Development of polymorphic microsatellite loci for conservation genetic studies of the coral reef fish Centropyge bicolor.

A total of 23 novel polymorphic microsatellite marker loci were developed for the angelfish Centropyge bicolor through 454 sequencing, and further tested on two spatially separated populations (90 individuals each) from Kimbe Bay in Papua New Guinea. The mean ± s.e. number of alleles per locus was 14·65 ± 1·05, and mean ± s.e. observed (HO ) and expected (HE ) heterozygosity frequencies were 0·676 ± 0·021 and 0·749 ± 0·018, respectively. The markers reported here constitute the first specific set for this genus and will be useful for future conservation genetic studies in the Indo-Pacific region.

Estimating connectivity in marine populations: an empirical evaluation of assignment tests and parentage analysis under different gene flow scenarios.

The application of spatially explicit models of population dynamics to fisheries management and the design marine reserve network systems has been limited due to a lack of empirical estimates of larval dispersal. Here we compared assignment tests and parentage analysis for examining larval retention and connectivity under two different gene flow scenarios using panda clownfish (Amphiprion polymnus) in Papua New Guinea. A metapopulation of panda clownfish in Bootless Bay with little or no genetic differentiation among five spatially discrete locations separated by 2-6 km provided the high gene flow scenario. The low gene flow scenario compared the Bootless Bay metapopulation with a genetically distinct population (F(ST )= 0.1) located at Schumann Island, New Britain, 1500 km to the northeast. We used assignment tests and parentage analysis based on microsatellite DNA data to identify natal origins of 177 juveniles in Bootless Bay and 73 juveniles at Schumann Island. At low rates of gene flow, assignment tests correctly classified juveniles to their source population. On the other hand, parentage analysis led to an overestimate of self-recruitment within the two populations due to the significant deviation from panmixia when both populations were pooled. At high gene flow (within Bootless Bay), assignment tests underestimated self-recruitment and connectivity among subpopulations, and grossly overestimated self-recruitment within the overall metapopulation. However, the assignment tests did identify immigrants from distant (genetically distinct) populations. Parentage analysis clearly provided the most accurate estimates of connectivity in situations of high gene flow.