A total of six isoprotein and isolipid diets for Atlantic salmon, Salmo salar L., were prepared substituting from 0 to 100% of fish meal protein (0–68% of diet by dry weight) with meal from Antarctic krill (Euphausia superba). The feed produced from high inclusion levels of krill meal had lower ability to absorb lipid during vacuum coating than fish meal. Both amino acid and fatty acid compositions of the diets were fairly similar. The experiment commenced using salmon averaging 500 g and ended at a mean weight of 1500–1800 g (140 days of feeding). Moderate amounts of krill meal (20–60% of krill protein) in the diets increased growth during the first 71 days of feeding compared with the fish meal control, while no growth difference was observed during the last 69 days of feeding. This may, at least in parts, be explained by a feed‐attractant function of the krill meal. Muscle dry weight and lipid concentrations were unaffected by the diet. Feed conversion rate increased with high levels of krill meal in the diets (e.g. for the last period from 0.94 in the 0% diet to 1.26 in the 100% diet). This indicates that the fish were able to compensate by eating more to maintain growth. The apparent digestibility coefficients of dry matter and protein were not influenced by diet, but both faecal moisture and lipid had a tendency to increase at the highest inclusion level (all protein from krill meal). This may be related to chitin in the krill diet that is known to decrease lipid absorption and induce diarrhoea (increased water content in faeces). Chitin was not utilized to any major extent. Welfare parameters such as blood haemoglobin, red blood cell counts, plasma protein, cholesterol, triacylglycerols and glucose levels were unaffected by diets. Clinical indicators of cellular damage (alanine aminotransferase and aspartate aminotransferase) were similar indicating no diet‐induced tissue damage during the trial.

Biochemical structure of protein (reactive SH content, content ratio of SH/SÐS and concentration of D-Asp as % of total (D L)-Asp) indicating digestibility of dietary protein was changed under different processing conditions. Based on ®sh crude enzyme extract, in vitro digestibility of different ®sh materials processed under different conditions correlated positively with reactive SH content and content ratio of SH/SÐS and negatively with D-Asp concentration. In vitro digestion of different experimental feeds, based on Atlantic salmon crude enzyme extracts, was studied in association with growth trials in order to investigate its value as a criterion for industrial strategy in predicting feed quality. Crude enzymes were extracted from the pyloric caeca before feeding. Signi®cant differences in in vitro digestibility between the experimental feeds were observed whereby there would be differences in feed conversion ef®ciency within 3 months of feeding. There were associations between the in vitro digestibility and other parameters for dietary quality, such as mink digestibility and the biochemical structure parameters of the dietary protein due to different processing conditions. Crude enzyme extracts from rainbow trout and European seabass were also used for in vitro digestibility study of different experimental feeds by standardising trypsin activity to that of Atlantic salmon crude enzyme extract. The results indicated that different ®sh species have different digestion ability to the same feed types, and the effective time for feed utilisation and growth is dependent on ®sh sensitivity and the extent of difference in digestibility between the feeds consumed as observed in the Atlantic salmon trials. For the species investigated, sensitivity ranking of the enzymes to feed quality under the condition studied was Atlantic salmon > rainbow trout > European seabass. The results indicated that in vitro digestibility study of experimental feeds using pyloric caecal crude enzyme extract from a speci®c species at an age of interest could be a practical, quick and reliable method for testing feed quality in growth trials. By standardising the crude enzyme extract with regards to trypsin activity, the in vitro digestibility values could be comparable not only within the same species but also between different species.

No abstract

The effect of processing conditions on protein digestibility and fluorodinitrobenzene (FDNB)-reactive (available) lysine in the production of fish meal and extruded fish feed has been studied under pilot and commercial conditions using mink as model animals. Fish meal produced under pilot-plant conditions at processing temperatures below 70-80 • C (FM1) had protein digestibility of 929 (grams of protein digested per 1000 g protein consumed) compared with 905 when processed at temperatures above 100 • C (FM2). A low-temperature-processed commercial fish meal (CFM1) had protein digestibility of 940 compared with 888 for a standard commercial fish meal (CFM2). Pilot-produced extruded fish feed had protein digestibility of 913 when based on FM1 as the main protein source (95% of total protein) compared with 892 when based on FM2. Commercial extruded fish feed had protein digestibility of 912 when based on CFM1 compared with 871 when based on CFM2. Varying extrusion conditions at the pilot scale, ie temperatures from 100 to 126 • C and moisture contents from 21 to 12%, did not affect protein digestibility. Similarly, under commercial conditions, variation in temperature from 89 to 110 • C and moisture from 24.5 to 19.5% did not affect FDNB-reactive lysine and protein digestibility. The FDNBreactive lysine content and protein digestibility of the extruded feed were less than the values calculated from the ingredient mixture before extrusion. Thus, despite different extrusion conditions not giving different FDNB-reactive lysine and protein digestibility, the total process, ie extrusion, drying and oil coating, caused a reduction.

The effects of partial replacement of fish meal (FM) with meal made from northern krill (Thysanoessa inermis), Antarctic krill (Euphausia superba) or Arctic amphipod (Themsto libellula) as protein source in the diets for Atlantic salmon (Salmo salar L.) and Atlantic halibut (Hippoglossus hippoglossus L.) on growth, feed conversion, macro‐nutrient utilization, muscle chemical composition and fish welfare were studied. Six experimental diets were prepared using a low‐temperature FM diet as control. The other diets included northern krill where 20, 40 or 60% of the dietary FM protein was replaced with protein from northern krill, and two diets where the FM protein was replaced with protein from Antarctic krill or Arctic amphipod at 40% protein replacement level. All diets were iso‐nitrogenous and iso‐caloric. Atlantic salmon grew from 410 g to approximately 1500 g during the 160 day experiment, and Atlantic halibut grew from 345 g to 500–600 g during the 150 day experiment. Inclusion of krill in the diets enhanced specific growth rate in salmon, especially during the first 100 days (P < 0.01), and in a dose–response manner in halibut for over the 150 day feeding period (P < 0.05). Feed conversion ratio did not differ between dietary treatments, and no difference was found in dry matter digestibility, protein digestibility and fish muscle composition. Good growth rates, blood parameters within normal ranges and low mortalities in all experimental treatments indicted that fish health was not affected either Atlantic salmon or Atlantic halibut fed the various zooplankton diets.