Sexual selection does not influence minisatellite mutation rate

Moller and Cuervo report a significant trend between minisatellite mutation rate and the frequency of extra-pair copulations in birds. This is interpreted as evidence that the high rate of evolution demanded by sexual selection has itself selected for a higher mutation rate in species where selection is strongest. However, there are good a priori reasons for believing that their method of calculating minisatellite mutation rates will be highly error prone and a poor surrogate measure of the evolutionary rate of genes. I therefore attempted to replicate their results using both their data and an independent data set based on papers they failed to locate.

I find that Moller and Cuervo's data set contains numerous errors that act somewhat to strengthen their key regression. More importantly, data from uncited papers fail to replicate their reported trend and one species in particular, Vireo olivaceus, is apparently deliberately omitted, yet its inclusion removes significance from the original correlation. Over the small number of cases were comparisons can be made, mutation rate estimates do not differ between species but do vary significantly depending on the laboratory/operator.

There appears to be no clear relationship between minisatellite mutation rate and EPC rate in birds. The previously reported trend can be attributed to data transcription errors and unfortunate data selection. My analysis highlights the importance of total methodological transparency when conducting meta-analyses.