Mice sense Moon and Sun

We treat the laboratory mouse as biology’s most controlled subject: same light-dark cycle, same food, same climate-controlled room for its entire life. Lunar and tidal entrainment of behaviour is well accepted in marine and amphibious species, but the idea that it might extend to a landlocked mammal in a windowless facility has mostly been met with polite scepticism.

Well, your mice may be quietly checking the lunar calendar!

Barde and colleagues take that scepticism on directly. They pulled together activity data from nine different groups of mice, spanning facilities across two continents, and asked one question: Does mouse activity still track the Moon, even in mice whose internal body clock has been genetically switched off? The answer was yes.

Circadian biology has given us a remarkably detailed picture of the 24-hour clock. Rhythms slower than a day, infradian rhythms, have had far less attention, and this preprint goes on to propose a mechanism for how a lunar signal might actually be getting in.

Nine datasets, one recurring signal

Rather than one new experiment, the authors leaned on scale. Six datasets came from their own “ColonyRack” system, an RFID-tracked home-cage network in which mice roam freely for months; the authors track Roaming Entropy  (a measure of how evenly an animal explores its environment) continuously. Two other datasets came from separate Dresden facilities that measured how much mice ran on exercise wheels. A ninth group came from a previously published Dallas study, using mice on a normal, unrestricted diet. These nine, together, are the backbone of the paper’s central claim.

Along with that, the authors reanalysed data from two further groups using mice with their circadian clock genetically disabled, and mice missing two proteins that are suspected of helping some animals detect magnetic fields. Multiple continents, multiple housing systems, multiple genotypes, unbiased experimental objectives and nearly two decades of collection. This is not a pattern easy to wave away as one lab’s fluke.

Activity rises at full moon and new moon, and it’s not a coincidence

Beneath the expected 24-hour and 12-hour rhythms, the authors broke the activity data down into its underlying repeating cycles, and in every dataset, one rhythm stood out as unusually strong. A recurring cycle of roughly 15 and 30 days. Lined up against the actual lunar calendar, mice were most active within about a day of a full or new moon (Figure 1from the preprint)

 

To rule out chance, the authors used a circular-shift test: they systematically shifted the mouse activity data thousands of times relative to the real lunar calendar, and asked how often a lunar-looking pattern turns up by luck alone. The real, unshifted data beat almost every one of those shifted versions. The shape of it also made biological sense: full and new moon are the two points in the lunar cycle where light and gravity are most extreme, so a genuine lunar effect should produce two activity peaks a month, not one. That’s exactly what showed up.

And it kept showing up. In mice in Germany and mice in the US, in males and females, across genetic strains, and on feeding schedules that had nothing to do with the Moon. That’s the detail that makes this hard to dismiss as a quirk of one cage room.

It doesn’t need a working body clock

The strongest test in the paper, to my eye, uses mice with their circadian clock genetically turned off. Take away light cues from a normal mouse’s disrupted clock and its daily rhythm collapses entirely. But the slower, Moon-linked rhythm held up anyway, meaning it isn’t riding on the same neural machinery that runs the 24-hour clock. It’s something else.

If it’s not gravity, what is it?

The next obvious question: Is the Moon acting directly, through its gravitational pull? The authors calculated this and the answer was No. The Moon’s actual gravitational pull on a single cell is roughly a billion times too weak to plausibly trigger any known biological sensor.

Instead, they point to something less obvious: Earth’s magnetic field. This field doesn’t just sit still; it fluctuates over time, in cycles of about 14 and 27-28 days, driven by a combination of the Sun’s rotation and the Moon’s orbit. Cross-referencing real geomagnetic observatory data against the behavioural recordings, the authors find that the two rhythms track each other closely, with mouse activity trailing magnetic field changes by roughly 0-2 hours, the kind of lag you’d expect from an actual sensory response rather than coincidence.

They then looked at mice missing two proteins (CRY1 and CRY2) that are suspected, in some other species, of helping detect magnetic fields. If these proteins were essential, removing them should have broken the mouse’s ability to track the magnetic field. But it didn’t, at least not completely, as the alignment with the magnetic field persisted, though these mice were also much less active overall, which softens how much weight that particular result carries.

Could this apply to humans, too?

This is where the preprint gets bold. Lunar synchronisation is increasingly well understood in marine species, but its relevance in terrestrial mammals, humans included, has stayed genuinely contested. The authors point to a scattering of existing human findings that have never quite added up to consensus: menstrual cycles occasionally, transiently syncing with the lunar cycle; mood cycling in some patients with rapid-cycling bipolar disorder aligning with specific lunar sub-cycles, and sleep-timing studies showing sleep starting later and running shorter on nights leading into a full moon.

None of that has ever settled into a clean, reproducible story. What this study offers isn’t proof that it applies to humans but perhaps implies a candidate mechanism. If terrestrial mammals really can sense geomagnetic fluctuations, that could be the missing thread tying together decades of inconsistent human lunar findings, rather than each of those studies capturing a different fluke.

Why this is worth your attention

What makes this preprint convincing is it’s the sheer consistency. Across nine separate mouse cohorts, different labs, different countries, different years, the same lunar-timed pattern was observed. Testing whether it survives in mice with no working body clock, rather than reporting the correlation and stopping, is genuinely good hypothesis-driven work.

What’s still missing is a direct cause-and-effect test. The geomagnetic link is, for now, a correlation. A convincing one, but a correlation, nonetheless.

The clean next step: shield or artificially alter the magnetic field around an otherwise identical group of mice and see if the rhythm shifts or disappears. Exactly how mice would sense it in the first place also remains an open question. The CRY1/CRY2 result hints they’re not strictly required, but doesn’t rule out a magnetite-based route, which has been proposed in other mammals like mole-rats and bats.

This preprint reopens a question mostly filed away as folklore and does it with enough scale and care to make “just a coincidence” a hard sell, even if exactly how mice are picking up the signal is still, delightfully, unresolved.

 

Questions for the authors

• Earth’s magnetic field varies in strength by location. Would you expect labs at very different latitudes to show a stronger or weaker version of this rhythm?
• Is there a practical way to shield or deliberately alter the magnetic field around a mouse cohort to move this from correlation to a direct causal test?
• The circadian clock-disabled mice still retain some peripheral clock activity outside the brain’s main pacemaker. Is there a way to rule out that residual activity as the source of the pattern?

 

Tags: #neuroscience #labtolunar #lunarcyclesync

Read preprint 1 Star (No Ratings Yet)