Methods development and implementation for the National Pesticide

David J. Munch, Robert L. Graves, Robert A. Maxey, and Tina M. Engel. Environ. ... MacCarthy , Ronald W. Klusman , Steven W. Cowling , and James A. Ri...
0 downloads 0 Views 8MB Size
Methods development and implementation for the National Pesticide Survey

. David J. Munch Robert L. Graves U S . Environmental Protection Agency Cincinnati, OH 45268 Robert A. Maxey US.Environmental Protection Agency Stennis Space Center, MS 39529-6000 Tina M. Engel Batrelle Columbus, OH 43201 Dunng the last decade, the amount of information indicating the presence of pesticides in groundwater has steadily increased. In 1988 EPA released data confirming that at least 46 pesticides had leached into the groundwaters of 26 states ( 1 ) . These data result from analy-

ses performed by universities, pesticide manufacturers, and state and federal governmental agencies. Although indicative of the extent of pesticide contamination of groundwater, these data could not be used to form valid conclusions concerning the significance of pesticide contamination of groundwater on a Mtional basis. In 1984, a joint project between EPA's office of Drinking Water (ODW) and the Office of Pesticide Programs (OPP) was initiated to conduct a statistically based survey of pesticide contamination of drinking-water wells. The National Pesticide Survey (NPS) has two primary objectives. First, the survey will provide statistically valid data which can be extrapolated to represent both rural domestic and community drinking-water

1446 Environ. Sci. Technol., Vol. 24. No. IO, 1990

i

wells nationally. Second, the survey will evaluate possible associations between pesticide contamination of drinking water wells and pesticide use and hydrogeological vulnerability. m e data ohtained in this survey will be used by the ODW to identify compounds for future regulation and will allow better assessments of drinking-water treatment methods for the removal of these compounds. The OPP may use these data in setting labeling and use restrictions for pesticides, &d to assist in defining monitoring requirements of pesticide registrants. A preliminary report on the results of the NPS is scheduled for release in the fall of 1990. The preliminary report will contain summaries of the data properly weighted for the statistical design and with variance statements for each of the

w 1 3 - ~ 3 6 w 9 o m ~ 2 ~ 1 ~ 6@ so 1990 ~.~ American m Chemical Society

parameters. A fat more detailed final report will be issued in 1991.

Analyte selection In 1984, beginning with a list of approximately 600 active ingredients used in 1982 for planning pesticide reregistrations, pesticide and pesticide decomposition products were identified as potential survey analytes. The number of potential analytes was reduced by choosing primarily those compounds that met certain selection criteria. These criteria included use of at least 1,ooO,ooOIbs in 1982, water solubility greater than 30 m a , and hydrolysis

half-life longer than 25 weeks. Pesticides and pesticide decomposition products previously detected in pundwater, as well as pesticides regulated under the Safe Drinking Water Act, were automatically included in the list of survey analytes. An initial list of 161 chemicals was developed in mid-1984. and was reduced by October of 1986 to a total of 100 pesticides and their degradation products identified as “priority” analytes. With the production of the fmt list of potential analytes in mid-1984, attempts to identify appropriate analytical methods were made by conducting literahue searches and consulting with pesticide analyses experts. Compounds were assigned to one of the following four categories:

a published analytical method was available or, based on the compound’s chemical properties, it could easily be incorporated into an existing method, no analytical method existed, but development of a method would require only a modest effort; extensive methods development was necessary, but the effort was deemed worthwhile because of the importance of the pesticide: or * the amount of effort necessary to develop an analysis method exceeded the need for inclusion of the uesticide in the survey. In addition to the needs of the NPS, it was also deemed necessary to revise EPA’s overall pesticide analytical methods, which required a significant methods development effort. To maximize the utility of the final methods, all additional pesticides listed in published methods identified previously and their significant degradation products were designated as “nonpriority” analytes. These analytes would be included in the fmal methods as long as their inclusion did not require significant additional expense. It was necessary that these methods use those state-of-the-art techniques and quality control measures that were amenable to future use by EPA and others for studies conducted beyond the NPS. By early 1985 research into the analyses of these compounds was necessary. Several methods development research needs were identified as follows. Methods development research. Analytical methods could not be identified that were capable of determining, in a routine labratory setting, the presence of several potential analytes. Therefore, methods development was necessary for these chemicals. Methods consolidations. By optimizing the extraction and analyte separation techniques for a broad range of pesticides and related compounds, analyses could be performed in the most costeffective manner. Research into confirmation techniques. To satisfy the need for highquality analyses, it was decided to maximize the use of capillary column gas chromatographic separations. This would facilitate both qualitative and quantitative analyte confiiations using a second dissimilar chromatographic column, and the use of gas chromatography/mass spectrometry (GC/MS) for positive qualitative confirmaton. Sample preservation studies. Little information was available on the susceptibility of the majority of the potential analytes to chemical or biological degradation in stored water samples. Therefore, research into methods of assuring analyte stability was necessary.

Ruggedness testing and validation

of final methods. The methods needed to be rugged enough to be used by a large number of laboratories of varying experience and still produce high quality data. Also, it was necessary to perform a series of precision and accuracy studies on reagent and field samples fortified at the full range of applicable analyte concentrations, in order to demonstrate each method’s performance. The methods development effort was initiated at the Battelle Columbus Division, which was under contract to the Environmental Monitoring and Systems Laboratory-Cincinnati (EMSL-CI). Work began at Battelle in April 1985 under the joint direction of the Technical Support Division of the ODW, and the EMSL-CI. Methods development and consolidation Reviews of the available literature indicated that analytical procedures existed for most of the priority pesticides. However, these methods varied widely in procedure and application. The primary goal of the methods development study became the consolidation of these many analytical procedures into a smaller group of methods. Because of the short amount of time available prior to initiation of the survey, efforts were directed towards retaining the proven aspects of previously published methods, while developing the procedures necessary to assure the quality of the analyses. In this manner, the original published methods were transformed into methods amenable to use in the NPS as well as for future pesticide in water analyses needs. The major work of determining proper sample preparation, separation, and detection techniques was accomplished withim the fvst three months of research. Only the method development for ethylene thiourea took substantially more effort. The six methods developed are as follows. NPS Method 1. Determination of Nitrogen Phosphorous Containing Pesticides in Ground Water by Gas chromatography with a Nitrogen Phosphorous Detector (39 analytes, Table 1). A measured volume of sample of approximately 1 L is solvent extracted with methylene chloride by mechanical or manual shaking. The methylene chloride extract is isolated, dried, and concentrated to a volume of 5 mL after solvent substitution with methyl tertiary butyl ether (MTBE). Chromatographic conditions are identified for the separation of the analytes using both primary [30 m x 0.25 mm inside diameter column 0.D.) bonded phase (DB)-5] and confmational (30 m x 0.25 mm LD. DB-1701)

Environ. Sci. Technoi., Vol. 24, No. I O , 1990 1447

fused silica capillary GC columns. NPS Method 2. Determination of Chlwinated Pesticides in Cmund Water by Gas Chromatography with an Elechon Capture Detector (25 analytes, Table 2). A measured volume of sample of approximately 1 L is solvent e x " J with methylene chloride by mechanical or manual shaking. The methylene chloride extract is isolated, dried, and wncentrated to a volume of 5 mL after solvent substihltion with MTBE. Chromatographic conditions are identified for the separationof the d y t e s using both primary (30m x 0.25 nun ID. DB-5) and confirmational ( 3 0 m x 0.25 nun ID. DB-1701) fused silica capillary Gc columns. NPS Method 3. Determination of Chlorinated Acids in Ground Water by Gas Chromatography with an Electron Capture Detector (13 analytes, Table 3). A measured volume of sample of approximately 1 L is adjusted to pH 12 with 6 N sodium hydroxide and shaken for one hour to hydrolyze acid derivatives. Extraneous organic material is removed by a methylene chloride wash. The sample is acidified, and the chlorinated acids are extracted with ethyl ether by mechanical or manual shaking. The acids are convened to their methyl esters using dimmethane as the derivatization reagent. The ethyl ether extract is concentrated to a volume of 10 mL after solvent substitution with MTBE. Chromatographic conditions are identified for the separation of the analytes using both primary (30 m x 0.25 mm LD. DB-5) and confmational (30 m x 0.25 mm I.D. DB-1701) fused silica capillary GC columns. The method provides a Florisil cleanup procedure to aid in the elimination of matrix interferences that may be encountered. NPS Method 4. Determination of Pesticides in Ground Water by High Performance Liquid Chromatography with an Ultraviolet Detector (16 analytes, Table 4). A measured volume of sample of approximately 1 L is solvent extracted with methylene chloride by mechanical or manual shaking. The methylene chloride extract is isolated, dried, and concentrated to a volume of 5 mL after solvent substitution with methanol. Chromatographic conditions are identified for the separation of the analytes using both primary (250 mm x 4.6 mm I.D. Du Pont Zorbax ODS) and confmational (250 mm x 4.6 mm LD. J+W Cyano) high performance liquid chromatographic columns. NPS Method 5. Measurement of Nmethyl Carbamoyloximes and N-methyl Carbamates in Ground Water by Direct Aqueous Injection High Performance Liquid Chromatography (HPLC) with Post-Column Derivatization (10 analytes, Table 5). The water sample is fil1448

Environ. Sci. Technol.. Vol. 24, No. 10, 1990

TABLE l

Characterlsticsof National Pesticide Survey Method 1 laboratory fortified field samples' Minlmum quantlfkatlon N u " ot limll(pg/L) nmplu

1.o 0.26 0.34 0.24 2.2 1.5 0.60 1.o 0.70 0.40 0.24 0.43 0.30 0.12 0.30 0.40 1.8 0.27 0.30

Aiachlor

Amelryn

Atraton Atrazine Bromacil htachlor

3%" Chlorpropham

Cycbate

DlChlONCS ~. . ..

Dii;gnamid Ethopmp

Fenamiphos Fenarimol Flurid one

Hexazimrne

Meth I paraoxon M .e..i d c h.i o.r

1.5

Metfibuzin

0.36 0.30 2.1 0.36 0.50 0.36 0.38 0.29 0.20 0.20 0.10 0.75 0.36 0.45 3.5 0.30 0.32 1.2 0.37

sti#lOS

Teffithturon Terbaoil

leI!Jutiyn

Triademefon Tricydazole

Vetnolate

30 26 25 30 25 24 27 26 23 25 27 26 25 28 28 27 27 26 30 26 27 26 26 25 24 27 26 27 27 27 26 24

30 27 26 29 27 26 26

Mnn pnmt

m o w

98 97

RSD' (%)

99 98 97 88 89 99 94 97 99 93 98

3.6 9.6 12 4.2 5.8 10 5.9 7.5 5.6 5.4 5.9 5.6 3.6 2.9

94

16

110

100 97 97 110

97 95 96 87 96 100 98 90 97 96 97 96 97 100

94 100 98

99 100

90

10 12 10 6.8 9.5 6.1 6.1 12 3.7 10 9.2 5.9 7.2 6.9 5.9 8.2 8.1 5.6 7.0 9.0 3.9 8.4 11

6.0

Mlnlnnnn X Maximum% movsfy movrry

89 81 71 89

I00

88

120 140 110 110

79 78 72

1M

83

1w

110 100 100 110 98 100 130 1.20 110 120 130

80

a2 88 85 92 60 73 66 70 92 81 77 80 62 87 77 77 77

97

120

100 110 100

83

100 130 120 96 120 110 110

73

110

83

120 120 100

81

76

88 81 81 89 74 79 76

120 110 120 1.20

1w

Characteristics of National Pesticide Survey Method 2 laboratory fortified field samples" Minimum

limit (FOIL)

samples

Mea" percent recovery

RSO' (96)

0.12 0.12 0.12 1.4 0.12 0.12 0.25 0.12 0.30 0.12 0.25 0.25 0.25 0.25 0.12 0.12 0.085 0.12 0.1 2 0.1 2 0.60 1.8 3.9 1.3 0.25

31 29 29 31 31 28 29 29 31 31 29 31 31 29 30 29 31 29 31 28 29 31 29 31 29

80 91 96 94 74 110 92 88 92 91 93 93 92 100 92 99 110 95 90 96 120 100 95 99 110

12 11 11 12 16 9.9 18 16 21 12 17 17 15 13 8.2 11 12 8.1 11 9.1 30 26 27 9.4 16

quantification Number of

Aldrin

a-Chlordane

Chlordane hloroneb EChlorothaionii DCPA

4,4-DDD 4,4-DDE 4.4-DDT

Dieldrin

Endosulfan sulfate Endrin Endrin aldehyde Etridiazole a-HCH P-HCH rHCH

Heptachlor Heptachlor-epoxide Hexachlorobenzene Methoxychlor cisPermethrin trans-Permethrin Propachlor

Trifluralin

*Spiking level 10 x minimum quantification limit in N g l l %Relative Standard deviation of the ~ P C O Y P N

Minimum % Maximum % recovery recovery

4BLE .,

hancteristicsof National Pesticide Survey Method 4 laborato icalion Number of it(pll/L) samples

Imine deethyhted 4.4 arban 3.8 arbolumnphenol 42

yamne

iumn

wmoxlde

luometumn Ketwubhran 1" nurm letnbuzin DA eburon

34

47

41 38 41

0.63

34

31

Maan

permnl ncovery

RSP

63 91 94

7.5 8.8 I1

93 94 94

6.5

92 110 110 110

5.1 6.7

76 83 79

110

94 90

7.8 7.1

79 74

It0

5.8 7.0

75 71 59 67

1W

90 85

loo

78 81

100 110

41 34

1.9 0.95 1.2

40

0.6

41

89 91 79 89

6.3 0.60 11 0.30

33 34 41 34

97 94 92 94

40

ronamlda

metabolic ropanil @am Wep

Minimum % Maxlmun rscovary m v a

67 61

9.4 0.95

41

(%)

$!king level 10 x minimum quantlcatwn limit In pgiL Wative standard deviation 01 the recovery. *(1 .ldtme1hylacetonyi)-3,5dohlombenzam~de

11

7.5 3.4 3.9 5.4 7.3

69

loo 100

100

90 100

99

tered, and a 400 fiL aliquot is injected onto a reverse-phase HPLC column. Separation of the analytes is achieved using gradient elution chromatography. After elution from the HPLC column, the analytes are hydrolyzed with 0.05 N sodium hydroxide at 95 "C. The methyl amine formed during hydrolysis is reacted with o-phthalaldehyde and 2-mercaptoethanol to form a highly fluorescent derivative which is detected using afluorescence detector. Chromatographic conditions are identified for the separation of the aualytes using both primary (150 mm x 3.9 mm I.D. Waters Novapak C18) and confiiational(250 mm x 4.6 mm 1.D. Supelco LC-1) HPLC columns. NPS Method 6. Determination of Ethylene Thiourea (ETU) in Ground Water by Gas Chromatography with a Nitrogen-Phosphorous Detector (1 analyte, Table 6). The ionic strength and pH of a measured 50-mL aliquot of sample are adjusted by the addition of ammonium chloride and potassium fluoride. The sample is poured onto an Extrelut column. ETU is eluted from the column with 400 mL of methylene chloride. The extract is solvent exchanged to ethyl acetate and concentrated to a volume of 5 mL. Conditions are identified for the chromatography of ETU using both primary (10 m x 0.25 mm LD. DB-Wax) and confinnational (5 m x 0.25 mm LD. DB-1701) fused silica capillary gas chromatography columns. Sample and extract preservation studies A limited preservation study was couducted during methods development. Samples of a groundwater from central Ohio were acquired for these experiments. First, this water was analyzed using each of the six NPS analytical methods to determine if any of the analytes of interest or analyte interferences were observed. Replicate samples were fortified with method analytes and then analyzed after 0, 14, and 28 days' storage at 4 "C. The extracts obtained during the analyses of the Day 0 samples were also analyzed after 14 and 28 days' storage. Serious analyte degradation upon storage was observed for 24 of the 145 analytes studied, including 22 Method l analytes, and two Method 2 analytes. Eighteen of the Method 1 analytes (Aspon, azinphos methyl, Bolstar, demeton0, dichlofenthion, EPN, ethion, ethyl parathion, famphur, fenitrothion, fensulfothion, fenthion, fonofos, malathion, methyl parathion, MGK 326, phorate, and phosmet) and two of the Method 2 analytes (chloropropylate and Dicloran) that degraded during storage had previously been designated as nonpriority and were therefore deleted from the Environ. Sci. Technol., VoI. 24,NO. IO, 1990 1449

scope of these two methods. The remaining four Method 1 analytes (disulfoton sulfoxide, prometon, Pronamide, and terbufos) were designated priority and were therefore kept within the method scope pending further preservation studies conducted in both the pilot and full surveys.

TABLE 6

Characteristicsof National Pesticide Survey Method 6 laboratory fortified field samplese

Ethylene thiourea

Pilot study

A pilot study was conducted in March and April of 1987. The main objective of this study was to test all aspects of the planned survey implementation. Eight domestic and eight community sites were sampled in each of three states: Califomia, Minnesota, and Mississippi. Duplicate samples from all sites were analyzed at one of two EPA laboratories-either the Technical Support Division of the Office of Drinking Water or the Environmental Chemistry Section of the Office of Pesticide Programs-and at Battelle’s Columbus Division. In addition to sample analyses, limited analyte preservation and extensive sample spiking programs were conducted. Due to analytical problems encountered early in the pilot study, limited information was obtained from the analyte preservation studies. However, the information that was obtained tended to confirm the results of the preservation studies conducted previously during development of the analytical methods. Except for Method 3, it was demonstrated during the sample-spiking program that good accuracy and precision could be obtained using all of the analytical methods. After introducing several procedural changes, primarily conceming the details of the derivatization step, good accuracy and precision were obtained using Method 3 for the remainder of the pilot study. For the pilot study, each laboratory separately purchased the “pure” compounds to be used for the production of analytical standards. To facilitate comparisons of data obtained at different laboratories, aliquots of each primary calibration standard were exchanged between Battelle and the appropriate EPA laboratory for analyses. Data resulting from these analyses generally agreed favorably. However, for a limited number of analytes, problems involving either the inaccurate production of primary calibration standards or the use of impure neat compounds were detected.

Full survey The key elements of the final design of the survey included: a sample site selection program based on obtaining a statistically valid sample set of approximately 600 community system and 800 rural domestic wells used as a source of 1450

9.0

36

7.4

92

80

112

Spiking level 10 x minimum quantitication limit in p@L %elalive standard deviation of the recovery.

TABLE 7

Characteristics of interlaboratory method validation studies NPS’ method

Study period

1 2

Jan.Juiy 1989 Nov.-Dec. 1988

3 4

5 6

July-Aug. 1988 Nov.-Dec. 1989 J u n d u l y 1988 Apr.-May 1990

Number 01 labs

9 11

8 10

8 ?

Collaborators EPNAOACb EPAJAOAC EPA EPNAOAC EPA EPNAOAC

Wational Pesticide SuNey. bAAssociation of Otticial Anaiyfical Chemist:

drinking water, stratified as to hydrogeological vulnerability and pesticide use; analytical methods capable of accurately quantifying I27 pesticides, pesticide degradation products, and related compounds; health advisories for 62 pesticides of primary interest to EPA; and information conceming well conshuction and pesticide use pattems, specific to each sample collection site. In April of 1988, sampling began for the full survey. Sampling of community water systems was completed in December 1989, and of the domestic systems in February 1990. Because of the problems encountered with the production of calibration standards in the pilot shtdy, all latmatories were provided with calibration standards from a common source. The purity of all compounds used as calibration standards was certified through the EPA Pesticide and Industrial Chemicals Repository. Using these certified materials, The Bionetics Corporation produced stock (12 mg/mL) standards, sealed in glass ampules, of each analyte, intemal standard, and surrogate compound. These standards were verified by the appropriate EPA laboratory each time that a new set of ampules was prepared. Following verification by the EPA laboratories, these ampules were supplied to each NPS laboratory. The identification of all pesticide and pesticide-related analytes detected was both qualitatively and quantitatively confmed. C o n f i t i o n s were conducted by reanalyses of all sample extracts using a second capillary gas chromato-

Environ. Sci. Technol., Vol. 24, No. 10, 1990

graphic or high performance liquid chromatographic column. These analyses provided both a preliminary qualitative confmation for all GC determinations and the final confirmation for HPLC analyses. The analytes detected using a GC-based method were also qualitatively confirmed using GC/MS. Presented in Figure 1 are the results of the evaluation of the selectivity of the primary and confirmational columns used in Methods 1-3. Using the Method 1 data as an example, the presence of an analyte was successfully c o n f i i e d using the confiiation column in 45.5% of the cases in which an analyte was detected using the primary column; of those analytes confimed by analyses using the confirmation column, 82.5% were successfully c o n f i i e d by GC/MS. Only the data for Methods 1-3 are presented because for Method 4 none of the analytes detected using the primary column were c o n f i e d upon second column analysis and too few detections were observed using Methods 5 and 6 to provide a meaningful evaluation. The Minimum Quantification Limit (MQL) was determined by multiplying the estimated detection limit (EDL) by a factor determined in part by the degree to which the method is subject to interferences (Method 1 = 4 x EDL, Methods 2 4 = 5 x EDL, and Methods 5-6 = 3 x EDL). The EDL was determined from results of seven or more replicate analyses of a reagent water sample fortified with the concenuation of the analyte that would yield a chromatographic peak with a

port Division of the Office of Drinking Water; Auhry E. Dupuy and the staff of the Environmental Chemisy Section of the Office of Pesticide Programs; and the staffs of the Environmental Monitoring Systems Lahoratory-Cincinnati and the Battelle Memorial Institute. This work would not have been possible without their talent and efforts.

% o f2nd miumn delenions succersluliy confirmed using GCIMS

Reference (1)

Pesticides in Ground Wafer Darn Base. 1988 Interim Report: Office of Pesticide Programs; U. S. Environmental Protection

Agency: Washington, Dc, 1988.

signal-to-noise ratio of approximately 5 to 1. The standard deviation of these data was then multiplied times the students’ t value appropriate for a 99% confidence level and a standard deviation estimate with n-l degrees of freedom. The EDL was defined as either the concentration of analyte yielding a 5-to1 signal-to-noise ratio or the calculated concentration, whichever was greater. Field samples were fortified at the primary laboratories throughout the course of the survey, with the analytes measured using each of the analytical methods. Approximately 40 different field samples were fortified at concentrations of 2 or 5 times the MQL of each analyte, while 20 to 30 samples were fortfied in duplicate at a concentration equal to 10 times the MQL for each analyte. Summarized in Tables 1-6 are the preliminary results of the IO x MQL concentration field sample fortification program. These data clearly demonstrate that these analytical methods can be used to obtain highly accurate and precise results over time in a variety of groundwaters analyzed by laboratones operating in a routine analysis mode. Reagent water fortification and analyte stability studies also were being conducted at the IO x MQL spiking level. In addition to using the fortified reagent water sample data for monitoring laboratory performance for each set of samples analyzed, comparisons of these data and data from the field sample fortification programs will be used as an indication of any matrix effects. At this time the analyte stability studies are still

in progress. However, the data evaluated to date appear to confirm those previously obtained, with the exception that no instability problems have been encountered for prometon. Current methods status In addition to the validation studies discussed above, a series of interlaboratory method validation studies is also being performed. Several of these studies are being conducted in collaboration with the Association of Official Analytical Chemists (AOAC). Table 7 is a listing of the method by number; the study period the number of collaborating laboratories: and whether the study was performed using paid EPA contractors or volunteer participants provided through the AOAC. In either case, the study designs and data evaluations are being performed by EPA. Four of these methods have been proposed for use in compliance monitoring under the Safe Drinking Water Act (SDWA): NPS I (EPA507), NPS 2 (EPA 5081. NPS 3 (EPA 515.11. and NPS 5 (EPA 531.1): These and’other methods proposed for use in conjunction with the SDWA are contained in a manual titled “Determination of Organic Compounds in Drinking Water,” order number PB-89-220461/AS, available through the National Technical Information Service, 5285 Port Royal Road, Springfield, VA 22161. Acknowledgments The authors would like to thank Herbert J. Brass and the staff of the Technical Sup-

David J. Munch ( I ) is EPA’s Technical Support Division project manager for the National Pesticide Survey. He was also the project manager for the development of the analytical methods described in this paper. Munch received a B.S. degree in from the University of Cincinbeen employed as a chemist %?%?has with EPA for the past 15 years. Robert L. Graves ( r ) is chief of the Development and Evaluation Branch, Qua& Assurance Research Division, Environmental Monitoring Systems Laboratory, EPA, Cincinnati, OH. He has B.S. and M.S. degrees in chemistry from the UniDayton, Dayfon, ,OH, a@ at versity M B A . om Xavier University in Cincin nati.

$’

Koben A. Marey is a chemisr and rhe National Pesticide S u n r y project leader ar the En, ironmental Chrmisrry Lahoraror). in Minissip i He has senvd as u rhemisr und ream /%der for numerous eni’irnnmenrul srudies involving ecricides and huiurdou chemicalc. He ios a B A . degree frnm the l’niirrsiry of Southern Mi.r. ckcippi. Tina M. Engel norked at BarreIIe for / I )ems as princfal research zcienritf. During rhur rime, er major inferem were in rhe deL,elupmem