Figure 1. Free running clocks and damped ‘hourglass’ clocks are equally good time-keepers in noiseless conditions but internal and external fluctuations reveal significant differences.

(a) Free running circadian clocks, such as the KaiABC protein clock in S. elongatus, show rhythms in both oscillating and constant light (top) or dark (bottom) conditions. (b) In contrast, damped circadian clocks, such as that in P. marinus which lacks Kai A, show rhythms only in changing light conditions and decay to a fixed state in constant conditions. (c) When subject to external noise (i.e., weather-related amplitude fluctuations in light), simulations of the free running clock show low population variance while the damped clock shows high variance. In contrast, Gillespie simulations with high internal noise due to low copy number of Kai molecule reveals that damped clocks are much more robust than free running clocks. (d) A systematic study of clock precision (i.e., mutual information between clock state and time) at fixed external noise but decreasing Kai protein copy number $N$ reveals that free running clocks are preferred at low internal noise but damped clocks are preferable at sufficiently high internal noise.