I have for a while been intrigued by sauropods, and how Atlantosaurus related to Apatosaurus. Of course because the former would be the senior synonym. But I didn’t want to code it into a large analysis. So I decided to place it in a purely sacral-based analysis. So I (somehow, honestly I’m not sure how, but I think it was a trial version) got PAUP* for free and added in my matrix. In the analysis, with the matrix based on my observations of photographs, with characters after multiple authors including Tschopp et al..
There were 25 characters, and 13 taxa, so not exactly the most comprehensive analysis. The outgrip was hypothetical, and all zeros. Taxa included were Cetiosaurus, Futalognkosaurus, Brachiosaurus, Camarasaurus, Elosaurus, Apatosaurus ajax, A. minimum, Atlantosaurus, and the three proposed species of Haplocanthosaurus in addition to H sp. I expected to see the normal group with Cetiosaurus as most basal, diplodocoidea formed, and macronaria together. However, something strange happened.
Like proposed by Mike Taylor and Matt Wedel, Apatosaurus minimus grouped with Camarasaurus and Brachiosaurus. Both the strict consensus and 50% majority trees were similar, with the strict consensus simply more dissolved. Cetiosaurus was most basal other than the outgroup. But up next was a polytomy of Futalognkosaurus and Atlantosaurus. This is strange, as normally these two fall into the ends of separate neosauropoda. Up next was what would normally be Neosauropoda. This split into two groups, Presumably diplodocoidea and macronaria. However, the split seemed to be more Rebbachisauridae/Diplomacronaria. In Rebbachisauridae was a clade of (Haplo delfsi, Haplo priscus (Haplo sp. (Haplo utterbacki + Tataouinea) ) ). And then there is Diplomacronaria, which is just a name I made up. It includes (Apato ajax ( Elosaurus ( Apato minimus (Camarasaurus + Brachiosaurus) ) ) ). This is really quite a weird result, but I think it could be expected with the small number of taxa and the purely sacral-based matrix. Both the strict consensus and 50% majority trees are shown below.
Things to note include that there were 15 possible trees for the strict consensus. In the analysis, when there was more than one species the first letter is directly following the generic name. I can provide people with the nexus file if they would like it.