One of the best new theropod workers out there is Roger Benson. He's been redescribing tons of megalosaurs (e.g. Megalosaurus, Duriavenator, Magnosaurus) and his cladistic analyses are of high quality. As his non-coelurosaur tetanurine analysis incorporates several controversial taxa (e.g. Marshosaurus, Monolophosaurus, Chuanodongocoelurus, Lourinhanosaurus, Shaochilong, Chilantaisaurus, Megaraptor, Fukuiraptor, Aerosteon) and resulted in novel clades (e.g. Neovenatoridae, Megaraptora), I've been interested to see what adding his data to my saurischian supermatrix would do. As a reminder, the saurischian supermatrix takes all the taxa, characters and codings from most of the relevent studies of non-coelurosaur theropods. These include Smith et al. (2007, 2008), Tykoski (2005), Carrano and Sampson (2007), Ezcurra and Novas (2006), Rauhut (2003), Allain (2002), Azuma and Currie (2000), Brusatte and Sereno (2008), Sereno and Brusatte (2008), and now Benson (2010) and Benson et al. (2010). It also has all the data from Langer and Benton (2006) and Yates (2006) to help outgroup comparisons with sauropodomorphs and basal avematatarsalians. Not all included taxa are coded for all characters yet, which is a flaw I noted in several recent analyses (but note I wouldn't publish mine as is), and codings are basically all taken from their matrices and assumed to be correct. The exception to the latter is when codings disagree, so I check them against the literature, and the 37 non-coelurosaur tetanurines which I've gone through the entire matrix with. All together it has 159 taxa and 925 characters. Here's Benson et al.'s (2010) cladogram-
|--"Megapnosaurus" kayentakatae
|--Dilophosaurus
`--+--+--Ceratosaurus
| `--Abelisauridae
`--Tetanurae
|*-Chuandongocoelurus
|--Megalosauroidea
| |--+--Xuanhanosaurus
| | `--+--Marshosaurus
| | `--+--Piatnitzkysaurus
| | `--Condorraptor
| `--+--Monolophosaurus
| `--Spinosauroidea
| |*-Piveteausaurus
| |--Spinosauridae
| | |--+--Baryonyx
| | | `--Suchomimus
| | `--+--Irritator
| | `--Spinosaurus
| `--Megalosauridae
| |--Eustreptospondylinae
| | |--Eustreptospondylus
| | `--+--Streptospondylus
| | `--Magnosaurus
| `--Megalosaurinae
| |--Duriavenator
| `--+--+--Afrovenator
| | `--Dubreuillosaurus
| `--+--Torvosaurus
| `--Megalosaurus
`--Avetheropoda
|--Carnosauria
| |--Sinraptoridae
| | |--+--Lourinhanosaurus
| | | `--Poekilopleuron
| | `--+--Sinraptor
| | `--Metriacanthosaurus
| `--+--Allosaurus
| `--Carcharodontosauridae
| |--+--Eocarcharia
| | `--+--Acrocanthosaurus
| | `--+--Shaochilong
| | `--+--Tyrannotitan
| | `--Carcharodontosaurinae
| | |--Carcharodontosaurus
| | |--Giganotosaurus
| | `--Mapusaurus
| `--Neovenatoridae
| |--Neovenator
| `--Megaraptora
| |*-Orkoraptor
| |--Chilantaisaurus
| `--+--+--Australovenator
| | `--Fukuiraptor
| `--+--Aerosteon
| `--Megaraptor
`--Coelurosauria
|--Compsognathus
`--+--Tanycolagreus
`--Guanlong
Within his trees, Chuandongocoelurus is excluded from Avetheropoda, Spinosauroidea and the Xuanhanosaurus+Piatnitzkysaurus clade. Piveteausaurus is excluded from Spinosauridae. And here's the results of my analysis once pruned (after running it) to include only those taxa present in Benson et al.'s matrix.
|--+--"Megapnosaurus" kayentakatae
| `--Dilophosaurus
`--+--+--Ceratosaurus
| `--Abelisauridae (9 taxa)
`--Tetanurae
|--Chuandongocoelurus
`--+--Xuanhanosaurus
`--+--+--Marshosaurus
| `--+--Piatnitzkysaurus
| `--Condorraptor
`--+--Megalosauroidea
| |--Poekilopleuron
| `--+--+--Piveteausaurus
| | `--Spinosauroidea
| | |--Torvosaurus
| | `--Spinosauridae
| | |--+--Baryonyx
| | | `--Suchomimus
| | `--+--Irritator
| | `--Spinosaurus
| `--Megalosauridae
| |--Afrovenator
| `--+--Megalosaurus
| `--Eustreptospondylinae
| |--Dubreuillosaurus
| `--+--Duriavenator
| `--+--Eustreptospondylus
| `--Magnosaurus
`--+--Monolophosaurus
`--+--+--Streptospondylus
| `--Lourinhanosaurus
`--Avetheropoda
|--Carnosauria
| |--Sinraptoridae
| | |--Sinraptor
| | `--Metriacanthosaurus
| `--+--Allosaurus
| `--Carcharodontosauridae
| |--Neovenator
| `--+--Eocarcharia
| `--+--Acrocanthosaurus
| `--Carcharodontosaurinae
| |--Shaochilong
| |--Tyrannotitan
| |--Carcharodontosaurus
| |--Chilantaisaurus
| `--Giganotosaurini
| |--Giganotosaurus
| `--Mapusaurus
`--Coelurosauria
|--Megaraptora
| |--Fukuiraptor
| `--+-Orkoraptor
| `--+--Australovenator
| `--+--Aerosteon
| `--Megaraptor
`--+--Guanlong
`--+--Compsognathus
`--Tanycolagreus
Chuandongocoelurus was identified as an elaphrosaur by me back on the DML several years ago, and as a megalosauroid related to Monolophosaurus in Benson's original (2008, 2010) analysis. Now Benson finds it to be a basal tetanurine of uncertain placement, while I place it as the basalmost tetanurine. Why the difference? The main one is that Benson only codes the holotype, which includes mostly hindlimb elements. Most of the Elaphrosaurus-like characters are in the vertebrae, which belong to the paratype. I coded both as separate OTUs and found the paratype to be sister to Elaphrosaurus.
Benson finds a clade consisting of Xuanhanosaurus, Piatnitzkysaurus, Condorraptor and Marshosaurus at the base of Megalosauroidea. The first three have traditionally been placed outside Megalosauroidea + Avetheropoda, while Marshosaurus has been controversial. My results are intermediate. All of these taxa are outside Megalosauroidea + Avetheropoda, but the latter three do form a clade with the same topology found by Benson. In my last post I wondered if Marshosaurus' position would be affected by including pelvic characters, and it seems the answer is no. One important question is if the snout, pelvis and partial skeleton belong to the same taxon. I may try running them as separate OTUs sometime. It should be noted that in my analysis Kaijiangosaurus, "Szechuanoraptor" and "Sezchuanosaurus" zigongensis form a clade with Xuanhanosaurus, while Shidaisaurus is sister to Marshosaurus. Erectopus is more basal than any of these, while Tugulusaurus is a tetanurine outside of Avetheropoda or Megalosauroidea.
Among the core megalosauroids, our topologies are different besides Spinosauridae, Megalosauridae (though mine excludes Torvosaurus) and placing Magnosaurus and Eustreptospondylus close. There are a few taxa which have varied between megalosauroids and carnosaurs in analyses. Monolophosaurus has traditionally been placed as a basal carnosaur, but Benson found it to be a megalosauroid. My analysis agrees with Smith et al.'s in placing it sister to Avetheropoda. Poekilopleuron is usually thought to be close to Megalosaurus, though Benson found it to be a sinraptorid carnosaur. My analysis agrees with Allain and Chure (2002), who proposed it may be a basal megalosauroid. Lourinhanosaurus has been found to be both a carnosaur (Mateus, 1998; Holtz et al., 2004; Benson et al., 2010) and a megalosauroid (Allain, 2002), but my analysis is intermediate in placing it sister to Avetheropoda. Finally, Streptospondylus has been generally viewed as a eustreptospondyline (Allain, 2001; Smith et al. 2007; Benson et al., 2010), but in my analysis it is sister to Lourinhanosaurus instead.
In Carnosauria, everyone has been focusing on carcharodontosaurids to the exclusion of more basal taxa (e.g. Yangchanosaurus, Saurophaganax). But my analysis does agree with the growing consensus that sinraptorids are basal to Allosaurus and carcharodontosaurids. Both Benson and I found that Metriacanthosaurus is sister to Sinraptor, providing yet more evidence this clade should be called Metriacanthosauridae instead of Sinraptoridae. Our topology within Carcharodontosauridae is basically identical, though mine is less resolved toward the tip due to Chilantaisaurus' presence. Notably, Siamotyrannus was resolved in my analysis as a carnosaur more derived than sinraptorids, but less than Allosaurus.
The main point of Benson et al.'s (2010) paper was the discovery of a new clade of carnosaurs- Neovenatoridae. This contained the basal carcharodontosaurid Neovenator, as well as the Megaraptora. Megaraptora contains many taxa which have been seen as spinosaur relatives (Megaraptor, Chilantaisaurus), carnosaurs (Chilantaisaurus, Fukuiraptor, Australovenator, Aerosteon, Megaraptor) or coelurosaurs (Chilantaisaurus, Fukuiraptor, Orkoraptor, Megaraptor) in the literature. The relationship between Australovenator and Fukuiraptor was noted in the former's description, while I noted the resemblence between Orkoraptor and Aerosteon (I viewed all four of these taxa as basal coelurosaurs). But combining all these genera into a carcharodontosaur clade was unexpected. My analysis agrees with Benson in the monophyly of Megaraptora (except Chilantaisaurus) and placing Megaraptor and Aerosteon as sister taxa within it. However, it places Megaraptora in Coelurosauria, with Neovenator and Chilantaisaurus as carcharodontosaurids. Within Coelurosauria, Megaraptora is outside Tyrannoraptora, but more derived than Gasosaurus.
Next time... the first of many Pickering taxa.
Roger Benson, not Robert.
ReplyDeleteYeah, his name is Roger... it's still a cool thought, Mickey. I'll be curious to see how this evolves with more taxa completely coded...
ReplyDeleteToo bad neovenatoridae seems to be just a wrong end of the stick... hope im wrong
ReplyDelete