Color discrimination and dichromacy in the horse (Equus caballus)

C.A. Hall¹, H.J. Cassaday², C.J. Vincent³ and A.M. Derrington³

¹School of Animal, Rural and Environmental Sciences, Nottingham Trent University, Nottingham, UK
²School of Psychology, University of Nottingham, Nottingham, UK
³Psychology, Brain and Behavior, University of Newcastle upon Tyne, Newcastle upon Tyne, UK

The findings of this paper demonstrate a correlation between the results of color discrimination training and cone excitation ratios in the horse. Six horses were trained to discriminate fifteen different colors from shades of gray. The behavioral data obtained was compared with cone excitation ratios that were calculated for each of the colors, based on spectral peak values of 429 nm and 545 nm. The horses were first trained to discriminate each color from the grays to a criterion of ten consecutive correct choices, with the order of presentation being determined by the spectral appearance of the colors to the trichromatic human. In order to make lightness cues irrelevant, the chromatic stimuli were randomly located within a chequerboard of shades of gray. All six horses successfully completed the training, reaching the learning criterion for all fifteen colors.

The relationship between the behavioral data obtained, and the spectral sensitivity of the cone photopigments of the horse, was then investigated. The reflectance spectrum of each of the colors was measured and cone excitations were calculated. Cone excitation ratios for the colors and the grays were compared and the difference between them used as an indication of the effect that each color had on the visual system of the horse. Significant correlations between the behavioral data (accuracy, repeated errors and latency of approach) and the cone excitation data were found for all colors.

In addition to providing further evidence for dichromacy, the results of this study demonstrate that cone excitation calculations can be used to predict the ease with which the horse can discriminate colors. Such information will facilitate further investigation into color perception in the horse as well as offering greater insight into the ways in which color vision in the horse differs from that of the trichromatic human.

Paper presented at Measuring Behavior 2005 , 5^th International Conference on Methods and Techniques in Behavioral Research, 30 August - 2 September 2005, Wageningen, The Netherlands.

© 2005 Noldus Information Technology bv