Moreover, wolf and prey populations operate at an even larger, “metapopulation” scale, with even broader landscape relationships. The Burwash caribou herd we studied in the years just before our current Algonquin research moved seasonally in and out of Kluane National Park in the southwest Yukon, periodically exchanging members with the adjacent Aishihik herd. Similarly, a huge Steese-Fortymile herd in Alaska numbering in the tens of thousands shrunk so dramatically that during my Ph.D. years we saw only about a dozen caribou on the northern sector of its range. Just a few years later, a picture in the Fairbanks newspaper showed an estimated ten thousand animals grazing there. They had come back.
There is only one caribou herd in northwestern North America if you accept a proper ecological time-space scale. It ebbs and flows across the land — eddies here, disperses there, declines here, and fills in there. Similarly, there is only one wolf population, the caribou’s constant companion, tracking its fate, responding to it both locally and regionally through adjusting densities.
That is how nature works on the large scale. Because of local immigration and emigration, population sizes and boundaries are ever-adjusting, kaleidoscopic, temporary. We know it now, with plenty of evidence, but can we adjust?
Onto this shifting mosaic of nature we have established a small set of disconnected parks and reserves, hoping they will protect the species within. They will, the smaller, less space-demanding species, providing patches of suitable habitat with opportunities for populations to move between them. Red squirrels and river otters are served well by a park such as Algonquin. But not wolves.
For wolves to be protected, first they need an adequate core. It is a fact of geometry that fully 50 per cent of Algonquin Park lies within only ten kilometres of a park boundary. With wolf-pack diameters averaging fifteen to twenty kilometres, only packs living in the centre of the park do not range routinely outside the park. That represents about sixteen to eighteen packs, or perhaps seventy-five animals in late winter. The other half of the population is unprotected part of the time, especially vulnerable to shattered social systems and coyote-gene invasion.
Put only a ten-kilometre-wide protection zone around the park, and wolf protection immediately doubles. Vacant lands where canoeists won’t hear wolves will be less common. Coyotes will be held a little more at bay.
But that is not enough. Also necessary are extensive linkages to other parks or wildlands, to provide for metapopulation flow, buffer the vagaries of nature that can lead to local population decline, and reduce both genetic isolation and coyote influence. Such a provision would require broad no-kill zones linking Algonquin with Killarney Park on Georgian Bay to the west, Temagami to the north, La Verendrye and Laurentide parks in Quebec to the east, and the Madawaska Highlands to the south. Call these broad bands “designated wolf protection zones,” a term coined by Monte Hummel, and redress to some extent the enormous imbalance that would still exist between wolf protection and exploitation in Ontario and eastern Canada.
The Foys Lake pack, stringing thirteen sets of footprints across frozen bogs, dividing home-chores at its rendezvous sites, forcing its boundaries against the Redpole pack to the west and the McDonald Creek pack to the east, made us rethink the concept of group selection and its implications for wildness. Although group selection is still not widely accepted, we join a growing number of biologists who argue its validity. Why should natural selection operate at only one level, the level of the individual? Better to see it in a nested, hierarchical way, like the very structure of nature it forges. Not only does the best-fit wolf survive to leave the most offspring, but so does the best-fit pack.
Survival of the best-fit group has been so obvious in our own evolutionary history that it should come as no surprise. We, like wolves, were a cooperative group-living animal exhibiting division of labour, competing with neighbouring tribes. Technological advances repeatedly gave one society after another the upper hand.
One form of group selection is more acceptable among biologists: “kin selection” occurs when individuals in the group are closely related. Then cooperation, even altruism, makes the most sense, because members of the group share the same genes — parents, offspring, grandparents, siblings, aunts, uncles, nephews, nieces. Among Algonquin wolves though, what we are seeing is not just kin selection, because the packs are not simply kin groups. We find too much dispersal for that. We had too many instances of adoption resulting in almost as much genetic relatedness between packs as within them.
In the Foys Lake pack, with its extensive, resource-rich territory, were two large adult males and one smaller one, at least two of them breeders according to genetic evidence. The pack must have been formidable to other packs who knew them around their boundaries, and dominant down in the farmlands where packs congregated each winter. But despite its strength, the pack did not survive.
It died largely because the group fitness it represented was for surviving in nature, not for surviving the onslaught of humans. In the end, it did not matter that the Foys Lake wolves were good at making a living, at cooperatively finding and dispatching prey, caring for and teaching their young, and actively or passively defending their boundaries.
What mattered more, but they were not adequately selected for, was avoiding humans: neck snares, bullets, and traps, and staying in the ditch when a vehicle approached instead of running out in front of it. Those are the things that really matter now, under the new world order.
It is a new world order, one where humans have taken over as a predominant selective force operating on most species on Earth, and we may not be aware of its frightening significance. It means that the species around us are incrementally, insidiously becoming “made by humans” instead of “made by nature.”
The quality of wildness inherent in nature is neither subtle nor obscure; it can be defined, even measured. A wild individual, pack, population, even ecosystem is simply one that is predominantly being continually shaped by natural selection, not by human selection.
A wolf pack with a rendezvous site in an alfalfa field near the Minnesota-North Dakota border, as biologist Eric Gese described to me, is not a wild, shaped-by-nature wolf pack. It is a conservation success but no replacement for truly wild wolves. That it can live there, that wolves have expanded in Minnesota and into northern Michigan and Wisconsin, and in Spain, Italy, Germany, Romania, and other European countries, is due to legal protection. Wolves can live almost anywhere, even in human-dominated landscapes, if there is prey, as long as they are not persecuted.
Contrary to the suggestion I made at the end of my book Wolves and Wilderness in 1975, wolves and wilderness are not inseparable. In light of this range expansion, I have reconsidered. Accepting the dose of reality that humans have and will continue to change and control much near-wild land, and that many human influences, such as the spread of contaminants, are so subtle, there is room for conservation policy to maintain wolf populations in places where humans impact them and select for new adaptive traits.
But real, wild wolf populations under the influence of natural selection are different, and they are inseparable from wilderness. To keep nature-forged wolves in parks and surrounding lands, we must maintain their wild, evolutionary crucibles predominantly under nature’s control. The wolf — highly social, group selected, cooperative — is not just numerically, but also qualitatively, sensitive to the effects of exploitation, more so than many other species. Social systems are fragile things, dependent as they are upon subtle interrelationships and the passing down of tradition. They can be easily broken.
Wild wolves, not made-by-human wolves, wild grizzlies, wolverines, cougars, great spotted cats, large primates, and all their dependants should be the goal of all large “protected” areas wherever these species are found and, in supportive ways, their surrounding lands. Having such places is a matter of humility.
Under this definition of “wild,” the Algonquin Park wolf population fails to qualify. The population declined by 43 per
cent between 1989 and 1993, then increased for three years to within 9 per cent of its previous high, then declined again by 28 per cent in 1997, and has stayed equally low into 1998. These changes in numbers have been driven predominantly by mortality of yearlings and adults, with highs of 61, 55, and 36 per cent in various years. Human killing has been the major cause of death, mainly snaring and shooting, and exceeds all other causes combined. It exerts the predominant selective force on wolves over one year old. The killing has resulted in many territory vacancies that have taken up to three years to refill.
Recruitment of pups to yearling age failed to offset losses in years of the heaviest killing. Recruitment has been roughly steady at 20 to 25 per cent, despite exploitation, meaning that the population cannot withstand combined annual mortality of more than that without declining. This level is low compared with other studies, indicating that the Algonquin Park population is particularly sensitive to exploitation. Causes of low yearling recruitment are unknown. Speculatively, canine parvovirus may be killing pups, or they may have low viability due to possible inbreeding depression (suggested by an unexpectedly low genetic variability), or parental care may be inadequate because of a predominance of young parents in the population due to exploitation. Average longevity of yearlings is only 3.6 years, and since wolves normally do not breed until they are two, or sometimes three, years of age, parental experience is limited.
The killing, besides causing numerical drawdown, has influenced social structure. It has resulted in small late-winter pack sizes averaging only between 3.5 and 4 animals. Vacant lands have driven dispersion to high levels (with consequent high genetic mixing within the park and possible increased interspecific pack tolerance). We have found less than expected coordinated pack hunting, less traditional use of den and rendezvous sites, less stable pack boundaries, and more lone wolves, especially adults.
This is an exploited, not a protected, population, a population predominantly under human-selection pressure. Only a handful of packs in the centre of the park live predominantly under nature’s influence. They represent less than half the park population, less than one hundred animals, too few to maintain the characteristics of their species in the long run. This constitutes a failure in park management. It rests with the MNR to change that by completing a zone of protection around the park.
We did not expect to find the group-tolerance system of land division that was just as obvious as the standard, defended territories normally described for wolves. Trespass rarely evoked active aggression. Nor did we expect the wolf migration and concentration, the biggest surprise of the study, and a dramatic example of behavioural plasticity and adaptability. In the deer yard each winter, packs showed an even greater level of tolerance by simple, short-range avoidance, and we concluded that this tolerance was largely due to the abundance of food.
But, paradoxically, this concentration resulted each winter, and still does, in population loss: while individuals maximize their own fitness by following their food supply, the population declines due to human killing. Similarly, the deer migration, while adaptive for the individual, exposes the population to greater human hunting pressure, often lowering its numbers too.
These examples show that maximum individual fitness does not necessarily translate, as you might expect, into maximum population fitness. This is especially evident when humans alter the system as they have in and around Algonquin Park. Natural selection, operating on individuals and packs, may not always bail out populations faced with environmental hazards. The broad consequence is that we cannot expect species to adjust automatically to a made-over human world. Species reach an evolutionary accord between individual, group, and population welfare. When environments and selection pressures change, we may break that accord. It is part of life’s vulnerability.
With great care we transported vials of wolf blood to the Pembroke Animal Hospital after wolf captures, but we had no idea that they would turn out as significant to wolf conservation as they have. We did not expect to raise the question of whether the wolves we were studying are an entirely different species. Nor did we suspect the possibility that the Algonquin Park population would turn out to be its last extensive stronghold. After a decade of research, a new avenue of inquiry suddenly emerged. We hope, research funds willing, to pursue a scientifically based conservation strategy for remaining lycaon range. With new interest we look at the red wolf recovery program in the eastern United States because we may be studying the same species, with many of the same pressures.
In the interests of not living in a biologically impoverished world, it matters if Algonquin wolves are distinctive from other gray wolves. Much of the case for preserving biodiversity rests at the level of genes. Subspecies have different percentages of genes, and species are even more different. Algonquin wolves trail their own evolutionary history. They may have their own unique ways of dealing with the contingencies of life in eastern forests.
We know that too much association with coyotes can result in gene swamping — the loss of wolf appearance and traits — such as was happening with red wolves just before they were extirpated. We also know that fragmentation of the wolf population by over-exploitation invites coyotes to invade wolf lands. It shocked us to find wolf-coyote hybrids that are distinctively different from wolves on wild lands where we thought there were wolves to the south and east of Algonquin Park. Possibly, gene swamping has happened almost everywhere else in the former Algonquin wolf’s range beyond the immediate vicinity of the park.
Will we do nothing, just let it happen, lose the Algonquin wolf, just when we have discovered its uniqueness?
Three prey species provide the Algonquin wolves with food. Of the three, white-tailed deer is by far the most influential, probably because it is easily caught and comes in optimum-sized packets of energy. Wolves show their preference for deer by following their migration and by selectively hunting any that remain on their territories.
The deer population began a slow recovery in the early years of our study, then fluctuated and declined. We estimate recent deer numbers to have ranged between 0.5 and 0.75 per square kilometre over our study area on the eastern side of the park. This figure is low by eastern North American standards, and even for Algonquin Park in the early 1960s when the population was at 5.8 per square kilometre.
These deer, spread out across 2,700 square kilometres in summer, migrate each winter into just a few hundred square kilometres outside the park. This is a large drainage area for a deer yard. Thirty-five years ago, when deer were more plentiful, other yards inside the park were used as well. Timing and completeness of the migration depend upon snowfall, temperature, and acorn abundance in various combinations.
While wolf predation is a significant cause of deer mortality, it is not limiting by itself. Deer make up only one-third of the wolf diet. We found that changes in deer numbers related primarily to human killing, and that has been important in preventing the deer population from increasing. Contributing to its lack of increase, as well, may be reduced cover due to the particular shelterwood system of pine and hemlock logging in the park.
Beaver is the unstudied component in the system, something we hope to rectify because they are important forest architects. We know from scat analysis that beaver is a seasonal food, most important when they are on land in early spring, late summer, and fall. We also know that beaver hair shows up surprisingly frequently in winter scats, and while wolves cannot dig into frozen beaver lodges, they always swing over to check them out. During mild spells, beaver emerge to restock their food supply. Sometimes they are vulnerable to wolves, travelling a hundred metres or more from a hole in the ice.
Moose are at a high density of about 0.5 per square kilometre. Partway through our study, the moose population fell when native people were allowed to hunt inside Algonquin Park for the first time since its establishment. However, after two years, the population recovered, possibly due to the counter-effect of logging that replaced extensive pine forest
s in the Bonnechere Valley, and to a lesser extent in the Petawawa drainage to the north, with plentiful red maple and poplar saplings, prime moose foods.
We found that wolves kill yearling or adult moose year-round, but scavenging is particularly common in late winter on animals that have died of the effects of winter tick. Moose calves make up a mere 15 per cent of the summer diet. Partly because of this low figure, and the scavenging, we conclude that wolf predation poses a relatively minor constraint on the moose population. More important are range conditions. Evidence of nutritional stress was shown by low twinning rates and, in some years, noticeably small yearlings in the population.
For wolves to be more significant in limiting the populations of either deer or moose, its population would need to be allowed to adjust upwards, rising to a level set by the availability of food. But the Algonquin Park wolf population is limited by heavy human killing and low recruitment.
In a natural system, forest conditions set ungulate densities, which set wolf densities, or wolf densities set ungulate densities, which influence forest conditions. The difference may relate to the frequency of large-scale events such as forest fires, ice storms, severe winters, or wildlife diseases. Both of these systems represent variants on how nature works, and should be allowed to operate in a park.
In or around Algonquin, however, human hunting influences the deer population, human killing influences the wolf population, and logging influences both. Behind the no-cut fringe of trees around lakes and portages is an ecosystem under human control.
Wolf Country Page 30
