Investigation of the effects of sensory salience on the refuge tracking performance of weakly electric fishes

Weakly electric fish species typically demonstrate hiding behaviors among plant roots and suspended objects in streams within their natural habitats. These fish employ the force generated by their anal ribbon-like fins to facilitate back-and-forth movements, thus modulating visual and electrosensory feedback. These movement abilities allow them to effectively maintain their position in a computer-controlled moving refuge. For our research, we conducted behavioral experiments involving wave-type gymnotiform species, namely, Apteronotus albifrons called black ghost knifefish and Eigenmannia virescens called glass knifefish, native to South America. We utilized an experimental setup consisting of a 3D-printed refuge, the motion of which was controlled using a linear actuator. The 3D refuges, made of PLA material, were prepared in three different lengths (7 cm, 14 cm, and 21 cm) and two surface feature variations (with and without windows). The refuge was moved by the sum of sinuses in the range of 0.10 Hz to 2.05 Hz with the help of a special code on the computer. The experimental conditions encompassed manipulations of four variables: 1) lighting levels (dark, dim, and light), 2) refuge structure (with or without windows), 3) refuge length (7 cm, 14 cm, 21 cm), and 4) conductivity levels (low, medium,and high). A total of 54 different sensory conditions were tested, (with N=5 individual fish) for each species. There was a statistically significant difference between species; with black ghost knifefish having lower RMSE values than glass knifefish species. The most important parameters for refuge tracking performance were illumination and length. Moreover, the conductivity had a significant impact on tracking performance. However, we determined that the window condition did not have the same effect on refuge tracking performance. Besides, we found statistically significant interactions between the 'Species and Length', 'Species and Window', 'Illumination and Length', 'Illumination and Window' conditions. The length of the refuge had a greater effect on the refuge tracking performance of the fish than whether the refuges are windowed or not. The presence of windows opened in refuges in well illuminated conditions had a minor effect on tracking performance, while the presence of windows in refuges in dark conditions had a positive effect on tracking performance, making a big difference in results. The results showed that refuge tracking behavior was much more successful at low conductivity. Fish weighted visual information when electrosensory salience was compromised at high conductivities.