About me
66 Million Years of Trophic Network Evolution in North America
Louis-Philippe Bateman, Hans C.E. Larsson
Trophic networks have been emerging in ecology for the past two decades. However, few neontological or paleontological studies have studied the long-term evolution of these networks, leaving the future of present networks in the face of climate change and anthropogenetic disturbance unclear. In this study, we reconstruct 264 Cenozoic mammal trophic networks in North America to tackle this knowledge gap. Our model predicts interactions using matching species traits like body size, broad dietary category, and life habit, and refines these predictions using phylogeny. Cross-validation using modern interaction databases shows it predicts most interactions accurately (true skill statistic between 0.75-0.95). We then track how network properties change over millions of years. Connectance, mean trophic level, proportion of omnivores, mean incoherence, and clustering all decrease through the Cenozoic, while other metrics show an increase, such as proportion of herbivores and mean interaction weight. Other metrics, such as modularity, mean trophic similarity, mean generality, and proportion of carnivores show variable or unvarying trajectories. This shows that while certain network properties may vary, others are very constant through geological time. We then group species into guilds based on body size, life habit, and broad dietary category. This reveals that while a handful of guilds and interactions between them have persisted through almost all the Cenozoic, the median persistence for interactions is only about 1 million years. We then cluster networks according to their properties and guilds, revealing the existence of at least three “states” throughout the Cenozoic: a post-K-Pg extinction recovery, a warm Paleogene state, and a post-Paleogene cool state. We discuss the factors that possibly mediated these state shifts, including biotic change and climate change. We also discuss the implications of these preliminary results for the future evolution of modern trophic networks.
