Ancient DNA re-opens the question of the phylogenetic position of the Sardinian pika Prolagus sardus (Wagner, 1829), an extinct lagomorph

Abstract

Palaeogenomics is contributing to refine our understanding of many major evolutionary events at an unprecedented resolution, with relevant impacts in several fields, including phylogenetics of extinct species. Few extant and extinct animal species from Mediterranean regions have been characterised at the DNA level thus far. The Sardinian pika, Prolagus sardus (Wagner, 1829), was an iconic lagomorph species that populated Sardinia and Corsica and became extinct during the Holocene. There is a certain scientific debate on the phylogenetic assignment of the extinct genus Prolagus to the family Ochotonidae (one of the only two extant families of the order Lagomorpha) or to a separated family Prolagidae, or to the subfamily Prolaginae within the family Ochotonidae. In this study, we successfully reconstructed a portion of the mitogenome of a Sardinian pika dated to the Neolithic period and recovered from the Cabaddaris cave, an archaeological site in Sardinia. Our calibrated phylogeny may support the hypothesis that the genus Prolagus is an independent sister group to the family Ochotonidae that diverged from the Ochotona genus lineage about 30 million years ago. These results may contribute to refine the phylogenetic interpretation of the morphological peculiarities of the Prolagus genus already described by palaeontological studies.

Figure 1

(a) Geographic location of the Cabaddaris Cave (Sardinia, Italy); (b) left hemimandible of Prolagus sardus (PR6 bone) used for DNA extraction and radiocarbon dating, from different perspectives (b₁: buccal, b₂: occlusal, b₃: lingual and b₄: occlusal view of the cheek teeth); (c) composite skeleton of P. sardus (Museo Sardo di Geologia e Paleontologia “D. Lovisato”, University of Cagliari); (d) palaeo-artistic reconstruction of P. sardus (drawing by D. Zoboli).

Figure 2

Prolagus sardus mtDNA consensus regions (inner circle) aligned against the Ochotona curzoniae complete mtDNA sequence. Only regions longer than 100 were reported here. Regions in red correspond to P. sardus mtDNA sequences used for phylogenetic analyses, instead blue regions were not considered. To simplify the graphical representation, tRNA features were not included in the reference sequence.

Figure 3

Maximum-likelihood trees obtained with: (a) ribosomal mtDNA contig sequences (16S + 12S); (b) D-loop and ribosomal contig sequences (D-loop + 16S + 12S). Node numbers represent confidence values. Branch lengths are proportional to the number of substitutions per site.

Figure 4

Divergence dates and 95% confidence intervals resulting from the analysis of the ~ 51 Ma rooted model (partial D-loop + partial 12S + partial 16S dataset) of the Bayesian relaxed molecular dating method implemented in BEAST2.