Enhanced predator detection by means of red fluorescence in a small fish species

点击翻译按钮获取中文摘要

For several years, we have studied the possibility that small, bottom-dwelling fish may be using their eyes to re-direct downwelling light to illuminate their immediate environment over distances of a few centimetres. We think this improves their ability to detect prey or predators and call this form of active sensing “diurnal active photolocation”. Previously, we were able to show that our model species, the triplefin Tripterygion delaisi, actively reflects downwelling sunlight sideways with its eyes. This effect was found to be sufficiently strong in shallow water (~ 10 m) to induce retroreflective eyeshine in the pupil of the scorpionfish Scorpaena porcus, a cryptic sit-and-wait predator. Behavioural experiments and modelling suggested that triplefins see this self-induced reflexion in the predator’s pupil over distances of 7-12 cm, and that this is a possible reason why they keep more distance. This distance is sufficient to escape an attack. Previous experiments also demonstrated that triplefins increase the frequency of bobbing behaviour (= repeated head-raising) when they see a scorpionfish relative to three other, harmless visual stimuli. This allows us to use bobbing as a measure of wariness. In this project, we assess whether – in addition light reflection in shallow water – red fluorescence can be used for local illumination at depth. The iris of T. delaisi fluoresces strongly. This is inconspicuous in shallow water, but stands out under dim, blue-green light at depths below 20 m. An experiment in 2022 with 122 triplefins showed more bobbing in the visual presence of a scorpionfish when seen through a neutral filter (all wavelengths pass), than through a blue-green filter that blocks reds. Their response to a stone as a control elicited much less bobbing and was unaffected by the filter treatment. In a follow-up experiment, and the subject of this application, we now want to specifically test whether it is the eyes of a scorpionfish that induce the effect, as implicitly assumed. Using a similar experimental design, we shall confront triplefins with 3D-printed scorpionfish models that are fitted with an artificial eye that is either retroreflective or not. We predict that triplefins show most bobbing to the model with the retroreflective pupil. The third visual stimulus will be a live scorpionfish, to confirm that the filter-treatment in the first experiment is reproduceable, as well as to assess how the response to a retroreflective eye compares to that to a live scorpionfish under the two filter treatments. With this approach, we hope to provide further evidence that the presence of a retroreflective eye is an important contributor to the differential response of triplefins to a visual target.