An Evolving View of Character Macroevolution.

Abstract

Phenotypes serve as the interface between organisms and their environments and are thus pivotal for comprehensive biological understanding. However, comparative analyses of species' phenotypes must account for the non-independence of characters imposed by the branching pattern of macroevolution. Methods to account for this phylogenetic non-independence have historically been conceived of as their own subfield: phylogenetic comparative methods (PCMs). In this conceptualization, a researcher takes a pre-existing phylogeny and uses it to correct for the non-independence of the character data that they wish to analyze. Here we argue that parallel developments in scientific philosophy, data availability, computational capacity, and model development have enabled a new paradigm of character evolution, where patterns of character evolution are seen as intrinsically related to the diversification process, and thus should be inferred jointly with the tree rather than reconstructed on an existing phylogeny in a two-step process. In the context of this paradigm shift, we review historical milestones in studies of character macroevolution and discuss major recent conceptual and methodological advances, with an emphasis on the opportunities and insights provided by the joint-inference perspective. We include primers on current topics in character evolution where joint inference is particularly effective, including: (1) state-dependent speciation and extinction models and the importance of cladogenetic change; (2) jointly modeling discrete and continuous characters; (3) accounting for hidden process variation in character evolution; (4) joint inference in divergence-time estimation; (5) joint inference of phylogeny and ancestral states, and; (6) joint inference of alignment and phylogeny. The article concludes with a reflection on the future trajectory of these methods, emphasizing the interconnectedness of character evolution with broader processes in biology.