Austin Allison—a PhD student at Colorado State University and recent MS graduate from the University of Idaho—discusses his recently accepted paper: “Why hibernate? Tests of four hypotheses to explain intraspecific variation in hibernation phenology.”
About the Paper
Why do animals hibernate? I—like many people—assumed I knew the answer to that question. Animals obviously hibernate to avoid seasonally unsuitable environmental conditions such as freezing temperatures and a lack of food, or so I thought. One of my first assigned tasks upon joining the University of Idaho’s northern Idaho ground squirrel research project in the fall of 2015 was to calculate the average date on which radio-collared adult northern Idaho ground squirrels began hibernating that year. To my genuine surprise, the date I calculated was 15 July! I was so concerned I had made a mistake that I recalculated the average date several more times. Each time, I came to the same conclusion: adult northern Idaho ground squirrels started hibernating months before the cold, snowy Idaho winter set in. This fact, of course, was not news to my supervisors, who had studied these rare squirrels for years at that point. Nor would this fact shock anyone who ever studied the numerous species of Holarctic ground squirrels inhabiting the plains and mountains of the western United States, many of which apparently enter hibernation during the warm, if not downright hot, summer months. But I was surprised. Why, I wondered, would these squirrels start hibernating when above-ground conditions were still apparently suitable for ground squirrel activity?
The answer I received from those in the know was that the squirrels likely entered hibernation when the plants on which the squirrels rely for food and water wilted under the hot, dry conditions that characterize the small region in west-central Idaho these federally threatened squirrels inhabit. This answer satisfied me at the time, but it became clear during the following years that seasonal vegetative senescence could not entirely explain observed patterns in ground squirrel hibernation behavior. For instance, juvenile northern Idaho ground squirrels remain active for weeks after adult squirrels settle into hibernation. If the squirrels entered hibernation when food and water became limiting factors, how did juvenile ground squirrels remain active and continue to grow rapidly for weeks following adult entry into hibernation? Moreover, extended bouts of torpor (a state of severely reduced body temperature and metabolism)—typical of mammalian hibernation—impose significant physiological costs on individual animals, including memory loss, reduced immune function, and cellular damage. Thus, it appeared that northern Idaho ground squirrels forewent energetically profitable foraging opportunities and accepted these non-trivial physiological handicaps by entering hibernation in mid-summer. But why?
Increasingly, we began to consider the possibility that squirrels may use hibernation not only to avoid energetically unprofitable foraging conditions, but to balance the costs and benefits of foraging under high levels of predation risk. Ground squirrels, after all, are prey for many predators inhabiting our study area, including coyotes, foxes, badgers, martens, weasels, hawks, and falcons. Of that formidable arsenal of squirrel hunters, only badgers are known to excavate and consume hibernating squirrels. Indeed, most mammals that hibernate—including northern Idaho ground squirrels—have significantly higher survival rates per unit time during hibernation than during the active season. If hibernation does provide an escape from predation, a better question than “Why hibernate?” might be “Why not hibernate?”. That is, what necessitates above-ground activity at all? Two obvious answers to this question are that squirrels must come above ground to reproduce and to forage for food that provides the energy necessary to survive and reproduce.
Thus, we deduced that if squirrels hibernate explicitly to avoid predation, then heavy squirrels with significant fat stores should enter hibernation earlier than lean squirrels. This is because excess fat stores allow squirrels to hibernate longer without risking starvation, thereby reducing predation risk and increasing the odds of surviving to reproduce during the following spring. Additionally, reproductive adult squirrels should emerge from hibernation as early in spring as possible to mate, because pups born earlier in the season have a higher probability of surviving to adulthood to reproduce themselves. However, emerging early also exposes squirrels to higher predation risk. Therefore, young squirrels that do not reproduce should emerge from hibernation later than reproductive adults to reduce predation risk and increase the odds of surviving to reproduce as adults the following year. Conversely, if squirrels are forced into accepting the physiological costs of hibernation by a lack of food and water, heavy squirrels should use their excess fat stores to prolong the active season, thereby mitigating reductions in survival and reproduction assumed to accompany the litany of negative physiological effects associated with prolonged hibernation. Similarly, squirrels seeking to avoid the costs of hibernation should emerge in spring as soon as environmental conditions allow, resulting in more synchronous emergence timing between reproductive adults and younger, non-reproductive squirrels.
About the research
To distinguish between these possibilities, we captured and weighed wild northern Idaho ground squirrels at several study sites throughout the species’ limited range. We also attached small data loggers to the squirrels. The data loggers recorded when squirrels were above and below ground by measuring the ambient light levels experienced by a squirrel. The data loggers also recorded squirrel body temperature, allowing us to determine when a squirrel entered and exited torpor during hibernation. Together, these data provided precise estimates of hibernation onset and termination dates and times (to the nearest five minutes!). We attached the data loggers to squirrels in summer, shortly before the squirrels entered hibernation. We then had to retrieve the data loggers the following spring to access the data stored on board. In many ways, this proved to be the most difficult aspect of the project. It took years to perfect a system of attaching the data loggers to radio-collars such that the loggers would not fall off the collars before we could recover them. We ultimately settled on a system that involved three types of glue in addition to two types of wrapping to secure the data loggers to the collars. This intensive attachment system has not yet let us down! The radio-collars to which we attached the data loggers helped us to locate the data loggers following hibernation. Unfortunately, radio-collars sometimes fail and squirrels wearing collars are frequently carried off by predators before we can retrieve the data. Another pitfall of this collar system is that ground squirrels burn a lot of fat during hibernation, meaning they lose a lot of weight and a lot of volume. In doing so, many squirrels slipped out of their collars before emerging from hibernation or shortly thereafter, often leaving the collars (and so our data!) in burrows several feet below the surface or under large rocks or tree roots. It therefore took several years for us to acquire sufficient data to test our hypotheses.
Despite these travails, capturing, collaring, and tracking squirrels is incredibly fun! Our study area in the mountains of western Idaho—just above Hell’s Canyon, where the Snake River has carved the deepest gorge in North America—is astoundingly beautiful. Spring wildflowers abound in the mountain meadows the squirrels inhabit, and a diverse array of Rocky Mountain wildlife coexist with the ground squirrels we study. Idaho is known as the gem state, but the state remains a hidden gem in many respects such that our remote study sites receive relatively little human use, even during the warm summer months. It is difficult to imagine a more idyllic setting in which to toil away at unravelling the mysteries of the natural world.
Regarding the mystery of why ground squirrels initiate hibernation in mid-summer, we found strong support for the hypothesis that northern Idaho ground squirrels hibernate to avoid predation. Heavy squirrels entered hibernation earlier and hibernated longer than did lean squirrels. One particularly plump female squirrel in 2021 entered hibernation on June 23rd and remained ensconced in her hibernation burrow until April 19th the following year—a hibernation period of 300 days! Reproductive males emerged from hibernation in spring about two weeks before adult females, with younger squirrels emerging about a week after the adult females on average. These findings suggest northern Idaho ground squirrels seek to maximize hibernation to avoid predation, but that those attempts are constrained by the need to forage and reproduce. This study is exciting to me because it challenges conventional wisdom and expands our understanding of the adaptive purpose and flexibility of one of the most remarkable traits in the animal kingdom.
Naturally, this study builds on the work of others. We certainly were not the first research group to entertain the possibility that some animals might hibernate to avoid predation; nor were we the first to empirically test this hypothesis. Extensive prior work on edible dormice in Austria (led by Claudia Bieber of the University of Veterinary Medicine, Vienna), which can hibernate for up to 11 months in non-reproductive years, suggests that those cousins to ground squirrels (both are rodents) also hibernate to avoid predation. However, we believe our study is the most rigorous treatment of the topic to date and provides a framework for future tests of the hypothesis in other study systems. More broadly, our work not only supports the predation avoidance hypothesis, but also supports the notion that animals must behaviorally and physiologically navigate trade-offs between survival and immediate reproduction to maximize lifetime reproduction—long a central idea in ecology and evolutionary biology. Often these trade-offs manifest as decisions regarding when and where to forage under the risk of predation. Consequently, although this paper may be most acutely of interest to other research groups studying the eco-physiology of hibernation, the paper also incorporates themes of broad interest to ecologists and evolutionary biologists studying animal behavior and life histories.
Many important questions relating to hibernation behavior and physiology and their role in life history evolution remain unresolved. Might some animals hibernate not to avoid predation, cold, or famine, but instead to avoid disease or competition? Why do hibernating mammals have slower life-histories than is predicted by their body size? How, if at all, do animals that maximize hibernation duration to avoid predation mitigate the physiological costs of extended torpor? We hope to make further contributions to this field, but we hope also that this paper will encourage others to tackle these questions or to challenge other bits of ecological conventional wisdom.
About the Author
Like many wildlife ecologists, I developed a fondness for animals and the outdoors at a young age. I grew up hiking and exploring the outdoors with my parents in the densely forested hills outside St. Louis, MO. I primarily reveled in the herpetological diversity of the region. I learned the basics of genetics in elementary school by breeding corn snakes in my bedroom with my dad. Perhaps unsurprisingly, I wanted to be a herpetologist. However, by the time I moved to Colorado for college, I had instead decided that I wanted to study bears and other large, ferocious mammals. I fell in love with the wide-open Western landscapes those charismatic animals inhabit. I certainly never planned to be a ground squirrel biologist. I hardly noticed the numerous ground squirrel species that surrounded me in Colorado during my undergraduate studies. But after graduating with a B.A. in biology from Colorado College, I discovered how difficult it is to find work on large mammal projects. I ended up working as a technician on a disease ecology study of prairie dogs, and I loved it! This experience led me to the northern Idaho ground squirrel project at the University of Idaho, where I started as a technician and eventually earned a M.S. studying the behavior and life history of those federally threatened squirrels. The paper discussed here was actually a side-project I put together after defending my thesis, but it might be my favorite paper to have so far come out of my time on this project.
I recently began a Ph.D. at Colorado State University studying the evolutionary integration of life history with behavior and physiology in Arctic and Wyoming ground squirrels. I hope my dissertation here in Colorado will further the work I completed as a M.S. student in Idaho. Ground squirrels, it turns out, are among the most tractable study systems in which to test hypotheses explaining patterns in physiology, behavior, and life history. Plus, they’re cute! I spend as much of my free time as I can outdoors, exploring the unfortunately few truly wild places remaining in this world—often by foot, sometimes by bicycle. I also greatly enjoy road tripping to such wild places and experiencing the great variety of human cultures found along the way.
Enjoyed the blogpost? Read the research here!