National Water-Quality Assessment (NAWQA) Project
U.S. Geological Survey Circular 1291
Appendix 7A. Statistical summaries of pesticide compounds in stream water, 1992-2001
Summary statistics for pesticide compounds in stream water are presented below in tables as a function of land-use class (agricultural, urban, undeveloped, and mixed land-use sites). In addition, summary statistics by stream-water site and by land-use class are provided at the end of the appendix as downloadable files. The summary statistics include time-weighted detection frequencies at selected thresholds (detections greater than or equal to 1 µg/L, detections greater than or equal to 0.1 µg/L, detections greater than or equal to 0.01 µg/L, and detections at any concentration) and selected time-weighted percentiles of concentration (50th, 75th, 90th, 95th, and maximum). Summary statistics for stream-water sites also include time-weighted mean concentration.
Statistical summaries by land-use class:
Table 7A-1. Agricultural land use.
Table 7A-2. Urban land use.
Table 7A-3. Undeveloped land use.
Table 7A-4. Mixed land use.
How the summary statistics were calculated:
All statistics are based on a 1-year time series of concentrations selected for each stream site. These statistics were calculated by weighting each concentration in the time series by the amount of time it was used to represent the pesticide concentration in the stream. The time-weighted frequency of detection was calculated as the sum of the sample weights of samples in which the specific pesticide compound was detected at or above the threshold. Time-weighted percentiles of concentration were computed by ranking concentrations from low to high and cumulatively summing the sample weights. Time-weighted mean concentrations were calculated as the sum of the products of the sample concentrations and the associated sample weights. A detailed description of analytical approach and methods for statistical summaries is provided in appendix 8A.
Notes for using the tables:
Frequency of detection: Detection frequencies indicate how often a compound was detected in samples. Four detection frequencies are provided in the tables: "All" detections, detections at concentrations greater than or equal to 0.01 µg/L, detections at concentrations greater than or equal to 0.1 µg/L, and detections at concentrations greater than or equal to 1 µg/L. The values for "All" detections provide the percentage of detections for a given compound, but should not be directly compared among compounds because reporting levels (detection capabilities) varied among compounds. Because maximum long-term detection levels varied, detection frequencies were calculated using three common detection thresholds (0.01, 0.1, and 1 µg/L). The use of these detection thresholds facilitates comparisons among compounds by censoring detections to a common reference concentration. Adjustments of this type are essential in order to answer questions like "is compound x detected more often than compound y?"
For most pesticides analyzed by GCMS, the lowest appropriate detection threshold for comparisons among compounds is 0.01 µg/L. Several compounds (azinphos-methyl for example) have maximum long-term detection levels that exceed the 0.01 µg/L threshold. In these cases, the detection frequency is preceded by a ">" symbol to indicate that the true percentage of samples with concentrations greater than the threshold probably is greater than or equal to that reported in the table. For most pesticides analyzed by HPLC, the lowest appropriate detection threshold for comparisons among compounds is 0.1 µg/L. Nonetheless, several compounds have maximum long-term method detection levels that exceeded the threshold. In these cases, the detection frequencies may underestimate the true percentage of samples with concentrations greater than the threshold and detection frequencies are preceded with a ">" symbol. All of the pesticides can be compared at the 1 µg/L threshold. Use of this threshold censors the majority of the pesticide detections, but allows comparisons among pesticides measured at high concentrations.
Detections in water samples were time weighted and the pooled detection frequency was calculated across all stream-water sites in a land-use group. As a consequence, the frequency of detection may be interpreted as an estimate of the percentage of time that pesticides are detected at NAWQA stream-water sites in that particular land use. For example, 54 stream-water sites in areas of agricultural land use had sufficient monitoring data to estimate an annual distribution of concentrations of acetochlor (table 7A-1). Acetochlor was detected 29.87 percent of the year at stream-water sites draining agricultural land use and was detected 21.95 percent of the year at concentrations greater than or equal to 0.01 µg/L.
Frequency of detection, in percent, was rounded to the hundredths place.
Percentiles of concentration: Annual time-weighted concentrations measured for each pesticide are summarized using percentiles. The 50th, 75th, 90th, and 95th percentiles of concentration by land-use category are provided. Percentiles provide information about the magnitude and duration of concentrations at selected points in the cumulative frequency distribution of the ranked, time-weighted concentrations. For example, concentrations of acetochlor at NAWQA stream-water sites draining agricultural land use were nondetections for 50 percent of the year, less than or equal to 0.008 µg/L for 75 percent of the year, less than or equal to 0.045 µg/L for 90 percent of the year, and less than or equal to 0.130 µg/L for 95 percent of the year (table 7A-1). If percentiles are detected concentrations, the percentiles also may be interpreted as the percentage of the year where concentrations were greater than a given concentration. For example, concentrations of acetochlor were greater than or equal to 0.008 µg/L for 25 percent of the year, greater than or equal to 0.045 µg/L for 10 percent of the year, and greater than or equal to 0.130 µg/L for 5 percent of the year (table 7A-1).
The tables include a column for "Maximum." The concentrations in this column are the maximum measured concentrations in the pooled data sets for each land-use class. Concentrations greater than the value reported as maximum in the tables may have been measured at stream-water sites but are not reported because they were not measured during the 1-year periods of data selected to describe the annual distribution of concentrations at these sites. In addition, the probability is low that a stream-water sample was collected at the time of the annual maximum concentration for any pesticide at any stream-water site. Consequently, the maximum concentrations reported in the tables should be interpreted as a lower bound for the true maximum concentrations.
All concentrations in the tables are rounded to the thousandths place. Nondetections are denoted as “ND.”
Pesticide names and analytical methods are presented in Appendix 1A.
Information on sampling sites and their characteristics is presented in Appendix 5A.
Downloadable concentration data are presented in Appendix 6A.
Additional information on analytical approach and methods is presented in Appendix 8A.
For more information, contact:
Jeffrey D. Martin
U.S. Geological Survey
NAWQA Pesticide Synthesis Project
5957 Lakeside Boulevard
Indianapolis , IN 46278-1996
voice: (317) 290-3333 x148
fax: (317) 290-3313