Combining DNA metabarcoding with macroscopic analysis increases the number of detected prey taxa in the estimated diet for harbour porpoises

Johanna Stedt ^1* Linnea Brokmar ² Aleksija Neimanis ³ William F Englund ² Per Carlsson ¹ Anna Roos ²

• ¹ Lund University, Lund, Sweden
• ² Swedish Museum of Natural History, Stockholm, Sweden
• ³ Swedish Veterinary Agency, Uppsala, Uppsala, Sweden

Knowledge of animal diets is fundamental in ecology as it can provide insight into the structure, function and resilience of entire ecosystems. In this study we investigate the diet composition of the harbour porpoise (Phocoena phocoena), a small marine top predator with great ecological importance, to provide updated dietary estimates for the species in Swedish waters. This is done by parallel use of macroscopic stomach content analysis and DNA based prey species determinations by DNA metabarcoding (sDNA analysis), which allows us to compare and evaluate the methods’ respective and combined performance. We show that harbour porpoises during 2017-2022 consumed a broad variety of both benthic and pelagic fish along the Swedish west coast. The combination of macroscopic and sDNA analysis for diet estimation yielded an almost two-fold increase in species and taxa detection compared to macroscopic analysis alone, with overall detection of 36 unique prey species from 21 prey families. Consistent with results from previous studies in the area, the main prey taxa were clupeids, gadoids and gobiids. In the macroscopic analysis these three taxa together represent more than 80% of the relative numerical contribution to the estimated diet. Using sDNA analysis the same three taxa dominate the relative read abundance, with clupeids detected in all sampled porpoise stomachs, gobiids in 86% of stomachs, and gadoids in 66%. The diet estimates from the two diet tracing methods are overall in high agreement, but sDNA analysis increased the number of detected prey taxa and also increases the sample size by allowing extraction of dietary data from apparently empty stomachs. The detection probability of some occurring prey taxa, however, appears to be method dependent. To facilitate combined use of sDNA and macroscopic analysis in diet studies, we establish and propose a new occurrence metric that can be used to merge data for more straightforward diet comparisons. We conclude that for diet studies on harbour porpoises and other cetaceans, which often present empty stomachs in post-mortem examinations, the use of sDNA analysis has the potential to recover valuable data from animals where stomach samples would otherwise be disregarded as unavailable for diet analysis.