Quantifying dispersal between marine protected areas by a highly mobile species, the bottlenose dolphin, Tursiops truncatus

Abstract

<jats:title>Abstract</jats:title><jats:p>The functioning of marine protected areas (<jats:styled-content style="fixed-case">MPA</jats:styled-content>s) designated for marine megafauna has been criticized due to the high mobility and dispersal potential of these taxa. However, dispersal within a network of small <jats:styled-content style="fixed-case">MPA</jats:styled-content>s can be beneficial as connectivity can result in increased effective population size, maintain genetic diversity, and increase robustness to ecological and environmental changes making populations less susceptible to stochastic genetic and demographic effects (i.e., Allee effect). Here, we use both genetic and photo‐identification methods to quantify gene flow and demographic dispersal between <jats:styled-content style="fixed-case">MPA</jats:styled-content>s of a highly mobile marine mammal, the bottlenose dolphin <jats:italic>Tursiops truncatus</jats:italic>. We identify three populations in the waters of western Ireland, two of which have largely nonoverlapping core coastal home ranges and are each strongly spatially associated with specific <jats:styled-content style="fixed-case">MPA</jats:styled-content>s. We find high site fidelity of individuals within each of these two coastal populations to their respective <jats:styled-content style="fixed-case">MPA</jats:styled-content>. We also find low levels of demographic dispersal between the populations, but it remains unclear whether any new gametes are exchanged between populations through these migrants (genetic dispersal). The population sampled in the Shannon Estuary has a low estimated effective population size and appears to be genetically isolated. The second coastal population, sampled outside of the Shannon, may be demographically and genetically connected to other coastal subpopulations around the coastal waters of the <jats:styled-content style="fixed-case">UK</jats:styled-content>. We therefore recommend that the methods applied here should be used on a broader geographically sampled dataset to better assess this connectivity.</jats:p>