A paper you may have missed

Images courtesy of Fernando Abdala

Today I’m showcasing a little paper on some very intriguing little teeth (shown above) from the Triassic of South Africa that appeared last year (Abdala et al. 2007). It didn’t make much of an impact partly because of its location (the local ‘South African Journal of Science’) and perhaps also because no firm conclusions could be reached. Nevertheless it deserves comment because whatever the teeth turn out to belong to they are significant, indeed they have the potential to be extremely significant.
The teeth in question were recognised when Helke Mocke, then an honours student at the BPI who was being supervised by Fernando Abdala, was attempting to find isolated cheek teeth of Bauria, a therocephalian therapsid with some striking convergences with mammals. Helke was attempting to test the hypothesis that Bauria was an herbivore by looking for dental microwear on the cheek teeth. Most of the Bauria in our collections either have the lower jaws tightly clamped against the upper jaws making it impossible to see the chewing surfaces of the teeth, or had the teeth mechanically prepared out of hard matrix with airscribes and pins, potentially adding a whole new set of scratches and microwear that had nothing to do with the diet of the animal. Bauria comes from the middle part of the Burgersdorp Formation of the Karoo Basin. This part of the formation is thought to be latest Early Triassic in age and is also the source of such well known extinct animals as Erythrosuchus, Cynognathus and Kannemeyeria. It is here that I play a very minor role in the story. I suggested to Helke that she might find good Bauria teeth in the older parts of the Burgersdorp Formation. The older parts (known informally as the ‘A zone’) are generally rich in aquatic or subaquatic taxa such as lungfish, hybodontid sharks and temnospondyl amphibians but a rich channel lag deposit found by palaeontologist/geologist John Hancox is crammed with all sorts of small vertebrate fossils, including some terrestrial creatures. Amongst them were teeth that had been identified as bauriid. These could be isolated by simple surface picking at the site, or screen washing in the lab. So Helke proceeded to look at the A zone bauriid teeth with a microscope. As it turned out this was the first time anyone had given these teeth anything more than a cursory glance. Surprisingly they weren’t bauriid teeth at all. Even more surprisingly the crowns of these teeth resembled nothing more than those of early mammals called haramiyidans. Compare the A zone teeth featured above with those of a haramiyidan below)

Haramiyavia, from Jenkins et al. 1997.

Haramiyidans themselves are an enigmatic group known only from teeth and incomplete jaws (apparently there are some postcranials of Haramiyavia but these are rather uninformative). Their cheek teeth bear double rows of small cusps, similar to those of multituberculates, which they are often grouped together with in the clade Allotheria. Multituberculates have a derived therian-like shoulder girdle, indicating that they share a more recent common ancestor with them than monotremes or other major Mesozoic mammal groups like eutriconodonts, docodonts and morganucodonts. The earliest haramiyidans are from the late Norian Stage of the Late Triassic. If they are truly early relatives of the multis then they would push back the origin of crown group mammals, and even more significantly advanced theriimorph mammals, to a time 45 million years or more before the next oldest appearance of this clade in the fossil record. If the A zone teeth are also regarded as allotherian then we have the appearance of theriimorph mammals before the appearance of probainognathians, at a time when only the earliest eucynodonts had just made their appearance. Clearly this is strongly at odds with the known fossil record. At this point I think there are three main hypotheses that can explain these teeth. I order them here from what is in my opinion the most likely to the least likely. 1) The A zone teeth are convergently similar to haramiyidans and do not belong to a mammal at all. 2) The A zone teeth are early haramiyidans but haramiyidans are themselves an early branch of cynodonts and are not mammals. 3) The A zone teeth are haramiyidans and haramiyidans are allotherian mammals. This last hypothesis implies a huge stratigraphic debt, with ghost lineages for chiniquodonts, probainoganthids, trithelodonts, tritylodonts, morganucodonts, Sinoconodon, docodonts, Hadrocodium, Kuehneotherium, australosphenidians and stem trechnotherians all extending back to the Early Triassic.
Multicusped rasping teeth have evolved multiple times in synapsid evolution with tritylodontids, multituberculates and ektopodontids all being independant examples (I am sure there are more). The A zone teeth are single rooted (like basal cynodonts but unlike almost all advanced cynodonts including mammals and haramiyidans) which supports hypothesis 1.
If they are only convergently mammal-like then what are they? We can't say now but I like the idea that maybe they are derived bauriids afterall. The answer will be forthcoming from the next field trip, I hope.
Oh, we still don't know what Bauria ate.

References

Abdala F, Mocke H, Hancox PJ (2007) Lower Triassic postcanine teeth with allotherian-like crowns. South African Journal of Science 103: 245-247.

Jenkins FA Jr, Gatesy SM, Shubin N, Amaral WW (1997) Haramiyids and Triassic mammalian evolution. Nature 385: 715-718.