Evolutionary history of Carnivora (Mammalia, Laurasiatheria) inferred from mitochondrial genomes

Abstract

The order Carnivora, which currently includes 296 species classified into 16 families, is distributed across all continents. The phylogeny and the timing of diversification of members of the order are still a matter of debate. Here, complete mitochondrial genomes were analysed to reconstruct the phylogenetic relationships and to estimate divergence times among species of Carnivora. We assembled 51 new mitogenomes from 13 families, and aligned them with available mitogenomes by selecting only those showing more than 1% of nucleotide divergence and excluding those suspected to be of low-quality or from misidentified taxa. Our final alignment included 220 taxa representing 2,442 mitogenomes. Our analyses led to a robust resolution of suprafamilial and intrafamilial relationships. We identified 21 fossil calibration points to estimate a molecular timescale for carnivorans. According to our divergence time estimates, crown carnivorans appeared during or just after the Early Eocene Climatic Optimum; all major groups of Caniformia (Cynoidea/Arctoidea; Ursidae; Musteloidea/Pinnipedia) diverged from each other during the Eocene, while all major groups of Feliformia (Nandiniidae; Feloidea; Viverroidea) diversified more recently during the Oligocene, with a basal divergence of Nandinia at the Eocene/Oligocene transition; intrafamilial divergences occurred during the Miocene, except for the Procyonidae, as Potos separated from other genera during the Oligocene.

Fig 1. Variation in base composition of the mitogenomes of Carnivora.

The PCG-mtDNA dataset was used to calculate the frequency of the four bases (A, C, G and T) at each of the three codon positions, and the 12 variables measured were then summarized by a principal component analysis (PCA). The main graph represents the individual factor map based on 220 taxa. The families of Carnivora are highlighted by different colours. The small circular graph at the top left represents the variables factor map.

Fig 2. Phylogeny of Carnivora based on mitogenomes.

The Bayesian tree was reconstructed using the mtDNA dataset (220 taxa and 14,892 bp) and GTR+I+G model. The two outgroup species are not shown. Species names follow the classification of the IUCN [1]; the taxa written in red highlight the taxonomic issues discussed in the main text. The accession numbers of the 42 mitogenomes of Carnivora specially sequenced for this study are indicated in red. The eight mitogenomes here assembled from SRA data are shown in green. The blue circle associated to Mustela nudipes MH464792 indicates that the mitogenome was originally misassigned to Viverra tangalunga. Fossil species are followed by the symbol “†”. For each terminal taxon, the number of similar mitogenome(s) found in GenBank (pairwise distance < 1%) is indicated after the accession number. Dash branches indicate nodes supported by posterior probability (PP) < 0.95. Black circles indicate nodes that are also monophyletic in the two following trees: SuperTRI bootstrap 50% majority-rule consensus tree; and Bayesian tree obtained from the analysis of the mtDNA-Tv dataset and JC69+I+G model. Grey circles show nodes that are not found to be monophyletic with one of the two methods detailed above. White circles indicate nodes that are not monophyletic in both mtDNA-Tv and SuperTRI bootstrap consensus trees. No information was provided for the nodes highly supported by the SuperTRI analyses, i.e. which were found monophyletic in all the 10 Bayesian trees reconstructed from the 10 half-overlapping sub-datasets of the mtDNA dataset. For nodes less supported by the SuperTRI analyses, the number of Bayesian trees (< 10) showing the nodes is indicated.

Fig 3. A molecular timescale for carnivoran evolution.

The estimates of divergence time were calculated under BEAST v.2.4.7 using the GTR+I+G model on the mtDNA dataset. The asterisks show the 21 fossil calibration points used for molecular estimation (see Table 1 for details). The chronogram, mean ages (values in black) and associated 95% confidence intervals (grey bars) were inferred using a uniform prior distribution for fossil calibration points (“U approach”). For comparison, the values in blue are mean ages estimated using a log normal prior distribution for fossil calibration points (“L approach”; see main text for details and discussion). Species names follow the classification of the IUCN [1]; the name of the taxa written in red have been changed following our results which suggest these taxonomic changes (see the discussion for details).

Fig 4. Comparison with published chronograms on Carnivora.

The mean divergence times were here estimated with two approaches for the prior distribution on the calibrated node ages: (1) a uniform distribution between maximum and minimum boundaries (“U approach”, blue histograms); and (2) a log-normal distribution (“L approach”, green histograms) (see main text for details). The results were compared with mean ages inferred in Eizirik et al. [9] (red histograms) and Nyakatura and Bininda-Emonds [10] (orange histograms).