Search

The growing world population and increasing consumption of fish per capita have driven the growth of the aquaculture sector. Fed-based aquaculture, however, is still dependent on capture fisheries for the supply of fishmeal and fish oil, which are valuable protein and lipid sources in fish feed. Due to their limited availability and high prices, alternative ingredients are needed such as insects. In particular, the black soldier fly (Hermetia illucens, BSF) has received increasing attention because of its mass-rearing potential and ability to convert biodegradable side streams into biomass high in protein and lipid. However, several challenges must be overcome before BSF can be included in fish feed on a global scale, including its low production volumes, limited content of certain amino acids, saturated fatty acid profile together with low n-3 levels, and presence of chitin. The main focus of this Ph.D. thesis is to address these challenges through four studies: three of them focusing on enhancing the BSF harvest yield and modifying the nutritional composition, and one applying sieving as a chitin fractionation method for BSF meal and investigating the effects of chitin on nutrient digestibility in fish.In Study I, the nutritional composition, including chitin, was investigated in all life stages of BSF. Six different commonly applied chitin quantification methods were used and compared to determine chitin contents including four direct methods (acid detergent fibre with/without amino acid correction, crude fibre, and neutral detergent fibre) and two indirect methods (glucosamine determination by spectrophotometry and ultra-performance liquid chromatography). It was found that the composition of most nutrients fluctuated between life stages and that the highest individual yield was obtained in the early prepupae stage. Furthermore, chitin was present in all life stages and its content generally increased throughout the life cycle. The chitin quantification method had a large effect on the determined chitin contents, with accuracy and precision being highest for the investigated indirect ultra-performance liquid chromatography method and lowest for the neutral detergent fibre method. These results underline that both the life stage and the quantification method can have a significant impact on the quantified nutritional composition of BSF.The effect of the rearing substrate on BSF macronutrient composition was investigated in Study II. Different locally available biodegradable side streams were used for rearing BSF larvae including brewer’s spent grain, mitigation mussels (Mytilus edulis), rapeseed cake, and shrimp waste meal (Pandalus borealis). Larval composition and performance on the different biodegradable side streams were compared to larvae reared on either a chicken feed control or a mixed feed consisting of chicken feed and biodegradable side streams, with a macronutrient composition similar to chicken feed. The type of rearing substrate affected both larval body composition and growth performance. There was a strong positive linear correlation between the ash content in the substrate and the larvae whereas larval lipid, protein, amino acid, and chitin content appeared to be more affected by the larval development stage. Interestingly, larvae reared on marine-based biodegradable side streams had a significantly higher proportion of n-3 fatty acids than larvae reared on the other substrates, indicating the ability of BSF to accumulate n-3 fatty acids from their rearing substrate. These findings demonstrated that the rearing substrate affects both larval composition and development; and that the larval body composition can, to a certain extent, be tailored according to further fish feed applications.To enhance harvest yield, the nutritional requirements of BSF need to be identified. In Study III, optimal dietary protein to non-fibre carbohydrate (P:C) ratios were established for BSF larvae. Larvae were reared on six isoenergetic and isolipidic experimental substrates formulated with P:C ratios ranging between 1:1 - 1:9, corresponding to protein to energy (P:E) ratios of 4.8 – 20.8 g/MJ, with comparison to a chicken feed control with a P:C ratio of 1:3. The results showed that P:C ratios of 1:2 – 1:3 (corresponding to P:E ratios of 11.2 – 14.4 g/MJ) were most optimal in terms of obtaining the highest larval harvest yield and ensuring the best utilisation of feed resources for BSF production whilst meeting larval macro nutritional requirements. Therefore, using these ratios to formulate rearing substrates can aid in improving the conversion efficiency of biodegradable side streams into BSF biomass.Study IV focused on the applications of partially defatted BSFLM as a novel feed ingredient for omnivorous Nile tilapia (Oreochromis niloticus) and carnivorous rainbow trout (Oncorhynchus mykiss). Partially defatted commercial BSFLM was sieved to obtain three size fractions with different chitin contents, being 1.8 (0 - 200 μm), 2.7 (200 - 400 μm), and 15.4 (>400 μm) % chitin on dry matter (DM) basis. Four isoenergetic and isonitrogenous diets were formulated for each fish species including an experimental reference diet and three diets replacing 25% of the experimental reference diet matrix with one of the three BSFLM size fractions. When compared to the other treatment groups, the results indicated that in both species the dietary inclusion of the high chitin BSFLM size fraction resulted in significantly lower apparent digestibility of crude protein, nitrogen-free extract, and DM while no similar negative effects were observed with lower chitin size fractions. It was discovered that both fish species could digest chitin and presented β-N-acetylglucosaminidase activity in their intestinal tract. However, fish fed the highest dietary chitin inclusion level had a decreased chitin digestibility despite an increased intestinal exochitinase activity. Altogether, these findings suggest that chitin acts as both an energy source and an anti-nutrient in Nile tilapia and rainbow trout.In conclusion, this Ph.D. thesis has addressed several challenges that currently limit large-scale BSF inclusion in fish feed, including low production volumes, limited presence of certain essential amino acids, saturated fatty acid profile combined with low n-3 levels, and presence of chitin. Harvest yield can be enhanced by the use of rearing substrates that meet BSF nutritional requirements, for example by providing larvae with rearing substrates containing P:C ratios of 1:2 – 1:3, and harvesting BSF directly when entering the prepupae phase, where the highest individual weight is achieved. Harvesting BSF in early larval instars can increase the relative crude protein content but neither harvest time nor rearing substrate seems to have a major impact on BSF amino acid profile. Therefore, supplementation of limiting essential amino acids to fish feed containing (high dietary inclusion levels of) BSF meal might be needed to meet fish requirements. On the contrary, the BSF fatty acid profile is largely dependent on the rearing substrate. Providing rearing substrates high in n-3 fatty acids can subsequently result in the accumulation of n-3 fatty acids in BSF, which could enhance further fish feed applications. Lastly, the presence of chitin in BSF meal may restrict its use in fish feed, as chitin has been shown to have anti-nutritional effects in fish. Since the relative chitin content increases particularly during the late prepupae stage, early BSF harvest can minimise the amount of chitin in the raw material. Furthermore, chitin content in BSF meal can be reduced by chitin fractionation.