Fossil Hunters

Testing the Hypothesis Datasets

In order to test our hypothesis, we reanalyzed the population genetic data by Assmann et al. (1994), Assmann and Weber (1997), and Reimann et al. (2002). Allozyme polymorphisms at five loci (Table 1) were obtained by electrophoretic separation using acrylamide gels. Standard methods (as described in Assmann and Weber 1997) and consistent naming of the alleles (Reimann et al. 2002) allowed data compilation and reanalysis. Only samples with a minimum sample size of 15 individuals were used. We used two different datasets (Table 1). Dataset A contains

Table 1 Population genetic datasets used for the reanalysis to localize the glacial refuge areas of C. auronitens

Sample No.

no. in this No. of alleles Sources (sample numbers

Table 1 Population genetic datasets used for the reanalysis to localize the glacial refuge areas of C. auronitens

Sample No.

no. in this No. of alleles Sources (sample numbers

a Data for polymorphisms at the ME locus partly unpublished: Samples 1-29 are monomorphic for one allele, samples 30-61 monomorphic for another allele a Data for polymorphisms at the ME locus partly unpublished: Samples 1-29 are monomorphic for one allele, samples 30-61 monomorphic for another allele

Fig. 2 Sample sites covering the Western part of the distribution area of C. auronitens. Renumbered after Assmann et al. (1994), Assmann and Weber (1997) and Reimann et al. (2002), see Table 1

information on a total of 1,771 individuals from 61 sample sites in the Western and Southwestern part of the distribution range of C. auronitens (including the Pyrenees) at five loci, whereas dataset B contains information from 48 samples with a total of 1,517 individuals at four loci. Dataset B comprises additional samples, mainly from Western France but lacks the samples from the Pyrenees used in dataset A. Sample sites were renumbered and are shown in Fig. 2.

3.2.2 Statistical Analyses

Nei's standard genetic distances (Nei 1978) were calculated from allele frequencies and phenograms (using neighbor-joining, Saitou and Nei 1987) were constructed (with PHYLIP, Felsenstein 2005) for both datasets. Principal component analyses (PCA) based on the arc-sin transformed allele frequencies were performed for both datasets using STATISTICA, version 7.1 (StatSoft Inc.). For the latter analyses rare alleles were omitted, thus only alleles with a minimum mean frequency of 10% were used. Private alleles (defined as alleles found in one or two neighboring sample sites only) were counted per sample.

Was this article helpful?