Seascape features, rather than dispersal traits, predict spatial genetic patterns in co-distributed reef fishes

Aim: To determine which seascape features have shaped the spatial genetic patterns of coral reef fishes, and to identify common patterns among species related to dispersal traits [egg type and pelagic larval duration (PLD)]. Location: Indian and Pacific Oceans, including the Indo-Australian Archipelago. Methods: We sampled coral reef fishes with differing dispersal traits (Pomacentrus coelestis, Dascyllus trimaculatus, Hailchoeres hortulanus and Acanthurus triostegus) and characterized spatial (mtDNA) genetic patterns using AMOVA-clustering and measures of genetic differentiation. Similarity in the spatial genetic patterns among species was assessed using the congruence among distance matrices method and the seascape features associated with the genetic differentiation of each species were identified using multiple regression of distance matrices (MRDM) and stepwise model selection. Results: Similar spatial genetic patterns were found for P. coelestis and H. hortulanus, despite their differing egg type (benthic versus pelagic). MRDM indicated that geographical distance was underlying their correlated genetic patterns. Species with pelagic eggs (A. triostegus and H. hortulanus) also had correlated patterns of genetic differentiation (D est ); however, a common underlying seascape feature could not be inferred. Additionally, the common influence of the Torres Strait and the Lydekker/Weber's line was identified for the genetic patterns of differentiation for P. coelestis and A. triostegus, despite their differing dispersal traits, and the uncorrelated spatial genetic patterns of these species. Main conclusions: Our study demonstrates the value of a quantitative, hypothesis-testing framework in comparative phylogeography. We found that dispersal traits (egg type and PLD) did not predict which species had similar spatial genetic patterns or which seascape features were associated with these patterns. Furthermore, even in the absence of visually similar, or correlated spatial genetic patterns, our approach enabled us to identify seascape features that had a common influence on the spatial genetic patterns of co-distributed species.