Field surveys were conducted in 2018 by personnel of the local Plant Protection Service (ERSA Friuli Venezia Giulia) in Cordenons commune (46.0082N, 12.6713E) as a part of routine monitoring of H. halys in the region. During these surveys, more than a dozen egg masses were found to be parasitized by T. mitsukurii (Sabbatini Peverieri et al. 2018). In Cordenons, on August 7, three H. halys egg masses collected on a Robinia pseudoacacia L. hedgerow near an IPM apple orchard appeared to be parasitized (visually detected by the dark color of the eggs). Collected egg masses were reared in a climatic chamber (26 °C, 65% RH, 16:8 L:D) until emergence of parasitoids. Emerged parasitoid specimens were stored in ethanol for further taxonomic and molecular analysis.

In the Veneto region, field pest surveys were conducted by personnel of the Department of Agronomy, Food, Natural Resources, Animals and Environment (DAFNAE), University of Padua, during the summers of 2017 and 2018. At three sites in the vicinity of Povegliano (45.7575N, 12.1872E), Montebelluna (45.7586N, 12.0174E) and Riese Pio X (45.7170N, 11.9397E), dozens of H. halys egg masses were collected in apple and kiwi orchards and vineyards implementing integrated pest management and in surrounding hedgerows of R. pseudoacacia, Acer campestris L., Sambucus nigra L. and Prunus sp. Collected egg masses were reared in a climatic chamber (26 °C, 65% RH, 16:8 L:D) until emergence of parasitoids. Emerged parasitoid specimens were stored in ethanol for further taxonomic and molecular analysis.

In 2018, during monitoring of H. halys, naturally laid egg masses of H. halys were collected in a parking area in the municipality of Ora (46.3620N, 11.2985E), in the province of Bolzano. Several species of maple trees (Acer platanoides L., Acer negundo L. and Acer pseudoplatanus L.), ailanthus [Ailanthus altissima (Mill.) Swingle] and linden (Tilia platyphyllos Scop.) were sampled for egg masses once or twice per week. Egg masses were collected between September 10 and October 26. Field-collected eggs were reared in a climatic chamber at 25 ± 1 °C, 65 ± 5% RH, 16:8 L:D until H. halys nymphs or parasitoids emerged. Emerged parasitoids were stored in 70% ethanol for further analysis.

In 2019, natural egg masses of H. halys (n = 11) and Palomena prasina L. (Hemiptera: Pentatomidae) (n = 6) were collected in an urban park area at the lake of Zurich (district Seefeld, 47.355912N, 8.550674E), where adventive H. halys populations were first detected in Europe in 2007 (Wermelinger et al. 2008). Trees inspected for eggs included Catalpa bignonioides Walter, Paulownia tomentosa (Thunb.) Steud., Tilia platyphyllos, and Liriodendron tulipifera L. Egg masses were collected on three occasions (June 28, July 2 and 10). Field collected egg masses were kept in small Petri-dishes and stored at 25 ± 1 °C, 60% RH, 16:8 L:D until nymphs and parasitoids had emerged. Emerged parasitoid specimens were stored in ethanol for further taxonomic and molecular analysis.

In 2017 and 2018, during monitoring activities of stink bugs and their egg parasitoids, two sites yielded naturally laid stink bug egg masses from which Acroclisoides emerged. One was an egg mass of H. halys collected from a pecan tree [Carya illinoinensis (Wangenh.) K. Koch] close to a commercial organic blueberry farm in Auburn (32.5325N, 85.4316W), Lee County, Alabama. Three egg masses of Chinavia hilaris (Say) (Hemiptera: Pentatomidae) were collected from a mimosa tree (Albizia julibrissin Durazz) in Irwin County, Georgia (31.3339N, 83.1926W). The egg masses were reared in a walk-in environmental chamber at 25 ± 2.0 °C, 50 ± 10% RH, and 12:12 L:D photoperiod for parasitoid emergence.

In 2017 and 2018, Project ‘Stink-be-Gone’, a citizen science project in collaboration with University of Maryland Extension and Master Gardeners, was established in central and western Maryland to monitor for stink bug egg masses, including H. halys. Sampling occurred for an average of one hour per week for six weeks (July and August 2017; late June to early August 2018). Participants recorded the time spent searching and collected GPS coordinates and general habitat characteristics (e.g., “private yard” or “park”) for all survey periods regardless of collection results. If egg masses were found, the host plant was identified to genus or species. Following collection, all egg masses and collection data were immediately sent to researchers at the University of Maryland, College Park for processing. Samples were placed in a growth chamber (Model 36LLVL, Percival Scientific, Perry, Iowa, USA) at 25 °C and 16:8 L:D. Egg masses were monitored daily for emergence of either bug nymphs or parasitoids. Parasitoids were transferred to 70% ethanol for later identification to species (all Trissolcus Ashmead and Telenomus Haliday and females of Anastatus Motschulsky) or genus (Anastatus males, see Burks 1967). After six weeks, unhatched eggs were dissected to ascertain their fate (e.g., unemerged bug nymph or parasitoid adult, partially developed parasitoid, infertile bug egg). Egg masses were identified to genus (unpublished key by Dieckhoff and Hoelmer). Master Gardeners collected egg masses of eight stink bug genera throughout both summers.

Other samples of Acroclisoides specimens used in the molecular analysis were collected in China and South Korea by Kim Hoelmer from 2014–2017 and by Lucian Fusu during field trips conducted in 2016 (see results, Table 4). Paratypes of A. solus were provided on loan from the National Museum of Natural History, Smithsonian Institution, Washington, D.C. The remaining sequences analyzed here were mined from GenBank.

The following keys and taxonomic works were used for the identification of Acroclisoides species: Grissell and Smith (2006), Sureshan and Narendran (2002), Xiao and Huang (2000), Huang and Liao (1988), Masi (1917), Girault and Dodd (1915).

The examined material listed below (all initially unidentified Pteromalidae, except paratypes of A. solus), including vouchers used in the molecular analyses, is deposited in the following institutions: CABI, Delémont (Switzerland); CREA, Florence (Italy); DAFNAE, University of Padua, Padova (Italy); EBCL, Montferrier le Lez (France); ERSA, Udine (Italy); FSCA, Florida State Collection of Arthropods, Florida (USA); Laimburg RC, Laimburg Research Centre, Vadena (Italy); MICO, Mitroiu Collection, Iași (Romania); NHMB, Natural History Museum Bern (Switzerland).

Italy: 19♀♀, 9♂♂, Cordenons, Friuli V. G., 46.0082N, 12.6713E, 8.viii.2018, Iris Bernardinelli, Giorgio Malossini & Luca Benvenuto leg., on Halyomorpha halys eggs on Robinia pseudoacacia (6♀♀, 3♂♂, CREA; 9♀♀, 6♂♂, ERSA; 1♀, FSCA; 3♀♀, MICO); 1♀, 8 unsexed, Riese Pio X, Veneto, 29.viii.2017, Paola Tirello and Davide Scaccini leg., on Halyomorpha halys eggs on Vitis vinifera (1♀, MICO; 8 unsexed DAFNAE); 3♀♀, 3 unsexed, Montebelluna, Veneto, 29.viii.2017, Paola Tirello and Davide Scaccini leg., on Halyomorpha halys eggs on Vitis vinifera (1♀, 3 unsexed, DAFNAE; 2♀♀, MICO); 4♀♀, 5 unsexed, Povegliano, Veneto, 16.viii.2017, Paola Tirello and Davide Scaccini leg., on Halyomorpha halys eggs on Actinidia sp. (1♀, 5 unsexed, DAFNAE; 3♀♀, MICO); 12♀♀, 2 unsexed, Ora, Trentino-Alto Adige/Südtirol, 46°21'43.3"N, 11°17'54.6"E, 27.ix.2018, Martina Falagiarda, on Halyomorpha halys eggs on Acer sp. (10♀♀, 2 unsexed, Laimburg RC; 2♀♀, MICO).

Switzerland: 5♀♀, 3♂♂, Zurich city, Canton Zurich, Lat. 47.351708 Long. 8.559493, 2.vii.2019, Tim Haye and Emily Grove leg., ex eggs of Halyomorpha halys on Tilia platyphyllos (ZP4), (1♀, MICO; 1♀, NHMB, 3♀♀, 3♂♂, CABI); 9♀♀, 6♂♂, Zurich city, Canton Zurich, Lat. 47.353213 Long. 8.553968, 10.vii.2019, Emily Lemke and Emily Grove leg., ex eggs of Halyomorpha halys on Liriodendron tulipifera (ZP6), (1♀, MICO; 3♀, NHMB; 5♀♀, 6♂♂, CABI); 12♀♀, 8♂♂, Zurich city, Canton Zurich, Lat. 47.353213 Long. 8.553968, 10.vii.2019, Emily Lemke and Emily Grove leg., ex eggs of Palomena prasina on Liriodendron tulipifera (ZP7), (1♀, 1♂, MICO; 1♀, 1♂, NHMB; 10♀♀, 6♂♂, CABI); 6♀♀, 6♂♂, Zurich city, Canton Zurich, Lat. 47.354905 Long. 8.535045, 10.vii.2019, Emily Lemke and Emily Grove leg., ex eggs of Palomena prasina on Catalpa bignonioides (ZP8), (1♀, 1♂, MICO; 1♀, 1♂, NHMB; 4♀♀, 4♂♂, CABI).

South Korea: 5♀♀ [GB] Gyeongsan-si, Daehak-ro, 280, Yeungnam Univ., 35°49'11.6"N, 128°45'53.6"E, 14.viii.2016, L. Fusu (MICO); 1♀ S. Korea: Chungbuk, Okcheon-gun, Bougimyeon, Soesan-li, 150 m, Malaise trap, 10.ix–03.x.2004, 36°16.594'N, 127°36.742'E, Tripotin rec. (MICO).

USA: 2♀♀ paratypes of A. solus, VA: Fairfax Co. near Annandale 38°50'N, 77°12'W Aug. 17–20 2004 Malaise trap David R. Smith; 2♀♀ Alabama, Lee Co., Auburn, ex BMSB eggs on pecan, R. Balusu and G. Tillman 25.vii.2018 (FSCA); 1♀ Georgia, Irwin Co., ex Chinavia hilaris eggs on mimosa tree, G. Tillman 24.vii.2017 (FSCA); 3♀♀ Georgia, Irwin Co., ex Chinavia hilaris eggs on mimosa tree, G. Tillman 21.vii.2017 (FSCA).

Abbreviations of morphological terms: F, funicular segment; GT, gastral tergite; MT, metasomal tergite; MV, marginal vein; OOL, ocello-ocular line; PV, postmarginal vein; POL, posterior ocellar line; SM, submarginal vein; SV, stigmal vein.

The DNA extraction, PCR amplification and sequencing of all the specimens listed in Table 4 were conducted in three different laboratories: the USDA-ARS – European Biological Control Laboratory (EBCL), the Florida Department of Agriculture and Consumer Service – Florida State Collection of Arthropods (FSCA), Division of Plant Industry (FDACS-DPI), and the Research Group in Invertebrate Diversity and Phylogenetics at the University of Iasi (UAIC). Genomic DNA was nondestructively isolated from the entire specimen using the Qiagen DNeasy kit (Hilden, Germany) at EBCL and FSCA as described in Sabbatini Peverieri et al. (2018) and at UAIC following the protocol developed by Cruaud et al. (2019). The barcode region of the mitochondrial Cytochrome Oxidase Subunit I (COI) was amplified using the universal barcoding primers LCO1490 and HCO2198 (Folmer et al. 1994). Amplification and sequence editing were done at EBCL and FSCA as described in Sabbatini Peverieri et al. (2018) and at UAIC as reported in Fusu and Ribes (2017).

Two A. solus samples required troubleshooting for successful COI barcoding. The samples USNMENT01335770 (MN018863.1) and FSCA00090246 (MN018864.1) both yielded Wolbachia COI sequences when using the universal barcoding primers (Folmer et al. 1994) and LEP-F1/LEP-R1 (Hebert et al. 2004) (see Smith et al. 2012 for an explanation of this phenomenon). These samples were alternatively amplified and sequenced with the primer pair C1-J-1632/C1-N-2191 (Kambhampati and Smith 1995; Simon et al. 1994). The thermocycling conditions for C1-J-1632/C1-N-2191 were: 1) initial denaturing at 95 °C for 2 minutes, 2) 98 °C for 20 seconds, 3) 40 °C for 30 seconds, 4) 72 °C for 30 seconds [steps 2–4 repeated for 30 cycles], and 5) final extension at 72 °C for 7 minutes. Slightly shorter (432 bp and 484 bp) COI barcode sequences were generated from these samples using this primer pair.

All sequences generated from this study are deposited in GenBank and all residual DNAs are archived at the place of the DNA extraction (Table 4). Voucher specimens which have been reexamined following the molecular analysis are presently deposited in public collections or in laboratory collections (Table 4). All barcode sequences were translated into amino acids to check for stop codons. The 45 sequences obtained were compared with sequences present in GenBank using the Basic Local Alignment Search Tool (http://www.ncbi.nlm.nih.gov/BLASTn) and aligned with the four barcode sequences of A. solus previously available. The alignment was performed using Clustal Omega (Sievers and Higgins 2014), implemented in Seaview version 4 (Gouy et al. 2010). Polymorphism information, haplotype diversity (Hd), nucleotide diversity (π) and number of haplotypes were determined using DnaSP version 6 (Rozas et al. 2017). The phylogenetic relationships among haplotypes were depicted using statistical parsimony in TCS (Clement et al. 2000) implemented in PopART (Leigh and Bryant 2015) which allows visualization of the frequency and geographical distribution of haplotypes.

The molecular distances between sequences of all individuals from different countries were calculated by the standard Kimura 2-parameter (K2P) measure (Kimura 1980) using Mega 6 (Tamura et al. 2013). The range of COI K2P distances found between A. sinicus haplotypes was compared to distances in the subfamily Pteromalinae as a whole using i) intraspecific distances, and ii) congeneric distances. The Pteromalinae DNA sequences used in this study were downloaded from GenBank and searched locally using software from the Blast+ toolkit. To create the Pteromalinae database used in this study, we targeted only genera represented by at least two identified species, and each species was represented by at least two sequences. The qualifying sequences were then extracted and aligned with the barcode sequence MK188331.1 of A. solus. The aligned Pteromalinae sequences were checked by eye and the edges trimmed using Seaview version 4 and imported into Mega 6 to calculate the molecular distances. The molecular distances were then split into intraspecific observations and interspecific (congeneric) observations.

Our taxonomic studies determined that the Acroclisoides specimens found in Cordenons in 2018 belong to the species A. sinicus. From the three collected H. halys egg masses a total of 28 specimens of A. sinicus emerged (Table 2). Specimens of Anastatus bifasciatus (Geoffroy) (Hymenoptera: Eupelmidae), a well-known European native parasitoid that can parasitize eggs of H. halys (Haye et al. 2015; Roversi et al. 2016), also emerged. Detailed observation of the parasitized eggs revealed consistent differences in the exit holes produced by A. sinicus and An. bifasciatus, enabling eggs to be associated with the species that parasitized them.

Males of An. bifasciatus are visibly smaller than females (Bolivar y Pieltain 1935), and the circular exit holes that they produce are smaller than those of the larger females (Fig. 1A, B). Acroclisoides sinicus, independent of sex, produced exit holes on H. halys eggs of similar size to those of females of An. bifasciatus, but with much more irregular margins (Fig. 1C, D). Trissolcus mitsukurii (associated with A. sinicus in the Veneto region), produced round exit holes similar in size to females of An. bifasciatus, but with less serrate margins and a conspicuous dark ring on the upper margin of the parasitized eggs, just below the exit hole and the operculum (Fig. 1E). These observations permitted us to assess the number and species of parasitoids that emerged prior to the field collection of the egg masses (see Table 2). Emergence holes produced by T. japonicus (associated with A. sinicus in Zurich city, see results of Switzerland) are similar to those produced by T. mitsukurii, except there is no dark ring below the operculum (Fig. 1F).

Figure 1. 

Typical shapes of exit holes produced by egg parasitoids of Halyomorpha halys in Europe: Anastatus bifasciatus male (A) and female (B) (NE-Italy); Acroclisoides sinicus (C, D) (NE-Italy); Trissolcus mitsukurii (E) (NE-Italy) and Trissolcus japonicus (F) (Switzerland).

Acroclisoides specimens found at the three localities of Povegliano, Montebelluna and Riese Pio X belong to the species A. sinicus as confirmed by our taxonomic study. In 2017, 24 A. sinicus specimens emerged from three collected H. halys egg masses. Egg masses parasitized by A. sinicus were collected on August 16 (one egg mass from Povegliano site on Actinidia sp.) and August 29 (two egg masses, one from Montebelluna and one from Riese Pio X both on Vitis vinifera L.). Parasitism of single egg masses by different parasitoid species was also observed: from the egg mass containing six A. sinicus individuals in Montebelluna, 15 individuals of T. mitsukurii also emerged; from the egg mass in Povegliano from which nine A. sinicus individuals emerged, 10 An. bifasciatus also emerged; but from the H. halys egg mass collected in Riese Pio X, only A. sinicus emerged (nine individuals). No Acroclisoides individuals were found in the monitoring campaign of 2018.

From the H. halys egg masses collected in Ora, only one egg mass (collected on September 27, 2018) containing 14 eggs produced 14 specimens of A. sinicus. No other parasitoid species or H. halys nymphs emerged from the egg mass.

Acroclisoides individuals were reared from two H. halys egg masses and two egg masses of the native stink bug, P. prasina (Table 2). The first A. sinicus emerged from a single H. halys egg mass collected from Tilia platyphyllos on July 2, 2019. From the same egg mass a single female of T. japonicus was reared (Fig. 1F). On July 10, 2019 single H. halys and P. prasina egg masses were found on the same L. tulipifera tree, both generating Acroclisoides individuals. From an additional P. prasina egg mass collected on the same date on C. bignonioides, 12 A. sinicus emerged along with 12 T. japonicus. No Acroclisoides individuals emerged from egg masses collected June 28, 2019.

Of the three H. halys egg masses collected in Auburn, AL, during the summer of 2018, one egg mass collected from pecan on July 25 produced six specimens of A. sinicus (Table 3). Nine Trissolcus euschisti (Ashmead) (Hymenoptera: Scelionidae) and seven Anastatus reduvii (Howard) (Hymenoptera: Eupelmidae), both of which are common indigenous parasitoids that sometimes parasitize H. halys in the USA (Cornelius et al. 2016), also emerged from this egg mass. One scelionid pupa died; four eupelmid late instars died. Of the nine C. hilaris egg masses collected in Irwin County, GA, during the summer of 2017, three egg masses collected from mimosa produced A. sinicus. Fifteen A. sinicus emerged from one C. hilaris egg mass collected from mimosa on July 21, 2107, and from this egg mass emerged 49 Trissolcus edessae Fouts (Hymenoptera: Scelionidae) and two An. reduvii. The second C. hilaris egg mass collected from this plant on the same date produced 31 A. sinicus and no other parasitoids. From the third egg mass of C. hilaris, collected on mimosa on July 24, 2017, one A. sinicus and 53 T. edessae emerged.

Acroclisoides sinicus emerged from two egg masses of Euschistus sp. (Hemiptera: Pentatomidae) collected on August 15 and 27, 2017 in Montgomery (39.2063N, 77.2079W) and Frederick (39.4760N, 77.2703W) Counties, respectively (Table 3). In Montgomery County, five A. sinicus emerged from the egg mass collected on Cornus sp. in a private home site with a wooded landscape of mixed deciduous trees, shrubs and herbaceous plants. In Frederick County, 14 A. sinicus emerged from two egg masses collected from Cercis canadensis L. located in the Fountain Rock Park and Nature Center (landscape has a mix of deciduous trees and shrubs and herbaceous plants). Six A. sinicus emerged from a Brochymena sp. (Hemiptera: Pentatomidae) egg mass collected in Allegany County (39.6786N, 78.7372W) (July 7, 2018) on Lonicera japonica Thunb. in the demonstration garden of the University of Maryland Extension (landscape with fruiting trees, managed and unmanaged trees, flowering herbaceous plants and shrubs, vegetable plants and mown turf). Fifteen of the emerged specimens were females, two were males, and the sex of the remainder could not be determined. No other parasitoids emerged during rearing and no stink bug nymphs emerged from any of the three egg masses.

The 5’-COI barcode fragment was successfully obtained from 45 individuals collected in numerous countries and from different hosts, plus four sequences from GenBank (Table 4). The lengths of these sequences varied from 432 bp for the paratype of A. solus to 652bp for three Italian, all Swiss, two American and three South Korean specimens. After edge trimming, the final data matrix of 49 sequences including the four GenBank accessions consisted of 429 characters with seven parsimony informative sites. As is typical for insect mitochondrial genes, the AT content in Acroclisoides was high (74.8%), similar to the AT content reported in Cynipidae (74.8%) at the higher end of the range observed for parasitic wasps (Rokas et al. 2002). A total of 7 haplotypes were recorded among the analyzed individuals. In Italy, we detected one unique haplotype (H1) which is shared with Switzerland. In Switzerland where H1 predominates (8 out of 9 specimens), we also found a second haplotype (H2) in one specimen. Although H1 and H2 did not match any of the 5 haplotypes found outside Europe, the network analysis showed that they are closely related as H1 differed only by one substitution from the haplotype H6 found in China and by two substitutions from the haplotype H3 found in South Korea and North America (Fig. 4). The two haplotypes found in North America, H3, which includes the A. solus paratype and H4, were shared with South Korea. Three haplotypes (H5, H6 and H7) were identified in China, H6 being only one mutational step away from the other two. The degree of divergence was found to be very low at the sampling level, with a mean K2P distance of 0.64%, ranging from 1.41% between the two haplotypes H3 and H4 both detected in North America and South Korea to 0.23% between H1 from Europe and H6 found in China or H2 in Switzerland. Haplotype diversity (Hd) was 0.118 (SD of 0.101) and 0.343 (SD of 0.128) in Europe and North America respectively. Nucleotide diversity (π) was 0.00027 (SD of 0.00024) and 0.0048 (SD of 0.00179) in Europe and North America respectively. In Asia, Hd was 0.824 (SD of 0.047), and π was 0.00662 (SD of 0.0007). In order to confirm that the range of divergence is what is expected at the intraspecific level and given that there are no data available at the genus level, we compared these data with other species in the subfamily. A total of 432 Pteromalinae barcodes representating of a minimum of two species in four genera (Lyrcus Walker, Mesopolobus Westwood, Pachyneuron Walker and Pteromalus Swederus) were downloaded and aligned. From this 283bp dataset (available upon request to authors), K2P distances for two classes (intraspecific and interspecific) were calculated. Figure 5 plots K2P values for the Pteromalinae, along with the divergence observed between Acroclisoides haplotypes, clearly confirming that divergence between Acroclisoides haplotypes is not significantly different from the Pteromalinae intraspecific divergence (Kruskal-Wallis test, p=0.294), but significantly different from the Pteromalinae congeneric divergence (Kruskal-Wallis test, p<0.0001). Species boundaries between A. solus and A. sinicus are not supported by the present mitochondrial data.

Figure 4. 

Haplotype network of the 49 Acroclisoides sinicus barcodes analyzed in this study. Each circle corresponds to one haplotype; circle size gives the proportion of individuals belonging to the haplotype. The color of the circles represents the geographical origin. Numbers correspond to the haplotype numbers. Hash marks symbolize the number of mutations between haplotypes.

Figure 5. 

Boxplot of pair-wise molecular distances between (left) the Acroclisoides haplotypes evidenced in this study, (center) conspecific individuals in the Pteromalinae, (right) individuals from different species of the same genus in the Pteromalinae.