Intertidal patterns and processes tracking the effects of coastline topography and settlement choice across life stages of the mussels perna perna and mytilus galloprovincialis

Within landscapes, spatial heterogeneity is common and specific landscape features can influence propagule dispersal by wind or water, affecting population connectivity and dynamics. Coastline topographic features, such as bays and headlands, have a variety of biophysical effects on nearshore oceanography, larval transport, retention and supply, and the processes of larval settlement and recruitment. Although this has been demonstrated in several parts of the world, engendering a perception of a general ‘bay effect’, few studies have investigated this generality in a single experiment or region, by replicating at the level of ‘bay’. The Agulhas biogeographic region of the south coast of South Africa is a useful system within which to test for such generality. Using the intertidal mussels Mytilus galloprovincialis and Perna perna as model organisms, patterns of adult distribution were surveyed across four large ‘halfheart’ bays and intervening stretches of open coast, providing replication at the level of ‘bay’ and duplication of ecologically similar species. In support of a general, pervasive influence of bays on intertidal populations, mussel cover was found to be greater in bays than on the open coast for both species, although the effect was strongest for M. galloprovincialis. To explain this adult distribution, settlement, post-settlement mortality and recruitment were examined over 12mo at the same sites, with the prediction that rates of each would favour larger bay populations. Contrary to this, an interaction between month and bay-status was found, with greater settlement and recruitment on the open coast than in bays reflecting extreme settlement and recruitment events at 3 westerly open coast sites during summer. Re-analysis excluding these outliers, revealed the expected effect, of greater settlement and recruitment in bays. While this indicates the broad generality of the bay effect, it highlights exceptions and the need for replication in time and space when examining landscape effects. Measuring post-settlement mortality required testing small-scale settlement behaviour on established and newly deployed settler collectors. It was found that all settlers preferred collectors with biofilm, but that primary settlers avoided conspecific settlers, while secondary settlers were attracted to them. With discrepancies in settler attraction to new and established collectors accounted for, initial (over 2d) and longer-term (over 7d) post-settlement mortality rates were found to be substantial (ca 60 %) for both species. No topographic effect on p-s mortality was evident. Finally, recruit-settler, adult-recruit and interspecies correlations were examined at regional and local scales. Synergistic (or neutral) effects maintained the initial settlement pattern in recruit and adult populations regionally, but not at local scales; striking interspecies correlations suggested the influence of common regional transport processes. Ultimately, the results emphasize the importance of the direction of effects in different life stages and at different spatial scales, and the possibility that antagonistic effects may mask even strong patterns