Aquatic Diagnostic Services, Atlantic Veterinary College,University of PEI, Charlottetown, Prince Edward Island, CanadaSummaryProlonged starvation resulting in sublethal condition factorvalues was hypothesized to have a detrimental effect on short-termgrowthcapacityuponrefeeding.Cod(Gadusmorhua)werefood-deprived and their length and mass measured beforerefeeding and after 3, 6, 9 and 12 weeks of ad libitum feeding.Total mass increase during the first 3 weeks of feeding wasgreatestinfishwithahigherinitialconditionfactor.Thereversesituation was observed during the last 3 weeks of feeding.Specificgrowthratepeakedintheperiodfromweek4toweek6,exceptincodwiththehighestconditionfactorforwhichasteadydecline in specific growth rate was observed, and was notinfluenced by the condition factor at the start of the feedingperiod. Total mass increase over 12 weeks was also notinfluenced by initial condition factor. Thus by the end of theexperiment, condition factors were lowest in fish with initiallylowcondition factors. The hepatosomatic indexandgonadoso-matic index did not differ at the end of the experiment, but theproportion of mature cod increased with increasing initialcondition factor. A disease outbreak caused significant mortal-ities among fish shortly after the start of the feeding period.Forty-one percent of the fish had died after 84 days. Nomortality was observed among fish that had started theexperiment with the highest condition factor. Mortality wasinversely related to initial condition factor. Growth was exam-ined for survivors exclusively. Poor condition in wild fish mayincreasevulnerabilitytodiseasesandcompensatorygrowthmaynot allow cod with low condition factors to fully recover unlessfood availability remains high over a long period of time.IntroductionWhen refed after a period of slow or negative growth caused bymoderate or severe starvation, fish will restore their energyreserves.Theymayexhibitashortperiodofexceptionallyfastergrowth known as compensatory or catch-up growth by whichfood-deprived fish will reach the size of continuously fedconspecifics. During early life stages food-deprived fish mayeven grow to larger sizes than continuously fed fish (Haywardet al.,1997;HaywardandWang,2001).Catch-upgrowthiswelldocumented in fishes, particularly in freshwater species andsalmonids, including Atlantic salmon Salmo salar (Nicieza andMetcalfe, 1997), rainbow trout Oncorhynchus mykiss (Dobsonand Holmes, 1984), andarctic charr Salvelinus alpinus (Miglavsand Jobling, 1989). Previous studies on catch-up growth havefocused on its impacts on life-history strategies (Nicieza andMetcalfe, 1997) and feeding hierarchies (Jobling and Koskela,1996) as well as on the underlying physiological processes(Be´langer et al., 2002). Hayward et al. (1997) examined com-pensatory growth in view of determining whether restrictedrations elicited hyperphagia and could be used advantageouslyto increase production in aquaculture facilities by using anoptimal feeding regime which would bank on catch-up growthmechanisms. In most studies, short periods of food deprivationwere considered, from a few days to 1 or 2 weeks (Dobson andHolmes, 1984; Ali and Wootton, 2001; Zhu et al., 2001, 2003),butothersusedperiodsaslongas100–120 days(Haywardet al.,1997; Hayward and Wang, 2001).Wild fish may experience prolonged periods of food-deprivation. For instance, Atlantic cod (Gadus morhua) ceasefeeding for as much as 70 days during prespawning and muchof the spawning period (Fordham and Trippel, 1999). Theyalso have to cope with periodic food shortages, particularly inwinter (Dutil et al., 2003). When shortages occur, cod mustdraw energy from body reserves accumulated during periodswhen food was available (Black and Love, 1986; Lambert andDutil, 1997b). As a result, cod exhibit marked seasonal andannual variations in energy reserves (Lambert and Dutil,1997a; Schwalme and Chouinard, 1999), and growth rate(Chouinard and Fre´chet, 1994; Dutil et al., 1999), with fastergrowth occurring in summer and autumn and peak somaticcondition factors being reached in autumn. These variationsare particularly pronounced in cold water stocks (Dutil et al.,1999; Dutil and Brander, 2003). Low condition factors reachedfollowing a prolonged period of fasting are expected topromote catch-up growth when food becomes more abundant.The success of the cod growout industry in eastern Canadamay build on this increased capacity for growth in post-spawning cod caught inshore in early summer; cod more thandouble their body weight in 4–6 months when provided with anearly unlimited source of raw wild mackerel and capelin(Murphy, 2002). On the other hand, metabolic capacities aredepressed and some tissues are significantly atrophied follow-ing an extended period of starvation (Dutil et al., 1998;Be´langer et al., 2002; Marti´nez et al., 2003) which maypotentially impair catch-up growth in more emaciated indi-viduals. Few studies have considered long periods of fooddeprivation in catch-up growth studies (Weatherley and Gill,1981; Pedersen and Jobling, 1989; Paul et al., 1995) and nonehas examined whether poor condition affected the time courseof catch-up growth upon refeeding.The present study assesses whether poor condition resultingfrom prolonged food deprivation may prevent catch-upgrowth to occur. Prolonged starvation resulting in sublethal