Environmental Monitoring Results of the Mediterranean Fruit Fly Eradication Program, Riverside County 1994

C. Ando¹, R. Gallavan¹, P. Wofford¹, A. Bradley¹, D. Kim¹, P. Lee², and J. Troiano<sup>1 

1</sup>Department of Pesticide Regulation Environmental Monitoring and Pest Management Branch
² California Department of Food and Agriculture, Center for Analytical Chemistry

September 1996


California Environmental Protection Agency Department of Pesticide Regulation Environmental Monitoring and Pest Management Branch Environmental Hazards Assessment Program 1020 N Street, Sacramento, California 95814-5624
Report Number EH 95-2

ABSTRACT 
To eradicate an infestation of the Mediterranean fruit fly, the staff of the California Department of Food and Agriculture (CDFA) aerially applied malathion-bait mixture to a 38 km² region in the residential areas of Corona and Norco in Riverside County, California. The Environmental Hazards Assessment Program (EHAP) staff of the Department of Pesticide Regulation monitored five of the eight CDFA aerial applications made at two week intervals from February 15, 1994 to May 24, 1994. Malathion and malaoxon concentrations were measured on land, air, water, soil, turf, green leafy lettuce, sand, and stainless steel sheet surfaces. Diethyl fumarate (DEF), a co-product of the malathion product manufacturing process and a potential skin irritant, was also measured on steel sheet surfaces collected immediately following aerial application.

The malathion mass deposited on land averaged 8.91 mg/m² for three monitored aerial applications and was equivalent to 87% of the expected malathion deposition rate. The 95% confidence interval for the combined average of the three sprays ranged from 7.91 to 9.90 mg/m². Higher deposition rates were measured in the previous Medfly eradication programs in 1990 and 1981, but they were a result of the higher amount of malathion formulation used, which was approximately two times greater. Malaoxon was not detected on mass deposition samples collected during the 1994 or 1981 studies but it was detected in 1990, averaging 0.13 mg/m².

Outdoor air samples were collected during the first, second, and seventh aerial spray events. At each event, malathion and malaoxon residues were measured before the aerial application, during the aerial application, and for two consecutive 24-hr periods after the aerial application. The highest average level for both compounds was measured during the first 24-hr period immediately following the spray at 0.069 µg/m³ (5.1 ppt) for malathion and 0.021 µg/m³ (1.6 ppt) for malaoxon. These air results were not adjusted to account for the fraction of malathion that oxidizes to malaoxon during sampling.

Malathion levels were measured in surface water immediately following three monitored sprays and were also measured in surface water during periods of rain runoff. The concentration of malathion exiting the treatment zone immediately following each of three monitored sprays ranged from 1.0 to 24.4 ppb. During periods of rainfall, malathion levels detected exiting the treatment zone ranged from 0.25 to 203 ppb. Malaoxon was not detected in surface water immediately following an aerial application, but it was detected frequently in rain runoff water. The highest malaoxon level, 34.4 ppb, was reported in runoff water 12 days after the fifth aerial application.

The average malathion residue level in soil collected 12 hr after aerial application was 1.30 ppm. Malaoxon residue was not detected in any soil sample. The major portion of malathion residue on turf and green leafy lettuce was dislodgeable residue. Dislodgeable malathion was 2.392 and 0.98 ppm for turf and lettuce, respectively, from samples collected 12 hr after aerial application. Malaoxon was only detected as dislodgeable residue on turf at 0.0181 ppm (mean) and not on lettuce. Internal malaoxon residue levels were not detected in either turf or lettuce media.

Dissipation half-lives for malathion on sand and steel sheet media were estimated. Data were collected on these surfaces over a 32-day period following one of the applications of baited malathion. Malathion half-life was six days on sand and four days on steel media. The maximum mean malaoxon concentration was 0.59 mg/m² on sand medium detected at day 24 and 0.05 mg/m² on steel medium, detected at day 2.

DEF was measured on steel sheet medium collected approximately 30 min after aerial spray with values ranging from 0.015 to 0.44 mg/m².



ACKNOWLEDGMENTS 

We would like to express our gratitude to the homeowners and other property owners in Riverside County who allowed us to sample on their property. We also thank Roger Sava, Carissa Ganapathy, Blanca Rodriguez, Kevin Bennett, Craig Nordmark, Heinz Biermann, and Joe Marade - all of the Department of Pesticide Regulation's (DPR) Environmental Monitoring branch who helped in the preparation and sample collection necessary to complete this study. We also gratefully acknowledge Pat McCool, Rosemary Neal, Min Poe, and Ted Younglove for their assistance in field sampling and graphics presented in this report. Many thanks are given to DPR staff, Nancy Miller and Cindy Garretson, who acted as laboratory liaisons and chemical quality assurance-quality control officers. We also thankfully acknowledge the California Department of Food and Agriculture's chemists - Cathy Cooper, Jorge Hernandez, Jean Hsu, Jim Echelberry, Jane White, and Karen Hefner for their dedication in processing the hundreds of samples.

DISCLAIMER

The mention of commercial products, their source or use in connections with material reported herein is not to be construed as an actual or implied endorsement of such product.


TABLE of CONTENTS

	Page
ABSTRACT	i
ACKNOWLEDGMENTS	iii
DISCLAIMER	iii
TABLE of CONTENTS	iv
LIST of FIGURES	v
LIST of TABLES	vi
INTRODUCTION	1
Background	1
Aerial Treatment Program	2
Environmental Monitoring Program	5
MATERIALS and METHODS	6
Media and Monitoring Sites - General Information	6
Mass Deposition on Kimbie Material	6
Ambient Air Concentrations	7
Surface Water	8
Soil	10
Turf	10
Green Leafy Lettuce	11
Sand	11
Stainless Steel Plates	12
DEF Analysis	13
RESULTS and DISCUSSION	13
Quality Control	13
Mass Deposition on Kimbie Material	14
Ambient Air Concentrations	18
Surface Water	21
Soil	25
Turf	26
Green Leafy Lettuce	27
Sand	29
Stainless Steel Plates	31
DEF Analysis	33
CONCLUSIONS	34
REFERENCES	35
APPENDIX A - Laboratory Quality Control Data
APPENDIX B - Statistical Analysis
APPENDIX C - Field Results
APPENDIX D - Chemical Analytical Methods

LIST of FIGURES

	Page
Figure 1. Map of the 1994 Mediterranean fruit fly aerial treatment site in Riverside County, California	4
Figure 2. Observed and expected mass deposition rates of malathion for the 1994, 1990, and 1981 Medfly eradication programs	15
Figure 3. Distribution of the malathion mass deposition rates in the 1994, 1990, and 1981 Medfly eradication programs, using data expressed as a fraction (ratio) of the expected rate	16
Figure 4. Ambient air concentration results for the 1994, 1990, and 1981 Medfly eradication programs during each sampled period	20
Figure 5. Malathion and malaoxon concentrations measured hourly in Temescal Creek (outflow site) during three rain runoff events	22
Figure 6. Dissipation of malathion and malaoxon residues on sand and stainless steel plate media	29

LIST of TABLES

	Page
Table 1. Percentage of impurities present in Clean Crop Malathion ULV®	3
Table 2. Mean, standard deviation, and 95% confidence interval for mass deposition of malathion on Kimbie material	14
Table 3. Mass deposition of malathion on Kimbie material at three sensitive sites within or outside the aerial treatment zone	18
Table 4. Summary of outdoor ambient air concentrations of malathion and malaoxon for aerial applications 1, 2, and 7	19
Table 5. Malathion and malaoxon residues measured in rain water runoff samples collected from Temescal Creek (outflow site)	24
Table 6. Malathion levels measured on 15 samples of soil. Results expressed on a ppm wet weight basis	25
Table 7. Mass deposition of malathion and malaoxon on 15 samples of turf. Results expressed on a land surface area basis	27
Table 8. Mass deposition of malathion and malaoxon on 15 samples of turf. Results expressed on a ppm wet weight basis	27
Table 9. Mass deposition of malathion on 15 samples of green leafy lettuce (wet weight)	28
Table 10. Means and confidence intervals for malathion and malaoxon residues on sand media expressed on a mg/m2 basis	31
Table 11. Means and confidence intervals for malathion and malaoxon residues on stainless steel sheet media	32
Table 12. DEF deposition on 10 samples of stainless steel sheet media	33

INTRODUCTION 

Background 

The Mediterranean fruit fly, Medfly (Ceratitis capitata), is a destructive pest with economic importance since it infests more than 260 fruits, flowers, vegetables, and nuts (Foote, 1993). To prevent this pest from becoming established in the state of California, the California Department of Food and Agriculture (CDFA) and United States Department of Agriculture (USDA) work cooperatively to eradicate any Medfly infestations detected in the State.

In 1993, members of the International Science Advisory Panel evaluated the CDFA/USDA Medfly eradication program and recommended that sterile Medflies be released over a 3603 km² area in the Los Angeles basin. The basin area which encompasses parts of Los Angeles, Orange, and San Bernardino Counties is a region in southern California where numerous Medfly detections have been reported over several years as well as in 1993. The Panel's objective was to treat this basin-wide area with sterile Medflies as an effective means to eradicate all existing Medfly infestations and prevent newly introduced infestations from developing in that region. This biological technique is based on the principle of breeding a wild Medfly with a sterile Medfly to produce infertile eggs that will not hatch. In accordance with the Panel recommendations, the weekly release of sterile flies began in March 1994 and will continue until February 1996.

In December 1993, a Medfly infestation was reported in Riverside County. Eradication of this infestation using sterile flies was not an alternative because additional sterile flies were unavailable for release. CDFA's sterile Medfly rearing facilities were at maximum production supplying 575 million sterile flies per week for the Los Angeles basin area. Consequently, treatment in Riverside County was to consist of ten aerial applications of baited-malathion made at two week intervals. The eradication program was later modified to consist of eight aerial applications made from February through May, 1994 to a 38 km² region encompassing the cities of Corona and Norco.

The Environmental Hazards Assessment Program (EHAP) of the Department of Pesticide Regulation monitored five of the eight scheduled aerial applications of baited-malathion in Riverside County. Objectives of this study were 1) to characterize the concentrations of malathion and malaoxon on land, in ambient air, and on surface water, and to compare these values with those reported in the 1990 and 1981 EHAP aerial Medfly monitoring studies; 2) to report levels of malathion and malaoxon on surfaces of soil, turf, and green leafy lettuce media that had not previously been examined during aerial application; 3) to determine the persistence of malathion and malaoxon over a 32-day period on playground sand and stainless steel plate media used to simulate outdoor surfaces that children may contact after an aerial application; and 4) to report the levels of diethyl fumarate (DEF), a co-product of the manufacturing process of malathion, on stainless steel plate surfaces collected immediately following aerial application.

Aerial Treatment Program 

In the 1994 Medfly eradication program, CDFA combined malathion in the form of Clean Crop Malathion ULV^® (Platte Chemical Company) with Nu-Lure^® insect bait, a protein-sugar based attractant. The mixture was composed of 88 ml malathion product and 789 ml insect bait applied per hectare (ha). Malathion was applied at a rate of 102 grams (g) active ingredient (ai) per ha, which is equivalent to a target deposition rate of 10.2 milligrams (mg) of malathion per square meter (m²).

The malathion applied in this eradication project is a technical formulation that has a specific gravity of 1.23 at 25 °C and is 95% ai by weight (1.16 kg ai/L). The inert ingredients in the malathion product was 5% of the formulation by weight. These compounds are either impurities or co-products from the malathion manufacturing process (Table 1) (Voss, 1990).

For each application, CDFA used three Bell 204 helicopters flying at a speed of 130 kilometers/hr and at an average above-ground height of 91 m. Helicopters were equipped with six stainless steel Tee Jet 8010 flat fan nozzles on the boom to apply the malathion-bait mixture in a 61 m total swath width. Treatments were made over a 38 km² area in Riverside County encompassing portions of the cities of Corona and Norco (Figure 1). Applications were made at night beginning at 2100 hr (9:00 pm) and generally ending by 0200 hr (2:00 am).

Table 1. Percentage of impurities present in Clean Crop Malathion ULV® (Platte Chemical Company)
Co-products	Percent by Weight
Diethyl fumarate (DEF)	0.90%
Diethylhydroxysuccinate	0.05%
O,O-dimethyl phosphorothioite	0.05%
O,O,O-trimethyl phosphorothioite (TMTP)	0.45%
O,O,S-trimethyl phosphorodithioate (TME)	1.20%
Ethyl nitrite	0.03%
Diethyl-bis (ethoxycarbonyl) mercaptosuccinate	0.15%
S-1,2,-ethyl-O,S-dimethyl phosphorodithioate (isomalathion)	0.20%
S-(1-methoxycarbonyl-2-ethoxycarbonyl) ethyl-O,O-dimethyl phosphorodithioate	0.60%
Bis-(O,O-dimethyl thionophosphoryl) sulfide (PSP)	0.30%
Diethyl methylthiosuccinate (DEMMS)	1.00%
S-ethyl-O,O-dimethyl phosphorodithioate	0.10%
S-1,2-bis (ethoxycarbonyl) ethyl-O,O-dimethyl phosphorothioate (malaoxon)	0.10%
Diethyl ethylthiosuccinate	0.10%
Water	0.07%
Acidity as sulfuric acid	0.05%

 

Figure 1. Map of the 1994 Mediterranean fruit fly aerial treatment site in Riverside County, California


Less malathion was used in the malathion-bait mixture in the current program than during the Medfly eradication programs monitored by EHAP in 1990 (Segawa et al., 1991) and 1981 (Oshima et al., 1982). The CDFA reduced levels of malathion because various studies showed that acceptable mortality levels for Medflies were achieved using lower malathion rates. In addition to changes in the malathion to insect bait ratio, the malathion product and application equipment differed between study years.

In this eradication effort, CDFA staff flagged three sensitive areas in the treatment zone: a 2.8 ha man-made, private lake; and two natural habitat regions located in the northeast and northwest corners of the treatment zone. These areas were flagged so that the spray crew would not directly treat these areas with malathion (Figure 1). The lake contained fish such as bass, white crappie, channel catfish, and brown bullhead while the two natural habitat areas sheltered several threatened or endangered species on the Department of Fish and Game's and the United States Fish and Wildlife Service's list. This list included organisms such as the Willow Fly Catcher, Western Yellow Billed Cuckoo, Bald Eagle, Least Bell's Vireo, Stephen's Kangaroo Rat, and the California Gnat Catcher.

Environmental Monitoring Program 

EHAP staff monitored media at five of the eight aerial applications in Riverside County. The monitoring plan involved measuring the mass of malathion and malaoxon deposited on the ground as well as in ambient air and natural water bodies. Values obtained from these three media were compared to values obtained from the previous Medfly monitoring studies conducted in 1990 (Segawa et al., 1991) and 1981 (Oshima et al., 1982). In addition, the conversion of malathion to malaoxon induced by the air sampling process (Segawa et al., 1991) was measured in this study.

At the request of the Office of Environmental Health Hazard Assessment (OEHHA) staff, deposition on additional media was measured after aerial application. Deposition of malathion and malaoxon was determined for soil, turf, and green leafy lettuce and the dissipation rate of malathion was estimated on sand and steel plate surfaces over a 32-day period following a single aerial application. These two media simulated play ground material commonly found in school yards or parks. Lastly, DEF, a possible skin irritant, was also measured on stainless steel plate surfaces immediately following a single aerial application.

MATERIALS and METHODS 

Media and Monitoring Sites - General Information 

Several media were used to monitor deposition of malathion, malaoxon, or DEF following an aerial application. These media included paper towel material, soil, turf, green leafy lettuce, sand, and stainless steel plates. Deposition monitoring sites were selected prior to the first aerial application and were located in open areas away from obstructions to reduce the interference with malathion droplet deposition. Media placed at sites within the treatment zone were used to characterize deposition within the spray area. A few media sites were also located outside the treatment area to monitor drift to sensitive natural habitat areas. In most instances, the same sites were used for monitoring during subsequent applications.

Malathion and malaoxon were also measured in air and water. Monitoring sites for these media were selected prior to the first aerial spray. Air monitoring sites were located in sheltered locations (e.g., beneath a carport or under a patio awning) so that droplet-sized material would not be captured by sampling equipment during aerial application. Surface water sites were selected based on the drainage patterns and presence of water bodies within the malathion treatment area.
Figure 1 shows sample sites and their specific identification numbers. Site identification numbers may be used to cross reference results for each medium (Appendix C).

Mass Deposition on Kimbie Material 

Mass deposition of malathion and malaoxon reaching the ground surface were measured with 30 Kimbie sheets placed within the treatment zone prior to the first, second, and third sprays. Kimbie sheets are paper towel material lined with a plastic backing, each measuring 22.9 cm x 40.6 cm. One Kimbie sheet was placed at each of the three sensitive sites in order to document and quantify the amount of malathion deposited at these flagged regions during each of the three applications monitored.

Kimbie sheets were raised approximately 20 cm above the ground by attaching a polyethylene- covered platform to a cement block and the Kimbie sheet secured to the top of the platform surface with four push pins. Sheets were collected 30 min after the helicopter had passed over the site. The Kimbie sheet was twice folded upon itself with the plastic side out, wrapped in aluminum foil, and then placed in an envelope. Samples were stored on dry ice in the field and remained frozen until chemical extraction.

Chemical analyses were conducted by the CDFA Chemistry Laboratory Services in Sacramento, California (Appendix D). The detection limit (DL) was 0.011 mg/m² for malathion and malaoxon.

Summary statistics include presentation of the mean, standard deviation, and the 95% confidence interval for the mean. Data for the three study years (1994, 1990, and 1981) are presented in a frequency histogram. These data were adjusted by dividing all deposition values by the theoretical deposition rate for each respective year to compensate for the different application rates used by CDFA. The median and 95% confidence interval for the median are presented in Appendix B for 1994 data, along with statistical comparison of the malathion deposited for the three study years.

Ambient Air Concentrations 

To determine concentrations of malathion and malaoxon in outdoor ambient air, air samplers were placed, for applications one, two, and seven, at five sites clustered within a 0.8 km radius region. Samplers were located near the center of the aerial treatment zone in order to increase the precision of estimates for concentrations.

At each site, one Anderson model SE-144 sampling pump was mounted with 15 ml XAD resin as a trapping medium and calibrated at 15 L/min for application one and at 25 L/min for applications two and seven. Air samples were collected for 76 consecutive hours with sampling partitioned into four periods: one 24-hr background sample collected immediately before the spray, one sample collected for the duration of the malathion application (application time plus one-half hour), one 24-hr post application sample, and another consecutive 24-hr sample. The air flow through each sampling tube was measured with a rotometer before and after each sampling interval to determine if the air flow had changed from the original setting.

A co-located air sampler was used at each site to determine the conversion of malathion to malaoxon that occurs during sampling. The co-located samplers operated at the same settings as the primary samplers and were mounted with malathion spiked resin tubes (20, 40, 60, 80, and 100 µg malathion) for each sample period. The average percentage of malathion breakdown to malaoxon in the spiked tubes was used to determine the overall conversion for each sample period.

Sample tubes were sealed with Teflon^®-lined rubber stoppers, placed in plastic bags, stored on dry ice in the field, and remained frozen until extraction. Chemical analysis (Appendix D) was conducted by CDFA Laboratory with a DL ranging from 0.003 to 0.03 µg/m³ depending upon the sampling time and sampling flow rate.

Summary statistics for each sample period include presentation of the mean and standard deviation for malathion and malaoxon concentrations. Graphics show total malathion levels for the three Medfly programs. Appendix C contains the adjusted air results for 1994 and also includes the unadjusted mean and median air concentrations for the three study years.

Surface Water 

For the first, second, and third aerial applications, malathion and malaoxon concentrations were measured in surface water sampled from 1) three sites (site numbers 37, 38, and 39) prior to entering the treatment area; 2) one outflow site (site number 35) outside of the treatment area; and 3) a sensitive lake site (site number 36) located within the treatment area (Figure 1). At all locations, background samples were collected four to six hr before each aerial application. Water at the inflow sites were sampled 0.8 km upstream from the perimeter of the malathion treatment zone and water from the outflow site was sampled 0.8 km downstream from the treatment boundary (Figure 1). At the lake, water was sampled off a dock. During the aerial application, water was sampled at the outflow site approximately five hr after application began. Water was also sampled at the lake site 30 min after the helicopter passed over this site.

During periods of rain, EHAP staff assembled as quickly as possible to arrive at the treatment zone to collect runoff water. For each rain period, water was sampled only once from each of the three inflow sites. After aerial applications one and two, staff collected water only once at the outflow site, but following the third, fourth, fifth, sixth, and seventh aerial applications, water was to be sampled for six consecutive hourly intervals during periods of rain. Samples were not collected at the lake site during periods of rain.

Samples were collected by submersing a clean stainless steel bucket into the water. A portion of the collected water was used to rinse four 1-liter amber bottles. The pH of each sample was recorded and adjusted within 3 to 3.5 to reduce malathion hydrolysis using a 3 N hydrochloric acid solution. Then, the bottles were filled and sealed with Teflon^®- lined caps. One field blank of de-ionized water was prepared during each sampling period to determine if any contamination had occurred during field sampling. All samples were stored on wet ice and maintained at 4 C until extracted.

Chemical analyses of water samples were conducted by CDFA Laboratory with a DL of 0.05 ppb for malathion and malaoxon (Appendix D). Split water samples were analyzed by Agriculture Priority and Pollutants Laboratories (APPL) of Fresno, California, with a DL of 0.1 ppb for malathion and malaoxon (Appendix D). Statistical analysis included inter-laboratory comparison between CDFA and APPL Laboratory results (Appendix A).

Soil 

Surface soil was used to determine the deposition of malathion and malaoxon at 15 sites during the second aerial application. Prior to the first aerial application, background soil samples were collected from eight areas in Corona. Soil was composited, air-dried, and pulverized with a mortar and pestle. To facilitate handling, the soil was then sifted through a 2 mm sieve. All material passed through the sieve. Particle size (Bouyoucos, 1962) and pH (Hausenbuiller, 1972) were determined on sifted soil which was not treated with malathion.

At each site, two polycarbonate containers (internal diameter 11 cm each) were placed on the ground, each holding 50 g of soil at approximately 0.5 cm depth. Samples were collected about 12 hr after the aerial application. Soil collected at each site was composited into a single sample and analyzed by CDFA Laboratory (Appendix D). At two of the 15 sites, two additional containers of soil were placed on the ground for inter-laboratory comparison purposes. The four containers of soil at these sites were combined into one container, then later equally divided into two separate samples: one sample analyzed by CDFA and one sample analyzed by APPL Laboratories.

All samples were stored on dry ice in the field and remained frozen until chemical extraction. CDFA's DL for malathion and malaoxon was 0.01 ppm. Soil samples analyzed by APPL Laboratory had a DL of 2 ppb. Results for malathion and malaoxon are reported on a ppm, wet weight basis, and mg/m². Inter-laboratory comparison of soil results are located in Appendix A.

Turf 

To quantify the levels of malathion and malaoxon deposited on turfgrass, one shallow aluminum baking tray containing 27.9 cm x 35.6 cm of pre-cut strips of living, dwarf fescue sod grass was placed at each of 15 sites during the second application. Approximately 12 hr after aerial application, four strips, each measuring 5.1 cm x 30.5 cm, were removed for sampling. The grass blades and thatch were cut with scissors and clipped into a 1 quart glass mason jar which was sealed with an aluminum lined lid. Samples were stored on wet ice in the field and maintained at 4 C until extraction.

Chemical analyses of turf samples for dislodgeable and internal residues were conducted by CDFA Laboratory (Appendix D). The DLs for dislodgeable and internal residues of malathion and malaoxon were 0.0040 mg/m² and 0.0047 ppm wet weight basis.

Green Leafy Lettuce 

On the second application, one head of green leafy lettuce was placed at each of 15 sites to measure the concentration of malathion and malaoxon residues deposited on the surface. Organically grown lettuce heads, stripped of excess wrapper leaves, were vertically supported on the surface of polyethylene-covered polystyrene containers. The morning following aerial application (about 12 hr after the spray), each lettuce head was collected and placed into a 2 L glass jar which was sealed with a Teflon^®-lined lid. Samples were stored on wet ice in the field and maintained at 4 C until chemical extraction.

Chemical analyses for dislodgeable and internal residues of malathion and malaoxon were determined by CDFA Laboratory (Appendix D). The DL for dislodgeable and internal residues was 0.001 ppm for both malathion and malaoxon. Dislodgeable and total residue results are presented on a ppm, wet weight basis for both compounds.

Sand 

Malathion and malaoxon residues were measured in sand over a 32-day period following the third aerial application. At five selected sites, a set of ten aluminum trays was placed on a large polyethylene-covered plywood surface. Each tray measured 19.1 cm x 29.2 cm and contained 500 g of playground sand. Samples were collected on days 0 (30 min after aerial application), 1, 2, 4, 6, 8, 13, 16, 24, and 32 following the application. Prior to subsequent aerial applications or inclement weather, media at each site were covered with a large portable shelter. The cover was removed after any of these events occurred. On each sample day, sand was collected from one aluminum tray from each site and transferred into a 1 quart glass mason jar. For inter-laboratory comparison purposes, sand samples were equally divided into two separate containers for analysis by CDFA and APPL Laboratories. This occurred at four sites on sampling days 4, 6, 8, and 16. All samples were stored on dry ice in the field and remained frozen until extraction.

Chemical analysis for malathion and malaoxon was conducted by CDFA Laboratory with a DL of 0.0021 ppm (0.019 mg/m²) and a DL of 2 ppb for APPL Laboratory (Appendix D). Results are presented on a ppm, wet weight basis, and as mg/m². Statistical presentation of results include the mean and 95% confidence interval for mean malathion and malaoxon concentrations. Description of the statistical methods used to determine half-lives are in Appendix B.

Stainless Steel Plates 

To study the dissipation rate of malathion and malaoxon on steel sheet surfaces, a set of 90 individual stainless steel plates was placed at five sites during the third aerial application. Plates each measured 15.2 cm x 6.8 cm and were placed on a large polyethylene-covered plywood surface. On each sampling day, nine steel plates were randomly selected off the platform, placed into a 1 quart mason jar, and sealed with a Teflon^®-lined lid.

Samples were collected on days 0 (30 min after aerial application), 1, 2, 4, 6, 8, 13, 16, 24, and 32 days after the aerial application. Prior to subsequent aerial applications or inclement weather, the remaining media at each site were covered with a large portable shelter. The cover was removed after any of these events occurred. Samples were stored on dry ice in the field and remained frozen until extraction.

Chemical analysis was performed by CDFA Laboratory for residues of malathion and malaoxon with a DL of 0.005 mg/m² for both compounds (Appendix D). Results are presented on a mg/m² basis. Statistical presentation of results include the mean and 95% confidence intervals for mean malathion and malaoxon concentrations. Description of the statistical methods used to determine half-lives are in Appendix B.

DEF Analysis 

DEF concentrations on stainless steel plate surfaces were measured on the second aerial application. At each of ten sites, nine stainless steel plates, each measuring 15.2 cm x 6.8 cm, were placed on the surface of a polyethylene-covered cardboard platform. Approximately 30 min after aerial application, samples were placed into a 1 quart mason jar which was sealed with a Teflon^®-lined lid. Samples were stored on dry ice in the field and remained frozen until chemical extraction. The DEF samples were analyzed by the CDFA Laboratory with a DL of 0.011 mg/m² (Appendix D). Results are presented on a mg/m² basis.



RESULTS and DISCUSSION 

The data presented and discussed in this section, unless otherwise mentioned, refer only to those results obtained by the CDFA Laboratory. When malathion or malaoxon residues were below detection limits, one-half the value of the detection limit was used to calculate residue concentrations for summary statistics in all media excluding surface water. Deposition data are presented in mg/m² units and may be converted to µg/ft² by dividing the mg/m² units by 0.01076.

Quality Control 

Prior to aerial monitoring, method validation studies for percent recovery of malathion and malaoxon, were completed by the CDFA and APPL Laboratories (Appendix A) on various media. DEF method validation results are also presented in Appendix A for CDFA Laboratory.

During chemical analyses of field samples, internal quality control information was obtained for malathion, malaoxon, and DEF for the various media (Appendix A). Overall CDFA recoveries for the blind spike samples submitted for analysis with the field samples ranged from 81 to 110% for the three compounds. Water, soil, and sand samples were also split and analyzed by two laboratories for an inter-laboratory comparison (Appendix A).

Mass Deposition on Kimbie Material 

Malathion was measured as mass deposition on Kimbie material for applications one, two, and three. Data collected from sites 1, 7, 8, and 20 were not included because the sites were either too close to the treatment boundary or were adjacent to the sensitive area in the northeastern corner of the treatment area in the "no spray buffer zone". Values at these sites were much less (mean of 2.8 mg/m²) than those measured at other sites within the spray zone. For the remaining sites in the spray boundary, the average malathion deposition rates for the first, second, and third aerial applications were 8.51, 9.46, and 8.74 mg/m², respectively, (Table 2) with malathion levels ranging from not detected to 23.19 mg/m². The means from each of the three sprays were not significantly different (Appendix B).

Table 2. Mean, standard deviation, and 95% confidence interval for mass deposition of malathion on Kimbie materiala
	Malathion (mg/m2)
			95% Confidence Interval
Application Date	Mean	Stdevb	Lower	Upper
Spray 1 (30 samples)	8.51	4.10	7.05	9.98
Spray 2 (30 samples)	9.46	5.43	7.51	11.4
Spray 3 (29 samples)	8.74	4.90	6.96	10.5
Combined Sprays (89 samples)	8.91	4.80	7.91	9.90
a Malaoxon was not detected. DL of 0.011 mg/m2.  b Standard deviation.

The overall average deposition rate from the three applications was 8.91 mg/m² which was 87% of the expected malathion deposition rate of 10.2 mg/m² (Figure 2). The 95% confidence interval for the mean malathion concentration of the three combined sprays ranged from 7.91 to 9.90 mg/m²; there is only a 5% chance that the true deposition rate falls outside this interval.

In a comparison of data from the three Medfly eradication programs, the average 1994 malathion deposition rate was well below the previous means of 21.8 and 14.9 mg/m², respective average deposition rates for the eradication programs in 1990 and 1981 (Figure 2). Deposition in the current study was anticipated to be lower because CDFA staff used less malathion in the malathion-bait mixture. 

Figure 2. Observed and expected mass deposition rates of malathion for the 1994, 1990, and 1981 Medfly eradication programs


Malathion concentration was observed to be more evenly distributed on the ground in 1994 and 1990 than in 1981 (Figure 3). The 1994 and 1990 observed distributions of malathion appear similar in location and shape with the highest observed proportion of samples in the target interval for both years. Approximately 50% of the samples were below the target rate interval for these two years. In the frequency histogram, all of the measured values with an observed deposition rate equal to or nearly equal to the expected rate are centered on the interval of 1 on the x-axis.

Figure 3. Distribution of the malathion mass deposition rates in the 1994, 1990, and 1981 Medfly eradication programs, using data expressed as a fraction (ratio) of the expected rate


In contrast, the 1981 distribution of malathion on the ground was dissimilar from either 1994 or 1990 data. Skewness was more pronounced, with nearly 60% of the samples lying below the target rate interval. There were some samples in 1981 which had unusually high deposition rates. These values were more than two and one-half times greater than the expected deposition rate. Equivalent deposition values exceeding the ratio of 2.50 were not seen in 1994 or 1990 data (Figure 3).

Although malaoxon is present in small concentrations in the malathion-bait spray mixture, it was not detected in any of the Kimbie samples. The expected malaoxon deposition rate of 0.011 mg/m² was near the laboratory's quantification limit which may explain why malaoxon was not found. Also, tank mix samples (diluted concentrate) taken the day of the spray showed that malaoxon was not detected in the samples collected by the Pesticide Enforcement staff for sprays one and three. The DL for malaoxon was 200 ppm (Andrews, 1994).

In 1990, the average malaoxon concentration was 0.13 mg/m²; 503% greater than the expected malaoxon deposition rate. Approximately two times more malathion was used in the malathion-bait mixture in that year than was currently used. Segawa et al. (1991) attributed this increase of the metabolite to possible oxidation of malathion to malaoxon during storage, transport, or mixing of the malathion product. No comparisons were made for malaoxon between study years because of the lack of detections in the current study and in 1981.

Mass deposition samples collected during three applications at the three sensitive sites averaged 2.16, 0.032, and 0.761 mg/m² malathion for the lake, northeast habitat region, and northwest habitat region, respectively. Although not sprayed, malathion was still found in each sensitive site. These levels were lower than the average deposition rate of 8.91 mg/m² for the sprayed treatment zone (Table 3) and the presence of malathion residue at the sensitive sites was probably due to drift. Ware et al. (1984) cited that drift can occur from all types of pesticide applications. Malaoxon was not detected in any of the samples from the sensitive sites.

Table 3. Mass deposition of malathion on Kimbie material at three sensitive sites within or outside the aerial treatment zonea
	Malathion (mg/m2)
Sample Date	Lake Site	Northeast  Habitat Regionb	Northwest  Habitat Regionb
Spray 1	4.09	0.062	1.97
Spray 2	0.469	0.019	0.170
Spray 3	1.92	0.013	0.143
a Malaoxon was not detected. DL of 0.011 mg/m2.   b Site located outside the treatment area.

Ambient Air Concentrations 

For the second and seventh aerial applications, the conversion rate of malathion to malaoxon induced by the sampling method ranged from 0.21 to 6.8%. The conversion rate was lowest during the spray period (4 to 5 hr sample time) and higher during either the first or second 24-hr sampling period (Appendix C). Conversion measurements were not obtained for the first aerial application because several of the air samplers failed to operate during the specific time period. Discussion of the following malathion and malaoxon air results refer to data which were not adjusted to account for the conversion.

Malathion and malaoxon concentrations were measured in outdoor ambient air collected at five sites during the first, second, and seventh aerial applications. Air samples were collected at each site for four consecutive periods: 24-hr immediately before the spray, during the spray period plus one-half hour, and for two consecutive 24-hr periods after the spray. The highest average malathion and malaoxon concentrations measured in air were 0.069 µg/m³ and 0.021 µg/m³ respectively, detected during the first 24-hr sampling period following aerial application (Table 4). Overall, malathion concentrations were greater than corresponding malaoxon concentrations for each sample.

Table 4. Summary of outdoor ambient air concentrations of malathion and malaoxon for aerial applications 1, 2, and 7a
		µg/m3
Sample Period		Malathion		Malaoxon
Background
	Number of Samples	14		14
	Mean	0.005		0.003
	Minimum	NDb		ND
	Maximum	0.017		0.006
	Stdev	0.005		0.002
Spray
	Number of Samples	14		14
	Mean	0.027		0.010
	Minimum	ND		ND
	Maximum	0.062		0.018
	Stdev	0.012		0.004
1st 24-hr Post Spray
	Number of Samples	15		15
	Mean	0.069		0.021
	Minimum	0.022		0.005
	Maximum	0.336		0.046
	Stdev	0.077		0.013
2nd 24-hr Post Spray
	Number of Samples	15		15
	Mean	0.056		0.018
	Minimum	0.018		0.006
	Maximum	0.337		0.075
	Stdev	0.079		0.017
a Results not adjusted for the conversion of malathion to malaoxon due to the sampling method.  b Not detected. DL varied from 0.003 to 0.03 µg/m3 depending upon sampling time and sampling flow rate.

Figure 4. Ambient air concentration results for the 1994, 1990, and 1981 Medfly eradication programs during each sampled period

Figure 4 shows a comparison between the average total malathion levels for 1994, 1990, and 1981. The results presented in Figure 4 are the average total malathion levels which are the mean values from the sum of malathion and malaoxon. It is presently considered more reliable for use than the unsummed values (Segawa et al., 1991). Statistical comparisons of malathion or malaoxon concentrations were not made between study years due to differences in the amounts of malathion applied, the number of applications made, the number of study locations monitored, and the number of samples collected.

Surface Water 

Surface water was sampled immediately before the first, second, and third aerial applications from three inflow sites. Malathion residues were not detected in water sampled from two inflow sites, the Temescal Wash and Arlington Channel. Main Street Canal, the third inflow site, delivered water only during periods of rain, and therefore was not sampled before the aerial sprays. Malaoxon was not detected in surface water prior to the three monitored applications.

Immediately following aerial applications one, two, and three, malathion concentrations in water sampled from the outflow site, Temescal Creek, contained malathion levels of 24.4, 2.02, and 1.04 ppb, respectively. Malaoxon was not detected. Background water samples collected at this site before each application had no detectable residues of malathion or malaoxon. This creek plays an integral part in the local habitat by supporting the California Gnat Catcher. It is also a tributary to the Santa Ana River.

During periods of precipitation, rain runoff water was sampled from the inflow and outflow sites following aerial applications one, two, three, and five. Water sampled at the inflow sites had malathion levels ranging from not detected to 0.72 ppb, indicating possible outside sources of malathion entering the treatment zone. These sources were not investigated, but the levels were much lower than the malathion levels detected at the outflow site.

At the outflow site, rain runoff water was sampled following aerial applications one and two. The maximum malathion level of 203 ppb was detected two days after application one (Table 5). The maximum levels detected in 1990 and 1981 runoff water were 44.1 and 583 ppb, respectively. Results from the three study years could not be statistically compared due to differences in study designs, the amount of malathion applied, and the amount of rainfall received.

For aerial applications number three and five, rain runoff water was sampled hourly for a six hr period at the outflow site to determine the range of malathion levels leaving the spray area. Within this period, malathion levels decreased to less than half the concentration initially detected. Malathion concentrations also appeared to lessen when precipitation occurred more than once within the two week spray period, possibly due to wash-off from cumulative rain-fall (Figure 5). These post-spray detections indicated that rain could wash malathion off various treated surfaces, then directly transport it to drainage systems. Giles (1970) reported that high rainfall can result in high pesticide concentrations in surface water, although the concentration may be decreased by dilution from increased stream volume and discharge.

Malaoxon levels were also measured in rain runoff water from the inflow sites and ranged from not detected to 0.32 ppb. Possible sources of malaoxon outside the treatment zone were not investigated. The highest malaoxon level detected at the outflow site was 34.4 ppb, 12 days after the fifth aerial spray.

Figure 5. Malathion and malaoxon concentrations measured hourly in Temescal Creek (outflow site) during three rain runoff events


Malaoxon concentrations measured in runoff water during this six hr period were higher than the corresponding malathion concentrations (Figure 5) and may have been due to the oxidation of malathion to malaoxon on treated surfaces. The highest malaoxon levels detected in 1990 and 1981 were 41.0 and 328 ppb, respectively. Malaoxon results, like malathion, could not be statistically compared for the three study years.

The malathion concentrations at the lake site were 3.62, 0.27, and 1.40 ppb for the first, second, and third aerial applications, respectively. These concentrations were detected in instantaneous grab samples taken from the lake.

Table 5. Malathion and malaoxon residues measured in rain water runoff samples collected from Temescal Creek (outflow site)
				ppb
Sample Date	Description	Sample Timea	Water pHb	Malathion	Malaoxon
2/16/94	Spray 1	0215	8.1	24.4	NDc
2/17/94	Rain - 2 Day Post	1000	7.6	203	1.61
3/02/94	Spray 2	0145	8.5	2.02	ND
3/06/94	Rain - 5 Day Post	0930	7.3	19.3	4.06
3/16/94	Spray 3	0320	8.1	1.04	ND
3/19/94	Rain - 4 Day Post	0654	7.1	13.9	1.36
		0810	6.3	17.9	1.97
		0915	7.2	15.3	1.40
		1010	7.0	11.0	1.09
		1115	7.0	4.01	0.39
		1210	7.0	5.16	0.68
3/25/94	Rain - 10 Day Post	0609	7.9	0.58	0.17
		0710	7.6	0.54	0.16
		0805	7.6	0.47	0.16
		0915	7.9	0.60	0.13
		1006	7.9	0.42	0.11
		1105	8.0	0.25	0.10
4/12/94	Spray 5	NSd	NS	NS	NS
4/24/94	Rain - 12 Day Post	0745	7.5	14.0	34.4
		0845	6.9	9.25	23.4
		0945	7.0	7.20	16.0
		1045	7.0	5.67	10.8
		1145	6.9	5.45	9.65
		1245	7.0	3.49	6.15
a Time given in military hours (2400 hr clock). On each spray day, spraying was generally initiated about 5 hr before sample was collected.   b Water pH before adjustment with acid.  c Malaoxon not detected. DL of 0.05 ppb.  d Not sampled.

Malathion was detected in the lake before the second application at 0.11 ppb, suggesting that residues were not completely decomposed 14 days after the first spray. Echelberger and Lichtenberg (1971) reported that malathion applied to river water at 10 ppb was found to have detectable residue over a two-week period. This detection differs from other authors who reported decomposition of malathion in the field within 1 to 3 days of application (Mulla, 1963; Guerrant et al., 1970).

Soil 

Particle size analyses of the Corona soil used in the monitoring study showed that the soil contained 58% sand, 26% silt, and 16% clay. The texture was a sandy loam. The soil pH was 7.2 and the average moisture content of the soil samples was below 1% (wet weight basis).

The mean mass of malathion deposited on local soil at 0.5 cm depth was 66% of the expected application rate of 10.2 mg/m ² (Table 6). Concentrations ranged from not detected to 16.6 mg/m².

Table 6. Malathion levels measured on 15 samples of soil. Results expressed on a ppm wet weight basisa
	Malathion
	mg/m2	ppm
Mean	6.78	1.30
Minimum	NDb	ND
Maximum	16.6	3.17
Stdev	5.59	1.08
95% Confidence Interval  for the Mean	3.69 to 9.88	0.71 to 1.90
aMalaoxon not detected. DL of 0.01 ppm.  bNot detected. DL of 0.01 ppm.

Several authors consider malathion a nonpersistent insecticide in soil, when applied at recommended rates (Van Middelem, 1965; Roberts et al., 1962). Hydrolysis may be responsible for degradation of this chemical and other organophosphate compounds (Walker and Stojanovic, 1973). Walker and Stojanovic (1973) also reported that malathion is hydrolyzed to various by-products and then metabolized by soil micro-organisms. Lichtenstein and Schulz (1964) found that 3 and 7 days after malathion application of 0.9 kg/ha, only 15 and 5% of the applied malathion, respectively, was recovered under field conditions in a silt loam soil.

In this study, malaoxon was not detected in soil 12 hr after malathion application. In a controlled study, Neal et al. (1993) reported the detection of malaoxon 4 days after malathion application to soil. The study conducted by Neal et al. involved using a different application rate, droplet size, and ratio of malathion to protein bait mixture than that used in the current study. Paschal and Neville (1976) found malaoxon half-life depends on soil pH and reported that at pH 8.3, the half-life was approximately 3 days while at pH of 6.2, the half-life increased to 7 days. They believed the dissipation of malaoxon in soil was due to chemical hydrolysis.

Turf 

The average moisture content of the dwarf fescue turf samples was 80% (wet weight basis). Approximately 99% of the recovered malathion was detected as dislodgeable residue on the exterior of the grass and thatch (Table 7). The dislodgeable residue is the residue fraction that was removed off the plant surface by the laboratory wash procedure. The total residue is the sum of the dislodgeable and internal residues. Dislodgeable malathion residues ranged from not detected to 4.249 mg/m², with an average dislodgeable concentration of 1.951 mg/m².

Dislodgeable residues of malaoxon were much lower than malathion and ranged from not detected to 0.0329 mg/m² with an average dislodgeable malaoxon level of 0.0157 mg/m². Internal malaoxon residues were not detected. Limited research has been conducted on malathion degradation by plants (Mulla et al., 1981), but some studies on stored grain showed that malathion can oxidize to malaoxon (Rowlands, 1964). Mulla et al. (1981) stated that conversion of malathion to malaoxon is possible through enzymatic oxidation, but that malaoxon is easily degraded to nontoxic metabolites.

Table 7. Mass deposition of malathion and malaoxon on 15 samples of turf. Results expressed on a land surface area basis
	Malathion (mg/m2)				Malaoxon (mg/m2)
	Dislodgeable	Internal	Total		Dislodgeable	Internal	Total
Mean	1.951	0.0284	1.979		0.0157	NDa	0.0184
Minimum	ND	ND	ND		ND	ND	ND
Maximum	4.249	0.0886	4.260		0.0329	ND	0.0359
Stdev	1.447	0.0225	1.456		0.0099	--	0.0099
aNot detected. DL of 0.0040 mg/m2.

When examining data on a weight of pesticide per weight of grass basis, dislodgeable malathion averaged 2.392 ppm and ranged from not detected to 5.880 ppm (Table 8). A small percentage of the applied malathion was found internally: the mean value was 0.0362 ppm. The mean dislodgeable malaoxon level was 0.0181 ppm and no internal residues of malaoxon were detected.

Table 8. Mass deposition of malathion and malaoxon on 15 samples of turf. Results expressed on a ppm wet weight basis
	Malathion (ppm)				Malaoxon (ppm)
	Dislodgeable	Internal	Total		Dislodgeable	Internal	Total
Mean	2.392	0.0362	2.428		0.0181	NDa	0.0214
Minimum	ND	ND	ND		ND	ND	ND
Maximum	5.880	0.1464	5.937		0.0360	ND	0.0390
Stdev	1.849	0.0357	1.868		0.0100	--	0.0098
aNot detected. DL of 0.0047 µg/g.

Green Leafy Lettuce 

Surface area of the lettuce heads were not measured prior to the spray, therefore, results are presented as ppm, wet weight basis. The moisture content of the green leafy lettuce heads ranged from 91 to 94%.

Dislodgeable malathion ranged from not detected to 2.30 ppm with an average of 0.98 ppm (Table 9). The average internal malathion residue detected was 0.10 ppm, indicating that only a small percentage was found internally 12 hr after aerial application. These levels were below 8.0 ppm which is the United States Environmental Protection Agency's tolerance for malathion (Sittig, 1980) in or on pre-harvest lettuce. Approximately 90% of the total malathion was detected as dislodgeable residue. These results indicate that a large percentage of malathion can be removed from the exterior of lettuce heads recently treated with malathion. Neal et al. (1993) conducted a controlled malathion dissipation study on unwashed lettuce heads treated with a malathion-bait mixture. Predicted malathion half-life was 6 hr.

Table 9. Mass deposition of malathion on 15 samples of green leafy lettuce (wet weight)a
	Malathion (ppm)
	Dislodgeable	Internal	Total
Mean	0.98	0.101	1.09
Minimum	NDb	ND	ND
Maximum	2.30	0.432	2.54
Stdev	0.713	0.113	0.801
aMalaoxon not detected as dislodgeable or internal residue. DL of  0.001 ppm.  bNot detected. DL of 0.001 ppm.

Malaoxon residues were not detected as dislodgeable or internal fractions from samples collected 12 hr after aerial application possibly because malaoxon is less stable than malathion and easily degraded to nontoxic metabolites (Mulla et al., 1981). Although there were apparent study differences between this and Neal et al. (1993) studies, Neal et al. reported that malaoxon was not detected during a 10-day sampling period on living lettuce heads which were treated with malathion-bait mixture.



Sand 

Dissipation of malathion and malaoxon residues were measured on sand over a 32-day period following one of the aerial applications of baited-malathion. Malathion concentrations declined over the month-long monitoring period (Figure 6; Table 10). The maximum mean malathion level detected during this time was 6.6 mg/m², detected on day 0, approximately 30 min after application.

Figure 6. Dissipation of malathion and malaoxon residues on sand and stainless steel plate media



Individual half-lives for the 5 sites ranged from 4.2 to 6.9 days. An estimated half-life of 6.0 days was calculated combining all data from individual sites (Appendix B).

Neal et al. (1993) reported that malathion dissipated rapidly on sand medium under controlled exposure conditions and the predicted malathion half-life was less than one hour. The study conducted by Neal et al. (1993) involved using a different application rate, droplet size, and ratio of malathion to protein bait mixture than that of the current study. Chowdhury et al. (1984) reported that both soil moisture and pH had a large influence on the persistence of malathion in soil media. They found that the pesticide residue level decreased as the soil moisture increased and soil pH increased. The sand used in the current study was found to contain less than 1% moisture on a wet weight basis for each sampling period.

Brown et al. (1993) reported that, in general, the absolute malaoxon levels following an aerial malathion application increased on three different types of surfaces over a nine-day sampling period. The authors suggested that the malaoxon rate of formation and degradation may depend upon surface type. Malaoxon half-life could not be determined in this study because simultaneous processes were occurring where 1) malaoxon concentrations were increasing due to degradation of malathion; and 2) malaoxon levels were also decreasing due to dissipation. Using data from all sites, an estimate of the maximum measured amount of malaoxon was calculated using a quadratic function to describe the mean malaoxon levels over time (Appendix B). The highest predicted mean estimate was 0.61 mg/m² occurring at 19.5 days post application.

Neal et al. (1993) reported the detection of malaoxon residue in sand approximately 10 days after a single pesticide application. In the current study, malaoxon was detected in sand 12 hr after aerial application. Again, there were several variables which differed between the two studies.

Table 10. Means and confidence intervals for malathion and malaoxon residue on sand media expressed on a mg/m2 basis
	Malathion (mg/m2)				Malaoxon (mg/m2)
		95% Confidence Interval				95% Confidence Interval
Days Post  Application	Mean	Lower	Upper		Mean	Lower	Upper
0a	6.6	1.64	11.62		NDb	--	--
1	5.9	1.43	10.42		0.029	0.01	0.05
2	4.6	0.94	8.29		0.14	0.09	0.18
4	3.3	0.19	6.34		0.38	0.24	0.51
6	2.8	0.04	5.57		0.33	0.19	0.48
8	2.2	0.05	4.39		0.51	0.32	0.70
13	1.9	0.00c	4.04		0.46	0.23	0.69
16	1.5	0.00c	3.26		0.55	0.20	0.91
24	0.39	0.00c	0.86		0.59	0.09	1.10
32	0.12	0.00c	0.27		0.38	0.07	0.70
a Samples collected 30 min after aerial application.  b Malaoxon not detected. DL of 0.019 mg/m2.  c Negative value determined. Value assumed to be zero.

Stainless Steel Plates 

Dissipation of malathion and malaoxon were measured on stainless steel medium over a 32-day period following one of the aerial applications. The highest mean malathion residue level detected was 6.7 mg/m² (Table 11). This level was reported at day 0, 30 min after the sites were aerially treated. By day 4, the residue level sharply decreased to 0.73 mg/m² and slowly dissipated from this point onward. The sharp decrease in residue level at day 4 (Figure 6) coincides with rainfall on this sample day (3/19/94). On day 4, staff observed water on the surface of the steel plates and also on the underside of the plastic tarp used to protect the media from rain. It is possible that residue may have been washed off the smooth, flat surface of the steel plates, resulting in lower detected values at all sites. This occurrence was not observed with the sand media, because the raised edges of the aluminum sand trays retained the residue and sand media within the tray.

Table 11. Means and confidence intervals for malathion and malaoxon residue on stainless steel sheet media
	Malathion (mg/m2)				Malaoxon (mg/m2)
		95% Confidence Interval				95% Confidence Interval
Days Post  Application	Mean	Lower	Upper		Mean	Lower	Upper
0a	6.7	1.69	11.6		NDb	--	--
1	6.0	1.37	10.6		0.03	0.02	0.03
2	5.7	1.04	10.3		0.05	0.03	0.07
4c	0.73	0.00d	1.60		0.03	0.02	0.05
6	0.28	0.00d	0.66		0.03	0.01	0.04
8	0.23	0.00d	0.57		0.02	0.01	0.03
13	0.09	0.00d	0.25		0.03	0.00	0.05
16	0.02	0.00d	0.04		0.01	0.01	0.02
24	0.006	0.00d	0.02		0.007	0.00	0.01
32	0.009	0.00d	0.03		0.007	0.00	0.01
aSamples collected 30 min after aerial application.  bMalaoxon not detected. DL of 0.005 mg/m2.  cDecreased malathion levels attributed to rain were observed on this day.   dNegative value determined. Value assumed to be zero.

To compensate for the sharp drop in malathion levels associated with interference from rain, samples were divided into two distinct groups: pre-rain samples and post-rain samples (Appendix B). Pre-rain samples were measured on days 0, 1, and 2, while post-rain samples were collected on days 4 through 32. Ignoring site differences (which are discussed in Appendix B), the overall dissipation curve indicated a pre-rain half-life of 3.5 days and a half-life of 4.3 days for post-rain data.

The half-life of malaoxon was not determined because it was being formed from oxidation of malathion as well as undergoing environmental degradation. The highest observed mean malaoxon level, however, was 0.05 mg/m² on day 2. A prediction for the occurrence of the maximum malaoxon levels could not be given as was done for malaoxon on sand medium due to the break in the dissipation curve caused by the rain event.

DEF Analysis 

Levels of DEF averaged 0.14 mg/m² on stainless steel sheet surfaces 30 min after aerial application. The DEF residues were 144% of the expected deposition rate of 0.096 mg/m² which was based on the theoretical percentage of DEF in the formulation (Table 12). Degradation of DEF was not studied in this monitoring program.

In 1990, DEF levels measured on Kimbie and Teflon sheet media were 0.023 and 0.026 mg/m², respectively (Segawa et al., 1991). Although these respective levels were much lower (only 10 and 11% of the 1990 expected deposition rate) than the current DEF values, results may not be comparable between the two studies due to dissimilar chemical extraction procedures used on the media. The 1990 Kimbie and Teflon sheet samples were extracted for malathion and malaoxon residues and later analyzed for DEF residue while the current study samples were extracted and analyzed solely for DEF residue.

Table 12. DEF deposition on 10 samples of stainless steel sheet media
	DEF (mg/m2)
Mean	0.14
Minimum	0.015
Maximum	0.44
Stdev	0.12
95% Confidence Interval for the Mean	0.037 to 0.24
Detection limit of 0.011 mg/m2.

CONCLUSIONS
Conclusion 1: The malathion mass deposition rate measured on Kimbie material in 1994 was approximately two times lower than the mean deposition rates reported in either 1990 or 1981. The low deposition rate was attributed to the decreased malathion level in the malathion-bait mixture used by CDFA in the current eradication project. Malaoxon was not detected in mass deposition samples measured on Kimbie material in 1994.

Conclusion 2: The total malathion level measured in air was highest during the first 24-hr post application sampling period.

Conclusion 3: Surface water malathion and malaoxon concentrations in 1994 were similar to levels detected in the previous Medfly studies. The highest malathion and malaoxon concentrations for all three study years were observed in rain runoff water when malathion and malaoxon were washed off treated surfaces.

Conclusion 4: Estimated half-lives of malathion on sand and steel sheet media were less than six days. Results tend to indicate that malathion is short-lived.

Conclusion 5: The average DEF level on steel sheet medium (30 min after the spray) was greater than the expected target rate for this compound.

REFERENCES 

Andrews, C. Memorandum to Ron Oshima dated February 25, 1994. Subject: Malathion sample analysis results. Department of Pesticide Regulation.

Bouyoucos, G.J. 1962. Hydrometer method improved for making particle size analysis of soils. Agro. J. 54: 464-465.

Bradman, M.A., M.E. Harnly, L.R. Goldman, M.A. Marty, S.V. Dawson, and M.J. DiBartolomeis. 1994. Malathion and malaoxon environmental levels used for exposure assessment and risk characterization of aerial applications to residential areas of southern California, 1989-1990. J. Expos. Anal. Environ. Epidemiol. 4: 49-63.

Brown, M.A., M.X. Petreas, H.S. Okamoto, T.M. Mischke, and R.D. Stephens. 1993. Monitoring of malathion and its impurities and environmental transformation products on surfaces and in air following an aerial application. Environ. Sci. Technol. 27: 388-397.

Chowdhury, A., R. Goswami, and S. Pal. 1984. Persistence of malathion in four soils of West Bengal. Indian J. Agric. Sci. 54: 763-766.

Echelberger, J.W. and J.J. Lichtenberg. 1971. Persistence of pesticides in river water. Environ. Sci. Technol. 5:541.

Foote, R.H., F.L. Blanc, A.L. Norrbom. 1993. Handbook of the fruit flies (diptera:tephritidae) of America north of Mexico. Comstock Publishing Associates. Ithaca, New York.

Giles, R.H. 1970. Wildlife Monographs. The ecology of a small forested watershed treated with the insecticide malathion -S³⁵. The Wildlife Society, Columbus, Ohio.

Guerrant, G.O., L.E. Fetzer, Jr., and J.W. Miles. 1970. Pesticide residue in Hale County, Texas, before and after Ultra-Low Volume aerial application of malathion. Pesticides Monitoring Jour. 4: 14-20.

Hausenbuiller, R.L. 1972. Soil science principles and practice. Third printing. Wm. C. Brown Co., Dubuque, Iowa.

Lichtenstein E.P. and K.R. Schulz. 1964. The effects of moisture and microorganisms on the persistence and metabolism of some organophosphorous insecticides in soils, with special emphasis on parathion. J. Econ. Ent. 57: 618-627.

Mulla, M.S. 1963. Persistence of mosquito larvacides in water. Mosq. News 23: 234.

Mulla, M.S., L.S. Mian, and J.A. Kawecki. 1981. Distribution, transport, and fate of the insecticides malathion and parathion in the environment. In Residue Reviews, Volume 81. Springer-Verlag, New York. .

Neal, Rosemary H., Patrick M. McCool, and Theodore Younglove. 1993. Assessment of malathion and malaoxon concentration and persistence in water, sand, soil and plant matrices under controlled exposure conditions. Department of Pesticide Regulation. EH 93-03.

Paschal, D.C. and M.E. Neville. 1976. Chemical and microbial degradation of malaoxon in an Illinois soil. J. Environ. Qual. 5: 441-443.

Oshima, R.J., L.A. Neher, T.M. Mischke, D.J. Weaver, and O.S. Leifson. 1982. A characterization of sequential aerial malathion applications in the Santa Clara Valley of California, 1981. Department of Pesticide Regulation. EH 82-01.

Roberts, J.E., R.D. Chisholm, and L. Koblitsky. 1962. Persistence of insecticides in soil and their effects on cotton in Georgia. J. Econ. Entomol. 55: 153-155.

Rowlands, D.G. 1964. The degradation of malathion on stored maize and wheat grain. J. Sci. Food Agr. 15:824.

Segawa, R.T., J.A. Sitts, J.H. White, S.J. Marade, and S.J. Powell. 1991. Environmental monitoring of malathion aerial applications used to eradicate Mediterranean fruit flies in southern California, 1990. Department of Pesticide Regulation. EH 91-3.

Sittig, Marshal, editor. 1980. Pesticide manufacturing and toxic materials control encyclopedia. Noyes Data Corporation. Park Ridge, New Jersey

Van Middelem, G.M. 1965. Fate and persistence of organic pesticides in the environment. In Organic pesticides in the environment. Advanc. Chem. Sr. 60.

Voss, H.J. 1990. Letter to Senator Art Torres, dated February 20, 1990. California Department of Food and Agriculture.

Walker, W.W. and B.J. Stojanovic. 1973. Microbial versus chemical degradation of malathion in soil. J. Environ. Quality 2: 229-232.

Ware, G.W., N.A. Buck, and B.J. Estesen. 1984. Deposit and drift losses from aerial ultra-low-volume and emulsion sprays in Arizona. J. Econ. Ent. 77: 298-303.




APPENDIX A - Laboratory Quality Control Data 

I. Method Validation Data

II. Continuing Quality Control Data

III. Inter-laboratory Quality Control



APPENDIX A 

I. Method Validation Data



APPENDIX A 

II. Continuing Quality Control Data

Inter-laboratory Quality Control 

APPENDIX A 


III. Inter-laboratory Quality Control

Water 

The CDFA and APPL Laboratories analyzed split water samples for inter-laboratory comparison. Split water samples included background samples collected before the spray, samples collected the day of the spray, and samples collected during periods of precipitation.

Analyses between both laboratories showed good agreement when both laboratories reported detectable residue levels for either compound. Of the 42 surface water samples split between CDFA and APPL Laboratories, there were seven samples that were in disagreement for the detection of either malathion or malaoxon (Table A-1). CDFA found more positive samples than did APPL Laboratory for these seven samples.

Table A-1. Inter-laboratory distribution of split water samples
	Malathion		Malaoxon
Variable	CDFA/APPL		CDFA/APPL
Number of Positive Split Samples in Agreementa	28		20
Number of Non-Positive Split Samples in Agreementb	11		18
Number of Split Samples in Disagreement	3		4
Total Number of Split Samples Analyzed	42		42
a Residue detected in split samples analyzed by both laboratories.   b Residue not detected in split samples analyzed by both laboratories.

Malathion was detected in 28 application day and rain runoff samples analyzed by CDFA and APPL Laboratories (Table A-2). Malathion concentrations ranged from 0.25 to 24.4 ppb. The range was determined excluding the results from one sample (CDFA; 203 ppb and APPL; 230 ppb) which were well beyond the range of other observed concentrations. These high values were detected in rain runoff samples. Regression analysis was used to investigate differences between the two sets of results. Unweighted regression procedures were used assuming that precision did not vary with concentration. In the regression of APPL concentrations on CDFA concentrations the result CDFA;203 ppb and APPL;230 ppb was flagged as an observation having substantial influence on the regression estimates. A log-log transformation on the data gave identical results. The regression analysis with this observation included, indicated evidence of systematic differences between the two sets of results for concentration of malathion in water samples. A regression of APPL concentrations on CDFA concentrations indicated that the calculated slope did not differ significantly from 1 and the calculated intercept did not differ significantly from 0 at the 5% level of significance. The correlation coefficient was 0.91. The result, CDFA; 15.3 ppb and APPL; 3 ppb was flagged as an observation having a large standardized residual in comparison to other results. The regression analysis, with or without this observation, indicated there was no evidence of systematic differences between the two sets of results for concentration of malathion in water samples. The regression results based on the reduced data set (influential case removed) were considered to describe the majority of the data observed.

Malaoxon was detected in 20 water samples analyzed by CDFA and APPL Laboratories (Table A-2). Malaoxon concentrations ranged from 0.09 to 16 ppb. The range was determined excluding two samples (CDFA; 23.4 ppb and APPL; 44 ppb and CDFA; 34.4 ppb and APPL; 56 ppb) because they were beyond the range of concentration common to the other 18 water samples. These high levels were detected in rain runoff samples. A regression of CDFA concentrations on APPL concentrations indicated that the calculated slope was not significantly different from 1 and the calculated intercept was not significantly different from 0 at the 5% level of significance. The correlation coefficient was 0.98. The sample results, CDFA; 6.15 ppb and APPL; 3.7 ppb and CDFA; 9.65 ppb and APPL; 3.7 ppb were flagged as observations having large standardized residuals in comparison to other observations. The regression analysis with or without these observations indicated there was no evidence of systematic differences between the two sets of results for concentration of malaoxon in water samples.

Table A-2. Inter-laboratory comparison of malathion and malaoxon results in surface watera
Malathion (ppb)			Malaoxon (ppb)
CDFA	APPL		CDFA	APPL
0.25	0.3		0.09	0.15
0.31	0.1		0.10	0.1
0.37	0.2		0.11	0.1
0.42	0.66		0.13	0.1
0.42	0.3		0.16	0.2
0.47	0.5		0.16	0.1
0.54	0.6		0.17	0.1
0.58	0.6		0.39	0.3
0.60	0.4		0.68	0.4
0.72	0.80		1.09	0.9
1.04	0.9		1.36	0.8
1.40	1.2		1.40	0.2
2.02	2.8		1.61	1.9
3.49	2.7		1.97	1.5
3.62	3.5		6.15	3.7
4.01	3.7		9.65	6.3
5.16	3.8		10.8	10
5.45	4.3		16.0	15
5.67	4.1		23.4	44
7.20	7.3		34.4	56
9.25	13
11.0	11
13.9	12
14.0	18
15.3	3.0
17.9	18
24.4	36
203	230
a Malathion and malaoxon results reported crosswise from each other are not necessarily cross-linked residue levels. Only positive values reported by both laboratories are presented.

Soil 

Analyses of two split soil samples suggested possible differences between reported results from CDFA and APPL Laboratories. The sample size was too small to perform a statistical test comparing differences between laboratory results. CDFA Laboratory found approximately 1.4 to 3 times more malathion on an area basis in both samples than did APPL Laboratory. Malaoxon was not detected by either laboratory.



Sand 

Comparison of CDFA and APPL Laboratories results for four split sand samples tended to vary for malathion, but were similar for malaoxon levels (Table A-3). The sample size was too small to compare statistically.

Table A-3. Inter-laboratory comparison of malathion and malaoxon concentrations on sanda
	Malathion (mg/m2)			Malaoxon (mg/m2)
Sample Date	CDFA	APPL		CDFA	APPL
Day 4	3.80	8.17		0.37	0.40
Day 6	3.96	6.92		0.43	0.60
Day 8	0.28	0.20		0.28	0.19
Day 16	2.17	2.34		0.65	0.80
a Individual values presented.

APPENDIX B - Statistical Analysis 

Statistical Analysis 


Mass Deposition on Kimbie Material 

The hypotheses of normality and homogeneity of variances were not rejected for any of the 1994 within spray samples at the 5% level of significance using the Ryan-Joiner procedure and Levene's test, respectively. Assuming that the population from which the samples were taken were normal and that the population variances were equal, the means from each of the three sprays did not differ significantly at the 5% level when tested using a one-way analysis of variance. Confidence intervals for 1994 data were calculated for the means assuming normality. Normal probability plots of the raw deposition data and logarithmically transformed data were completed for each spray within each year (i.e., 1994, 1990, and 1981) and for combined data of all sprays within each year. Results were inconclusive using the Anderson-Darling Normality test to determine the form of the underlying within spray distribution or the combined yearly deposition distribution. Given that the underlying distribution was uncertain, non-parametric estimation techniques for data analyses are also provided.

Median values are presented for each spray (Table B-1) with corresponding 95% confidence intervals.

Table B-1. Median and the 95% confidence interval for mass deposition of malathion on Kimbie materiala
	Malathion (mg/m2)
		95% Confidence Interval
Application Date	Median	Lower	Upper
Spray 1 (30)b	8.15	6.14	10.5
Spray 2 (30)	9.73	6.18	12.6
Spray 3 (29)	9.16	5.94	11.1
Combined Sprays (89)	9.11	6.94	10.4
a Malaoxon was not detected. DL of 0.011 mg/m2.  b Spray day followed by number of samples in parenthesis.

The H test of Kruskal and Wallis was used to compare the 1994 spray medians. These results indicated that there was no reason to reject the null hypothesis that the three samples came from a common population at the 5% level of significance. The overall median for the three sprays combined was 9.11 mg/m². 1994 deposition data was graphically displayed using boxplots (Figure B-1). Each boxplot consists of two boxes (a transparent box and a solid black box), a set of whiskers, and any outliers. The number of samples for each spray interval or combined spray interval is proportional to the width of the transparent box. The first quartile, represented by the bottom line of the transparent box, and the third quartile, represented by the top line of the transparent box, show the theoretical distribution of individual samples.

A horizontal line drawn across the transparent box represents the median value while the solid black box represents the confidence interval for each median. The whiskers are the vertical lines that extended from the top and bottom of the transparent box to the adjacent values which are the lowest and highest observations that are still inside the region. Outliers are malathion deposition values outside the lower and upper limits and are plotted with asterisks. For 1994 data, only two outlier samples with high deposition rates were observed during the three sprays monitored (Figure B-1).

Figure B-1. Boxplot of the observed and expected mass deposition rate of malathion for the 1994 Medfly eradication program

The boxplot of the deposition data encompassing the three study years (Figure B-2) shows that 1981 data differs in location when compared to 1994 and 1990 data with regards to the expected malathion target rate (indicated by a dashed horizontal line). The three study years are similar in spread; however, the 1981 malathion distribution is skewed to the left with many samples (outliers) well above the expected target rate.

Figure B-2. Boxplot of the distribution of the malathion mass deposition rate in the 1994, 1990, and 1981 Medfly eradication programs, using a fraction (ratio) of the expected rate

For each study year, the distribution was further collapsed into three categories: the percentage of samples within + 25% of the expected application rate, the percentage of samples more than 25% below the expected application rate, and the percentage of samples more than 25% above the expected application rate. Visual comparison of results show that similar deposition patterns were observed in 1994 and 1990. Table B-2 shows that approximately 40% of the samples were within 25% of the expected target rate. This contrasts with 1981 results which only had 25% of its samples falling near the expected rate. Malathion appeared to be under applied (59%) to the vast majority of the 1981 samples.

Table B-2. Relative frequency of malathion samples within 25% of expected application rate for the 1994, 1990, and 1981 Medfly eradication programs
	Expected Malathion Application Rate  (mg/m2)a	Percent of Samples
Year		<ara - b	ar +	>ar +
1994	10.2	40.5	40.5	19.1
1990	23.78	35.0	46.3	18.7
1981	19.76	58.9	24.5	16.6
a Expected application rate.  b= 0.25 x ar.

A statistical comparison of the similarity between the three independent distributions of frequency data for the three study years was made using a x² test. The chi-square test of whether the 1994, 1990, and 1981 target distributions could have come from the same population was rejected at the p < 0.01 level of significance. On further comparison, the chi-square test of whether the 1994 data could have come from a population having similar percentages to the 1990 target distribution was not rejected (p = 0.61). The chi-square test of whether the 1994 data could have come from a population having similar percentages to the 1981 target distribution was rejected at the p < 0.01 level of significance; and the chi-square test of whether the 1990 data could have come from a population having similar percentages to the 1981 target distribution was rejected at the p < 0.01 level of significance. All critical bounds for multiple comparisons were determined from a Bonferroni x² table.

Sand 

To determine a malathion half-life, a linear least squares regression analysis was performed on logarithmically transformed malathion means (n=5) across sample days. A multiple comparison of slopes for the five sites was used to test whether malathion dissipation curves were parallel.

Initially, a separate dissipation curve was estimated for each of the five sites monitored. Linear least squares regression analysis was used to describe the relationship of logarithmically transformed malathion concentrations over time. An analysis of variance was used to compare the five slopes at the five sites and the subsequent F-test rejected the null hypothesis of no slope differences (p < 0.01). A multiple comparison of site slopes was done using the Newman-Keuls procedure. Results indicated that the dissipation curve for site 9 showed a faster decline than did sites 6, 41, and 42 (p < 0.05). There was no difference in the rate of decline for sites 9 and 29 at the 5% level of significance. Individual half-lives for the 5 sites ranged from 4.2 to 6.9 days. Combining similar site data, two separate half-lives of 4.6 days and 5.9 days resulted. Ignoring site differences, an overall dissipation curve resulted in a half-life of 6.0 days.

Malaoxon levels were found to increase and then decrease over time. A quadratic function was used to describe the relationship of malaoxon mean concentrations over time. Ignoring possible site differences (mentioned earlier with malathion residue detected on sand), an estimate was calculated for the occurrence of the maximum amount of malaoxon observed during the 32-day monitoring period using data from all sites. The estimate for the highest mean malaoxon level was 0.61 mg/m² occurring at 19.5 days post application.

Stainless Steel Plates&nb