Compensatory growth following various time lengths of restricted feeding in rainbow trout (Oncorhynchus mykiss) under summer conditions By H. Sevgili 1 , B. Hos ßsu 2 , Y. Emre 1 and M. Kanyılmaz 1 1 Mediterranean Fisheries Research, Production and Training Institute, Antalya, Turkey; 2 Ege University, Faculty of Fisheries, Department of Aquaculture, Izmir, Turkey Summary This study was conducted to determine the effects of various time lengths of restricted feeding at 0.5% of body weight on compensatory growth (CG) in rainbow trout under summer conditions. Seven treatments with triplicate tanks consisted of control (C) fed to satiation over 98 days and the remain- der being one (R1) to six (R6) weeks of restriction and then refeeding for the remaining 8 weeks of the experiment. At the end of the experiment R1 and R2 were able to catch up with C. Repeated measures ANOVA suggested a convergence in body mass but not in body length (structure), whereas there was an association between mass and structural CG responses. Hyperphagia and transiently better food utilisa- tion were main mechanisms of the observed CG. Organoso- matic indices of the restricted groups were significantly reduced at the end of the restriction periods, but were restored to the control fish levels by the end of the refeeding period. There was a linear increase in body moisture and a decrease in lipid and lipid/lean body mass ratio with the severity of the restriction periods, but these trends vanished by the end of refeeding. The findings of the present experi- ment suggest that restricted feeding and the following reali- mentation to elicit CG as a management tool can be used in rainbow trout, but for no more than 2 weeks under summer conditions. Introduction Compensatory growth (CG), a growth spurt phenomenon following a growth retardation period, has been suggested as a rearing tactic in numerous aquatic organisms (Ali et al., 2003). Dietary restriction and starvation with a single phase and subsequent satiation feeding or cycled restriction and full feeding have been principally applied strategies (Hayward et al., 1997; Nikki et al., 2004; Eroldo gan et al., 2006; Ali et al., 2010; Xiao et al., 2011). During either restriction or starvation, substantial changes occur in body nutrient levels and organ sizes as a result of the mobilization of energy sources, namely, lipids (Rueda et al., 1998; Gaylord and Gatlin, 2000; B elanger et al., 2002; Ali et al., 2003; Cho, 2005; Cho et al., 2006). Upon refeeding, fish show good appetites or better utilisation of food to compensate their weight loss (Bull and Metcalfe, 1997; Hayward et al., 1997; Boujard et al., 2000; Qian et al., 2000; Ali et al., 2003; Nikki et al., 2004; Xiao et al., 2011). Water temperature has been found to have a remarkable effect on eliciting CG response in growth-depressed fish (Mortensen and Damsg  ard, 1993; Wang et al., 2000; Maclean and Metcalfe, 2001; Mylonas et al., 2005; Person-Le Ruyet et al., 2006). Attempted food deprivations or restrictions to invoke CG in fish at high summer water temperatures may, however, create permanent depressions in the fish body due to a high metabolic rate (Morgan and Metcalfe, 2001). In support of this notion, Hokanson et al. (1977) and Ma¨kinen (1994) found that the optimum temperature for rainbow trout growth was between 14 and 17.3°C, and that weight loss due to fasting was linearly elevated with the temperature, in agreement with findings in other species (Brett et al., 1969; Wang et al., 2000; van Dijk et al., 2002). Moreover, there is evidence indicating that fish that undergo a period of low nutrition show a tendency of lower temperature optima than well-fed fish (Brett et al., 1969; van Dijk et al., 2002, 2005) and that during realimentation the restricted fish select lower temperatures for some time (van Dijk et al., 2002, 2005). So far, the CG in rainbow trout has been studied mostly between suboptimum and optimum temperatures (3–16°C) using food deprivation (Dobson and Holmes, 1984; Kindschi, 1988; Nikki et al., 2004; Mylonas et al., 2005; Blake and Chan, 2006; Blake et al., 2006; Bhat et al., 2011; G€ uzel and Arvas, 2011). Nevertheless, there appear to be differences between starvation and restricted feeding in terms of invoking CG; the former stimulates mostly hyperphagia, and the latter a combination of hyperphagia and improved feed utilisation (Miglavs and Jobling, 1989; Johansen et al., 2001; Ali et al., 2010; Jiwyam, 2010; Imsland and Gunnarsson, 2011). Restricted feeding at near maintenance level for a certain time period appears to be more advantageous as compared to starvation, but the degree of restriction severity for subse- quent CG in rainbow trout, particularly during high summer temperatures, has not been documented. This is particularly important for countries like Turkey, where supra or maxi- mum critical temperatures can pose significant problems, including the suspension of operations (Atasoy and S ßenes ß, 2004; Alpaslan and Pulatsu¨, 2008). The present experiment was therefore designed to deter- mine the effects of various lengths of single-phase food restriction on the subsequent CG response, feed intake, organ indices and whole body composition of rainbow trout under summer conditions. Materials and methods Fish and rearing conditions The study was conducted at the Kepez Unit of Mediterra- nean Fisheries Research and Training Institute, Antalya, U.S. Copyright Clearance Centre Code Statement: 0175-8659/2013/2906–1330$15.00/0 J. Appl. Ichthyol. 29 (2013), 1330–1336 © 2013 Blackwell Verlag GmbH ISSN 0175–8659 Received: August 7, 2012 Accepted: November 25, 2012 doi: 10.1111/jai.12174 Applied Ichthyology Journal of