Can smolting be assessed by food intake in steelhead trout, Oncorhynchus mykiss (Walbaum)?

Salmonid smolts under cultivationmust often be able to tolerate abrupt transfer from freshwater (FW) to seawater (SW). Traditionally, smolt readiness for SW transfer has been estimated using terminal methods. In view of the increasing regard for animal welfare it would be desirable to develop a reliable, non-lethal method for assessing smolt readiness in hatcheries and for research. One potential non-lethal smolt index could be the measurement of changes in food intake. In Atlantic salmon, Salmo salar L., only some individuals of a population are able to feed soon after transfer to sea (Usher, Talbot & Eddy 1991; Arnesen, Johnsen, Mortensen & Jobling 1998), likely those individuals that have a better capacity to osmoregulate in SW. However, earlier studies did not reveal a correlation between feed intake and physiological parameters in SW (Damsg rd & Arnesen1998; Arnesen et al.1998). Nobody has heretofore attempted to correlate feed intake and smolting physiology in FW, or feed intake in FWand physiology in SW.The purpose of the present experiment was to monitor whether feed intake could be used as an indicator of smolting with respect to physiology in steelhead trout, Oncorhynchus mykiss (Walbaum), during the spring. The experiment was carried out at the Hat¢eld Marine Science Center (Newport, OR, USA), Oregon State University, between14 April and 22 June1999. The ¢sh were 11 steelhead from Eagle Creek (WA) strain, o¡spring of wild parents. Fish were held in two holding tanks with single pass, £owthrough dechlorinated (by activated carbon) tap water. Each of the ¢ve separate experiments (referred to as runs) consisted of a 10-day acclimation period of 30 ¢sh in each of the six experimental tanks (1m in diameter, water depth 0.33m) with a water £ow of 6 Lmin . The sampling days for each run are given in Table 1. The ¢sh weight ranged from 50.2 to 149.8 g during the ¢rst run in mid-April and from 88.8 to 311.3 g during the last run in mid-June. On the10th day after acclimation for each run, three tanks were switched to running UV-sterilized SW. Maximum salinities (Fig. 1) were achieved in 2 h. Salinity and temperature were measured using aYSI model 30 (YSI,Yellow Springs, OH, USA). Because of the di¡erences in oceanic upwelling, the temperature and salinity varied between runs; the temperature of FW also rose between spring and summer (Fig. 1). The ¢sh were fed by hand twice daily according to a hatchery feeding table. Sampling was accomplished by netting 15 ¢sh from each tank as quickly as possible into bu¡ered (NaHCO3) tricainemethanesulphonate (MS-222, 200mg L ). SW-exposed ¢sh were sampled 6 h after SW exposure started. FW control tanks were sampled immediately before the SW tanks. After anaesthetization, ¢sh were dried, weighed (to 0.1g) and measured (total length to1mm). A blood sample was then withdrawn from the caudal vessel via a vacutainer. Plasma was separated and stored at 80 1C for lateranalysis of sodium, potassium (NOVA 1 Na, K analyser, Nova biomedical,Waltham,MA, USA) and cortisol. The plasma cortisol concentration was analysed in duplicate by radioimmuno-assay