A topic in the Open Knowledge Graph — a free, open map of 15,290 topics and the order to learn them in.
Comparative phylogenetic methods use evolutionary trees to test hypotheses about trait evolution, adaptation, and diversification. Phylogenetically independent contrasts (PIC) correct for non-independence of species due to common ancestry. These methods reveal correlated evolution of traits and identify drivers of speciation and ecological divergence.
From phylogenetics and cladistics, you know how to build evolutionary trees that depict the branching relationships among species. From your study of adaptation, you understand that natural selection shapes traits to fit ecological demands. Comparative phylogenetic methods sit at the intersection of these ideas: they use the tree itself as an analytical framework to ask rigorous questions about how and why traits evolve. The central problem these methods solve is deceptively simple — species are not independent data points.
Consider a classic question: do larger-bodied mammals have larger brains? You could plot brain size against body size for 100 mammal species and run a correlation. But this analysis treats each species as an independent observation, which it is not. All primates share a recent common ancestor who probably already had a relatively large brain, so including 20 primate species in your dataset is not like having 20 independent tests of the brain-body relationship — it is largely measuring the same evolutionary event 20 times. Phylogenetically independent contrasts (PIC), introduced by Joseph Felsenstein in 1985, solve this problem by transforming species data into contrasts calculated at each node of the phylogeny. Instead of comparing species directly, you compare sister lineages at each branching point — each contrast represents an independent evolutionary divergence. The correlation is then run on these contrasts, and the result reflects genuinely independent evolutionary changes rather than shared ancestry.
Beyond correcting for non-independence, phylogenetic methods enable powerful tests of evolutionary hypotheses. Ancestral state reconstruction uses the tree and the traits of living species to estimate what extinct ancestors looked like — did the common ancestor of whales and hippos live on land or in water? Correlated evolution analysis tests whether two traits tend to change together across the tree — do species that evolve bright coloration also tend to evolve toxicity? Models of trait evolution compare whether traits evolve by random drift (Brownian motion), are pulled toward an optimum (Ornstein-Uhlenbeck model), or show bursts of change associated with lineage splitting. Each model makes different predictions about how trait variation should be distributed across the tree, and statistical model comparison identifies which evolutionary process best explains the observed pattern.
These methods also illuminate macroevolutionary dynamics — the tempo and mode of diversification itself. Diversification rate analysis tests whether certain lineages speciate faster or go extinct less often, and whether these rate shifts correlate with the evolution of key traits (like the evolution of flowers in angiosperms or flight in bats). Phylogenetic signal statistics (like Pagel's λ or Blomberg's K) quantify how much of the variation in a trait is predicted by phylogeny versus independent ecological factors. A trait with high phylogenetic signal (close relatives are similar) behaves differently from one with low signal (trait value is driven by current ecology regardless of ancestry). Together, these tools have transformed comparative biology from a descriptive exercise — cataloguing similarities and differences — into a hypothesis-testing discipline that can distinguish adaptation from phylogenetic inertia, convergent evolution from shared ancestry, and evolutionary constraint from ecological opportunity.