Skip to main content

South African Tursiops aduncus ddRAD genepop file

Cite this dataset

Hoelzel, Rus (2021). South African Tursiops aduncus ddRAD genepop file [Dataset]. Dryad.


Many marine species exhibit fine-scale population structure despite high mobility and a lack of physical barriers to dispersal, but the evolutionary drivers of differentiation in these systems are generally poorly understood. Here we investigate the potential role of habitat transitions and seasonal prey distributions on the evolution of population structure in the Indo-Pacific bottlenose dolphin, Tursiops aduncus, off South Africa’s coast, using double-digest Restriction-site Associated DNA sequencing (ddRADseq). Population structure was identified between the eastern and southern coasts and correlated with the habitat transition between the temperate Agulhas (southern) and subtropical Natal (eastern) Bioregions, suggesting differentiation driven by resource specialisations. Differentiation along the Natal coast was comparatively weak, but evident in some analyses and varied depending on whether the samples were collected during or outside the seasonal sardine (Sardinops sagax) run. This local abundance of prey could influence the ranging patterns and apparent genetic structure of T. aduncus. These findings have significant and transferable management implications, most importantly in terms of differentiating populations inhabiting distinct Bioregions and seasonal structural patterns within a region associated with the movement of prey resources.


Data were generated using the ddRADseq method (Peterson et al. 2012), sequenced on an Illumina HighSeq 2500 platform.  Genotype data were generated in the program STACKS (Catchen et al., 2011).  Single-nucleotide polymorphisms (SNPs) were detected using ref_map (to the turTRU reference sequence). During short-read alignment the minimum depth of coverage to create a stack was 3, and the maximum distance (in bp) between stacks was 2. The number of mismatches allowed when building a catalogue was 2. Loci were filtered using Populations, removing those that had a stack depth below eight, or were not present in at least 70% of individuals or in all five groups according to sampling areas.