Skip to main content
Dryad

Data from: Bayesian stable isotope analysis as a validation approach for marine food web models: a coastal Louisiana, USA case study

Data files

Jun 30, 2026 version files 459.71 KB
Jun 30, 2026 version files 459.74 KB

Click names to download individual files

Abstract

Marine food web models can provide a range of potential outcomes for managers to consider so they can make more informed fisheries management decisions. It is well known, however, that food web models require large amounts of disparate data, which can lead to the amplification of uncertainties in model output. Previous studies have leveraged stable isotope data from the same system to compare the trophodynamics between modeled and in situ data. Here, we provide a new iterative framework to aid users in evaluating the statistical agreement between food web model generated trophic levels and trophic levels calculated from stable isotope data. Using a previously developed food web model from coastal Louisiana as a case study, the validation process included a series of Spearman-rank correlation analyses to statistically compare the level of agreement between food web model and stable isotope derived trophic levels. Then, we applied a Bayesian statistical approach to derive trophic levels from the stable isotope data along with trophic probability curves. To provide a more accessible method to carry out a validation analysis, we developed the first open-source food web model validation visualization tool called EcoTuneR. This webtool provides a seamless approach to visualizing the Bayesian derived trophic level probability curves collectively with the food web model trophic levels. Results of the case study showed that the correlation between the stable isotope trophic levels and the food web model trophic levels were comparable to previous studies from others. Results of the Bayesian probability visualizations showed that 47% of species functional groups in the food web model fell within the 95% credibility intervals of the Bayesian derived stable isotope generated trophic levels. We demonstrate the importance of careful interpretation of the validation process through a deep analysis of the data sources for both the food web model and stable isotope data. Finally, we provide a suite of considerations and suggestions for researchers to consider for future applications of this framework.