Fish gut content analysis: robust measures of diet composition

Abstract Trophic studies are fundamental components of our understanding of biology and ecology, from observing individual organisms to modelling ecosystem function. When measuring fish gut contents, we rely on collecting samples that represent snapshots in time. Many limitations in extrapolating from these snapshots are well understood. However, there seems to be a widespread belief that when quantifying the composition of gut contents, more detail always provides more information. We highlight some fundamental problems with the apparently more quantitative approaches (i.e. ‘bulk’ methods measuring biomass or volume of each prey type) and suggest that frequency of occurrence (% F ) provides the most robust and interpretable measure of diet composition. The additional information provided by bulk methods contains unquantifiable and potentially significant error from a variety of sources. In our experience, the contents of most guts cannot be unambiguously separated into prey categories for quantification because of the presence of unidentifiable and inseparable partially digested material. Even where separation is possible, the composition of a gut at one point in time is affected by many unquantifiable factors unrelated to the actual composition of the diet. Consequently, bulk methods provide ambiguous interpretations from superficially quantitative models. Where research questions require more detail, these problems mean there is little alternative to time‐consuming approaches like prey reconstruction. However, for the descriptions of dietary composition presented in many studies, % F provides robust data that overcome many of the limitations of the more detailed approaches and provides considerable logistical and economic benefits.