Genetic diversity and landscape genetic structure of otter (Lutra lutra) populations in Europe
|
|  Yüklə 1.4 Mb.
|
H11 H9 H5 H13 H14 H18 H2 H20 one distinct haplotype every 4.75 individuals. Overall haplotype diversity was high (0.79 ± 0.037, SD), but nucleotide diversity (0.0014 ± 0.00012) and average number of pairwise differences (2.25) were low, suggesting that extant otter mtDNA lineages originated recently. The network was star-shaped and did not show any obvious signal of phylogeographic structuring (Fig. 2), with a H1 H6 H7 H8 H10 H12 H3 H15 H16 H4 H17 H19 central haplotype (H3) that is widespread in all European countries, with the exception of populations from southern Italy, which showed only haplotype H10 (Table 2). The additional shorter sequences (from 15 faecal samples), confirmed that only H10 is present in the southern Italian otters. The widespread haplotype H3 corresponds to H1 and Lut1, as reported respectively by Ferrando et al. (2004) and Stanton et al. (2009). Haplotypes H8 and H14 were Fig. 2 Minimum spanning network connecting the otter mtDNA haplotypes included in this study (n = 95). Each circle represents a single haplotype and the area of the circle is proportional to its frequency. The length of the lines are proportional to the number of mutations (short segments = one mutation; long segments = two mutations). The central haplotype (H3) is widespread in all Europe (except in southern Italy). Other haplotypes, differing in one or two mutations respect to the H3 haplotype, are distributed in the European countries. H1, H2 (Denmark); H5 (Germany, Hungary); H17 (Hungary); H19, H20 (Austria); H8, H9, H12, H18 (Germany); H7 (Ireland); H4 (UK); H13 (Spain and Portugal); H10 (Italy); H11 (Finland, Norway); H6, H16 (Sweden); H14, H15 (Finland) found only in Germany and in Finland with frequencies of 67 and 59%, respectively. Most of the haplotypes differed from H3 by one mutation (Fig. 2). The unimodal mismatch distribution indicated a sudden demographic expansion at Tau = 2.5 (90% confidence interval = 1.1–3.7). The rag- gedness index was R = 0.382; the probabilities to obtain by simulations SSD and R values[than the observed were = 0.001. Tajima’s D = -2.187 was significantly negative (P \ 0.01), consistent with a signal of demographic expan- sion. Assuming a star-shaped genealogy and that the
Table 2 Distribution of the otter mtDNA haplotypes in the sampled countries: P Portugal, SP Spain, FR France, UK United Kingdom, IR Ireland, DK Denmark, D Germany, A Austria, SK Slovakia, HU Hungary, B Belarus, FIN Finland, SW Sweden, NW Norway, IT Italy  Haplotype P SP FR UK IR DK D A SK HU B FIN SW NW IT Total
otter mtDNA evolved at the ‘‘standard’’ mammalian mtDNA rate of ca. 2% sequence divergence/million years (Avise 1986), the estimated average number of pairwise differences (2.5/1580 bp) could have been generated in ca. 100,000– 150,000 years of evolution from a common ancestral mtDNA genome. This date is congruent with a simple phylogeo- graphic model assuming that all extant European otter pop- ulations derive from the postglacial expansion of a single small population that survived isolated in a refuge area during the last glaciation.
11 microsatellites in 616 otter samples are shown in Table 3. All loci were polymorphic in all sampled popu- lations. The average allele number per locus was Ao = 4.9, and ranged from four (Czech Republic) to seven (Norway) alleles, with the exception of populations sampled in southern Italy and Denmark, which showed only 2.5 and 2.9 alleles per locus, respectively. The average effective allele number per locus was Ae = 2.8, ranging from the lowest values in Denmark and Italy (Ae = 1.6) to the highest value in Sweden (Ae = 3.7). Observed and expected heterozygosity was moderate (Ho = 0.50; He = 0.58), on average across populations. Observed Table 3 Genetic diversity in sampling groups as estimated at 11 microsatellite loci
heterozygosity was lowest in samples from Denmark and Italy (Ho = 0.35 and 0.37, respectively), and highest in some northern populations (i.e. otters from Latvia, Belarus, Finland and Sweden, showing Ho [ 0.65). Values of He were always greater than values of Ho, with the exception of England, Serbia-Montenegro, and Latvia-Belarus where He \ Ho, and Slovakia and Italy where Ho and He did not differ. Thus, the FIS values were significantly greater than zero in 10 of 17 locations, indicating widespread signifi- cant departures from HWE in those sampling groups, while samples from isolated populations (i.e. southern Italy, Slovakia, Czech Republic, Ireland) were in HWE (Table 3). Therefore, otters from the sampled geographical locations showed moderate levels of microsatellite allelic diversity, which was particularly low in some isolated groups, such as otters in Denmark and southern Italy. Moreover, most of the sampling groups assembled on the base of their country of origin, were not in HWE, showing less than expected heterozygotes, which possibly indicates a Wahlund effect (Wahlund 1928). Global genetic population structure The FCA plot of multilocus genotypes showed limited geographical differentiation among most of the sampled otter populations (Fig. 3a). The FCA did not provide evi- dence of clearly distinct clusters, although the individuals sampled from the various populations were not evenly distributed in the multivariate space, but showed a sharp tendency to clump according to their geographical origins. Along the first FCA component (that described 3.76% of the total variance) all Iberian otters were located towards the upper right corner of the plot, while most of the otters sampled from Germany were located towards the upper left corner. Along the second component (explaining 2.72% of the total variance) a group of the samples from Norway (the southern Norwegian samples) were distributed separately in the lower region of the FCA (Fig. 3a). Along the third component (explaining 2.70% of the total variance) all the Italian otters clustered in the lower region of the plot, sep- arately from all the other samples (Fig. 3b). The PCA performed with PCA-GEN confirmed these results, showing that populations from Iberia, Italy and Germany clustered separately from all the others (Fig. 3c). In this analysis there were two significant principal components, which explained 26.69% (PC I) and 19.50% (PC II) of the total inertia. This means that the populations separated along the first and second PCA components respectively showed FST = 0.034 and FST = 0.024, corresponding to 27 and 19% of the total n sample size, Ao average number of alleles per locus, Ae average number of effective alleles per locus, He expected heterozygosity, Ho observed heterozygosity, FIS fixation index; significance of the FIS probability value (P-value); n.s. not significant; * P \ 0.05, ** P \ 0.05 FST value of 0.126. The same results were obtained by an FCA on populations computed using GENETIX (not shown). An analysis of molecular variance (AMOVA), calcu- lated in GENALEX, showed that 17% of the total genetic |
