Unravelling population genetic structure with mitochondrial DNA in a notional panmictic coastal crab species: Sample size makes the difference

Background: The extent of genetic structure of a species is determined by the amount of current gene flow and the impact of historical and demographic factors. Most marine invertebrates have planktonic larvae and consequently wide potential dispersal, so that genetic uniformity should be common. However, phylogeographic investigations reveal that panmixia is rare in the marine realm. Phylogeographic patterns commonly coincide with geographic transitions acting as barriers to gene flow. In the Mediterranean Sea and adjoining areas, the best known barriers are the Atlantic-Mediterranean transition, the Siculo-Tunisian Strait and the boundary between Aegean and Black seas. Here, we perform the so far broadest phylogeographic analysis of the crab Pachygrapsus marmoratus, common across the north-eastern Atlantic Ocean, Mediterranean and Black seas. Previous studies revealed no or weak genetic structuring at meso-geographic scale based on mtDNA, while genetic heterogeneity at local scale was recorded with microsatellites, even if without clear geographic patterns. Continuing the search for phylogeographic signal, we here enlarge the mtDNA dataset including 51 populations and covering most of the species' distribution range. Results: This enlarged dataset provides new evidence of three genetically separable groups, corresponding to the Portuguese Atlantic Ocean, Mediterranean Sea plus Canary Islands, and Black Sea. Surprisingly, hierarchical AMOVA and Principal Coordinates Analysis agree that our Canary Islands population is closer to western Mediterranean populations than to mainland Portugal and Azores populations. Within the Mediterranean Sea, we record genetic homogeneity, suggesting that population connectivity is unaffected by the transition between the western and eastern Mediterranean. The Mediterranean metapopulation seems to have experienced a relatively recent expansion around 100,000 years ago. Conclusions: Our results suggest that the phylogeographic pattern of P. marmoratus is shaped by the geological history of Mediterranean and adjacent seas, restricted current gene flow among different marginal seas, and incomplete lineage sorting. However, they also caution from exclusively testing well-known biogeographic barriers, thereby neglecting other possible phylogeographic patterns. Mostly, this study provides evidence that a geographically exhaustive dataset is necessary to detect shallow phylogeographic structure within widespread marine species with larval dispersal, questioning all studies where species have been categorized as panmictic based on numerically and geographically limited datasets.