Revisiting a Key Innovation in Evolutionary Biology: Felsenstein's "Phylogenies and the Comparative Method"

Abstract

The comparative method has long been a fundamental exploratory tool in evolutionary biology, but this venerable approach was revolutionized in 1985, when Felsenstein published "Phylogenies and the Comparative Method" in The American Naturalist. This article forced comparative biologists to start thinking phylogenetically when conducting statistical analyses of correlated trait evolution rather than simply applying conventional statistical methods that ignore evolutionary relationships. It did so by introducing a novel analytical method (phylogenetically "independent contrasts") that required a phylogenetic topology with branch lengths and that assumed a Brownian motion model of trait evolution. Independent contrasts enabled comparative biologists to avoid the statistical dilemma of nonindependence of species values, arising from shared ancestry, but came at the cost of needing a detailed phylogeny and of accepting a specific model of character change. Nevertheless, this article not only revitalized comparative biology but even encouraged studies aimed at estimating phylogenies. Felsenstein's characteristically lucid and concise statement of the problem (illustrated with powerful graphics), coupled with an oncoming flood of new molecular data and techniques for estimating phylogenies, led Felsenstein's 1985 article to become the second most cited article in the history of this journal. Here we present a personal review of comparative biology before, during, and after Joe's article. For historical context, we append a perspective written by Joe himself that describes how his article evolved, unedited transcripts of reviews of his submitted manuscript, and a guide to some nontrivial calculations. These additional materials help emphasize that the process of science does not always occur gradually or predictably.

Keywords: adaptation; allometry; comparative method; comparative physiology; phylogeny; statistics.