LEVERAGING LARGE DATASETS AND PALEOGRAPHIC MAPS TO EVALUATE SAUROPOD ECOLOGY AND ECOSYSTEM INTERACTIONS

Modern organisms shape their ecosystems through a variety of direct and indirect interactions with other organisms and the landscape. Although the impacts of organisms on their ecosystems can be readily studied in modern landscapes (e.g., the influence of African elephants on savannah habitats); identification and analysis of biological interactions is difficult in the fossil record. This work explores the utility of large dataset analyses in order to detect impacts of Earth’s largest herbivores, sauropods, on their habitats. The Late Jurassic Morrison Formation is used as a case study to examine this phenomenon because it is geographically expansive, stretching across approximately 1,000,000 km² of the western United States and Canada, and because it represents a geologically stable depositional system for eight to ten million years. For this project, we created three new paleoenvironmental/paleogeographic maps representing the major depisodes of the Morrison Formation. These environments contain an incredibly diverse dinosaur fauna, including over two dozen sauropod species and numerous plant taxa. A preliminary dataset of over 850 dinosaur and plant occurrences from almost 300 fossil localities is leveraged in order to assess the impacts of sauropods on Morrison landscapes. The size of this dataset permits us to employ/use/leverage several analytical approaches typically not available to terrestrial and vertebrate paleontological analysis. Co-occurrence analysis at differing spatial scales is applied to determine potential fossil plant associations and delineate habitats during three intervals of Morrison deposition. The prevalence of different sauropod taxa within these regions is also assessed. In addition, paleo environmental niche modeling (paleo-ENM) is used to generate habitat suitability maps of both plant and sauropod taxa through the duration of Morrison Formation deposition. The habitat suitability maps are then compared among taxa and through time to look for associations of vegetation, sauropod taxa, and physical environment and any spatiotemporal shifts in these associations. This work provides a roadmap for the use of large vertebrate and terrestrial paleontological datasets and potential applications of co-occurrence analysis and paleo-ENM to detect impacts of large herbivores on their ecosystems in deep time.