This approach will allow valid technical evaluations
of individual projects. For example, the monitoring
requirements established by the EPA Clean Lakes
Program have been published in the Federal Register
(1980b). The lack of a complete and uniform
database has limited the effectiveness of evaluations
of the Model Implementation Program (MIP),
RCWP, and (by current indications) the present
USDA Demonstration Hydrologic Unit Areas as well
as Management Systems Evaluation Areas (MESA)
water quality project
The paired watershed approach involves
monitoring two or more similar subwatersheds
before and after BMP implementation in one of the
watersheds. This design is the most technically
sound and reliable method available to document
water quality changes in the shortest time period (3
to 5 years). The Vermont RCWP project employed
the paired watershed approach successfully and
372
-------
S.W. COFFEYETAL
demonstrated that winter storage of manure (instead
of winter spreading) was an effective nutrient
management strategy.
Land treatment information for high- and
medium-level projects should be reported and linked
directly to the water quality monitoring data. For ex-
ample, each observation should be paired hydrologi-
cally to a water quality monitoring station on an
annual or seasonal basis. All significant land use
changes and other nonpoint and point source control
efforts should be documented. The monitoring
design should include multiyear monitoring of both
land treatment or use and water quality before and
after BMP implementation.
Several RCWP projects had strong water quality
monitoring programs emphasizing pre- and post-
BMP monitoring and above and below site testing in
combination with, a large land treatment effort.
These projects were able to document substantial
water quality improvements. In the Utah RCWP
project, animal waste management systems reduced
phosphorus concentrations leaving the watershed by
75 percent and reduced nitrogen and fecal coliform
by 40 to 90 percent. In the Florida RCWP project,
fencing, water management, and animal waste
management systems reduced phosphorus con-
centrations in water entering Lake Okeechobee by
45 percent.
In the Oregon RCWP project, animal waste
management systems installed on dairies reduced
bacterial contamination of oyster beds,by about 40 to
50 percent. Sites in the bay restricted to shellfishing
based on Food and Drug Administration classifica-
tion decreased from 12 in 1979-80 to 1 in 1985-86. In
the Idaho RCWP project, water management and
sediment control BMPs reduced sediment loads in
return flows from irrigated land by 70 percent. Trout
fishing has been partially restored to this coldwater
trout stream.
Likewise, the Idaho and Nebraska RCWP
projects realize that a substantial effort would have
been saved if they had established clear protocols in
the beginning for documenting water quality and
land treatment on a subbasin and annual basis such
that the two databases could be linked hydrological-
ly and temporally. Both projects have taken the initia-
tive to reconstruct and link the two databases. For
these projects, the land treatment databases were
the most difficult to reconstruct.
The Vermont project used extensive monitoring
of BMP implementation and agricultural activities to
establish a link between cows under BMP manure
management and bacteria levels in streams. The
Minnesota RCWP project used vadose zone monitor-
ing to establish the relationship between agricultural
practices, best management practices, and ecologi-
cal niches to groundwater contamination.
Explanatory variables, which should be
monitored in the high-level projects, can include
other land-use changes, the seasons, stream dis-
charge, precipitation, groundwater table depth, im-
pervious land surface area, and others. In Alabama,
technicians were unable to determine the cause of a
sudden increase in fecal coliform levels in a par-
ticular stream until they determined that beavers
had built a dam upstream of the sampling site. The
Florida RCWP project confirmed that the changes in
cow numbers and water table depth affected the
water quality monitoring results and that documenta-
tion and adjustment for these changes allowed valid
conclusions to be made regarding changes in water
quality.
Evaluation and Reporting
Regular review of progress helps ensure that the
project is working toward its goals and that its ac-
tivities are on track. As part of the evaluation
process, regular meetings must be held by the local
coordinating committee to keep the project team in-
formed and to coordinate activities. In addition,
quarterly meetings of technical groups can help
guide the project to water quality improvements.
Both the South Dakota and the Vermont RCWP
projects used frequent meetings of technical staff to
identify needs and document progress.
Annual progress reports on the projects create
an opportunity to compile and analyze findings; an-
nual progress reviews by the national coordinating
committee and the national technical support group
can help projects meet their goals.
Feedback Loop
Regular meetings are a must for project staff, if water
quality and land treatment monitoring are to be used
to make mid-project adjustments. The project
manager can facilitate communication by scheduling
local coordinating committee meetings on a quarter-
ly basis. The State and local coordinating commit-
tees should meet jointly at least once each year.
Regional workshops should be scheduled to pro-
vide information transfers between projects with
similar hydrology and agriculture. National work-
shops are helpful and especially beneficial if all
projects are represented. Some of the most impor-
tant RCWP lessons were learned from projects that
were seldom represented at national RCWP
workshops.
373
-------
Proceedings of National RCWP Symposium, 1992
In the early part of the Virginia RCWP project,
extremely high levels of coordination and coopera-
tion existed among the different agencies, and com-
munication was excellent. However, after BMP
implementation, which occurred about five years
into the project, both the State and local coordinating
committees stopped meeting, which caused a break-
down in communication between the land treatment
and water quality groups.
Conclusion
The Rural Clean Water Program has demonstrated
that nonpoint source pollution control programs can
be successful in protecting and restoring water
resources if they are carefully structured and based
on the findings of previous programs. The model
program we propose requires administrative and
technical support from all levels — Federal, State,
and local. The States and their local counterparts
need guidance on project implementation. Much of
this guidance can best be communicated through
program regulations similar to the regulations writ-
ten for the RCWP (Federal Register, 1980a). A nation-
al technical support group, independent of
designated cooperating agencies, should be in place
to help develop program guidance, provide technical
assistance, and conduct project evaluations.
Water quality monitoring is required to docu-
ment the problem and track project effectiveness.
We suggest minimum monitoring requirements to
guide the development of the monitoring program
design. BMP systems must be targeted to treat criti-
cal areas and specific pollutants responsible for the
present or potential problem. Finally, a project
manager and a core project staff (from various coor-
dinating agencies) are needed to implement the
project. Greater accountability among project staff
and incentives to avoid turnover will improve the
likelihood of meeting project goals. Information and
educational efforts should be expanded to en-
courage greater adoption of BMPs.
Continual evaluation of programs and projects
and full communication of technical information are
key factors in controlling nonpoint source pollution
and achieving water quality goals.
References
Brichford, S.L and M.D. Smolen. 1991. AManager's Guide to NFS
Implementation Projects. NCSU Water Qual. Group, Biolog.
Agric. Eng., North Carolina State Univ., Raleigh.
Coffey, S.W. and M.D. Smolen. 1990. Results of the experimental
Rural Clean Water Program: methodology for on-site evalua-
tion. /» Rural Clean Water Program 1990 Workshop Proc.
Nonpoint Source Branch, U.S. Environ. Prot Agency,
Washington, D.C.
Coffey, S.W. and T.J. Hoban. 1992. Rural Clean Water Program
Methodology for Evaluation, Short Answer Questionnaire.
Coop. Ext Serv., Dep. Biolog. Agric. Eng., North Carolina
State Univ., Raleigh.
Federal Register. 1980a. 1980 Rural Clean Water Program
(RCWP). 7 CFR Part 700. March 4, 1980 (45 F.R. 14006).
Reprinted by Agric. Stabil. Conserv. Serv., U.S. Dep. Agric.,
Washington, DC.
. 1980b. Cooperative Agreements for Protecting and Restor-
ing Publicly Owned Freshwater Lakes. 40 CFR Part 35.
February 5,1980. U.S. Environ. Prot. Agency, 45 (25) :7788-99.
Hoban, TJ. and R.C. Wimberley. 1992. Farm operators' attitudes
about water quality and the RCWP. /» Proc. National Rural
Clean Water Symposium, 1992. Orlando, Florida.
Maas, R.P., M.D. Smolen, C.A. Jamieson, and AC. Weinberg. 1987.
Setting Priorities: The Key to Nonpoint Source Control. Off.
Water, Reg. Stand., U.S. Environ. Prot. Agency, Washington,
DC
Newell, AD., L.C. Stanley, M.D. Smolen, and R.P. Maas. 1986.
Overview and Evaluation of Section 108a Great Lakes
Demonstration Program. EPA-905/9-86-001. U.S. Environ.
Prot Agency, Washington, DC
National Water Quality Evaluation Project and Harbridge House,
Inc. 1983a. The Model Implementation Program: Lessons
Learned from Agricultural Water Quality Projects — Execu-
tive Summary. Biol. Agric. Eng. Dep., North Carolina State
Univ., Raleigh.
. 1983b. An Evaluation of the Management and Water
Quality Aspects of the Model Implementation Program: Final
Report. Biol. Agric. Eng. Dep., North Carolina State Univ.,
Raleigh.
Smolen, M.D. 1988. Targeting critical water quality areas. Pages
131-41 in Rural Clean Water Program 1988 Workshop Proc.
Natl. Water Qual, Eval. Proj., Agric. Ext. Serv., North
Carolina State Univ., Raleigh.
Spooner, J. 1992. Linking land treatment to water quality. In Proc.
1992 National Monitoring and Evaluation Conference,
Chicago, IL U.S. Environ. Prot Agency, Washington, DC. In
review.
U.S. Environmental Protection Agency. 1991. Watershed Monitor-
ing and Reporting for Section 319 National Monitoring Pro-
gram Projects. Assess. Watershed Prot Div., Office Water,
Washington, DC
Young, RA, C.A Onstad, D.D. Bosch, and W.P. Anderson. 1987.
AGNPS: Agricultural Nonpoint source Pollution Model, A
Watershed Analysis Tool. Conserv. Res. Rep., No. 35. U.S.
Dep. Agric., Washington, DC.
374
-------
Extending the RCWP Knowledge
Base to Future Nonpoint Source
Control Projects
Paul D. Robillard
Department of Agricultural and Biological Engineering
Penn State University
University Park, Pennsylvania
ABSTRACT
The Rural Clean Water Program (RCWP) and its associated database and practice implementation
experiences represent a wealth of information that can and should be used in current and future
nonpoint source watershed projects. This paper presents an expert system configuration based on
lessons learned from the RCWP experience. The RCWP EXPERT system emphasizes design of
nonpoint source control systems. Design methods include site-specific characterization of
problems. Water quality monitoring information, contaminant properties, transport, and mobility
are accounted for in the practice design process. Examples and data from RCWP projects are in-
cluded in the system. Powerful linkages to AGNPS, climate data, the GRASS geographic informa-
tion system, and Soil Conservation Service practice design modules add computational power to
RCWP EXPERT control system design solutions.
The experimental Rural Clean Water Program
(RCWP) was initiated in 1981 in 21 agricul-
tural watersheds located throughout the
United States. To date, this effort represents the
most intensive water quality monitoring and im-
plementation of nutrient, sediment, and pesticide
reduction practices undertaken in the United States.
With almost 10 years of monitoring data from a
variety of hydrologic and physiographic regions,
RCWP projects represent state-of-the-art approaches
for implementing best management practices
(BMPs) and controlling nonpoint sources of water
pollution.
Before RCWP, nonpoint source control practices
involved applying traditional soil and water conserva-
tion practices to cropland primarily to control
erosion and improve surface water quality. In many
cases, participation was based solely on willingness
to cooperate in cost-sharing programs. RCWP
projects have greatly advanced these early efforts by
achieving several generations of practice designs,
site selection criteria, and direct linkages with
specific water quality goals.
The RCWP Knowledge Base
and Technology Transfer
The completion of the RCWP provides an oppor-
tunity to extend project results to current and
planned nonpoint source watershed projects, includ-
ing hydrologic unit areas and demonstration water-
shed projects. This paper presents two components
of the RCWP experience that will be critical to the
success of future nonpoint source watershed
projects. The first, documentating the RCWP
375
-------
Proceedings of National RCWP Symposium, 1992
knowledge base, involves identifying and recording
successful components of nonpoint source planning,
design, and implementation. This knowledge base
includes engineering designs, monitoring systems,
innovative water quality control practices, animal
waste systems, beneficial uses, and successful im-
plementation methods. Other areas of knowledge,
such as institutional and economic factors, are equal-
ly important and should be documented.
The second topic is technology transfer. This
transfer can occur in several ways, for example,
through published RCWP reports, conferences,
databases, traditional training, and workshops. But
another method of technology transfer uses state-of-
the-art computer technology to develop a system for
designing nonpoint source control systems. This
type of system uses the RCWP knowledge base and
directly facilitates improved control practice design
procedures. The RCWP EXPERT system is
presented here as an example of a computer-aided
design for technology transfer.
Documenting the Knowledge
Base
Engineering Design
The RCWP has initiated new or modified design
methods for several types of practices to achieve
water quality objectives or adapt to site-specific con-
ditions. For example, improved design procedures
for surface and subsurface controls in high source
areas (such as confined livestock areas, conserva-
tion tillage, and off-site sediment retention) were in-
itiated in several projects. In particular, animal waste
system designs involved two and three generations
of practice designs at some RCWP sites, resulting in
more efficient floor designs, better pump selection
with higher reliability, and improved manure storage
unloading ramps, augers, and pumps. Because
animal waste disposal is a serious water quality prob-
lem in many agricultural regions, the engineering
design and implementation experience gained from
RCWP projects can be a cornerstone for future non-
point source control projects.
• Animal Waste Systems. As many RCWP
projects demonstrated, simply storing manure is not
an effective BMP. The design and development of in-
tegrated collection, storage, unloading, and field ap-
plication systems is the key to success. The
Tillamook, Oregon, RCWP project developed dual
animal waste handling systems for both collection-
storage and field application of solid and liquid
animal waste systems. The transfer and unloading
systems were changed and improved to incorporate
special farm requirements. Other examples include
the following:
• animal waste collection, storage, and transfer
systems using both gravity and pressurized
components were implemented in RCWP
projects in Vermont, Oregon, Maryland, and
Alabama;
• integrated animal waste storage and disposal
systems were implemented in Oregon, Utah,
and Idaho;
• swine floor design and collection systems
were implemented in Virginia; and
• a solid separator to remove heavy organic ag-
gregates from confined livestock runoff is
now in use in Maryland.
Special attention was focused on animal waste
systems in Virginia. A flushing system was devel-
oped together with lined lagoons and spray irrigation
systems to provide total system control. Because of
the proximity of water supply reservoirs, the irriga-
tion disposal systems were specifically managed to
minimize any effects on nearby streams.
The Alabama RCWP project improved animal
waste uses by creating waste management systems
with holding ponds, lagoons, and spray irrigation
systems. An associated program in which the con-
servation district provided equipment loans or
helped purchase used equipment was particularly
helpful for developing farm capabilities in lagoon un-
loading and spray irrigation of stored manure. Thus,
the control system included technical assistance and
supporting practices design through the pollutant
delivery phase.
Monitoring Systems
Before RCWP projects were initiated, the design and
selection of nonpoint source control practices were
conducted without comprehensive and accurate
water quality monitoring data. Without such data, no
scientifically defensible method existed for judging
the effectiveness of control practices. RCWP projects
introduced event-based, long-term water quality
monitoring strategies to nonpoint source programs.
Florida and Vermont developed comprehensive and
specialized monitoring systems to characterize the
many and varied sources of contaminants. Projects
in both States used monitoring data to select and
place BMPs and to evaluate the effect of nonpoint
source pollutant loads on the water quality of receiv-
ing lakes and streams. The long-term experience
and advances in monitoring design and evaluation
376
-------
P.O. ROBILLARD
methods are the RCWP's unique and vital contribu-
tion to the scientific basis of nonpoint source con-
trols.
Perhaps the most important element in design-
ing effective control systems is how water quality
monitoring data is used to modify practice design
and site selection. The extent to which water quality
data are used to identify high source areas or critical
seasonal loads is essential for developing effective
control systems. In fact, if a BMP does not change
contaminant mass loading, it is, by definition, not an
effective practice. Here are other examples of RCWP
projects that used monitoring data to isolate or iden-
tify high source areas:
• St. Albans Bay, Vermont, used four levels of
water quality monitoring to identify spatial and
temporal mass contaminant loads, including
bay sampling, trend stations, mobile monitor-
ing, and random sampling. The St. Albans Bay
monitoring effort resulted in the isolation of a
subwatefshed Qewett Brook) and the iden-
tification of several farms that had high direct-
to-bay loading.
• Also in St. Albans Bay, similar watersheds
were paired to provide comparative loading
and effectiveness data. This approach is par-
ticularly promising for watersheds with
similar uncontrolled variables, such as soil,
cropping, and livestock systems.
• The Rock Creek, Idaho, RCWP project used
extensive biomonitoring methods to identify,
track, and evaluate contaminant effects and
the effectiveness of control practices.
• The Alabama RCWP project used bacteria
counts to isolate problem areas.
• The Virginia RCWP project developed com-
prehensive integrated worksheets to charac-
terize farms with high loading potential. This
effort focused BMPs on farms with a high im-
pact on nearby water supply reservoirs.
• The Snake River, Utah, RCWP project
recorded dramatic reductions in nutrient load-
ing and related effects on the Deer Creek
Reservoir after intensive treatment of a prob-
lem watershed, which their monitoring pro-
gram had identified as critical.
Innovative Practices
RCWP projects have inspired the design of innova-
tive practices and systems that address the special
implementation problems and site-specific design re-
quirements of control practices that are effective and
compatible with modern agricultural systems. Ex-
amples of innovative practices in RCWP projects in-
clude:
• Several projects practiced integrated control
and disposal of surface and subsurface water
flows in high intensity areas, such as confine-
ment facilities and satellite pasture areas, by
specifically designing these areas for produc-
tion and water pollution control.
• Vegetative filter treatment systems were
designed for milkhouse wastes in Vermont.
• Short- and long-term manure storage in-
tegrated with land application systems,
milkhouse waste disposal, and pasture man-
agement to achieve water quality loading ob-
jectives evolved during the RCWP project in
Vermont In each case, special implementation
problems or farm limitations were incor-
porated into practice design over time.
• Conservation tillage systems to reduce con-
taminant loading to streams were developed
in Alabama, Idaho, Virginia, and Maryland.
• Roof runoff control structures were con-
structed in confinement areas and high rain-
fall zones in Oregon.
• Intensive treatment of high loading areas
resulted in maximum reductions of contam-
inant impact in Florida, Virginia, Oregon, and
Utah.
• Small tributary control and relocation through
high intensity areas were practices imple-
mented in Utah.
• In Idaho, off-site sediment retention basins
decreased delivery of sediment and adsorbed
pollutants to receiving streams.
• Modified drop inlets for safe runoff disposal
were constructed in Virginia wherever eleva-
tion differences of 2 to 5 meters existed be-
tween edge of field and stream discharge
points.
Beneficial Uses
Because of advances in monitoring system design at
a few sites, the RCWP has begun to establish a
relationship between land treatment and water
quality improvements and beneficial uses. RCWP
projects were the first in the Nation to directly relate
the application of nonpoint source control practices
to specific water quality objectives in receiving lakes,
streams, and estuaries. RCWP projects in Florida,
Vermont, Oregon, Maryland, and Idaho are excel-
377
-------
Proceedings of National RCWP Symposium, 1992
lent examples of projects that relate nonpoint source
control programs in upland watersheds to direct ef-
fects on drinking water supplies, fish populations,
shellfish industries, and recreational uses.
• Farm Water Supplies. One example of a benefi-
cial use that evolved from RCWP activities relates to
improved quality of farm water supplies. Of all rural
water supply systems, farm supplies are often the
most at risk. Poor siting, well construction, protec-
tion, and maintenance cause many systems to be in-
adequate. Robillard and Walter (1982), Sharpe et al.
(1985), and others have found the incidence of con-
tamination of farm wells to be high. Surprisingly, at a
time when modern production practices and com-
puter-assisted farm management is common, many
farms are still using 19th century water supply tech-
nology. Not only is the productivity of livestock af-
fected; the health of the farm family is also at risk.
Four RCWP projects solved dual water quality
problems by developing and protecting farm water
supplies. Although in some applications protection
measures resulted in improved quality of the entire
farm system, improved water quality for livestock
consumption was the principal objective. In all in-
stances, high source loading of nutrients was
eliminated as livestock were moved away from
streambanks. As part of its high intensity use areas
program, the Taylor Creek, Florida, RCWP project
included water troughs, shade, and feeding areas in
satellite pastures. Low-cost trough units were
designed in Tillamook Bay, Oregon, to improve live-
stock water quality and decrease associated loading
to streams. Other examples include limited access
gravity and pumped livestock water supply systems
(Maryland) and combined livestock watering and
streambank protection systems (Alabama).
The Maryland RCWP project developed and
protected livestock water supplies with trough sys-
tems that not only improved livestock water quality
but also greatly decreased biological oxygen
demand, nitrogen, and bacteria loading to streams.
Alabama also developed farm water supply systems
to improve access to potable water and improve the
quality of water for livestock consumption. By limit-
ing the access of livestock to streams, localized
streambank erosion was also reduced.
Technology Transfer
From their inception, RCWP projects have been the
source of emerging technology for nonpoint source
control programs. Two examples of the far-reaching
influence of these projects are Oregon and Idaho.
New and innovative designs for animal waste sys-
tems in Oregon have been directly incorporated into
the Oregon Engineering Handbook (U.S. Soil Con-
serv. Serv. 1989), a state-of-the-art reference for en-
gineers who design nonpoint source pollution
control practices. Idaho's project provides another
example of RCWP technology that has attained na-
tional importance. In Idaho, intensive work with dif-
ferent types of water quality monitoring systems
resulted in the development of effective low-cost
biomonitoring methods that complement and, in
some cases, substitute for traditional inorganic and
organic water sample analysis techniques. Technol-
ogy transfer of animal waste systems, vegetative fil-
ter strips, conservation tillage systems, and high
intensity water control practices also occurred in the
Vermont, Florida, Virginia, Alabama, and South
Dakota projects.
The massive, long-term implementation of con-
trol systems through the RCWP is an excellent
source of new field-verified technology. Tillamook
Bay, Oregon, has used several methods of dissemi-
nating technology that emerged from the RCWP ef-
fort. Most important, new plans and calculation
procedures for waste storage, facilities, waste stack-
ing facilities (covered and uncovered), covered liq-
uid manure storage, and watering troughs have been
included in the Oregon Engineering Handbook. In
addition, RCWP project personnel are used in State
and regional workshops to assist in training and
other educational programs oriented toward new
water quality initiatives.
Extending the Technology
RCWP EXPERT
Those who design and select BMPs to control
nutrients, pesticides, and sediment loads in agricul-
tural watersheds must consider the complex interac-
tions between hydrologic, soil, crop, tillage, and
management variables. In particular, when con-
taminants exhibit different physical and chemical
characteristics resulting in varying mobility and
transport patterns in surface and groundwater
regimes, the effectiveness of control practices is dif-
ficult to predict. One aspect of technology transfer
related to applications of the RCWP knowledge base
will be the development of design procedures and
computational methods to improve practice effec-
tiveness as it relates to specific water quality loading
objectives.
One example of this type of emerging technol-
ogy is the development of expert systems that help
design and select nonpoint source controls. The
RCWP EXPERT computer program is being
developed to solve problems and aid technical train-
378
-------
P.D. ROBILLARD
ing. It is also being developed to help Soil Conserva-
tion Service engineers design control systems and
support and train State and Federal water quality
staff. The interactive, user-friendly nature of the
Apple™ Macintosh computer system provides an ex-
cellent focus for small group training activities.
The complex problem of designing practices to
control contaminants of different solubilities, soil ad-
sorption properties, toxicity, and persistence re-
quires consideration of numerous, and sometimes
competing, factors. For example, control practices
that decrease soluble nutrients in surface runoff
may increase loading to groundwater for certain soil
types. Similarly, the design of control systems re-
quires that water quality monitoring information be
used to evaluate and quantify contaminants as well
as focus on the BMP implementation programs in
watersheds. RCWP EXPERT was developed to deal
with these complex design and implementation
problems by providing a problem-solving format and
supporting data. The problem-solving format
specifies rules by which control practices are recom-
mended for certain contaminants, hydrologic condi-
tions, contaminant loadings, and time-of-year.
Other modules provide data and information
about contaminant characteristics, monitoring
methods, and how contaminant transport processes
influence the effectiveness of control practices. This
latter information is particularly important as a
bridge to help design teams reach a common under-
standing of the mechanics of water quality monitor-
ing and the design of control measures. Often, an
engineer will have specific training in the design of
control practices, but not in water quality monitoring
systems or how data from such systems can be used.
In many RCWP projects, the opposite case was also
true because members of the monitoring teams had
limited experience with practices or structures that
decrease or control contaminant loading. RCWP EX-
PERT is designed to provide both types of users with
procedures, data, and information to help design and
select nonpoint source control systems.
System Overview and Status
The RCWP EXPERT system configuration combines
RCWP's extensive monitoring and practice im-
plementation and computer-aided design features to
improve control practice design and siting on a field
and watershed basis (Fig. 1). Version 2.0 of RCWP
EXPERT, which will be completed in 1992, incor-
porates links to a Nonpoint Source Database and the
AGNPS simulation model (Young et al. 1987). This
linkage allows actual field practice data and simu-
lated contaminant losses to be evaluated for site-
specific variables. In addition, technical reference
modules (contaminants, transport, monitoring, and
case studies) are available throughout the design
procedure.
During 1993, geographic information systems,
practice design modules, and climate linkages will
be initiated. All control system design options and
tools will be accessed from the control systems
module. System output will include calculated con-
taminant mass load to surface, subsurface, and
groundwater. To the extent possible, mass loading
on a watershed and subwatershed basis will be spa-
tially represented and used to evaluate alternative
siting of control systems.
All reference modules, calculation procedures,
and database interfaces are organized to help users
design and evaluate nonpoint source control systems
on both a field and watershed scale (Fig. 1). The
knowledge base includes both heuristic data and .
knowledge from field applications of key RCWP
projects.
• Computer-aided Design Procedures. The con-
trol systems module assembles complementary sets
of practices, including a control practice for each
stage of contaminant transport (source-transfer-field-
delivery) and associated practices that enhance the
overall effectiveness of the control system.
. RCWP EXPERT allows the user to design con-
trol systems for three primary purposes:
• Control systems design assembles alternative
systems for total contaminant pathway control
(source-transfer-field-delivery);
• Simulation models (AGNPS is the first one)
and a nonpoint source database will be used
as evaluation tools to compare the effective-
ness (contaminant load reduction) of specific
practice applications and control systems; and
• Through linkages with climate data and a
geographic information system application,
the effectiveness of control systems can be im-
proved with respect to their siting in a water-
shed and the control of contaminant source
availability.
Control Systems Module
The control systems module (Fig. 2) is the platform
for design evaluations and comparison of manage-
ment operations. RCWP EXPERT provides users
with total control over the sequencing of practice
design operations. Reference modules are available
at all stages of the design process.
379
-------
Proceedings of National RCWP Symposium, 1992
Climate Data
Data for Simulation
Models and
Event Forecasts
Simulation
Models
Geographic
Information
System
Site Selection on a
Watershed Basis
Reference Modules
Contaminants
Monitoring
Transport
'XXXXXXXXXXXXXXXXXXXXXXXXXXXX
Case Studies
xxxx.
r f , r ,'
Interfaced at all
stages of design
process
Nonpoint
Source
Database
Field and Watershed
Simulation (Surface
and Groundwater
Effects)
Design of
Control Systems
BMP Matrix
Storet and other
WQ Databases
Input Data Files
Water Quality
Impact of
Practice Design
Variables
Practice
Design
Modules
Practice Design
Practice Impact
Figure 1.—Example of the RCWP EXPERT system (computer-aided design of nonpolnt source control systems).
The control systems module consists of a site-
specific selection process and a control systems
mode. The site-specific selection mode matches
practice contaminant control mechanisms with pol-
lutant transport variables. Each control practice is
evaluated for important contaminant, hydrologic,
and soil variables. The output of these evaluations,
called "Dependency Networks," screens practices
for their effectiveness under specific field conditions.
RCWP EXPERT currently contains 16 dependency
networks — one for each of the 16 types of practices
used in RCWP projects.
• Nonpoint Source Database. The nonpoint
source database allows users to access research data
on the effectiveness of practices. All 16 practice
categories and all individual practices (referenced by
SCS Technical Practice Codes) are available for in-
dividual contaminants and site-specific variables,
such as climatic region and soil type.
380
-------
P.O. ROBILLARD
/Site Specific
/ Practice
' Selection
AGNPS
©
©
Event-based
Simlulations
Control
Systems
NFS Database
Practice Design
Calculations
Process
The control systems module allows the design
engineer to:
A. Return to the Site Specific Sub-module to
choose control practices for different
hydrologic- soil-contaminant loading conditions.
B. User can directly access the NPS
database to observe related research
effectiveness data.
C. Direct access to Practice Design Calculations
(DOS based versions for selected practices in
CAMP system).
D. Import and export of data files and
results from AGNPS model.
Practice
Effectiveness
Data
Reference
Modules
•Contaminants
•Monitoring
•Transport
•Case Studies
Water Quality Impact of
Practice Variables
Figure 2.—Example of the control systems module.
• Simulation Models. AGNPS is currently linked
to RCWP EXPERT. This linkage facilitates evaluation
of practices under varying conditions (22 input vari-
ables) as defined by the user. The nonpoint source
database also allows verification and comparison of
AGNPS results with published field effectiveness
data.
• Climate Data. Three types of climate data will be
used in the completed expert system, including
• historical precipitation and temperature data,
• 10-day antecedent precipitation and
temperature data, and
• five-day forecast data.
Precipitation data is the key climate variable in all
cases. Temperature is important for snowmelt
routines and rate constants, such as mineralization
of nitrogen.
Climate data, data interpretations, and examples
of data use will help users estimate losses in surface
and subsurface flows and estimate the probability of
381
-------
Proceedings of Hatlonal RCWP Symposium, 1992
losses occurring during a five-day period on a real-
time basis. Climate data will be used for both field
and watershed simulations. These data will also be
an element of the monitoring and transport modules.
In locations that have detailed data available, special
and spatially varied precipitation data applications
will be used, including variations in intensity, droplet
size, and other parameters affecting the detachment
and mobility of contaminants.
Three reference modules (monitoring, trans-
port, and case studies) will provide expert input and
climate databases to use for different applications
and as an evaluation of the usefulness and quality of
the databases.
• Geographic Information System (GIS). CIS
linkage (using GRASS) will provide the basis for site
selection and control system designs that minimize
contaminant loading on both farm and watershed
bases. Although several components of the GIS
linkage are useful for designing nonpoint source
control systems, the ability to spatially orient prac-
tices to achieve contaminant loading objectives is
perhaps the most important feature.
• Practice Design Modules. These modules pro-
vide standard engineering procedures for practice
design. Typically, they will be used in conjunction
with the Practice Impact Module. The user can
change design variables to achieve specific water
quality improvements for a site-specific application.
• Water Quality Impact of Practice Design Vari-
ables. Each design practice variable can affect
source availability, the partitioning of contaminants
between surface and groundwater flow, soil erosion
and sedimentation, and other factors relating to con-
taminant transport This module provides guidance
about how practice design variables can be used to
affect contaminant loadings.
• Reference Modules. At any stage of the design
procedure, users can consult reference modules that
provide information, guidance, and data about con-
taminant properties, transport variables, and ex-
amples of applications from RCWP projects.
Four reference modules are available in RCWP
EXPERT:
• Contaminants;
• Monitoring;
• Transport; and
• Case Studies.
All four modules use graphics to demonstrate
design procedures and contaminant control proc-
esses.
The contaminants module provides information
about 11 categories of pollutants cited in various
RCWP projects and their effects on surface and
groundwater and on drinking water supplies. This
module includes four types of information:
• Impact — the effects specific contaminants
have on stream and lake water quality. If ap-
propriate, human health effects are also given.
Sixteen pesticides are currently included in
the module.
• Standards — concentrations of various con-
taminants are given relative to their environ-
mental impacts on aquatic systems and to
standards of the Safe Drinking Water Act.
• Background Concentrations — typical con-
centration ranges are supplied for those sub-
stances existing naturally in soil-aquatic
systems.
• Case Studies — examples of how water
quality impacts, standards, and background
concentrations were used in RCWP projects.
The monitoring module describes different
aspects of water quality sampling and analysis sys-
tems. In addition, methods for calculating mass
loads and statistical analyses are presented. Key ref-
erences in this section include the recent EPA
Monitoring Guide and examples from RCWP
projects. This module contains sk types of informa-
tion:
• System Design,
• Parameter Selection,
• Sampling,
• Handling Preservation and Analysis,
• Loading Calculations, and
• Statistical Analyses.
The transport module describes contaminant
pathways in surface and groundwater. In particular,
the solubility and adsorption processes of each con-
taminant are described and compared in the context
of control mechanisms and options. Contaminant
loading and critical loading periods are also
described in this module. The transport module in-
cludes six types of information:
• Surface Flow,
• Subsurface and Groundwater Flow,
• Adsorption and Solubility,
• Critical Events,
• Delivery, and
• Control Mechanisms.
382
-------
P.D. ROBILLARD
The case states module presents detailed ex-
amples from key RCWP projects. Both practice
selection and implementation aspects of BMP sys-
tems are covered. Regional precipitation, tempera-
ture, soil, groundwater, and common agricultural
practices data as well as descriptions and data from
individual RCWP sites are provided.
The RCWP provides a key source of information
for agricultural nonpoint source control programs.
The massive implementation of BMPs has resulted
in successes and failures that should be docu-
mented. The case studies module includes the fol-
lowing examples of implementation topics;
• continued emphasis on protection of farm
water supplies;
« design of total control systems for high con-
taminant source areas;
• use of monitoring data to target subwater-
sheds and high source areas for treatment;
• application of technology transfer efforts that
result in increased engineering assistance and
staff training to provide expertise in animal
waste systems, or other specialty areas;
• use of an experimental design phase that al-
lows new technology to be tested on farms,
which in turn leads to improved design proce-
dures;
• use of on-farm monitoring to evaluate changes
in concentration or load before and after prac-
tice implementation;
• further development of biomonitoring as a
method for identifying problems and evaluat-
ing practice effectiveness;
• use of watershed and stream assessments to
provide guidelines and public awareness of
water quality improvements;
• establishment of flexible cost-share incentives
so farms with more serious problems can be
treated more intensively;
• minimization of technical staff turnover dur-
ing design phase of project; and
« implementation of practices most likely to
reduce water quality loading.
Conclusion
RCWP projects have accumulated valuable data and
information about the design of control practices,
monitoring systems, and successful implementation
strategies; they also have data on institutional and
economic successes and failures, which are an im-
portant knowledge base for future nonpoint source
projects. Successes resulted from improved practice
designs and widespread implementation of BMPs by
the farm community. Failures are associated with in-
adequate design and the lack of water quality
monitoring data. In either case, useful guidance can
be gained from these RCWP lessons.
An important element in future technology
transfer efforts from RCWP to new nonpoint source
watershed projects will be the development of com-
puter-aided procedures to integrate data and infor-
mation into helpful tools. RCWP EXPERT is an
example of such a tool.
References
Robillard, P.D. and M.F. Walter. 1982. Lake Ontario Phosphorus
Evaluation Project Dep. Agric. Eng., Cornell Univ., Ithaca,
NY.
Sharpe, W.E., D.W. Mooney, and R.F. Adams. 1985. An analysis of
groundwater quality obtained from private individual water
supplies in Pennsylvania. Northe. Environ. Sci. 4(314):155-
59.
U.S. Soil Conservation Service. 1989. Oregon Engineering Hand-
book. Portland, OR.
Young, R.A., CA Onstad, D.D. Bosch, and W.R Anderson. 1987. A
Watershed Analysis Tool: AGNPS, Agricultural Nonpoint
Source Pollution Model. U.S. Dep. Agric., Washington, DC.
383
-------
-------
Research Needs and Future Vision for
Nonpoint Source Projects
Paul D. Robillard
Department of Agricultural and Biological Engineering
Penn State University
University Park, Pennsylvania
John C. Clausen
Department of Natural Resources Management and Engineering
University of Connecticut
Storrs, Connecticut
Eric G. Flaig
South Florida Water Management District
West Palm Beach, Florida
Donald M. Martin
U.S. Environmental Protection Agency
Boise, Idaho
Region X
ABSTRACT
The importance of the Rural Clean Water Program (RCWP) to future nonpoint source control
projects depends on careful identification of its experimental nature and results, and perhaps most
important, on the analysis of research that should be accomplished to correct and enhance those
results before future projects are undertaken. Consequently, this paper reports on research areas
and technology that will lead to better correlations between agricultural best management prac-
tices and water quality results in future projects. The review of key research areas is essential to an
understanding of the likely evolution of nonpoint source programs over the next 10 years.
The experimental nature of the Rural Clean
Water Program (RCWP) provides an excel-
lent opportunity to assess research needs for
future nonpoint source projects. During the 10-year
RCWP effort, the effectiveness of best management
practices (BMPs) in reducing contaminant loading
to groundwater became increasingly important.
Widespread adoption of reduced tillage systems was
observed in most RCWP project areas; control of
pesticide losses to surface and groundwater flows
redirected efforts in several RCWP watersheds; and
intensive treatment of high source contributing
areas (such as livestock confinement facilities) was
another example of the changing focus and technol-
ogy that can be derived from RCWP implementation
efforts. This paper has two objectives: it reviews the
need for research in key areas and helps provide a
vision of the future and how nonpoint source
programs might evolve during the next 10 years. Re-
search topics related to the RCWP's recent ex-
perience and future nonpoint source projects include
• monitoring networks,
• tillage systems,
• critical area delineations,
• control system design, and
• improved implementation strategies.
385
-------
Proceedings of National RCWP Symposium, 1992
Research Needs
Monitoring Networks
Several RCWP projects (Vermont, Idaho, Florida,
South Dakota, Minnesota) developed and used both
conventional and innovative water quality monitor-
ing systems. However, the use of water quality
monitoring systems to judge the success of the land
treatment program and improve the siting and effec-
tiveness of control practices was a formidable chal-
lenge for many RCWP projects (Dickinson et al.
1990; Chadderton and Kropp, 1985; Humenik et al.
1987). In several cases, baseline (or pre-BMP im-
plementation) monitoring was not accomplished.
Other sites established trend monitoring stations
only for the entire project watershed; thus, little or
no change in water quality parameters resulted from
BMP implementation programs.
Similarly, the dynamic nature of cropping and
livestock systems required complementary monitor-
ing of watershed changes in livestock numbers and
location, acreage, and types of crops. Among the
projects that established monitoring stations, limited
resources often prevented comprehensive coverage
of the watershed. In response to these limitations,
different levels and types of water quality monitoring
were designed. For example, biological monitoring
offered a low-cost alternative to conventional techni-
ques and produced data that were more meaningful
to the public.
Nonpoint source projects will need to use in-
tegrated site-specific methods for monitoring land
treatment systems. Designs for many of these
monitoring systems were developed and refined
during RCWP watershed projects. Generally, the
design of water quality monitoring systems needs to
meet the following criteria:
1. Although baseline water quality requirements
are dependent on numerous hydrologic and
land use variables, a preproject water quality
status must be established and control water-
sheds (subwatersheds) maintained during
the project and postimplementation period.
From a practical viewpoint, one year of data is
often the most that can be obtained for the en-
tire watershed, while use of control or paired
subwatersheds may allow for the more
desirable three-to-five-year baseline data
needed to evaluate overall effectiveness of the
land treatment program.
2. Design of monitoring systems should en-
courage the use of multilevel networks. Al-
locating all resources to long-term trend
monitoring is not recommended. Field and
subwatershed monitoring systems should be
used to evaluate individual practices or prac-
tice systems. These types of monitoring sys-
tems will vary considerably — from mobile
units (including rainfall simulation devices) to
noncontinuous stations of evaluation of ex-
perimental practices, to permanent contin-
uous stations for longer-term trend analyses.
The monitoring system must develop a
method for relating water quality parameters
to land treatment. Land use information that
can be linked to water data must be collected.
Collection of accurate land use data is ex-
tremely difficult; it is also uncertain how long
monitoring should continue (Clausen et al.
1992). A land use information collection sys-
tem should be designed parallel with the
water quality monitoring system. For ex-
ample, significant changes in nutrient loading
may not occur as a result of BMPs imple-
mented on a watershed basis (Meals, 1992).
The time required for changes (improve-
ments) in water quality parameters after BMP
implementation is likewise uncertain.
Based on these criteria, the following
design requirements have been identified:
• Determination of the frequency of sam-
pling needed to achieve water quality
trend or evaluate practice objectives.
Current guidelines provide insufficient
advice regarding sample size.
• Better methods for collecting accurate
land use monitoring data. It is difficult
to be field-specific about practices and
timing of operations; however, this in-
formation is critical for event-based
evaluations.
• Improved techniques to link land use
and water quality data, especially at the
watershed level, are needed to show
causal relationships. At a minimum,
land use and in-stream water quality
monitoring must be measured simul-
taneously.
• Development and improvement of
field-validated computer models that
can simulate the outcomes of control
systems must continue. Models can be
important tools in selecting alternative
management practices. Similarly,
models can be used to estimate the
time needed for a practice to be effec-
tive.
386
-------
P.O. ROBILLARD, J.C. CLAUSEN, E.G. FLAIG, & D.M. MARTIN
• Linking bioassessment to other water
quality parameters and land use ac-
tivities with statistical confidence. The
relationship between bioindicators and
traditionally used chemical and physi-
cal parameters has not been firmly es-
tablished.
• Use of monitoring data to determine
the effectiveness of existing BMPs and
to develop new and innovative BMPs
must be better defined and docu-
mented for a variety of conditions. For
example, the water quality benefits of
field nutrient management are largely
assumed.
• Time needed for water quality
parameters to respond to BMPs.
Knowledge of this response time will
improve management decisions and
result in more accurate and realistic ex-
pectations of the impact of control sys-
tems on changes in monitored
parameters.
3. The use of monitoring methods not directly
related to practice effectiveness or trend
analysis must be developed or improved.
Monitoring water column chemistry alone
has limited applications. The health of the
broader aquatic ecosystem including both the
biological and physical habitats and aquatic
chemistry must also be measured. Land treat-
ment practices and the aquatic ecosystem
must be evaluated at the same level of inten-
sity.
Biological monitoring such as that used
during RCWP projects provided low-cost al-
ternative techniques that were a basis for
quantifying changes in aquatic life and the
quality of riparian zones. In addition, this type
of monitoring data is often more meaningful
to users of the water resource and the general
public. Well water data can also be used to
supplement effectiveness evaluations and in-
form the public about water quality. These
data are particularly helpful since ground-
water quality directly affects farms and other
users of water resources in the project area.
4. Because we must continually define and re-
state agricultural water quality goals, both
conventional and biological monitoring are
useful because they have enough flexibility to
assist in the short-term evaluation and
redirection of the land treatment program.
Tillage Systems
Widespread adoption of reduced tillage systems by
RCWP project participants resulted in increased em-
phasis on several research topics — for example, the
impact of increased pesticide applications associated
with reduced tillage and tillage-manure systems on
water quality.
Control of Pesticide Losses
Changes in tillage and cropping systems require
changes in weed and insect control methods. A prin-
cipal reason for the success of reduced tillage sys-
tems has been the development of effective pesticide
control techniques. Research continues to inves-
tigate the effect of pesticide application rates, timing,
and techniques on surface and groundwater quality.
In particular, computational methods that can
predict biological degradation rates and the attenua-
tion of pesticides between the field and delivery to a
waterbody are important topics. Finally, with in-
creased emphasis on the impact of pesticides on
groundwater quality, the cycling and transport of
pesticides in fractured bedrock and deep confined
aquifers are topics of increasing interest
Tillage Manure Options
An inherent conflict exists in the maintenance of sur-
face residue in tillage-manure systems. One objec-
tive clearly indicates that surface residue should be
conserved to reduce erosion. On the other hand,
manure applications should be incorporated as soon
as possible, specifically to minimize contaminant
losses and retain nutrients in the soil profile to meet
crop needs. Experimental results indicate that
nutrient and pesticide losses associated with tillage-
manure systems vary widely depending on applica-
tion rates, incorporation depths, and times of
application. The following principles are derived
from studies that provide criteria for the develop-
ment of tillage-manure systems:
• The time between initial application of
nutrients or pesticides and the first precipita-
tion event that induces runoff is a key factor in
determining the magnitude of contaminant
losses.
• Drying conditions during the period between
manure application and precipitation events
will vary greatly between seasons and the
probability of losses changes accordingly.
• Minimal incorporation of applied manure
greatly decreases nutrient losses.
387
-------
Proceedings of National RCWP Symposium, 1992
* After one or two wet-dry cycles following ap-
plication, the benefits of manure amendments
to soil can be observed in terms of greater
surface storage, infiltration, and lower runoff
volumes.
• The amount of bedding in manure will in-
fluence residue cover for surface manure ap-
plications. The benefits of this residue are
similar to plant residues associated with
reduced tillage systems. However, the quality
of this cover is different, and much of the bed-
ding will be transported off the application site
over time.
The number of possible combinations of manure
disposal systems associated with various tillage prac-
tices on different soils is myriad. In addition, in-
dividual farm objectives, resources, and limitations
will influence the choice of tillage-manure systems.
For these reasons, the development of a single set of
criteria will probably not lead to an efficient im-
plementation program.
The monitoring needs discussed in the previous
section directly apply to tillage-manure systems. The
mobile rainfall simulator method for evaluating
nutrient and pesticide losses for various tillage-
manure systems is a promising option. Because
mobile units can be used on operating dairy farms,
the number of specific tillage-manure systems that
can be evaluated is much greater than traditional
plot studies provide. Most important, mobile unit
treatments on different farms in different areas can
be directly compared and evaluated, since all
precipitation variables are constant. Finally, the unit
can be used for evaluations during the late winter
and early spring periods when losses are typically
high.
Contaminant Mobility and Transport
Contaminant pathways are difficult to predict in time
and space. One aspect of the transport process with
implications for control of pesticides and nutrients
through residue management involves a zone of in-
teraction at the soil surface. Precipitation striking
the surface reacts with a shallow depth (typically < 1
cm) of soil. Water mixing with soil in the shallow
zone of interaction is the source of contaminant
mobility and losses. The concentration of pesticides
and nutrients leaving the field will be determined by
the complex and dynamic equilibrium between the
dissolved phase and suspended sediment interac-
tions in this shallow zone. There are several oppor-
tunities for improving our understanding of these
processes.
B Effects of Tillage Practices on Contaminant
Transport and Delivery. Several researchers
(Mueller, 1979; Oloya and Logan, 1980; Robillard and
Walter, 1991) have interpreted experimental results
that are consistent with the concept of a shallow
zone of interaction originally proposed by Sharply et
al. (1981). In particular, the mixing of soil and water
in the zone of interaction can help explain losses of
soluble nutrients and pesticides in surface runoff.
Tillage practices change the depth of the interaction
zone, redistribute pesticides and nutrients, and
change surface runoff and infiltration conditions.
The depth of the interaction zone is affected by til-
lage because inversion or stirring of the soil by til-
lage operations changes surface density and
porosity. This effect is typically short-lived, as con-
solidation and compaction decrease the zone of inter-
action over time.
• Soluble Contaminant Losses. In the physical
process of contaminant availability and loss, a mass
of soil in the interacting zone mixes with a volume of
water from a precipitation event. The release rates
and amount of contaminant made available or lost
can be related to initial contaminant concentration,
pH, and other parameters. Consistent with the con-
cept of a shallow zone of interaction, the mass of pes-
ticides or nutrients remaining will, in part, determine
subsequent losses. The relatively high concentration
of nutrients and pesticides in the shallow zone as-
sociated with surface applications validates this con-
cept. Experimental results also indicate that residual
soluble contaminant concentration sustains some-
what higher losses during subsequent precipitation
events. Therefore, a larger pool of contaminants
must be available in reduced tillage systems than in
conventional tillage.
Additional important research is needed to char-
acterize the physical and chemical changes in the
zone of observed interaction when pesticides or
manure are applied to the soil surface or incor-
porated into the soil profile. The depth of interaction
and the soil-to-water ratio should be observed or
derived for the rapidly changing conditions as-
sociated with biological degradation and stabilization
in this zone. This relationship is similarly important,
since manure applications directly influence soil pH
levels (Westerman et al. 1981; Muck et al. 1978) and
make available a large pool of soluble nutrients. In
fact, manipulation of pH in the zone of interaction
should be investigated as a control option. Residue
cover and its effect on surface storage and mixing in
the zone of interaction should be examined more
closely. Other related aspects of prediction and con-
trol of soluble pesticide and nutrient losses that
should be investigated include
388
-------
P.O. ROBILLARD, J.C. CLAUSEN, E.G. FLA1G, & D.M. MARTIH
• the effect of storm intensity on mixing depth
and efficiency;
• the effect of quantity and quality of residue
on soluble contaminant losses seasonally;
• the influence of precipitation patterns on
mixing, particularly in wet-dry cycles; and
• differences in soluble contaminant
availability during the nongrowing season.
Critical Area Delineations
The definition of and emphasis on critical area treat-
ment changed considerably during the RCWP. How-
ever, a direct link is needed between critical area
delineations and specific water quality parameters.
The use of water quality monitoring data should be a
key element in this process (Koehn and Stanko,
1988). An example of critical area identification and
development of control options can be associated
with land uses where the potential for high loss rates
is evident.
High Source Contributing Areas
The changing technology in agricultural systems
during the past 50 years has increasingly em-
phasized more intensive utilization of inputs and
more control of production-related environmental
conditions. These crop and animal production sys-
tems create areas with a high nonpoint pollution
source contamination potential — for example,
animal confinement facilities, manure, pesticide and
fertilizer storage areas, livestock and material han-
dling areas, and transport lanes. Methods for inten-
sive control of surface and subsurface flow from
these areas are required. For example, the Florida
RCWP designed and implemented "high intensity
use areas" to provide maximum control over surface
and subsurface phosphorus losses. Confinement
areas were integrated with satellite pastures to pro-
vide animal control until all manure collection,
storage, and spray irrigation systems had been
designed to meet nutrient requirements and achieve
water quality concentration standards.
Control Systems
Initially, RCWP focused efforts on control of sedi-
ment and nutrients in surface runoff. Then increas-
ing concern about gfoundwater contamination
forced a more integrated evaluation of the effective-
ness of control practices at certain project sites. Con-
sequently, future design of control practice systems
will need to incorporate the estimated change in total
mass load delivered to surface and groundwater as a
result of BMP implementation.
During the initial stages of the RCWP, specific
control practices were installed to reduce contamina-
tion, primarily at the field level. The concept of prac-
tice systems designed to control all phases of
contaminant transport began to evolve at a few
RCWP watersheds. Thus, where only manure
storage or lagoons were installed at some sites, other
projects designed practices for source-field and
delivery control. For example, manure storage may
have been an element of a control system incorporat-
ing nutrient management, conservation tillage, and
detailed land application procedures. The practices
used in a control system can be management
measures, vegetative controls, or structural com-
ponents. Identification of associated practices is one
element of control systems design that improves
overall effectiveness. This type of control can include
animal waste transfer systems, improved farm roads,
or livestock confinement facilities. Associated prac-
tices can extend to implementation of control sys-
tems through equipment modification, equipment
rental, and consulting activities used collectively by
several RCWP sites to enhance the overall quality ef-
fectiveness of control systems.
Field Evaluation and Monitoring of
Control Systems
The use of different levels and types of monitoring
systems to evaluate water quality impacts is an im-
portant nonpoint source activity that was lacking at
most RCWP sites. In particular, monitoring systems
to evaluate practice effectiveness were almost nonex-
istent. Additional research is needed to develop low-
cost, practical methods of using these tools (Laflen
et al. 1991). For example, mobile monitoring resour-
ces are an excellent method for collecting practice
effectiveness data. Two primary methods of mobile
monitoring are grab sampling and rainfall simulation
units. Grab sampling systems can be designed to iso-
late the water quality impact of control practices or
land uses in both space and time.
Mobile rainfall simulators can be used to
evaluate practice effectiveness. Control of precipita-
tion variables in order to compare practices or treat-
ments provides an opportunity to make direct
comparisons. The mobile units can evaluate prac-
tices for actual farm conditions, and they are a more
efficient way to evaluate multiple sites and obtain
measurements in a short time. For example, a com-
bination of laboratory and field evaluations of tillage-
manure systems can be undertaken for controlled
conditions as well as more realistic field conditions.
389
-------
Proceedings of National RCWP Symposium, 1992
If additional site-specific details of rate, depth, and
time interactions are needed, field soils can be col-
lected and prepared for evaluation in the laboratory.
Constant or variable precipitation intensities can
then be used to meet pesticide or nutrient evaluation
needs.
Rainfall simulators can also be used to evaluate
contaminant losses for particular precipitation pat-
terns. For example, historical precipitation data can
be used to estimate the probability of specific storm
intensities and dry periods for different seasons.
These intensities and dry periods can be physically
simulated under field or laboratory conditions for
various rate, depth, and time treatments. Finally, a
distinct advantage of using a mobile simulator unit is
the accumulation of a baseline data file. This file con-
sists of laboratory and field runs that can be refer-
enced to specific rate, depth, and time values.
Therefore, a field run can be related to a set of meas-
urements made for precisely the same treatment on
other fields or soil types. A scale or index value can
be calculated to provide probable pesticide or
nutrient loading for each case. In addition, a series of
runs can help determine the statistical variability for
a particular field or soil. Finally, the soil can be
evaluated under laboratory conditions for a par-
ticular treatment, time series, or controlled environ-
mental condition, such as a freeze-thaw cycle.
Implementation Strategies
Selection and design of control systems will increas-
ingly involve more efficient use of inputs and waste
products. Whether integrated pest management,
nutrient management, or improved land application
of waste methods, the objectives of nonpoint source
systems are consistent with improved farm opera-
tions and management. Several RCWP sites
demonstrated this compatibility. Additional research
is needed to quantify and verify this approach, and
establish its economic implications and agricultural
sustainability (Karlen, 1990; Onstad etal. 1991; High-
feld, 1983),
RCWP provides numerous examples of practice
implementation techniques that increase adoption
rates within a voluntary program. Typically, these
techniques involve both water quality impacts and
related farm improvements. Three examples of
these implementation methods include
• development of farm water supplies to de-
crease farm animals' direct access to streams
and improve livestock drinking water quality;
• design of reduced tillage systems to decrease
nutrient losses along with labor and energy
costs; and
• improved farmstead equipment and livestock
access and control integrated with structural
practices that decrease total contaminant load
and allow more efficient use of heavily traf-
ficked areas within a farmstead.
An institutional aspect of nonpoint source water-
shed projects that requires additional research in-
volves the sequencing of these activities. In most
instances, three primary activities occurred during
RCWP: land treatment, monitoring, and cost-shar-
ing. For a project to accomplish water quality objec-
tives, these activities need to be sequenced; for
example, monitoring data should be used to select
practices, target critical areas, and evaluate effective-
ness of the land treatment program. Similarly, cost-
sharing should be used early to encourage the
adoption of the most effective water quality control
practices in the critical areas requiring treatment.
Because of the experimental nature of RCWP, this
sequencing and coordination was difficult to ac-
complish and often not achieved.
Future Projects
Progress toward nonpoint source water quality goals
during the past 20 years has been modest. RCWP
represents the most extensive, long-term nonpoint
source implementation effort in the United States
(Larsen et al. 1988; Wall et al. 1989). The completion
of the RCWP provides an opportunity to extrapolate
these accomplishments to the future.
Although the original qualitative water quality
goals are not likely to change in the next 10 years,
methods for achieving "fishable and swimmable"
status are likely to change dramatically. New
strategic approaches will include
• the integration of water quality control
programs to account for surface, subsurface,
and groundwater effects; similarly, simul-
taneous, basinwide evaluation and control of
point and nonpoint contaminant sources will
be emphasized;
• water quality parameters will become more
broadly based and include composite physical,
chemical, and biological indicators;
• advances in monitoring networks and com-
puter applications will develop the links be-
tween land treatment activities and their
impact on water quality parameters; and
« use of spatially oriented watershed data, based
on geographic information system concepts,
will help to target critical areas and guide
more intensive management measures for im-
390
-------
RD. ROB1LLARD, J.C. CLAUSEH, E.G. FLA1G, & D.M. MARTIN
plementation in high source cells (McNain
and Prutz, 1990).
As nonpoint source projects begin to use re-
search findings and the practice implementation
techniques realized in RCWP projects, it is possible
to distinguish emerging methods and technologies.
Artificial Intelligence Application
The power of computers "will be enhanced by interac-
tive systems using both heuristic guidance and com-
putational tools. For example, the links between
expert systems and simulation models are ad-
vantageous because they provide the user with a
flexible, intelligent interface to the deep knowledge
encoded in simulation models (Stone et al. 1986; Par-
saye et al. 1989). Such integration exemplifies a
growing trend toward the coupling of artificial intel-
ligence methodology (e.g., expert systems) and con-
ventional computer science (e.g., simulation models)
for problem-solving in agriculture and natural
resource management (Whittaker and Tieme, 1990;
Robillard et al. 1990; DeCoursey, 1985). Further, this
integration allows the design of practices to make
site-specific calculations for alterative control sys-
tems. A further extension of this concept can include
access to weather data.
(Lehman, 1989). These estimates are then expressed
as the probability of load or concentration thresholds
that can be exceeded for various nutrient or pes-
ticide application rates on a given day. Such prob-
ability calculations can guide timing and rate
decisions by farm operators. With the increasing im-
portance of event-based monitoring coupled with the
increasing cost of establishing and maintaining
monitoring networks, real time control systems can
also be used to activate monitoring networks for
more intensive coverage of selected events.
Access to radar and satellite data on real time
can also be advantageous. These products can take
the form of spatially-depicted precipitation coverage
areas, similar to geographic information system
databases, and they greatly enhance single-point rain
gage measurements by providing detailed rate and
total precipitation information. Localized storm cells
that create heavy precipitation are often missed by
low-density gage networks; however, they can be in-
stantly captured by radar and satellite database infor-
mation. When these data systems are coupled with
forecasted events they provide powerful tools to im-
prove the management of nutrient and pesticide ap-
plications. The new, almost unlimited potential of
computer graphics allows for development of practi-
cal user interfaces in most cases.
Real Time Control Systems
The use of computer-aided design creates an oppor-
tunity to provide real time guidance and recommen-
dations for farm operators to minimize the effect of
agricultural operations on surface and groundwater
quality. As research improves our understanding of
contaminant mobility and transport, it will be pos-
sible to focus resources on control systems or
decision aids that can be utilized by water quality
control personnel and farm operators. In developing
real time control systems, two primary applications
are possible:
• the use of antecedent and forecast climate
data to minimize losses of nutrients, pes-
ticides, and sediment; and
• the use of microclimate data to improve the
application efficiency of nutrients and pes-
ticides and reduce losses to surface and sub-
surface flows.
In both cases, the timing and rate of application
are sensitive to both water quality impact and ineffi-
cient utilization of farm inputs. Real time control sys-
tems assemble 5 to 10 days of antecedent precip-
itation and temperature data coupled with short-term
forecast data to derive contaminant loss estimates
Concluding Observations
Lessons learned from the RCWP must be incor-
porated into new nonpoint source programs. Oppor-
tunities exist to extend monitoring, land treatment,
practice design, and implementation methods to new
projects through research and technology transfer.
These efforts will lead to efficient and standardized
methods for achieving nonpoint source goals.
Without this advantage, new projects may repeat er-
rors made in the RCWP and the effectiveness of the
projects will remain uncertain.
References
Chadderton, R.A. and I.S. Kropp. 1985. An Evaluation of Eight
Wasteload Allocation Methods. Water Resour. Bull. October.
21(5):83341.
Clausen, J.C., D.W. Meals, Jr., and E.A. Cassell. 1992. Estimation of
lag time for water quality response to BMPS. In Proc. Natl.
RCWP Symp. Orlando, FL
DeCoursey, D.G. 1985. Mathematical models for nonpoint water
pollution control. J. Soil Water Conserv. 40(5):408-15.
Dickinson, W.R., R.P. Rudra, and GJ. Wall. 1990. Targeting
remedial measures to control nonpoint source pollution.
Water Resour. Bull. 26(3) :499-507.
391
-------
Proceedings of National RCWP Symposium, 1992
Highfcld, RE. 1983. Modern terrace systems. J. Soil Water Con-
serv. (July-August). 38(4):336-38.
Humcnik, FJ., M.D. Smolen, and SA Dressing. 1987. Pollution
from nonpoint sources. Environ. Sci. Technol. 21(8) :737-42.
Karlen, D.L 1990. Conservation tillage research needs. J. Soil
Water Conserv. (May-June). 45 (3) :365-69.
Kochn, J.W. and G.H. Stanko. 1988. Groundwater Monitoring. En-
viron. Sci. Technol. 22ftl):1262-64.
Laflen, J.M. et al. 1991. WEPP, soil credibility experiments for ran-
geland and cropland soils. J. Soil Water Conserv. (January-
February). 46(l):34-39.
Larson, D.P., D.R. Dudley, and R.M. Hughes. 1988. A regional ap-
proach for assessing attainable surface water quality. J. Soil
Water Conserv. (March-April). 43(2):171-76.
Lehman, R. 1989. QPMO (Quick Monthly Projected Outlook).
Nail. Weather Serv., Natl. Oceanic Atmos. Admin.
Washington, DC.
McNaln, H. and R. Prutz. 1990. Evaluation of the use of satellite im-
agery to measure crop residue cover on fields in the Lake
Erie watershed. /« Proc. Remote Sensing and CIS Applica-
tions to Nonpoint Source Planning. Terrene Inst.
Washington, DC.
Meals, D.W., Jr. 1992. Water quality trends in the St. Albans Bay
watershed following RCWP land treatment. In Proc. Natl.
RCWPSymp. Orlando, FL
Mueller, DM., T.C. Daniel, B. Lowery, and B. Andraski. 1979. Ef-
fects of Conservation Tillage on the Quality of Runoff Water.
ASAE Pap. 82-2022. Am. Soc. Agric. Eng., St Joseph, MI.
Muck, R.E. 1978. The Removal of Nitrogen and Phosphorus from
Poulty Manure with Simulated Rainfall. Ph.D. dissert Cor-
nell Univ., Ithica, NY.
01oya,,TO. andTJ. Logan. 1980. Phosphate desorption from soils
and sediments with varying levels of extractable phosphate.
J. Environ. Qual. 3(10):20-31.
Onstad, C A, M.R. Burkart, and G.D. Bubenzer. 1991. Agricultural
research to improve water quality. J. Soil Water Conserv.
(May-June). 46(3):184-89.
Parsaye, K, M. Chignell, S. Khoshafian, and H. Wong. 1989. Intel-
ligent Databases. Wiley & Sons, New York.
Robillard, RD., MA Foster, and PA Heinemann. 1990. Expert
systems for water quality control. Tech. Pap. Am. Soc. Agric.
Eng. St Joseph, MI.
Robillard, P.O. and M.E Walter. 1991. The Influence of Tillage on
Phosphorus Losses from Manured Cropland. Great Lakes
Nafl. Prog. Off., U.S. Environ. Prot Agency. Chicago, IL.
Stone, N., R.N. Coulson, R.E. Frisbie, and D.K. Loh. 1986. Expert
systems in entomology: three approaches to problem solv-
ing. Bull. Entomol. Soc. Am. 32:161-66.
Wall, G J., TJ. Logan, and J.L. Ballantine. 1989. Pollution control in
the Great Lakes basin: an international effort J. Soil Water
Conserv. (January-February). 44(1):12-15.
Westerman, P.W., T.L Donnelly, and M.R. Overcash. 1981. Erosion
of Soil and Manure after Surface Applications of Manure.
' ASAE Pap. 81-2108. Am. Soc. Agric. Eng., St Joseph, MI.
Whittaker, DA and R.H. Thieme. 1990. Integration of knowledge
systems into agricultural problem solving. Computer and
Electron. Agric. 4:271-73.
392
-------
Synoptic Survey of Dairy Farms in the
Lake Okeechobee Basin: Post-BMP
Water Quality Sampling
Gregory J. Sawka, Paul Ritter, Boyd Gunsalus, and Thomas Rompot
South. Florida Water Management District
Okeechobee, Florida
ABSTRACT
High phosphorus loadings from dairy operations in the Lake Okeechobee Basin threaten Lake
Okeechobee's water quality; therefore, best management practices have been implemented to con-
trol nutrient runoff to surface waters. An extensive water quality monitoring network operated by
the South Florida Water Management District monitors the water quality of lake inflows, basin
tributaries, and off-site dairy farm discharge. Synoptic surveys have been developed to give a more
comprehensive on-farm evaluation of water quality problems that cannot be identified by routine
water quality monitoring at dairy farm outflows along property boundaries. Synoptic surveys are
performed on selected dairies where phosphorus concentrations greater than 1.2 mg/Lare consis-
tently detected. Water quality samples are taken within property boundaries to identify high
nutrient sources and identify problem areas that influence off-site runoff. The data are reviewed
with landowners and used as a tool to correct design problems and develop on-farm nutrient
management strategies. This paper presents a case study in which field scale problems were iden-
tified to help the landowner properly manage the operation and maintain the animal waste disposal
system.
Lake Okeechobee covers 189,216 hectares and
serves as a source of drinking water for area
municipalities, irrigation for agriculture, local
tourism, and recreation. The lake supports a large
commercial and sport fishing industry and provides
habitat for a variety of wildlife. The eutrophic lake
has experienced large algal blooms driven by high
nutrient inputs—particularly phosphorus (Joyner,
1971; Davis and Marshall, 1975; Federico et al.
1981). Northern inflows to the lake contribute a
large portion of phosphorus from primarily dairy
and beef agricultural operations (Allen et al. 1976;
Federico, 1977; Ritter and Allen, 1982). Nutrient dis-
charge is high during the spring and summer rainy
season when the water table is within 25 cm of the
ground surface in many areas.
Under the authority of Section 700.3 (Public Law
96-108, 93 Stat. 821, 835), Rural Clean Water Pro-
gram (RCWP) projects were initiated in the Taylor
Creek-Nubbin Slough basin in 1981 and in the Lower
Kissimmee River basin north of the lake in 1987. The
purpose was to assist landowners in implementing
best management practices (BMPs) designed to
control nutrient discharges, including
• fencing to keep cows out of waterways,
• improved water use efficiency,
• lagoon systems to capture barn wash, and
« pasture management (Stanley et al. 1988;
Stanley and Gunsalus, 1991).
393
-------
Proceedings of National RCWP Symposium, 1992
In 1987, the Florida Department of Environmen-
tal Regulation initiated a technology-based dairy
rule: the Florida Administrative Code Chapter 17-
670. This rule required additional BMPs specifying
that each dairy operation must collect waste from its
milking parlor and runoff from denuded herd pas-
ture areas around the barns in which a high volume
of animal waste is produced. The effluent from this
high intensity area is diverted into waste storage
ponds. Effluent from the ponds is applied to crops
through sprayfield irrigation for nutrient uptake
(Fla. Dep. Environ. Reg. 1987).
Thirty dairies are involved in the dairy rule
BMP program. As of April 1992, BMPs are complete
and operational at 29 dairies with one dairy under
construction. In support of the FDER dairy rule,
water quality monitoring by the South Florida Water
Management District provides nutrient concentra-
tion measurements at dairy outflows. Concentration-
based standards for total phosphorus were set for
dairy off-site discharges (1.2 mg/L) and basin out-
lets to the lake (0.18 mg/L). These concentrations,
based on phosphorus assimilation of wetlands and
streams in the flow path to the lake (S. Fla. Water
Manage. Distr. 1989a), were established to meet the
Surface Water Improvement Management Plan's
phosphorus-loading reduction goal of 42 percent
from controllable sources (S. Fla. Water Manage.
Distr. 1989b).
Water quality monitoring consists of grab and
automated sampling at off-site discharges from dairy
operations, tributary, and structure locations. A
revision in the water quality network, implemented
in October 1991, included synoptic surveys of dairy
farms.
Review of the literature indicates that synoptic
surveys were included in river quality data programs
by the U.S. Geological Survey (Hines et al. 1977).
These surveys are problem-oriented studies that
result in interpretive reports performed to make
sound resource decisions. The surveys—more
flexible than rigidly structured routine monitoring—
are tailored to fit the needs of each particular project
area. In the dairy program, synoptic surveys are on-
site water qualify investigations designed to give a
more detailed look at the sources of phosphorus
within the farm operation than is possible from
routine off-site discharge monitoring. The surveys
are intended to be a "snapshot" look at the dairy.
Data collected are not comparable with other long
term routine monitoring data or intended to deter-
mine trends. An important function of this program
is to provide a tool to landowners to correct on-farm
nutrient management problems and improve opera-
tion and maintenance plans. In addition, this survey
gives information on the efficiency of specific BMPs.
Methods
Dairies are selected for synoptic surveys when off-
site discharges are consistently greater than the 1.2
mg/L target standard. Surveys are performed when
off-site discharge is occurring. Sampling is con-
ducted within the property boundaries at various
locations upstream from off-site discharge points
and areas where high phosphorus sources are
suspected, including field ditches, major water
courses, sprayfield outlets, primary and secondary
water retention structures, wetlands, and animal
lounging areas.
The first step in conducting the survey on a
specific dairy is a background search of all literature
pertinent to the operation. Historical water quality
discharge data are reviewed along with computer
maps of the dairy operation. Engineering design
drawings, aerial photographs, and soil survey infor-
mation are used to determine flow paths and soil
drainage. Samples collected in the field are analyzed
for soluble reactive phosphorus in the laboratory
using a HACK DR2000 spectrophotometer. To ascer-
tain other environmental factors that may be occur-
ring (such as groundwater infiltration and dilution),
physical parameters, including pH, conductivity, dis-
solved oxygen, and temperature, are measured in
situ using a Hydrolab Surveyor II.
Following field investigations, phosphorus con-
centrations are plotted on maps of the dairy for
presentation to the landowner and research staff.
The landowner and South Florida Water Manage-
ment District staff discuss the survey results, that is,
the problems it identifies, any farm practices that
may have contributed to them, and potential solu-
tions to the problems within the operation.
Case Study
The following case study of a dairy operation along
the Lower Kissimmee River illustrates the process
and its value. A description of the study area, dairy
statistics, historical water qualify data, survey sam-
pling scheme, and survey data are given to docu-
ment the usefulness of the synoptic approach.
Study Area Description
This synoptic survey was conducted on a large dairy
operation in the Lower Kissimmee River project area
located approximately 14 km north of Lake Okee-
chobee. Surface drainage leaves the dairy primarily
through a marsh wetland system that joins a
tributary and canal before reaching the lake. This
394
-------
G J. SAWKA, P. RITTER, B. GUNSALUS, & T. ROMPOT
dairy was selected because high phosphorus con-
centrations were consistently detected in the dis-
charge waters (Table 1).
Table 1.—Total phosphorus period of record
averages of pre- and post-BMP values at routine
monitoring sites.
SITE
32B
49
1R1
IR2
33
32C
P.O.R.1
3.64
7.44
1.81
2.81
2.80
2.67
PRE-BMP
4.77
7.41
NA
NA
2.142
—
POST-BMP
2.78
7.50
NA
NA
4.053
2.673
% CHANGE
-^1.7*
1.2
NA
NA
—
1 Period of record for each site is within the date range of 4/87-3/92
Historical data were not continuous
"Sites initiated 10/1/91 (n=2)
*Data were significant at P=0.0001
The dairy consists of two barns and encom-
passes about 995 hectares with 2,340 cows. (Cow
numbers include milk herd, dry herd, calves, and
heifers.) On-farm BMPs include a waste manage-
ment system consisting of a first stage anaerobic
lagoon (approximately 1 hectare) and a second stage
aerobic waste storage pond (approximately 6 hec-
tares) at each barn. Effluent is pumped from the
waste storage pond and applied to the sprayfield
using a center pivot irrigation system. The largest
sprayfield is approximately 94 hectares; smaller
sprayfields range from 20 to 77 hectares. Sprayfields
are used for hay and sorghum production. Waste
management BMP construction was completed in
February 1990.
Dominant soils in the area are poorly drained; a
high seasonal water table within 25 cm of the surface
and other soils creates occasional ponding above the
surface. Most soils have a spodic horizon, or organic
hardpan, occurring within 50 to 125 cm of the sur-
face. The hardpan varies in thickness and consisten-
cy and promotes lateral subsurface drainage. These
soils are very permeable and somewhat responsive
to drainage. Other soils on lower landscape positions
lack subsurface restrictive layers and are peri-
odically ponded. Most of these areas lack drainage
outlets and are used for native range or as watering
places for livestock (McCollum and Pendleton,
1971). Soils on this dairy are similar to soils found
within the two basins. Soils within the region have a
low phosphorus retention capacity and do not inhibit
phosphorus transport.
Water Quality Historical Data and
Site Locations
Phosphorus discharge concentrations are collected
biweekly from sites 32B, 32C, 32D, 33, and 49 (see
Map 1). Sprayfield discharge has also been
monitored since 1988 at sites IR1 and IR2. Sites 32C
and D were added to the routine monitoring pro-
gram in October 1991 to monitor oncoming water
from a dairy operation to the north of the property
and outer pastures.
Table 1 lists the period of record discharge
averages, pre- and post-BMP phosphorus values at
the routine monitoring sites, and the percent of
change between the pre- and post-BMP values. The
data indicate that high phosphorus concentration
discharges were common to this dairy operation.
Some of these high phosphorus values can be at-
tributed to poor soil drainage in the area, past prac-
tices of animal lounging in wet areas, residual soil
effects, and animal density.
Although phosphorus concentrations have
steadily decreased since implementation of BMPs,
high phosphorus concentration discharges are still
observed (Fig. 1).
Synoptic Survey Sampling
Field investigations were conducted on October 15,
16, and 22, 1991. Before and during the field inves-
tigations, 5.5 cm of rainfall were recorded between
October 10 and 17. Soils in the pasture and interior
farm roads were saturated during the first two sur-
vey visits but were considerably dryer on the last
visit. Sampling was performed at 39 individual sites
during the three days of investigation. Map 1 gives
an overview of the property boundaries, location of
BMPs, hydrologic flow pathways, and routine
monitoring sites. Maps 2 and 3 (enlarged sections of
Map 1) show the synoptic survey site locations.
• Barn 1. Off-site discharges from Barn 1 are
routinely collected at sites 49, 32B, and 32C (Map 1).
The main source of discharge 'from site 49 (Map 2)
originates from the four smaller sprayfields serving
Barn 2. Sprayfield discharge flows through a series
of conveyance ditches before reaching site 49. Other
discharge sources from site 49 originate from animal
lounging areas located south of the Barn 1 high in-
tensity area and pasture areas east of the two
southern-most sprayfield pivots. Off-site discharge
from site 32B (Map 3) originates from lounging
areas north of Barn 1's high intensity area, the large
center pivot sprayfield located north of Barn 1, outer
pasture areas and wetlands on the northern bound-
ary of the property, and inflow from a dairy opera-
tion north of the property. Discharge at site 32C
originates from wetland and pasture areas in the
northeast area of the property.
• Barn 2. Sampling was conducted upstream from
site 33 near the south end of the dairy property
395
-------
Proceedings of National RCWP Symposium, 1992
1
a.
o*
a.
TOTAL PHOSPHORUS
Barn
- Site 32B
16.O
i.
a.
o*
OL
TOTAL PHOSPHORUS
4O
35
Barn #2 - Site 49
88
89
9O
Figure 1.—Time series graphs of total phosphorus concentrations at the primary outflows of dairy Barns 1 and 2.
(Map 2). The main source of off-site discharge from
site 33 originates from areas adjacent to the waste
storage pond and outer pastures. Runoff from pas-
lures southeast of Barn 2 is treated with an alterna-
tive nutrient management system (labeled on Map 2)
known as an "eco-reactor," which uses chemical
precipitation methods to remove nutrients from sur-
face water runoff. Discussion of this experimental
system (being developed by private consultants) is
beyond the scope of this paper.
Results and Discussion
As indicated by the pre- and post-BMP data, phos-
phorus concentrations at farm discharge sites have
decreased since the implementation of BMPs.
Average concentrations have decreased and con-
centrations are less variable within the post-BMP
data. However, high total phosphorus concentration
discharges above 1.2 mg/L were still detected. Con-
centrations of soluble reactive phosphorus from the
396
-------
G J. SAWKA, P. RITTER, B. QCJHSALUS, & T. ROMPOT
JH
t
WASTE
STORAGE
PONDS
SPRAYFIEU)
WITH
CENTER PIVOTS
LEGEND:
INSET MAP
___ PROPERTY LINE
-* FLOW DIRECTION
A ROUTINE
w MONITORING SITE
WASTE STORAGE PONDS
I
HIGH INTENSITY AREA
Map1.
synoptic survey ranged from 0.24 mg/L (sample
#26) to 30.24 mg/L (sample #20). Routine off-site dis-
charge points from Barn 1 at sites 49, 32B, and 32C
had soluble reactive phosphorus values of 9.35, 3.72,
and 1.50 mg/L respectively. Off-site discharge from
Barn 2 at site 33 was 2.57 mg/L. Values obtained
during the synoptic survey are similar to phos-
phorus concentrations observed through the routine
monitoring network—all above the 1.2 mg/L stand-
ard.
• Barn 1. Upstream of site 49 (Map 2), soluble
reactive phosphorus concentrations (samples #04
and #06) were greater than 13 mg/L. Inflows from
the east-west lateral ditch that originate from the
small sprayfields and the lounging area north of
397
-------
Proceedings of National RCWP Symposium, 1992
WASTE
STORAGE
PONDS
HIGH
INTENSITY
AREA
LEGEND:
... PROPERTY UNE
-» FLOW DIRECTION
• SYNOPTIC
SAMPLING SITE
fiff SAMPLE NO.
(mo/1) P-CONC.
- J
Map 2.
Barn 2 had a concentration of 15.50 mg/L (sample
#05). West of the confluence, samples from ditches
draining animal lounging areas south of the Barn 1
high intensity area (samples #11 and #13) had con-
centrations of 9.25 and 13.85 mg/L. Ditches from
pastures near the small sprayfields exhibited ex-
tremely light flow to ponded conditions, making it
difficult to accurately assess impacts from this area.
Survey results indicate that the primary sources
of high phosphorus concentrations at site 49 are
from the east-west lateral ditch (sample #05) and
ditches draining the animal lounging areas south of
the Barn 1 high intensity area. Further investigation
of lounging areas north of Barn 2 and the sprayfield
ditches is needed.
398
-------
Q.J. SAWKA, P. RITTER, B. GUHSALUS, & T. ROMPOT
(ADJACENT DAIRY OPERATION)
(3.66)
4
X
8
bl
8
1
0
1
1
1
1
1
1
1
1
I
1
1
SPRAYFIELD
WITH
CENTER PIVOTS
I
LEGEND:
PROPERTY UNE
FLOW DIRECTION
SYNOPTIC
SAMPLING SITE
Oft SAMPLE NO.
(mg/l) P-CONC.
Map 3.
Off-site discharge at site 32B (Map 3) northwest
of Barn 1 had a soluble reactive phosphorus con-
centration of 3.72 mg/L. Flows from the large
sprayfield north of site 32B were 1.06 mg/L or less.
While no other samples were taken in this area,
potentially high concentrations may originate from
animal lounging areas northeast of site 32B.
Along the northeast property line, site 32C had a
soluble reactive phosphorus concentration of 1.50
mg/L. Flows originate from a small wetland and pas-
tures to the south and a larger wetland with adjacent
pastures to the north. Discharge from the northern
wetland had a soluble reactive phosphorus con-
centration of 4.65 mg/L at site 32D, which receives
inflow from another dairy operation located to the
north. South of the smaller southern wetland, the
ditch discontinues, resulting in minimal flushing
while receiving input from surrounding pastures.
399
-------
Proceedings of National RCWP Symposium, 1992
This area had a soluble reactive phosphorus value of
3.38 mg/L. In addition to applying soil amendments,
the landowner is considering fencing out the heifers
that currently graze along the north sections of the
property and in areas of inflow to site 32C.
II Barn 2. Water leaving the property at site 33
(Map 2) measured 2.57 mg/L. Upstream, outflows
from the waste storage pond measured 30.25 mg/L
(sample #20). During the survey, water from the
storage pond flowed through adjacent wetlands into
the ditch leading to the off-site discharge point. Sur-
vey results infer that the high concentrations of
soluble reactive phosphorus at off-site discharge site
33 result from overflow from the waste storage pond.
A decision was made with the landowner that main-
taining proper storage capacity of the waste storage
pond would help eliminate high off-site nutrient dis-
charges. Further investigation of drainage ditches
from the Barn 2 waste management area is sug-
gested.
Conclusion
Following BMP installation, proper management of
dairy farms is necessary to achieve the desired
phosphorus reduction results. Long-term monitor-
ing does not provide the data needed to identify site-
specific high sources of phosphorus within the farm
and assist in nutrient management decisions. Unlike
routine monitoring, synoptic surveys are investiga-
tive surveys that provide detailed information to as-
sess on-farm water quality problems and improve
nutrient management strategies. These surveys also
provide the opportunity to better relate and assess
water quality improvements resulting from BMP im-
plementation. In this study, intense outer pasture
grazing, animal lounging in wet areas, improper
waste storage pond operation, and phosphorus-
contaminated ditches were identified as potential
source contributions that result in high phosphorus
discharge.
References
Allen, L.H. Jr., E.H. Stewart, W.G. Knisel Jr., and RA Slack. 1976.
Seasonal variation in runoff and water quality from the Taylor
Creek watershed, Okeechobee County, Florida. Proc. Soil
Crop Sci. Soc. Fla. 35:126-38.
Davis, F.E. and M.L. Marshall. 1975. Chemical and Biological In-
vestigations of Lake Okeechobee, January 1973-June 1974,
Interim Report Tech. Publ. #75-1. Resour. Plann. Dep. Cen-
ter. S. Fla. Flood Control Distr., West Palm Beach, Fla.
Federico, AC. 1977. Investigations of the Relationship Between
Land Use, Rainfall, and Runoff Quality in the Taylor Creek
Watershed. Tech. Publ. #77-3. Resour. Plann. Dep., S. Fla.
Water Manage. Distr., West Palm Beach, FL.
Federico, AC., KG. Dickson, C.R Kratzer, and F.E. Davis. 1981.
Lake Okeechobee Water Quality Studies and Eutrophication
Assessment. Tech. Publ. #81-2. Resour. Plann. Dep., S. Fla.
Water Manage. Distr., West Palm Beach, FL.
Florida Department of Environmental Regulation. 1987. Chapter
17-670, Florida Administrative Code. Dairy Rule. Tallahas-
see, FL.
Hines, W.G., DA Rickert, and S.W. McKenzie. 1977. Hydrologic
Analysis and River-quality Data Programs. U.S. Geo. Surv.,
Portland, OR.
Joyner, B.F. 1971. Appraisal of Chemical and Biological Conditions
of Lake Okeechobee, Florida, 1969 - 1970. Open File Rep.
71006. Geo. Surv., Water Resourc. Div., U.S. Dep. Treasury,
Tallahassee, FL.
McCollum, S.H. and RE Pendleton. 1971. Soil Survey of
Okeechobee County, Florida. U.S. Govt. Print Off.,
Washington, DC.
Ritter, G.J. and L.H. Allen, Jr. 1982. Taylor Creek Headwaters
Project, Phase I Report: Water Quality. Tech. Publ. #82-8.
Resour. Plann. Dep., S. Fla. Water Manage. Distr., West Palm
Beach, FL
South Florida Water Management District. 1989a. Document in
Support of Chapter 40E-61, Works of The District Basins.
September 1989. West Palm Beach, FL.
. 1989b. Interim Surface Water Improvement and Manage-
ment (SWIM) Plan for Lake Okeechobee. March 1. West
Palm Beach, FL.
Stanley, J., V. Hoge, L. Boggs, and G. Ritter. 1988. Taylor Creek-
Nubbin Slough RCWP 1988 Annual Report. Okeechobee
Agric. Stabil. Conserv. Serv., Okeechobee, FL.
Stanley, J.W. and B. Gunsalus. 1991. Ten Year Report Taylor
Creek-Nubbin Slough Project RCWP 1981-1991. Okeechobee
Agric. Stabil. Conser. Serv., Okeechobee, FL.
*US. GOVERNMENT PRINTING OFFICE: 19 9 2 - 6 <. a - o o *6 o o l <.
400
-------
-------
------- |