Phase III – Small-Scale versus Large Scale Coral Genetic Connectivity within Southeast African Marginal Reefs

Marine protected areas (MPAs) have been established globally to protect coral communities from human and natural disturbances. Effective MPAs should be able to primarily preserve the genetic integrity of marine resources. Their success is dependent on the degree to which such populations are self-seeding or otherwise connected. Despite their significance to the effectiveness of MPAs, ecological genetic connectivity is yet to be determined for southeast African MPAs. This study was thus initiated to provide needed information on contemporary genetic connectivity in southeast African MPA. The study had the following objectives: (1) To assess genetic diversity and differentiation in A. austera and P. daedalea populations within and between southeast African MPAs. (2) To estimate levels of gene flow in A. austera and P. daedalea populations within and between MPAs. (3) To investigate within-reef patterns of spatial and temporal genetic connectivity in populations of A. austera and P. daedalea. (4) To explore the implications of the estimated genetic connectivity of A. austera and P. daedalea for the effective management of southeast African MPAs. This study was conducted in five countries within the WIO, namely South Africa, Mauritius, Mozambique, Reunion, Tanzania, and Kenya. Targeted beneficiaries of this research included the scientific community, MPA managers, and the international community, including donors, decision-makers, and policy makers. 

Coral samples of the two species (A. austera and P. daedalea), as well as data on population density and colony size, were collected along belt transects by divers. Molecular techniques were deployed to assess aspects of genetic diversity and estimate levels of gene flow and genetic connectivity. Assessment of genetic diversity was done using multiple, highly polymorphic molecular markers. The rarefaction method was used to estimate mean allelic richness (Ar(g)) and private allelic richness (Ap(g)) to account for differences in sample size. The Bayesian clustering method was implemented in BAPS to detect genetic heterogeneity in the sample. Analysis of Molecular Variance (AMOVA) and F- and G-statistics implemented in GenAlEx to assess the magnitude of genetic structure and differentiation among clusters.  Mantel test in GenAlEx was used to test the hypothesis that distance alone is having an effect in the levels of subdivision observed, that is. Gene flow is reduced by increasing the distance between reefs. Assignment tests were used to estimate recent migration among the identified clusters. Spatial autocorrelation analyses assessed the presence of non-random spatial distribution of genetic variation.

Results revealed a low contribution of asexual reproduction and a prevalence of gametic recombination via sexual reproduction in the coral populations of both species along the sampled area. Genetic diversity estimates of A. austera and P. daedalea showed that in both species, the northern most tropical sampled reefs (Reunion Island for A. austera and Chagos archipelago for P. daedalea) had the lowest allelic richness of all reefs, while NMR in South Africa had, in both species, allelic richness that was higher than any other reef. Results suggest that southeast African reef populations of A. austera and P. daedalea along the sampled area comprised several genetically distinct populations. Bayesian clustering analysis identified four genetically distinct populations (i.e., clusters) of A. austera and five of P. daedalea in the sampled area. The arrangement of these genetic populations appears to be related to the nature of reef development that is characteristic of the latitude at which they are found. Estimates of several migrants per generation between genetic clusters suggest a considerable amount of long-term migration between populations of A. austera and P. daedalea located in southern Mozambican and northern South African reefs. This migration was, however, not relevant at ecological time scales. Overall results suggest that northern South African reefs have a source of individuals that does not correspond to any of the sampled reefs or identified genetic populations (i.e., clusters). Our findings suggest genetic discontinuity at the reefal scale in both species. We found significant SPG between individuals within the reef in A. austera; this being the first study to demonstrate the occurrence of this, and the estimate of gene dispersal distance in this species was 8.1 to 24.2 m. P. daedalea, on the other hand, displayed limited dispersal. 

The existence of discrete populations of the two coral species in the southeast African region found in this study may correspond to management units (MUs) or evolutionarily significant units (ESUs). Furthermore, the inclusion of the full extent of the South African populations of the two species in iSWP provides grounds for some confidence in the functioning of this MPA for their protection. The evidence of limited dispersal in these species, their significant spatial genetic structure at the reef scale, and the potential role that northern reefs play as“stop-overs” for putative migrants suggest that the scope of management strategies for their protection should not be too broad; each reef should be seen only as a single unit, part of the whole rather than a representative of the whole. With some reefs acting as “stop-overs” for migrants (NMR) and others showing significant within-reef genetic structure (Two-mile Reef, South Africa), an adaptive management framework is recommended as the best option for this MPA. 

Analysis of genetic diversity and differentiation provided evidence for the existence of four discrete genetic populations of A. austera and five of P. daedalea in the sampled area. Higher genetic diversity was found on northern South African reefs (NMR and RAB), and the unassigned migrants that were detected in assignment tests suggested that South African reefs might have a source of gene flow that was not sampled. Similarly, genetic discontinuity observed in the area appears to be influenced by limited dispersal of larvae, inferred from significant spatial genetic structure found on Two-mile Reef.  Altogether, the results are consistent with the isolation observed in other studies using fewer variable markers and support the hypothesis that there is demographic discontinuity between the coral populations along the southeast African coast. The management implications and recommendations of the findings are discussed. 

Future studies may investigate contemporary gene flow between southern Madagascar and South African reefs to ascertain whether reefs from Madagascar are a more important source of coral larvae for South African reefs. Furthermore, future studies need to assess temporal genetic variations in the genetic structure of corals on South African reefs and the influence of environmental impacts to which the reefs are exposed.