Regional differences in photoreceptor performance in the eye of the praying mantis

1. Photoreceptor spacing, and angular, spectral, absolute and relative sensitivities have been measured across the compound eye of the praying mantisTenodera australasiae using optical and electrophysiological techniques. 2. Together, the two compound eyes cover virtually all spatial directions. The huge binocular fields extend vertically 240° with a maximum horizontal overlap of 35° in the frontal part of the eye (Fig. 2). 3. Interommatidial angles (Δφ) range from 0.6° in the frontal eye region up to 2.5° in edge regions of the eye (Fig. 5; Table 1). The eye region with minimumΔφ-values is defined as a fovea. 4. The acceptance angles (Δρ) of light-adapted photoreceptors are almost equal to the interommatidial angles over the whole eye (Δρ is 0.7° in the fovea and 2.5° in the edge of the eye) (Fig. 6; Table 1). The measured values ofΔρ are close to those predicted by the theories of Snyder (1977) and Horridge and Duelli (1979) from the optical and anatomical dimensions of the eye. In this context, the facet diameters are larger and the crystalline cones are longer in the fovea than elsewhere, whereas the rhabdom diameters are smaller. It is concluded that diffraction limitsΔρ in the fovea, whereas the acceptance function of the rhabdom limitsΔρ in eye regions outside the fovea (Fig. 16). 5. The angular sensitivity depends on the state of light adaptation and the time of day. In a defined foveal region the photoreceptors have mean acceptance angles of 0.74° (S.D. = 0.1°) when light-adapted, 1.1° (S.D. = 0.2°) when dark-adapted in daytime, and 2° (S.D. = 0.4°) when dark-adapted at night. The corresponding angles for a defined dorsal eye region are 2.4° (S.D. = 0.3°), 3.2° (S.D. =0.3°), and 6° (S.D.= 0.5°) (Fig. 10). 6. All units recorded from have similar spectral sensitivities, with a maximum in the wavelength range 500 nm to 520 nm, and a weak secondary peak around 370 nm (Fig. 12). 7. Dark-adapted photoreceptors produce bumps whose mean amplitude varies from cell to cell from 1 to 3 mV. Retinular cells in the defined foveal and dorsal eye regions have almost identical quantum capture sensitivities (defined as number of bumps per incident peak axial photons per cm2). Consequently foveal photoreceptors must have a lower quantum capture efficiency (defined as the number of bumps per incident peak axial photons perfacet), because their ommatidia have larger facet diameters. The quantum capture efficiencies are 0.04 (S.D. = 0.02) in the fovea and 0.10 (S.D. = 0.02) in the dorsal eye (Fig. 13; Table 2). This finding supports theoretical predictions that the acceptance function of the rhabdom of foveal photoreceptors is narrowed to decreaseΔρ. 8. Absolute and relative sensitivities of photoreceptors, defined as the reciprocal of the quantal flux required to generate a voltage response of 50% maximum, were determined in the defined foveal and dorsal eye regions. To a point source dark adapted retinula cells from both eye regions have almost identical sensitivities (Fig. 14b, c; Table 3). However, to a large diffuse source, dark-adapted foveal photoreceptors with their relatively small acceptance angles are less sensitive than receptors in the dorsal eye with their largeΔρ values, and retinula cells of both eye regions are more sensitive at night than during the day because their fields of view are broadened (Fig. 14d, e; Table 4). 9. The primary visual task of the praying mantis is the recognition and localization of prey moving against a disrupted background. Prey capture mainly occurs during the day, while the sexual behaviour takes place at night. Some relations between the eye performance and the visual behaviour are discussed.