Divergent Positive Selection in Rhodopsin from Lake and Riverine Cichlid Fishes

Studies of cichlid evolution have highlighted the importance of visual pigment genes in the spectacular radiation of the African rift lake cichlids. Recent work, however, has also provided strong evidence for adaptive diversification of riverine cichlids in the Neotropics, which inhabit environments of markedly different spectral properties from the African rift lakes. These ecological and/or biogeographic differences may have imposed divergent selective pressures on the evolution of the cichlid visual system. To test these hypotheses, we investigated the molecular evolution of the dim-light visual pigment, rhodopsin. We sequenced rhodopsin from Neotropical and African riverine cichlids and combined these data with published sequences from African cichlids. We found significant evidence for positive selection using random sites codon models in all cichlid groups, with the highest levels in African lake cichlids. Tests using branch-site and clade models that partitioned the data along ecological (lake, river) and/or biogeographic (African, Neotropical) boundaries found significant evidence of divergent selective pressures among cichlid groups. However, statistical comparisons among these models suggest that ecological, rather than biogeographic, factors may be responsible for divergent selective pressures that have shaped the evolution of the visual system in cichlids. We found that branch-site models did not perform as well as clade models for our data set, in which there was evidence for positive selection in the background. One of our most intriguing results is that the amino acid sites found to be under positive selection in Neotropical and African lake cichlids were largely nonoverlapping, despite falling into the same three functional categories: spectral tuning, retinal uptake/release, and rhodopsin dimerization. Taken together, these results would imply divergent selection across cichlid clades, but targeting similar functions. This study highlights the importance of molecular investigations of ecologically important groups and the flexibility of clade models in explicitly testing ecological hypotheses.