UC Irvine

Fish use their sensory systems to detect and engage with the world around them. Fish use different sensory cues from their environment to direct their own behaviors. Understanding how these cues affect the behavior of fish guides our thinking about fish behaviors. I studied three sensory systems in fish (vision, the lateral line, and the vestibular system) in the context of two behaviors important to fish (predator evasion and schooling).

My first dissertation chapter explored the use of vision in predator evasion. I created an automated experimental setup that projected looming visual stimuli of different approach velocities on the wall of a tank that contained an experimental fish. By analyzing the escape response of the fish in reaction to these stimuli, I was able to discern the threshold-visual angle (and rate of change of this angle) to which the fish were likely responding. I then ran experiments with live predators and found strong evidence that these threshold stimuli were accurate. Lastly, I used a mathematical model to explore the functional significance of these threshold stimuli. This model predicted that these threshold stimuli were most useful to the fish when the predator is slower than the prey. This work demonstrates how fish may use the threshold-visual angle (and angle rate) to evade predators.

My second dissertation chapter explored how fish use their lateral line and vestibular systems to detect flows which may indicate an approaching predator. The lateral line contains two types of flow sensors: canal neuromasts (CNs), which detect pressure differences along the body that correlate with the acceleration of water flow, and superficial neuromasts (SNs), which detect the velocity of water flow with respect to the body. The vestibular system is a sensory system that detects the linear and rotational acceleration of the body. Using a neomycin sulfate bath, I was able to temporarily ablate the CNs and SNs on fish, leaving only the vestibular system functional. Using a more precise technique with neomycin sulfate, I was able to temporarily ablate the SNs, leaving the CNs largely intact. Thus, I created three groups of fish: those with their lateral lines intact (CNs+SNs), those with only their SNs ablated (CNs-only), and those with their lateral lines completely ablated (no-LL). I exposed each of these groups to a randomized set of flows and used the frequency of escape as a proxy for detection of the flow stimulus. I found that fish were able to escape from these flows with no lateral line, indicating the use of their vestibular system. Additionally, I found the CNs+SNs had an increased probability of detection over CNs-only and no-LL. This study demonstrates that fish are capable of using multiple sensory systems to detect flow stimuli.

My third chapter explored the use of the lateral line and vision in fish schooling behaviors. I recorded schools of five fish swimming after adjusting to different light levels. By measuring the distance between neighbors and the amount of polarization in the group, I was able to determine how well fish were able to school at each light level. I found that a minimum illuminance of $>$1.5 lux allowed fish to swim with sufficient polarization and distance between neighbors to be schooling. I then repeated the experiment with fish whose lateral line I had chemically ablated with neomycin sulfate. I found that these fish still schooled, but with a lower polarization and increased distance between neighbors. These results demonstrate that vision is required for schooling, and flow sensing modulates the quality of the schooling behavior.