Macronutrient regulation in adult spotted‐wing Drosophila and nutritional impacts on susceptibility to three commonly‐used insecticides

Nutrition provides the resources necessary to carry out all the physiological functions organisms require. This has important implications for insect pest management, as polyphagous species can experience considerable nutritional variability, which may impact their response to management tactics. Many studies have shown that dietary protein and carbohydrates play a primary role in insect fitness and that insects regulate their intake of these macronutrients to reach an optimal balance or intake target. Importantly, this intake target can vary across developmental stages, gender, and environmental conditions as physiological requirements change. In this study, we determined the intake target for adult spotted‐wing Drosophila, Drosophila suzukii, an invasive polyphagous fruit pest. We found that adult D. suzukii regulate their intake to meet a 1:4–1:5 protein‐to‐carbohydrate ratio, which is slightly more carbohydrate‐biased than the larval intake target. We also examined the effect of nutritional state on susceptibility to three commonly‐used insecticides by rearing flies on diets with different field‐relevant protein‐to‐carbohydrate ratios and exposing adults to an LC50 dose of zeta‐cypermethrin, spinetoram, or pyrethrin. Overall, female flies had higher survivorship than males, and diet effects varied across insecticides. Diet had no impact in the spinetoram treatments, while the extreme carbohydrate‐ and protein‐biased diets produced the lowest survivorship and reproductive performance in the zeta‐cypermethrin treatments. Survivorship increased with dietary protein in the pyrethrin treatments but reproductive performance did not differ across diets. Our results suggest that tactics, which reduce fly access to dietary protein sources, such as sterilisation, could increase fly susceptibility to insecticides.


INTRODUCTION
All physiological processes are constrained by the availability of compounds needed to fuel their underlying biochemical reactions.In this way, an organism's nutritional state can strongly impact its ability to carry out necessary physiological responses.This relationship makes nutrition a strong mediator of plasticity and highlights the broad relevance that nutritional ecology has to a multitude of other disciplines (Deans et al., 2016;Raubenheimer et al., 2009).In order to understand an organism's nutritional ecology, one must first establish a nutritional baseline of what is required for optimal function.Just as important, is understanding the physiological repercussions of not meeting these requirements and the likelihood of this occurring under different field conditions.This pursuit has led to the development of many nutritional paradigms that utilise different physiological currencies, such as energy or single nutrients.The geometric framework (GF) of nutrition has been successful in showing that the balance of multiple nutrients, rather than the absolute amounts of single nutrients, has the greatest impact on animal fitness (Raubenheimer et al., 2009;Simpson & Raubenheimer, 1995, 2012).In particular, the balance of macronutrients, specifically protein and carbohydrates, have been shown to play a primary role in animal fitness and that animals regulate their intake of these macronutrients to reach an optimal balance, termed an intake target.Importantly, this intake target can vary across developmental stages, gender, and environmental conditions as physiological requirements change (Behmer, 2009;Simpson & Raubenheimer, 2012).
Several GF studies have demonstrated the importance of macronutrient balance in the management of economically-important insect pest species.For instance, Cease et al. (2012Cease et al. ( , 2015) ) established a link between excessive grazing and locust outbreaks in China, mediated by changes in soil nitrogen that shift plant macronutrient profiles towards the locusts' optimal balance.Similarly, Le Gall et al. (2020Gall et al. ( , 2020) ) showed that crop fertilisation can also have strong nutritional implications for locust populations in West Africa.In addition to land management impacts on pest nutritional ecology, other studies have highlighted the important effects that nutrition can have on pest control efficacy.Nutrition can impact the effectiveness of both microbial and insecticide control tactics, via physiological and behavioural factors.Many studies have shown that dietary protein plays an important role for caterpillars in mounting effective immune responses to bacterial or fungal infections (Cotter et al., 2011;Cotter and Al Shareefi, 2022;Lee et al., 2006;Wilson et al., 2019).For fruit flies, however, it appears that feeding on low-protein diets increases survival during infection (Ponton et al., 2020).Diet can also affect insect susceptibility to pesticides (Deans et al., 2016;Orpet et al., 2015;Shikano & Cory, 2014a).Deans et al. (2017) showed that the median lethal dose of Bt endotoxin Cry1Ac could be increased by almost 100-fold for cotton bollworm larvae by simply changing the proteinto-carbohydrate (P:C) ratio of their rearing diet.Additionally, insects can use nutrition to mitigate infection and/or toxin exposure by altering their feeding behaviour.Povey et al. (2009) showed that Spodoptera exempta larvae infected with Bacillus subtilis bacteria increase their intake of protein-rich foods as a method of self-medication, and Povey et al. (2014) found a similar response upon viral infection.Orpet et al. (2015) found that cotton bollworm larvae were more willing to consume Bt toxin when it was present in protein-rich diets.It is also well-established that H. zea larvae can detect Bt toxins and actively avoid feeding on transgenic plants expressing them (Braswell et al., 2019;Gore et al., 2002;Orpet et al., 2015).
While the threat of genetic resistance poses perhaps the most eminent threat to insecticide viability, the role that geneby-environment interactions play in reducing efficacy is gaining more attention (Brevik et al., 2018;Deans et al., 2016;Guedes et al., 2009;Piiroinen, S., Lyytinen, A. & Lindström, 2013).Not only can plasticity locally reduce the effectiveness of insecticides, but the intermediate phenotypes they produce may play a key role in the subsequent evolution of insecticide resistance (Brevik et al., 2018;Pigiluicci, 2005;Pigliucci et al., 2006).Because of this, it is important to explore the role that mediators of plasticity, such as nutrition, play in pest dynamics in agricultural systems.These investigations are particularly probative for polyphagous generalist species that feed on a wide variety of hosts and contend with extreme nutritional variability.One such pest is spotted-wing Drosophila (Drosophila suzukii), an invasive fruit fly from East Asia that has successfully spread throughout the continental United States in less than 5 years after its initial detection in California in 2008.It has also become an established pest throughout Europe, North and South America (Asplen et al., 2015;Hauser, 2011;Lee et al., 2011;Tait et al., 2021), most recently Africa (Boughdad et al., 2021).Moreover, D. suzukii has already developed resistance to several commonly used insecticides, where foliar applications are targeted at the adult stage (Deans & Hutchison, 2022a;Ganjisaffar et al., 2022;Gress & Zalom, 2019;Van Timmeren et al., 2018).
Drosophila suzukii is highly mobile and feeds on hosts from over 25 different plant families, including native and invasive species throughout the United States (Asplen et al., 2015;Lee et al., 2011; MU Spotted wing Drosophila host plant list, 2017).Female D. suzukii possess a serrated ovipositor that allow them to lay eggs in ripening intact fruit, rather than split or over-ripe fruit that other drosophilid species utilise (Asplen et al., 2015;Lee et al., 2011).This unique niche, along with the ability to overwinter in some areas, has allowed them to quickly establish in fruit-producing regions, with yield losses being as high as 80%-100% (Bolda et al., 2010;DiGiacomo et al., 2019;Farnsworth et al., 2017), given the stringent damage thresholds on fresh market crops.In the Western U.S., economic losses to D. suzukii are estimated at $500 million a year (Goodhue et al., 2011).The fruit hosts targeted by D. suzukii vary substantially in both P:C ratio and total macronutrient content (P + C), with P:C ratios ranging from 1:23 in grapes to 1:3 in blackberries and macronutrient contents from 40 to 200 g/L (Deans & Hutchison, 2021;Young et al., 2018).The nutritional ecology of D. suzukii is further complicated by their dependence on microbial resources for dietary protein (Deans & Hutchison, 2021;Hamby et al., 2012;Young et al., 2018).Larval nutrition has been studied more extensively than adult nutrition (Deans & Hutchison, 2021;Silva-Soares et al., 2017;Young et al., 2018), and while intake target does exist for D. suzukii larvae (Silva-Soares et al., 2017), an adult target for D. suzukii has not yet been determined, likely due to difficulties in measuring adult diet consumption.Adult flies are much more mobile than larvae and likely have different nutritional requirements.Adult nutrition is also more complex because the host fruits that maximise adult fitness and larval fitness are not the same (Lihoreau et al., 2016;Olazcuaga et al., 2019;Silva-Soares et al., 2017).Consequently, oviposition preference may draw adult flies away from optimal hosts.Adult flies also deal with more environmental variability than larvae, likely leading to more nutritional plasticity in adults (Shu et al., 2022).Extrapolating from larval choice tests, the intake target for D. suzukii falls around a 1:3 P:C ratio (Deans & Hutchison, 2021;Silva-Soares, 2017), and similar studies with D. melanogaster documented a similar intake target (between 1:2 and 1:4) (Lee, 2015;Lee et al., 2008;Rodrigues et al., 2015).Given that current management tactics are focused on adult (Tait et al., 2021), rather than larval, control, understanding adult nutritional requirements is essential for determining the potential impacts of host variability on insecticide efficacy.
In this study, we used choice bioassays to document macronutrient regulation in D. suzukii adults and determine their optimal balance of dietary protein and carbohydrates.We then used this information, along with macronutrient data for different fruit hosts, to design artificial diets that encompass the range of P:C ratios and P + C contents encountered by flies in the field.We measured survivorship, adult lifespan, and female reproductive performance of flies reared on these diets and exposed to a median lethal dose of three commonly used insecticides in the Midwest U.S. region, including the following: zeta-cypermethrin (Mustang Maxx ® ), spinetoram (Delegate ® ), and organic pyrethrin (Pyganic ® ) (e.g., Deans & Hutchison, 2022a).These data will be useful for understanding (1) how host nutritional variability may impact fly nutritional state and (2) how nutritional state may affect fly susceptibility to insecticidal control methods.These questions are particularly pertinent for polyphagous species that forage in a highly heterogenous nutritional landscape, such as D. suzukii.Further documenting the extent to which nutrition mediates changes in insecticide susceptibility is important for advancing our understanding of plasticity and the role that may play in pest population dynamics.

Fly culture
All flies used in the experiments were from a lab colony established in 2018 from infested raspberry samples collected from the University of Minnesota's UMORE Station in Rosemount, MN.They were reared on a standard cornmeal-based oligidic diet in narrow polystyrene vials (Genesee Scientific Corporation) and transferred to a new diet every 2-3 days.Colonies were kept in a walk-in chamber at ambient lab temperature, which ranged from 20 to 22 C, under a 14:10 light-daycycle (Tran et al., 2020).

Choice bioassay protocol
Eight liquid diets were prepared using the recipe in Lee et al. (2008).
Diets consisted of distilled water, hydrolysed yeast (65% protein, 24% carbohydrates, and 11% fibre), sucrose, and phosphoric acid (0.01%) and propionic acid (0.1%) as antifungal agents.The diets encompassed four different P:C ratios, 1:20, 1:2, 1:1, and 2:1, and two different macronutrient concentrations, 45 and 180 g/L.These parameters were selected based on the P:C ratio and macronutrient concentrations found in various fruit utilised by D. suzukii in the field (Deans & Hutchison, 2021;Young et al., 2018).The relative amounts of yeast and sucrose were varied to achieve the desired diet P:C ratio, while the total amount of yeast and sucrose was varied to achieve the desired diet macronutrient concentrations.Diet was stored under refrigeration until used.
A choice study was used to determine whether adult flies actively regulate their intake of dietary protein and carbohydrates and what their self-selected intake target, or optimal P:C ratio, is for male and female flies.For the choice study, newly-eclosed unmated adult flies were allowed to simultaneously feed on two diets with different P:C ratios for a period of 48 h.Three different diet pairings were tested to encompass a broad nutritional space: 1:20 and 1:2, 1:20   and 1:1, and 1:20 and 2:1.These diet ratios pairings were tested at both macronutrient concentrations, resulting in six total treatments with 12-15 male and female fly replicates each.Figure 1 shows the diet pairings across nutritional space and the experimental setup.Choice tests were conducted in individual 20 mL glass scintillation vials.A volume of 150 μL of each diet was placed inside of its own receptacle made out of the lids of 1.5 mL Eppendorf tubes and placed inside the vial cap.Different colour lids were used to identify specific diet P:C ratios and were randomly assigned to reduce potential behavioural confounds associated with the colour attraction/avoidance.A small metal bead was placed inside each receptacle so that the flies could drink from the outer edge of the lid but could not climb into the diet vessels.This was done to prevent diet loss from contact with the flies or spillage.Consumption of the liquid diet was determined gravimetrically.The individual diet receptacles were weighed at the start of the choice test and then 48 h later to determine how much of each diet had been consumed.The amount of protein and carbohydrates consumed by each fly was then calculated by multiplying the amount of each diet consumed by its macronutrient profile (% P and %C).All vials were housed in a 30 L clear plastic container with wet filter paper to maintain high humidity and minimise evaporation throughout the experiment.This container was then placed in a growth chamber set at 27 C with a 14:11 L:D cycle and >65% humidity.The higher experimental temperature of 27 C was chosen to replicate common summer temperatures in Minnesota.Over a period of 17 days three identical trails were performed containing 10-18 replicates per treatment.Fly mortality, as well as average consumption of each diet, the amount of protein and carbohydrates consumed, and the P:C ratio of total diet consumed was calculated.

Dose-response assays
We tested responses to three insecticides commonly used to control

Diet-insecticide experimental protocol
To determine the potential impact of fly nutritional state on susceptibility to three commonly-used insecticides, we reared flies on diets ranging in P:C ratio and macronutrient concentration and exposed them to sub-lethal doses of insecticide.Because these experiments lasted the entire lifespan of the flies, and we were not measuring consumption, we used the agar-based artificial diet described in Lihoreau et al. (2016) to create our experimental diets.These diets consisted of a 1:1 ratio of whey and casein, sucrose, and nutritional yeast as the primary macronutrient sources.Vanderzant vitamin mixture provided micronutrients, propionic acid and methylparaben were added as antimicrobials, and all ingredients were set in 2% agar.The amounts of yeast, Vanderzant mix, and anti-microbials were standardised across diets, while the relative proportions of whey/casein and sucrose were adjusted to achieve five diets with the same total macronutrient content (45 g/L) but different P:C ratios: 1:12, 1:6, 1:3, 1:1, and 3:1.These diet P:C ratio were selected based on the results of our choice study, the range of P:C ratios found in different fruit hosts utilised by D. suzukii (Deans & Hutchison, 2021;Young et al., 2018), and the reported intake target for larval D. suzukii (1:3) documented in Silva-

Soares et al. (2017).
Newly-eclosed (within 24 h) adult flies were transferred, under CO 2 anaesthetisation, from colony vials to individual vials containing their respective diet.Ten male and ten female replicates were used for each treatment and all flies were housed individually.Three days later they were exposed to the LC 50 amount of one of four insecticide treatments: control (no exposure), zeta-cypermethrin (0.2 ppm), spinetoram (20 ppm), or pyrethrin (0.1 ppm).Exposure was carried out by moving flies into 20 mL glass scintillation vials coated with insecticide for 4 h.Control flies were placed in vials with no insecticide coating to standardise for CO 2 anaesthetisation and vial conditions.After the exposure period, all flies were transferred back to their individual rearing vials.Flies were only exposed to insecticides for one 4-h period.
Twenty-four hours after exposure, two male flies from the lab culture were added to each female vial for mating (experimental males were unmated).Mating males were replaced as needed throughout the experiment.Mortality was checked daily and flies were transferred to fresh diet as needed.All female vials were saved and the number of emerging pupae was recorded weekly.Vials were maintained/ collected throughout the entire lifespan of each adult.
F I G U R E 1 Bar graph showing consumption (±SE) in grams of liquid diet for each diet in the paired-diet treatments at each total macronutrient concentration.Diet 1 refers to the first diet listed in the treatment label on the x-axis (1:20 in all treatments).Diet 2 refers to the second diet listed (variable depending on the treatment).Different letters indicate significant differences in diet consumption, with post-hoc letters for diet 1 being lowercase and diet 2 uppercase.Asterisks indicate significant differences in consumption between diets at p ≤ 0.05.

Data analysis
For the choice test, we used a three-way ANOVA to determine whether total consumption differed between diets (1:20 vs. 1:2, 1:20 vs. 1:1, and 1:20 vs. 2:1), paired-diet treatments (1:20-1:2, 1:20-1:1, and 1:20-2:1), diet macronutrient concentrations (45 and 180 g/L), and across fly sex (male or female).This test showed whether the flies were feeding non-randomly on the two diets offered to them within each diet pairing and whether consumption differed across diet pairing treatments.A three-way ANOVA was also used to determine if the P:C ratio of consumed diet varied across paired-diet treatments, macronutrient concentrations, and fly sex.This test showed whether the flies were actively regulating their macronutrient intake to meet a specific intake target.A Tukey's test was used for all post-hoc comparisons with a Bonferroni correction.A one-sample Wilcoxon Signed Rank Test (non-parametric) was also used to determine if the median P:C ratio of consumed diet differed from the line of random feeding, which would result if flies were feeding randomly across diets.Data were rank-transformed when necessary to meet normality assumptions.
For the diet-insecticide experiments, we used a probit analysis to calculate the LC 50 values for each insecticide dose-response curve.
Each diet and sex effects were analysed separately for each insecticide.A two-way ANOVA was used to assess differences in lifespan across sex and diet treatments.A Kaplan-Meier survival analysis (Mantel-Cox) was used to explore differences in time-dependent mortality for each sex and diet treatment.A one-way ANOVA was used to determine if female reproductive performance (pupae/female) varied across diet treatments.Tukey HSD tests were used as post-hoc tests for all significant ANOVA effects or interactions.All statistics were done using IBM SPSS (v.27).

Nutrient regulation
Diet consumption was impacted by many factors.Table 1 shows consumption of each diet across treatments.Table 2 shows that there were several significant interactions between factors, including a diet treatment*macronutrient concentration, a diet*treatment, and a sex*macronutrient concentration interaction.Overall consumption was lower for the 180 g/L diets than for the 45 g/L diets (Table 1).
Figure 1 shows that across paired-diet treatments macronutrient concentration did not impact the consumption of diet 1, which had a P:C ratio of 1:20 in all treatments.For both macronutrient concentrations, consumption of the 1:20 was significantly lower in the 1:20 versus 1:2 diet treatment but was comparable across the 1:20 versus 1:1 and 1:20 versus 2:1 pairings (Figure 1).Consumption of diet 2, which varied in P:C ratio depending on the treatment, was different across macronutrient concentrations but displayed the same consumption pattern (Figure 1, SI Table 1).For the 180 g/L diets, the consumption of the 1:2 diet was significantly greater than either the 1:1 and 2:1 diets at each macronutrient concentration (Figure 1, SI Table 1).There were differences in consumption between males and females for the 45 g/L diets but not the more concentrated 180 g/L diets (Table 2; SI Table 1).Female flies consumed a significantly larger total amount of the 45 g/L diets than males T A B L E 1 Summary of diet consumption data for the choice experiment.T A B L E 2 Three-way ANOVA results for differences in consumption between each diet (1:20 vs. 1:2, 1:20 vs. 1:1, and 1:20 vs. 2:1), across diet pairings (1:20-1:2, 1:20-1:1, and 1:20-2:1), diet macronutrient concentration (45 and 180 g/L), and between fly sex (male or female).(0.057 g for females compared to 0.055 g for males).Figure 1 shows that the relative consumption of diet 1 and 2 was significantly different in each diet-pairing treatment, indicating that feeding across diets was non-random (Table 2).
These differences in diet consumption resulted in differences in protein and carbohydrate intake.Figure 2 shows that the average P:C ratio of the diet consumed across all treatments was more carbohydrate-biased than expected by random feeding.The results of a one-sample Wilcoxon Signed Rank test, shown in Table 3, confirm that the median P:C ratio for each treatment was significantly lower than the ratio for lines of random feeding for both diet macronutrient concentrations.Table 4 shows that P:C intake varied across paireddiet treatments and diet macronutrient concentration but not sex.All paired-diet treatments had different intake ratios, indicating that they did not converge on a single shared intake target.Flies fed on the more concentrated 180 g/L diets had significantly lower P:C ratios than the same treatments on the 45 g/L diets.The intake P:C ratios for the 1:20 versus 1:2 pairing ranged from 0.206 (1:4.8) to 0.231 (1:4.3), while the 1:20 versus 1:1 and 1:20 versus 2:1 treatments ranged from 0.272 (1:3.7) to 0.306 (1:3.3) and 0.326 (1:3.7) to 0.408 (1:2.5)respectively.The average P:C ratio across all diets was 0.293 (1:3.4).

Lifespan and survivorship
There were several significant interactions between insecticide, diet, and sex on fly lifespan.Table 5 shows significant insecticide*diet, insecticide*sex, and diet*sex interactions.Overall, lifespan was most strongly impacted by exposure to spinetoram, with both male and female flies showing significantly shorter lifespans than controls and the other insecticide treatments (Figure 3a; SI Table 2).Insecticide F I G U R E 2 Carbohydrate and protein intake for flies across paired-diet treatments.Dashed lines indicate the lines of random feeding for each treatment.Solid lines show the limits of nutrient space (1:20 and 2:1 diets).Empty circles show consumption for the 45 g/L diets and filled circles for the 180 g/L diets.
T A B L E 3 Results from a one-sample Wilcoxon Signed Rank test to determine if the median P:C ratio of consumed diet for each paired-diet treatment was different from the median for the line of random feeding, shown in parentheses.exposure had no impact on fly lifespan in the pyrethrin treatment and only reduced male lifespan in the zeta-cypermethrin treatment (Figure 3a; SI Table 2).Male and female lifespans were similar across diets, except for the 1:3 diet where female longevity was significantly longer than males (Figure 3b, SI Table 3).Statistically-significant diet effects varied across insecticide treatments, but there was a common trend of lifespan increasing with dietary protein-content.For controls, flies on the carbohydrate-biased 1:12 had lower lifespans than the 1:6, 1:3, and 1:1 diets, while flies on the protein-biased 3:1 diet exhibited intermediate longevity (Figure 4, SI Tabel 4).In the zetacypermethrin treatments, the 1:1 diet produced the longest lifespan, with the 1:6, 1:3, and 3:1 diets being similar.The 1:12 diet again produced the lowest lifespan.Diet had no impact on lifespan for the spinetoram or pyrethrin treatments (Figure 4, SI Table 4).
A Kaplan-Meier survival analysis was done to detect differences in the time course of mortality for treatment populations.Table 6 shows the mortality rate varied significantly between insecticides and diet treatments but not between the sexes.Significant diet effects were found in all insecticide treatments except spinetoram (Table 7).
For the controls, the mortality rate was the highest on the extreme 1:12 and 3:1 diets and similar across intermediate P:C ratios (Figure 5a, SI Table 5).A similar pattern was found for the zetacypermethrin treatments, with the 1:3 and 1:1 diets showing delayed mortality compared to the other diets (Figure 5b, SI Table 5).The pyrethrin treatments showed a different pattern, where the mortality rate decreased as dietary protein increased.Here, the 1:3, 1:1, and 3:1 diets maintained higher survivorship than the 1:12 and 1:6 diets (Figure 5d, SI Table 5).

Female reproduction
Female reproduction, as measured by the number of pupae produced by each female, varied across insecticides and diets.Table 8 shows there was a significant insecticide* diet interaction for reproductive performance.Figure 6 shows that pupal rates were consistently low on the 1:12 diets across all insecticide treatments and highest on the 1:1 and 3:1 diet treatments, with the 1:6 and 1:3 diets being intermediate (SI Table 6).Despite variability, pupal rates were statistically similar across the control, zeta-cypermethrin, and pyrethrin treatments.
However, the zeta-cypermethrin treatment showed a significantly higher pupal rate in the 1:1 diet treatment, while the other insecticide treatments had less variability across diets (Figure 6; SI Table 6).

Discussion
Our nutrient regulation data show that, when given a choice between two diets with different P:C ratios, adult D. suzukii actively regulate their macronutrient intake, rather than feeding randomly across diets.
Although the intake targets in each diet-pairing treatment did not converge on a single ratio, the macronutrient intakes of flies in all treatments was more carbohydrate-biased than the respective lines of random feeding.The intake target for the two most protein-biased pairings, the 1:20 versus 1:1 and 1:20 versus 2:1 treatments, moved much farther than the 1:20 versus 1:3 pairing from the lines of random feeding towards the high-carbohydrate portion of nutrient space.
However, even the flies in the most carbohydrate-biased 1:20 versus 1:3 pairing, regulated for a more carbohydrate-biased ratio.This suggests that while the average P:C intake across all treatments was $1:3, the adult intake target is likely more carbohydrate-biased, being closer to the 1:4-1:5 ratio observed in the 1:20 versus 1:3 treatment.
This intake target is more in-line with what has been documented for D. melanogaster adults (Lee et al., 2008).Given that regulation was only measured over 48 h, it is possible that there was not enough time for the flies in the more protein-biased pairings to reach their optimal target on the diets available, resulting in higher P:C intakes.In any case, the intake targets recorded across all treatments for adult D. suzukii in this study were slightly more carbohydrate-biased than that found for larvae in Silva-Soares et al. (2017).Although they did not explicitly calculate an intake target, the Silva-Soares et al. (2017) data suggest that D. suzukii larvae regulate for a ratio of $1:3.2 T A B L E 4 Three-way ANOVA for P:C ratio intake across paireddiet treatments, diet macronutrient concentration, and fly sex, as well as Tukey's post-hoc results for significant main effects.T A B L E 5 Three-way ANOVA results for insecticide, diet, and sex on fly lifespan.(Deans & Hutchison, 2021).Rodrigues et al. (2015) reported an even more protein-biased larval intake target of 1:2 for D. melanogaster.
Several studies found that adult female Drosophila prefer to oviposit on carbohydrate-biased substrates, similarly to other drosophilid species (Rodrigues et al., 2015;Silva-Soares et al., 2017;Young et al., 2018).While it is possible that oviposition preference could have affected adult feeding in our study by drawing flies towards highercarbohydrate diets, all female flies used in the choice test were unmated.We also did not observe any differences in consumption or macronutrient regulation between male and female flies, suggesting that oviposition preference had little influence on feeding behaviour and/or nutrient regulation in this study.
We observed many interactions between insecticide, diet, and sex on lifespan, survivorship, and reproductive performance.Insecticide effects on fly lifespan varied across diet and sex, while the effect of diet was also sex-specific.The spinetoram treatments exhibited unexpectedly high mortality after exposure to an LC 50 diagnostic dose, showing between 78% and 90% mortality across diets.This limited our ability to detect diet effects for this insecticide, as average Female flies were more robust than males when exposed to zetacypermethrin and pyrethrin but not spinetoram.Across diets, females showed significantly longer lifespans than males only when reared on the 1:3 diet.Several other insecticide studies have documented higher female survivorship in D. suzukii across sub-lethal exposures to a range of different insecticides (Blouquy et al. 2021;Smirle et al., 2017;Gress and Zalom, 2019;Van Timmeren et al., 2020).These results suggest that physiological differences may exist between male and female flies that affect susceptibility.Female reproduction and detoxification are both protein-intensive and energetically-expensive processes, and given that female flies have higher protein, glycogen T A B L E 6 Kaplan-Meier (Mantel-Cox) results for differences in mortality curves across insecticides, diets, and male/female flies.and lipid stores than males (Tochen et al., 2016), females may have more flexibility for allocating protein and/or energy away from reproduction and towards detoxification when needed.In the choice study, female flies did ingest a greater amount of the more dilute 45 g/L diet than males, suggesting that they regulate their intake to meet stricter energy quotas than males, likely to produce the fat stores needed for egg development (Wheeler, 1996).Protein intake, on the other hand, did not differ between male and female flies, despite differences in susceptibility and differential requirements for egg production.
The overall effects of diet on lifespan and mortality rate were similar for the controls and zeta-cypermethrin treatments.The intermediate diets produced the longest lifespans and slowest mortality rates, while the carbohydrate-biased 1:12 diet treatment had consistently lower survivorship.The protein-biased 3:1 diet produced intermediate survivorship in the control and zeta-cypermethrin treatments, but played a different role in the pyrethrin treatments.While lifespan and survivorship were lower for flies in the 3:1 diet treatment compared to the 1:3 and 1:1 diets in the control and zeta-cypermethrin treatments, both parameters remained high for the 3:1 diet in the pyrethrin treatments.As mentioned, detoxification is a proteinintensive process (Berenbaum & Zangerl, 1994), and the beneficial effects of feeding on a high-protein diet observed in the pyrethrin treatment suggests that D. suzukii may be better able detoxify pyrethrin than zeta-cypermethrin.In a study on cotton bollworm, Helicoverpa amigera, it was shown that several different detoxification enzymes degraded pyrethrin-resembling compounds more readily than the modified synthetic versions (Tian et al., 2021).Given that survivorship was comparable across the zeta-cypermethrin and pyrethrin treatments, it is not likely that these differences were due to overall differences in toxicity.
In general, the diet effects on reproductive performance mirrored that of the effects on survivorship across insecticides.Pupal rates in the zeta-cypermethrin treatments exhibited an optimum at the intermediate 1:1 diet and then decreased slightly, though not significantly, on the 3:1 diet.Pupal rates were low and similar across all spinetoram treatments.Pupal rates in the pyrethrin treatments trended higher for the protein-biased 1:3, 1:1, and 3:1 diets than the more carbohydratebiased 1:12 and 1:6 diets, but these differences were not statistically significant.Overall, the high variability in pupal rates made identifying significant differences across diets and insecticides difficult.
This study shows that nutritional state can impact D. suzukii susceptibility to neurotoxic insecticides.Most research on nutritioninsecticide interactions in insects have focused on gut-acting toxins, such as Bt (Deans et al., 2016(Deans et al., , 2017;;Orpet et al., 2015Orpet et al., , 2015;;Shikano & Cory, 2014a), but this study shows that nutrition can also affect insecticides with broader modes of action.The mechanisms underlying the observed diet effects are still unknown, but they are most likely related to nutritional constraints on the allocation of resources to different processes that affect toxicity.These can include processes related to toxin activation, such as detoxification, or processes involved in cellular repair.The physiological effects of a particular toxin, its mode of action, and the evolutionary history between the toxin and the insect will undoubtedly direct these responses.Berenbaum and Zangerl (1994) documented a trade-off between growth and detoxification capacity in the parsnip webworm, where protein allocation to detoxification was prioritised at the expense of growth when caterpillars were exposed to xanthotoxin.However, the prioritisation of physiological processes in a generalist, such as D. suzukii, is likely to be different than for a specialist.Low dietary protein was shown to constrain esterase and glutathione transferase activity in the generalist gypsy moth but had no impact on monooxygenase activity and increased quinone reductase activity (Lindroth et al., 1990).
Drosophila suzukii feed on a wide range of different hosts that vary in macronutrient profiles (Deans & Hutchison, 2021;Young et al., 2018).Additionally, the requirement for colonisation of hosts to provide a source of dietary protein adds an additional layer of complexity (Deans & Hutchison, 2021;Hamby et al., 2012).Our choice data suggest that the intake target for adult D. suzukii is between 1:4 and 1:5, which is slightly more carbohydrate-biased than that found for larvae (Silva-Soares et al., 2017).The data from the diet-insecticide experiment also show that carbohydrate-biased diets tend to produce greater mortality and reduced reproductive performance for flies exposed to insecticides, with negative effects occurring on diets with P:C ratios below 1:3.Although the results of this study encompass the cumulative effects of diet on both larval and adult stages, our data suggest that feeding on resources with a ratio below 1:3 can increase adult susceptibility to insecticides.Additional research will be needed to determine how differences in larval versus adult nutrition may impact adult susceptibility, particularly because adults are much more mobile and have access to a greater diversity of resources.In any case, management tactics that can reduce microbial colonisation, such as sterilants, may bolster the efficacy of insecticides by reducing dietary sources of protein for larvae and/or adults.In fact, Van Timmeren et al. (2020) found that when a sterilant was applied to blueberry shoots alone and in conjunction with a zeta-cypermethrin application, infestations of D. suzukii were reduced to the same extent, or lower than, a conventional insecticide rotation.The sterilant had no impact on acute adult mortality or oviposition (under no-choice conditions) but did reduce oviposition preference on sterilised fruits.The reduction in the number of larvae found within the fruit suggests that the sterilant had negative impacts on larval growth.More research will be needed to determine if additional impacts on adult nutrition exist.In addition, understanding the interplay between nutrition and insecticide resistance will be important, as D. suzukii has demonstrated a propensity to develop resistance to spinosad and zeta-cypermethrin (Deans & Hutchison, 2022a, 2022b;Ganjisaffar et al., 2022;Gress & Zalom, 2019).Elucidating how nutritional plasticity can affect the evolution of resistance but also whether resistant population have different intake targets will be necessary for better understanding mechanisms of resistance and for making sure relevant diets are used in resistance monitoring bioassays (Deans et al., 2017;Shikano & Cory, 2014a, 2014b).

D
. suzukii in the field: Mustang Maxx ® , a pyrethroid containing the active ingredient zeta-cypermethrin, Delegate ® , containing spinetoram, and Pyganic ® , an organic pyrethrin formulation.Initial doseresponse assays were conducted in August 2020 for each insecticide to determine baseline susceptibilities and calculate concentrations for the median lethal dose (LC 50 ) of each insecticide to be used in the subsequent experiments.Mortality was recorded after 4 h of exposure to six concentrations of Mustang Maxx ® , seven concentrations of Delegate ® , and eight concentrations of Pyganic ® .Exposures were carried out by moving flies into 20 mL glass scintillation vials coated with insecticide and dried.Mustang Maxx ® formulations were acetone-based, while Delegate ® and Pyganic ® solutions were waterbased.Six replicate vials were used for each concentration and each vial contained 3 male and 3 female flies.
Bolded values indicate statistical significance at p ≤ 0.05.Data were rank-transformed to meet normality assumptions.

F
I G U R E 3 Average adult lifespan (±SE) for (a) male and female flies exposed to an LC 50 dose of different insecticides (pooled across diets) and (b) reared on different diets.Different letters indicate significant differences between bars (Tukey's test, p ≤ 0.05).F I G U R E 4 Average adult lifespan (±SE) for flies reared on different diets and exposed to an LC 50 dose of different insecticides.Different denote significant differences between bars (Tukey's test, p ≤ 0.05).
lifespan and reproductive performance were very low.The LC 50 value tested for spinetoram was similar to the LC 50 values recorded in field populations byVan Timmeren et al. (2018), which were between 13.7 and 25.9 ppm.However, D. suzukii has been documented to develop resistance to spinetoram over short periods of exposure(Deans & Hutchison, 2022a;Van Timmeren et al., 2018).The initial doseresponse assays were conducted in August 2020 while the dietinsecticide experiments were performed in May 2021.It is possible that unforeseen selection or drift occurred within our lab colony that may have impacted susceptibility to spinetoram.All experiments were performed at the same time, using the same fly colony, and all replicates were housed in the same location during the experiment.It is unknown why the flies in the spinetoram treatments exhibited increased susceptibility to the median lethal dose.

F
I G U R E 5 Mortality curves for diet treatments across insecticides: (a) controls, (b) zeta-cypermethrin, (c) spinetoram, and (d) pyrethrin.Different letters denote significant differences between diets within each insecticide (Mantel Cox, p ≤ 0.05).T A B L E 8 Three-way ANOVA results for female reproductive performance (pupae/female) across insecticide and diet treatments.

F
I G U R E 6 Reproductive performance (pupae/female) (±SE) for female flies reared on different diets and exposed to different insecticides.Different letters indicate significant differences between bars (Tukey' test, p ≤ 0.05).
Bolded values indicate significance at p ≤ 0.05.
Bolded values indicate significance at p ≤ 0.05.Date for ranktransformed to meet normality assumptions.