Using photographs to study animal social cognition and behaviour: Do capuchins' responses to photos reflect reality?

Behavioural responses to photos are often used to infer what animals understand about their social environment, but are rarely validated against the same stimuli in real life. If subjects' responses to photos do not reflect responses to the same live stimuli, it is difficult to conclude what happens in reality based on photo responses alone. We compared capuchins' responses to photos versus live stimuli in an identical scenario within research cubicles. Subjects had the opportunity to approach food placed in front of an alpha group member and, in a separate condition, photos depicting the same individual. Subjects' latencies to approach food when placed in front of the real alpha negatively correlated with time subjects spent in close proximity to the alpha in their main enclosure. We therefore predicted subjects' latencies to approach food in the presence of photos would positively correlate with their latencies to approach food in the presence of the real alpha inside the cubicles, but negatively correlate with time they spent in proximity to the alpha in their enclosure. Neither prediction was supported. While not necessarily surprising, we explain why these results should be an important reminder that care is needed when interpreting results from photo studies.