Data from: Widespread variation in stable isotope trophic position estimates: patterns, causes, and potential consequences

Stable isotope analysis is one of the most widely used techniques to estimate trophic position and provides fundamental insight into the structure and management of ecological communities. To account for the effects of geographic variation in isotope levels, trophic position is typically estimated relative to an isotope “baseline” (i.e., material representing geographic variation) using a methodology such as a formula or statistical analysis. There is, however, remarkable variation in the baselines and methodologies used to estimate trophic position from stable isotopes. The consequences of this lack of standardization are unknown but could result in biased or erroneous conclusions. We conducted a literature review to quantify the variation in baselines and methodologies used to estimate trophic position from stable isotopes. Next, we assessed the consequences of this variation on individual species estimates and food web structure by extracting published trophic positions and applying various baselines and methodologies to existing datasets. We identified ten baselines and eight methodologies, the use of which varied by ecosystem studied. Moreover, we found that different baselines and methodologies yield significantly different trophic position estimates for individual species, as well as different conclusions about food web structure. Authors should avoid biological interpretations of absolute, stand-alone trophic positions (as these are prone to a number of biases). We recommend pairing stable isotope analysis with other techniques for more robust conclusions. Increased sample size may mitigate some of the variation caused by different baselines and methodologies; however, an alarmingly large proportion of studies collected only one sample in at least one trophic group (41% of all reviewed studies). Authors should collect a minimum of five samples per trophic group (but ten for best-practices) from as many trophic groups as possible to increase statistical power and redundancy in comparisons. When sample size is unavoidably constrained, we recommend using compound-specific isotope analysis with a taxon-specific trophic discrimination factor, because it may be more accurate and require fewer samples to maintain appropriate statistical power. Implementing our recommendations will increase the robustness and accuracy of conclusions based on stable isotopes, resulting in better management decisions and a more accurate understanding of ecological communities.