RPM and Kambiz comment on a paper which argues for the utility of mtDNA in phylogeography.
Remember,
- there is lots of mtDNA that is easy to extract because there are so many mitochondria within eukaryotic cells
- it mutates fast, building up a lot of genetic variation, in a neutral matter (i.e., perfect "clock")
- it is uniparental, passed from mother to daughter, making the coalescent model congenially tractable
It's not so much the phylogentic utility of mtDNA that's in question, it's whether mtDNA polymorphism adequately captures population size changes.
it mutates fast, building up a lot of genetic variation, in a neutral matter
But that's what's being called into question! Any selection at any site in the molecule will affect polymorphism everywhere else on the molecule.
It's not that mtDNA is a poor locus for popgen analysis. mtDNA is like any other locus. We need to sample multiple independent loci to understand demographic history.