Climate change may differentially affect the efficiency of insecticide treatments against pests and beneficial insects through a wide range of factors in the field. This study first wants to assess the effect of temperature on three insecticide toxicity in the controlled condition in the laboratory to avoid the complex interactions between factors, including the behaviour or ecology of arthropods in natura. It provides a first step of whether the efficacy of these insecticides may change with the climate using the example of a pest and two of its natural enemies used in IPM programmes. Indeed, the consequences of climate change on biological control caused by a differential influence of temperature on insecticide toxicity between pests and their natural enemies have never been considered, although such an influence is known to be species-specific. This topic is especially relevant for organic farming and IPM strategy, which rely on both natural enemies and insecticides. The worst scenario is that insecticide toxicity decreases with global warming for a pest but increases for its natural enemies.
In this study, we observed that higher temperature (i) decreased the toxicity of emamectin and spinosad but did not influence the toxicity of chlorantraniliprole against the pest C. pomonella, while it (ii) increased the toxicity of these three insecticides on the parasitoid, M. ridens, and (iii) increased the toxicity of emamectin on the predator, F. auricularia. Our results suggest a possible problem in future C. pomonella population control in the field because of a decreased efficiency of insecticides on the pest coupled with the opposite effect on natural enemies’ populations.
Our results on C. pomonella agree with those reported for spinosad on another pest species, the cotton mealybug Phenacoccus solenopsis Tinsley (Mansoor et al. 2014). However, our results on C. pomonella contrast with those obtained on P. solenopsis (Mansoor et al. 2014) and Plutella xylostella L., on which toxicity of emamectin increased with temperature as for chlorantraniliprole on P. xylostella (Li et al. 2004; Teja et al. 2018). Indeed, interspecific differences in temperature-mediated insecticides’ toxicity between pests have been frequently reported (Boina et al. 2009), although underlying mechanisms still have to be investigated.
Compared to pests, few studies have been conducted on natural enemies. Still, increased toxicity at high temperatures was reported for chlorantraniliprole on the parasitoid Bracon nigricans Szépligeti (Abbes et al. 2015), as we observed for M. ridens when the substance was ingested. However, in our study, spinosad toxicity increased with temperature for M. ridens. At the same time, it was not influenced by temperature in B. nigricans (Abbes et al. 2015) and decreased with increasing temperature in the lacewing Chrysoperla carnea Stephens (Mansoor et al. 2015). The same interspecific differences for pests were also observed for natural enemies.
The positive relationship between toxicity and temperature observed for the three substances in M. ridens and emamectin ingested by F. auricularia may result from several mechanisms that should be investigated in further studies. First, high temperature may influence detoxification enzymes (Yan et al. 2013; Zhang et al. 2015; Liu et al. 2017) or Heat Shock Protein (Ge et al. 2013; Su et al. 2018) activities and expression. Second, metabolic rate increases with temperature (Brown et al. 2004), consequently increasing food consumption, insect locomotion (Gillooly 2001; Medrzycki et al. 2010), and substance penetration in the insect's body (Boina et al. 2009).
This study was conducted in a simplified system where insects are "forced" into contamination. In natura, other complex behavioural or ecological parameters might limit contact between insects and insecticide. For example, F. XXXuricularia is a nocturnal species that hunt at dusk, which could allow them to avoid diurnal substances spraying (Vancassel 1973). Moreover, some parasitoid species also avoid hosts resistant to some insecticides to minimise contamination (Alfaro-Tapia et al. 2021).
Whether the two natural enemies ingested the insecticides or were exposed by contact, M. ridens was more sensitive to substances than F. auricularia. The three insecticides caused 100% mortality in M. ridens above 28°C at the authorised field dose, while no mortality was detected in F. auricularia. This interspecific difference may be due to differences in the composition or thickness of their cuticle (Fernandes et al. 2010) and/or insecticide penetration, depending on the affinity between the cuticle and the substance (Leite et al. 1998). The smaller size of M. ridens compared to F. auricularia may also account for this difference, as the specific target area of the insecticides decreases with increasing body size, resulting in reduced insecticide exposure (Picanco et al. 1997; Bacci et al. 2007).
Such differences may also explain that toxicity was mediated by temperature for the three insecticides for M. ridens, while it was only true for emamectin in F. auricularia. These interspecific differences in the relationship between temperature and insecticides’ toxicity underline the need to study specifically each pest-natural enemy’s system. This approach would allow for selecting insecticides that will be the most reliable with global warming. Ideally, insecticides used in IPM programmes and organic agriculture should be less toxic to natural enemies than to target insect pests (Zhao et al. 2012) and remain harmless as temperature increases.
From our results, spinosad was as toxic to the codling moth as it was to the European earwig and was even more toxic to M. ridens, regardless of the temperature. It was also more toxic to natural enemies than emamectin and chlorantraniliprole, considering the tiny dose needed to kill 50% of parasitoids compared to the two other insecticides. Lethal and sublethal effects of spinosad and other spinosyns on beneficial arthropods have been previously reported in several studies (Biondi et al. 2012; Abbes et al. 2015), and yet, this product is authorised and widely used in organic farming while emamectin and chlorantraniliprole are not (Biondi et al. 2012). Moreover, our results indicate that higher temperatures emphasise differences in spinosad toxicity between the two trophic levels. Among the insecticides evaluated, it is likely to cause problems managing codling moth populations under high-temperature conditions. The combination of spinosad with strategies of regulation based on natural enemies should be avoided. Its approval for organic production may raise questions in the future considering its increased toxicity on some beneficial organisms as temperature increases.
Surprisingly, chlorantraniliprole may be the most promising insecticide in a warming world for strategies relying partly on biological control by conservation. The toxicity of this substance indeed remained very effective at high temperatures against C. pomonella while it did not cause any mortality on F. auricularia, regardless of temperature or exposure method. Moreover, it was relatively safe against M. ridens when parasitoids ingested the substance. Despite the product's increased toxicity at high temperatures, the lethal doses for 50% of individuals remained high (equivalent to 10–50× the recommended field dose). However, the substance caused 100% mortality during contact for M. ridens. Thus, this substance may be appropriate for classical biological control relying on the introduction of parasitic wasps, although field experiments should confirm our laboratory observations.
For both natural enemies, the LD50 remained very high at low temperatures for emamectin, indicating that a high quantity of this insecticide should be applied to kill 50% of the individuals. These results are consistent with the literature on other natural enemies for this substance (Argentine et al. 2002; Khan et al. 2018). Emamectin has been recommended in IPM strategies for a long time because of its low toxicity on beneficial organisms, high selectivity for pests and rapid environmental degradation (Argentine et al. 2002; López et al. 2010). However, we observed that its toxicity increased rapidly with temperature for F. auricularia and M. ridens and decreased for the codling moth. In this sense, its efficacy may decrease with global warming despite being the most effective insecticide against C. pomonella among the three tested in terms of effective doses at high temperatures.
Contrary to chlorantraniliprole, emamectin toxicity on F. auricularia increased strongly with temperature. Alternating the use of these insecticides in conservation biological control, with preferential use of emamectin during the colder periods (e.g., early spring) and the use of chlorantraniliprole (e.g., during the summer) could limit the risks of resistance appearance in pest’s populations while limiting the impact of chemical substances on the European earwig. However, its use in CBC programmes involving the introduction of M. ridens should be avoided considering its toxicity at high temperatures.
The present study highlights opposite cross effects of temperature and insecticides (spinosad, emamectin, chlorantraniliprole) on C. pomonella and two of its natural enemies, one exotic (M. ridens) and one native (F. auricularia) in southern France. According to the future climate change predictions, we recommend using these insecticides to preserve natural enemies associated with C. pomonella. The use of spinosad should be avoided, while emamectin can be used in conjunction with the conservation of F. auricularia. Only chlorantraniliprole appears suitable for controlling C. pomonella, conserving F. auricularia and establishing M. ridens. The present study remains a simplified but essential tool to understand the cross effects of temperature and three insecticides on the target pest and two of its natural enemies, allowing us to propose an adaptation of strategies relying on insecticides in apple orchards in a changing climate. We focused on the mortality of natural enemies, but the effects of sublethal exposures, which affect insects’ behaviour and life history traits, should also be investigated since they may be critical to the long-term conservation and/or establishment of natural enemies (Saber 2011; Poorjavad et al. 2014).