Creating Reproducible Workflows to Relate Changes in Small Pelagic Fish Diets to Prey Distribution and Environmental Factors across the Northeast U.S. Shelf

The Northeast U.S. Shelf Long Term Ecological Research (NES-LTER) team in partnership with NOAA Northeast Fisheries Science Center (NEFSC) is studying the diets of small pelagic fish collected on trawl surveys from the Mid-Atlantic Bight to the Gulf of Maine. The goal of this project is to understand long term changes in the diets of these fish and how environmental variability correlates with these changes in the NES region. Established in 2017, one challenge facing this new LTER is the standardization of data collection and analyses. Considering the intent to make comparisons of fish diet across cruises, it becomes critical to establish foundational workflows for iterative analyses. In this project, we prototyped workflows for analyses of data previously published for the NES region and tested the reproducibility of these analyses with recent, unpublished data. Parsing by fish species and cruise, publicly accessible scripts were composed in R that analyzed relative prey abundance, richness, and size, and performed a canonical correspondence analysis (CCA) to elucidate the interplay of diet composition and environmental parameters. Producing these analyses required datasets to be supplemented with external data, such as bottom temperature from the NEFSC Ecosystems Survey Branch, which served as an explanatory variable for CCA. Essential for analyzing prey abundance, but also important for harmonization with NEFSC EcoMon and other regional data for zooplankton (prey) distribution, an ITIS taxonomic validation workflow was created to resolve, classify, and bin prey types. In application, the workflows created for this project demonstrate how long-term biological observations can be integrated into ocean observation studies, relevant to other regional analyses. Small pelagic fish represent a crucial link in the food web which influences the greater NES ecosystem. With collaboration, this project’s output can be wielded for purposes beyond diet analyses, such as informing stock assessment models for commercial fisheries.