Scope for sympatric giant-dwarf speciation driven by cannibalism in South-American annual killifish ( <i>Austrolebias</i> )

Abstract A trophic radiation in the South-American annual killifish genus Austrolebias has led to the evolution of large specialized piscivores from small generalized carnivores. It has been proposed that this occurred in a single series of vicariant speciation events. An alternative hypothesis is denoted giant-dwarf speciation: piscivores would have evolved in sympatry by character displacement and cannibalism. We test the plausibility of both scenarios using size measures combined with distributional data and new phylogenetic trees based on mitochondrial and nuclear molecular markers. Our analysis uses historical biogeography models and Ornstein-Uehlenbeck processes describing trait evolution across the posterior distributions of phylogenetic trees. Large species most likely evolved three times from small ones. For the clade containing A. elongatus , we argue that vicariance was not involved in the origin of these large and specialized piscivores. They experience stabilizing selection with an optimum shifted towards larger bodies and longer jaws. The branch leading to this clade has the fastest evolving jaw lengths across the phylogeny, in agreement with expectations for giant-dwarf speciation. For A. wolterstorffi , the support for giant-dwarf speciation is weaker. When the species is placed at the root of Austrolebias , ancestral reconstructions are unreliable and vicariance cannot be ruled out. For the remaining large species, we can reject vicariance and giant dwarf speciation. Our results give rise to two new additional scenarios for the evolution of specialized piscivores. In the first, two successive speciation events in sympatry or parapatry produced large and piscivorous species. In the second, the immigration of a different annual killifish genus ( Cynopoecilus ) in the Patos area of endemism has contributed to in-situ diversification of Austrolebias species.