Modeling electrosensory and mechanosensory images during the predatory behavior of weakly electric fish

Abstract

Black ghost knifefish (Apteronotus albifrons) are nocturnal, weakly electric fish that feed on insect larvae and small crustaceans in the freshwater rivers of South America. In the absence of visual cues, prey detection and localization in this species is likely to rely on weak electrosensory and mechanosensory cues generated by the prey. In this paper, a modeling approach is used to estimate contributions to prey capture behavior from three octavolateralis modalities: the high- (tuberous) and low- (ampullary) frequency components of the electric sense and the high-frequency (canal neuromast) component of the lateral line mechanosensory system. For each of these modalities, the physical stimulus generated by the prey is approximated using a simple dipole model. Model parameters are constrained using previously published data as well as new empirical data on the electrical impedance characteristics of Daphnia magna. Models of electrosensory and mechanosensory stimuli are combined with actual prey strike trajectories from infrared video recordings to reconstruct spatial images of the prey along the sensory surface of the fish during the behavior. Modeling results suggest that all three modalities might contribute and that the relative contributions may change as a function of environmental conditions (e.g., water conductivity) and as a function of time over the course of the prey capture event.