Polarized-light sensitivity in rainbow trout (Oncorhynchus mykiss): characterization from multi-unit responses in the optic nerve

Integrated spike activity of axons from the optic nerve was measured in an investigation of the e-vector sensitive mechanism underlying the ability of rainbow trout (Oncorhynchus mykiss) for orientation in downwelling, linearly-polarized light. In anaesthetized, immobilized fish, one eye was exposed to incremental light flashes which were superimposed over closely controlled background conditions. Under scotopic and various photopic conditions, intensity/response curves were generated from the on-response of the optic nerve. Relative sensitivity curves were then obtained as a function of e-vector direction for the 5 kinds of receptor cells in this trout's retina: rods, ultraviolet cones (UV), short wavelength cones (S), medium wavelength cones (M), and long wavelength cones (L).Under scotopic conditions, no sensitivity to e-vector was apparent: thus, rods do not mediate polarization sensitivity. Under photopic conditions, parr weighing 8–10 g exhibited e-vector sensitivity in two orthogonal channels. A UV stimulus (380 nm) on a white background evoked a three-peaked response (0°, 90°, and 180°) to the e-vector orientations presented in 30° increments between 0° and 180°. In contrast, when the background was illuminated with appropriate short and long wavelength cut-off filters, M-and L-cones showed maximum responses only to the horizontal (90°) plane whether they were stimulated at their -absorption band or their -absorption band in the near UV. Isolated UV-cones gave maximum responses to the vertical (0° and 180°) e-vector, thus corresponding to a second channel. The blue sensitive, S-cones, did not show evidence of polarization sensitivity. As well, no evidence of the polarization sensitivity was observed under UV isolating background conditions in larger individuals, 50–78 g smolts, although the other cone mechanisms responded as in smaller individuals.