Across all bioassays carried out, both earwig populations and life stages exhibited few cases of differences in the susceptibility to the tested insecticides (Fig. 1). Adults from the AR population were less susceptible than the PD population when exposed to lambda-cyhalothrin or thiamethoxam through either dried residues or ingestion of contaminated prey. However, indoxacarb dried residues allowed greater adult survival for the PD population than the AR population, whereas nymphs from the AR population survived better than nymphs from PD population when exposed to indoxacarb.
Exp. #1. Nymphal and adult exposures to dried residues
The three-way ANOVA for survival of the ring-legged earwig exposed to dried-residues of insecticides testing for predator population (F1, 156 = 2.90, P = 0.0901), predator life stage (F1, 156 = 2.2, P = 0.1359), and treatments (insecticides and control), resulted in significant effect only for insecticide treatments (F12, 156 = 431.90, P < 0.0001), among these main sources of variation. The average survival was 54.4% and 52.3% for individuals from populations AR and PD, respectively. Likewise, the survival of nymphs and adults was 54.3% and 52.4%, respectively, irrespective of population and insecticides. Across the two- and three-factor interactions, only those interactions involving the insecticide treatments were significant: predator population and treatments (F12, 156 = 5.10, P < 0.0001), predator life stage and treatments (F12, 156 = 8.51, P < 0.0001), predator life stage, treatment, and population (F12, 156 = 7.90, P < 0.0001). Based on that, the three-way statistical model was reduced to a one-way model considering only the treatments as a source of variation, thus enhancing the degrees of freedom of the error. The insecticides significantly affected the earwig survival (F12, 195 = 185.06, P < 0.0001; Fig. 2), and hence we grouped them with respect to their toxicity levels. Pyriproxyfen, pymetrozine, chlorantraniliprole, cyantraniliprole, and spinetoram were classified as low-impact, allowing predator survival of >90%, which was similar to control. In contrast, chlorfenapyr, chlorpyrifos, dimethoate, and malathion caused high mortality of the earwig (survival <3%), and these products were classified as a high-impact group. The remaining insecticides, i.e., thiamethoxam, indoxacarb and lambda-cyhalothrin, allowed 20–60% earwig survival and were classified as of intermediate impact.
The chronic toxicity evaluation (survival or molting into adult across 20 d exposure) in Exp. #1 showed different results (Fig. 3). Over 70% of adult earwigs from both AR and PD populations survived exposure to chlorantraniliprole, pymetrozine, spinetoram, and pyriproxyfen residues, which was significantly greater than the ~27% survival after exposure to cyantraniliprole (Fig. 2). The remaining insecticides caused 100% mortality of both earwig populations. With respect to nymphs, over 78% of both populations exposed to pymetrozine or spinetoram residues and the nymphs of the PD population exposed to chlorantraniliprole, successfully molted into adults within 30–40 d (Fig. 3). Survival of nymphs in these treatments was similar to the control treatment and significantly greater than the rate observed for the AR population exposed to chlorantraniliprole (Fig. 3). Regardless of the population, nymphs exposed to pyriproxyfen did not reach adulthood, although they lived for up to 80 d. All remaining insecticides caused 100% nymphal mortality.
Within treatments that allowed survival of nymphs until the adult molt, the time required to complete nymphal stage varied by earwig population (F1, 221 = 30.0, P < 0.0001), insecticides (F1, 221 = 12.0, P < 0.0001), and interaction of these factors (F3, 221 = 11.7, P < 0.0001). Nymphs from the PD population exposed to spinetoram or pymetrozine residues delayed their development about 2 d in comparison to nymphs from the AR population (Table 1). Furthermore, nymphs from the AR population exposed to chlorantraniliprole residues delayed their development in comparison to nymphs exposed to pymetrozine and spinetoram residues (Table 1).
Exp. #2. Ingestion of contaminated prey
Ring-legged earwigs from the AR and PD populations consumed similar quantities (average = 5.6 and 5.4 mg, respectively) of contaminated prey during a 24 h confinement period (F1, 985 = 0.9, P = 0.3402), with lack of significant effect predator population and treatment interaction (F12, 985 = 1.6, P = 0.084). However, contaminated prey consumption was significantly affected by earwig life stage (F1, 985 = 474.9, P < 0.00001), treatments (F1, 985 = 43.4, P < 0.0001), and by the interactions of predator population and treatments (F1, 985 =159.9, P < 0.0001), life stage and treatments (F1, 985 = 159.9, P < 0.0001), and predator population, life stage and treatments (F12, 985 = 2.0, P = 0.016). Prey consumption by nymphs was similar to control values for eggs treated with chlorfenapyr, spinetoram, lambda-cyhalothrin, pyriproxyfen, indoxacarb, and chlorantraniliprole, but was reduced for prey treated with chlorpyrifos, pymetrozine, thiamethoxam, cyantraniliprole, malathion, and dimethoate (Fig. 4 - nymphs). For adult earwigs, the highest and the lowest prey consumption rates were observed in the spinetoram and thiamethoxam treatments, respectively, with the other compounds resulted in intermediate consumption levels (Fig. 4 – adults).
The developmental time (days to molt into adult) of 3rd-instar nymphs preying on contaminated prey was affected by earwig population (F1, 166 = 222.7, P < 0.0001), insecticide treatments (F5, 166 = 38.1, P < 0.0001), and the interaction of these factors (F5, 166 = 23.2, P < 0.0001). Nymphs from the PD population that consumed prey treated with chlorantraniliprole, cyantraniliprole, lambda-cyhalothrin, pymetrozine, or those eating the untreated (control treatment) took longer to reach adulthood compared to those from AR population in these same treatments (Table 1). Among the insecticides that allowed the PD population nymphs to molt into adults, indoxacarb and lambda-cyhalothrin caused developmental delays (Table 1). In contrast, nymphs from the AR population were most affected by eating prey contaminated with cyantraniliprole, chlorfenapyr, and spinetoram (Table 1).
Survival of nymphs (that molted into adults) and of adult earwigs did not vary as a function of the population. However, survival of nymphs after the consumption of contaminated prey was affected by life stage, insecticide treatment, and the interaction of these factors. Consumption of prey contaminated with pyriproxyfen, thiamethoxam, chlorpyrifos, dimethoate, or malathion caused ≥97% mortality to nymphs irrespective of population. The same high level of mortality was also observed for the PD population after eating chlorfenapyr-contaminated prey and for the AR population after eating indoxacarb-contaminated prey (Table 1, Fig. 5). Nevertheless, the lowest level of nymphal mortality was observed with pymetrozine and spinetoram, which were similar to controls (Fig. 5). Furthermore, nymphs of the PD population were more affected than nymphs of the AR population following consumption by lambda-cyhalothrin, chlorfenapyr, and cyantraniliprole-contaminated prey. Yet, nymphs of the AR population suffered high mortality from eating prey contaminated with cyantraniliprole (Fig. 5).
Mortality of adult earwigs from the PD population, when followed for 20 days was 20% for pymetrozine, spinetoram, pyriproxyfen, chlorantraniliprole, lambda-cyhalothrin, and the control, whereas mortality were 100% for malathion, chlorpyrifos, dimethoate, cyantraniliprole, and thiamethoxam. For indoxacarb and chlorfenapyr, mortality was intermediate (Fig. 6). Similarly, for the AR population, there was <20% mortality following exposure to spinetoram or the control, but exposure to malathion, chlorpyrifos, dimethoate, or indoxacarb caused 100% mortality, with the remaining insecticides causing intermediate levels of mortality (Fig. 6).
Exp. #3. Mixed exposure to dried residues on treated plants and contaminated prey
The survival of adult females of E. annulipes that were simultaneously exposed to treated cotton plants and contaminated prey for five days was significantly affected by the insecticide treatments (F5, 128 = 32.26, P < 0.0001, Fig. 6). The greatest and the lowest survival were observed in the thiamethoxam and malathion treatments, respectively, whereas the control, pymetrozine, lambda-cyhalothrin, and indoxacarb showed intermediate rates of survival.
Repeated measures ANOVAs detected an effect of time on the level of prey consumption (24 h vs. 120 h) (Wilks’ lambda = 0.71, F = 40.53, P < 0.0001, DFnum = 1, den = 100), the insecticide treatments [F5, 100 = 8.20, P < 0.0001 (24 h); and F5, 100 = 2.49, P = 0.006 (120 h)], and the interaction of these factors (Wilks’ lambda = 0.72, F = 7.90, P < 0.0001, DFnum = 5, den = 100). In the 24h-evluation period, >85% of the predators in the control and indoxacarb treatments had foraged on the plant canopy and consumed ca. 82% of the available prey, which was significantly greater than that observed in the thiamethoxam, pymetrozine, and malathion treatments, followed by lambda-cyhalothrin, for which the percentage of predators foraging on the plant was lowest (Fig. 7). By the last 120h-evaluation, the prey consumption level and the percentage of predators that foraged on the plants for the indoxacarb and malathion treatments were not different that the first assessment, as mentioned before. Nevertheless, a significant increase of the percentage of predators that foraged on the plants and consumed prey was observed over the time in the pymetrozine and lambda-cyhalothrin treatments. For thiamethoxan, the increase in these parameters over the trial was very slight.