Effects of boric acid,fipronil, hydramethylnon,and diflubenzuron baits on colonies of ghost ants (Hymenoptera: Formicidae)

Laboratory colonies of the ghost ant, Tapinoma melanocephalum (F.) were administered sugar solution (10%) baits containing the insecticides boric acid, fipronil (REGENT), hydramethylnon (SIEGE), or diflubenzuron (DIMILN). Colonies were exposed to the baits for 21 d, and development of workers, queens, and brood (larvae and pupae) was observed for 4 wk. Fipronil (0.05%) caused 100% mortality in all colonies the first week. With boric acid (0.5%), 100% mortality of workers, queens, and brood was reached at the end of the third week. With hydramethylnon (2%), 83% of the colonies disappeared at the end of the fourth week, but some queens were still alive 9 wk after the trial started. Diflubenzuron (1%) behaved similarly to the control, although in some colonies, the brood production increased, whereas in other colonies, the queens disappeared. In the control colonies, workers, queens, and brood were always observed even up to 9 wk.     

Delayed toxicity as a critical factor in the efficacy of aqueous baits for controlling Argentine ants (Hymenoptera: Formicidae)

Boric acid, imidacloprid, and thiamethoxam in sucrose aqueous baits had different delayed toxicities to worker Argentine ants, Linepithema humile (Mayr). The concentrations required to produce an LT50 (time required to produce 50% mortality) within 1-4 d were 3.63-0.55% boric acid, 9.2 x 10(-3) to 7.1 x 10(-4)% imidacloprid, and 3 x 10(-4) to 2 x 10(-5)% thiamethoxam. The three toxicants were not repellent. Other laboratory trials showed that 1% boric acid, 5 x 10(-4) to 5 x 10(-3)% imidacloprid, and 1 x 10(-5) to 1 x 10(-3)% thiamethoxam had delayed toxic effects, whereas 0.5% boric acid and < 5 x 10(-3)% imidacloprid did not. Baits that provided an LT50 between days 1 and 4 were considered to have delayed toxic effects. The utility of aqueous sucrose baits and toxicants soluble in such systems and the negative impact of fast-acting toxicants on trail following, recruitment, trophallaxis, and control of Argentine ants are discussed.     

Influence of toxic bait type and starvation on worker and queen mortality in laboratory colonies of Argentine ant (Hymenoptera: Formicidae)

The efficacy of toxic baits should be judged by their ability to kill entire ant colonies, including the colony queen or queens. We studied the efficacy of four toxic baits to the Argentine ant, Linepithema humile (Mayr) (Hymenoptera: Formicidae). These baits were Xstinguish that has the toxicant fipronil, Exterm-an-Ant that contains both boric acid and sodium borate, and Advion ant gel and Advion ant bait arena that both have indoxacarb. Experimental nests contained 300 workers and 10 queen ants that were starved for either 24 or 48 h before toxic bait exposure. The efficacy of the toxic baits was strongly influenced by starvation. In no treatment with 24-h starvation did we observe 100% worker death. After 24-h starvation three of the baits did not result in any queen deaths, with only Exterm-an-Ant producing an average of 25% mortality. In contrast, 100% queen and worker mortality was observed in colonies starved for 48 h and given Xstinguish or Exterm-an-Ant. The baits Advion ant gel and Advion ant bait arena were not effective against Argentine ants in these trials, resulting in <60% mortality in all treatments. Because of the strong influence of starvation on bait uptake, control efficacy may be maximized by applying bait when ants are likely to be starved. Our results suggest queen mortality must be assessed in tests for toxic bait efficacy. Our data indicate that of these four baits, Xstinguish and Exterm-an-Ant are the best options for control of Argentine ants in New Zealand.     

Toxicity and repellency of borate-sucrose water baits to Argentine ants (Hymenoptera: Formicidae)

The oral toxicity of boron compounds to the Argentine ant, Linepithema humile (Mayr), was evaluated in laboratory tests. The ants were provided 25% sucrose water containing 0.5 and 1% boric acid, disodium octaborate tetrahydrate, and borax. Lethal times of these solutions were a function of the concentration of boron. In field tests, the ants showed no discrimination between disodium octaborate tetrahydrate and boric acid. There was a significant reduction in consumption of sucrose water with > 1% boric acid.     

Determining the foraging range and origin of resurgence after treatment of Argentine ant (Hymenoptera: Formicidae) in urban areas

The foraging range and distribution of Argentine ant, Linepithema humile (Mayr), colonies in urban areas of southern California extended at least 61 m (200 feet) from feeding stations and structures. Ants were fed at 25% sucrose feeding stations containing 0.01% fluorescent brighter (FB28). Within 14 d, from 77-90% of the ants sampled next to the feeding stations were positive for FB28. The percentage of ants with FB28 declined gradually to approximately 55% 61 m away from the feeding station. The percentage of marked ants in the controls didn't change over the 4-wk-test period. There were approximately 290,000 ants visiting the monitored stations each night before treatments. The 0.0001% fipronil baits and 0.06% fipronil SC sprays provided significant reductions in at least 4 wk. The percentage of ants marked with FB28 decreased significantly in both bait and spray treatments over 4 wk suggesting that the resurgence of ants in treated areas were because of immigration from untreated areas. It is likely that much larger areas will need to be treated to control Argentine ants in urban settings, especially if baits are being used.     

Aggressive interactions between Solenopsis invicta and Linepithema humile (Hymenoptera: Formicidae) under laboratory conditions

The Argentine ant, Linepithema humile (Mayr), and the red imported fire ant, Solenopsis invicta Buren, are natural agonists in their country of origin. Since the first report of L. humile in California in 1907 its range expanded statewide, displacing native ant species wherever it spread. Since the discovery of established populations of S. invicta in southern California in 1998, it has been restricted to discrete areas of southern California. However, as these discrete populations expand, they are encountering large populations of L. humile, which are effective competitors for available resources and are particularly aggressive in their encounters with other ant species such as S. invicta. Most Dolichoderine ants such as L. humile do not prefer to forage on baits made with defatted corn grit and soybean oil typically used in red imported fire ant control programs. Applications of these baits in areas where distributions of these species overlap might selectively affect populations of S. invicta and give L. humile a competitive advantage. Three laboratory experiments were conducted to determine the competitive outcomes between S. invicta pitted against L. humile: 1) agonistic behavior of workers in small arenas, 2) colony interactions with different population ratios, and 3) the effects of pyriproxyfen on the competitiveness of S. invicta against L. humile. Populations of S. invicta consisting of major workers killed more L. humile than did minors or a mixture of majors and minors. When paired against L. humile colonies consisting of 1,100 workers, colonies consisting of 38 S. invicta workers were easily defeated by L. humile. Colonies consisting of 450 S. invicta workers plugged their nest entrances, but they were ultimately defeated by L. humile after 13 d. The S. invicta colonies consisting of 1,100 workers took control of the bridge connecting the colonies, invaded the L. humile colony, killed the Argentine ant queens, and removed their brood. Pyriproxyfen-treated fire ant workers took significantly longer to chase the Argentine ants from a connecting bridge than did untreated fire ants. Thus, fire ant baits may have long-term effects on intercolonial aggression between S. invicta and L. humile, especially when Argentine ant populations are high in the summer.     

Transfer of ingested insecticides among cockroaches: effects of active ingredient, bait formulation, and assay procedures.

Foraging cockroaches ingest insecticide baits, translocate them, and can cause mortality in untreated cockroaches that contact the foragers or ingest their excretions. Translocation of eight ingested baits by adult male Blattella germanica (L.) was examined in relation to the type of the active ingredient, formulation, and foraging area. Ingested boric acid, chlorpyrifos, fipronil, and hydramethylnon that were excreted by adults in small dishes killed 100% of first instars within 10 d and >50% of second instars within 14 d. Residues from these ingested baits were also highly effective on nymphs in larger arenas and killed 16-100% of the adults. However, when the baits and dead cockroaches were removed from the large arenas and replaced with new cockroaches, only residues of the slow-acting hydramethylnon killed most of the nymphs and adults, whereas residues of fast acting insecticides (chlorpyrifos and fipronil) killed fewer nymphs and adults. Excretions from cockroaches that ingested abamectin baits failed to cause significant mortality in cockroaches that contacted the residues. These results suggest that hydramethylnon is highly effective in these assays because cockroaches that feed on the bait have ample time to return to their shelter and defecate insecticide-laden feces. The relatively high concentration of hydramethylnon in the bait (2.15%) and its apparent stability in the digestive tract and feces probably contribute to the efficacy of hydramethylnon. To control for differences among baits in inert ingredients and the amount of active ingredient, we compared 1% chlorpyrifos with 1% hydramethylnon in identical baits. Again, hydramethylnon residues provided greater secondary kill, but the results highlighted the importance of the inert ingredients. We conclude that, in the absence of cannibalism and necrophagy, translocation of baits and secondary kill are most effective with slow acting insecticides in palatable baits that can traverse the digestive tract and be deposited within and around the cockroach aggregation.     

Efficacy of commercial baits and new active ingredients against firebrats and silverfish (Zygentoma: Lepismatidae)

Commercial baits containing boric acid, hydramethynon, and indoxacarb were tested against the firebrat, Thermobia domestica (Packard), and silverfish, Lepisma saccharina L. (both Zygentoma: Lepismatidae), under laboratory conditions. Three boric acid baits were consumed in significantly smaller amounts than untreated control food and did not effectively control either species. Baits containing hydramethylnon and indoxacarb were consumed in greater amounts than boric acid baits, but were relatively ineffective, with LT50 values >9 d. Presence or absence of competitive untreated food did not consistently affect bait efficacy. A ground oat matrix was used to evaluate the potential effectiveness of abamectin, chlorfenapyr, dinotefuran, fipronil, hydramethylnon, metaflumizone, and novaluron baits. The most effective compound was chlorfenapyr. At 0.05 and 0.20% (wt:wt) rates, chlorfenapyr baits produced LT50 values, for both species, ranging from 2 to 4 d. All other compounds had LT50 values >7 d.     

Preference of food particle size among several urban ant species

Appropriate particle size may be a critical characteristic for effective granular ant baits. We examined the particle size preference of six species of pest ants to an anchovy-based bait. We also examined head capsule widths of Argentine ants, Linepithema humile (Mayr) (mean = 0.54 mm), California harvester ants, Pogonomyrmex californicus (Buckley) (mean = 1.63 mm), red imported fire ants, Solenopsis invicta Buren (mean = 0.9 mm), and southern fire ants, Solenopsis xyloni McCook (mean = 0.76 mm) and compared them with the first and second most preferred particle size. There were differences between particle size of which the most mass was removed and of which there were more particles removed by ants. California Argentine ants, southern fire ants, and Alabama Argentine ants removed more 840 to 1,000-microm particle mass of the anchovy diet but had more visits to dishes containing 420 to 590 microm particles. California harvester ants and Allegheny mound ants, Formica spp., removed more >2,000 microm particle mass but visited dishes containing 1,000 to 2,000 microm particles more often. Red imported fire ants also removed more >2,000 microm particle mass but visited dishes with 590 to 840-microm particles most often. Pharaoh ants, Monomorium pharaonis (L.), removed and visited 420 to 590-microm particles more than any other size. A linear regression model determined that particle size preferred by each ant species relates to forager head width. The majority of particles of commercial ant bait, including Amdro, Ascend, Award, Bushwhacker, Max Force with fipronil, and old and new formulations of Max Force with hydramethylnon, were 1,000 to 2,000 microm, but the majority of Niban particles were <420 microm. Altering the size of particles of toxic ant baits to fit the particle size preference of each pest ant species may increase the efficacy of ant baits.     

Testing baits to control Argentine ants (Hymenoptera: Formicidae) in vineyards

Liquid baits were evaluated for control of the Argentine ant, Linepithema humile (Mayr) (Hymenoptera: Formicidae), and associated mealybug and soft scale pests in California vineyards. In 2003, liquid baits with small doses ofimidacloprid, boric acid, or thiamethoxam dissolved in 25% sucrose water resulted in lower ant and mealybug densities and fruit damage, compared with an untreated control. Similar treatments in a soft scale-infested vineyard showed only a reduction of ant density and fruit infestation in only the boric acid and thiamethoxam treatments. In 2004, commercial and noncommercial formulations of liquid baits reduced ant densities in three separate trials, but they had inconsistent effects on mealybug densities and fruit infestation; granular protein bait had no effect. Using large plots and commercial application methodologies, liquid bait deployed in June resulted in lower ant density and fruit infestation, but it had no effect on mealybug density. Across all trials, liquid bait treatments resulted in lower ant density (12 of 14 trials) and fruit damage (11 of 14 sites), presenting the first report of liquid baits applied using commercial methodologies that resulted in a reduction of ants and their associated hemipteran crop damage. For commercialization of liquid baits, we showed that any of the tested insecticides can suppress Argentine ants when properly delivered in the crop system. For imidacloprid, bait dispensers must be protected from sunlight to reduce photodegradation. Results suggest that incomplete ant suppression can suppress mealybug densities. However, after ant populations are suppressed, there may be a longer period before hemipteran populations are effectively suppressed. Therefore, liquid baits should be considered part of a multiseason program rather than a direct, in-season control of hemipteran pest populations.     

Acceptance and intake of gel and liquid sucrose compositions by the Argentine ant (Hymenoptera: Formicidae)

Liquids and gels are common delivery forms used in commercial ant baits, but the relative effectiveness of each is unknown. We compared the feeding responses of the Argentine ant, Linepithema humile (Mayr), to liquid and gel compositions of sucrose. In choice assays, more workers were counted on gel than liquid; however, substantially more liquid was consumed. Because workers could stand on the gel, more workers could feed simultaneously on the gel. The feeding bouts of individual workers, however, were much less efficient at extracting sucrose in gel form. Workers fed eightfold longer on the gel, yet removed fivefold less sucrose than workers feeding on liquid. This potential bias should be considered during attraction and palatability studies that use physically different bait compositions. When the toxicant fipronil was added to the compositions, a greater proportion of the colony died after workers had fed on liquid than gel baits. This finding suggests that liquid formulations may provide more effective control of Argentine ants due to the greater speed and abundance in which it is ingested.     

Strategies for controlling the ghost ant,Tapinoma melanocephalum (Hymenoptera: Formicidae) with liquid bait

The ghost ant, Tapinoma melanocephalum (F.), is a household pest and a considerable nuisance. The aim of this study is to demonstrate the toxicity and control efficacy of boric acid in liquid bait against queen and worker ghost ants. The LT₅₀ values for workers fed with 0.5%–2.5% boric acid and 2% chicken peptone in 20% sucrose water solutions were 4.3–2.4 days. The lethal times (LT₅₀ = 5.2–7.6 days) for queen ghost ants fed with various concentrations of a boric acid solution depended on the feeding behavior of the queens. The high boric acid (4%) content solutions were not repellent to the ghost ant workers. The liquid bait formulation of 1% boric acid, which caused a 100% worker, brood, and queen population reduction in 4, 4, and 5 weeks, respectively, was significantly more effective than the solid bait formulation containing the same concentration (p ≦ 0.05). The simulation tests involved using chicken peptone and sucrose as the attractant, and colonies were provided an alternate food source (20% sucrose solution and dog food) to achieve a more accurate assessment of bait acceptability in screening for the efficacy of the liquid boric acid bait. The control efficiency attained 99.9% in week 4. The results demonstrated that liquid bait, containing 2% chicken peptone, 20% sucrose as a food attractant, and 1% boric acid as the toxin, is efficient and highly recommended for ghost ant control.     

Control of Vespula germanica (Hym. Vespidae) populations using toxic baits: bait attractiveness and pesticide efficacy

Abstract:  Our aim was to test a set of baits and pesticides for the control of yellowjacket populations in NW Patagonia (Argentina), through field trials. We tested the attractiveness of protein (fresh and freeze-dried beef) and carbohydrate (corn syrup and honey) baits (alone and mixed) and three pesticides commonly employed to control terrestrial domestic arthropods (hydramethylnon 2%, permethrin 0.3% and chlorpyrifos 0.25%). Our results show that beef proved to be the most attractive bait tested, for Vespula germanica wasps in NW Patagonia. Honey and corn syrup alone or mixed with beef did not attract foraging wasps as did beef-only baits throughout the wasp season. Additionally, we show that the attractiveness of lyophilized beef baits is similar to that of fresh beef. The efficacy of the insecticides tested was limited. In fact, only baiting with hydramethylnon 2% reduced wasp populations (54%) and this occurred after 72 h since poisoning. However, given the potential of insects to develop resistance to a consistent use of a single product (e.g. fipronil 0.1%), the use of hydramethylnon may allow the alternating of insecticides in specific situations. The information provided here contributes to the existent knowledge on baits and insecticides for the control of yellowjackets using toxic baits.     

Development of virtual bait stations to control Argentine ants (Hymenoptera: Formicidae) in environmentally sensitive habitats

A novel bait station referred to as a virtual bait station was developed and tested against field populations of the invasive Argentine ant, Linepithema humile (Mayr) (Hymenoptera: Formicidae), at White Beach, Camp Pendleton, in Oceanside, CA. White Beach is a nesting habitat for an endangered seabird, the California least tern (Sterna antillarum browni Mearns). The beach is heavily infested with Argentine ants, one of the threats for the California least tern chicks. Conventional pest control strategies are prohibited because of the existence of the protected bird species and the site's proximity to the ocean. The bait station consisted of a polyvinyl chloride pipe that was treated on the inside with fipronil insecticide at low concentrations to obtain delayed toxicity against ants. The pipe was provisioned with an inverted bottle of 25% sucrose solution, then capped, and buried in the sand. Foraging ants crossed the treated surface to consume the sucrose solution. The delayed toxicity of fipronil deposits allowed the ants to continue foraging on the sucrose solution and to interact with their nestmates, killing them within 3-5 d after exposure. Further modification of the bait station design minimized the accumulation of dead ants in the sucrose solution, significantly improving the longevity and efficacy of the bait station. The virtual bait station exploits the foraging behavior of the ants and provides a low impact approach to control ants in environmentally sensitive habitats. It excluded all insects except ants, required only milligram quantities of toxicant, and eliminated the problem of formulating toxicants into aqueous sugar baits.     

Let them eat termites—prey‐baiting provides effective control of Argentine ants,Linepithema humile,in a biodiversity hotspot

Invasive ants threaten biodiversity, ecosystem services and agricultural systems. This study evaluated a prey‐baiting approach for managing Argentine ants in natural habitat invaded by Argentine ants. Blackmound termites (Amitermes hastatus) were topically exposed to fipronil and presented to Argentine ants (Linepithema humile). In laboratory assays, L. humile colonies were offered fipronil‐treated termites within experimental arenas. The termites were readily consumed, and results demonstrate that a single termite topically treated with 590 ng fipronil is capable of killing at least 500 L. humile workers in 4 days. Field studies were conducted in natural areas invaded by L. humile. Fipronil‐treated termites scattered within experimental plots provided rapid control of L. humile and ant densities throughout the treated plots declined by 98 ± 5% within 21 days. Results demonstrate that the prey‐baiting approach is highly effective against L. humile and may offer an effective alternative to traditional bait treatments. Furthermore, prey‐baiting offers environmental benefits by delivering substantially less toxicant to the environment relative to current control methods which rely on commercial bait formulations and may offer greater target specificity.     

Evaluations on the Lethal, Choice and Secondary Effects of Four Insecticidal Baits against the German Cockroach (Blattaria, Blattellidae)

ABSTRACT The laboratory studies were carried out for evaluating control effects of four commercial insecticidal baits such as two different hydramethylnon [2.0%(AI)] products (DBK^® and Combat‐Gold^®), fipronil [0.05 %(AI)] (Combat‐Power^®) and (0.6% chlorpyrifos [0.6%(AI)] (Raid‐Roachbait^®) against German cockroaches (Blattella germanica). The control rates of four kinds of toxic baits were all 100.0% mortality of German cockroaches in 5 days after treatment. The results of chlorpyrifos and fipronil brought 100.0% mortalities in 2 to 3 days after treatment, respectively. After 3 day treatment, there was no significant difference of control effect among the four toxic baits. As a result of this study, fipnonil and chlorpyrifos showed faster killing action against German cockroaches than the two hydramethylnon formulation products. In the choice test, DBK^® (hydramethylnon) (average 17.0 ind.) significantly attracted more German cockroaches than Combat‐Gold^® (hydramethylnon) (avg. 7.0 ind.), Combat‐Power^® (fipronil) (avg. 5.3 ind.) and Raid‐Roachbait^® (chlorpyrifos) (avg. 3.3 ind.). The difference in attraction effects came 10 minutes after treatment. In order to evaluate the secondary killing effect of toxic baits by coprophagy against adult Blattella germanica, the comparative test was carried out. The adult mortality rates were significantly different among the four toxic baits showing 86.7% mortality for fipronil, 60.0% for hydramethylnon (DBK^®), 30.0% for chlorpyrifos, and below 13.3% for hydramethylnon (Combat‐Gold^®) in 6 days. During the first 12 days, fipronil showed the highest mortality rate (90.0%), followed by hydramethylnon (DBK^®) (60.0%). The secondary killing effect of toxic baits by coprophagy appeared in all of the baits against adult Blattella germanica.     

Effects of interspecific competition between two urban ant species, Linepithema humile and Monomorium minimum, on toxic bait performance

We evaluated the effects of interspecific competition on ant bait performance with two urban pest ants, the Argentine ant, Linepithema humile (Mayr), and the little black ant, Monomorium minimum (Buckley). In a laboratory study, the impact of a solid sulfluramid bait on M. minimum was diminished when L. humile were present, whereas the presence of M. minimum reduced the performance of a liquid fipronil bait against L. humile. Argentine ants were not adversely affected by sulfluramid bait at any time, whereas M. minimum was unaffected by fipronil bait until 14 d of exposure. In field studies measuring diel foraging activity, M. minimum seemed to delay L. humile foraging to food stations by approximately 30 min during summer 2001. However, L. humile subsequently recruited to food stations in very high numbers, thereby displacing M. minimum. L. humile visited food stations over an entire 24-h period, whereas M. minimum was only observed visiting food stations during daylight hours. Adjusting the timing of bait placement in the field may minimize any negative effects of interspecific competition between these two species on toxic bait performance.     

Secondary transmission of toxic baits in German cockroach (Dictyoptera: Blattellidae)

The death rates of three toxic gel baits (fipronil bait, 0.05%; hydramethylnon, bait 2.15%; and abamectin B1 bait, 0.05%) were estimated in Blattella germanica (L.). After ingestion, all three baits killed similar proportions of cockroaches, but they died more rapidly after ingestion of fipronil bait than after ingestion of abamectin B1 bait or of hydramethylnon bait. Laboratory experiments evaluated the impact of secondary transmission of fipronil bait and hydramethylnon bait. Mortality rates caused by secondary transmission were higher for fipronil bait than for hydramethylnon bait. Under controlled laboratory conditions, secondary transmission occurred mainly through direct contact with, or ingestion of, traces of baits dispersed in the environment by contaminated cockroaches. These traces were either deposited by "trampling" in the environment or on dead contaminated cockroaches. Social interactions and cannibalism played a minor role in secondary transmission of these two baits.     

Foraging activity and dietary spectrum of the Argentine ant (Hymenoptera: Formicidae) in invaded natural areas of the northeast Iberian Peninsula

We analyzed the foraging activity and the dietary spectrum of the Argentine ant (Linepithema humile Mayr) and select native ants on cork oaks from Mediterranean open cork oak (Quercus suber) secondary forests. The study areas included invaded and noninvaded zones in close proximity. The Argentine ant's daily foraging activity was correlated to the abiotic factors studied, whereas the seasonal foraging activity was related not only to the variations in the average air temperature, but also to the trophic needs of the colony. Argentine ant workers focused their attention on protein foods during the queens' oviposition periods and during the larvae development phase, and on carbohydrate foods, such as honeydew, when males and workers were hatching. There were no significant differences over the entire year in the quantity of liquid food collected by the Argentine ant workers in comparison with the native ants studied. The solid diet of the Argentine ant on cork oaks is composed of insects, most of which are aphids. Our results have clear applications for control methods based on toxic baits in the invaded natural ecosystems of the Iberian Peninsula.     

Laboratory and field performance of consumer bait products for German cockroach (Dictyoptera: Blattellidae) control.

Bait formulations containing boric acid, chlorpyrifos, or hydramethylnon were evaluated in field and laboratory experiments against the German cockroach, Blattella germanica (L.). In continuous exposure tests, chlorpyrifos formulations produced 100% mortality in less than 24 h, boric acid formulations had LT50's of approximately 5.5 d, and hydramethylnon had an LT50 of 0.86 d. LT50's estimated in Ebeling choice boxes were 1.6 to 10.8 times greater than those from continuous exposure tests for chlorpyrifos formulations and 2.5 times greater for the hydramethylnon formulation. The same LT50's were produced by choice box and continuous exposure tests for boric acid formulations. Relative repellency, measured as the mean percentage of live cockroaches in the light side of the choice box, was greatest for chlorpyrifos formulations, followed by boric acid formulations, then hydramethylnon. A performance index (PI) that combined choice box repellency and mortality data indicated that hydramethylnon (Combat) had the greatest potential for field effectiveness. Only hydramethylnon significantly reduced German cockroach populations in the field. Apartments treated with the IT WORKS formulation of boric acid bait had significantly greater cockroach trap catches. The PI is an accurate predictor of the field performance of toxic baits for German cockroach control.