Trophic relationships of hake (Merluccius Capensis Castelnau, 1851 and M. Paradoxus Franca 1960) from the Nothern Benguela current ecosystem (Namibia): Inferences from stable isotopes and fatty acids

A thesis submitted in fulfilment of the requirements for the Degree of Doctor of Philosophy (Science)

Two species of hake (Merluccius capens is and Merluccius paradoxus) account for most of Namibia's fisheries catch, and they are important secondary consumers in the Benguela Current ecosystem. Inferences on their trophic relationships have been based mainly on stomach content analyses. However, such data are limited temporally because they represent only snapshots of recent feeding, and are quantitatively biased because of variation in the digestion rates of different prey. The principal aim of the thesis was to understand the trophic relationships of two hake species relative to each other, their known prey and top predators (demersal sharks) in the northern Benguela Current ecosystem (Namibia), using time-integrating trophic biomarkers. By using stable isotope (carbon and nitrogen) and fatty acid signatures of their muscle tissues, my overall objectives were to produce new knowledge about 1) hake ontogenic trophic relationships, 2) the contributions of different prey to hake diets, 3) hake dietary differences, and 4) some aspects of hake's trophic relationships with demersal sharks. Tissues of hake (n=358), their potential prey (n=455), and demersal sharks (n=42) were collected between 2008 and 2012 during demersal bottom trawl surveys offNamibia, for stable isotope and fatty acid analyses. Nitrogen isotopes (8 15N) showed a significant positive relationship with size in both M capensis and M paradoxus; however, the correlation slopes of size and 8 15N in the two species were significantly different, with M paradoxus displaying a steeper trophic shift. A significant increase in carbon isotope signature (8 13C) with size was observed in M capensis but not in M paradoxus. In all size classes (except 20-29 em) M capensis had significantly enriched 13C values, generally matching their respective adult shelf-slope distribution. Smaller hake from both species (20-39 em) were trophically indistinguishable, with trophic levels (TL) of3.3 indicating predominant zooplanktivory. The largest M capensis (60-70 em) had TLs of approximately 3.5- 3.6, whereas M paradoxus of the same size. were slightly higher at 3.7-3.8 TL, indicating greater piscivory in the latter. A Bayesian isotope mixing model indicated that teleosts Synagrops microlepsis and Chlorophthalmus agazizi, the shrimp Plesionika martia, myctophids and euphausiids dominated the diets of both hake species. The greatest contributor to the diet of M paradoxus changed depending on the body size of the hake, while the diet of all size classes of M cap ens is was dominated by myctophids. Cannibalism accounted for less than 6% and less than 4% of the diets of M capensis and M paradoxus respectively. Isotope-based estimates of prey dietary contributions differed from those of published accounts of gut content analyses, an indication of overestimation of the time-integrated diet by the latter. The fatty acid (FA) profiles of the two species (M paradoxus and M cap ens is) and potential prey were similar to those typically abundant in marine fish. Palmitic acid (16:0) was the dominant saturated fatty acid (SFA), oleic acid (18: ln-9) was the dominant monounsaturated fatty acid (MUFA) and docosahexaenoic acid (22:6n-3) was the dominant polyunsaturated fatty acid (PUF A) in both hake species. Fatty acid (FA) profiles in the neutral storage lipids of M paradoxus and M capensis were significantly different, an indication of hake species' dietary resource partitioning, although some overlap was apparent. There was a significant temporal effect on hake FA compositions, reflecting temporal variations in quantity and quality of dietary resources. Feasible contributions of hake to the diets of three demersal sharks indicated that hake are not major prey of Centrophorus squamosus, whose diet was dominated by crustaceans. However, there was an increased contribution of hake to the diets of Deania sharks. All three sharks studied (Deania calcea, D. profundorum and C. squamosus) potentially consumed upwards of 18 % (208 960 tonnes per year) of the hake biomass [2011 hake biomass estimates (1111 196.8 tonnes)]. The two Deania species were isotopically indistinguishable in terms of 815N, and fed on a prey composition that overlapped with that of M paradoxus. 813C values indicated that hake and sharks fed on prey from similar basal resources, with C. squamosus being significantly different from M paradoxus but not from M capensis. The range of 813C and 815N indicated that the two hake species had a higher range of carbon sources as well as enhanced trophic diversity in their feeding patterns than did the three shark species. This study is the first combined usage of stable isotope and fatty acid techniques to analyse trophic relationships of M capensis and M paradoxus in the northern Benguela Current (Namibia). The study contributes towards understanding of trophic interactions in Namibian waters, and can aid the implementation of a science-based ecosystem approach to fisheries management in the Northern Benguela Current region (Namibia). The extension of this analysis throughout hake distribution ranges will provide additional insights into the trophic dynamics of hake species in future, and my study represents a solid foundation from which to develop such extensions.