Crypsis increases survival by reducing predator detection. Xenopus laevis tadpoles decode light properties from the substrate to induce two responses: A cryptic coloration response where dorsal skin pigmentation is adjusted to the colour of the substrate (background adaptation) and a behavioural crypsis where organisms move to align with a specific colour surface (background preference). Both processes require organisms to detect reflected light from the substrate. We explored the relationship between background adaptation and preference and the light properties able to trigger both responses. We also analysed which retinal photosensor (type II opsin) is involved. Our results showed that these two processes are segregated mechanistically, as there is no correlation between the preference for a specific background with the level of skin pigmentation, and different dorsal retina-localized type II opsins appear to underlie the two crypsis modes. Indeed, inhibition of melanopsin affects background adaptation but not background preference. Instead, we propose pinopsin is the photosensor involved in background preference. pinopsin mRNA is co-expressed with mRNA for the sws1 cone photopigment in dorsally-located photoreceptors. Importantly, the developmental onset of pinopsin expression aligns with the emergence of the preference for a white background, but after the background adaptation phenotype appears. Furthermore, white background preference of tadpoles is associated with increased pinopsin expression, a feature that is lost in pre-metamorphic froglets along with a preference for a white background. Thus, our data show a mechanistic dissociation between background adaptation and background preference, and we suggest melanopsin and pinopsin, respectively, initiate the two responses.

Physiological skin pigmentation indices were measured as described previously (Bertolesi et al 2015). Briefly, pictures of the dorsal head of tadpoles were taken with identical conditions of light, time exposure and diaphragm aperture, and converted to binary white/black images using NIH ImageJ (U. S. National Institutes of Health, Bethesda, MD) public domain software.

Behaviour assay of embryos was performed as described previously (Viczian & Zuber, 2014). Briefly, a minimum of nine embryos were tested simultaneously, each in individual boxes (6.5 x 9.0 x 2.5 cm) set over a platform with light from above (1200 lux) in a quiet room. Embryos were recorded using a Logitech C920S camera. Each individual box was divided into equal half area white and black surfaces, and each embryo was recorded for 7-10 cycles of 2 min each (total 840-1200 sec).