Detecting the invisible through DNA metabarcoding: The role of gelatinous taxa in the diet of two demersal Antarctic key stone fish species (Notothenioidei)

Gelatinous zooplankton (GZP), i.e., ctenophores, cnidarian medusae, chaetognaths, appendicularians and salps, are considered climate change winners. This becomes particularly obvious in the Southern Ocean, which has undergone a significant shift from a krill-based to a salp-based ecosystem over the last decades. A better knowledge on the role of gelatinous invertebrates as prey is needed to predict the impact of such a gelatinous shift. Until recently, GZP was considered as a “trophic dead end”. However, their true importance in diets has remained unresolved due to the rapid digestion of their watery and soft tissues in predators' stomachs. In this study, we want to validate the paradigm shift from GZP being considered as “survival food” to be considered a “regular” prey item for two demersal fish species (Notothenia rossii and N. coriiceps) of Potter Cove, South Shetland Islands, using a multimarker (COI and 18S) metabarcoding approach. We found that GZP taxa commonly occurred in the diets of both species, represented by pelagic tunicates (appendicularians, salps), cnidarians, chaetognaths and ctenophores. Salps were the most abundant prey group, preyed upon by each individual of both species, reaching 98.7% relative read abundance for 18S. We recovered a wide range of different taxa in their diets, from primary producers to highly abundant invertebrates, thus the two nototheniid species can be regarded as “natural samplers” of the ecosystem in study. Finally, we want to point out the importance of multimarker metabarcoding approaches for broad ecological assessments, given the differential amplification and sequencing success of different markers for specific groups and the unequal taxonomic coverage of the reference databases. The output of each marker was highly complementary, since an important prey item such as salps, was only detected with 18S, while other taxa (e.g., Arthropoda) were represented with a higher taxonomic resolution with COI.