A novel methodology for assessing reinforcement choice in primates under human care

Mandrillus sphinx shows a very strong preference for sweet potato over banana, grape, and apple in primate choice experiments under human care.

• Why it matters: Reinforcement training in zoological settings often assumes a single “best” reward, but species-specific preferences can be missed, reducing effectiveness and welfare alignment.
• What they did: Three non-human primate species (Mandrillus sphinx, Pan troglodytes, and Hylobates lar) chose among four foods (banana, grape, apple, sweet potato), and latent variable models estimated which foods best fit reinforcement choice.
• The result: The species-dependent preference patterns indicate that natural foraging ecology may be linked to fruit selection and can guide reinforcement and training decisions for each species.

AI summary of the abstract · may contain errors · not medical advice · read the paper.