Key message

  • Spotted wing drosophila is an invasive fly pest that damages soft- and thin-skinned fruit crops.

  • A classical biological control program was initiated by surveys for spotted wing drosophila parasitoids in South Korea during 2013 and 2014.

  • Seven known and one new parasitoid species were reared from spotted wing drosophila, with field percent parasitism by larval parasitoids ranging from 0 to 17.1 %.

  • Quarantine studies are ongoing to select the most suitable natural enemies for release in North America and Europe.

Introduction

Spotted wing drosophila, Drosophila suzukii Matsumura (Diptera: Drosophilidae), is a pomace fly distributed natively across eastern China, Japan, the Korean Peninsula, and other regions in Southeastern Asia (Hauser 2011; Kanzawa 1939). In North America, the fly was first detected in 2008 in California (Bolda et al. 2010) and was subsequently reported in most fruit growing regions in the continental United States (US) and Canada (Asplen et al. 2015; Emiljanowicz et al. 2014; Walsh et al. 2011). Drosophila suzukii was also detected in Spain and Italy in 2008, and soon thereafter reported in other European countries (Calabria et al. 2012; Cini et al. 2014); most recently, the fly was reported in Brazil (Deprá et al. 2014). Drosophila suzukii is considered a key pest of soft- and thin-skin fruits such as blueberries, cherries, figs, raspberries, and strawberries in all of these newly invaded regions (Burrack et al. 2013; Kinjo et al. 2013; Lee et al. 2011; Mitsui et al. 2006; Yu et al. 2013).

The pest status of D. suzukii is attributed to the female’s serrated ovipositor (Atallah et al. 2014) which enables oviposition in ripening, rather than over ripe or rotting fruit. While D. suzukii is unable to oviposit in fruits with thick, hard, or fuzzy skin, it can oviposit and complete development in some of these fruits when the skin is compromised (e.g., pathogen or insect damage, hail) (Lee et al. 2015; Poyet et al. 2014; Steffan et al. 2013; Stewart et al. 2014; Ioriatti et al. 2015). Additionally, D. suzukii’s fast development (a generation in ≈10 days at 22 °C) and high reproductive potential (>600 eggs per female) (Emiljanowicz et al. 2014; Tochen et al. 2014) can result in an explosive population increase (Wiman et al. 2014) and significant economic losses to commercial crops (Beers et al. 2011; Goodhue et al. 2011; Walsh et al. 2011). Control efforts in North America currently rely on the use of insecticides that target adult D. suzukii (Beers et al. 2011; Bruck et al. 2011). However, insecticide-based programs can be limited by the fact that many host fruits in non-crop habitats act as reservoirs for D. suzukii and support its reinvasion into commercial fields (e.g., Klick et al. 2014, 2015).

From this perspective, biological control in non-crop habitats may help suppress regional D. suzukii population densities and provide a new tool for reduced risk and environmentally sound management strategies. Worldwide, over 50 hymenopteran parasitoid species attack drosophilid species. In most cases, these parasitoids attack their host in the late-larval or pupal developmental stages (Carton et al. 1986). The majority of the larval parasitoids are braconids in the genus Asobara and figitids in the genera Leptopilina and Ganaspis, and they predominantly attack host fly species found in fermenting substrates (Asplen et al. 2015). There are also two common pupal parasitoids, Pachycrepoideus vindemiae Rondani (Pteromalidae) and Trichopria drosophilae Perkins (Diapriidae). These larval and pupal parasitoid species undoubtedly play a role in population suppression of some drosophilid species. The effectiveness of these parasitoids against D. suzukii in its invaded regions has, however, not been adequately studied and, where surveys have been conducted, there appears to be limited parasitism rates of D. suzukii in terms of economic suppression (Miller et al. 2015).

Kacsoh and Schlenke (2012) studied 24 strains or populations of parasitoid species for their effectiveness against D. suzukii, and report that among 12 larval parasitoid species tested, only Asobara japonica Belokobylskij (from Japan), A. citri (from Ivory Coast), and two strains of Ganaspis sp. (from Hawaii and Florida) could complete development on D. suzukii. Of the tested larval parasitoid species, A. japonica had the highest rate of successful development. In most cases, the larval parasitoid species’ failure to develop in D. suzukii was explained by the host’s immune resistance (Kacsoh and Schlenke 2012; Poyet et al. 2013). Similarly, in another laboratory study, Chabert et al. (2012) trialed five parasitoid species that are commonly found in Europe attacking drosophilids (A. tabida Nees, Leptopilina heterotoma (Thomson), L. boulardi Barbotin et al., T. cf drosophilae, and P. vindemiae) and found that only the two pupal parasitoids (T. cf drosophilae and P. vindemiae) successfully developed on D. suzukii. Field surveys also suggest that parasitoids attacking drosophilid species resident in the US and Europe have not readily adapted to the invasive D. suzukii. Typically in field surveys, only the pupal parasitoids T. drosophilae and P. vindemiae have been reported to attack D. suzukii in Italy (Rossi Stacconi et al. 2013), Spain (Gabarra et al. 2015), and the US (Miller et al. 2015; Wang et al. 2016). One exception is a strain of L. heterotoma from northern Italy that was recovered from D. suzukii (Rossi Stacconi et al. 2015).

The lack of effective biological control in the newly invaded range of D. suzukii led to the initiation of a classical biological control program. Genetic analyses suggest that East Asia is the region of origin for D. suzukii populations that invaded North America (Adrion et al. 2014; Chiu et al. 2013; Ometto et al. 2013), and this is the focal region for our initial and planned collections. Information on parasitoids associated with D. suzukii and other drosophilids in Asia is however limited to reports from Japan. For example, Mitsui et al. (2007) reported 15 parasitoid species collected from fruit-baited traps, these included A. japonica, A. tabida, A. rossica Belokobylskij, A. rufescens (Forster), A. pleuralis (Ashmead) A. leveri (Nixon), L. heterotoma, L. victoriae Nordlander, Ganaspis xanthopoda (Ashmead), and P. vindemiae. Among the 15 species, only A. japonica, A. tabida, and G. xanthopoda emerged from D. suzukii. The aim of this study was to determine the presence and biological suitability of Asian parasitoids of D. suzukii to be considered for quarantine examination and potential field release in North America and Europe. In South Korea, no reports of D. suzukii-related damage have been made and this species has not been considered a pest (Asplen et al. 2015). We describe herein the first collection efforts in South Korea and the initial quarantine studies on the imported material.

Materials and methods

Collection sites and methods

The South Korean sample locations were selected based on a pre-collection exploration in 2011 (J. Miller, H. Riedl and Y. Song, pers. comm.). Surveys for frugivorous Drosophila parasitoids in South Korea were conducted at 10 locations and three provinces during August 2013 (J. Miller, B. Miller, H. Riedl and P. Shearer), and at 18 locations and four provinces from June to July 2014 (J. Miller and B. Miller), aided in each year by collaborators in South Korea (Fig. 1).

Fig. 1
figure 1

Collection locations for drosophilid parasitoids in 2013 (black), 2014 (gray) or both years (white) in South Korea: a Namhae Marina; b Boriamsa; c Ungbongsan; d Mangunsan; e Kumosan; f Sacheon; g Guam-Imdo; h Jinju city and Gyeongsang; i Jinju; j Uigoksa and Jinju Heights; k Munsusa; l Jagulsan; m Jeongryeongchi; n Namwon; o Baekyeonsan; p Poyeongsa; q Gamaksan; r Geochang; s Deogyusan; t Jeoksangsan; u Sangju; v Namjangsa; w Gunjae Bong and Ban Suk San. Close locations (<5 km) for both years were merged

Full size image

During the expeditions, parasitoids were field-surveyed using uninfested-fruit-baited traps deployed for 4–7 days and via the collection of commercial or wild host fruits with a high probability of infestation as determined primarily by the presence of adult flies. The uninfested-fruit-baited traps were based on previous collections by Rossi Stacconi et al. (2013). For each trap, 100–200 g of mixed fruits (banana, blueberry, fig, melon, and peach) was placed in a 162-mL plastic cup. Each cup was fitted with a lid and ten 0.8-cm diameter entrance holes along the cup rim to allow entry of ovipositing flies and parasitoids, but prevent the entry of other animals or larger insects. The plastic cup was placed inside a 500-mL plastic square tub and the ‘cup-in-tub’ set-up was then placed inside an orange delta trap (Suterra, LLC, Bend, OR). The cup-in-tub method provided a dry microhabitat for larvae to pupate when the fruit started decaying. Upon collection of the baited traps, the square tub was covered with a screened lid and the tubs were kept at room temperature (22–28 °C) until pupation. Field-collected fruits were separately placed in 162-mL cups on filter paper inside a 500-mL plastic tub until pupation.

In 2013, we primarily used the uninfested-fruit-baited traps, deployed at inland or coastal forests or along the perimeter of crop fields (Table 1). At each collection site, 8–12 traps were hung at shoulder height (1.5 m) from sturdy branches and deployed for 4–7 days. Commercial blackberries (approximately 10 L), damaged pears, apples, peaches, or split grapes were also sampled in four different locations (Table 1). In 2014, we predominantly used fresh-fruit collections of 2–10 L of wild Rubus fruits per collection site, as these fruits were still available in these unmanaged habitats during June and July (Table 1). In addition, commercial blackberries were collected from one location (Geochang) and 10 uninfested-fruit-baited traps were placed at another location (Boriamsa) where wild Rubus fruits were also collected (Table 1).

Table 1 Number of D. suzukii and other drosophilids collected during August 2013 and June to July 2014 explorations in South Korea using different collections methods (fruit-baited traps or collection of fruit from crops or wild habitats) and in different geographic locations
Full size table

Drosophilid pupae that developed from the field-collected larvae were sorted according to collection or trap location and placed on moist filter papers in 5-cm diameter Petri dishes. The dishes were kept at 4 °C before being packaged for hand-carry to the US. During transportation, the puparia were placed inside a Styrofoam box and kept cool with packaged ice.

Quarantine of imported puparia

Once the Styrofoam box passed through US Homeland Security at San Francisco International Airport, and inspection at the USDA, APHIS, PPQ Plant Inspection Station (permit: P526P-11-03867), imported puparia were processed under controlled conditions (22 ± 3 °C, 12L: 12D, 40–70 % RH, and natural light) in quarantine facilities at the University of California, Berkeley. Hosts for imported parasitoid colonies and quarantine evaluation of imported parasitoid species were obtained from laboratory colonies of D. suzukii and D. melanogaster Meigen (a common host for most parasitoid species attacking drosophilids). The fly colonies originated from field collections in Parlier, California and were maintained on standard corn-meal diets using methods similar to Dalton et al. (2011).

Upon arrival into quarantine, the puparia were sorted to separate those bearing visible signs of parasitism, i.e., visible parasitoid pupae inside the fly puparia. Individual puparia were placed in Eppendorf tubes with a streak of honey-water provided for emerging adults. Emerged flies were killed and preserved with 95 % alcohol along with empty pupal cases in Eppendorf tubes for later species identification. Emerging female parasitoids were individually placed into vials containing artificial diet and larvae or pupae of D. suzukii or D. melanogaster. Male wasps emerging from the same location-based samples were coupled with their respective females. Honey-water droplets were provided for the parasitoids and host larvae/pupae were renewed every 2–3 days. Upon death, individuals of the parental generation of parasitoids were preserved in 95 % alcohol along with their pupal case. Pupae of D. suzukii were distinguishable from other Drosophila pupae by the presence of a pair of distinct respiratory tubes on the anterior end that was used as a diagnostic character (Kanzawa 1939). Vouchers of the figitids are deposited with the U.S. National Entomological Collection (Washington, D. C.), while other species are deposited in the Institute for Sustainable Plant Protection (Portici, Italy) or in the Essig Museum at the University of California (Berkeley, California). Guerrieri et al. (2016) identified all Asobara species using both morphological and molecular methods, while all other species were confirmed or identified by ML Buffington.

Additionally, all dead (no adult emergence) pupae were first reconstituted in water and then dissected under the microscope to assess the presence or absence of recognizable fly or parasitoid cadavers. However, it was often difficult to distinguish parasitized pupae from unparasitized pupae when they died at early stages, and in these cases the pupae were classified as ‘unknown.’ The parasitism rate was estimated as the sum of emerged parasitoids and dead parasitized pupae divided by the sum of emerged flies and parasitoids plus dead parasitized and unparasitized pupae.

Quarantine bioassays

Three parasitoid species, later identified as A. japonica, T. drosophilae, and P. vindemiae, from the collections in 2013, and two additional larval parasitoid species, Leptopilina japonica japonica Novković & Kimura and G. brasiliensis, obtained during the 2014 sampling (see Results section), were briefly tested to confirm their ability to attack and develop from D. suzukii or from D. melanogaster, chosen as a common alternative host. Two or three-day old larvae (for A. japonica, L. j. japonica, and G. brasiliensis) or two or three-day old pupae (for T. drosophilae and P. vindemiae) were offered to the parasitoids for 2 days. For each parasitoid species, one female and one male were released into ventilated plastic vials (40 mL) containing hosts infesting artificial diet, or on plastic cups (330 mL) with three D. suzukii-infested fruits (blueberries or cherries). Fruits were infested by exposing them, in groups of ten, for 3 h into fly rearing cages containing several adult flies; fruits were then observed under the microscope and the number of eggs per fruit was standardized to three by removing excess eggs using a needle probe. The host density in diet vials varied between 10 and 30 for D. suzukii and 40–80 for D. melanogaster. Exposed hosts were reared until the emergence of flies or parasitoids.

The most common parasitoid species (see “Results” section), A. japonica, was subjected to a no-choice test to determine the parasitoid’s performance on two different-sized hosts. One-week-old mated female wasps were individually exposed to 15 young larvae of either host species in vials with artificial diet for 2 days. The number, sex, and developmental time of emerged parasitoids were recorded. A sub-sample of emerged females was also measured for body size (ovipositor and hind tibia length). Tests consisted of 14 and 24 replicates for D. suzukii and D. melanogaster, respectively.

Data analyses

Infestation percentages by D. suzukii or other drosophilids (estimated as number of emerged adults), and parasitism rates in sampled fruits or in fruit-baited traps were subjected to Analysis of Variance (ANOVA) separately for each collection year. The effect of host species on A. japonica’s juvenile development, fertility, and body size was also compared with ANOVA. Prior to the Analysis of Variance (ANOVA), percentage data were logit transformed as needed to normalize the data distribution. All analyses were performed using JMP V11 (SAS 2011, Cary, NC, USA).

Results

Field collections

In 2013, a total of 3266 Drosophila puparia were collected in South Korea and imported to the quarantine facilities at Berkeley (Table 1). Drosophila suzukii adults emerged mainly from collected blackberries, whereas the uninfested-fruit-baited traps primarily captured other drosophilid species (Table 1). There were 97 dead parasitized puparia, 728 dead unparasitized puparia, and 1,085 dead puparia for which parasitism status was unknown. Seven parasitoid species (44 individuals) were recovered from drosophilids collected during 2013: A. japonica (45.5 % of the recovered parasitoids), A. leveri (25 %), L. j. japonica (9.1 %), L. j. formosana Novković & Kimura (2.3 %), L. boulardi (2.3 %), P. vindemiae (6.8 %), and T. drosophilae (9.1 %). Among these, L. j. formosana and L. j. japonica emerged only from D. suzukii; A. japonica, A. leveri, and T. drosophilae emerged from both D. suzukii and other drosophilids; and L. boulardi and P. vindemiae emerged from other drosophilids (Table 2).

Table 2 Total number and number of female (in parenthesis) parasitoid species that emerged from collected D. suzukii and other drosophilids in South Korea in different geographic locations
Full size table

In 2014, a total of 20,358 puparia were collected from a variety of habitats, including wild Rubus located in inland forests and blackberries at one commercial farm (Table 1). Collections of wild Rubus in five of the locations yielded no drosophilid pupae (Table 1). Drosophila suzukii comprised the majority of flies that emerged from Rubus but no D. suzukii emerged from fruit-baited traps in the same localities (Table 1). From this material, there were 116 dead parasitized puparia, 5271 dead unparasitized puparia, and 8383 dead puparia for which parasitism status was unknown. A total of 181 wasps emerged, representing six species of larval parasitoids: A. japonica (60.2 % of the recovered parasitoids), A. leveri (2.2 %), A. brevicauda (2.8 %), L. j. japonica (14.9 %), L. j. formosana (1.7 %), and G. brasiliensis (18.2 %). Among these, A. brevicauda is a new species. Most important for this survey is that A. brevicauda, L. j. japonica, and G. brasiliensis emerged only from D. suzukii, whereas A. japonica and A. leveri emerged from both D. suzukii and other drosophilids, and L. j. formosana emerged only from other drosophilids (Table 2). The two northern Korean locations (Munsusa and Baekyeonsan) yielded no parasitoids (Table 2). Both the braconids and figitids co-existed in seven locations, but the figitids also occurred in another two locations (Table 2). Up to four different larval parasitoids were reared from D. suzukii in a single geographic location.

The percentage of D. suzukii among all emerged flies was higher from fruit samples than from fruit-baited traps in both 2013 (F 1,76 = 68.7, P < 0.001) and 2014 (F 1,192 = 1545, P < 0.001) (Fig. 2a). In 2013, mean parasitism of D. suzukii in natural fruit samples (range: 0–17.4 %) was significantly lower than in fruit-baited traps (range: 0.0–33.3 %) (F 1,34 = 4.5, P = 0.042), but mean parasitism of other drosophilids was similar between these two sampling methods (range: 0–33.3 % in fruit samples, and 0–39.1 % in fruit traps) (F 1,46 = 1.3, P = 0.26) (Fig. 2b, c). In 2014, parasitism of D. suzukii was 2.8 ± 0.4 % (n = 171) (range: 0–23.5 %) in fruit samples (no D. suzukii was found in fruit traps). Parasitism of D. suzukii in fruit samples highly varied among locations, but the overall parasitism by braconids (range: 0–10.5 %) was similar to that by figitids (0–8.8 %) (Fig. 3). Mean parasitism of other drosophilids was higher in fruit traps (range: 0–33.3 %) than in fruit samples (range: 0–12.5 %) (F 1,48 = 22.0, P < 0.001) (Fig. 2b, c).

Fig. 2
figure 2

Effect of sampling methods (fruit trap vs. fruit sample) on (a) the percentage of D. suzukii that emerged, (b) parasitism of D. suzukii or (c) other drosophilid species during the 2013 and 2014 exploration in South Korea. Data were pooled from all locations; values are mean ± SE and different letters above the bars indicate significant difference between the two different methods within each year (ANOVA, P < 0.05)

Full size image
Fig. 3
figure 3

Mean (±SE) parasitism of D. suzukii by braconids (Asobara) and figitids (Leptopilina and Ganaspis) from wild Rubus in different locations during the 2014 exploration in South Korea; numbers in brackets after each collection site refer to number of samples collected

Full size image

Quarantine rearing and Asobara japonica performance

Of the eight parasitoid species, we were able to establish colonies of five species for further testing in quarantine: A. japonica, L. j. japonica, G. brasiliensis, T. drosophilae, and P. vindemiae. We were unable to develop colonies of A. leveri, L. j. formosana, and A. brevicauda, because we either did not collect any females (A. brevicauda) or collected too few females during any single collection period (A. leveri, and L. j. formosana) (Table 2). We additionally reared some males and one female L. boulardi from drosophilids other than D. suzukii. The limited reared material prevented the development of a colony.

Each of the five tested parasitoid species (A. japonica, L. j. japonica, G. brasiliensis, T. drosophilae, and P. vindemiae) was able to attack and develop from both D. suzukii and D. melanogaster cultured on artificial diet or fruits (Table 3). Detailed quarantine studies are ongoing, but initial screening showed that the numbers of offspring produced per A. japonica female were similar between D. suzukii (8.0 ± 0.7) and D. melanogaster (6.9 ± 0.7) (F 1,37 = 1.1, P = 0.311). However, female parasitoids developed faster on D. melanogaster (27.1 ± 0.2 days) than on D. suzukii (25.9 ± 0.1 days) (F 1,303 = 19.5, P < 0.001), but perhaps as a consequence females reared from D. melanogaster larvae were smaller than those reared from D. suzukii (hind tibia length: F 1,50 = 28.4, P < 0.001; ovipositor length: F 1,50 = 21.6, P < 0.001, Fig. 4).

Table 3 Numbers of host flies and parasitoids emerged in quarantine rearing of three larval parasitoids (A. japonica, L. j. japonica, and G. brasiliensis) and two pupal parasitoids (P. vindemiae and T. drosophilae) imported from South Korea when exposed to hosts in different media (artificial diet or different fruit species, n is the number of replicates for each parasitoid, host and media combination)
Full size table
Fig. 4
figure 4

Body size of developed female A. japonica from D. suzukii and D. melanogaster: different letters above the bars indicate significant difference between different host treatments (ANOVA, P < 0.05)

Full size image

Discussion

The current study is the first report of a foreign exploration effort for Asian parasitoids of D. suzukii in part of its native range. Six species of larval parasitoids (A. brevicauda, A. japonica, A. leveri, L. j. japonica, L. j. formosana, and G. brasiliensis) and one pupal parasitoid (T. drosophilae) were recorded from D. suzukii in South Korea. Another pupal parasitoid species, P. vindemiae, was field-collected from drosophilid species other than D. suzukii in South Korea, but it was able to attack and develop on D. suzukii in quarantine. Colonies of five out of eight parasitoid species were successfully established using D. suzukii as host. G. brasiliensis was recently re-described as a new combination by Buffington and Forshage (2016).

The estimated parasitism rates varied largely without any clear pattern of location, sampling method, or site. However, parasitism by these larval parasitoids could be underestimated for several reasons. First, at the time when fruits were field-collected, parasitized larvae could have already left the fruits and pupated in the soil, while eggs and young larvae may not have yet been attacked. When these fruits were brought to the South Korean laboratories where the pupae were isolated for shipment to California, there would have already been a bias against pupal parasitoids because many of the larval-pupal parasitoids might have already exited the host fruit. A more precise estimate of the parasitism would be to collect only those pupae that pupated within 1–2 weeks (i.e., when they were at the suitable larval stages at the time of collection) following the collection. Second, mortalities of the imported pupae were high, partly because all collected materials had to be completely sealed during the transposition resulting in mold growth within the Petri dishes and our observation of numerous dead pupae from unknown factors.

In our South Korean collections, A. japonica was the most commonly recovered parasitoid. It was collected in four South Korean provinces and in particular it was recovered in 3 of 10 sampled sites in 2013 and 6 of 18 sampled in 2014. Moreover, 128 A. japonica specimens were obtained out of the 141 total Asobara spp. that emerged in the quarantine laboratory. These results agree with those obtained in the few previous studies of Asian drosophilid parasitoids, A. japonica is the most common and widespread parasitoid and is capable of parasitizing a number of Drosophila species (Ideo et al. 2008; Mitsui et al. 2007; Mitsui and Kimura 2010). Nomano et al. (2015) investigated the host use of eight Japanese Asobara species; six of these species (A. japonica, A. tabida, A. rossica, A. rufescens, and two unidentified Asobara spp.) emerged from D. suzukii breeding on wild Cerasus and Vaccinium fruits. In our survey, A. japonica was commonly collected from D. suzukii in wild Rubus. Interestingly, based on adult emergence in the quarantine, the A. japonica populations collected in our study were 99.8 % female. Similarly, populations of A. japonica on the main islands of Japan were parthenogenetic, whereas those in the subtropical islands were sexually reproducing (Mitsui et al. 2007). However, earlier work by Mitsui et al. (2007) and our study may not be directly comparable as the earlier Japanese survey used banana fruit-baited traps, which tend to attract numerous drosophilids in general, but fewer D. suzukii (XG Wang pers. observ.). On the other hand, recent studies reported variations in genetic, physiological, or ecological traits among different geographic populations of A. japonica, showing a high degree of geographic adaptations to various host habitats (Murata et al. 2009).

With the exception of A. japonica, there is little information on the parasitoid species reared in terms of their geographic distribution in Asia or their host species range. Kasuya et al. (2013) reported three species (L. japonica, Ganaspis sp., and one unidentified Asobara species) from D. suzukii on wild cherry fruit (Prunus donarium Sieb.). The Ganaspis sp. strain obtained from D. suzukii exhibits a high level of specificity for D. suzukii and was referred to suzukii-associated type of G. xanthopoda (Kasuya et al. 2013), while other stains do not appear to have the same level of specificity (Mitsui and Kimura 2010). The two generalist pupal parasitoids T. drosophilae and P. vindemiae are cosmopolitan (Carton et al. 1986; Chabert et al. 2012; Gabarra et al. 2015; Kacsoh and Schlenke 2012; Rossi Stacconi et al. 2015). As expected, these parasitoid species were found in South Korea and readily developed on D. suzukii in the quarantine. In the current study, all three species of figitids were collected in South Korea from D. suzukii collected in fruit samples, especially G. brasiliensis, which was found in multiple locations. We also report one new species (A. brevicauda) and two other species (A. leveri and L. j. formosana) with new host-association records on D. suzukii. It must still be verified whether or not the new species are the same as one of those undescribed species in Japan (Kasuya et al. 2013; Nomano et al. 2015).

Interestingly, we did not recover A. tabida or G. xanthopoda, which were reported to occur over a relatively wide range of the Japanese islands (Mitsui et al. 2007; Nomano et al. 2015). In earlier surveys, a few individuals of A. tabida were collected from D. suzukii in Japan (Mitsui et al. 2007; Nomano et al. 2015) and this host was the focus of our collection efforts in South Korea. However, several laboratory experiments have showed that A. tabida could oviposit but did not survive in D. suzukii (Chabert et al. 2012; Kacsoh and Schlenke 2012). Nomano et al. (2015) suggested several possible causes for this inconsistency in A. tabida and G. xanthopoda host use across different geographic ranges with relatively similar environmental conditions, including misidentification of host species in the earlier collections, host stage influence on parasitoid survival, and kleptoparasitism. Previous studies have shown that host age affected the survival of A. tabida in D. melanogaster (Van Alphen and Janssen 1982), and A. tabida survived better in D. simulans (Sturtevant) previously parasitized by L. boulardi (Kraaijeveld 1999). It is also possible that geographic variations occur in the flies’ immune resistance against parasitoids or in parasitoid species for their preference or effectiveness against flies (Rossi Stacconi et al. 2015).

While uninfested-fruit-baited traps provided a quick means of monitoring the presence of adult parasitoids that attack frugivorous Drosophila, it appeared in our collections to be relatively unattractive to South Korean D. suzukii. We found that the uninfested-fruit-baited traps captured mainly other drosophilid species, thus disrupting the search for specialist parasitoid species. In contrast, our collections of field-infested fresh fruits yielded primarily D. suzukii and with this more host-specific parasitoids, which was similar to previous collections in Japan (Kasuya et al. 2013; Mitsui et al. 2007; Nomano et al. 2015), although the authors did not always emphasize the differences in collection methods. In Japan, Kasuya et al. (2013) reported that D. suzukii was the only fly species breeding on wild cherry fruit. Both the host-specific Asobara species (Nomano et al. 2015) and D. suzukii-associated type of G. xanthopoda (Kasuya et al. 2013) were collected from wild fruits. Fruit traps could also influence the estimated rate of parasitism as both entering flies and parasitoids could reside in the traps for considerable periods of time, permitting repeated oviposition by the same species.

In the Berkeley quarantine, A. japonica, L. j. japonica, G. brasiliensis, T. drosophilae, and P. vindemiae readily attacked D. suzukii. This assessment was important to verify that, at least in the laboratory, the North American strain of D. suzukii (from Parlier, CA) did not show complete resistance to the five tested parasitoid species, as was shown in earlier studies (Kacsoh and Schlenke 2012; Poyet et al. 2013). In these preliminary quarantine studies, we also used D. melanogaster as a potential alternative and widespread host. This host was physiologically suitable for all the tested parasitoid species, thus potentially increasing the chances of these parasitoids to survive in the wild even without the availability of the pest. We also found a positive correlation between the host size and the size of emerged A. japonica, thus potentially favoring D. suzukii over other drosophilids as a preferred host. However, a more accurate assessment of the ecological suitability of D. melanogaster for these parasitoid species is needed before predicting its real potential as an alternative host species in the field (Desneux et al. 2012). To this goal of finding specialized parasitoid species that exclusively attack or strongly prefer D. suzukii, future extensive quarantine work is designed to test host preference on a series of non-target fly species. A. japonica was the more common parasitoid reared from D. suzukii (Table 2). If strains of this, or other species, have clearly better performance on or preference for D. suzukii compared with other fly species, an argument could be made for their release in some regions.

In conclusion, some specialized D. suzukii parasitoids in the genera of Asobara, Leptopilina, and Ganaspis seem to be present in South Korea and their specialization may be preferable as biological control candidates for D. suzukii populations that occur in North America and Europe. However, further detailed evaluations are needed to determine their effectiveness and safety with regard to non-target risk in order to obtain release permits in the US and/or in Europe. Moreover, the highest levels of D. suzukii parasitism recorded in South Korea were by the more generalist parasitoid species, particularly A. japonica. Therefore, this initial collection shows promise for improved biological control of D. suzukii in some of its invaded regions. Furthermore, additional collections in Asia will likely result in novel species and species associations with D. suzukii. For example, in 2013 collections in South Korea (K. Hoelmer, E. Guerrieri, M. Giorgini, D.S. Choi) using banana-baited traps infested by D. suzukii yielded A. japonica, A. leveri, G. xanthopoda, and one new species Asobara brevicauda Guerrieri and Van Achterberg (Guerrieri et al. 2016), while collections in Yunnan Province, China using uninfested-banana-baited traps yielded A. japonica, A. leveri, G. xanthopoda, one undescribed Leptopilina sp. and four Asobara species that have been described as new including A. brevicauda, A. elongata Guerrieri and Van Achterberg, A. mesocauda Guerrieri and Van Achterberg, A. unicolorata Guerrieri and Van Achterberg and A. triangulata (Guerrieri et al. 2016). While the material was not processed in quarantine and screened for effectiveness against D. suzukii, these collections highlight the need for further exploration and quarantine screening, as well as more detailed studies of parasitoid-host associations in situ in the native range of D. suzukii. The wide geographic distribution of D. suzukii in Asia also suggests that surveys in other regions of Asia, habitats, and over longer periods and throughout the seasons will be needed to fully discover the diversity of its natural enemies.

Author contribution

KMD, XGW, JCM, PWS, KAH, and VMW conceived and designed the project. KMD and XGW wrote the initial manuscript, all co-authors helped editing the manuscript thereafter. XGW and AB analyzed the data and conducted the quarantine work. JCM, BM, PWS, YS, TK, CJ, DWL, BC, and HR conducted the foreign exploration in South Korea. EG, MG, and KvA identified all Asobara species, while MB identified all other parasitoid species. All authors read, revised, and approved the manuscript.