|dc.description.abstract||Current New Zealand fisheries management policy is directed towards a more ecosystem-based approach to fisheries management. Any ecosystem-based approach to fisheries management requires an understanding of the feeding interactions within and between fish species; both those commercially and non-commercially exploited. Despite the existence of several decades-old deep-sea commercial fisheries for orange roughy, little is known of the feeding interactions of non-commercially exploited mid-slope deep-sea fishes in New Zealand waters. As a consequence, understanding the roles these fish species play in this ecosystem is of great importance.
To investigate the feeding interactions between New Zealand’s deep-sea fish, the diets of 2888 fish attributed to 76 species, of a combined weight of 779 kg, were procured from two New Zealand mid-slope locations: northeastern Chatham Rise 901–1196 metres depth; Puysegur Bank, southern New Zealand, 1020–1072 metres depth; and one upper slope area off the Wairarapa coast, 239–448 metres depth. The stomachs of 1078 and the intestines of 695 of these fish contained food.
Fish with everted stomachs tend to be excluded from dietary analyses, resulting in an underestimation of their role in the ecosystem. Investigation of the intestinal contents permitted partial reconstruction of the diets of fish species with high rates of stomach eversion, or infrequent feeding periodicity. Many of the fish with everted stomachs were benthic feeders, suggesting that their exclusion from dietary studies could underestimate the significance of benthic trophic pathways in the deep-sea: this was highlighted by differences between feeding guilds based on stomach and all gut contents from the northeastern Chatham Rise.
Overall fish diets were from a variety of benthic and/or benthopelagic sources: with some species exhibiting ontogenic dietary shifts, larger bodied food items (often fish, prawns or squid) became more important in the diets of larger fishes. There was some evidence for scavenging in several species, and its role in the diets of these deep-sea species is likely to be underestimated. There were high levels of dietary overlap within species size classes, but not between most species. The fish species at either upper or mid-slope depths do not appear to compete very much for food, indicating a high level of niche partitioning. Niche breadths also were generally low for most species. Scampi was not an important source of food for benthic feeding upper-slope fish species from off the Wairarapa coast.
Stomach parasite analysis and the percentage of empty stomachs (PES) could both be useful dietary metrics. Serrulate rattail had significantly more stomach parasites than four-rayed rattail. The parasite loading correlated with the composition of the diet in both species, and an ontogenic shift in diet for serrulate rattail. Parasite analysis could be a useful addition to stomach content analysis in the absence of stomach contents. Several species did not appear to have fed recently, having a high percentage of individuals with empty stomachs. The percentage of empty stomachs (or intestines) could be a useful indicator of the surrounding food supply, and as such a proxy for the productivity of surface waters.
This thesis addresses some of the gaps in knowledge that are impeding an ecosystem-based approach to the management of New Zealand orange roughy fisheries. This approach is possible for New Zealand orange roughy fisheries, but is still some way in the future, and will require further investigation regarding the interactions between these fish species and their environment.||en_NZ