Identifying source populations for the reintroduction of the Eurasian beaver, Castor fiber L. 1758, into Britain: evidence from ancient DNA

Abstract

Establishing true phylogenetic relationships between populations is a critical consideration when sourcing individuals for translocation. This presents huge difficulties with threatened and endangered species that have become extirpated from large areas of their former range. We utilise ancient DNA (aDNA) to reconstruct the phylogenetic relationships of a keystone species which has become extinct in Britain, the Eurasian beaver Castor fiber. We sequenced seventeen 492 bp partial tRNAPro and control region sequences from Late Pleistocene and Holocene age beavers and included these in network, demographic and genealogy analyses. The mode of postglacial population expansion from refugia was investigated by employing tests of neutrality and a pairwise mismatch distribution analysis. We found evidence of a pre-Late Glacial Maximum ancestor for the Western C. fiber clade which experienced a rapid demographic expansion during the terminal Pleistocene to early Holocene period. Ancient British beavers were found to originate from the Western phylogroup but showed no phylogenetic affinity to any one modern relict population over another. Instead, we find that they formed part of a large, continuous, pan-Western European clade that harbored little internal substructure. Our study highlights the utility of aDNA in reconstructing population histories of extirpated species which has real-world implications for conservation planning.

Figure 1

Map of sampling localities for ancient Castor fiber individuals. The Western clade (pink) and Eastern clade (blue) have two Holocene contact zones; one in Eastern Europe and one in NW Scandinavia. A now extinct clade existed in the Danube Basin (green) until the late Holocene. Sampling localities for the British data are indicated with dotted square icons, all other points refer to those generated by Horn et al.. Only localities from where successful sequences were retrieved are shown. These represented 17 specimens from the SW and SE of England (SI Table 2). Approximate locations of extant relict populations (1 to 8) from where modern C. fiber haplotypes derive are shown in black. Map created using ArcMap v10.5, accessed at www.esri.com. Data was sourced from Natural Earth Data, www.naturalearthdata.com.

Figure 2

Neighbour-net network showing relationships among ancient British and modern and ancient beavers. Data is comprised of partial (492 bp) tRNAPro-Control regions mitochondrial haplotypes and comprise haplotypes generated here (prefixed ‘MM’), ancient haplotypes generated by Horn et al.^, and modern relict population haplotypes previously described by Ducroz et al., Durka et al. and Senn et al.. British data are clustered closely together within the Western ESU and show no level of genetic distinctiveness from other Western populations.

Figure 3

Phylogenetic relationships among ancient British and modern and ancient beavers. Shown is a maximum clade credibility (MCC) tree, generated in BEAST with partial (492 bp) tRNA-Pro/control region haplotypes, with British samples highlighted in the dashed box. Only node and branch posterior probability values above 0.5 are shown. The genealogy was generated using a Strict Molecular Clock with a fixed rate of nucleotide substitution set to 5.0 × 10⁻⁷ and a coalescent expansion model of demography. The tree shows the British samples group within the large Western clade with no posterior support for genetic distinctiveness.