k'-S Environmental
Protection Agency
Region 3, Philadelphia, PA
Drinking Water/Ground Water Protection (3WM40)
Information Resources Management (3PM50)
EPA 903-K-94-002
December 1994
Use of GIS to Assess Ground
Water Vulnerability to
Pesticide Contamination
(
Jefferson County, WV Pilot Study
-------
-------
Use of GIS to Assess
Ground Water Vulnerability
to Pesticide Contamination
Jefferson County, WV PUot Study
DECEMBER 1994
-------
ACKNOWLEDGEMENTS
This document was written by Cynthia Kranz Greene of the Ground Water Protection
Section in EPA Region ffl. GIS analyses, maps and Appendix B (technical CIS
documentation) were produced by Donald C. Evans, of R.O.W. Sciences, Inc., for EPA
Region Hi's Information Resources Management Branch. Appendices C and D (survey
questionnaire, enumerator training manual, data entry software and users manual) were
developed for EPA Region m by ICF, Inc. under contract number 68-C8-0003. EPA
Region ffl's MERITS fund, and the WV Department of Natural Resources ground water
protection grant provided funding for contractual support. Funding for additional data
collection and automation was provided by State and Federal agencies who cooperated with
EPA. Much gratitude is extended to the following people for all the support they provided
for the pilot project and the development of this document:
Jon Capacasa, Deputy Director of EPA's Chesapeake Bay Program provided the
initial inspiration for the project. Stuart Kerzner, Chief of the Drinking Water/Ground
Water Protection Branch, and Virginia Thompson, Chief of the Ground Water Protection
Section, provided support and encouragement. Sumner Crosby of the Water Quality
Planning Section provided valuable technical guidance, map edits and computer programming
assistance. Robert Braster, Chief of the EPA's Information Resources Management Branch
(ERMB) provided essential staff support. David West of IRMB provided valuable technical
assistance and support with map development. Robert Schnably of the Soil Conservation
Service and Craig Yohn of the Cooperative Extension Service hi Jefferson County established
and maintained excellent rapport with the farming community during the Agricultural
Practices Survey, and provided invaluable review and editing support for the development of
the Survey instrument and the interpretation of results. Steve Carpenter and Robert Schnably
of the Soil Conservation Service spent many patient hours compiling, converting and
transmitting to EPA a substantial amount of die digital data, such as soil types and land use,
which was essential for this project. Ken Green, Executive Director of the Eastern
Panhandle Regional Planning and Development Council, served as the primary individual
responsible for administering the Agricultural Practice Survey, and provided many inspiring
ideas throughout the project. Mark Kozar and Bill Hobba of the U.S. Geological Survey
provided invaluable insight regarding the nature of ground water movement in the County, as
well as important data on water levels and the locations and attributes of karst features. Pete
Lessing of the WV Geologic and Economic Survey provided an up to date geologic coverage
for the County. Mike Sienkewitz of the Agricultural Stabilization and Conservation Service
provided essential acreage data. Doug Hudson of the WV Department of Agriculture
sampled wells for pesticides in Jefferson County, and provided helpful guidance on the
project. John Means, Jwewis Baker and Pat Campbell of the WV Department of Natural
Resources provided the DRASTIC Ground Water Vulnerability map. Stephen Hammel of
the University of Pennsylvania researched valuable information on GIS data processing and
analysis. Donald Lott, Chief EPA Region JH's Pesticides Grants Section (PCS), Gerard
Florentine of PGS and Karen Angulo of the Office of Pesticide Programs (EPA
Headquarters) provided essential information on EPA's pesticide regulatory programs.
iii
-------
DISCLAIMER
This document describes a pilot project initiated by EPA Region m in Jefferson
County, WV to explore the use of Geographic Information Systems (GIS) for purposes of
identifying areas where pesticides pose a potential threat to ground water quality within the
County. As part of the project, detailed information was obtained from landowners
regarding their pesticide usage and other agricultural practices. Landowners agreed to
participate in the survey if their identities remained confidential.
Although this report contains maps which display survey results, great care was taken
in the map creation to present survey data in a manner that demonstrates the types of
analyses which can be conducted, without compromising the confidentiality of the survey
participants. For example, wherever farm locations and survey data are mapped, farm parcel
boundaries and roads are not displayed, for purposes of confidentiality. In addition, EPA
Region in and the West Virginia workgroup will only use the survey information as a
research tool, and will not use the data to support any specific enforcement action targeted to
survey participants.
The types of GIS analyses conducted and the vulnerability assessment techniques
utilized in the project and presented hi this report are examples of the types of analyses and
assessments which can be conducted to identify threats to ground water quality from
pesticides, the inclusion of specific techniques in this report is not an endorsement of that
technique; every effort is made to explain the pros and cons, assumptions and limitations of
the methods utilized. For development of their State Management Plans, States should
choose methods and select appropriate GIS analyses based on the ground water protection
philosophy, hydrogeology, pesticide usage patterns, agronomic practices and available
resources of the State.
IV
-------
TABLE OF CONTENTS
TITLE PAGE
EXECUTIVE SUMMARY x
CHAPTER 1: INTRODUCTION 1
A. PURPOSE OF DOCUMENT 1
B. HISTORY OF THE PESTICIDES IN GROUND WATER CONCERN . 1
C. EPA's PESTICIDES AND GROUND WATER STRATEGY 3
D. PURPOSE OF JEFFERSON COUNTY PILOT PROJECT 6
E. GENERAL DESCRIPTION OF PROJECT AREA 6
F. RATIONALE FOR SELECTION 8
CHAPTER 2: GEOGRAPHIC INFORMATION SYSTEMS 9
A. DEFINITION AND CONCEPTUAL DISCUSSION 9
B. APPLICATION OF GIS TO THE DEVELOPMENT OF STATE
PESTICIDE AND GROUND WATER MANAGEMENT PLANS ... 12
CHAPTER 3: BUILDING THE JEFFERSON COUNTY DATA BASE 14
A. ORGANIZATION OF THE WORKGROUP 14
B. DETERMINATION OF OBJECTIVES, IDENTIFICATION OF
DATA NEEDS 14
C. DESCRIPTION OF DATA LAYERS, INCLUDING SOURCES
AND METHODS USED TO OBTAIN AND AUTOMATE DATA ... 16
SECTION I: BASE MAP, SOILS, GEOLOGY 16
SECTION II: DRASTIC, SPISP 28
SECTION HI: LAND USE, LAND COVER 41
CHAPTER 4: AGRICULTURAL PRACTICES SURVEY 63
A. BACKGROUND 63
B. SURVEY DESIGN: DEVELOPMENT OF QUESTIONNAIRE .... 63
C. SURVEY DESIGN: SELECTION OF SAMPLE 64
D. CONFIDENTIALITY POLICY 66
E. ADMINISTRATION OF SURVEY 66 .
F. SURVEY EDITS, ESTIMATES AND ASSUMPTIONS 67
G. AUTOMATION OF SURVEY DATA 74
H. PROBLEMS ENCOUNTERED, OBSERVATIONS MADE IN
THE FIELD 75
I. SURVEY RESULTS 76
-------
TITLE EASE
CHAPTER 5: USING CIS TO IDENTIFY AREAS
SUSCEPTIBLE TO PESTICIDE
CONTAMINATION OF GROUND WATER 82
A. NONPOINT SOURCE THREATS TO GROUND WATER
FROM LEACHING : 82
B. POINT SOURCE THREATS TO GROUND WATER AND
DRINKING WATER WELLS/SPRINGS FROM CONDUIT
FLOW RECEPTORS 98
CHAPTER 6: SUMMARY OF PROBLEMS IDENTIFIED,
MANAGEMENT RECOMMENDATIONS AND
INSTITUTIONAL RESPONSE
TO THE PROBLEMS 107
A. POTENTIAL PROBLEM AREAS IN JEFFERSON COUNTY
IDENTIFIED THROUGH THE AGRICULTURAL
PRACTICES SURVEY 107
B. MANAGEMENT RECOMMENDATIONS TO
ADDRESS PROBLEMS IDENTIFIED BY THE
SURVEY 109
C. POTENTIAL PROBLEM AREAS IN JEFFERSON COUNTY
IDENTIFIED BY THE GIS ANALYSES, AND
RECOMMENDED SOLUTIONS Ill
D. INSTITUTIONAL RESPONSE: EPA's PESTICIDE
REGISTRATION PROCESS, PROPOSED RULES,
REGULATIONS, POLICY, TECHNICAL
AND FINANCIAL ASSISTANCE, STATE AND LOCAL
PROTECTION AND WELL SAMPLING INITIATIVES 113
CHAPTER 7: LESSONS LEARNED, KEY CONSIDERATIONS,
BENEFITS, COSTS, NEXT STEPS 133
A. LESSONS LEARNED, KEY CONSIDERATIONS 133
B. BENEFITS OF PILOT PROJECT 137
C. FINANCIAL SUMMARY 139
D. FUTURE PLANS AND NEEDS 141
BIBLIOGRAPHY 143
VI
-------
APPENDICES
APPENDIX A:
APPENDIX B:
DETAILED SURVEY RESPONSES FOR JEFFERSON COUNTY,
WV INDIVIDUAL PROPERTY QUESTIONNAIRE AND
OBSERVATION RECORD
TECHNICAL CIS DOCUMENTATION: STEPS INVOLVED IN
AUTOMATING SOURCE DATA, INCLUDING PRE-PROCESSING,
FORMATTING, IMPORTING AND MANIPULATING DATA FILES,
DOCUMENTATION OF PROBLEMS AND SOLUTIONS,
INCLUDING IMPORT/EXPORT PROBLEMS BETWEEN GRASS
AND ARC/INFO, DOCUMENTATION AND DESCRIPTION OF
EACH GIS COVERAGE, INCLUDING DATA SOURCE, SCALE,
CREATION PROCESS AND ATTRIBUTES
APPENDK C:
APPENDK D:
AVAILABLE UPON REQUEST:
JEFFERSON COUNTY AGRICULTURAL PRACTICES SURVEY
QUESTIONNAIRE BOOKLET AND TRAINING MANUAL FOR
ENUMERATORS
SOFTWARE AND DOCUMENTATION FOR DATA ENTRY OF
SURVEY RESPONSES
APPENDK E:
APPENDK F:
UNDER DEVELOPMENT:
SOFTWARE AND USER'S MANUAL FOR CONVERSIONS OF
PESTICIDE TRADE NAME PRODUCT APPLICATION RATES TO
ACTIVE INGREDIENT APPLICATION RATES
COMPARISON OF RESULTS FROM JEFFERSON COUNTY
AGRICULTURAL PRACTICES SURVEY, VERSUS COUNTY
AGRICULTURAL AGENT AND COMMERCIAL DATA BASE
ESTIMATES
Vll
-------
LIST OF MAPS
MAP TITLE PAGE
1. LOCATION OF STUDY AREA 7
2. PLANIMETRIC AND HYDROLOGIC FEATURES 17
3. USGS 7.5' QUADRANGLES 18
4. GENERALIZED BEDROCK GEOLOGY 19
5. GEOLOGY, FAULTS, FRACTURES AND SINKHOLES 21
6. WATER TABLE CONTOURS AND WELLS 23
7. MAJOR SOIL TEXTURES 26
8. DRASTIC GROUND WATER VULNERABILITY 31
9. DRASTIC WITH STREAMS 36
10. DRASTIC, FAULTS, FRACTURES AND SINKHOLES 37
11. SPISP SOIL LEACHING POTENTIAL 40
12. GENERALIZED LAND USE 42
13. JEFFERSON COUNTY FARMS 44
14. SURVEYED FARMS 45
15. ROCK TYPES AND SURVEYED FARMS 47
16. PRIVATE DRINKING WATER WELLS AND SPRINGS 48
17. TOTAL PESTICIDE USAGE PER FARM (1988) 50
18. TOTAL PESTICIDE USAGE PER FARM (1989) 51
19. TOTAL POUNDS OF PESTICIDES APPLIED PER ACRE (1988) 52
20. TOTAL POUNDS OF PESTICIDES APPLIED PER ACRE (1989) 53
viii
-------
LIST OF MAPS (Continued)
MAP TITLE PAGE
21. PRIORITY LEACHING PESTICIDE USAGE PER FARM (1988) 55
22. PRIORITY LEACHING PESTICIDE USAGE PER FARM (1989) 56
23-27 PRIORITY LEACHER USAGE RANK 1 THROUGH 5 (1989) 58-62
28. SURVEYED vs. NON-SURVEYED CROPLAND 65
29. LEACHER USE LOCATIONS / SOIL POTENTIAL (1989) 83
30. LEACHER LOADINGS / SOIL POTENTIAL (1989) 84
31. DIVERSITY OF LEACHER USE / SOIL POTENTIAL (1989) 87
32. ATRAZINE USE / SOIL POTENTIAL (1989) 89
33. LEACHER LOCATIONS / DRASTIC (1989) 92
34. LEACHER LOADINGS / DRASTIC (1989) 93
35. LEACHER USE LOCATIONS / DEPTH TO GROUND WATER 94
36. LEACHER LOADINGS / DEPTH TO GROUND WATER 96
37. LEACHER LOADINGS / DEPTH TO GROUND WATER
SOIL POTENTIAL 1 97
38. WELLS OR SPRINGS WITH PESTICIDE POINT SOURCES
WITHIN 1000 FEET 99
39. WELLS OR SPRINGS WITH PESTICIDE POINT SOURCES
WITHIN 200 FEET 100
40. PROXIMITY OF PESTICIDE DISPOSAL SITES TO
PRIMARY WELL OR SPRING 101
41. VULNERABLE WELLS AND SPRINGS 103
42. POINT SOURCE THREATS TO GROUND WATER
CONDUIT RECEPTORS 105
ix
-------
EXECUTIVE SUMMARY
USE OF CIS TO ASSESS GROUND WATER VULNERABILITY
TO PESTICIDE CONTAMINATION:
Jefferson County, WV Pilot Study
This document describes a pilot project which EPA Region HI implemented in
Jefferson County, WV in conjunction with an interagency workgroup. The purpose of the
project was to explore the use of Geographic Information Systems (GIS) as a tool to assess
the relative vulnerability of ground water to pesticide contamination.
Pesticides have been detected in the ground water of at least 42 States in the nation,
according to EPA's Office of Pesticide Program's Pesticides in Ground Water Database.
This data base is a compilation of the results of pesticide hi ground water monitoring studies
which have taken place between 1971 and 1991 by different agencies, manufacturers and
university researchers. The data base shows that 24% of the wells tested had detections,
nearly 60% of these detections were greater than or equal to EPA's drinking water standards,
and 83% of the detections could be attributed to normal field use. Overall, 132 different
pesticides were detected.
To address the issue of pesticide detections hi ground water, EPA released a
Pesticides in Ground Water Strategy on October 31, 1991. The Strategy requires States to
develop and implement State Management Plans (SMPs) to protect their ground waters from
pesticides. The use of specific pesticides could be restricted or canceled hi States which lack
an EPA approved SMP.
To assist States with the development of then- SMPs, EPA Region ffl's Ground Water
Protection Section selected two areas hi the Region (Jefferson County, WV and Lancaster
County, PA) to demonstrate how GIS can be used to target priority areas for management
measures to prevent ground water contamination from pesticides. Interagency workgroups of
Federal, State, and local environmental, agricultural and health agencies were established hi
both Counties to carry out the pilot projects. These agencies shared expertise, resources and
data, hi an effort that turned out to be a showcase of cooperation. In addition to the insights
gained on the use of GIS to help develop SMPs, the pilot projects provide a good example of
how to organize and mobilize a multi-disciplinary workgroup, including how to identify
needs and work together on solutions. This document provides the results of the Jefferson
County project, discusses the use of GIS to support the development of SMPs, and discusses
some approaches which are being taken currently hi the Lancaster County project, which
differ from the Jefferson County project. This document also describes lessons learned
through the pilot project experiences.
In Jefferson County, detailed data on pesticide usage and drinking water wells was
obtained by the Region 9 Planning and Development Council through interviews with 118
fanners hi the Spring of 1990. In Lancaster County, the Conservation District interviewed
over 250 farmers in 1991 and 1992. Through these Agricultural Practices Surveys, the
-------
following data was collected for both Counties: pesticide application rates, crop types, crop
acreage, pesticide storage, disposal, mixing, loading and equipment cleaning practices, well
location and construction, well water monitoring data, animal waste management practices,
fertilizer usage, conservation practices and information regarding on-lot septic systems and
underground storage tanks on the properties.
EPA Region Hi's GIS was used to enter, store and analyze Jefferson County survey
data on pesticide usage and drinking water wells, in combination with data on soils and
hydrogeology, to identify areas where ground water is susceptible to contamination.
Specifically, analyses were conducted within the following two categories:
1. Nonpoint sources and normal leaching: Areas were identified where pesticides
classified as "priority teachers" are used over soils with high, medium and low
leaching potentials. Soil and pesticide leaching potentials were based on the Soil
Pesticide Interaction Screening Procedure (SPISP) method developed by the Soil
Conservation Service. Data from the National Pesticide Survey was also used to help
assign pesticide leaching potentials. The diversity of pesticide usage per farm was
evaluated, in addition to the total application rates. Since atrazine had the highest and
most widespread use as compared to other priority leachers in 1989, an analysis was
also conducted just for atrazine. In addition to determining soil leaching potentials,
pesticide usage was evaluated in conjunction with ground water vulnerability data,
based first on the DRASTIC scoring procedure, and then by using a depth to ground
water map for the County. A comparison of the soil teachability versus ground water
vulnerability evaluations was made, in order to identify areas where pesticides are
likely to not only leach to the root zone, but also to leach to ground water.
2. Point source threats to ground water and drinking water wells/springs from conduit
flow receptors: GIS was used to display the locations of wells and springs on the
surveyed farms, and to highlight those farms where potential point sources of
contamination were found within 200 feet and 1,000 feet of the well or spring.
. Potential point sources of contamination were considered to be pesticide storage,
disposal, mixing, loading, equipment cleaning or spill sites, which met specified
criteria (such as pesticide containers are buried in the ground, or the mixing and
loading takes place over a permeable surface). Wells were identified which were
either poorly constructed, old, or of shallow depth. In addition, both wells and
springs were identified which were poorly protected at the surface. Finally, farms
were targeted which had conduit flow receptors, such as abandoned wells, sinkholes,
faults or fractures, as well as potential point sources of contamination within 1,000
feet of a vulnerable drinking water source on the farm. This analysis was conducted
to identify sites with the highest likelihood of pesticide point source contamination of
the drinking water supply, either due to the construction of the well or spring or due
to the proximity of potential conduit receptors and pesticide sources.
XI
-------
All of the analyses conducted underscored the importance of focusing on vulnerability
assessments at the rr»ost refined resolution possible. For example, it is more useful to utilize
soil teachability ck ;fications for polygons of a few acres in size (using SPISP), than to
utilize ground water vulnerability classifications for areas of 100 acres or larger in size
(using DRASTIC). It is essential to utilize detailed soil data to determine the potential of a
pesticide to leach to the root zone, hi order to prevent contamination of ground water. It is
also critical to know how pesticides are being handled, stored, and disposed of right at the
farmstead, near drinking water wells. Data on the construction of wells and proximity to
sinkholes and other potential avenues for contamination is essential for conducting a localized
vulnerability assessment.
Both the Agricultural Practices Survey and the GIS analyses uncovered areas of
potential problems in Jefferson County. The following is a partial listing of some of the
highlights of the analyses and the survey:
* Surveyed farms with the highest likelihood of pesticide leaching to ground water are
concentrated hi the middle and eastern parts of the County's carbonate valley.
* Corn was identified as the crop which used the highest quantity of priority leaching
pesticides. The top three crops which relied most heavily on priority leachers as a
percentage of overall pesticide use were: truck crops, pasture and corn. Other crops,
such as apples, relied heavily on pesticides, but less than 50% of overall use was of
priority leaching pesticides. Other crops which were less reliant on priority leachers
included alfalfa, peaches, and nectarines.
* The five pesticides used hi highest quantity in the County are all priority leachers
(atrazine, metolachlor, alachlor, simazine, and 2,4-D).
* A total of 40,882 pounds of pesticides were applied to 105 surveyed parcels hi 1989;
priority leaching pesticides made up 63% of the total pesticide use.
* Approximately 16% of the fanners who used pesticides disposed of 47 different
unwanted pesticides or their containers in a farm dump or by burning on the property.
Fanners expressed frustration over not knowing what to do with unwanted pesticides
or containers.
* Most pesticide mixing and loading takes place over permeable surfaces.
* Less than 2% of the fanners participate in an Integrated Pest Management program.
* Over half (57%) of the properties surveyed have either unprotected springs or wells
which are susceptible to contamination due to their age, construction and/or depth,
and approximately 40% of the pesticide application areas are 200 feet or less from the
drinking water source.
xii
-------
* 21% of the properties surveyed have wells or springs susceptible to contamination
from pesticide point sources.
* A total of 65 different pesticides are stored by farmers who participated in the survey.
* Approximately 70% of the respondents use commercial applicators for all or part of
their pesticide spray ing. Many commercial applicators had incomplete pesticide
application records, no record of actual location of applications and were unaware of
critical ground water resource areas to protect (such as avoiding sinkholes with the
sprayer).
* Over 30% of the respondents have underground petroleum storage tanks (USTs)
which are made of corrosive material; half of these tanks have been in the ground 15
years or longer. In addition, of the 89 farms which reported USTs, close to 50%
have tanks located 150 feet or less from the primary drinking water source.
Local USDA offices in Jefferson County are using the survey results and GIS
analyses to target farmers in need of technical and financial assistance to better manage
pesticides and other contaminant sources on their properties. In addition to the valuable
insights gained regarding agricultural practices and potential risks to ground water in
Jefferson County, many important lessons were learned through the pilot project.
Chapter 7 of this document contains information on "lessons learned" which should prove
helpful to anyone embarking upon a similar endeavor.
Some of the lessons learned in Jefferson County have been put to use hi the Lancaster
County project. For example, in Lancaster County, a watershed was selected as the study
area, rather than using a political boundary. A statistically representative sample of farmers
was surveyed and crop types are being determined for the entire study area, using aerial
photo-interpretation. Survey results, including estimated pesticide application rates, will be
extrapolated to the non-surveyed population, in order to conduct modeling on an entire
watershed basis. The Lancaster project will provide a comparison of the accuracy and costs
associated with two different methods used to obtain pesticide usage data, specifically, on-
farm interviews versus literature estimates for crops identified through aerial photo-
interpretation. Other differences between the two projects include: in Lancaster County, the
workgroup directed the type of hydrogeologic characterization which was conducted in the
beginning of the project, local agencies administered the survey (rather than working as
subcontractors on an EPA contract) and public wellhead protection areas were delineated in
the first stage of the project. Finally, the selected area dovetails with a USDA water quality
initiative which can provide funding to assist the farmers in areas targeted by the GIS
analyses.
Xlll
-------
In addition to the many lessons learned, the Jefferson County pilot project has proved
that the use of GIS to develop a State Pesticide in Ground Water Management Plan has many
benefits, including:
1. GIS ultimately saves managers time and money, because it accelerates the whole
decision making process.
2. Even though initial automation of data is expensive, once data is automated, there are
a multitude of applications.
3. GIS can handle complex queries to target vulnerable areas. Targeting saves time and
money, because it allows resources to be spent more efficiently, by focusing
inspections, technical assistance, enforcement and monitoring in the high risk areas.
4. GIS becomes a bridge linking many sources of data and disciplines. This is
particularly noteworthy for the applications described hi this document, because of the
need for a cooperative cross programmatic workgroup to develop SMPs.
5. GIS allows for the storage and updating of data that changes over the years, such as
pesticide usage, inspection and enforcement records, management practices and
property boundaries.
This document includes the following: a brief history of the pesticides and ground
water concern, a general discussion of basic capabilities of Geographic Information Systems,
a description of how the Jefferson County data base was built, results of the Agricultural
Practices Survey, an explanation of the design and administration of the Survey, a description
of the types of analyses which were conducted, a summary of the areas targeted as potential
problem areas, recommendations for management measures to address problems identified, a
summary of federal, State and local initiatives to protect ground water from pesticides,
information on lessons learned and benefits of the project, a financial summary, and a
description of future plans and needs for technical assistance and additional analyses hi the
study area and in Lancaster County.
Specific products which are available from the Jefferson County project, as
appendices to this document include: the questionnaire and enumerator training manual,
detailed survey results, data entry software, and a technical description of the mathematical,
logical and cartographic procedures used to automate the source data, conduct the analyses
and create the maps. Products under development include: brand name to active ingredient
conversion software and a comparison of County Agricultural Agent pesticide usage and data
base estimates versus on farm interviews.
xiv
-------
CHAPTER 1: INTRODUCTION
A. PURPOSE OF DOCUMENT
B. HISTORY OF THE PESTICIDES IN GROUND WATER CONCERN
C. EPA's PESTICIDES AND GROUND WATER STRATEGY
D. PURPOSE OF JEFFERSON COUNTY PILOT PROJECT
E. GENERAL DESCRIPTION OF PROJECT AREA
F. RATIONALE FOR SELECTION
A. Purpose of Document
This report describes a pilot project which EPA Region in implemented hi Jefferson
County, WV in conjunction with an interagency workgroup. The purpose of the project was
to explore the use of GIS as a tool to assess the relative vulnerability of ground water to
pesticide contamination hi Jefferson County. This document provides the results of the
project and discusses the use of GIS to support the development of State Pesticides and
Ground Water Management Plans. Results of GIS analyses and ground water vulnerability
assessments are presented hi the document. The inclusion of specific techniques hi the report
is not an endorsement of that technique; every effort is made to explain the pros and cons,
assumptions and limitations of the methods utilized. For development of Pesticides and
Ground Water State Management Plans, States should choose methods and select appropriate
GIS analyses based on the ground water philosophy, hydrogeology, pesticide usage patterns,
agronomic practices and available resources of the State.
B. History of the Pesticides in Ground Water Concern
In 1985, U.S. farmers applied approximately 400,000,000 kilograms (close to 900
million pounds) of pesticide active ingredients to agricultural lands (14). Increased and
continued pesticide use is associated with human health risks, adverse effects of non-target
organisms and contamination of air, soil and water. Less than 0.1% of applied pesticides are
estimated to actually reach targeted pests, leaving a large percentage available for potentially
contaminating environmental resources (22). While the contamination of any of our
environmental resources is a critical issue, this document will focus on the concern of ground
water contamination by pesticides.
To obtain a statistically accurate assessment of the frequency and concentration of
pesticide contamination of drinking water wells across the country, EPA carried out the
National Pesticide Survey between 1988 and 1990. EPA sampled approximately 1300 wells
for the presence of 126 pesticides or their degradates. The results statistically represent
approximately 94,600 community water system (CWS) wells and over 10.5 million rural
domestic wells throughout the United States. In November of 1990, EPA reported that the
ix-sults of sampling indicate 10.4% of all CWS wells and 4.2% of all domestic wells hi the
U.S. have detectable residues of at least one pesticide. A total of 16 different pesticides or
then- metabolites were detected at least once.
-------
The National Pesticide Survey was not a ground water study, but rather an evaluation
of currently used drinking water wells. In addition, the survey was not designed to detect all
the possible "hot spots" where localized contamination may exist. Therefore, survey results
should be evaluated in conjunction with EPA's Office of Pesticide Program's (OPP's) data
base. In August of 1992, EPA published a report entitled, "Pesticides in Ground Water Data
Base: A Compilation of Monitoring Projects: 1971-1991". The report contains a summary
and analysis of all the pesticides in ground water monitoring data that OPP currently had
available as of August 1992. The data was assembled from numerous sources, including
federal, State and local agencies, universities and private institutions, chemical companies
and consulting firms.
OPP's report summarizes data from the sampling of 68,824 wells across the country
for the presence of 302 pesticides or pesticide metabolites. The results show that pesticides
have been found in the ground water of at least 42 States. There were 132 pesticides (or
their metabolites) detected in at least one well. One or more pesticides were detected in 24%
of the wells sampled. Out of those wells with detections, nearly 60% had concentrations
greater than or equal to EPA's drinking water standards (MCLs). Most (96%) of the
detections which were greater than or equal to MCLs were found in drinking water wells
(versus monitoring or observation wells). Normal field use (versus spills, and other point
sources) was considered the source of 83% of all the pesticide detections.
There are some limitations associated with the use of OPP's data base. Most notably,
there are differences between studies in sampling practices, analytical methods, limits of
detection, and study design, which renders comparisons between studies, extrapolations, or
trend analyses fairly meaningless. The data is not representative of general drinking water
quality in the United States. For example, many of the monitoring efforts which yielded data
for the report were initiated in response to suspected problems, and resulted in a high
number of positive samples for a local area; these results cannot be extrapolated to represent
a larger region or State.
Given their limitations, the National Pesticide Survey and OPP's data base still
provide useful insights regarding the frequency of contamination at levels of health concern
(at or above an EPA Maximum Contaminant Level or Health Advisory Level). In the
National Pesticide Survey, there were no detections above levels of health concern in
Community Water Supply wells and less than 1 % of domestic wells are estimated to have
concentrations above levels of concern to human health. OPP's data base, however,
indicates a higher frequency of detections above levels which pose a health concern. In
almost 30% of the 42 States which had pesticide detections, concentrations were greater than
the MCL in 25% of the wells with detections. The majority (over 90%) of the MCL
exceedances in the data base, however, appear to be limited to four States (New York,
Florida, California and Connecticut).
Since a statistically representative survey of pesticide detections in the nation's ground
water (as opposed to drinking water wells) has not been conducted, conclusions from the
-------
National Pesticide Survey and OPP's data base can be misleading. The available data,
however, does appear to indicate that although the majority of drinking water wells
throughout the United States do not contain pesticides above levels of concern to human
health, there are several "hot spots" throughout the country where numerous wells are
contaminated with pesticides at levels that do pose a health concern. In addition, both the
National Pesticide Survey and OPP's data base show that the most frequently detected
pesticides are limited to approximately 25 compounds. This information helps to narrow the
focus on dealing with the top priority "bad actors" first. In addition, several of these
compounds are no longer registered for use in the United States.
Both the National Pesticide Survey and OPP's data base tell us that pesticide
detections in ground water are occurring, and as more monitoring studies are conducted and
compiled over time, as evidenced by the data base, more detections are found. The data
shows that there are areas throughout the country where our ground water resources are
either at risk or are already contaminated by pesticides. Once contamination of ground water
has occurred, it is usually not economically or technically feasible to restore the resource. In
addition, while the national survey indicates a low frequency nationwide of pesticide
detections in drinking water above the MCL, there has been little analysis of either the
potential for these levels to increase over time, or the potential risk to surface water
ecosystems from the levels commonly found hi ground waters. Ground water discharges to
wetlands or other bodies of surface water can have a significant impact on species which rely
on high quality surface waters.
C. EPA's Pesticides and Ground Water Strategy
EPA responded to the above concerns by taking action to prevent further
contamination of ground water from pesticides. On October 31, 1991, EPA released it's
Pesticides and Ground Water Strategy. The goal of the Strategy is to prevent contamination
of ground water resources that could cause unreasonable risk to human health and the
environment resulting from the normal, registered use of pesticides, by taking appropriate
actions where such risks may occur. This goal of the Pesticides Strategy is consistent with
EPA's overall Ground Water Policy, which is articulated in EPA's May 1991 report entitled,
"Protecting the Nation's Ground Water: EPA's Strategy for the 1990s". EPA's Ground
Water Policy is to prevent adverse effects on human health and the environment and to
protect the environmental integrity of the nation's ground water resources. In determining
appropriate prevention and protection strategies, EPA will consider the use, value and
vulnerability of the resource, as well as it's social and economic values. Priority for
protection is placed on currently used and reasonably expected sources of drinking water and
ground water that are closely hydrologically connected to surface waters.
To define adverse effects, the agency's Ground Water Policy uses MCLs, or health'
advisory levels where MCLs are not available, as "reference points" when the ground water
is a potential source of drinking water. Water Quality Standards under the Clean Water Act
are used as reference points where ground water is closely connected hydrologically to
-------
surface water. Reaching these reference points is considered a failure of prevention.
Due to the mandate of the Federal Insecticide, Fungicide and Rodenticide Act
(FIFRA), which was first passed in 1947, EPA must balance the risks versus the benefits of
all pesticide regulatory decisions. Therefore the agency's Pesticides and Ground Water
Strategy incorporates this risk/benefit analysis in deciding whether to restrict or prohibit the
use of a pesticide. FIFRA mandates that the determination of "unreasonable risk to man or
the environment" include consideration of economic and social benefits. The agency believes
that the risk-benefit balancing mandates of FIFRA are compatible with the agency's overall
Ground Water Protection Policy.
EPA is pursuing the following approach to carry out the Pesticides and Ground Water
Strategy:
1. Determine a pesticide's potential to leach into ground water.
2. Determine whether national label restrictions, such as classifying the pesticide as a
restricted use product, and additional required training for users will adequately
address leaching concerns.
3. Determine whether providing States with the opportunity to develop a Management
Plan to protect the State's ground water resources from pesticides will effectively
address the contamination risk.
If a determination is made that a pesticide poses a threat to ground water, EPA will
initiate the above steps to prevent unreasonable risk to human health and the environment. If
passage of national restrictions alone is deemed inadequate to address the concern, the State
Management Plan approach will be pursued.
Under the State Management Plan (SMP) approach, legal sale and use of the selected
pesticides would be confined to States with an EPA approved SMP. EPA will be issuing
proposed rules to classify some pesticides for the SMP approach as part of the process of re-
registering existing pesticide products. Once the rules are final, States have the option of
responding to EPA's decision by developing a SMP if they wish to continue the use of the
product(s) in then- jurisdiction. Some States may opt not to develop a SMP if, for example,
the product is of little economic importance to their State. In some rare cases, there may be
pesticides which pose such significant risks that EPA would have to resort to national
cancellation, and the SMP approach would not be followed.
The Pesticides and Ground Water Strategy provides States with an opportunity to
tailor management of their ground water resources to local conditions of pesticide usage and
ground water vulnerability. The alternative would be national decision making based on a
crude assessment of national risk, resulting in over-protection hi some areas and under-
protection in other areas. There are several areas, however, where EPA is pursuing national
-------
restrictions, in addition to SMPs to better protect ground water. For example EPA is
developing regulations which will govern the storage, disposal and transport of pesticides and
their containers. These proposed regulations, as well as other national, regional, State and
local initiatives to protect ground water from pesticides are discussed in Chapter 6.
EPA can only require SMPs through a chemical-specific regulatory action. However,
the agency is encouraging States to voluntarily take the initiative to develop generic SMPs
which would form the basis of the State's chemical specific SMP. Funding is being provided
under FIFRA and the Clean Water Act to State agricultural and environmental agencies to
get started in the development of generic plans. Detailed procedures for the development,
approval and subsequent oversight of both generic and chemical-specific SMPs have been
developed by EPA, and are found in the following guidance documents:
1. "Guidance for Pesticides and Ground Water State Management Plans" (EPA 735-D-
93-005A, December 1993) discusses the appropriate components of both generic and
chemical specific State Pesticide Management Plans.
2. Appendices to the Guidance, which cover more detailed guidelines for technical
aspects of SMPs such as the design of monitoring programs, response programs,
procedures for approval of SMPs, and the process for evaluating the effectiveness of
SMPs. These appendices include Appendix A: "Review, Approval and Evaluation of
State Management Plans" (EPA 735-B-93-005D, February 1994) and Appendix B:
"Assessment, Prevention, Monitoring and Response Components of State Management
Plans" (EPA 735-B-93-005C, February 1994).
3. "A Review of Methods for Assessing Aquifer Sensitivity and Ground Water
Vulnerability to Pesticide Contamination" (EPA 813-R-93-002, September 1993) was
prepared by EPA with the assistance of USDA, USGS and various State agencies. It
is a Technical Assistance Document which describes methods for evaluating the
potential for ground water to become contaminated by pesticides.
The first document, "Guidance for Pesticides and Ground Water State Management
Plans", identifies the following 12 essential components of a SMP:
1. State philosophy and goals for ground water protection
2. Roles and responsibilities of State agencies
3. Legal authority
4. Resources
5. Basis for assessment and planning
6. Monitoring
7. Prevention actions
8. Response to detections of pesticides
9. Enforcement mechanisms
10. Public awareness and participation
-------
11. Information dissemination
12. Records and reporting
These 12 elements should be included in a generic SMP, with more detailed coverage
in a chemical specific SMP. The extent to which each element must be addressed will
depend on each State's unique hydrogeologic and pesticide use characteristics.
D. Purpose of the Jefferson County Pilot Project
In order to provide further guidance and technical assistance to States which are
beginning to develop SMPs, EPA Region ffl embarked upon a pilot project in Jefferson
County, West Virginia in 1989. The purpose of the project was to address specific concerns
raised by States regarding development of their Pesticides in Ground Water SMPs. Concerns
included how to: obtain good pesticide usage data, conduct ground water vulnerability
assessments, and target potential problem areas for ground water monitoring. Since EPA
grant money for SMP development is limited, and insufficient resources was an additional
concern expressed by the States, Region ffl decided it was critical to provide technical
assistance. The Region's technical assistance focuses on the areas the States identified as
concerns, most of which fall under elements 5, 6, 7, and 12 of a SMP, as noted above. This
report supplements EPA Headquarter's Technical Support and Assistance Documents for the
development of SMPs.
The Jefferson County pilot project demonstrates how to identify areas where
pesticides pose a potential threat to ground water quality within the County. Data on soils,
hydrogeology, and pesticide use was obtained and analyzed to conduct the vulnerability
assessment. An investigation into sources and methods for obtaining pesticide use data was a
critical component of the pilot study, in keeping with the State's request for assistance in this
area. State officials are now targeting ground water monitoring efforts in areas identified as
potentially vulnerable, and EPA Region HI, State and local agencies have developed
suggestions for protecting these areas through management techniques, education and
financial assistance. These suggestions are included in Chapter 6 of this report.
E. General Description of Project Area
Jefferson County, an area of approximately 212 square miles, is situated in the
Eastern Panhandle of West Virginia (map number 1 shows location). Land use in the County
is predominantly agricultural (approximately 54%), however development pressures and
population (approximately 36,000 people at the time of the study), have been rapidly
increasing over the past 30 years due to the close proximity of the Baltimore and Washington
metropolitan areas. The County includes scenic and historic Harpers Ferry, which i.s located
approximately 50 miles northwest of Washington, D.C. Most of the County is hi the
Shenandoah Valley of the Valley and Ridge physiographic province. The Blue Ridge
Mountains, found along the southeastern border, comprise approximately one fifth of the
County.
6
-------
Ground Water Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, West Virginia
Region III
LOCATION OF STUDY AREA
flap: JC001
July. 1992
-------
F. Rationale for Selection
Jefferson County, West Virginia was selected as the first site to conduct a pilot
project because this area met the following criteria:
1. Vulnerable ground water: approximately 86% of the County is underlain by folded
and faulted carbonate rocks.
2. Population largely dependant on ground water for drinking water: 82% of the
drinking water comes from wells, and over half of the ground water used for drinking
is for public water supplies.
3. Predominantly agricultural: over 50% of the County is in cropland.
4. Documented detections of pesticides in ground water, including the organochlorines
DDE, dieldrin, endrin and heptachlor. In addition, cyanazine, simazine, parathion,
prometone, atrazine, methyl parathion, DDT, DDE, DDD and endirin have been
detected in springs and surface water hi the County.
5. Several key data layers were automated by the Soil Conservation Service, and made
available to use for conducting analyses with Geographic Information System
technology.
6. Local, State and Federal agencies in the area were interested in working with EPA
and using results of the project.
7. Hydrogeologic conditions had been characterized by USGS, including ground water
levels, flow directions, recharge areas and ground water quality. In addition, the
geology was recently mapped by the West Virginia Geologic and Economic Survey.
8
-------
CHAPTER 2: GEOGRAPHIC INFORMATION SYSTEMS
A. DEFINITION AND CONCEPTUAL DISCUSSION
B. APPLICATION OF CIS TO THE DEVELOPMENT OF STATE PESTICIDE AND
GROUND WATER MANAGEMENT PLANS
A. Definition and Conceptual Discussion
A Geographic Information System (GIS) is an information technology which captures,
stores, checks, manipulates, analyzes and displays data which is spatially referenced to the
earth's surface (16). The main function of an information system is to improve one's ability
to make decisions. An information system is a chain of operations that proceeds from
planning the observation and collection of data, to storage and analysis of the data, to use of
the data to solve problems (2). A map is a type of information system, in that it is a
collection of stored and analyzed data, which is used to make decisions. A geographic
information system can be either manual, such as a map, or automated (based on a digital
computer). The use of an automated GIS will be the focus of this Chapter.
A GIS stores and manipulates both spatial and nonspatial data. Spatial data is
represented as points, lines or areas (polygons) that are geographically referenced to a
coordinate system, such as latitude and longitude. Using coordinates that define its location,
data is stored in the computer. Attribute information associated with spatial features is also
stored in the GIS data base and used as part of the decision making process. An example
would be data regarding the quality of water (attribute) associated with a lake (polygon).
The software used by EPA Region HI for this project was ARC/INFO. Locational
information related to map features is stored and manipulated in a topological structure
known as ARC, and non-spatial, non-geographic attributes are stored hi the relational data
base known as INFO. The Soil Conservation Service, which provided much of the
automated data used for this project, used GRASS software in an AT&T 6386E hardware
environment. The primary difference between the two types of GIS software is that
ARC/INFO is vector based, while GRASS is raster based.
A raster data structure is formed by partitioning a study area into a set of grid cells
that are usually square. Each cell is assigned a code describing the feature contained within
the cell, such as elevation or highway name. The size of the cell is selected based on the
resolution needed or the resolution available from the source data, such as remote sensing
satellite pixel data. Explicit x and y coordinates are not given to each cell because the cell
location is implicit in it's row and column location on the grid (8). For example, the origin
of the raster cell is frequently the upper left corner of the grid, and all the cell locations are
relative to this origin (18).
Vector data structures use a series of x and y coordinates, such as latitude and
longitude to describe the point, line and polygon features. In addition, information about
-------
connections and relationships among the features portrayed on the map, such as which
polygons share a boundary, etc., is calculated and stored with the coordinates. The
relationship among map features is called topology. Because of the smooth lines associated
with the vector format, information such as the length of a shoreline can be calculated, which
would be difficult with a raster format. The raster format however, allows for easier
programming and analytical operations that require calculations based on quantifiable cell
attributes (8).
Without the benefit of an automated GIS, environmental planners have to struggle
with relating key information from different data layers, or maps, of different scales, map
projections, coordinate systems, and different levels of distortion. For example, a planner
with the Soil Conservation Service routinely uses a topographic map, a soils map and an
aerial photograph to plan management practices, such as strip crops, waterways and terraces
on a farm. All three data layers have critical information needed as part of the decision
making process. If the planner wants to identify the slope of the field, the topographic map
is needed, the soil map is needed to determine credibility of soils and the aerial photograph
is needed to identify vegetation and farm boundaries. If all three data layers were automated
for use in a GIS, the computer could analyze all the data simultaneously, and respond to
queries, such as, "shade all the areas on the farms in the County where all of the following
conditions are met: slopes greater than 5%, erodible soils, existence of intermittent
drainageways and fields where row crops are grown". If all four criteria have to be met in
one site in order for a fanner to qualify for financial assistance for grassed waterways, the
planner would then have a map showing all the areas which qualify. In addition, a GIS can
plot the map at any scale the planner desires.
Once data layers are automated, GIS can be used as a very powerful decision making
tool. Instead of juggling between different scale maps and the literature that describes the
map features to slowly obtain answers about individual areas, one can use GIS to efficiently
and accurately respond to multiple queries over large geographic areas in seconds. The key
ingredient to success with GIS is careful attention to collecting the right kind of data for the
uses envisioned, preprocessing the raw data in a manner that ensures a high degree of
locational accuracy, and obtaining accurate attribute data. Put simply, poor quality data
yields incorrect analyses.
The five essential elements of a GIS are: data acquisition, preprocessing, data
management, manipulation and analysis, and product generation (18).
Data acquisition involves identifying and gathering the data necessary to meet the
objectives of the application. Acquiring data is one of the greatest operational problems and
costs in the field. Data are acquired from maps, aerial photography, site visits, interviews,
existing files and documents. There is also a great deal of spatial digital data available in the
public domain from Federal agencies such as USGS, USDA, NASA-NOAA, and FJ>A.
10
-------
After the data set has been collected it must be preprocessed for conversion into a
form suitable for use in a GIS database. Data entered into a GIS usually needs processing
and manipulation to make it conform to a data type, a georeferencing system and a data
structure compatible with the system. The following are eight essential preprocessing
procedures (18):
1. Data structure conversion from-raster to vector format, and conversion of non-
computer compatible data, such as maps, into a digital form. The latter process,
called digitizing, can be very expensive and time consuming. It involves using a
digitizing tablet and tracing features of interest on a map, which allows the map
features to be converted to a computer compatible digital signal. Once completed,
geographic coordinates of all the features on the map are stored in the GIS.
2. Data reduction and generalization involves changing from a data set that records, for
instance, all crop types in an area, such as com, apples, and alfalfa, to a listing of
areas ^here only three groupings of crop types are found, such as row crops,
orchards and hay land. This type of aggregation of data is called generalization. Data
reduction might involve simplifying a complex coastline by reducing the density of
points that describe it to achieve an approximation suitable for the study.
3. Error detection and editing includes correcting problems such as polygon errors,
slivers, unjoined unions and gaps.
4. Merging points into lines, and points and lines into polygons is the process of
building a more complex topology from simpler elements.
5. Edge matching is matching and joining separate portions of an area of interest which
fall on different map sheets, to form a seamless data set. Edge matching of adjacent
maps is necessary where errors exist from improper georeferencing before digitizing
or from digitizing off of paper which experienced expansion or shrinkage.
6. Rectification and registration: rectification is the transformation of non-georeferenced
and non-dimensioned spatial data into a georeferenced, dimensional map. Registration
is the adjustment of spatial data from one map projection and coordinate system to
another. There are also approximate approaches to rectification, such as rubber
sheeting, that must be attempted when the data does not conform to any standard
projection or georeferencing system.
7. Interpolation: geographic data, such as elevation control points, demographic
surveys, vegetation biomass estimates, etc., cannot possibly be taken at all desired
places, therefore, sampling techniques, which employ statistical principles, are used to
maximize the amount of information gained for the least expended effort. To estimate
values for locations (points) that have no measurements, interpolation is done by
applying the normal principle which says that objects that are near are more important
11
-------
than objects that are far away. If one has a set of measured elevations in an area, a
statistical model can be used to interpolate values of elevation at places where there
are no measurements. In any interpolation the basic decision is to choose a model for
the statistical relationships between points. For example, a common model for such
relationships is a weighting function. Weighting functions specify what each
neighboring point's value will contribute to the determination of an unknown data
point (9).
8. Photo-interpretation is the extraction of useful information from photographs or other
images using an iterative process by which evidence is collected, hypotheses are
tested and interpretations are refined to reach conclusions within an acceptable error
range (20).
Preprocessing of data for use in a GIS can be very time consuming and costly, since
it involves extracting large volumes of information from maps, photographs and records and
then recording this information hi a computer data base. It is critical to establish a consistent
system, adhere to strict quality control procedures and document the processes used to
automate the data.
Data management is the third essential element of a GIS. It involves managing the
way data is entered, stored, updated, deleted and retrieved from the system. It allows
different users to have different kinds of access to the database.
Manipulation and analysis of the data stored in a GIS to solve problems and make
decisions is the fourth essential element of a GIS, and is usually the major focus of attention.
A description of the types of GIS analyses conducted for the Jefferson County project, such
as intersection, proximity and Boolean operation analyses can be found in Chapter 5.
The final phase of using GIS to solve problems is the generation of a final product.
Outputs from GIS analyses may be statistical reports, maps or graphics (such as bar charts).
Often outputs are tapes or disks for transmission to other systems. The output of one
particular analysis can always be placed back into the GIS for future analysis, such as to
evaluate trends over time. Once data has been automated for use in the system, it provides a
powerful foundation upon which numerous additional analyses can be performed. The
products generated are snapshots in time; the ongoing analysis is a dynamic process, which
is dependant upon keeping the system updated with the most recent information.
B. Application of GIS to the development of State Pesticide and Ground Water
Management Plans
As described in Chapter 1, EPA's Pesticides and Ground Water Strategy provides
States with an opportunity to tailor the management of their ground water resources to local
conditions of pesticide usage and ground water vulnerability, through the development of a
State Management Plan (SMP). To develop these plans, States must first identify where and
12
-------
to what degree ground water is susceptible to contamination from pesticides. The next step
is to devise management strategies for protecting the vulnerable areas.
The Jefferson County pilot project demonstrates how to use GIS to identify areas
where pesticides pose a potential threat to ground water quality. Data on soils,
hydrogeology, drinking water wells and pesticide use was obtained and processed for use hi a
GIS. Because GIS can analyze multiple data layers simultaneously and respond to complex
queries, it has proven to be an extremely valuable tool for this type of application.
Several different types of analyses were conducted to identify vulnerable areas. For
example, the overlay analysis technique was used frequently. To use this method, the
computer is programmed to identify all locations where specific conditions are met. By
using a polygon intersection algorithm, for example, all areas of the County where pesticides
labeled as "priority leachers" are used over soils which rank high for leachability (using a
USD A screening procedure) are displayed hi red. This analysis shows where pesticides are
likely to migrate past the root zone, which is the first step in identifying areas of potential
ground water vulnerability. Details of analyses conducted are provided hi Chapter 5 with
accompanying maps. A technical description of the mathematical, logical and cartographic
procedures used to automate the source data, conduct the analyses and create the maps can be
found hi Appendix B. All analyses were conducted using ARC/INFO on a Data General
AViiON workstation.
13
-------
CHAPTER 3: BUILDING THE JEFFERSON COUNTY DATA BASE
A. ORGANIZATION OF THE WORKGROUP
B. DETERMINATION OF OBJECTIVES, IDENTIFICATION OF DATA NEEDS
C. DESCRIPTION OF DATA LAYERS, INCLUDING SOURCES AND METHODS
USED TO OBTAIN AND AUTOMATE DATA
*SECTION I: BASE MAP, SOILS, GEOLOGY, HYDROGEOLOGY
*SECTION H: DRASTIC, SPISP
*SECTION ID: LAND USE
A. Organization of the Workgroup
At the beginning of the project a workgroup was established with representation from
the following ten different agencies:
1. WV Department of Agriculture
2. USD A Soil Conservation Service
3. USDA Cooperative Extension Service
4. WV Division of Environmental Protection
5. WV Department of Health
6. U.S. Geological Survey
7. WV Geological and Economic Survey
8. Eastern Panhandle Regional Planning and Development Council (Region 9)
9. WV University
10. EPA Region III
The workgroup turned out to be a showcase in cooperation between government
agencies. The purpose of the workgroup was to identify the following: types of questions
that GIS could be used to answer to achieve project objectives, what data was necessary in
order to answer the questions, who had the necessary data, what form the data was in and
what protocol should be followed to share data among agencies.
B. Determination of Objectives, Identification of Data Needs
The first step was to obtain consensus on the primary objective of the project, namely
to identify the relative susceptibility of ground water in different areas to contamination from
pesticides. The second step was to ensure that all workgroup members had a good
understanding of the factors which influence fate and transport of pesticides in the
environment, in order to determine what data was needed to identify susceptible ground
water areas.
Factors which influence the probability that a pesticide will contaminate groundwater
during normal agricultural use fall into the following categories: 1) soil properties,
2) pesticide chemical characteristics, 3) climatic and management variables, and
14
-------
4) hydrogeologic characteristics. The following is a discussion of each of these categories.
1. Soil properties: texture, porosity, organic matter content, pH, soil structure,
thickness of soil horizons, permeability and moisture holding capacity all influence the
fate and transport of pesticides in the subsurface environment. A coarse textured soil
like a sandy loam, for example, with a low percent organic matter offers a higher
potential for pesticide leaching than a fine textured clay loam with a high percent of
organic matter. Adsorption of pesticides onto organic matter and clays is a very
important factor affecting pesticide mobility in soils. Many pesticides tend to bind
with organic carbon found hi soils, and are then tied up in an organic complex,
unavailable for leaching. Clay soils, however, can have macropores and fissures,
which enhance the potential for the preferential flow of water and dissolved
pesticides. A soil environment conducive to microbial degradation is another critical
factor which affects the persistence and availability of pesticides to leach. In addition,
slopes of fields and soil credibility influence infiltration is more likely than runoff.
2. Chemical properties of pesticides that affect the likelihood of leaching include:
solubility, soil adsorption, soil persistence and vapor pressure. Soil adsorption is
generally correlated to the organic matter content of soils, and expressed as the
organic carbon partitioning coefficient, or Koc, which measures the amount of
pesticide that can be adsorbed by organic carbon. Soil persistence is generally
expressed as the pesticide's half life hi a soil environment, which is the amount of
time required for a pesticide to degrade to one half the original mass.
3. Climatic and management variables: the variability at a given site due to weather and
crop management techniques influences the degree of pesticide migration to
groundwater. The amount and temporal variation of soil moisture due to
precipitation, irrigation, evapotranspiration, runoff and recharge also affects leaching.
The larger the volume of water available to percolate thorough the soil, the farther
distance a water soluble pesticide can be transported. Soil moisture and temperature
also affect the persistence or half life of a pesticide hi soil. The types of vegetation
which are grown, tillage practices which are used (no-till versus moldboard plowing,
for example), pesticide application rates, and methods of pesticide application (soil
incorporation versus backpack sprayer, for example) all influence the likelihood of
leaching.
4. Hydrogeologic factors: If a pesticide is mobile and persistent enough to migrate past
the root zone, important considerations which influence the potential of a pesticide to
contaminate groundwater include: depth to ground water, hydraulic conductivity of
the unsaturated zone, depth to bedrock, type of aquifer media (limestone, shale, etc.)
and ground water recharge rates. The presence and degree of fractures, faults,
sinkholes, or other avenues for preferential flow of solutes to the aquifer are also
important considerations.
15
-------
Although it is not an exhaustive list of all factors or processes which influence the
fate and transport of pesticides in the soil environment, the above list does provide many key
factors to consider when identifying data needs for determining susceptibility of ground water
to pesticide contamination. The Jefferson County workgroup agreed on the need to obtain
data on geology, hydrogeology, soils, drinking water wells, pesticide usage and general land
use, with a focus on collecting as many of the parameters described above as possible in each
of these categories. Where available, data from other agencies was shared with EPA, and
where necessary, EPA collected raw data in the field. The following is a description of each
of the data coverages obtained for the County, with accompanying maps. Chapter 5
describes the GIS analyses conducted with this data.
C. Description of Data Layers, Including Sources and Methods Used to Obtain and
Automate Data
Section I
This section discusses single theme data coverages obtained for the County, such as
roads, streams, geology, soils, etc.
1. Planimetric and Hvdrologic Features: Map number 2 is the base map which was
obtained from the U.S. Geological Survey. The scale of the source map was
1:100,000. The map shows the County is bounded on the northwest by the Opequon
Creek, and on the northeast by the Potomac River. The County shares the northeast
border with Maryland, and is bordered on the southeast and southwest by Virginia.
Harpers Ferry is located at the confluence of the Potomac and Shenandoah Rivers.
The town of Ranson, roughly in the center of the County, is the location of the local
USDA Soil Conservation Service office, which houses the agency's GIS workstation
for the County. The County includes parts of ten 7.5 minute USGS quadranglesr as
shown by map number 3.
2. Geology: Map number 4 shows the geologic formations of the County. The geology
was mapped by the WV Geological and Economic Survey. The Survey published the
map in 1991 and EPA Region HI digitized the paper map. The scale of the source
map was 1:24,000, and was 45" by 61" in size. The map was developed through
several years of field work augmented by aerial photography and radar imagery.
Most of Jefferson County is in the Shenandoah Valley of the Valley and Ridge
physiographic province. The Blue Ridge Mountains are found along the southeastern
border, covering approximately one fifth of the County. The valley, which is °
bounded by the Shenandoah River to the east and the Opequon Creek to the west,
covers about 86% of the County (15). The valley is smoothly rolling and underlain
predominantly by carbonate aquifers, composed of highly fractured, faulted and
folded limestone and dolomite. The carbonate valley of the County is composed of
the following formations: the Tomstown Dolomite; the limestones and dolomites of
16
-------
Ground Water Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, West Virginia
j~] Roads
Railroads
p~| H j
d r o 1 o
PLANIMETRIC & HYDROLOGIC FEATURES
Map: JCQ2
Data Source: U.S.G.S.
July, 1992
-------
Ground Water Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, West Virginia
Region III
U.S.G.S. 7.5' QUADRANGLES
Mao: JC03
Data Source: U.S.G.S
July,
-------
Ground Kater Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, tfest Virginia
Ordovician System
g] Mirlinibnrj Formation
Chambenkurj limeitone
Ke» Market Limestone
Pintsfcorj StiiicD Dolomite
Kackdtle Run formjtioc
Stoaehenje Limestone
Sloafferstota Member
Cambrian System
| Conocoche»5ue Formition
g^'j Elbrook PormitioD
| | Iijneiboro Form«lioQ
^| Tcmitoin Dolomite
g Antietim Formation
t- j Barpers Formation
fe»ertoD Formation
Precambrian System
! Catoclin Formation
Region 1 1 I
GENERALIZED BEDROCK GEOLOGY
Hap: JCOf
Data Source : IVGiES
July, 1992
-------
the Waynesboro, Elbrook, and Conococheague Formations; the Beekmantown group
(Stoufferstown Member, Stonehenge Limestone, Rockdale Run Formation, and
Pinesburg Station Dolomite), the New Market Limestone and the Chambersburg
Limestone. On the southeastern border of the County, the Blue Ridge Mountains are
underlain by metamorphosed shales, sandstone and quartzite, found in the Harpers,
Antietam and Weverton Formations. The Martinsburg Shale on the western border is
the only other noncarbonate area of the County.
3. Geology. Faults. Fractures and Sinkholes: Fracture traces and faults were mapped
from aerial photographs, satellite imagery and available geologic maps, by USGS and
the State Geologic and Economic Survey. EPA digitized the fractures and faults from
a 1:50,000 scale map developed by USGS. Sinkholes and cavernous zones were
mapped onto topographic maps by a USGS ground reconnaissance survey, and the
locations were digitized by SCS.
As can be seen from map number 5, the faults, fractures and sinkholes concentrate in
the carbonate valley, from the Tomstown Dolomite to the western border. The
Chambersburg Limestone, the Beekmantown Group and the Conococheague
Formation have the greatest density of mapped sinkholes per square mile of outcrop.
The density ranges from 3.5 sinkholes per square mile in the Chambersburg
Limestone to 5.0 sinkholes per square mile in the Beekmantown Group (15).
Limestone, and to a lesser extent dolomite, are both soluble in rainwater that is
slightly acidic. Rainwater percolates into and through these carbonate rocks,
dissolving rock material and enlarging fractures in the rocks as it moves. The
dissolution of calcite or dolomite along fractures, faults and bedding planes creates
solution enlarged channels that allow ground water to flow at rapid velocities. In
addition, sinkholes, which are direct openings on the land surface, are also formed by
this dissolution process. Rainwater either runs directly off into sinkholes or
percolates through the soil and then flows downgradient as ground water through a
complex interconnection of the solution enlarged channels. Ground water then
discharges to springs and streams, in some cases forming huge cavernous zones (up to
1500 feet in diameter) in these discharge areas, as can be found in the northeastern
border of the County (15). As much as 85% of stream flow is estimated to be
derived from ground water discharge where the underlying aquifers are limestone
(12). The carbonate valley in Jefferson County contains a complex network of
sinkholes, enlarged fractures, caves, springs and disappearing streams. Areas with
these characteristics are commonly referred to as karst settings.
Permeabilities of the geologic formations in the valley generally decrease from west
to east, with the Chambersburg Limestone being the most permeable (far west), the
Beekmantown group the second most permeable and the Conococheague the least
permeable (11). The permeability of these formations is primarily a function of
secondary porosity from solution enlarged channels, rather than primary porosity
through unconsolidated materials.
20
-------
Ground Water Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, West Virginia
Ordovician System
Martinsburg Formation
Chambersburj Limestone
New Market Limestone
Pinesburg Station Dolomite
|g Rockdale Run Formation
m Stonehenge Limestone
j \ Stoufferstown Member
Cambrian System
m Conoeocheague Formation
! El brook Formation
| [ lajnesboro Formation
^H Tomstown Dolomite
Antietam Formation
Harpers Formation
leverton Formation
Precambrian System
^H Catoctin Formation
Faults
[7\71 Fractures
[7] Sinkholes
Region III
GEOLOGY, FAULTS, FRACTURES k SINKHOLES
Map: JCQQ5
Data Source: ffVG&ES k USGS
July, 1992
-------
The noncarbonate areas under the Blue Ridge Mountains to the east, and in the
Martinsburg Shale to the west, are much more resistant to weathering, resulting in
less percolation of rainwater and lower yields in the aquifers (15). In general, ground
water is most vulnerable to contamination in the carbonate valley, due to the
dissolution process described above. Contaminants can be easily transported with
rainwater into sinkholes and through the underground channels at high velocities, and
can affect a large part of the aquifer hi a short period of time.
3. Water Table Contours: Map number 6 is a map of ground water elevations. "Ground
water levels were measured by USGS in 196 wells during a two week period of low
flow conditions in September 1974. USGS plotted water levels on a 1:50,000 scale
topographic map and drew contours by interpolating between measured levels, using
the topographic map as a guide. Contours were digitized by EPA from mylars used
by USGS to create the original prints of the 1:50,000 scale map.
Ground water levels fluctuate in response to recharge to, of discharge from, the
aquifers. The range hi annual water level fluctuations is generally greater hi the
recharge areas than the discharge areas (15). Water levels were measured at the same
wells again hi the Spring of 1975 during a recharge period. Fluctuations from the
September levels ranged as much as 32 feet (12). Water levels will also fluctuate due
to pumping. Of the 196 wells which had water level measurements taken, there were
18 wells which had been pumped recently. Water levels in another 15 wells were
based on levels reported at the date of well drilling, since the well caps were not
accessible.
Geologic and topographic setting influence the depth to water. In the gently rolling
carbonate valley, land surface elevations range from approximately 400 to 600 feet
above sea level. In the Blue Ridge Mountains on the southeastern edge of the
County, elevations range from approximately 1,100 to 1,700 feet above sea level (1).
The depth to water ranges from 5 feet to 295 feet below land surface hi the County.
In the carbonate valley, the average depth to water from hillsides and hilltops is over
twice the average depth to water below flat areas or depressions (15). In the Blue
Ridge Mountains, the average depth to water is less than it is hi the lower-relief
hillsides and hilltops of the carbonate valley. This is due to lower hydraulic
conductivity of the rocks in the mountains, which results hi slower water movement
and allows water levels to remain at higher levels than hi the carbonate rocks (13).
Under conditions of diffuse flow hi a water table aquifer, a water table map can be
very useful for determining ground water flow directions, since flow is assumed to be
perpendicular to equipotential, or contour, lines. In a karst setting such as Jefferson
County, however, where a combination of diffuse and conduit flow exists, it is
extremely difficult to predict ground water flow directions. Dye tests have been done
by the USGS which show a general ground water divide around the middle of the
County separating regional flow either east to the Shenandoah or west to the Opequon
22
-------
Ground Water Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, West Virginia
Indei Contours X7 'W/>7 / <^?%>
\^*/// / ^
Intermediate Contours
fell Locations
Contours Indicate Height of later Table Above Sea Level Region III
IATER TABLE CONTOURS & IELLS
Map: JC06
Data Source : USGS
July. 1992
-------
Creek. Permeability increases where there is a higher density of fractures and faults.
Preferential flow has been documented along bedding planes parallel to strike (parallel
to the water table contours). Although dye tests also showed ground water moved
perpendicular to water table contours (normal to the strike of the rocks), the median
hydraulic conductivity was found to be approximately five times greater parallel to
strike than perpendicular to strike. The primary reason for this phenomenon is a
greater degree of dissolution which occurs along bedding planes and faults parallel to
strike, resulting in solution cavities that transmit water at higher velocities in this
direction (15).
Because of the complexities of the ground water flow system, the focus of this pilot
project's susceptibility analysis is on identifying where pesticides are most likely to
leach to ground water, versus what direction or speed the pesticides would ultimately
move in. The primary goal, then, is one of pollution prevention: to identify areas
where it is most critical to prevent pesticide leaching from occurring hi the first place.
To prevent pesticides from reaching ground water, it is important to know the
characteristics of the unsaturated zone, such as texture, porosity and permeability. It
is also critical to know the thickness of the unsaturated zone, or distance to ground
water. The longer the distance, the more time and material that is available for
attenuation and ultimately degradation of contaminants hi the unsaturated zone. To
determine the distance to ground water, water table elevations are evaluated in
relation to the land surface. The water table contour map included a table which
provided measurements of water levels in wells, given as the distance measured below
land surface. Water level measurements in these wells formed the basis upon which
the water table contours were created. However, since the water table contours only
provide elevation above mean sea level, a digital topographic coverage would also be
needed in order for GIS to calculate zones of varying depths to water. Since a digital
topographic coverage was not available at an appropriate scale, the next alternative
was to use the measurements of depth of water below land surface, taken by USGS at
individual well points, and use GIS to interpolate between these points. The result is
portrayed hi Chapter 5, map number 35 which shows polygons of different ranges of
depth to water. This map of the depth of the unsaturated zone was developed using
GRID, an ARC/INFO module, to interpolate between points where well data exists.
From this map it is evident that the shallowest depths to ground water are found hi
the carbonate valley.
4. Soils: Soils were digitized by the Soil Conservation Service (SCS), based upon
original 1:20,000 scale County Soil Survey maps. SCS transferred the line work
from the Soil Survey maps to stable base mylars of USGS topographic maps.
Original soil survey maps were ratioed to a scale of 1:24,000 to match the 7.5 minute
1:24,000 scale topographic maps, to ensure a more accurate transfer of all soil lines.
To facilitate the recompilation of linework, the topographic map mylars were ordered
with vegetation removed, and hi areas of high relief the elevation lines were thinned.
24
-------
After the boundaries of the soil types were transferred to the mylars, the lines were
then digitized, using the line segment data capture program in GRASS GIS software.
All soil compilation work was done in the Ranson, WV SCS office using an AT&T
6386E workstation. The soil data layer is a very costly and tune consuming coverage
to automate manually (3).
Although automating detailed soil survey maps is a tremendous effort, the data is
critically important to many environmental planning decisions. The detailed soil
coverage was extremely valuable for the applications described in this document.
Since soils form the first barrier to pesticide leaching, a majority of the GIS analyses
in Chapter 5 focus on the soils layer to identify areas of high potential for pesticide
leaching below the root zone.
There are three levels of soil geographic data bases. The County Soil Survey is the
SSURGO level, the State aggregated coverage is the STATSGO level and the national
aggregated coverage is the NATSGO level. As of June 1991 less than 50 Counties
nationwide had SSURGO level soil series maps digitized. The SSURGO data base is
complete for Jefferson County.
There are 74 mapping units hi the Jefferson County Soil Survey. Five of these units
are land types, rather than soils, including the following: Alluvial land, Alluvial land
with marl substratum, Marl pit, Quarries and Steep rock land. The rest of the map
units belong to 21 different soil series. Each soil series has a distinctly different soil
profile, as described by the texture, structure, and color of the soils, as well as the
thickness of each horizon and number of horizons hi the profile. A generalized soil
survey map of the County aggregates the soil series into nine soil associations. A soil
association is a landscape that has a distinctive proportional pattern of soils. It is
named for the major soil series found within the association. It may have more than
one major soil series, and at least one minor soil series that occurs hi a smaller
percent of the area. The soil association map can be used for general planning
purposes, but is inadequate for planning at the farm level.
Another approach to aggregating soils was taken with map number 7. This map
shows the major soil types or series grouped according to texture, from clays to sandy
loams. The map color scheme displays areas in violet/gray where the finest textured
soils (clays) are found, then areas where medium textured loams are located are
displayed in orange, and more coarsely textured soils (sandy loams) are displayed in
red. Land types are displayed hi various colors, such as alluvial land (pale orange,
found hi the valley) and steep rock land (dark gray/brown, found in the Blue Ridge
Mountains). The aggregation of soils was done by EPA based upon the SCS Soil
Survey for Jefferson County. Table 1 describes the groupings and associated colors
displayed by map number 7. This aggregation was chosen since soil texture plays an
important role hi soil permeability.
25
-------
Ground Water Vulnerability to Contamination by Agricultural Chemicals
Jefferson C o u n t v , West Virginia
Major Soi Is
Clay Loams
Silt loame
Lca11"
Sondy Loami
Uarl Pit
CZ3
Steep Rock Lund
Water
Region
MAJOR SOIL TEXTURES
Map: JC07
Data Source: l.S.D.A. S.C.S
Julv, 1993
-------
Table 1
MAJOR SOIL TEXTURES
SOIL OR LAND TYPE AS
CLASSIFIED IN THE JEFFERSON
COUNTY SOIL SURVEY
clay
silty clay
very rocky clay
very rocky silty clay
silty clay loam
cherty silty clay loam
very rocky silty clay loam
silt loam
shaly silt loam
cherty silt loam
very stony silt loam
very rocky silt loam
extremely rocky silt loam
alluvial
loam
gravelly loam
very stony loam
fine sandy loam
channery fine sandy loam
marl pit
quarry
steep rock land
water
GROUPING OF
SOILS FOUND IN
MAP #7
CLAYS
CLAY LOAMS
SILT LOAMS
ALLUVIAL
LOAMS
SANDY LOAMS
MARL PIT
QUARRY
STEEP ROCK
LAND
WATER
COLOR
ASSOCIATED WITH
SOIL GROUPING
VIOLET/GRAY
GREEN
YELLOW
PALE ORANGE
DARK ORANGE
RED
BLACK
LIGHT GRAY
BROWN/GRAY
DARK
BLUE/BLACK
27
-------
There is a close relationship between soil associations and geology. For example, the
Berks-Weikert and Weikert series, which consist of shaly silt loams, are found
throughout the Blue Ridge Mountains and over the Martinsburg Shale. This soil
series is formed in material weathered mainly from shale: Soils hi the carbonate
valley are predominantly medium textured, consisting mostly of clay and silt loams
weathered from limestone, limy shale and dolomite. A representative group of soils
in the carbonate valley includes the Hagerstown-Frederick-Huntington association,
which consists of deep, medium textured soils, formed from weathered limestone,
found in level areas to moderately steep slopes. The soils are fertile, moderately
permeable and have a high available moisture capacity. Because of slopes and rock
outcrops, these soils are mostly used for orchards, dairy farming and raising
livestock. There is only a small percent of coarse textured soils hi the County, which
are found predominantly in alluvial areas along streams and major water courses, such
as the Shenandoah River, and are scattered throughout the slopes and foothills of the
Blue Ridge Mountains. These soils are coarse grained because they weathered from
sandstone and quartzite. Although there are soils overlying the metamorphic rocks of
the Blue Ridge Mountains which are more permeable than the soils over the carbonate
valley, runoff is significant in the mountains, but negligible hi the valley (10).
Section IT
This section describes two techniques used to rank the potential for ground water to
become contaminated by pesticides. Maps of the County which show the results of the
evaluations by the two techniques are also displayed hi this section. The first method,
DRASTIC, uses a parameter weighting and scoring system to evaluate aquifer sensitivity to
contamination. Final DRASTIC scores are based on the intrinsic characteristics of many of
the data coverages described hi Section I, such as soils and geology. The second method is
called SPISP, which is a screening technique that focuses on characteristics of soils and
pesticides used on the soils. SPISP evaluates the likelihood for pesticides to leach below the
root zone. Land use data, such as type, location and quantity of pesticides applied, must be
combined with DRASTIC scores and SPISP soil ratings hi order to evaluate potential threats;
this analysis is done hi Chapter 5.
DRASTIC
1. GENERAL DESCRIPTION
The DRASTIC method (Aller et al., 1985 and 1987) ranks the relative potential of
ground water to become contaminated. To create the map, seven elements which
influence ground water flow are rated and weighted to generate scores. The higher
the score, the more vulnerable the ground water is to contamination. The seven
elements which are evaluated include:
D - Depth to water
28
-------
R - Recharge (net)
A - Aquifer media
S - Soil media
T - Topography
I - Impact of the vadose zone
C - Conductivity (hydraulic)
Different weights (between 1 and 5) assigned to each element of DRASTIC reflect
their relative importance in influencing pollution potential of ground water. For
example, depth to water has a weighting of 5 and topography has a weighting of 1.
An altered form of DRASTIC, called AG DRASTIC has higher weights for some
parameters, such as topography and soil media, due to the importance of these factors
in determining the likelihood of pesticide leaching to ground water.
Different ranges encountered for each DRASTIC factor are assigned ratings between
1 and 10. The higher the rating, the higher the vulnerability. For example, the
rating is 7 for a range of 15 to 30 feet for depth to water in alluvial mountain valleys,
compared to a rating of 5 for a range of depth to water that is 30 to 50 feet on
mountain slopes. These are two different hydrogeologic settings in the non-glaciated
Central Heath region.
DRASTIC utilizes Ground Water "Heath" regions to characterize the areas to be
mapped. The 15 Heath regions in the U.S. are based on similarities in hydrogeologic
features which influence ground water occurrence and availability, and therefore
pollution potential.
Within each Heath region are found characteristic hydrogeologic settings. These
settings are described with narrative and block diagrams in the publication
"DRASTIC: A Standardized System for Evaluating Ground Water Pollution Potential
Using Hydrogeologic Settings" (April 1987, Aller et. al.). The settings have common
hydrogeologic characteristics, and as a consequence, common vulnerabilities to
contamination by introduced pollutants. The ranges and their associated ratings
provided for each setting can be used as a guide where site specific field data is not
available.
After mapping each element, the rating is multiplied by the weighting to obtain a final
score for each element. Scores for each element within a specific mapped area are
added to obtain the total DRASTIC score, or "pollution potential index" for that area.
Cumulative DRASTIC scores are classified into eight ranges, with a corresponding
29
-------
national color code as follows:
* Violet < 79
* Indigo 80-99
* Blue 100-119
* Dark Green 120-139
* Light Green 140-159
* Yellow 160-179
* Orange 180-199
* Red > 200
The higher the number, the greater the pollution potential of the mapped area. Scores
represent relative rankings, not absolutes.
2. HOW THE JEFFERSON COUNTY DRASTIC MAP WAS DEVELOPED
Map number 8 was developed by John Means of the West Virginia Division of
Environmental Protection and was digitized by EPA. Source material used to create
the map came predominantly from the coverages previously mentioned, such as the
geology and soil maps for Jefferson County. The following is a description of the
methods used by WVDEP to create this map, including the degree of accuracy
associated with each mapped and scored parameter of DRASTIC (17).
To create the Jefferson County DRASTIC map, all seven parameters of DRASTIC
were drawn on one base map and one overlay trace map, beginning with the heaviest
weighted parameter that varied the most across the study area: depth to water. Slopes
and soil types were closely associated with depth to water and in most cases fit into
the same zones. Since geologic formations had only four different hydrologic areas,
they were mapped last over the depth-to-water zones. The 1978 County Groundwater
Hydrology map (scale 1:50,000) was used as the base map (12).
Depth to- Water: The Groundwater Hydrology map for Jefferson County gave depth
to water values and locations for approximately 200 wells and provided surface
contours and water table contours. Isobars for depth to water were drawn onto this
map by doing visual interpolation between land surface and water table contours.
Each depth to water category was colored and labelled with the appropriate DRASTIC
score (weight X rating). Since well data was taken in September, a time of low water
table conditions, a correction of five feet was made across the map, based on the 20
year hydrograph for the area (12). Accuracy of this parameter was affected by ten
year old data and since the source of the data was water table contour interpolations
of well measurements, the depth to water zones were interpolations of interpolations.
However, this method is acceptable, since the DRASTIC manual notes that it is
important to lump generalities and not to "split" unnecessarily.
30
-------
Ground ftater Vulnerability to Contamination by Agricultural Chemica
Jefferson County, ttest Virginia
DRASTIC Index
H < 79
80 - 99
100 - 119
120 - 139
140 - 159
HO - 179
180 - 199
> 200
R e g i o. Q 11
DRASTIC Groundwater Vulnerability
Mao: JC08
Dala Source: lest Virginia 1M
July, 1 i)
-------
Net Recharge: Because no direct data was available for this category, it was
estimated. Background information which was used included: (1) DRASTIC manual
general recharge estimates for different hydrogeologic settings, (2) Washington
County, Maryland, Geological Survey data for average annual ground water
discharge, and (3) USGS (Morgantown, West Virginia) data on precipitation and
evaporation. For karst limestone, which covers approximately 75% of the county, the
difference between precipitation and evaporation was multiplied by 85%, which is the
percent of rainwater estimated by USGS to enter the groundwater in Jefferson
County. This gave a value of approximately 12 inches for net recharge. The net
recharge estimate was consistent with the DRASTIC manual estimate and with the
Washington County discharge value. It was assumed that the average yearly value for
ground water discharge would be approximately equal to ground water recharge. For
the Blue Ridge Mountain slopes, the DRASTIC manual estimate of 2 to 4 inches of
ground water recharge was used, since the area is steeply sloped and heavily wooded.
Aquifer media: The 1968 Geologic Map for West Virginia and the Groundwater
Hydrology Map for the Potomac River Basin (13) were used with field observations
to determine locations and compositions of individual formations. The karst area of
Jefferson County extends from the border between the Elbrook formation and the
Waynesboro limestone and dolomite to the western border of the County. Non-karst
dolomite and limestone is found to the east of this border, up to the border of the
Antietam formation. The Antietam, Harpers and Weverton formations contain the
metamorphic rocks of the Blue Ridge Mountains, including quartzite and phyllite.
Martinsburg shale, which is found in a small area along the western border of the
County, falls in the massive shale category.
It was difficult to differentiate between the karst and massive limestone categories.
Because geologic sources showed moderate to high fracturing and solution cavities in
all limestone areas, the lower end of the value range was used for the karst areas and
the higher end for the massive limestone areas in order to avoid great "leaps" in
category values across limestone formation borders. For the metamorphic zone, the
higher end of the range was chosen in order to place it closer to values given for
sedimentary rocks, due to the presence of sandstones and shales in these areas.
Soil Media: The 1973 Jefferson County Soil Survey prepared by the USDA Soil
Conservation Service was used as a source for this parameter. The General Map of
soil associations, at a scale of 1:126,720, was used instead of the detailed soil series
map. The scores for soil media were added to the depth to water zones on the base
map. Weighted parameter values for soil varied only a little, from loam to clay loam,
since the aggregation of the soil associations reduces the variation found in the soils
of the County. In addition, a relatively small weighting of 2 is assigned to the soil
parameter in DRASTIC.
Topography: The 1973 detailed 1:20,000 scale Soil Survey was used to rate
32
-------
topography, since soils are classified partly based on the percent slope of the ground
they are found on. The depth to water zones on the base map were used as a
reference, and as the polygons within which to make the average slope
determinations. Lumping and generalization of the data was done because of the low
weighting (1) assigned to topography.
Impact of the Vadose Zone Media: To evaluate the unsaturated zone above the water
table, geologic maps and the Soil Survey were utilized. Depth to water was
compared to depth of soil profiles. In most areas the vadose zone was found to lay
mostly within the rock formations rather than just within the soil profiles. Most karst
valleys were well drained, and underlying rock affected the vadose zone more than
did the soil profile, so this parameter was handled the same as aquifer media, with the
following exception: because some lower portions of the soil profiles contained silty
clay, lower or more conservative values were chosen for the limestone and
metamorphic categories.
Hydraulic Conductivity: The DRASTIC manual contains a table entitled, "Ranges of
Values of Hydraulic Conductivity and Permeability", which was originally developed
by Freeze and Cherry in 19?9. This table was used to assign values for the different
geologic formations. High values were assigned to the highly fractured karst areas in
the carbonate valley, moderate to low fractured limestone was given a mid-range
value, and massive shale was given a low conductivity value. Estimation from this
table was crude, since the ranges are broad and overlap.
Final Mapping and DRASTIC Index Compilation: On the groundwater hydrology
base map were drawn the depth to water zones, which were labeled with numerical
scores for the depth to water, soil and topography parameters. This was called the
DST map. Then a vellum was placed over the base map and the DST zones were
traced without the numerical values. A chart was constructed with the remaining
values for recharge, aquifer media, impact of the vadose zone and conductivity
parameters. This was called the RAIC chart. These four parameters were keyed in
each case into a particular geologic formation, yielding one RAIC additive value for
each of these four formations: shale, karst limestone, limestone dolomite and
metamorphic formations. With the vellum over the colored DST base map, and the
RAIC chart at hand, final DRASTIC index values were computed and recorded in
each zone on the vellum. The 1:50,000 scale vellum map was reduced by one-half
onto bond to produce a 1:100,000 scale copy.
3. IMPORTANT CONSIDERATIONS. ASSUMPTIONS AND LIMITATIONS
ASSOCIATED WITH DRASTIC
It should be noted that out of the seven parameters in DRASTIC which determine the
susceptibility of groundwater to contamination, some parameters, such as geology,
have a greater influence on the final index score than other parameters, such as
33
-------
topography. Bedrock geology influences aquifer media, vadose zone and hydraulic
conductivity for a combined weight of 11, and slope has only a weight of one. In
Jefferson County, since the geology is primarily limestone, it had to be determined
whether it was massive or karst. Data on fracturing, porosity and permeability were
important variables affecting groundwater pollution potential. The final DRASTIC
index varied by three categories between massive and karst, arid from moderately to
highly susceptible because of the heavy weighting of geology.
Also of importance to note, is that hi many cases site specific field data is lacking. In
Jefferson County insufficient data for the categories of depth to water, net recharge
and hydraulic conductivity, led to guesswork and tentative conclusions regarding the
pollution potential of the study area. In addition, very generalized maps were used to
estimate soil and topography scores.
Examples of intended uses of DRASTIC include: prioritizing protection, monitoring
and clean up efforts, and directing investigations and resource expenditures. A
specific example of an intended use of DRASTIC is to combine die DRASTIC map
with information on pesticide usage to estimate the degree of threat pesticides may
pose to ground water.
DRASTIC was never meant to be used for facility siting decisions, such as where to
allow septic systems or landfills. It does not replace on site investigations.
There are many assumptions of DRASTIC, including the following:
- Pollutant has the mobility of water.
- Pollutant is introduced at the surface.
- Pollutant is carried to ground water via recharge from precipitation.
There are also many limitations of DRASTIC, such as the following:
- The area to be evaluated must be a minimum of 100 acres.
- Cannot evaluate semi-confined or leaky aquifers: can only evaluate confined or
unconfined aquifers.
While it is true that EPA funded the development of the DRASTIC method, and
utilized DRASTIC for the design of the weU sampling network hi the National
Pesticide Survey, the following should be noted:
1. DRASTIC is considered by EPA as just one of many methods described hi the
agency's Technical Assistance Document (TAD) entitled, "A Review of
Methods for Assessing Aquifer Sensitivity and Ground Water Vulnerability to
Pesticide Contamination" (September 1993). The purpose of the TAD is to
provide States with information on the uses, assumptions and limitations of
34
-------
each of the methods. The TAD should help States make informed decisions
regarding what techniques they find most appropriate given each State's
hydrogeologic and pesticide use characteristics, protection goals and resource
constraints.
2. EPA's National Pesticide Survey showed that at the County level there was no
positive correlation between overall DRASTIC scores and pesticide detections
in individual wells. No association was found between overall subCounty
DRASTIC scores and detections either, except the "impact of the vadose zone"
subscore component of DRASTIC was found to be associated with detections
in the manner expected.
3. EPA recognizes that DRASTIC was developed as a general model to indicate
relative sensitivity for areas 100 acres or larger in size, and does not address
chemical characteristics of contaminants. DRASTIC cannot by itself
adequately predict contamination. In addition, DRASTIC was designed to
predict the potential for contamination in ground water rather than in drinking
water wells. Although the National Pesticide Survey sampled wells in a
variety of settings, including confined aquifers, water table aquifers and
different depths below the land surface and the water table, there was
insufficient funding to field verify the exact hydrogeologic setting of each well.
Therefore, it is impossible to use the results of the Survey alone to validate or
invalidate DRASTIC.
DRASTIC must be combined with data on pollutant loadings and land use. The
DRASTIC map developed for Jefferson County is useful as a first cut screening tool
when combined with land use information. However, given the limitations described
above, its use should be limited to a screening tool. For these reasons, the use of
DRASTIC in the Jefferson County report is primarily for the purpose of comparison
of DRASTIC with more site specific methods such as SPISP.
4. DRASTIC WITH STREAMS: There is a good correlation between the location of
streams and high DRASTIC scores, as shown by map number 9, because of the
influence of shallow ground water and flat topography in the floodplain of these areas.
This poses a concern for ground water vulnerability in these areas, as well as the
possibility of contaminated ground water discharging to surface water.
5. DRASTIC. FAULTS. FRACTURES AND SINKHOLES: Since the presence of
faults and fractures influenced the DRASTIC ratings for hydraulic conductivity,
aquifer media, and the vadose zone, there is a good correlation between these
structural features, the extent of the karst limestone formations and the areas scored
high by DRASTIC, as shown by map number 10. DRASTIC scores drop off to the
east of the border between the Elbrook and Waynesboro formations, consistent with
the rationale presented earlier, since this is the break between karst and massive
35
-------
Ground Water Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, West Virginia
DRASTIC Index
H < 79
80-99
100 - 119
120 - 139
160 - 179
180 - 199
> 200
Region III
DRASTIC WITH STREAMS
Mao: JC09
Data Source: lest Virginia DNR &USGS
July, 1992
-------
Ground Water Vulnerabi 111\ to Contamination bv Agricultural Chemicals
1 v o
Jefferson C o u n t\ , West Virginia
DRASTIC Index
< 79
80-99
100 - 119
120 - 139
140 - 159
160 - 179
> 200
|^| Faults
HO Fractures
[T] Sinkholes
Region
DRASTIC, FAULTS, FRACTURES & SINKHOLES
lap: JCOlO
Data Source: ffVDNR, TOES & USGS
July, 1992
-------
limestone and the beginning of the less permeable dolomite in the Waynesboro
formation. Where streams cut through, the DRASTIC rankings increase in all the
formations. The metamorphic rocks of the Blue Ridge Mountains on the eastern
boundary and the Martinsburg shale on the western boundary have the lowest
permeability, and the lowest DRASTIC scores.
Since sinkholes often act as direct conduits for contamination to reach ground water,
they violate the assumptions of DRASTIC (that a contaminant be introduced at the
surface and be carried through the soil profile with recharge before reaching ground
water). Therefore, since favorable ratings can be overruled by sinkholes, they should
be considered in conjunction with the DRASTIC map.
SOIL PESTICIDE INTERACTION SCREENING PROCEDURE
1. GENERAL DESCRIPTION
Since DRASTIC is only useful as a gross screening tool for areas 100 acres or
greater, and since DRASTIC uses a very generalized small scale soil association map,
the decision was made to use a technique which would allow a more refined and
detailed analysis, based on specific soil types. Soils were found in this study to be an
extremely important data layer. Not only are they often mapped in polygons of a few
acres in size, they also form the first potential barrier to pesticide leaching in a
farmer's field. From a pollution prevention standpoint, this initial migration process
is the most critical to evaluate.
Don Goss of the Soil Conservation Service in Fort Worth, Texas, developed a
screening technique to evaluate the likelihood for different pesticides to leach below
the root zone, or to be carried in surface runoff water away from the site of
application, under different field conditions. This technique, known as SPISP (Soil
Pesticide Interaction Screening Procedure) is based on the GLEAMS (Ground Water
Loading Effects of Agricultural Management Systems) model. The GLEAMS model
was developed by the Agricultural Research Service to evaluate the effects of
agricultural management systems on the movement of agricultural chemicals within
and through the plant root zone. GLEAMS calculates the grams per hectare of
pesticides that would be lost from different soil types, by simulating the following:
migration of up to 10 pesticides, routed through up to 12 layers contained within up
to five soil horizons for a simulation period that can be as long as 50 years (7).
In order to develop SPISP, pesticide loss was estimated from over 40,000 runs of the
GLEAMS model, using different combinations of soils and pesticides with a wide
range of properties. Soil input parameters used in the model simulations included:
thickness and percent organic matter of the surface soil horizon, and the surface and
subsurface soil texture and drainage characteristics. Pesticide parameters included:
solubility, soil half-life, and the organic carbon partitioning coefficient (Koc).
38
-------
Climatic variables, such as soil moisture and temperature, and the type of
management or pesticide application practices employed by the farmer were not
included as variables in the model. A consistent 4% slope and considerable
precipitation after application was assumed. Specifically, the precipitation assumption
was as follows: a 3.5 inch precipitation event was generated every second day after
application for five events, and then a 1.0 inch event every other day for at least four
times the half life of the pesticide (7). Considerable precipitation was assumed in
order to produce the most likely situations for pesticide loss. The primary goal was
to determine the capacity of a soil to retain a pesticide at the point of application,
regardless of management or climatic inputs. The meteorological inputs were not
intended to represent any climatic zone, although the probability of rainfall soon after
application should be considered in most climates.
To analyze and categorize the results of the 40,896 simulations, a statistical regression
analysis was used to select just those soil and pesticide input parameters that weighted
most heavily for estimating pesticide losses. Soil parameters selected were soil
hydrologic group, depth of the first soil horizon, K factor and organic matter content.
Hydrologic group is a soil interpretation used by SCS to categorize soils based on
their potential to allow water infiltration. The soil K factor is a measure of the
credibility of a soil, or the likelihood of detachment and transport of soil particles due
to rainfall. Pesticide parameters selected were half life, Koc and solubility (as
defined in the beginning of this chapter). These soil and pesticide parameters were
then used with various algorithms to group pesticide loss. For example, one of the
soil leaching loss algorithms states that if the hydrologic group = B, the % organic
matter x horizon #1 depth is <. 9 and the soil K factor is _<_ 0.48, then the potential
for leaching is high (7). Using all the algorithms developed, soils are classified into
high, intermediate, low and very low leaching potential, and pesticides are classified
into large, medium, small and extra small pesticide leaching potential. Soil attributes
used in the SPISP algorithms are available through queries of the Soil Interpretations
Record (SCS-SOI-S) data set at Ames, Iowa for soils found throughout the country.
A computer program is available which runs the attributes through the appropriate
algorithms to generate ratings for each soil type. In addition, West Virginia has a
State Soil Survey Database which contains soil attribute data for digitized soil maps in
the State. Where soil leaching and runoff ratings are not available for a specific soil
type, the soil attributes can be used in the SPISP algorithms to obtain the soil
potential rating for the soil type. Ratings for over 230 commonly used pesticides are
available through the Soil Conservation Service's Pesticide Data Base. Pesticide loss
ratings have been assigned to all pesticides in the data base using algorithms
developed through the GLEAMS model simulations.
2. SPISP RANKING OF JEFFERSON COUNTY SOILS
Map 11 shows soils ranked for leachability, using the SPISP screening method.
There are no soils ranked high because of the influence of the large percent of fine
39
-------
Ground Water Vulnerability tc Contamination by Agricultural Chemica
Jefferson Count v, West Virginia
N om i n a I
| | In t e rme d i i t e
gT| N omi n 11 / I n t e rme d i i t e
I Non-Soil Areti
Region II
SPISP SOIL LEACHING POTENTIAL
Man- JC011
Data Source: U.S.D.A. S.C.S
July, 1992
-------
textured soils (silts and clays) in the County, as well as a predominance of soils which
have a high percent of organic matter.
The highest ranked soils are those with an intermediate potential for leaching; these
soils are in the parts of the County shaded yellow, which includes most of the
carbonate valley, where the medium to coarse textured soils predominate, and the
gently rolling terrain allows for infiltration.
The intermediate/nominal potential is a split ranking, because the mapped soil units in
these areas consist of two different soil types. A common example is the Hagerstown
and Frederick soil map units which contain deep, well drained, medium to moderately
fine textured soils formed from weathered limestone. Soils in this unit are either all
Hagerstown, all Frederick or include characteristics of both soils. The
intermediate/nominal potential is shaded light green, and covers most of the western
half of the carbonate valley.
Soils with a nominal leaching potential are found predominantly in the Blue Ridge
Mountains, and over the Martinsburg Shale. A high percent organic matter
contributes to low rankings in the mountains. Fine textured soils in the mountains, as
well as over the Martinsburg Shale, also play a role in keeping rankings low hi these
areas. Soils with a nominal leaching potential are also found in alluvial areas along
tributaries of the rivers. This is most likely due to the high percent of organic matter
associated with soils found along the tributaries. In addition, poorly sorted alluvial
material, poor drainage, high water tables and flooding associated with these soils can
reduce leaching potentials.
It should be noted that the SPISP classifications for soils that were utilized in
Jefferson County are based on the original version of SPISP. These classifications
have now been replaced with the categories of high, intermediate, low and very low.
The use of the original SPISP I versus the newer SPISP n, as well as an analysis of
soil leachability hi conjunction with pesticide use is presented in Chapter 5.
Section HI
Data discussed up to this point, with the exception of the base map, has included
soils, geology, hydrology and hydrogeology of the County. The series of data coverages
which follow include various aspects of land use, such as general land use classifications,
pesticide use and characteristics of private drinking water sources.
3
1. LAND USE. LAND COVER: The Agricultural Stabilization and Conservation
Service (ASCS) of USD A conducts annual aerial flights for program compliance, and
develops slides from these flights. Color slides developed by ASCS hi 1989 were
used to establish land cover and land use patterns. SCS projected the slides against
the 1:24,000 scale topographic mylar base map, and using a light table, made
41
-------
Ground Vater Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, West Virginia
Commerciil/lDduitrii
Region II!
GENERALIZED LANDUSE
Hap: JC012
Data Source: U.S.D.A. S.C.S
July. 1992
-------
approximations to adjust the photographic information to fit the base map scale.
Boundaries of different land uses and land cover were drawn by SCS onto the stable
base map and digitized. The SCS conservation plan maps, which are aerial
photographs (scale 1" = 660') upon which planners have drawn boundaries of
cropped fields, were also used to guide the delineations of different land uses. The
land use categories that were used are consistent with the National Resources
Inventory. Map 12, which is the SCS land use map, shows that most of the County
is hi agriculture, and there is very little land in commercial or industrial use. The
blue areas on the map depict locations of two agricultural research facilities (orchards)
and two fisheries research sites. The greatest concentrations of residential areas of
the County (shown hi red) are found predominantly along the Shenandoah River
valley, the Blue Ridge Mountains and in Charles Town, the center of the County.
Because of the distortion found hi the 35 mm color slides, the land use/land cover
data layer does not have as high quality locational accuracy as the farm parcel
coverage.
2. FARM PARCELS: The 1987 Census of Agriculture hi West Virginia reported a total
of 363 farms hi Jefferson County. ASCS and SCS developed the farm tract coverage
of over 90% of these farms. The coverage was created by transferring farm tract
boundaries, which were drawn by conservation planners hi the field on aerial
photographs (1" = 660' scale SCS farm plan maps), to mylars of 7.5 minute
topographic maps. Infrared aerial photographs provided a photobase to help make
accurate delineations. ASCS aerial photos and other aerial photography flown around
1982 was also used to support the delineations of farm tract boundaries. The tract
delineations were manually digitized from the mylars. These tracts are displayed by
map number 13, which shows a high density of farms throughout the County, with
the exception of the Blue Ridge Mountains and the Martinsburg Shale area.
3. SURVEYED FARMS: In the Spring of 1990, EPA funded an Agricultural Practices
Survey which was conducted on 118 farms by EPA's consultant, ICF, and the Eastern
Panhandle Regional Planning and Development Council. Since no data on pesticide
usage existed at the farm level, raw data had to be collected and automated. The 118
farms surveyed represent approximately one third of the farms hi the County. Map
number 14 shows the farms where interviews took place. Close to 90% cooperation
was achieved, many interviews were taped and some interviews lasted up to one and a
half hours hi length. Collection of pesticide usage data was a major effort. Chapter
4 contains details on the design and administration of the survey and the automation
of survey data.
Participants were assured their names and addresses would remain confidential, and
results would not be used for enforcement, but rather for technical assistance. To
help ensure confidentiality, all maps which display farm locations or survey
information do not have road locations or tract boundaries displayed.
43
-------
Ground ffater Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, West Virginia
1! Firms
Region III
JEFFERSON COUNTY FARMS
Map: JC013
Data Source: U.S.D.A. S.C.S.
July.
-------
Ground ffater Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, West Virginia
Sur reje d F arms
Non-Snr veje d F arms
Non- Agrica 1 I ura 1 Areai
\
s
Region III
SURVEYED FARMS
Map: JC014 Data Source: EPA Agricultural Practices Survey July. 1992
-------
4. ROCK TYPES AND SURVEYED FARMS: Map number 15 shows geology
aggregated by the major rock types found in the geologic formations, with an overlay
of farm parcels which were surveyed. Faults, fractures and sinkholes are also
displayed on this map. Most of the farms surveyed are located over predominantly
limestone formations. There is generally a high density of sinkholes within surveyed
parcels.
5. PRIVATE DRINKING WATER WELLS AND SPRINGS: As part of the
Agricultural Practices Survey, locations of wells and springs used by respondents as
their primary drinking water source, were obtained in the field, marked on low
altitude aerial photographs (1" = 660'scale), then transferred to stable base mylars of
7.5 minute topographic maps. The locations were then digitized, and are displayed
on map number 16. Through the survey, information was obtained on well depth,
well or spring construction, the date the well was drilled, and results of any well or
spring sampling.
There was a wide range in variation of well depth, from 22 to 760 feet. Over half
the wells were 160 feet or less in depth. Wells were drilled (or dug) between 1832
and 1989; 50% of the wells are over 25 years old. There were 28 farmers who had
tiieir wells or springs tested for nitrates - concentrations ranged from 4 to 250 ppm.
Since 25 ppm is the next highest response below 250 ppm, it is suspected that this
reading is the calculation of nitrate versus nitrogen. The drinking water standard, or
MCL, is 10 ppm for nitrate, recorded as nitrogen and 45 ppm for nitrate recorded as
nitrate. Only seven farmers had their wells or springs tested for pesticides, and there
were no detections reported. All of this type of information is stored in the Region's
CIS as attribute files for the well and spring locations.
6. PESTICIDE USAGE BY FARM: Most of the Agricultural Practices Survey focused
on questions related to pesticide usage. It was determined by the workgroup in the
beginning that pesticide usage information was lacking and was a critical data layer
necessary to conduct any type of pesticides in groundwater susceptibility analysis.
Furthermore, State agencies in Region HI expressed the concern that they have little
information on pesticide usage. This is confirmed by a recent survey by the
Resources for the Future (RFF) which cites the lack of a comprehensive data base on
pesticide use that could be used for risk assessment studies. RFF reported that nine
States, some of which are largely agricultural, have absolutely no records of pesticide
use, and nine other States have published reports on pesticide usage, but their data has
not been regularly updated. According to the RFF report, only Hawaii, Oregon,
Ohio and New Hampshire produce regular up-to-date reports (6). The pesticide usage
survey in Jefferson County focused on the agricultural use of pesticides. This
decision was made because the Pesticides and Ground Water Strategy addresses
agricultural use only and because more than 70% of all the pesticides used in the
United States are applied to agricultural lands. (14).
46
-------
Ground Water Vulnerability to Contamination by Agricultural Chemicals
Jefferson County , West Virginia
Rock Type
limestone
Dolomite
Quartzite
Shale
Phyllite
^? J Mela Lava
| | Composite (shale, sandstone, dolomite, limestone)
\fij\ Faults
|/\/| Fractures
[V] Sinkholes
VTA Surveyed Farms
w
o
^
Region II
ROCK TYPES & SURVEYED FARMS
Uap: JC015
Data Source: WG k ES. USGS & USDA S.C.S
July, 1992
-------
Ground Water Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, West Virginia
later Wells
Water Springs
w
5?
Region III
PRIVATE DRINKING WATER FELLS AND SPRINGS
Mao: JC016 Data Source: EPA Agricultural Practices Survey
July,
-------
All survey data was tracked through unique identification numbers which were
assigned to the farms. The numbers are a combination of the ASCS tract number, the
watershed number and the order of the property on the SCS list. Unique
identification numbers allowed the attribute information associated with the survey to
be displayed geographically.
The remainder of this chapter describes the pesticide use information from the survey
which is displayed in maps numbered 17 through 27. Since many of these maps are
sequential and relate to each other, the map keys sometimes have more symbols than
appear on any one map; this has been done to allow the reader to compare maps that
are found hi a series (such as a comparison of pesticide use in 1989 versus 1988). In
addition, to reduce the number of classifications found on each map, the "no data"
and "zero pesticide usage" categories are not displayed. Readers should be aware that
there were some surveyed farms which either did not use the pesticide(s) in question
or application rates could not be calculated.
Maps numbered 17 and 18 show total pesticide usage per farm for 1988 and 1989.
Total pesticide usage ranged from 1.0 to 4,096 pounds of active ingredient applied
per farm.
Respondents reported using a total of 69 different pesticides on their crops in the
years 1988 and 1989. The top five pesticides reported as having the highest use hi
1989 (total pounds of active ingredient) were as follows (in descending order):
Atrazine, Metolachlor, Alachlor, Simazine and 2,4-D. In 1989 the major crops
grown by respondents (hi descending order) were com, alfalfa, soybeans., wheat,
apples, peaches, pasture, hay, truck crops, sorghum and nectarines. In addition to
pesticide application information, questions were asked on management practices,
such as integrated pest management (IPM). animal waste practices, septic systems,
underground storage tanks, and pesticide storage and disposal practices. Detailed
survey results are provided in Chapter 4 and in Appendix A.
RATIO OF PESTICIDE USAGE TO CROPLAND: The purpose of maps numbered
19 and 20 is to compare total pesticide usage on a farm with the amount of tillable
acres, for the years 1988 and 1989. By displaying the data hi terms of pounds of
pesticides applied per tillable acre, the value reflects intensity of use. A comparison
of these maps with the previous maps, reveals that some farms which had moderate or
low total pesticide usage on the farm, now show up as having high application rates
per acre. In most cases, however, if the total amount of pesticides applied on the
farm were high, the rates per acre are also fairly high. Application rates of individual
pesticides ranged from .0075 to 19 pounds of active ingredient per acre. The crop
with the highest intensity of overall pesticide use per acre was nectarines, which used
a little over 21 pounds per acre of pesticides in 1989.
49
-------
Ground Water Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, West Virginia
o
500
1000
1500
2000
Total Lb s
500
1000
1500
2000
4096
w
Region 111
Total Pesticide Usage per Farm- 1988
(Lbs. of Active Ingredient)
Mao: JC017
Ma Source: EPA A£ricultural Practices Survey
July, 1992
-------
Ground Water Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, lest Virginia
\
o
Region III
Total Pesticide Usage per Farm- 1989
(Lbs. of Active Ingredient)
Map: JC018
Data Source: EPA Agricultural Practices Survey
July, 1992
-------
Ground Water Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, West Virginia
Greater than
\
o
Region III
Total Lbs of Pesticides Applied per Acre - 1988
Map: JC019
Data Source: EPA Agricultural Practices Survey
July, 1992
-------
Ground Water Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, West Virginia
w
Region III
Total Lbs of Pesticides Applied per Acre - 1989
Map: JC020
Data Source: EPA Agricultural Practices Survey
July, 1992
-------
PRIORITY LEACHING PESTICIDE USAGE BY FARM: Priority leachers have
specific properties that make these pesticides good candidates for leaching to ground
water. For this project, a pesticide is considered a priority leacher if it meets any of
the following criteria:
* Detected in drinking water wells by EPA's National Pesticide Survey,
* Ranked as a high potential for leaching by the SPISP (version II) algorithm, or
* Listed as a priority leacher on EPA's list of analytes sampled for in the National
Pesticide Survey.
The SPISP (version II) technique rates a pesticide with a high probability of leaching
if the conditions of the following algorithm are met:
* log(half-life) x (4-log (Koc)) _>_ 2.8
The National Pesticide Survey selected potential analytes for sampling by applying the
following criteria. First pesticides and pesticide degradates previously detected in
ground water, as well as pesticides regulated under the Safe Drinking Water Act,
were automatically included if:
* the pesticide's chemical/physical parameters, including for example water
solubility', partition coefficients, field half life and hydrolysis half life, indicated a
potential to leach to ground water; and
* at least one million pounds of the pesticide was estimated to have been used
nationwide in 1982 (23).
Additional analytes added to the list were chemicals that the multi-residue testing
methods would identify and quantify for little or no added expense. A final list of
127 analytes were chosen for National Pesticide Survey sampling (23). Approx-
imately 55 of these pesticides were classified as priority leachers based on then-
physical and chemical characteristics, as well as their presence in ground water (24).
Maps numbered 21 and 22 show just those farms which use priority leachers in 1988
and 1989. A total of 25,983 pounds of priority leachers were applied in 1989, and
25,170 pounds were applied in 1988 to the surveyed parcels. Crops which used the
highest amounts of priority leaching pesticides in 1989, in descending order, were:
corn, soybeans, alfalfa, peaches, pasture, fallow land, apples, truck crops, sorghum,
wheat, nectarines and hay. Priority leaching pesticides made up approximately 63%
of the total amount of pesticides applied to the surveyed parcels. When just priority
leacher use is portrayed, the total pounds per farm generally drops from the amount
portrayed hi maps 17 and 18, which showed total pesticide use per farm. For
54
-------
Ground Water Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, West Virginia
w
Region III
Priority Leaching Pesticide Usage per Fam- 1988
(Lbs. of Active Ingredient)
Map: JC021
Data Source: EPA Agricultural Practices Survey
July, 1992
-------
Ground tfater Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, West Virginia
Ed
O
Region III
Priority Leaching Pesticide Usage per Farm- 1
(Lbs. of Active Ingredient)
Map: JC022
Data Source: EPA Agricultural Practices Survey
July, 1992
-------
example, there are no farms which use over 2,000 pounds of priority leachers per
farm. A few farms which had high pesticide usage now show up as low or moderate
use of priority leachers. Appendix A contains information on what specific priority
leachers were used hi Jefferson County.
9. LOCATION AND AMOUNT OF USE FOR TOP FIVE PRIORITY LEACHERS: .
Maps numbered 23 through 27 show the parcels which used the top five priority
leaching pesticides for the 1989 growing season. Out of all the pesticides classified as
priority leachers, the compounds which had the highest use hi 1988 (hi descending
order) were: Metolachlor, Atrazine, Simazine, Alachlor and Cyanazine. In 1989 the
order was: Atrazine, Metolachlor, Alachlor, Simazine and 2,4-D ranked number five
instead of Cyanazine. These maps show the total pounds used per parcel in 1989. In
both years, Metolachlor and Atrazine use is the most widespread and the amounts
used per farm tend to be higher than for the other compounds. Regardless of the year
evaluated, the top four pesticides used in the County were Atrazine, Metolachlor,
Alachlor and Simazine, all of which are priority leachers.
57
-------
Ground Water Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, West Virginia
Region III
Priority Leacher Use -
Usage Rank 1 - ATRAZ I NE
Total Pounds - County: 7479.25
1989
Map: JC023
Data Source: EPA Agricultural Practices Survey
July, 1992
-------
Ground Water Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, West Virginia
Region III
Priority Leacher Use - 1989
Usage Rank 2 - METOLACHLOR
Total Pounds - County: 6846.05
Map: JC024
Data bource: EPA Agricultural Practices Survey
July, 1992
-------
Ground Water Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, West Virginia
Region III
Priority Leacher Use - 1989
Usage Rank 3 - ALACHLOR
Total Pounds - County: 3174.35
Map: JC025
Data bource: EPA Agricultural Practices Survey
July, 1992
-------
Ground later Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, West Virginia
Region 111
Priority Leacher Use - 1989
Usage Rank 4 - S1MAZINE
Total Pounds - County: 2134.21
Map: JC026
Data Source: EPA Agricultural Practices Survey
July, 1992
-------
Ground Water Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, West Virginia
Priority Leacher Use - 1989
Usage Rank 5 - 2,4-D
Total Pounds - County: 1898.32
Mao: JC027
Data Source: EPA Agricultural Practices Survey
July, 1992
-------
CHAPTER 4: AGRICULTURAL PRACTICES SURVEY
A. BACKGROUND
B. SURVEY DESIGN: DEVELOPMENT OF QUESTIONNAIRE
C. SURVEY DESIGN: SELECTION OF SAMPLE
D. CONFIDENTIALITY POLICY
E. ADMINISTRATION OF SURVEY
F. SURVEY EDITS, ESTIMATES AND ASSUMPTIONS
G. AUTOMATION OF SURVEY DATA
H. PROBLEMS ENCOUNTERED, OBSERVATIONS MADE IN THE FIELD
I. SURVEY RESULTS
A. Background
EPA Region III funded a survey in Jefferson County, West Virginia, which was
carried out in the Spring of 1990 through EPA's contractor, ICF, arid the Eastern Panhandle
Regional Planning and Development Council, located in Martinsburg, WV. In addition,
assistance for the survey was provided by the local offices of the USDA Soil Conservation
Service and Cooperative Extension Service. Information was obtained from Jefferson County
farmers on agricultural practices. While the primary focus of the survey was to obtain
pesticide usage information, information was also obtained on private and public wells,
fertilizer usage, animal waste management, underground storage tanks, septic systems and
conservation practices. All data was automated by ICF in the Summer of 1990 and is
available on dBASE HI Plus files.
EPA Region ffl used their Geographic Information System (GIS) to enter, store and
analyze survey data on pesticide usage and drinking water wells in combination with data on
ground water vulnerability. The GIS was used to target priority areas in the County which
have a high potential for ground water contamination from pesticides, and to identify where
populations are potentially at risk. In order to utilize the survey data in GIS, all data had to
be geo-referenced. Each questionnaire was assigned a unique farm tract identification
number for the farm the questionnaire was administered on. These numbers are the farm
tract numbers assigned by the Agricultural Stabilization and Conservation Service (ASCS).
The tract numbers and the locations of the farm parcels they correspond to are stored in GIS.
B. Surrey Design: Development of Questionnaire
The questionnaire used by EPA for the National Pesticide Survey (conducted between
1988 and 1990) was modified to meet the needs of the Jefferson County pilot project. The
final questionnaire used in Jefferson County incorporated the comments and suggestions of
the Jefferson County Workgroup. Examples of some changes which were made include the
following: expanded the number and complexity of questions dealing with conservation
practices, such as nutrient management, conservation tillage, integrated pest management,
and manure management, added more specific questions on water wells and springs and
63
-------
added a section on underground storage tanks.
Whenever the federal government funds a survey to be administered to 10 or more
people, approval has to be obtained from the Office of Management and Budget (OMB) on
the design of the questionnaire and on the overall survey. The purpose of OMB's approval
is to ensure compliance with the Paperwork Reduction Act, namely to prevent duplication of
effort within the government and undue burden on respondents to complete government
surveys. OMB required that the Jefferson County Survey be shortened in length. The OMB
review process was very time consuming and a very laborious procedure. In the process of
condensing the questionnaire to comply with OMB review requirements, the question which
asked respondents about the acreage of each crop field was inadvertently deleted. This
accidental omission created numerous problems later, which are discussed under Section F.
It was found that one of the most difficult aspects of questionnaire design involved
determining the correct skip sequences to use to control the flow of questions. The skip
sequences were critical to prevent inappropriate questions from being asked and to ensure
that all relevant questions were asked. During the pre-test and administration of the full
survey, some minor edits were made to skip sequences, which did not sacrifice the
completeness of the survey results. However, several phone calls and some follow-up visits
were made to some farmers because of the problem associated with sorting out questions that
were germane to the farmers versus the commercial applicators. For example, if a
commercial applicator applies all the pesticides on a property, the enumerator is instructed to
skip past all the questions which relate to pesticide storage, mixing, loading and disposal, and
integrated pest management on the property. However, in some cases, even though a farmer
did not apply pesticides on the property, the farmer knew more about these practices than the
commercial applicator, and therefore should have been originally asked the questions
pertaining to the practices. Close attention to development of optimum skip sequences
reduces the need for time consuming follow up.
C. Survey Design: Selection of Sample
The 1987 Census of Agriculture indicated a total of 363 farms in Jefferson County.
Due to the difficulty, time and expense of accessing and verifying the data base used by the
Census of Agriculture, a random sample was not selected from this data base. Instead, the
USDA's list of farmers who are cooperators with the County Conservation District was used,
since close to 90% of the farmers in the County are cooperators with the District. The list
was initially edited by Cooperative Extension Service and Soil Conservation Service
personnel, to exclude individuals from the list that they believed would be non-cooperative in
the survey, and possibly detrimental to the success of the overall data collection effort.
Using a sample of cooperators proved helpful in overcoming some respondent reluctance
situations. The edited list provided 147 farm tracts and 115 farm operators within 12 of the
13 watersheds hi the County. Maps 14 and 28 show that the farms surveyed are distributed
hi a manner that is spatially representative of the agricultural land in the County.
64
-------
Ground ffater Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, lest Virginia
H^ Surrejed Cropland
Q Non-Surrejed Cropland
|| Non Agricultural Areas
Region III
Surveyed vs. Non-Surveyed Croplands
Mao: JC028
Data Source: U.S.D.A. s.U.
-------
There were only 16 refusals out of the original list, which is a response rate of close
to 90%. Surveys were taken on 118 tracts. The remaining tracts were either no longer in
farming, were surveyed during a pre-test, or were duplicates on the original list. There were
91 farmers who signed a form giving permission to use the collected information; the rest
gave their verbal consent.
Because the survey design did not yield a statistically representative sample, the
results cannot be used to make inferences regarding the population in the County which was
not surveyed. For example, one cannot assume that average pesticide application rates for
specific crops on surveyed acres will be the same for those same crops on non-surveyed
acres. The results of the survey do, however, provide good insights regarding the practices
of almost one third of the farmers in Jefferson County and many valuable lessons were
learned regarding the collection of agricultural land use data.
D. Confidentiality Policy
Participants were assured their names and addresses would remain confidential and
survey results would not be used for enforcement, but rather for an informational data base
to support research and to identify areas where growers may need technical assistance.
As mentioned above previously, 91 respondents signed a form granting permission to
use the information collected on their farms, and the remaining 17 farmers gave their verbal
consent. The permission form specifies that, although names and addresses will not be
released, the survey data may be mapped or graphed and subsequently displayed.
In addition, participants were informed that the survey data may be shared with State
and local agencies. The data is only shared with agencies that agree not to use the data in
any manner which would violate the confidentiality policy.
E. Administration of the Survey
The Eastern Panhandle Regional Planning and Development Council (EPRPDC)
provided six people to conduct the survey interviews. These individuals attended a one day
training session which was conducted by ICF. The training covered: purpose of the survey,
how to make initial contacts with respondents, how to conduct the interview, how to control
bias through the manner in which questions are asked, attitude and body language, how to
probe to get specific answers, and how to establish a good rapport with the respondents. In
addition, a slide show was presented which covered technical considerations, such as what to
look for at the site in terms of well construction, septic systems, agricultural practices,
pesticide mixing, loading, storage and disposal sites, potential sources of contamination, etc.
Critical definitions were provided on terminology used in the Survey, such as the Safe
Drinking Water Act definition of public water supplies and Maximum Contaminant Levels,
explanations of types of pesticides, brand names versus active ingredients, formulations and
use of correct units for recording application rate information. Instructions were given
66
-------
regarding how to locate well sites on topographic maps and sketch the wellhead area.
One person from the survey team was put in charge of making contacts and arranging
interview schedules. This arrangement worked well, made efficient use of tune and
prevented duplications in making initial contacts.
Interviewers were provided with tape recorders and instructed to ask respondents if
they objected to the taping of the interview. Use of the tape recorder proved very valuable,
since it allowed the survey manager to assess the interviewer's field work and provide
immediate feedback. In addition, the recorded information was used to edit the surveys
where data was inadvertently left out by the enumerator. :
Interviewing time was extremely varied, from 20 to 90 minutes in length, depending
upon the interviewer, the attitude of the respondent, the knowledge of the respondent, the
complexity of the respondent's operation and the use or non-use of commercial pesticide
applicators.
Overall, the interview team did an exceptional job of interacting with the respondents
and the cooperation of the farmers was excellent.
F. Survey Edits, Estimates and Assumptions
Edits made to Acreage
Since fanners were only asked for total tillable acres, and were not asked to give the
acreage for each crop grown, this information had to be obtained from the Agricultural
Stabilization and Conservation Service (ASCS). Every year, ASCS obtains information from
farmers on acres of crops grown. This information is used to determine the amount of
commodity payments the farmer is eligible for. ASCS also conducts an aerial flight every
summer, and uses the subsequent photographs to determine the size and locations of the
fields tilled and planted, in order to validate the fanner's verbal reports. The following list
provides an explanation of how the 1989 ASCS photography and report was used to obtain
acreage and crop type data for the project:
1. The ASCS report often indicated a higher amount of total acres of cropland than was
provided by the respondents in the survey. As per the recommendation of the ASCS
office in Jefferson County, where there was a discrepancy between numbers, the
ASCS total acreage was deferred to, since these numbers are used for commodity
payment decisions and are validated through aerial photo interpretations.
2. If all the crops a respondent reported in the survey also appeared in the ASCS report,
the acres of each crop were simply considered identical to the ASCS report. If only
one crop did not show up in the ASCS report, the acreage for all crops reported
through ASCS were subtracted from the total tillable acres respondents provided in
67
-------
the survey, and the remaining acres were assigned to the crop in question.
3. If a fanner only grew one crop, but there was no ASCS report, the total tillable acres
reported by the farmer in the survey was assigned to the crop.
4. If the ASCS report only had wheat or small grain reported for a tract, and the
respondent reported in the survey that he or she grew soybeans, it was assumed that
the soybeans and small grain were double cropped, and the acres of wheat or other
small grain were considered the same as the acres of soybeans. This assumption was
also made in reverse.
5. If sorghum was reported as grown in the ASCS report, but the respondent only stated
that corn was grown on the property, it was assumed that a sorghum field one year
was likely a cornfield the next year, or the previous year, and the acreages were
assumed to be the same.
6. The ASCS report uses the acronym CUPCP for many farms, which stands for
"Conservation Use Acreage, Considered Planted". This category of crop type covers
a variety of relatively permanent ground covers, such as clover, alfalfa, hay, pasture
or grass. If a farmer reported any of these types of crops hi the survey, the CUPCP
category was used to determine acreage, unless the specific crop was stated in the
ASCS report. If there was just a total acreage for CUPCP, but the farmer reported a
variety of different crops in the survey, such as alfalfa, clover and pasture, the total
CUPCP acreage was divided by the three crops, to obtain three equal sized fields.
7. For some crops, such as orchards, there was never an ASCS report, but the
respondents volunteered the information on acreage in the survey.
Edits made to Application Rates Provided bv Respondents:
1. If a response was given as a rate per acre of active ingredient applied, the
recommended application rate of the product was checked against the rate the
respondent gave. This check was made since it is unusual for responses to be
provided in terms of active ingredient application rates. In most cases the rate given
by the respondent was the same as the recommended rate of the product. For these
cases it was assumed that the respondent meant product, not active ingredient. The
recommended product rate was then converted to the pounds per acre of active
ingredient that the product rate was equivalent to. Recommended product application
rates were obtained from the 1988 Crop Protection Chemicals Reference (4).
Medium textured soils with 2-3% organic matter were assumed, when selecting a
recommended rate from a range of rates. In addition, for situations where the
respondent gave an application rate in liquid units for a solid formulation, and
specified it was the rate of the active ingredient, it was assumed that the respondent
meant the product (not the active ingredient) was applied at the specified rate.
68
-------
2. In some cases the computer program converted liquid product application rates to
liquid units of active ingredient applied. Where information was available on the
pounds of active ingredient per liquid unit (gallon) of product, conversions were made
to pounds of active ingredient applied per acre. Information on the chemical
formulation of products was obtained from the 1988 and 1990 Crop Protection
Chemicals References (4), the State of Maryland Pesticide Statistics for 1988 (19),
The Farm Chemicals Handbook (21) and documentation provided by ICF for the trade
name to active ingredient conversion software.
3. If the respondent provided a rate, but no units, the rate was checked against the
recommended rate. If the rates were the same or very similar, the corresponding
units from the literature were used:
4. If a respondent did not give any numerical rate, or simply said that they used the
label rate, the recommended rate for the reported crop and pesticide was used. Once
again, medium textured soils and 2-3% soil organic matter was assumed in
determining the recommended rate. In addition, guidance was obtained from the
County Agricultural Agent regarding recommended rates.
5. If the respondent did not give a formulation, and more than one formulation existed
for the product, the formulation that was consistent with the given rate was chosen.
In addition, the County Agricultural Agent provided guidance regarding the most
common formulations used in the County for the pesticides in question.
6. Although multiple applications are common to orchards, only one respondent
indicated more than one application per year. For this respondent, the application
rate was multiplied by the number of applications to obtain the total amount applied
for the year. If growers with multiple applications did not provide this information to
the enumerators in the field, total applications for those sites would have been
recorded as lower than the actual amounts applied.
7. The software used to convert from brand name to active ingredient application rates
did not break out individual amounts of each active ingredient applied for those
compounds which had multiple active ingredients. An "add-on" program was
developed to do this for all the compounds with multiple active ingredients which
farmers specified they applied to their crops. These compounds included: Bicep,
Bronco, Dowfume, Marksman, Crossbow, Dithane, Dikar, Manzate and Penncozeb.
8. The conversion software had some incorrect mathematical formulas for some
compounds, which were noted and fixed manually. In some cases the "form factor"
used the wrong percentage of active ingredient for calculating amount of active
ingredient, in other cases the "unit factor" had the wrong number to multiply by to
convert from the original units to pounds of active ingredient per acre.
69
-------
9. Farms with a high ratio of total pounds of pesticides applied, as compared to the total
number of tillable acres, were checked to determine first if the properties had
orchards, since orchards tend to have multiple applications of many pesticides. None
of the properties were orchards. Next a check of the conversion program was done
and it was determined that conversions from product to active ingredient application
rates was done correctly. Finally, the data base was checked to determine if the
properties were sprayed by a commercial applicator or by the respondent. In all cases
the properties were sprayed by a commercial applicator. Raw data was checked for
these properties, and it was determined that the commercial applicator records for
several tracts were being assigned to one tract. This problem was due to the fact that
the commercial applicator records were kept in aggregated form for all the tracts
owned by a particular farmer. It is therefore impossible to separate out which
applications apply to which tract. In order to estimate total pesticide use for an
individual tract that was part of the survey sample, the problem was addressed in the
following way:
* Where commercial applicator or respondent records showed more than one field
for the same crop, the total acres of that crop which belonged to the tract in
question were divided by the total number of different fields of the crop indicated
by the commercial applicator's or the respondent's records. For example, if a
fanner had 200 acres of corn on a tract which was included hi the survey, and the
commercial applicator's records showed four different cornfields were sprayed, and
each cornfield was sprayed with a slightly different combination of pesticides, for
a total of 10 different pesticides, not all 10 pesticides are sprayed over the 200
acres belonging to the survey tract. In reality, the four cornfields indicated are
either all on different tracts of land, or there are four different fields on one tract,
which when added up equal 200 acres. It was impossible to tell from the records
which field corresponded to the tract which was surveyed. The solution therefore,
using the above example, was to divide the 200 acres of corn by four, resulting in
four fields, each 50 acres in size. The original respondent data was then used to
indicate which subset of the 10 pesticides was applied to each 50 acre field. This
results in keeping total loadings within a realistic range, as portrayed by the
examples in tables 2 and 3.
70
-------
Table 2
ORIGINAL DATA FOR ONE TRACT WITH ONE 200 ACRE FIELD OF CORN
Commercial
Applicator
Records
Field #1
Field #2
Field #3
Field #4
Total pounds of
Pesticides Applied
to Farm
Pesticide Applied
Aatrex 4L
Gramoxone 1.5S
Aatrex 90 DF
Princep 4L
2,4-D Ester
Extrazine
90 DF
Weedone LV4
Lasso L
Bicep 6L
Banvel 4L
Pesticide
Application
Rate (lb.
a.i./A)
0.4
0.3
1.0
0.8
0.5
2.9
1.0
3.0
3.5
0.3
Field Size
(acres)
200
200
200
200
200
200
200
200
200
200
Subtotal (lb.
a.i./field)
80
60
200
160
100
580
200
600
700
60
2,740
71
-------
Table 3
REVISED DATA FOR SAME TRACT, ASSUMING FOUR 50 ACRE
CORNFIELDS
Commercial
Applicator
Records
Field #1
Field #2
Field #3
Field #4
Total Pounds of
Pesticides Applied
to Farm
Pesticide Applied
Aatrex 4L
Gramoxone 1.5S
Aatrex 90 DF
Princep 4L
2,4-D Ester
Extrazine
90 DF
Weedone LV4
Lasso L
Bicep 6L
Banvel 4L
Pesticide
Application
Rate (Ib.
a.i./acre)
0.4
0.3
1.0
0^8
0.5
2.9
1.0
3.0
3.5
0.3
Field Size
50
50
50
50
50
50
50
50
50
50
Subtotal (Ib.
a.i./field)
20
15 .
50
40
25
145
50
150
175
15
685
72
-------
* The difference between the original data in Table 2 versus the revised data in
Table 3 is 2,055 total pounds of pesticides which would have been incorrectly
assigned to the surveyed tract. The new figure of 685 pounds reflects the total
amount applied when the respondent's tract is divided up into four fields, as
indicated by the commercial applicator records. If there are cases where the four
fields are actually four different tracts, the surveyed tract would be considered one
field of 200 acres with only two to four different pesticides applied to that tract.
For example, if field #2 was the surveyed tract, the total would be 460 pounds of
pesticides applied to the tract, as opposed to the total of 685 pounds of pesticides
applied to the surveyed tract as calculated hi Table 3. However, since it was
impossible to tell if the four fields belonged to four different tracts, the approach
described above was used.
* It should be noted, that by using this approach, a higher diversity of pesticides may
be assumed to be all applied to one farm, which would be an inaccurate
assumption if the four fields are located on different tracts, but would be a correct
assumption for those cases where all four fields are located on the surveyed tract.
There are 30 farms where these edits were made, and this consideration should be
kept hi mind if the pesticide application data is used to guide the selection of
analytes for ground water monitoring on or around these farms. In addition, if
these compounds are priority leaching pesticides, GIS analyses may indicate the
threat of contamination of ground water from these compounds at sites where they
are not even used. It is probably safe to assume, however, that if all the
compounds commonly used on any one crop are not used on one farm which
grows this crop during the two year period covered by the survey, the compounds
probably were used hi previous years or will be used hi subsequent years, simply
because of the rotation of pest and weed problems, and because the compounds are
used on the same crop on neighboring farms. For this reason, the possible
misrepresentation of diversity of pesticides should be considered minor, hi
comparison to the alternative, which would be a gross miscalculation of total
pounds of pesticides applied per farm.
Edits Made during Survey Administration
As the survey was being conducted, ICF and their subcontractor FJRPDC reviewed
completed questionnaires as they were returned from the field. The survey forms were
reviewed for completeness and to determine if skip sequences were followed correctly. In
addition, the responses were evaluated with logic check software to identify where responses
fell outside of a logical range. Where data was missing or incomplete and vague responses
were provided, respondents were called to obtain more information. All edits made to the
questionnaires were done hi the margins with blue pencil, so data entry personnel could
identify the correct response to enter. Original responses were never erased.
73
-------
G. Automation of Survey Data
ICF developed a data entry program in dBASE ffl Plus which was used to automate
all the survey data, and to serve as an interface between the raw survey data and the GIS
software (ARC/INFO). The data entry software and documentation as well as Standard
Operating Procedures for coding survey data can be found in Appendix D.
There are three types of questions in the survey: pre-coded, short answer and open-
ended. The pre-coded questions are multiple choice and the code number which corresponds
to the response selected is easily entered into the data base by the data entry personnel. The
other two types of questions, however, are either coded prior to data entry or entered as
character strings. Coding was done as much as possible, since coding decreases the chances
of error in data entry. The following method was used to code the questionnaires:
1. Most all multiple choice questions had an "other (specify)" category, for those
situations where none of the choices given would suffice. The "other" choice given
by the respondent was written in the questionnaire. The coders rewrote these
responses on separate pages in a list. There was a separate list for all the "other"
choice responses for each question. Then the coder wrote hi blue pencil in the
margin next to the question on the survey form, the number that corresponded to the
order of the "other" choice on the list. This number was then entered into the data
base.
2: For those questions requiring a lengthy explanation or description, such as "describe
the pesticide storage area", the response was entered into the data base as it appeared
on the survey form, as a character string.
3. All pesticide names, pesticide formulations, crop types, and units received unique
numerical codes. For question H-5, which asks about pesticides used and their
application rates, a separate coding sheet was provided and all the information
provided by the respondent for question H-5 was translated into code numbers and
written on the.separate coding sheet. The code "C" was used if a commercial
applicator was the source of the records, the code "R" was used to indicate the
respondent applied his own pesticides. In addition, if the formulation or rate had
been edited by the County Agricultural Agent, this was noted with a "Y" in the flag
column of the coding sheet. For questions which asked about pesticides stored or
disposed of on the property, the code numbers associated with the pesticides were
written in blue pencil in the margins next to the questions.
4. The Observation Record sketch was coded as follows: first a mylar grid was placed
over the sketch, and a phi was used to fix the origin of the X-axis and Y-axis
precisely over the well as marked on the sketch with an X. The grid was oriented to
make sure the Y-axis was parallel with the direction of the north arrow. Next the
distance between features drawn on the sketch and the well were measured by
74
-------
counting the blocks on the grid on the X-axis, and then the blocks on the grid on the
Y-axis to reach the feature. The number of blocks were then converted to meters,
using the scale indicated on the sketch. The code number which corresponded to the
feature, and it's X and Y coordinates were then entered onto a separate coding sheet.
Since the X and Y coordinates are relative to the primary drinking water well or
spring on the property, which was previously digitized, the location of the features
can be displayed using GIS.
H. Problems Encountered, Observations Made in the Field
Problems Encountered:
One problem with administration of the survey had to do with the completion of
information associated with commercial applicator data. A small number of respondents
either did not accurately remember or purposefully deferred the interviewer to a commercial
applicator service when responding to the question, "Who applies the pesticides on this
property?". In some cases the commercial applicator firm had no record of the respondent,
and in other cases the respondent had not used the firm in the past two years. In other cases,
commercial applicator records were found for farmers who did not indicate that they had
used the firm, which infers that some of the respondents may not have given a full
accounting of their pesticide use practices, or commercial applicator records could be
incomplete. In many cases the commercial applicator records could not be associated with a
particular tract of land, instead the records were aggregated for all the tracts of land that a
particular farmer owned. This situation made the task of assigning specific applications to a
tract extremely difficult, as described in Section F. Commercial applicators in West
Virginia, Maryland and Virginia were utilized by survey respondents. Unfortunately, at the
time of the survey, there were no standardized systems in any of these States for
commercial applicators regarding record keeping of pesticide applications tied to specific
locations.
Observations Made in the Field:
Most commercial applicators were not aware of the locations or the importance of
critical features, such as wells and sinkholes on properties that they treat. No special
instructions are given to applicators to protect these areas. In general, however, commercial
applicators were very willing to cooperate and provide records.
Questions about pesticide spills, storage, disposal, mixing or loading, and questions
about underground storage tanks caused both verbal and body language changes in many
respondents. This could indicate that some respondents hedged or became creative in their
responses regarding these questions.
Farm condition may have some utility as an indicator of how well pesticides are
managed. It was often observed that respondents at farmsteads that were in a condition of
75
-------
deterioration produced less written documentation or fewer records and were more elusive,
vague, and/or resentful in their responses to questions about farm practices. In addition,
these farms were often cluttered with trash and empty containers.
I. Survey Results
Owners or operators of 118 farm parcels were interviewed in the Spring of 1990.
Questions were asked about farming practices over the 1988 and 1989 growing seasons. Part
I of the survey was an interview with the respondents, Part n entailed recording observations
of the wellhead and farmstead area.
Part I: Summary of Interview Results:
Most of the farms have their source of drinking water on the farm; 90 parcels had
wells and 8 had springs. Of the 28 farms which tested for nitrates, only three farms reported
levels at or above the drinking water standard of 10 mg/1; most of the other farms did not
know their testing results. Seven farms tested their drinking water for pesticides, none of the
farms knew if any were detected. Drinking water wells were drilled between 1832 and
1989; 83% of the respondent's wells were drilled before 1984. The significance of the
year 1984 is that in 1984 the County passed an ordinance which required the grouting of
drinking water wells.
Well depths ranged from 22 to 760 feet; four wells were less than or equal to 50 feet
deep. At least 38 wells had one or more characteristic of poor construction or potential
susceptibility to contamination, including the following: no grout, inadequate or no well cap,
inadequate or no casing, a depth of 50 feet or less or drilled (dug) before 1916. An
additional 25 wells just met the criteria of constructed prior to 1984 (indicating no grout) but
after 1915. In addition, there were three springs which had no cover and one spring which
was poorly covered. There is one or more abandoned well within 200 feet of the drinking
water well or spring on 16 of the properties. Six of these wells are open at the surface and
only three wells have been filled in - soil and gravel was used as fill material.
Most of the properties (80%) have an on lot septic system or cesspool and 14% of
these systems are located less than 100 feet from the drinking water source on the property.
Four respondents admitted then- system had malfunctioned between 1989 and 1990.
The number of tillable acres farmers have in production ranges from zero to 805
acres; half of the farms had 130 acres or less in production. In 1989 the major crops grown
by the respondents (in descending order) were corn, alfalfa, soybeans, wheat, apples,
peaches, pasture, hay, truck crops, sorghum and nectarines. The 1991 ASCS report of total
cropland reported by farmers in the County also shows corn as the number one crop, but the
order of wheat and alfalfa is reversed. Assuming the ASCS Report acreages are close to the
total cropland acreages in the County, the survey covered 40% of corn cropland, 23% of
soybean cropland and 11% of wheat cropland in the County.
76
-------
Pesticides are used on 105 of the 118 parcels in the survey. A majority of the farms
(70%) utilize commercial applicators to do all or part of the pesticide spraying on their
farms. Respondents reported using a total of 69 different pesticides on their crops in the
years 1988 and 1989. In 1989 the top ten pesticides reported as having the highest use (in
descending order) were: Atrazine, Metolachlor, Alachlor, Simazine, 2,4-D, Metiram,
Ethylene bisdithiocarbamate, Dikar, Paraquat and Cyanazine.
The top ten priority leaching pesticides, which had the highest use hi 1989 according
to the County survey included the following (hi descending order): Atrazine, Metolachlor,
Alachlor, Simazine, 2,4-D, Cyanazine, Dicamba, Carbofuran, Carbaryl, and Diuron.
It is interesting to note that the top five pesticides used hi the County are all priority
leachers, and are all herbicides. In addition, regardless of the year evaluated, the top four
pesticides used hi the County are the priority leaching herbicides atrazine, metolachlor,
alachlor, and simazine. The remaining pesticides in the list of top ten are either fungicides
(Metiram and Ethylene bisdithiocarbamate), herbicides (Paraquat, Cyanazine, Dicamba and
Diuron) or the following combinations: nematicide, insecticide and fumigant (Carbofuran),
insecticide and growth regulator (Carbaryl) or fungicide and miticide (Dikar).
Crops which used the highest amounts of pesticides hi 1989, hi descending order,
were: corn, apples, soybeans, alfalfa, peaches, nectarines, fallow land, pasture, truck crops,
sorghum, wheat and hay. There was a total of 40,882 pounds of pesticides applied to
10,953 acres.in 1989, and 40,222 pounds of pesticides applied to 10,227 acres hi 1988. It is
important to consider total crop acreage when evaluating pesticide use. Although corn had
the highest overall pesticide use, the pesticide application rate per acre, or intensity of use,
ranks number four out of the 12 crops, while orchard crops rank higher hi intensity of total
pesticide use.
Priority leaching pesticides made up approximately 63% of the total amount of
pesticides applied to the surveyed parcels. Crops which used the highest amounts of priority
leaching pesticides in 1989, hi descending order, were: corn, soybeans, alfalfa, peaches,
pasture, fallow land, apples, truck crops, sorghum, wheat, nectarines and hay. A total of
25,983 pounds of priority leachers were applied hi 1989, and 25,170 pounds were applied in
1988 to the surveyed parcels.
When total leacher use is evaluated hi conjunction with acreage information, com still
ranks number one, with close to an average of four pounds per acre of leacher pesticides
applied to corn crops, compared to less than one pound per acre for apples. It is interesting
to note that less than 6% of the pesticides applied to apples were priority leachers, as
compared to 93% for corn. Although 100% of the pesticides applied to truck crops are
priority leachers, total pesticide use for truck crops is low, due to low acreage. Peaches
rank second, just after corn hi intensity (pounds per acre) of priority leacher use. Although
peach acreage is substantially lower than corn, rates per acre are higher and 43% of the
applications to peaches were priority leachers. Nectarines have the highest intensity of
77
-------
pesticide use (over 21 pounds per acre) but the lowest percent of leacher use. Wheat, hay
and alfalfa have a low intensity use of priority leaching pesticides (less than 0.6 pounds per
acre) and the percent of leacher use, compared to the total pesticide use is 61% or below.
Application rates of individual pesticides ranged from .0075 to 19 pounds of active
ingredient per acre. Total applications per farm ranged from 1.0 to 4,096 pounds of
pesticide active ingredient applied per farm. In general, there were very few cases where
farmers did not know their application rates, or gave responses like, "used the label rate".
In most cases actual amounts applied per acre were provided, although fanners were less
likely to know the exact formulations of the products they used.
Only 24% of the farmers who.use pesticides referred to their records when providing
information on pesticide applications to the property.
The distance between pesticide application areas and drinking water wells ranges from
one foot to 3000 feet; on 40% of the properties the distance was 200 feet or less.
At least 86% of the fanners calibrate their pesticide application equipment. Most of
those who do not calibrate use a computerized or pre-set calibration system. For those
fanners who do calibrate, the frequency of calibration varies; approximately one third
calibrate once or twice a season, another one third calibrate once a week or every time the
sprayer is used, and the remainder gave miscellaneous responses, such as, "calibrated once
when purchased or every 8,000 acres".
The top sources of information for farmers regarding the proper use of pesticides (hi
descending order) are: the County Agricultural Agent, sales representatives and the label.
The fourth most popular choice was a combination of personal knowledge and neighbor's
knowledge.
Approximately 24% of the farmers who use pesticides store them year-round. A total
of 65 different pesticides are stored by these farmers. The eight most commonly stored
pesticides (active ingredients), in descending order, are as follows: Dicamba, Glyphosate,
2,4-D, Manganese, Zinc, Carbaryl, Carbamate, and Azinphos-methyl. Most farmers store
pesticides in a completely enclosed building; only four farmers had partially enclosed
buildings, one of which had a dirt floor. Approximately half of the storage areas are 100
feet or less from the drinking water source on the property.
Regarding unwanted or excess pesticides, most of the respondents either return excess
pesticides to the manufacturer, or hold onto the pesticides and use them up on the property.
Four respondents burn pesticides on the property, five take unwanted pesticides to the
landfill, and one buries the pesticides on the property.
The majority of the respondents take pesticide containers to the landfill or triple rinse
the containers and dispose of mem in the trash service. There were, however, several
78
-------
disturbing responses, such as the 17 fanners who burn pesticide containers, and six farmers
who bury the containers on the farm. Overall, 17 farmers disposed of 47 different pesticides
or theu: containers on their property. Less than half the farmers triple rinse the containers
before burying them on the property. Pesticides were not buried on any properties at
distances closer than 900 feet from the property's drinking water source. For properties that
burned pesticides, the burn site ranged from 100 feet to two miles from the drinking water
source. Several farmers expressed the desire for better information regarding what to do
with unwanted pesticides and containers.
The top 14 pesticides or containers which are most frequently disposed of, in
descending order, are as follows: Methyl Parathion, Atrazine, Chlorpyrifos, Methomyl,
Paraquat, Carbofuran, Metolachlor, Carbamate, Azinphos-methyl, Metiram, Glyphosate,
Manganese, Zinc and Ethylene bisdithiocarbamate. There were five or more positive
responses to disposal for each of these compounds.
Pesticide equipment cleaning areas range from one foot to two miles from the
property's drinking water source; 40% of the sites are a distance of 400 feet or less. All
pesticide mixing and loading sites are over permeable surfaces, such as bare ground or grass,
except five sites which are over cement or asphalt. A little over half of the mixing and
loading sites are within 500 feet of the drinking water source, with the closest a distance of
80 feet.
None of the farmers reported any accidental spills of pesticides. In addition, none of
the farmers were aware of any back-siphoning incidents of pesticide contamination to the
drinking water supply.
Only three farmers participate in a professional scouting program for Integrated Pest
Management (IPM). All three of the farmers, however, were able to reduce their pesticide
application rates and reduce the number of times they had to spray their crops by using IPM.
Pheromone traps to catch insects were the most popular IPM technique. All three farmers
had practiced IPM for at least six years on 25% to 100% of their properties.
Regarding cultivation practices, no-till farming is very popular - 85% of the
respondents who farm use no-till practices, compared to 59% of the respondents who till
conventionally with a moldboard plow.
Only seven farms irrigate their crops. These farms only irrigate 50 acres or less,
except for one farm which irrigates 170 acres.
Most of the farms use nitrogen fertilizer on their crops but usually do not store the
fertilizer for more than one month on the property. For those farms who do store fertilizer
more than one month, the amount stored ranges between 7 and 40,000 pounds. Almost all
the storage sites are over 100 feet from the property's drinking water source. Some fertilizer
application areas, however, are as close as one foot from the drinking water well or spring;
79
-------
20% are less than or equal to a distance of 75 feet.
Over half the farms have their soil tested every year. Of the remaining farmers, 20
test every other year, 11 test every three years, and 11 others test between every three to
five years. Five farmers do not test their soil at all. Only 15 farmers had tested their
livestock manure for it's nutrient value within the past five years. Most of the farmers
(67%) use commercial soil test recommendations to decide on the amount of nitrogen
fertilizer to apply to their crops. The other most popular sources of information that are
used to make fertilizer application rate decisions are (in descending order): personal
judgement, yield goals, agricultural dealer recommendations and University Extension
Service recommendations. Only 20 farmers said they were able to decrease their fertilizer
application rates due to the results of testing soil, manure or plant tissue.
More farms had beef cattle than any other type of livestock, followed by dairy cattle,
horses, hogs, sheep and chickens. Most of the farms with dairy cattle used manure storage
structures on the property, but only about 30% of the beef cattle farms used manure storage
structures. The most common type of manure storage structure is concrete liquid, followed
by solid and earthen liquid structures. The distance between manure storage structures and
drinking water sources ranged from 75 to 2,900 feet; 60% were 500 feet or less from the
drinking water source on the property. Of the 33 farms which use manure storage
structures, only three farms had a higher number of livestock using the structure than it was
designed to handle. Two farms were dairy farms and one raised beef cattle. For these three
farms, the structures were cleaned out at intervals ranging from three to six months, and the
distances to the drinking water sources on the property were 100, 300 and 500 feet.
The storage of petroleum products on the properties was the topic of the final part of
the interview. More respondents stored diesel fuel on their property than any other type of
petroleum product, followed by gasoline, heating oil, lubricating oils, kerosene and propane.
There are 89 farms which have one or more underground storage tank on the property.
Most of the underground tanks store 1,000 gallons or less, and are made of bare steel. Out
of those who knew what their tank was made of, 66% reported bare steel, and 30% said
their tank was made of coated steel. Over half the bare steel tanks have been in the ground
15 years or longer. Approximately half of all the underground storage tanks are 150 feet or
less from the property's drinking water source. Four farmers stated they had a leak or spill
at some time from the underground tank on their property.
Part II: Observation Record:
For this part of the survey, conditions were observed around 91 wells and 8 springs
which serve as primary drinking water sources for the interviewed farmers. The topographic
setting for most of the well and spring sites was either a hilltop, hillside or level ground.
Only three sites were in a depression. At seven sites the well or spring was open at the
surface, which indicates no well cap or spring house. At five sites the only protection for
the well or spring was either a metal grate, a metal lid or a partial concrete block
80
-------
impoundment.
At two sites, obvious misuse of pesticides around the drinking water supply was
observed. At one site pesticide containers were lying around nearby and at another site
agricultural chemicals are applied right over the well.
The most commonly observed features within 200 feet of the well or spring (in
descending order) are as follows: cropland, septic systems, petroleum storage tanks,
livestock confinement areas, and pesticide storage areas. In addition, one manure storage
structure, five pesticide equipment cleaning areas, six mixing and loading areas and two
pesticide disposal sites were observed within 200 feet of the primary drinking water source
for the property. Respondents hi general gave greater distances of features from their
drinking water source during the interviews than the enumerators observed hi the field.
81
-------
CHAPTER 5: USING CIS TO IDENTIFY AREAS SUSCEPTIBLE TO
PESTICIDE CONTAMINATION OF GROUND WATER
A. NONPOINT SOURCE THREATS TO GROUND WATER FROM LE.ACHING
B. POINT SOURCE THREATS TO GROUND WATER AND DRINKING WATER
WELLS/SPRINGS FROM CONDUIT FLOW RECEPTORS
The following is a description of the types of analyses conducted in Jefferson County
using GIS. These analyses combine land use, soils and hydrogeological data to identify areas
where ground water is susceptible to pesticide contamination. There are two general
categories of analyses. Section A focuses on nonpoint sources of pesticides and normal
leaching through the unsaturated zone. Section B analyzes the potential for point sources of
pesticides to enter ground water through conduit receptors, such as abandoned wells or
sinkholes, and an evaluation is made regarding the risk posed to drinking water sources.
Both categories of analyses include a series of steps taken to refine the analysis from an
initial first stage screening to a focus on just those areas which meet a more stringent set of
criteria, hi order to assess potential risk.
A. Nonpoint Source Threats to Ground Water from Normal Leaching
The first step of this analysis is to use the overlay procedure to select just the portions
of parcels which are cropped. This is done by intersecting the polygons classified as
cropland, grazing or research (agriculture experimental stations) from the land use data
coverage with the farm parcel coverage. Map number 28 shows the cropped portion of all
the farm parcels hi the County.
The second step combines soils data with pesticide use data, by first identifying where
priority leaching pesticides are used hi relationship to a soils map which has been classified
into different categories of soil leaching potential. This classification is based on the SPISP
technique developed by USD A, which was discussed hi Chapter 3. Instead of just selecting
those areas where the polygons of interest intersect, map number 29 shows a physical
overlay of the coverage of parcels which use priority leachers with the coverage of soil
leaching potentials.
The third step of this analysis is to identify parcels that use priority leaching
pesticides (as defined hi Chapter 3), and classify the usage for the 1989 growing season as
high (> 1,000 Ibs. active ingredient/farm), medium (500 - 999 Ibs. active ingredient/farm),
or low (0 - 499 Ibs. active ingredient/farm). Next, areas where high, medium or low use of
priority leaching pesticides occurs hi combination with intermediate, intermediate/nominal
and nominal soil classifications are selected. Map number 30 portrays the selected areas and
Table 4 provides a key to the map colors associated with the different combinations of soils
and pesticide use.
82
-------
Ground Water Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, West Virginia
j^B N om i n 11
[ I 1 n I e rme d i 11 e
|p| Nomini1/1ntermediite
Non-Soil Areas
teacher Use
\
Region III
on-Point Source Threats to Ground later - Normal Leaching
Leacher Use Locations / Soil Potential -
July. 1992 I
Man- JC029
Data Source: SCS & AP Practices Survey
-------
Ground Water Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, West Virginia
SOILS
Nominal
N om/ I n t
Intermedi a te
LOADINGS
Low lied High
u
Region II
on-Point Source Threats to Ground later - Normal Leaching
Leacher Loadings / Soil Potential - 1989
Mao: JC030 Data Source: SCS & A£ Practices Survey
-------
Table 4
SOIL LEACHING
POTENTIAL
Nominal
Nominal/
Intermediate
Intermediate
PRIORITY LEACHING PESTICIDE USE '
Low
Blue
Blue
Yellow
Medium
Blue
Yellow
Red
High
Yellow
Red
Red
Areas in red have the greatest potential for leaching, due to die combination of
intermediate soil leaching potentials and high amounts of priority leaching pesticides used in
these areas. Yellow represents a medium threat and blue the lowest threat. There are at
least 10 parcels where a large portion of the parcel shows up red, indicating a potential
problem area for pesticide leaching. By looking at maps 22, 29 and 30 together, it is clear
that 8 of the 10 farms which show up red hi map number 30 have soils with an intermediate
potential for leaching, yielding an overall Potential 1 (high) when high or medium quantities
of priority leachers are used on these soils (see Table 5). Although the other 2 farms have
soils with a lower leaching potential, the high amount of priority leachers used on these soils
causes these areas to show up red on map number 30. Overall, the highest threat areas show
up in the middle and eastern parts of the carbonate valley and the lowest threat areas are
concentrated in the southwestern part of the County.
The next SPISP evaluation looks at diversity of pesticide use. The first step identifies
where priority leaching pesticides are used, then displays the diversity of use (for 1989)
relation to soil leaching potentials. Since there are often several pesticides with high leaching
potentials used on one parcel, the number of pesticides is reflected in the intensity of the
color used to shade the polygons. The colors associated with the different combinations of
soil loss potentials and diversity of use of priority leaching pesticides is as follows:
Soil loss Potential 1 with
pesticide diversity of 1 -3 (pale pink), or
pesticide diversity of 4 - 6 (medium pink), or
pesticide diversity of 7 - 9 (red)
Soil loss Potential 2 with
pesticide diversity of 1 - 3 (orange), or
pesticide diversity of 4 - 6 (dark yellow), or
pesticide diversity of 7 - 9 (yellow)
85
-------
Soil loss Potential 3 with
pesticide diversity of 1 - 3 (blue), or
pesticide diversity of 4 - 6 (dark gray), or
pesticide diversity of 7 - 9 (green)
For example, as demonstrated by map number 31, if 7 - 9 pesticides with a high
ranking for leaching potential are used over a soil ranked intermediate for leaching loss
potential, the shade of red will be deep red, compared to a pale red if only 1-3 pesticides
are used over the same soil type. Diversity of pesticide use is fairly high (4-6 priority
leaching pesticides per farm) on 7 of the 10 parcels which had medium to high pesticide
loadings in the previous analysis. However, there is even more diversity of pesticide use on
several of the farms which showed up yellow on the previous analysis. These farms had
overall loadings that ranged from low to moderate to high. Some of the farms with low
loadings have a high diversity of pesticides used, which shows it is good to utilize both maps
30 and 31 together.
The previous analyses are variations of how SPISP was designed to be used, since a
group of pesticides (priority leachers) were all analyzed together. The next step is to
determine leaching potentials for specific pesticide and soil combinations, which is necessary
if one wants to compare the potential risks associated with using one pesticide versus using
another pesticide. A composite ranking for the polygons where specific pesticide use and
soil types intersect is obtained using the SPISP ranking scheme, as it was designed by SCS to
be used. For example, a pesticidersuch as atrazine has a large leaching potential, and a soil
type such as Frankstown shaly silt loam has an intermediate leaching potential. When
atrazine is used over Frankstown shaly silt loams, the composite or overall SPISP leaching
potential is called Potential 1, which is considered to give a high probability of pesticide loss
from the field. A medium pesticide loss potential combined with an intermediate soil loss
potential yields a Potential 2, which is an intermediate probability of pesticide loss from the
field. A matrix has been developed by USD A to combine the pesticide and soil rankings
manually, for use in conservation planning with farmers. Jefferson County SCS is using the
matrix to advise farmers on alternative pesticides to use which have lower leaching
potentials. Table 5 provides an example of how the SPISP matrix works.
86
-------
Ground Water Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, West Virginia
SP1SP
Potential
1 2 3
w
u
Region III
!on-Point Source Threats to Ground later - Normal Leaching
Diversity of Leacher Use / Soil Potential - 1989
Mao: JC031
Data Source: SCS & A? Practices Survey
July, 1992
-------
Table 5
SOIL
Duf-
F,^\A
tieiu
silt
loam
Lind-
side
silt
loam
CROP
Corn
Cora
TARGET
PEST
Foxtail,
Pigweed
and
Lambs-
quarter
Corn
Root-
worm
PESTICIDE
APPLIED
Atrazine
Cyanazine
EPTC
Terbufos
Fonofos
Carbo-furan
PESTICIDE
LOSS
POTENTIAL
L
M
S
S
S
L
SOIL LOSS
POTENTIAL
I
I
I
L
L
L
OVER-
ALL
SCORE
1
2
3
3
.3
2
PESTICIDE:
L = LARGE
M = MEDIUM
S = SMALL
X = EXTRA SMALL
SOIL:
H = HIGH
I = INTERMEDIATE
L = LOW
V = VERY LOW
OVERALL SCORE:
1 = HIGH
2 = MEDIUM
3 = LOW
4 = VERY LOW
As noted in Chapter 3, the soil ratings used for Jefferson County are based on the
original version of SPISP. The newer version (SPISP II) has categories for both adsorbed
and solution losses of pesticides for surface runoff. In addition, SPISP n, as can be seen in
Table 5, changed the soil leaching categories to high, intermediate, low and very low.
Along with the category changes, the algorithms used to derive the soil ratings changed as
well. Because of this, it is very important not to mix SPISP I pesticide ratings with SPISP II
soil ratings, and vice versa. At the time of the Jefferson County study, SPISP n pesticide
ratings were available and used, however, SPISP II soil ratings were not available from the
Ames, Iowa data base. Therefore, the pesticide leaching algorithms were evaluated to
determine differences between the old and new versions of SPISP. It was found that the only
change was the development of the new category "extra small" for pesticide leaching.
Therefore, SPISP version n pesticide ratings for the large, medium and small categories
were used with SPISP version I soil ratings for the Jefferson County study.
Instead of presenting a separate coverage for each pesticide used in the survey, this
report presents an SPISP analysis which was done just for atrazine. Atrazine was chosen as
an example, since it had higher use than any of the other priority leaching pesticides which
were used in the County in 1989. Map number 32 shows the composite SPISP scores, or
-------
Ground Water Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, West Virginia
on-Point Source Threats to Ground later - Normal Leaching
ATRAZINE Use / Soil Potential - 1989
Data Source: SCS & A
-------
leaching potentials for locations where atrazine was used.
Even though atrazine has a high (or large) leaching loss potential, if it is used over
soils ranked nominal (by SPISP I) or low (by SPISP H) for leaching, the overall potential is
medium, which is considered Potential 2. Atrazine use over soils ranked intermediate for
leaching yields a Potential 1, or high loss potential. If soils were present in the County with
a high leachability rating, atrazine use over those soils would also result in a Potential 1
score. Table 6 presents the SPISP I leaching potential matrix.
Table 6
SOIL
LEACHING
POTENTIAL
HIGH
INTERMEDIATE
NOMINAL
PESTICIDE LEACHING POTENTIAL
LARGE
1
1
2
MEDIUM
1
2
3
SMALL
2
3
3
As described in Chapter 3, there is an intermediate/nominal rating for much of the
soils in the western half of the carbonate valley in Jefferson County. This is due to the fact
that the mapped soil units in this area consist of two different soil types. In many cases the
soils present include characteristics of both soil types. Because of the presence of the
intermediate/nominal soils rating in Jefferson County, the following matrix (Table 7) was
used to assign ratings and color codes for Map 32. This variation of SPISP was chosen hi
order to differentiate between the nominal and intermediate/nominal soil ratings.
Table 7
PESTICIDE
LEACHING
POTENTIAL
L
L
L
SOIL LEACHING
POTENTIAL
Nominal
Intermediate/Nominal
Intermediate
MAP
COLOR
Blue
Yellow
Red
POTENTIAL
3
2
1
Map 32 displays how the paper matrix process can be automated; by using GIS all
the variations in pesticide use and soil type which exist hi the County can be evaluated, so
farmers can see maps of the best combinations of pesticides to use on soil found in their
fields. On map number 32, parts or all of at least 20 farms show up as red; these are
locations where soils have an intermediate leaching potential (the highest leaching potential
90
-------
found in the County), and where atrazine is used. Since Atrazine ranks high for leaching,
the composite soil/pesticide rankings in these areas is high (Potential 1). All or parts of
another approximately 20 farms show up with medium leaching potentials (Potential 2).
Very few farms have low leaching potentials (blue) when atrazine use is considered.
The SPISP procedure is a screening technique; it does not quantify concentrations
and it does not address every variable that affects pesticide losses from farm fields. It does,
however, evaluate alternative pesticides, focusing on initial migration of a compound to the
root zone, which allows managers to make well informed decisions to prevent leaching to
ground water. In addition, SPISP allows for a much more detailed analysis than DRASTIC
(as described below).
The next maps display the DRASTIC method in combination with pesticide use
information. DRASTIC, as described in Chapter 3, is a method developed by EPA to
classify the relative sensitivity of an aquifer to contamination, based on the hydrogeologic,
soil and topographic characteristics of a site. The objective is to identify areas'where ground
water is vulnerable to pesticide contamination. In order to accomplish this objective, first
the coverage showing locations of priority leaching pesticide use is overlaid onto the
DRASTIC map, as illustrated by map number 33. Second, as illustrated by map number 34,
polygons of high (> 1,000 Ibs. of active ingredient/farm), medium (500 to 999 Ibs. of active
ingredient/farm) and low (0 - 499 Ibs. of active ingredient/farm) application quantities of
priority leaching pesticides are intersected with polygons scored high, medium and low by
the DRASTIC method. Map number 34 shows that the majority of the areas ranked highest
(red) are in a center strip of the County, and there is some similarity to the areas ranked high
by SPISP (map number 30).
Even though the western part of the County ranks high for DRASTIC, leacher use is
low enough to keep the composite score medium to low. If just those areas where high or
medium use and high or medium DRASTIC occur (red polygons) are compared to just those
areas ranked highest by the SPISP analysis (map number 30), there are 10 parcels where
identical parts of the same parcels rank high (red) for both analyses. By comparing the
results of the soil teachability analysis (SPISP) with the ground water vulnerability analysis
(DRASTIC), it is possible to target the areas displayed in red, where pesticides have the
greatest potential not only to leach beyond the root zone, (as identified by the SPISP
analysis), but also to reach ground water, (as identified by the DRASTIC analysis).
The final stage of identifying sites where there are nonpoint source pesticide threats to
ground water via normal leaching, is to display the relative depth to ground water across the
County in conjunction with pesticide usage information. Map number 35 shows the locations
of where priority leaching pesticides are used in relation to the depth to water polygons.
The depth to ground water polygons were created using data on water level
measurements made in 196 wells in September 1974 by USGS (12). The measured wells
were digitized from a mylar used to create the map entitled "Ground Water Hydrology of
91
-------
Ground Water Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, West Virginia
teacher Uie
Region
on
-Point Source Threats to Ground fater - Normal Leaching
Leacher Locations / DRASTIC - 1989
i- JP033
Data Source: WV DNR & EPA Ae Practices Survey July. 1992
-------
Ground Water Vulnerability to Contamination bv Agricultural Chemicals
^ u O
Jefferson County, West Virginia
DRASTIC
0 119
120 179
LOADINGS
Low Med High
w
o
Region 111
!on-Point Source Threats lo Ground Water - Normal Leaching
Leacher Loadings / DRASTIC - 1989
Data Source: IV DNR & EPA As Practices _S_urvej
July, 1992
-------
Ground Water Vulnerability to Contamination by Agricultural Chemicals
^ Jefferson County, West Virginia
0
3 1
4 6
6 1
76
101
3 0
4 5
6 0
7 5
1 00
.292
DEPTH TO
GROUND IATER (FT
Leacher Use
\
Region 111
Non-Point Source Threats to Ground Water - Normal Leaching
Leacher Use Locations / Depth to Ground Water
TS Data Source: U.S.G.S & EPA Af Practices Survey July, 1992
-------
Jefferson County, West Virginia" (Hobba, 1978). In ARC/INFO, each well point had an
associated attribute file for depth to water found hi that well. ARC/INFO GRID was used to
interpolate between the measured points to create the depth to ground water map. Utilizing
the depth data associated with each well, the inverse distance weighted command was' used to
assign values of depth to ground water within individual grids across the County. Using this
command, the computer is given thresholds for maximum distances to search for neighboring
points, and weights to assign to values based on their distance from the primary points. The
result is an interpolated surface of different polygons, with each polygon having a known
value of depth to ground water (additional information on the creation of this map can be
found hi Appendix B).
It should be noted that there are three primary areas on map 35 where ground water
is found at depths of 200 feet or greater. These areas include the two quarries found along
the Shenandoah River, and the cavernous zone at the northern tip of the County. The water
table is artificially drawn down at the quarries due to man made conditions. At the northern
tip of the County, as the Potomac River continually cuts a deeper channel into the limestone,
the hydraulic gradient hi the vicinity of the river becomes steeper, resulting hi higher
velocities and increased dissolution of the limestone adjacent to the river. Due to this natural
process, springs pour out of caves as waterfalls into the river and the water table is lowered
in the area.
Map number 36 shows polygons hi red where high or medium loadings of priority
leachers intersect zones of shallow (45 feet or less below land surface), or medium (46 - 75
feet below land surface) depth ground water. Since the distance to ground water influences
the degree of attenuation of leaching pesticides, as well as the length of time it may take for
the pesticide or its degradate to reach ground water, this analysis selects those sites where a
pesticide is most likely to reach ground water, due to pesticide and site characteristics.
Approximately 10 parcels show up at least half shaded red, indicating a higher potential for
pesticides to reach ground water in these areas. Next, depth to ground water information is
combined with soil leachability SPISP rankings. Map number 37 shows the intersection of
the previous map with just those polygons where soils have the highest ranking for
leachability (intermediate loss potential).
This last step shows the intersection of polygons from three different data coverages,
depicting hi red where it is most likely for pesticides not only to leach below the root zone,
but also to reach ground water. Depth to ground water was selected from the other
DRASTIC parameters for this analysis because it has a high weighting (5) and relatively site
specific data is available for it. This map shows very few areas in red, as compared to the
10 parcels described earlier which ranked high for both SPISP potential (map number 30),
DRASTIC (map number 34), and priority leaching pesticide use. Both methods are an
attempt to identify where pesticides are likely to migrate below the root zone to ground
water.
95
-------
Ground Water Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, West Virginia
DEPTH
0 45
46 75
76 - 292
LOADINGS
flijh Ued Low
w
o
Region III
Non-Point Source Threats to Ground later - Normal Leaching
Leacher Loadings / Depth to Ground later
Map: JCQ36 Data bource: U.S.G.S & EPA Af Practices Survey July, 1992
-------
Ground tfater Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, West Virginia
DEPTH
0 - 45
46 - 75
76 292
LOADINGS
High Med Lo*
Region III
Non-Point Source Threats to Ground later - Normal Leaching
Leacher Loadings / Depth to Ground Water
Soil Potential 1
Mao:
Data Source: U.S.G.S k EPA Ag Practices Surve
-------
B. Point Source Threats to Ground Water and Drinking Water Wells/Springs
from Conduit Flow Receptors
The first step of this analysis is to identify all primary drinking water wells and
springs which have a potential for pesticide contamination from point sources. Data used for
the analysis comes from the Agricultural Practices Survey. A pesticide point source is
described as a pesticide storage, disposal, mixing and loading, equipment cleaning or spill
site. This analysis is portrayed by map number 38. Any well or spring which appears on
this map has a pesticide point source within 1,000 feet. For comparative purposes, map
number 39 shows wells or springs with pesticide point sources within 200 feet. There are 33
wells and two springs with point sources within 1000 feet; 20 of the wells and one of the
springs have point sources within 200 feet. The radius of 1,000 feet was chosen for this
analysis because USGS studies hi the County document conduit flow rates through fracture
and fault channels as high as 840 feet per day (15). The radius of 200 feet was also chosen
since this was the area evaluated by the Observation Record portion of die Agricultural
Practices Survey.
The next step of the analysis, also shown by map number 38 is to identify just those
parcels where any of the pesticide point sources meet any of the following criteria, in
addition to being 1,000 feet or less from the well or spring:
1. Priority leachers are stored at the site.
2. The pesticide storage facility is a partially enclosed building with a dirt floor.
3. The disposal site has buried pesticides.
4. The disposal site has pesticide containers.
5. The pesticide mixing and loading site has a permeable surface, such as bare
ground, grass, stone or gravel.
6. Obvious misuse of pesticides around the well or spring was noted by the
enumerator.
If any of the above criteria are met, the feature is considered to be a "poor pesticide
point source". Map number 38 shows there are 27 wells and two springs with a poor point
source within 1,000 feet, and map number 39 shows 14 wells with poor point sources within
200 feet. Map number 40 shows the locations of six properties which provided information
on pesticide disposal locations hi relation to the primary drinking water well or spring. All
the properties dispose of pesticides or containers by burying them hi the ground on the farm,
with the exception.of one of the properties, where the containers are burned. There are three
disposal sites within 1000 feet of the primary drinking water well or spring; two sites
between 1000 and 2,640 feet and one site that is one mile away.
98
-------
Ground Water Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, West Virginia
fells Springs
Point Sources
Poor Point Sources
u
Point Source Threats to Drinking Water
If ells or Springs With Pesticide
Point Sources Within 1000
Mao: JC038
Data Source: EPA A? Practices Survey
July, 1992
-------
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^HHMHil^^M
Ground Water Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, tfest Virginia
fells Springs
Point Sources
Poor Point Sources
°
Region III
Point Source Threats to Drinking later
ffells or Springs With Pesticide
Point Sources ffithin 200'
Map: JC039
Data Source: EPA A? Practices Survey
July, 1992
-------
Ground Water Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, tfest Virginia
1000' or Less
1000' 2640
More than 2640
w
o
Region II
Point Source Threats to Drinking later
Proximity of Pesticide Disposal Sites
to Pr imary We 11 or Spring
Mao: JC040
Data Source: EPA AE Practices Survey
July, 1992
-------
The third stage of this analysis is to identify those sites which use a poorly
constructed well or a vulnerable spring as a drinking water source, since the wells and
springs pose a potential threat to human health. The attributes for the drinking water source
point data were obtained through the Agricultural Practices Survey. Map number 41 shows
just those wells which meet the definition of "poorly constructed", or springs which are
inadequately protected. In order for a well or spring to appear on this map, it must meet one
or more of the following criteria:
* Well is not grouted, or only gravel is used for grout
* Well is not capped or spring is not covered
*
Inadequate cap on well, or cover on spring (metal grate, broken concrete block
cover, etc.)
* No casing for well
* Inadequate casing for well (tile, brick, or stone)
* Well is 50 feet deep or less
* Well was drilled (or dug) between 1832 and 1983
The significance of whether a well was drilled prior to 1984 is due to the fact that the
County passed an ordinance requiring new wells to be grouted in 1984. Approximately 83%
of the wells (from 64 responses) were drilled prior to 1984.
From map 41 it can be seen that there are 63 wells and four springs which show up
as vulnerable due to construction, age, depth and/or degree of protection. There were four
wells which were dug between 1832 and 1915. Two of these wells are less than or equal to
50 feet hi depth. Another two newer wells are also less than or equal to 50 feet hi depth.
Four wells have no cap and three springs have no cover. Four additional wells and one
spring are poorly protected at the surface by materials such as loose fitting metal grates or
lids. Three wells have no casing, another three wells have poor casing, such as tile, brick or
stone. At least 28 wells have no grout.
It is important to note that while there is no particular spatial pattern associated with
the most vulnerable wells and springs, there are several sites which have multiple
characteristics of vulnerability. Examples include one well with no casing, less than or equal
to 50 foot depth, and construction before 1916. Another example is a well with no cap, no
grout and constructed before 1983. Sites where wells and springs have either no protective
cover or a poor cover, and those sites with multiple vulnerable attributes are high priorities
for providing owners with drinking water source protection assistance.
The fourth stage of this analysis identifies potential conduits for contamination to
enter aquifers used for drinking water. First, those parcels which have an abandoned well
within 200 feet of the primary drinking water source are identified. This information is
based on responses given by well owners in the Agricultural Practices Survey. Abandoned
wells, particularly if improperly sealed, can serve as conduits for contamination of ground
water and nearby drinking water sources. Abandoned wells are displayed as red dots on map
102
-------
Ground Water Vulnerability to Contamination by Agricultural Chemicals
Jefferson County, West Virginia
fell Cap / Spring Cover
N o o e
Poor
Well Grout
None
O Gravel
We 11 Casing
; None
6"
1 o o r
Depth
0' -50'
50 ' +
Con>s true ted
Before 1916 - After
1916 1983 1983
X
Spring
X
N/A
Region 111
Vulnerable fells and Springs
Mao: JC041
Data Source: Af Practices Survey
July, 139 !
-------
42. Next those parcels which utilize a poorly constructed well as a drinking water source are
identified. Poorly constructed wells which can serve as conduits for contamination are
displayed as black X's on map 42.
The next stage of this analysis involves identifying other possible avenues of
contamination to ground water through conduit receptors within 1000 feet of the primary
drinking water well or spring. This stage uses the proximity operation in ARC/INFO to
buffer the area within a 1000 foot radius of the well or spring. Then the system searches for
features such as sinkholes, faults or fractures within the buffered area. These features are
found on coverages developed by USGS and the West Virginia Geologic and Economic
Survey, as described in Chapter 4. Map number 42 displays parcels which meet the sinkhole
criteria with a small red circle and those that meet the faults and fractures criteria with a
larger red circle. Six parcels meet the sinkhole criteria and 41 parcels meet the faults and
fractures criteria. Overall, the most commonly observed potential conduit receptors are
poorly constructed wells, followed by faults and fractures, then sinkholes and abandoned
wells.
Once the potential conduit receptors are identified the final stage of the analysis is to
locate potential sources of pesticide contamination within 1000 feet of the drinking water
source and evaluate the proximity of the contaminant sources to the conduit receptors and to
the drinking water sources. Map 42 displays sites with "poor potential point sources of
pesticides" (as defined in the beginning of Section B) with a black cross.
There are at least 25 sites selected which have potential point sources of pesticide
contamination in addition to potential conduit receptors - all within 1000 feet of the drinking
water source for the property. Out of the surveyed farms, these sites have the highest
likelihood of pesticide contamination of the drinking water source via conduit receptors. At
least seven sites have four or five of the criteria showing up all on one site. For example,
one site has a vulnerable well, a poor pesticide point source, an abandoned well, sinkhole(s),
fault(s) and fracture(s) all within 1000 feet of the drinking water source. At a minimum,
drinking water wells and springs at all 25 sites should be sampled for the presence of
pesticides used within the buffered area, with an emphasis on those sites which meet multiple
citeria, including pesticide point sources.
There are several different analytical techniques used in the decision making process
to select the final areas. The first through fifth stages use the logical, or Boolean operation
(18). This procedure selects only those sites which meet specific criteria, such as (a), (b),(c)
OR (d), or (a),(b),(c), AND (d). Since the enumerators measured and recorded distances of
pesticide point sources in the field, and queried respondents on distances, GIS does not have
to do measurement analyses for the first two stages. The final stage uses the proximity
analyses. Within a 1,000 foot buffer of the wells or springs, the system searches for points
(sinkholes) and lines (faults and fractures) from other data layers, to identify whether they
are present. This proximity analysis can only be done within a buffered area of a known
location, such as the point which marks the location of the well or spring. Of the 98 farms
104
-------
Ground Water Vulnerability to Contamination by Agricultural Chemicals
Je f fer son County, West Virginia
Condui t Receptors
Abandoned fell within 200'
S i nkho I e s wi Ihi n 1000'
Fau] Is 4 Fractures »ithin 1000
X Vulnerable fell
-)- Poor Pesticide Point Source
*Z.
C3
Reg i o'n III
Point Source Threats to Ground Water
Conduit Receptors
JC042 Data Sources: yVG&ES. USGS, 4 EPA Ag Practices Survey July, 1992
-------
which use a well or spring as their drinking water source, there are recorded locations for 87
of these sites. Therefore, if faults, fractures or sinkholes exist within 1,000 feet of the 11
well or spring sites that have not been digitized, they will not show up on map number 46.
GIS can be used to compare the frequency of different attributes which appear in
conjunction with each other, such as monitoring data versus well and spring attributes. For
example, queries such as "what was the frequency at which pesticides were detected in wells
that were less than 50 feet deep versus wells greater than 50 feet deep?" can be answered.
For a complete tabulation of frequency distributions of attribute data versus monitoring data,
please refer to Chapter 6.
106
-------
CHAPTER 6: SUMMARY OF PROBLEMS IDENTIFIED, MANAGEMENT
RECOMMENDATIONS AND INSTITUTIONAL RESPONSE TO
THE PROBLEMS
A. POTENTIAL PROBLEM AREAS IN JEFFERSON COUNTY IDENTIFIED
THROUGH THE AGRICULTURAL PRACTICES SURVEY
B. MANAGEMENT RECOMMENDATIONS TO ADDRESS PROBLEMS
IDENTIFIED BY THE SURVEY
C. POTENTIAL PROBLEM AREAS IN JEFFERSON COUNTY IDENTIFIED BY
THE CIS ANALYSES, AND RECOMMENDED SOLUTIONS
D. INSTITUTIONAL RESPONSE: EPA's PESTICIDE REGISTRATION PROCESS,
PROPOSED RULES, REGULATIONS, POLICY, TECHNICAL AND FINANCIAL
ASSISTANCE, STATE AND LOCAL PROTECTION INITIATIVES
A. Potential Problem Areas in Jefferson County Identified through the Agricultural
Practices Survey
1. According to the respondents, 57% of the properties surveyed have a drinking
water well or spring which has at least one of the following characteristics: no
grout (28 wells), no cap or springhouse (7 sites), poor cap or springhouse (5
sites), no casing (3 wells), poor casing (3 wells), less than 50 feet deep (4
wells), drilled before or during 1915 (4 wells) or drilled prior to the 1984
grout ordinance (53 wells). These wells or springs are susceptible to
contamination due to a lack of surface protection, their age, construction
and/or depth.
2. 13.5% of the properties have an abandoned or non-operating well within 200
feet of the primary drinking water source; six wells are open at the surface
and only three of the wells have been filled in.
3. 20% of the fertilizer application areas are 75 feet or less from the drinking
water source; 40% of the pesticide application areas are 200 feet or less from
the drinking water source. Some application areas are as close as one foot
from the drinking water source.
4. Approximately 70% of the fanners use commercial applicators. Records kept
by these applicators were often incomplete; in particular, information on the
locations of tracts which were sprayed was usually missing. In addition, many
commercial applicators are not aware of locations of critical areas, such as
sinkholes and drinking water wells.
5. Only 26% of the farmers who use pesticides referred to records when
answering questions about pesticide use on the property. Many of the farmers
had difficulty providing information on formulations of pesticides and exact
107
-------
application rates.
6. A total of 65 different pesticides are stored by the fanners in the survey, and
some of the most commonly stored pesticides are priority leachers, such as
Dicamba, Carbaryl and 2,4-D.
7. 16% of the farmers who use pesticides dispose of 47 different unwanted
pesticides or containers either in a farm dump (over one third) or by burning
the containers or pesticides on the property. Five respondents take unwanted
pesticides to the landfill. Less than half of the farmers triple rinse containers
before disposing of them in the farm dump. Some of the pesticides commonly
disposed of are priority leachers, such as Atrazine, Carbofuran and .
Metolachlor. Many farmers expressed frustration over not knowing what to
do with unwanted pesticides or their containers.
8. Most pesticide mixing and loading takes place over permeable surfaces, such
as bare ground or grass.
9. Some equipment cleaning areas are as close as one foot from the drinking
water source, several equipment cleaning areas and mixing and loading areas
were observed within 200 feet of the drinking water source.
10. A total of 40,882 pounds of pesticides were applied to 105 surveyed parcels in
1989; priority leaching pesticides made up 63% of the total pesticide use.
11. Corn was identified as the crop which used the highest quantity of priority
leaching pesticides. The top three crops which relied most heavily on priority
leachers as a percentage of overall pesticide use were: truck crops, pasture
and corn. Other crops, such as apples, relied heavily on pesticides, but less
than 50% of overall use was of priority leaching pesticides. Other crops
which were less reliant on priority leachers included alfalfa, peaches, and
nectarines.
12. There are very few respondents (only 3) who participate in an Integrated Pest
Management program. Many farmers indicated that they were involved in an
initial EPM program in the County, but quit the program due to problems with
how it was managed. Information from the County Agricultural Agent
revealed that problems with the initial IPM program were due to the fact that
it relied heavily on college students who were taking final exams at the times
most critical for scouting. In addition, an artificially low cost was charged to
the producer.
13. Only 15 farmers test their livestock manure for its nutrient value and only 20
fanners decreased fertilizer application rates due to soil, manure or plant tissue
108
-------
testing recommendations.
14. Three farms had a higher number of livestock using their manure storage
structure than the structure was designed to handle; the distance of the
structures to the drinking water sources on the properties ranged from 100 to
500 feet.
15. Of the 56 fanners who described an abandoned or in-use underground storage
tank (UST) on their property, 66% have a tank which is made of corrosive
material (bare steel), and over half of these tanks have been in the ground 15
years or longer. These tanks pose a threat to ground water quality. Since 89
farms have one or more UST on the property, there are potentially more tanks
of concern, since only 63% of tank owners described their tanks. About half
of the USTs are 150 feet or less from the property's drinking water source.
Four farmers said they had a leak or spill from the tank.
16. At two sites, apparent misuse of pesticides was observed around the drinking
water source; at one site empty pesticide containers were lying near the well,
at another site agricultural chemicals were applied directly over the well.
B. Management Recommendations to Address Problems Identified by the Survey
1. Wells which are old, poorly constructed, and/or shallow, and all wells and
springs which are improperly protected at the surface should be tested (at a
minimum) for: pesticides, nitrates, bacteria and/or petroleum products,
depending on the potential sources of contamination on or near the site. Well
testing is critical at these drinking water sources to determine if there is any
human health risk. Farmers should be educated on: how to protect the
surface of the well or spring from contamination, whether it is in their best
interest to drill a new well, and whether they need to treat their drinking
water.
2. Abandoned or non-operating wells should be properly sealed so they do not
serve as potential conduits for contamination.
3. Farmers and commercial applicators need to be educated more extensively on
the need to protect critical areas, such as drinking water wells, sinkholes,
streams, abandoned wells, etc., during pesticide application, and hi overall
pesticide handling, such as mixing and loading, storage and disposal. °
4. Commercial Applicators should be required to keep consistent, uniform
recqrds for all pesticides applied, including the locations of where the
pesticides were applied.
109
-------
5. Farmers need to be educated on the importance of, and in some cases the
requirements for, keeping pesticide application records, including
formulations, active ingredients, application rates and locations where
applications were made.
6. Unwanted pesticides and containers should be collected from farmers and
disposed of properly. A program should be developed to support collection,
transport and disposal costs.
7. Fanners need to be educated regarding the proper methods of disposal of
leftover pesticides and their containers. Alternatives, such as buying products
in bulk quantity containers and using refillable and recyclable containers
should be encouraged. >ua-
.r iif-
8. Financial assistance to farmers to construct impermeable pesticide mixing and
loading pads with spill collection systems should be explored. Another
technique is the use of closed sprayer systems, combined with the practice of
adding chemicals to the system in the field. The degree of ground water
protection associated with either of the two practices should be explored.
9. The economic and environmental feasibility of major crop shifts and pesticide
use shifts hi the County to reduce reliance on priority leaching pesticides
should be evaluated. For example, if less corn and truck crops and more
wheat, alfalfa, and apples were grown, what would be the net effect on
farmer's incomes and the local economy? What are the environmental risks
and benefits associated with the corresponding shifts in pesticide use? Are
there feasible alternative chemical or biological control methods that can be
used on the current crops which pose less of a threat to ground water?
10. Farmers should be educated on pesticide equipment cleaning practices. In
particular, pesticide equipment should not be cleaned in close proximity to
drinking water wells, springs, sinkholes and other areas where ground water or
drinking water is susceptible to contamination.
11. Many farmers indicated an interest hi IPM, and discussed its benefits, but
stated they would only participate if the program was run well. A new IPM
program should be re-instated with a focus on preventing problems associated
with the earlier program from re-occurring.
12. More education on nutrient management is needed, to get higher participation
hi manure, soil and plant tissue testing programs, and to ensure that over-
fertilization is not occurring.
110
-------
13. Farms with animal waste structures that are filled to capacity or filled beyond
design standards should be visited, to ensure that the facilities are either
cleaned at frequent enough intervals to prevent contamination, or enlarged to
accommodate the livestock which use the structures.
14. Abandoned, leaking, or high-risk underground storage tanks should be
removed and replaced with non-metal or coated metal tanks. Drinking water
wells potentially at risk from leaking USTs should be sampled for petroleum
products.
C. Potential Problem Areas Identified in Jefferson County by the CIS Analyses, and
Recommended Solutions
1. Analysis A identified those areas where nonpoint sources of pesticides pose a
potential threat to ground water from normal leaching. Three different
methods were used to identify these areas: SPISP, DRASTIC, and depth to
ground water. All three methods were used in conjunction with survey data
on use of priority leaching pesticides. Although there were differences in
areas targeted by each method, a common theme was found hi all analyses:
surveyed farms with the highest potential for pesticide leaching tend to
concentrate in the middle and eastern parts of the carbonate valley, and
surveyed farms with the lowest leaching potential tend to concentrate in the
southwestern part of the carbonate valley.
Recommendations: Work with the farmers in the middle and eastern parts of
the carbonate valley where the highest risk of ground water contamination
from pesticides exists (based on all three methods of analysis and survey data)
to decrease the use of priority leachers. Decreased use of priority leachers
could be accomplished through a successful IPM program, investigating
alternative pesticides for crops currently grown and by encouraging farmers to
grow alternative crops which are less dependant on priority leachers. The
study indicates, for example, that priority leaching pesticides are used more to
cultivate corri than to grow apples and other orchard crops. While some
farmers may be interested in starting orchards, it is not reasonable to
recommend that a majority of the farmers make a shift from growing corn to
growing apples. Orchards take years to establish, and there are numerous
other factors to consider, such as market demand and environmental and
human health risks (other than through the ground water route of exposure)
associated with pesticides used on orchards. It would be worthwhile,
however, to further evaluate what crop shifts would be reasonable from both
an environmental and economic standpoint. For example, the study shows that
alfalfa and other types of hay, as well as small grains, such as wheat, are far
less reliant on priority leaching pesticides than com. For example, in 1989,
an average of 4.28 pounds of pesticides per acre was applied to corn, versus
111
-------
0.17 pounds per acre to wheat and 1.25 pounds per acre to alfalfa. The
percentage of the total average pesticide applications per acre per crop that
represent priority leacher use is as follows: 93% for corn versus 46% for
alfalfa and 53% for wheat. The data indicates that if it is economically
attractive to do so, farmers should make a shift to grow less corn and more
hay and small grains to reduce the likelihood of pesticide leaching to ground
water.
Growing the same crops but using alternative pesticides should be evaluated as
well. For example, leachers used in large quantities throughout the County,
such as atrazine and metolachlor, are prime candidates for replacement, if
possible, with pesticides that have a lower propensity to leach.
Since many of the farms targeted are in the eastern part of the valley near the
Shenandoah River, sampling of ground water in springs near the river, as well
as river base flow, should be conducted to identify whether pesticides are
migrating to the river.
In the southwestern part of the County, use of priority leaching pesticides is
lower and the soils are not as conducive to leaching. The final recommend-
ation from this analysis is to prevent any increase in the use of priority
leachers in this area. This is particularly important since DRASTIC scores are
. high and depth to water is shallow in the southwestern part of the County.
2. Analysis B identified at least 25 surveyed farm parcels where pesticide point
sources pose a potential threat to the farm's drinking water well or spring
through conduit receptors, such as sinkholes, abandoned wells, faults, fractures
or poorly constructed wells. These sites are scattered throughout the County,
although a higher percent of the sites appear to be found in the southern and
western parts of the carbonate valley, which can be partly attributed to the
high density of faults and sinkholes in this area. Most (76%) of the 25 parcels
have a poorly constructed well used for a drinking water source; over half of
the parcels have faults or fractures within 1,000 feet of the primary drinking
water well or spring; close to one third of the parcels have abandoned wells
within 200 feet and one fourth have sinkholes within 1,000 feet of the primary
drinking water source. All 25 sites have one or more of the following: a
pesticide mixing and loading, equipment cleaning, storage or disposal site
which are used or have been constructed in a way that poses a threat to ground
water.
Recommendations: The GIS analyses provide the first step for identifying
farms which are good candidate sites for field investigations. The analyses
show that, at a minimum, close to a quarter of the surveyed farms should be
visited in the field to investigate the area within 1,000 feet of the farm's
112
-------
drinking water source. "Farm-A-Syst", which was developed by the
Cooperative Extension Service at the University of Wisconsin, would be an
ideal tool to use to assess these sites. Farm-A-Syst utilizes worksheets to
assess the potential risk of drinking water contamination by scoring factors
such as pesticide handling practices, drinking water well construction, and
distances to potential contaminant sources. Ideally, personnel from the
Cooperative Extension Service, Soil Conservation Service and/or Conservation
Districts would help the farmers score the area around their drinking water
source. If the final scores indicate problems, the Farm-A-Syst worksheets can
be referred to, since they describe practices fanners can implement to protect
these drinking water sources. At a minimum, drinking water wells and
springs at the sites with the highest scores should be sampled for nitrates and
for any pesticides used within 1000 feet.
Many of the management recommendations described under Section B of this
chapter also apply here, such as the need to properly seal abandoned wells,
and to properly dispose of pesticides and containers. In addition, measures to
protect sinkholes, such as the installation of biofilters to filter contaminated
stormwater that enters sinkholes should be conducted where proven effective.
D. Institutional Response: EPA's Pesticide Registration Process, Proposed and
Existing Regulations, Policy, Technical and Financial Assistance, State and Local
Protection and Well Sampling Initiatives
EPA's primary response to the pesticides and ground water concern was to
develop the Pesticides and Ground Water Strategy in 1991 and the State Management
Plan Guidance and Technical Assistance Documents in 1993 and 1994. As described
hi Chapter 1, the main thrust of the Strategy consists of State implementation of
Management Plans to protect ground water. In addition to the State Management
Plan approach, however, there are many regulations, programs and policies which
address ground water concerns from a national perspective and are intended to
supplement, and in some cases be a part of, the State Pesticides and Ground Water
Management Plans. The following is a brief description (Section Dl) of federal
statutes, existing or proposed regulations, policies, programs, and technical and
financial assistance available to provide better protection of ground water from
pesticide contamination. Section D2 describes initiatives which the State and local
governments are pursuing to protect the ground water in Jefferson County as well as
the State of West Virginia from agricultural chemicals.
113
-------
Section Dl - EPA and other Federal Initiatives to Protect Ground Water from
Pesticide Contamination
1. Registration Process
EPA regulates the use of pesticides under the authority of FLFRA - the Federal
Insecticide, Fungicide, and Rodenticide Act. FIFRA gives EPA the authority
and responsibility for registering pesticides for specified uses. No pesticide may
be legally sold in the U.S. unless it is properly registered with EPA.
Approximately 25,000 formulated pesticide products that contain 750 different
active ingredients are registered for marketing and use hi the U.S. (5). FIFRA
requires EPA to balance human health and environmental risks against the
benefits of pesticide use to society and the economy when making registration
decisions. To make these decisions, EPA requires applicants to submit data on
acute and chronic toxicity, physical and chemical properties, and environmental
fate of the compound, to determine if the pesticide poses a threat to the
environment, including a threat to surface and ground water quality.
EPA is required by law to reregister existing pesticides that were originally
registered before current scientific and regulatory standards were established.
As of May 1991, EPA had issued registration standards for 350 active
ingredients which account for 85 to 90% of the total volume of pesticides used
hi the U.S. (5). A registration standard includes a comprehensive review of the
available data for the chemical, a list of additional data needed for full
reregistration and the Agency's current regulatory position on the pesticide. The
reregistration process may result in changes to the label, the registration, and/or
the tolerances set for food; it may also result hi the agency revoking the
registration of some pesticides, due to hazards associated with the use of the
pesticides which were not known at the original time of registration.
If EPA seeks to revoke a pesticide registration, the agency must first announce
its reasons and offer the manufacturer an opportunity for a formal hearing to
present opposing evidence. A less formal and more efficient approach, known
as the Special Review process, offers opportunities for interested parties on all
sides to comment and present evidence on the risks and benefits of a pesticide.
This process often results hi an agreement to modify the registration to
sufficiently reduce the risk associated with use of the pesticide, and avoids the
need for a formal hearing. If however, the issues cannot be resolved through
the Special Review process, EPA may issue a proposed notice of intent to cancel
the product. If a hearing is requested, it is conducted as a trial-like
administrative proceeding before an EPA Administrative Law Judge, who issues
a recommended decision to the EPA Administrator. The cancellation process
often takes two or more years. During this process, the product will stay jon the
market unless EPA issues a suspension order that bans the sale or use of the
114
-------
pesticide while the cancellation decision is under review. A suspension order is
issued in those cases where the continued sale and use of the product poses an
imminent hazard.
Through the registration, reregistration and Special Review process, EPA will
identify specific risks associated with the pesticides under review, including
threats to ground water quality. Product labels will be used to communicate to
applicators which pesticides can only be used in States with EPA approved
Pesticide and Ground Water Management Plans.
2. Labeling: It is a violation of FIFRA to use a pesticide in a manner inconsistent
with its label. There are several pesticide products which have special language
to address ground water concerns. Examples include atrazine and aldicarb
which have restrictions on mixing, loading or apply ing the product within a
specified distance from any well. At least 10 products have ground water
advisory statements on the label. Since 1978 States have been given the primary
responsibility to take enforcement action for pesticide use activities which are a
violation of the label. EPA provides oversight of the enforcement.
3. Pesticide Mixing. Loading. Transport. Storage and Disposal Rules: The 1988
amendments to FIFRA significantly expanded EPA's authority and responsibility
to regulate the packaging, storage, transportation and disposal of pesticides.
EPA has developed proposed regulations to address many of these areas. Phase
I of the proposed regulations includes requirements governing the storage,
transport and disposal of pesticides that have been recalled. Phase n focuses on
pesticide container standards. These standards address nonrefillable container
design, residue removal, re-using of containers, recycling and labeling
requirements. Refillable containers are also addressed by the rule, including
specifications on how these containers are refilled, the type of mixing and
loading facilities required, and record keeping. A definition of bulk storage of
pesticides and secondary containment standards are also included for commercial
facilities. Phase ffl regulations, which are expected to follow, will address
disposal, storage, transport, mixing and loading of all pesticides not covered
under Phase I.
4. Ground Water Restricted Use Rule: as part of the FIFRA amendments of 1972,
Congress gave EPA authority to classify certain pesticides for restricted use if
unrestricted use could cause unreasonable adverse effects on the environment or
to the applicator. The restrictions placed on these pesticides are designed to
protect health and the environment and avoid the need for national cancellation
of the product. EPA requires that restricted use pesticides be used only by or
under die direct supervision of a certified applicator. Certified applicators are
required to attend regular training sessions to maintain their certification
licenses.
115
-------
On April 29, 1991, EPA proposed to add new criteria for selection of pesticide
products as candidates for restricted use classification based on the potential of
the products to contaminate ground water. Criteria used in the current
regulations for restricted use classification of pesticides are based largely on the
acute toxicity a pesticide product presents to man or non-target wildlife, or the
chronic toxicity a pesticide may pose to man.
The proposed rule outlined two options: the first option would have added
criteria consisting of either the measured persistence and mobility of the
pesticide or the detection of the pesticide in ground water in at least three
different counties. The second option, which was chosen, considers the
pesticide's persistence and mobility and/or whether the pesticide has been
detected in ground water in at least three counties at levels greater than 10% of
the Maximum Contaminant Level (MCL) or lifetime Health Advisory established
under the Safe Drinking Water Act, or in 25 or more wells in 4 or more States.
5. Chemigation Rule: On March 11, 1987, EPA issued a notice to manufacturers,
formulators, producers and registrants of pesticide products requiring label
revisions to address chemigation for any pesticide product shipped after April
30, 1988. Chemigation is the application of pesticide products through irrigation
systems to crops. Risks associated with chemigation include direct human
exposure from the use of irrigation water unknowingly for domestic purposes, as
well as the risk of ground water contamination due to back-siphoning of the
irrigation water. The required label language addresses what types of irrigation
systems are acceptable for use in chemigation, prohibits the connection of a
chemigation system to a public water supply well system unless prescribed safety
devices are in place, and requires posting of the area to be treated. In addition,
there are extensive requirements addressing types of check valves and other
safety devices which different types of irrigation systems must have to prevent
the water source (such as a well or river) from becoming contaminated by
backflow.
6. Registrant Product Stewardship: As part of the implementation of EPA's
Pesticides and Ground Water Strategy, registrants will be required to conduct
more representative monitoring of ground water where pesticide use occurs in
areas that may be susceptible to contamination. Registrants will also be
expected to play a greater role in product stewardship, i.e., informing
distributors and applicators about how their products should be managed to
prevent degradation of ground water quality. An example of expanded
monitoring requirements is demonstrated by the National Alachlor Well Water
Survey (NAWWS) which was conducted by Monsanto from June 1987 to May
1989. A total of 1,430 wells from 89 counties in 26 States were sampled, to
comply with EPA's re-registration requirements for alachlor. The purpose of
the study was to obtain statistically representative estimates of alachlor
116
-------
occurrence in all rural wells in Counties where the product was used. The study
showed that less than 1% of the wells had alachlor detections, although atrazine
was present hi close to 12% of the wells.
7. Certification and Training: In cooperation with the States and USDA's
Cooperative Extension Service, EPA is improving training and certification
programs so that users are better trained on management measures to protect the
ground water resource. Both EPA and USDA fund, develop, and distribute
training materials for certified applicators. EPA grants help fund State extension
service training programs. Although certification requirements vary from State
to State, all States must meet minimum federal requirements established by
EPA.
8. Pesticide Use Record Keeping: There are currently over 100 federally
registered restricted use pesticides and approximately 1.25 million certified
applicators (5). Applicators include both "private" applicators (mostly farmers)
and "commercial" applicators, who are hired to apply pesticides. There are no
record keeping requirements under FIFRA for private or commercial applicators,
however, EPA has established guidance on minimum record keeping
requirements for commercial applicators. The guidance suggests the records
include application rates, location of application, crops treated, target pest, and
name and license number of the applicator. In addition, the 1990 Farm Bill does
require private applicators of restricted use pesticides to keep pesticide use
records, including product name, amount applied and location of application (see
item number 12 for additional record keeping requirements under the Farm Bill).
States can adopt more stringent record keeping requirements than EPA or USDA
(see Section D2).
FIFRA requires that restricted use pesticides only be sold to certified
applicators. States have developed regulations to ensure that this requirement is
met. Dealers have to keep records regarding the products they sell, the quantity
sold and who the product is sold to. In addition, many States have regulations
which require private and commercial applicators of restricted use pesticides to
keep records. EPA's record keeping requirements are being revised under
FIFRA to clarify and coordinate federal record keeping requirements (40 CFR
Part 171).
9. Pest Control Alternatives: EPA is encouraging the development of safer
chemical and non-chemical pest control alternatives and the adoption of
environmentally sound agricultural practices. As evidence of this support,
USDA and EPA jointly sponsored a workshop on Integrated Pest Management
(IPM) hi June of 1992 to discuss collaborative efforts to promote and implement
IPM methods nationwide. IPM is a system of preventing pest populations from
reaching damaging levels by predicting pest population outbreaks, and using
117
-------
natural control techniques such as introducing or encouraging natural predators,
planting pest resistant plants, rotating crops, and adjusting planting and
harvesting dates. When these practices fail to keep pests from becoming an
economic threat, carefully timed applications of small quantities of pesticides are
used for pest control. IPM is one part of Integrated Crop Management (ICM)
which involves the systematic use of economical practices that achieve relatively ,
high yields without damaging the environment. A major goal of ICM is to grow
plants that are vigorous enough to withstand pests and diseases. Increasing plant
vigor and protecting the environment is achieved through practices such as crop
rotations, conservation tillage, mulching, growing cover crops, optimizing soil
organic content and fertility and managing pests with minimal chemical usage.
The use of ICM practices will help to ensure crop production which is both
economically sound and environmentally safe.
EPA has registered over 20 naturally occurring microbial pesticides, which are
currently used in over 100 products in agriculture, forestry, mosquito control,
and home and garden applications. In addition, EPA has registered over 30
biochemical pesticides, that are either naturally occurring or identical to
naturally occurring substances, such as hormones, pheromones and enzymes (5).
10. Health Advisories and Drinking Water Standards: In 1989 EPA published health
advisories for 55 pesticides, which are used by federal, State and local officials
to provide guidance to well owners when incidents of pesticide contamination of
drinking water occur. The health advisories describe health risks associated with
drinking water containing specified concentrations of pesticides, establish safe
concentrations and provide guidance on treatment options. In addition, EPA has
established enforceable drinking water standards, or maximum contaminant
levels (MCLs) for 23 pesticides. EPA regulates public drinking water supplies
to ensure these levels are not exceeded.
11. EPA Grants: Under FEFRA, State lead agencies (usually State Departments of
Agriculture) receive funds to carry out pesticide management, training,
monitoring and enforcement programs. A portion of the grant is devoted to the
development and implementation of SMPs. In addition, under Section 106 of
the Clean Water Act, EPA provides grants to State environmental agencies to
develop Comprehensive Ground Water Protection Programs. Key elements of
these programs include characterizing the ground water resource, identifying
sources of potential contamination and controlling the sources. Part of the 106
grant is also used to support the development and implementation of State
Pesticides and Ground Water Management Plans (SMPs).
12. Farm Bill: The 1990 Farm Bill contains provisions for incentive payments to
farmers to implement plans to protect water quality. Only farmers with
approved Conservation Plans are eligible to have a Water Quality Incentive Plan
118
-------
developed for their farm. Water Quality Incentive Plans are long term
agreements (from three to five years), which are designed to protect surface and
ground water through measures such as efficient fertilizer use, IPM, safe
storage, and mixing and loading of agricultural chemicals. Farmers who receive
financial assistance to implement water quality protection plans on their
properties must comply with certain requirements, including the yearly reporting
of well test results, and fertilizer and pesticide application rates. Federal, State
and local agricultural agencies are responsible for ensuring compliance with the
provisions of the Farm Bill.
Section D2 - State and Local Initiatives to Protect Ground Water from Pesticide
Contamination in Jefferson County and West Virginia
1- Pesticide Disposal: West Virginia Department of Agriculture is developing a
pesticide container disposal outreach training program, and working to secure
funds to support a program to collect surplus unwanted pesticides.
2. Pesticide Use Record Keeping: At the time the pesticide use survey was taken
in Jefferson County, West Virginia law required commercial applicators of both
general and restricted use pesticides to keep records for a period of three years
from the date of application on the following information:
1. The pesticide used.
2. The formulation used, including the quantity of formulation.
3. The date and place of application.
4. The pest against which the pesticide was used.
5. The applicator's name.
At the time the survey was taken, applicators only had to record information on
amounts of specific pesticides applied for a property owner in order to comply
with the "place of application" requirement in the law. Since the records did not
have to specify locations of multiple properties owned by one person or
corporation, all applications made for one owner were often aggregated.
Proposed revisions to the State regulations will require records to be kept for
more precise locations.
3. Soil Conservation Service Technical and Financial Assistance: The Soil
Conservation Service's Jefferson County office has initiated the following
activities designed to address concerns raised by the Agricultural Practices.
Survey and GIS analysis:
119
-------
Sinkhole protection
* As of August 1993 signs have been placed in farmer's fields marking the
locations of 40 sinkholes. Vegetative filter strips are also placed around the
sinkholes to help trap sediment, nutrients and pesticides. A total of 250
signs have been made for placement in the field.
* Biofilters have been installed for three major sinkholes in the County. The
innovative design for biofilters, which originated from the Jefferson County
SCS field office, utilizes vegetation and a mixture of soil, gravel and fabric
to filter pollutants hi stormwater draining to sinkholes. A concrete collar
within the sinkhole supports the soil and gravel. SCS is currently
monitoring the sites to determine the effectiveness of the biofilters for
improving water quality.
Well Identification
* Abandoned, buried and dug wells which are not visible from the surface of
the ground can serve as conduits for pollutants to enter aquifers. These
wells are being marked with signs to identify their locations. In addition,
farmers are being educated to avoid using agricultural chemicals in a manner
that could cause entry of contaminants into these wells.
Pesticide Applications
* Using the SPISP rankings for soils (as described hi Chapters three and five),
fanners are advised regarding what pesticides are safest to use in order to
protect ground water. The assessment evaluates the combination of soils
found hi the farmer's fields and alternative pesticides available to meet crop
protection needs. SPISP rankings for soils are all automated and available
from the Jefferson County SCS office's GIS.
Fuel Storage Tanks
* A draft design standard was approved by the State of West Virginia for the
construction of above ground fuel storage tank spill collection systems. The
local SCS office is waiting for national SCS approval to use cost share funds
for the installation of these systems on farms.
Nitrogen Management
* As of August 1993, over 4,000 acres of cornfields in the County received
soil tests through the "nitrate quick test". Based on the results of these tests,
farmers are reducing their fertilizer applications, which results hi a reduction
120
-------
of soil nitrate available for leaching to ground water. This practice is saving
farmers money while protecting ground water.
4. Wellhead Protection: The Eastern Panhandle Regional Planning and
Development Council, Region 9, received a grant from EPA in 1992 to develop
a demonstration Wellhead Protection Program for Jefferson, Berkeley and
Morgan Counties. The goals of the project are to delineate protection areas for
public wells and springs, design proposed local and regional management
strategies to protect the areas and develop programs to generate revenue to
implement the overall protection plan.
5. Ground Water Sampling: Between September 7, 1992 and September 21, 1992
the West Virginia Department of Agriculture sampled 28 wells and two springs
in Jefferson County. The wells and springs were all on farms that were part of
the Agricultural Practices Survey. Properties with wells or springs that were
suspected to be vulnerable to pesticide contamination were selected for sampling.
Sampling in Jefferson County was conducted as part of an initial Statewide effort
to characterize both ambient ground water quality and potential "hot spots" hi
terms of the presence of pesticides. The results will be used to help implement
the State Pesticides and Ground Water Management Plan. The following
compounds were analyzed for:
1. Atrazine
2. Metolachlor
3. Cyanazine
4. Alachlor
5. 2,4-D
6. Triclopyr
7. Diazinon
8. 2,4,5-T
9. Linuron
10. .Chlorpyrifos
11. Simazine
Table 8 provides information regarding what compounds were detected,
concentrations detected, minimum detection levels and methods used for
analyses.
121
-------
TABLE 8
COMPOUND DETECTED
Atrazine
Metolachlor
Picloram
Triclopyr
Chlorpyrifos
RANGE IN CONCENTRATION
DETECTED lug/11
0.2-5.0
0.47 8.51
0.011 033
0.021
0.38 9.39
MCL
lug/11
3.0
100.0 IHAU
500.0
NA
20.0 IHAU
MINIMUM DETECTION LEVEL
lug/11
< 0.142
< 1.790
< 2.100
< 0.039
< 0.057
METHOD OF ANALYSIS
Method 507, G.C.
Method 507, G.C.
Method 515.1, G.C.
Method 515.1, G.C.
Method 507, G.C.
GC - Gas Chmmatognph
The analytical methods used by the Department of Agriculture were the same
as those used by EPA's National Pesticide Survey and are used by EPA and
State agencies for analyzing drinking water under the Public Water Supply
Supervision Program. In some cases concentrations were reported which fall
below the established minimum detection levels. In these cases, the West
Virginia Department of Agriculture laboratory director is confident the
compound is present, but the concentration listed cannot be confirmed.
It was beyond the scope of the West Virginia Department of Agriculture's
sampling effort to conduct a relational analysis of site characteristics versus
well and spring water quality. In addition, the study was not designed to yield
any statistically valid infor- mation regarding ground water quality in the
County.
A very preliminary analysis was conducted, however, of the factors which
occur most frequently at the sites which had pesticide detections, as compared
to the sites where pesticides were not detected. The following are some
reasons why great caution should be exercised in conducting this type of
analysis:
1. Several site characteristics which were evaluated, such as, "no casing" and
"no grout" are based on survey responses and have not been field
validated. Therefore, these factors may not be representative of actual
field conditions.
2. Contaminant sources identified could be downgradient of the tested well or
spring. For example, a survey response of "buried pesticides on'the
farm" does not provide any location in relation to ground water flow
directions.
122
-------
3. Other pesticides may be present in the well or spring water which were
not analyzed as part of the sampling program.
The following tables and discussion are intended to summarize the available
well sampling data. Tables 9,10 and 11 summarize the data that was
collected for each of the wells sampled. Tables 12 and 13 provide frequency
distributions of the factors found to occur hi wells with detections versus wells
without detections. A thorough hydrogeologic investigation and farmstead
assessment is necessary to actually identify causes of pesticide detections.
123
-------
RESULTS OF SEPTEMBER 1992 WELL AND SPRING SAMPLING IN JEFFERSON COUNTY
SITE CHARACTERISTICS. TYPE AND CONCENTRATION OF PESTICIDES DETECTED
TABLES
SITEt
DETECTION
lugK
DATE WELL
DRILLED
WELL DEPTH
< SO FEET
NO CASING
NO GROUT
POOR WELL CAP
PESTICIDES
USED < 200
FT FROM DW
WELL
PRIORITY
LEACHING
PESTICIDES
USED ON
PROPERTY
PJL PESTICIDES
USED
< 200 FT
FROM DW WELL
ABANDONED
WELL
< 200 FT
FROM DW WELL
SINKHOLE ON
PROPERTY
UNUSED
PESTICIDES
BURIED ON
PROPERTY
PESTICIDE
CONTAINERS
DISPOSED OF
ON PROPERTY
647
P. 011
1954
X
X
X
X
X
sotr
023
A JO
I960
X
X
X
X
107
TJ021
1970
X
X
X
X
<.ar
X
650
A JO
1971
X
1000"
072
A. 94
1965
X
<200-
FROM
DW
WELL
184
AJ4
OK
X
X
X
X
X
X
X
BURN
2001
FROM
DW
WELL
135
M.47
AJ8
1988
X
X
X
X
<. sr
sot
M8.51
A 5.0
C9J9
1970
X
X
X
X
X
326
M.75
A. 45
CJ8
1832
X
X
X
X
X
005
PJ3
1978
X
X
124
-------
SfTEl
PESTICIDE
MIXING AND
LOADING
WITHIN 200 FT
OF DW WELL
PESTICIDE
EQUIPMENT
CLEANING
WITHIN 200 FT
OF OW WELL
PESTICIDE
STORAGE AREA
WITHIN 200 FT
OF DW WELL
647
023
107
X
1 MILE
650
072
X
X
X
184
X
X
135
601
326
X
005
X
X
X
A ATRAZ1NE
P PICLORAM
T- TRICLOPYR
H METOLACHLOR
C - CHLORPYRIFOS
DW - DRINKING WATER
125
-------
SITE CHARACTERISTICS AROUND WELLS OR SPRINGS
WHERE PESTICIDES NOT DETECTED
TABLE 10
SITE It
DATE WELL
DRILLED
WELL DEPTH
< 50 FEET
HO CASING
HO GROUT
POOR WELL CAP
OR SPRING
COVER
PESTICIDES
USED < 200
FT FROM DW
WELL
PRIORITY
LEACHING
PESTICIDES
USED ON
PROPERTY
P.L. PESTICIDES
USED
< 200 FT
FROM DW WELL
ABANDONED
WELL
< 200 FT
FROM DW WELL
SINKHOLE ON
PROPERTY
UNUSED
PESTICIDES
BURIED ON
PROPERTY
PESTICIDE
CONTAINERS
DISPOSED OF
ON PROPERTY
PESTICIDE
MIXING AND
LOADING
WITHIN 200 FT
OF DW WELL
217
1940
X
woo-
198
1930
408
OK
--
X
X
160
1985
X
X
X
X
X
273
1927
X
BURN
wo-
X
378
1986
X
X
X
< SO1
X
488
1989
X
X
X
X
< Sff
X
499
OK
X
X
X
X
X
471
spring
. X
X
X
X
X
433
spring
X
2B4ff
126
-------
SITE it
PESTICIDE
EQUIPMENT
CLEANING
WITHIN 200 FT
OF DW WELL
OBVIOUS
MISUSE OF
PESTICIDES
NEAR WELL OR
SPRING
PESTICIDE
STORAGE AREA
WITHIN 200 FT
OF DW WELL
217
198
X
X
408
X
160
X
X
273
X
378
438
499
X
X
471
433
127
-------
SITE CONDITIONS AROUtJD WELLS OR SPRINGS
WHERE PESTICIDES NOT DETECTED
{CONTINUED)
TABLE It
SITES
DATE WELL
DRILLED
WELL DEPTH
<_ SO FEET
NO CASING
NO GROUT
POOR WELL CAP
PESTICIDES
USED < 200
FT FROM DW
WELL
PRIORITY
LEACHING
PESTICIDES
USED ON
PROPERTY
P.L. PESTICIDES
USED
< 200 FT
FROM OW WELL
ABANDONED
WELL
< 200 FT
FROM DW WELL
SINKHOLE ON
PROPERTY
UNUSED
PESTICIDES
BURIED ON
PROPERTY
PESTICIDE
CONTAINERS
DISPOSED OF
ON PROPERTY
PESTICIDE
MIXING AND
LOADING
WITHIN 200 FT
OF DW WELL
BIB
OK
X
X
240
1981
X
X
X
X
X
X
X
X
236
1945
X
X
2B4ff
125
1933
247
1978
X
X
X
X
X
X
200"
052
DK
X
X
X
<_.sr
028
1983
X
X
X
X
X
070
1840
X
X
X
074
1963
X
X
X
<. 50-
X
X
0
030
1970
X
X
X
128
-------
SITE It
PESTICIDE
EQUIPMENT
CLEANING
WITHIN 200 FT
OF DW WELL
OBVIOUS
MISUSE OF
PESTICIDES
NEAR WELL OR
SPRING
PESTICIDE
STORAGE AREA
WITHIN 200 FT
OF DW WELL
616
X
240
236
125
X
X
247
X
X
052
028
070
074
030
X
X
129
-------
TABLE 12
SITE CHARACTERISTIC ISC)
WELL DEPTH < SO FEET
HO CASING
NO GROUT
POOR WELL CAP OR SPRING HOUSE
WELL DRILLED BEFORE 1300
WELL DRILLED BETWEEN 1900-1970
WELL DRILLED BETWEEN 1971-1984
WELL DRILLED AFTER 1984
SPRING
ABANDONED WELL < 200 FT FROM WELL/SPRING
SAMPLED
SINKHOLE ON PROPERTY
PESTICIDES USED < 200 FT FROM WELL/SPRING
SAMPLED
PRIORITY LEACHING PESTICIDES USED ON
PROPERTY
PJL. PESTICIDES USED <_ 200 FT FROM
WELL/SPRING SAMPLED
PJL PESTICIDES USED <_50FT FROM
WELL/SPRING SAMPLED
PESTICIDES BURIED ON PROPERTY
PESTICIDE CONTAINERS DISPOSED ON PROPERTY
PESTICIDE CONTAINERS DISPOSED
<_200FT FROM WELUSPR. SAMPLED
PESTICIDE MIXING AND LOADING
<_200fT FROM WELUSPR. SAMPLED
EQUIPMENT CLEANING <_ 200 FT FROM
WELUSPR. SAMPLED
PESTICIDE STORAGE SITE < 200 FT FROM
WELUSPR. SAMPLED
OBVIOUS MISUSE OF PESTICIDES, SUCH AS
EMPTY CONTAINERS ADJACENT TO WELUSPRING
% WELLS/SPRINGS WITH
DETECTIONS WHERE SC
PRESENT
20%
10%
50%
10%
10%
SO*
20%
10%
0
10%
30%
80%
70%
70%
20%
10%
30%
10%
30%
40%
30%
0
% WELLS/SPRINGS WITH NO
DETECTION WHERE SC
PRESENT
10%
10%
20%
20%
5%
35%
15%
15%
10%
15%
30%
70%
50%
50%
20%
0
25%
5%
35%
35%
30%
10%
HIGHER FREQUENCY IN
WELUSPRING WITH
DETECTS? (Y/NI
Y
N
Y
N
Y
Y
Y
N
N
N
N
Y
Y
Y
N
Y
Y
Y
N
Y
N
N
130
-------
TABU 13
SfTE CHARACTERISTIC ISC)
WELL DEPTH <. SO FEET *
HO CASING
NO GROUT
POOR WELL CAP OR SPRING HOUSE
WELL DRILLED BEFORE 1300
WELL DRILLED BETWEEN 1300-1970
WELL DRILLED BETWEEN 1971-1984
WELL DRILLED AFTER 1384
SPRING
ABANDONED WELL < 200 FT FROM WELL/SPRING
SAMPLED
SINKHOLE ON PROPERTY
PESTICIDES USED <_ 200 FT FROM WELL/SPRING SAMPLED
pRiORrrY LEACHING PESTICIDES USED ON PROPERTY
PL PESTICIDES USED < 200 FT FROM WELL/SPRING
'SAMPLED
PL PESTICIDES USED < 50 FT FROM WELUSPRING
SAMPLED
PESTICIDES BURIED ON PROPERTY
PESTICIDE CONTAINERS DISPOSED ON PROPERTY
PESTICIDE CONTAINERS DISPOSED
<. 200 FT FROM WELL/SPRING SAMPLED
PESTICIDE MIXING AND LOADING
<. 200 FT FROM WELUSPRING SAMPLED
EQUIPMENT CLEANING <_ 200 FT FROM WELUSPRING
SAMPLED
PESTICIDE STORAGE SITE <_ 200 FT FROM WELUSPRING
SAMPLED
OBVIOUS MISUSE OF PESTICIDES. SUCH AS EMPTY
CONTAINERS ADJACENT TO WELUSPRING
% WELLS/SPRINGS THAT MET SC CRITERIA
WHICH HAD DETECTION
50%
33%
55%
20%
50%
42%
40%
25%
0%
25%
33%
36%
41%
41%
33%
100%
37%
50%
30%
30%
33%
0%
GREATER THAN OR EQUAL TO 50%?
Y
N
Y
N
Y
N
N
N
N
N
N
N
N
N
N
Y
N
Y
N
N
N
N
131
-------
Since there are very few wells or springs which meet certain site characteristic criteria
(such as "drilled prior to 1900"), these criteria are not adequately addressed by Table 12.
Therefore, the analysis shown by Table 13 was done to supplement Table 12 and help
determine what percent of all wells or springs which met a specific criterion had a detection.
Factors which occurred at a higher frequency hi the set of wells with detects are the
following:
1. In the set of wells which had detections, the following site characteristics occurred in
50% or more of the wells:
* no grout
* well drilled between 1900 - 1970
* pesticides used <_ 200 feet from the sampled well/spring
* priority leaching pesticides used on the property
* priority leaching pesticides used j<. 200 feet from sampled well/spring
2. 50% or more of the wells/springs which met the following criteria had detections:
* well depth .<. 50 feet
* no grout
* pesticides buried on property
* well drilled before 1900
* pesticide containers disposed of <. 200 feet from sampled well/spring
This is a first step to a more refined analysis. Due to the limitations of the study, as
described earlier, these factors may or may not be the primary factors influencing detections,
and may or may not be statistically significant contributors to detections. In addition, those
factors which are not singled out could be important in influencing detections. The factors
listed above, however, warrant further study regarding their use as indicators of good
candidate sites for detecting pesticides.
132
-------
CHAPTER 7: LESSONS LEARNED, KEY CONSIDERATIONS, BENEFITS,
COSTS, NEXT STEPS
A. LESSONS LEARNED, KEY CONSIDERATIONS
B. BENEFITS OF PILOT PROJECT
C. FINANCIAL SUMMARY
D. FUTURE PLANS AND NEEDS
A. Lessons Learned, Key Considerations
The following is a listing of key considerations for anyone planning a multi-agency
pesticides and ground water GIS project. These considerations are a combination of lessons
learned from mistakes which were made in Jefferson County, as well as approaches taken
from the onset of the project which appeared to work well. A very valuable aspect of the
Jefferson County pilot project was that different approaches were tried out. While some
approaches worked and some did not, the successes, as well as mistakes, are documented
here so others can benefit from what was learned. Listed first are general items to consider
when embarking on this type of a project, and listed second are specifics related to the
design and administration of an Agricultural Practices Survey.
General Lessons Learned/Key Considerations:
1. Start out as a pilot project in a small area where as much data as possible is already
automated in GIS.
2. For water quality and/or quantity studies select a watershed or ground water basin as
the study area.
3. Beware of tremendous project delays associated with converting between different
software or operating systems, especially if methods have not yet been established to
ease the conversion process. Ensure compatibility between software and operating
systems of different agencies.
4. Substantial advance planning is required if raw data needs to be collected, or existing
data needs to be preprocessed for use in GIS, since these steps are the most time
consuming.
5. If collection of raw data is necessary, it is critical to have a statistically representative
sample, as well as data entry software and good public communication regarding the
objectives the data collection effort. Before collecting new data, it is helpful to
determine the usefulness, overlaps or gaps hi data available from historical sampling
conducted by different individual agencies.
6. Combine objectives, resources and expertise from as many complementary programs
133
-------
as possible, since this increases efficiency in reaching ultimate goals. For example,
the Wellhead Protection Program, Safe Drinking Water Act vulnerability assessments
for monitoring waivers, and State Pesticides and Ground Water Management
Programs all have some overlapping goals. Coordination provides for a higher
quality product and gets a multi-disciplinary group "bought into" the project, the
process, the results and the solutions. In addition, agencies can work together to
purchase data, software and equipment which will benefit all.
7. For multi-agency projects it is essential to:
* identify analyses to meet multiple""program objectives,
* determine who is best suited to do what,
* get commitments from each agency,
* agree on timeframe; since GIS analyses are usually the planning stage, need to
agree on when implementation of recommendations from planning stage is
expected,
* ensure a minimum level of quality control and spatial standards are agreed upon
from the beginning,
* work off of the same base map layers,
* expect delays and plan for them, especially if raw data needs to be automated or
cooperating agencies have little experience with GIS, and
* coordinate future routine data collection efforts between agencies to keep the GIS
data base up to date.
8. Combine efforts with programs that have funding to help implement management
practices within areas targeted by the GIS analysis. An example is the USDA Water
Quality Initiative Program, which provides funding for technical and financial
assistance to fanners to implement best management practices within Hydrologic Unit
Area watersheds.
9. Guide the direction of hydrogeolbgic assessments and ground water monitoring to
support the development of the SMP. Utilize the results of ongoing studies such as
the USGS National Water Quality Assessment Projects.
10. In addition to FIFRA funding, States can combine portions of many different FJPA
grants to fund analyses and the development of SMPs. For example, both the Clean
Water Act Section 106 funding for Comprehensive Ground Water Protection
Programs and Section 319 funding for the Non-point Source Program have goals
consistent with the need to protect ground water from pesticides.
11. It is important to collect statistically representative pesticide usage data, so inferences
can be made to the unsurveyed population. This allows estimates of potential
pesticide loadings to ground water to be made over a larger area, such as an entire
aquifer, watershed, or State.
134
-------
12. In order to make inferences to the unsurveyed population, obtain aerial photography
for the same growing season(s) covered by the survey. Hire an experienced photo-
interpreter to determine crop types from the aerial photography. Infer pesticide
application rate information from the surveyed to the non-surveyed areas by crop
type. Use the survey data for field validation to check the degree of error associated
with the photo-interpretation.
Lessons Learned in Design and Administration of the Survey:
1. If a consultant is hired, ensure that the consultant has substantial knowledge of, and
access to, information regarding pesticide formulations, active ingredients, label
information and recommended application rates.
2. Obtain information directly from the respondent on total acres in cropland, number of
different fields of each crop and acreage of each field. Obtain pesticide application
information by field.
3. Utilize State Departments of Agriculture to obtain records from commercial
applicators. Ensure that only the application information germane to the surveyed
property is recorded for that property.
4. To increase efficiency and reduce badgering respondents with multiple surveys, it can
be very useful to incorporate the objectives of multiple programs and agencies into
one questionnaire. If this approach is taken, however, caution must be exercised in
the following areas:
* Ensure that each program clearly defines the objectives behind the questions they
want to ask.
* Require each program to assist in training the enumerators, to ensure their
program's questions are clearly understood by the enumerators, and correctly
communicated to the respondents.
* Care must be taken in the wording of all questions to ensure clear understanding,
avoid bias and obtain the answers which satisfy the objectives of the question.
* Obtain agreement among all programs regarding the primary purpose of the
survey. The final questionnaire should clearly meet these primary objectives.
* Do not let the questionnaire get too long or the respondents will get restless and
give short, vague answers, and the primary objectives will be lost.
5. Before full implementation, the entire questionnaire should be tested on a subset of
the survey population to identify problems in areas such as wording and skip
135
-------
sequences.
6. For those questions suspected of covering sensitive areas, such as pesticide disposal
and accidental spills, there should be a component to the questions which allows
enumerators to note any changes in the respondent's attitude, verbal messages or body
language that might indicate uneasiness, or the possibility of hedging on the part of
the respondent.
7. If water quality modeling will be done as part of the project, obtain data through the
survey which may be needed as input parameters for the model(s), such as method of
pesticide application, tillage methods, field slopes, etc. In addition, if pesticide
application rates will be estimated, obtain pertinent data needed for estimates, such as
percent organic matter in the soil, spacing of row crops and soil texture (by field).
8. Ask farmers how many applications of a pesticide were made per growing season.
9. Ask respondents first if they keep records, then ask the questions regarding what
pesticides were applied to what crops, and at what rate. If incomplete or vague
responses are provided, ask respondents ifahey can look up their records to answer
the questions.
10. Conduct the survey in the winter when farmers are least busy.
11. Provide farmers with information and brochures regarding the type of technical and
financial assistance which can be made available to them, and the appropriate agencies
to contact with questions. Leave a request form with them to fill out and mail in to
obtain assistance. Obtain their input on ideas for new programs, or needs they have
which are not indicated on the form. Use this input to determine the level of potential
support from the farming community for new programs, such as an amnesty day to
pick up and properly dispose of unwanted pesticides.
12. Approval is needed from the Office of Management and Budget (OMB) to conduct a
federally funded survey of 10 or more people in order to ensure compliance with the
Paperwork Reduction Act. Because of the time consuming nature of the approval
process, agencies may want to seek alternative funding sources when conducting
surveys.
13. Ensure that skip sequences in the questionnaire do not prevent enumerators from
asking farmers pertinent questions. An example of this type of problem would be to
skip over questions on pesticide storage and disposal practices for farms sprayed by
commercial applicators, when farmers or owners of the property often know more
about these practices than the commercial applicators.
14. Only get x and y coordinates for locations of potential sources of contamination
136
-------
around fanner's drinking water wells or springs (from field observations) if the other
data layers, such as sinkholes, faults and topography will have similar measures of
resolution. By obtaining x and y coordinates, one has the flexibility to use GIS to
conduct proximity analyses. If other data layers, however, do not have the same
degree of accuracy at the same large scale resolution, obtaining distances only of the
sources from the drinking water well or spring will suffice.
B. Benefits of the Pilot Project
In addition to the lessons learned, federal, State and local agencies across the nation
will be able to benefit from the methods used and specific products developed for the
Jefferson County Pilot project. The methods and products which are transferable to other
areas and projects are listed below.
Methods were developed to:
1. Obtain detailed pesticide usage data, through designing a questionnaire, selecting a
sample, and administering the survey.
2. Automate and preprocess data for use in GIS, including methods to:
* develop software to automate survey data,
* transfer linework between aerial photographs, slides and mylars for the digitization
process, and
* utilize software to interpolate between points of known values.
3. Create the maps and conduct the analyses in GIS, including the development of
mathematical, logical and cartographic procedures.
4. Conduct vulnerability assessments using GIS; including determination of data needs,
building coverages and utilizing and comparing different assessment methods to
conduct the analyses.
5. Convert between raster and vector GIS software systems, such as GRASS and
ARC/INFO, and to transport data from UNIX to DOS systems.
Products:
1. Questionnaire and enumerator training manual.
2. Software used to automate survey data, and Standard Operating Procedures used to
edit and code data.
3. Final maps and report, including management recommendations for targeted areas.
137
-------
4. Information on current pesticide storage and disposal practices, to help guide the
direction of guidance and regulation development.
5. Comparison of pesticide usage estimates from different data collection methods (on-
farm surveys, County Agricultural Agent estimates and.existing data bases).
6. Brand name to active ingredient conversion software, manual and bibliography of
source information.
7. Comprehensive data base with multiple applications.
Benefits of Using GIS:
The pilot project proved that the use of GIS to develop a State Pesticides hi Ground
Water Management Plan has many benefits, including:
1. Even though initial automation of data is expensive, once the data is automated, there
are a multitude of applications possible from one comprehensive data set. GIS
ultimately saves managers time and money, because it speeds up the whole decision
making process, and handles multiple complex queries simultaneously to target
priority areas.
2. Targeting saves time and money, because it allows resources to be spent more
efficiently, by focusing activities such as inspections, technical assistance,
enforcement and monitoring in the high risk areas.
3. GIS becomes a bridge linking many sources of data and disciplines. This is
particularly noteworthy for the application described hi this report, because of the
need for a cooperative cross programmatic workgroup to develop SMPs.
4. GIS allows for the storage and updating of data that changes over the years, such as
pesticide usage, inspection and enforcement records, management practices and
property boundaries.
5. Through the use of GIS, trends in attributes such as land use, water quality and the
type and degree of BMP implementation can be easily evaluated over time.
In addition to the methods and products developed, and the value of the project hi
demonstrating the benefits of using GIS, insight was provided through the pilot project on
how to organize and mobilize a multi-disciplinary workgroup, to identify needs and work
together on solutions. The results of the survey and GIS analyses are being used by
members of the workgroup to target technical and financial assistance to farmers. In
addition, the survey was a useful educational tool; it resulted in increased community
awareness and provided detailed information on the use, storage and disposal of pesticides
138
-------
which cannot be found in existing data bases.
C. Financial Summary
There will be a substantial difference in the cost of a GIS project which utilizes
existing digital data, versus the cost of a project where all the data must be collected and
automated for use in the GIS. Digital data can either be obtained for free from an agency
willing to share it, or it can be purchased relatively inexpensively, leaving the cost of the
analyst, and program staff as the main remaining expenditures. Existence of detailed digital
data for soils and farm parcels made Jefferson County an attractive site to conduct the pilot
project. However, there was still a substantial effort involved in collecting land use data, as
well as importing and preprocessing existing digital data. Differences in software and
operating systems resulted in the expenditure of additional staff time to import existing digital
data. In addition, it is important to note that there are high start up costs associated with
implementing a pilot project, since it is a learning process.
In order to provide a complete picture of all the costs associated with .this pilot
project, the following two tables were developed. It is critical to note that these expenditures
are not for this project alone: most of the costs are for equipment and data necessary to
meet a multitude of planning objectives, ranging from siting septic systems to delineating
wetlands to regulating housing density. For example, soils were digitized primarily to meet
the conservation planning needs of the local SCS office. In addition, USD A and EPA
purchased their workstations to serve a variety of planning functions, not just to meet the
objectives of this project. Finally, USGS characterized ground water flow and quality to
serve the planning objectives of their clients, the County Commissioners, who funded the
hydrogeologic study before the pesticides and ground water GIS project was ever initiated.
These examples are not all inclusive, they are simply listed here to help show that most of
the data collection and automation efforts were embarked upon to address a multitude of
planning needs.
The following is a compilation of rough estimates of the costs associated with
collecting, automating and analyzing all the data which was used to carry out this project.
Table 14 has an itemized listing of hours of labor associated with specific tasks and Table 15
has a listing of expenditures for salaries, equipment and contracts. It is important to note
that the following tables do not list the USDA costs associated with collecting raw data such
as soils, farm parcels, roads and sinkholes and preparing that data for digitizing. Preparation
of data for digitizing far exceeds the costs associated with digitizing.
139
-------
Table 14
AGENCY
USDA
USDA
USDA
USDA
EPA
TASK
Manually digitize soils (1200 polygons) @ 37
hours per one 7.5 minute quadrangle
Manually digitize roads and streams @ 6 hours
per one 7.5 minute quadrangle
Manually digitize farm parcels for the County
Manually digitize sinkholes for the County
(170)
Manage contracts, coordinate data transfers,
digitize geology, water table and DRASTIC
maps, import and preprocess digital data,
determine types of analyses to conduct,
develop methods and conduct analyses
ESTIMATED
HOURS
148
24
5
4
3,120
Table 15
CATEGORY
Equipment
Salaries
Salaries
Contractual
Contractual
Contractual
Contractual
TOTAL
TOTAL
TASK or ITEM
Workstations(2)
USDA - 20 months professional staff and 20
months assistant staff to automate data
EPA - program staff and computer analyst,
25% of time for three years
Agricultural Practices Survey
Develop data entry software
Edit, code and enter survey data into data base
USGS Hydrogeologic Study
FOR ALL TASKS LISTED ABOVE
FOR PESTICIDES AND GROUND WATER
ANALYSES ONLY
ESTIMATED
COST
$70,000
$62,000
$80,000*
$32,500*
$30,000
$30,000*
$200,000
$504,500
$142,500
140
-------
In order to properly interpret the above costs, those items which are a one time cost
with multiple applications are separated from those items (with an asterisk) which were
earned out predominantly to conduct pesticide/ground water vulnerability analyses for the
County. The items with asterisks, specifically, EPA staff time, and administration and
automation of the Agricultural Practices Survey, total $142,500. Data from the Agricultural
Practices Survey can also be used for other applications, such as a County Wellhead
Protection Program. In addition, there are other alternative methods of obtaining pesticide
usage data, as opposed to interviewing individual farmers. For example, County
Agricultural Agents and University and Extension literature can be used to estimate what
pesticides are used, and in what quantities. Photo-interpretations of crop types can provide
locational information for assigning pesticide usage estimates. The costs versus the benefits
of these alternatives as compared to individual fanner interviews is being explored; the final
analysis will be available as Appendix F to this document.
D. Future Plans and Needs
Future plans for the County include monitoring ground water in areas targeted as
having a high potential for pesticide leaching. Initial monitoring took place in September
1992 at 30 sites with suspected problems. Pesticides were detected at 10 sites. Additional
monitoring is needed in areas representative of different classifications of ground water
vulnerability. GIS predictions could then be compared with actual ground water quality, to
help determine which factors play the greatest role in influencing pesticide migration to
ground water. In addition, extensive monitoring is needed to determine what percent, if any,
of the population of the County may be drinking water with pesticides at unsafe levels.
Additional future plans include development and implementation of a Wellhead
Protection (WHP) program for Jefferson County. A pilot WHP project has been initiated in
the County to delineate zones of contribution around public water supply wells, identify
sources of potential contamination to the wells and devise control measures. Hydrogeologic
and contaminant source data from the pesticide project will be used in GIS along with data
collected for the WHP project to achieve these objectives.
There are many more GIS analyses which can be conducted in Jefferson County using
the data collected through this pilot effort. For example, it would be useful to do analyses of
farms with the highest application rates per acre of specific priority leaching pesticides,
(versus total pounds per farm of all priority leachers) hi relation to soil and hydrogeologic
characteristics. This analysis would be useful for targeting locations to sample ground water,
identifying compounds to analyze for, and identifying farmers which may benefit from
educational assistance.
Another pilot project is being carried out hi the Pequea and Mill Creeks watershed hi
Lancaster County, Pennsylvania to implement recommendations that arose from lessons
learned in the Jefferson County project. Examples of recommendations being implemented
include the following: the Pequea-Mill project surveyed a statistically representative sample,
141
-------
crop types in nonsurveyed areas are being identified through photo-interpretation of the entire
study area, ground water monitoring is being conducted to compare GIS predictions of
vulnerability with actual water quality, the project is watershed based, and the regolith
(surficial deposits between the root zone and bedrock) is being characterized through
geophysical logging, to help determine susceptibility. In addition, USDA is working hi the
same watershed to provide technical and financial assistance to farmers as part of their
Hydrologic Unit Area Project.
142
-------
Bibliography
1) Beiber, P.B., USGS, Ground Water Features of Berkeley and Jefferson
Counties. WV. WV Geological Survey Bulletin 21, 1961.
2) Calkins, H.W. and R.F. Tomlinson, Geographic Information Systems:
Methods and Equipment for Land Use Planning. International Geographic
Union Commission on Geographical Data Sensing and Processing, Resource
and Land Investigations Program, U.S.G.S., Reston, VA, 1977.
3) Carpenter, Stephen G., Geographic Information Systems and Natural Resource
Assessment: Using SCS GRASS 3.0 as a Comprehensive GIS at the County
level for Ground Water Quality Assessment and Conservation Planning.
Annual GRASS Users Conference, 11/7/89.
4) Crop Protection Chemicals Reference. Chemical and Pharmaceutical Press and
Publishing Company and John Wiley and Sons, 1988.
5) EPA's Pesticide Programs. U.S. EPA, Pesticides and Toxic Substances
(H-7506-C), May, 1991.
6) Gianessi, L.P., Lack of Data Stymies Informed Decisions on Agricultural
Pesticides. Resources, 89, 1, 1987.
7) Goss, D., and Wauchope, R. Don, The SCS/ARS/CES Pesticide Properties
Database II: Using it with Soils Data in a Screening Procedure. 11/90.
8) Guptill, Stephen C. and Nystrom, David A., Geographic Information Systems:
An important Tool for Spatial Analysis.
9) Hammell, Stephen R., Groundwater Contamination Remediation Planning.
Regional Planning Masters Thesis, University of Pennsylvania, Philadelphia,
PA, 12/91.
10) Hatfield W.F. and Warner, J.W., Soil Survey of Jefferson County. WV: U.S.
Department of Agriculture, Soil Conservation Service, 1973.
11) Hobba, W.A., Jr., USGS, personal communication, 3/28/91.
12) Hobba, W.A., Jr., USGS, Ground Water Hydrology of Jefferson County.
WV. Environmental Geology Bulletin, EGB-16, 1981.
143
-------
13) Hobba, W.A. Jr., Friel, E.A. and Chisholm, J.L., Water Resources of the
Potomac River Basin. WV: WV Geological Survey River Basin Bulletin 3,
1972.
14) Kinsinger, A.E., Project on Agricultural Conservation and Tax Policy. Natural
Resources Defense Council, Washington, D.C., 1987.
15) Kozar, M.D., Hobba, W.A. Jr., Macy, J.A., Geohvdrologv. Water
Availability and Water Quality of Jefferson County. WV. U.S. Geological
Survey Water Resources Investigations Report 90-4118, 1991.
16) Maguire, David J., et al., Geographic Information Systems Principles and
Applications. Longman Scientific and Technical, Essex, England, 1991.
17) Means, John, Methods and Conclusions for Application of DRASTIC to
Jefferson County. WV. WV Division of Environmental Protection, 1988.
18) Star, Jeffrey and John Estes; GIS. An Introduction. Prentice Hall; Englewood
Cliffs, NJ, 1990.
19) State of Maryland Pesticide Statistics for 1988. MD Department of
Agriculture, June 1990.
20) Stoltzfus, Jeffrey, Penn State Cooperative Extension Service Press Release:
Pequea-Mill Creek Water Quality Project. New Holland, PA, June 1991.
21) The Farm Chemicals Handbook. Meister Publishing Company, 1991.
22) Young, A.L. Minimizing the Risk Associated with Pesticide Use: An
Overview in Pesticides - Minimizing the Risks. N.N; Ragsdale and R.J. Kuhr
(Eds.), ACS Symposium Series 336, American-Chemical Society, Washington
D.C., 1987.
23) National Survey of Pesticides in Drinking Water Wells. Phase I Report. U.S. EPA,
EPA 570/9-90-015, November 1990.
24) Pesticides Included in the EPA National Pesticide Survey. U.S. EPA fact sheet, April
14, 1988.
144
------- |