Phylogeography and epifauna of two intertidal seaweeds on the coast of South Africa

Southern African biogeographic boundaries delimit the phylogeographic distribution of some coastal and estuarine invertebrates. This study investigated the impact of these boundaries on the phylogeographic distribution of two intertidal red seaweeds, Gelidium pristoides and Hypnea spicifera using the mitochondrial Cox2-3 spacer and the nuclear ITS1 regions. G. pristoides spores have short distance-dispersal, while long distance-dispersal is more likely in H. spicifera via spores and drifting fertile thallus fragments. Both markers revealed a south-western and south-eastern lineage within G. pristoides but the breaks between lineages do not coincide with any recognised biogeographic limits. The Cox2-3 spacer revealed a boundary between the two lineages at the Alexandria Coastal Dunefield (ACD) and ITS1 at the Gamtoos-Van Stadens Dunefields (GVD) which is approximately 80km west of the ACD. The minor difference between the two markers regarding location of the phylogeographic boundary is probably due to the dating differences between the two dunefields. The ACD as developed currently is superimposed on the ancient dunefields which formed during the Pleistocene, coinciding with the Cox2-3 spacer sequences divergence which dates back 500,000 - 580,000 years. The GVD formed during the Holocene (6,500 - 4,000 years ago), coinciding with the ITS1 sequences divergence which dates 4,224 - 4,928 years ago. Thus, these phylogeographic boundaries probably appeared without the influence of biogeographic boundaries, but rather due to the lack of suitable habitat in the dunefields, coupled with short dispersal-distances of the spores. Analysis of the ITS1 and Cox2-3 spacer regions in H. spicifera revealed that the species is characterized by uniform genetic structure along the coastline. This reflects the species`s potential for long range expansion as it inhabits both the intertidal and subtidal zones; and this presumably leads to high gene flow among populations. The ITS1 sequences showed minimal genetic variation of one substitution between the gametophyte and tetrasporophyte generations within H. spicifera. This suggests the predominance of asexual reproduction, which reduces gene flow and fixes alleles between generations. ANOSIM and Bray-Curtis cluster analyses showed scale-dependant variation in the abundances of epifauna (mainly amphipod, isopod, mollusc and polychaete species) on G. pristoides. At small local (within site) and large (among sites) scales, there were weak and no structure in epifaunal abundances respectively. However, at larger, biogeographic scales, samples from the same biogeographic region tended to be clustered together. Thus, there was a group containing predominantly south coast samples and a group containing east coast samples mixed with the remaining south coast samples. Such scale-dependant variation in epifaunal abundances is probably due to the effects of factors driving species richness at small local (within site) scales (e.g. wave exposure, seaweed biomass) and at larger, biogeographic scales (e.g. surface sea temperature). Moreover, at very small (individual samples) scales; there was no correlation between epifauna composition and genotype of the seaweed. Seaweed samples characterized by distinct ITS1 or Cox2-3 spacer sequences did not show any significant differences in epifaunal composition. Although the distributional pattern of the epifaunal community observed at large biogeographic scale is not clear, it seems to be associated with the biogeographic regions. However, phylogeographic distribution of Gelidium pristoides is not connected to biogeographic regions. Thus, at larger, biogeographic scales, there is no correlation between phylogeographic distribution of G. pristoides and distribution of the associated faun