Why one light cycle can produce two activity rhythms
The suprachiasmatic nucleus, a small cluster of cells in the brain's hypothalamus, is the body's master clock. It is not a single uniform timer. It has a light-sensitive region called the ventrolateral part and a light-insensitive region called the dorsomedial part, and these two regions can oscillate on their own schedules [4].
When rats are housed in a 22-hour light-dark cycle (11 hours light, 11 hours dark), their wheel-running and general locomotion split into two distinct bouts at the same time. One bout stays entrained to the light cycle, meaning it keeps pace with the lights going on and off. The other dissociates, drifting with a period longer than 24 hours [4]. Each bout is associated with independent clock gene oscillations in one of the two suprachiasmatic regions, so the split in behavior mirrors a split inside the clock itself [4].
The two regions are only weakly coupled under the 22-hour cycle, and the period of the drifting bout depends on how strongly they are coupled. Mathematical models treating the two regions as mutually coupled oscillators reproduced both the observed behavior and the clock's response to light pulses, supporting the idea that the whole system's period and light sensitivity emerge from that coupling [4].
What happens when the odd cycle ends
The split is reversible. When rats from the 22-hour cycle were released into constant darkness, their activity rhythms rapidly resynchronized into a single fused rhythm. That fused rhythm had a longer free-running period, the natural cycle length the clock adopts without light cues, than the rhythm of rats released from a standard 24-hour cycle [4].
The same difference appeared outside the body. Tissue from the suprachiasmatic nucleus of rats desynchronized by the 22-hour cycle kept a longer period in a dish than tissue from normally entrained rats, matching the behavioral result [4]. This suggests the altered light schedule leaves a lasting change in the clock's intrinsic timing, not just a temporary behavioral quirk.
Melatonin responds to light itself, not by splitting
Melatonin, the hormone the pineal gland releases mostly at night, behaves differently from activity in these studies. Brief exposure to dim red light at night strongly activated the suprachiasmatic clock, rapidly suppressed melatonin secretion, and produced a subsequent shift in when daily activity began [1]. That is a direct suppression response to light, not a second rhythm running alongside the first.
Longer-term light conditions also move melatonin levels as a whole. Fifteen days of constant darkness raised serum melatonin in rats by up to about 22 percent, while constant light did not lower it compared with a normal light-dark cycle [2]. The authors suggest that melatonin normally lost from the pineal gland under prolonged constant light may be compensated by melatonin produced in other organs, though this is proposed as an explanation rather than demonstrated [2]. Iron overload also reduced melatonin, consistent with melatonin being consumed while scavenging the free radicals that iron generates [2].
These melatonin findings come from constant light or constant darkness rather than from a 22-hour cycle, so they describe how melatonin tracks overall light conditions rather than how it behaves when activity splits in two. None of the studies here measured activity and melatonin rhythms together under the same non-24-hour cycle, so the claim that one splits while the other does not is an inference from separate experiments, not a directly observed contrast.
The red light in the room is not neutral
A practical wrinkle matters for anyone reading rat rhythms: the dim red light long used in rodent rooms because it was assumed the animals could not see it is not invisible to them. Brief dim red light across a range of wavelengths strongly activated the suprachiasmatic clock and suppressed melatonin, and in a behavioral task rats could detect long-wavelength red light, though not near-infrared light [1].
So a red-lit room is still a light cue that can shift the clock and flatten melatonin, which means the light conditions in an experiment are part of the experiment. This is worth keeping in mind when interpreting any rhythm measurement in rats, including the split activity bouts seen under a 22-hour cycle.
Sources used in this answer
Red Light Sensitivity of Non-image and Image Forming Visual Systems of Laboratory Rodents: Circadian Disruption and Behavioral Detection
Brief dim red light at night strongly activates the suprachiasmatic clock, rapidly suppresses melatonin, shifts daily activity onset, and is detectable by rats at long wavelengths but not near-infrared.
Effect of iron on rat serum melatonin levels under different light/dark cycle patterns
Fifteen days of constant darkness raised rat serum melatonin by up to about 22 percent, constant light did not lower it versus a normal cycle, and iron overload reduced melatonin.
Circadian rhythm sleep disorders in aging
Pineal melatonin and physical activity are discussed as agents relevant to non-24-hour sleep-wake cycles, and melatonin response was lower in old rats than young rats.
Coupling Between Subregional Oscillators Within the Suprachiasmatic Nucleus Determines Free-Running Period in the Rat
In a 22-hour light-dark cycle, rat locomotion splits into a light-entrained bout and a dissociated bout with a period longer than 24 hours, tied to independent clock gene oscillations in the ventrolateral and dorsomedial suprachiasmatic nucleus.