Number 47                                                     March 1991            

The Oakwood Lakes - Poinsett project hosted an excellent Rural Clean Water Project (RCWP) workshop held September 18-20, 1990, in Brookings, South Dakota. The workshop focused on preparation of the 10-year report required of each project. This report is a comprehensive look at project accomplishments since implementation. and will be an important reference on the findings of each RCWP project.

Workshop participants discussed methods for completing each section of the standardized report outline. All topics for project documentation were covered, including institutional arrangements, information and education, land treatment, and water quality results. Project representatives presented different methods for reporting accomplishments. Several papers were presented that showed methods for reporting land treatment and linking land treatment to water quality monitoring results. Presentations were followed by an open discussion of the strengths and weaknesses of reporting methods. Several important points emerged from the discussions:

• Each project has unique reporting requirements that depend on objectives. The way in which a project "tells its story" is important so that key concepts are developed and findings supported.
  

• Close monitoring of project activities is critical for documenting accomplishments.
  

• It may be difficult to anticipate all reporting needs at the beginning of a project. However, sitting down with the people who represent different project components during the early stages of the project helps to avoid problems later.
  

The methodology for on-site evaluation of the RCWP projects was presented at the workshop (see next item).

Methodology

The methodology for the on-site evaluation of the RCWP has been developed by an interagency work group consisting of the National Water Quality Evaluation Project at the North Carolina State University Water Quality Group, the U.S. Environmental Protection Agency (EPA), the Agricultural Stabilization and Conservation Service (ASCS), the Soil Conservation Service (SCS), the Extension Service, and team members from several RCWP projects. The methodology was presented on September 20, 1990, at the RCWP Workshop in Brookings, South Dakota. The objectives of the evaluation, the evaluation procedure, and the potential importance of the findings were determined by representatives of NWQEP, ASCS, SCS, and EPA. The following review of the methodology is designed to inform all project team members about the evaluation procedure and the status and scheduling of on- site visits.

Objectives of the evaluation are 1) to assess: a) cooperation among project team members, committees, and agencies; b) the agreement between the documented water quality problem and the choice of solutions; c) the achievements of the project and individuals in relation to RCWP objectives; d) monitoring and the assessment of project impacts; and e) the effectiveness of the project and its progress toward improving water quality and 2) to compile lessons learned.

An evaluation team will visit project sites, interview project team members, review project activities, and report results. Interviews with individual RCWP team members will be conducted on a one-to-one basis by evaluation team members. The survey questions were distributed at the RCWP Workshop and included in the workshop proceedings. They will also be sent to the appropriate state ASCS office during the planning stages of each on-site evaluation.

Survey questions are designed to gather specific information on project elements,including administrative oversight, local and state coordinating committees, information and education, land treatment, and water quality monitoring and evaluation. A separate survey of producers and project team members will be used to determine factors that influence participation and perceptions of program effectiveness.

Evaluation Team

The evaluation team will usually consist of one SCS staff person from outside the project who has worked in the RCWP for two years, one NWQEP staff member, and possibly a member of the interagency work group. For comprehensive monitoring and evaluation project evaluations and general projects that require more detailed evaluation, selected RCWP personnel from other RCWP projects or EPA staff may be needed. Composition of the evaluation team will be determined on a project-by- project basis by NWQEP.

Project Preparation for the Site Visit

ASCS headquarters in Washington, DC has informed each state ASCS office of the evaluation and has provided a copy of the survey questions in the draft proceedings of the 1990 RCWP Workshop. NWQEP will contact each state ASCS office to schedule the site visit and interviews. Each project team will be asked to list team members, their agency, and all project elements in which each team member participated. Cooperation by individual RCWP projects is encouraged and expected.

Project team members should review the methodology before the evaluation team arrives. Please call Steve Coffey at NWQEP (919-737-3723) if your have any questions. Ideally, the preparations made prior to the arrival of the review team will help reduce anxiety and an understanding the methodology should help respondents feel at ease about the nature of the evaluation.

Names will not be used for data gathering or data analysis. We will also supply a separate data sheet if the respondent wants to make a written statement about some feature of RCWP not covered in the in- person interview. Tape recording will most likely be necessary to assure that all the response data are gathered. Each interview will have a target length of about one hour and we will try to stay on schedule.

The following on-site evaluations have taken place to date: Two or three evaluations are being planned for May, 1991.

We have been very pleased with the cooperation of all who have participated in site visits. We look forward to visiting each project and compiling lessons learned.

Nonpoint source pollution (NPS) currently poses the greatest threat to surface and ground water quality in Wisconsin. Over ten years ago, the state enacted legislation that resulted in the Wisconsin Nonpoint Source Water Pollution Abatement Program. The program provides a comprehensive, coordinated approach to the state's nonpoint pollution problems.

Because the program is largely voluntary, education is a key component. Curricula for public education on NPS pollution should stress 1) increasing awareness of NPS impacts and the diversity of sources of NPS pollution and 2) the importance of individual attitudes and practices in successful NPS pollution control programs. Also, since NPS is only one of many important issues facing the public today, ongoing public education and awareness activities are needed to ensure that the issue receives its proper share of attention.

A report released in August of 1990 by the Environmental Resources Center of the University of Wisconsin at Madison provides guidance for developing information and education (I & E) strategies for priority watershed projects of the state's Nonpoint Source Water Pollution Abatement Program. The report is the result of a significant interagency cooperative effort by staff of the state's departments of Natural Resources and Agriculture and area water quality Extension staff to improve the effectiveness of educational programs related to Wisconsin's NPS projects. The report is organized around the following main topics: 1) first steps, 2) I & E roles of agency staff, 3) I & E audiences, 4) I & E objectives, 5) sequence of target audiences, education objectives, and activities, and 6) recommendations for improvement of the program. The following information is drawn from the report, copies of which are available from Gary Jackson, Environmental Resources Center, 216 Agriculture Hall, University of Wisconsin, Madison, WI 53706, Tel: (608) 262-0020.

The most important phase in creating a successful I & E program lies at the beginning of a project, when communication lines are established, roles are defined, relationships are developed, and expectations set. Suggestions for successful project staff organization include conducting individual meetings with people responsible for organizing the I & E work group in order to initiate good working relationships, promote teamwork, and identify potential concerns or conflicts. (In the case of the Wisconsin program, participants ideally include county and conservation project managers; Department of Natural Resources (DNR) project coordinator; county Extension agents; University of Wisconsin Extension (UWEX) area water quality agent; Department of Agriculture, Trade and Consumer Protection (DATCP) project coordinator; and DNR district public information coordinator.) Conducting organizing meetings of the work group involving all staff who have current or future roles in the project can help establish a sense of teamwork and acknowledge the importance of each person's contribution to the team effort. Definition of roles and setting realistic expectations of each team member are among the most important functions of an organizing meeting. Poorly defined roles are common barriers to successful I & E programs. Part of the discussion should address how roles may change during the course of the project. Assignment of specific responsibilities and building accountability into the process are also key. Written documentation of roles and responsibilities, schedules, and other elements of the planning phase can assist the team in evaluating project progress and performance.

Administrative rules require the creation of a citizen advisory group to provide input and advice during the planning phase of a watershed project. Organization, composition, and roles of subcommittees of the advisory board are discussed with the goal of creating a dynamic, well-informed, committed local citizen advisory group. This group should also provide ongoing guidance for education on maintaining best management practices, home care practices, public works programs, and other activities.

A guiding principle of the Wisconsin program has been that I & E programs should be as localized as possible, with area and state staff serving in a support role. Another important guideline is that the I & E strategy be developed and implemented using a team approach. UWEX plays a lead role in I & E activities, working at three levels: county extension office, area water quality agent, and state-level specialists and staff. Additional technical and educational resources are provided by the university's Nutrient and Pest Management Program and Central Wisconsin Groundwater Center.

As the state's lead water quality agency, the DNR oversees the administration of the Nonpoint Source Water Pollution Abatement Program. Ideally, the DNR's district nonpoint source coordinator works closely with the area water quality education agent and others to integrate the I & E strategies with other elements of the watershed project. DNR Bureau of Information and Education develops nonpoint source materials for use by watershed project staff.

The County Land Conservation Department (LCD) is the lead watershed project management agency at the county level. Depending on the type and level of involvement of county Extension offices in a watershed, county land conservation departments will be involved in the project to varying extents. They may also organize demonstration projects and play an educational role in on-farm visits by distributing literature, collecting data, and responding to landowner questions.

Basic generalized information on common agricultural practices for training purposes is provided by the DATCP.

U.S. Soil Conservation Service staff provide training in engineering practices and design to LCD staff and may serve on watershed committees and contribute to I & E committees. SCS educational programs in areas such as soil testing and pesticide application also contribute to the educational aspects of the NPS program.

Because the Nonpoint Source Water Pollution Abatement Program is largely voluntary, I & E strategies must be action-oriented, motivating people to adopt practices that will reduce nonpoint source pollution. Ideally, education strategies will also result in public support for needed public policy changes. To achieve these goals, the following sequence of objectives is suggested for the educational process: 1) problem recognition, 2) project awareness and planning involvement, 3) acceptance of solutions identified through the project, 4) technical assistance, and 5) impact recognition.

The report includes a table summarizing target audiences, education objectives, and potential I & E activities for the planning, sign-up, and implementation phases of a watershed project.

Suggestions for improving the state Nonpoint Source Water Pollution Abatement Program include better audience targeting, more effective measurement and communication of the economic benefits of Best Management Practices, a sustained state-wide media campaign to increase public awareness of NPS, facilitation of information sharing among the state's 40 priority watershed projects, initiation of homeowner education programs, expanded educational strategies for addressing urban nonpoint source pollution, increasing public awareness of the potential linkage of nonpoint source pollution and groundwater pollution.

In mid-December, 1990, we bid a fond farewell to Sarah Brichford, editor of NWQEP NOTES for the past four years. Sarah has taken a position as Extension Water Quality Specialist at Purdue University in West Lafayette, Indiana. In her new job, Sarah is responsible for the Extension I & E program in the Upper Tippecanoe and Upper Kankakee Rivers Hydrologic Unit Area (HUA) Water Quality Projects.

In addition to her outstanding work on NWQEP NOTES, Sarah also contributed to the efforts of the Water Quality Group through her work on the Rural Clean Water Program and the Gaston County, North Carolina Water Quality Data Base Development and Assessment Project (highlighted in NWQEP NOTES No. 46). While she was a member of the Water Quality Group, she also wrote a manager's guide to nonpoint source implementation projects for the U.S. Environmental Protection Agency to support implementation of projects receiving Section 319 funds and a training video script describing streamside management zone best management practices (BMPs) for control of forestry NPS pollution.

Sarah was a tremendous asset to the NCSU Water Quality Group. We truly miss her and wish her the very best in her new job!

Judith Gale joined the NCSU Water Quality Group staff in February, 1991. Her responsibilities include coordinating and editing NWQEP NOTES. Judith brings many skills and experiences relevant to the work of the group. She has been involved in the analysis and dissemination of information about coastal, water, and wetland resources and policy for the past ten years. Judith has managed projects, coordinated multi-disciplinary research teams, performed field work, and written and edited technical documents and publications for the general public and policy makers on a variety of natural resource issues. Her professional work includes positions as a water resource planner with the North Carolina Division of Water Resources and as principal investigator, project coordinator, and editor with the North Carolina State University Center for Environmental Studies. Judith also worked as a Research Scientist at the Battelle-New England Marine Research Laboratory in Duxbury, Massachusetts. For the past three years, she worked as a self-employed environmental consultant. Issues addressed in her work include cumulative impacts of peat mining; wetlands regulation; river basin management; effects of salinity changes on coastal fisheries; estuarine ecology and water quality; ocean dumping; and local options for air and water quality protection.

Much of Judith's work has involved the presentation of scientific information in a format accessible and interesting to citizens. Her most recent publications include a citizen's guide to estuarine ecology and a guide to environmental protection options for community leaders.

In addition to her work with U.S. fresh, estuarine, and marine resources, Judith also maintains an interest in international resource and development issues. During 1988-90, she spent two months in Bolivia coordinating the work of a voluntary organization linking a state in Bolivia with North Carolina. As part of a two-year fellowship in international leadership development, she will visit and learn about the natural resources of Mexico, Brazil, and the Caribbean and carry out a small community-based watershed management project in Bolivia.

Judith received a Master of Forest Science in natural resource management from the Yale University School of Forestry and Environmental Studies in 1982. Her training includes work at the Marine Biological Laboratory in Woods Hole, Massachusetts and the University of Oregon's Institute of Marine Biology in Charleston, Oregon.

Jean Spooner
NCSU Water Quality Group

Description

A water quality sample may have a pollutant concentration below or above the detection limit (DL) of the available analytical equipment. Instrument readings beyond the DL are considered inaccurate because of their high variability. Censoring occurs when the observation value is reported as the DL. The percentage of censored values in a data set determines the censoring intensity.

Censoring is most common when toxics and pesticides are being analyzed. It is not as common with other NPS pollutants such as sediment, phosphorus, and nitrogen. Generally, values for sediment, phosphorus, and nitrogen are above their DL's. In addition, levels considered to be an environmental problem are usually greater than the DL's.

Data sets censored at the low values are considered to be `left-censored.' These values are referred to in several ways: samples below the detection limit, non-detects, below limit of detection (LOD), less-thans, censored values, and trace data. Left-censored data sets are very common in water quality monitoring data.

Although not as common in water quality data, `right-censoring' can occur when observations are extremely high or out-of-range. Common examples include mis-estimated dilution amount for coliform analysis resulting in TNTC (too numerous to count) and exceedance of flow gauge limit during floods. In other disciplines involving analysis of survival and failure times, research data commonly have values that are longer than the time of measurement, which results in a data set that is censored at the high values.

There are two classes of censoring. Type I censoring occurs when the DL is a fixed value and the number of censored variables is a random variable. Most censoring in environmental studies is of Type I. Type II censoring fixes the number of censored values while the censoring value (e.g., DL) is a random variable. Type II censoring is common in survival analysis where the data are right-censored.

The usual reporting convention is to report the concentration as the DL value with an associated flag variable indicating that the real concentration is below or above the reported value. For example, this reporting procedure is used in the U.S. EPA's STORET (STOrage and RETreival ) water quality data base.

The detection limit may vary between samples (Millard and Deverel, 1988), resulting in multiple detection limits. There are several reasons multiple detection limits may occur in a data set. Several analytical methods for the same pollutant may be used based on the expected sample concentration. Also, the dilution of the sample prior to analysis will affect the detection limit. Finally, the detection limit can change over time (usually decreasing) as the instrumentation becomes more sensitive. The detection limit may therefore vary within and/or between laboratories.

Effect of Censoring on Statistical Trend Analyses

When sample values are beyond detection limits, information contained in these values is lost (Cohen, 1959; Gilliom et al., 1984). This loss of information and the treatment of censored values will affect estimation of the mean and variance as well as the results and interpretations of statistical trend analyses.

In statistical analyses, it is common to set the concentration of each censored value to a constant value. The choice of an arbitrary value to represent concentrations below (above) the detection limit has a dramatic effect on estimates of distribution parameters such as the mean and standard deviation (or variance). It can be difficult to obtain the unbiased and minimum variance estimates of these distribution parameters when the data are censored (Rajagopal et al., 1987; Gilbert, 1987; Newman et al., 1989; Gilliom and Helsel, 1986; Helsel and Gilliom, 1986; El-Shaarawi, 1989). Methods that provide unbiased and minimum variance estimates of a population's mean and standard deviation are available and discussed below. However, most of these methods require extensive programming, a knowledgeable statistician who can apply probability distribution theory to the data set with DL samples (this can be performed with the use of the appropriate SAS functions), or the use of specialized computer programs designed to handle censored data sets (see, for example, Newman et al., 1989; Newman and Dixon, 1990).

The choice of an arbitrary replacement value for censored data also has an effect on the significance level of a parametric trend test (Helsel, 1987). The magnitude and direction of this effect is a function of the replacement value chosen and is discussed below for each method of handling censored data values. In addition, the presence of many non-detects in a data set sampled from a normal population causes the data set to be non-normal, which may violate the normality assumption required in most parametric trend analysis.

Some nonparametric statistics are valid with data sets having censored values because these procedures 1) do not require the residuals to follow a normal distribution and 2) treat all less-than values equivalently as ties that are less than all noncensored values (i.e., values within detection limits) (Hirsch et al., 1982; Helsel, 1987; Gilliom et al., 1984; Millard and Deverel, 1988). The magnitude chosen for these values does not affect the significance level of most nonparametric trend tests. These tests are appropriate when there is only one detection limit value for the entire data set. Also, the nonparametric trend procedure must be able to handle ties properly. The nonparametric equivalent of the mean, the sample median, will be an unbiased estimate of the population median as long as the censoring intensity does not exceed 50 percent (Gilbert, 1987).

Two problems occur when the detection limit decreases over time: 1) how to choose the value to use for each non-detect observation; and 2) how to compare observations between the two time periods. The data will show an artificial decreasing trend if the detection limits for each time are used. Using the least sensitive DL value results in a more accurate trend estimation, but results in a large loss of information from the samples analyzed with the more sensitive DL.

References

Cohen, A.C., Jr. 1959. Simplified estimators for the normal distribution when samples are single censored or truncated, Technometrics 1(3):217-237.

El-Shaarawi, A.H. 1989. Inferences about the mean from censored water quality data, Water Resources Research 25(4):685-690.

Gilbert, R.O. 1987. Statistical Methods for Environmental Pollution Monitoring. Van Nostrand Reinhold Company, New York, New York. 320 p.

Gilliom, R.J. and D.R. Helsel. 1986. Estimation of distributional parameters for censored trace level water quality data. 1. Estimation techniques, Water Resources Research 22(2):135-146.

Gilliom, R.J., R.M. Hirsch, and E.J. Gilroy. 1984. Effect of censoring trace-level water-quality data on trend-detection, Environmental Science and Technology 18(7):530-535.

Helsel, D.R. 1987. Advantages of nonparametric procedures for analysis of water quality data. Hydrological Sciences J. 32(2):179-190.

Hirsch, R.M, J.R. Slack, and R.A. Smith. 1982. Techniques of trend analysis for monthly water quality data, Water Resources Research 18(1):107-121.

Millard, S.P. and S.J. Deverel. 1988. Nonparametric statistical methods for comparing two sites based on data with multiple nondetect limits, Water Resources Research 24(12):2087-2098.

Newman, M.C. and P.M. Dixon. 1990. UNCENSOR: A program to estimate means and standard deviations for data sets with below detection limit observations, American Environmental Laboratory 2(2):26-30.

Newman, M.C., P.M. Dixon, B.B. Looney, and J.E. Pinder, III. 1989. Estimating mean and variance for environmental samples with below detection limit observations, Water Resources Bulletin 25(4):905-916.

Rajagopal, R., K.V. Ramana, and A.M. Pitchford. 1987. Assessment of variations in ground-water quality in Iowa. p. 209-223. In: Symposium on Monitoring, Modeling, and Mediating Water Quality. American Water Resources Association, Bethesda, Maryland.

Coffey, Steven W., Jean Spooner, and Jon Arnold. 1990. Grazing Best Management Practices: Annotated Bibliography. National Water Quality Evaluation Project, Biological and Agricultural Engineering Department, North Carolina State University, Raleigh, NC.

An annotated bibliography of grazing best management practices prepared by NWQEP staff was published in December of 1990. The bibliography is based on a topic search of the NWQEP computerized Annotated Nonpoint Source Library. The document contains 49 references with short annotations printed by first author in alphabetical order.

Copies of the bibliography are available from the NCSU Water Quality Group for $5 per copy.

Construction sites are widely known to exhibit serious problems in controlling erosion and sedimentation. A new videotape shows how to employ various methods and devices to control on-site erosion and prevent sediment from leaving construction sites. Produced by the NCSU Water Quality Group, Principles of Erosion and Sedimentation Control on Construction Sites shows practices as used on actual construction projects, including measures that are working well and those that are failing. The 17-minute video provides useful information for public erosion control programs and teaches erosion control techniques which cannot easily be conveyed via the printed word.

Copies of the video can be ordered from the NCSU Water Quality Group. The price is $20.

U.S. EPA. 1990. RCWP: Lessons Learned from a Voluntary Nonpoint Source Control Experiment. EPA 440/4-90-012. U.S. Environmental Protection Agency, Nonpoint Source Control Branch, Washington, DC.

Ten years experience with the Rural Clean Water Program (RCWP) is detailed in this report based on 22 projects representing a wide range of agricultural pollution problems and impaired uses throughout the U.S. The report describes how the RCWP has worked so far and synthesizes its successes and failures into lessons that can help state and local managers put together their own management programs for controlling agricultural nonpoint source pollution.

The findings indicate that BMPs, when properly implemented, have improved water quality in some of the RCWP projects. Because these findings are backed by adequate water quality data and sound statistical analyses, they begin to fill the voids in our understanding of how BMPs affect water quality in watersheds.

Copies of the report can be obtained from the NPS Information Exchange, (WH- 553), U.S. EPA, 401 M St., SW, Washington, DC 20460 or the NCSU Water Quality Group.

U.S. EPA. 1991. The National Pesticide Survey: Phase I Report. U.S. Environmental Protection Agency, Washington, DC.

The National Pesticide Survey was undertaken to evaluate the presence of pesticides, pesticide degradates, and nitrate in drinking water wells in the United States. The Survey lasted over five years and cost more than $12 million. It involved collection of extensive data, including both samples of well water and detailed questionnaire responses, about community water system wells and rural domestic wells throughout the country. Well water samples were analyzed for the presence of 127 analytes.

The Phase I Report provides a detailed summary of Survey procedures and results. The background, design, and implementation of the Survey; analyte selection and analytic methods; and quality assurance/quality control procedures are described. Detailed descriptions of the statistical methodology and Survey procedures, as well as copies of the questionnaires used to gather data, are included in the report.

Copies can be requested from the National Technical Information Service, 5285 Port Royal Road, Springfield, VA 22161, Tel: (703) 487-4650. Cost per copy is $35, plus $3 for handling.

Office of Technology Assessment. 1990. Beneath the Bottom Line: Agricultural Approaches to Reduce Agrichemical Contamination of Groundwater. OTA, Washington, DC.

The report provides a comprehensive review (though on a relatively elementary level) of the following topics: 1) contamination of the hydrogeological system; 2) technologies to improve nutrient and pest management; and 3) farmer decisionmaking and technical assistance to reduce agrichemical contamination of groundwater. Also addressed are public policy and policy options related to agrichemical contamination of groundwater.

Copies of the full report (337 pp.) or the report summary (79 pp.) may be ordered from the U.S. Government Printing Office, Washington, DC 20402-9325, Tel: (202) 783-3238. The GPO numbers and prices are 052-003-01190-5 ($15) for the full report and 052-003-01191-3 ($4) for the summary.

A 20-minute video entitled Environmentally Safe Handling of Fertilizers is available for $20 from TVA, Technical Library, NFE 1E Muscle Shoals, AL 35660- 1010.

NWQEP NOTES is issued bimonthly. Subscriptions are free (contact: Publications Coordinator at the address below or via email at wq_puborder@ncsu.edu). A list of publications on nonpoint source pollution distributed by the NCSU Water Quality Group is included in each hardcopy issue of the newsletter.

I welcome your views, findings, information, and suggestions for articles. Please feel free to contact me.

Judith A. Gale, Editor
Water Quality Extension Specialist
North Carolina State University Water Quality Group
Campus Box 7637
North Carolina State University
Raleigh, NC 27695
Tel: 919-515-3723
Fax: 919-515-7448
email: notes_editor@ncsu.edu