The precision of circadian clocks: assessment and analysis in Syrian hamsters.

Locomotor activity recordings of Syrian hamsters were systematically analyzed to estimate the precision of the overt circadian activity rhythm in constant darkness. Phase variation, i.e., the standard deviation of phase markers around the regression line, varied with the definition of phase. Smallest phase variation was found in the onset of wheel running activity defined by 1h running means of the raw data. Both lower and higher degrees of smoothing lead to decreased precision measured in the overt rhythm. With passive infrared recordings, the midpoint of activity defined by 3h running means was the least variable. This demonstrates that the choice of phase marker should vary between recording methods. Phase variation decreased with increasing activity and was larger in females than in males. By calculating the average cycle variation and serial covariance of consecutive cycles, we estimated the contribution of 'clock' and 'non-clock' related processes to the overt rhythm variability. Variance in precision between phase markers could be shown to be attributable mainly to nonclock processes. Variance in pacemaker cycle length appeared reduced in wheel running activity records compared with passive infrared sensing records, suggesting feedback from running activity onto pacemaker function.