Spatial vision in diverse invertebrates

Abstract

Light-detection provides incomparabaly rapid information at a range of time scales and, consequently, directional photoreception is found in almost every major metazoan clade. Conversely, a much smaller cohort have evolved sophisticated high resolution vision, which facillitates complex tasks via the detection of fast or small objects,. such as courtship and high-speed pursuit of prey. A few such species are the focus of most vision research. In contrast, low resolution image-forming vision is little investigated and, in some cases, whether vision is present has not been directly evidenced. To ameliorate this, I tested the image-forming capability (spatial resolution) of a variety of eye types from diverse animal groups: the camera eye of a velvet worm, Euperipatoides rowelli, (paper I), the dispersed visual system of a diadematid sea urchin, Diadema africanum, (paper II), which lacks discrete eyes, the cup eye of a planarian flatworm, Schmidtea lugubris, (paper III) and the compound eye of a millipede, Cylindroiulus punctatus, (paper IV). For each study animal, I used behavioural experiments in which the animals responded to dark visual cues to test whether they could response to visual stimuli of different sizes. The animals were placed in circular arenas under bright light and the direction they moved in relation to the stimulus was recorded. As a negative phototaxis response is present in each species, attraction towards the stimulus, if it could be resolved provided a measure of resolution. The angular sensitivity of an eye required to see a given visual signal was modelled (making assumptions about their contrast sensitivity; discussed in paper I).. I compared the efficacy of the various visual stimulus types for this purpose in paper II. To account for mutiple effects and low response rates to stimuli I apply logistic regression models using Bayesian inference in papers III and IV , to make more robust estimates which better express uncertainty. I also used imaging techniques, including x-ray tomography and transmission electron microscopy, to relate the visual performance of these animals to their visual systems. We found that these animals each had coarse vision with a spatial resolution of 20-30° to for the velvet worm E. rowelli, 29-69° for the sea urchin D. africanum (in respect of object taxis), 63° for the millipede C. punctatus and 73° for the flatworm S. lugubris. These modest proposed ranges of spatial resolution are nonetheless sufficient for the visual tasks employed by these animals.

Type