An Insight into the Role of Trissolcus mitsukurii as Biological Control Agent of Halyomorpha halys in Northeastern Italy

Sustainable strategies such as classical or augmentative biological control are currently being evaluated for the long-term management of the alien invasive pest Halyomorpha halys (Stål) (Hemiptera: Pentatomidae). A three-year study carried out in northeastern Italy was performed to investigate the distribution and field performance of the H. halys egg parasitoid Trissolcus mitsukurii (Ashmead) (Hymenoptera: Scelionidae), in comparison with other parasitoid species. In the study area, adventive populations of T. mitsukurii were present since 2016, representing the earliest detection of this species in Europe. Trissolcus mitsukurii was the most abundant parasitoid and showed a higher “parasitoid impact” (i.e., number of parasitized eggs over the total number of field-collected eggs) compared to the other species, i.e., Anastatus bifasciatus (Geoffroy) (Hymenoptera: Eupelmidae), Trissolcus basalis (Wollaston) and Trissolcus kozlovi Rjachovskij (Hymenoptera: Scelionidae). The hyperparasitoid Acroclisoides sinicus (Huang and Liao) (Hymenoptera: Pteromalidae) was also recorded. Phylogenetic analysis of T. mitsukurii population distinguished two clades, one covering samples from Italy, Japan and China, the other from South Korea. The present study provides promising results for the biological control of a pest that is having a dramatic impact on a wide range of crops worldwide.


Introduction
The brown marmorated stink bug, Halyomorpha halys (Stål) (Hemiptera: Pentatomidae), is a polyphagous pest native to eastern Asia, detected in North America in the 1990s [1], in Europe in the mid-2000s [2], and in South America in 2017 [3]. Feeding on more than 170 host plants, H. halys causes serious damage to agricultural crops, ornamental plants, and urban and forest trees [4][5][6][7]. Moreover, H. halys can represent a nuisance problem in residential areas during fall and winter, when large numbers of adults invade buildings searching for overwintering sites [1,8,9]. In Italy, H. halys was reported in the Emilia Romagna region in 2012, in the Piedmont region in 2013, in the Veneto region and in the Friuli-Venezia Giulia region in 2014, and two years later in the Trentino-Alto Adige region [10][11][12][13][14]. The species is now distributed all over the country [13,15].
How and when these exotic parasitoids were introduced to Europe is still unclear, and this represents an essential aspect in tracking both H. halys and its parasitoid spread [23]. Genetic analyses of European H. halys populations revealed multiple invasions from different sources in their native areas [27][28][29], which could potentially be involved in the distribution pathways of egg parasitoids. To test this hypothesis, we analyzed material on H. halys egg parasitoid complexes collected in the northeast of Italy, in areas not covered by previous surveys and colonized by a diverse set of H. halys haplotypes [28,29]. Assessments of the parasitoid complexes of H. halys eggs in northeastern Italy started in 2016, and we compared the impact of the different parasitoid species across three years. Data revealed that T. mitsukurii was the dominant species among egg parasitoids of H. halys, and we characterized the genetic structure of its population.

Sampling Sites and Laboratory Study
Eight sites in the Veneto region were surveyed every 7-10 days, from June 2016 to October 2018 ( Table 1). The selection of sites was based on H. halys population levels. Additional observations were performed in other sites located in northeastern Italy: from the beginning of September to October 2018 in the Trentino-Alto Adige region, and in August 2018 in two sites in the Friuli-Venezia Giulia region (Table 1). Field surveys were conducted by searching for H. halys eggs on the vegetation, and sampling the canopy between 0.5 to 2 m from the ground, inspecting in particular fruits and leaves; egg masses were hand-collected on the leaves or fruits and transferred to the laboratory. Field-collected egg masses were reared in a climatic chamber at 26 ± 1 • C, 65 ± 5% RH, and 16:8 L:D until parasitoid adult emergence. Emerged parasitoids were stored in 70% ethanol for morphological and molecular identification.
The "exploitation efficiency" (number of parasitized eggs divided by the total number of eggs in the parasitized egg masses [30]) was calculated for each site. On data collected from sites showing a high presence of H. halys parasitoids and a steady sampling (i.e., Sites 3, 5, 8 and 9; Table 1), the "discovery efficiency" (number of egg masses from which at least one parasitoid emerged on the total number of sampled egg masses) and the "parasitoid impact" (number of parasitized eggs divided by the total number of field-collected eggs) were calculated [30]. Indexes obtained from data at the latter sites were used to compare performances of the parasitoid species emerged from H. halys eggs. These indexes were analyzed using a generalized linear model, assuming a binomially distributed model with a logit link function with the GENMOD procedure of SAS ® ver. 9.4 [31]. The effect of parasitoid species was tested using a Wald χ 2 test (α = 0.05) followed by a Tukey's test (α = 0.05) on the least-square means. Only species observed in more than one site were included in the analyses. In the case of the Auer (Ora) site (Site 3), eggs from which no H. halys nymphs or parasitoids emerged were further assessed for other causes of mortality [32] under a Leica stereomicroscope (series MZ6) according to Morrison et al. [33], and then ascribed to predation or to other causes of mortality (e.g., deformed and discolored eggs). For eggs where H. halys hatching or parasitoid emergence took place in the field before sampling, the category was assessed according to Jones et al. [34], and eggs were classified as parasitized or hatched eggs.

Insect Identification and Molecular Analysis
Ethanol-stored specimens were dried and glued on card points for morphological analyses. A Wild M3 stereomicroscope with magnification up to 200× and a 2700 k spotlight was used for morphological diagnosis. The Palaearctic genera of Scelionidae and Trissolcus species were respectively determined using keys from Kozlov and Kononova [35] and Talamas et al. [36]. Moreover, Trissolcus specimens were compared with pictures of holotypes and paratypes in Hymenoptera Online (HOL), provided by Talamas et al. [36]. Anastatus species were identified using the key by Askew and Nieves-Aldrey [37]. Acroclisoides (Hymenoptera: Pteromalidae) specimens were kindly identified by Dr. Mircea-Dan Mitroiu [38]. The parasitoids used for morphological identification were deposited in the Department of Agronomy, Food, Natural Resources, Animals and Environment (DAFNAE), Legnaro, Padua, Italy, in Dipartimento di Scienze Agrarie, Forestali e Alimentari (DISAFA), Grugliasco, Turin, Italy, and in Laimburg Research Center, Auer, Bolzano, Italy.
To confirm the results of the morphological identification for parasitoid species and to study phylogenetic relationship for the most common non-native parasitoid species, molecular analyses were conducted. Total DNAs from three Trissolcus adults, singularly analyzed, collected in Trentino-Alto Adige were extracted by homogenizing each sample in 400 µL of CTAB buffer (CTAB 2.5%, Tris pH 8 100 mM, NaCl 1.4 M, EDTA 50 mM pH 8, PVP-40 1%, Proteinase K 10 mg/mL) in a microcentrifuge tube containing a 5-mm tungsten carbide bead (Qiagen, Valencia, CA, USA). Samples were disrupted using a Retsch Mixer Mill MM 400, at 30 Hz for 3 min. After disruption, DNA was extracted using the DNeasy Blood and Tissue kit (Qiagen) following the instructions of the provider. Moreover, DNA of 14 parasitoid samples from Veneto and of three samples from Friuli-Venezia Giulia, singularly analyzed, was extracted according to a previously described salting-out protocol [39]. A partial region of the cytochrome c oxidase subunit I (COI) gene was amplified with primers HCO2198 (5 -GGTCAACAAATCATAAAGATATTGG-3 ) and LCO1490 The produced amplicons were purified through a QIAquick PCR and Gel Cleanup Kit following the instructions of the provider (Qiagen). The purified amplicons were sequenced by LGC Genomics GmbH (Berlin, Germany). PCR products from Veneto and Friuli-Venezia Giulia samples were purified using exonuclease and Antarctic Phosphatase (GE Healthcare, Chicago, IL, USA) and sequenced at the BMR Genomics Service (Padua, Italy).
Sequences were edited and aligned using MEGA X software [41]. To assess the identity of the sequences obtained, the integrated bioinformatics platform Barcode of Life Data System database [42] was used, and a nucleotide Blast was performed using the Basic Local Alignment Search Tool (National Center for Biotechnology Information-NCBI). Similar sequences retrieved from both databases were added to our dataset. Additionally, sequences of the same target gene obtained in previous studies [25,26] were added to our dataset.
To exclude the presence of stop codons in the coding sequences, all COI sequences were translated with Transeq (EMBOSS [43]).
Phylogenetic relationships among sequences of non-native species were estimated with an approximate maximum-likelihood (ML) analysis, using MEGA X software. A GTR + I + G model was applied, with neighbor-joining starting tree and 100 bootstrap replications. A haplotype parsimony network of the final dataset with a probability cut-off at 90% was reconstructed following the TCS method in PopART [44].

Results
In this study, 251 egg masses and 6527 eggs were collected ( Table 2). Egg masses were mostly found on leaves and fruits of crop plants in Veneto, on maple leaves and fruits (93.7%), ailanthus and linden leaves in Trentino-Alto Adige, and on fig and soybean leaves in Friuli-Venezia Giulia (Table 1). In all the three years, H. halys egg masses were found in particular from mid-June to the end of September.
Five species of hymenopteran parasitoids emerged from 46 H. halys egg masses, and in five cases the egg mass was parasitized by more than one parasitoid species (Table 2). There were no sites with all five parasitoid species together, and, in many cases, only one species was found ("Relative Abundance per Site" in Table 2). These parasitoids emerged mainly between July and September, and some individuals also in June and October.
Based on morphological analyses, Trissolcus individuals were identified as Trissolcus basalis (Wollaston), T. kozlovi Rjachovskij and T. mitsukurii. The latter was the most abundant and it emerged from collected egg masses from June (2018) or August (2016) to October, and from May to August in 2017. Both T. basalis and T. kozlovi were less frequently found parasitizing H. halys eggs ( Table 2) Table 2). Other Hymenoptera emerged from eggs were morphologically identified as A. bifasciatus and the hyperparasitoid Acroclisoides sinicus (Huang and Liao) (senior synonym of A. solus Grissell and Smith [38]). The first species was found in six sites across Veneto and Trentino-Alto Adige regions, while A. sinicus individuals emerged from H. halys egg masses in three sites, on 29 August 2017 (nine individuals: Site 8; six individuals: Site 9) and on 27 September 2018 (14 individuals: Site 3; Table 2). These three egg masses were probably primary parasitized by T. mitsukurii, as indicated by the presence of round exit holes and the dark ring on the upper border [38].
Parasitoid performances were compared among A. bifasciatus, T. basalis and T. mitsukurii during the three years of observation. No differences were recorded in parasitoid discovery efficiency in the three years (Table 3). The exploitation efficiency among species did not differ in 2016, while, in 2017 and 2018, T. mitsukurii showed a higher exploitation efficiency than A. bifasciatus (Table 3). Exploitation efficiency varied from 32.3% to 100% for T. mitsukurii, and from 7.4% to 46.4% for A. bifasciatus, while for the other species it was always below 18.0% (Table 2). Parasitoid impact varied among species in all years, with T. mitsukurii showing always the highest value (Table 3). In Site 3, 327 eggs (out of 1520) were found with signs of sucking or chewing by predators. Deformed or discolored eggs were also observed; 939 H. halys eggs (61.8%) hatched (Supplementary Table S1). a As number of parasitoids for each species over the number of all parasitoids found in the site; b As number of parasitized eggs by a species over the total number of eggs of the parasitized egg mass; c In this site, some parasitoids emerged from egg masses were not identified (emerged before the collection of the egg mass). Trissolcus mitsukurii was recorded in seven sites in Veneto, and in one site in Trentino-Alto Adige and in Friuli-Venezia Giulia. In the Veneto region, the first three records of this parasitoid occurred in 2016 (Site 8) and 2017 (Sites 9 and 11; Figure 1). The presence of T. mitsukurii emerging from H. halys eggs was recorded in these three sites also in the following year(s). To date, the distribution of T. mitsukurii in its non-native area covers many cultivated and natural areas in northern Italy, across at least four regions (Figure 1). and Sabbatini Peverieri et al. [26] are included.

Molecular Analysis
The results of the molecular analysis were consistent with the morphological identification of T. basalis, T. kozlovi and T. mitsukurii. A total of 20 Trissolcus specimens-14 from Veneto, 3 from Trentino-Alto Adige and 3 from Friuli-Venezia Giulia-were successfully amplified and sequenced. A fragment of 534 bp of COI was obtained for all sequences. A BLAST search of sequences from 18 adult insects showed significant alignment with the deposited T. mitsukurii sequence AB971831.1, giving an E-value of 0.0, 100% query coverage and 99.32% sequence identity. Similarly, BOLD Systems databases showed a >99% similarity with T. mitsukurii. The COI sequence of the T. kozlovi specimen in this study had 100% sequence identity to those of Moraglio et al. [25], while T. mitsukurii specimens have a perfect match with data in Sabbatini Peverieri et al. [26]. Sequences were deposited in the NCBI database with the GenBank accession numbers MT345599-MT345602.
The phylogenetic tree, inferred with sequences of this study and sequences retrieved from GenBank database, distinguished between two highly supported clades. One of them grouped the T. mitsukurii samples from Italy, Japan and China, while the other clade contained specimens from South Korea (Figures 2 and 3). The TCS Network showed the presence of five haplotypes (Figure 2). The haplotype H5 grouped all sequences of this study and five sequences previously obtained from Italy (MK097188, MK097189 and MK097190 in Sabbatini Peverieri et al. [26], and two sequences from Lombardy-kindly provided by S.T. Moraglio). The other four haplotypes corresponded to sequences from Asian samples. Two of them, H1 and H2, were separated from H5 by four and five mutational steps, respectively, and included samples from Japan and the Yunnan province of China [26,45]. The other two haplotypes, H4 and H3, differed from H5 by several mutational steps (12 and 13 respectively) and grouped sequences from South Korea [26].  [25] and Sabbatini Peverieri et al. [26] are included.

Molecular Analysis
The results of the molecular analysis were consistent with the morphological identification of T. basalis, T. kozlovi and T. mitsukurii. A total of 20 Trissolcus specimens-14 from Veneto, 3 from Trentino-Alto Adige and 3 from Friuli-Venezia Giulia-were successfully amplified and sequenced. A fragment of 534 bp of COI was obtained for all sequences. A BLAST search of sequences from 18 adult insects showed significant alignment with the deposited T. mitsukurii sequence AB971831.1, giving an E-value of 0.0, 100% query coverage and 99.32% sequence identity. Similarly, BOLD Systems databases showed a >99% similarity with T. mitsukurii. The COI sequence of the T. kozlovi specimen in this study had 100% sequence identity to those of Moraglio et al. [25], while T. mitsukurii specimens have a perfect match with data in Sabbatini Peverieri et al. [26]. Sequences were deposited in the NCBI database with the GenBank accession numbers MT345599-MT345602.
The phylogenetic tree, inferred with sequences of this study and sequences retrieved from GenBank database, distinguished between two highly supported clades. One of them grouped the T. mitsukurii samples from Italy, Japan and China, while the other clade contained specimens from South Korea (Figures 2 and 3). The TCS Network showed the presence of five haplotypes (Figure 2). The haplotype H5 grouped all sequences of this study and five sequences previously obtained from Italy (MK097188, MK097189 and MK097190 in Sabbatini Peverieri et al. [26], and two sequences from Lombardy-kindly provided by S.T. Moraglio). The other four haplotypes corresponded to sequences from Asian samples. Two of them, H1 and H2, were separated from H5 by four and five mutational steps, respectively, and included samples from Japan and the Yunnan province of China [26,45]. The other two haplotypes, H4 and H3, differed from H5 by several mutational steps (12 and 13 respectively) and grouped sequences from South Korea [26].

Discussion
In many areas of the invaded range, H. halys is causing severe damage to crops. Current pest management strategies are mostly based on chemical control and have often failed in controlling this pest. The implementation of biological control needs to be considered for the management of this invasive pest [5] because the lack of effective natural enemies is one of the main explanations for the high impact of exotic species in invaded territories [46]. In this three-year study, five different parasitic wasp species emerged from H. halys eggs: A. sinicus, A. bifasciatus, T. basalis, T. kozlovi and T. mitsukurii, and the last was the most common species found.
Trissolcus mitsukurii is an Asian species reported to develop in H. halys eggs [20,36]. In this work, for the first time, the performances of T. mitsukurii across multiple years were studied outside its native area. It should be noted that information on the performances of this egg parasitoid on H. halys in its native range are also limited. Trissolcus mitsukurii turned out to be the dominant species among H. halys parasitoids, and it showed the highest performances (i.e., exploitation efficiency and parasitoid impact) compared to the other egg parasitoids. In particular, the parasitoid impact of T. mitsukurii ranged from 7.7% to 15.1%, while this parameter for the native A. bifasciatus was always lower than 4%. Bearing in mind previously published data on other H. halys egg parasitoids, the data observed here for T. mitsukurii confirm that this exotic parasitoid can have similar or even higher performances as compare to the best native ones, i.e., A. bifasciatus, which displayed yearly average parasitism levels from 0.5% to 13.4% [25,47,48]. The performances of T. mitsukurii are comparable to those of T. japonicus, the other exotic H. halys egg parasitoid observed in Europe, which showed a parasitoid impact from 10.5% to 21.5%, measured in one site in northwestern Italy [25]. However, in another study, the parasitoid impact recorded for T. japonicus was no more than 2% in three sites [24].
The distribution area of T. mitsukurii is wider than previously defined and includes all regions of northern Italy (except Aosta Valley), covering the area across the Friuli-Venezia Giulia, Veneto, Trentino-Alto Adige, Lombardy and Piedmont regions [25,26,49]. All samples of T. mitsukurii found in Veneto share the same haplotype of the samples from Trentino-Alto Adige, Lombardy, and Friuli-Venezia Giulia, suggesting a single introduction event of the parasitoid, which possibly followed the invasion pathways of its invasive host. However, multiple introductions of individuals with the same COI haplotype cannot be excluded. The low genetic diversity in T. mitsukurii European populations found here and in previous studies [26] may suggest that this species has spread from a limited area in the Veneto region in 2016 to different sites across the whole northern part of Italy, as recorded two years later [25,26]. This information is of importance for tracking the distribution of the parasitoid outside its native range and for a better understating of the evolution of host-parasitoid population dynamics. Samples of T. mitsukurii retrieved in Italy are genetically close to the ones from Japan and China, suggesting that this may be their origin, even though some missing haplotypes separate them. Further analyses using other molecular markers and expanding collection sites in their native area will help to improve knowledge of the genetic variability of this parasitoid and delineate its patterns of invasion in Europe. This kind of study has already been performed with H. halys [27][28][29][50][51][52], showing high genetic variability in invaded areas and suggesting multiple introduction events.
The occurrence of adventive populations of T. mitsukurii in Europe was set one year before the other Asian parasitoids T. japonicus [24]. The latter was found emerging from H. halys eggs both in Europe and North America [23][24][25][26]53], where this species seems to coexist with A. bifasciatus, possibly acting in synergy in the control of H. halys [54]. Trissolcus mitsukurii and T. japonicus are considered key parasitoids of H. halys in its native area [18,[20][21][22], and their occurrence where H. halys is now considered a major pest offers interesting perspectives for its control. However, little is known about T. mitsukurii, and the present study provides information on this species in Europe, adding data on the parasitoid impacts in field conditions. Trissolcus mitsukurii represents an example of incidental introduction of an exotic parasitoid of H. halys, since no classical biological control program has been performed in Europe. This parasitoid emerged from H. halys eggs collected in orchards (organic and conventional), vineyards, and ornamental plants in urban areas, and its potential to track H. halys in different habitats is advantageous because H. halys is highly polyphagous and is present in these environments [5].
Among the other species that emerged from H. halys eggs, A. bifasciatus was the second-most frequently found. This is not surprising, since it is a common egg parasitoid of H. halys in Europe [25,47,48,64]. Anastatus bifasciatus exhibits a wide host range that includes 30 known host species belonging to the orders Hemiptera and Lepidoptera [65,66], and its use has been proposed for augmentative biocontrol against H. halys [66][67][68]. However, parasitism rates of A. bifasciatus in field conditions are generally low [25,47,48,64,68], and data here reported show a lower impact of this parasitoid compared to T. mitsukurii.
In this study, other Trissolcus species emerged from H. halys eggs. In particular, T. kozlovi was obtained from a H. halys egg mass in summer 2017. This species is morphologically and genetically similar to T. japonicus [25,36], and was first recorded in Italy emerging from eggs of H. halys and other stink bugs in 2016 [25]. The acceptance of H. halys eggs is here confirmed, but its occurrence in the field was limited, and it showed a very low exploitation efficiency. Furthermore, in two sites, T. basalis emerged from H. halys egg masses. Despite the low exploitation efficiency of the species, the ability of T. basalis to develop on H. halys eggs has to be taken into account because of its wide distribution in Europe and USA [36,69], and it has been reported to parasitize live H. halys eggs in the southeastern USA [70].
The hyperparasitoid A. sinicus was found to emerge from H. halys eggs collected in northeastern Italy, associated with other parasitoid species [25,38]. Pteromalidae belonging to the genus Acroclisoides were recorded in different studies acting as hyperparasitoids of Scelionidae [71][72][73]. For instance, Gariepy et al. [71] recorded A. sinicus from an egg mass of the pentatomid Chinavia hilaris (Say) (= Acrosternum hilare), which was found to be primarily parasitized by Trissolcus sp. Similarly, Acroclisoides sp. emerged from egg masses of N. viridula, primarily parasitized by T. basalis in Australia [71]. Acroclisoides sinicus may also act as hyperparasitoid of Anastatus spp. [38]. Little is known about the biology of A. sinicus and its influence on the primary parasitoids. Still, its occurrence should be considered in the definition of long-term biocontrol strategies, as showing potential impacts on the primary parasitoid species.
The overall parasitoid impact ranged from 11.2% in 2018 to 19.9% in 2016. However, it should be noted that, in this assessment, we did not take into account damaged and unhatched eggs. This can be the result of predatory activity by natural enemies [32,34,74], but also may derive from the parasitoid-induced host egg abortion or be the result of non-emerged parasitoids [74][75][76][77]. For example, Konopka et al. [77] showed that native parasitoids are not always able to develop inside the eggs of a new host species. In their study, Trissolcus euschisti (Ashmead), an indigenous parasitoid of Podisus maculiventris (Say) in the US, readily attacks H. halys eggs, but it is unable to complete its development. Since we did not assess the level of predation and egg abortion, we can expect that the overall impact of H. halys natural enemies is higher than what we estimated in this study. Further research should elucidate these aspects to obtain a full picture of the extent of biological control of H. halys by the parasitoid complex considered here. Additionally, the present study was designed to investigate the parasitoid complex of H. halys on crops and habitat characterized by a high presence of the pest, but the parasitoid abundance and its performances can vary during the season and differ among habitats and crops; thus, future studies should investigate the effects of these factors.

Conclusions
Ecological interactions between H. halys and its parasitoids in the field need to be studied for the selection of effective species as biocontrol agents. The present study, performed on the egg parasitoid complex of H. halys in northern Italy, reported high parasitoid impacts by T. mitsukurii, and provides promising results for the control of a pest that is having a dramatic impact on a wide range of crops. Data here presented, coupled with those of previous research, provide information on the expanding distribution range of T. mitsukurii. However, its distribution appears spottier than evenly distributed, since there are sites in the Veneto region where only a limited number of native parasitoids are recorded. Biological control with egg parasitoids is likely to have an impact on H. halys populations, but further studies are needed to assess the host range of T. mitsukurii in Europe and which factors influence the distribution of this exotic species.