DO CLADES HAVE A CARRYING CAPACITY: A NEW PERSPECTIVE FROM (MOLECULAR) PHYLOGENIES

Molecular phylogenies (in fact any phylogeny) offer a fresh and powerful perspective on the debate over whether logistic (indicative of a carrying capacity) or exponential growth best characterizes the process of diversification. Molecular phylogenies can be classified into one of two types, those with constant rates of diversification (type I radiations, indicative of exponential growth), and those with decreasing diversification rates (type II radiations, which imply a carrying capacity). Significantly, about half of molecular phylogenies analyzed to date are type II, implying that density dependence is common. But what about the other half, those that show a type I radiation? At face value these support exponential growth. However, several lines of evidence, including the results of computer simulations and the analysis of the fossil record, suggest that the meaning of type I radiations has been misinterpreted. Rather than indicating a constant rate of diversification, it appears they indicate a constant rate of change in the diversification rate: for type I radiations the underlying rate could be positive (exponential growth), or zero (perhaps as the clade undergoes turnover at its carrying capacity), or negative (indicative of a clade that is decreasing in diversity). Key to understanding the meaning of type I radiations is the use of paleontological data to help determine empirically what proportion of type I radiations really represent exponential growth. For example, for the Cetacea, while its molecular phylogeny suggests exponential growth, its fossil record unequivocally indicates a steady loss of diversity since the mid-Miocene. Finally, we note that even if logistic growth is found to be ubiquitous at lower taxonomic levels, the entire biota could still be diversifying exponentially, as long as the per lineage rate of origin of new clades is constant, and if there is no change in the average carrying capacities of the new clades. Thus, determining the relative importance of logistic versus exponential growth requires a hierarchical approach; the large-scale pattern of diversification is not necessarily reflected in the dynamics of smaller clades.