Introduction

Careless use of pesticides may prevent the success of biological control because of their direct and indirect toxic effects on natural enemies (Jacobson 1993; Michaud and Grant 2003). Therefore, several tactics should be considered and implemented in practice in order to minimize the adverse effect of pesticides on beneficials. Such tactics may expand the role of biological control further. The choice of chemicals that are compatible with IPM is the goal, and requires data on their toxicity to beneficials. In this respect, toxicity studies on natural enemies provide valuable information.

A number of studies have suggested that Orius species may be highly effective for biological control of the western flower thrips (WFT), Frankliniella occidentalis (Pergande), in greenhouses (Jacobson 1993; Kıman and Yeargan 1985; Lykouressis and Perdikis 1997; Zhang and Shipp 1998). Various Orius species such as Orius majusculus (Reuter) are being produced commercially and used successfully in biological control of F. occidentalis in several countries. However, applications of pesticides that are toxic to O. majusculus may significantly compromise the effectiveness of biological control (Jacobson 1993).

The goals of the present study were to obtain data on topical and residual toxicity of some insecticides and fungicides to adults, nymphs and eggs of O. majusculus and to determine the toxicity of the insecticides on adults and nymphs, as well as oviposition behavior of females, in choice and no-choice tests. Topical, residue and systemic uptake methods were also compared to determine the differences in the toxicity levels of imidacloprid, a systemic insecticide.

The pesticides chosen for this study are used for many greenhouse pathogen and pest species in Turkey and other countries. In Turkey, abamectin and spinosad are recommended mainly for spider mites, thrips, leafminers and aphids in vegetables; imidacloprid and endosulfan are recommended mainly for Bemisia tabaci Genn., Helicoverpa armigera Hübner and some other lepidopterous pests in both vegetables and cotton. Benomyl and copper salts + mancozeb are recommended for Sclerotinia sclerotiorum Lib., Pseudoperonospora cubensis Berk and Curt, Fusarium spp., Pythium spp., Rhizoctonia spp. and some other fungal pathogens.

Materials and methods

Insect rearing

The O. majusculus populations used in this study were collected from Cucurbita spp. in Antalya in 2004. The stock colony was reared in 500 ml glass jars using Ephestia kuehniella Zell. eggs as food and bean pods (Phaseolus vulgaris L.) as egg-laying substrate. Insects were reared and maintained in a walk-in growth chamber at 26 ± 1°C and a photoperiod of 16 h:8 h (L:D).

Chemicals

Active ingredients, trade names and manufacturers of the chemical pesticides used in this study are: Abamectin (Agrimec EC, Syngenta, Turkey), endosulfan (Hektionex 36EC, Hektaş, Turkey), spinosad (Laser SC, Syngenta), imidacloprid (Confidor 350 SC, Bayer), Benomyl (Pilben 50WP, Hektaş) and copper salts + mancozeb 21% + 20% (Tri-miltox Forte WP, Syngenta). Chemicals were used at the recommended field doses in the bioassays. The test concentrations were 7.2, 720, 96 and 300 mg l −1 as active ingredient (a.i.) for abamectin, endosulfan, spinosad and benomyl, respectively. Imidacloprid was tested at two concentrations: 350 mg a.i. l −1 (recommended field dose) and 35 mg a.i. l −1 As the mixture pesticide, copper salts + mancozeb 21% + 20% was tested at the formulation dose of 3,000 mg l −1.

Topical and residue bioassays

Topical (direct spray) and residue bioassays were conducted to determine the toxicity of the pesticides to O. majusculus adults and nymphs. For the topical bioassay, adults and nymphs were collected from stock cultures using a mouth aspirator and anaesthetized with CO2. They were gently released onto bean leaf disks (30 mm diam) and pesticides (at the concentrations described earlier) or water (control) were sprayed using a Potter spray tower (Burkard, Uxbridge, UK), resulting in a deposit of 2.7 mg cm−2 at a pressure of 0.84 atm. The residue bioassay followed a similar protocol except the insects were placed on bean leaf disks 1, 3 or 5 days after the disks had been sprayed. This method was used for all chemicals except imidacloprid, which was tested using a different protocol (described below). Sandwich-type plexiglass cells were used to keep the test organisms on the treated disks; E. kuehniella eggs were added as food. There were four replicates per treatment and approximately ten insects per replicate. Test duration was 24 h; mortality was determined at the end of the test.

Chronic toxicity to adults and nymphs

Persistency tests were conducted in the same manner as the residue tests except the exposure period was extended. Adults and nymphs were placed on the disks 1 day after chemical application; the exposure period was 4 days. Each treatment had four replicates, with 10 insects per replicate.

Toxicity to eggs

Bean pod pieces (∼3 cm) were covered with parafilm except for a 2 cm2 area. The covered pod pieces were placed in 40 ml jars. O. majusculus females were added to the jars and allowed to lay their eggs onto the uncovered bean pod surfaces for 24 h. E. kuehniella eggs were supplied as food. Insecticides and water (control) were sprayed on the bean pods where O. majusculus eggs had been laid. The number of eggs hatched was recorded after 6 days. There were six replicates per treatment with >20 eggs per replicate.

Choice and no-choice tests for oviposition behavior

Bean pod pieces (∼3 cm length) were dipped in the respective insecticides (or water) at the recommended doses, and allowed to dry for approximately 2 h. For the choice tests, one insecticide-treated pod and one water-treated pod were placed in a 40 ml jar. For the no-choice tests, only an insecticide-treated or a water-treated pod (but not both) was placed in a jar. One 7–10-d-old female O. majusculus was released into each jar and E. kuehniella eggs were supplied as food. The number of eggs laid on the bean pods was recorded after 48 and 96 h. Each treatment consisted of ten replicates.

Bioassays for imidacloprid

Topical, leaf residue and systemic uptake methods were used to assess the toxicity of imidacloprid. Topical methods were as described above. For the leaf residue assay, whole bean plants were dipped into imidacloprid solutions (350 and 35 mg a.i l −1) for 5 s. Disks were prepared from detached leaves 1 day after treatment and insects were exposed to these disks. To evaluate the toxicity of imidacloprid taken up systemically, 10–15-d-old bean plants were removed from pots and the roots were cleared of soil. Root and stem parts of the plants were placed in imidacloprid solutions (350 and 35 mg l −1) for 24 h. Leaf disks were prepared from detached leaves and insects were exposed to the disks. Mortality was recorded after 48 h.

Data analysis

The data were analyzed using one-way General Linear Model ANOVA. Tukey’s multiple comparison test was used to differentiate among significantly different means (SAS 1998).

Results

Topical and residue toxicity

Percent mortality of O. majusculus adults exposed to abamectin, endosulfan and spinosad was 95%, 95% and 100%, respectively; nymph mortality was 57%, 100% and 88%, respectively, in the topical bioassays (Tables 1 and 2). Mortality of adults and nymphs was significantly lower in the residue tests (Tables 1 and 2). O. majusculus mortality was much lower following exposure to the fungicides benomyl and copper salts + mancozeb, and was not significantly different from the control response in the topical assays (Tables 1 and 2).

Table 1 Mean percent mortality (± S.E.) of Orius majusculus adults in topical and residue bioassays
Full size table
Table 2 Mean percent mortality of Orius majusculus nymphs in topical and residue bioassays
Full size table

Chronic toxicity to adults and nymphs

In the short-term topical and residue tests the exposure duration was only 24 h, whereas a 4-d exposure was used in the persistency tests. This extended exposure period increased the mortality of adults and nymphs (Table 3).

Table 3 Mean percent mortality of Orius majusculus adults and nymphs following a 4-day exposure to insecticide residues
Full size table

Toxicity to eggs

Although the hatch rate in controls was somewhat higher than in the chemical treatments, the difference was not significant (Table 4).

Table 4 Mean hatch rate of Orius majusculus eggs following topical exposure to insecticides
Full size table

Choice and no-choice tests

Choice and no-choice test results are shown in Table 5. In choice tests, the total number of eggs from each small jar including treated and untreated bean pods did not differ significantly from that of the control treated with water only. Similar results were recorded for no-choice tests.

Table 5 Results of no-choice and choice tests to determine the oviposition rate of Orius majusculus in treated and non-treated bean pods
Full size table

Imidacloprid toxicity

At 350 mg a.i l −1 of imidacloprid, all test organisms were dead after 24 h, and there were no significant differences among the different exposure methods (Table 6). At 35 mg l −1, topical application of the pesticide also caused 100% mortality after 48 h. In the residue and systemic methods, mortality was high (>76%) but some insects were still alive after 10 days (Table 6).

Table 6 Results of topical, residue and systemic uptake tests using imidacloprid on Orius majusculus adults
Full size table

Discussion

Orius species have been used in several toxicity studies, most of those with O. insidiosus (Say) (e.g. Ashley et al. 2006; Bakker et al. 2000; Bostanian and Akalach 2004; Contreras et al. 2006). Elzen (2001) reported 55–63% mortality of O. insidiosus females during egg-feeding bioassays using imidacloprid, spinosad or endosulfan. In the current studies, topical exposure to imidacloprid, spinosad and endosulfan resulted in greater mortality, but mortalities during residue tests were lower (except for imidacloprid), when compared to the results of Elzen (2001). Michaud and Grant (2003) reported that, when exposed to field application rates of bifenthrin, carbaryl and methidathion, mortality to second instar nymphs of O. insidiosus was between 95% and 100% after 24 h in a leaf residue bioassay. Mortality was only 38%, however, upon indoxacarb exposure using the same test methods. The mortalities observed in the current residue studies with abamectin, endosulfan and spinosad were lower than those reported by Michaud and Grant (2003). Studebaker and Kring (2003) reported that test method significantly impacted the toxicity of nine insecticides to O. insidiosus. They tested the chemicals in the field, in pots and in glass petri dishes. Spinosad and imidacloprid caused higher mortality to O. insidiosus in petri dish studies than in field or greenhouse tests using treated cotton leaves. The low level of spinosad toxicity in the field or greenhouse tests was similar to our residue bioassays. The toxicity of abamectin in the current studies was lower than the toxicity reported by Studebaker and Kring (2003).

Data from both our study and prior investigations indicate substantial variability in the reported responses. There are many reasons for this variability, one of which is the test method used. It is also likely that sensitivity to the tested insecticides varies among Orius species. The length of exposure is an important factor in the degree of response observed for any particular test organism. Although acute responses in short-term tests can be substantial, toxicity during long-term exposures may be greater since continuous exposure promotes additional mortality, or cumulative effects require several days to be manifested, thus increasing the observed mortality after several days. Bostanian and Akalach (2004) evaluated toxicity of abamectin, endosulfan and other insecticides to O. insidiosus, with the insects being monitored for 1–9 days. Effects after a 1-d exposure to abamectin and endosulfan were similar to our results, with 44% and 8% mortality, respectively, in residue bioassays. However, mortality caused by abamectin and endosulfan increased to 100% and 37%, respectively, on day nine. Our data also showed higher mortality when exposure time was extended. Therefore, toxicity data from short-term (e.g. 24 h) exposures are likely to underestimate the actual toxicity of an insecticide. Topical, residual and systemic uptake methods using plant parts were the primary focus in our study. We believe these methods may provide some advantages for our aims when compared to other methods using an artificial surface such as a glass vial, petri dish, etc. A topical bioassay using a spray tower is similar to the field application of pesticides and reflects the “worst-case” conditions. Exposure occurs both through direct body contact as well as ingestion of residues on leaf surfaces. Furthermore, since plant metabolism and interactions with biological ligands on plant surfaces may alter chemical form and bioavailability, toxicity may also change, and any such change would not be apparent on artificial materials, as evidenced by the different exposure methods and significantly different responses reported by Studebaker and Kring (2003).

Field assays are required to evaluate the importance of the results obtained under our laboratory assays. However, the results from this study provide some important clues on the toxicities of the pesticides to O. majusculus. Topical results suggest that the field application of abamectin, endosulfan and spinosad should be avoided when O. majusculus is present, as they may cause significant negative effects, possibly approaching 100% mortality. Despite the high toxicity when the insecticides were applied topically, toxicity from residues was low after 24 h. The low residual toxicity may be an advantage for biological control systems using Orius spp. After a 4-d exposure, however, mortality ranged from 38% to 100%. Imidacloprid toxicity was consistently high, despite the test/exposure method, causing 100% mortality in all cases (Table 6). In addition, imidacloprid toxicity was persistent (Table 6). The very high initial toxicity of imidacloprid, even at 10% of the recommended dose rate, as well as its persistence, indicates imidacloprid is not a good candidate for IPM. The fact that abamectin, endosulfan and spinosad were not toxic to O. majusculus eggs (Table 4) may help re-establish the next generation of O. majusculus, even if there was 100% mortality of adults during insecticide application. The presence of insecticides on bean pods did not affect the rate at which eggs were laid on those surfaces (Table 5). This apparent lack of discrimination between treated and untreated surfaces may prove to be disadvantageous for O. majusculus, since avoidance would reduce adult exposure to the chemical as well as exposure for the juveniles, once hatched. Toxicity of the two fungicides, benomyl and copper salts + mancozeb, to O. majusculus was low. These chemicals, therefore, may be appropriate for IPM, and compatible with the use of O. majusculus.