Assessing genetic structure of thornback ray, <i>Raja clavata</i>:A thorny situation?

Between 1956 and 1995, the thornback ray, Raja clavata, disappeared from the waters along the continental coasts of the North Sea; and severely declined in most North European waters. Because of its slow growth rate, long generation time, low fecundity and no pelagic larval phase, the time of recovery is long; and thus the alarming decrease has led to concern about the sustainability of skate populations, and whether depleted areas can be naturally replenished from adjacent areas. Because tagging data suggested a small “home range” and philopatric behaviour, relatively strong population differentiation was expected. The aim of this thesis was to test this hypothesis by assessing the genetic population structure of thornback rays. There were two initial questions: What is the spatial scale and connectivity of genetically defined populations or stocks of the thornback ray, Raja clavata in the North Sea/Northern Atlantic? How well do indirect estimates of migration and connectivity, based on genetic data, correspond to direct estimates based on tagging data? The first step of this study was the development of microsatellite loci, which proved to be difficult but was finally successful. At a regional scale, in British waters, population differentiation was assessed using the five newly developed microsatellite loci and significant global population differentiation was found (FST=0.013 P&lt;0.001). Pairwise comparisons revealed no relationship with geography or with time and locations sampled in Spring/Summer were more genetically differentiated than locations collected in Autumn/Winter. Altogether, these results strongly suggest that the population genetic structure of thornback rays is temporally and spatially dynamic, which may be due to species behavior. Genetic data indicated also that migration distance is substantially larger than suggested by earlier tagging studies. Such discrepancies between tagging and genetic data are common given that tagging studies follow individuals for a maximum of 4-5 years (half a generation time for thornback ray), whereas population genetic studies integrate time over several generations. The main conclusions, in any case, are that dispersal capacity of thornback rays is greater than what had been previously reported. At the European scale, the population structure was assessed in Raja clavata using the microsatellite loci and a fragment of 290 bp of the mitochondrial cytochrome b (mtDNA). Strong population differentiation was found among the Mediterranean basin (Corsica, Adriatic and Black Seas), the Azores and the European continental shelf (Britain, Gulf of Biscay and Portugal), as well as a significant isolation by distance. Depth (continental margins), the Gibraltar Strait/Almeria-Oran front and geographic distance act as barriers to gene flow. The genetic diversity distribution and the haplotype network suggested the Iberia Peninsula and the Azores were former glacial refugia, and British waters a secondary contact zone between these two refugia. Population genetic structure was also investigated in Amblyraja radiata across the Northern Atlantic. Amblyraja radiata has similar life-history traits than R. clavata, and thus using the same mtDNAmarker, patterns and degree of genetic structure were expected to be similar marker than for R. clavata. Genetic differentiation was not significant across the Northern Atlantic, i.e. at a scale of several thousand kilometers and showed that A. radiata as a higher dispersal capacity than R. clavata. The conclusions here are 1) life history type is not a priori a good predictor of dispersal capacity in Rajidae; and 2) dispersal of Amblyraja radiata unlike Raja clavata is not restricted by bathymetry. An important parameter for conservation is the effective population size Ne and the Ne/N ratio (N= census population size). Based on temporal samples from 1965 and 2003-2004, the Ne was estimated of between 3 to 4 orders of magnitude smaller than N, showing that very few individuals mate and contribute to the next generation. For fisheries management, this low Ne and Ne/N ratio cannot be ignored and strongly suggest the need to protect adults and the mating grounds to avoid further reductions of Ne. Finally mating system of a species is also an important parameter for long term conservation and is not well known for thornback rays. We investigated the possibility that polyandry is part of the mating system in R. clavata by analyzing egg clutches of four females. A minimum of 4 to 6 fathers was necessary to explain the progeny array and thus showed that females are polyandrous. As R. clavata has a long egg-laying season, polyandry could maximize the likelihood of viable sperm being available during the egg-laying season and/or multiple paternity could ensure further fitness advantages for the females. Protecting nurseries and mating grounds are fundamental for efficient conservation in Elasmobranchs, and the low Ne and Ne/N of thornback ray support the need to protect adults and mating grounds. Juveniles and immature rays must also be protected as their morphology and size makes them vulnerable to virtually all classes of nets. The Thames Estuary is known for the high abundance of juveniles and thus constitutes one or several nurseries. Individuals seasonally migrate in and out of the Thames Esturary and the higher level of population differentiation observed among Spring/Summer populations correlates with this seasonal migration; and suggests a certain level of philopatry, related to mating or egg-laying behaviour. Flatfish fisheries take place in the same area that includes nurseries for thornback rays and the economic value of sole and other fisheries is far higher than that of skates. Thus protected area design with closure areas will be difficult in the Thames estuary. Moreover, even if one type of design proves to be efficient for thornback rays, there is a strong chance that it will not be for another species. This last remark is one of the main reasons for a recent shift from single-species based management to ecosystem based management from which thornback ray may benefit more. Efforts to establish protected areas at the ecosystem level are still difficult. In any case, it is certain that compromises will have to be found between exploitation of a single species and ecosystems management