Selective feeding by pelagic fish in the Belgian part of the North Sea

Pelagic fish and their planktonic prey are susceptible to a changing climate, giving rise to mismatches and planktonic bottlenecks. A detailed examination of the feeding ecology of pelagic fish can provide valuable insights in the causes and consequences of these phenomena. The present study investigated the diets of both juvenile and adult herring, sprat, horse mackerel, and adult mackerel in the Belgian part of the North Sea (BPNS) in relation to the distribution of zooplankton and ambient abiotic conditions. A study sampling pelagic fish and zooplankton simultaneously every month during consecutive years, and spanning nearshore to offshore sampling locations, is unprecedented in the southern North Sea. In all, 71 prey taxa were found in 725 stomachs of fish gathered at ten stations, sampled monthly in 2009 and 2010. The proportion of fish with empty stomachs was low (11%), and the number of prey species ranged from 0 to 21 sp. per stomach. The diet of herring and sprat was dominated by calanoid copepods, but herring stomachs also contained many decapod larvae, amphipods, cumaceans, and mysids. Mackerel added sandeels to an otherwise planktivorous diet. Horse mackerel consumed both benthic and pelagic prey. The highest frequency of occurrence in the stomachs was observed for the calanoid copepods Temora longicornis (33 408 of all 55 004 prey items identified) and Centropages hamatus (5003 times found). The fullness index ranged between 0 and 20.6, and averaged highest for sprat (0.86), followed by herring (0.60), horse mackerel (0.26), and mackerel (0.24). We observed a different composition of zooplankton species and life stages in the plankton samples compared with those in the fish stomachs. More adult and female copepods were eaten than the plankton samples would suggest. Also, the calanoid copepod Acartia clausi, the most common calanoid species in the BPNS, was barely eaten, as was the case for fish eggs and larvae, and for common planktonic species known to be preyed upon elsewhere (e.g. Oikopleura dioica, Evadne nordmanni, Euterpina acutifrons). Additionally, plankton densities averaged highest in spring and at midshore (20–30 km from shore) stations, but fullness index was highest nearshore (<12 km from shore) and (apart from sprat) in summer. A significant correlation between fullness index and total density of planktonic prey species was not observed, indicating that zooplankton densities were not restrictive. Yet the fact that more than 100 plankton species occurred in the plankton samples and just two of these (T. longicornis and C. hamatus) accounted for nearly three-quarters of all ingested prey items leads us to conclude that even minor changes in the ecology or phenology of these dominant zooplankters could have profound effects on pelagic fish stocks.