DOI: 10.1111/eea.12041 Invasion genetics of American cherry fruit flyinEurope and signals of hybridization with the European cherry fruit fly Jes Johannesen1*, Nusha Keyghobadi2, Hannes Schuler3, Christian Stauffer3 & Heidrun Vogt4 1Institute of Zoology, Ecological Department, University of Mainz, Johann-Joachim-Becherweg 13, Mainz 55128, Germany, 2Department of Biology, Western University, London, Ontario, ON N6A 5B7, Canada, 3Department of Forest & Soil Sciences, Institute of Forest Entomology, Forest Pathology & Forest Protection, BOKU, University of Natural Resources & Applied Life Sciences, Hasenauerstrasse 38, Vienna 1190, Austria, and 4Julius Kuhn-Institute,€ Federal Research Centre for Cultivated Plants, Institute for Plant Protection in Fruit Crops and Viticulture, Schwabenheimer Strasse 101, Dossenheim 69221, Germany Accepted: 2 January 2013

Key words: cingulata, Rhagoletis indifferens, Rhagoletis cerasi, microsatellites, COI, cross- species amplification, hybridization, invasion biology, pest species, Diptera, , wing pattern

Abstract The American cherry fruit fly is an invasive pest species in Europe, of serious concern in tart cherry production as well as for the potential to hybridize with the European cherry fruit fly, Rhagoletis cerasi L. (Diptera: Tephritidae), which might induce new pest dynamics. In the first European reports, the question arose whether only the eastern American cherry fruit fly, Rhagoletis cingulata (Loew) (Di- ptera: Tephritidae), is present, or also the closely related western American cherry fruit fly, Rhagoletis indifferens Curran. In this study, we investigate the species status of European populations by com- paring these with populations of both American species from their native ranges, the invasion dynamics in German (first report in 1993) and Hungarian (first report in 2006) populations, and we test for signals of hybridization with the European cherry fruit fly. Although mtDNA sequence gene- alogy could not separate the two American species, cross-species amplification of 14 microsatellite loci separated them with high probabilities (0.99–1.0) and provided evidence for R. cingulata in Europe. German and Hungarian R. cingulata populations differed significantly in microsatellite allele frequencies, mtDNA haplotype and wing pattern distributions, and both were genetically depauper- ate relative to North American populations. The diversity suggests independent founding events in Germany and Hungary. Within each country, R. cingulata displayed little or no structure in any trait, which agrees with rapid local range expansions. In cross-species amplifications, signals of hybridiza- tion between R. cerasi and R. cingulata were found in 2% of R. cingulata individuals and in 3% of R. cerasi. All putative hybrids had R. cerasi mtDNA indicating that the original between-species mating involved R. cerasi females and R. cingulata males.

Doody et al., 2009; Luan et al., 2012), whereas indirect Introduction effects between native and invasive species can arise via the Invasive species have serious impacts on agricultural crops transmission of new pathogens (Kozubikova et al., 2009). and biodiversity worldwide (Shine, 2007; Vie et al., 2009). A third potentially serious impact of invasive species Deleterious effects may take place by direct interaction involves altering the genetic constitution of native species between species for resources, such as predation, herbiv- through hybridization and introgression. This may for ory, or competitive exclusion (Reitz & Trumble, 2002; example lead to species erosion (Elledge et al., 2008) or new disease syndromes (Brasier, 2001; Nunney et al., *Correspondence: Jes Johannesen, Institute of Zoology, Ecological 2010). Department, University of Mainz, Johann-Joachim-Becherweg 13, The eastern American cherry fruit fly, Rhagoletis cingu- Mainz 55128, Germany. E-mail: [email protected] lata (Loew) (Diptera, Tephritidae), is an invasive species

© 2013 The Netherlands Entomological Society Entomologia Experimentalis et Applicata 1–12, 2013 Entomologia Experimentalis et Applicata © 2013 The Netherlands Entomological Society 1 2 Johannesen et al. in Europe that causes direct negative effects in agricultural 1997), of which the two most widespread are R. cingulata cherry production and potentially poses the threat of and Rhagoletis indifferens Curran. In the first reports from hybridizing with the native European cherry fruit fly, Europe, it was debated whether one or both of these two Rhagoletis cerasi L. Direct negative effects are caused by lar- species were present (Merz, 1991). The two species are val infestation of cherries, particularly in Prunus cerasus L. genetically closely related (Berlocher & Bush, 1982; Smith (tart cherry; Rosaceae) (Lampe et al., 2005). The infesta- & Bush, 1997) and morphologically very similar (Bush, tion of tart cherries poses a new pest syndrome because 1966). In the native range, the main identification trait of important economic varieties like ‘Schattenmorelle’ and these very similar species is the geographical distribution. ‘Gerema’ mature roughly 3–4 weeks later than the main The eastern cherry fruit fly R. cingulata occurs in eastern sweet cherry varieties, Prunus avium L., and, in doing so, and midwestern USA, Canada, and Mexico, whereas the they have avoided high infestation rates by R. cerasi, whose western cherry fruit fly R. indifferens is found in western peak flight period is mainly timed to the phenology of North America. Morphologically, R. cingulata has a yellow middle- to late-ripening sweet cherry. By contrast, the fore coxa and the anterior apical crossband on the wing is flight period of R. cingulata coincides with the above- often reduced to an isolated spot, whereas R. indifferens mentioned tart cherry varieties causing infestation rates up has posteriorly black shaded fore coxa and the anterior to 30% (Vogt, 2007; Vogt et al., 2010). Reported for the apical crossband is rarely reduced to an isolated spot first time in Europe in 1983 in Switzerland and in 1993 in (Foote et al., 1993). Although the morphological determi- Germany, R. cingulata is now established in Germany nation is most easily based on the frequency of the (Lampe et al., 2005; this study) and The Netherlands crossband’s spot-line pattern, it remains a frequency based (Smit & Dijkstra, 2008), whereas recent reports from Hun- and not a decisive trait and this has perhaps led to confu- gary (Szeoke,€ 2006) and Slovenia (OEPP/EPPO, 2007) sion in species determinations outside the core distribu- indicate new pest status in these countries. tion ranges in USA (CABI, 2012) and in the invasive range The potential threat of hybridization posed by Ameri- in Europe (Merz, 1991; van Aartsen, 2001). can cherry fruit flies is the possibility for introgression into In this study, we asked the following four questions to R. cerasi populations and the evolution of novel pest explore the species status and population dynamics of dynamics. In the Tephritidae, plants for oviposition also American cherry fruit flies in Europe. (1) Which species is/ act as mating locations (Zwolfer,€ 1974). Both the Ameri- are present in Europe? (2) How genetically structured can and the European cherry fruit fly use cherry trees as are European populations of American cherry fruit flies? mating location, and this creates the possibility for contact (3) Are there signs of multiple introductions in the Euro- and hybridization between them. In North America, inde- pean populations? (4) Are there signals of hybridisation pendent host plant shifts of two native Rhagoletis species, with the European cherry fruit fly? The second question Rhagoletis mendax Curran and Rhagoletis zephyria Snow, addresses the relative rate of dispersal in the introduced to the same introduced plant, Lonicera spec., has led to the range. We answer these questions by comparing allele fre- evolution of a hybrid Rhagoletis species (Schwarz et al., quency distributions at 14 microsatellite loci, mtDNA 2005). Although American and European cherry fruit flies haplotypes, and wing patterns of North American cherry have different peak flight and mating periods, they show fruit flies sampled in Germany and Hungary and compare considerable overlap (Lampe et al., 2005; Vogt et al., them with reference populations of R. cingulata and R. in- 2010). The phenological overlap, and with it the potential differens from eastern and western North America, respec- for increasing mating interactions between the species, can tively, and R. cerasi collected in Germany and Austria. The be extended by prolonging the sweet cherry season with European populations of American fruit fly in this study late varieties or by different management practices (Teix- represent a succession of population establishments: ‘old’ eira et al., 2007). The observation that some R. cerasi do populations in western Germany, ‘middle-aged’ popula- not diapause (Koppler,€ 2008) could also lead to a second tions in eastern Germany, and ‘young’ populations in generation in the late flight period of R. cingulata. Finally, Hungary. contact between the species may be facilitated by the wide distribution in Europe of the native host plant Prunus serotina Ehrh. (black cherry). Thus, both the possibility Materials and methods of bridging flight seasons and the introduction of a new Sampling host plant to R. cerasi (i.e., P. serotina) may facilitate were collected in North America and Europe in 2009 hybridization. and 2010 in P. avium (sweet cherry) and P. cerasus (tart The cherry-infesting R. cingulata species complex is cherry) orchards and in stands of P. serotina (black divided into four species (Bush, 1966; Smith & Bush, cherry), with yellow sticky traps or as larvae from infested American cherry fruit fly in Europe 3

Table 1 Populations investigated for American and European cherry fruit flies (Rhagoletis spp.). The population number (#) refers to Figure 1. n, number of analysed flies. Wing pattern numbers represent the number of flies with spot/line patterns and are based on flies with intact wings

Wing Host plant Species Country Region/state Population (#) n pattern (Prunus spp.) Sampled by R. cingulata USA Michigan Coloma (16) 9 8/1 P. cerasus LAF Teixeira Laubach (17) 4 2/1 – N Rothwell Evans (18) 8 3/2 – N Rothwell Watervliet (19) 34 30/4 P. serotina LAF Teixeira CB Setterbo (20) 16 12/1 – N Rothwell Hungary Pest Erd-Elvira (12) 57 11/44 P. cerasus, P. avium E Voigt Veszprem Felo~ors€ (13) 32 –– C Stauffer Fejer Agard (14) 28 11/16 P. cerasus, P. avium E Voigt Pest Sosk ut (15) 8 2/4 P. cerasus E Voigt Eastern Thuringen€ Ammern (1) 186 125/61 – E Maring Germany Dollst€ €adt (2) 10 7/3 – E Maring Brandenburg Frankfurt-Oder-N (3) 63 42/21 P. serotina U Holz Frankfurt-Oder-Q (4) 71 41/29 P. serotina U Holz Sachsen Dresden (5) 10 –– A Trapp Thuringen€ Kleinfahrer (6) 17 15/2 – E Maring Western Baden-Wurttemberg€ Dossenheim (7) 17 12/7 – H Vogt Germany Mannheim (8) 50 38/7 P. serotina H Vogt Rheinland-Pfalz Ingelheim (9) 74 40/34 P. cerasus G Hensel Heidesheim-1 (10) 16 – P. cerasus G Hensel, W Dahlbender Heidesheim-2 (11) 25 – P. cerasus G Hensel, W Dahlbender R. indifferens USA Utah Kaysville (22) 19 0/15 P. cerasus D Alston Washington Cle Elum (21) 30 8/19 P. avium W Yee Canada British Columbia Creston (23) 11 – P. emarginata H Thistlewood Salmon Arm (24, 25) 6 – P. emarginata H Thistlewood WCFF (26) 10 0/9 – H Thistlewood R. cerasi Germany Dossenheim (27) na P. avium H Vogt Austria Vienna (28) na P. avium C Stauffer

–, no data; na, not applicable. fruits. One population from Canada was collected on Pru- Erlangen, Germany). The extracted DNA was diluted to 1 nus emarginata Eaton (bitter cherry). The flies were 50 ng llÀ and stored at 20 °C until use. À brought to the laboratory, removed from the traps, and The 14 microsatellite loci used in this study (Maxwell stored in 70–90% ethanol at 20 °C until genetic analysis. et al., 2009) were amplified in four multiplex reactions À When possible, all flies were sexed and the wing pattern (primer mix A: WCFF-024, -111, -061B; primer mix B: (spot or line) was recorded (Table 1). The distance by air WCFF-031, -083, -048; primer mix C1: WCFF-084A, between western and eastern German populations was -086A, -049, -105; and primer mix C2: WCFF-007, -093, 250–530 km, between eastern Germany and Hungary ca. -011, -057). The PCR and multiplex conditions in the 850 km, and between western Germany and Hungary ca. present study differ from those of Maxwell et al. (2009); a 800 km. comparison of the methods is presented in Appendix S1. The PCR was carried out using a total volume of 10 ll DNA amplification procedures (5 ll29 QIAGEN Multiplex PCR Master Mix, 1 ll pri- DNA was extracted with the High Pure PCR Template mer mix, 2.5 ll RNAse-free water, and 1.5 ll DNA tem- Preparation Kit (Roche Diagnostics, Grenzach-Wyhlen, plate) and with a final primer concentration of 0.2 lM in Germany) following the manufacturer’s protocol. DNA all four multiplex sets. The cycling protocol followed the concentrations and purity were measured with a Nanodrop manufacturer’s standard with an annealing temperature of 1000 UV/VIS-Spectrometer (PEQLAB Biotechnologie, 60 °C for all four multiplex sets. An initial activation step 4 Johannesen et al.

(15 min at 95 °C) was followed by a 3-step cycling: dena- in diversity estimates between species-pairs were tested turation 30 s at 94 °C, annealing 90 s at 60 °C, extension across the 14 loci with paired t-tests. The estimate of spe- 60 s at 72 °C, with 30 iterations, and a final extension of cies-specific allelic richness was based on 22 individuals, 30 min at 60 °C. The fragment analysis was performed on the lowest number of individuals scored for any locus and a GA3130XL Genetic Analyzer (Applied Biosystems, species. Expected heterozygosity was based on the Darmstadt, Germany). For the analysis, 1 ll of the PCR unweighted mean across loci within each species. product was added to 11.7 ll of Hi-Di Formamide We used Bayesian analysis to investigate the number of (Applied Biosystems) and 0.3 ll of GeneScan 500 ROX- species, and for inferences regarding founder events and Standard (Applied Biosystems) in an ABgene 96well-plate multiple vs. single origin of American fruit fly in Europe. (Thermo Fisher Scientific, Waltham, MA, USA). The solu- In a first analysis including all individuals, we used the tion was denatured at 95 °C for 3 min using a BIOER program STRUCTURE 2.3.3 (Pritchard et al., 2000) as an Mixing Block MB-102 (BIOER Technology, Hangzhou, assignment test to test the hypothesis that the three cherry China). Genotypes were scored with GeneMapper v. 4.0 fruit fly species, R. cingulata, R. indifferens, and R. cerasi, (Applied Biosystems) using the automatic scoring func- are genetically different by fixing the number of genetic tion. All genotypes were checked and edited manually. clusters to K = 3. In this analysis, both European and ApartialsequenceofthemtDNAgenecytochromeoxi- reference populations of the American species were dase subunit 1 was amplified with primers C1-J-1751 included. The analysis was performed twice, with correlated (5′-GAG-CTC-CTG-ATA-TAG-CTT-TTC-C-3′)andC1-N- and uncorrelated allele frequencies. The analyses were done 2776 (5′-GGA-TAA-TCA-GAA-TAT-CGT-CGA-GG-3′) with all 14 loci, and for a data set excluding locus 007 (Hedin & Maddison, 2001) using the amplification proto- because it did not amplify in R. cerasi (see Results). In a sec- col therein. Purified PCR products were sequenced on an ond analysis, we used STRUCTURE to evaluate the number ABI 3130xl capillary sequencer. Sequences were aligned of genetic clusters of the American fruit fly species found in using the program Sequence Navigator (ABI, Applied Bio- Europe. In this analysis, we included both American and systems). Subsequently, all aligned sequences were checked European populations, but excluded putative hybrid indi- manually. viduals. Structure was run with a burn-in of 5 000 genera- tions and MCMC of 50 000 generations. Twenty runs Microsatellite genetic analysis were carried out for each K to quantify the variance of the Microsatellite analyses were based on 57 R. cingulata from likelihood estimate. The likelihood distribution was analy- Michigan, 75 R. indifferens from the western USA and sed by the method of Evanno et al. (2005) where the Canada, 319 American cherry fruit flies collected in wes- inferred K (except for K = 1, which cannot be detected by tern and eastern Germany and in Hungary, and 27 R. cer- this method) is a peak at the second rate of change of the asi collected in Germany and Austria. Microsatellite allele likelihood function with respect to K (DK). frequencies and genetic diversity indices were estimated as The genetic relationship among populations was calcu- expected heterozygosity (He) and allelic richness (AR) per lated with the UPGMA method and compared with the population or species with FSTAT v. 2.9.3.2 (Goudet, Bayesian analysis. In effect, Bayesian analysis is a phyloge- 1995). The frequency of null alleles per locus and popula- netic inference in the sense that the membership of popu- tion was calculated with GENEPOP v. 4.0 (Rousset, 2008). lations to successive genetic clusters, as K increases, Differences in mean genetic diversity among three Euro- correspond to significant branches between these clusters pean regions (West and East Germany, Hungary) and spe- in the phylogenetic tree. The difference and therefore the cies in North America were examined with ANOVA, and reason for using both methods concerns the way the compared between pairs of populations with Tukey HSD (same) data are computed. Corroboration of the methods tests (VassarStats: Website for Statistical Computation, therefore strengthens the generality of the results. http://vassarstats.net/). Genetic differentiation among populations, FST,was Microsatellite allele frequencies at the species level were estimated among populations within Europe and for estimated to determine species-specific and diagnostic European and ancestral American populations with alleles. Species-level allele frequencies were calculated by ARLEQUIN version 3.5 (Excoffier & Lischer, 2010). In including all individuals belonging to each species. We three hierarchical analyses of molecular genetic variance defined species-specific alleles as those alleles that were (AMOVA), we estimated the amount of differentiation present above 0.10 in the focal species and lacking in non- explained by mean differentiation among populations focal species. Diagnostic alleles were defined as alleles that within geographic regions, FSR, and that among geo- were present in non-focal species with a frequency of less graphic regions, FRT. The three hierarchical analyses were than 0.05 and above 0.20 in the focal species. Differences performed to evaluate differentiation (1) between western American cherry fruit fly in Europe 5 and eastern Germany, (2) between Germany and Hungary there was complete amplification failure at locus 007. In (European level), and (3) among Germany, Hungary, and R. cerasi, genotypes at locus 011 were generally either USA (global level). homozygous for allele 239 or were distributed around an allele length of 265. The genotype pattern in 011 R. cerasi mtDNA analyses might indicate sex-linkage, but unfortunately we do not We sequenced 800 bp of the COI gene from 43 R. indiffe- know the sex of the sampled individuals. Because rens and 33 R. cingulata from North America, 89 Ameri- genotypes of 011 (and all other loci) in R. cerasi were can cherry fruit flies collected in Europe, and five R. cerasi. reproducible, they were considered in the species-level The genealogy of mtDNA haplotypes was calculated from analysis. The mean frequency of null alleles per locus per the number of mutational differences and visualized with population in R. cingulata was equal in USA and Europe a minimum spanning network. Genetic diversity of (0.078) and similar to the mean frequency estimated mtDNA was estimated as the number of haplotypes and as in R. indifferens (0.081), and higher than that estimated nucleotide diversity per population (ARLEQUIN version for R. cerasi (0.016). The allele sizes in this study fall 3.5). Differences in mean mtDNA diversity among the within the range reported in Maxwell et al. (2009) three European regions and the species in North America (Appendix S1). were tested with t-tests. Different frequency distributions between two regions/species were tested with Fisher’s exact Species status and diversity – microsatellites tests using the Freeman-Halton extension for 2 9 3 tables The STRUCTURE analysis for species affiliation with fixed (VassarStats). In this test, the frequencies of the two most K = 3 correctly assigned all North American individuals common haplotypes were considered with the frequency of R. cingulata and R. indifferens, and European R. cerasi, of a third haplotype, which was the pooled frequency of all and assigned American cherry fruit flies in Europe to rare haplotypes. R. cingulata with high membership assignments of 0.99– 1.00 (Figure 1). (K = 3 was also inferred as the optimal Wing pattern analysis number of genetic clusters in a sequential test from K = 1– Analyses of wing patterns were based on 64 R. cingulata 12; results not shown). Exceptions to the high membership and 52 R. indifferens from North America and 572 Ameri- assignments were found in five German and two Hungar- can cherry fruit flies collected in Europe. The wing pattern ian (total = 7) R. cingulata and in one R. cerasi (light was quantified as whether the anterior apical crossband was grey/black bars in Figure 1). These eight individuals had unbroken (line pattern) or discontinuous (spot pattern). assignment probabilities intermediate between R. cingula- The spot-line ratio between regional populations and/or ta and R. cerasi, with assignment probabilities to R. cingu- species was tested with Fisher’s exact tests (VassarStats). lata ranging from 0.11 to 0.77. The divergence between R. cingulata and R. indifferens was confirmed in Bayesian analyses including these two species only (fixed K = 2, Results probability scores >0.99) (results not shown) and by UP- Cross-species amplification of microsatellite loci GMA analysis (Figure 2). Cross-species amplification of 14 microsatellite loci devel- Membership assignments above 0.99 were achieved oped for R. indifferens (Maxwell et al., 2009) was successful through many specific and diagnostic alleles: we found 10 for all loci in R. cingulata and for 13 loci in R. cerasi, where specific alleles at seven loci and one diagnostic allele at one

Germany Hungary USA USA/Canada D/A 1.00

0.80

0.60

0.40

0.20

0.00 1234567 8 9 10 11 12 13 14 15,16,17,18 19 20 21 22 23 26 27 28 24,25 R.cingulata R.indifferens R.cerasi

Figure 1 Bayesian STRUCTURE analysis based on 14 microsatellite loci for inference of individual affiliation to Rhagoletis cingulata (light grey), R. indifferens (dark grey), and R. cerasi (black). All individuals except seven R. cingulata and one R. cerasi were assigned to the predicted species with probabilities >0.99. The eight intermediate individuals (light grey/black bars) are putative hybrids between R. cingulata and R. cerasi; all had R. cerasi mtDNA. Population names are given in Table 1. 6 Johannesen et al.

Ammern FF-N C6 Mannheim Kleinfahrer C7 FF-Q C2 Dossenheim 90 Ingelheim Germany

Heidesheim-2 C4 C1 I3 I2 Heidesheim-1 Dresden Doelstadt C3 C5 CB USA Evans Figure 3 Minimum spanning network of nine mtDNA Erd Elvira haplotypes of the COI gene (800 bp) found in Rhagoletis 95 Veszprem Hungary cingulata (white), R. indifferens (black), or both species (grey). 73 Agard Soskut 100 Coloma 93 Species status and diversity – mtDNA R. cingulata Watervliet USA We observed seven haplotypes in R. cingulata and three in Laubach R. indifferens (GenBank accession numbers JX315331- Cle Elum JX315339). In contrast to microsatellite data, the mtDNA 100 Creston sequence genealogy did not separate R. cingulata and R. indifferens Kaysville WCFF R. indifferens into two distinct units. Both species shared haplotype C1, which was the dominant haplotype in both Figure 2 UPGMA phenogram with bootstrap scores (>70) based species in North America (Table 3, Figure 3). The second on 14 microsatellite loci showing monophyly of Rhagoletis most common R. cingulata haplotype, C2, was not cingulata and R. indifferens, the basal position of American observed in our R. indifferens populations, which were all, R. cingulata populations relative to European populations, and a except the population Cle Elum, mostly monomorphic for separate clade of Hungarian populations within Europe. C1. The population Cle Elum had the R. indifferens-spe- cific haplotypes I2 and I3, but lacked C1. Hence, in North locus in R. cingulata, 11 specific alleles at six loci and six America, the haplotype frequency distribution differed sig- diagnostic alleles at six loci in R. indifferens, and six spe- nificantly between the species; this was caused by (1) the cific alleles at four loci and two diagnostic alleles at one population Cle Elum (including Cle Elum: P<0.0001) and locus in R. cerasi (Appendix S2). In addition, complete (2) monomorphism of C1 (excluding Cle Elum: P<0.01) amplification failure at locus 007 in R. cerasi makes this (Freeman–Halton extension of Fisher’s exact tests). locus diagnostic for detecting hybrids or introgression in We observed a single haplotype in R. cerasi (accession presumed R. cerasi, i.e., amplification of the locus indi- number JX315340; identical to R. cerasi DQ006858.1; M cates that R. cingulata or R. indifferens alleles are present. Pfunder, B Frey & JE Frey, unpubl.). The haplotype Alternatively, lack of amplification at 007 in R. cingulata differed by 0.11 sequence divergence from the North or R. indifferens is only semi-diagnostic for detecting American species. introgression because 007 had an amplification failure in All the eight individuals with microsatellite membership 3% of individuals in these two species. assignments intermediate between R. cingulata and R. ce- Allelic richness was significantly higher in R. cingulata rasi (0.11–0.77) had the R. cerasi mtDNA haplotype. By than in R. indifferens (mean AR = 6.977 vs. 5.042; paired contrast, all 115 R. cingulata with microsatellite member- t-test: t = 2.4, d.f. = 13, P = 0.03), but the species did not ship assignments >0.99 had R. cingulata haplotypes. The differ in mean heterozygosity (He = 0.68 vs. 0.60; distribution of haplotypes differed significantly between t = 1.15, d.f. = 13, P = 0.27). Rhagoletis cerasi was signifi- R. cingulata of intermediate (0.11–0.77) and pure (>0.99) cantly less polymorphic at the loci investigated, with mean nuclear genetic status (Fisher’s exact test: P<0.0001). He = 0.346 (R. cingulata:t= 4.13; R. indifferens: t = 3.31, both d.f. = 13, P<0.01), and mean AR = 3.01 Population structure of Rhagoletis cingulata – microsatellites (R. cingulata:t= 5.23, d.f. = 13, P<0.001; R. indifferens: Three genetic clusters of R. cingulata were identified with t = 2.64, d.f. = 13, P = 0.02). the method of Evanno et al. (2005) (Appendix S3). The American cherry fruit fly in Europe 7 three clusters corresponded to populations in Germany, there was no significant differentiation between the East

Hungary, and USA, with mean membership probabilities and West German regions (FRT = 0.004, P = 0.098). to each cluster of 0.64, 0.67, and 0.76, respectively. Genetic differentiation among all populations in Europe STRUCTURE analyses of K 1 and K+1 were studied to was also significant (F = 0.040, P<0.001), of which 68% À ST give inferences of the phylogenetic relationships leading to of the variance was explained by differentiation between the optimal number of clusters. K = 2 indicated that Hun- German and Hungarian regional populations (FRT = garian and American populations (membership probabili- 0.0275, P<0.001). The global level of differentiation among ties to cluster 1 = 0.74 and 0.70) were more related to each European and American R. cingulata populations other than to German populations (membership probabil- remained significantly greater than zero, but including ity to cluster 2 = 0.76). For K = 4, the membership proba- American populations did not increase the differentiation bility of American populations to a separate cluster stayed estimate relative to that observed in Europe alone constant, 0.67, whereas the membership probability of the (FST = 0.040, P<0.001; FRT = 0.025, P<0.001). European populations collapsed to a maximum probabil- ity of 0.41–0.50 of belonging to one of the three remaining Population structure of Rhagoletis cingulata – mtDNA clusters. Phylogenetic inference with UPGMA corrobo- For the mtDNA data set, the frequency distribution of the rated the STRUCTURE analysis by showing a basal posi- two common haplotypes C1, C2, and the pooled haplo- tion of American populations and a separate regional set type C3–C7 differed significantly between R. cingulata in of populations in Hungary (Figure 2). USA and Germany (Freeman-Halton extension, Fisher Microsatellite genetic diversity per population in exact test: P<0.001), Hungary and Germany (P<0.001), R. cingulata differed significantly among the four regions but not between USA and Hungary (P = 0.08). The distri- of western Germany, eastern Germany, Hungary, and butions in western and eastern Germany did not differ sig-

USA (AR:F3,19 = 16.56, P<0.001; He:F3,19 = 8.75, nificantly (Fisher’s exact test: P = 0.82) (Table 3). The P<0.01). The difference among regions was caused by mean nucleotide diversity per population in R. cingulata greater diversity in the area of origin, Michigan, USA in USA (0.00102) was about twice the diversity found in

(AR = 3.90, He = 0.72), than in the invasive area (Euro- R. cingulata in Hungary (0.00054) and Germany pean regions: AR = 3.37–3.60; He = 0.64–0.69; Tukey (0.00065), although this difference was not statistically sig- HSD tests, see Table 2). In Europe, microsatellite genetic nificant due to the lack of variation in the American popu- diversity was slightly higher per population in Hungary lation Evans (Table 3). The higher diversity was caused by than in Germany, but only significantly so for allelic rich- more haplotypes in USA (n = 6) than in Hungary (n = 3) ness between Hungary and western Germany (Table 2). and Germany (n = 2). Genetic differentiation among all populations in Ger- many was low, but significant (FST = 0.016, P<0.001) but Population structure of Rhagoletis cingulata – wing patterns The spot/line ratio did not differ significantly between western (0.65) and eastern (0.66) German populations of Table 2 Mean microsatellite genetic diversity in Rhagoletis cingu- R. cingulata (Fishers exact test: P = 0.83). The ratio in lata in four geographic regions calculated across 14 microsatellite Germany was significantly lower than in the American loci, and Tukey HSD tests for differences between pairs of R. cingulata populations (0.85; P<0.001). In Hungary, the regions line pattern was the most frequent pattern; thus the ratio 0.27 differed significantly from that in Germany and USA Diversity (both P<0.001), but not from the ratio in investigated estimate Tukey HSD AR/He R. indifferens from western North America, 0.16 Eastern Western (P = 0.14) (Figure 4). Region n AR He Germany Germany Hungary USA 5 3.90 0.72 **/* **/* **/ns Eastern 6 3.53 0.69 ns/ns ns/ns Discussion Germany In this study, we genetically confirmed with microsatellites Western 5 3.37 0.66 */ns the presence of the eastern American cherry fruit fly, Germany R. cingulata, in Europe, as assumed by previous authors, Hungary 4 3.60 0.64 while finding no evidence for the presence of R. indiffe- n, number of populations per region; AR, allelic richness; He, rens. Our study successfully achieved cross-species amplifi- expected heterozygosity; ns, not significant. cation of R. indifferens microsatellite loci (Maxwell et al., *0.01

Table 3 Genetic diversity in populations of Rhagoletis cingulata and R. indifferens estimated for 14 microsatellite loci and 800 bp of the mtDNA gene COI. The putative R. cingulata-R. cerasi hybrids were excluded from the analyses

Microsatellites mtDNA

3 Species Country Population nHo He AR n C1 C2 C3 C4 C5 C6 C7 I2 I3 Nd (910À ) R. cingulata USA Coloma 9 0.56 0.71 3.81 8 1 4 2 1 1.875 Laubach 4 0.57 0.79 4.07 4 4 0 Evens 8 0.63 0.72 4.00 –– Watervliet 23 0.61 0.69 3.72 13 8 2 2 1 1.026 CB Setterbo 14 0.61 0.69 3.91 8 5 2 1 1.161 Hungary Erd-Elvira 29 0.54 0.69 3.67 8 4 4 0.714 Felo~ors€ 32 0.59 0.69 3.67 8 7 1 0.313 Agard 26 0.46 0.65 3.48 8 6 1 1 0.625 Sosk ut 8 0.55 0.71 3.57 5 4 1 0.500 Eastern Ammern 30 0.57 0.65 3.46 8 1 7 0.313 Germany Dollst€ €adt 10 0.50 0.65 3.44 6 4 2 0.833 Frankfurt- 27 0.55 0.65 3.53 4 2 2 0.833 OderN Frankfurt- 14 0.57 0.70 3.72 –– OderQ Dresden 8 0.54 0.66 3.38 7 2 5 0.595 Kleinfahrer 10 0.48 0.69 3.63 4 1 3 0.625 Western Dossenheim 17 0.60 0.66 3.55 –– Germany Mannheim 17 0.55 0.62 3.34 7 2 5 0.595 Ingelheim 30 0.54 0.63 3.33 11 4 7 0.636 Heidesheim-1 28 0.61 0.64 3.41 –– Heidesheim-2 13 0.49 0.63 3.24 6 3 3 0.750 R. indifferens USA Kaysville 19 0.57 0.59 3.02 8 8 0 Cle Elum 30 0.44 0.56 2.87 15 9 6 0.595 Canada Creston 11 0.47 0.54 3.02 11 11 0 Salmon Arm 6 0.61 0.62 3.19 5 4 1 0.500 WCFF 10 0.49 0.50 2.61 4 4 0 n, sample size; Ho, observed heterozygosity; He, expected heterozygosity; AR, allelic richness; Nd, nucleotide diversity. –, not tested.

n = 64 n = 346 n = 138 n = 88 n = 53 abbcc 100 90 80 70 Figure 4 Percentage flies showing line 60 (black) and spot (grey) wing patterns in 50 % four regional populations of Rhagoletis 40 cingulata (USA, East and West Germany, 30 Hungary) and in the sample of 20 R. indifferens. Different letters capping bars 10 indicate significantly different wing pattern 0 ratios (Fisher’s exact test: P<0.05). n, USA Germany East Germany West Hungary R. indifferens sample size. and diagnostic alleles and loci in each species (Appendix for hybridization in Germany and in Hungary between the S2), individuals of all three species were assigned and iden- invasive R. cingulata and native R. cerasi. Together with tified with probabilities above 99%. We also found signs reported cross-species amplification in R. cingulata and American cherry fruit fly in Europe 9

R. cerasi in an independent study (Drosopoulou et al., allozyme loci (Berlocher & Bush, 1982). However, as we 2011) including other loci, the multilocus microsatellite show in this study, founder effects led to frequency shifts genotypes combined with mtDNA sequencing provide in both genetic and morphological markers in invasive powerful tools for studying population structure and European populations. Could founder events also explain hybridization in the three species. the discrepancy in North America between mtDNA, on In the invasive range in Germany, R. cingulata displayed the one hand side, and microsatellite and wing patterns, relatively little structure, which agrees with the rapid range on the other, as species markers? The common occurrence expansion witnessed since 2004 (Lampe et al., 2005). The of C1 in both species may result either from unfinished similar frequency distributions of microsatellite alleles, lineage sorting in young species (e.g., Diegisser et al., 2004, mtDNA haplotypes and wing patterns in eastern and wes- 2006) or from range expansion of R. cingulata followed by tern German populations provide evidence for a single introgression into R. indifferens populations in the latter’s common introduction of these populations followed by native range. Although our data cannot decide between rapid range expansion that did not lead to a loss of genetic the alternatives, monomorphic mtDNA distributions variation after the initial founding event. In contrast, Hun- across species boundaries have been linked to introgressive garian and German populations likely have separate ori- hybridization and selective sweeps (Jiggins, 2003; Hurst & gins. Populations in the two countries differed in Jiggins, 2005; Gompert et al., 2008). The interpretation of frequencies of microsatellite alleles, mtDNA haplotypes, R. cingulata being misidentified in western USA and Can- and wing patterns. The distributions of these three mark- ada (Foote et al., 1993; CABI, 2012) might follow from ers are corroborated by Wolbachia infection rates, which R. cingulata expanding naturally in North America, and are similar in the two German regions, but different will give evidence to the invasive potential of this species. between Germany and Hungary (H Schuler, unpubl. The late flight period of R. cingulata has led to high lar- data). The incongruity between German and Hungarian val infestation levels in tart cherry orchards that are mostly populations in four traits indicates that the young unaffected by the European R. cerasi due to this species’ Hungarian population was not founded via stepping stone earlier flight period (Vogt et al., 2010). Potentially as wor- dispersal from the older German population. Under such rying as the presence of R. cingulata alone are signs of a scenario, we would also expect the Hungarian popula- hybridization between R. cingulata and R. cerasi. The tion to have less genetic variation. In fact, it had slightly seven potential hybrid R. cingulata and the single hybrid more. R. cerasi represented 2 and 3%, respectively, of the studied Both genetic markers linked populations in USA and European specimens. It is noteworthy that the frequency Hungary more together than USA and Germany. This sug- of hybrids in this study is based on a random sample, i.e., gests that either the Hungarian population has a more we did not specifically test individuals that were morpho- recent ancestry with the American population (Szeoke,€ logically aberrant. The level of introgression between 0.11 2006), or, alternatively, has experienced a less strong foun- and 0.77 implies that some hybrid individuals are F2 or der effect. There is, however, a lack of concordance in wing backcrosses and thus that hybridization between the two patterns between USA and Hungary. This might result species results in fertile F1 hybrids. The possibility of falsely from genetic drift in a polymorphic quantitative trait. identifying parental individuals as hybrids is unlikely Given this lack of trait concordance, we cannot determine because each species has a multilocus assignment of 0.99– whether the recent introduction of R. cingulata in Hun- 1.00. Although the CO1 sequence divergence of 0.11 might gary is caused by direct colonization from Northern Amer- argue against reproductive compatibility, similar CO1 ica or from a secondary source in Europe. divergence has been reported within species (Cognato, Our study highlights the problematic use of wing pat- 2006), as has a lack of complete mating isolation between tern for discriminating between the eastern and western allopatric species in Drosophila with similar high genetic American cherry fruit fly. In the European populations of divergence (Coyne & Orr, 1997) as between R. cingulata R. cingulata, the spot-line ratio has shifted towards that and R. cerasi (Berlocher & Bush, 1982). In particular, observed in R. indifferens, and was not statistically differ- reproductive isolation tends to be delayed between allopat- ent in Hungarian R. cingulata. These results may have ric species, as in R. cingulata and R. cerasi, relative to sym- implications for the distinction of R. cingulata and patric pairs (Coyne & Orr, 1997). Nevertheless, R. indifferens in the native range where the species’ spot- hybridization should be confirmed in controlled mating line ratios and geographical distributions are correlated. experiments. The species distinction is corroborated by highly differenti- Viable introgressed flies are potentially very worrying ated microsatellite loci, but not by mtDNA (Smith & Bush, because genetic exchange between the species might lead 1997; this study) and the species are little differentiated at to new pest dynamics. Dissemination of hybrid strains is 10 Johannesen et al. well known in bacteria and viruses (e.g., Smith et al., 2009; References Sheppard et al., 2011) and introgression has also been put van Aartsen B (2001) Rhagoletus indifferens, een nieuwe boorvlieg forward for specialization in populations (Scriber, voor de Nederlandse fauna (Diptera: Tephritidae). Neder- 2002). The potential of hybridization to facilitate new landse Faunistische Mededelingen 14: 19–22. dynamics in is profoundly witnessed in hybridiza- Berlocher SH & Bush GL (1982) An electrophoretic analysis of tion between the American species R. mendax and Rhagoletis (Diptera: Tephritidae) phylogeny. Systematic Zool- R. zephyria on the introduced plant Lonicera sp. (Schwarz ogy 31: 136–155. et al., 2005) where a new hybrid Rhagoletis species has Brasier CM (2001) Rapid evolution of introduced plant patho- evolved. For R. cerasi, hybridization with R. cingulata gens via interspecific hybridization. BioScience 51: 123–133. most certainly will include the exchange of alleles involved Bush GL (1966) The , cytology, and evolution of the in flight period. Temperature-dependent alleles for flight, genus Rhagoletis in North America (Diptera, Tephritidae). flight-related mating success, and/or eclosion time is com- Bulletin of the Museumof Comparative Zoology 134: 431–562. – monplace in insects and has a genetic basis (Watt, 1977; CABI (2012) Centre for Agricultural Bioscience International invasive species compendium. Available at http://www. Watt et al., 1985; Scriber & Ording, 2005; Feder et al., cabi.org/isc (acessed 25 June 2012) 2010). Introgression of ‘late flight alleles’ from R. cingulata Cognato AI (2006) Standard percent DNA sequence difference into R. cerasi could lead to infestation of tart cherry by the for insects does not predict species boundaries. Journal of Eco- latter species. Until now, the earlier flight season of R. cer- nomical Entomology 99: 1037–1045. asi relative to the phenology of suitable tart cherries for Coyne JA & Orr HA (1997) ‘Patterns of speciation in Drosophila’ oviposition has largely prevented infestation of tart cher- revisited. Evolution 51: 295–303. ries by this species. Diegisser T, Johannesen J, Lehr C & Seitz A (2004) Genetic and Teixeira et al. (2007) showed that the peak flight period morphological differentiation in Tephritis bardanae (Diptera: of R. cingulata in USA in unmanaged orchards (June– Tephritidae): evidence for host-race formation. Journal of Evo- July) was earlier than in managed orchards (June–August), lutionary Biology 17: 83–93. whereas the flight period in natural host populations of Diegisser T, Seitz A & Johannesen J (2006) Phylogeographic pat- terns of host-race evolution in Tephritis conura (Diptera, Teph- P. serotina extended the season (June–September). This ritidae). Molecular Ecology 15: 681–694. observation has two implications for hybridization Doody JS, Green B, Rhind D, Castellano CM, Sims R & Robinson between R. cingulata and R. cerasi in the European T (2009) Population-level declines in Australian predators populations. First, it agrees with where we found most caused by an invasive species. Conservation 12: 46–53. hybrids, namely in abandoned orchards around Ingelheim Drosopoulou E, Augustinos AA, Nakou I, Koppler€ K, Kounatidis and Heidesheim in western Germany. Second, because the I et al. (2011) Genetic and cytogenetic analysis of the American natural resource P. serotina is a common plant in many cherry fruit fly, Rhagoletis cingulata (Diptera: Tephritidae). cherry-producing regions of Europe, it may not only act as Genetica 139: 1449–1464. a reservoir for R. cingulata (Smit & Dijkstra, 2008; Vogt Elledge AE, Allen LR, Carlsson BL, Wilton AN & Leung LKP et al., 2010; this study) but also has the potential to (2008) An evaluation of genetic analyses, skull morphology and visual appearance for assessing dingo purity: implications generate contact between R. cingulata and R. cerasi; for dingo conservation. Wildlife Research 35: 812–820. although in recent fruit collections only infestations by Evanno G, Regnaut S & Goudet J (2005) Detecting the number of R. cingulata have been found in P. serotina (H Vogt, clusters of individuals using the software STRUCTURE: a sim- unpubl.; U Holz, pers. comm.; A Trapp, pers. comm.). ulation study. Molecular Ecology 14: 2611–2620. The sign of R. cerasi/R. cingulata hybrids warrants an Excoffier L & Lischer HEL (2010) Arlequin suite ver 3.5: A new extensive survey of the common host plant range of these series of programs to perform population genetics analyses two species. under Linux and Windows. Molecular Ecology Resources 10: 564–567. Feder JL, Powell THQ, Filchak K & Leung B (2010) The diapause Acknowledgements response of Rhagoletis pomonella to varying environmental We are grateful to LAF Teixeira, H Thistlewood, N Roth- conditions and its significance for geographic and host plant- related adaptation. Entomologia Experimentalis et Applicata well, D Alston, WL Yee, E Voigt, E Maring, U Holz, W 136: 31–44. Dahlbender, and G Hensel for collecting and sharing Foote RH, Blanc FL & Norrbom AL (1993) Handbook of the specimens for this study, and to D Klebsch, C Gross, and C Fruit Flies (Diptera: Tephritidae) of America North of Mexico. Sturzbecher€ for invaluable help in the laboratory. This Comstock, Ithaca, NY, USA. study was funded by Stiftung Rheinland-Pfalz fur€ Innova- Gompert Z, Forister ML, Fordyce JA & Nice CC (2008) Wide- tion, grant 0861. spread mito-nuclear discordance with evidence for introgres- American cherry fruit fly in Europe 11

sive hybridization and selective sweeps in Lycaeides. Molecular Schwarz D, Matta BM, Shakir-Botteri NL & McPheron BA Ecology 17: 5231–5244. (2005) Host shift to an invasive plant triggers rapid animal Goudet J (1995) FSTAT (Version 1.2): A computer program to hybrid speciation. Nature 436: 546–549. calculate F-statistics. Journal of Heredity 86: 485–486. Scriber JM (2002) Evolution of insect-plant relationships: chemi- Hedin MC & Maddison WP (2001) A combined molecular cal constraints, coadaptation, and concordance of insect/plant approach to phylogeny of the jumping spider subfamily Den- traits. Entomologia Experimentalis et Applicata 104: 217–235. dryphantinae (Araneae: Salticidae). Molecular Phylogenetics Scriber JM & Ording GJ (2005) Ecological speciation without and Evolution 18: 386–403. host plant specialization; possible origins of a recently Hurst GDD & Jiggins FM (2005) Problems with mitochondrial described cryptic Papilio species. Entomologia Experimentalis DNA as a marker in population, phylogeographic and phyloge- et Applicata 115: 247–263. netic studies: the effects of inherited symbionts. Proceedings of Sheppard SK, McCarthy ND, Jolley KA & Maiden MCJ the Royal Society of London B 272: 1525–1534. (2011) Introgression in the genus Campylobacter: generation Jiggins FM (2003) Male-killing Wolbachia and mitochondrial and spread of mosaic alleles. Microbiology-SGM 157: 1066– DNA: selective sweeps, hybrid introgression and parasite pop- 1074. ulation dynamics. Genetics 164: 5–12. Shine C (2007) Invasive species in an international context: IPPC, Koppler€ K (2008) Occurrence of non-diapausing pupae in Rha- CBD, European Strategy on invasive alien species and other goletis cerasi L. TEAM - Tephritid Workers of Europe and the legal instruments. EPPO Bulletin 37: 103–113. Middle East, Newsletter 6: 6–8. Smit JT & Dijkstra EGM (2008) De invasieve Oost-Amerikaanse Kozubikova E, Filipova L, Kozak P, Duris Z, Martin MP et al. kersenboorvlieg Rhagoletis cingulata in Nederland (Diptera: (2009) Prevalence of the crayfish plague pathogen Aphanomy- Tephritidae). Nederlandse Faunistische Mededelingen 28: ces astaci in invasive American crayfishes in the Czech Repub- 1–16. lic. Conservation Biology 23: 1204–1213. Smith JJ & Bush GL (1997) Phylogeny of the genus Rhagoletis Lampe I, Krauthausen HJ & Burghause F (2005) Introduction (Diptera: Tephritidae) inferred from DNA sequences of mito- and distribution of the American eastern cherry fruit fly (Rha- chondrial Cytochrome Oxidase II. Molecular Phylogenetics goletis cingulata) in the Rhine Valley, Germany. Plant Protec- and Evolution 1: 33–43. tion and Plant Health in Europe: Introduction and Spread of Smith GJD, Vijaykrishna D, Bahl J, Lycett SJ, Worobey M et al. Invasive Species (ed. by DV Alford & GF Backhaus), pp. 135– (2009) Origins and evolutionary genomics of the 2009 140. British Crop Protection Council, Alton, UK. swine-origin H1N1 influenza A epidemic. Nature 459: 1122– Luan JB, Xu J, Lin KK, Zalucki MP & Shu-Sheng L (2012) Species 1125. exclusion between an invasive and an indigenous whitefly on Szeoke€ K (2006) Announcement about the occurrence of the host plants with differential levels of suitability. Journal of Inte- Eastern American Cherry Fruit in Hungary. Nov€ enyvede- grative Agriculture 11: 215–224. lem 42: 470. (In Hungarian) Maxwell SA, Rasic G & Keyghobadi N (2009) Characterization of Teixeira LAF, Isaacs R & Gut LJ (2007) Habitat-specific flight microsatellite loci for the western cherry fruit fly, Rhagoletis period in the cherry fruit fly Rhagoletis cingulata (Loew) indifferens (Diptera: Tephritidae). Molecular Ecology (Diptera: Tephritidae). Environmental Entomology 36: 1339– Resources 9: 1025–1028. 1348. Merz B (1991) Rhagoletis completa Cresson und Rhagoletis indiffe- Vie J-C, Hilton-Taylor C & Stuart SN (2009) Wildlife in a Chang- rens Curran, zwei wirtschaftlich bedeutende Nordamerikanis- ing World – An Analysis of the 2008 IUCN Red List of Threa- chen Fruchtfliegen, neu fur€ Europa (Diptera: Tephritidae). tened Species. IUCN, Gland, Switzerland. Mitteilungen der Schweizerischen Entomologen Gesellschaft Vogt H (2007) Short information about an invasive Rhagoletis 64: 55–57. species in Germany. TEAM - Tephritid Workers of Europe and Nunney L, Yuan X, Bromley R, Hartung J, Montero-Astua M the Middle East, Newsletter 4: 6–7. et al. (2010) Population genomic analysis of a bacterial plant Vogt H, Koppler€ K, Dahlbender W & Hensel G (2010) pathogen: novel insight into the origin of Pierce’s disease of Observations of Rhagoletis cingulata, an invasive species from grapevine in the US. PLoS ONE 5: e15488. North America, on cherry in Germany. IOBC/WPRS Bulletin OEPP/EPPO (2007) EPPO Reporting Service 2007 No. 8 (148). 54: 273–277. OEPP/EPPO, Paris, France. Watt WB (1977) Adaptation at specific loci. I. Natural selection Pritchard J, Stephens M & Donnelly P (2000) Inference of popu- on phosphoglucose isomerase of Colias butterflies: biochemical lation structure using multilocus genotype data. Genetics 155: and population aspects. Genetics 87: 177–194. 945–959. Watt WA, Carter PA & Blower SM (1985) Adaptation at spe- Reitz SR & Trumble JT (2002) Competitive displacement among cific loci. IV. Differential mating success among glycolytic insects and arachnids. Annual Review of Entomology 47: 435– allozyme genotypes of Colias butterflies. Genetics 109: 157– 465. 175. Rousset F (2008) Genepop’007: a complete reimplementation of Zwolfer€ H (1974) Das Treffpunkt-Prinzip als Kommunikations- the Genepop software for Windows and Linux. Molecular strategie und Isolationsmechanismus bei Bohrfliegen (Diptera: Ecology Resources 8: 103–106. Trypetidae). Entomologica Germanica 1: 11–20. 12 Johannesen et al.

Supporting Information R. indifferens (USA, Canada) and R. cerasi (Germany, Austria) Additional Supporting Information may be found in the online version of this article: Appendix S3. Inference of the optimal number of Appendix S1. A comparison of microsatellite protocols genetic clusters (K = 3) for the combined American and and allele lengths European sample of Rhagoletis cingulata using the method of Evanno et al. (2005). The three clusters corresponded to Appendix S2. Allele frequencies in regional populations populations in Germany, Hungary, and USA of Rhagoletis cingulata in Germany, Hungary, and USA,