The importance of sampling standardization for comparisons of insect herbivory in deep time: a case study from the late Palaeozoic.

Sampling standardization has not been fully addressed for the study of insect herbivory in the fossil record. The effects of sampling within a single locality were explored almost a decade ago, but the importance of sampling standardization for comparisons of herbivory across space and time has not yet been evaluated. Here, we present a case study from the Permian in which we evaluate the impact of sampling standardization on comparisons of insect herbivory from two localities that are similar in age and floral composition. Comparisons of insect damage type (DT) diversity change dramatically when the number of leaves examined is standardized by surface area. This finding suggests that surface area should always be taken into account for comparisons of DT diversity. In addition, the three most common metrics of herbivory-DT diversity, proportion of leaves herbivorized and proportion of leaf surface area herbivorized-are inherently decoupled from each other. The decoupling of the diversity and intensity of insect herbivory necessitates a reinterpretation of published data because they had been conflated in previous studies. Future studies should examine the divergent ecological factors that underlie these metrics. We conclude with suggestions to guide the sampling and analysis of herbivorized leaves in the fossil record.