How predictable is evolution?

Imagine 500 to 1,000 species of cichlids living in one of the African Great Lakes, one of the largest freshwater habitats in the world. The degree of complexity is unimaginable. Even the genealogical relationships between the cichlid species living in these lakes have only partially been resolved

For every two species of mammal there is one species of cichlid fish, which shows that biodiversity is distributed rather unevenly among animals. The question is why and to what extent can this be predicted? We know that a number of factors play a role in the evolution of this. One of them are ecological conditions, i.e. the number of different habitats and the similarity of ecological niches available. The demographic history of a population can also influence biodiversity. A higher level of genetic variation in a population is beneficial in a sense that it allows - given sufficient time - adaptation to more ecological niches. Quantifying all potential factors that contribute to biological diversity, even for only one group of animals, is not easy, not to mention that comparing mammals with a group of fish would be like comparing apples and oranges. 

A new study coming from the lab of my PhD supervisor, Axel Meyer shows some of the factors that contribute to recurrent patterns of diversity and similarity in cichlids. The colleagues aimed to determine factors that led to similar outcomes and thereby help to predict evolution. As any African Great Lake harbours an incredibly species diversity, the team studied a more simple system involving parallel species of Midas cichlids, which occur in two great lakes as well as in a chain of crater lakes in Nicaragua. 

The more similar the habitat of the crater lake is to that of the large source lake, the more similar the fish are to each other. Habitat seems to be more important than demographic criteria when it comes to predictability of diversity.  The data collected by the colleagues also shows that, compared to the source population, the morphology of all crater lake populations has diversified mostly in the same direction. Crater lake fish all very quickly evolved body shapes that are longer and more slender than those of their cousins from the bigger lakes. In addition body shapes of the crater lake populations correlate with the average depth of the lakes. It makes sense. The deeper a lake is, the more likely it is to provide various ecological niches, including in the deep open water. 

In summary parallel morphological divergence in allopatry and the propensity for diversification in sympatry across the entire Midas cichlid fish radiation is partly predictable and mostly driven by ecology.