Quantitative Phylogenomics of Within-Species Mitogenome Variation: Monte Carlo and Non-Parametric Analysis of Phylogeographic Structure among Discrete Transatlantic Breeding Areas of Harp Seals (Pagophilus groenlandicus)

• Published in: PloS one 2015-08, Vol.10 (8), p.e0134207-e0134207
• Main Authors: Carr, Steven M, Duggan, Ana T, Stenson, Garry B, Marshall, H Dawn
• Format: Article
• Language: English
• Subjects: Age composition; Analysis; Animals; Bayesian analysis; Biology; Breeding; Climate change; Computer simulation; Deoxyribonucleic acid; Dispersal; DNA; DNA, Mitochondrial - genetics; Evolution; Fast ice; Gene sequencing; Genetic analysis; Genetic aspects; Genome - genetics; Genomes; Genomics; Glaciation; Glaciology; Haplotypes; Homogeneous structure; Human populations; Hypotheses; Ice; Ice front; Laboratories; Marine mammals; Mathematical models; Mitochondrial DNA; Molecular Biology; Monte Carlo Method; Monte Carlo methods; Monte Carlo simulation; Nonparametric statistics; Nucleotide sequence; Pagophilus groenlandicus; Parametric analysis; Phylogenetics; Phylogeny; Phylogeography; Population genetics; Populations; Seals; Seals (Animals); Seals, Earless - genetics; Species; Statistics, Nonparametric
• ISSN: 1932-6203; 1932-6203
• DOI: 10.1371/journal.pone.0134207

Phylogenomic analysis of highly-resolved intraspecific phylogenies obtained from complete mitochondrial DNA genomes has had great success in clarifying relationships within and among human populations, but has found limited application in other wild species. Analytical challenges include assessment of random versus non-random phylogeographic distributions, and quantification of differences in tree topologies among populations. Harp Seals (Pagophilus groenlandicus Erxleben, 1777) have a biogeographic distribution based on four discrete trans-Atlantic breeding and whelping populations located on "fast ice" attached to land in the White Sea, Greenland Sea, the Labrador ice Front, and Southern Gulf of St Lawrence. This East to West distribution provides a set of a priori phylogeographic hypotheses. Outstanding biogeographic questions include the degree of genetic distinctiveness among these populations, in particular between the Greenland Sea and White Sea grounds. We obtained complete coding-region DNA sequences (15,825 bp) for 53 seals. Each seal has a unique mtDNA genome sequence, which differ by 6 ~ 107 substitutions. Six major clades / groups are detectable by parsimony, neighbor-joining, and Bayesian methods, all of which are found in breeding populations on either side of the Atlantic. The species coalescent is at 180 KYA; the most recent clade, which accounts for 66% of the diversity, reflects an expansion during the mid-Wisconsinan glaciation 40~60 KYA. FST is significant only between the White Sea and Greenland Sea or Ice Front populations. Hierarchal AMOVA of 2-, 3-, or 4-island models identifies small but significant ΦSC among populations within groups, but not among groups. A novel Monte-Carlo simulation indicates that the observed distribution of individuals within breeding populations over the phylogenetic tree requires significantly fewer dispersal events than random expectation, consistent with island or a priori East to West 2- or 3-stepping-stone biogeographic models, but not a simple 1-step trans-Atlantic model. Plots of the cumulative pairwise sequence difference curves among seals in each of the four populations provide continuous proxies for phylogenetic diversification within each. Non-parametric Kolmogorov-Smirnov (K-S) tests of maximum pairwise differences between these curves indicates that the Greenland Sea population has a markedly younger phylogenetic structure than either the White Sea population or the two Northwest Atlantic pop