|Steiper and Seiffert, Fig 1. PNAS 2012|
Steiper and Seiffert report in PNAS on a very elegant study of molecular mutation rates throughout primate evolution. The impetus for this study was the discordance between previous date estimates for the origins of primates (about 82 million year ago) and the earliest fossil evidence for true primates (the omomyid Teilhardina, which appears in Europe, Asia, and North America at about 56 million years ago). That is a huge discrepancy of about 26 million years! This could mean that we are missing relevant fossils, or that we are failing to recognize true primates in the fossil record.
The authors of this study argue that, in fact, previous molecular clock dates are systematically incorrect. Previous dates of primate origins have been way too old because they haven't taken into account the fact that primate molecular evolution has convergently slowed down in multiple lineages through geological time. This makes molecular clock estimates which assume a more uniform rate overestimate apparent divergence dates.
But why would evolution slow down through time in multiple lineages, and how do they know? Steiper and Seiffert demonstrated what within primates, rates of molecular evolution are inversely related to body mass, absolute endocranial volume, and relative endocranial volume.
|Steiper and Seiffert, Fig 3. PNAS 2012|
The mechanism by which this works is likely generation time. Big animals with big brains also have longer generation times, so over a given time period there are fewer generations and thus fewer opportunities for the molecular substitutions needed to produce date estimates. Critically, these parameters can be estimated from fossils, so the authors were able to estimate, based on fossil evidence and phylogenetic comparative methods, the rates of molecular evolution in the past. They argue that early primates were mouse lemur sized and small brained, so they would have had fast generation times and fast mutation rates. Primates in multiple groups have gotten bigger and brainer through time, so molecular rates have slowed along with them. Using this combination of molecular data, fossils, and phylogenetic comparative methods, they provide calibrated estimates for all the major events in primate evolution, and their dates for the origin of primates match the fossil evidence quite well!
Thanks to Chris Kirk for bringing this paper to my attention.