Ability of Muscidifurax raptorellus and Other Parasitoids and Predators to Control Drosophila suzukii Populations in Raspberries in the Laboratory

Drosophila suzukii is an invasive pest and economic threat to berry crops in Europe and the Americas. Current methods of control of this pest rely primarily on frequent applications of insecticides; therefore, there is a need for alternative control methods to reduce insecticide reliance. In this study, we evaluated the biological control potential of three parasitoid wasps: Diglyphus isaea, Muscidifurax raptorellus and Pachycrepoideus vindemmiae, and four predators: Chrysoperla carnea, Dicyphus hesperus, Orius insidiosus and Podisus maculiventris. Experiments were conducted for 15 days under controlled conditions in experimental arenas with D. suzukii females and raspberries, allowing for all life stages of D. suzukii to be available to natural enemies. Results showed the first evidence of M. raptorellus’s ability to parasitize D. suzukii, resulting in a 40% reduction. Orius insidiosus, P. vindemmiae and C. carnea were also efficient, reducing D. suzukii numbers by 49%, 43% and 32%, respectively. Predator preferences for each D. suzukii life stage were assessed. The clutch size, sex ratio and adult size variability of D. suzukii pupal parasitoids were also evaluated. This study expands the list of species that can effectively parasitize D. suzukii and provides new insights into the biological responses of M. raptorellus to D. suzukii pupae.


Introduction
Drosophila suzukii Matsumura (Diptera: Drosophilidae) is an invasive pest from Asia that is now considered a major economic threat to soft fruit crops in Europe and the Americas, causing up to 80% yield loss in some crops [1]. Unlike other vinegar flies which prefer decaying fruit, D. suzukii females damage ripening fruits by laying their eggs with a serrated ovipositor and most damage is caused by larvae feeding on the fruit pulp [2]. Raspberry, strawberry, blackberry, blueberry and cherry have been the main crops affected by D. suzukii since its invasion in 2008 in North America [1,3,4]. Raspberries are D. suzukii's preferred host, with approximately $40 million in losses from 2009 to 2014 in California [5].
Synthetic insecticides are the primary control method for D. suzukii but repeated interventions may promote insect resistance and most products target only the adult stage, which is outside the fruit [6,7]. Spinosad (Entrust SC) is an effective biological insecticide for organic growers, but preharvest intervals and the number of allowable applications per season per crop make management difficult [8]. There are

Insect Sources
Drosophila suzukii was obtained from a colony maintained at 23 ± 1 • C; 16 L: 8 D; 50 ± 10% RH in plexiglass cages in the Institut de recherche et de développement en agroenvironnement (IRDA) in Saint-Bruno-de-Montarville, Quebec. The colony was started in summer 2014 from infested blueberry collected near St-Jude, Quebec. Eggs were laid in raspberries placed directly on the diet, in which larvae migrated. The diet consisted of a mixture of 70 g of brewer's yeast, 150 g of carrot powder, 2.5 g of sodium benzoate, 2 g of methylparaben, 50 mL of 1 N HCl and 800 mL of distilled water. Flies were also provided with a mixture of white sugar and baker's yeast (1:1) in a Petri dish and water on a dental cotton roll.
The parasitoid wasp M. raptorellus was purchased from Bugs for Bugs (Guelph, ON, Canada). The parasitoid was sold as a mix in pupae of Musca domestica Linnaeus (Diptera: Muscidae), from which three parasitoid wasp species (Pteromalidae) emerge. Emerged parasitoids were sorted and male and female M. raptorellus were transferred using an insect vacuum to a separate plexiglass cage. Prior to use, parasitoids were provided with honey and water (1:1) on a dental cotton roll.
The parasitoid wasp P. vindemmiae was from a colony started in summer 2017 and maintained at 22 ± 1 • C; 16 L: 8 D; 55 ± 10% RH in a plexiglass cage in IRDA's laboratory. The colony was reared on D. suzukii pupae and originated from a P. vindemmiae colony at Agriculture and Agri-Food Canada (AAFC) in Agassiz, British Columbia, which was started from infested cherries collected near Summerland, British Columbia (BC), in 2015.
The larval predator C. carnea, the adult predator O. insidiosus and the parasitoid wasp D. isaea were purchased from PlantProducts (Leamington, ON, Canada) and kept in their shipping package with an ice pack for one day prior to use.
The adult predator D. hesperus and the nymph predator P. maculiventris were purchased from Anatis Bioprotection (Saint-Jacques-le-Mineur, QC, Canada) and kept in their shipping package with an ice pack for one day prior to use.

Parasitoid and Predator Experiment with Multiple D. suzukii Life Stages
The experimental arena was a clear plastic box (1 L) with a 40-cm length mesh sleeve with an opening on one end. Each arena contained five store-purchased red raspberries, washed in water, and placed in a clear plastic dish (177 mL, 10-cm diameter) with a cotton square to absorb excess liquid during the experiment.
On day 1 of the experiments, five 3-day old D. suzukii mated adult females were placed in each plastic box to lay eggs in raspberries (22 ± 1 • C; 16 L: 8 D; 55 ± 10% RH). Two raspberries were added to each plastic box every three days throughout the experiment in order to avoid the presence of decaying egg laying sites.
Eighteen females of each parasitoid species (3 to 5 days old and previously mated) or 18 individuals of each predator species were placed in arenas (one species per arena) on day 4 for larval parasitoids, on day 9 for pupal parasitoids and on day 3 for predators of the experiment, to ensure larvae and/or pupae were available. Eighteen parasitoids or predators were used based on the results of a pre-test with 12, 18 or 24 individuals and the densities used in a related study [36].
The experiments were conducted in two blocks, with the three parasitoid species and a control in the first and the four predator species and a control in the second. Each parasitoid and predator species was replicated eight times. After 15 days (time required to complete one D. suzukii life cycle under optimal laboratory conditions), the live predators, live parasitoids, D. suzukii larvae, pupae and live and dead adults were counted. The clear plastic dish with raspberries was filled with water so that D. suzukii larvae and pupae floated and could easily be removed and counted. The design of this experiment allowed age structure to be established so natural enemies were in the presence of all life stages of D. suzukii (egg, larva, pupa, adult) in each arena at the same time and the control (no natural enemies) was used as the base for comparisons among treatments.
For arenas with larval parasitoids, recovered D. suzukii larvae were placed in a 500-mL plastic container with a lid containing fine mesh for ventilation, filled with drosophila diet and kept in growth chambers for 35 days to determine the D. suzukii pupation rate or emergence rate of adult parasitoids. For arenas with pupal parasitoids, D. suzukii pupae recovered were placed singly in wells of an ELISA plate and kept in growth chambers for 35 days to determine the emergence rates of adult parasitoids or D. suzukii.
The success rate of parasitism (parasitoid adult emergence), the clutch size (number of parasitoids from one pupa), sex ratio and adult size variability were determined. The last three parameters are used in biological control studies to predict the efficacy of insect parasitoids and their foraging behavior [37]. While the clutch size and sex ratio are foraging decisions based on host quality, the progeny body size, directly influenced by the clutch size, is a direct indicator of fitness [37].
The adult parasitoid size was estimated by measuring the length (mm) of the left hind tibia [38]. For M. raptorellus, the measurements of 15 males and 15 females that emerged from M. domestica pupae served as a control. For individuals from D. suzukii pupae recovered in the 15-day experiment, 15 males and 15 females were measured when there was one parasitoid per pupa and 15 pupae were randomly selected when there were two or more parasitoids per pupa. For P. vindemmiae, the measurements of 15 males and 15 females that emerged from D. suzukii pupae in the colony served as a control. For individuals from D. suzukii pupae recovered in the experiment, 15 males and 15 females were measured, as P. vindemmiae is a solitary pupal parasitoid. The measurements were made by removing the hind tibia from the body of the insect, photographing and measuring it with an electronic stereomicroscope, Zeiss SteREO Discovery V12 with Zeiss Axiocam 503 3 Mpix video camera, using software, ZEN 2012 SP2 blue edition (© Carl Zeiss Microscopy GmbH, 2011).

Predator Experiment with Single D. suzukii Life Stage
The experimental arena was a clear, plastic container with a lid (177 mL, 10-cm diameter) and a raspberry leaf (cv. 'Polana') from unsprayed research plots that was laid flat and had its petiole inserted into a water-filled 2-mL microcentrifuge tube. Each arena contained six individuals of one D. suzukii life stage (eggs, larvae, pupae or adults) which were placed on top of the leaf which served as a support. Eggs were collected from a Petri dish filled with store-purchased raspberry jam that was placed in a D. suzukii colony cage for two hours. Eggs were rinsed with distilled water then placed on the central vein of the leaf with a paintbrush. Larvae and pupae were collected from the artificial diet used in the colony cage, rinsed with distilled water and placed onto leaves with a paintbrush. Adults were collected from the emerged pupae that were isolated in a cage to ensure that the flies were all five days old and three males and three females were placed into the experimental arenas using an insect vacuum. A predator (either adult O. insidiosus, adult D. hesperus, nymph P. maculiventris or larval C. carnea), starved for 24 h, was placed in each arena. Arenas were held at 22 ± 1 • C; 16 L: 8 D; 55 ± 10% RH in a growth chamber. No food was provided to D. suzukii adults or larvae to prevent predators from feeding on it. After 24 h, predators were removed and the number of live D. suzukii individuals for each life stage was recovered from the leaves except for the pupae which were held seven additional days for adult emergence. There were 15 replications of each predator species and a control without a predator, for each of the four D. suzukii life stages.

Data Analysis
Data were tested with the Shapiro-Wilk test for normality and error variance for homoscedasticity, and analyses were conducted using JMP software (version 12.0.1, SAS Institute Inc., Cary, NC, USA). Data from the parasitoid and predator experiments with multiple D. suzukii life stages were analyzed using a one-way analysis of variance (ANOVA), followed by Tukey-Kramer's HSD (p < 0.05) for post-hoc comparisons effects of natural enemy species on the multiple D. suzukii life stages, except for D. suzukii live adult data, which were non-parametric and analyzed using a Kruskal-Wallis test, followed by post-hoc Dunn's test (p < 0.05). Data from the predator experiments with a single D. suzukii life stage were non-parametric and analyzed using a Kruskal-Wallis test, followed by post-hoc Dunn's test (p < 0.05) to compare means of live D. suzukii among predators. Data from parasitized D. suzukii pupae for each pupal parasitoid wasp species were analyzed using a Student's t-test (p < 0.05) to compare the parasitism rates. Data from individual parasitoid wasps that emerged from D. suzukii pupae were analyzed using a Chi-square test (p < 0.05) to compare male: female ratios for each clutch size. The leg measurements of M. raptorellus that emerged from D. suzukii pupae were non-parametric and analyzed using a Kruskal-Wallis test, followed by a post-hoc Dunn's test (p < 0.05). However, the leg measurements of individuals that emerged from P. vindemmiae were normally distributed and analyzed using a Student's t-test (p < 0.05).

Parasitoid and Predator Experiment with Multiple D. suzukii Life Stages
At the end of the 15-day experiment, M. raptorellus and P. vindemmiae significantly reduced D. suzukii numbers by 40% and 43%, respectively compared to the control (F 3,27 = 15.97, p < 0.0001) ( Figure 1). Only P. vindemmiae caused a decrease in the mean number of D. suzukii live adults (χ 2 = 25.54, df = 3, p < 0.0001). Both pupal parasitoids decreased the number of D. suzukii pupae (F 3,27 = 8.81, p = 0.0003). Larval parasitoid D. isaea did not have an impact on any of the D. suzukii life stages. At the end of the experiment, there were 7.0 ± 4.2 live M. raptorellus and 14.3 ± 2.7 live P. vindemmiae (mean ± SD), but all individuals of D. isaea were dead.

Predator Experiment with Single D. suzukii Life Stage
Predation occurred primarily on D. suzukii eggs and larvae ( Figure 3). Predators C. carnea, D. hesperus and O. insidiosus were able to feed on D. suzukii eggs (χ 2 = 47.88, df = 4, p < 0.0001) and predators C. carnea, O. insidiosus and P. maculiventris were able to feed on D. suzukii larvae (χ 2 = 34.54, Figure 2. Means (± SE) of D. suzukii cumulative life stages (larvae, pupae, and dead and alive adults) recovered after 15 days when exposed to predators. Each arena (n = 8 per predator species) contained five D. suzukii mated females and raspberries at the beginning of the experiment, with 18 predators added on day 3. Cumulative means with the same letter were not significantly different (Tukey's HSD, p < 0.05).

Predator Experiment with Single D. suzukii Life Stage
Predation occurred primarily on D. suzukii eggs and larvae (Figure 3). Predators C. carnea, D. hesperus and O. insidiosus were able to feed on D. suzukii eggs (χ 2 = 47.88, df = 4, p < 0.0001) and predators C. carnea, O. insidiosus and P. maculiventris were able to feed on D. suzukii larvae (χ 2 = 34.54, df = 4, p < 0.0001). None of the four predators were able to predate significantly D. suzukii adults and pupae (χ 2 = 10.25, df = 4, p = 0.0363; χ 2 = 8.54, df = 4, p = 0.0737). At the end of the experiment, all predators were found alive except for two D. hesperus. Most of the different life stages were recovered alive in the control treatments except for adults, nearly 50% of which died in the absence of food.

Discussion
Our results provided the first evidence of M. raptorellus's ability to parasitize D. suzukii pupae in the laboratory and demonstrated a level of efficacy comparable to P. vindemmiae. As with P. vindemmiae, M. raptorellus females were able to lay viable eggs in D. suzukii pupae and larvae readily fed on D. suzukii pupae, resulting in successful adult emergence. Both pupal parasitoid species also showed the same success rates of parasitism, although P. vindemmiae's rate in our study (45%) is lower than that reported in Europe (57% and 53%) [20,39].
Diglyphus isaea was not able to parasitize D. suzukii, as no parasitoids emerged from D. suzukii larvae kept in growth chambers and most D. suzukii pupated successfully. We also observed no parasitoid adults alive in experimental arenas at the end of the experiment. Cheah and Coaker [40] reported that D. isaea females ignored live host larvae of Chromatomyia syngenesiae isolated from their mines, suggesting that the host-finding process required host plant stimuli. Diglyphus isaea is primarily a parasitoid of leafmining insects [30] and probably would simply not recognize and assess D. suzukii larvae at the surface of fruits as potential hosts [41].
Up to six gregarious M. raptorellus adults survived and emerged from a single D. suzukii pupa in this experiment, while Harvey et al. [42] reported up to 15 adults emerging from a single M. domestica pupa. However, the clutch size frequency decreased as the number of individuals emerging per host increased for both host species [42]. The most common clutch sizes in D. suzukii were 1-3 eggs, which was the same as in M. domestica [42]. Similarly, Petersen and Currey [29] noted that M. raptorellus laid 2-3 eggs per host when there was a high availability of M. domestica. Geden and Moon [43] reported that M. raptorellus produced 2.2-2.9 progeny per parasitized host from five dipteran hosts of different sizes, comparable to the production of an average of 1.9 parasitoids per D. suzukii pupae in this study. It appears that M. raptorellus does not strongly discriminate hosts for quality and does not adjust the clutch size according to the host size, knowing that D. suzukii pupae are more than ten times smaller than M. domestica pupae ( Figure 5).

Discussion
Our results provided the first evidence of M. raptorellus's ability to parasitize D. suzukii pupae in the laboratory and demonstrated a level of efficacy comparable to P. vindemmiae. As with P. vindemmiae, M. raptorellus females were able to lay viable eggs in D. suzukii pupae and larvae readily fed on D. suzukii pupae, resulting in successful adult emergence. Both pupal parasitoid species also showed the same success rates of parasitism, although P. vindemmiae's rate in our study (45%) is lower than that reported in Europe (57% and 53%) [20,39].
Diglyphus isaea was not able to parasitize D. suzukii, as no parasitoids emerged from D. suzukii larvae kept in growth chambers and most D. suzukii pupated successfully. We also observed no parasitoid adults alive in experimental arenas at the end of the experiment. Cheah and Coaker [40] reported that D. isaea females ignored live host larvae of Chromatomyia syngenesiae isolated from their mines, suggesting that the host-finding process required host plant stimuli. Diglyphus isaea is primarily a parasitoid of leafmining insects [30] and probably would simply not recognize and assess D. suzukii larvae at the surface of fruits as potential hosts [41].
Up to six gregarious M. raptorellus adults survived and emerged from a single D. suzukii pupa in this experiment, while Harvey et al. [42] reported up to 15 adults emerging from a single M. domestica pupa. However, the clutch size frequency decreased as the number of individuals emerging per host increased for both host species [42]. The most common clutch sizes in D. suzukii were 1-3 eggs, which was the same as in M. domestica [42]. Similarly, Petersen and Currey [29] noted that M. raptorellus laid 2-3 eggs per host when there was a high availability of M. domestica. Geden and Moon [43] reported that M. raptorellus produced 2.2-2.9 progeny per parasitized host from five dipteran hosts of different sizes, comparable to the production of an average of 1.9 parasitoids per D. suzukii pupae in this study. It appears that M. raptorellus does not strongly discriminate hosts for quality and does not adjust the clutch size according to the host size, knowing that D. suzukii pupae are more than ten times smaller than M. domestica pupae ( Figure 5).
The clutch size influenced the parasitoid progeny size, as reported in Harvey et al. [42], and an M. raptorellus clutch size of three or more eggs, adult female and male wasp size decreased with D. suzukii as the host. Smaller female wasps may result in reduced fitness, as the ability to control dispersal and lifetime egg production and reproductive success may be reduced [38]. However, as there was no significant difference between the sizes of M. raptorellus that emerged from M. domestica (control) and from D. suzukii (1-2 eggs per host), D. suzukii, as the host, may not cause reduced fitness for most of the M. raptorellus progeny. The clutch size influenced the parasitoid progeny size, as reported in Harvey et al. [42], and an M. raptorellus clutch size of three or more eggs, adult female and male wasp size decreased with D. suzukii as the host. Smaller female wasps may result in reduced fitness, as the ability to control dispersal and lifetime egg production and reproductive success may be reduced [38]. However, as there was no significant difference between the sizes of M. raptorellus that emerged from M. domestica (control) and from D. suzukii (1-2 eggs per host), D. suzukii, as the host, may not cause reduced fitness for most of the M. raptorellus progeny.
However, regardless of the clutch size (1-5), we observed high variability in the size of M. raptorellus siblings of both sexes within the same pupae. Furthermore, most pupae with a single emergent parasitoid also contained small dead larvae of non-emerged parasitoids, which would explain the different sizes of singly emerged adults. Size variability, irrespective of the clutch size, is typical in gregarious idiobiont parasitoids, as many offspring can develop inside a single host [44]. For idiobiont species such as M. raptorellus that paralyze their host, host development is stopped at oviposition and the resources are then immutable, as the amount of food available inside the pupae does not increase [28,38]. Therefore, space and resources are limited, and competition increases as more parasitoid larvae feed on the host.
The low female ratio of M. raptorellus from D. suzukii hosts may be explained by the general sex-allocation theory that parasitoid wasps will invest in female offspring in larger hosts and use male eggs for less-than-optimal hosts [41,45]. Harvey and Gols [46] reported that M. raptorellus had 60% female progeny in M. domestica host and 76% female progeny in a larger host, Calliphora vomitoria Linnaeus, and as the clutch size increased, the percentage of female progeny increased in the latter host from 52% with a clutch size of one and two eggs to 81% with a clutch size of five eggs or more. We observed the opposite trend in this experiment with the smaller D. suzukii host-as the clutch size increased, the percentage of female progeny decreased from 44% to 27%. However, as there was no significant difference from a 1:1 sex ratio of M. raptorellus progeny from D. suzukii when the clutch size was 1-2 eggs, D. suzukii pupa may be a suitable host for most of the M. raptorellus progeny.
As expected, the solitary parasitoid P. vindemmiae produced one parasitoid per D. suzukii pupa in the control and experiment. The sex ratio of progeny was not different from 1:1, and the percentage of female progeny (53%) was similar to the parent colony at AAFC Agassiz [47].
Idiobiont ectoparasitoids such as M. raptorellus and P. vindemmiae are often associated with concealed hosts [14]. Therefore, D. suzukii pupae hidden among decomposed raspberries inside the experimental arenas may represent a suitable host for these parasitoids. Both ectoparasitoid species tend to be more adaptable in their biological response to different hosts and are likely to search for a specific type of habitat or host location instead of a specific type of host itself [14]. This suggests that M. raptorellus may be a good pupal parasitoid of D. suzukii pupae fallen from fruits and buried in the soil.
The results from our predator experiments with single and multiple D. suzukii life stages showed that O. insidiosus and C. carnea were efficient D. suzukii predators. By conducting a, no-choice experiment in a simple environment, we determined which easily accessible D. suzukii life stages were attacked by selected predators. We found that P. maculiventris fed only on larvae, and D. hesperus fed only on eggs, while C. carnea and O. insidiosus both fed on D. suzukii eggs and larvae. None of the predators appeared to be able to attack pupal or adult D. suzukii, but we observed opportunistic feeding by some predators on adults that were weak or dead.
Prey finding and acceptance can be affected by predator hunting and mobility capacities, nutritional benefits and the chemical composition of the prey cuticle [41]. Though destroyed D. suzukii pupal cases were found in experimental arenas, pupal cuticles may be impenetrable by predatory bugs and lacewings with piercing-sucking mouthparts. Nearly all predators mentioned in studies that report predation of D. suzukii pupae have chewing mouthparts [23,48,49]. Adult flies may be difficult for predators to capture in small experimental arenas where flying insects are too confined. Cuthbertson et al. [22] reported significant predation on adult flies by Orius spp. in a small experimental arena but only after 72 h. We also observed that D. suzukii eggs may be too small for P. maculiventris which possess a large proboscis to stab and carry their prey.
By conducting a more complex experiment with all D. suzukii life stages simultaneously present in raspberries, we determined predator efficiency that mimics field situations. Podisus maculiventris and D. hesperus reduced D. suzukii numbers by 15% and 16% respectively, but suppression was not significantly different from the control. It is likely that these two predators were not able to prevent exponential D. suzukii population growth because they only prey on one life stage, as shown in the no-choice trial. In contrast, C. carnea and O. insidiosus reduced D. suzukii numbers by 32% and 49%, respectively. The efficacy of C. carnea and O. insidiosus in a more complex environment may be explained by their ability to prey on two D. suzukii life stages, as shown in the no-choice trial. Therefore, it appears that predators that feed primarily on early D. suzukii life stages, such as eggs and larvae, can suppress D. suzukii population increases inside the experimental arena over the course of the experiment, while those that are only capable of consuming one life stage are less effective. The presence of raspberries in the experimental arenas may also have played a role in the efficacy of the predators and their foraging behavior. We observed O. insidiosus and C. carnea foraging on the surface of fruits for D. suzukii eggs and larvae, as was reported for O. insidiosus in Woltz et al. [25]. As the raspberries started to decompose, all the predators gained easier access to D. suzukii life stages. In addition, the suppression of D. suzukii may also be attributable to nonconsumptive predator effects [50]. Predators may have physically disturbed D. suzukii females during oviposition which would have resulted in reduced D. suzukii numbers.

Conclusions
Our study provided new additional information on the host range of the pupal parasitoid M. raptorellus and its ability to parasitize invasive D. suzukii. Muscidifurax raptorellus had similar parasitism rates to P. vindemmiae, and both reduced D. suzukii numbers in raspberries in the laboratory by 40 and 43%, respectively. Studies have shown that P. vindemmiae parasitized D. suzukii in the laboratory and field. However, in a natural environment, P. vindemmiae was not sufficiently abundant to prevent exponential D. suzukii population growth [39,51,52]. As a potential solution, M. raptorellus is commercially available in Canada and the United States and could be released in sufficient numbers to suppress D. suzukii. Further study on the growth and development of M. raptorellus in D. suzukii pupae is needed to better understand the host-parasitoid interactions.
Our results also confirmed that predatory natural enemies, O. insidiosus and C. carnea, were effective at reducing D. suzukii by almost 50%. In conclusion, all four of the effective natural enemies are present in agroecosystems in North America, were complementary in the life stages of D. suzukii they attacked and may play a role in the suppression of D. suzukii in fruit crop fields. Further research is needed to precisely measure the combined impacts of multiple natural enemies under field conditions, to evaluate a possible preference for alternative preys or hosts, and to optimize and support the augmentative releases of commercially-available species.