Abstract
Mitochondrial DNA remains one of the most widely used molecular markers to reconstruct the phylogeny and phylogeography of closely-related birds. It has been proposed that bird mitochondrial genomes evolve at a constant rate of ~0.01 substitution per site per Million years, i.e. that they evolve according to a strict molecular clock. This molecular clock is often used in studies of bird mitochondrial phylogeny and molecular dating. However, rates of mitochondrial genome evolution vary among bird species, and correlate with life-history traits such as body mass and generation time. These correlations could cause systematic biases in molecular dating studies that assume a strict molecular clock. In this study, we overcome this issue by estimating corrected molecular rates for birds. Using complete or nearly complete mitochondrial genomes of 475 species, we show that there are strong relationships between body mass and substitution rates across birds. We use this information to build models that use bird species' body mass to estimate their substitution rates across a wide range of common mitochondrial markers. We demonstrate the use of these corrected molecular rates on two recently-published datasets. In one case, we obtained molecular dates that are twice as old as the estimates obtained using the strict molecular clock. We hope that this method to estimate molecular rates will increase the accuracy of future molecular dating studies in birds.