The “placental explosions”

By Razib Khan | February 9, 2013 11:30 pm

Spinger et al. doi: 10.1073/pnas.0334222100

By this time I’m sure you’ve encountered articles about the reconstructed last common ancestor of all placental mammals. Greg Mayer at Why Evolution is True has an excellent review of the implications, along with a link to a moderately skeptical piece by Anne Yoder in Science. Yoder’s piece is titled Fossils vs. Clocks, while the original paper is The Placental Mammal Ancestor and the Post–K-Pg Radiation of Placentals. The results clearly support the “Explosive Model” in the figure to the left for the origination of placentals. That might prompt the thought: “isn’t this what we knew all along?”

The standard story for the last generation in the popular imagination is that a massive asteroid impact was the direct cause of the extinction of all dinosaurs  (and of course a host of other groups) except the lineage which we now term birds. And yet it turns out that there is actually some debate about this, though at least in some form it seems likely that the impact is going to be important (see this Brian Switek piece for exploration of this issue, and the general opinion of the scientific literature as of now). The second aspect to focus on is timing. Contrary to the intuition of many, over the past 20 years molecular phylogenetics has inferred a very definite (on the order of tens of millions of years) pre-K-T boundary coalescence for the common ancestors of the disinct mammalian lineages. A plausible explanation for this is that these lineages diversified through allopatry, as the Mesozoic supercontinent fragmented. Morphological diversification of these mammalian lineages also may have occurred after the K-T event.

The reality is that I know little more about this domain than the “typical person off the street,” so why does this matter? It matters because the difference between model a and model b above impacts our assessment of the nature of the K-T event, and the construction of the “tree of life” in general. For example, if model a, that the diversification of placental mammals was an explosive event after the K-T boundary, with extant lineages derived from one common ancestor, is correct, was the selection of this ancestor a matter of chance, or did it have exaptations which allowed it to traverse the boundary? In contrast, if there were many different mammalian lineages which traversed the boundary than perhaps the nature of the diversification is less contingent? In regards to the K-T boundary issue, it seems on the face of it rather peculiar that anything but extreme circumstances which included the scale of an asteroid impact (perhaps with other necessary conditions). These aren’t trivial questions, and it is within this context that we need to frame our understanding of how life on earth came to be the way it came to be.

An issue which is of more direct interest to me is the long standing conflict between molecular phylogenetics and paleontology. In short, the paleontology seems to support the explosive model. This paper, using morphological traits in a phylogenetic context does just that. In contrast molecular phylogeneticists have long perceived there to be relationships on may genes which pre-date the boundary. So many in fact that it simply isn’t plausible from a genetic perspective that these groups were not reproductively isolated tens of millions of years before the K-T boundary. Yoder points out that the methods of the molecular phylogeneticists is not entirely comparable with that of this paper. And, critically, she suggests that estimation of the branch lengths in the nodes is a weak link within their argument. If you believe that paleontology speaks truly, then you believe in this paper. If you accept the robustness of molecular phylogenetic calibrations of the evolutionary rates of genetic change, than this paper’s dating of the diversification is unpersuasive.

Where do I stand? In more recent questions of human evolution (e.g., the divergence between chimps and humans) the molecular phylogeneticists won, and the bones contingent were wrong. But that is just one lineage. I have a hard time believing that the paleontologists could be wrong about such much. Obviously there’s an error in the assumptions somewhere within these scientific disciplines, but I can’t pinpoint it.

Citation: M. O’Leary et al., DOI: 10.1126/science.1229237

  • Mike Keesey

    Fossil data on its own only tells you the latest possible divergence date. See: The Signor-Lipps effect.

  • Felis

    The gaps between divergence dates estimated from the fossil record and molecular clock studies can be much larger than for chimps and humans, sometimes on the order of 100s of millions of years. The fossil record has faults, but it is not that bad. Plus, once such a gap is produced, the race is on to find the Jurassic elephant for publication in Nature or Science. In the case of placental mammals, paleontologists have been looking for these animals for decades.

    In any case, there are some issues to think about when comparing the fossil record and molecular clock dates. One issue is that most molecular clock studies assume steady rates of evolution along the entire tree when estimating branch lengths. Therefore, estimated dates may be far older than reality when evolutionary rates were highly elevated and major divergences fast and furious early on in the history of the group but eventually slowed way down. This is what is expected to happen after mass extinction events, and is inferred to have occurred for placental mammals here.

    Additionally, molecular clock methods often assume a gradual transition between origin of the total group and origin of the crown, even though large diverse stem-groups may have been filling all the same niches for a prolonged period of time (e.g. multituberculates and mammals; stem-birds and crown aves).

    Another issue is that molecular clock studies also rarely take into account mass extinction boundaries, or the likelihood that over a dozen placental mammal lineages would have survived when their multituberculate relatives were decimated, and other terrestrial faunas reduced to a few survivors.

    Lastly, while fossils are used as calibration points in molecular clock studies, their actual data is often ignored in estimations of rates, with reconstructed ancestors at nodes given full say (even if these diverge from the fossil data – as reconstructed body sizes sometimes do by as much as 8x). This study actually uses fossil groups to reconstruct the ancestors at various points.

    It should also be noted that many theories of vicariant speciation through continental drift have fallen in recent years, as more and more molecular phylogenies have found that divergences post-date the break-up of Pangaea. Animals are also far more mobile than previously realized (one recent study found that it takes only 8 days for garbage to travel from Africa to South America!). Continental drift and allopatry as a explanation for the origin of all major living lineages is on its way out.

    Finally, it should be noted that molecular clock studies do not always agree with each other or produce exact dates. Gaps between mean divergences and 95% confidence intervals within single studies can be tens or hundreds of millions of years long on deep divergences within large clades. Within such ranges, some molecular clock studies do come into agreement with the record. For example, a molecular study by dos Reis et al. 2012 already found crown placentals to be a post K-T radiation, so there is not always strife between the fields!

  • Skadhi_the_Raverner

    If there was an explosion then biogeography poses a huge problem. Where was the common ancestor living? How did its descendants get everywhere so quickly?

    I don’t believe its even possible, never mind probable.

  • Felis

    Well, this is an explosion on a geological time scale – it still took over 5-15 million years (normal timespan for recoveries from mass extinction). Dispersal is easy when there’s little competition in your new land. So, you get lots of founder events, leading to further divergence.

  • timmyRI

    Well, the molecular biologists definitely won the first round with paleontoligists back in the 80s, proving that the man-chimp diverengence wasn’t nearly as old (15-20mya) as thought, however for much of the last decade they were pushing a pretty hard 4 million year mark that paleontoligists resisted in favor of something in the range of 6-7 million years ago and they seem to have been right this time.


Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Gene Expression

This blog is about evolution, genetics, genomics and their interstices. Please beware that comments are aggressively moderated. Uncivil or churlish comments will likely get you banned immediately, so make any contribution count!

About Razib Khan

I have degrees in biology and biochemistry, a passion for genetics, history, and philosophy, and shrimp is my favorite food. In relation to nationality I'm a American Northwesterner, in politics I'm a reactionary, and as for religion I have none (I'm an atheist). If you want to know more, see the links at


See More


RSS Razib’s Pinboard

Edifying books

Collapse bottom bar