Abstract
Following the Late Jurassic, megaherbivore communities in North America undergo a dramatic turnover in faunal composition: sauropods decline to the point of becoming relatively minor components of ecosystems, stegosaurs become extinct, and hadrosaurids, ceratopsids and ankylosaurs rise in diversity and abundance. Although a variety of causes have been proposed to account for the dramatic decrease in sauropod diversity following the Late Jurassic and could have also been applicable to the disappearance of stegosaurs, the potential for competitive replacement of sauropods by hadrosauroids as an explanation has been previously dismissed due to morphological differences without further investigation. Using twelve ecomorphological correlates of the skull, this study provides a preliminary investigation into ecomorphospace occupation of major megaherbivore clades from the Late Jurassic through to the Late Cretaceous of North America and assess if morphological differences were enough to have potentially facilitated dietary niche partitioning between sauropods and iguanodontians and stegosaurs and ankylosaurs. Overlap in reconstructed ecomorphospace was observed between sauropods (particularly non-diplodocid sauropods) and iguanodontians, as would be expected if morphological differences were not enough to facilitate niche partitioning, contrary to original claims used to dismiss the competitive replacement hypothesis. Overlap was also observed between stegosaurs and ankylosaurs, particularly between Late Cretaceous ankylosaurs. Whether this overlap is reflective competitive replacement or opportunistic occupation of recently vacated niches will require further assessment as sampling of some clades prior to the Late Cretaceous is too poor to make a reliable assessment and several underlying assumptions necessary for competition to occur (e.g., resource limitation) still need investigation. Teasing out the cause(s) of the 'sauropod decline' and extinction of stegosaurs in North America following the Late Jurassic will require future research not only into the competitive exclusion hypothesis, but other hypotheses as well with better sampling from Early Cretaceous and Late Jurassic intervals.