Abstract
Evolutionary innovation contributes to the spectacular diversity of species and phenotypes across the tree of life. 'Key innovations' are widely operationalized within evolutionary biology as traits that facilitate increased diversification rates, such that lineages bearing the traits ultimately contain more species than closely related lineages lacking the focal trait. In this article, I briefly review the inference, analysis and interpretation of evolutionary innovation on phylogenetic trees. I argue that differential rates of lineage diversification should not be used as the basis for key innovation tests, despite the statistical tractability of such approaches. Under traditional interpretations of the macroevolutionary 'adaptive zone', we should not necessarily expect key innovations to confer faster diversification rates upon lineages that possess them relative to their extant sister clades. I suggest that a key innovation is a trait that allows a lineage to interact with the environment in a fundamentally different way and which, as a result, increases the total diversification-but not necessarily the diversification rate-of the parent clade. Considered alone, branching patterns in phylogenetic trees are poorly suited to the inference of evolutionary innovation due to their inherently low information content with respect to the processes that produce them. However, phylogenies may be important for identifying transformational shifts in ecological and morphological space that are characteristic of innovation at the macroevolutionary scale.This article is part of the themed issue 'Process and pattern in innovations from cells to societies'.