Abstract Somphospondylan titanosauriforms have a rich, global fossil record. However, the early evolutionary history of this sauropod dinosaur clade, including that of the titanosaurian lineage, is poorly understood. Purported titanosaurs have been described from the earliest Cretaceous of South America and Eurasia, which would make them the stratigraphically oldest known representatives of this clade, with implications for early titanosaur biogeography. However, their remains are fragmentary and their phylogenetic affinities require further testing. Among these is Triunfosaurus leonardii, known from caudal vertebrae, haemal arches, and an ischium from the Berriasian–lower Hauterivian Rio Piranhas Formation, Triunfo Basin, north-eastern Brazil. We present a redescription of the holotype of Triunfosaurus, including a number of anatomical re-interpretations, which enables us to provide a new diagnosis. Its inclusion into the largest phylogenetic data matrix for eusauropods so far assembled, comprising 162 OTUs (including 100 titanosauriform species) scored for 576 characters, recovers Triunfosaurus as a somphospondylan that is either close to, or an early-diverging member of, the titanosaur radiation. Additional results of our phylogenetic analyses indicate that the contemporaneous Argentinean sauropod, Ninjatitan zapatai, is probably a non-titanosaurian somphospondylan, closely related to the late Early Cretaceous taxon Chubutisaurus insignis, but that it could alternatively be a diplodocoid or represent chimeric remains of both clades. The stratigraphically earliest known titanosaur with unequivocal phylogenetic support comes from the Valanginian of eastern Eurasia, with our analyses reinforcing the previously proposed eutitanosaurian affinities of Tengrisaurus starkovi. Non-titanosaurian somphospondylans from the latest Jurassic–earliest Cretaceous are rare, but include Gondwanan and Eurasian occurrences. With the possible exception of Triunfosaurus, the stratigraphically earliest known titanosaurs do not belong to the earliest diverging titanosaurian lineages; members of the latter, as well as their closest relatives (e.g. Andesaurus delgadoi, Diamantinasauria, and Huabeisaurus allocotus), are from the mid-Cretaceous and globally widespread. We would expect the first titanosaurs to be morphologically closer to these taxa, rather than displaying ‘typical’ titanosaurian features, such as strongly procoelous anterior caudal vertebrae. Collectively, this implies that we are yet to sample representatives of the initial titanosaur radiation. The biogeographic origins of both Somphospondyli and Titanosauria are, therefore, probably obscured by spatiotemporal sampling failure.
Mannion et al. (Wed,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: