Molecular Clock Calibration in Vertebrates

Abstract

Molecular clock dating -- inferring the timing of evolutionary divergences from molecular sequence data calibrated against the fossil record -- underpins our understanding of vertebrate diversification, biogeographic history, and the tempo of macroevolution. However, the accuracy of divergence time estimates depends critically on the quality, placement, and interpretation of fossil calibrations, and systematic disagreements between molecular and palaeontological timelines for major vertebrate clades persist despite decades of methodological refinement. This study conducted a comprehensive re-evaluation of molecular clock calibration quality for 14 major vertebrate nodes using 847 published molecular phylogenetic studies (1990-2024), 312 fossil calibration constraints drawn from the Paleobiology Database and FossilCalibrations repository, and Bayesian total-evidence dating analyses on concatenated alignments of 248 nuclear protein-coding genes from 847 vertebrate species. Calibration constraint quality was assessed using the CladeAge fossil probability density approach and ranked by minimum age reliability, clade completeness, and biogeographic ambiguity. Revised divergence time estimates for 14 key vertebrate nodes showed mean uncertainty reduction of 28.4% relative to prior published estimates from single-calibration analyses. The amniote crown origin was estimated at 317.4 +- 8.4 Ma (Carboniferous), reptile-mammal split at 312.8 +- 7.2 Ma, and bird-crocodilian divergence at 247.4 +- 12.8 Ma (Triassic). Substitution rate heterogeneity among vertebrate lineages was substantial (coefficient of variation: 0.84), confirming the necessity of relaxed clock models over strict clock assumptions. These calibrated divergence time estimates and the associated calibration quality assessments are released as a standardised reference resource for vertebrate phylogenetics.