The Free-Running Circasemilunar Period Is Determined by Counting Circadian Clock Cycles in the Marine Midge Clunio Marinus

Abstract

Semilunar rhythms are found in numerous marine organisms, but the molecular mechanism and functional principles of endogenous circasemilunar clocks remain elusive. Here, we explore the connection between the free-running circasemilunar clock and the circadian clock in the marine midge Clunio marinus with three different chronobiological assays. First, we found that the free-running circasemilunar period of the adult emergence rhythm in C. marinus changes linearly with diel T-cycle length, supporting a day-counting mechanism. Second, under LD 6:6, periods of circasemilunar and circadian emergence were comparable to those under LD 12:12, indicating that the circasemilunar counter in C. marinus relies on endogenous circadian oscillations rather than external T-cycles. Finally, when desynchronizing the circadian clock with constant light, the free-running circasemilunar emergence rhythm disappeared as well, suggesting that it requires a synchronized circadian clock. These results oppose the long-held view that C. marinus' free-running circasemilunar clock operates independently of the circadian clock. In a broader evolutionary context, our results strengthen the idea that the circasemilunar clocks of dipterous insects are based on different functional principles compared to the circasemilunar or circalunar clocks of marine annelids and algae. These divergent clock principles may indicate multiple evolutionary origins of circasemilunar and circalunar clocks.

Keywords: T-cycle; beat hypothesis; circalunar clock; counter hypothesis; oscillator hypothesis.

Figure 1.

Observed circasemilunar periods of C. marinus’ emergence rhythm under different T-cycle lengths are compared with periods as predicted by three hypotheses (counter = yellow, oscillator = blue, and beat = red). (a) Period of the free-running circasemilunar emergence rhythm increased linearly with T-cycle length between 22 and 26 hours. Emergence is arrhythmic under T-cycles of 28 and 30 hours. n represents the number of replicates conducted. The number of significant values used for downstream analyses is depicted in brackets. The linear model fitted to the observed periods is depicted as the gray solid line (ci = 0.95). Overlapping periods observed under the T-cycle of 22 hours were jittered along the x-axis for better visualization. Replicates are labeled according to the year they were conducted. 2023b refers to the control of the constant light experiment. (b) The residuals (difference between observed and expected periods) are plotted for all three hypotheses. The AICc value was the lowest for the counter hypothesis, indicating that it is the best-fitting model. All residuals were jittered along the x-axis for visualization of overlapping data.

Figure 2.

Circadian emergence phenotype of C. marinus under LD 12:12 and LD 6:6 after entrainment in LD 16:8. The experiment was conducted in 2022. (a, b) Hourly number of emerged midges for 91 days in (a) LD 12:12 and (b) LD 6:6. Missing values (NA) are displayed in red. (c-d) Numbers of emerged midges were summed per hour over all days. (c) Emergence occurred right before the light-dark transition in LD 12:12. (d) Emergence occurred only at one of the two light-dark transitions in LD 6:6, indicating frequency demultiplication of the circadian clock. (e-g) Lomb-Scargle periodogram analysis reveals a 24-h period of emergence under both LD 12:12 and LD 6:6.

Figure 3.

(a, b) Daily number of emerged midges for C. marinus under LD 6:6 and circasemilunar free-run after circasemilunar entrainment in LD 12:12. Experiments were conducted in 2021 (a) and 2023 (b). For time series analysis, only the colored bars were considered. (c) The circasemilunar period remains close to 15 days (dashed line) under LD 6:6 for both replicates (a): ${\tau }_{circasemilunar}=14.06d,p=5.73e-04$ ; (b): ${\tau }_{circasemilunar}=13.08d,p=3.16e-02.$

Figure 4.

Circadian emergence of C. marinus in LD 12:12 and under constant light. (a) Under LD 12:12 midges emerge synchronized in phase around ZT 18, at the light-dark transition. (b) In constant light, midges emerge throughout the 24-h day. (c, d) Number of emerged midges per ZT was summed up over all 83 cycles to visualize the phase of emergence. (e-g) Lomb-Scargle periodogram analysis identifies a significant 24 h-period under LD 12:12 (e) but not under LL (f), suggesting that C. marinus’ circadian clock desynchronizes in constant light.

Figure 5.

The circasemilunar emergence pattern of C. marinus becomes arrhythmic under constant light. (a) Number of emerged midges per day under LD 12:12. (b) Number of emerged midges under LL. (c, d) Lomb-Scargle periodogram analysis reveals that free-running circasemilunar emergence was rhythmic under LD 12:12 but not in constant light. (e) meta2d() detects a significant circasemilunar period only under LD 12:12.