PubMed · 2026-06-12
Scientists are proposing a major shift in how we study how plant and animal traits evolve over millions of years — instead of treating the family tree of life as a fixed backdrop, they argue we should figure out the tree and the traits together, simultaneously, because each shapes the other.
Traditional phylogenetic comparative methods treat the evolutionary tree as fixed and separate from trait data, but this two-step approach misses important feedback between trait evolution and species diversification.
Joint inference — simultaneously estimating the evolutionary tree and how traits changed along it — produces more accurate reconstructions, especially for understanding speciation and extinction linked to specific character states.
Six major methodological frontiers are identified where joint inference is most powerful: state-dependent diversification, mixed discrete/continuous traits, hidden process variation, divergence-time estimation, ancestral state reconstruction, and alignment-plus-phylogeny inference.