The SWIO is an area of great biodiversity and included in the diverse species that occupy the region are five species of sea turtles that include green turtles, hawksbills, leatherbacks, loggerheads and olive ridleys. Despite considerable conservation efforts at sea turtle rookeries in the South Western Indian Ocean, only green and loggerhead turtle populations have shown an increase in population size in recent years (<10 years), whereas leatherbacks remained stable and hawksbills and olive ridleys declined. This begs the question if fisheries (or other offshore pressures) are responsible for slowing the recovery of these populations in the region, and if so, which specific fisheries are responsible for this trend? Several offshore (mostly industrial) and coastal (mostly artisanal) fisheries overlap with sea turtle distribution at sea. Industrial fisheries that are globally known to have a demonstrable impact on sea turtle populations are longline and to a lesser extent purse seine fisheries, whilst prawn trawl, gillnet and beach seine fisheries are coastal fisheries with a known negative impact on sea turtle populations. Holistic conservation strategies should be developed that include both land and sea protection for sea turtle species. It is thus necessary to identify and manage offshore threats including fisheries activities, particularly those fisheries that are showing the highest risk to sea turtle populations. This prompted an investigation into the bycatch rates and mortality of all sea turtle species that occur in the SWIO region in several offshore and coastal fisheries including both industrial (longline, purse seine and prawn trawl) and artisanal (including gillnet and beach seine) fisheries. The specific aims were (i) to identify and quantify the interactions (and if possible mortality) of sea turtle species in fisheries and (ii) to identify vulnerable species/populations to fishing operations using a semi-quantitative Ecological Risk Assessment (ERA) in the form of a Productivity-Susceptibility Analysis (PSA). Published information, online databases and technical reports were used as data sources to establish a database containing essential information regarding fishing effort and sea turtle bycatch in the region. The existing information was used to map fisheries extent and effort within the region, and to perform bycatch calculations. Interactions and mortality rates for sea turtles in five fisheries were quantified using bycatch rates from regional studies. Between 2000 – 2011, industrial longline and purse seine fisheries captured sea turtles at a rate of 4 388 indiv.y-1, with the mortality rate being 189 indiv.y-1. The bulk of these interactions were in the longline industry that captured 4 129 ± 1 376 indiv.y-1, with a corresponding mortality rate of 167 ± 53 indiv.y-1. The most commonly caught species (in longlines) were loggerheads and leatherback turtles, but the greatest impact is expected to be on the leatherback population due to the high interaction rate relative to population size. The bycatch (259 ± 34 indiv.y-1) and mortality (20 ± 2 indiv.y-1) rates of sea turtles in the purse seine fishery was considerably lower than the longline fishery. The purse seine fishery thus does not seem to have a significant impact on sea turtle populations in the SWIO. The impact of all forms of fish aggregation devices were excluded from the analysis as the impacts of these are poorly documented. Coastal prawn trawl, gillnet and beach seine fisheries captured an estimate of between 50 164 - 52 370 indiv.y-1 from 2000-2011. The highest bycatch rate was estimated for gillnet fisheries (40 264 indiv.y-1) followed by beach seine fisheries (9 171 indiv.y-1) and prawn trawl fisheries (at 1089 – 2795 indiv.y-1). The gillnet fishery could be responsible for slowing the recovery rate of green turtle and leatherback populations in the SWIO due to the high capture rates in this fishery compared to the population sizes of the species. Beach seine and prawn trawl fisheries are not expected to be hamper the recovery rate of any of the populations in the SWIO due to the low levels of interactions and low mortality rates compared to the population sizes. There are however very few data available regarding the bycatch of sea turtle species within these fisheries, highlighting the need for further research regarding this. A productivity-Susceptibility Analysis (PSA) was used to evaluate the relative vulnerability of species to fisheries, and is frequently applied in data poor situations. Limited data on sea turtle life history characteristics and population dynamics of species in the SWIO prompted the use of a PSA to determine the species most vulnerable to fisheries in the region. Results of the PSA indicated that gillnet fisheries poses the largest fishery-related threat to sea turtle populations, specifically the green and leatherback populations. The longline fishery that poses a particular threat to the leatherback population in the SWIO is also a particular concern. A cumulative impact assessment (combining fisheries and other threats) indicated that the SWIO leatherback population is extremely vulnerable to the combination of threats that influence this population in the SWIO. Even though individual fisheries may pose a small threat, the cumulative impacts of the fisheries can lead to severe impacts on populations such as slowing the recovery rate of populations. There are however significant data gaps that require attention in order to fully assess the impact of these fisheries on sea turtle populations. Despite the fact that fisheries are not implicated as a mayor reason for the decline in the hawksbill and olive ridley populations in the region, these two species are in decline indicating that there are other factors responsible for the decline not yet identified. It however remains imperative to reduce the mortality from all sources to ensure the continued viability of sea turtle populations in the region.

Deep-water prawns Haliporoides triarthrus and langoustines Metanephrops mozambicus are endemic to the South West Indian Ocean (SWIO) region and make up the largest proportion of deep-water crustacean trawl catches in Mozambique and South Africa. Despite their economic importance to these fisheries, little is known about their distribution, biology and genetic population structure. The metapopulation genetic variation of H. triarthrus and M. mozambicus was assessed from 220 specimens per species collected from three sites in Mozambique (Bazaruto A, Boa Paz and Inhaca), two sites in western Madagascar (Morombe and Tulear) and one site in eastern South Africa (Durban). Two fragments of the mitochondrial region were amplified using universal primers ribosomal 16S subunit (16S) and mitochondrial cytochrome oxidase subunit I (COI). From H. triarthrus, fragments of 569 base pair (bp) (16S) and 1300 bp (COI) were amplified. A total of 207 sequences (16S) and 151 sequences (COI) were recovered, and 69 and 78 haplotypes identified, respectively. Metanephrops mozambicus 16S and COI genes produced similar fragment lengths, and 112 (16S) and 127 haplotypes (COI) were recovered. Both species demonstrated high genetic diversity and significant population differentiation in the SWIO region. Two sister-species (or subspecies) of H. triarthrus were identified, one occurring along the African continental shelf and the other off western Madagascar. Furthermore, individual populations making up each lineage were genetically structured, as indicated by the absence of shared haplotypes, and should be recognized as demographically distinct subspecies. Both species have undergone recent population expansions, likely since the late Pleistocene. The large anti-cyclonic and cyclonic eddies prevalent in the Mozambique Channel, and the boundary area between these eddies and upper Agulhas Current are likely factors driving larval retention or return process, thus giving rise to the observed genetically structured populations. The findings from this study are unique for the SWIO region, and may lead to a paradigm shift in the way that deep-water crustacean stocks are perceived by fisheries managers – instead of single shared stocks, they comprise of many isolated ones, in spite of the dispersal potential of larvae in strong ocean current regimes. Thus stocks should be managed as small independent units. / Thesis (M.Sc.)-University of KwaZulu-Natal, Durban, 2013.