Spotted wing drosophila, Drosophila suzukii (Matsumara) (Diptera: Drosophilidae) is a large vinegar fly that has spread from its center of origin in Asia to Hawaii, North America, and Europe, and is now beginning its invasion of fruit-producing areas of South America and Africa (Asplen et al. 2015). Currently, crops in Australia and New Zealand show no signs of D. suzukii activity. However, these nations' proximity to the home range of D. suzukii may soon lead to an invasion by this fly species (Markow & O'Grady 2006; Biosecurity Australia 2010). In 2008, D. suzukii began to inflict 50 to 100% berry losses in parts of the western U.S., primarily in California, Washington, and Oregon (Lee et al. 2011; Cha et al. 2014). Two years later, eastern and southern U.S. states began losing fruit to D. suzukii. Four to 5 yr after that, D. suzukii damage accounted for an estimated 6% loss in blueberry yield and 34% loss in bramble yield, and with time, these losses are expected to rise because of the increased likelihood of fruit rejections during or after harvest (Price et al. 2009; Goodhue et al. 2011; Burrack 2014).

In exotic habitats, D. suzukii has become a major harvest pest of berry crops. The females possess long, sclerotized, and deeply serrated ovipositors capable of inserting eggs into the mesocarp of fresh intact berries (Atallah et al. 2014), a behavior unknown in the common saprophagous species, Drosophila melanogaster (Meigen) (Diptera: Drosophildae). Consequently, D. suzukii's niche outside of Asia overlaps little with any other native North American drosophilid, and together with its wide polyphagy (100+ fruit-bearing wild and cultivated plant species from 25 families), has fostered this species' extreme invasiveness (Biosecurity Australia 2010). The females of D. suzukii prefer ovipositing into bumpy fleshy fruits with a thin skin, e.g., fruits that have drupes or achenes (blackberries, raspberries, and strawberries), but dozens of wild berry species can serve as alternative hosts for D. suzukii to breed almost year-round in the U.S. (Lee et al. 2015; Poyet et al. 2015). Two dozen of D. suzukii's crop hosts yield berries with an exocarp thin enough for ovipositors to puncture. Some of these crops include globally important and high-value fruit species including stone fruits (peaches, cherries, and apricots), blueberries, overripe grapes, citrus, apples, pears, persimmons, figs, and pomegranates (Price et al. 2009; Beers et al. 2011; Stewart et al. 2014).

With any new invasive insect species, the development of optimal diets is crucial for the success of any integrated pest management program and one of the first steps in obtaining adequate live material for study (Vanderzant 1974). Rearing drosophilid flies is well established. However, rearing D. suzukii for research is more difficult than culturing D. melanogaster or other drosophilids for two reasons. First, D. suzukii females preferentially oviposit into firm thin-skinned fruits, not into soft media or overly soft and decomposing berries. Second, and perhaps more importantly, D. suzukii cultures lacking solid fruit in their diet are highly susceptible to microbial disease (B. J. Sampson, personal observation). Although D. suzukii related articles based on laboratory tests briefly describe diet methodology, none mentions incidences of disease, use of antibiotics, or the contributions of the diet's ingredients to fly nutrition and rearing success. Fruit-amended diets for D. suzukii commonly have included purees of fresh fruit, either peach or raspberry, with (Stewart et al. 2014) or without the addition of an agar thickener (Hardin et al. 2015). Whole berries (e.g., raspberries), one or perhaps more, can be dropped into a rearing vial onto the soft diet media (Kacsoh & Schlenke 2012). These authors did not describe the quantities of ingredients or microbial contamination caused by adding or pureeing fruit in the D. suzukii diet. These methods also do not mention supplementing the diet media with symbiotic yeasts, which are recommended for use in D. melanogaster media to increase fly yield and curb microbial contamination. Drosophila species typically vector a community of yeasts, of which many species, once introduced into host tissues, serve as vital sources of nutrients for ovipositing females and developing larvae. The presence of yeast, which can outcompete harmful microbes, can stimulate oviposition in many fruit-breeding species of Drosophila, and perhaps yeast benefits reproduction in D. suzukii (Hamby et al. 2012).

The rearing method described here is similar to the fruit-based diet described by Kacsoh & Schlenke (2012). However, we discuss problems encountered while rearing locally captured D. suzukii and solutions found to optimize this species' diet. We chose D. melanogaster as a control in some of our studies because of its close taxonomic relationship with D. suzukii and its ease of management. We tested three hypotheses related to improving the management of captive D. suzukii populations. The first hypothesis tested was that incorporating blueberry or blackberry fruit into standard D. melanogaster diet media increases D. suzukii's oviposition rate, culture productivity, disease resistance, and adult fitness based on body size (Promislow et al. 1998). Second, inoculating fresh media with dry baker's yeast further promotes D. suzukii's reproductive output. Third, D. suzukii adults poorly tolerate the fly anesthetic 50% triethylamine (FlyNap®) when compared with their taxonomic relative D. melanogaster.

Materials and Methods

Results

INCLUSION OF WHOLE FRUIT IN DIET MEDIA INCREASES CULTURE PRODUCTIVITY AND HEALTH

Fitness of reared D. suzukii based on oviposition success and culture productivity varied with the fruit crop chosen as an oviposition substrate. Confined with blueberry fruit, a single mated female laid 10 eggs per berry in 3 d, and 90% of these eggs were inserted fully into the mesocarp (mean ± SE = 3.30 ± 0.66 mesocarpic eggs per day, n = 243 berries). The tiny remainder of eggs sat fully exposed atop the fruit surface (mean ± SE = 0.40 ± 0.08 eggs per day). The D. suzukii females were opportunistic egg-layers, ovipositing upwards of 17 eggs into blueberry picking scars or other deep blemishes.

By addition of fruit to diet media, the probability of mature 21-dold cultures reaching mean peak pupal production of 73 pupae per vial doubled. Cultures of D. suzukii produced equivalent numbers of pupae after 21 d of culture growth, regardless of the fruit species used, or lack thereof (Fig. 1; F = 1.08; df = 5, 35; P > 0.38). However, for 14-d-old D. suzukii cultures (Fig. 1) with diets containing a blackberry fruit, pupal yields were comparable to those obtained on diets containing black cherry (1 fruit per vial, Wilcoxon 2-sample test; Z = 0.72; P > 0.45), blueberry (3 fruits per vial, Wilcoxon 2-sample test; Z = 1.52; P > 0.05), and strawberry (0.5 fruit per vial, Wilcoxon 2-sample test; Z = 0.72; P > 0.45). These yields were also twice those of vials with diets containing a white grape (Wilcoxon 2-sample test; Z = 2.16; P < 0.05) or no fruit (Wilcoxon 2-sample test; Z = 2.16; P < 0.05).

Adult quality, based on body size, within disease-free D. suzukii cultures equaled that of wild-caught adults. Females whether wild or reared had the same body length (2.80 ± 0.04 mm; F = 2.37; df = 1, 39; P > 0.13) and eye width (0.46 ± 0.01 mm; F = 0.01; df = 1, 39; P > 0.90). Similarly, reared and wild males were approximately the same size (2.22 ± 0.03 mm vs. 2.29 ± 0.04 mm; F = 2.03; df = 1, 36; P > 0.16). Reared males, however, had larger eyes or, more precisely, a larger area of red pigment (0.43 ± 0.01 mm vs. 0.37 ± 0.01 mm; F = 17.07; df = 1, 36; P < 0.001). For laboratory-reared D. suzukii, body size for male and female flies varied little between diets with or without blackberry fruits (fruit vs. no-fruit diet; F = 0.00; df = 1, 89; P > 0.9400). Males were the same size irrespective of diet (1.40 ± 0.02 mm from anterior of eye to posterior of scutellum) and the same was true for the larger females (1.63 ± 0.02 mm, male vs. female body length: F = 73.45; df = 1, 89; P < 0.0001).

Fig. 1.

Dietary manipulation of Drosophila suzukii cultures based on the use of 5 berry species (blackberry, strawberry, black cherry, blueberry, and grape) with a no-fruit (No Fruit) control (grey bars), and cultures to which yeast was added (+Y) or omitted (-Y) from 4–24® drosophila media (white bars). Asterisks indicate mean differences from 2 respective baseline controls (B), i.e., Blackberry-Y (14 d) and Blackberry+Y (21 d), according to multiple Wilcoxon 2-sample tests.

Without fruit, 2 to 100% (average = 43%) of new D. suzukii cultures (n = 122 cultures) contained a fully developed biofilm after 4 to 7 d. Cultures containing a fruit sterilized in 80 to 90% ethanol exhibited 0 to 13% (average 4%) spoilage, a rate similar to that in D. melanogaster cultures during the same period (n = 55; Chi-squared test; P < 0.005). Using deionized water to remove ethanol from sterilized fruit was intended to prevent the inadvertent poisoning of adults of D. suzukii, a relatively alcohol-intolerant species (Sampson et al. 2016; Fig. 2). Ethanol concentrations between 0 and 2.5% led to a 6-fold increase in adult turnover after 21 d of culture growth (6.11 ± 0.98) for D. melanogaster, twice that of D. suzukii (adult turnover rate = 2.85 ± 0.57; F = 8.55; df =1, 59; P < 0.005). In addition, fruit-enriched media prolonged culture productivity in D. suzukii; adult production continued within 93% of 48-d-old cultures (n = 27), 80% of 73-d-old cultures (n = 5), 38% of 81-d-old cultures (n = 13), and 29% of 101-d-old cultures (n = 7). Sixty percent of cultures without fruit were producing adults after 67 d (n = 25). Overall, 39% more mature cultures continued producing adults for 60 to 67 d when diet contained a blueberry or blackberry fruit (n = 70 cultures; Chi-squared test; P < 0.05).

INCLUSION OF A SUPPLEMENTAL YEAST IN DIET INCREASED CULTURE PRODUCTIVITY

Inoculating media with dry baker's yeast further increased D. suzukii pupal production by ∼30% (Fig. 1; Wilcoxon 2-sample test; Z = 2.00; P < 0.05). Moreover, in Jul 2015, D. suzukii vials (n = 54) containing moistened media lacking yeast and fruit were 98% free of microbial infection, yet, 1 mo later, a random sample of these same vials were found to produce, on average, 33% fewer pupae and adults (92.5 ± 8.2) than did cultures fed a yeast+fruit diet (138.7 ± 8.2; F = 13.38; df = 18; P < 0.005). These latter fruit-enriched cultures were 100% free of microbial contamination (n = 44 vials).

Fig. 2.

Effects of dietary ethanol (normal environmental ethanol range ≈ 0–9% ethanol) on Drosophila suzukii survival when compared with D. melanogaster tolerance to ethanol (top graph) and sex-specific sensitivities of D. suzukii adults to ethanol (bottom graph). Summary of probit analyses are provided and statistics with b_o = intercept estimate, b₁ = estimated slope estimate for each fly species, with b₁ = 0 for the baseline control in each graph (D. suzukii [top graph], D. suzukii males [bottom graph]).

ADULT ANESTHETIZATION WITH TRIETHYLAMINE CAN BE ACUTELY TOXIC TO D. SUZUKII

Recovery rates of D. suzukii adults after exposure to triethylamine for 15, 30, 60, 90, and 120 s averaged 31%, with 20% or less adults recovering after ≥30 s of exposure (Fig. 3; Cochran-Armitage trend analysis; Z = -8.78; P < 0.0001). For wildtype D. melanogaster adults, overall recovery rate for the same lengths of time was 3-fold higher (94%), varying little despite longer exposures to triethylamine (Fig. 3; Cochran-Armitage trend analysis; Z = -0.68; P > 0.20). Exposure to triethylamine anesthetic for more than 30 s could kill D. suzukii or paralyze 70% of adults for more than 3 h. Most immobilized D. suzukii adults recovered after 15 s of triethylamine exposure. This short exposure time, however, left some flies active long enough to escape the vials.

Fig. 3.

Effect of triethylamine anesthetic exposure on the recovery of Drosophila adults: D. melanogaster (white circles, n = 181 adults), D. suzukii (black circles, n = 175 adults).

Discussion

A solid food diet for raising laboratory cultures of D. suzukii includes some of the same basic ingredients used in other drosophilid diets, e.g., sugar, fruit, and yeast (Dalton et al. 2011; Follett et al. 2014; Jaramillo et al. 2014; Hardin et al. 2015). Although our D. suzukii diet is a very simple formulation, the proper ratio of its ingredients is vital for the fly's balanced nutrition. However, the rapid spread of D. suzukii and little knowledge of this pest's basic biology and nutrition afforded researchers little time to optimize D. suzukii diets. Outbreaks of deadly microbial infections commonly occurred in D. suzukii cultures, whether fruit were added or not. Mold commonly originated from fruit but was effectively controlled after fruit surfaces were sterilized with ethanol. These chemically sterilized fruit, however, were highly toxic to flies and, therefore, required pre-washing in sterile water before their addition to diet media.

The D. suzukii cultures reached peak pupal production when started with ∼20 adults and provided with a carbohydrate-rich diet containing 7 parts clean water to 1.5 parts dry diet to which was added 1 part fruit (w/w/w). Age structure for healthy populations of cultured D. suzukii and D. melanogaster appeared stable with 2 pupae produced for every emerging adult (Emiljanowicz et al. 2014). Inoculating culture media with dry baker's yeast further increased D. suzukii adult and pupal production by one-third. In two other studies, yeast augmentation increased D. suzukii oviposition and pupation by a similar rate (Jaramillo et al. 2014; Hardin et al. 2015), indicating yeasts are symbiotic and probably provide D. suzukii larvae with the protein that their natural habitats (e.g., ripe or ripening berries) lack (Hamby et al. 2012; Hardin et al. 2015). However, adding far more than 5 grains (>0.80 mg) of exogenous yeast to the diet could harm cultures if fermentation produces toxic levels of carbon dioxide that sterilize, incapacitate, or kill flies (Markow & O'Grady 2006). Moreover, the diverse fungal community associated with D. suzukii is a possible source of endosymbionts that might someday be cultivated and introduced to a rearing system to enhance fly fitness or exploited in novel ways to reduce fly survival and reproduction (Hamby et al. 2012).

In D. melanogaster cultures, artificial media alone induces prompt egg laying. In D. suzukii cultures, earlier oviposition occurs on natural food substrates, i.e., intact thin-skinned fruits, a discovery made by Kacsoh & Schlenke (2012), who added thawed, frozen raspberries to standard drosophila diet media. Our findings concur with those of Kacsoh & Schlenke (2012); adding intact fruit to drosophila media prompted earlier oviposition by D. suzukii. These researchers, however, reported no concomitant increase in overall fly production. In contrast, our fruit-based diet enhanced fly output by (1) hastening larval survival, (2) increasing pupal production, and (3) preventing outbreaks of disease in the cultures. Interestingly, cultures of D. suzukii supplied with fruit experienced faster population growth without incurring a concomitant reduction in adult body size. This result suggests cultures had unlimited food resources for the first month of culture growth. Fruit added to diet media probably provides D. suzukii adults and larvae with additional macronutrients, micronutrients (e.g., phytosterols), and carbohydrate (Jaramillo et al. 2014; Hardin et al. 2015). Fruit composed 40% of the diet by weight and accounted for a one-third increase in early pupal output. Fortuitously, the addition of whole berries to the drosophila diet controlled disease, which not only saved cultures from destruction, but also enhanced culture productivity by 30%.

Our observation of D. suzukii females ovipositing into blueberry picking scars indicates that this species has the capacity to expand its host choice to damaged fruit of host plants and perhaps non-host plants. For instance, picking scars and other forms of harvest damage to non-host fruits such as peaches can induce D. suzukii females to lay 4 times more eggs than insect damage does and 10 times more eggs than simulated damage does (Stewart et al. 2014). Hence, fruit marred from hail and during harvest might possibly increase market losses to D. suzukii infestation. Therefore, reducing D. suzukii population size and fruit injury during harvest could lessen overall yield loss.

Not all intact fruits species stimulate D. suzukii oviposition. For instance, white grapes produced no more pupae that did the no-fruit diet. Likewise, muscadine grapes have skins too thick for D. suzukii females to oviposit. When given no alternative hosts, D. suzukii females misplace or abort their eggs by positioning them horizontally atop un-punctured unblemished muscadine fruits (B. J. Sampson, personal observation). These misplaced eggs withered without larval hatch (Stewart et al. 2014). Drosophila suzukii females can easily lay eggs in blueberry and blackberry fruits, two hosts with thin skins. Interesting, this is perhaps the first study that describes females misplacing or aborting a small fraction of their eggs on blueberry, a known cultivated host of D. suzukii. This behavior suggests blueberry is a less preferred reproductive host, and does explain lower mean yield losses for blueberry when compared with other cultivated small fruits, notably brambles (Burrack 2014). Successful oviposition occurs only after the female vertically inserts eggs into the mesocarp with their paired respiratory (oxygen) filaments projecting above the puncture site (Stewart et al. 2014). Hence, the ratio of viable eggs (successful oviposits) to aborted or misplaced eggs might serve as an excellent index for measuring host plant resistance.

Berries added to diet media stimulates captive D. suzukii females to oviposit, yet in their raw state, fruits can greatly increase the probability of fungal disease (mold). To prevent such fungal infections, the exocarp of whole berries or fruit chunks, whether fresh or thawed, require chemical sterilization with ethanol. Triple rinsing berries with deionized or distilled water then removes or dilutes ethanol to a safe level and prevents the inadvertent poisoning of alcohol-sensitive adults (Sampson et al. 2015). In their study of D. suzukii associated yeasts, Hamby et al. (2012) pre-sterilized flies with 70% ethanol to remove any cuticular yeasts. Afterward, they rinsed away any residual alcohol that, if swallowed, could harm microbial endosymbionts. This prudent pre-washing of flies probably prevented mass death among their experimental flies.

Ethanol has higher acute oral toxicity to D. suzukii when compared with saprophagous species like D. melanogaster (Sampson et al. 2015). The LC50 value for D. suzukii adults imbibing ethanol was ∼9%. Adults of D. melanogaster after drinking 10% alcohol experienced no apparent mortality and ethanol's LC50 value for this species was twice that observed for D. suzukii (Sampson et al. 2015). Detoxification of low concentrations of ethanol (≤2.5%) by dehydrogenase enzymes produces acetate, a metabolite used by D. melanogaster as a supplemental high-energy food source to double pupal production as well as increase the rate of reproduction to a level twice that of D. suzukii (Sampson et al. 2015). Drosophila suzukii cannot convert low levels of ethanol to food, suggesting a very low activity of alcohol detoxification. Increasing rates of environmental ethanol destabilized sex ratios of D. suzukii, whereas sex ratios of D. melanogaster remained unaffected even at dietary ethanol concentrations as high as 40% (Sampson et al. 2015). A female-biased sex ratio for D. suzukii cultures probably results from the greater sensitivity of males to dietary ethanol. Without dietary alcohol, the sex ratio for D. suzukii stabilized near that of D. melanogaster, ∼1:1. Hence, using fruit and yeast to rear D. suzukii could produce sufficient dietary ethanol to depress the size of male populations without reducing overall culture longevity. Intolerance to higher concentrations of ethanol perhaps evolved in D. suzukii, as adult and larval habitats contain little if any environmental alcohol.

Berries sitting in media can inhibit microbial growth by enticing D. suzukii oviposition, which in turn may further inoculate food with competing symbiotic microbes (Bergomaz & Boppré 1986), disrupt biofilm formation through larval feeding (Jefferson 2004), and acidify media via fruit decomposition. Small D. suzukii cultures with less than 20 adults and lacking a fruit for oviposition are prone to microbial infection and biofilm formation. These sticky biofilms invariably kill every adult fly in a culture. Whether microbial communities associated with D. suzukii exude these biofilms is unknown. Clearly, sugar-rich media can spur growth in some harmful microbial species (Chandler et al. 2014). If the presence of a fruit spurs D. suzukii eggs to be laid and larvae to hatch within 3 d of culture installation (parents introduced to fresh media), then resulting larvae should have ample time to feed and bury harmful microbes as they move throughout the food matrix. Hence, adding a blueberry or blackberry confers to D. suzukii flies resistance to microbial contamination of their food. Likewise, cleaning metal worktables with 100% ethanol and adding ∼300 μL of penicillin-streptomycin antibiotic to the media further curbs microbial outbreaks. We have used this antibiotic mixture in the spring months when some microbial outbreaks seem more frequent, especially those outbreaks associated with deadly biofilm formation. However, antibiotics are only effective until biofilm-producing microbes develop resistance to them (Jefferson 2004). Adding fruit to the diet and pre-sterilizing drosophila vials overnight in 5% or 10% bleach solution are more cost-effective measures for suppressing microbial growth. Further microbial control is possible by cleaning foam sponge stoppers (sponges) in soapy water, not bleach, as bleach reacts chemically with foam to darken, shrink, and reduce stopper elasticity.

Adult flies and pupae at risk of drowning in soggy media might benefit from fruit in another way; berries can serve as edible perches. Generally, D. suzukii prefers a firmer diet (not dry) to a soggy diet. Therefore, the proper balancing of water-to-diet ratios would depend on humidity, temperature, fruit size, and berry species. Excessive residual water leaking from decomposing fruit could combine with condensation inside small vials to create droplets or pools large enough to drown flies individually or en mass. However, flies reared in standard (68 mL) plastic vials faced no serious moisture buildup. Fruit degradation seems to preserve media moisture and fly production for approximately 60 to 80 d.

Adults will require immobilization before being examined. Drosophila suzukii showed greater sensitivity to triethylamine anesthetic than did D. melanogaster. The D. suzukii adults could safely tolerate approximately 15 s of exposure to triethylamine; any exposure longer than 15 s caused fly recovery rates to drop dramatically. However, 15 s might be too brief of an exposure to immobilize every fly in a vial. Drosophila suzukii and mosquito adults respond similarly to FlyNap® and both species are highly sensitive to this anesthetic (Chen & Hillyer 2013).

Reared and wild D. suzukii adults raised on a fruit-based diet were similar in quality based on body size. The relatively larger eyes of reared males perhaps occurred because eyes of wild-caught flies lost some of their ocular pigment to leaching into the preservation fluid. In fact, Jaramillo et al. (2014) assessed body size for D. suzukii fed a diet containing either fruit or yeast and discovered that mean body size for males and females remained unchanged.

Current Drosophila rearing technology can now be used to prepare a disease-free D. suzukii diet. A suitable ratio for mixing diet ingredients was, by weight, 1 part berry for oviposition, 1.5 parts dry diet media for carbohydrate, 7 parts clean water for moisture, and ∼5 grains of dry yeast for protein. In 68 mL vials, these proportions of ingredients translate into 2.5 g fresh or thawed blackberries (preferred) or blueberries, 3.75 g dry diet media, 20.0 mL water, and 0.8 mg of dry baker's yeast. Protein-rich yeasts and nutrient-rich fruits are important dietary supplements for D. suzukii, a drosophilid species whose larval niche is initially devoid of protein and large cultures of symbiotic microflora. Fruits require sterilization with doses of ethanol deadly to D. suzukii (∼80–90%) followed by its thorough removal before introducing D. suzukii adults. The use of chemical anesthetics such as triethylamine to knockdown adults can be extremely toxic to these flies.