The responses of the digestive proteases trypsin and chymotrypsin and protein metabolism to differences in feed protein quality were investigated in Atlantic salmon (Salmo salar L.). Two sets of experimental feeds were produced. Each set of high and low quality feeds was provided to either 150 g or 2 kg salmon. Protein in the high quality feeds had significantly higher percentages of free (reactive) sulphydryl (SH) groups than the corresponding feeds based on low quality meals. After 90 days feeding, groups given high and low quality feeds did not differ in their specific growth rates (SGR) in either experiment. However, feed conversion efficiency (FCE) was significantly different between the high and low quality feed groups in 2 kg salmon, where the difference between the high and low feed protein qualities was larger, 10% versus 4% SH/[SH + (S-S)] in 150 g salmon. Higher FCE was preceded by significantly higher trypsin and chymotrypsin specific activities on day 60. SGR, in general, changed after the first month and was stable during the last 2 months in both experiments. Concurrently, both trypsin (T) and chymotrypsin (C) decreased with an increased activity ratio of trypsin to chymotrypsin (T/C ratio), and resulting in significantly lower T/C ratio on day 90 in salmon feeding on high quality feeds in both sizes of fish. Differences in FCE were associated with significant differences in levels of total free amino acids (TFAA) in the plasma and the white muscle, as well as in the ratio of essential to non-essential free amino acids (EAA/NEAA ratio), free hydroxyproline, and RNA in the white muscle. Interestingly, after 3 days starvation (day 93), 5-7 h postprandial EAA/NEAA ratio in the plasma was significantly lower in the high quality diet groups in both experiments. Trypsin specific activity inversely correlated with muscle TFAA levels in 2 kg salmon, concurrent with higher muscle levels of RNA, lower free hydroxyproline and higher FCE in fish fed higher quality diets. KEY WORDS

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 dioxin, dioxin-like polychlorinated biphenyl (DL-PCB), fat, and dry matter partitioning during fishmeal production have been studied in pilot scale. Most of the dry matter and lipid content in the fishmeal could be ascribed to the press cake intermediate product. Dioxins and DL-PCBs are fat-soluble compounds, and the process partitioning is reflected by the fat partitioning data. Enzyme and heat treatment of the press cake and stickwater concentrate did not improve fat separation. Soybean oil extraction of the press cake reduced the dioxin and DL-PCB content by 97%. Less exchange of fatty acids was observed (56-72%). Combined with fat separation of the stickwater concentrate, the applied process conditions were able to give a fishmeal decontamination rate higher than hexane and isopropanol extraction of the fishmeal. Quantification of fat content based on chloroform/methanol extraction was found to be the best protocol to estimate fat partitioning and decontamination effects. The oil extraction process requires further optimization, but has several advantages compared to organic solvent extraction. These include easy implementation in an existing fishmeal processing line, use of a safe and nonflammable extraction medium, and expected lower investment and operation costs. A new integrated fishmeal and fish oil production and decontamination process line is proposed.

No abstract

The racemization kinetics of aspartic acid in heat-treated whole herring have been studied under conditions of treatment comparable to those that may occur in processing of fish meal. D-Aspartic acid content in the samples has been measured by RP-HPLC with precolumn automatic derivatization. The major parameters affecting the rate of racemization of aspartic acid k(Asp) have been demonstrated to be temperature (elevation of temperature from 95 to 120 degrees C resulted in an increase of k(Asp) from 0.46 to 3.39x10(-3) min(-1)), moisture of the raw material (reduction of the moisture content of the raw material from 80 to 15% lowered k(Asp) measured at 95 degrees C from 0.46 to 0.06x10(-3) min(-1)), and to a lesser extent, pH (k(Asp) at 95 degrees C was lowered from 0.46 to 0.37x10(-3) min(-1) following a decrease of pH from 7.0 to 4.0). No significant effects on the racemization rate of aspartic acid was observed for reducing the oxygen pressure to 0.8%. The results from the present study show that the content of D-aspartic acid in fish material is a function of heat exposure and may be used to predict the thermal history of fish meal.

No abstract