For compass orientation many insects rely on the pattern of sky polarization, but some species also exploit the sky chromatic contrast. Desert locusts, Schistocerca gregaria, detect polarized light through a specialized dorsal rim area (DRA) in their compound eye. To better understand retinal mechanisms underlying visual navigation, we compared opsin expression, spectral and polarization sensitivities and response-stimulus intensity functions in the DRA and main retina of the locust. In addition to previously characterized opsins of long-wavelength-absorbing (Lo1) and blue-absorbing visual pigments (Lo2), we identified an opsin of an ultraviolet-absorbing visual pigment (LoUV). DRA photoreceptors exclusively expressed Lo2, had peak spectral sensitivities at 441 nm and showed high polarization sensitivity (PS 1.3-31.7). In contrast, ommatidia in the main eye co-expressed Lo1 and Lo2 in five photoreceptors, expressed Lo1 in two proximal photoreceptors, and Lo2 or LoUV in one distal photoreceptor. Correspondingly, we found broadband blueand green-peaking spectral sensitivities in the main eye and one narrowly tuned UV peaking receptor. Polarization sensitivity in the main retina was low (PS 1.3-3.8). V-log I functions in the DRA were steeper than in the main retina, supporting a role in polarization vision. Desert locusts occur as two morphs, a day-active gregarious and a night-active solitarious form. In solitarious locusts, sensitivities in the main retina were generally shifted to longer wavelengths, particularly in ventral eye regions, supporting a nocturnal lifestyle at low light levels. The data support the role of the DRA in polarization vision and suggest trichromatic colour vision in the desert locust.

A very well-documented case of flower-beetle interaction is the association in the Mediterranean region between red bowl-shaped flowers and beetles of the family Glaphyridae. The present study examines the visual mechanisms by which Pygopleurus israelitus (Glaphyridae: Scarabaeoidea: Coleoptera) would perceive the colors of flowers they visit by characterizing the spectral sensitivity of its photoreceptors. Our measurements revealed the presence of three types of photoreceptors, maximally sensitive in the UV, green and red areas of the spectrum. Using color vision space diagrams, we calculated the distribution of beetle-visited flower colors in the glaphyrid and honeybee color space and evaluated whether chromatic discrimination differs between the two types of pollinators. Respective color loci in the beetle color space are located on one side of the locus for green foliage background, whereas in the honeybee the flower color loci surround the locus occupied by green foliage. Our results represent the first evidence of a red sensitive photoreceptor in a flower-visiting coleopteran species, highlighting Glaphyridae as an interesting model group to study the role of pollinators in flower color evolution.

In many insect species, photoreceptors of a small dorsal rim area of the eye are specialized for sensitivity to the oscillation plane of polarized skylight and, thus, serve a role in sky compass orientation. To further understand peripheral mechanisms of polarized-light processing in the optic lobe, we have studied the projections of photoreceptors and their receptive fields in the main eye and dorsal rim area of the desert locust, a model system for polarization vision analysis. In both eye regions, one photoreceptor per ommatidium, R7, has a long visual fiber projecting through the lamina to the medulla. Axonal fibers from R7 receptors of the dorsal rim area have short side branches throughout the depth of the dorsal lamina and maintain retinotopic projections to the dorsal medulla following the first optic chiasma. Receptive fields of dorsal rim photoreceptors are considerably larger (average acceptance angle 33°) than those of the main eye (average acceptance angle 2.04°) and, taken together, cover almost the entire sky. The data challenge previous reports of two long visual fibers per ommatidium in the main eye of the locust and provide data for future analysis of peripheral networks underlying polarization opponency in the locust brain.

Studies on the generation of rhythmic motor patterns have shown that peripheral sensory input may contribute substantially to the rhythm generating network. A prominent example is the wing beat frequency of desert locusts, which can be entrained to rhythmic mechanosensory input, but also to the frequency of periodic light flashes. To further analyze the entrainment by light, tethered flying locusts were presented with periodic light flashes, while the position of the forewing was filmed. We show that entrainment of wing beat occurs both in the UV and green range of light. Animals maintained a characteristic phase relationship to the light stimulus with the most elevated wing position occurring at the end of the dark phase. Speed and time course of entrainment varied greatly and ranged from the duration of a single wing beat cycle to several seconds. To identify the visual system mediating entrainment, synchronization to UV light was tested after cutting the optic stalks to the optic lobes/compound eyes or the ocellar nerves. The results show that light entrainment of the locust flight pattern is largely and perhaps exclusively mediated via the fast ocellar pathway and may have a role to stabilize flight with respect to the horizon.