Social synchronization of circadian rhythmicity in female mice depends on the number of cohabiting animals.

Communal animals often engage in group activities that require temporal synchrony among its members, including synchrony on the circadian timescale. The principles and conditions that foster such collective synchronization are not understood, but existing literature hints that the number of interacting individuals may be a critical factor. We tested this by recording individual circadian body temperature rhythms of female house mice housed singly, in twos (pairs), or in groups of five (quintets) in constant darkness; determining the daily phases of the circadian peak for each animal; and then calculating the cycle-to-cycle phase relationship between cohabiting animals over time. Significant temporal coherence was observed in quintets: the proportion of quintets (4/7), but not pairs (2/8), that became synchronized was greater than could be achieved by the complete simulated reassortment of all individuals. We speculate that the social coupling of individual circadian clocks of group members may be adaptive under certain conditions, and we propose that optimal group sizes in nature may depend not only on species-specific energetics, spatial behaviour and natural history but also on the mathematics of synchronizing assemblies of weakly coupled animal oscillators.