Introduction
The brown marmorated stink bug Halyomorpha halys (Stål, 1855) is an invasive insect pest introduced from Asia into several countries in the world (Cianferoni et al., Reference Cianferoni, Graziani, Dioli and Ceccolini2018). Outside its native range it was first described in Pennsylvania, USA in 1996. It subsequently spread across the continent, invading most states in the USA before it was intercepted in Canada in 2010 (Gariepy et al., Reference Gariepy, Haye, Fraser and Zhang2014; Xu et al., Reference Xu, Fonseca, Hamilton, Hoelmer and Nielsen2014). In Europe, H. halys was introduced initially into Switzerland and Liechtenstein in 2004 (Arnold, Reference Arnold2009; Haye et al., Reference Haye, Abdallah, Gariepy and Wyniger2014) and was subsequently recorded in neighboring countries including Germany, France, Italy and Greece in the following years (Heckmann, Reference Heckmann2012; Callot and Brua, Reference Callot and Brua2013; Maistrello et al., Reference Maistrello, Dioli, Vaccari and Caruso2014; Milonas and Partsinevelos, Reference Milonas and Partsinevelos2014). H. halys has invaded Russia, Abkhazia, Georgia (Gapon, Reference Gapon2016) and Chile (Faundez and Rider, Reference Faundez and Rider2017), and was intercepted multiple times in New Zealand and Australia but no populations have established in these latter countries (Gariepy et al., Reference Gariepy, Haye, Fraser and Zhang2014; EPPO 2019).
Detailed knowledge of invasion routes is of considerable interest to prevent and control introductions and establishments of novel invasive species (Estoup and Guillemaud, Reference Estoup and Guillemaud2010). In particular, the use of molecular and population genetic tools has shown to provide substantial information about the reconstruction of invasion routes. Several studies have described the invasion history and genetic diversity of H. halys using molecular tools with a particular focus on various mitochondrial regions (Cesari et al., Reference Cesari, Maistrello, Ganzerli, Dioli, Rebecchi and Guidetti2014, Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018; Gariepy et al., Reference Gariepy, Haye, Fraser and Zhang2014; Xu et al., Reference Xu, Fonseca, Hamilton, Hoelmer and Nielsen2014; Zhu et al., Reference Zhu, Ye, Du, Zhang, Zhen, Zheng, Zhao, Li and Bu2016; Valentin et al., Reference Valentin, Nielsen, Wiman, Lee and Fonseca2017; Lee et al., Reference Lee, Guidetti, Cesari, Gariepy, Park and Park2018; Kapantaidaki et al., Reference Kapantaidaki, Evangelou, Iii, Leskey, Brodeur and Milonas2019).
The comprehensive characterization of the genetic diversity of H. halys showed a high diversity within Asian populations and a significant genetic structure between populations from China, Korea and Japan (Xu et al., Reference Xu, Fonseca, Hamilton, Hoelmer and Nielsen2014). In contrast, invasive populations in the USA showed much lower genetic diversity than in the native range with only two haplotypes present before 2008 (Xu et al., Reference Xu, Fonseca, Hamilton, Hoelmer and Nielsen2014; Valentin et al., Reference Valentin, Nielsen, Wiman, Lee and Fonseca2017) and three additional haplotypes described in 2016 (Valentin et al., Reference Valentin, Nielsen, Wiman, Lee and Fonseca2017). In particular, the presence of a single haplotype (H1) in the east coast suggests that the invasion resulted from a single introduction event of a few founder individuals (Xu et al., Reference Xu, Fonseca, Hamilton, Hoelmer and Nielsen2014). Since in Asia H1 is exclusively present in China, it is assumed that the first individuals on the east coast were introduced from China (Valentin et al., Reference Valentin, Nielsen, Wiman, Lee and Fonseca2017). In contrast, higher levels of haplotype diversity in the west coast indicate that multiple introduction events have occurred (Xu et al., Reference Xu, Fonseca, Hamilton, Hoelmer and Nielsen2014; Valentin et al., Reference Valentin, Nielsen, Wiman, Lee and Fonseca2017). Europe, in contrast shows a different picture: while populations in Switzerland belong mainly to haplotype H3 and the less common haplotype H8, individuals in Northern Italy across different populations in the Emilia Romagna region are mainly H1 (Cesari et al., Reference Cesari, Maistrello, Ganzerli, Dioli, Rebecchi and Guidetti2014; Gariepy et al., Reference Gariepy, Haye, Fraser and Zhang2014, Reference Gariepy, Bruin, Haye, Milonas and Vétek2015). This suggests that while the first introduction of H. halys into Switzerland resulted from a direct interception from China, a second introduction occurred in Northern Italy, probably from an eastern US population (Valentin et al., Reference Valentin, Nielsen, Wiman, Lee and Fonseca2017). Moreover, a high haplotype diversity was found in Greece. Since several haplotypes – including the most common haplotype H33 – were outside of their native range only described in Greece, it is assumed that populations in Southeastern Europe resulted from an independent introduction event from China (Cesari et al., Reference Cesari, Maistrello, Ganzerli, Dioli, Rebecchi and Guidetti2014; Gariepy et al., Reference Gariepy, Haye, Fraser and Zhang2014, Reference Gariepy, Bruin, Haye, Milonas and Vétek2015; Valentin et al., Reference Valentin, Nielsen, Wiman, Lee and Fonseca2017).
A recent characterization of populations across Italy compared the haplotype frequency across different years and provided new insight into invasion dynamics on a smaller scale (Cesari et al., Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018). In 2013 just two haplotypes were identified in Italy, H1 in the Emilia Romagna region and H3 in the Lombardy region (Cesari et al., Reference Cesari, Maistrello, Ganzerli, Dioli, Rebecchi and Guidetti2014, Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018). While the population in Lombardy likely resulted from an introduction from Switzerland, H. halys in Emilia Romagna was probably introduced from Eastern USA (Valentin et al., Reference Valentin, Nielsen, Wiman, Lee and Fonseca2017; Cesari et al., Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018). The number of haplotypes increased in just a few years with a total of 22 haplotypes recorded in 2016, likely as a result of multiple introduction events and range expansions from other European populations (Cesari et al., Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018).
Although the invasion history of H. halys on a global (Xu et al., Reference Xu, Fonseca, Hamilton, Hoelmer and Nielsen2014; Valentin et al., Reference Valentin, Nielsen, Wiman, Lee and Fonseca2017; Lee et al., Reference Lee, Guidetti, Cesari, Gariepy, Park and Park2018; Kapantaidaki et al., Reference Kapantaidaki, Evangelou, Iii, Leskey, Brodeur and Milonas2019), continental (Cesari et al., Reference Cesari, Maistrello, Ganzerli, Dioli, Rebecchi and Guidetti2014; Gariepy et al., Reference Gariepy, Bruin, Haye, Milonas and Vétek2015) and national (Cesari et al., Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018) scale has been well-investigated, studies describing the invasion dynamics on a smaller scale are currently scarce. Here we focus on the genetic diversity of H. halys in South Tyrol (Northern Italy). This region is of special interest as it is one of the largest contiguous areas of apple production in Europe and therefore threatened by this emerging agriculture pest. The first exemplars of H. halys were found in South Tyrol in 2016 on a cargo from Bergamo (Lombardy, Italy) (Unterthurner, Reference Unterthurner2016). Since 2016 H. halys was frequently found in South Tyrol suggesting a successful establishment in this region (Unterthurner et al., Reference Unterthurner, Ladurner, Zelger, Fischnaller, Walcher and Haye2017). Moreover, from 2016 the frequency of H. halys detected in a monitoring program increased across the region, confirming the successful establishment and rapid spatial spread in this region (Fischnaller et al., unpublished). Whether the current population in South Tyrol resulted from a single introduction in 2016 and a subsequent spread or from multiple introduction events across different years is currently not known. We therefore aim to study the genetic diversity across different populations sampled across 4 years and compare our data to previously published global, continental and national data to understand the population dynamics of H. halys in this area.
Materials and methods
Insect collection, DNA extraction and sequencing
A total of 156 individuals of H. halys collected from 2016 to 2019 across different localities in South Tyrol were analyzed. Localities within an area of 3 km were clustered into populations resulting in a total of nine different geographic populations (table 1). The number of individuals per population ranged on average from 14 to 16 individuals with the exceptions of Klausen (population 9, n = 18), Montan (population 9, n = 5) and Laimburg (population 7, n = 43). In total 21 individuals were sampled in 2016, 17 in 2017, 33 in 2018 and 85 in 2019. Additionally, six individuals that were analyzed previously in South Tyrol by Cesari et al. (Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018) were included in our dataset (population *Brixen).
Table 1. Sampling sites, geographical coordinates, year captured and abundance of different haplotypes
After taxonomic determination all individuals were stored in absolute ethanol at −20°C. DNA was extracted from a single hindleg using the GenElute Mammalian Genomic DNA miniprep kit (Sigma-Aldrich, Steinheim, Germany). DNA was eluted in 100 μl elution solution (10 mM Tris, 1 mM EDTA) and stored at 4°C. A 615 bp fragment of the mitochondrial COI was polymerase chain reaction (PCR) amplified using primers LCO1490 and HCO2198 (Folmer et al., Reference Folmer, Black, Hoeh, Lutz and Vrijenhoek1994). This region has been characterized already in other studies (Gariepy et al., Reference Gariepy, Haye, Fraser and Zhang2014, Reference Gariepy, Bruin, Haye, Milonas and Vétek2015; Xu et al., Reference Xu, Fonseca, Hamilton, Hoelmer and Nielsen2014; Zhu et al., Reference Zhu, Ye, Du, Zhang, Zhen, Zheng, Zhao, Li and Bu2016; Cesari et al., Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018; Lee et al., Reference Lee, Guidetti, Cesari, Gariepy, Park and Park2018) allowing a direct comparison with previously published data.
Each PCR reaction was performed in a total volume of 25 μl containing 7.5 μl H2O, 12.5 μl UCP HiFidelity PCR Master Mix (Qiagen, Hilden, Germany), 0.25 μM of each primer and 2.5 μl template DNA. All reactions were performed in an Eppendorf Mastercycler Gradient (Eppendorf; Hamburg, Germany) with the following conditions: 35 cycles with 30 s at 98°C, 10 s at 98°C, 10 s at 51°C, 15 s at 72°C followed by a final extension at 72°C for 2 min. All PCR products were run on a 1% agarose gel stained with SYBRsafe (Thermo Fisher Scientific) and Sanger-sequenced by Eurofins MWG Operon (Ebersberg, Germany).
Population genetic analysis
Sequences were edited manually and aligned with CodonCode Aligner v8.0 (CodonCode Dedham, MA, USA). Our sequences were compared to sequences available in Genbank using the BLAST algorithm in NCBI. To exclude PCR artifacts, sequences with ambiguous polymorphisms and individuals with unique haplotypes (see ‘Results’ section) were confirmed by sequencing the product of two independent PCR runs. New haplotypes were deposited in the Genbank under the accession numbers MT199091–MT199094. The number of haplotypes (H), haplotype diversity (h) and nucleotide diversity (p) were determined using the program DNAsp v6 (Librado and Rozas, Reference Librado and Rozas2009). Neighbor-joining tree analysis was performed with Mega X (Kumar et al., Reference Kumar, Stecher, Li, Knyaz and Tamura2018) using the Tamura 3-parameter model (T92+G).
Results
We found a total of 15 different haplotypes across all populations of H. halys in South Tyrol. Eleven haplotypes have been described already in several other native and invasive populations, whereas four haplotypes (H164–167) were detected here for the first time. The most prevalent haplotype H1 was found in 86 out of 162 individuals (53.37%) in each population in South Tyrol (fig. 1; table 1). Haplotype H3 was found in 24 individuals (14.72%) across all populations, and was absent only in Montan (population 8). Haplotype H153 was found in 17 individuals (10.43%) across all populations except for Montan and Brixen (Cesari et al., Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018). The other 12 haplotypes were found in low frequencies in one (0.61%; H155 and H167) to seven (4.29%; H78) individuals.
Figure 1. Geographical distributions of H. halys haplotypes across South Tyrol. Population localities are listed in table 1, the pie chart indicated by an asterisk represents a population studied by Cesari et al. (Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018). The larger pie chart represents the total frequency of all haplotypes.
Overall, we found a high genetic diversity of H. halys in South Tyrol with a haplotype diversity of 0.684 ± 0.00128 and a nucleotide diversity of 0.00455 ± 0.00037(table 2). The number of haplotypes per population ranges from two in Montan (population 8) to eight in Klausen (population 9). Apart from the low haplotype and nucleotide diversity in Brixen in 2016 (Cesari et al., Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018; table 2), the lowest haplotype diversity in our study was found in Montan (population 8) with a haplotype diversity of h = 0.4 ± 0.237. The lowest nucleotide diversity was found in Laimburg (population 7) with p = 0.0024 ± 0.00065 (table 2). In contrast, the highest haplotype diversity was found in Meran (population 2) and Terlan (population 5) with h = 0.85 ± 0.057, while the highest nucleotide diversity was found in Meran with p = 0.00694 ± 0.0008 (table 2).
Table 2. Genetic diversity of H. halys in South Tyrol
n, number of individuals analyzed; N, number of haplotypes; h, haplotype diversity; p, nucleotide diversity; data of *Brixen were taken from Cesari et al. (Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018).
The comparison between individuals collected in different years shows an increasing number of haplotypes across the different years: while in 2016 six haplotypes and in 2017 four haplotypes were found, the number of haplotypes increased to nine in 2018 and 14 in 2019 (fig. 2). This is reflected by a lower haplotype diversity (0.453 ± 0.116 and 0.419 ± 0.141) and nucleotide diversity (0.00321 ± 0.00101 and 0.00341 ± 0.00137) in 2016 and 2017, respectively, compared to 2018 (h = 0.716 ± 0.067 and p = 0.00368 ± 0.00073) and 2019 (h = 0.768 ± 0.039 and p = 0.00538 ± 0.00044; table 2).
Figure 2. Frequency of H. halys haplotypes across different years.
The most widespread haplotype across all years and regions in South Tyrol is H1. This haplotype is the most common haplotype in the native range in Asia but also in most countries of the world. The haplotype is present in China and Japan but not in Korea, it is the most widespread in the USA, Canada and most European countries (Gariepy et al., Reference Gariepy, Bruin, Haye, Milonas and Vétek2015; Cesari et al., Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018) (table 3). In 2016 five other haplotypes were already present in South Tyrol. Haplotype H3 has been described in its native range in China (Gariepy et al., Reference Gariepy, Bruin, Haye, Milonas and Vétek2015) and is present in Hungary, Greece and Italy, and is described as the most prevalent haplotype in Switzerland and France (Gariepy et al., Reference Gariepy, Bruin, Haye, Milonas and Vétek2015; Cesari et al., Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018; Lee et al., Reference Lee, Guidetti, Cesari, Gariepy, Park and Park2018). In Italy this haplotype is frequent in Lombardy and has been recorded also in Piedmont, Emilia Romagna, Lazio and Calabria (Cesari et al., Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018). Haplotype H23 was described in Switzerland, France (Gariepy et al., Reference Gariepy, Haye, Fraser and Zhang2014) and in Italy where it was exclusively found in Veneto (Morrison et al., Reference Morrison, Milonas, Kapantaidaki, Cesari, Di Bella, Guidetti, Haye, Maistrello, Moraglio, Piemontese, Pozzebon, Ruocco, Short, Tavella, Vétek and Leskey2017) while H78 was described in China (Zhu et al., Reference Zhu, Ye, Du, Zhang, Zhen, Zheng, Zhao, Li and Bu2016) and is especially present in Liguria and southern Piedmont. In contrast, H153 has been solely found in Italy where it was exclusively recorded in Veneto (Cesari et al., Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018; table 2). One individual in 2016 belonged to a haplotype (H164) which has not been described yet.
Table 3. Overview of the haplotypes found in South Tyrol compared with descriptions in their native range (Xu et al., Reference Xu, Fonseca, Hamilton, Hoelmer and Nielsen2014, Valentin et al., Reference Valentin, Nielsen, Wiman, Lee and Fonseca2017), invasive range in Europe and North America (Gariepy et al., Reference Gariepy, Haye, Fraser and Zhang2014, Reference Gariepy, Bruin, Haye, Milonas and Vétek2015; Morrison et al., Reference Morrison, Milonas, Kapantaidaki, Cesari, Di Bella, Guidetti, Haye, Maistrello, Moraglio, Piemontese, Pozzebon, Ruocco, Short, Tavella, Vétek and Leskey2017, Lee et al., Reference Lee, Guidetti, Cesari, Gariepy, Park and Park2018, Kapantaidaki et al., Reference Kapantaidaki, Evangelou, Iii, Leskey, Brodeur and Milonas2019) and different regions in Italy (Cesari et al., Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018)
Countries highlighted in bold represent the country/region where the corresponding haplotype is predominant.
Haplotypes H1, H3, H151 and H153 were still present in 2017 and all six haplotypes described in 2016 were found again in 2018. Three additional haplotypes were found in 2018: haplotype H8 has been described in low frequency in France and Switzerland as well as in Italy in low frequencies in Lombardy, Piedmont, Emilia-Romagna and Veneto (Cesari et al., Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018; table 3). Haplotype H152 was exclusively found in Piedmont in low frequencies, whereas H41, a haplotype that originates in Japan, has been described outside its native range exclusively in Veneto (Cesari et al., Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018; table 3).
In 2019, besides H164 all haplotypes have been found again, whereas six haplotypes appeared for the first time. Haplotype H33 was described already in China and outside its native range exclusively in Greece (Gariepy et al., Reference Gariepy, Bruin, Haye, Milonas and Vétek2015). In 2019 this haplotype was found in Meran (population 2) and Klausen (population 9). Haplotype H180 occurs in low frequencies in Italy, in Lombardy, Piedmont, Veneto and has been found in Meran (population 2) and Bozen (population 6), whereas H155 outside its native range occurs only in low frequencies in Veneto, Trentino (Cesari et al., Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018) and has been found in Klausen (population 9). Finally, three haplotypes (H165, H166 and H167) that have not been previously characterized in any other region in the world are scattered across South Tyrol.
Phylogeny and network analysis of the different COI haplotypes described in previous studies showed that haplotypes described in China and Korea formed a different clade than those found in Japan (Zhu et al., Reference Zhu, Ye, Du, Zhang, Zhen, Zheng, Zhao, Li and Bu2016; Valentin et al., Reference Valentin, Nielsen, Wiman, Lee and Fonseca2017; Lee et al., Reference Lee, Guidetti, Cesari, Gariepy, Park and Park2018). The incorporation of the 15 haplotypes in South Tyrol revealed that six haplotypes (H1, H3, H8, H78, H80 and H152) clustered with haplotypes described in China, one haplotype (H33) clustered with a branch containing individuals from China and Japan, and eight haplotypes (H23, H41, H153, H155, H164, H165, H166 and H167) clustered with haplotypes described primarily in Japan (fig. 3).
Figure 3. Neighbor-joining tree of H. halys haplotypes described in Xu et al. (Reference Xu, Fonseca, Hamilton, Hoelmer and Nielsen2014) and Valentin et al. (Reference Valentin, Nielsen, Wiman, Lee and Fonseca2017) with the corresponding Genbank accession number, haplotypes present in South Tyrol are highlighted in bold. Black branch colors represent haplotypes predominantly found in China whereas red branches include haplotypes mainly found in Japan based on Xu et al. (Reference Xu, Fonseca, Hamilton, Hoelmer and Nielsen2014) and Valentin et al. (Reference Valentin, Nielsen, Wiman, Lee and Fonseca2017).
Discussion
The invasion history of the brown marmorated stink bug H. halys in North America and Europe has been investigated in several recent studies (Gariepy et al., Reference Gariepy, Haye, Fraser and Zhang2014, Reference Gariepy, Bruin, Haye, Milonas and Vétek2015; Valentin et al., Reference Valentin, Nielsen, Wiman, Lee and Fonseca2017; Cesari et al., Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018; Lee et al., Reference Lee, Guidetti, Cesari, Gariepy, Park and Park2018). By comparing mitochondrial haplotypes of native populations from Asia with populations outside its native range, the invasion history was studied on a global (Gariepy et al., Reference Gariepy, Haye, Fraser and Zhang2014; Valentin et al., Reference Valentin, Nielsen, Wiman, Lee and Fonseca2017; Lee et al., Reference Lee, Guidetti, Cesari, Gariepy, Park and Park2018), continental (Gariepy et al., Reference Gariepy, Bruin, Haye, Milonas and Vétek2015; Cesari et al., Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018) and national (Cesari et al., Reference Cesari, Maistrello, Ganzerli, Dioli, Rebecchi and Guidetti2014, Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018; Xu et al., Reference Xu, Fonseca, Hamilton, Hoelmer and Nielsen2014) scale. In the present study, we document for the first time the invasion history of H. halys on a regional scale in a time period of 4 years. We describe a remarkably high haplotype diversity across the whole range in South Tyrol with a high number of haplotypes that were increasing across the different years. The comparison of the haplotypes found in South Tyrol with previously published studies highlight a complex invasion history in this region, challenging the previously believed introduction hypothesis of H. halys in Europe.
The most striking result of our study is the high number of haplotypes present in South Tyrol. General invasive populations express a lower genetic diversity than populations in their native range after experiencing a strong population bottleneck (e.g. Dlugosch and Parker, Reference Dlugosch and Parker2008). The founder effect was particularly strong in H. halys populations in North America where initially just one to two haplotypes were described in the USA (Gariepy et al., Reference Gariepy, Haye, Fraser and Zhang2014; Xu et al., Reference Xu, Fonseca, Hamilton, Hoelmer and Nielsen2014) and just three haplotypes in Canada (Gariepy et al., Reference Gariepy, Haye, Fraser and Zhang2014; Valentin et al., Reference Valentin, Nielsen, Wiman, Lee and Fonseca2017). Although the number of haplotypes in North America increased likely because of additional introductions (Morrison et al., Reference Morrison, Milonas, Kapantaidaki, Cesari, Di Bella, Guidetti, Haye, Maistrello, Moraglio, Piemontese, Pozzebon, Ruocco, Short, Tavella, Vétek and Leskey2017; Valentin et al., Reference Valentin, Nielsen, Wiman, Lee and Fonseca2017; Lee et al., Reference Lee, Guidetti, Cesari, Gariepy, Park and Park2018; Kapantaidaki et al., Reference Kapantaidaki, Evangelou, Iii, Leskey, Brodeur and Milonas2019), genetic diversity of H. halys in North America is still limited with a maximum of five haplotypes described (Valentin et al., Reference Valentin, Nielsen, Wiman, Lee and Fonseca2017). Similarly, the number of haplotypes in various European countries is low, especially in Romania (one haplotype), Hungary (two haplotypes), France (three haplotypes) and Switzerland (four haplotypes) (Gariepy et al., Reference Gariepy, Bruin, Haye, Milonas and Vétek2015; Morrison et al., Reference Morrison, Milonas, Kapantaidaki, Cesari, Di Bella, Guidetti, Haye, Maistrello, Moraglio, Piemontese, Pozzebon, Ruocco, Short, Tavella, Vétek and Leskey2017; Valentin et al., Reference Valentin, Nielsen, Wiman, Lee and Fonseca2017; Cesari et al., Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018; Lee et al., Reference Lee, Guidetti, Cesari, Gariepy, Park and Park2018). In contrast, populations in Greece show considerably high genetic diversity with up to ten haplotypes described (Gariepy et al., Reference Gariepy, Bruin, Haye, Milonas and Vétek2015; Valentin et al., Reference Valentin, Nielsen, Wiman, Lee and Fonseca2017; Lee et al., Reference Lee, Guidetti, Cesari, Gariepy, Park and Park2018; Kapantaidaki et al., Reference Kapantaidaki, Evangelou, Iii, Leskey, Brodeur and Milonas2019). Moreover, the use of multiple molecular markers revealed up to 22 haplotypes in Italy (Cesari et al., Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018).
On a regional scale up to ten different haplotypes were described in Piedmont and Lombardy based on a combined analysis of the mitochondrial COI and COII regions (Cesari et al., Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018). Considering only the COI locus used in our study a maximum of eight haplotypes where described in Piedmont, seven in Lombardy and six in Veneto (Cesari et al., Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018). Thus, with 15 haplotypes described in South Tyrol, this region represents the highest number and diversity of haplotypes outside the native range of H. halys.
H. halys was recorded in Italy for the first time in 2012, in the province of Reggio Emilia (Maistrello et al., Reference Maistrello, Dioli, Vaccari and Caruso2014). The first individuals in South Tyrol were found four years later in March 2016, with some individuals detected in open fields and remarkably high numbers found in packing materials in shipments from Bergamo, Northern Italy (Unterthurner et al., Reference Unterthurner, Ladurner, Zelger, Fischnaller, Walcher and Haye2017). By haplotyping 27 individuals (including six individuals studied by Cesari et al. (Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018)) from 2016 we found six haplotypes present in South Tyrol: H1, H3, H8, H78, H153 and H164. By that time just two haplotypes, H1 and H3, were described in Bergamo, whereas H78 and H8 have been described in Brescia and in Como, respectively, approximately 50 km next to Bergamo. However, neither H153 nor H164 have been found in this geographic area (Cesari et al., Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018). Although these findings are generally in concordance with a primary introduction of H. halys from Lombardy, the presence of haplotype H153 that has been described exclusively in Veneto (Morrison et al., Reference Morrison, Milonas, Kapantaidaki, Cesari, Di Bella, Guidetti, Haye, Maistrello, Moraglio, Piemontese, Pozzebon, Ruocco, Short, Tavella, Vétek and Leskey2017; Cesari et al., Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018) and H164, a haplotype that has not been described outside of South Tyrol indicates that additional introductions, likely secondary invasions from other areas have occurred.
Generally, we could not find a significant impact of geography on the haplotype composition, with H1 present in the majority of individuals from all populations and H3 and H153 that were present in most populations in minor frequencies. The other haplotypes were scattered across the region with no spatial structure. The haplotype diversity was generally higher in cities (Meran, Bozen and Klausen) than in villages with a particularly low nucleotide diversity in remote areas in Völlan, Laimburg and Montan. This emphasizes the important role of active and passive human-mediated transportation for the dispersal of H. halys but might also highlight a general preference for urban areas where a higher diversity of host plants and the presence of suitable overwintering shelter locations might benefit the insect (Hoebeke and Carter, Reference Hoebeke and Carter2003; Leskey et al., Reference Leskey, Hamilton, Nielsen, Polk, Rodriguez-Saona, Christopher Bergh, Ames Herbert, Kuhar, Pfeiffer, Dively, Hooks, Raupp, Shrewsbury, Krawczyk, Shearer, Whalen, Koplinka-Loehr, Myers, Inkley, Hoelmer, Lee and Wright2012; Wallner et al., Reference Wallner, Hamilton, Nielsen, Hahn, Green and Rodriguez-Saona2014; Leskey and Nielsen, Reference Leskey and Nielsen2018).
Essentially all haplotypes described in different years were still present in the following years suggesting that introduced H. halys individuals were able to establish into the new territory. However, the number of haplotypes and the general haplotype diversity increased across the different years. While the number of haplotypes in 2016 and 2017 with six and four haplotypes was relatively low, it increased to nine and 14 haplotypes in the following two years. Although the higher sample size in 2019 might have affected this outcome, both haplotype and nucleotide difference increased across the years (table 2). This is in line with Cesari et al. (Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018) who described an increasing haplotype diversity in Italy within a few years. We interpret this result as a consequence of repeated introduction of H. halys from different populations across Italy and other countries. Especially the detection of H33, a haplotype that outside its native range was found exclusively in Greece indicates that secondary invasions likely due to long-range human-mediated transportation might have occurred.
It is unclear, why the populations in a restricted area in South Tyrol show a higher haplotype diversity than other areas previously studied. Most of the studies describing the haplotype diversity of H. halys in invasive areas in Europe and North America describe populations sampled between 2011 and 2016 (Gariepy et al., Reference Gariepy, Haye, Fraser and Zhang2014; Morrison et al., Reference Morrison, Milonas, Kapantaidaki, Cesari, Di Bella, Guidetti, Haye, Maistrello, Moraglio, Piemontese, Pozzebon, Ruocco, Short, Tavella, Vétek and Leskey2017; Valentin et al., Reference Valentin, Nielsen, Wiman, Lee and Fonseca2017; Cesari et al., Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018; Lee et al., Reference Lee, Guidetti, Cesari, Gariepy, Park and Park2018; Kapantaidaki et al., Reference Kapantaidaki, Evangelou, Iii, Leskey, Brodeur and Milonas2019) or earlier (Xu et al., Reference Xu, Fonseca, Hamilton, Hoelmer and Nielsen2014). It might be possible that additional introductions from the native range into Europe occurred recently and increased the haplotype diversity in other areas as well. Valentin et al. (Reference Valentin, Nielsen, Wiman, Lee and Fonseca2017) analyzed the invasion history of H. halys and concluded that most invasive populations in North America and Europe established from a direct introduction of H. halys populations from China with separate introductions into Eastern and Western USA and Canada, as well as into Switzerland and Greece. In contrast, the population in Northern Italy was established by a secondary introduction from Eastern USA into Emilia Romagna and the spread of the founder population in Switzerland to the neighboring countries. The authors found little evidence that populations outside China such as Japan and Korea established outside their native range (Valentin et al., Reference Valentin, Nielsen, Wiman, Lee and Fonseca2017). Since half of the haplotypes found in South Tyrol have not been described in their native range, we performed a phylogenetic analysis to reconstruct their potential origin. Surprisingly, half of the haplotypes described in South Tyrol clustered with haplotypes that were previously described in Japan (Xu et al., Reference Xu, Fonseca, Hamilton, Hoelmer and Nielsen2014). This includes all four newly described haplotypes as well as four haplotypes that have been described previously in other regions in Italy (Cesari et al., Reference Cesari, Maistrello, Piemontese, Bonini, Dioli, Lee, Park, Partsinevelos, Rebecchi and Guidetti2018). As limited commercial relations between South Tyrol and Japan exist, we consider a direct introduction of H. halys into South Tyrol unlikely. Nevertheless, our results suggest that an introduction of H. halys individuals from Japan might have occurred in Italy or neighboring countries recently, a hypothesis that needs to be tested in future studies.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S0007485320000553.
Acknowledgements
We thank Evi Deltedesco, Hannes Elsler, Markus Ladurner, Manuel Messner, Toni Riegler, Anna Rottensteiner, Arnold Schuler, Tanja Schuler, Raffael Tiziani and Michael Unterthurner for providing insect specimens, Manfred Wolf, Thomas M. Wolfe, Christian Stauffer and two anonymous reviewers for helpful comments and suggestions and Michele Cesari for clarification of previously published data. This study was funded by the Austrian Science Fund (project number P31441) and internal funds of the Free University of Bozen-Bolzano to HS and internal funds of the Research Centre Laimburg and the South Tyrolean Apple Consortium (SAK) to SF. The authors thank the Department of Innovation, Research and University of the Autonomous Province of Bozen/Bolzano for covering the Open Access publication costs.
Author contributions
This work is part of the bachelor thesis of DE, supervised by HS and SF. HS and SF designed the study, HS, DE and SF collected and taxonomically identified the samples, DE and HS conducted lab work and analyzed the data. HS wrote the first draft of the manuscript and all authors contributed to the final version.
Conflict of interest
The authors declare no competing interests.
Data availability
Sequence data will be deposited in public databases after manuscript acceptance.
Open access
