Defence behaviour against brood parasitism is deeply rooted in mainland and island scrub-jays BRIAN D. PEER * †, STEPHEN I. ROTHSTEIN * , KATHLEEN S. DELANEY ‡ & ROBERT C. FLEISCHER † *Department of Ecology, Evolution and Marine Biology, University of California, Santa Barbara yGenetics Program, National Museum of Natural History, Smithsonian Institution zDepartment of Ecology and Evolutionary Biology, University of California, Los Angeles (Received 28 August 2005; initial acceptance 12 November 2005; final acceptance 28 June 2006; published online 13 November 2006; MS. number: A10233) When selection pressures for an adaptation relax, the trait may decline, or it may be maintained if there are no fitness costs. The interactions between avian brood parasites and their hosts are dynamic, with selec- tion pressures changing as host or parasite ranges shift, or as parasites switch to new hosts once old hosts evolve defences. The extent to which hosts retain defences in the absence of parasites has important consequences to parasiteehost coevolution. We tested whether island scrub-jays, Aphelocoma insularis, and western scrub-jays, A. californica, have maintained egg ejection behaviour in the absence of brood par- asitism and provide an estimate of how long they have maintained ejection. Island scrub-jays and western scrub-jays ejected 100% of foreign eggs placed into their nests, and genetic analyses revealed no evidence of conspecific brood parasitism that could maintain ejection. Extreme variation in intraclutch egg appear- ance may result in hosts ejecting their own oddly coloured eggs, which would select against the mainte- nance of ejection. However, island, western and Florida scrub-jays, A. coerulescens, also an ejecter, showed less variation than the common grackle, Quiscalus quiscula, a species that has an extremely high level of intraclutch egg variation and may have lost its rejection defence, because it rejects its own oddly coloured eggs. Based on molecular clock analyses of mitochondrial DNA control region and ND2 sequences, island and western scrub-jays split approximately 140 000e151 000 years ago, western and Florida scrub-jays 1 000 000 years ago, and island and Florida scrub-jays 1 250 000 years ago. Ejection behaviour may have been maintained this long in the absence of parasitism, but it is possible that scrub-jays were parasitized as recently as the end of the Pleistocene 10 000 years ago, when cowbirds were more abundant. Neverthe- less, these results indicate that (1) egg ejection is not a costly trait to maintain in scrub-jays, (2) brood par- asites may not be able to alternate between well-defended hosts and hosts that have lost ejection defence following past episodes of parasitism and (3) brood parasites may have to deal with host communities that have well-developed defences by evolving specialized adaptations for a single host species or a small set of species. Ó 2006 The Association for the Study of Animal Behaviour. Published by Elsevier Ltd. All rights reserved. Keywords: Aphelocoma californica; Aphelocoma coerulescens; Aphelocoma insularis; brood parasitism; egg ejection; Florida scrub-jay; host defence; island scrub-jay; molecular clock; western scrub-jay Adaptations that have no current utility are relics of an organism’s evolutionary past. These adaptations persist because the costs of maintaining them are minimal. Conditional behavioural traits that are expressed only in response to specific stimuli are most likely to be retained after selection has relaxed, because once the necessary stimuli have been removed, their lack of expression makes them neutral or nearly so. This has been demonstrated in prey species that show defences for thousands of years in the absence of particular predation pressures (Coss & Goldthwaite 1995; Byers 1997; Rydell et al. 2000). When such adaptations persist, they will be expressed when the selection pressures recur even though the pressures may appear to be novel ones from the perspective of hu- man history. Correspondence: B. D. Peer, Department of Biological Sciences, Western Illinois University, Macomb, IL 61455, U.S.A. (email: bd-peer@wiu. edu). 55 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, 2007, 73, 55e63 doi:10.1016/j.anbehav.2006.06.005