Skip to main content
Dryad logo

Female preference for colour-enhanced males: a test of the sensory bias model in medaka, a drab fish


Downer-Bartholomew, Benjamin; Rodd, F Helen (2021), Female preference for colour-enhanced males: a test of the sensory bias model in medaka, a drab fish, Dryad, Dataset,


Sexual selection research has long focused on the evolution of female mate preferences. Most of the models that have been developed posit that mate preferences evolve in a mating context. In contrast, the sensory bias model proposes that mate choice preferences arise in a non-mating context, as a by-product of natural selection acting on a female’s perceptual system. Recent research has shown that many species of fishes, from across a large clade including poeciliids, goodeids, and medaka, have a bias for long wavelength (LW) colors (yellow, orange, red) in a non-mating context. Even species that do not have LW-colored ornaments, apparently because they have been lost secondarily, retain this latent bias for LW colors. Here, we predicted that female Oryzias latipes (Japanese medaka), a drab species with a latent preference for LW colors, would show a mate choice preference for males with an artificial secondary sexual trait—a colored stripe added to their flank. We confirmed that females were more responsive to red and orange objects in a non-mating context than to other colors. We also showed that females were less resistant towards males with a LW-colored stripe than to those enhanced with a non-LW stripe and that, for many females, responses towards specific LW colors were consistent across these non-mating and mating contexts. Therefore, our results provide support for the sensory bias model by providing a link between a sensory bias in a non-mating context and a mate choice preference in a drab species like medaka.


Data set we collected using JWatcher while watching behavioural observations. Data processing was done in JMP©14.0 (SAS Institute, Inc. 2018) for the divided tank data, and in RStudio (ver.1.4.1103) for the free swim data and the comparison across two behavioral contexts; the disc test data were analysed using both software programs.

Usage Notes

ReadMe file has been uploaded