Bear denning runs on two clocks, and only one of them is allowed to drift

Photoperiod cannot drift. Day length on a given date at a given latitude is fixed while temperature on that date is not, and the researchers name that asymmetry directly: it is the reason climate change can pull the two cues apart rather than simply slide the whole calendar later ^[9].

That decoupling does not act evenly across the season, and the paper's own results say why. At the onset of hibernation, temperature alone influenced behaviour; in the other stages, temperature and day length worked together ^[6]. Read as a forecast rather than a finding, that means the entry into hibernation is the cheapest place for warming to buy extra active weeks, because there is no light cue braking it. Emergence, where the two signals interact, is partly buffered. Warm autumns should therefore show up in the conflict data before warm late winters do.

The evidence base is deep and narrow at the same time. More than 22,000 hours of continuous video, four bears, 45 coded behaviours ^[4] works out to roughly 5,500 hours per animal, about 229 days of unbroken watching each ^[5]. Very few hibernation datasets have that much within-animal detail. What they do not have is spread. All four animals were pregnant females, and the footage came from Virginia Tech's Black Bear Research Center rather than from dens scattered across a management unit ^[4]. The study can establish that the two cues interact, which is the part that overturns the temperature-first assumption ^[7]. It cannot tell a state biologist how many bears in a given county cross a given threshold in a given warm December.

That gap is the operational problem. A manager who wants to know when to expect nuisance calls has, in practice, been reasoning from temperature, which is the assumption the paper says was never sufficient ^[7]. Substituting a two-variable rule is harder than it sounds: photoperiod is perfectly predictable, which means all the uncertainty in the joint cue sits in the weather, and the interaction term decides how much of that uncertainty reaches the bear's behaviour. Nobody has that coefficient at population scale yet.

The consequence the authors flag is not that bears sleep less. It is that activity and food supply stop lining up. Holcombe wrote in the study that as temperatures rise, bears may become more active during periods when they would traditionally remain dormant ^[2], and the paper's stated concern is that this produces a mismatch with seasonal food availability, pushing bears toward human-associated resources ^[3]. A bear awake in February is not awake into abundance. It is awake into a landscape where the reliable calories are in bins, feeders and livestock, which converts a phenology result into a municipal budget line.

The paper is pitched as giving managers better tools to anticipate behavioural shifts ^[10]. On this evidence, what it gives them first is a reason to stop trusting the one variable they already had.