Experimental evidence of homing to site of incubation by mature sockeye salmon, Oncorhynchus nerka THOMAS P. QUINN, IAN J. STEWART & CHRISTOPHER P. BOATRIGHT Fisheries Research Institute, School of Aquatic and Fishery Sciences, University of Washington (Received 5 October 2005; initial acceptance 21 January 2006; final acceptance 14 March 2006; published online 6 September 2006; MS. number: A10262R) Fish of the family Salmonidae (salmon, trout and charr) are famous for their ability to learn (‘imprint’) olfactory information as juveniles and to use those odour memories as adults to home to their natal site for reproduction years later. However, the spatial scale at which salmonids home has not been determined, and this is important not only for an understanding of the imprinting and homing processes but also be- cause homing determines the spatial scale at which salmon populations are structured. To test the hypoth- esis that salmon home to specific habitat units within a single small creek, we induced banding patterns on the otoliths (ear bones) of pre-hatch sockeye salmon embryos by altering the temperatures at which they were incubated. Thermally marked embryos were buried in a small pond associated with Hansen Creek, in southwestern Alaska, and allowed to complete development and migrate. In the year when the salmon were expected to return to the creek, examination of otoliths from carcasses revealed that 12 of 324 salmon in the pond were marked, whereas all 138 salmon examined in the creek were unmarked. This distribution of marked fish indicates that more of the experimental salmon returned to the pond than would have occurred by chance, although we cannot be certain that all surviving salmon did so. These results contrib- ute to the growing literature indicating the capacity for very fine-scale populations structure in salmon. Ó 2006 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved. The tendency of animals to migrate back to their natal site for reproduction after feeding migrations is widely distrib- uted among diverse animal taxa (Papi 1992; Dingle 1996). In animals with lengthy migrations, such homing behav- iour is a prerequisite for the structuring of the species into discrete populations. In some cases homing must be inferred from studies of population genetics because defin- itive marking studies are difficult to accomplish (e.g. sea turtles: Bowen et al. 1993). The teleost family Salmonidae (salmon, trout and charr) has been one of the models for research on the mechanisms of homing and the differen- tiation of populations, although many fish show this behaviour (Lucas & Baras 2001). Anecdotal evidence indi- cates that centuries ago astute observers realized that salmon returned to their natal stream to spawn (Nordeng 1989), and by the 1930s convincing scientific evidence was accumulating to support this idea (e.g. Foerster 1936). In addition to recognizing the homing ability of salmon, these early investigators (e.g. Rich 1939) under- stood that the homing of salmon generates populations with unique dynamics (i.e. levels of abundance and recruitment patterns) and life history traits, necessitating management and conservation as discrete units. Evidence has accumulated that salmon populations evolve differences in a wide variety of adaptive traits (reviewed by Ricker 1972; McDonald 1981; Taylor 1991; Quinn 1999), and research has also revealed many aspects of the mechanisms of olfactory imprinting and homing (Nevitt et al. 1994; Dittman et al. 1997). However, almost all studies of homing have evaluated the behaviour of salmon released from hatcheries at the smolt (seaward migrant) stage (e.g. Quinn et al. 1991; Candy & Beacham 2000). A hatchery is an unnatural environment for incubation and rearing (hence imprinting) and it also con- strains the exploratory behaviour of returning adults (Griffith et al. 1999); unfortunately, there is little informa- tion on the homing patterns of wild salmon. In particular, it is unclear how the process of homing grades into the processes of spawning site selection by females (Hoopes 1972; Kondolf & Wolman 1993) and search for mates by males (Healey & Prince 1998). Spawning grounds are heterogeneous, and it is unclear whether the perennial Correspondence and present address: T. P. Quinn, Fisheries Research Institute, School of Aquatic and Fishery Sciences, Box 355020, Univer- sity of Washington, Seattle, WA 98195, U.S.A. (email: tquinn@u. washington.edu). I. J. Stewart is now at the National Marine Fisheries Service, Northwest Fisheries Science Center, 2725 Montlake Boulevard East, Seattle, WA 98112, U.S.A. 941 0003e 3472/06/$30.00/0 Ó 2006 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved. ANIMAL BEHAVIOUR, 2006, 72, 941e949 doi:10.1016/j.anbehav.2006.03.003