We report the first detection of Trissolcus japonicus, an exotic Asian egg parasitoid and the primary candidate for classical biological control of the invasive brown marmorated stink bug, Halyomorpha halys, in Canada. Twenty-eight Trissolcus japonicus emerged from an H. halys egg mass from a site heavily infested by H. halys in Chilliwack, British Columbia, in 2018. This egg mass was deployed and retrieved as part of ongoing sentinel egg mass surveys for natural enemies of H. halys from 2017–2018 in coastal and interior British Columbia (total of 1,496 egg clusters at 16 sites). The identification of T. japonicus was based on biology (high levels of successful emergence from H. halys eggs), morphology, and mitochondrial DNA sequences. Trissolcus japonicus was not detected at any other survey sites in 2017–2018; however, three species of indigenous egg parasitoids were found attending or emerging from H. halys egg masses at low levels (<4%) at several sites. The origin of the detected T. japonicus, the extent of its establishment in British Columbia, and its ultimate impact on H. halys populations remain to be determined. Nonetheless, the detection of this exotic biological control agent in Canada concurrently with regulatory review of its intentional importation and release is emblematic of the current uncertainty around regulatory control on the movement of biological control agents across borders.

adventive establishment, classical biological control, brown marmorated stink bug, Halyomorpha halys

Classical (= importation) biological control of invasive pests, where natural enemies are imported and intentionally introduced from a pest’s area of origin, involves years of research to assess risks and benefits of proposed introductions, followed by regulatory approval (Bigler et al. 2006, Cock et al. 2016, Heimpel and Cock 2018). However, there is increasing recognition that unintentional introductions of natural enemies are probably common, introducing a high level of uncertainty to the regulatory process for biological control introductions (Mason et al. 2017a).

The samurai wasp, Trissolcus japonicus (Ashmead) (Hymenoptera: Scelionidae), has become a prominent case study for the establishment of a candidate biological control agent outside of its native range prior to a decision by regulatory authorities on the appropriateness of release (Servick 2018). Its host, Halyomorpha halys (Stål) (Hemiptera: Pentatomidae), also native to Asia, is an invasive alien pest that has caused extensive economic damage and increases in insecticide use in a wide range of crops in areas of the United States and Europe where it has become established (Leskey and Nielsen 2018). Halyomorpha halys is also a nuisance pest in human dwellings. A classical biological control program for H. halys was initiated in the USA and T. japonicus was identified as the most promising candidate for introduction based on high parasitism rates of H. halys (typically ~60–90%) in Asia (Qiu et al. 2007, Yang et al. 2009, Zhang et al. 2017). However, in 2014, while non-target host range testing was still underway, a population of this parasitoid was recovered in nature from sentinel egg masses in Maryland, USA, indicating that it had been introduced accidentally (Talamas et al. 2015a). Other adventive populations were found in Oregon and Washington State (USA) in 2015 (Milnes et al. 2016, Hedstrom et al. 2017). More adventive populations of T. japonicus in both the Pacific Northwest and the northeastern USA continue to be discovered (stopbmsb.org), and intentional redistributions are now taking place within some states (K.A. Hoelmer, personal communication). Unexpectedly, in 2017 and 2018, adventive populations of T. japonicus were also detected in Switzerland (Stahl et al. 2018) and Italy (Sabbatini et al. 2018), suggesting that like its host, T. japonicus is becoming a “global invader”.

Trissolcus japonicus has not been detected previously in Canada, where H. halys populations have established recently (Gariepy et al. 2014, Abram et al. 2017a). Thus, intentional introduction of T. japonicus to control H. halys populations in Canada would require regulatory approval based on review of a petition for release (Mason et al. 2017b). Here we report the detection of T. japonicus in British Columbia, Canada, representing another remarkable instance where this parasitoid has arrived in a country before a regulatory decision has been made regarding the appropriateness of its introduction.

From May to September in each of 2017 and 2018, a total of 1,496 H. halys sentinel egg masses (= 41,351 eggs) were set out at 16 field sites in coastal and interior British Columbia where large, established breeding populations of H. halys are present (Table 1). Eggs were then retrieved to measure parasitism levels and parasitoid species composition. All sentinel sites were in urban, suburban, and backyard settings with mixed woody and herbaceous vegetation. Halyomorpha halys egg masses, laid on Reemay® polyester fabric (Avintiv, USA), were collected from H. halys laboratory colonies within 24 hours of being laid and either placed in the field the same day or stored at 10 °C to delay development for up to a week before they were deployed. The fabric substrate holding the eggs was stapled to the undersides of the leaves of wide variety of host plants infested by H. halys, mostly woody trees (e.g. Prunus spp., Davidia spp., Acer spp., Gleditsia spp., Ailanthus spp., Sorbus spp.) and shrubs (e.g., Rubus sp., Mahonia spp., Symphoricarpos spp., Rosa spp.). Sentinel egg masses were retrieved from the field within four days, before the emergence of H. halys nymphs. Parasitoids found attending egg masses at recovery (indicating post-oviposition brood guarding behavior; see Abram et al. 2014, Cornelius et al. 2016) were also collected. Egg masses were then kept in Petri dishes (50 mm diameter, 9 mm depth) under ambient laboratory conditions to assess parasitoid emergence. Attending and emerging egg parasitoids were preserved in 95% EtOH, then point-mounted and identified to species using the key of Talamas et al. (2015b). Finally, all egg masses were dissected under a stereomicroscope to verify that non-emerged eggs did not contain parasitoids.

While three species (26 total individuals) of indigenous egg parasitoids [Trissolcus euschisti (Ashmead), Trissolcus cosmopeplae (Gahan), and Telenomus podisi (Ashmead) (Hymenoptera: Scelionidae)] were found attending H. halys sentinel egg masses upon recovery, successful emergence of parasitoids from H. halys eggs was rare (Table 1). Of these species, only T. euschisti successfully emerged. Less than one fourth of the eggs in each of these masses were parasitized successfully (average of 22.5 ± 1.4%; mean ± SE, n = 6), and emerging parasitoids produced few or no offspring when subsequently offered H. halys eggs in the laboratory. These findings are consistent with past surveys and laboratory trials in other areas of North America and Europe showing that attack of H. halys egg masses by indigenous egg parasitoids is probably common (Gariepy et al. 2018), but their offspring are usually unable to complete development successfully (Abram et al. 2014, Haye et al. 2015, reviewed in Abram et al. 2017b). In contrast, all 28 eggs (100%) of one egg cluster deployed at a site highly infested by H. halys in Chilliwack, BC on August 23, 2018 were parasitized, and emerging offspring completely parasitized a number of H. halys egg masses offered in the laboratory with >90% successful offspring emergence.

Specimens were identified to species using the key to Nearctic Trissolcus by Talamas et al. (2015b) and are fully congruent with the concept of T. japonicus presented by Talamas et al. (2015b, 2017). Specifically, the presence of 4 clypeal setae (Fig. 1A) and well-defined episternal foveae that extend from the postacetabular sulcus to the mesopleural pit (Fig. 1B) unambiguously separate T. japonicus from the Nearctic fauna. Additionally, the absence of rugae on the mesoscutum (Fig. 1C) and the absence of a smooth area below the median ocellus (Fig. 1A) confirm that it is neither of the Palearctic species closest to T. japonicus, T. kozlovi and T. plautiae (Talamas et al. 2017). Voucher specimens are deposited in the Florida State Collection of Arthropods and the Canadian National Collection of Insects (Table 2). The collection data, including host associations, for all voucher specimens are deposited in the Hymenoptera Online Database (hol.osu.edu).

Figure 1. 

Trissolcus japonicus female (FSCA 00033107) from Chilliwack, British Columbia: A head, anterior view, cs: clypeal setae B head and mesosoma, anterolateral view, ats: postacetabular sulcus, eps: episternal foveae, mpit: mesopleural pit C head, mesosoma, metasoma, dorsal view. Scale bars in millimeters.

DNA was extracted from 5 specimens using a chelex DNA extraction protocol, and the universal primers LCO-1490 and HCO-2198 (Folmer et al. 1994) were used for amplification and sequencing of the DNA barcode region of the Cytochrome Oxidase I (COI) gene (as described by Gariepy et al. 2014). All 5 specimens yielded identical COI sequences of 643-bp in length, and a representative sequence was submitted to GenBank (Accession number: MK188349) and uploaded to the North American Scelionidae DNA barcode database (project NSCEL) available on the Barcode of Life Datasystems (BOLD, http://www.boldsystems.org). In comparison to public sequences available in the NSCEL database, these specimens shared 100% sequence similarity with voucher T. japonicus collected from established populations in the USA, and shared 99–100% sequence similarity with voucher specimens collected in Asia. The small amount of variation among T. japonicus specimens is likely due to intraspecific variation between individuals collected from geographically distinct regions (e.g., Stahl et al. 2018).

This detection of T. japonicus in Canada occurred while a petition for the release of this biological control agent was under review by the Canadian Food Inspection Agency (CFIA), the national regulatory authority in Canada. We are not aware of any historical cases where this has occurred, in Canada or elsewhere, and its implications for the prospects of intentionally importing and releasing T. japonicus in Canada remain to be seen. It is important to note that because T. japonicus has been detected only at a single site in one year, we cannot yet definitively conclude that this species is established in Canada. However, given the relative proximity (<400km) of the closest known established populations in Washington State (Milnes et al. 2016, stopbmsb.org), it is plausible that this detection is indicative of a range expansion of adventive T. japonicus populations, and that the parasitoid is in the early phases of establishment. Very low parasitism levels, as we observed here, were also characteristic of the initial detections in other areas where adventive T. japonicus populations have since been confirmed and are spreading (Talamas et al. 2015a, Milnes et al. 2016). In addition, conservative climate suitability modeling has predicted that several areas of southern Canada, including British Columbia, are suitable for T. japonicus establishment, survival, and reproduction (Avila and Charles 2018). Continuing field surveys and extensive phylogeographic analyses using microsatellite DNA markers are underway to track the establishment and biological control impact of T. japonicus in Canada, and to reconstruct potential pathways of introduction.

We thank Warren Wong, Emily Grove, Kennedy Bolstad, Peggy Clarke, Jason Thiessen, Jessica Leung, Nemo DeJong, Keith DeGlow, Kandace Zurowski-Tiffin, Gary Judd, Laura Keery, Don Magnusson, and Allison Bruin for support in the laboratory and field. We also thank Tracy Hueppelsheuser for referring us to sites infested by H. halys in Coastal BC, and all homeowners who granted access to their property. Susan Halbert provided helpful edits on an earlier version of the manuscript. Funding for this research was provided to PKA, TDG, and PGM by Agriculture and Agri-Food Canada (A-BASE #2362); funding to SA was from was from Investment Agriculture Foundation of BC, Canadian Agricultural Partnerships, BC Cherry Association, BC Wine Grape Council, and BC Fruit Growers Association. We also thank the Florida Department of Agriculture and Consumer Services – Division of Plant Industry for their support on this contribution.