The anatomy ofStratesaurus(Reptilia, Plesiosauria) from the lowermost Jurassic of Somerset, United Kingdom

Marine and terrestrial animals show a mosaic of lineage extinctions and diversifications during the Jurassic-Cretaceous transition. However, despite its potential importance in shaping animal evolution, few palaeontological studies have focussed on this interval and the possible climate and biotic drivers of its faunal turnover. In consequence evolutionary patterns in most groups are poorly understood. We use a new, large morphological dataset to examine patterns of lineage diversity and disparity (variety of form) in the marine tetrapod clade Plesiosauria, and compare these patterns with those of other organisms. Although seven plesiosaurian lineages have been hypothesised as crossing the Jurassic-Cretaceous boundary, our most parsimonious topology suggests the number was only three. The robust recovery of a novel group including most Cretaceous plesiosauroids (Xenopsaria, new clade) is instrumental in this result. Substantial plesiosaurian turnover occurred during the Jurassic-Cretaceous boundary interval, including the loss of substantial pliosaurid, and cryptoclidid diversity and disparity, followed by the radiation of Xenopsaria during the Early Cretaceous. Possible physical drivers of this turnover include climatic fluctuations that influenced oceanic productivity and diversity: Late Jurassic climates were characterised by widespread global monsoonal conditions and increased nutrient flux into the opening Atlantic-Tethys, resulting in eutrophication and a highly productive, but taxonomically depauperate, plankton. Latest Jurassic and Early Cretaceous climates were more arid, resulting in oligotrophic ocean conditions and high taxonomic diversity of radiolarians, calcareous nannoplankton and possibly ammonoids. However, the observation of discordant extinction patterns in other marine tetrapod groups such as ichthyosaurs and marine crocodylomorphs suggests that clade-specific factors may have been more important than overarching extrinsic drivers of faunal turnover during the Jurassic-Cretaceous boundary interval.

Previous attempts to resolve plesiosaurian phylogeny are reviewed and a new phylogenetic data set of 66 taxa (67% of ingroup taxa examined directly) and 178 characters (eight new) is presented. We recover two key novel results: a monophyletic Plesiosauridae comprising Plesiosaurus dolichodeirus, Hydrorion brachypterygius, Microcleidus homalospondylus, Occitanosaurus tournemirensis and Seeleyosaurus guilelmiimperatoris; and five plesiosaurian taxa recovered outside the split between Plesiosauroidea and Pliosauroidea. These taxa are Attenborosaurus conybeari, 'Plesiosaurus'macrocephalus and a clade comprising Archaeonectrus rostratus, Macroplata tenuiceps and BMNH 49202. Based on this result, a new name, Neoplesiosauria, is erected for the clade comprising Plesiosauroidea and Pliosauroidea. Taxon subsamples of the new dataset are used to simulate previous investigations of global plesiosaurian relationships. Based on these simulations, most major differences between previous global phylogenetic hypotheses can be attributed to differences in taxon sampling. These include the position of Leptocleididae and Polycotylidae and the monophyly or paraphyly of Rhomaleosauridae. On this basis we favour the results recovered by our, larger analysis. Leptocleididae and Polycotylidae are sister taxa, forming a monophyletic clade within Plesiosauroidea, indicating that the large-headed, short-necked 'pliosauromorph' body plan evolved twice within Plesiosauria. Rhomaleosauridae forms the monophyletic sister taxon of Pliosauridae within Pliosauroidea. Problems are identified with previous phylogenetic definitions of plesiosaurian clades and new, stem-based definitions are presented that should maintain their integrity over a range of phylogenetic hypotheses. New, rank-free clade names Cryptoclidia and Leptocleidia are erected to replace the superfamilies Cryptoclidoidea and Leptocleidoidea. These were problematic as they were nested within the superfamily Plesiosauroidea. The incongruence length difference test indicates no significant difference in levels of homoplasy between cranial and postcranial characters.

Invasion of the open ocean by tetrapods represents a major evolutionary transition that occurred independently in cetaceans, mosasauroids, chelonioids (sea turtles), ichthyosaurs and plesiosaurs. Plesiosaurian reptiles invaded pelagic ocean environments immediately following the Late Triassic extinctions. This diversification is recorded by three intensively-sampled European fossil faunas, spanning 20 million years (Ma). These provide an unparalleled opportunity to document changes in key macroevolutionary parameters associated with secondary adaptation to pelagic life in tetrapods. A comprehensive assessment focuses on the oldest fauna, from the Blue Lias Formation of Street, and nearby localities, in Somerset, UK (Earliest Jurassic: 200 Ma), identifying three new species representing two small-bodied rhomaleosaurids (Stratesaurus taylori gen et sp. nov.; Avalonnectes arturi gen. et sp. nov) and the most basal plesiosauroid, Eoplesiosaurus antiquior gen. et sp. nov. The initial radiation of plesiosaurs was characterised by high, but short-lived, diversity of an archaic clade, Rhomaleosauridae. Representatives of this initial radiation were replaced by derived, neoplesiosaurian plesiosaurs at small-medium body sizes during a more gradual accumulation of morphological disparity. This gradualistic modality suggests that adaptive radiations within tetrapod subclades are not always characterised by the initially high levels of disparity observed in the Paleozoic origins of major metazoan body plans, or in the origin of tetrapods. High rhomaleosaurid diversity immediately following the Triassic-Jurassic boundary supports the gradual model of Late Triassic extinctions, mostly predating the boundary itself. Increase in both maximum and minimum body length early in plesiosaurian history suggests a driven evolutionary trend. However, Maximum-likelihood models suggest only passive expansion into higher body size categories.