One of my favorite things I wrote in 2019 was about a research study that taught rats how to drive. It was an activity that the rats seemed to enjoy. Today we have another story about lab animals learning how to drive. However, this time, the motorists were not mammals. They were goldfish who learned how to drive a fish-operated vehicle within a terrestrial environment.
At this point, the most common question people ask is “Why?” The driving-rat study was conducted in 2019. It was a study that sought to determine environmental stress. Driving is an activity that reduced stress levels in the rats. This studyShachar Givon and colleagues from Ben-Gurion University of the Negev, Israel, conducted the study and published it in Behavioral Brain Research. The goal was to find something a little more.
The idea was to test if fishes can navigate in unfamiliar environments and if they are able to do so universally. This concept is known as domain transfer methodology. You have to admit that driving a tank in a lab is not an ideal environment for a goldfish.
Nice fish tank-how fast can it go?
The FOV is basically a fishtank with wheels. The fish can learn to use the FOV without any physical controls, unlike the rat-mobile. Instead, a downward-looking cam tracks the fish’s position within the tank. If the fish is located near one of the tank walls, and facing outwards (using an onboard Raspberry Pi 3B+), the fish-control algorithm uses that to move the FOV in that direction. If the FOV crosses within 20cm from the walls of the terrestrial environment (a 4×3 meter enclosure), the fish control algorithm is overridden by a lidar sensor mounted on the same mast.
Six fish participated in the experiment. They were taught to drive in 30 minute sessions three times a week. Each session was conducted every two days. To start, the fish were rewarded if they navigated to the target (a pink corrugated wooden board) in the enclosure’s center.
The fish was able to learn how to drive the FOV and became more proficient over time. This was both in terms of completing tasks (navigating to a target) as well as the time it took to complete the task. To ensure that the fish were not just memorizing movements to earn rewards, the authors repeated tests with a different starting position or added decoy targets of a different color.
The goldfish were able not only to adapt to these changes but they also were able approach targets from different angles, suggesting an internal representation of the world. This is despite fact that the interface between water and air, as well as the tank’s plexiglass walls and the water, would create nonlinear refractive effects that would presumably make fish-vision look different to the environment in the which the fish eyes evolved.
Maybe goldfish aren’t quite as forgetful than we think.
Behavioral Brain Research. 2022 DOI: 10.1016/j.bbr.2021.113711(About DOIs).
