Wolf Country

by John Theberge

  Occasionally, wolves have responded to prey that are concentrated and vulnerable by killing more than they need — surplus killing. They reportedly did this when encountering snow-disadvantaged caribou in northern Saskatchewan, and again among wolf-naïve elk that had no experience fending off predators after wolf reintroduction into Yellowstone National Park. Biologist Doug Smith explained that the Yellowstone killing in 1997 happened with the highest elk density of the century near the end of the most severe winter of the century. In both Saskatchewan and Yellowstone, surplus killing was transitory.

  We did not find surplus killing. While the kill rate was difficult to ascertain because the wolves were often on private property and moved so rapidly, we found that packs of four to six wolves appeared, on average, to kill a deer every four to six days. The wolves indicated that periodicity when, occasionally after killing and consuming a deer, they would go back to their territories in the park. We found that at times we could predict when they would return for another deer — unless they obtained food on territory.

  That frequency of predation is somewhat longer than Doug Pimlott’s estimated three days. Deer were much more plentiful thirty-five years ago, however, when he did his study, and average pack sizes slightly larger. Many carnivores, including wolves, respond to more prey by increasing their consumption and end up in better condition. Wisconsin biologist Lloyd Keith reviewed various wolf studies and described a three-fold difference in consumption per wolf depending upon the density of prey. Our lower consumption may indicate some difficulty in obtaining prey compared to thirty-five years ago, but not so great to represent a food shortage when translated into pounds of meat per day per wolf and compared to reported requirements.

  All but two of the deer carcasses we examined were completely consumed, either when we found them or before the pack left. The two exceptions were revisited and consumed a few days later. Typically, only hide and blood remained, or a thoroughly chewed skull, or a femur, or part of a vertebral column. Consumption took between eighteen and thirty-two hours, during which the wolves typically stayed close to their carcass, possibly concerned about losing it to other wolves. After that, they would wander around but normally not hunt immediately, because kills in close succession were so rare. Ravens and foxes shared the feast, but most meat ended up in wolf stomachs.

  Even though we found no surplus killing, food could still be abundant enough to turn off aggression. Surplus killing is aberrant behaviour not favoured by natural selection because it may result in a lack of food later. So we had to look for other evidence that the deer in the yard were easily available. We hoped it would come from our pool of descriptive events — anything that showed how well deer or wolves succeeded or failed in the prey-predator drama, especially how well or poorly the migration and yarding phenomenon served them both.

  Colin Fabian, living near Bonnechere, has been an observer of this deer migration for many years. His map, showing all the trails used by deer, looks like a stretched-out spider web with strands funnelling into a distorted centre. Once, Colin killed wolves, but he says they are too interesting. Now he prefers to track their movements as he does the deer. He has witnessed changes in deer distribution and movements over the years. Originally, the deer concentrated in a recently logged area just outside the park ten kilometres to the west, but they adjusted as browse declined. Also stimulating their move was the feeding of hay promoted by the MNR as part of a province-wide program to avoid winter deer starvation. The hay brought the deer right into the towns of Round Lake Centre and Bonnechere. People enjoyed watching them, like watching birds at a feeder.

  Years ago, the deer migration drained a much smaller area. Other deer yards were used throughout Algonquin Park. Then deer numbers crashed. As the population began recovering through the 1980s, the area draining into the Round Lake yard increased to cover most of the eastern side of the park. In 1991, the MNR stopped providing hay, a wise reversal of policy, and attempted through posters and meetings to explain why. Artificial feeding promotes local overbrowsing. As well, deer become dependant on the hay even though they have difficulty breaking it down in their digestive tracts.

  Each winter, two or three packs always position themselves on the main deer runways leading into the Round Lake yard. Other packs do the same for runways into the auxiliary “Germanicus” yard separated from the Round Lake yard by Round Lake itself. Their runway kills, together with kills made earlier in the migration, show that migratory deer are especially vulnerable to wolves. This conclusion supports a similar one based on limited evidence from radio-collared deer in Minnesota. The time wolves normally spend searching for prey is reduced by the concentration of deer in one place. It seems a stupid strategy for deer to set themselves up in this way, but they appear to be intent on reaching the yard and willing to run the wolf gauntlet to get there.

  Wolves undoubtedly remember from experience where to position themselves to intercept these converging deer, but, as well, have an ability to find them almost immediately. Whatever the deer do differently, the wolves quickly detect. In 1993 and again in 1997 when acorns were plentiful and the deer hung up longer in the hills, so did the wolves.

  To understand the risks of migration to deer better, we wanted to contrast their survival with those who stayed in the park. Small pockets of deer normally stayed in, especially when the migration was late. Sometimes we would track a pack to a deer carcass even when we could find no trace of any deer from air or ground.

  That happened one blustery day on Sec Lake. We were anxious to find this kill because the Sec Lake pack has never visited the deer yard. It has been the only pack not to, although its territory is only one pack-distance away. The wind-driven snow fast-froze our faces as we headed out on the lake by snowmobile. On our list of items to buy are plastic visors for our snowmobile helmets.

  Rounding a point, we came upon the partially drifted-over skeletal remains of a deer. The wolves had jumped it in the pines on the point and pursued it about two hundred metres from shore before bringing it down.

  On another excursion, the Hardwood Lake pack had attacked a moose on the ice of Lone Creek, leaving slush-frozen tracks and broken branches along the shore. The moose escaped, but only fifty metres away, surprisingly, was a dismembered deer.

  From these and similar experiences we concluded that the wolves selectively hunt deer wintering in the park over moose, making these deer especially vulnerable. These non-migratory deer may be inadvertently gambling on low snow all winter, or may have been caught by deep snow and forced to stay. Whatever the reason, their vulnerability may help perpetuate the migration tradition. Few of the deer that stay may survive to build up a sub-population of non-migratory deer.

  After the migration subsides, the deer gradually shift into the lowland, conifer-dominated core of the Round Lake yard. German student Doerte Poszig constructed computer maps based upon intensive tracking in 1996-97, and they illustrated this shift. The maps also showed an accompanying shift in most of the wolf locations. Over the winter, most deer were killed by wolves in areas with the highest deer concentration. Doerte used the information for a bachelor’s thesis at the University of Marburg.

  Her observations run counter to those of Steve Fritts and Dave Mech in Minnesota, who found more wolf-killed deer in areas of low deer density. Their findings had made us speculate about a “domino effect,” where deer at high density running away from wolves cause other deer to run, leaving wolves perpetually in a deer void. At times we did track wolves into dense pockets of deer that simply scattered, but wolves still were successful often enough.

  Occasionally we could read a snow story that explained the wolves’ success. Most commonly one or two wolves would run right behind a deer, while other wolves acted as flankers. Sometimes the deer would turn to avoid deadfall and encounter a flanker; other times the deer would pause at heavy brush, or a fence, and wolf tracks would converge there.

  The deer might have had a better chance of escape had t
hey stayed on their well-established trails, particularly when the snow was deep, but they didn’t. They simply bounded off anywhere in a panic. That observation contradicted one of the supposed benefits of yarding to deer.

  Other benefits speculated upon by various biologists were not apparent either, most notably the “dilution factor.” In other studies, wolf packs stayed on their summer territories, allowing deer to reduce the predator-prey ratio by migrating. The Algonquin wolves kept the ratio up by migrating too.

  Nor did individual deer benefit greatly from the anonymity of being part of a tightly formed group. Many species form flocks or herds because individuals try to obtain an internal position and by that means reduce their chances of being the ones selected for pursuit. The predator, whether a hawk flying into a flock of swallows or a grey whale swimming into a school of herring, benefits from a “sure” meal. Even though predation is made easier, often resulting in more prey being killed, that does not matter to the individual prey animal as long as the meal isn’t it.

  In comparison with these tight groupings, even in yards, deer are scattered mostly in small groups of twos or threes over many square kilometres — except where they are being artificially fed. Not only is anonymity limited, but so is any potential benefit from earlier detection of wolves.

  Yarding, like migration, does, however, reduce search time for wolves, setting them up for more frequent encounters with deer. Similar predator “set-ups” resulted in increased predation in central Labrador, where, with a previous graduate student, Kent Brown, we found caribou aggregated because of deep, soft snow. On Vancouver Island, wolves reportedly benefit where black-tailed deer are concentrated by a patchy pattern of logging.

  If yarding reduces search time and does not benefit deer greatly through the “domino effect,” or enhanced escape along established trails, or predator dilution, or earlier prey detection, or anonymity, how might it have evolved as a predator defence? The answer has to involve snow, because snow, not the presence of wolves, causes deer to yard. Deer do not yard on red wolf range in North Carolina and Tennessee. In Minnesota, deer were shown to concentrate less in winters of low snow. Yarding occurs only where snow is deep enough to make moving around energetically costly, with or without wolves.

  Yarding still could have evolved as an anti-predator strategy as well. There may be dual benefits for deer. You have to see it from the individual deer’s perspective. Being running-impaired in deep snow, and so stripped of its major defence against wolves, individual deer head for places where the snow is shallower, as it is in conifer-dominated deer yards. Whatever deer choose to do, detection may be inevitable, not only in the yard but even if they had stayed on territory — wolves on territory course their lands repeatedly. An ability to run may be the key to an individual deer’s survival.

  Nevertheless, more deer appear to end up in wolf stomachs because of yarding. Supporting that conclusion, in Minnesota, Mike Nelson and Dave Mech found that wolf predation was lower in the low snowfall winters when the deer stayed more dispersed.

  It seems contradictory that individual deer would benefit by yarding at the expense of the population. However, like all tight groupings of prey, decisions are made by individuals for their own welfare, and the population as a whole simply takes the consequences.

  We concluded that because we found no underutilized carcasses, deer were not so abundant and easy to obtain that the situation overcame the biological wolf taboo on overkilling. But, because of reduced search time on migration and in the yard, deer were sufficiently easy to catch to explain the high level of tolerance among packs. Possibly the high degree of genetic relatedness may have helped too, but knowing the genetic make-up of only part of the population, we could not fully work out its significance.

  We wondered if the Algonquin Park wolves were allowed to concentrate in the Round Lake deer yard because no resident wolves lived there. In the summers of 1995 and 1996, John Pisapio and various partners radio-collared twelve resident canids in four packs. To our surprise, these wolves varied considerably in size. Smallest were two sharp-muzzled, coyote-like, 14.5-kilogram (32-pound) animals, smaller than anything we had collared in the park. The largest were thirty and thirty-six kilograms (sixty-seven to eighty pounds) — typical Algonquin Park wolves.

  The southern portion of the deer yard was occupied by the Byers Creek pack, with a twenty-kilogram (forty-four-pound) lactating female. To the north near the village of Bonnechere was the Cybulski pack, and extending beyond the northern fringe of the yard, the Acorn Lake pack. The two coyotes ranged to the east.

  Possibly some of these animals were wolf-coyote hybrids, with even a mixture of genotypes within packs. Summer territories of the Byers Creek and Cybulski packs covered less than sixty square kilometres, more typical of coyotes than wolves. Also coyote-like were the howls of the Byers Creek female we recorded late one warm summer night along a concession road.

  With coyote-like animals present, we wondered if they would get out of the way when the migratory park wolves appeared. Canadian Wildlife Service biologist Lu Carbyn found that coyotes avoided wolves at Riding Mountain National Park in Manitoba; some that didn’t were killed. The same thing happened in Yellowstone National Park after wolves were reintroduced there.

  Starting the winter field work in 1995-96, we found the signal of one of the coyote-like animals coming from a pond beside the highway where someone had thrown the collar. This was a significant loss because much of that winter the migratory Travers pack was resident in the woods where it had lived.

  Over that winter, both the Byers Creek and Cybulski packs reduced their summer territories by one-third. The Cybulski and Acorn Lake packs were well positioned to take advantage of the migrating deer and, after that, benefited from pockets of deer that stayed in their territories the rest of the winter. The Byers Creek pack did not move to intercept the migrating deer, but because it was well situated in the core of the deer yard, it shifted within its territory to stay with the greatest concentration of deer.

  After the deer migration, the migratory wolf packs seemed to avoid the resident packs, except single collared wolves that moved everywhere just as they do in the park. Between the Byers Creek and Cybulski packs was a gap of about two kilometres, and there most of the fixes of the migratory packs fell. Use of this gap was even more striking the following winter when the Byers Creek pack consisted of only three wolves. They drew back from the northeastern segment of their territory next to the gap, and it was used like the gap itself by most of the migratory packs.

  Notably apparent in the maps is the more limited use of the core of the deer yard by the migratory packs in recent winters than in the previous winters of heavy wolf killing. In those earlier years, any resident packs may have been fractured or annihilated too. The obvious conclusion, in hindsight, is that in the earlier winters without these resident packs, the deer yard was not being defended, or was defended less, and so was open for more complete wolf migration. The presence of more migratory wolves back then, the result of people killing resident wolves, probably resulted in more dead deer.

  With the enhanced susceptibility of the yarded deer to wolf predation, we wanted to compare the importance of predation with other possible causes of deer death. Tracking wolves to more than a hundred deer carcasses over the years, we found only one deer that had died of malnutrition. Bone marrow for all but that one was typically fat-rich, even in late March, the only exception being a pinkish tinge in big bucks, expected because male deer suffer greater energy stress during the rut.

  The one starved deer was discovered by the Travers pack. The deer was frozen on its brisket just like the many moose we found over the years being scavenged by wolves. Its bone marrow was bright red. It had died in a large cedar swamp, a type of forest rare in the park but the preferred habitat for wintering deer in our study area. Cedar swamps are dark places under dense tree crowns that hold the calm, cold winter air as if it were delicate. The sound of just a coat sleeve b
rushing on a twig carries a long way; these are silent places too, with few birds. Every green cedar-twig was browsed to deer height.

  We came upon that starved deer in midwinter 1995-96, a year when the deer travelled to the yard early because of mid-November snows. The finding was a forewarning of possible starvation ahead. MNR biologist Dennis Voigt found that more deer died of malnutrition in a variety of Ontario deer yards when deer entered the yard early. However, shortly after we found the starved deer, a big thaw bared the fields, and the deer took immediate advantage of the exposed forage. We watched dozens of deer grazing out in the open. The vegetation had been fast-frozen in the fall so was likely highly nutritious. They made out fine after that.

  Reducing the effects of any overbrowsing at Round Lake, too, has been a shift in the yard. In 1996-97, the yard expanded southeast into an area unused before, and more deer went to Germanicus. The whole region has the same requisites of a good yard — low snowfall, abundant conifer cover, cedar swamps, and hardwood browse — so undoubtedly it could provide for many more deer.

  We concluded that these deer killed by wolves were not about to starve, so wolf predation was resulting in fewer deer. Roughly ninety fewer deer. This calculation was based upon an estimated five-day periodicity of deer kills per pack, multiplied by a winter average of six packs in the deer yard for an average of two-thirds of the time between mid-December and late March. To err on the high side, if wolf packs killed deer every four instead of five days, the estimate would be 108 deer.

  Inflate these two figures to 110 and 124 to include the scattering of deer killed in the park and that makes up around 6 to 8 per cent of the estimated park herd. An unknown additional number of local non-park deer augment their numbers and lower this percentage.

  During the rest of the year, predation apportioned among the three species of prey may account for another 10 or 12 per cent of the park deer. The combined total is a significant but not excessive kill compared to deer recruitment estimates in Ontario of about 35 per cent each year.