Will a Jurassic placental mammal make the molecular clock make sense?

A new paper in Nature by Zhe-Xi Luo and colleagues Luo:Jurassic:2011 reports the discovery of a 160-million-year-old early mammal, Juramaia, which they attribute to the placental mammal lineage. The news aspect is that this extends the chronology of fossil placental and marsupial mammals (the sister clade of placentals) by some 40 million years. That's a big chunk of time, but it's a really nice fossil which seems pretty clear in its morphology.

I'm reading this closely because of the effect on the interpretation of mutation rates and the molecular clock. Obviously, if the earliest evidence for placental mammals used to be 120 million years ago, and now it's 160, that should affect the way we approach the genetic divergence of mammal lineages. In particular, when it comes to primates, some modern lineages are represented by fossils relatively early in the Cenozoic, suggesting that the common ancestor of all the primates may have been much earlier, deep in the Cretaceous period. But there is no fossil evidence of that ancestor, and until recently molecular comparisons seemed to suggest a recent chronology with a common ancestor just before the Cretaceous-Tertiary (K-T) boundary. That is, until direct estimates of the human mutation rate started to suggest a much lower rate of mutations per generation than had previously been assumed.

I've written about these issues several times, both with respect to hominins and other primates. For example my (unfinished) series from 2010:

"Were there Cretaceous anthropoids? Part 1. The problem in a nutshell"

"Were there Cretaceous anthropoids? Part 2: What is an anthropoid?"

"Were there Cretaceous anthropoids? Part 3: Ghost lineages"

And last month's "More on the mutation rate", pointing to my review from late last year, "What is the human mutation rate?" It's a key scientific problem right now, and genetic evidence may be approaching the point of a solution. Finding older and older fossils tends to confirm a lower rate of mutations, and a long chronology for the extant lineages.

The current paper by Luo and colleagues addresses the molecular clock and suggests how a 160-million-year-old placental mammal may affect things:

Timing of the divergence of marsupials and placentals is critical for calibrating the rates of evolution in therian mammals, especially for molecular evolutionary studies and comparative genomics 2, 10, 13. Previously, some molecular time estimates for marsupial and placental divergence postulated significantly older windows for this divergence than the then-oldest fossil records3, 7. However, these and other previous molecular estimates differed widely. Several were compatible with relatively young placental intraordinal divergences (for example, ref. 10), and just about all showed wide error margins (reviewed by ref. 13). Regarding the marsupialplacental split, recent molecular rate studies provided estimates of 147.7??5.5?Myr (ref. 11), or 160?Myr (median) with a 95% highest posterior distribution of 143178?Myr (ref. 12), or a window of 193186?Myr (ref. 9). This new eutherian fossil age is now similar to the age of placentals at 160?Myr with 95% posterior distribution from 143 to 178?Myr by the latest molecular estimate12. The age of Juramaia has now set the minimal divergence time by the fossil to coincide with the range of molecular time estimates, serving as a corroboration of the newest fossil record with the molecular clock of evolution. The 160-Myr-old Juramaia also has important implications for mammalian evolution as a whole. Eutherian mammals are nested in the more inclusive Mesozoic boreosphenidan clade (Fig. 3, node 1), for which the previously earliest record had been entirely Early Cretaceous1, 27. The eutherian Juramaia requires that the ghost-lineages of boreosphenid and cladotherian mammals would also extend to the Middle Jurassic. Therefore the magnitude of the mammalian faunal turnover from the Early to Middle Jurassic is greater than previously known, and the EarlyMiddle Jurassic is a critical transition for the appearance of more of the derived mammalian clades1, 2.

Reference 10 from that quote is a paper by Kitazoe and colleagues (2007) Kitazoe:2007. In that paper, the divergence of New World and Old World monkeys is up over 55 million years ago, and the divergence of anthropoids and strepsirrhines was around 85 million years. In other words, that "fast" chronology paper predicted anthropoids at or around the K-T boundary. Reference 11 is by Bininda-Emonds and colleagues (2007) Bininda-Emonds:2007, in which primates were estimated to have originated 91 million years ago, with a haplorhine-strepsirrhine divergence 87 million years ago. The other references here don't discuss within-primate divergences together with the more ancient mammalian representatives. I discussed more focused comparisons of primate divergences last year ("Were there Cretaceous anthropoids? Part 1. The problem in a nutshell").

It looks to me like an earlier origin of placental mammals will elevate the likely divergence dates for primates to some degree, which will make a difference to interpretation of fossils like Altiatlasius or Algeripithecus. I think it's consistent with a lower mutation rate within the hominoids also, but it's unclear whether we need the within-family rate of change to be consistent with the longer term rate of change among orders of mammals.