Discriminating whether the most appropriate model to reconstruct the demographic history of a species is structured or unstructured should be the first step in empirical population genetics investigations, particularly when targeting species of conservation concerns. Even when an extensive spatial sampling is lacking, an ABC model selection approach can actually distinguish whether the sampled deme belongs or not to a meta-population (similarly to previous studies(Maisano Delser et al., 2019; Peter, Wegmann, & Excoffier, 2010)). Among the four species considered here, the tiger shark is the only panmictic. The three other species conversely are best described by the SST model, i.e., the sampled populations belong to a meta-population exchanging migrants following a stepping stone matrix. Our results reflect the tight link between the level of meta-population structure (or its absence) and life history traits. The panmictic G. cuvierunsurprisingly can accomplish transoceanic movements and has the largest body size among the sharks here considered (Table S1). In the three other species, the estimated number of migrants (Nm) remarkably follows the increase of movement range (Table 1 and S1) and it is consistent with their behaviour and habitat use. Indeed, C. melanopterus, a strongly lagoon dependent species, displays the lowest level of connectivity among the studied species (Tables 1 and S1). These results bring meaningful hints about the influence of life history traits on population structure in sharks, but more studies addressing this topic will be needed to accurately detect which traits best predict its extent.