898—JOURNAL OF FOOD SCIENCE—Volume 62, No. 4, 1997 Handling Stress and Storage Temperature Affect Meat Quality of Farmed-raised Atlantic Salmon (Salmo Salar) T. SIGHOLT, U. ERIKSON, T. RUSTAD, S. JOHANSEN, T.S. NORDTVEDT, and A. SELAND ABSTRACT Salmon slaughtered by standard routines (control) or stressed by con- finement for 10 min before stunning and then stored at 0.4 or 3.3°C for 9 days were compared. Handling stress led to lower muscle phospho- creatine (p0.001), adenosine-5'-triphosphate (p0.05) and shorter pre-rigor period. Storage temperature affected external quality index, white muscle pH and K-value (degradation products of ATP). Stress produced a softer fillet (p0.001). A lower breaking strength (p0.01) was found in fish stored at 0.4°C. Sensory tests distinguished the control/ stress groups within the 0.4°C chilling regime and the 0.4°C/3.3°C chill- ing groups within the control regime. Stress caused a lower score for texture (p0.05) both at 0.4 and 3.3°C and for odor at 3.3°C in a de- scriptive sensory test. No detectable effects of stress or storage temper- ature were found on flavor or color. Key Words: slaughtering, stress effects, Salmo salar, high energy phos- phates, sensory quality INTRODUCTION AFTER DEATH, fish pass through several stages: rigor mortis, dissolution of rigor mortis, autolysis and bacterial spoilage. Au- tolytic changes occur as a result of enzymatic changes within the muscle. Losses of respiratory capability and membrane breakdown are primary factors initiating autolysis (Bremner and Hallett, 1989). Storage temperature is the main factor affecting freshness and quality of fish. Fish should be chilled quickly to the temperature of melting ice soon after capture and maintained at that temperature (FAO/WHO 1977; FAO 1982). However, ante-mortem handling has appeared to be a factor affecting fish quality. This is a well-known effect in mammals and birds (Gregory, 1994). It is more important due to increased produc- tion of cultured fish which enables influencing ante-mortem han- dling procedures to a much greater extent than in traditional fisheries (Haard, 1992). Capture, handling and transportation are traumatic procedures and may cause considerable physiological reactions in fish. This general adaptation syndrome (GAS) like stress syndrome (Selye, 1956) causes release of adrenalin and cortisol. This is followed by secondary changes such as increased muscle activity, mobi- lization of energy stores in muscle and liver and changes in acid- base balance. There may also be increased blood plasma ion concentrations and reduced tissue water content for fish in sea- water (Mazeaud et al., 1977; Mazeaud and Mazeaud, 1981; Wood, 1991). Stress and muscle activity during transport, netting and stun- ning of fish may shorten the time for onset of rigor mortis. Stressed fish may develop stronger rigor mortis (Nakayama et al., 1992) which may also affect texture. A more rapid rise in K-value in stressed fish has been reported (Izquierdo-Pulido et al., 1992; Lowe et al., 1993). Handling and processing of fish Authors Sigholt and Erikson are with SINTEF Applied Chemistry, Aquaculture Centre, N-7034 Trondheim, Norway. Author Rustad is with the Dept. of Biotechnology, Norwegian Univ. of Science & Technology (NTNU), N-7034 Trondheim, Norway. Authors Jo- hansen and Nordtvedt are with SINTEF Energy, Refrigeration & Food Engineering, N-7034 Trondheim, Norway. Author Seland is with SINTEF Hydrotechnical Laboratory, N-7034 Trondheim, Nor- way. during rigor can result in loss of quality and lower fillet yield (Lave ´ty, 1984). This investigation was carried out in a commercial salmon slaughtery. Our main objectives were to test the effects of pre- slaughtering stress, and of 0.4°C and 3.3°C chilling, on the quality of farmed Atlantic salmon after 9 days ice storage. Measurements of white muscle high-energy phosphates and pH together with the development of rigor mortis were used to evaluate handling stress. An external quality index (based on odor, flesh firmness, mucus, skin appearance, eyes and gills in raw fish), muscle pH, water- and salt-soluble proteins, water-holding capacity, meat texture, K- value and sensory analysis were used to evaluate effects of treat- ment on meat quality. MATERIALS & METHODS Facilities and fish Atlantic salmon (Salmo salar) from a third-generation bred stock (Gjedrem et al., 1991) were used. The experiment was done in a com- mercial salmon slaughtery and the fish were sampled from the first fish harvested in production of iced salmon on May 10, 1994. The fish had been fed a diet containing 40% protein, 30% fat with a gross energy content of 23 MJ kg -1 (digestible 21.8 MJ kg -1 ) according to manufac- turer specifications. Mean weight (whole fish) was 5.1  1.1 kg (n=33). Mean fat content in a sample taken from the fillet in the center of the fish including red muscle and belly flap was 19  2% (n=12) (Bligh and Dyer, 1959, as modified by Hardy and Keay, 1972). Before slaugh- ter, the fish had been starved for 12 days. The fish were netted from the cage and transported in a commercial well-boat for 4 hr and kept in the boat additional 4 hr at the quay before slaughtering. The well boat contained 25 MT of salmon at a density of 125 kg m -3 . Continuous seawater exchange provided a measured mini- mum O 2 level of 6.5 mg L -1 and NH 3 of 4.6 μgL -1 during transport. Based on measurement of total carbonate, pH and salinity, maximum CO 2 (aq) was calculated as 1.38 mg L -1 according to Gieskes (1974). Seawater temperature was 6.5–7.4°C. Experimental design A2  2 factorial design with two levels of handling stress and two storage temperatures was used. Control groups were netted from the well-boat with water-filled nets, stunned by CO 2 in two parallel 1.5 m 3 tanks, then both gill arches were cut and the fish were bled in seawater (5.8°C, 31.7 ppt salinity). Stunning was effective within 2 min and the fish were kept in the tank for a maximum of 5 min. Water quality in the stunning tank was: salinity 29.9 ppt, O 2 3.04–4.58 mg L -1 , CO 2 (aq) 706– 820 mg L -1 and pH 3.2–3.8. Immediately before bleeding, the fish were individually marked regarding experimental treatment. The fish on the 0.4°C chilling regime were sampled after being chilled with refrigerated seawater while those on the 3.3°C chilling regime were sampled before passing through the refrigeration tank. Both groups were packed (four fish/box) in styropore boxes, iced and transported to the laboratory and then stored in temperature-controlled rooms. All fish were packed before any noticeable rigor mortis. Mean temperatures in the storage periods were 0.4 (-0.1 to 6.8)°C and 3.3 (-0.6 to 5.4)°C in the two chilling regimes (Fig. 1). During the first 2 days, including transport, the mean temperature was 0°C and 1.2°C in the 0.4°C and 3.3°C chilling groups. Stressed groups were handled as control groups except that they were netted into a 1 m 3 tank filled with seawater and kept there for 10 min before they were netted into the stunning tanks. The netting and crowding in stagnant water induced vig- orous muscular activity. Measured oxygen level during the stress episode was 4.6–6.8 mg L -1 . The treatment was intended to simulate conditions