Persistence and Bioavailability of Termiticides to Subterranean Termites (Isoptera: Rhinotermitidae) from Five Soil Types and Locations in Texas

by

R.E. Gold, H.N. Howell, Jr., B.M. Pawson, M.S. Wright, & J.C. Lutz'

ABSTRACT

The results of a 5 year study with 6 termiticides including: bifenthrin, chlorpyrifos, cypermethrin, fenvalerate, permethrin, and isofenphos indicate significant differences in effectiveness among products applied to different soil types in Texas. Each of the 5 field test locations represent very different soil types and environmental conditions. Test locations within Texas included: Lubbock, Dallas, Overton, Corpus Christi and College Station. Tenniticides were applied in 1990 with soil samples taken from the replicated treatment plots at 1 and 6 months, and then annually through 5 years. Soil residues of termiticides were measured with gas chromatography. The amount of pesticide remaining in each sampling period indicates a significant loss of termiticide by the fifth year of the test. The bioavailability of termiticide remaining in the soil was estimated through bioassays utilizing field collected subterranean termites (Reticulitermesflavipes (Kollar)). Both tunneling distance and mortality were used as indicators of termiticide activity and availability. The results of the bioassays confirmed the findings of the residue analysis portion of the project. The most stable termiticides, through 5 years, were permethrin and fenvalerate. Isofenphos was the least stable with significant loss of activity within 24 months post-application. The most challenging conditions, in terms of effective termiticide residuals retained through time, were alkaline soils with high clay content and organic compositions greater than 1%. The most favorable soils were those that are acidic with low clay and organic content.

INTRODUCTION

This research project was initiated at the request of representatives of the Texas Pest Control Association, based on their perspective that termiticides available in 1988-90 were not as effective in controlling termites as were aldrin, chlordane and heptachlor. The professional pest control industry believed that retreatment rates during the first year were higher with the modern termiticides than they had been in the past. Efforts to decrease the need for retreatments have been made (Potter 1994). but the problem persists. Unfortunately, records which we examined were incomplete and direct comparisons between retreatment rates were not possible. It was agreed that we would undertake a research project to examine the efficacy and residual activity of termiticides under Texas conditions. This report provides the results through the fifth year of the study. Preliminary findings are available in Gold et at. (1996).

Subterranean termites have been collected from all regions of Texas (Howell et at. 1987). They are considered to be of major economic importance in terms of the damage they do, the cost of repairs, and the costs associated with tern-lite prevention and control. McLIveen et at. (1993) estimated that the value of termites to the economy of Texas exceeded $250 million/year based on a survey of 3000 licensed pest control companies. Total loss due to termites in the United States was estimated at $1.7 billion/year (Gold et al. 1993). It has been estimated that termites do more damage than all tornadoes, hurricanes, and wind storms combined and involve 5 times as many houses as fire (Granovsky 1979, 1983, Granovsky & Sadberry 1983). In the coastal regions of Texas, by the time a structure is 40 years old, the probability of infestation with one or more species of termites exceeds 90%.

The modem prevention and control of tcrmite damage has relied primarily on the use of termiticides (Gold et al. 1994, 1996). There have been relatively few changes associated with the chemical control of termites for over 50 years. The basic concept of establishing a chemical barrier around a structure is apparently as effective today as it has been in the past. There is considerable information available about the merits of soil applied termiticides which act as barriers to tunneling termites (Su & Scheffrahn 1990, Su et al. 1995, Jones 1990, Smith & Rust 1990, 1991, 1992, Grace 199 1, Gold et al. 1994, 1996 and Forschler 1994). The major changes have come in the form of different chemical groups of termiticides, particularly organophosphates (chlorpyrifos and isofenphos) and pyrethroids (bifenthrin, cypermethrin, fenvalerate, and permethrin). These pesticides were originally developed for soil applications in agricultural situations where they were effective for a single season. It was believed that with an increase in application rates these agri-chemicals would be effective as termiticides. The first of the new generation of termiticides registered with the United States Environmental Protection Agency (EPA) was chlorpyrifos (Dursban TC(tm)) which by 1991 had an estimated 65. 1% of the termiticide market share (Mix 199 1). With a total market value exceeding $ 100 million/year, it was no surprise that other pesticides were labeled and marketed for termite control (Gold et al. 1994).

In order to register a pesticide as a termiticide, the EPA has required efficacy data from field tests conducted by the United States Department of Agriculture, Forest Service (FS). It was assumed that 5 years of data were required to register a pesticide as a termiticide. The testing was done at the Southern Forest Experiment Station in Gulfport, Mississippi or on their other test sites (Kard et al. 1989). In these tests, proposed termiticides were to provide 100% control (protection) of subterranean termites for 5 years in at least 3 of 5 field sites. Different types of testing were done as part of the Forest Service program (Stanley 1994) including concrete slabs and ground boards. The results of the testing have been made available to the public (Kard et al. 1989, Kard & McDaniel 1993, Kard & Mauldin 1994). It appears that there are differences in the persistence and efficacy of the termiticides included in their tests.

State regulatory agencies have demonstrated a great deal of interest in sampling soils treated with termiticides to determine the residues remaining through time (Kard & McDaniel 1993, Mix 1995). The Association of -Structural Pest Control Regulatory Officials (ASCPRO) sponsored research involving soil sampling from structures treated with termiticides. They determined that through proper sampling procedures, it is possible to use results from soil sampling for regulatory actions if sufficient pesticide is not present following treatment. Again, there were differences in termiticide concentration through time, indicating that within 180 days all pesticides included in the tests had significantly decreased in concentration.

Independent studies which compare the persistence and efficacy of pesticides in field trials used for termite control are somewhat limited; however, Suet al. (1993) and Gold et al. (1994, 1996) both reported that termiticides lost effectiveness through time. The present study was conducted to carefully compare termiticide persistence and efficacy under varying soil types and environmental conditions.

MATERIALS AND METHODS

Study Sites

The termiticide tests were conducted at 5 locations in Texas, each of which represented a specific soil type and climatic condition (Tables I and 2). All test locations are on properties owned and managed by the Texas A&M University System. It was initially anticipated that the tests would be conducted for a 15 year period. The specific locations in Texas are: Lubbock Research and Extension Center; Dallas Research and Extension Center; Overton Research and Extension Center; College Station (Easterwood Airport); and the Corpus Christi Research and Extension Center. All of the termiticides included in these tests were included in replicated test plots at each of the five locations.

Table 1. Soil characteristics for Texas termiticide test sites

Termiticide Applications

At each test location, the termiticides were applied as per the recommendations of the manufacturers including: isofenphos (Pryfon 6 insecticideTM: -Bayer @ 0.75% A.I.); bifenthrin (Biflex T.C. (tm): FMC @ 0.062% A.I.); chlorpyrifos (Dursban TC(tm): DowElanco @ 1.0% A.I.); cypermethrin (Demon TC(r): Zeneca @ 0.25% A.I. and Prevail FT(r): FMC @ 0.30% A.I.); fenvalerate (Tribute R : AgroEvo @ 0.50% A.I.); and permethrin (Dragnet FTR: FMC @ 0. 50% A. I. and Torpedo r: Zeneca @ 0.5 A. Q. Each termiticide and recommended test concentration was replicated 3 times at each of the 5 locations. Each replication consisted of four applications arranged in a cluster. The clusters were formed by auguring soil out of 4 holes, each was 25cm in dia. and 30.5cm deep. All 4 holes per cluster were dug at the same time. The soil removed was batched, screened through a 1 cm hardware cloth, and placed in a concrete mixer. The same weight of soil was used in each treatment batch (approximately 58kg). With the mixer turning, termiticides was applied uniformly to the soil. The mixing process continued until thorough mixing had been accomplished (15 minutes per batch). The mixer was thoroughly cleaned between batches to minimize carry over of termiticide between treatments.

Table 2. Mean meteorological data for Texas termiticide test sites.

The amount of termiticide applied to the soil in the mixer was calculated to provide the same concentration as would be found in a post- construction treatment at the concentration and rate recommended by the manufacturer. A layer of white sand was placed in the bottom of each hole. This contrasting band was used during sampling as an indication of the boundaries for the treatments. After the soil and termiticide had been thoroughly mixed, the holes were refilled with the treated soil, tamped (to a density of approximately 3 g/CM3), and leveled to surrounding grade levels. Into the first hole in the cluster was placed a 20cm. pine stake; into the second, a pine stake was driven, but in addition the hole was covered with a precast concrete block (5 x 30.5 x 30.5cm); the third hole was left exposed and was used for residue analysis; and the fourth hole was covered with a similar concrete block and was also used for soil sampling. A metal spike (25.4cm) was driven in the 'Center of each filled hole to facilitate finding the center of the plot at each sampling period. The pine stakes were used to monitor termite activity and efficacy of the treatments. Records were kept of as to which of these pine stakes were infested by termites since the prior sampling period. Control (non-treated) plots were included in equal numbers with treatments at all of the 5 test sites. Soil from the control clusters were handled and analyzed as were the termiticide treated soils.

Soil Samples

Soil sampling was done with a 2.5cm dia. soil probe pushed 30.5cm in the sampling site. The resulting core was carefully placed in a plastic specimen bag. Soil sampling was done at specific times: pretreatment (time 0), 1, 6, 12, and 18 months, and then annually through 4 years. As each sample was taken, the resulting hole was filled with a contrasting color of sand to insure that the area was not resampled at a later time. Soil samples were frozen and transported to the analytical laboratory in College Station, Texas, where they were held at -50C until extraction, chromatographic analysis, and bioassay.

Sample Preparation

In the laboratory, the soil core was separated into 3 distinct sections (top, middle and bottom). Each section was blended to produce an homogeneous mixture. Three 5g subsamples were taken from each sample, and each subsample was then placed in individual 25mL plastic scintillation vials to which was added 20mL of acetone or other appropriate solvent. All samples were then agitated for 30min on a horizontal shaker, after which time they were allowed to settle over night. The following day, supernatant was diluted with the appropriate solvent to an acceptable level for electron capture detection (ECD) on a gas chromatograph. Extraction efficiencies and moisture determinations were also performed and incorporated into the concentration calculations.

Chemical Analysis

A series of Perkin Elmer Autosystem Gas Chromatographs (Model 9000) fitted with ECD detectors and auto injectors were used in this work. Carrier gas was Zero grade helium, and make up gas was a mixture of 5% methane:95% argon. The capillary column used for all samples was a Restek XTI-5 megabore 0.53mm I.D., 30 meters long, 1. 5mm df, 5% diphenyl-95% dimethyl polysiloxane. Data collection and analysis was performed on a Perkin Elmer model 1020S digital data analysis system. Performance of the methods and instrumentation was monitored through the use of analytical standards (ChemServices, West Chester, Pa.) inserted before each successive plot number, and linear regression curves were constructed for each termiticide used. The overall method and instrument sensitivity was 0. 1 mg/g with an extraction efficiency of 91.4% based on spiked soil samples.

Instrument conditions for bifenthrin, permethrin, fenvalerate, and chlorpyrifos were as follows: Carrier gas=7ml/min; makeup gas=23ml/ min; attentuation=32: temperatures=inJection port @ 250 C detector @ 375 C column (isothermally) @ 225 C Instrument conditions for cypermethrin were as stated above except that the injector temperature was 2700C and column temperature was 245 C Instrument conditions for isofenphos were the same as permethrin, except that the column was thermally programmed from an initial temperature of 1600C and then ramped at +5C/minutes to 2200C and held for 5 minutes, then ramped at +30C/min to 2400C and held for 5 minutes.

Bioassay Analysis

Two types of bioassays were used to evaluate termiticide effectiveness. The first involved determining if the termiticide afforded protection to wood. At the time termiticides were mixed and placed in the ground, pine stakes were placed in each of the treatment plots. With each soil sampling period, the pine stakes were removed and examined for evidence of termite infestation or feeding. If a stake was attacked by termites, the termiticide applied to the surrounding soil failed to protect the wood from termites. If the stake was not attacked, the termiticide may have protected the wood from attack, or the wood may have been missed by termites.

The second type of bioassay tested the effects of termiticides directly on termite behavior. The soil bioassays presently being used were derived from earlier work by Su et al. (1 993) and Gold et al. (1994, 1996). It involved taking ca. 12g of soil from each sample bag (a soil core) which represented one replication of termiticide per time interval. The specific soil used in the bioassays was taken from the middle section of the cores from the covered plots. Three replications were done for each termiticide at each site and time interval. All soil samples were air dried overnight. The following day 1 ml of distilled water was added to each soil sample and then thoroughly mixed. A 2cm agar plug was placed ca. 3cm from one end of a 1.6cm O.D. x 15 cm glass tube and was designated the top. The moistened soil was carefully placed in the bottom of the tube and lightly packed to remove air pockets within the soil. After 5cm of soil was packed into the tube, a second 2cm agar plug was inserted and pushed into the bottom of the glass tube until it contacted the soil. Once the soil and agar plugs were in place, a 3cm pieces of wooden applicator stick was placed in the bottom of the tube. The end of the tube was then covered with a piece of aluminum foil. To the top of the glass tube, 30 pseudergates (Reticulitermes flavipes [Kollar]) were added and then sealed with another piece of aluminum foil. Pieces of aluminum foil were held in place by orthodontic rubber bands. After assembly, bioassay tubes were placed in an upright position in a rack and held at 250C with a 12:12 (L:D) photoperiod.

Each bioassay tube was checked for termite tunneling after 24 hours. Termite tunneling was recorded from 0 mm (soil/agar interface at the top of the bioassay tube) to 50 mm (soil/agar interface at the bottom of the tube). After 5 days, final termite tunneling was recorded and the bioassay tubes were carefully dissembled to determine the number of surviving termites. All values presented are means of three replications.

Controls were performed in conjunction with the bioassay tests and involved using untreated soil from the same geographical location as the test plots. All controls were set up similarly to treated soils. Control soils were used to indicate whether the termite tunneling and survival could be attributed to the termiticides being tested instead of the behavior of the termites themselves. Bioassay results for the 60 month samples are reported.

Statistical Analysis

Pesticide residue data are expressed as mg/g remaining at each test site per time period. Data were analyzed by analysis of variance (SAS Institute 1987). Significant differences in treatment means were calculated by LSD at p=0.05.

RESULTS

Chemical Analysis

The methods and procedures developed for this research project proved to be effective in the extraction, detection, and quantification of the termiticides applied to soils in the field plots (Gold et al. 1996). Extraction efficiencies varied slightly between pesticides, but had an overall mean of 91.4t3.2% with limits of detection at 0. 10 mg/g. We recognize that it is possible to develop procedures which are even more sensitive to individual pesticides, however, the methods used allowed us sufficient data to draw conclusion from this work. The methods were consistent throughout the five years of the study, making it possible to compare data from year to year. The results of the pretesting of soils indicated that there were no residues of pesticides in the plots at initiation of the study.

The results of the soil analysis are presented in Tables 3- 7 (at the end of this article). It is apparent from these summaries that all of the termiticides were significantly degraded through the 5 years of the study. There were differences between termiticide products in terms of residues remaining in the fifth year. Part of these differences are explained by the chemical makeup of the individual products used in these trials, while other differences may have been due to soil type, pH, and organic content (Table 1). Weather may also have played a role in affecting termiticide degradation (Table 2). RainfW1 varied from 42 1 mm/ year in Lubbock to 1189 mm/year in Corpus Christi. Mean temperatures, particularly in the winter months, varied markedly. The ground was frozen in Lubbock while the mean temperature in Corpus Christi was relatively warm.

There were no significant differences in residues remaining in the top, middle and bottom of the soil samples, so the data present were combined (Tables 3-7). There were no significant differences between those plots that were covered with concrete as compared to those that were exposed to the elements. It was determined there was a difference between covered and uncovered for isofenphos at 24 months in the Lubbock plots, but this difference was not noted in later sampling periods or from other locations (Gold et al. 1994).

Estimates of the half-life of the termiticides in these tests varied markedly, but with all products less than 50% of the active ingredients were present at I year posttreatment. Isofenphos had the shortest halflife (less than 90 days) as compared to the other termiticides (estimated at 9 months overall). By 36 months isofenphos had degraded (both isofenphos and isofenphos oxon) to undetectable levels at all 5 sites. All other termiticides had detectable residues at least through 48 months posttreatment. Degradation was the slowest at the Lubbock and Overton sites, and was most rapid at the Corpus Christi and Dallas locations. The College Station site was most similar to Corpus Christi and Dallas, but had a slightly slower rate of degradation.

The most persistent of the termiticides tested were permethrin and fenvalerate (Tables 3-7). Both of these pyrethroids were recovered at consistently higher levels than the other termiticides through 60 months. The least persistent was isofenphos, which as indicated above, was undetectable at 36 months. Chlorpyrifos appeared to be more persistent at the Lubbock and Overton sites as compared to Corpus Christi and Dallas where the soils had higher clay contents.

There were variations in the initial concentration (time 0) of the termiticides at all sites. We made every effort to keep the variations to a minimum, but the results show that certain products were more difficult to measure, mix and apply than others (Tables 3-7). The most variation occurred with isofenphos and the two perniethrin termiticides. This was due in part to formulations differences which made it difficult to agitate the products in the original containers and obtain a consistent, uniform technical product to measure and apply. Certain soils were generally more difficult to treat, particularly those in Corpus Christi and Dallas which were rich in clays. It proved to be difficult to get even mixing of termiticide with the heavier soil throughout the study.

Bioassay Analysis

Termite bioassay results are reported for the 60 month samples (Figs. 1-5). Results have been grouped by test location for ease of comparison. Results from earlier sampling periods are reported in Gold et al. (1996). Termiticide effectiveness as a barrier to termites varied with location and active ingredient. It was apparent that two different responses of the termites to the treated soils were taking place. Chlorpyrifos, an organophosphate insecticide, caused mortality to the tunneling termites particularly at the Lubbock (Fig. 4) and Overton (Fig. 5) sites. Termites tunneling in chlorpyrifos treated soils were exposed to terniiticide and died within the 5 days of the tests. Pyrethroid termiticides were repellent to termites and generally inhibited tunneling, but had limited effects on mortality (Figs. 1- 5). Both termite responses were dependent on the concentration of termiticide present at 60 months. This result supports the findings of Su & Schaffrahn (1990).

Bifenthrin inhibited R.flavipes tunneling in all but the Dallas (Fig. 3) and Overton (Fig. 5) sites where the concentration of termiticide remaining at 60 months was 0.9 and Oppm, respectively. The mean concentration at the other sites was 1.2ppm; however, this was sufficient to minimize tunneling by the termites.

Fig. 1. The mean tunneling distance and percentage (%) survival of 30 subterranean termites (R. flavipes) held 5 days in bioassay tests with soils from Corpus Christi, TX treated 60 months prior with the termiticides indicated.

As indicated above, chlorpyrifos continued to provide an effective barrier to termites in the bioassay tests in the sandy soils represented by the Lubbock and Overton sites (Figs. 4 and 5). The concentration of chlorpyrifos at 60 months for these sites was 8 and 66ppm, respectively. Chlorpyrifos was not effective in preventing tunneling or causing mortality at Corpus Christi (Fig. 1), College Station (Fig. 2), or Dallas (Fig. 3) where the concentrations were 0, 2, and 0.5ppm, respectively. . Cypermethrin is sold as both Demon TCR (Zeneca) and Prevail FIR (FMC). In the bioassay tests, this termiticide was effective as a barrier treatment except at the Dallas site (Fig. 3) where the mean concentration was less than 0. 1 ppm. There was considerable tunneling activity in the soils from the Corpus Christi (Fig. 1) and College Station (Fig. 2) repellent termiticide was best demonstrated at the Lubbock (Fig. 4) and Overton (Fig. 5) sites where only slight to moderate tunneling activity was noted. As with the other pyrethroid termiticides, cypermethrin caused only sites where the mean concentration was less than lppm remaining at 60 months. Efficacy of cypermethrin as alight mortality to the termites.

Fig. 2. The mean tunneling distance and percentage (%) survival of 30 subterranean termites (R. flavipes) held 5 days in bioassay tests with soils from College Station, TX treated 60 months prior with the termiticides indicated.

Fenvalerate (TributeR) was effective as a barrier to tunneling termites at all test sites through 60 months (Figs. 1-5). More tunneling occurred in the Dallas (Fig. 3) soils where the mean concentration was 7ppm, however, even at this site the termites did not breach the 50mm barrier. At Overton (Fig. 5), it was noted that fenvalerate may have caused mortality to the tunneling termite population based on a 30% survival rate at a mean termiticide concentration of 66ppm.

Permethrin is sold as Dragnet Fr (FMC) and TorpedoR (now Prelude R from Zeneca). This pyrethroid termiticide was persistent and efficacious through 60 months at Corpus Christi (Fig. 1), Lubbock (Fig. 4) and Overton (Fig. 5) where the mean concentration at 60 months was 7.5, 2 1, and 65ppm, respectively. At the College Station site, the termites breached the treatment barrier in the bioassay test where the mean concentration of TorpedoR was 4ppm as compared to Dragnet FTRat I ppm. Apparently in the soil type represented by College Station, the minimum inhibitory concentration (MIC) required to prevent tunneling is between 4 and 7ppm. In Dallas (Fig. 3), the Dragnet FV barrier was breached (mean concentration of 0.4ppm), while TorpedoR remained effective even though there was considerable tunneling (40 mm) at a mean termiticide concentration of 3ppm. The results of the bioassay tests indicated that permethrin is among the most persistent and efficacious of the termiticides used in the field trials.

Fig. 3. The mean tunneling distance and percentage (%) survival of 30 subterranean termites (R. flavipes) held 5 days in bioassay tests with soils from Dallas, TX treated 60 months prior with the termiticides indicated.

The other organophosphate in the study was isofenphos (Pryfon 6 Insecticidem. Results from the 36 month bioassays supported the results obtained from the chemical analysis (Gold et al 1996). Soils treated with isofenphos from the five test locations failed to prevent termites from tunneling, nor did it kill them. Reticulitermesfiavipes tunneled completely through all but one location, Lubbock, and high survival occurred in the bioassays from all 5 locations. Because isofenphos was not detected at 36 months (Tables 3-7), 60 month bioassays were not performed.

Fig. 4. The mean tunneling distance and percentage (%) survival of 30 subterranean termites (R. flavipes) held 5 days in bioassay tests with soils from Lubbock, TX treated 60 months prior with the termiticides indicated.

Bioassay results from untreated controls were consistent for the 60 month tests (Figs. 1-5). Termites were able to tunnel completely through all soil types during the test period, and termite survival was good. Controls are important for indicating the health of the termites being subjected to termiticides so that true behavior can be assessed and proper termiticide evaluations can be made.

The results of bioassays which utilized pine stakes driven into the test plots to measure termite infestation were inconclusive. There was no uniform infestation by native termites in any of the 5 test sites. Neither the control or treatment plots were satisfactorily infested within the 60 months of the field trials. It must be noted, had we depended on the infestation data, we could have erroneously concluded that all termiticides were equal in efficacy and persistence through 5 years.

Fig. 5. The mean tunneling distance and percentage (%) survival of 30 subterranean termites (R. flavipes) held five days in bioassay tests with soils from Overton, TX treated 60 months prior with the termiticides indicated.

DISCUSSION

The results of this research clearly indicate that modem termiticides applied to soils in Texas are unlikely to last as long as aldrin, chlordane or heptachlor. The chlorinated hydrocarbon termiticides have long been recognized as providing long term protection of structures from termite invasion. The retreatment rate with the organophosphate and pyrethroid termites appears to vary markedly depending on the application procedures used as well as the specific products and rates of application. Based on the results we obtained, all the termiticides we tested should have provided excellent control of termites through at least two years (Tables 3-7). This is based on the fact that all the termiticides residues were well above the minimum threshold levels needed to either kill or repel invading termite foragers (Su & Scheffrahn 1990, Gold et al. 1996). There was little doubt however, that the termites were significantly degraded from the initial application levels within 6 months.

Several questions have been raised which need clarification. We recognize that the results of our testing procedures are different than those of the USDA Forest Service (Kard et al. 1989, Kard & Mauldin 1994, Stanley 1994), but more closely approximate those represented by other published studies (Mix 1995, Kard & McDaniel 1993, Su et al. 1993). Our testing procedures were basically different in that we took soil samples from the test plots which were analyzed for termiticide residues with a chromatograph as well as with bioassays utilizing live subterranean termites.

We recognize that the utilization of "post hole tests" may not represent what happens to a termiticide applied in and around a structure. We believe the methods used represents an attempt at a reductionist approach to managing variables. By thoroughly mixing weighed amounts of soil, and adding known amounts and concentrations of terniiticide, we presented the best possible situation as compared to applications made in the field with customary application equipment. We have done this type of field work (Gold et al. 1993), and can report that the methods used in this research resulted in significantly more uniform applications of pesticides than those occurring in normal commercial situations.

The termiticides tested performed the best, in terms of efficacy and persistence, in sandy soils, which are at least slightly acidic, with low organic content (see Overton site in table 7). The worst situation was an alkaline soil, with high levels of both clay and organic matter(see Dallas site in table 5).

The results of the bioassays reinforced the importance of applying a uniform chemical barrier for termite prevention and control. We confirmed the work of Su & Scheffrahn (1990) and conclude that with the organophosphate termiticides, termites require contact with the treated soil in order to be killed. When using pyrethroids, the termites are repelled and survive inclusion in the bioassays. It is our understanding, based on the results of this work, that if there is sufficient concentration of termiticide present, treated soils will provide an effective barrier to invading termites regardless of the mode of action of the chemicals. However, if there are gaps in the chemical barrier (Forschler 1994). if the barrier layer is too thin (Su et al 1995), if the termiticide has degraded (Gold et al 1996), or if the termiticide is not available to foraging termites, then the barriers can be breached (Su et al 1993, Gold et al. 1993, Forschler & Townsend 1996), and termite invasion and infestation of structures will result.

It is our recommendation that persons interested in utilizing termiticides to protect wooden structures obtaining the data and information that is most relevant to their local conditions including soil types, pH,and organic content. We further recommend that professional pest control operators utilize the maximum label rates for termiticide applications, and that the labels be followed as carefully as possible. They should be conservative in their contracts with clients regarding offers of termite control over extended periods of time. We also recognize the need for very thorough inspections of treated structures on a regular basis to detect as early as possible any missed areas in the chemical barrier, or the degradation of termiticide below the minimum inhibitory concentration (MIC).

ACKNOWLEDGMENTS

We express our appreciation to representatives from AgroEvo, Bayer Corporation, FMC, DowElanco, and Zeneca Professional Products for their financial and other support of this research project. We also recognize the contributions of A.A. Collins and E.A. Jordan, both of whom were instrumental in the development of the extraction and analytical methods utilized in this work.

REFERENCES

Forschler, B.T. 1994. Survivorship and tunneling activity of Reticulitermes flavipes (Kollar) (Isoptera: Rhinotermitidae) in response to termiticide soil barriers with and without gaps of untreated soil. J. Entomol. Sci. 29(l): 4354.

Forschler, B.T. & M.L. Townsend. 1996. Mortality of Eastern subterranean termites (Isoptera: Rhinotermitidae) exposed to four soils treated with termiticides. J. Econ. Entomol. 89(3): 678-681.

Gold, R.E., H.N. Howell, Jr. & E.A. Jordan. 1993. Horizontal and vertical distribution of chlorpyrifos termiticide applied as liquid or foam emulsions, pp 140-155. In: Racke, K.D. & Leslie, A.R., edits. ACS Symposium Series 522. Pesticides in Urban Environments - Fate and Significance. Washington D.C.: American Chemical Society.

Gold, R.E., A. A. Collins, B.M. Pawson & H.N. Howell, Jr. 1994. Termiticide technology-theisofenphos dilemma. Technology: J.of The Franklin Institute 33 1 (A): 189-198.

Gold, R.E., H.N. Howell, Jr., B.M. Pawson, M.S. Wright & J.C. Lutz. 1996. Evaluation of termiticides residues and bioavailability from five soil types and locations in Texas, pp 467-484. In: Wildey, K.B. Edit. Proceedings of the 2ndInternational Conference on Insect Pests in the Urban Environment.

Grace, J.K. 1991. Response of Eastern and Formosan subterranean termites (Isoptera: Rhinotermitidae) to borate dust and soil treatments. J. Econ. Entomol. 84(6): 1753-1757.

Granovsky, T.A. 1979. Termites. Tierra Grande 8: 19-2 1.

Granovsky, T.A. 1983. Economic impact of termites. Pest Control 50(3): 14, 16, 20.

Granovsky, T.A. & P. A. Sadberry. 1983. The termite invasion. Tierra Grande 22: 12-15.

Howell, H.N., Jr., P.J. Hamman & T.A. Granovsky. 1987. The geographical distribution of the termite genera Reticulitemies, Coptotemies, and Incisitemies in Texas. Southwest. Entomol. 12(2): 119-125.

Jones, S.C. 1990. Effects of population density on tunneling by Formosan subterranean termite (Isoptera: Rhinotermitidae) through treated soil. J. of Econ. Entomol. 89(3): 875-878.

Kard, B.M., J.K. Mauldin & S.C. Jones. 1989. Evaluation of soil termiticides for control of subterranean termites (Isoptera). Sociobiology 15(3): 285297.

Kard, B.M. & C.A. McDaniel. 1993. Field evaluation of the persistence and efficacy of pesticides used for termite control. Racke, K.D. & Leslie, A.R., eds, ASC Symposium Series 522.'Pesticides in urban environments-fate and significance. Washington D.C., American Chemical Society.

Kard, B.M. & J.K. Mauldin. 1994. Gulfport: The 1994 update on termiticide efficacy. Pest Control 62(5): 52, 56, 58.

McIlveen, G, Jr., P.J. Hamman & R.E. Gold. 1993. A profile of Texas structural pest control industry. Urban Entomology Survey Report No. 1, Department of Entomology, Texas Agricultural Extension Service, Texas A&M University.

Mix, J. 1991. Dursban TC tops termiticide survey. Pest Control 59(2):22-23.

Mix, J. 1995. Termiticide soil sampling may be in your future. Pest Control 63(,2): 40-41.

Potter, M.F. 1994. Still fighting retreats. Pest Control 62(2): 45-48.

SAS Institute. 1987. SAS/STAT guide for personal computers, version 6 ed. Cary, NC: SAS Institute.

Smith, J.L. & M.K. Rust. 1990. Tunneling response and mortality of the western subterranean termite (Isoptera: Rhinotermitidae) to soil treated with termiticides. J. Econ. Entomol. 83(4): 1395-1401.

Smith, J.L. & M.K. Rust. 1991. Vapor activity of insecticides used for subterranean termite (Isoptera: Rhinotermitidae) control. J. of Econ. Entomol. 84(l): 181-184

Smith, J.L. & M.K. Rust. 1992. Activity and water-induced movement of termiticides in soil. J. Econ. Entomol. 85(2):430-434.

Stanley, R. 1994. Stringent testing as Gulfport yields a few good termiticides. Pest Control 62(2):54-57.

Su, N-Y. & R.H. Scheffrahn. 1990. Comparison of eleven soil termiticides against the Formosan subterranean termite and Eastern subterranean termite (Isoptera: Rhinotermitidae). J. Econ. Eritomol. 83(5):1918-1924.

Su, N-Y., R. H. Scheffrahn & R.H. Ban. 1993. Barrier efficacy of pyrethroid and organophosphate formulations against subterranean termites (Isoptera: Rhinotermitidae). J. Econ. Entomol. 86(3):772-776.

Su, N-Y., G.S. Wheeler & R.H. Scheffrahn. 1995. Subterranean termite (Isoptera: Rhinotermitidae) penetration into sand treated at various thickness with termiticides. J. Econ. Entomol. 88(6):1690-1694.

Table 3. Concentrations of termiticides in parts per million (mg/g) and percent of initial concentration remaining at the end of the specified post-treatment period and treatment conditions at College Station, Texas.

a) Uncovered = plot not covered with concrete slab
b) Covered = plot covered with concrete slab
c) Limit of detection = 0.10 micrograms/gram

Table 4. Concentrations of termiticides in parts per million (mg/g) and percent of initial concentration remaining at the end of the specified post-treatment period and treatment conditions at Corpus Christi, Texas.

a) Uncovered = plot not covered with concrete slab
b) Covered = plot covered with concrete slab
c) Limit of detection = 0.10 microgramstgrarn

Table 5. Concentrations of termiticides in parts per million (mg/g) and percent of initial concentration remaining at the end of the specified post-treatment period and treatment conditions at Dallas, Texas.