Evolution of rampant hybridisation as a "key innovation" in Acropora

An important hypothesis in evolutionary biology posits that an increase in speciation rate often results from the acquisition of a "key innovation", a trait permitting new or more efficient resource use. Acropora is the most speciose genus of scleractinian corals, comprised of nearly one quarter of reef-inhabiting species. This group is thought to have attained its current high diversity via an ability to frequently hybridise (van Oppen et al. 2001 Mol Biol Evol 18 1315) or through the acquisition of unique morphological features, such as dimorphic corallites, which then enabled a rapid adaptive diversification of colony morphology (Wallace 1999 Staghorn Corals of the World CSIRO). We tested the idea that these characters are responsible for a uniquely high speciation rate in Acropora. Using maximum likelihood estimation and an Akaike weighting procedure, we identify a confidence set of evolutionary models from which we draw several novel inferences about the evolution of diversification rate in corals. First, Acropora speciates at a rate no greater than that of a con-familial genus Montipora, whose large size (ca. 10% of reef corals) does not appear to be primarily a consequence of interspecific hybridisation (van Oppen et al. 2004 Mar Biol 144 9). Second, a model is only slightly favoured in which high diversification arose before divergence by Acropora from its low-diversity sister group. However, consideration of the derived status of known large syngamea within Acropora renders this scenario unlikely. Further, we can reject a model in which a single high speciation rate and an attendant key innovation arose but once in the most recent ancestor common to Acropora and Montipora and were subsequently lost in these genera’s respective low-diversity sister groups. Hence, the high diversification rate in acroporid corals does not appear to have been a long-term feature of this family or, by extension, Cenozoic coral reefs. These results are insensitive to uncertainties in divergence times between acroporid clades. Taken together, these data rule out the frequently quoted hypothesis that Acropora possesses a uniquely high diversification rate and that the group’s radiation must be due to putative key innovations unique to the group, such as rampant interspecific hybridisation or corallite dimorphism. Below, is shown a subset of tested models Mij with the ith number of rate parameters and jth colouring of rates.