Journal of Fish Biology (2013) 82, 1423–1432 doi:10.1111/jfb.12063, available online at wileyonlinelibrary.com Shifts in δ 15 N signature following the onset of exogenous feeding in rainbow trout Oncorhynchus mykiss : importance of combining length and age data A. L. Liberoff*†, C. M. Riva Rossi*, M. L. Fogel‡, J. E. Ciancio* and M. A. Pascual* *Centro Nacional Patag´ onico–CONICET, Boulevard Brown 2915, 9120, Puerto Madryn, Chubut, Argentina and ‡Geophysical Laboratory, Carnegie Institution of Washington, Washington, DC, 20015, U.S.A. (Received 24 March 2012, Accepted 16 January 2013) The δ 15 N isotopic change of recently emerged rainbow trout Oncorhynchus mykiss due to diet shift from yolk sac to exogenous feeding was evaluated in a field study. The fit of a general model including both fish length and age in days as co-variables indicates that the specific δ 15 N of individual fish at any given time along the ontogeny is determined by its growth trajectory. The results suggest that estimations based on fish size alone could bias data interpretation and maternal origin determinations in partially migratory salmonids.  2013 The Authors Journal of Fish Biology  2013 The Fisheries Society of the British Isles Key words: fry; maternal origin; partial anadromy; salmonids; stable isotopes. Stable isotope analysis is a powerful tool to assess energy flow and nutrient sources in food webs (Finlay et al ., 2002), estimate trophic interactions (Post, 2002), track animal movements (Hobson, 1999; Rasmussen et al ., 2009) and study diet shifts (Vander Zanden et al ., 1998). In order to study ecological processes through isotope signatures, ecologists have devoted much effort in constructing models that describe the pattern of isotopic change in different tissues (Boecklen et al ., 2011). To date, there are four major classes of isotope incorporation models: growth-dependent mod- els which track changes in isotopic ratios as a function of the growth of the consumer (Fry & Arnold, 1982; Maruyama et al ., 2001), time-dependent models which track changes in isotopic ratios as a function of time (Hobson & Clark, 1992; O’Brien et al ., 2000; Martínez del Rio & Wolf, 2005), growth-and-time-dependent models which are the latest modelling attempts at examining joint contributions of growth and metabolism to isotopic turnover (Hesslein et al ., 1993; Carleton & Martínez del Rio, 2010) and multi-compartment models which use distinct isotope pools to exam- ine isotopic turnover (Ayliffe et al ., 2004; Cerling et al ., 2007; Carleton et al ., 2008). †Author to whom correspondence should be addressed. Tel.: +542804451024 ext. 332; email: liberoff@ cenpat.edu.ar 1423  2013 The Authors Journal of Fish Biology  2013 The Fisheries Society of the British Isles