Response of Doru luteipes (Dermaptera: Forficulidae) to insecticides used in maize crop as a function of its life stage and exposure route

The present study aimed to evaluate insecticide toxicity to Doru luteipes (Scudder), a major predator of maize pests. Lethal and sublethal effects were assessed on nymphs and adults exposed to the insecticides through contact (maize leaves) and ingestion (prey eggs) routes. Tested insecticides included a biopesticide (Spodoptera frugiperda multiple nucleopolyhedrovirus, SfMNPV), modern (flubendiamide and metaflumizone), and older neurotoxins (imidacloprid + β-cyfluthrin). The imidacloprid/β-cyfluthrin mix was highly toxic (100% mortality) to the predator, regardless of the exposure route and predator stage. Metaflumizone caused mortality higher than 95% and 45% of nymphs and adults. Flubendiamide and SfMNPV were the least toxic insecticides, not differing from the untreated control in any of the assessed endpoints. Adult tibial length did not differ among treatments. Metaflumizone impaired egg consumption by nymphs and walking distance of adult D. luteipes. Overall, the insecticides caused a more pronounced effect on D. luteipes nymphs than on adults and were more toxic by the contact route. From these findings, flubendiamide and SfMNPV are safer for D. luteipes and should head insecticide choice in integrated pest management programs in maize.


Introduction
Maize (Zea mays L.) is one of the most important cereals worldwide and is widely grown for human and animal consumption and other industrial uses (Ranum et al. 2014). High-yielding fields are needed to meet global demand for maize products but their productivity is often challenged by damage from phytophagous insects (Badji et al. 2004;Silva et al. 2018). Spodoptera frugiperda (Smith) (Lepidoptera: Noctuidae), Dalbulus maidis (DeLong & Wolcott) (Hemiptera: Cicadellidae), and Diceraeus melacanthus (Dallas) (Hemiptera: Pentatomidae) are the major pests infesting maize fields in Brazil (Santos-Amaya et al. 2016;Fernandes et al. 2022;Neves et al. 2022).
Nymphs and adults of D. luteipes are generalist voracious predators; their diet includes aphids, leafhoppers, and immature (eggs and small larvae) lepidopterans (Fernandes et al. 2007;Pacheco et al. 2021). This species remains hidden during the day and forages at night (Naranjo-Guevara et al. 2017). Throughout its lifespan, D. luteipes can consume more than 8300 Helicoverpa zea (Boddie) (Lepidoptera: Noctuidae) eggs and 2800 S. frugiperda larvae (Reis et al. 1988;Cruz et al. 1995). A field study revealed a negative correlation of D. luteipes with S. frugiperda and D. maidis (Badji et al. 2004), which highlights the importance of this predator for sustainable maize agroecosystems. Hence, preservation of this natural enemy in the field is important for pest population regulation (Castle and Naranjo 2009).
The application of broad-spectrum insecticides (i.e., organophosphates, pyrethroids, and carbamates) is commonplace in maize crops because they are generally cheap and fast-acting (Laiz et al. 2018). Most of these compounds are non-selective to natural enemies (predators and parasitoids), which causes a reduction in the populations of these beneficial organisms and consequent biological imbalances (e.g., pest resurgence or secondary pest outbreaks) (Torres and Bueno 2018). Therefore, the adoption of integrated pest management (IPM), aiming at the harmonious integration of control methods (e.g., biological and chemical control), is key to the conservation of natural enemies in agroecosystems (Castle and Naranjo 2009;Cutler 2020). Within this framework, the assessment of insecticide impact on key natural enemies is central to the development of sound IPM programs (Desneux et al. 2007;Torres and Bueno 2018;Carvalho et al. 2019). Bioinsecticides (e.g., Bacillus thuringiensis and baculovirus-based insecticides), modern neurotoxins (e.g., metaflumizone, diamides, and spinosyns), and insect growth regulators, among others, present themselves as potentially safer options for IPM programs (Horowitz and Ishaaya 2004;Campos et al. 2011;Torres and Bueno 2018).
Pesticides adversely affect natural enemies through direct mortality and/or sublethal effects (e.g., impaired development, sex ratio, fecundity, locomotion, and predation ability) (Desneux et al. 2007;He et al. 2019;Soares et al. 2020;Williams et al. 2020), both of which lead to reduced efficacy of biocontrol agents. These adverse effects are mediated by the route of exposure (e.g., contact and ingestion), as reported for parasitoids (Haseeb et al. 2005;Ruiz et al. 2008;Costa et al. 2022) and predators (He et al. 2012;Barbosa et al. 2017;Wang et al. 2018). Life cycle stage is another factor that influences the susceptibility of natural enemies to insecticides; overall, earlier stages are more susceptible, as demonstrated for dermapterans (Potin et al. 2022) and other predators (Prabhaker et al. 2017;Balanza et al. 2021). Insecticides have been shown to be lethal and to reduce predation activity and locomotor activity of earwigs, as shown to Forficula auricularia L. (Dermaptera: Forficulidae) and D. luteipes (Campos et al. 2011;Malagnoux et al. 2015;Fountain and Harris 2015;Jana et al. 2021); chlorpyrifos and, to a lesser extent, spinosad were consistently non-selective in these studies. However, the effects of some insecticide groups (e.g., bioinsecticides and novel neurotoxins) on D. luteipes remain unstudied.
In this study, the selectivity of insecticides labeled for maize was investigated for D. luteipes. The lethal and sublethal effects of the compounds, which include different insecticide classes (biopesticide, novel, and old neurotoxins), were studied in nymphs and adults exposed to the compounds via the contact and ingestion routes.

Insects
Adults of D. luteipes (ca. 500 individuals) were collected in a maize field located in the municipality of Lavras, Minas Gerais, Brazil (21° 14′ 43 S, 4° 59′ 59 W) and taken to the laboratory. Occasionally, new adults were collected in the field and added to the laboratory rearing to minimize inbreeding.
The insects were placed in polyvinyl chloride (PVC) cages (20 cm wide × 20 cm high). In the rearing cages, pieces of black cardboard folded in an accordion-like fashion were placed as shelters for the insects. A 5-cm diameter Petri dish containing a piece of cotton moistened with water was placed inside the cage, and a plastic straw with a piece of cotton in one of its ends was used as the oviposition substrate. The artificial diet (containing cat food, wheat bran, beer yeast, and milk powder) described by Cruz (2009) was offered in Petri dishes (5-cm diameter). Insect rearing maintenance was performed three times a week on alternate days, by changing the diet and moistening the cotton. The eggs laid by the predator were removed from the rearing cage and placed together with the progenitor in other cylindrical PVC cages (20 cm wide × 20 cm high), supported on circular plastic trays 25 cm in diameter until the nymphs hatched. The ends of the PVC cages were sealed with voile fabric secured with rubber bands, and the cages were kept in a climate-controlled room (25 ± 2 C, 70 ± 10% RH, with a 12-h:12-h L:D photoperiod).
We also maintained a S. frugiperda colony that served as prey for D. luteipes in the bioassays. Field-collected larvae were brought to the laboratory and reared on artificial food (Greene et al. 1976). Adults were placed in cylindrical PVC cages (30 cm wide × 40 cm high) for mating and fed a 10% honey solution. They were allowed to lay their eggs on sulfite paper on the inner walls of the cages. The eggs were collected daily and incubated in plastic bags with moistened filter paper until hatching. Neonates (L1) were fed an artificial diet containing cooked dry beans, wheat germ, soybean protein, and casein (Greene et al. 1976). Second-generation insects (F2) were used in the bioassays. This colony was kept at 25 ± 2 C, 70 ± 10% RH, and a 12-h:12-h L:D photoperiod.

Bioassays
All bioassays were performed at 25 ± 2 C, 70 ± 10% RH, and a 12-h:12-h L:D photoperiod. The experiments were performed using two routes of exposure of D. luteipes to the insecticides. In the contact bioassay, the predator was exposed to insecticide-treated corn leaves. Conventional maize (SHS4070) plants were grown in 10-L PVC pots containing soil and cattle manure (3:1 v:v), and received N-P-K 4-14-8 chemical fertilizer according to the results of the soil analysis. The plants were treated at stage V4 by manual spraying (with a 20-L backpack sprayer, Jacto, model PJH 20, Máquinas Agrícolas Jacto SA, Brazil) of the insecticides to the runoff point (60 mL plant −1 ), and allowed to dry at room temperature.
The second exposure route consisted in offering insecticide-sprayed S. frugiperda eggs to the predator. Newly laid eggs of S. frugiperda from the laboratory rearing were treated with insecticides using a Potter spray tower (Burkard, Uxbridge, UK) calibrated a 15 lb. pol −2 to deposit 1.5 ± 0.4 mg cm −2 , according to guidelines of the International Organization for Biological Control (IOBC) (Hassan et al. 1985). Eggs were allowed to dry at room temperature for 1 h before being placed in the Petri dishes (ca. 300 eggs per dish) containing the predator.

Survival of D. luteipes nymphs and adults
Nymphs and adults of D. luteipes were tested. Second-instar nymphs were placed in Petri dishes (10-cm diameter) and exposed to treated maize leaves (10-cm-long piece) or insecticide-sprayed S. frugiperda eggs. We set an exposure time of 72 h, which means that the nymphs were in contact with the sprayed maize leaves or S. frugiperda eggs for that long. In the contact bioassay, nymphs were fed an artificial diet ad libitum, offered in small plastic containers, and a moistened cotton pad was given as the water source (Cruz 2009;Silva et al. 2022). In the ingestion bioassay, nymphs did not receive the artificial diet but treated S. frugiperda eggs (ca. 300 eggs per Petri dish) and after 72 h, a daily supply of untreated S. frugiperda eggs ad libitum. In both experiments, the experimental design was completely randomized, with six treatments (SfMNPV, flubendiamide metaflumizone, imidacloprid + β-cyfluthrin, chlorpyrifos, and the untreated control) and 16 replicates, each consisting of five nymphs per Petri dish. The daily mortality of nymphs (up to 15 days) and the tibial length of adults (n = 70) developing from the treated nymphs were assessed. The tibial length was measured using ImageView Software and a UA1000CA camera coupled to a computer for image capture. Due to the low survival of chlorpyrifos-and imidacloprid + β-cyfluthrin-treated insects, tibial length assessments were not performed for these insecticides.
Similar to nymphs, adults (24 h old) were exposed to two exposure routes (contact and ingestion). In the contact experiment, an adult pair (male and female) was placed in Petri dishes (10-cm diameter) and exposed to treated maize leaves (10-cm-long piece). Adults were fed ad libitum with an artificial diet (Cruz 2009), and a moistened cotton pad was used as the water source. In the ingestion bioassay, an adult pair was placed in Petri dishes (10-cm diameter) with moistened cotton pads, and insecticide-treated S. frugiperda eggs were offered ad libitum. The survival of each insect was monitored daily for up to 15 days (360 h). In both bioassays, the experimental design was completely randomized, with six treatments and five replicates, each consisting of five pairs, for a total of 50 insects per treatment.

Egg consumption by D. luteipes nymphs and adults
Insecticides were assessed for their effect on reducing egg consumption by D. luteipes. Before the experiments, the insects fasted for 24 h. Adults and nymphs were exposed to the insecticides through the contact and ingestion routes, as described in the previous sections. Then, they were offered S. frugiperda eggs (ca. 300 eggs placed in small plastic discs), which had been previously weighed. Egg consumption (in milligrams) was determined by subtracting the initial weight of treated eggs from the final weight after 72 h. To account for egg weight loss due to desiccation, an untreated control (without water application) was also included, referred to here as blank. Egg consumption (EC) was then corrected using Abbott's formula (Abbott 1925): EC = (CT-LB)/(1-LB), where CT = egg consumption in a given treatment and LB = weight loss in the blank treatment 72 h after the start of the experiment. We were unable to evaluate egg consumption by chlorpyrifos-and imidacloprid + β-cyfluthrin-treated insects because of their low survival as determined in the previous bioassays. The experimental design was completely randomized, with four treatments (flubendiamide, metaflumizone, SfMNPV, and untreated control) and 10 replicates, each consisting of five nymphs or an adult pair (one male and one female) placed in Petri dishes.

Walking distance of D. luteipes adults
Adult D. luteipes (up to 48 h of age) were exposed to the insecticides through contact and ingestion (as described in the previous sections), and had their walking distance recorded. Twenty adults of the predator (10 females and 10 males) were treated per treatment. Twenty-four hours after exposure to the insecticides, each predator was transferred to a 10-cm Petri dish, with the edges covered with Teflon powder to avoid escape. The walking distance (in centimeters) of each insect (placed individually in Petri dishes) was monitored for 10 min by using a digital camera connected to a computer equipped with Ethowatcher tracking software (Campos et al. 2011;Crispim Junior et al. 2012;Andreazza et al. 2020). Due to the low survival of chlorpyrifos-and imidacloprid + β-cyfluthrin-treated insects, walking distance studies were not performed for these insecticides.

Data analysis
The analyses were performed using R software (V. 3.5.3) and RStudio (V. 1.2.5001) (R Core Team 2019). All statistical tests were based on a significance level of 0.05. Nymph and adult survival data were analyzed by applying the Weibull model with the survreg function from the survival package (Weibull 1951;Therneau 2020). The fit of the data to the Weibull distribution was checked using the Kolmogorov-Smirnov test. After selecting the appropriate mathematical model according to the residuals, contrast analyses were performed to evaluate the similarity between treatments, and congener groups were formed in the case of the occurrence of statistically similar treatments (Passos et al. 2017;Pares et al. 2021). The survival probability function for each group was estimated using the following formula: S x = exp(-(x/δ) α ), where S x = survival probability at a given time x; δ = shape parameter; and α = scale parameter. Both δ and α parameters were estimated by fitting Weibull models (Pares et al. 2021). The median lethal time (LT 50 , the time required to kill 50% of the population) was calculated for each congener group by algebraic manipulation of the survival function to find the value of x (time) that gives S = 50.
Data on tibial length of adults developed from treated nymphs, egg consumption, and walking distance were checked for normality (Shapiro-Wilk test) and homoscedasticity of residuals (Bartlett test). The effect of insecticide on egg consumption and tibial length was checked through one-way ANO-VAs (lm function from stats package). A two-way ANOVA was performed to verify the effect of insecticide, predator sex, and their interaction on the walking distance of adult D. luteipes and, where appropriate, a Tukey's post hoc test was performed (function glht from multcomp package) (Hothorn et al. 2008).

Discussion
This study reports the lethal and sublethal effects of insecticides from different classes (biopesticide and novel and oldgeneration neurotoxins) on D. luteipes. The predator susceptibility to insecticides varied with the life stage exposed (nymph and adult) and exposure route (contact and ingestion).
Chlorpyrifos (positive control) was the most lethal compound, followed by imidacloprid + β-cyfluthrin (neonicotinoid + pyrethroid), regardless of the predator stage and exposure route. Imidacloprid + β-cyfluthrin is a mix of compounds that act on insects' nervous system, in axonal and synaptic transmissions, which can potentiate their toxic effect (Coats 2012). Neonicotinoids and pyrethroids, applied individually or in mix, comprises roughly 80% of foliar insecticide sprays in Brazil's maize crops (Laiz et al. 2018), and the present study indicates that these compounds can disrupt D. luteipes populations. Our findings are corroborated by earlier reports of the harmfulness of neonicotinoids and pyrethroids to earwigs and other natural enemies (Redoan et al. 2013;Barros et al. 2018;Skouras et al. 2021).
Metaflumizone (a semicarbazone) was the most toxic compound among the novel insecticides; it impaired nymph feeding and adult locomotion. In addition, metaflumizone significantly reduced the survival of D. luteipes, with total mortality greater than 95% and 45% of nymphs and adults, respectively. The survival curve for this insecticide was  (a and b) and adults (c and d) exposed to insecticides through contact (a and c) and ingestion routes (b and d). Contrast analyses were performed between the treatments to identify similarities and four congener groups were formed: group 1 = control (distilled water), flubendiamide and SfMNPV; group 2 = metaflumizone; group 3 = imidacloprid + β-cyfluthrin; group 4 = chlorpyrifos Fig. 2 Tibial length (mean ± SEM) of Doru luteipes adults developing from nymphs exposed to insecticides through (a) contact and (b) ingestion routes. NS = non-significant difference among treatments flatter compared to the more toxic compounds, indicating that metaflumizone has a slow action on the predator. Metaflumizone is active against a broad spectrum of insects (e.g., Coleoptera, Diptera, Hemiptera, and Hymenoptera) (Salgado and Hayashi 2007). This compound exhibits lethal and sublethal effects on natural enemies, including Orius laevigatus (Fieber) (Hemiptera: Anthocoridae) (Biondi et al. 2012), Nesidiocoris tenuis (Reuter) (Hemiptera: Miridae) (Wanumen et al. 2016), and Harmonia axyridis (Pallas) (Coleoptera: Coccinellidae) ). This molecule blocks insect sodium channels by binding selectively to the slow-inactivated state of the channels, which leads to flaccid paralysis, feeding cessation, and ultimate death (Takagi et al. 2007). This explains its effects on D. luteipes (reduced feeding and locomotion and delayed mortality).
Metaflumizone decreased adult mobility and prey consumption of D. luteipes. Other studies have shown that insecticides affect these biological traits of dermapterans; spinosad caused a marked reduction in locomotor activity of adult D. luteipes (Campos et al. 2011), whereas both spinosad and chlorpyrifos reduced aphid consumption by F. auricularia (Malagnoux et al. 2015). Impaired movement of predators could make them more susceptible to desiccation and intraguild predation (Ffrench-Constant and Vickerman 1985;Salerno et al. 2002). The reduction in prey consumption likely resulted from the diminished mobility (which limits prey search and capture), although changes in the neural and sensory systems may also play a role (Salerno et al. 2002;Desneux et al. 2007). Either way, when insecticides interfere with prey consumption by predators, they reduce the ecosystem services exerted by these beneficial insects, which can lead to community Bars with the same letters do not differ by Tukey's test (α = 0.05). NS = non-significant difference among treatments imbalance and increases in pest populations (Naranjo et al. 2015;Torres and Bueno 2018).
Few studies have examined the sex-dependent responses of earwigs to insecticides. Redoan et al. (2013) found no differences in survival between male and female D. luteipes exposed to selected insecticides, whereas Malagnoux et al. (2015) found that F. auricularia males were more affected than females in terms of predatory activity. Here, we found that male D. luteipes walked longer than females in the ingestion bioassay, while this response was not sex-dependent in the contact bioassay. We hypothesize that oral uptake of insecticides may have influenced one of the sexes (either by decreasing female walking performance or by increasing male activity) by triggering responses that were not elicited by contact.
Flubendiamide and SfMNPV did not cause detrimental effects on D. luteipes, not differing from the untreated control in any of the assessed endpoints. Flubendiamide is a narrow-spectrum insecticide highly active against lepidopteran pests (Tohnishi et al. 2005) which acts as a modulator of ryanodin receptors to block the muscular contraction of insects (Lahm et al. 2009). Diamides in general are quite selective for several groups of natural enemies. Stecca et al. (2014) found flubendiamide and chlorantraniliprole to be harmless to adult D. lineare. Another study found chlorantraniliprole and cyantraniliprole to differ in their lethality to Euborellia annulipes (Lucas) (Dermaptera: Labiduridae); cyantraniliprole caused 100% mortality of nymphs and adults up to 38 days through the contact route, while chlorantraniliprole led to a markedly low lethality (< 35% mortality) (Potin et al. 2022). As for SfMNPV (a disruptor of the midgut peritrophic membrane, Rao et al., 2004), baculovirus-based insecticides have been consistently shown to be innocuous to natural enemies (Simões et al. 1998;Amaro et al. 2015;Gutiérrez-Cárdenas et al. 2020).
Irrespective of the predator stage (adult and nymph), D. luteipes was more susceptible to the insecticides by the contact pathway, as indicated by the lower LT 50 s of contact bioassays. This occurred especially for insects exposed to metaflumizone and imidacloprid + β-cyfluthrin, which killed twice as fast by contact. These results are in line with those of Potin et al. (2022), who found insecticides to be more toxic through contact than by ingestion against the earwig E. annulipes. The contact route may have a more pronounced effect since earwigs exhibit self-and allogrooming behavior (Weiß et al. 2014), which may increase the insecticide assimilation into the body and result in higher mortality. Besides, ingested insecticides face salivary and digestive enzymes, which may break down these molecules into less toxic forms (Potin et al. 2022).
This work provides a better understanding of the impact of insecticides on D. luteipes, a ubiquitous predator in maize crops. Based on our findings, exposure to imidacloprid + β-cyfluthrin and metaflumizone reduces predator Fig. 4 Walking distance (mean ± SEM) of adult Doru luteipes (female and male) exposed to insecticides through (a) contact and (b) ingestion routes. Control = distilled water, Flub = flubendiamide, Metaf = metaflumizone. Within each predator sex, bars with the same letters do not differ by Tukey's test (α = 0.05) survival and hinders its locomotion and prey consumption. These compounds have a stronger impact on contact-treated D. luteipes nymphs, although their effects in adults and by the ingestion route cannot be disregarded. Because of their selectivity toward D. luteipes, flubendiamide and SfMNPV are safer options for IPM programs in maize.