Sunday, August 9, 2009

Two Cent History Tour of Sebecosuchia

Sebecosuchia was first named by Edwin (not Stephen) Colbert for Sebecus and Baurusuchus (Colbert 1946). Turner and Calvo (2005) recognized Stratiotosuchus, Ayllusuchus, Ilchunaia, Bretesuchus, Bergisuchus, Pabwehshi, Iberosuchus, Eremosuchus and their new taxon, Pehuenchesuchus as being sebecosuchians. Their analysis put placed Sebecosuchia as the sistergroup to the taxa we usually would call 'notosuchians' (though technically, the phylogenetic definition of Notosuchia would cover sebecosuchians as well). They did not recognize a distinct division of the Sebecosuchia into two separate clades (e.g. baurusuchids and sebecids), but instead for 'sebecids' to form a paraphyletic assemblage to the Baurusuchidae (incl. Baurusuchus, Pabwehshi, and Bretesuchus).

Larsson and Sues (2007) offered a very different arrangement in which sebecids (represented by Sebecus and Bretesuchus) were the sistergroup to the Peirosauridae forming a clade that they called "Sebecia". "Sebecia" was the sistergroup to the Neosuchia, and Baurusuchus was more basally placed relative to these groups, as was Araripesuchus, and the notosuchians (only represented by Notosuchus and Malawisuchus). They also offered two new phylogenetic definitions for the Sebecidae and the Peirosauridae which would not be very useful in other topologies since both were basically defined to exclude each other (e.g. Sebecidae as Sebecus but not Uberabasuchus or Peirosaurus and Peirosauridae as Uberabasuchus and Peirosaurus but not Sebecus), and in current more global analyses of crocodyliforms (where sebecids are derived members of a clade with notosuchians and peirosaurids are basal neosuchians), it would effectively make all notosuchians to be 'sebecids' and all neosuchians to be 'peirosaurids'. It seems very strange to me that Larsson and Sues (2007) also found Stolokrosuchus (an odd longirostrine form) to be a peirosaurid as well, this taxon is otherwise a basal neosuchian (Turner and Buckley 2008), but more on Stolokrosuchus later...since you can count on more of me giving my two-cents on crocodyliforms to come in the future.

Wargosuchus and Cynodontosuchus are two other baurusuchids which haven't been treated within a phylogenetic analysis to my knowledge, but are both from the Late Cretaceous of Argentina. And of course, there are the three Eocene Venezuelan sebecids... Langstonia, Zulmasuchus, and Barinasuchus. Knowing more about how these taxa really fit in would be good. The same goes for the poorly known trematochampsids for which we have no idea if they have anything to do with sebecosuchians or if they're closer to peirosaurids or something entirely different.



AddThis Social Bookmark Button Subscribe to Feed

2 comments:

Ville Sinkkonen said...

Sebecid could do with a major revision. I'm kinda worried that some of the characters used may just simply be convergence due to boxy skull and ziphodont dentition.
I think these studies could greatly benefit on comparing all the known zipho crocs (everything from Pristichampsus to some Mekosuchines)and actually see which of the features are just adaptations to that particular skull form.

BTW have you played with any of these recent croc phylogenies? Which charachters support the Notosuchia?

Nick Gardner said...

It's entirely possible, most of these taxa are rarely sampled in recent croc phylogenies. The only major tweaking I've tried to do is merging Turner and Buckley (2008) with Pol et al. (2009), unfortunately, I have trouble running their data sets in PAUP (with and without modifications), I never seem to be able to recover their topology (I suspect PAUP gets bogged down).

There's a few key papers I'd like to read (e.g. Pol's recent one on Dakosaurus) but haven't had the time. There are definitely some areas that the AMNH matrix is weak in, and that would be within the crown clade. The position of those taxa seems highly affected by what the conditions are throughout the rest of the tree (specifically the topology of basal Neosuchia wrt peirosaurids).