Skip to main content
. 2019 Oct 7;116(43):21843–21853. doi: 10.1073/pnas.1910807116

Fig. 3.

Fig. 3.

Polarotactic behavior depends on the spectral composition of polarized light. (A) Transmittance of the filters used to control the spectra of light reflected from the black spheres (solid and dashed lines), together with the spectral sensitivities of the central photoreceptors (shaded areas: R7H and R7V, dark violet; R8H, light blue; R8V, light green). Full-spectrum filter, UBG (transmitting UV, blue B, and green G); UV cutoff filter, BG (transmitting B and G); UV and blue cutoff, G (transmitting G); blue bandpass, B (transmitting B). (B) RGB photograph of the combinations of spheres and filters. (C) Relative numbers of horsefly visits to the shiny balls reflecting full-spectrum polarized light (UBG), green polarized light (G), blue- and green-polarized light (BG), and blue-polarized light (B), compared to nonpolarized control (matte UBG). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, not significant (ns) (P > 0.05) (Bonferroni-corrected Student’s t test). (D–F) Interpretation of experimental results (D) Attenuation of reflected UV and B reduces the signal in all central photoreceptors except R8V; hence, a yellow-reflecting ball is equally as unattractive as a matte ball, which reflects no polarized light. (E) Attenuation of reflected UV enhances the difference in excitation within the orthogonal analyzer receptor pair of R7H and R8H, leading to increased polarotaxis. (F) Attenuation of reflected UV and G simultaneously increases the differential signal in the polarization analyzer pair and reduces the excitation in R8V, thereby increasing the polarotaxis and decreasing the inhibition by long-wavelength light.