U.S. Department of the Interior Water-Quality Assessment of the U.S. Geological Survey Trinity River Basin, Texas Pesticides in By L.F. Land Urban and Agricultural Streams, 1993-95

Nonpoint sources of pesticides are a national issue and are ot major Ten insecticides were detected in the urban streams and 10 in the interest in the U.S. Geological Survey National Water-Quality Assess­ agricultural streams. Although the number of compounds detected was ment (NAWQA) Program because of their potential toxicity to humans the same, the frequency of detections was much greater in the urban and aquatic life. Using the water-quality data collected by the Trinity streams, with carbaryl. chlorpyrifos, and diazinon occurring in more River Basin NAWQA study-unit team, an analysis of pesticides in than half the urban and agricultural streams was made by aggregating and grouping urban samples. data from several networks. The samples were collected during March Diazinon was HERBICIDES , t i | i i , i ^j^^^^j^^ 1993 through September 1995. About 70 samples were collected from detected in all Atrazine -' Metolachlor - ~ ' ^ {^ I^^BH urban streams and 70 from agricultural streams. The urban dataset con­ the samples Prometon - iHMHHMM sists of sampling sites in the Dallas-Fort Worth metropolitan area where Simazine - ^ H from urban Tebuthiuron - more than 3.5 million people lived in 1990. Land use in the sampled streams and Fluometuron - M^M^M Alachlor - urban watersheds is primarily residential, but there are commercial in about 60 2,4-D - areas, such as shopping centers and office complexes along major high­ DCPA - - percent of the Pendimethalin - ways and streets. The agricultural dataset consists of sampling sites in Metribuzin - M samples from Trifluralin -* the watershed of Richland-Chambers Reservoir, an area in the west- agricultural Thiobencarb - - central part of the Trinity River Basin where fertile soils are extensively Molinate ^ - streams. Other Diuron - - cultivated. The land use is mostly cropland and pasture with some Cyanazine - than diazinon, Dichlorprop - Urban Agricultural - rangeland. In the cultivated areas, the major crops are corn, cotton, and Oryzalin - streams « streams ~ all insecticides Propanil - ; sorghum. In the rangeland areas, cattle are the dominant livestock. In in the agricul­ MCPA - the pasture areas, hay is grown along with grass for cattle feed. None of DNOC - tural streams Pronamide - , the crops are irrigated. Propachlor - occurred in less EPTC - Dicamba - I The material in this fact sheet is condensed from a report by Land than 10 percent Napropamide - 1 ' and Brown (1996) where the study is presented in greater detail. of the samples. Dichlobenil - I Bentazon -, ,', 1 , , , , 1 ,-..,.....!...., 1 , ,....,.-- 100 50 0 50 1C)0

EXPLANATION INSECTICIDES ' ' ' ' ' ' ' WftfWaM ' ' *-i Land use or land cover Diazinon -l IB Urban ^ Water Chlorpyrifos - I' Carbaryl - - I | Agricultural | | Wetland Malathion - I ~- Dieldrin - 1 | | Rangeland | j Barren land Carbofuran - -

p,p'-DDE - 1 ~ | | Forest land Methomyl - Urban Agricultural - methyl-Azinphos - streams f streams ~ Fonofos - Sampling site in cis-Permethrin - i urban area alpha-HCH -, . , , 1 , , j , J , , , ,-i ,,,.,-< Sampling site in 100 50 0 50 1C)0 agricultural area PERCENT DETECTION

FREQUENCY OF DETECTION OF PESTICIDES

Study area What Were the Concentrations of Commonly Detected Pesticides? Land use and land cover modified from U.S. Geological Boxplots of the five most frequently detected Survey, 1990, and Hitt, 1994 herbicides show that concentrations of atrazine and 80 KILOMETERS metolachlor generally were greater in the agricul­ tural streams, while prometon and simazine gener­ ally were greater in the urban streams. Tebuthiuron Which Pesticides Were Detected? concentrations generally were greater in the urban streams, but the max­ Twenty-four herbicides were detected in the urban streams and 19 in imum concentration was in an agricultural stream. the agricultural streams. Fifteen herbicides were detected in both areas. Three herbicides atrazine, 2,4-D, and metolachlor had maxi­ Atrazine is the most commonly detected herbicide, occurring in all sam­ mum concentrations equal to or greater than 1.0 microgram per liter ples from both areas. Metolachlor was detected in about 80 percent of (|lg/L) in both the urban and agricultural streams. In the agricultural the samples from the urban streams and in all samples from the agricul­ area, two other herbicides, alachlor and fluometuron, had maximum tural streams. Two other commonly detected herbicides are prometon concentrations greater than 1.0 JJ.g/L. Atrazine not only was the most and simazine. These occurred in about 90 percent of the urban samples frequently detected herbicide but also had the greatest median concen­ and in about 60 percent of the agricultural samples. trations for both urban and agricultural streams. Boxplots of the three most frequently detected insecti­ LU ioo - - cides show that the concentrations of each were greater in I t I I 1 1 \> 1 1 1 1 1 the urban streams than in the agricultural streams. Car- o £ 10 1 A AA Atrazine baryl and diazinon had maximum concentrations greater I- Q. < W \ A A HA * A H ^ 1 than 1.0 Jlg/L in urban streams. 2 < A * A A "VA A 4 ^ AA * ~ u] cc A A * A A A o o i A A » A The U.S. Environmental Protection Agency has estab­ §g 0.1 o o A A * lished a maximum contaminant level (MCL) and a health 1 j I I i ! i I I ! 1 . . z 0.01 advisory (HA) level for many pesticides (Nowell and Resek, 1994). Atrazine was the only herbicide with con­ CC 10 1 1 i I f 1 1 1 ! 1 1 1-HI centrations greater than the applicable water-quality stan­ - -1 Diazinon dards. The MCL and HA for atrazine are 3 |Llg/L. In the Q uj A

agricultural streams, approximately 20 percent of the sam­ L±cc £22 U01-' 2 < ples exceeded this level. Diazinon was the only insecticide 111 rn Z O °-01 - A A A& 4 | Ai*± A * ^ A ^ A ^ AA A - with concentrations greater than the applicable water- o cc tJfc ^fc 00 quality standards. The HA for diazinon is 0.6 |Llg/L. (No 2 0 001 ' , 1 1 1 1 1.1 1 A .1 1 ! ...... I...... MCL for diazinon has been set.) In the urban streams, z JFMAMJJASOND approximately 15 percent of the diazinon samples EXPLANATION exceeded this level. Urban streams A Agricultural streams A No detection, agricultural streams

SEASONALITY OF PESTICIDES References Hitt, K.J., 1994, Refining 1970's land-use data with 1990 Herbicides population data to indicate new residential develop­ S ment: U.S. Geological Survey Water-Resources Inves­ CONCENT ICROGRAM tigations Report 94^-250, 15 p. Land, L.F., and Brown, M.F., 1996, Water-quality assess­ 1I 1 ment of the Trinity River Basin, Texas Pesticides in Atrazine Metolachlor Prometon Simazine Tebuthiuron streams draining an urban and an agricultural area, 0.001 1993-95: U.S. Geological Survey Water-Resources

10 EXPLANATION Investigations Report 96^114, 22 p. Urban Agricultural Nowell, L.H., and Resek, E.A., 1994, Summary of

i Maximum value national standards and guidelines for pesticides in 0.1 75th percentile water, bed sediment, and aquatic organisms and their Median applications to water-quality assessments: U.S. Geo­ 0 CC 0.01 logical Survey Open-File Report 94^55, 42 p. 0 0 25th percentile Carbaryl Chlorpyrifos Diazinon z 0.001 Minimum value U.S. Geological Survey, 1990, Land use and land cover digital data from 1:250,000- and 1:000,000-scale maps: Detection limit U.S. Geological Survey Geodata Users Guide 4, 33 p.

CONCENTRATIONS OF THE MOST COMMONLY DETECTED PESTICIDES Information on technical reports and hydrologic data related to the NAWQA Program can be obtained from: When Were the Concentrations Greatest? Project Chief Seasonal concentrations of the pesticides are illustrated by graphing Trinity River Basin NAWQA Study U.S. Geological Survey the concentrations for the most commonly detected herbicide (atrazine) 8011 CameronRoad and the most commonly detected insecticide (diazinon) versus sampling Austin, Texas 78754-3898 date without regard to year. E-mail: [email protected] The atrazine graph shows a great amount of scatter in concentration Any use of trade, product, or firm names is for descriptive in both urban and agricultural streams. During April, the median con­ purposes only and does not imply endorsement by the U.S. centration of atrazine was 0.8 jlg/L in the urban streams and 6.0 |Llg/L in Government. the agricultural streams. During September to January, when concentra­ The National Water-Quality tions tended to be low, atrazine concentrations were about one-tenth of Assessment Program the concentrations in April in both types of streams. The U.S. Geological Survey began the National The diazinon graph shows a great amount of scatter in concentration Water-Quality Assessment (NAWQA) Program in 1991 to assess the status of and trends in the quality of in the urban streams but not in the agricultural streams. In the urban the Nation's streams and aquifers. The program is streams, concentrations were greatest in April and were relatively high designed to enhance understanding of natural and human factors that through August. During April, the median concentration was 0.21 Jlg/L. influence water quality, and consists of studies in 60 major river basins During September to January, when concentrations tended to be low, and aquifers of the United States. Together, the 60 studies compose about one-half of the land area of the United States and 60 to 70 per­ diazinon concentrations were about one-half of the concentrations dur­ cent of the water use and population served by public water supplies. ing April. Most of the diazinon concentrations in the agricultural The similar design of each study and use of consistent methods allow streams were at or near the method detection limit; thus, no seasonality comparisons at regional and national scales. This information is being is evident. used to manage water resources at the national, State and local levels.

July 1996 Fact Sheet FS-178-96