Non-numerical strategies used by bees to solve numerical cognition tasks

Area of Science:

• Comparative Cognition
• Animal Behaviour
• Neuroscience

Background:

• Studies on numerical cognition in animals often use visual discrimination tasks.
• Previous research suggested some animals possess rudimentary number sense.
• Potential for non-numerical cues to influence task performance is a concern.

Purpose of the Study:

• To investigate how bees solve visual discrimination tasks.
• To determine if bees utilize numerical or continuous (non-numerical) cues.
• To model bee behavior using a biologically plausible network.

Main Methods:

• Bees were trained on a visual discrimination task.
• Additional tests were conducted to identify learned cues.
• A neural network model simulated task performance using visual feature filtering.

Main Results:

• Bees performed well but learned continuous, non-numerical cues.
• The network model successfully learned the task using only continuous cues.
• The model replicated behaviors previously interpreted as numerical cognition.

Conclusions:

• Bees' performance may not indicate true numerical cognition.
• Continuous cues can be mistaken for numerical processing in animal studies.
• Magnitude sense might be more fundamental than number sense in evolution.
• Improved experimental designs are crucial for studying animal numerical cognition.