Abstract
Sharks represent one of the most diverse groups of predators, playing important functional roles in coastal and oceanic ecosystems. They are also one of the most threatened groups because of their vulnerability to anthropogenic pressures due to their particular life history traits. Shark populations are therefore collapsing with drastic decrease in abundance in all marine ecosystems. Even relatively common species are near- threatened. Despite the deployment of important resources for shark population assessments, 41% of the 482 shark species on the International Union for Conservation of Nature (IUCN) Red List of Threatened Species lack a conservation status due to data deficiency. Improving our knowledge on such species is thus crucial for efficient protection to slow down their decline. More particularly, there is a necessity for a better characterization of presence, structure and connectivity of shark populations to define their conservation status, prioritize spatial management and optimize conservation efforts. This thesis relies on the emergence of new technologies to fill knowledge gaps on tropical coral reef sharks and to suggest conservation measures for better management. First, a method to survey shark communities has been developed during this thesis, based on the collection and sequencing of DNA present in the environment (environmental DNA metabarcoding; eDNA). Then, this method has been compared to exhaustive surveys of reef shark communities with traditional methods. This quick and non-invasive approach detected at least 21 shark species in waters of two distinct biogeographical areas (Caribbean and New Caledonia). Moreover, diversity and abundance patterns of DNA reads match with anthropogenic impact gradients and protected status of the sampled areas. The analysis of 22 eDNA samples detected more species in both remote reefs and impacted areas of the New Caledonian archipelago than 2758 scientific dives conducted during nearly 30 years and 385 baited remote underwater videos deployed over two years. Then, population structure and connectivity of a more common reef shark species, Carcharhinus amblyrhynchos, have been characterized using a seascape genomics approach. This thesis is based on a substantial genetic sampling in the archipelago of New Caledonia but also in several other sites in the Indo-Pacific (515 sharks in total). An isolation-by-resistance approach using circuit theory has been developed to explore what parameters are driving the genetic differentiation of C. amblyrhynchos. Here I show that deep oceanic areas act as strong barriers and proximity to habitat is a facilitator for dispersal. High-resolution modelling of genetic differentiation at the entire distribution range of the species (Indo-Pacific) led to the definition of hierarchical conservation units and a high number of isolated sites. Then, an approach taking into account the decline of abundance in impacted reefs showed an important fragmentation of shark populations and allowed the identification of remote reefs as refuges but also sources through dispersal towards impacted areas, insuring population persistence at a regional scale. This thesis demonstrates the potential of eDNA analysis for unveiling the presence of rare and elusive species such as sharks and for filling knowledge gaps in the conservation status of sharks. It also reveals the persistence of residual populations in impacted areas, that could show behavioral alterations like shifts in habitat use towards deeper waters or increased nocturnality. Finally, this thesis not only describes the population structure of a near-threatened species at high resolution and global scale, but also identifies conservation units and areas of high conservation priority that could help in the near future for the spatialization of marine management at multiple scales.