Figure 4.

Spectral location of pigments in Metriaclima zebra is tuned to allow large chromatic contrast between fish pattern and background. (A-H) Median ΔC between fish pattern and background. ΔC was modeled for opponent channels that compare the outputs of one single cone and one double cone (A-D), and channels that compare the outputs of two double cones (E-H). The various colors and tags in the contour plots represent different median ΔC values. The maxΔCmag pigment pairs comprised pigments with λmax that were far apart, and resembled the λmax of the real pigments in M. zebra (368, 484, and 523 nm). In contrast, the minΔCvar pigment pairs consisted of pigments with λmax that were closer together. This trend held for both opponent channel types and for both 1 m and 10 m depth. (I-L) Scatterplots illustrating the relationship between ΔC magnitude and ΔC variation for the real (black), minΔCvar (red), and maxΔCmag (green) pigment pairs. ΔC magnitude produced by the real pigment pair resembled the ΔC amplitude produced by the maxΔCmag pigment pair better than it resembled the ΔC amplitude produced by the minΔCvar. This was observed at both 1 m and 10 m depth, and for both opponent channel types (I,J and K,L). ΔC variation at 10 m depth was smaller than at 1 m depth. Regardless of water depth, however, the real and maxΔCmag pigment pairs produced variation in ΔC that was larger than in the minΔCvar pigment pairs. Sample size, n = 696 for either the real, minΔCvar, or maxΔCmag pigment pairs, for each of the opponent channels and water depths examined.

Sabbah and Hawryshyn BMC Biology 2013 11:77   doi:10.1186/1741-7007-11-77
Download authors' original image