Peppermint essential oil inhibits Drosophila suzukii emergence but reduces Pachycrepoideus vindemmiae parasitism rates

Spotted Wing Drosophila (Drosophila suzukii; Matsumura) is an invasive fruit fly with the ability to oviposit in a broad range of agriculturally valuable fruits. Volatile organic compounds (VOCs) produced by botanical oils may reduce D. suzukii’s attraction to hosts and decrease survival, but it is unknown whether their efficacy varies across D. suzukii life stages or affects the survival and success of higher trophic levels. Through a series of laboratory bioassays, we evaluated the effects of peppermint (Mentha arvensis L.) oil produced VOCs on D. suzukii survival and the survival of and parasitism rates by a pupal parasitoid wasp, Pachycrepoideus vindemmiae (Rondani). First, we determined whether fumigation with peppermint oil VOCs at the pupal stage reduced adult emergence, and whether this depended on environmental conditions (i.e. soil moisture). Second, we evaluated whether fumigation with peppermint oil VOCs reduced or enhanced parasitism by the pupal parasitoid and whether this depended on the timing of peppermint oil VOC exposure (i.e. before, during, or after parasitoid access). Fumigation with VOCs of 4.5 mg of peppermint oil reduced D. suzukii emergence under moist soil conditions but dry soil had a similar effect on reducing adult emergence as peppermint oil presence. Peppermint oil VOC fumigation was toxic to adult P. vindemmiae, but developing P. vindemmiae were unaffected by peppermint oil VOC fumigation. Using peppermint essential oil as a fumigant may reduce D. suzukii emergence from the pupal stage. However, this could negatively impact P. vindemmiae dependent on the timing of application.

www.nature.com/scientificreports www.nature.com/scientificreports/ the dispenser by 47.6% 15 . Botanical oils which contain high concentrations of VOCs, such as peppermint (Mentha × piperita, Lamiaceae) oil, can repel adults or, in the case of thyme oil, increase adult mortality 19 . Extending this, Renkema et al. 20 found reduced larval infestations in strawberries that were surrounded by menthol-infused polymer flakes. While peppermint essential oil has been reported to repel adult D. suzukii 19,20 , it has yet to be determined if these botanical extracts are toxic to other D. suzukii life stages.
As adults only comprise 8% of the population at any given time 21 , management of D. suzukii could be improved through targeting larvae and pupae. This can include maintaining bare ground or using black plastic around host plants, which can create a hostile microclimate for all life stages for D. suzukii (e.g. high surface temperatures 22 and low humidity 23,24 ), or using pesticides that target other life stages 25 . The use of natural enemies for biological control, including generalist predators and parasitoids, is another strategy for D. suzukii control 26 . Pupae may be especially vulnerable to natural enemies as 82-93% of D. suzukii pupae develop in soil and not in the fruit 27 , similar to the behaviour of other fruit-targeting Drosophilids 28 . In the field, predatory hemipterans have been found to occur in D. suzukii-infested fruit 29,30 and Woltz and Lee 27 found exposure to predators in a blueberry field to decrease the abundance of pupae by 61-91% and larval survival by 19-49%, due to multiple predators, including ants and spiders. Similarly, in a laboratory setting, Garbarra et al. 29 found high levels of D. suzukii larval and pupal predation by the earwig (Labidura riparia Pallas).
Parasitic wasps are another biological control option for D. suzukii management, including the generalist pupal parasitoid Pachycrepoideus vindemmiae (Rondani; family: Pteromalidae). P. vindemmiae is able to parasitize over 60 fly species, including D. suzukii 29,31 and several other Drosophila species 32 . P. vindemmiae has a broad native range across North America and Europe 29,33,34 , and natural populations of P. vindemmiae often overlap with D. suzukii-infested areas [35][36][37] . In the laboratory, individual P. vindemmiae can successfully lay 7.5 ± 0.64 eggs in D. suzukii pupae each day 29 , yet current in-field parasitism rates are reported to be less than 10% in wild D. suzukii populations 38 . Furthermore P. vindemmiae experience high mortality and had lower parasitism rates when exposed to insecticides, including spinosyns, abamectin, neonicotinoids, organophosphates and pyrethroids 39 . However, it may be that optimising the use of resident parasitoids, such as P. vindemmiae, could reduce D. suzukii as part of an integrated pest management program.
Integrated pest management programs use multiple methods to reduce pest populations, including chemical and biological control 40 . Ideally, the multiple strategies used for pest control should work in synergy; however classical pesticide regimes can reduce the efficacy of biological control. Similarly, botanical oils and botanical oil VOCs may have both direct effects on natural enemy survival and performance, and also indirect effects through changes to host D. suzukii quality and susceptibility. We evaluated whether peppermint oil produced VOCs reduce D. suzukii pupal survival and whether this depended on soil moisture. We additionally assessed the compatibility of fumigation with peppermint oil VOCs with biological control, by evaluating the effects of peppermint oil produced VOCs on parasitoid wasp performance (survival, oviposition success, and adult emergence).

Materials and methods
Insects. Drosophila suzukii were kept in an incubation chamber set to 16:8 light:dark cycle with temperatures at 24.5 °C during light hours and 22.5 °C during dark hours. We reared D. suzukii on commercially available Drosophila diet (Ward's Instant Drosophila medium) supplemented with instant yeast (Red Star). Our laboratory D. suzukii colony was founded in June 2017, by individuals sourced from another colony started in the summer 2016 with D. suzukii collected from blueberries in Agassiz, BC, Canada. The laboratory D. suzukii colony is supplemented annually with wild-caught individuals.
The laboratory P. vindemmiae colony began in May 2018, from a colony established in 2015 from collections in sweet cherry at Summerland, BC. Pachycrepoideus vindemmiae were kept in separate rearing containers in the same incubator as our D. suzukii colony. We fed our P. vindemmiae colony 10% sugar water solution and a protein-lipid supplement (high protein patty, MegaBee). Drosophila suzukii pupae were provided for P. vindemmiae colony oviposition three times a week.
For all experiments involving these insects, we used 2-3-day-old D. suzukii pupae and 3-5-day-old P. vindemmiae adults. To remove excess diet, we washed all D. suzukii pupae under tap water and allowed them to air dry for one hour before any experiments. Both D. suzukii and P. vindemmiae used in the experiments were naïve to peppermint oil produced VOCs, and the P. vindemmiae had not previously oviposited.
Essential oil GC-MS analysis. Volatiles were collected from the headspace of a commercially available Japanese peppermint essential oil (Mentha arvensis L., Bulk Apothecary), with no dilution, and analyzed by headspace GC/MS at the Wine Research Centre Mass Spectrometry Core Facility (UBC) (see supplementary material for specific GC-MS methodology). All peaks found were integrated and tabulated as a percent value ( Supplementary Fig. 1), and identified based on the Wiley/NIST (National Institute of Standards and Technology) Library matching. Sixteen individual compounds were identified, with the primary compounds (>2% of the total amount) being Menthone (55.61%) and its isomer (16.40%), Menthol (16.83%) and its isomer (3.42%), and Menthol acetate (2.741%). All other compounds and their percentages are listed in Table 1.

Insect bioassays.
We conducted four bioassays to assess the effects of fumigation with VOCs from peppermint essential oil on D. suzukii pupal survival and P. vindemmiae parasitism success. All bioassays with D. suzukii were conducted in an environmentally-controlled growth room at the University of British Columbia, Vancouver, Canada. Within the growth room, we maintained a 16:8 light:dark schedule with cool white lights. Temperatures within the growth room were 22.6 ± 0.4 °C and relative humidity was 46.5 ± 3.2%.
We prepared our peppermint essential oil fumigation treatments as described by Renkema et al. 19 . We prepared three peppermint essential oil stock solutions for use in all bioassays: (1) acetone (0 mg, control), (2) 15 g peppermint essential oil/L acetone, and (3) 30 g peppermint essential oil/L acetone. As determined through preliminary experimentation, we cut eRoma Microfiber Absorbing Pads into 1 × 1 cm squares and treated each pad with 150 μL of their respective treatment, via stock solution, before each experiment. All treated pads were placed in a fume hood for one hour to allow the acetone to evaporate before use in bioassays 19 . After the acetone evaporated, we were left with three fumigation treatments which we refer to by the mass of peppermint oil applied to the absorbing pad: 1) 0 mg (control), 2) 2.25 mg and 3) 4.5 mg.
Bioassay 1: Effects of peppermint oil VOC fumigation and soil moisture on D. suzukii emergence. As soil moisture and relative humidity can be important for D. suzukii survival, we sought to determine whether peppermint oil VOC fumigation would reduce D. suzukii emergence, and whether this depended on soil moisture. We placed two grams of dry potting soil in a 30 mL polystyrene cup (Solo), into which we added 2 mL of distilled water to half of the cups (moist soil treatment) and the others received no distilled water (dry soil treatment). We then covered the soil with a circle of Mosquito No-See-Um Ultra-Fine Netting (Skeeta) to prevent D. suzukii pupae and emerged adults from burying into the soil. Within each soil cup, we placed ten D. suzukii pupae. A treated pad with one of the three peppermint oil VOC fumigation treatments (0 mg (control), 2.25 mg, and 4.5 mg) was placed on the opposite side of the cup from the D. suzukii pupae (n = 20 for each treatment). Bioassay cups were covered with a perforated plastic lid, which allowed ventilation. We monitored and removed adult flies emerging within the cups every 24 hours until no more adult D. suzukii emerged (5-7 days).

Bioassay 2: Effects of peppermint oil VOC fumigation on parasitoid mortality and emergence.
To determine whether peppermint oil VOC fumigation had a negative effect on D. suzukii emergence and P. vindemmiae parasitism, we conducted a bioassay with peppermint oil (0 mg (control) vs. 4.5 mg) and P. vindemmiae (male/female pair or none). We used these two concentrations based on preliminary analysis of Bioassay 1. We placed ten D. suzukii pupae along the side of a 120 mL clear plastic jar (Uline, Model S-9934) with ventilated lid, and placed treated pads on the opposite side of the jar from the D. suzukii pupae. Half of each peppermint VOC treatment received no wasps or one male/female P. vindemmiae pair (n = 15 for each treatment).
We recorded P. vindemmiae mortality every 24 hours and replaced dead wasps with new living wasps to maintain similar parasitism pressure. After 72 hours, we removed the remaining P. vindemmiae. Drosophila suzukii began emerging three days after the experiment start date with the majority of emergence happening on days four and five. We recorded D. suzukii emergence every 24 hours for one week until no more adult D. suzukii emerged. We continued to incubate the jars under the same laboratory conditions until P. vindemmiae emergence 23 days later. We counted and removed all emerging P. vindemmiae every 24 hours for six days, at which point no additional P. vindemmiae emerged.
Bioassay 3: Effects of peppermint oil VOC fumigation before parasitism. To identify whether VOCs produced by peppermint oil influenced P. vindemmiae host quality, we conducted a bioassay where we fumigated D. suzukii pupae with three different weights of peppermint oil (0 mg (control), 2.25 mg and 4.5 mg) prior to P. vindemmiae oviposition. We suspended a treated pad (0 mg (control), 2.25 mg or 4.5 mg peppermint oil) from the foam plug of a polystyrene Drosophila vial (VWR International, LLC.), into which ten D. suzukii pupae were placed. After 24 hours of peppermint oil VOC fumigation (or control), we removed the suspended treated pad and randomly placed either a single female P. vindemmiae or no wasp (control) into each peppermint oil VOC fumigation treatment (n = 15 per treatment). www.nature.com/scientificreports www.nature.com/scientificreports/ We recorded D. suzukii emergence every 24 hours for one week until no more adult D. suzukii emerged. We continued to incubate the jars under the same laboratory conditions until P. vindemmiae emergence 23 days later. We counted and removed all emerging P. vindemmiae every 24 hours for six days, at which point no additional P. vindemmiae emerged.
Bioassay 4: Effects of peppermint oil VOC fumigation after parasitism. We conducted an additional bioassay to examine the effects of peppermint oil VOC fumigation (0 mg, 2.25 mg or 4.5 mg peppermint oil) on D. suzukii pupae after parasitoid oviposition. We placed ten D. suzukii pupae and one presumably mated P. vindemmiae female into a polystyrene Drosophila vial (VWR International, LLC.). After 24 hours for oviposition, we removed the P. vindemmiae female. We then suspended a treated pad (0 mg, 2.25 mg or 4.5 mg peppermint oil) from the foam plug of the bioassay vial (n = 15 for each fumigation treatment). As in previous experiments, we monitored D. suzukii total emergence and continued to incubate the jars under laboratory conditions until parasitoid wasp emergence. We counted and removed all emerging wasps every 24 hours until the end of wasp emergence.
Statistical analysis. For D. suzukii and P. vindemmiae emergence, we used a generalized linear model (GLM) with a quasibinomial distribution and a log link function, after detecting overdispersion in preliminary models with binomial distribution. We then used F-tests to compare the effects among treatments 41 . For Bioassay 1, we found that D. suzukii emergence was zero for all replicates within one of the treatments (4.5 mg, dry soil). This confounded the model's standard error estimates and our ability to conduct posthoc tests on interactive effects. To counter these issues, we treated peppermint oil weight as a continuous factor and soil moisture as a categorical factor for this analysis. For Bioassays 2-4, we analyzed the proportion of D. suzukii pupae that emerged as adults, dependent on peppermint oil weight, P. vindemmiae presence or absence, and their interaction as categorical factors. For P. vindemmiae treatments, we analyzed the proportion of P. vindemmiae emerged as adults from D. suzukii pupae, with peppermint oil weight as a categorical factor. To determine the effect of peppermint essential oil fumigation on P. vindemmiae death (Bioassay 2), we used a linear model with number of dead P. vindemmiae as the response variable and peppermint oil (as a categorical factor) as the explanatory variable. All analyses were conducted in R 42 .

Results
Bioassay 1: Effects of peppermint oil VOC fumigation and soil moisture on D. suzukii emergence. The proportion of adult D. suzukii emergence decreased with exposure to peppermint oil VOCs through fumigation (peppermint: F 1,116 = 6.58, p = 0.002; Fig. 1). The proportion of D. suzukii adults emerging was lower with dry soil compared to moist soil (soil moisture: F 1,116 = 39.38, p < 0.0001; Fig. 1). In our control treatment, D. suzukii emergence decreased by 22% in dry compared to moist soil treatments. The effects of peppermint oil VOC fumigation was marginally significant in its interaction with the soil moisture treatment (peppermint x soil moisture: F 1,116 = 2.89, p = 0.059, Fig. 1).  Fig. 2A). The interaction between peppermint oil VOC fumigation and P. vindemmiae presence marginally affected D. suzukii emergence (peppermint x P. vindemmiae: F 1,56 = 3.94, p = 0.052; Fig. 2A). When P. vindemmiae was present, we observed a difference in D. suzukii emergence of 15% between peppermint oil treatments, and a difference of 3% between peppermint oil treatments when P. vindemmiae was absent. Pachycrepoideus vindemmiae reduced D. suzukii emergence by 48% in 0 mg (control) and 30% in 4.5 mg peppermint oil VOC fumigation treatments. Fumigation with peppermint oil VOCs increased the number of dead P. vindemmiae during the oviposition phase of this trial (F 1,28 = 6.86, p = 0.014; Fig. 2B) resulting in a marginal effect on subsequent P. vindemmiae emergence (F 1,28 = 3.94, p = 0.057; Fig. 2C). In our 4.5 mg peppermint oil VOC fumigation treatment, subsequent P. vindemmiae emergence was reduced by 6%.

Discussion
Our study suggests that under certain conditions fumigation with peppermint oil VOCs may be effective at reducing D. suzukii during its pupal stage (Fig. 1), and moreover may be compatible with biological control (Figs. 3-4). We found fumigation of pupae with peppermint oil VOCs to be compatible with P. vindemmiae before and after oviposition; however, fumigation with peppermint oil VOCs was not compatible with adult P. vindemmiae actively parasitizing pupae (Fig. 2B,C).
While the specific chemical composition varies between essential oils, there is evidence that oils with higher concentrations of monoterpenoids, such as peppermint essential oil, are efficient adulticides and larvicides for insect pests 44 . However, while VOCs from peppermint oil has been shown to be repellent to adults 19,20 , its toxicity to Drosophila adults varies from no effect 19 , to increased mortality of Drosophila auraria (Peng) exposed to pennyroyal oil (Mentha pulegium L.) 45 . The effects of peppermint oil produced VOCs on D. suzukii were inconsistent across our bioassays, suggesting mixed efficacy of using peppermint oil VOCs to control pupal stages of D. suzukii. We found the strongest effect of peppermint oil fumigation within our first bioassay (Fig. 1), with marginal to no effect within the last three bioassays on D. suzukii emergence. Our results suggest that more intensive fumigation effects were seen in smaller chambers like the one used in our first bioassay. We suspect that these www.nature.com/scientificreports www.nature.com/scientificreports/ mixed results were dose dependent effects of peppermint oil VOCs on pupal survival that we were not able to fully capture with the current set of experiments. Similarly, although Renkema et al. 19 found peppermint essential oil produced VOCs to be repellent for up to six days, a later study by Renkema et al. 20 observed lesser peppermint oil VOC effects as they found that male repellence was lost at 24 hours and that peppermint VOCs were ineffective against females.
Our results suggest that soil moisture may be more important for D. suzukii pupal mortality than fumigation effects with peppermint oil VOCs (Fig. 1), and D. suzukii pupae on moist soil had 22% higher emergence rates compared to pupae on dry soils. As we did not measure humidity within the emergence cups directly, any difference seen between the moist soil and dry soil treatment could be due to either the change in soil moisture or the relative change in general humidity between the moist soil versus dry soil treatment. The effect of soil moisture we observed is consistent with others' findings on the importance of microclimate and relative humidity is a key Figure 3. (A) The proportion of adult Drosophila suzukii emerging from ten pupae after 24 hours of pupal fumigation with VOCs from absorbance pads treated with three weights of peppermint essential oil (0 mg (control), and 4.5 mg) and subsequent exposure to the parasitoid Pachycrepoideus vindemmiae (presence or absence), with no significant interaction between peppermint oil and wasp presence; (B) The proportion of P. vindemmiae emerging from ten possible pupal hosts exposed to three peppermint essential oil fumigation treatments pre-oviposition. Means and standard errors shown. www.nature.com/scientificreports www.nature.com/scientificreports/ factor in D. suzukii success 23,24,43,46 . In the field, increased relative humidity increased D. suzukii fecundity and longevity while reduced humidity correlated with reduced trap captures 23 . Similarly, Diepenbrock and Burrack 24 observed higher infestations of blackberry (Rubus L.) within the center canopy and around field edges with higher humidity levels. Under extreme dry conditions (0% moisture), larvae desiccate or pupate on the soil surface and pupate at a shallow soil depth (1-6 mm) when moist conditions are present 46 .
Irrigation regimes are regionally tailored for specific crops. In berries, irrigation methods are often implemented for quality (i.e. firmness 47 ) or yield 48,49 , with consideration for disease incidence such as root rot 50,51 . In regard to D. suzukii, higher pupal emergence is associated with overhead sprinklers compared to drip irrigation as pupae located on the soil surface are less likely to desiccate with overhead irrigation systems 43 . In our study, there was a marginal interactive effect between soil moisture and peppermint oil VOC fumigation. Exploiting the potential general interaction between soil moisture and other control efforts may reduce D. suzukii in the field, as we found that fumigation effects of peppermint oil were especially effective in dry conditions. During suboptimal conditions, like drought, fumigation with peppermint essential oil may be a better tactic than when more optimal conditions are present. When conditions are optimal for soil pupation, such as when growers utilize overhead sprinkler irrigation, peppermint oil fumigation could still potentially reduce D. suzukii levels, but more work is needed to determine an effective application rate beyond what we tested and determine any potential non-target effects.
We did not test how moisture levels influenced P. vindemmiae's response to the presence of peppermint oil VOCs. Moreover, the P. vindemmiae in our three bioassays did not have a water or sugar source during their oviposition time, potentially influencing their behaviour and effect on D. suzukii mortality. For example, Da Silva et al. 52 found that water-deprived P. vindemmiae increased their host-feeding behaviour, resulting in increased D. suzukii mortality. Furthermore, this host-feeding behaviour lead to increased parasitism and had no effect on P. vindemmiae offspring mortality 52 . In a different study, Da Silva et al. 53 found that a sugar food source (i.e. honey) for young P. vindemmiae females, like those we used in our bioassays, did not affect D. suzukii host mortality.
Integrated pest management plans which use VOCs may be more compatible with biological control options than conventional insecticides 15,54 , motivating our study to observe the effects of peppermint oil VOC fumigation on P. vindemmiae performance. Our results show that for peppermint oil produced VOC, the timing of application is important for its compatibility with biological control. We found that direct fumigation with peppermint oil VOCs to be toxic to adult P. vindemmiae, while indirect effects of peppermint oil VOCs on P. vindemmiae were minimal (i.e. through potential changes in pupal host quality from peppermint oil VOCs exposure before and after parasitism). This may be because P. vindemmiae is able to complete development even within dead pupae, as we have found that P. vindemmiae successfully emerge from previously frozen D. suzukii pupal hosts (Reut, personal observation). Additionally, these results suggest that any persistence of peppermint oil VOCs on pupal hosts is short-lived and unlikely to directly expose P. vindemmiae to residual fumigation. Overall, high mortality and decreased parasitism rates occur in P. vindemmiae when directly exposed to insecticides, including spinosyns, abamectin, neonicotinoids, organophosphates and pyrethroids 39 . Furthermore, spinosad-treated pupae result in increased mortality of adult P. vindemmiae females 55 . We found fumigation with peppermint oil VOCs to be toxic to adult P. vindemmiae indicating that peppermint oil VOC fumigation should not occur when adult P. vindemmiae are active within the field.
While adult P. vindemmiae were negatively affected by peppermint oil VOC fumigation, we found developing P. vindemmiae receive some protection during peppermint oil VOC fumigation while within D. suzukii pupal cases. Other immature parasitoids benefit from insecticide protection while maturing within their hosts as has been shown with the housefly, Musca domestica L. 56 , aphids and mealybug mummies [57][58][59] . Insecticide protection during development within D. suzukii hosts has also been shown in P. vindemmiae but that degree of protection is dependent on timing of development 55 . Further understanding of the timing of peppermint oil VOC fumigation and associated trade-offs are needed in order to further develop D. suzukii management plans.
More research is needed to determine how peppermint oil VOC fumigation reduces D. suzukii emergence from its pupal stages in field applications. More localized applications of peppermint oil VOCs at the soil surface may be accomplished by using laminate flake technology. For example, laminate polymer flakes treated with peppermint oil have been shown to decrease D. suzukii oviposition in strawberries with recommendations for frequent applications (at least every four days) 20 . However, the effects of laminate flake technology treated with essential oils on D. suzukii pupae in field applications have yet to be determined. This technology could potentially provide an additional management strategy through more localized applications of peppermint oil VOCs. Furthermore, it could be especially useful for farmers using overhead sprinkler irrigation, as this would exploit the potential interactive effect of soil moisture with peppermint oil produced VOCs.

conclusion
Our results provide further support that microclimatic effects (e. g. humidity) are important for D. suzukii control and reducing humidity in the field will help suppress D. suzukii populations and potentially amplify effects of other control efforts. Using peppermint essential oil as a source of VOCs may reduce D. suzukii emergence from the pupal stage, and these techniques could potentially be used alongside a biological control program with P. vindemmiae or other natural enemies.

Data availability
The datasets generated during the current study are available from the corresponding author upon reasonable request.