How and why are uniformly polarization-sensitive retinae subject to polarization-related artefacts? Correction of some errors in the theory of polarization-induced false colours

Abstract

If the photoreceptors of a colour vision system are polarization sensitive, the system detects polarization-induced false colours. Based on the functional similarities between polarization vision and colour vision, earlier it was believed that a uniformly polarization-sensitive (insect) retina (UPSR)—in which receptors of all spectral types have the same polarization sensitivity ratio and microvilli direction—cannot detect polarization-induced false colours. Here we show that, contrary to this belief, a colour vision based on a UPSR is subject to polarization-related artefacts, because both the degree and the angle of polarization of light reflected from natural surfaces depend on wavelength. Our second goal is to correct certain errors in the theory of polarizational false colours. The quantitative estimation of the influence of polarization sensitivity on colour vision was recently motivated by the suggestion that certain Papilio butterflies detect such false colours. The theoretical basis of this subject is to calculate the colour loci in the colour space of a visual system from the quantum catches of polarization-sensitive receptors of different spectral types. Horváth et al. (J. Exp. Biol. 205 (2002) 3281) gave the first exact mathematical and receptor-physiological derivation of formulae for these calculations. Here we prove that the two formulae given earlier by others are inappropriate or erroneous. This, however, does not influence the validity of the experimental data and the principal conclusions drawn about the colour vision and polarization sensitivity in Papilio butterflies.