Territoriality in Female Wolverines

For reasons that I’ll get into in a future post, I feel suddenly compelled to write up a quick literature review on the topic of territoriality in female wolverines.

Let’s get this out of the way immediately: female wolverines are territorial. This is evident throughout the recent literature. Out of concern for the safety of resident adult wolverines in the US, researchers generally don’t publish maps of wolverine home ranges, but anyone who has had the chance to look at the minimum convex polygons of study animals can see a pretty clear pattern. Adult female wolverines don’t overlap with other adult female wolverines. The territorial boundaries are pretty strict, often down a drainage that both animals will approach and traverse, but that neither will cross. There are instances of incursions and small amounts of overlap, but this is generally less than 2% of any female’s home range. Adult males are also territorial, but with varying degrees of territorial incursion or excursion. The territoriality of wolverines is one of the most important factors for understanding wolverine distribution, their natural rarity, and, importantly, conservation concerns and strategies as we look at a warming world.

Strictly speaking, a territory is a patch of ground that an animal defends, and a home range is a patch of ground that an animal occupies and uses for subsistence. In the wolverine research community, including in published papers, these terms are used in a conflating fashion, although some papers do make a distinction. In presentations and talks and casual conversation, we often use the words “home range” to refer to the area that a wolverine (male or female) occupies, but we mean this as a defended home range, which is actually a territory.

Territoriality in the Literature

While the territoriality of wolverines is widely understood at this point, I was surprised to note that there are few American or Canadian papers or book chapters that deal explicitly with the question of how territoriality functions, or the ways in which it potentially limits population in areas of restricted habitat where the population fragments into a meta-population structure. The Scandinavians, on the other hand, do tend to deal with territoriality, probably because the management imperatives are more urgent in places where depredation on domestic reindeer is a major problem. A Norwegian master’s thesis from 2014 looks at genetic sampling as a method to determine territoriality, in comparison to GPS collar data. Jens Persson’s dissertation deals in some depth with possible explanations for territoriality, and how those explanations may differ between males and females. Proving that the Scandinavians more or less own this topic, a 2017 thesis by Malin Aronsson closely focuses on the dynamics of territoriality and dispersal among female wolverines (and lynx). But the territoriality of the species is taken for granted in most publications, especially the North American publications.

I was also reminded of the absence of peer-reviewed publications focusing specifically on several of the major American wolverine research projects. Doug Chadwick’s popular-science book The Wolverine Way remains the best published account of the Glacier wolverine project. The Rocky Mountain Research Station has put out a number of impressive papers that draw on the Glacier data to ask large-scale questions about habitat relationships, population genetics, dispersal, and climate change effects, but the annual project reports remain the best source of information for the park population itself. The same is true of the Absaroka-Beartooth project, the project with which I got my start in the wolverine world, working as a volunteer. Peer-reviewed papers for that project were somewhat limited by the low numbers of detected wolverines within the study area – sample sizes and statistics being the perennial priority for most journals these days – but the eerie emptiness of prime modeled wolverine habitat deserves some consideration, hopefully in future publications.

Back in 1981, Hornocker et al published a study asserting that wolverines were not territorial. This was one of the earliest studies of wolverines in the lower 48, lasting from 1972-1977, on the Flathead National Forest. It was a telemetry study in which the researchers observed overlap among many different wolverines, and concluded that wolverines were tolerant of fellow wolverines. They reported “no intraspecific strife” and discussed how a wounded female wolverine, whose injuries they first attributed to another wolverine, were likely caused by a mountain lion. (Despite assertions that wolverines are not territorial, the home range maps that are included in this paper do show a familiar pattern – two male wolverines who don’t overlap, one female wolverine who sticks to a fairly tight home range, and a second female, overlapping the first, who makes wider movements that include the home range of the male who overlaps the first female. If I were to guess, many years later, what was up with this scenario, I’d suggest a male-female pair and their juvenile daughter, preparing to disperse. The second male looks like an unrelated individual as his range does not overlap with any of the other animals.) Earlier observational studies of wolverines, and books about the species, also fail to make note of territorial behavior. The same is largely true for the Mongolian hunters and herders who I interviewed; they were well aware of the wolverine’s rarity and ability to travel over long distances, but only a few noted that a wolverine would reappear in a particular spot at intervals. None of those interviewees made a leap to territoriality as an explanation, but some did refer (maybe jokingly) to a wolverine’s nutag, which is a Mongolian concept denoting an individual’s homeland.

Home range maps showing seasonal movements for four wolverines in NW Montana, from Hornocker et al 1981

Home range maps for four wolverines in NW Montana, showing seasonal movements, from Hornocker et al 1981

Observation, track surveys, and radio telemetry, of course, are limited in important ways. These methods allow glimpses of an animal only at the moments when the observer or listener happens to have their eyes or telemetry antenna trained on the animal. Finding dens is more difficult, which – in the absence of DNA techniques – makes understanding relatedness more difficult as well. With the advent of GPS collars and DNA analysis, we were able to observe wolverines more closely and consistently. Hundreds of locations for multiple animals, taken over months, in combination with VHF locations over the course of years, made the territorial behavior of the species clear. Den locations and kit collaring showed that wolverines will tolerate their own offspring within their territories for up to two years after birth. This could account for Hornocker’s and others’ observations of multiple wolverines sharing the same home range, and photos like those that Igor Shpilenok took in Kamchatka, of up to six wolverines on a bear carcass at the same time.

The Wildlife Conservation Society’s Yellowstone Wolverine Project likewise relied on annual reports and white papers to convey results for many years. The project director, Bob Inman, was in the process of getting his PhD, so his dissertation eventually yielded several published papers. Again, though, these deal mostly with larger-scale questions about habitat relationships and conservation priorities. Several of his papers do discuss territoriality, but again, for obvious reasons, most of the published papers don’t include home range maps.

What all of these papers and sources do include, however, is an assumption that territoriality is important. Some imply territoriality; for example, in a summary of draft papers from 2007, the WCS project notes in the abstract for a chapter on wolverine space use:

Mean annual (1 Mar–28 Feb) 95% fixed kernel home range size was 453 km2 for adult females (n = 15 wolverine years) and 1,160 km2 for adult males (n = 13 wolverine-years). Mean percent area overlap of same-sex adults was < 1% (SE = 0.00, range = 0–2%, n = 12 pairs) using annual 100% minimum convex polygon home ranges.

The “<1% overlap” suggests territoriality, although it’s not explicitly stated here. In a 2012 review of wolverine reproductive chronology, however, Inman et al do note this:

Throughout its distribution, the wolverine displays extremely large home ranges, territoriality, low densities, and low reproductive rates (Copeland 1996; Inman et al. 2012; Krebs et al. 2007; Lofroth and Krebs 2007; Magoun 1985; Mattisson et al. 2011a; Persson et al. 2006, 2010). These adaptations are necessary for exploiting a cold, low-productivity niche where growing seasons are brief and food resources are limited (Inman et al. 2012).”  

Another 2012 Inman et al paper on spatial ecology does deal explicitly with the question of territoriality vs. undefended home ranges and is probably the most extensive discussion of this topic in the literature on wolverines in the lower 48. This paper does include some home range maps, with the locations stripped out.

Spatial distribution patterns of the Mustelidae are typically described as intra-sexual territoriality, where only home ranges of opposite sexes overlap (Powell 1979). Wolverine-specific reports exist for both intra-sexual territoriality (Magoun 1985, Copeland 1996, Hedmark et al. 2007, Persson et al. 2010) and for a high degree of spatial overlap but with temporal separation (Hornocker et al. 1983). Arguments against territoriality by wolverines include the lack of ability to defend such a large home range (Koehler et al. 1980). Our data on movement rates in relation to home range size, temporal development of the home range, minimal overlap of same-sex adults, and relatively immediate shifts upon a death suggest that wolverines are capable of patrolling a large territory and provide further support for intra-sexual territoriality. Reproductive success is closely correlated to the amount of energy that a female wolverine can obtain (Persson 2005), and for predators that are capable of individually acquiring prey, the presence of conspecifics reduces foraging efficiency (Sandell 1989). Since wolverines feed on individually obtainable prey and occupy relatively unproductive habitats, it follows that behaviors for maintaining exclusive access to resources would likely have selective advantage. Frequent marking behavior (Pulliainen and Ovaskainen 1975, Koehler et al. 1980) is likely part of an adaptive strategy that involves maintenance of exclusive territories within sexes so that feeding and breeding opportunities are monopolized by dominant individuals and their immediate offspring.

Female wolverine territories in Wyoming, with takeover of one home range by a female kit after the death of the resident adult female. From Inman et al 2012

The Scandinavian literature is also rife with references to territoriality among both male and female wolverines. Jens Persson, who works on wolverines in Sweden, reflects in his dissertation on the reasons for territorial behavior in female wolverines, and concludes that it’s related both to food, and also potentially to the need to protect kits against infanticide. Historically, it was widely believed that male wolverines would kill any kits they encountered, even their own; this has since been proven false, but the idea that males kill unrelated kits persists. Persson is the first researcher I’m aware of to suggest that female territoriality may actually be a defense against other females intent on infanticide.

Females could also gain from infanticide by eliminating non-related progeny to decrease future competition for territories or denning areas for her and her progeny. In addition, the death of an unrelated infant could also reduce the net reproductive success of a competitor (Hrdy & Hausfater, 1984). Competition for territories determine dispersal behaviour in female wolverines (Paper IV), suggesting that there is strong competition for territories among female wolverines.

 Wolff and Peterson (1998) hypothesized that a primary function of female territoriality in solitary mammals could be to protect vulnerable young from infanticidal conspecific females. Four predictions can be deduced from their offspring-defence hypothesis: 1) Female territoriality should be associated with young that are vulnerable to infanticide. 2) Female territoriality should be associated with defence of offspring, and therefore most pronounced during the offspring-rearing season. 3) Defence will be greatest against the segment of thepopulation that commits infanticide and against those individuals that females can dominate. 4) Optimal territory size should be a function of intruder pressure, intruder detectability, female response distances and offspring vulnerability, and changes in food abundance and distribution should not affect territory size directly unless they are correlated with the other factors. In concordance with predictions 1-3, wolverines have altricial young that are vulnerable from late winter until late summer (March – August) and female territoriality seem to be strongest during this period (Magoun, 1985; Landa, Lindén & Kojola, 2000). We lack data to evaluate prediction 4. However, in contrast to prediction 4, I believe that food actually is an important determinant of territory size in wolverine females (see Banci, 1994).

Malin Aronsson’s 2017 thesis examines the territorial dynamics of female wolverines, using the vast dataset from Swedish studies dating back to the 1990s. She makes some interesting observations about the counter-intuitive conclusion that wolverines are territorial despite living in low-resource environments, which is the opposite of what studies on other carnivores would suggest:

Wolverines are highly territorial (Persson et al. 2010), and by comparing space use overlap between years for the same individual I found that wolverines show high territorial fidelity resulting in a stable distribution of resident individuals. Interestingly, territorial fidelity in general is predicted to be low in habitats where food resources are low, variable, unpredictable or deplete fast (Wauters et al. 1995; Kirk et al. 2008; Edwards et al. 2009), which corresponds to the characterization of wolverine habitat in general (Inman et al. 2012b), and particularly in this study area (Person 2005). However, scavenging and caching are integral parts of wolverine biology (Inman et al. 2012b; Mattisson et al. 2016), which increase resource predictability, decrease depletion rate and create a valuable resource (i.e. cache sites) to defend, promoting high territorial fidelity despite the unpredictable environment (Tye 1986; Eide et al. 2004). In addition, occurrence of more efficient predators, such as the Eurasian lynx (Lynx lynx), provide carcasses for direct consumption and caching (Mattisson et al. 2011b). Furthermore, both males and females showed higher between-year fidelity at the territory level (i.e. 90% isopleth) compared to the core areas (i.e. 50% isopleth). That fidelity was lower at the core area compared to territory level suggests that it is critical to maintain the outer territory boundary to secure long-term resources, while the most used area within the territory may vary between years due to spatial fluctuations in key resources, or, for females, location of den sites may vary between years.

Aronsson also documents a few cases of territorial adult females shifting territories after successfully reproducing. Why this happened, and how frequently such moves occur, would be interesting questions for further investigation.

Another Scandinavian paper, a master’s thesis from 2014 by Espen Gregersen, compares home range and territoriality derived from GPS data to those derived from scat analysis. The question in this thesis was not so much “are wolverines territorial?” as “can wolverine territoriality be detected using non-invasive methods like scat analysis?” It’s an interesting question and one of relevance to those of us who have limited resources for large-scale trapping efforts in places like, say, Mongolia. Gregersen concludes that yes, we can indeed determine home range size and observe territoriality using scat samples – but it takes a very large number of samples. This thesis also addressed the question of territorial turnover after the death of a resident adult, which is particularly interesting at the southern edge of distribution, where wide separation of habitat patches makes recolonization less certain.

Finally, a forthcoming book chapter from Copeland et al. proposes a slightly different take on territoriality among male and female wolverines. That chapter will be out soon and I’ll look at it in depth once it’s published, but it too reinforces the idea that female wolverines are highly territorial, and maybe even more strictly territorial than males.

There are many other papers out there that include brief mentions of territoriality and intrasexual exclusion in home ranges. These are just a sample, and this write-up fairly cursory, but I hope they’re adequate to illustrate that wolverines – both male and female – are territorial.

Territoriality and Conservation

At this point, territoriality in wolverines is accepted as an important feature of their life history and ecology. The question of why hasn’t yet been answered, but the fact that wolverines require such large territories, and the fact that their reproductive rates are so low, accounts for their natural scarcity on any landscape – let alone one in which suitable habitat is located only at certain elevations in widely scattered patches across a sea of non-habitat.

Female territories structure the wolverine population. Females must have adequate resources to meet their needs and, hopefully, to reproduce. They occupy and defend territories that allow them to do this. Males in turn seem to select for territories that overlap with resident females. Whether this represents a territorial strategy for sexual monopoly, or whether it’s defined by the male’s capacity for paternal investment, or some combination, is worth investigation (male wolverines seem not to always entirely encompass a female’s territory, leaving her open to potentially overlap with other males, which raises questions about the accepted narrative of males “controlling” access to females). Both males and females disperse over long distances, although the longest movements have been observed in males like M56.

Habitat availability for females is the limiting factor on wolverine population growth and range expansion in the US Rockies. I’d hypothesize that keeping a certain number of territories occupied is critical to the long-term persistence of wolverines in the lower 48, and that there is some distinction to be made between female population numbers, strictly speaking; the percentage of habitat that’s occupied; and where that habitat is located in relation to other habitat. There’s been an enormous focus on “connectivity”– concurrent with the fashion for corridors among conservationists –  but a surprising lack of attention paid to the population nodes themselves. For example, the question that Gregersen raises about recolonization of vacant territories is interesting and important, especially given the observed disappearance of wolverines from places like the Tetons. Presumably this disappearance represents some kind of natural cycle of die-off for a relatively isolated population node, but how long does it take before those territories are reoccupied? And how is time-to-recolonization related to population density and occupancy of the next-nearest population nodes? Questions about functional connectivity among wolverine population nodes are important, but connectivity as a single conservation strategy for wolverines seems like an odd allocation of resources; wolverines don’t migrate, they disperse, and their dispersal patterns are unique and erratic. They are likely to benefit from the broad and intense focus on connectivity and road-crossing structures for other species, but trying to preserve wolverine-specific corridors seems like a good recipe for driving oneself nuts. As one of my Mongolian interviewees once said when discussing wolverines, “One day it’s here, the next day it’s 50 kilometers away. It could turn up anywhere!” I hope to see a greater focus on what’s going on within habitat in the future, including investigation of questions about what drives territoriality and territory size, and how territorial turnover works in a meta-population.

That’s it for now. If any of you have any thoughts about the function of territoriality in female or male wolverines, if you want to point out an obvious resource on this question that I overlooked, or if you just want to say hi, please feel free to comment.

References (with apologies for lack of consistent style formatting and for referring to multiple authors as “et al” instead of writing them out. Time constraints!)

Aronsson, M. 2017. ‘O Neighbour, Where Art Thou?’ Spatial and social dynamics in wolverine and lynx from individual use to population distribution. Doctoral dissertation. Swedish University of Agricultural Sciences. Uppsala. ISBN (electronic version) 978-91-576-8822-4

Chadwick, D. 2010. The Wolverine Way. Patagonia press.

Copeland, J. P., Landa, A., Heinemeyer, K., Aubry, K. B., van Dijk, J., May, R., Persson, J., Squires, J., and Yates, R. 2017. Social ethology of the wolverine. In: Biology and Conservation of Musteloids. Edited by David W. Macdonald, Christopher Newman, and Lauren A. Harrington: Oxford University Press. DOI 10.1093/oso/9780198759805.003.0018

Gregersen, E. 2014. Assessing territoriality in wolverines (Gulo gulo) using non-invasive genetic sampling. Master’s thesis. Norwegian University of Life Sciences.

Hornocker, M and H Hash. 1981. Ecology of the wolverine in northwestern Montana. Canadian Journal of Zoology. V. 59. pp. 1286-1301.

Inman et al. 2007. Wolverine space use in greater Yellowstone. In Greater Yellowstone Wolverine Program Cumulative Report. Wildlife Conservation Society.

Inman R. et al. 2012. The wolverine’s niche: linking reproductive chronology, caching, competition, and climate. Journal of Mammalogy 93(3):634-644.

Inman, R. et al. 2012. Spatial Ecology of Wolverines at the Southern Periphery of Distribution. Journal of Wildlife Management. 76(4). 778–792. DOI: 10.1002/jwmg.289

Persson, J. 2003. Population Ecology of Scandinavian Wolverines. Doctoral dissertation. Swedish University of Agricultural Sciences. Uppsala.



New Study on Food Storage and Reproduction

Wolverine kits, at least several weeks old. Borrowed without permission from care2.com. Too cute not to repost.

For years now, we’ve known that wolverines are found in regions of deep spring snowpack and low summer temperatures – Copeland et al’s 2010 paper elucidated this elegant model of wolverine distribution by mapping known wolverine locations from all over the globe and placing these locations onto a map of global snowpack on May 15th, and maximum August temperatures of less than 22 degrees Celsius. The paper showed that wolverines are tied to the cold regions of the northern hemisphere, and linked this dependence on cold to the fact that wolverines give birth in snow dens. The paper was groundbreaking and its publication eagerly anticipated, because it provided enough evidence of climate change threat to support the USFWS’ 2010 decision that wolverines are warranted for listing.

A new paper in the Journal of Mammalogy by Bob Inman, Audrey Magoun, Jens Persson, and Jenny Mattisson expands our understanding of the links between wolverines and the cold, exploring the complex reasons for the evolution of wolverine reproductive timing and behavior.  If the Copeland paper told us that wolverines are indeed climate sensitive due to denning requirements and a cold-adapted physiology, this paper asks why those denning requirements and physiological limits are so strict – in other words, what adaptive advantage does cold-climate specialization offer to the species?

Inman and his co-authors suggest that the wolverine’s strategy is driven by the nutritional needs of the species, and of reproductive females in particular. Pregnancy and nursing are the most nutritionally demanding activities that any wolverine – any mammal, in fact – undertakes, and the successful rearing of young requires a secure source of food. The timing of wolverine births early in the year, according to the paper, allows females to take advantage of a flux of winter-killed ungulates that they have cached, and the nursing and weaning periods encompass the spring surge in baby ungulates and the brief summer explosion of small mammal populations. Persistently cold climates allow wolverines to cache food in an environment where it won’t go bad, allowing them to store sparse nutritional sources and “be efficient in channeling available food resources towards reproduction.” And, suggest the researchers, by living at the outer – or upper, in the case of the Rockies – margins of the ranges of other predators such as wolves and bears, wolverines avoid direct competition with much larger and better equipped rivals. Put all of this together, and wolverines obtain a neat set of advantages by living in a harsh and desolate landscape.

The paper relies on synthesis of existing research, and contains a great section reviewing all the data on reproduction. Wolverines mate in summer, but implantation of the fertilized embryos is delayed until winter. We generally say that wolverines give birth around February 14th (Valentine’s Day) and that the kits are weaned by May 15th (Mother’s Day), because these dates are easy to remember and are, on average, accurate. But in the details of known wolverine birth dates, we see a much wider range, with some wolverines giving birth as early as January, others as late as April. This means that implantation – nidation, in scientific parlance – also occurs over a range of dates, from November through January, with a 45-day gestation. Most of the births do occur mid-February to early March, but the range offers the possibility that female wolverines are responding to environmental factors on a year-by-year basis (several of the females were monitored over several years and had different chronologies in different years.) Does this mean that wolverines might possess some latitude in timing births to correspond to changing snow and nutrition availability over the longer term? Like almost everything else about wolverines, we have no definite answer, but it’s interesting to think about.

The paper also summarizes reports in some studies of very high levels of wolverine pregnancy (implanted embryos). This, too, is interesting, since female wolverines seem to raise very few litters. The data suggest that many females who give birth lose their litters early. Nutritionally, this is more adaptive than struggling to keep a litter alive and then losing it later, since it represents a much smaller investment of resources. Losing a litter early during a year when conditions are sub-optimal gives females a chance to maintain better body condition for next year’s litter, when conditions might be better. All of the attention to this question of reproduction is critical, since we absolutely must understand these dynamics in order to determine effective conservation strategies.

This paper received a fair amount of attention in the press, most of it focusing on food storage: here at the Examiner, at the Huffington Post, at LiveScience, at USA Today (there is, perhaps, a wolverine fan on staff there, because this makes two articles on gulos in two months), at National Geographic, and at MSNBC. I’m sure that there are others out there, too. Most of these articles quote lead author Bob Inman, of the Wildlife Conservation Society, sounding very respectable and scientific – but his secret identity is revealed in an article discussing whether or not the ubiquitous presence of Wolverine the X-Men superhero may also be an adaptation to climate change, in which Wolverine uses his extensive social connections as a food storage strategy: “Understanding why and how Wolverine exists where he does and the various adaptations he has evolved to eke out a living will better inform population management strategies and conservation of the comic industry,” said researcher Robert Inman. I’d heard that Bob was away in Sweden for several months, but now we know the truth.

My own summary of this paper is delayed; I usually like to talk to the scientists to make sure I have all of the implications correct, but the paper came out just as I left for Mongolia, and I haven’t had the chance to talk to anyone. So this is just a basic summary. More later.