An almost unbroken fringing reef runs along the east coast of Africa, the lagoon inside the reef is the foundation of almost all artisanal fisheries. It is a low-tech fishery conducted by many people. Some areas can have up to 19 fishermen per square kilometer. High fishing pressures, coupled with declining fish stocks has led to changes in mean size and reproductive age of many exploited species. There is a vital and urgent need for scientifically based management systems, including the utilization of genetic information to guide management practices.

This thesis aims to investigate the presence of genetic structures in the western Indian Ocean. In order to do that we first investigated the historical patterns of connectivity throughout the region (paper I). In papers II and III we focused on local scale connectivity in Kenya and Tanzania and finally in paper IV we investigate the large-scale contemporary gene flow throughout the Western Indian Ocean. In paper III we also investigate the temporal genetic variation at one site and compare it to the small-scale genetic variation along a stretch of the Kenyan coastline. Some overall conclusions that can be drawn from my body of work are: there are genetic structures present in the western Indian Ocean even though the apparent lack of physical barriers. Major oceanic currents aid evolutionary dispersal patterns. A single geographic site need not be genetically homogenous or temporally stable. Island sites are genetically more homogenous than mainland sites.

In conclusion, there are clear and distinct genetic structures present especially in Siganus sutor, the most targeted fish for the artisanal fishery in East Africa.

Studies on genetic connectivity are important for the management of fisheries. In this study we used AFLP to investigate population structure of the endemic Spinefoot shoemaker, Siganus sutor, from 6 countries, Kenya, Tanzania, Comoros, Seychelles and Mauritius in the Western Indian Ocean. We collected 506 samples from 20 fish landing sites, 171 variable loci were used in the statistical analysis. Global F_ST was significant and showed a pattern of isolation by distance, mostly influenced by remote oceanic islands. In a previous study we have described the temporal variation of Siganus sutor to be about 1/5 of the global variation, and by applying a 1/5 of the global variation cut of value for the pair-wise comparisons we were able to account some of the pair- wise genetic variation as temporal fluctuations. A STRUCTURE analysis was also preformed that corroborates the pair-wise F_ST comparisons. Overall these results show that S. sutor is genetically diverse and subdivided throughout the region, but also that the current management regime might not be optimal. 

Artisanal fisheries in Kenya have been in decline since the 1980’s and are currently managed by gear restriction and no take zones. The fishery is a mixed species fishery but the Shoemaker spinefoot (Siganus sutor) comprise a large portion of the total catches. The sustainable use of these resources is dependent on informed assessment and management of the harvested species. In Kenya there is a lack of critical information about landings, fish stock productivity and genetic stock structure, and there is no knowledge of populations size or genetic variation of S. sutor. In this study we used the molecular marker, AFLP to investigate the genetic variation within and between sites of S. sutor landed along the 200 km coast of Kenya. We compared the spatial genetic variation among sites with the within site temporal genetic variation from a single site, adjacent to a number of spawning aggregations. Our results show that the there is genetic variation among the sites (spatial variation) and that the temporal genetic variation with in a six week period was about 1/5 of the spatial genetic variation. We believe these findings to be an important aspect to considered for both future scientific research as well as management.

There is a growing demand for wild caught juvenile fish to supply the market for aquaculture. However, little is known about the genetic effects of juvenile collection from wild populations. There are a number of imminent threats to both aquaculture systems and wild fish populations. Juvenile collection from a single population can for example reduce population’s evolutionary potential as well as the disease resistance within an aquaculture pond. In this study, we investigated the local genetic structure of juvenile mullets collected from five sites around Bagamoyo (Tanzanian mainland) and Zanzibar Island, East Africa. Fish were caught in low tide using a seine net. The fish were morphologically identified, and then genetically identified using direct sequencing of the CO1 gene with cross referencing with the Barcode of Life Database (BOLD) systems.  Molecular variance analyses were used to infer genetic subdivision based on geographic sampling site as well as inferring population structure through the Bayesian assignment test implemented in STRUCTURE 2.3. Our results showed that samples morphologically identified as Mugil cephalus where in fact Valamugil buchanani and we also found evidence of an introgression genome event, where the gene flow from one species may have affected the general gene pool. The Bayesian analysis revealed a clear genetic population structure among the sampled fish; the main difference was the presence of a unique mainland cluster. Our findings may have important implications for management and conservation of mullet fishes in the region and elsewhere.

Abstract—Uncontrolled growth of sea urchin populations may have a negative effect on coral reefs, making them barren. To avoid this, different methods of sea urchin reduction have been developed but, without knowledge of their genetic structure and connectivity, these methods may be ineffective. The aim of this study was to examine the fine-scale genetic structure and connectivity in the sea urchin, Diadema setosum, population around Unguja, Zanzibar, using AFLP. We found evidence of different genetic clusters, high migration between the sites and high genetic diversity within the sites. These findings indicate that a manual reduction of sea urchins with similar genetic connectivity, implemented on the same geographic scale as our study, would be ineffective since sites are probably repopulated from many sources.

This study showed that population differentiation in P. damicornis varied over spatial scales and that this variability occurred at both evolutionary and ecological time scales. This paradox is discussed in light of stochastic recruitment and small scale population structures found in other species of coral. The study also identifies potential source reefs, such as those within Mnemba Conservation area near Zanzibar and genetically isolated reefs such as those within Malindi Marine National Park and Reserve in northern Kenya.

Studies on genetic connectivity are essential for the design of management strategies for coral reef fisheries. In this study we used a mitochondrial DNA marker to investigate population structure of the reef-associated parrotfish, Scarus ghobban, from four countries, Kenya, Mauritius, Seychelles and Tanzania, in the western Indian Ocean. We obtained nucleotide sequences of the mitochondrial control region for 117 individuals. Measures of haplotype diversity were relatively high. Pairwise population differentiation (F (ST)) was low, but not always non-significant. Analysis of molecular variance (AMOVA) showed genetic differentiation between groups, when the data was partitioned into two groups consisting of samples from Mauritius and Tanzania in one group, and samples from Kenya and Seychelles in another group. Direction of gene flow was estimated using a Bayesian approach. Migration was sometimes asymmetric or directional, coinciding with the flow of major oceanic and coastal currents in the region. Mismatch distributions, based on the observed number of differences among haplotype pairs, produced a unimodal distribution, indicative of recent demographic expansion. Phylogenetic analyses revealed three clades without any geographic structure, suggesting recent migration between historically isolated lineages. We reconstructed the historical demography of S. ghobban and examined it in the context of Pleistocene climate stages and changes in relative sea level. Overall, these results showed that populations of S. ghobban are genetically diverse and have relatively high gene flow, with some genetic structuring in the western Indian Ocean.