Understanding the population dynamics of highly mobile, widely distributed, oceanic sharks, many of which are overexploited, is necessary to aid their conservation management.  We investigated the global population genomics of tiger sharks (Galeocerdo cuvier), a circumglobally distributed, apex predator displaying remarkable behavioral versatility in its diet, habitat use (near coastal, coral reef, pelagic), and individual movement patterns (spatially resident to long-distance migrations).  We genotyped 242 tiger sharks from 10 globally distributed locations at more than 2000 single nucleotide polymorphisms. Although this species often conducts massive distance migrations, the data show strong genetic differentiation at both neutral (F_ST=0.125-0.144) and candidate outlier loci (F_ST=0.570-0.761) between western Atlantic and Indo-Pacific sharks, suggesting the potential for adaptation to the environments specific to these oceanic regions. Within these regions, there was mixed support for population differentiation between northern and southern hemispheres in the western Atlantic, and none for structure within the Indian Ocean. Notably, the results demonstrate a low level of population differentiation of tiger sharks from the remote Hawaiian archipelago compared to sharks from the Indian Ocean (F_ST=0.003-0.005, P<0.01). Given concerns about biodiversity loss and marine ecosystem impacts caused by overfishing of oceanic sharks in the midst of rapid environmental change, our results suggest it imperative that international fishery management prioritize conservation of the evolutionary potential of the highly genetically differentiated Atlantic and Indo-Pacific populations of this unique apex predator. Furthermore, we suggest targeted management attention to tiger sharks in the Hawaiian archipelago based on a precautionary biodiversity conservation perspective.

Genotyping-by-Sequencing library preparation was performed per Elshire et al. (2011) and raw sequence reads were generated on an Illumina HiSeq.

These data were sequenced on three Illumina HiSeq lanes, and the data from each lane has been split into two separate fq.gz files for upload (i.e., split1, and split2).

Key file associating barcodes with sample IDs included.