TY - JOUR T1 - Community structure and sample size affect estimates of predation intensity and prey selection: A model-based validation JF - bioRxiv DO - 10.1101/2022.07.18.500550 SP - 2022.07.18.500550 AU - Madhura Bhattacherjee AU - Devapriya Chattopadhyay Y1 - 2022/01/01 UR - http://biorxiv.org/content/early/2022/07/20/2022.07.18.500550.abstract N2 - Predation estimates inferred from the preserved records of predation traces are essential in evaluating the evolutionary effect of ecological interactions. It is, however, rarely investigated how sampling intensity and community composition of an assemblage influence the reliability of these measures.Using a resampling technique, we evaluated the effect of evenness and sampling intensity of a community on the inferred predation estimates. We theoretically simulated model communities representing different levels of evenness (ET), predation intensity (PIT), and predatory behavior (selective, non-selective). These communities were resampled without replacement; we noted the variation in the inferred predation intensity (PIT.inf) and the number of prey species (Sprey.inf) during resampling. Our results demonstrate that ET does not influence PIT.inf for non-selective predation. However, communities with highly selective predation are sensitive to evenness and sampling intensity; PIT.inf of these assemblages can substantially deviate from the actual value. Sprey.inf is also influenced by ET, predation selectivity, and PIT. For non-selective predation, sampling intensity heavily influences communities with low ET and low PIT; Sprey.inf is underrepresented at smaller sample size. For prey-abundance guides selective predation, the sensitivity depends on the nature of selection; for communities of low ET and PIT where rare species are attacked preferentially, the Sprey.inf deviates significantly at small sample size. We proposed a post-facto standardization method for comparing predation estimates of discrete communities. We validated its utility using the published predation data of the Plio-Pleistocene molluscan fossil assemblage. This approach provides critical insight into the biological reliability of predator estimates across time and space.Competing Interest StatementThe authors have declared no competing interest. ER -