Molecular phylogeny of South African gobiid fishes of the genus Caffrogobius Smitt, 1900 and intraspecific genetic variation of Caffrogobius gilchristi boulenger,1900

Abstract

The molecular phylogeny of the southern African endemic genus Caffrogobius and the variability in intraspecific DNA sequences of the prison goby, Caffrogobius gilchristi, from the South African coastline, were investigated. The genus Caffrogobius (Gobiidae) consists of seven nominal western Indian Ocean species, six of which occur in the southern African region (from Mozambique to Namibia) with the seventh species, C. dubius, from Seychelles and do not form part of this study. These are small, benthic fishes that generally inhabit shallow-water habitats in estuaries while their larvae are marine. There is a high degree of morphological similarity among the species of this genus, hampering species identification using traditional dichotomous morphological keys. In a study of the phylogeny of the genus,the cytochrome b (cyt-b) and cytochrome oxidase subunit I (COI) loci were employed to resolve taxonomic problems.Six Caffrogobius species (C. gilchristi, C. nudiceps, C. saldanha, C. agulhensis, C. caffer and C. natalensis) were used for molecular phylogenetic analyses. Two partial mitochondrial DNA genes, cyt-b (281 bp) and COI (535bp) were amplified for 28 specimens using Polymerase Chain Reaction (PCR). Three phylogenetic tree reconstruction methods were employed namely, maximum parsimony (MP), maximum likelihood (ML) and Bayesian inference (BI) using Bathygobius sp. 1 and Bathygobius sp. 2 as outgroups. According to the estimated topologies from the family Gobiidae, the genus Caffrogobius is monophyletic and consists of three clades, the C. agulhensis and C. saldanha clade with C. natalensis as sister group, the sister groups C. gilchristi and C. caffer and the basal clade C. nudiceps. Phylogeographic patterns of the prison goby, C. gilchristi using sequences from cyt-b (425 bp from 63 specimens) and COI (524 bp from 109 specimens) may reveal patterns related to intraspecific population genetic structure. Demographic history was investigated with neutrality tests (Tajima‘s D, Fu‘s Fs and Ramos-Onsins & Rozas R2), mismatch distributions and haplotype networks which all indicated population expansion. Spatial groupings of populations were examined using AMOVA. Coalescent analyses were used to estimate gene flow between sampling locations and indicated a high level of gene flow among localities. AMOVA showed no consistent iv patterns of differences in the genetic structure within populations and thus AMOVA and parsimony network analyses revealed panmictic populations (i.e. no significant population structure detected), suggesting a recent divergence. Furthermore, parsimony networks support recent population expansion of only one population amongst all the localities analysed. Several possible life-history adaptations could be responsible for maintaining gene flow across phylogeographical distribution range. These may include a long pelagic larval stage and larval behavior, as well as oceanographic features such as the flow of sea currents. The presence of a marine pelagic larval stage in gobies may account for this lack of population structuring. It is also reasonable to assume that estuaries farther away from a particular spawning site would receive fewer recruits from that site, than would be the case for estuaries situated at a closer proximity to the breeding event. This means that conservation should not be focused on individual estuaries, but on a wider geographical scale. However this must still be tested further