FLYING SQUIRRELS IN A HUDDLE - Winter World: The Ingenuity of Animal Survival - Bernd Heinrich

Winter World: The Ingenuity of Animal Survival - Bernd Heinrich (2003)


One April I found young northern flying squirrels (Glaucomys sabrinus) with still-closed eyes, and I adopted one of the litter. Fed on Similac baby formula with an eyedropper, the tiny waif grew quickly. It often slept in my shirt pocket when I carried it to my office and occasionally to the campus dairy bar, where I enticed it out onto the countertop to lap up ice cream. Outgrowing my pocket, it later lived in a spare bedroom, where it slept all day in a hollow log. When I entered the room after dark, it ran up to a ceiling beam and jumped off to glide through the air and land on my chest with a light thump. When jumping from hundred-foot trees, northern flying squirrels can glide over three hundred feet, given suitable slope and wind.

Like my father’s pet weasel, my squirrel died in an unfortunate accident. I had put sprigs of geranium into a jar with water. One evening when the squirrel was free in the living room, it crawled down the cut geranium stems and drank from the water they were in. A little later the little animal was retching: it had been poisoned by the plant’s chemical defenses. The next morning my charming pet was stone dead. Animals that to us appear to have astounding toughness, such as surviving in the winter world, are also extraordinarily fragile, each in its own way.

Northern flying squirrels are common across North America from the Canadian maritimes all the way to Alaska, and they survive the harshest winters. Whatever it is that flying squirrels do to live through northern winters it does not involve the usual tricks of storing food, getting fat, or hibernating. Furthermore, my tame ice-cream-lapping flying squirrel notwithstanding, these animals are normally strictly nocturnal. One might predict instead that they “should” try to avoid night activity to avoid low temperatures by then resting in their snug nests, yet in the wild they sleep away the day even when temperatures are reasonable. They come out of their snug nests only when the sun goes down and temperatures dip sharply. I have no answer to why they do this, but a comparative perspective gives hints of selective pressures that have made flying squirrels nocturnal.

It is likely indicative that all of the about thirty species of the world’s flying squirrels are nocturnal, while none of the one hundred or so species of day-active squirrels are adapted for gliding flight. The fact that none of the day-active mammals are fliers or gliders cannot be attributed to dietary specializations. Does it concern predation? Gliding flight saves much energy for moving around, yet it makes the animal conspicuous to predators, with the additional drawback that the wing membranes compromise agility. (Bats, the best mammalian fliers, are among the poorest runners.) Maybe squirrels flew to save energy, then had to become nocturnal to escape predators, then had to fly even more, because the noise of scampering on the forest floor would be the owls’ hunting cue. Hence, the need for energy economy in these mammals would positively reinforce a nocturnal lifestyle that encourages gliding and flying.

Flying squirrels don’t leave being active at night up to mere chance; their circadian clock ensures that they get up and going only after sundown. That does not mean that they do not heed light cues from the environment. They do use light cues to synchronize their internal clock to keep to the daily twenty-four-hour rhythm so that they can get up and go out of their dark daylight hiding places soon after it gets dark outside. How then do we know whether a night-active animal becomes active after dark because it is the right time, or merely because it is dark then (and vice versa in a day-active animal)?

Flying squirrels were important in answering this fundamental question. They are one of the first mammals that were shown to be able to become active at the right time independent of external cues. The pioneering and now-classic experiments that revealed the fascinating world of chronobiology in the southern flying squirrels (Glaucomys volans), and subsequently in almost all other organisms examined, were conducted by Patricia J. DeCoursey from the Zoology Department at the University of Wisconsin.

DeCoursey’s research was based on sixty-eight squirrels that she trapped and raised in Wisconsin. The squirrels were individually housed in cages, each equipped with a running wheel mounted on a bicycle axle. An eccentric cam attached to the axle momentarily closed a microswitch circuit at one point in each wheel revolution to leave a mark on a chart moving at a uniform rate of 18 inches per day. A continuous record of both the number and time of wheel revolutions were thus displayed for later analysis. The continuous record over many yards of paper was cut into daily strips that were aligned by time and then pasted one day beneath the other in sequences of days extending over months. From these records DeCoursey could determine within two minutes when the squirrel had run in the twenty-four-hour cycle and how the activity on one day compared with others.

From the stack of numerous twenty-four-hour records one on top of the other, she saw at a glance that, not surprisingly, the flying squirrels are night-active (unless, of course, they can safely come out to the dairy bar for ice cream). They began running shortly after dark, and then they ran either sporadically or almost continuously (depending on the individual) until dawn, when they ceased all running activity until the next night, or they have two activity periods, one right after dark, and another before daylight.

The above now and perhaps even then rather prosaic results were the preconditions for the real experiment, when DeCoursey next put the caged squirrels into constant darkness. Would they now run continuously or sporadically? The answer was: neither. Surprisingly to DeCoursey, each squirrel ran on the wheel at nearly the same times as it had previously when it had experienced a twenty-four-hour light-dark cycle. That is, the squirrel knew when it was time to be active because it apparently consulted an internal timer. Skeptics cautioned that perhaps the squirrels were instead responding to some unknown external or exogenous signal that was associated with evening, rather than keeping to their previous schedule by using an internal or endogenous time sense.

Ultimately, DeCoursey proved with her squirrels that the timing originated internally, and it was no small irony that her best proof came from the squirrels’ small errors in timing. For example, Squirrel Number 131 on average started to run (in total darkness) every 23 hours and 58 minutes, plus or minus 4 minutes, while another under the same dark conditions in the same room ran 21 minutes later each day; i.e., it had an activity cycle of 24 hours and 21 minutes. That is, under constant dark conditions, one squirrel lost 2 minutes each day while another gained 21 minutes per day. Within ten days of “free running” in constant darkness, one squirrel started activity 20 minutes before external evening, while the other was then 210 minutes late, or 3.5 hours out of synchrony with the external world. If both squirrels had consulted an exogenous or external timer, then they both would have run as if responding to the same drummer; they would have kept the same time.

The squirrel’s clock-running speed is genetically determined, but the time at which the beginning of the animal’s running activity is read off is determined by frequently resetting the clock in reference to an external signal. In squirrels, the signal to which their internal clock is synchronized is the moment of lights-out. We now know, for day-active animals, that light hitting the eye causes the pineal gland of the brain to reduce its production of melatonin, a sleep-inducing hormone that is normally produced rhythmically, on an approximate (but not exact) twenty-four-hour schedule. Hence melatonin pills to combat jet lag. A flying squirrel would have to take them in the morning. When DeCoursey reintroduced a one day light-dark cycle in their environment, then the off-schedule squirrels that had been “free-running” in continuous darkness reset their activity regimen to again start running right after lights-out the next day. Ordinarily the squirrels therefore reset their clocks when subjected to the normally occurring light-dark cycle. Superficially they act as though they respond only directly to darkness or light, and without the experiments that is all one could know. DeCoursey could, of course, have waited to make her observations at total eclipses of the sun. But it would have delayed her conclusions because of difficulties for replicating of observations. Experiments involve making things happen and then applying keen observations of the results.

DeCoursey’s demonstration of internal time-keeping was simple, elegant, and irrefutable. It brought a closure to the debate of whether or not a mammal had an internal circadian clock, and it opened up an area of research in cellular mechanisms. We now have volumes of information on circadian clocks since DeCoursey’s experiments in the 1950s, and the information is becoming of tremendous relevance to medicine. For example, effective dosages of many drugs depend strongly on the time in our own circadian rhythms when they are administered. The molecular mechanisms whereby circadian clocks operate have lately been traced to a number of genes, and the most popular “model organisms” in which they are now studied are no longer flying squirrels but mice and fruit flies.

The circadian clock has many potential uses. It allows hibernating ground squirrels, for example, to measure the daily light-dark durations, and from that data the squirrel can derive information about the changing seasons. Correct seasonal responses are crucial for winter survival. Indeed, the circadian clock mechanisms are necessary for all organisms that must prepare for winter, whether by pupating (insects), migrating (insects, birds, some mammals), or hibernating and physiologically preparing (most northern organisms).

BEING ABLE TO GLIDE from tree to tree is a very efficient way of locomotion, but in flying squirrels the shift to nocturnal activity is costly from the perspective of energy supplies required to keep warm. Energy is saved by gliding, but the ability to glide precludes the laying up of fat stores, such as practiced by their relatives, woodchuck and other ground squirrels, that may get to be obese by fall. Also unlike ground squirrels, northern flying squirrels do not avail themselves of the huge potential energy savings of torpor, since they also don’t have a buffer of energy stores in body fat or food caches, nor change into a more insulating winter coat. As regards energy balance in winter, much seems to be stacked against them. One wonders what solution they might have that counterbalances their numerous presumed shortcomings for winter survival.

I am on the lookout for squirrel nests in my search for overnighting sites of kinglets, so I habitually bang on any tree that holds a nest to see if a kinglet seeking shelter might fly out. All the northern flying squirrel nests I’ve found were in dense spruce-fir thickets. I’ve never chased out a kinglet, but on occasion I’ve been rewarded with seeing one or two individual flying squirrels pop out of a nest, glide off the nest tree, and land on a neighboring tree. Assuming that the squirrels spend half or more of their time in winter in their nests, nest insulation should be of great relevance to energy balance. One nest that I examined in December 2000 was an unfinished framework of dry spruce twigs that contained no lining. Confined between several upward-bending branches, it had probably been abandoned before being finished because the space was too small. It showed, however, that the squirrel starts its nest structure by first making a globe of dry twigs, then inserts the lining. That December I found six other nests that had the same magpielike frames of small dry twigs but that did however contain the nest proper. (One had been torn open, and nest lining had been pulled out.)

Flying squirrel nest, covered by cushions of snow.

The nest linings varied from nest to nest. In one I found a mixture of moss, lichens, grass, and shredded birch bark. In two the lining was almost exclusively finely shredded birch bark. In a fourth it was almost all moss. In a fifth it was exclusively shredded cedar bark, and in the sixth the lining was in two distinct layers of shredded birch and cedar bark. (Many cedar trees in these woods show evidence of some of the outer bark having been stripped off, presumably collected by squirrels although bears also collect cedar bark.) When thoroughly dried this last football-size nest weighed 17 ounces, 12 ounces of which was lining, with a thick 8-ounce shell of densely packed usnea (“old man’s beard”) lichen and a 4-ounce layer of soft, finely shredded cedar bark within that. A good choice—the Northwest Indians used such shredded cedar bark to diaper their babies.

Even after heavy rainstorms the insides of the nests remained dry. Normally in winter these nests are also insulated on top when they are roofed-over with cushions of snow. All the nests had two entrances, one each on opposite sides. These entrances were not visible. They were, like the elastic ends of our mittens and socks, closed. Thus, in structure, each nest was like an old-fashioned hand muff. (In none of these, nor in seven additional red squirrel nests, was there one speck of bird feces, making it unlikely that they serve as kinglet overnighting sites.)

To get a rough idea of whether the flying squirrel’s nest indeed affords much insulation, I heated a potato to simulate the body of a squirrel and examined its cooling rates. At an air temperature of -13°C, a hot potato (60°C) cooled to only 42°C in thirty-five minutes when within the nest, and to 15°C in the same time period when outside it. My rough experiment only says that the nest indeed affords effective insulation. Of course the value of insulation would be much greater in wind, and it would be even more effective in a snow-covered nest. Furthermore, a squirrel, with its downy fur and a bushy tail wrapped around itself, would lose heat much more slowly than a potato. And the slower it cools, the less energy it would have to use up to shiver and maintain a stable and elevated body temperature.

In Jack London’s story “To Build a Fire,” the newcomer to the North was ultimately killed because he got his feet wet. He broke through thin ice under a thick insulating layer of snow on Henderson Creek. His fire that was snuffed by the avalanche of snow under a spruce only made it impossible for him to correct his initial bad luck, or mistake. Ironically, in an insulated sock, mitten, or a squirrel’s nest, a tiny bit of moisture is far more dangerous than deep cold; because wetness destroys insulation. Thus rain, at near 0°C, can be lethal, while snow at -30°C can ensure comfort because it won’t wet and destroy insulation. Without dryness, all lifesaving insulation is for naught, and nest construction or placement must provide for it. Nowhere was this more evident to me than when examining a gray squirrel’s nest in winter.

Gray squirrels’ nests, or dreys as they are often called, appear as haphazard brush-piles of leaves and twigs when we see them piled up high up in trees. All fall and winter I saw one in the branches of an oak tree along our driveway. In mid-January after a heavy rainstorm, the nest blew down, and when I examined it I found it to be anything but haphazard in construction. It was a functionally crafted thing. The outside layer of the 30-centimeter diameter globular nest was of red oak twigs with leaves still attached. The twigs had therefore been chewed off the tree during the summer. Inside this rough exterior I found layer upon layer (twenty-six in one spot where I counted) of single flattened dried green oak leaves. The multiple sheets of leaves served as watertight interlocking shingles, because the nest was dry inside. The leaf layers sheltered a 4-centimeter-thick layer of finely shredded inner bark from dead poplar and ash trees. This soft upholstering enclosed a round, cozy 9-centimeter-wide central cavity. I could not imagine a more efficient functional design from simple common materials. However, not all gray squirrels’ nests are as natty as this. Many that I have inspected were mere piles of junk, as though they might have been fake nests to distract predators so that the real nest could escape being raided.

Gray squirrel nests incorporate leaves, unlike hawks’ nests.

Nests require effort to build, and not all squirrels bother to build one, as I found out with a little help from four of my friends. In the winter of 2000 we saw fresh signs of red squirrels almost everywhere we looked in the spruce forests of Maine where these squirrels live. Yet, we found few red squirrel nests. I wondered if (as is reported in the literature) their winter nests are underground, since I found many red squirrel tracks leading into underground tunnels. I saw numerous tunnels leading under roots of a big rotten pine stump and thought that if a red squirrel nest is anywhere, it should be here. Would this nest be less insulated than that of a flying or a gray squirrel?

Inasmuch as biology is a sterile undertaking until one gets hands-on experience, five of us armed ourselves with spades, pickaxes, plain axes, saws, and a digital thermometer, and then after a good breakfast approached the stump that was in the spruce thicket opposite my swimming hole in Alder Stream. The three entrances going into the ground under the stump had been used within the last day. Bracts of red spruce cones lay in piles on the top of the stump and had been recently chewed. All signs were promising.

It was 8 A.M. and a brisk -14°C on the morning of December 21, 2000, when we started excavating. After only five minutes of work, following tunnels in the spongy duff and soft rotten wood, we were apparently getting somewhere because a red squirrel shot out from one of the three exit holes. We dug deeper and also farther around the periphery of the stump, pulling off huge chunks of frozen humus that, like a carapace, covered the almost dry duff and soil underneath. Then we dug all the way down onto the bedrock under the stump. After about another half hour (and the first signs of skepticism from my friends), a second red squirrel darted out. Now freshly motivated, we dug ever harder, and after another hour and a half we had thoroughly excavated a fifteen-square-foot area all around and under the stump.

We found no food stores and no sign of a nest. Negative data is not considered good evidence, and is generally not reported. However, we had dug all this up so thoroughly that the negative data sure felt like a positive result: No nest. Given the lack of tracks around the stump in the morning, the two squirrels had spent the night in the tunnels under this stump whose spongy humus felt warm to the touch (mostly because it was dry, but measured -0.02°C). The several dry maple leaves that we found in our excavation may have been carried down by the rodents in a weak motivation to build a nest, but in their heavy winter coats they would probably not have needed one.

Flying squirrels may also not bother to build much of a nest when they can snuggle up next to other warm bodies. On November 19, 2000, I was banging yet another tree, this one a dead red maple in the woods near my home, in my continuing quest for birds overnighting in tree holes. I happened to look up in time to see a flying squirrel scamper to the top of the twenty-foot stump and stop there, as if frozen in place. Its flat tail was flush against the bark, and it didn’t move a muscle. I saw then a second squirrel peeking out of the tree hole above me. I looked again, and saw two more scampering up the tree, one behind the other. As my companion and I walked around the tree, the top squirrel jumped off, gliding about fifty feet to the bottom of a live red maple. Seconds later, another squirrel “flew” off in another direction. We left quickly, because we did not want to disturb them further. It had snowed a couple of days earlier, and there were deep nightly frosts. It stayed below freezing all day. Would any of the four return to their shelter?

Flying squirrel in the hole made by a hairy woodpecker in a dead sugar maple.

The next day around 3:00 P.M., after it had snowed hard, I went back to the tree, hoping to see them leave their communal den. I hunkered under the low spreading branches of a red spruce, and there I waited until 4:45 P.M. (forty minutes past sunset), when it was too dark for me to see any more. According to DeCoursey’s experiments, the squirrels should have known it was time to get up, even in the constant darkness of their den, yet I saw no squirrels exit the hole.

Four days later I went to the stump again. This time I lightly jiggled it as I had done before. Nothing came out of the hole. Then I banged the tree hard with an ax. One came out. No more. I banged it again, harder. Three more came out. So—they do come back to the same place. However, a month later, on December 17, when I checked again, no amount of banging raised a squirrel. I climbed up and examined the hole. Surprise! There was no nest at all, and the hole was only three or four inches deep. It contained dry rotted wood and several tiny wisps of dried green moss that must have been carried in. The four squirrels would have nearly filled the whole cavity, and either there was no need for nest insulation or there was no room for it.

From tracks, I knew that these or other flying squirrels were still nearby. My cabin at the edge of a one-acre clearing was within about three hundred feet of where I had seen the four. One seldom sees the tracks of a flying squirrel in the woods, when they land on tree trunks rather than on the snow. But on March 16 that winter I saw where a flying squirrel didn’t quite make it across my acre-sized clearing the night before. The squirrel had hopped out into the field from the south, climbed the maple in the middle, and then twenty snowshoe lengths (sixty-five feet) farther it hit the field again, almost at the edge on the other side. Another, similar flying squirrel track commenced across the clearing from the east, also in the direction of the same big birdbox that I had put up years earlier for some kestrels. The two tracks converging at the sugar maple tree with the birdbox were a clue I could not ignore.

I hit the tree with an ax. One flying squirrel with huge black eyes and soft gray pelage popped its head out of the birdbox. After I started to climb the tree I saw three heads looking out. No—it was four! Coming near the nest box itself, I saw several more squirrels climb out and scamper up ahead of me. Definitely more than five. I counted again—four, five, six—and when I was up under to the box itself I saw even more climb up to the very tip top of the maple. They lined up one behind the other, like a queue of planes taxiing down the runway waiting for takeoff. A couple were still close enough for me to touch. One jumped off, flying toward the field, then veering in midair, changing direction while still airborne and gliding to the right. It made a perfect landing at the base of another maple at the edge of the field. I counted again—there were nine more squirrels on the tree with me. Ten squirrels in all! I reached into the birdbox and felt a flimsy structure of shredded plant material that was warm to the touch. No more squirrels. Not wanting to disturb the animals on the tree just feet above my head, which had pressed themselves immobile onto the tree trunk, I quickly climbed down and then watched as one after the other of the nine squirrels ran headfirst back down the maple trunk to rush back into their birdbox. The tenth that had jumped stayed away for the time being. All the squirrels were adult, and it was close to mating time. I had even seen enlarged testicles on one who’d been inches from my face.

I learned later that the winter sleeping aggregations of flying squirrels had been described, although the northern flying squirrel had not been reported to bunch up to the same extent as the southern species. Curiously, the communal aggregations are sex-specific (Osgood 1935; Maser, Anderson, and Bull 1981). The squirrels huddle for warmth, but why should males not huddle with females, or vice versa?

When I revisited the birdbox in early May right after the snow had melted in the woods, it was empty, at least of squirrels. When I reached in to pull the flimsy nest structure out to examine it more closely, my hand encountered a deep, foul layer of slimy material that was all to easy to identify. Apparently the squirrels had used the nest box not only as a sleeping place. There was not a bit of lichen in the nest, although lichens are a significant portion of the winter diet of flying squirrels, and lichens were a main component of some of their tree nests that I had found. Thus, although some flying squirrels in winter in effect live in a gingerbread house where they may find insulation, others opt for body-warmers and/or a convenient indoor toilet.