Sublethal and Lethal Effects of Low-Dose Prothioconazole Alone and in Combination with Low-Dose Lambda-Cyhalothrin on Carabid Beetles in a Field-Realistic Scenario

Abstract

Environmental risk assessment (ERA) for pesticide approval in the context of predatory insects remains inadequate as it often overlooks the influence of agricultural practices. An increasing number of studies have shown that prolonged and synergistic pesticide exposure can elevate insect mortality. However, such effects remain largely unstudied in non-target predatory carabid beetles. The carabid beetle Platynus assimilis was subjected to repeated oral and continuous contact exposure to low doses of prothioconazole (20 g ha⁻¹), lambda-cyhalothrin (0.4 g ha⁻¹), or their combination over a 64-day period. The food consumption rate, body mass, locomotor activity, and mortality were monitored throughout the experiment. All pesticide-treated groups showed significantly increased final mortality, with median lethal times (LT₅₀) of 51.6 days for prothioconazole, 60.3 days for lambda-cyhalothrin, and 12.2 days for their combination. A significant synergistic effect on mortality was observed in the combined treatment group, with the highest synergistic ratio detected 20 days after the first exposure. Pesticide-treated beetles exhibited significant abnormalities in locomotor activity and body mass compared to the untreated group. These findings demonstrate that both time-cumulative mortality and potential synergistic interactions, reflecting field-realistic conditions, must be considered in ERA. Failure to do so may lead to an underestimation of pesticide toxicity to predatory carabids.

1. Introduction

The widespread contamination of terrestrial ecosystems with agricultural pesticides [1,2,3] has significantly contributed to the global decline in insect species richness and abundance [4,5,6,7]. Populations of non-target beneficial insects, which provide essential ecosystem services in agricultural food production, are also at serious risk [5,8,9]. Despite the rigorous processes involved in evaluating the environmental impact of pesticides for authorisation in the European Union [10], the current environmental risk assessment (ERA) procedures are insufficient. These regulatory frameworks fail to prevent authorised and widely used pesticides from exerting unacceptable, detrimental effects on non-target arthropods, biodiversity, and the broader environment [10,11,12]. Notably, current ERA schemes often overlook realistic field conditions, where time-cumulative, synergistic, and sublethal effects of pesticides may occur [11,13,14,15], adversely affecting the health, performance, survival, and population dynamics of non-target beneficial insects [16,17,18,19]. Without incorporating such realistic exposure scenarios, the toxicity of pesticides to non-target insects is likely underestimated.

In light of a growing global human population [20] and the ongoing biodiversity crisis, a substantial reduction in pesticide use is essential, returning to the foundational principles of integrated pest management (IPM) [11,16,21]. IPM combines biological control strategies, including the use of natural enemies of crop pests, with restricted pesticide application. Epigeal predatory carabids are key agents in IPM programmes [22], providing ecosystem services as effective natural enemies of various crop pests [23,24,25] and as voracious consumers of weed seeds [26,27]. A prerequisite for the successful integration of biological and chemical controls is that pesticide applications must not impair the effectiveness of biological control agents.

Standard toxicity testing methods for pesticides in non-target predatory insects typically span 10–14 days [28,29]. However, this duration is inadequate for adequate environmental risk assessment, as evidence demonstrates time-cumulative mortality effects with extended exposure periods lasting several months. Such effects have been observed for fungicides, such as boscalid and pyraclostrobin, in honeybees [30,31], and for low-dose neonicotinoids in bees, ants, and termites [32,33]. Pesticide-induced physiological responses, including detoxification and toxic stress, may elevate energetically costly metabolic activity [34,35], resulting in physiological exhaustion and compromising insect health and survival [30,36]. To date, no studies have addressed the effects of prolonged, field-realistic exposure to widely used fungicides and/or low-dose insecticides on the survival of predatory carabid beetles.

Furthermore, data on the synergistic effects of pesticide combinations in natural enemies are limited, making it difficult to predict the magnitude of such interactions under realistic exposure conditions. In agricultural systems, the co-application of insecticides and fungicides, either sequentially or as tank mixtures, is commonplace [37]. When sterol biosynthesis-inhibiting (SBI) fungicides (FRAC code 3) are applied with pyrethroid, neonicotinoid, or butenolide insecticides, acute synergistic lethal and sublethal effects have been documented in bees [38,39,40,41,42,43] and parasitoid wasps [18]. These synergistic effects are believed to result from the SBI fungicide-induced inhibition of cytochrome P450 monooxygenases, which play a crucial role in detoxifying various xenobiotics, including insecticides [37,44]. Therefore, similar synergistic toxic effects are also plausible in predatory insects, such as carabids, although empirical data are currently lacking.

In the present study, we evaluated the potential toxic effects of the low-dose SBI fungicide prothioconazole (Prot), the low-dose pyrethroid insecticide lambda-cyhalothrin (LCyh; IRAC class 3A, sodium channel modulators), and their combination on the ground-dwelling predatory carabid beetle Platynus assimilis under a prolonged, field-realistic exposure scenario. Prot and LCyh are both extensively used to protect various crops, including cereals, vegetables, oilseed rape, orchards, and cotton [45,46,47]. The maximum recommended field doses (MRDs) for LCyh and Prot are 40 g ha⁻¹ and 200 g ha⁻¹, respectively, with application frequencies ranging from one to four times per season depending on crop type, pest pressure, and fungal infection levels. In our 64-day laboratory experiment, we assessed the lethal and sublethal effects of low-dose Prot, low-dose LCyh, and their combination on adult P. assimilis. The exposure model included repeated oral ingestion of contaminated prey and continuous dermal contact with pesticide residues, reflecting real-world conditions in intensively managed agricultural landscapes. Based on previous studies in other insect taxa, we hypothesised that co-exposure would result in toxic synergism, wherein the combined effect would exceed the additive effects of individual treatments. We also predicted that prolonged Prot exposure alone would significantly increase mortality due to accumulation, toxic stress, and physiological exhaustion. Our primary objective was to demonstrate that simultaneous or prolonged pesticide exposure can lead to elevated mortality and sublethal effects in beneficial predatory carabid beetles, outcomes not detected under standard ERA procedures.

Prot is characterised by low persistence in soil, with dissipation times (DT₅₀) ranging from 1.7 to 9.9 days [48,49,50]. However, two of its primary metabolites—prothioconazole-desthio (49.4%) and prothioconazole-S-methyl (14.6%)—are more persistent, with DT₅₀ ranges of 15.8–72.3 days and 16.6–99.6 days, respectively [48,49]. These compounds exhibit low mobility, largely remaining within the top 10 cm of soil [48,49]. Toxicity data for Prot and its metabolites in terrestrial arthropods are scarce. Existing studies have suggested a low risk to foliar-dwelling non-target arthropods, including honeybees (Apis mellifera) and predatory insects (Coccinella septempunctata (Coccinellidae), Chrysoperla carnea (Chrysopidae)), and to ground-dwelling predatory beetles such as Poecilus cupreus (Carabidae) and Aleochara bilineata (Staphylinidae) [48,51]. However, recent work has shown that contact exposure to Prot can induce oxidative stress in A. mellifera [52,53]. These studies, however, do not consider potential long-term or synergistic effects with other pesticides.

In comparison, LCyh shows a soil DT₅₀ of 14 to 163 days under laboratory conditions [54,55,56], and a half-life of 3 to 40 days in field conditions [57,58]. It is also immobile in soil and highly hydrophobic, suggesting greater toxicity through dermal contact than ingestion [37]. Accordingly, both Prot and Lcyh, along with their degradation products, are likely to persist in the topsoil during the growing season, posing long-term risks to non-target, ground-dwelling carabids via multiple exposure pathways.

The medium-sized Euro-Siberian carabid P. assimilis (10–13 mm) predominantly inhabits woodland, but also migrates into agricultural lands during spring and summer [23,59,60,61]. It also inhabits crop fields [62,63,64] and grasslands [65]. Trophically, it belongs to the generalist predator guild [66,67]. The beetles hatch from the pupa from July until September, overwinter, and reproduce in the following spring from April to June [60,68]. Their life span ranges from one to two years. Only a part of the old generation dies during the winter. Adults from the previous year form about half of the breeding population [68]. Therefore, prolonged exposure to low doses of these pesticides within several months is a field-realistic scenario. Previous ecotoxicological research has identified P. assimilis as a suitable model organism for laboratory testing [69,70,71].

2. Materials and Methods

2.1. Test Insects

The adults of P. assimilis were collected from their preferred overwintering sites in brown-rotted wood at a forest margin in Tartu County, southern Estonia (58.36770 N, 26.35106 E) in October, 2021. This site was near conventional crop fields. The beetles were kept in 3 L plastic containers, with 50 beetles in each, containing moist pieces of brown-rotted wood, and stored in a refrigerator at 5 °C for several months until the experiment. During storage, no signs of aggression or mortality were observed.

Five days prior to the toxicological experiment, the beetles were placed individually in 50 mm acrylic Petri dishes containing moistened Whatman No. 1 filter paper disks (300 µL distilled water) at the bottom. The dishes were then transferred to a Versatile Environmental Test Chamber MLR-35 1H (SANYO Electric Co., Ltd., Osaka, Japan) set to 20 °C, with a 16L:8D photoperiod and 80% relative humidity, for acclimation to the experimental conditions. During this short acclimation period, the beetles were not fed. In the present study, the sex and age of the beetles were not determined.

2.2. Treatments

Pesticide treatments were conducted in 50 mm acrylic Petri dishes, each lined with a 50 mm diameter Whatman No. 1 filter paper disk. The experiment included an untreated control (distilled water), Prot-, LCyh-, and Prot + LCyh mixture-treated variants, with 20 beetles in each group. Beetles in the pesticide treatment groups were exposed to both continuous contact exposure via pesticide-treated filter paper and repeated oral exposure through pesticide-contaminated food (Figure 1).

2.2.1. Preparation of Filter Paper Disks for Contact Exposure

To identify suitable low doses of Prot and LCyh for investigating their synergistic dynamics in a long-term experiment, 60 individuals of P. assimilis (three Petri dishes with 20 beetles each) were exposed to Prot at 0.1 of the maximum recommended field dose (MRD) (20 g ha⁻¹) and LCyh at 0.1 MRD (4 g ha⁻¹), as well as to a control treatment (distilled water). One hundred percent mortality was observed in the LCyh treatment within a week, suggesting that this dose was unsuitable for examining the synergy dynamics. Consequently, we subsequently reduced our experimental dose of LCyh to 0.01 MRD (0.4 g ha⁻¹). No mortality was observed in the Prot-treated and control beetles within a week.

Contact exposure was performed using pesticide-treated filter paper disks placed in acrylic Petri dishes. Each disk was treated with 300 µL of one of the following: distilled water (control), 0.1 MRD of Prot, 0.01 MRD of LCyh, or a mixture of both. Freshly treated disks were prepared every four days, and beetles were immediately placed individually in these dishes. The dishes were maintained in sealed plastic containers (30 × 20 × 10 cm) under controlled conditions (20 °C, 16L:8D, 80% RH) for the 64-day duration of the experiment.

2.2.2. Food Contamination for Oral Pesticide Exposure

For oral administration of LCyh, Prot, and their combination, the beetles were fed blue bottle fly larvae (Calliphora vomitoria) (third instar, 85–100 mg in body weight) that had been previously contaminated with the respective pesticides. Larval contamination was carried out in 50 mm Petri dishes lined with filter paper disks. The same pesticide doses and volumes of water were applied to the filter paper disks for both the contamination of the fly larvae and the contact exposure of the carabid beetles. The larvae were exposed to either pesticide-treated or untreated (distilled water) filter paper disks for 20 min, with 20 larvae per dish. This procedure yielded fly larvae contaminated with low doses of Prot, LCyh, or their combination, as well as the untreated control larvae. Then, the larvae were placed in tightly sealed plastic boxes and kept in a freezer at −18 °C until needed for beetle feeding. A couple of hours before each feeding event, the larvae were warmed up to 18–20 °C and cut into two pieces. In the pesticide-treated variants, the pesticide-treated and untreated larvae were alternatively presented to the carabid beetles for feeding for 2 h, as shown in Figure 1, with a four-day interval between feeding events. The control beetles were fed untreated larvae only. The consumption of clean and pesticide-treated food was measured by weighing individual beetles immediately before and after each feeding event using an analytical balance (AS 220/X, RADWAG Wagi Elektroniczne, Radom, Poland). The difference in weight was used to calculate the food consumption rate.

2.3. Video Recordings and Video Data Analysis

Video tracking of beetle locomotor activity was conducted prior to each feeding event and the replacement of Petri dishes and treated filter paper disks (Figure 1). When video tracking and feeding occurred on the same day, feeding, and the replacement of Petri dishes and filter paper disks always followed the video recording. The activity of up to 80 beetles was recorded simultaneously at a resolution of 1920 × 1080 pixels and a frame rate of 5 frames per second using two parallel video tracking systems. These systems comprised a USB Logitech HD Pro Webcam C920 camera (Logitech Inc., San Jose, CA, USA) and Debut Video Capture software (NCH Software, Greenwood Village, CO, USA). The Petri dish arenas were illuminated from above using diffused light provided by four MR16 LED lamps (12 V, 6 W, 400 lm, 3000 K). Illumination at the level of the Petri dishes containing the test beetles (3000 lux) was measured using a Digital Light Meter TES-1335 (TES Electrical Electronic Corp., Taipei, Taiwan). All video recordings were conducted for one hour, between 10:00 and 11:00 a.m., under laboratory conditions at a temperature of 20–21 °C. The measured locomotor activity parameters included the total distance moved (TDM, mm h⁻¹), velocity (mm s⁻¹), and time not moving (TNM, min), which were extracted offline from the recorded video files using EthoVision XT Version 11 software (Noldus Information Technology, Wageningen, The Netherlands). The following data profile settings were applied: bin width: 1 min; start velocity: 0.4 mm s⁻¹; stop velocity: 0.04 mm s⁻¹; and averaging interval: one sample.

2.4. Mortality Assessment

Mortality was assessed every 4 days up to 64 days after the first treatment, before feeding events. Beetles were considered dead if they showed no response to gentle stimulation with a dissection needle.

2.5. Statistical Analyses

Statistical analyses were carried out in R statistical software (version RStudio 2024.12.1+563). We compared the daily food consumption rates of the beetles in each pesticide-treated group to those of the control group over time using a general linear model (GLM; glm function in R) followed by pairwise comparisons with Fisher’s least significant difference (LSD) test implemented via the ‘agricolae’ package in R. Before fitting the GLM, the data were examined to ensure compliance with its key assumptions, including linearity, independence of observations, homogeneity of variances and normality of residuals. The normality of residuals was assessed using quantile–quantile (Q–Q) plots and the Shapiro–Wilk test. Homoscedasticity was evaluated through residuals versus fitted values plots and tested statistically using Levene’s test. Within each pesticide-treated group, the pesticide-treated and clean food consumption rates were compared using the Wilcoxon matched-pairs test. The total food consumption rates (combining clean and pesticide-treated food) over the 64-day experimental period were compared across the Prot-treated, LCyh-treated, and control groups using the Kruskal–Wallis test, because the data did not meet the assumption of normality. As the number of live beetles in the Prot and LCyh mixture-treated group dropped below three after 20 DAFT, their total food consumption rates were compared with that of the control group over the 20-day period using the Mann–Whitney U test. The body mass and locomotor activity parameters (TDM, Velocity, and TNM) of the beetles were compared across the pesticide-treated and control groups over time using a GLM (glm function in R) followed by Fisher’s LSD test.

We plotted Kaplan–Meier survival curves using the R packages ‘survminer’ and ‘survival’ within RStudio for each pesticide treatment and the control. We then tested for statistically significant synergistic interactions between the pesticides using a modified binomial proportion test for additivity (BPA) [41,72,73]. The BPA test applies the Bliss independence criterion [74], which is calculated using the following formula:

$$p_{AB}^{exp}= p_A+p_B-p_A \cdot p_B$$

where $p_A$ and $p_B$ represent the observed probabilities of mortality resulting from treatment substances A and B, respectively, and $p_{AB}^{exp}$ denotes the expected probability of mortality from the combination of A and B, assuming the two substances act independently.

To evaluate the potential synergistic effects of treatment substances A and B over the course of the experiment, we calculated the synergistic ratios (SR) at 4-day intervals immediately prior to each feeding event (Figure 1). The SR was defined as follows:

$$SR= {\displaystyle \frac{p_{AB}^{obs}}{p_{AB}^{exp}}}$$

where $p_{AB}^{obs}$ represents the observed probability of mortality resulting from the combined application of substances A and B, and $p_{AB}^{exp}$ represents the expected probability of mortality resulting from the combination of A and B. An SR > 1 indicates a synergistic effect, SR = 1 suggests additive or independent action, and SR < 1 indicates an antagonistic effect.

To assess and compare the relative speed and intensity of toxic effects across the pesticide treatments, the median lethal time (LT₅₀) and its 95% confidence intervals were estimated for each treatment group using the glm function, Abbot’s correction for control mortality, and a probit link in the MASS package in RStudio. Differences in the LT₅₀ values were considered statistically significant if their confidence intervals did not overlap.

3. Results

3.1. Food Consumption

Across all treatment groups, a general decline in food consumption was observed over time, with up to a four-fold reduction from approximately 20 mg to 5 mg per beetle (Figure 2). When comparing the food consumption rates of the LCyh- and Prot-treated beetles to those of the control group (Figure 2a–c), no consistent or considerable differences were detected. Only an occasional, statistically significant increase (p = 0.03) in food consumption occurred in the Prot-treated group (16.8 ± 0.9 mg per beetle) at 20 DAFT (Figure 2b) compared to that of the control group (13.9 ± 0.9 mg per beetle) (Figure 2a). In the LCyh-treated group, a small decrease (p = 0.01) in food consumption (11.7 ± 1.3 mg per beetle) was only observed at 16 DAFT (Figure 2c) compared to the control group (15.6 ± 0.9 mg per beetle). In contrast, a notable reduction (p = 0.03) in food consumption was observed at 12 DAFT in the group treated with the mixture of Prot and LCyh (6.3 ± 1.6 mg per beetle) (Figure 2d) compared to the control group (16.6 ± 1.3 mg per beetle).

Clean and contaminated food consumption rates within each pesticide-treated group were also compared (Figure 3). The consumption ranged between 9.1 ± 1.3 and 12.5 ± 3.1 mg per beetle, but the differences between clean and contaminated food consumption were not statistically significant (p = 0.07, p = 0.12 and p = 0.71 for the Prot-, LCyh-, and their mixture-treated groups, respectively). Interestingly, beetles in the Prot-treated group consumed significantly more food in total, with an increase of 15% compared to the control group (p = 0.003, Figure 4). In contrast, no significant difference in total food consumption was observed between the LCyh-treated and control groups (p = 0.39). Similarly, no difference in total food consumption was found between the Prot and LCyh mixture-treated group and the control group (p = 0.6).

3.2. Body Mass

At the start of the experiment, prior to the application of treatments, the body weight of the beetles ranged from 24.6 to 58.3 mg (mean: 38.8 ± 0.73 mg), with no significant differences observed between the treatment groups (p ≥ 0.05, Kruskal–Wallis test). Throughout the experiment, body weight increased in all treatment groups due to feeding, and significant differences were detected between the groups (Figure 5). In particular, the body mass of beetles in the Prot-treated group was significantly higher (p < 0.001) compared to the control group, with the percentage increase ranging from 1.0% to 21.4%. In contrast, beetles in the LCyh-treated group exhibited significantly lower body mass (p < 0.001), showing a percentage decrease of 2.0% to 12.9% compared to the control group. No statistically significant difference in body mass was observed between the Prot and LCyh mixture-treated group and the control group (p = 0.96).

3.3. Locomotor Activity

All three pesticide treatments significantly affected the locomotion of the beetles (F_{(3, 862)} = 8.13; p < 0.001). In the control group, at 8 DAFT, the track parameter TDM exhibited a pronounced, temporal peak, reaching up to 2303 ± 824 mm h⁻¹ (Figure 6a). By contrast, in all three pesticide-treated groups, the parameter values remained significantly lower (p < 0.001) than those of the control group, ranging from 193 ± 34 to 587 ± 154 mm h⁻¹, indicating that the beetles became strongly hypoactive. In the LCyh-treated group, only a few significant deviations from the control were observed thereafter. By contrast, beetles in the Prot-treated group (465.6 ± 30.4 mm h⁻¹) exhibited significantly lower TDM values (p < 0.001) compared to the control group (886 ± 71.8 mm h⁻¹), indicating that the Prot-treated beetles were predominantly hypoactive during the 64-day experimental period. In the Prot and LCyh mixture-treated group, at 11–16 DAFT, the TDM values exhibited a pronounced and significant peak, reaching up to 2435 ± 1307 mm h⁻¹, indicating that the beetles were in a state of strong hyperactivity during this period.

The track parameter TNM, which characterises insect mobility, indicated that beetles in both the Prot- and LCyh-treated groups were significantly less mobile than those in the control group throughout most of the 64-day experimental period (p < 0.001), demonstrating hypoactivity (Figure 6b). In contrast, beetles in the group treated with the Prot and LCyh mixture (17.6 ± 1.73 min h⁻¹) exhibited significantly higher mobility (p < 0.001) than those in the control group (20.8 ± 0.67 mm h⁻¹), indicating hyperactivity.

Velocity was the third track parameter used to quantify the locomotor activity of the beetles. At 3 days after the first treatment (DAFT), significantly higher velocities were observed in both the Prot-treated (0.65 ± 0.08 mm s⁻¹; p = 0.002) and LCyh-treated (0.79 ± 0.10 mm s⁻¹; p = 0.003) groups compared to the control group (0.57 ± 0.13 mm s⁻¹). However, no significant difference (p = 0.71) was found between the Prot and LCyh mixture-treated group (0.49 ± 0.67 mm s⁻¹) and the control group (Figure 6c). In contrast, at 8 DAFT, the velocities in all three pesticide-treated groups were significantly lower than in the control group (1.03 ± 0.24 mm s⁻¹), with values of 0.61 ± 0.07 mm s⁻¹ for Prot (p = 0.05), 0.53 ± 0.04 mm s⁻¹ for LCyh (p = 0.02), and 0.47 ± 0.06 mm s⁻¹ for the mixture (p = 0.006). Subsequently, the velocity values in both the Prot- (0.66 ± 0.03 mm s⁻¹) and LCyh-treated (0.85 ± 0.05 mm s⁻¹) groups remained significantly higher (p < 0.001) than those in the control group (0.52 ± 0.03 mm s⁻¹). In contrast, no significant difference (p = 0.57) was observed between the mixture-treated group (0.65 ± 0.14 mm s⁻¹) and the control (0.48 ± 0.03 mm s⁻¹). Overall, the three quantified track parameters consistently revealed significant differences in locomotor activity between the pesticide-treated and control beetles throughout the 64-day experimental period.

3.4. Mortality

Mortality in the control group remained very low, reaching up to 5% at 64 DAFT (Figure 7). All three pesticide treatments significantly increased insect mortality compared to the control group (Figure 7a and

Table 1. One-tailed binomial proportion z-test results comparing the mortality rates of the pesticide-treated groups to the control group (n = 20 per group).

Treatment	Final Mortality (%)	Z-Statistic	p
LCyh	50	z = 2.67	p < 0.01
Prot	75	z = 3.89	p < 0.001
LCyh + Prot mixture	100	z = 5.66	p < 0.0001

). During the first 32 DAFT, mortality rates in the Prot- and LCyh-treated groups were relatively low, at 15% and 10%, respectively. A rapid time-cumulative mortality increase occurred afterward, however, resulting in final mortality rates of 75% and 50%, respectively, but the difference between the treatments was not statistically significant (Chi-squared test, p > 0.05). The combined LCyh and Prot treatment resulted in complete mortality by 36 DAFT.

The LT₅₀ for the LCyh- and Prop-treated groups were 60.3 and 51.6 days, respectively; however, overlapping 95% confidence intervals indicate that this difference was not statistically significant (

Table 2. Median lethal time (LT₅₀) of P. assimilis beetles in pesticide-treated groups, estimated using probit analysis with Abbott’s correction for control mortality. Each group initially contained 20 beetles.

Treatment	LT50 (Days)	95% Confidence Interval
LCyh	60.26	49.52–71.01
Prot	51.57	39.89–63.25
LCyh + Prot mixture	12.21	3.24–21.19

). In the group treated with the LCyh and Prot mixture, the LT₅₀ was 12.2 days, which was significantly shorter than those in the LCyh- and Prot-treated groups.

We detected a significant synergistic interaction between Prot and LCyh during the 64-day experiment (Figure 7a). Early time points (12 to 20 DAFT) exhibited extremely high synergy ratios (up to 16), which gradually declined to 1.25 by the end of the experiment (60 DAFT), consistent with the expected increase in mortality observed in the BPA test (Figure 7b).

4. Discussion

In the present study, we assessed, for the first time, the sublethal and lethal effects of low-dose Prot, low-dose LCyh, and their combination on P. assimilis under prolonged, field-realistic conditions. Feeding is essential for the maintenance of all life processes. Therefore, it is of particular interest in ecotoxicological studies. In agricultural systems where pesticides are widely used, the feeding behaviour of insects may be adversely affected by repellents, antifeedants, or impaired olfactory perception when food sources are contaminated with sublethal doses of pesticides [16,70,75,76,77,78], including pyrethroid insecticides [16,79] and fungicides [31,80]. An antifeedant effect and reduced food consumption were also demonstrated in the predatory carabid Nebria brevicollis feeding on deltamethrin-treated aphids [81]. Insects have also been shown to reduce feeding after exposure to pesticides, even when they subsequently return to clean food [70,82,83,84,85,86]. Abnormalities in the clean food consumption rate serve as indicators of physiological stress resulting from pesticide exposure [87]. Insects often employ behavioural strategies as a primary defence mechanism to limit exposure to toxic substances. For example, they may detect olfactory or gustatory cues before lethal exposure, enabling avoidance. Such behavioural resistance has been documented for various classes of insecticides, including pyrethroids [88,89,90], and can include altered movement patterns, knockdown effects, or cessation of feeding and area abandonment.

By contrast, our study revealed that P. assimilis beetles exposed to prolonged, continuous contact and repeated oral exposure to LCyh (0.01 MRD) and Prot (0.1 MRD), applied either alone or in combination, consumed similar amounts of clean and pesticide-contaminated food. This suggests that the beetles were likely unable to distinguish pesticide-treated food from clean food and avoid it. A temporary reduction in feeding was observed in the LCyh and Prot mixture-treated group at 12 DAFT, coinciding with a pronounced locomotor hyperactivity peak. This likely reflects heavy toxic stress [91,92]. Intriguingly, however, beetles in the Prot-treated group consumed significantly more food in total compared to the control group. To our knowledge, there is no prior information on the impact of pyrethroid insecticides and fungicides on food consumption in other carabid species. Mauchline et al. [93] investigated the feeding responses of Pterostichus madidus, P. melanarius, and N. brevicollis to aphids contaminated with the organophosphate dimethoate in ‘no-choice’ and ‘two-choice’ tests, and found no evidence of avoidance or discrimination between treated and untreated prey. Under laboratory conditions, carabids readily attack prey and seeds contaminated with insecticides [93,94,95]. Although food consumption in the field is poorly quantified, both larvae and adults are known to be voracious feeders [60]. This implies that carabids could experience prolonged exposure to pesticides in treated fields through contaminated soil, plant seeds, and prey, leading to the bioaccumulation of insecticides and fungicides in their bodies [96].

Xenobiotics are known to exert direct effects on the locomotor system of arthropods, inducing changes in locomotor activity depending on concentration [91,92]. Altered locomotion, including hyperactivity or hypoactivity, reflects overall toxic stress and is a sensitive biomarker in ethotoxicological studies [70,97,98,99,100]. In our study, prolonged contact and repeated oral exposure to low doses of LCyh, Prot, or their combination, caused significant disturbances in the locomotor activity of P. assimilis. Prolonged periods of both hypoactivity and hyperactivity indicate that the beetles experienced substantial toxic stress throughout the entire 64-day exposure period.

In beetles of P. assimilis, exposure to low-dose LCyh significantly reduced body mass compared to the controls, aligning with previous findings that pesticide-induced stress reduces fitness through impaired growth and development [101,102,103]. Insecticides, including neonicotinoids, organophosphates, and pyrethroids, disrupt normal neural, endocrine, and cellular functions, increasing energy expenditure for detoxification. These detoxification processes, involving cytochrome P450 monooxygenases, glutathione S-transferases, and esterases, are metabolically costly and divert energy away from growth and development [16,104,105].

In contrast, toxic stress can be advantageous for insects in some cases, enhancing their fitness [106,107]. This phenomenon, known as hormesis, describes a biphasic response in which low levels of a stressor induce stimulatory biological effects, while higher levels are detrimental or lethal [106]. Such responses may confer adaptive advantages and can be transmitted to offspring through transgenerational mechanisms [107,108,109,110]. For instance, Leptinotarsa decemlineata exposed to low doses of deltamethrin exhibited increased body mass [109], and exposure to a fluazinam-based fungicide resulted in similar outcomes [110]. In line with these findings, the present study demonstrated that prolonged exposure to a low dose of Prot significantly increased body mass in P. assimilis beetles. An increase in body mass in response to pesticide-induced stress may provide fitness benefits, such as enhanced survival and developmental rates, which can be inherited by subsequent generations [109,110]. These transgenerational effects highlight the potential for hormetic responses to influence population dynamics and drive adaptation under sustained pesticide exposure.

Prolonged exposure to Prot (0.1 MRD), LCyh (0.01 MRD), and their combination significantly increased mortality in P. assimilis. During the first 32 DAFT, mortality among beetles exposed to Prot and LCyh remained below 20%; however, a sharp increase was observed thereafter, resulting in final mortality rates of 75% and 50% by 64 DAFT, respectively. The estimated LT₅₀ values (52 and 60 DAFT) indicate delayed lethality, likely attributable to pesticide bioaccumulation [30,32,33] and time-reinforced toxicity (TRT), a phenomenon in which toxicity increases with prolonged low-dose exposure [30,33,111]. TRT has been extensively studied in bees, and recent EFSA guidelines recommend its consideration in risk assessment [111,112]. However, TRT has yet to be systematically assessed in carabids. If ERAs for pesticides do not account for delayed mortality effects, the risks posed to non-target carabid beetles may be substantially underestimated.

We observed a strong synergistic effect between LCyh and the SBI fungicide Prot, with maximum synergy ratios of 13–16 observed at 12–20 DAFT. This effect likely results from the Prot-induced inhibition of cytochrome P450 monooxygenase activity, impairing the detoxification of pyrethroids [73]. In other insect species, certain SBI fungicides have been shown to markedly increase pyrethroid toxicity, while others have elicited only minor effects. For example, in an acute toxicity assay, the combination of the pyrethroid bifenthrin with the SBI fungicide difenoconazole led to a slight increase in bumblebee mortality, whereas co-application with the SBI fungicide myclobutanil significantly enhanced mortality, yielding a synergy ratio of 11 [73]. Similarly, LCyh exhibited a strong synergistic effect on the acute mortality of honeybees when combined with the SBI fungicide propiconazole, resulting in a synergy ratio of 16.2 [113]. Therefore, the synergistic interaction observed in P. assimilis is consistent with previously reported pesticide–fungicide interactions and may be considered strong. The total lethality observed by 36 DAFT, with an LT₅₀ at 12.2 DAFT for the LCyh and Prot mixture, suggests that these pesticides, used jointly or sequentially, could drastically reduce P. assimilis numbers during the growing season in agricultural fields.

In conclusion, for the first time in this study, we demonstrated in P. assimilis that prolonged exposure to Prot and LCyh at field-relevant low doses may result in high levels of delayed mortality over the course of a growing season. This also represents the first evidence of pyrethroid–SBI fungicide synergism in a carabid species under a field-realistic exposure duration, potentially leading to total mortality within a month. If delayed mortality and synergistic effects between pesticides are not adequately captured by standard ERAs [28,29], the toxicity of pesticides to non-target carabid beetles may be significantly underestimated.

However, some limitations of our study should be acknowledged. Although several studies have demonstrated that insects may exhibit varying levels of susceptibility to pesticide exposure depending on their sex and age [114,115,116], these factors were not accounted for in the present study. Some bias may have arisen due to the omission of these important confounding variables. Understanding these patterns is essential for optimising pest control strategies and minimising non-target impacts. While our exposure regimes were designed to mimic realistic field scenarios, the experiments were conducted under controlled laboratory conditions, which inherently lack the ecological complexity of natural environments. In the field, multiple biotic and abiotic factors—such as weather variability, microhabitat heterogeneity, predator–prey interactions, and resource availability—can influence pesticide exposure, detoxification, and insect behaviour. These factors may either mitigate or exacerbate the effects observed under laboratory conditions. Furthermore, our findings are based on a single carabid species, and although the cytochrome P450 system is widespread and important in insects, synergistic effects are often species-specific, influenced by interspecific differences in metabolism, behaviour, and ecology [34,117]. Thus, while similar mechanisms may exist across carabid beetles, further empirical testing under semi-field or field conditions is essential to confirm the broader ecological relevance and generalisability of our findings.