The work presented in this thesis is a study of the spatial genetic structure of two species of scleractinian corals, Pocillopora damicornis and Platygyra daedalea. The overall objective was to investigate the current capacity of MPAs to preserve and restore coral reefs and to contribute information to improve management decisions. Samples were taken from 11 sites along the coast of Kenya. In addition P. daedalea was sampled from one site on each of the islands of Unguja (Zanzibar archipelago) and Mafia in Tanzania.

Populations of P. daedalea were largely panmictic within two lagoonal clusters. Samples obtained from two off-shore sites were more similar to the samples from the island samples in Tanzania than they were to proximal lagoonal samples. Off-shore and island samples were also significantly more genetically diverse in both the level of heterozygosity and allelic richness. Migration predominantly occurs from off-shore and island reefs toward the lagoonal reefs. The investigated populations of P. damicornis exhibited limited genetic neighbourhoods and highly localised recruitment and a significant pair-wise differentiation was found between a majority of the sampling sites. These results are probably a cause of the genetic bottleneck caused by the high mortality of this species during the 1998 coral bleaching event and reflects subsequent founder events and variable recruitment patterns, causing genetically distinct populations.

P. damicornis also displayed plastic reproduction, with some sites being dominated by large numbers of identical multilocus genotypes. The clonal colonies showed signs of being genotypcially distinct as they were all found to be homozygotes for a specific allele at one of the microsatellite loci. Due to the decreased genotypic diversity, associated with the high rate of clonal reproduction, the level of genetic diversity and allelic richness tended to be lower at unmanaged reefs. Also, the probability of encountering clonally produced individuals was significantly greater on unprotected reefs compared with protected reefs.

Canonical discriminant analysis of measurements of ten skeletal characters identified two distinct morphological types of P. daedalea. Analysis of molecular variance showed that morphotype explained a larger proportion of the genetic divergence than did the geographic distribution of sites sampled within MMP. However, phylogenetic analysis of rRNA sequences showed no evidence of sequence divergence, neither between morphotypes nor between samples of P. daedalea from MMP and samples of Platygyra sinensis from Hong Kong.

In summary, these data demonstrate the importance of incorporating multiple species and aspects when addressing the issue of connectivity and its implications for management. P. damicornis would benefit from localised management efforts to preserve its genotypic diversity, whereas P. daedalea would benefit from larger reserves to minimise the risk of loosing rare alleles through genetic drift which would further reduce its genetic diversity on lagoonal reefs. It also brings the subject of morphological characteristics and local adaptation into focus by revealing possible genetic divergence between two morphotypes of P. daedalea and two reproductive modes in P. damicornis.