Memory for time of training modulates performance on a place conditioning task in marmosets

In rodents, the expression of a reward-conditioned place preference (CPP) is regulated in a circadian pattern such that the preference is exhibited strongly at the circadian time of prior training but not at other circadian times. Because each animal is trained only at a single circadian phase, the concept of time as a context cue is derived from a rhythmic internal state rather than learned explicitly from the external cues. We now report that the same “time memory” is expressed following context conditioning in the common marmoset (Callithrix jacchus). Animals were trained at a specific time to discriminate between an unpaired context and a context paired with food reward. Marmosets were then tested for preference at circadian times that were either the same or different from the training time. Preference was expressed only when training and testing times matched. The results show that time of day learning can be generalized to this new world primate implying that a similar circadian mechanism might regulate craving for reward in diverse mammals including human beings.