- Several dolphin species have global distributions. The extent of their radiation and limits to gene flow are presumably a product of oceanographic features both recent and historical, behavioral specializations and social organization. Rough-toothed dolphins (Steno bredanensis) are globally distributed in tropical and subtropical waters and are generally found in depths greater than 1,500 meters making them challenging to comprehensively sample. Although it has been assumed that pelagic dolphins range widely due to the lack of apparent barriers and unpredictable prey distribution, recent evidence suggests rough-toothed dolphins exhibit fidelity to some oceanic islands. A small number of photo-identification and genetic studies conducted to date on rough-toothed dolphins show regional population structure and stable associations in groups, with some individuals observed repeatedly in the same groups over several years. The aim of this dissertation is to describe patterns of phylogeography over evolutionary time on a global scale and expand studies of population and social structure on a regional level. The dataset contains 351 rough-toothed dolphin biopsies, tissue and teeth samples collected from the Pacific and Atlantic Oceans and limited samples from the Indian Ocean.
To evaluate the phylogeography and test for possible species or subspecies level delineation between oceans, I used mitochondrial DNA sequences from the control region (350 bp) and 12 concatenated protein-coding genes from the whole mitogenome, as well as six nuclear introns. Although I found support for two Pacific clades and a private North Atlantic clade in the whole mitogenome, there were no genealogical patterns consistent across multiple loci, allowing me to reject species level delineation. To further evaluate the amount of gene flow and test for divisions below the species level, I used population level indices and found significant genetic differentiation for rough-toothed dolphins between the Atlantic Ocean with both the Indian/Western Pacific and Central/Eastern Pacific for both the mitochondrial datasets and the intron dataset. Significant differentiation between the Indian/Western Pacific Ocean and Eastern Pacific Ocean was found for the mitochondrial but not nuclear datasets. From these results I recommended the Atlantic Ocean basin be considered a separate evolutionary significant unit. This reflects that these populations are on independent evolutionary trajectories, but are not diagnosable species or subspecies.
To further evaluate population structure on a regional scale, I used a subset of these samples from three archipelagos in the Central Pacific Ocean including the Hawaiian Islands, the Society Island of French Polynesia and the Samoan Islands. Using a 450bp portion of the mtDNA control region and 15 microsatellite loci, an overall AMOVA indicated strong genetic differentiation among islands within the main Hawaiian Islands (mtDNA F[subscript ST]=0.165; p<0.001; microsatellite F[subscript ST]=0.043 p<0.001) and the Society Islands of French Polynesia (F[subscript ST]=0.499; p<0.001; microsatellite F[subscript ST]=0.079 p<0.001) as well as among the three archipelagos (mtDNA F[subscript ST]=0.299; p<0.001; microsatellite F[subscript ST]=0.055 p<0.001). My results corroborate the photo-identification and the genetic studies for three archipelagos, confirming population structure on the regional level. Lastly, to test the hypothesis that social structure observed in rough-toothed dolphins is kinship based, as in other delphinid species such as killer whales, I used a subset of the main dataset from groups of living and mass stranded dolphins. I found multiple matrilines in more than half the groups, allowing me to reject a strictly matrilineal group structure, such as that observed in some killer whales. Instead I found rough-toothed dolphin groups showed weak matrilineality, where some groups are more matrilineal than expected by chance. Although group structure is stable, is not determined primarily by kin-based relationships. These analyses provide new insights into a little studied species. The use of worldwide datasets allowed me to evaluate population structure on different temporal, spatial and regional scales and delineate populations for future conservation and management.