Hibernation

Why does an animal facing a freezing winter survive it by letting its own body go cold?

It reads as a surrender. The threat is cold; the response is to become cold. A ground squirrel spends its active life defending a body temperature near 37 °C against every gust of wind, then, facing the worst of it, gives that up and lets its core fall to a few degrees above freezing. If warmth is worth defending in October, why is it abandoned in December?

Ask first what actually kills a small mammal in winter. Not the cold directly — its fur and its metabolism handle far worse in a November storm. What is missing is food. Seeds are under snow, insects are gone, green shoots will not return for months, so the animal must run its body for six or seven months on fuel it has already banked.

Now the size of that bill. A small body has a great deal of surface for its volume, so it sheds heat fast and must burn fuel continuously to replace it. Affordable when you can eat; arithmetically hopeless when you cannot, because no ground squirrel can carry six months of full metabolism as fat. The question stops being "how do I stay warm" and becomes "how do I need less."

Here is the lever. Chemistry runs slower when it is cooler — across the ordinary biological range, reaction rates fall roughly two- to three-fold for every 10 °C drop. Fall from 37 °C to 5 °C and the cost of being alive collapses: a deep hibernator's metabolic rate drops to a few percent of its resting rate, its heart from hundreds of beats a minute to a handful. Now the fat lasts.

So far this suggests the animal lets the cold in and the slowdown follows. That is the intuitive story, and the measurements say it is backwards. Metabolic rate falls first, and body temperature follows it down; animals can enter torpor at warm ambient temperatures, where nothing is forcing them to cool. The suppression is actively initiated and centrally controlled, not a passive consequence of chilling. Cold is not the cause of the shutdown — it is the result. Does that change what you think is being defended? The animal has not lost temperature regulation. It has moved the setpoint. Push a hibernator's core toward freezing and it will burn fuel to warm itself back up; the arctic ground squirrel, which supercools its core below 0 °C, is the extreme edge of a regulated range, not a failure of regulation.

Then the finding that reorganises everything. A hibernator does not stay down. Every few days to a couple of weeks it rewarms all the way to normal body temperature, holds it for a few hours, and then descends again — and those brief arousals consume the large majority of the entire winter's energy budget, despite occupying a small slice of the time. Rewarming thirty degrees by shivering and by burning brown fat is enormously expensive, and the animal chooses to do it dozens of times.

Why? This is genuinely unsettled. The oddest line of evidence is the strongest: the EEG during those arousals looks like the recovery sleep of a sleep-deprived animal, which suggests deep torpor, whatever else it is, does not do sleep's job. Other candidates — clearing metabolic waste, restoring depleted substrates, letting an immune system function, repairing synaptic connections that torpor retracts — have support and are not exclusive. We know the arousals are necessary, because animals prevented from having them do badly; we do not agree what for.

That reframes the trade. Hibernation is not free rest but a state so metabolically compromised that the animal must periodically pay a fortune to come out of it and do maintenance — on top of the immune competence it costs, the inability to flee, and something like reperfusion stress on every rewarming.

Where this sits next to its neighbours: the collection's piece on countercurrent heat exchange is about staying warm cheaply while remaining fully awake — the strategy hibernation gives up on. And the piece on sleep asks why an animal accepts hours of vulnerability each day; the arousal puzzle here suggests torpor is not an extension of that bargain but a different state that sleep must still be paid for on top of.

Three refinements hold the reasoning together. First, hibernation is not long sleep but a distinct physiological state, which is exactly why sleep appears to have to be repaid during arousals. Second, "letting the body go cold" describes the outcome, not the method — suppression comes first and temperature is its consequence, so an animal in deep torpor is regulating at a new setpoint, not surrendering. Third, the low temperature is not the saving so much as the visible sign of it: the slowed chemistry is what preserves the fat.

And a limit on how far to generalise. Bears are the famous case and the awkward one — a denning bear's temperature drops only a handful of degrees, yet its metabolism falls far more than that drop alone would explain, which is another reason to treat suppression and cooling as separable things rather than one thing.

• Why hibernators do not lose bone and muscle over months of immobility, when bed-rested humans do.
• How the timing is set when an animal underground has no cues about the season above it.

Every term the collection defines is gathered in the glossary.