Orius insidiosus (Say) is an important predator of several economic pests in cotton. Laboratory-reared males, females and third instar nymphs were exposed to residues of nine insecticides applied to cotton plants in the field, in potted plants in the greenhouse and glass Petri dishes in the laboratory. Insects were exposed for 24-hours and then removed to determine mortality. Insecticides tested were spinosad, indoxacarb, imidacloprid, tebufenozide, methoxyfenozide, abamectin, emamectin benzoate, fipronil and λ-cyhalothrin. Differences were observed in mortality as measured by different methods. Spinosad, imidacloprid and indoxacarb induced significantly higher mortality with treated Petri dishes than on treated cotton plants in the field or greenhouse. No differences in mortality were observed between methods with fipronil or λ-cyhalothrin, and in only one instance with abamectin. Spinosad was not toxic in the field or greenhouse bioassays, but was highly toxic in the Petri dish bioassay. Imidacloprid was moderately toxic in the field and greenhouse, but was highly toxic in the Petri dish bioassay. Indoxacarb had variable toxicity, with low to moderate toxicity in the field and greenhouse, and high toxicity in the Petri dish bioassay. It is apparent that multiple testing methods should be used in evaluating the effects of pesticides on beneficial arthropods.
An increasing number of scientists are evaluating the toxicity of new pesticide chemistries on beneficial arthropods. Although a considerable number of studies have been published, it is sometimes difficult to compare results among researchers. The variety of methods used in bioassays is as varied as the number of individuals conducting the work. Scientists have used direct topical applications of the pesticide to the insect (Yu 1988; De Cock et al. 1996; Trisyono et al. 2000) or injected the insect with insecticide (Yu 1988), or fed treated prey (De Cock et al. 1996; Trisyono et al. 2000; Elzen 2001). These methods insure that a specific insecticide dose makes contact with the insect, either topically or internally, and will give an accurate indication of the actual toxicity of the pesticide to the insect in question. It is likely that if the insect species survives the topical or injection application, it will also survive any exposure in the field. However, the reverse may not always be true. High mortality resulting from a topical or injection bioassay may not be related to mortality observed under field conditions.
Researchers commonly use an inert substrate such as glass vials, Petri dishes or slides to test the toxicity of various insecticide residues to predatory or parasitic insects (Plapp & Bull 1978; Mizell & Schiffhauer 1990; Bayoun et al. 1995; Oetting & Latimer 1995; De Cock et al. 1996) or may use treated plastic cups (Mizell & Schiffhauer 1990). There are several possible errors that could occur in using treated substances such as glass or plastic to evaluate the toxicity of any insecticide to an insect. While data from these bioassays will give an indication of the toxicity of a pesticide to an insect, relating this toxicity to that which may be encountered in the field is difficult. The activity of a pesticide may be affected by the substrate upon which it is deposited (Cogburn 1972; White 1982). Jain and Yadav (1989) found that some insecticides persisted much longer when applied to a plastic substrate as compared with glass or painted wood.
Potentially more realistic testing methods use treated excised leaves (Samsoe-Petersen 1985; Oetting & Latimer 1995; Jones et al. 1997; Elzen & Elzen 1999), or treated potted plants grown in the greenhouse (Brown & Shanks 1976; Pietrantonio & Benedict 1999). These methods should provide a more realistic picture of actual toxicity from contact with residues on a natural substrate. Environmental effects (e.g., solar radiation or insect movement within the plant canopy) which may effect actual toxicity are not addressed.
Numerous researchers have evaluated insecticide effects on beneficial arthropods by making evaluations in the field from treated field plots (Brown & Shanks 1976; Stoltz & Stern 1978; Young et al. 1997; Simmons & Jackson 2000). Generally, the results from field studies express toxicity as the resulting presence or absence of the insect from a treated plot in comparison with an untreated plot or with pretreatment counts. These data are often taken within a few days to a week after treatment, depending on the researcher and the experimental design. In many instances the studies were designed to evaluate mortality induced in the target pest, with beneficial arthropod counts being made as a secondary goal to the study. The test plots in this latter case are not designed to accurately evaluate the induced mortality in the beneficial arthropods naturally present in the study.
In evaluating the effects of pesticides on any insect, the method used may have an effect on the final results. Confounding factors include solar radiation, rainfall, substrate treated, temperature, etc. Under field conditions, the effectiveness of a properly-applied insecticide may be diminished by high temperatures, sunlight and rainfall events. Similarly, the same tests may underestimate mortality caused by those insecticides that are systemic in the plant tissue, particularly on plant feeding insects. Therefore, it would be important to compare these effects as measured through various methodologies.
Materials and Methods
Orius insidiosus (Say) were collected from host plants (crimson clover, vetch and corn) early in the season of each year and used to start a lab colony maintained on green bean pods and Helicoverpa zea (Boddie) eggs. H. zea pupae were obtained from a colony maintained at the University of Arkansas Agricultural Research and Extension Center, Fayetteville, AR. Once adult moths emerged, they were placed in aquariums covered with a layer of cheesecloth onto which the females could oviposit. Wild adult moths were also collected and added to the colony during the growing season when they were abundant. O. insidiosus were reared at a photoperiod of 14:10 (L:D) at 25°C in an illuminated incubator (Precision Scientific® model 818, Winchester, VA). Green bean pods were not only a source of food and moisture, but also served as a substrate into which females would readily oviposit. Green beans and H. zea eggs were replaced daily. Pods with O. insidiosus eggs were placed into separate containers to allow nymphs to hatch. Fresh bean pods and H. zea eggs were provided to nymphs as well.
Field Plots
Plots of SureGrow 125 cotton were planted at the University of Arkansas Northeast Research and Extension Center, Keiser, AR during the growing seasons of 2000 and 2001. Fertility and weed control recommendations outlined by the University of Arkansas Cooperative Extension Service were followed (Baldwin et al. 2001). No insecticides were applied to plots with the exception of the insecticide treatments outlined in this study (Table 1). Also, no in-furrow insecticides were applied at planting to insure insecticide-free plants. Plots were 4 rows by 7.6-m long arranged in a RCB design with 4 replications. Insecticides were applied using a CO2 powered backpack sprayer. The sprayer was calibrated to deliver 10 gpa at a pressure of 40 psi through 2-TX8 hollowcone nozzles per row. Water alone was applied to the untreated control plots. Only the center 2 rows of each plot were treated to give a buffer of 2 rows between each pair of treated rows. Treatments were applied early in the morning, just after sunrise, when wind conditions were negligible to avoid spray drift. The spray boom was cleaned between each treatment by rinsing with a water and bleach solution, followed by water.
O. insidiosus were caged on plants as soon as sprays had dried (approximately 1-h after application). Cages were placed on the fourth leaf down from the plant’s terminal. Insects were caged on the plants for 24 h and then removed to evaluate mortality. Cages were constructed from 11.5 cm hair clips that were bent to fit around 6 cm diameter polystyrene Petri dishes. Each cage was constructed of either 2 Petri dish bases or 2 Petri dish tops so that the edges would meet forming an enclosure. Strips of foam were glued to the edges of each dish so that a seal would form when the cage was closed. A hole 3.2-cm in diameter was cut in each side of the cage and a piece of organdy cloth was glued over the opening to allow for air flow through the cage. Males, females and third instar nymphs were evaluated separately to determine the effects on gender and insect stage. Data were arcsine transformed and means from all bioassays were subjected to analysis of variance and separated by least significant difference test (LSD, P ≤ 0.05). Detransformed means are reported.
Greenhouse
SureGrow 125 cotton was grown in pots in the greenhouse at the University of Arkansas Northeast Research and Extension Center, Keiser, AR. Potted plants were treated in a DeVries model SB8 spray chamber. The chamber was calibrated to deliver 11.5 gpa through a single TX8 hollowcone nozzle. Potted plants were treated individually with insecticide and then placed back into the greenhouse. The spray chamber nozzle was cleaned between each treatment by rinsing with a water and bleach solution, followed by pure water. O. insidiosus were caged on plants as soon as sprays had dried (approximately 1 h after application). Insects were caged on the plants for 24 h and then removed to evaluate mortality. Cages were the same as those used in the field study (20 per replicate). Males, females and third instar nymphs were evaluated separately to determine the effects on gender and insect stage. Data were arcsine transformed and means from all bioassays were subjected to analysis of variance and separated by least significant difference test (LSD, P ≤ 0.05). Detransformed means are reported.
Laboratory
Glass Petri dishes 6-cm in diameter were treated with the insecticides listed in Table 1. Dishes were treated in the same spray chamber as the potted plants at the same rate (20 dishes per replicate, 4 replications). Individual O. insidiosus were placed in each dish as soon as sprays had dried (approximately 1-h after application), which was then covered with a piece of parafilm to keep insects from escaping. Mortality was checked after 24 h. Data were arcsine transformed and means from all bioassays were subjected to analysis of variance and separated by least significant difference test (LSD, P ≤ 0.05). Detransformed means are reported.
Results and Discussion
Abamectin, emamectin benzoate, fipronil and λ-cyhalothrin were consistently the most toxic of the tested insecticides to O. insidiosus as measured by all three methods during 2000 (Table 1, Table 2, Table 3) and 2001 (Table 4, Table 5, Table 6). Mortality from λ-cyhalothrin ranged from 95% to 100%, fipronil 77% to 100%, emamectin benzoate 61% to 100% and abamectin 56.3% to 100%. No differences in mortality were observed for any of the three methods with fipronil or λ-cyhalothrin.
In all instances, mortality induced by abamectin and emamectin benzoate was significantly higher than that in the untreated control. In three instances, the mortality measured after treatment with these two products using the Petri dish bioassay was significantly lower than that in the field and greenhouse bioassays. In two instances mortality measured in the Petri dish bioassay was significantly higher than that in the field bioassay. In all other instances mortality was not significantly different among methods with these two pesticides.
Mortality induced by tebufenozide and methoxyfenozide was not significantly different from that of the untreated control when measured by field or Petri dish bioassays. However, in a few instances, mortality was significantly higher with these two insecticides when measured by the greenhouse bioassay.
Differences in mortality measured between methods was most consistent with spinosad, imidacloprid and indoxacarb. In every instance, mortality was much higher in the Petri dish bioassay compared with both the field and greenhouse bioassays. This was most pronounced with spinosad. While no significant mortality was observed in the field and greenhouse bioassays, mortality was very high in the Petri dish bioassay with spinosad. Mortality was also quite high in the Petri dish bioassay with imidacloprid and indoxacarb, but the difference was not as great because mortality was approximately 50% in the field and greenhouse bioassays.
Overall, there were few differences in mortality as measured by the field and greenhouse bioassays. The majority of significant differences were between these two plant bioassays and the Petri dish bioassay. The field and greenhouse bioassays would be expected to be similar in the fact that both use the same substrate, treated cotton leaves. The only differences between the two would be environmental conditions (e.g., solar radiation, temperature, relative humidity).
Croft (1990) defines that mortality or sublethal effects of pesticides occur through three avenues: 1) direct contact with the insecticide, 2) residual uptake (contacting pesticide residues on another surface), and 3) food chain uptake (consuming prey or host plants containing the pesticide). In this study, obviously O. insidiosus could only take up pesticide through the residual uptake avenue in the Petri dish bioassay. However, because of this insect’s omnivorous habit, the possible uptake on treated plants used in the field and greenhouse bioassays could be through residual uptake and/or food chain uptake. With uptake possibly occurring through two avenues, one would expect mortality to be higher in the greenhouse and field bioassays as compared with the Petri dish bioassay. Although this was not true in the majority of cases in this study. Imidacloprid, indoxacarb and spinosad had much higher mortality in the Petri dish bioassay. Possibly, O. insidiosus did not receive a toxic dose in every instance in the treated plant bioassays (field and greenhouse) by not feeding on the treated plant during the 24-h exposure time used in this study. Another explanation offered is that the plant surface somehow altered or bound the pesticide deposits making them less available for uptake by the test insects. Imidacloprid and indoxacarb are known to have good translaminar movement into the leaf and would therefore move the pesticide away from direct contact to the insect in the plant bioassays. Because glass is an inert substance, it is not likely that the pesticide deposits would be altered or somehow bound to the substrate, leaving them free for uptake by an insect. Also, the entire inside surface of the dish was treated, making the parafilm cover the only area in which the insects could avoid the pesticide. In this study, test insects were observed on the inside of the parafilm cover only occasionally. In both the field and greenhouse bioassays, the clip cages offered a greater surface area on which the insects could avoid the pesticide. Even if the insect was not attempting to avoid the pesticide deposits, the chances of picking up a lethal dose would have been greater in the Petri dish. The most interesting results from this study were with spinosad. In both the field and greenhouse bioassays, mortality was not significantly different from that found in the untreated control, indicating that this pesticide is not toxic to O. insidiosus. However, mortality was very high in the Petri dish bioassay with this pesticide (100% in some instances). This leads one to think that the plant surface somehow makes this compound unavailable to this insect. Although spinosad is reported to have some translaminar movement into the leaf (Bret et al. 1997), this does not adequately explain the low toxicity in the plant bioassays.
Obviously, experimental design can have a pronounced effect on the outcome of a study and may offer some explanation on the disparity of results sometimes observed in the literature. In this study, particularly with spinosad, one would come to the conclusion that this pesticide would not be a good fit in a cotton IPM program with O. insidiosus when looking at the Petri dish bioassay alone. However, no effects were observed in the caged field and greenhouse studies, leading one to the opposite conclusion. It is apparent that merely evaluating mortality of pesticides on beneficial arthropods by only one method does not give an accurate depiction on how those pesticides would fit into IPM programs. This study concurs with Banken and Stark (1998) and Croft (1990) in that multiple testing methods should be used in evaluating pesticide effects on beneficial arthropods. However, this may not always be feasible. Often, lab studies utilizing an artificial substrate, are the quickest and least expensive means of obtaining data. However, particularly when working with omnivorous predators such as O. insidiosus, utilizing field studies or potted plants grown in the greenhouse would be the preferred method for bioassays. Bioassays utilizing artificial substrates, while providing important information, should not be the sole means of evaluating the effects of pesticides on beneficial arthropods.
References Cited
Table 1.
Percent mortality in O. insidiosus males measured by three methods in 2000.
Table 2.
Percent mortality in O. insidiosus females measured by three methods in 2000.
Table 3.
Percent mortality in O. insidiosus nymphs measured by three methods in 2000.
Table 4.
Percent mortality in O. insidiosus males measured by three methods in 2001.
Table 5.
Percent mortality in O. insidiosus females measured by three methods in 2001.
Table 6.
Percent mortality in O. insidiosus nymphs measured by three methods in 2001.
