ORIGINAL PAPER A ´ . Miklo´si ® Zs. Gonda ® D. Osorio ® A. Farzin The effects of the visual environment on responses to colour by domestic chicks Accepted: 15 January 2002 / Published online: 13 February 2002 Ó Springer-Verlag 2002 Abstract It is well known that development of vision is affected by experience, but there are few studies of en- vironmental effects on colour vision. Natural scenes contain predominantly a restricted range of reflectance spectra, so such effects might be important, perhaps biasing visual mechanisms towards common colours. We investigated how the visual environment affects colour preferences of domestic chicks (Gallus gallus), by training week-old birds to select small food containers distinguished from an achromatic alternative either by an orange or by a greenish-blue colour. Chicks that had been raised in control conditions, with long-wavelength- dominated reflectance spectra, responded more readily to orange than to blue. This was not due to avoidance of blue, as increasing saturation enhanced the chicks’ preference for the same hue. The advantage of orange was, however, reduced or abolished for chicks raised in an environment dominated by blue objects. This indi- cates that responses to coloured food are affected by experience of non-food objects. If colours of ordinary objects in the environment do influence responses to specialised visual signals this might help explain why biological signals directed at birds are often coloured yellow, orange or red; long-wavelength-dominated spectra being more prevalent than short-wavelength- dominated spectra. Keywords Chicken ® Colour ® Learning ® Conditioning ® Visual environment Introduction It is a familiar observation that reflectance spectra of visual signals directed at birds are often dominated by long wavelengths, giving colours that we see as red, yellow or orange. Examples include fruit (Willson and Whelan 1990), flowers (Grant and Grant 1966), warning colours of toxic animals (Schuler and Roper 1992), plumage used in sexual displays (Olson and Owens 1998), and nestlings’ gape (Kilner and Davies 1998). At least for flowers this bias is not probably not due to a general constraint on their coloration, as flowers polli- nated by bees are much more likely to be blue or violet (Giurfa et al. 1995). A possible reason for the predominance of long- wavelength colours in visual signals is that birds find them especially noticeable or memorable. This may de- pend in part upon a genetic predisposition; selective breeding can modify the colour preferences of quails, C. coturnix (Kovach 1971, 1980; Kovach and Wilson 1993). However, the visual environment could influence the efficacy of colour signals in two ways. Signals will be conspicuous if they differ from their background, gen- erally green vegetation, leaf-litter and so forth, but conversely, if the eye and vision are adapted to operate with such background stimuli unusual signals may be ‘overlooked’. For mammals it is clear that the environment affects development of vision. Thus, restricting kittens’ experi- ence to stimuli of one orientation leaves them less sen- sitive to others (Blakemore and Cooper 1970). In natural environments the distribution of orientation stimuli is not uniform, with horizontals and verticals being com- moner than oblique orientations (Baddeley and Han- cock 1991). This anisotropy could account for humans being better at detecting verticals and horizontals than oblique lines (Campbell and Kulikowski 1966). J Comp Physiol A (2002) 188: 135–140 DOI 10.1007/s00359-002-0284-z A ´ . Miklo´si ® Zs. Gonda ® D. Osorio (&) ® A. Farzin School of Biological Sciences, University of Sussex, Brighton, BN1 9QG, UK E-mail: d.osorio@sussex.ac.uk Tel.: +44-1273-877440 Fax: +44-1273-678433 Present address:A ´ . Miklo´si Department of Ethology, Eo¨tvo¨s University, Go¨d, Ja´vorka S.u.14, 2131 Hungary Present address: Zs. Gonda National Institute of Chemical Safety, 1450 Budapest. Pf. 22, Hungary