Science —

Re-re-re-thinking the rise of mammals and death of the dinosaurs

Placental mammals may have waited for the asteroid to hit after all.

Grandma? An artist's conception of the appearance of the first placental mammal.
Grandma? An artist's conception of the appearance of the first placental mammal.
Carl Buell

The end of the era of dinosaurs and rise of the mammals has held a fascination for me since I was young (and I suspect I'm not alone). But it's a tale that has been retold many times now.

In the earliest versions, I recall ideas about clever, fast-moving mammals outcompeting the slow, lumbering dinosaurs. With time, however, that story changed. The dinosaurs became quicker and actually survived, albeit as birds (papers now refer to the loss of "non-avian dinosaurs"). The mammals became less clever and more lucky, in that it took a freak hit from an asteroid to trigger the mass extinction. As DNA data came in, the amount of luck involved seemed more and more significant. Data indicated some of the lineages of modern placental mammals had been around for millions of years before the dinosaurs died, but didn't really do much until after the extinction event.

A new analysis, published in Science now pushes back against the molecular data. A large team of authors tracked thousands of individual traits in more than 125 species (40 of them known only from fossils) to build the biggest reconstruction of the history of mammals ever attempted. In doing so, they find the first placental mammal probably didn't exist until after the non-avian dinosaurs were gone, the study even provides some hints of what it might have looked like.

Reconstructing the history of mammals is challenging for a lot of reasons. To begin, their history in the fossil record is incomplete. For example, we know the lemurs are related to lorises, but diverged before their home island, Madagascar, was separated from nearby continents. However, there simply isn't any hint of lemurs in the fossil record; they were what's termed a "ghost lineage" for tens of millions of years.

Other complications come from the fact that the first big radiation of placental mammals looks a bit like an explosion. It took millions of years for ecosystems to recover after the death of the dinosaurs but, once they did, new species and major groups of mammals appeared in quick succession, making the precise order in which new groups arose difficult to determine.

All of which made DNA work rather challenging. The timing of some splits can be estimated using the degree of difference between the DNA of existing species, but there still needs to be something to anchor these "molecular clocks." In most cases, that involves finding a split in the fossil record that we can assign a precise date to. That's difficult when the fossils tell a somewhat confused tale.

The new tree is informed by molecular data—the authors now have more than 25 mammalian genomes to work with—but is mostly focused on morphological traits (the shapes of bones and organs). Although individual traits can get confused as some are lost and others evolve independently, a large catalog like the one used by the authors can provide a clear picture of how species relate to each other. Reaching back through fossils, it can also provide an indication of when different species branched off.

A mammal that dates to the Cretaceous and has a mix of marsupial and placental traits.
Enlarge / A mammal that dates to the Cretaceous and has a mix of marsupial and placental traits.
AMNH/S. Goldberg and M. Novacek

The authors conclude some mammalian lineages are quite old. Monotremes like the platypus and echidna, for example, may go back roughly 200 million years to the Triassic. Marsupials split off in the Jurassic. There are a few extinct lineages that are more closely related to us than marsupials, but the new tree clearly excludes them from the placentals (and none of them survived the mass extinction). The ancestor of the placentals seemed to evolve within a couple of million years of the mass extinction, and the diversification really got going about five million years after.

Because the tree is based on shared traits, the authors were able to reconstruct what the ancestral species probably looked like. These include some remarkable details, including the paths traveled by specific nerves (which can be preserved on bone surfaces) to major features like the presence of large regions of the brain devoted to smell, and the fact that males carried their testes inside their abdomen. More generally, the animal was likely to be a small insectivore, looking a bit like a hybrid of a North American opossum and a rat.

The unexpected aspect of this late arrival on the scene is the first placentals appeared after Pangea was well on its way to breaking up, meaning mammals would have had to cross some significant bodies of water. It also means some of them moved around quite a bit. Fossil evidence, for example, indicates the Afrotheria, a group that includes animals like elephants and aardvarks, probably got its start in North or South America, although later extinctions removed most members of the group from those continents.

Although this study is impressively comprehensive, it shouldn't be viewed as the last word. There's still some work to do to get this tree in better accordance with the DNA data, which itself is changing with the development of more sophisticated methods for generating a molecular clock. Meanwhile, because of the importance of fossil data to the tree, a significant find could potentially alter its shape. The story of mammals may end up being retold again.

Science, 2013. DOI: 10.1126/science.1229237  (About DOIs).

Channel Ars Technica