The ecology of adaptive radiation of dabbling ducks (anas spp.)

The ecological theory of adaptive radiation proposes that three processes are responsible for the evolution of a single ancestor into a clade of species: divergence in phenotype between contrasting environments; divergence in phenotype caused by negative interspecific interactions; and ecological speciation. I tested for evidence of these processes among sympatric dabbling ducks (Anas spp.), a putative adaptive radiation. Divergent selection between environments requires a trade-off in the ability of phenotypes to exploit resources in different environments. I tested whether variation in bill morphology imposes a performance trade-off when ducks filter-feed in environments containing different size-frequency distributions of prey and indigestible detritus. Experiments demonstrated that ducks could avoid ingesting detritus when prey and detritus differ in size. Foraging models based on filter-feeding biomechanics predict prey size selection causes a decline in filtration rates and that the form of this trade-off depends on interspecific differences in bill morphology. To test these models, I used them to predict the results of manipulative foraging experiments on 2 species reported in the literature. There was overall agreement between model predictions and reported differences in filtration rates, particle retention probabilities and ingestion rates, both between species and due to variation in prey size, presence of detritus and surgical manipulation of bill morphology. Extension of these models to five additional species predicts that interspecific variation in the foraging trade-off should result in interspecific partitioning of prey by size when detritus is present. To determine if phenotypic divergence is the result of negative interspecific interactions, I tested for a negative correlation between frequencies of interspecific aggression and phenotypic divergence. Comparison of observed frequencies to the predictions of a null model indicated aggression differed between species pairs. Divergence in body size, body length, lamellar density and divergence of species along a prey size axis predicted by the biomechanical models were all negatively correlated with frequency of aggression, even after controlling for phylogenetic distance. Variation in aggression accounted for by phenotype and phylogeny were additive, indicating ecology and evolutionary history contribute independently to species interactions. These results provide evidence that dabbling ducks represent an adaptive radiation.