Disentangling ecological and taphonomic signals in ancient food webs

Abstract Analyses of ancient food webs reveal important paleoecological processes and responses to a range perturbations throughout Earth's history, such as climate change. These can inform our forecasts future biotic similar perturbations. However, previous analyses rarely accounted for key differences between modern community data, particularly selective loss soft-bodied taxa during fossilization. To consider how fossilization impacts inferences structure, we (1) analyzed node-level attributes identify correlations ecological roles potential (2) applied information procedures web data extant systems. We found that organisms has predictable effects on the trophic structure “artificially fossilized” because these occupy unique, consistent positions. Fossilized misleadingly appear less stable (i.e., more prone cascades), with predation an overrepresentation generalist consumers. also soft- hard-bodied taxa—indicated by distinct positions in webs—are recorded early Eocene web, but not Cambrian webs. This suggests groups have existed ?48 Myr. Our results indicate accounting is vital accurate depictions consistency trends across means it possible predict affects metrics, which permit better modeling communities.