Phenotypic flexibility may incur a selective advantage in changing and heterogeneous environments, and is increasingly recognized as an integral aspect of organismal adaptation. Despite the widespread occurrence and potential importance of rapid and reversible background-mediated color change for predator avoidance, knowledge gaps remain regarding its adaptive value, repeatability within individuals, phenotypic correlates, and whether its expression is context dependent. We used manipulative experiments to investigate these issues in two fish species, the three-spined (Gasterosteus aculeatus) and nine-spined stickleback (Pungitius pungitius). We sequentially exposed individuals to dark and light visual background treatments, quantified color change from video recordings, and examined associations of color change with phenotypic dimensions that can influence the outcome of predator-prey interactions. G. aculeatus expressed a greater degree of color change compared to P. pungitius. In G. aculeatus, the color change response was repeatable within individuals. Moreover, the color change response was independent of body size but affected by sex and boldness, with males and bolder individuals changing less. Infection by the parasite Schistocephalus solidus did not affect the degree of color change, but it did modulate its association with sex and boldness. G. aculeatus adjusted the expression of color change in response to predation risk, with enhanced color change expression in individuals exposed to either simulated attacks, or olfactory cues from a natural predator. These results provide novel evidence on repeatability, correlated traits, and context dependence in the color change response and highlight how a suite of factors can contribute to individual variation in phenotypic flexibility.

Data was generated by quantifying color change from 4K video recordings on stickelbacks subjected to different backgrounds (dark and light) in laboratory settings. Two experiments were conducted: the first investigated the association between the degree of color change and other phenotypic dimensions (infection of Schistocephalus solidus, boldness, body length and sex) as well as repeatability of color change be repeated exposure to the same background; the second (the manipulation experiment) focused on whether two proxys of predation risk (predator olfactory cues and simulated attacks) influenced the degree of color change in rlation to a control treatment.

The data includes some missing values (NA) regarding the asscoaiton between color cahgne and phenotypic dimensions whne these (for inastnce sex) could not be determined. Also there are no data for phenotypic dimensions on indivduals involved in the manipulation experiemtn or the Pungitius pungitius subjects.