This research investigated the cause of parallel evolution in certain fishes inhabiting the rivers of tropical South America and Africa. In the 90 million years that have passed since separation of those continents, the fishes of each continent evolved one group with astounding morphological and ecological diversity, and another closely related group of detritus-eaters that are all very similar in morphology and ecology. Research on these fishes synthesized measurements of morphological diversity, evolutionary simulations, trees-of-life, and ecological information to compare and explain the parallel evolutionary histories of the two faunas. The main findings were that 1) on each continent a different innovation in the structure the lower jaw appears to have catalyzed the evolution of many further changes in skull structure and feeding ecologies and 2), detritus-eaters are not necessarily more constrained in their ability to evolve novelty than are fishes that eat other types of food, such as invertebrates. This research also developed phylomorphospace analysis, a new technique that combines trees-of-life and scatterplots of morphological diversity to test for differences in the underlying evolutionary process in different groups of related organisms. Other products included the construction of a tree-of-life for more than 200 species of fishes from the union of more than a dozen morphological datasets, the description of a new species of South American headstander, and collaborative papers applying the phylomorphospace approach to other groups of organisms, including triggerfishes and 300 million year old fossil tetrapods.