Coevolution-induced selection for and against phenotypic novelty shapes species richness in clade co-diversification.

Coevolution can occur because of species interactions. However, it remains unclear how coevolutionary processes translate into the accumulation of species richness over macroevolutionary timescales. Assuming speciation occurs as a result of genetic differentiation across space due to dispersal limitation, we examine the effects of coevolution-induced phenotypic selection on species diversification. Based on the idea that dispersers often carry novel phenotypes, we propose and test two hypotheses. (1) Stability hypothesis: selection against phenotypic novelty enhances species diversification by strengthening dispersal limitation. (2) Novelty hypothesis: selection for phenotypic novelty impedes species diversification by weakening dispersal limitation. We simulate clade co-diversification using an individual-based model, considering scenarios where phenotypic selection is shaped by neutral dynamics, mutualistic coevolution, or antagonistic coevolution, where coevolution operates through trait matching or trait difference, and where the strength of coevolutionary selection is symmetrical or asymmetrical. Our key assumption that interactions occur between an independent party (whose individuals can establish or persist independently, e.g. hosts) and a dependent party (whose individuals cannot establish or persist independently, e.g. parasites or obligate mutualists) yields two contrasting results. The stability hypothesis is supported in the dependent clade but not in the independent clade. Conversely, the novelty hypothesis is supported in the independent clade but not in the dependent clade. These results are partially corroborated by empirical dispersal data, suggesting that these mechanisms might potentially explain the diversification of some of the most species-rich clades in the Tree of Life.