In recent years, there has been an increase in the adventive establishment and spread of parasitoid wasps outside of their native range. However, lack of taxonomic tools can hinder the efficient screening of field-collected samples to document the establishment and range expansion of parasitoids on continent-wide geographic scales. Here we report that Leptopilina japonica (Hymenoptera, Figitidae), a parasitoid of the globally invasive fruit pest Drosophila suzukii (Diptera, Drosophilidae), is now widespread in much of North America despite not having been intentionally introduced. Surveys in 2022 using a variety of methods detected L. japonica in 10 of 11 surveyed USA States and one Canadian Province where it was not previously known to occur. In most surveys, L. japonica was the most common species of D. suzukii parasitoid found. The surveys also resulted in the detection of Ganaspis cf. brasiliensis (Hymenoptera, Figitidae), the recently-released biological control agent of D. suzukii, in six USA States where it had not previously been found. These new detections are likely a result of intentional biological control introductions rather than spread of adventive populations. A species-specific multiplex PCR assay was developed as a rapid, accurate and cost-effective method to distinguish L. japonica, G. cf. brasiliensis, the closely-related cosmopolitan parasitoid Leptopilina heterotoma (Hymenoptera, Figitidae) and other native parasitoid species. This dataset and the associated molecular tools will facilitate future studies of the spread and ecological impacts of these introduced parasitoids on multiple continents.

Adventive establishment, DNA barcoding, Drosophila suzukii, Ganaspis brasiliensis, Hymenoptera, molecular diagnostics, multiplex PCR, spotted-wing drosophila, unintentional biological control

The rate of global introductions of non-native insect species due to human activities continues to accelerate (Seebens et al. 2017; Seebens et al. 2021). Many of these introductions result in widespread negative environmental, human health and economic impacts (Bellard et al. 2016; Diagne et al. 2021). A minority of these introductions may have some positive consequences or a mixture of positive and negative effects (Schlaepfer et al. 2011; Vimercati et al. 2020; Sax et al. 2022). The potential for desirable or mixed effects is particularly likely with unintentionally introduced parasitoids and predators of invasive insects (Roy et al. 2011; Weber et al. 2021). While it is increasingly recognised that unintentional introductions of species occupying higher trophic levels are probably common (Weber et al. 2021), many are likely undetected and, as such, there are relatively few examples that have been well-documented on continental scales.

Parasitoid wasps are one of the most biodiverse groups of animals on the planet (Forbes et al. 2018) and are the most common group of biological control agents used to target invasive insects (Heimpel and Mills 2017). Parasitoid wasps have frequently been intentionally introduced as biological control agents worldwide, successfully controlling numerous invasive pest species in forestry, agriculture, and conservation contexts (Mason 2021). Host range is the strongest filter used to select parasitoid species for modern importation biological control programmes. To minimise the potential for non-target ecological impacts, only natural enemies with relatively narrow host ranges are selected from the native range of an invasive pest (Bigler et al. 2006; Hoddle et al. 2021). However, parasitoids from the native range of an invasive pest, even those with broader host ranges that may be less suitable for biological control programmes, can occasionally establish themselves in regions previously invaded by their host. These unintentional introductions can occur outside the scope of the standard biological control regulatory process (e.g. Frewin (2010); Talamas et al. (2015); Mason et al. (2017); Peverieri et al. (2018); Weber et al. (2021)). There is growing interest in documenting the spread of unintentionally introduced parasitoids and assessing the balance of their positive and negative ecological and environmental impacts (Mason et al. 2017; Weber et al. 2021).

One major challenge for documenting the spread of introduced parasitoids in their new geographic ranges is a taxonomic impediment. For many parasitoid groups, the native versus non-native parasitoid fauna itself may not be entirely known and may be difficult to distinguish based on morphology. Further, it may be unfeasible or inefficient for the limited number of taxonomic experts on a given group of parasitoids to provide identification services for numerous research teams conducting large-scale surveys. The use of molecular tools, including species-specific PCR primers, for parasitoid species identification can help alleviate this problem (Gariepy et al. 2005; Traugott et al. 2006; Gariepy et al. 2007, 2008; Rugman-Jones et al. 2011, 2020; Gariepy and Messing 2012; Shariff et al. 2014; Furlong 2015; Shimbori et al. 2023). However, their development requires DNA sequences that are tied to authoritatively identified specimens (Lue et al. 2021; Shimbori et al. 2023), which is often not the case, given that many of the world’s parasitoid fauna remain undescribed (Forbes et al. 2018). Fortunately, when an economically-important invasive pest is the focus of an importation biological control programme, the specimens and taxonomic knowledge may be available to facilitate the development of molecular diagnostic tools that can be used to document the spread and impact of unintentionally introduced parasitoids.

These dynamics are currently playing out on a global scale for parasitoids of spotted-wing drosophila, Drosophila suzukii Matsumura (Diptera, Drosophilidae), an invasive vinegar fly that has become a significant pest in all major fruit production areas in the Americas, Europe and parts of Africa (Asplen et al. 2015; Tait et al. 2021). Although the first report of D. suzukii establishment outside of its native range was in Hawaii in 1980 (Kaneshiro 1983), subsequent spread was not reported until 2008 when D. suzukii was found in California (Hauser 2011). By 2011, populations of D. suzukii were well-established in fruit-growing areas in western and eastern North America (Tait et al. 2021), presumably as a result of movement and spread of already-established populations from the western US (Fraimout et al. 2017). Surveys for parasitoids soon after the invasion of North America and Europe did not detect any specialised parasitoids of D. suzukii (Lee et al. 2019; Wang et al. 2020). Foreign exploration for parasitoids of D. suzukii in its native range of Asia documented a total of 21 parasitoid species (summarised in Wang et al. 2020), with three dominant parasitoids attacking D. suzukii larvae: Asobara japonica Belokobylskij (Hymenoptera, Braconidae), Ganaspis cf. brasiliensis Ihering and Leptopilina japonica Novković & Kimura (Hymenoptera, Figitidae) (Mitsui et al. 2007; Daane et al. 2016; Girod et al. 2018a; Giorgini et al. 2019). One strain of G. cf brasiliensis (called “G1”) was selected and approved for intentional biological control releases in 2021 in the USA and Italy because of its specificity to D. suzukii (Biondi et al. 2021; Daane et al. 2021; Lisi et al. 2021; Fellin et al. 2023). Another genetically distinct lineage of G. cf. brasiliensis (referred to as G3) co-occurs with G1 in East Asia; however, the two strains are reproductively incompatible (Seehausen et al. 2020) and will be described as two distinct species (M. Buffington, in prep). Ganaspis cf. brasiliensis G3 has a broader host range that includes other Drosophila species and was, therefore, not considered for intentional release (Girod et al. 2018a; Giorgini et al. 2019; Seehausen et al. 2020; Daane et al. 2021). Similarly, approval for the release of L. japonica and A. japonica was not pursued because their host range included several species of Drosophilidae other than D. suzukii (Girod et al. 2018b; Giorgini et al. 2019; Daane et al. 2021; Seehausen et al. 2022).

However, just prior to intentional releases of G. cf. brasiliensis G1, field surveys in the Pacific Northwest of North America revealed the presence of adventive (presumed to be accidentally introduced) populations of L. japonica (as early as 2016) and G. cf. brasiliensis G1 (in 2019) in British Columbia, Canada (Abram et al. 2020). By 2020, it was clear that both species were well-established in British Columbia, parasitising D. suzukii in diverse habitats (Abram et al. 2022a). Both parasitoid species were subsequently found in Washington, USA in 2020 and 2021, respectively (Beers et al. 2022). Similarly, adventive populations of L. japonica were discovered in northern Italy in 2019 (Puppato et al. 2020) and Germany in 2021 (Martin et al. 2023). These observations suggest that L. japonica and G. cf. brasiliensis may be rapidly spreading to areas outside their native range due to unintentional introductions. In North America, however, the distribution of adventive populations of these parasitoids outside of the Pacific Northwest was unknown.

During 2022, multiple research groups independently initiated new surveys to document the parasitoid community of D. suzukii as part of the intentional release programme for G. cf. brasiliensis G1. However, given the large number of samples involved in these surveys, efficiently documenting the presence of L. japonica, in particular, was challenging and time-consuming because it is exceedingly difficult to distinguish from related native fauna, especially the cosmopolitan species L. heterotoma Thompson (Abram et al. 2020; Abram et al. 2022b). In fact, before having access to specimens of L. japonica from Asia, the leading taxonomic expert on Figitidae in North America misidentified the first L. japonica found in North America as L. heterotoma (Abram et al. 2020).

In this study, we consolidated the findings of research groups surveying for parasitoids of D. suzukii in 12 USA States and one Canadian Province. We used this combined dataset to assess how geographically widespread L. japonica and G. cf. brasiliensis G1 are in the surveyed areas of North America. In addition, to address the difficulty of parasitoid identification in these and future surveys, DNA barcodes were generated from field-collected specimens to confirm the identity and to generate an authoritative DNA barcode library for those species likely encountered in D. suzukii parasitoid surveys. Then, a single-step multiplex PCR assay was developed for G. cf. brasiliensis G1, L. japonica and L. heterotoma to facilitate screening of large numbers of field-collected samples for the presence of adventive and intentionally introduced parasitoid species. We anticipate that continent-wide survey data, combined with new molecular diagnostic tools, will help provide information for future strategies to release and re-distribute D. suzukii parasitoids and will serve as a baseline for measuring the balance of positive and negative ecological effects of unintentionally introduced parasitoid species.

The geographic range of L. japonica was historically restricted to Asia and although this species was not being considered as a candidate agent for intentional biological control introductions, this study reveals that it is now present throughout a large part of North America and appears to be the dominant parasitoid associated with invasive D. suzukii at this point in time. This represents one of few documented cases of what appears to be a rapid (< 10 years) continent-wide spread of an unintentionally introduced parasitoid wasp, with the most similar example being the ongoing spread of unintentionally introduced egg parasitoids of Halyomorpha halys Stål (Hemiptera, Pentatomidae) in North America and Europe (Talamas et al. 2015; Stahl et al. 2019a).

This study reports a considerable North American range expansion for L. japonica, with new detections of L. japonica in 10 USA States (Delaware, Georgia, Maine, Maryland, Michigan, New Jersey, New York, North Carolina, Oregon and Pennsylvania) and one Canadian Province (Ontario). We also confirmed its establishment in Washington State, where it had already been reported (Beers et al. 2022) following its initial detection in adjacent British Columbia, Canada in 2016 (Abram et al. 2020). Only one USA State surveyed, California, did not detect any L. japonica; however, 2022 collections in this State were focused on a single location. Although the timeline of L. japonica’s unintentional introduction and spread in North America is somewhat uncertain, it seems most likely that it has occurred relatively recently, i.e. since the widespread establishment of D. suzukii (~ 2011). This is supported by both published (Thistlewood et al. 2013; Miller et al. 2015; Lue et al. 2016; Wang et al. 2016; Huang et al. 2023) and unpublished survey results (see Suppl. material 4) from several locations in North America that took place between 2011 and 2021 that did not recover any evidence of the presence of L. japonica. The only prior reports of L. japonica were recent; in the Pacific Northwest by Abram et al. (2020) and Beers et al. (2022) and, in Oregon, where parasitoids recovered from an extensive fruit collection in September 2021 were later identified as L. japonica (Suppl. material 4). Similarly, in Europe, where there has been a considerable amount of field research on parasitoids of Drosophila spp. (Rossi Stacconi et al. 2013; Englert and Herz 2016; Mazzetto et al. 2016; Knoll et al. 2017; Kremmer et al. 2017; Shaw et al. 2023), it seems that L. japonica has only arrived since the introduction of D. suzukii (Puppato et al. 2020; Fellin et al. 2023; Martin et al. 2023). These observations support the “receptive bridgehead hypothesis” (Weber et al. 2021), in which unintentional introductions of biological control agents to new areas may be more likely following invasions of those areas by suitable hosts from their native range. As L. japonica has a host range that includes several native and cosmopolitan species of Drosophilidae other than D. suzukii, including the abundant and widespread Drosophila melanogaster Meigen (Girod et al. 2018b; Giorgini et al. 2019; Daane et al. 2021; Fellin et al. 2023), the L. japonica “invasion” has the potential to have both positive economic and environmental impacts [via suppression of D. suzukii populations (and/or other pestiferous Drosophila spp.)] and potential ecological harm (direct and indirect effects of attacking native and cosmopolitan Drosophilidae) that should be evaluated in the coming years.

Unintentionally introduced populations of G. cf. brasiliensis G1 (the more specialised parasitoid of D. suzukii recently approved for intentional releases in North America and Europe) do not appear to be nearly as widespread as L. japonica. However, it is important to note that this species was only recently released and is in the early stages of establishment in the locations surveyed in the present study. In Washington, populations of G. cf. brasiliensis were detected prior to intentional releases and the present study suggests additional spread of adventive populations to new locations within the state. However, the detection of G. cf. brasiliensis in the additional 6 US States (Delaware, Georgia, Maryland, New Jersey, Pennsylvania and Oregon) reported here only occurred in 2022, following intentional releases. There are records of G. cf. brasiliensis from tropical and sub-tropical regions of the Palearctic (Uganda), Nearctic (Mexico), Neotropical (Panama, Brazil) and Oceania (Hawai’i) (Buffington and Forshage 2016), but these represent strains of this species other than G1 that are likely to be described as distinct species (M. Buffington, in prep.). Given the absence of previous records of G. cf. brasiliensis outside of the Pacific Northwest and the very low abundance of G. cf. brasiliensis G1 relative to L. japonica outside of this area, we conclude that most of the G. cf. brasiliensis G1 identified in the present study represent detections of this parasitoid resulting from the intentional releases (and not adventive populations). Although it is possible that the G. cf. brasiliensis G1 detections in Oregon were from adventive populations (given the proximity to Washington where populations were present before intentional releases took place; Beers et al. (2022)), extensive releases of G. cf. brasiliensis G1 took place in Oregon in 2022, with the only detections occurring after release. As the DNA barcodes for both adventive and intentionally-released populations are identical (Abram et al. 2020; Lue et al. 2021), the methods used in the present study cannot distinguish between post-release detections and the spread of already-established adventive populations. Other methods and continued surveys will be needed to tease apart the contribution of released and adventive populations of G. cf. brasiliensis G1 to its range expansion in North America in the coming years.

The present study also yielded new information on the distribution of native and cosmopolitan parasitoids associated with Drosophilidae. Native and cosmopolitan Leptopilina species (L. maia , L. leipsi and L. heterotoma) were only found in the vinegar and Scentry traps and not in fruit collections or sentinel baits, as they are not known to parasitise D. suzukii in the field and these traps are not specific to D. suzukii. Leptopilina maia was identified from trap catches in North Carolina (USA) and Ontario (Canada). The occurrence of L. maia in North Carolina is consistent with the distribution of this species in eastern North America (Lue et al. 2016); however, this is the first time this species has been reported in Canada. Similarly, although known from New Hampshire, Illinois, Maryland and Virginia (Lue et al. 2016), this is the first report of L. leipsi in Canada. Both species were recently described by Lue et al. (2016) and the present study provides updated distribution records. The records of the cosmopolitan species L. heterotoma from Ontario are also the first official records of this species in Canada, although re-inspection of historically collected specimens from British Columbia (reported erroneously in Thistlewood et al. (2013) as “Ganaspis sp.”) indicates that it has been present in Canada since at least 2011 (P.K. Abram and D.R. Gillespie, unpublished data). The amount of new information about figitid geographic distributions gathered by this single study highlights the paucity of data on this important group of parasitoids and the uncertainty associated with inferring recent spread and establishment from survey data on understudied species groups.

Although the routes of invasion have been reconstructed for the global movement and spread of D. suzukii (Fraimout et al. 2017), the pathways of adventive introduction of L. japonica and G. cf. brasiliensis G1 are currently unknown and it is not yet clear whether they represent a single introduction with subsequent spread of the same introduced population or whether multiple introduction events have occurred. Given the large geographic distance between the initial introductions in the Pacific Northwest and the more recently-discovered establishments in eastern North America, it is possible that more than one introduction event took place and/or that a bridgehead event occurred, in which one introduced population served as the source for rapid dispersal and establishment of additional populations in North America (Guillemaud et al. 2011). However, further population genetic analysis of native and introduced populations would be necessary to identify the source population(s) and to trace the potential routes of entry and spread. Although the present study generated DNA barcode sequences for adventive populations in North America, sequences from additional genes and populations from across the native (and recently invaded) range could clarify source populations and infer pathways of entry. Further, additional genetic approaches may provide a higher resolution of population differences. For example, when COI barcode sequences were not adequate to detect population-level differences, ddRADseq was used to clarify relationships amongst adventive populations of Trissolcus japonicus Ashmead (Hymenoptera, Scelionidae), a parasitoid of H. halys (Abram et al. 2023) and between different populations of D. suzukii in Canada (Nelson et al. 2023).

Detection of introduced and adventive parasitoids in an invasive pest population is critical to document the establishment and spread of biological control agents, assess their effectiveness and evaluate their potential for long-term suppression of an invasive pest species (Gariepy et al. 2007; Furlong 2015; Lue et al. 2022). To do this, large numbers of samples may need to be collected and processed to detect parasitism and parasitoid species composition. The availability of a molecular diagnostic tool can facilitate the delivery of these data and provide information for subsequent pest management strategies, including continued release and redistribution of biological control agents (Gariepy et al. 2007, 2008). Monitoring efforts to detect new adventive populations and to assess the post-release establishment following intended releases can be facilitated using the multiplex PCR approach described here. Our method provides a rapid, accurate and cost-effective option for screening field-collected samples that does not require the additional expense of DNA sequencing. It also facilitates conclusive identification of species that would otherwise be difficult to separate, based solely on morphological features (e.g. L. japonica versus L. heterotoma; G1 versus G3 G. cf. brasiliensis). This is an important consideration when screening large numbers of samples for these species, as the cost of identifying a sample through bidirectional sequencing is approximately tenfold more than identification using the multiplex PCR protocol described here. Furthermore, the species-specific primers more reliably provided species identifications than sequencing the entire barcode region (Fig. 3), likely because they generate a shorter fragment of DNA which can improve identification of samples that fail to sequence due to DNA degradation (Strutzenberger et al. 2012; Hebert et al. 2013; Mitchell 2015). This multiplex PCR may also be useful to detect ecological interactions that are difficult to address using conventional techniques, including potential competitive interactions between parasitoids, detection of host-parasitoid associations and evaluation of the host range in the area of introduction (e.g. Gariepy and Messing (2012); Gariepy et al. (2019); Lomeli-Flores et al. (2019); Stahl et al. (2019b); Hepler et al. (2020)). It is important to note that, in the current context of screening figitid parasitoids that emerge from D. suzukii in ripe and sentinel fruit and species attracted to baited traps, the multiplex PCR protocol is effective and accurate. However, it may be prudent to screen additional verified species of Figitidae with this multiplex PCR protocol to ensure specificity is retained if this protocol is used in a different context with a more diverse parasitoid community (for example, figitids associated with other drosophilid species or for use in other countries with a different parasitoid complex associated with D. suzukii).

Our findings raise several questions relevant to interpreting the recent trend of unintentionally introduced populations of parasitoid wasps being detected during the course of biological control programmes. For example, is the apparently more widespread establishment of L. japonica compared to G. cf. brasiliensis G1 due to the fact that it has been established in these locations for a longer period of time? Or is it due to its broader host range and/or a broader climatic tolerance? If so, does that mean that parasitoids that are less likely to be approved for intentional biological control releases are more likely to be unintentionally introduced and become established? If the routes of introduction needed to spread L. japonica to multiple areas of the world are present and this parasitoid can attack Drosophilidae species other than D. suzukii, why did it seemingly not establish outside of Asia before the D. suzukii invasion? If a broader host range makes a parasitoid species more likely to be unintentionally introduced to new areas, why is A. japonica, a very common parasitoid of D. suzukii in some areas of its native range and which attacks a wide range of Drosophilidae species (Daane et al. 2021), not yet present in North America or Europe? The relative contributions of several factors to unintentional parasitoid establishment (e.g. parasitoid host range; parasitoid prevalence in source areas; climatic tolerance; patterns of global trade driving pest and parasitoid propagule pressure; host abundance in space and time) should be the focus of future work. Indeed, our study contributes to the increasing recognition that parasitoid wasps are part of the rising number of non-native insect introductions and the potential for them to have widespread positive, negative and mixed ecological impacts is considerable. However, the factors that contribute to their global spread and what this means for biological control of invasive species and risks to native ecosystems, deserve more attention.

The authors would like to thank the following people for their assistance with sample collection and processing: L. Jocius, A. Weber, J. Collins, E. Desbois, F. Drummond, A. Fischer, G. Wiggins, N. Flanagan, N. Oderkirk, C. Purser, C. Dial, N. Lasala, K. Beckwith, S. Hesler, O. McCall, Y. Grynyshyn, M. Freeman, E. Janasov, S. Harrington, I. McEvoy, K. Wallner, S. Olsen, T. Preston, K. Day, S. Ill, C. Lamberty, A. Young, V. Rogers, R. Saroka, M. Brubaker, K. Robel. M. Huizingh, S. Snowden, G. Johnson, J. Klakulak, R. Chave, R. Turner, C. Molokwu, R. Bryan, A. Antelman, A. Kemmerer, K. Kettmann, S. Rivas, A. Yee, B. Diehl, J. Gray, J. Johnson, M. Smytheman, L. Smytheman and R. Czokajlo. We would also like to thank the following people and organisations for facilitating the work: Maryland grower cooperators and University of Maryland farm staff (D. Price, B. Spielman and L. Simmons); TJ Hafner (Agricare), S. Robbins and D. Hinds-Cook (OSU Lewis Brown Farm); Ontario Ministry of Agriculture, Food, and Rural Affairs (OMAFRA) (H. Fraser, E. Pate) and the Berry Growers of Ontario (BGO), blueberry and cherry growers of Washington.