I don't know the validity of this because those who published the research had a vested interest (Color-C-Lector) in it, but I recall there being some very compelling data put out in the 1980s about bass being able to recognize colors. Supposedly, a device with 3 compartments was placed in a large aquarium with some bass. Each compartment had a different-colored door. Food was placed behind one of these doors and only one. When the bass bumped the door with their noses, the door opened and they got the food. After several days, the bass learned to immediately go to the specific door where the food was located. After several days, the food was moved either to the left or right but the door color remained the same. Supposedly, the bass followed that door, recognizing the color of the door more than its location. As I recall, no matter where that particular color of door was moved, the bass went directly to it.
Hopefully I'm recalling this accurately. Pretty sure I am.