Causes and consequences of geographical variation in demography and larval exchange in reef fishes

Many species vary in their ecology across their geographic ranges in response to environmental variation. I document changes in the ecology of a common reef fish in the Galapagos Islands across regions that differ in their environmental conditions and test potential causal mechanisms. Individuals are larger, occur in higher densities, and live longer in the coldest region of the islands. Food quality is highest in the coldest region and a model of reproductive output based on temperature suggests that fish in the warmest region are allocating more available energy to reproduction, resulting in apparent regional life-history tradeoffs. Regional demographic differences appear to be driven by variation in food availability and an environmentally-mediated life-history tradeoff. Regional differences in reproductive output may influence rates of larval exchange among regions, but levels of larval exchange among populations are unknown for any marine species with a pelagic larval phase. I sought to use natural chemical variation in otoliths (earstones) of fishes as natural tags to measure larval exchange. To do this successfully, a reference collection of pre-pelagic, natal otoliths are required, as are measurements of the chemical signatures of cores of otoliths of recruits. I analyzed the chemical composition of natal otoliths from benthic eggs of S. beebei from around the Galapagos Archipelago and from two species of reef fishes from the Hawaiian Islands. I found no consistent variation in chemical signature at the largest scales, either among regions in Galapagos or among islands in Hawaii, but found significant variation among sites and among clutches. In addition, I examined cores of newly settled recruits and found that they are different chemically from natal otoliths, which ultimately comprise the otolith core as the fish grows, and non-core otolith material. Variation in natal otolith chemistry at smaller spatial scales but not larger ones implies that otolith chemistry will be usable only if larvae disperse short distances, but may not be usable if larvae regularly disperse larger distances. While additional validation work is required, otolith chemistry has the potential to address questions of larval connectivity, and provide information on the relationship between reproductive output and larval connectivity.