bs_bs_banner
Biological Journal of the Linnean Society, 2012, 106, 893–909. With 3 figures
Genetic structure of Eurasian badgers Meles meles (Carnivora: Mustelidae) and the colonization history of Ireland DENISE B. O’MEARA1,2*†, CEIRIDWEN J. EDWARDS3,†‡, D. PADDY SLEEMAN1, TOM F. CROSS4, MARK J. STATHAM2,5, JAN R. MCDOWELL6, EILEEN DILLANE4, JAMIE P. COUGHLAN4, DAVID O’LEARY4, CATHERINE O’REILLY2, DANIEL G. BRADLEY3 and JENS CARLSSON4 1
School of Biological, Earth and Environmental Sciences, University College Cork, Cork, Ireland Department of Chemical and Life Sciences, Waterford Institute of Technology, Cork Road, Waterford, Ireland 3 Molecular Population Genetics, Smurfit Institute, Trinity College Dublin, Dublin 2, Ireland 4 Beaufort Population Genetics Group, Aquaculture and Fisheries Development Centre, School of Biological, Earth and Environmental Sciences, Distillery Fields, University College Cork, Cork, Ireland 5 VGL-Canid Diversity and Conservation Laboratory, University of California, Davis, CA 95616, USA 6 Virginia Institute of Marine Science, College of William and Mary, PO Box 1346, Gloucester Point, VA 23062, USA 2
Received 11 January 2012; revised 6 March 2012; accepted for publication 6 March 2012
bij_1927
893..909
The present study examined the contemporary genetic composition of the Eurasian badger, Meles meles, in Ireland, Britain and Western Europe, using six nuclear microsatellite loci and a 215-bp fragment of the mitochondrial DNA control region. Significant population structure was evident within Europe (global multilocus microsatellite FST = 0.205, P < 0.001; global mitochondrial control region FST = 0.399, P < 0.001). Microsatellite-based cluster analyses detected one population in Ireland, whereas badgers from Britain could be subdivided into several populations. Excluding the island populations of Ireland and Britain, badgers from Western Europe showed further structuring, with evidence of discrete Scandinavian, Central European, and Spanish populations. Mitochondrial DNA cluster analysis grouped the Irish population with Scandinavia and Spain, whereas the majority of British haplotypes grouped with those from Central Europe. The findings of the present study suggest that British and Irish badger populations colonized from different refugial areas, or that there were different waves of colonization from the source population. There are indications for the presence of an Atlantic fringe element, which has been seen in other Irish species. We discuss the results in light of the controversy about natural versus human-mediated introductions. © 2012 The Linnean Society of London, Biological Journal of the Linnean Society, 2012, 106, 893–909.
ADDITIONAL KEYWORDS: genetic structure – Irish – microsatellites – mitochondrial DNA – origins – phylogeography.
INTRODUCTION The last glacial maximum (LGM) ended around 15 000 years ago in Europe (Edwards & Brooks, *Corresponding author. E-mail:
[email protected] †These authors contributed equally to this work. ‡Present Address: Research Laboratory for Archaeology, Dyson Perrins Building, South Parks Road, Oxford OX1 3QY, England
2008). It is assumed that temperate European fauna and flora survived that period in refugial areas in southern and eastern regions, and that recolonization of Northern Europe occurred as the ice retreated (Hewitt, 1999). It has been suggested that Britain (England, Scotland, and Wales) was colonized after the Pleistocene glaciations via land connections with Central Europe, and that Ireland was subsequently colonized via one, or several, land
© 2012 The Linnean Society of London, Biological Journal of the Linnean Society, 2012, 106, 893–909
893
894
D. O’MEARA ET AL.
bridges from Britain (Scharff, 1899; Beirne, 1952; Mitchell & Ryan, 1997). However, there are a number of warm climate-adapted species found in Ireland and the Northern Iberian peninsula that are absent in Britain (‘Lusitanian distribution’), which have been used to suggest an alternative hypothesis by which certain Irish flora and fauna shared a common ice age refugium with Iberian species (Kerney & Cameron, 1979; Yalden, 1999) rather than British species. It is thought that much of Ireland’s Lusitanian fauna is a result of the accidental introduction of these species on boats sailing from Iberia to Ireland in the past. With the development of molecular tools (DNA markers), as well as advances in associated statistics, genetics has been used successfully to infer phylogeographical origins of many species. It has become clear that the recolonization of Europe from glacial refugia is complex and varies considerably among species (Taberlet et al., 1998), mainly as a result of biological differences such as life-history, migratory capacity, and the size of home ranges. Furthermore, it is evident that the colonization history of species influences their contemporary population structure. This is particularly evident in peninsular and island populations such as Ireland. There are three known native (i.e. present in Pleistocene and colonized naturally) terrestrial mammals currently present in Ireland: the Irish stoat (Mustela erminea hibernica Thomas & Barrett-Hamilton), the Irish hare (Lepus timidus hibernicus Bell), and the Irish otter (Lutra lutra roensis Ogilby). The sea level was substantially lower (by 120 m) during the LGM and the Irish landmass extended further south-west than at present (Hoarau et al., 2007; Provan & Bennett, 2008; Teacher, Garner & Nichols, 2009), and so it is possible that these species may have survived the arctic conditions during the LGM on land linked to Ireland that has subsequently been submerged (Fairley, 2001; Randi et al., 2003; Hamill, Doyle & Duke, 2006, 2007; Martínková, McDonald & Searle, 2007). There is also evidence of an Irish refugium, with wood ant (Formica lugubris Zetterstedt) and common frog (Rana temporaria L.) possibly surviving the LGM in Ireland (Mäki-Petäys & Breen, 2007; Teacher et al., 2009). The Eurasian badger (Meles meles L.) is a medium-sized, social carnivore that is locally abundant across Eurasia (Aulagnier et al., 2008). Badgers have attracted considerable scientific attention as reservoirs of the disease bovine tuberculosis (bTB), which can ‘spill-over’ to cattle (Fairley, 2001; Griffin et al., 2005; Aulagnier et al., 2008) and cause economic problems. Although there have been many studies of badgers, there is uncertainty not only about their post-Pleistocene colonization of Europe,
in general, but also of Britain and of Ireland in particular. Findings to date are equivocal. Historically, badgers have been frequently transported by humans. For example, during what might be called the ‘peak fox hunting period’ of British history (approximately 1850–1939), badgers were translocated to areas where there were no fox earths to dig setts that would suit foxes (Blakeborough & Pease, 1914; Wentworth Day, 1937). Also, in prehistoric times, there is evidence that many were taken off shore of British islands, along with other furbearing species that were transported as trophies or pelts during the Bronze Age (Fairnell & Barrett, 2007). Badgers are considered to have entered into Ireland later than other mammals because, although they are mentioned in the 7th/8th Century Irish Law tracts, remains have only be found in a secure context at medieval sites (McCormick, 1998). Because, historically, badgers have been hunted for their hair, leather, meat, and grease (Griffiths, 1993; Griffiths & Kryštufek, 1993; McCormick, 1998; Searle, 2008), some studies suggest that badgers may have been introduced to Ireland intentionally for use by humans for similar reasons (Lynch, 1996; Yalden, 2010). Although the status of the different badger species is now resolved (Del Cerro et al., 2010), the contemporary relationships within European populations of M. meles, and the phylogeographical origins of badgers, are still not clear. There have been recent genetic studies that have included Irish badger samples in the context of a wider Europe. A microsatellite-based study (Pope et al., 2006) demonstrated that English badgers have a similar degree of genetic diversity to those in the rest of Western Europe, whereas badgers from Ireland (N = 25), Scandinavia, and Scotland exhibit significantly lower levels of diversity. They used their results to argue that the Irish badger may be most closely related to English and possibly Scottish badgers. Mitochondrial (mt) work (Marmi et al., 2006), using only two Irish samples (from Ulster), found that Irish samples grouped with Scandinavia and Luxembourg. However, inferences about the origin of Irish badgers have been limited as a result of small Irish sample sizes. A better understanding of the genetic structure of European badgers, especially badgers from Ireland, would help elucidate their origin, as well as assist in a broader understanding of their ecology. The present study is unique in badger phylogeography research because it combines microsatellite DNA data and mitochondrial (mt)DNA control region sequence data to investigate the contemporary relationships between Irish, British, and Western European badgers, and to make inferences about the colonization history of Irish badgers.
© 2012 The Linnean Society of London, Biological Journal of the Linnean Society, 2012, 106, 893–909
GENETIC STRUCTURE OF EURASIAN BADGERS
MATERIAL AND METHODS MICROSATELLITE DNA DETAILS Badger tissue samples were obtained from three countries: Holland (N = 29), Ireland (N = 28), and Spain (N = 16). Dutch and Spanish badgers were road-kill individuals from a wide geographical range, whereas the Irish samples were collected from a Department of Agriculture bTB cull in County Kilkenny, as well as from road-kill from the wider Co. Kilkenny area. Data were included from a previously published study (Pope et al., 2006); Ireland (N = 19), Britain (N = 330), and across Europe (N = 107), including Austria, Italy, Luxembourg, Norway, Poland, Spain, Sweden, and Switzerland. Tissue samples (either muscle or liver) were preserved in 70% ethanol and total genomic DNA extracted by Chelex-100 resin (Bio-Rad) extraction, in accordance with the methods described by Dillane et al. (2008). Samples were screened for allelic variability at six microsatellite loci: Mel1 (Bijlsma et al., 2000), Mel110, Mel111, Mel113, Mel114, and Mel116 (Carpenter et al., 2003). Polymerase chain reaction (PCR) details are available on request from the authors. All amplified PCR products were run on 6% denaturing polyacrylamide gels using a LI-COR 4300 automated DNA sequencer (Licor) in accordance with the manufacturer’s instructions. To directly compare our results with previously published microsatellite data, eight samples from Pope et al. (2006) were amplified for cross-calibration at the six loci. Scoring was compared with reference samples from the present study. When there was a difference in the expected allele size for Pope et al. (2006), all samples from the present study were adjusted accordingly. The loci used in the present study were all dinucleotides with perfect allele repeat structure (i.e. no allele showed sizes that did not fit a dinucleotide pattern), and allelic size ranges were within a very tight range (with the maximum difference in size between any two alleles within a locus being 26 bp). Therefore, it was straightforward to unambiguously apply the cross calibration, and the combined data set of 529 individuals was used for joint analysis.
MICROSATELLITE
DATA ANALYSIS
MICROCHECKER, version 2.2.3 (van Oosterhout et al., 2004) was used to identify possible genotyping errors, including the presence of null alleles, large allele dropout, and scoring errors as a result of stutter peak (using default settings). Populations were divided geographically and groups with at least 18 individuals were described as a population and assessed for descriptive statistics (Table 1). A number
895
of areas analyzed by Pope et al. (2006; Austria, Italy, Poland, and Sweden) had too few individuals to describe as a population for this analysis. Observed (HO) and expected (HE) heterozygosities, FST and allelic richness per locus per sample (RS) were estimated sensu Weir & Cockerham’s (1984), whereas the unbiased estimator of Wrights’ F-statistic FST was implemented in MSANALYZER, version 4.05 (10 000 permutations; Dieringer & Schlötterer, 2003). Gametic phase linkage disequilibria by Fisher’s method (1000 dememorizations and 5000 iterations) and deviations from Hardy–Weinberg equilibrium were assessed (default settings, exact tests) using GENEPOP, version 4.0.10 (Rousset, 2009). GENEPOP was also used to test for allele frequency heterogeneity among samples (exact G-tests with 100 000 dememorizations, 1000 batches, and 50 000 iterations). The genetic relationships among populations were visualized in a Factorial Correspondence Analysis as implemented in GENETIX, version 4.05.2 (Belkhir et al., 2000). To further investigate population substructure within the dataset, the number of genetic clusters present was assessed using the Bayesian clustering algorithm implemented in STRUCTURE, version 2.3.1 (Pritchard, Stephens & Donnelly, 2000; Falush, Stephens & Pritchard, 2003). The entire dataset of 529 individuals was analyzed using the default settings with a burn-in period of 10 000, followed by 100 000 replicates with no prior population information. We compared the likelihood of K equals 1 to 20, with each K-value replicated ten times. The most likely K was assessed by implementing the DK method, as described by Evanno, Regnaut & Goudet (2005). The software CLUMPP, version 1.1.2 (Jakobsson & Rosenberg, 2007) was used to account for cluster label switching and to assign which clusters each run corresponded to (search options: greedy G′, using random input orders and 100 repeats). The STRUCTURE-defined clusters were used for subsequent analyses, excluding 33 individuals that had less than 50% assignment to any one cluster or were obviously misassigned (i.e. individuals that were sampled in one geographical region but assigned to a geographically remote region). BOTTLENECK, version 1.2.02 (Piry, Luikart & Cornuet, 1999) was used to evaluate whether any population had undergone a recent bottleneck. This was investigated using the two-phased model, which has been shown to be the best fit for most microsatellite data sets (Di Rienzo et al., 1994). Seventy percent of the mutations were assumed to follow a stepwise mutation model, whereas the remaining 30% followed the two phase model, using 1000 iterations. The probability of heterozygote excess in each population was evaluated with a two-tailed Wilcoxon
© 2012 The Linnean Society of London, Biological Journal of the Linnean Society, 2012, 106, 893–909
896
D. O’MEARA ET AL.
Table 1. Microsatellite summary statistics Locus Mean across loci
Mel1
Mel110
Mel111
Mel113
Mel114
Mel116
Ireland (Ir) N a Rs as HE HO HW
49 2 2.0 282–284 0.287 0.306 1.000
47 5 4.3 324–332 0.479 0.468 0.084
49 2 2.0 132–138 0.359 0.306 0.417
48 5 4.5 118–130 0.636 0.604 0.791
49 3 3.0 231–235 0.639 0.571 0.538
46 8 6.8 113–139 0.793 0.717 0.440
48.0 4.2 3.8
Scotland (Sc) N a Rs as HE HO HW
40 4 4.0 276–286 0.634 0.500 0.007
42 4 3.4 322–328 0.475 0.429 0.637
42 2 2.0 132–138 0.486 0.452 0.750
42 4 3.8 118–130 0.315 0.286 0.170
42 2 2.0 231–233 0.490 0.381 0.206
33 5 4.9 113–137 0.714 0.242 < 0.001
40.2 3.5 3.4
North England (NE) N a Rs as HE HO HW
140 5 4.1 276–284 0.634 0.479 < 0.001
135 6 5.3 324–334 0.697 0.630 0.178
140 3 2.7 130–138 0.526 0.407 0.097
140 6 5.6 118–130 0.791 0.729 0.021
137 3 3.0 231–235 0.597 0.438 0.001
126 8 7.6 113–137 0.841 0.476 < 0.001
136.3 5.2 4.7
North Wales (NW) N a Rs as HE HO HW
40 2 2.0 282–284 0.255 0.250 1.000
40 4 4.0 324–330 0.718 0.625 0.159
40 3 3.0 130–138 0.478 0.475 0.350
40 6 5.7 118–130 0.733 0.650 0.023
40 4 3.4 231–237 0.508 0.500 0.824
36 7 6.4 113–135 0.716 0.500 0.001
39.3 4.3 4.1
South Wales (SW) N a Rs as HE HO HW
54 4 3.0 276–284 0.233 0.259 1.000
54 4 4.0 324–330 0.743 0.815 0.490
54 2 2.0 132–138 0.475 0.667 0.005
54 5 4.3 120–130 0.652 0.630 0.368
54 4 3.2 231–237 0.345 0.370 0.618
43 6 5.1 113–131 0.598 0.512 0.015
52.2 4.2 3.6
Gloucester UK (Gl) N a Rs as HE HO HW
20 3 3.0 280–286 0.226 0.200 0.241
20 5 5.0 324–334 0.739 0.550 0.052
20 3 3.0 130–138 0.604 0.700 0.184
20 4 4.0 118–130 0.579 0.550 0.643
20 4 4.0 231–237 0.714 0.700 0.814
20 5 5.0 113–135 0.645 0.450 0.001
20.0 4.0 4.0
Essex UK (Es) N a
25 4
24 5
25 4
25 7
25 4
22 6
0.532 0.496
0.519 0.382
0.681 0.526
0.568 0.500
0.508 0.542
0.584 0.525
24.3 5.0
© 2012 The Linnean Society of London, Biological Journal of the Linnean Society, 2012, 106, 893–909
GENETIC STRUCTURE OF EURASIAN BADGERS
897
Table 1. Continued Locus Mean across loci
Mel1
Mel110
Mel111
Mel113
Mel114
Mel116
Rs as HE HO HW
3.8 280–286 0.286 0.200 0.054
4.6 324–334 0.639 0.792 0.138
4.0 130–140 0.662 0.360 0.002
6.2 118–130 0.616 0.600 0.475
3.6 231–237 0.220 0.240 1.000
6.0 113–133 0.814 0.591 0.016
Holland (Ho) N a Rs as HE HO HW
27 5 4.6 276–286 0.605 0.667 0.207
19 5 4.8 324–334 0.615 0.579 0.684
29 4 3.6 130–140 0.604 0.483 0.214
26 6 5.3 118–130 0.673 0.500 0.293
29 4 3.9 231–237 0.568 0.690 0.740
28 6 5.0 113–133 0.612 0.536 0.315
26.3 5.0 4.5
Luxembourg (Lu) N a Rs as HE HO HW
34 8 6.3 272–286 0.708 0.647 0.226
34 7 6.3 322–324 0.774 0.765 0.351
34 4 4.0 132–142 0.568 0.441 0.008
34 6 5.9 120–130 0.732 0.765 0.982
34 4 3.8 211–233 0.541 0.471 0.527
33 7 6.0 113–135 0.754 0.788 0.828
33.8 6.0 5.4
Switzerland (Swz) N a Rs as HE HO HW
18 7 6.9 272–286 0.769 0.722 0.513
17 6 6.0 324–334 0.697 0.706 0.505
18 4 3.9 132–142 0.446 0.556 0.802
18 4 4.0 120–130 0.591 0.556 0.074
17 4 4.0 211–233 0.382 0.353 0.065
18 5 5.0 113–135 0.696 0.667 0.495
17.7 5.0 5.0
Spain (Sp) N a Rs as HE HO HW
30 6 5.2 274–284 0.659 0.600 0.729
28 7 6.7 322–336 0.782 0.429 < 0.001
33 3 3.0 132–138 0.466 0.273 0.004
34 4 3.3 120–130 0.529 0.324 0.011
33 4 3.5 211–235 0.338 0.303 0.130
33 8 7.0 119–133 0.782 0.758 0.548
31.83
Norway (No) N a Rs as HE HO HW
20 2 2.0 268–276 0.375 0.300 0.541
20 5 5.0 320–332 0.721 0.700 0.844
20 4 4.0 134–140 0.446 0.450 0.413
20 3 3.0 124–130 0.649 0.650 0.616
20 5 5.0 211–235 0.715 0.650 0.451
20 4 4.0 119–129 0.645 0.450 0.094
20.0 3.8 3.8
4.7 0.539 0.464
0.613 0.576
0.680 0.646
0.597 0.593
4.76 0.593 0.448
0.592 0.533
Summary statistics for the six microsatellite loci among the geographically segregated badger populations. N, number of individuals; a, number of alleles; Rs, allelic richness per locus and sample; HO, observed heterozygosity, HE, expected heterozygosity; HW, probability values of concordance with Hardy–Weinberg expectations. Values in bold indicate P-values considered to be significant after false discovery rate correction for six multiple tests.
© 2012 The Linnean Society of London, Biological Journal of the Linnean Society, 2012, 106, 893–909
898
D. O’MEARA ET AL.
signed-rank test (a = 0.05). We also used the M-ratio test (calculated as M = k/r, where k is the number of alleles and r the overall range in fragment sizes) and tested for effective population size reductions (Garza & Williamson, 2001). M-values (10 000 replicates) were calculated across loci using the software M_P_Val (http://swfsc.noaa.gov/textblock.aspx? Division=FED&id=3298). To determine whether any of the loci showed evidence of being under selection, we searched for outlier loci by employing a simulation-based approach implemented in LOSITAN, version 2 (Antao et al., 2008). Simulations were employed using ‘neutral mean FST’ and ‘force mean FST’ with 10 000 simulations, and both the infinite allele and stepwise mutation model. If the distribution of the loci lay outside the expected range, then those loci might be considered to be influenced by balancing or directional selection. In addition, to determine whether families had been sampled during the tissue collection process, the presence of half-sibs and full-sibs within clusters was investigated by applying the maximum-likelihood method implemented in COLONY, version 2.0.0.1 (Jones & Wang, 2009). The analysis was based on the full likelihood method and the ‘long length of run’ and ‘high likelihood precision’ options. Effective population sizes were calculated using three different methods. LDNe (Waples & Do, 2006) was used to estimate the linkage disequilibriumbased estimator of effective population within each cluster. LDNe implements a bias correction for estimates of effective population size (Waples, 1998). The second method used an effective population size (Ne) estimator included in COLONY. Ne estimates were generated for each cluster and 95% confidence intervals (CIs) were obtained via bootstrapping. In addition, MIGRATE, version 3.2.6 (Beerli & Felsenstein, 2001; Beerli, 2010) was used to estimate the relative effective population size through the calculation of theta (Q); Q is equal to 4Nem, where Ne is the longterm (inbreeding) effective population size and M is the ratio of migration rate, m, to mutation rate, m. From the formula Q = 4Nem, the mutation rate will have a very large impact on Ne estimates and, because the mutation rates at the loci used in this study are not known, we used Q as a proxy for effective population size. We ran MIGRATE five times to generate replicates using ten short chains (sampling 10 000 trees) and three long chains (sampling 100 000 trees), with a burn-in period of 10 000 trees, using an adaptive heating scheme.
MITOCHONDRIAL DNA
DETAILS
We collected 84 tissue samples from Ireland (N = 53), England (N = 13), Wales (N = 10), Scotland (N = 6),
Luxembourg (N = 1), and Spain (N = 1) for mtDNA analysis. The Irish samples were from the following counties: Armagh (N = 1), Cork (three locations; N = 22), Carlow (N = 3), Derry (N = 1), Dublin (N = 9), Kilkenny (N = 6), Laois (N = 1), Waterford (N = 9), and Wexford (N = 1). Total genomic DNA was isolated using the QIAGEN dNeasy Tissue Kit, in accordance with the manufacturer’s instructions. A section of the mtDNA control region was amplified (PCR conditions will be supplied upon request) using the primers: LRCB1 5′-TGG TCT TGT AAA CCA AAA ATG G-3′ (Davison et al., 2001) and H16498M 5′-CCT GAA GTA AGA ACC AGA TG-3′ (Statham, Turner & O’Reilly, 2005) and sequenced by Macrogen Inc. (http://www. macrogen.com). In addition, 75 European sequences from Marmi et al. (2006) were included, from Austria, England, Finland, Germany, Ireland, Italy, Luxembourg, Holland Norway, Poland, Spain, Sweden, and Switzerland. This gave a total mitochondrial dataset of 159 European samples.
MITOCHONDRIAL
DATA ANALYSIS
To combine our new mtDNA data with the previously published dataset of Marmi et al. (2006), our 368-bp sequence was truncated to 215 bp. Sequences were aligned with MEGA, version 5.0 (Tamura et al., 2007). A median-joining network was constructed using the median algorithm of Bandelt, Forster & Röhl (1999) in NETWORK, version 4.6 (http://www.fluxusengineering.com). Simulation studies have demonstrated that this method provides reliable estimates of the true genealogy (Cassens, Mardulyn & Milinkovitch, 2005). In addition, we conducted Bayesian analyses using MrBayes, version 3.1.2 (Ronquist & Huelsenbeck, 2003). Individual haplotypes were included from the present study (H1–H19), with two haplotypes from the Japanese badger (Meles anakuma: AB538981 and AB538976; Tashima et al., 2011). We used the most appropriate substitution model (GTR+G), as determined using the Akaike information criterion in JMODELTEST, version 0.1.1 (Guindon & Gascuel, 2003; Posada, 2008). Ten million generations were run, with two runs of four independent chains (with one heated) running simultaneously, and sampling every thousand generations. We checked runs for convergence and stationarity, and an appropriate burn-in of 30% was removed. We then constructed a phylogenetic tree, and calculated Bayesian posterior probability (BPP) support values at the nodes, based on the remaining 14 000 trees. Pairwise FST values (a statistic that takes into account the divergence between haplotype sequences; Statham et al., 2012) were generated according to Slatkin (1995), using the ARLEQUIN, version 3.5 (Excoffier & Lischer, 2010) with 10 100 replicates.
© 2012 The Linnean Society of London, Biological Journal of the Linnean Society, 2012, 106, 893–909
GENETIC STRUCTURE OF EURASIAN BADGERS Population affinities were visualized using multidimensional scaling (Kruskal, 1964) as implemented in the SPSS, version 17.0 (SPSS Inc.). ARLEQUIN was also used to estimate mismatch distributions (SDD; 1000 replicates) to determine whether the data differ from a population expansion model. DNASP, version 5.10.01 (Librado & Rozas, 2009) was used to estimate haplotype diversity (h), nucleotide diversity (p), Harpending’s raggedness index (Hri), Fu’s Fs, and tau (t). Harpending’s raggedness index (Hri) tests whether the data deviates significantly from a population expansion model (Harpending, 1994). Fu’s Fs tests whether mutations are selectively neutral, although significant negative estimates also suggest population expansion (Fu, 1997). Tau (t) can be used as a relative measure of time since population expansion. The false discovery rate method (Benjamini & Yekutieli, 2001) was used to correct significance values for multiple tests. The clusters that had previously been identified in the STRUCTURE analysis were examined with respect to the mitochondrial data, resulting in a total of eight populations: Ireland, England (samples sizes did not permit further subdivision of the English samples), Wales, Scotland, Scandinavia (comprised of samples from Norway and southwest Sweden), Central Europe (Austria, Germany, Italy, Luxembourg, Holland, Poland and Switzerland), Spain, and Finland, the latter for which we did not have any microsatellite data. A spatial analysis of molecular variance was calculated using SAMOVA, version 1.0 (Dupanloup, Schneider & Excoffier, 2002) to define groups of populations that were geographically homogeneous and maximally differentiated. The method uses a simulated annealing procedure to maximize the proportion of total genetic variance as a result of differences between groups of populations. The analysis was run for K = 2 through K = 8, with 100 simulated annealings, where K refers to the number of groups.
RESULTS MICROSATELLITE
ANALYSIS
Genetic variability among sampling populations The number of alleles per locus within samples varied from two at loci Mel1, Mel111, and Mel114 to eight at Mel1 and Mel116 (Table 1). Low levels of allelic richness per locus and per sample (Rs = 2.0) were observed at Mel1 in Ireland, North Wales, and Norway; at Mel11 in Ireland, Scotland, and Switzerland; and at Mel114 in Scotland. The highest levels were 7.6 at Mel116 in North England (Table 1). Mean expected heterozygosities varied from 0.22 at Mel14 in Essex to 0.841 at Mel116 in North England, and mean observed heterozygosities ranged from 0.20 at
899
Mel1 in Gloucester and Essex to 0.82 at Mel110 in Switzerland (Table 1). Five loci showed deviations from Hardy–Weinberg expectations; Mel1: Scotland and North England; Mel110: Spain; Mel111: Essex, Luxembourg, and Spain; Mel114: North England; Mel116: Scotland, North England, North Wales, and Gloucester (Table 1). We continued to use all loci for further analyses because MICROCHECKER found no evidence to suggest systematic presence of null alleles, large allele dropout or stutter at any locus. In addition, a recent study suggests that null alleles (even at high frequency) have little effect on STRUCTURE-based assignment tests or on multilocus FST estimates (Carlsson, 2008). There was significant genetic structure among samples, with a global multilocus FST estimate of 0.205 (P < 0.001), and all global single locus FST estimates were significantly different than zero, with a range from 0.149 at locus Mel110 to 0.322 at locus Mel1. Pairwise multilocus FST ranged from 0.076 (between Luxembourg and Spain) to 0.360 between Essex and Norway (Table 2). The graphical relationship among identified clusters, based on a factorial correspondence analysis, is shown in the Supporting information (Fig. S1). The British samples largely clustered together, whereas the Irish, Norwegian, Spanish, and Central European samples showed further differentiation. STRUCTURE The entire data set was genetically subdivided using the method of Evanno et al. (2005). With the STRUCTURE data, K = 2 was found to be the most likely estimate of K, which separated Britain (by which is meant England, Scotland, and Wales) from Ireland and the rest of Europe. At K = 4, Ireland and the rest of Europe formed unique clusters and the British group was divided by into two groups: a Scottish/ English group and a Welsh/English group. By calculating the value of K with the largest posterior probability, we found support for finer-scale clustering with K = 9, corresponding more closely to individual sample sites. We were satisfied that STRUCTURE populations were correct because, when the runs were replicated ten times, the alpha and FST values remained stable. The replicate STRUCTURE runs were sorted in CLUMPP (Fig. 1A) and produced a mean symmetric similarity coefficient G′ of 0.986, which indicated that STRUCTURE was performing efficiently. Individuals were grouped largely by geographical area, with unique clusters found in Ireland, Central Europe, Spain, and Scandinavia (Norway and Sweden). However, a high degree of admixture was found within Britain (Fig. 1B). A widespread British cluster was found to be common within England and Wales, and unique clusters were also found in
© 2012 The Linnean Society of London, Biological Journal of the Linnean Society, 2012, 106, 893–909
900
D. O’MEARA ET AL.
Table 2. Microsatellite FST matrix Population
Ir
Sc
NE
NW
SW
Gl
Es
Ho
Lu
Swz
Sp
Ir Sc NE NW SW Gl Es Ho Lu Swz Sp No
0.323 0.141 0.307 0.296 0.303 0.338 0.255 0.223 0.205 0.273 0.354
0.141 0.221 0.245 0.192 0.168 0.281 0.217 0.176 0.221 0.351
0.086 0.125 0.084 0.131 0.113 0.108 0.112 0.154 0.210
0.121 0.102 0.076 0.268 0.194 0.273 0.223 0.318
0.178 0.093 0.245 0.146 0.267 0.218 0.338
0.101 0.246 0.194 0.230 0.250 0.346
0.264 0.154 0.266 0.208 0.360
0.083 0.086 0.170 0.240
0.223 0.076 0.194
0.176 0.304
0.278
No
Pairwise multilocus FST values for microsatellite data among badger populations. P-values were all < 0.001, and thus considered to be significant after false discovery rate correction for 36 multiple tests. Abbreviations are as given in Table 1.
Scotland, North England, North Wales, and South Wales, with admixture occurring in all locations. The clusters were further examined to explore and explain the observed results. They were not biased by closely-related individuals or an artefact of sampling only a few families because the COLONY analysis indicated that the largest full-sib family in any cluster comprised only four individuals. The BOTTLENECK analysis suggested that two of the clusters may have been affected by bottlenecks (Central Europe, P = 0.008; Scandinavia, P = 0.016; Wilcoxon one tailed test). However, none remained significant after corrections for multiple tests (false discovery rate correction for nine tests). There was no evidence that any population had undergone a bottleneck when analysing the data with M-ratio because all M-values were estimated to be greater than 1.000. LOSITAN did not indicate that any locus was under selection. Effective and relative effective population size and gene flow within clusters A total of 17 loci/population combinations showed significant linkage disequilibrium. However, only two loci (Mel110 and Mel111 in the Irish and Welsh samples) remained significantly linked after Sequential Bonferonni correction (corrected for 135 tests). The lack of consistent linkage across samples most likely indicates that the observed linkage is a result of gametic phase linkage disequilibrium rather than physical linkage. The effective population size as estimated by LDNe varied from 10.9 (95% CI = 3.1–42.8) in Scandinavia to 49.6 (95% CI = 21.6–613.2) in Iberia, with the lowest allowed frequency set at 1%. LDNe failed to estimate Ne for South Wales because
the value generated was negative. Upper 95% confidence level values also included infinity for South Wales, North Wales, and North England (Table 3). The effective population size as estimated by COLONY varied from 13 (95% CI = 7–30) in Scandinavia to 37 (95% CI = 24–62) in Central Europe (Table 3). MIGRATE estimated the mean (over five replicates) relative effective population size, Q, to range from 0.543 (95% CI = 0.453–0.633) in Ireland to 1.337 (95% CI = 0.949–1.726) in Central Europe (Table 3). The Ne estimates from both LDNe and COLONY all showed overlapping 95% confidence intervals for all clusters. However, Q estimates from the MIGRATE analysis (Table 3) showed that Central Europe had higher Q estimates (non-overlapping 95% CIs) than all other clusters, excluding Spain, North Wales/North England, and Britain. Similarly Spain and North Wales/North England showed higher Q than did both Scotland and Ireland. Widespread Britain also had a higher Q than South Wales, North England, Scotland, Ireland, and Scandinavia. The number of migrants per generation (from multiplying Q by M) showed no significant difference in pairwise directionality. However, there were several differences in the scale of gene flow among populations, which ranged from 0.148 (from Scandinavia to North England) to 1.858 (from North Wales to widespread Britain).
MITOCHONDRIAL
ANALYSIS
The mitochondrial analysis resulted in a total of 159 sequences, exhibiting 19 haplotypes over a 215-bp sequence. The 19 haplotypes (H1 to H19) have
© 2012 The Linnean Society of London, Biological Journal of the Linnean Society, 2012, 106, 893–909
Figure 1. Microsatellite STRUCTURE distribution map. A, CLUMPP graphical plot of K = 9 from ten replicate STRUCTURE runs. Each column represents one individual and each colour represents the genetic composition of that individual. B, Microsatellite distribution map, with clusters derived from STRUCTURE and mapped geographically. Pie charts do not correspond to sample size.
GENETIC STRUCTURE OF EURASIAN BADGERS
© 2012 The Linnean Society of London, Biological Journal of the Linnean Society, 2012, 106, 893–909
901
902
D. O’MEARA ET AL.
Table 3. Microsatellite-based population estimates
Cluster
LDNe Ne
Ir Br SW NW NE Sc Sca CE Sp
34.0 69.9 -103.9 372.0 88.1 31.3 10.9 68.3 49.6
(15.6–121.6) (38.6–177.2) (348.2 – infinite) (66.8 – infinite) (35.5 – infinite) (13.7–123.4) (3.1–42.8) (42.2–134.6) (21.6–613.2)
Colony Ne
Migrate Q
20 36 23 30 25 18 13 37 25
0.543 0.982 0.702 0.841 0.751 0.552 0.588 1.337 0.954
(12–39) (23–59) (14–42) (19–52) (15–45) (10–36) (7–30) (24–62) (15–46)
(0.453–0.633) (0.892–1.071) (0.572–0.833) (0.671–1.010) (0.642–0.861) (0.461–0.643) (0.387–0.790) (0.949–1.726) (0.776–1.131)
Effective and relative population size for the badger clusters as defined by STRUCTURE analysis. Results estimated by LDNe, COLONY and MIGRATE, with confidence levels given in parentheses. Additional abbreviations are from Tables 1 and 2; Br, Britain; CE, Central Europe.
been deposited in GenBank (Accession numbers: JQ734951–JQ734969). The median-joining network revealed that the Irish population contained four haplotypes (H2, H9, H14, and H15). Haplotype H9 was the geographically most widespread Irish haplotype, also occurring in Scandinavia and in low numbers in Scotland, Central Europe, and Spain. H2 was found in Ireland and in Central Europe. H14 was found in a single sample from the southwest of Ireland. In Britain, H1 and H4 were found to be the most commonly distributed haplotypes, occurring in England, Scotland, and Wales, although they were also common within Central Europe and one individual from Spain. Haplotypes H1, H2, H4, and H5 were found throughout Central Europe, whereas H3, H5, H6, H7, H8, H9, and H10 were less common (Fig. 2). Within Spain, H15 was shared with Ireland, and this haplotype was closely related to H9. Three unique haploypes were found in Finland (H17, H18, and H19). A multidimensional scaling plot (Fig. 3) also simplified the clustering of haplotypes by grouping Ireland, Spain and Scandinavia on one axis and Central Europe and British groups on the other axis. Bayesian phylogenetic analysis determined that the European samples in the present study belonged to a single main clade, separate from that of Japanese badgers (BPP 1.0; see Supporting information, Fig. S2). One subclade (BPP 0.68) consisted of three haplotypes (H6, H10, and H16) primarily found in Spain but extending northward into Central and Northern Europe. Similarly, there was support for the grouping of two haplotypes found in Finland (H18 and H19; BPP 0.68), and two haplotypes found in England (H11 and H12; BPP 0.56). This final grouping conflicts with the positioning of these two haplotypes in the network, albeit with low support, which
reflects the high level of homoplasy seen in the sequence data. The nucleotide diversity (p) ranged from 0.002 in Scandinavia to 0.013 in Finland, whereas haplotype diversity (h) ranged from 0.356 in Scandinavia to 1.000 in Finland (Table 4). The SAMOVA analysis maximizes the percentage of variation among groups at the same time as minimizing the amount of variation as a result of variance among samples within groups at any given K. This analysis indicated that the values were optimized when K = 4 (Table 5), which created four groups; the first of which included Central Europe and British samples; the second comprised Ireland, with Scandinavia; whereas Spain and Finland constituted two individual groups respectively. ARLEQUIN analyses (Table 6) indicated global FST at 0.40 (P < 0.001) and pairwise FST ranged from –0.01 (between Central Europe and Wales) to 0.77 (between Wales and Scandinavia). Populations within Britain were not significantly differentiated from each other and the British samples were not significantly differentiated from Central Europe (Table 6). Only the Central European population showed any evidence of rapid population expansion, as indicated by a significant negative Fu’s Fs estimate (Table 4). No sample showed significant mismatch (SDD) or significant raggedness indexes (Hri; Table 4). Tau (t; Table 4) ranged from 0.356 in Scandinavia to 2.667 in Finland.
DISCUSSION It is not currently known whether the badger colonized Ireland naturally or if it was anthropogenically introduced. With this in mind, we investigated the contemporary relationships between Irish, British, and Western European badgers with both nuclear and mtDNA markers. Our results indicated
© 2012 The Linnean Society of London, Biological Journal of the Linnean Society, 2012, 106, 893–909
GENETIC STRUCTURE OF EURASIAN BADGERS
903
Figure 2. Mitochondrial DNA median-joining network. Constructed from 215 bp of control region mitochondrial DNA from 159 badger sequences from across Europe. Circles represent sequence haplotypes numbered H1 to H19, with the area being proportional to the frequency of the haplotypes, with H19 being an example of sample size equal to 1. Branches between the haplotypes represent single nucleotide mutations.
Table 4. Mitochondrial DNA summary statistics Population
N
nh
h
p
t
*Fu’s Fs
Hri
SSD
Ir E W Sc Sca CE Sp Fi
55 23 9 6 10 39 13 3
4 4 3 3 2 10 5 3
0.483 0.597 0.417 0.733 0.356 0.808 0.821 1.000
0.004 0.004 0.003 0.004 0.002 0.007 0.008 0.013
0.275 0.767 0.429 0.867 0.356 1.541 1.667 2.667
0.344 -0.474 -0.532 -0.427 0.417 -3.697 -0.504 n/a
0.136 0.167 0.169 0.347 0.210 0.061 0.086 0.444
0.013 0.019 0.003 0.063 0.004 0.002 0.011 0.122
*Fu’s Fs was significantly negative. t and SDD were estimated under a demographic expansion model. Table of European badger mitochondrial DNA control region sequence variability, showing: number of individuals (N); number of haplotypes (nh); haplotype diversity (h); nucleotide diversity (p); tau (t); Fu’s Fs; Harpending’s raggedness index (Hri); and the sum of squared differences from mismatch analyses (SSD). Bold values indicate significant differences. E, England; Fi, Finland. © 2012 The Linnean Society of London, Biological Journal of the Linnean Society, 2012, 106, 893–909
904
D. O’MEARA ET AL.
Figure 3. Mitochondrial DNA multidimensional scaling (MDS) plot. A two-dimensional MDS plot, drawn using data from a 215-bp region of the control region, summarizing FST genetic distances among eight populations of badgers. The proportion of the data explaining the first two principal coordinates (i.e. the r2 value) is 0.984. Table 5. Mitochondrial groupings
DNA
SAMOVA
population
Groups (K)
Structure recovered
FCT
2 3 4 5 6 7 8
[Ir, E, W, Sc, Sca, CE, Sp] [Fi] [Ir, E, Sca, Sp] [W, Sc, CE] [Fi] [Ir, Sca] [E, W, Sc, CE] [Sp] [Fi] [Ir] [E, W, Sc, CE] [Sca] [Sp] [Fi] [Ir] [E, W, CE] [Sc] [Sca] [Sp] [Fi] [Ir] [E, W] [Sc] [Sca] [CE] [Sp] [Fi] [Ir] [E] [W] [Sc] [Sca] [CE] [Sp] [Fi]
0.400 0.409 0.444 0.434 0.422 0.418 0.400
SAMOVA mitochondrial DNA analytical results describing the structure of populations and the percentage of variation among groups, among populations within groups, and within populations.
genetic disparities between Ireland and Britain, suggesting a complicated colonization history for Irish badgers, which we discuss in light of these findings. The microsatellite analysis revealed strong phylogeographical structure of the badger within Europe, as signified by the unique badger populations within Ireland, Britain, Spain, Norway, and Central Europe. One highly differentiated population was evident within Ireland, whereas several populations were
present within Britain (although this may be partly a result of the larger sampling effort). It might be expected that island or peripheral populations, such as Ireland, Scotland, and Scandinavia, would exhibit genetic differentiation as a result of isolation, but the FST values were also high in populations within Britain, supporting previous genetic studies (Pope et al., 2006). Because of a lack of shared haplotypes, the mtDNA data revealed that Ireland supports a different population of badgers to that found in Britain. That is, no Irish haplotypes were shared with English or Welsh samples, despite both regions sharing haplotypes currently found in Central Europe and Spain. However, it must be noted that one haplotype, H2, which occurred in Ireland and Central Europe, was more closely related to the geographically widespread H1 haplotype, which included most of the British samples, than it was to other Irish haplotypes (H9, H14 and H15). All three analyses of mtDNA (FST, SAMOVA and the network) suggested that Irish and Scandinavian badgers belonged to lineages originating in Spain, suggesting elements of an Atlantic fringe. This has been found in other Irish mammals, such as the pygmy shrew (Searle et al., 2009; McDevitt et al., 2011), where replacement events may have occurred in the southern part of Britain as the first wave of colonizers moved north and into peripheral areas. In
© 2012 The Linnean Society of London, Biological Journal of the Linnean Society, 2012, 106, 893–909
905
GENETIC STRUCTURE OF EURASIAN BADGERS Table 6. Mitochondrial DNA FST matrix Population
Ir
Ir E W Sc Sca CE Sp Fi
0.531 0.572 0.377 0.119 0.404 0.327 0.718
E
W
Sc
Sca
CE
Sp
Fi
< 0.001
< 0.001 0.444
0.004 0.308 0.129
0.034 < 0.001 < 0.001 0.001
< 0.001 0.097 0.464 0.325 < 0.001
< 0.001 < 0.001 < 0.001 0.002 0.005 < 0.001
< 0.001 < 0.001 0.004 0.015 0.005 0.002 0.003
0.002 0.011 0.672 0.029 0.542 0.629
0.164 0.772 -0.007 0.516 0.619
0.590 0.011 0.382 0.495
0.469 0.378 0.755
0.378 0.411
0.504
Pairwise FST estimates in the lower diagonal and significance values in the upper diagonal among badger populations. P-values in bold were considered to be significant at the P = 0.05 level after false discovery rate correction for 28 multiple tests.
a similar manner, the single endemic mtDNA control region haplotype found in Irish pine martens is highly divergent from the single British haplotype, and belongs to a group predominantly found in southern Europe, including Spain (Davison et al., 2001). Although this suggests an anthropogenic origin of Irish martens, recent genetic analysis of Welsh museum specimens identified a Southern European haplotype, closely related to that found in Ireland (C. O’Reilly, unpubl. data; Mullins et al., 2010), and thus identifying a potential natural overland colonization source. The Irish stoat is also a good example of a colonization pattern that occurred in two waves, with one population colonizing Ireland naturally overland via Britain, and then a subsequent wave replacing the original colonizers in Britain after Ireland became isolated (Martínková et al., 2007). Previously, it had been suggested that badgers, similar to many mammals (Corbet, 1961), may have been introduced into Ireland via Britain (Pope et al., 2006; Yalden, 2010). There is evidence that some mammals were introduced during the Neolithic into Ireland (Mascheretti et al., 2003; Mattiangeli, McEvoy & Bradley, 2008; Edwards & Bradley, 2009), and it is possible that badgers were also introduced to Ireland during this period, possibly for their secondary products. If people had been actively moving animals between Spain, Ireland, and Scandinavia along the western areas of Europe, this would explain the Atlantic fringe element seen here. However, using two independent molecular marker sets, within the constraints of our sampling, we have not found evidence to suggest this. In addition, securely dated evidence for badgers in Ireland is absent until Medieval times (McCormick, 1998). From our results, it is highly unlikely that badgers were introduced into Ireland from Britain. Had the Irish badger population been translocated from a par-
ticular location, we would have expected the microsatellite and mitochondrial data to cluster with a putative source population or be composed of a mixture of shared clusters and haplotypes. We would also expect to see lower genetic diversity as a result of a founder effect, as well as a lower effective population size. We do not see this; our mitochondrial results indicate that the badgers that colonized Ireland were different from those that now occupy much of Britain. By contrast, there was no differentiation at the mtDNA level between British and Central European populations, indicating natural colonization of Britain. If animals were introduced to Ireland from Britain in the past, they have had no substantial impact on the gene pool of the Irish population. Because of its isolated position in the Atlantic, Ireland may be a reflection of the genetic legacy of the European badger, which has been replaced naturally in Central Europe and Britain, similar to the pattern seen in Irish stoat (Martínková et al., 2007). We predict that archaeological specimens, and possibly even additional British badger samples, may identify further individuals that contain the Spanish haplotype (H15; present in Ireland), which has possibly been displaced in England but may be residual in peripheral areas such as Scotland. Despite the uncertain colonization mechanism of the badger into contemporary Ireland, it is clear that the badger is now another species to add to the growing list of Irish mammals that exhibit genetic disparities from the mammals present in the closest geographical landmass of Britain. As Ireland’s geological history becomes clearer, the mechanisms for colonization events may be better understood, although we consider that there is sufficient evidence to suggest that the Irish badger may have been a natural colonizer in Ireland.
© 2012 The Linnean Society of London, Biological Journal of the Linnean Society, 2012, 106, 893–909
906
D. O’MEARA ET AL.
CONCLUSIONS The effective population size and genetic variability in Irish badgers were found to be comparable to British and Central European populations, although the population was genetically distinct from both regions. This difference is an important finding given the economic and ecological importance of badgers, and we recommend that future bTB culls in Ireland should carefully consider the genetic repercussions to Ireland’s unique fauna. We note that the differences detected in data gathered on badger ecology (e.g. diet; Cleary et al., 2009) and incidence of visits to farmyards (Sleeman, Davenport & Fitzgerald, 2008), as well as tuberculosis levels in Ireland and Britain, may be influenced by genetic components. The effects of host genotype to susceptibility to tuberculosis has been overlooked in cattle and deer until recently (Allen et al., 2010), and there is equally little known about this in badgers. Given the considerable genetic structuring revealed in the present study, there is potential to investigate such links in badgers further. In addition, more detailed studies are needed to determine whether the Irish badgers constitute one or more panmictic populations, and additional sampling within mainland Europe would allow further insight into the location of LGM refugial areas for badgers. In conclusion, we recommend that the genetic integrity of the Irish badger population be conserved to maintain the diversity of what is most likely a native species.
ACKNOWLEDGEMENTS We thank Lisa Pope, Allan McDevitt, and Andrew Byrne for constructive criticisms on earlier versions of this manuscript, and the three anonymous reviewers for their helpful comments. We also thank Kate Lynch (Trinity College Dublin) for help with mitochondrial laboratory work; Paulo Prodöhl for analysis advice; Lisa Pope, Alain Frantz and Terry Burke for access to previously published microsatellite and unpublished sequence material; and the following individuals for access to badger tissue samples: Sandrine Lesellier and Eamonn Gormley (University College Dublin); Donal Toolan (Regional Veterinary Laboratory, Kilkenny); Peter Turner (Waterford Institute of Technology); and Hugh Jansman. C.J.E. was supported by the Irish Research Council for Science Engineering and Technology Basic Research Grant Scheme (project numbers SC/1999/409 and SC/2002/ 510) and a Welcome Trust VIP Award (2010–2011). J.P.C. and J.C. acknowledge funding from the Beaufort Marine Research Award in Fish Population Genetics funded by the Irish Government under the Sea Change Programme.
REFERENCES Allen AR, Minozzi G, Glass EJ, Skuce RA, McDowell SWJ, Woolliams JA, Bishop SC. 2010. Bovine tuberculosis: the genetic basis of host susceptibility. Proceedings of the Royal Society of London Series B, Biological Sciences 277: 2737–2745. Antao T, Lopes A, Lopes RJ, Beja-Pereira A, Luikart G. 2008. LOSITAN: a workbench to detect molecular adaptation based on a FST-outlier method. BMC Bioinformatics 9: 323. Aulagnier S, Haffner P, Mitchell-Jones AJ, Moutou F, Zima J. 2008. Mammals of Europe North Africa and Middle East. Paris: Delachaux & Niestlé. Bandelt H-J, Forster P, Röhl A. 1999. Median-joining networks for inferring intraspecific phylogenies. Molecular Biology and Evolution 16: 37–48. Beerli P. 2010. MIGRATE: documentation and program, Version 3.0.3. Available at: http://evolution.genetics. washington.edu/lamarc.html. Beerli P, Felsenstein J. 2001. Maximum likelihood estimation of a migration matrix and effective population sizes in n subpopulations by using a coalescent approach. Proceedings of the National Academy of Sciences of the United States of America 98: 4563–4568. Beirne BP. 1952. The origin and history of the British fauna. London: Methuen. Belkhir K, Borsa P, Chikhi L, Raufaste N, Bonhomme F. 2000. Genetix v. 4.02, logiciel sous Windows pour la génétique des populations. Available from: Laboratoire Génome, Populations, Interactions, CNRS UMR 5000, Université de Montpellier II, Montpellier, France. Benjamini BY, Yekutieli D. 2001. The control of the false discovery rate in multiple testing under dependency. Annals of Statistics 29: 1165–1188. Bijlsma R, van de Vliet M, Pertoldi C, van Apeldoorn RC, van de Zande L. 2000. Microsatellite primers from the Eurasian badger, Meles meles. Molecular Ecology 9: 2216– 2217. Blakeborough JF, Pease AE. 1914. The life and habits of the badger. London: The Foxhound. Carlsson J. 2008. Effects of microsatellite null alleles on assignment testing. Journal of Heredity 99: 616–662. Carpenter PJ, Dawson DA, Greig C, Parham A, Cheeseman CL, Burke T. 2003. Isolation of 39 polymorphic microsatellite loci and the development of a fluorescently labelled marker set for the Eurasian badger (Meles meles) (Carnivora: Mustelidae). Molecular Ecology Notes 3: 610– 615. Cassens I, Mardulyn P, Milinkovitch MC. 2005. Evaluating intraspecific ‘network’ construction methods using simulated sequence data: do existing algorithms outperform the Global Maximum Parsimony approach? Systematic Biology 54: 363–372. Cleary GP, Corner LAL, O’Keeffe J, Marples N. 2009. The diet of badgers in the Republic of Ireland. Mammalian Biology 74: 438–447. Corbet GB. 1961. Origin of the British insular races of small
© 2012 The Linnean Society of London, Biological Journal of the Linnean Society, 2012, 106, 893–909
GENETIC STRUCTURE OF EURASIAN BADGERS mammals and of the ‘Lusitanian’ fauna. Nature 191: 1037– 1040. Davison A, Birks JD, Brookes RC, Messenger JE, Griffiths HI. 2001. Mitochondrial phylogeography and population history of pine martens (Martes martes) compared with polecats (Mustela putorious). Molecular Ecology 10: 2479–2488. Del Cerro I, Marmi J, Ferrando A, Chashcin P, Taberlet P, Bosch M. 2010. Nuclear and mitochondrial phylogenetics provide evidence for four species of Eurasian badgers (Carnivora). Zoologica Scripta 39: 415–425. Di Rienzo A, Peterson AC, Garza JC, Valdes AM, Slatkin M, Freimer NB. 1994. Mutational processes of simple sequence repeat loci in human populations. Proceedings of the National Academy of Science of the United States of America 91: 3166–3170. Dieringer D, Schlötterer C. 2003. Microsatellite analyser (MSA): a platform independent analysis tool for large microsatellite data sets. Molecular Ecology Notes 3: 167– 169. Dillane E, McGinnity P, Coughlan JP, Cross MC, De Eyto E, Kenchington E, Prodöhl P, Cross TF. 2008. Demographics and landscape features determine intrariver population structure in Atlantic salmon (Salmo salar L.): the case of the River Moy in Ireland. Molecular Ecology 17: 4786–4800. Dupanloup I, Schneider S, Excoffier L. 2002. A simulated annealing approach to define the genetic structure of populations. Molecular Ecology 11: 2571–2581. Edwards CJ, Bradley DG. 2009. Human colonisation routes and the origins of Irish mammals. In: McCartan S, Schulting R, Warren G, Woodman P, eds. Mesolithic Horizons: papers presented at the seventh International Conference on the Mesolithic in Europe, Belfast 2005. Oxford: Oxbow, 217–224. Edwards R, Brooks A. 2008. The island of Ireland: drowning the myth of an Irish land-bridge? In: Davenport J, Sleeman DP, Woodman P, eds. Mind the Gap: postglacial colonization of Ireland. Irish Naturalists’ Journal Special Supplement, 19–34. Evanno G, Regnaut S, Goudet J. 2005. Detecting the number of clusters of individuals using the software STRUCTURE: a simulation study. Molecular Ecology 14: 2611–2620. Excoffier L, Lischer HL. 2010. Arlequin version 3.5: a new series of programs to perform population genetic analyses under Linux and Windows. Molecular Ecology Resources 10: 564–567. Fairley JS. 2001. A basket of weasels. Belfast: James Fairley. Fairnell EH, Barrett JH. 2007. Fur-bearing species and Scottish islands. Journal of Archaeological Science 34: 463– 484. Falush D, Stephens M, Pritchard JK. 2003. Inference of population structure using multilocus genotype data: linked loci and correlated allele frequencies. Genetics 164: 1567– 1587. Fu Y-X. 1997. Statistical tests of neutrality against population growth, hitchhiking and background selection. Genetics 147: 915–925. Garza C, Williamson EG. 2001. Detection of reduction in
907
population size using data from microsatellite loci. Molecular Ecology 10: 305–318. Griffin JM, Williams DH, Kelly GE, Clegg TA, O’Boyle I, Collins JD, More SJ. 2005. The impact of badger removal on the control of tuberculosis in cattle herds in Ireland. Preventative Veterinary Medicine 67: 237–266. Griffiths HI. 1993. The Eurasian badger (Meles meles, L., 1758) as a commodity species. Journal of Zoology 230: 340–342. Griffiths HI, Kryštufek B. 1993. Hunting pressures and badgers (Meles meles): patterns and possible futures. Lutra 36: 49–61. Guindon S, Gascuel O. 2003. A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Systematic Biology 52: 696–704. Hamill RM, Doyle D, Duke EJ. 2006. Spatial patterns of genetic diversity across European subspecies of the mountain hare (Lepus timidus L). Heredity 97: 355–365. Hamill RM, Doyle D, Duke EJ. 2007. Microsatellite analysis of mountain hares (Lepus timidus hibernicus): low genetic differentiation and possible sex-bias in dispersal. Journal of Mammalogy 88: 784–792. Harpending RC. 1994. Signature of ancient population growth in a low-resolution mitochondrial DNA mismatch distribution. Human Biology 66: 591–600. Hewitt GM. 1999. Post-glacial recolonization of European biota. Biological Journal of the Linnean Society 68: 87– 112. Hoarau G, Coyer JA, Veldsink JH, Stam WT, Olsen JL. 2007. Glacial refugia and recolonisation pathways in the seaweed Fucus serratus. Molecular Ecology 16: 3606– 3616. Jakobsson M, Rosenberg NA. 2007. CLUMPP: CLUster Matching and Permutation Program. Available at: http:// rosenberglab.bioinformatics.med.umich.edu/clumpp.html. Jones OR, Wang J. 2009. COLONY: a program for parentage and sibship inference from multilocus genotype data. Molecular Ecology Resources 10: 551–555. Kerney MP, Cameron RAD. 1979. A field guide to the land snails of Britain and NW Europe. London: William Collins Sons & Co. Ltd. Kruskal JB. 1964. Multidimensional scaling by optimising goodness to fit to a nonmetric hypothesis. Pyschometrika 29: 1–27. Librado P, Rozas J. 2009. DnaSP v5: a software for comprehensive analysis of DNA polymorphism data. Bioinformatics 25: 1451–1452. Lynch JM. 1996. Postglacial colonization of Ireland by mustelids, with particular reference to the badger (Meles meles L. Journal of Biogeography 23: 179–185. Mäki-Petäys H, Breen J. 2007. Genetic vulnerability of a remnant ant population. Conservation Genetics 8: 427–435. Marmi J, Lopez- Giraldez F, Macdonald DW, Calafell F, Zholnerovskayas E Domingo-Roura X. 2006. Mitochondrial DNA reveals a strong phylogenetic structure in the badger across Eurasia. Molecular Ecology 15: 1007– 1020. Martínková N, McDonald RA, Searle JB. 2007. Stoats
© 2012 The Linnean Society of London, Biological Journal of the Linnean Society, 2012, 106, 893–909
908
D. O’MEARA ET AL.
(Mustela erminea) provide evidence of natural overland colonization of Ireland. Proceedings of the Royal Society of London Series B, Biological Sciences 274: 1387–1393. Mascheretti S, Rogatcheva MB, Gunduz I, Fredga K, Searle JB. 2003. How did pygmy shrews colonize Ireland? Clues from a phylogenetic analysis of mitochondrial cytochrome b sequences. Proceedings of the Royal Society of London Series B, Biological Sciences 270: 1593–1599. Mattiangeli V, McEvoy B, Bradley DG. 2008. Irish genetic affinities: the mosaic genome. In: Davenport J, Sleeman DP, Woodman P, eds. Mind the Gap: postglacial colonization of Ireland. Irish Naturalists’ Journal Special Supplement, 127–133. McCormick F. 1998. Early evidence for wild animals in Ireland. In: Benecke N, ed. The Holocene history of European vertebrate fauna. Berlin: Archäologie in Eurasien 6, 355–371. McDevitt AD, Vega R, Rambau RV, Yannic G, Herman JS, Hayden TJ, Searle JB. 2011. Colonization of Ireland – revisiting ‘the pygmy shrew syndrome’ using mitochondrial, Y chromosomal and microsatellite markers. Heredity 107: 548–557. Mitchell F, Ryan M. 1997. Reading the Irish landscape. Dublin: Town House. Mullins J, Statham MJ, Roche T, Turner PD, O’Reilly C. 2010. Remotely plucked hair genotyping: a reliable and non-invasive method for censusing pine marten (Martes martes, L. 1758) populations. European Journal of Wildlife Research 56: 443–453. van Oosterhout C, Hutchinson W, Willds D, Shipley P. 2004. MICROCHECKER: software for identifying and correcting genotyping errors in microsatellite data. Molecular Ecology Notes 4: 535–538. Piry S, Luikart G, Cornuet JM. 1999. Bottleneck: a computer program for detecting recent reductions in the effective population size using allele frequency data. Journal of Heredity 90: 502–503. Pope LC, Domingo-Roura X, Erven K, Burke T. 2006. Isolation by distance and gene flow in the Eurasian badger (Meles meles) at both a local and broad scale. Molecular Ecology 15: 371–386. Posada D. 2008. jModelTest: phylogenetic model averaging. Molecular Biology and Evolution 25: 1253–1256. Pritchard JK, Stephens P, Donnelly P. 2000. Inference of population structure using multilocus genotype data. Genetics 155: 945–959. Provan J, Bennett KD. 2008. Phylogeographic insights into cryptic glacial refugia. Trends in Ecology and Evolution 23: 564–571. Randi E, Davoli F, Pierpaoli M, Pertoldi C, Madsen AB, Loeschcke V. 2003. Genetic structure in otter (Lutra lutra) populations in Europe: implications for conservation. Animal Conservation 6: 93–100. Ronquist F, Huelsenbeck JP. 2003. MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19: 1572–1574. Rousset F. 2009. Genepop’007: a complete re-implementation of the Genepop software for Windows and Linux. Molecular Ecology Research 8: 103–106.
Scharff RF. 1899. The history of the European fauna. London: Scott. Searle JB. 2008. The colonisation of Ireland by mammals.In: Davenport J, Sleeman DP, Woodman P, eds. Mind the gap: postglacial colonization of Ireland. Irish Naturalists’ Journal Special Supplement, 109–115. Searle JB, Kotlík P, Rambau RV, Marková S, Herman JS, McDevitt AD. 2009. The Celtic fringe of Britain: insights from small mammal phylogeography. Proceedings of the Royal Society of London Series B, Biological Sciences 276: 4287–4294. Slatkin M. 1995. A measure of population subdivision based on microsatellite allele frequencies. Genetics 139: 457–462. Sleeman DP, Davenport J, Fitzgerald A. 2008. Incidence of visits by badgers to farmyards in Ireland. Veterinary Record 163: 724. Statham M, Turner PD, O’Reilly C. 2005. Use of PCR amplification and restriction enzyme digestion of mitochondrial D-loop for identification of mustelids in Ireland. Irish Naturalist’s Journal 28: 1–6. Statham MJ, Sacks BN, Aubry KB, Perrine JD, Wisely SM. 2012. The origin of recently established red fox populations in the United States: translocations or natural range expansions? Journal of Mammalogy 93: 52–65. Taberlet P, Fumagalli L, Wust-Saucy A-G, Cossons J-F. 1998. Comparative phylogeography and postglacial colonisation route in Europe. Molecular Ecology 7: 453–464. Tamura K, Dudley J, Nei M, Kumar S. 2007. MEGA4: Molecular Evolutionary Genetics Analysis (MEGA) software version 4.0. Molecular Biology and Evolution 24: 1596– 1599. Tashima S, Kaneko Y, Anezaki T, Baba M, Yachimori S, Abramov AV, Saveljev AP, Masuda R. 2011. Phylogeographic sympatry and isolation of the Eurasian badgers (Meles, Mustelidae, Carnivora): implications for an alternative analysis using maternally as well as paternally inherited genes. Zoological Science 28: 293–303. Teacher AGF, Garner TWJ, Nichols RA. 2009. European phylogeography of the common frog (Rana temporaria): routes of postglacial colonization into the British Isles, and evidence for an Irish glacial refugium. Heredity 102: 490– 496. Waples RS. 1998. Separating wheat from the chaff: patterns of genetic differentiation in high gene flow species. Journal of Heredity 89: 438–450. Waples RS, Do C. 2006. LDNe: a program for estimating effective population size from data on linkage disequilibrium. Molecular Ecology Research 8: 753–756. Weir BS, Cockerham C. 1984. Estimating F-statistics for the analysis of population structure. Evolution 38: 1358– 1370. Wentworth Day J. 1937. Sporting adventures. London: George and Harrap. Yalden DW. 1999. The history of British mammals. London: Poyser. Yalden DW. 2010. Conclusion. In: O’Connor T, Sykes N, eds. Extinctions and invasions. Oxford: Windgather Press, 190– 196.
© 2012 The Linnean Society of London, Biological Journal of the Linnean Society, 2012, 106, 893–909
GENETIC STRUCTURE OF EURASIAN BADGERS
909
SUPPORTING INFORMATION Additional Supporting Information may be found in the online version of this article: Figure S1. Microsatellite factorial correspondence analysis (FCA) plot. Factorial correspondence plot illustrating the genetic structure observed based on microsatellite variability in European badgers. Abbreviations within clusters: No, Norway; CE, Central Europe; Sp, Spain; Br, Britain (England, Scotland and Wales); and Ir, Ireland. The other coloured abbreviations: Ne, North England; Gl, Gloucester; Es, Essex; Ho, Holland; Lu, Luxembourg; Swz, Switzerland; and W, Wales. Figure S2. Mitochondrial DNA 50% majority rule consensus tree. Constructed from 215 bp of the control region mtDNA of the 19 European badger haplotypes. Meles anakuma are Japanese badgers (Tashima et al., 2011) used to root the tree. Values at the nodes are Bayesian Posterior Probabilities (see main text). Please note: Wiley-Blackwell are not responsible for the content or functionality of any supporting materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.
© 2012 The Linnean Society of London, Biological Journal of the Linnean Society, 2012, 106, 893–909