Evolutionary Consequences of Selection on Circadian Accuracy

In nature, circadian clocks align behavior and physiology of organisms with predictable environmental cycles, making behavioral timing a frequent target of natural selection (Pittendrigh, 1993). In contrast to approaches isolating individual traits to study their biology, the study of evolution in selection experiments within and outside of the lab reveals interrelationships between traits, which is often under-appreciated. Since circadian traits span across multiple levels of organization from molecular oscillations to neural circuits to overt physiology and behavior, they evolve as integrated phenotypes under sustained selection pressures, making them well-suited to probing trait correlations using laboratory selection approaches (Dani et al., 2023). Maintaining a stable phase-relationship is linked to organismal fitness (Pittendrigh, 1993; Jabbur et al., 2024), and Balaji et al. (2025), following prior studies, asked if and how greater stability of phasing co-evolves with other traits such as light sensitivity and circadian robustness.

Eclosion is a circadian-gated developmental event for flies, measurable as a population-level rhythm. The selection regime permitted flies emerging in a narrow window of time, that is, 1 h after lights on to form the next generation, thus reducing variance around the targeted phase (Kannan et al., 2012). Prior studies reported that selection for accuracy (reduced phase-angle variation) of eclosion timing led to shorter free-running period, increased precision (reduced variation in free-running period), and narrower eclosion window independent of developmental variation (Kannan et al., 2012; Varma et al., 2019; also see Dani et al., 2023). Balaji et al. examined activity rhythms and sleep of these populations after 335 generations of selection under different lighting conditions, along with differences in circadian gene expression and larval light avoidance.

The direct response to selection remains robust: flies from precision populations (PP) exhibited higher emergence within the selected ZT01-ZT02 window compared to control populations (CP), accompanied by a reduced fraction of flies emerging before the window, a change in emergence waveform, as well as a much narrower emergence gate. The correlated changes in activity-rest rhythms appear to be sex-specific: PP males were more active in mornings and evenings but less active late at night compared to controls under light-dark (LD) cycles and had a shorter free-running period under constant darkness, consistent with earlier reports (Kannan et al., 2012). Female flies, in contrast, did not show differences in activity patterns or most sleep parameters. The authors checked activity and sleep under low- and high-light-intensity LD cycles and found low-intensity LD to enhance sleep differences with PP males having significantly increased sleep and longer sleep episodes than controls. These are subtle differences, but since they are seen across replicate large, outbred populations they may reflect consistent population-level effects. While variation in developmental timing between sexes is known, recent evidence also suggests that such sex-specific differences could arise from differences in how clock signals are processed by sexually differentiated neural circuits such as subsets of circadian neurons expressing fruitless and Clock (Deluca et al., 2025). Thus, whether the same mechanism underlies sex-dependent circadian differences in the selected populations should be explored in future studies.

The effects of this selection extend into pre-adult stages, where PP larvae exhibit preference to light compared with controls. This is informative because larval light avoidance is distinct from adult locomotor and sleep parameters, with only a subset of clock neurons existing at that stage. While CRY-positive lateral neurons are present in larvae, light avoidance has been shown to be mediated by CRY-negative DN2 and 5th LNv neurons (Keene et al., 2011). PP, having phase-advanced cry, per, and tim oscillations, may have evolved altered integration of photic input resulting in larval light avoidance.

The authors acknowledge that continued selection beyond ~70 generations did not change eclosion accuracy or activity timing proportionally, suggesting a selection plateau. A point to note is even though sleep differences have been recently characterized (~335 generations), they may have existed earlier as divergence in activity-rest rhythms was reported after ~70 generations of selection. A limitation for the current study remains that one cannot unambiguously attribute the correlated changes in adult sleep to evolution of the circadian oscillator in the PP. Since the phenotype of accuracy mechanistically depends on input and output pathways along with the oscillator, sleep differences in PP may be a result of correlated changes in those due to selection on accuracy of eclosion timing.

Evolutionary change in multiple circadian outputs should also not be interpreted as an indicator of change in loci associated with input pathways or core oscillator exclusively. Convergent changes across multiple outputs could arise from shared downstream control points such as neural, hormonal, or motor pathways. Interestingly, recent theoretical studies have provided conceptual support for the possibility of circadian output systems being independent targets for evolving precision rather than the oscillator itself (Kaji et al., 2023). Viewed in the context of discrete versus continuous circadian gating mechanisms (Paajanen et al., 2025), whether selection on accuracy or precision of activity/sleep can result in similar changes in clock properties of eclosion remains an open avenue for future research. In the same way, differences in downstream clock-associated output circuits (Deluca et al., 2025) may also have caused sex-specific circadian responses observed in the selected populations. As correlated evolution of sleep parameters due to selection on clock properties is understudied, future studies should investigate the causal mechanisms and organizational basis of how evolutionary change in circadian phasing and light input can result in changes in the waveform and architecture of sleep.