5
Wolf and Bear Management: Experiments and Evaluations

INTRODUCTION

In addition to many studies on the population ecology of wolves, bears, moose, and caribou, a number of experiments have been conducted in Alaska and elsewhere, in which wolf and/or bear numbers were reduced or their behavior changed, and responses of caribou and/or moose populations were monitored. These control activities have been targeted to specific areas that cover a relatively small part of the state. In most of these experiments, wolves were killed, but some used translocation or diversionary feeding. These experiments provide the best data with which to evaluate the biological basis of control as a management tool to increase ungulate numbers (Theberge and Gauthier 1985; Boutin 1992). In this section, the committee analyzes and evaluates these control experiments.

Although, the primary goal of wolf control and/or bear management in Alaska is to increase the availability of moose and caribou for human harvest, most management actions have not been monitored to directly assess whether, in fact, this goal was achieved. Instead wildlife managers have relied on less-expensive and short-term measurements, such as changes in birth rates (calf:cow ratios) or changes in adult population sizes, to assess the results of predator management experiments. The following cases are presented in order from the most to the least direct measurements of whether predator management resulted in increased human harvest of moose or caribou (table 5.1; figure 5.1).



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 90
Wolves, Bears, and their Prey in Alaska: Biological and Social Challenges in Wildlife Management 5 Wolf and Bear Management: Experiments and Evaluations INTRODUCTION In addition to many studies on the population ecology of wolves, bears, moose, and caribou, a number of experiments have been conducted in Alaska and elsewhere, in which wolf and/or bear numbers were reduced or their behavior changed, and responses of caribou and/or moose populations were monitored. These control activities have been targeted to specific areas that cover a relatively small part of the state. In most of these experiments, wolves were killed, but some used translocation or diversionary feeding. These experiments provide the best data with which to evaluate the biological basis of control as a management tool to increase ungulate numbers (Theberge and Gauthier 1985; Boutin 1992). In this section, the committee analyzes and evaluates these control experiments. Although, the primary goal of wolf control and/or bear management in Alaska is to increase the availability of moose and caribou for human harvest, most management actions have not been monitored to directly assess whether, in fact, this goal was achieved. Instead wildlife managers have relied on less-expensive and short-term measurements, such as changes in birth rates (calf:cow ratios) or changes in adult population sizes, to assess the results of predator management experiments. The following cases are presented in order from the most to the least direct measurements of whether predator management resulted in increased human harvest of moose or caribou (table 5.1; figure 5.1).

OCR for page 90
Wolves, Bears, and their Prey in Alaska: Biological and Social Challenges in Wildlife Management TABLE 5.1 Predator Reductions Discussed in Chapter 5   Duration (Years)     Method of Predator Reduction* Wolves Bears Prey Response Measured Air-assisted East-central AK (GMU 20A)a 7 Not done Calf survival, adult mortality Finlayson, Yukonb 6 Not done Calf:cow ratios, adult mortality, population densities, hunting success Southwest, Yukon c 5 5 Population densities, survival rates Aishihik, Yukon d 4 Not done Calf:cow ratios Northern BC e 10† Not done Calf:cow ratios; population densities Quebec f 4 3 Calf:cow ratios East-central AK (GMU 20E) g 3 Not done Calf:cow ratios; calf mortality South-central AK (GMU 13) h 3 ‡ 1 Calf:cow ratios Ground-based Kenai Peninsula, AK i 3 Not done Population densities Vancouver Island, BC j 4 Not done Hunting success East-central Saskatchewan k Not done 1.1 (different areas) Calf:cow ratios * See text for explanation of grouping by methods. Several experiments involved multiple periods of predator reduction. Primary sources: a, Boertje and others 1995; b, Farnell and Hayes 1992; Farnell and others, in preparation; Larsen and Ward 1995; c, Larson and others 1989a, 1989b, Hayes and others 1991; d, Hayes 1992, Yukon Fish and Wildlife Branch 1994, 1996: e, Bergerud 1990, Elliott 1986a, 1986b, 1989; f, Crête and Jolicoeur 1987; g, Gasaway and others 1992; h, Ballard 1991; i, Peterson and others 1984; j, Archibald and others 1991; k, Stewart and others 1985. † The areas in which wolf populations were reduced spanned a 10-year period, but was not done in the same place every year. ‡ A combined aerial shooting and poisoning program is also described under this case study. AIR-ASSISTED WOLF CONTROL East-central Alaska (Delta, GMU 20A) The best documented and most successful example of wolf control in Alaska was conducted from 1976 to 1982 in Game Management Unit 20A (GMU 20A), south of Fairbanks. The 17,000 km2 study area included 7300 km2 in the poorly drained lowlands of the Tanana Flats and 9700 km2 to the south in the foothills

OCR for page 90
Wolves, Bears, and their Prey in Alaska: Biological and Social Challenges in Wildlife Management FIGURE 5.1 Locations of case studies discussed in this chapter are indicated by gray dots. Predator control in the northern British Columbia case study was carried out over several distinct areas, as indicated by the 4 smaller dots. and mountains of the Alaska Range (Gasaway and others 1983; Boertje and others 1995; figure 5.2). Before wolf control was begun in 1976, the average age of the moose population was very low. Brown and black bear populations were judged to be low. Between 1965 and 1975, moose and the Delta caribou herd in this area were overharvested by humans (Gasaway and others 1983). A record snowfall in the winter of 1970–71 had caused substantial mortality of moose. Local trappers were taking approximately 20% of the wolves annually, but this harvest was not enough to reduce the population of wolves in the area. The caribou hunting season was closed in beginning in 1973. From 1976 until 1982, a 7-year air-assisted wolf control program was conducted by both ADFG staff and private hunters. Each year during this period, the wolf population was reduced to 55–80% below pre-control numbers. Regular harvest of wolves by private trappers continued throughout the period of wolf control. In 1976 there was an estimated 14 wolves per 1000 km2, in 1982 at the end of the control period there was an estimated 8 wolves per 1000 km2. During the 7 years of wolf control, survival of moose calves and yearlings increased and mortality of adults, especially middle-aged and old adults, declined. The moose population increased from 183 to 481 per 1000 km2, a mean

OCR for page 90
Wolves, Bears, and their Prey in Alaska: Biological and Social Challenges in Wildlife Management FIGURE 5.2 Study area in interior Alaska (GMU 20A) where wolves were controlled during 7 winters, 1975–76 through 1981–82. Wolves were controlled in a 10,000 km2 portion of the 17,000 km2 study area during winters 1993–94 and 1994–95. annual rate of increase of 15%. After the wolf control program was terminated, the moose population continued to increase for 12 more years, reaching 1020 moose per 1000 km2 by 1994, a mean annual increase of 5%. During the period of wolf control (1976–1982), calf survival in the Delta caribou herd increased and adult mortality declined, contributing to an average annual rate of 16% which then continued to increase at 6% for 7 years after wolf control ended (1983–1989). The peak density was 890 caribou per 1000 km 2. The subsequent decline in caribou in the early 1990s coincided with several severe winters that ended the previous 20 years of mild winters. Two adjacent, low-density caribou herds also declined during this period. By 1985, 3 years after the control programs were ended, wolf numbers had recovered to near precontrol levels in most of the area. After ADFG terminated this wolf control program, private hunters continued to take up to 25% of the autumn wolf population each year. This level of harvest presumably had some impact on the wolf population, even though census data did not reveal a decline in the number of wolves (Peterson and others 1984; Gasaway and others 1992). Human harvest of moose during the 20-year period was restricted to maintain at

OCR for page 90
Wolves, Bears, and their Prey in Alaska: Biological and Social Challenges in Wildlife Management least 30 males for every 100 females. Board of Game regulations prohibited the harvesting of female moose. Caribou harvest was kept to loss than 6% during the 20-year period, except for 1983–1986 when it was 11%. With the decline of caribou populations in the early 1990s, legal human harvest of caribou was reduced and then eliminated. Documentation of the responses of wolves, moose, and caribou in GMU 20A was based on intensive sampling by ADFG personnel and analysis of information from hunter moose harvest reports. Data for wolves were based on aerial surveys; information from local trappers, hunters and pilots, and from tracking radio-collared wolves (Gasaway and others 1992). Estimates of moose abundance were based on data from aerial surveys (Gasaway and others 1992) and estimates of harvest rates. Birth rates of radio-collared female moose and caribou were compared to estimates of birth rates in other surveys. Ten-month-old calves were weighed after they had received immobilizing doses of anesthetics. Estimates of caribou abundance were based on aerial photographs, total aerial searches, and radio-search techniques. Juveniles of known ages were weighed. Data were also collected from areas where wolf control was not conducted, but they are difficult to interpret. For instance, calf:cow ratios were high for migratory moose that calved in the Tanana Flats (with wolf control), and wintered in the Chena and Salcha areas (no wolf control). But in the Denali, Fortymile, or Tok areas, where wolves were not controlled, calf:cow ratios of moose did not increase in the late 1970s. During this same period, caribou herds in Macomb and Denali were stable. During and after the 7-year air-assisted wolf control program from 1976 to 1982, moose and caribou populations increased. Further, during the 7 years of wolf control human harvests of moose and caribou were curtailed, and it is difficult to separate the combined effects of wolf control and reduced human harvests. Nevertheless, the positive responses of both moose and caribou lasted longer than in any other control program and no such responses were seen in the untreated areas. In the winters of 1989–90, 1990–91, and 1992–93, snow reached critical depths and yearling moose survival was low. The growing season in 1992 was particularly short, and no twin moose were reported the following spring. Except for 1992, caribou birth and recruitment were reduced in GMU 20A and in the comparison Denali and Macomb herds. The Delta caribou herd declined from 1989–93 at an annual average rate of 0.78. In contrast, the moose population apparently increased between early winter 1988 and 1994. During the winters of 1993–94 and 1994–95, a ground-based wolf control program was conducted in the Tanana Flats to determine whether wolf control would reverse the decline of the Delta caribou herd and allow their numbers to increase again. Before control there were 262 wolves in the area. this was reduced by 62% in 1993–94 and by 56% in 1994–1995. The Delta caribou herd appeared to increase in numbers after the winter of 1993–94. However, 2 years of

OCR for page 90
Wolves, Bears, and their Prey in Alaska: Biological and Social Challenges in Wildlife Management ground-based wolf control in a 600 km2 area in the early 1990s apparently had no effect on caribou calf:cow ratios. Finlayson, Yukon Territory The 14-year Finlayson wolf reduction experiment included 6 years of wolf reductions and 8 years when populations were monitored in the absence of reductions. It is a very important experiment because although it was not conducted with an appropriate experimental design, it provides the best available data to test the two equilibrium or ''predator pit" hypothesis. In addition, wolves were reduced to lower levels for more years than most other experiments and the population ecology and behavior of wolves was studied in detail after control ceased. The Finlayson Caribou Herd (FCH) lives on a 23,000 km2 area in the south-central Yukon Territory (figure 5.3). FCH wintered in a complex of hills and small plateaus separated by a several broad, glaciated valleys. The area is typical northern boreal forest with black and white spruce, lodgepole pine, and occasional stands of aspen. The large mammal system includes caribou, moose, 200–300 mountain goats, less than 100 Dall's sheep, a few mule deer, wolves, brown bears, black bears, coyotes, wolverine, lynx, and red fox. Beaver and Arctic ground squirrels were common and the cyclic snowshoe hare populations declined in 1983 and again in 1992. An all-season highway that bisects the FCH winter range facilitates hunting, particularly during the winter when the caribou distribution straddles the road. Another road provides summer and fall access for hunters to the northern portion of the area. Conditions Before Wolf Reduction In 1982 the caribou population was crudely estimated to be between 2,000 and 2,500 animals. In October 1982, there were 17 calves/100 cows, a ratio thought to be insufficient to support the estimated harvest of 250 adults per year. In October 1983, a ratio of 34 calves/100 cows was recorded, giving a pre-treatment average of 25.5 calves/100 cows for those two years. In 1982 and 1983, the annual mortality rate of radio-collared adult caribou was 28% (n=5/18). In the winter of 1983, an estimated 215 wolves lived in the study area (9.3/1000 km2) with an average pack size of 9.6 wolves. Wolf Reduction Wolves were systematically reduced by aerial shooting in late winter, with an effort to remove entire packs. In the winter of 1983, 49% of the 215 wolves that lived in the study area were shot. Over the next 5 winters, from 80 to 85% of the wolves were shot each year. Of the 454 wolves removed, 77% were shot from

OCR for page 90
Wolves, Bears, and their Prey in Alaska: Biological and Social Challenges in Wildlife Management FIGURE 5.3 Location of the Finlayson study area in the Yukon, Canada.

OCR for page 90
Wolves, Bears, and their Prey in Alaska: Biological and Social Challenges in Wildlife Management a helicopter. In addition, hunter harvest of caribou was reduced from about 250 per year prior to 1983 to about 25 thereafter. Response When Wolf Numbers Kept Low In 1986 there were estimated to be 3073 ± 333 (90% confidence limits) caribou. In 1990, the year following the end of 6 years of wolf control, there were 5950 ± 1055 (90% confidence limits) caribou, an average rate of increase of 18% per year. Mortality rate of radio-collared adult caribou, which was 24% in 1983 after a 49% wolf reduction, decreased further to an average of 11% between 1984 and 1987. The pre-treatment average of 25.5 calves/100 cows almost doubled to 50.2 (42, 50, 45, 55, 47, 62) calves/100 cows during the years of wolf reduction. Two neighboring untreated caribou herds had between 15 and 28 calves/100 cows in 1985 and 1986 when the FCH had 45 and 50 calves/100 cows. Moose were not the focal prey species of this project and few pre-treatment data are available. Censuses in 2 portions of the treatment area, North Canol (NC) and Frances Lake (FL), were conducted in 1987 during the fourth year of the wolf control program. At that time the moose population was estimated to be 516 in NC and 741 in FL, and there were 67 and 66 calves/100 cows respectively. In 1991, the number of moose had increased to 950 in NC and 1475 in FL for rates of increase of 16% and 18% per year. The number of days it took a hunter to kill a moose decreased from 26 in 1979–1984 to 18 in 1985–1991 in NC. In FL, the number of hunter days required to kill a moose decreased from 31 to 23. Response When Wolf Control Stopped Wolves increased from 29 survivors at the end of wolf control (March 1989) to 240 in March 1994, and 260 in 1996, 17% higher than the pre-control level of 215 wolves. The number of packs increased from 7 in 1989 to between 26 and 28 after 1991. Mean number of wolves per pack increased from 4.4 in 1990 to 7.8 in 1994 and 9.3 in 1996; almost identical to the pre-treatment pack size average of 9.6 wolves per pack. Detailed monitoring of wolf predation rates on moose demonstrated that wolves in small packs had greater kill rates per wolf than wolves in large packs. In addition, wolves were efficient predators when moose were at low densities. Kill rate was density-independent when there were between 250 and 430 moose/1000 km2 for all sizes of wolf packs. This result is important because it demonstrate that in Finlayson study area, when moose densities are low (less than 250 moose/1000 km2 wolves can maintain the population at low densities (estimated to be at 120 moose/1000 km2). Whether or not there is also a high equilibrium remains unknown, but the efficiency of wolf predation suggests it is unlikely for the interior of the Yukon. After 1989 and the end of wolf control, caribou calves/100 cows gradually

OCR for page 90
Wolves, Bears, and their Prey in Alaska: Biological and Social Challenges in Wildlife Management declined from an average of 50.2 when wolves were low to about 44, 32, 20, and 30 in the following 4 years; about the same as in adjacent herds where there was no wolf control. The caribou population decreased from about 5950 ± 1055 to 4550 ± 550 for an average rate of decrease of 5% per year. A moose census of both NC and FL census areas in 1996 found that the number of calves/100 cows had decreased to 29 and 30 from the 1991 ratios of 52 and 44 respectively. The estimated number of moose in NC had decreased from 1001 in 1991 to 810 in 1996, an average annual rate of decrease of 4%. Similarly, the estimated number of moose in FL decreased from 1475 in 1991 to 1320, an average rate of decrease of 2%. To summarize, although hunting was reduced during wolf control in Finlayson, the responses of both moose and caribou were clear. The removal of approximately 85% of the wolf population for 6 years and reducing the harvest rate appeared to result in an increase in adult caribou survival, an increase in calves/100 cows of both caribou and moose, and increases of both caribou and moose populations at a rate of approximately 18% per year. Once wolf control ceased, the wolves rapidly increased and pack size recovered in 4 years. Wolf predation rate on moose was independent of moose density, suggesting that a low moose/wolf equilibrium existed. Seven years after wolf control stopped and 3 years after wolves recovered, moose and caribou numbers appeared to be decreasing. This decrease suggests either that a higher equilibrium does not exist in the area, or, if it does exist, it is at a higher density than the ungulates reached during the wolf control years. However, because there was a great increase in mineral exploration in the area in the 1990s, human activity may have prevented moose and caribou from attaining even higher population densities. Southwest Yukon In the early 1980s hunter demand for moose in southwestern Yukon exceeded the supply; moose populations either declined or remained stable at low numbers. To assess the effects of brown bear and wolf predation on limiting moose populations, the causes and rates of moose mortality were documented between 1983 and 1987. The effects of hunting, weather, moose reproduction, forage availability, and emigration on moose numbers were also evaluated. The study included 3 experimental wolf reduction areas: Haines Junction (4890 km2), Rose Lake (6310 km2) and Lorne Mountain (1020 km2), and 2 areas in which wolves were not manipulated: Teslin (2580 km2) and the Auriol Range (1190 km2) (figure 5.4). Survival rates and causes of mortality of adult moose and calves were compared among experimental and control areas. Conditions Before Wolf Reduction Before wolf reduction, moose populations in two of the three experimental

OCR for page 90
Wolves, Bears, and their Prey in Alaska: Biological and Social Challenges in Wildlife Management FIGURE 5.4 Location of the wolf control study area in the southwest Yukon, Canada. areas (Haines Junction and Lorne Mountain) appeared to be declining, but were stable or slightly increasing in the third experimental area (Rose Lake). The moose population in one unmanipulated area appeared to be gradually decreasing while in the other unmanipulated area the moose population was stable or slightly increasing. Calf survival varied greatly between the two years before wolf reduction: In autumn 1980 and 1981, respectively, there were 40 and 11 calves/100 cows in Haines Junction, 20 and 26 in Rose Lake, and 37 and 9 in Lorne Mountain. The overall average in the two years before wolf reduction was 23.8 calves/100 cows. In the unmanipulated areas there was an average of 22 calves/100 cows before treatment. Wolf densities in the 3 experimental areas averaged 12.5 wolves per 1000 km2 before removal and averaged 15 wolves per 1000 km2 in the areas where wolves were not removed.

OCR for page 90
Wolves, Bears, and their Prey in Alaska: Biological and Social Challenges in Wildlife Management Wolf and Bear Reduction Wolf populations in the experimental areas were reduced over a 5-year period through aerial hunting between the winters of 1982–83 and 1986–87 but the level of reduction varied among years and treatment areas. Reduction of the wolf population by more than 40% was thought to be sufficient to influence moose populations, but this level of reduction was attained for only 3 years in Haines Junction and Lorne Mountain and for 5 years in Rose Lake. Population reductions of 80% were attained for 2 years at Rose Lake and for one year at Lorne Mountain. Liberalized brown bear hunting regulations were used to reduce the number of brown bears in the same experimental areas. The average number of bears removed each year increased from 3 to 6 in Haines Junction and from 2 to 8 in Rose Lake. No bears were removed from Lorne Mountain. The bear population was estimated to be reduced by 7–9% but bears are difficult to census, and this estimate may not be correct. Response to Wolf and Bear Reduction Because the size of the moose population was not monitored in either Haines Junction or Lorne Mountain after predator removal, the effects of wolf and bear reduction on moose numbers could be assessed only in the Rose Lake experimental area. The moose population in Rose Lake did not increase significantly after 5 years of reducing the wolf population by more than 66% and 4 years of liberalized bear hunting regulations (607 ± 109 in 1981 and 582 ± 163 in 1982 compared to 717 ± 143 in 1986). In the 3 treatment areas, the average number of calves/100 cows was 23.8 before predator reduction and 22 during experimental reductions. There was no difference in mean annual female moose survival rate in areas and years with wolf reduction (92%) and areas and times without wolf reduction (88%). However, 8 of the 16 female moose that died of known causes were killed by wolves, 4 by brown bears, 2 by either bears or wolves, and 2 from an unknown predator. Eighty percent of 132 collared moose calves died in their first year, and brown bears killed 58% and wolves 27% of these. The percent of calves that survived to 6 months of age was significantly greater in areas with wolf reductions (31%) than in areas without (21%). Multiple regression analysis also indicated that calf survival was greater when wolf population sizes were low. When wolf reduction years were omitted from this analysis, mean calf survival was significantly higher when maximum snow depth was less than 80 cm (22%) compared to when it was deeper (11%). Thus, although moose populations increased slightly, reducing wolf numbers by more than 60% for 5 years and liberalizing brown bear harvest in southwest Yukon failed to produce a substantial increase in moose numbers. Wolves were responsible for half of the adult female moose mortalities and 27% of the

OCR for page 90
Wolves, Bears, and their Prey in Alaska: Biological and Social Challenges in Wildlife Management outcome of interactions between large mammalian predators and their prey is an important wildlife management issue. Any intentional perturbation is an experiment, but the reliability of inferences about whether the perturbation caused the observed effect depends on the design of the study. During wolf control and bear management, causal relationships have been assumed to be already known, that is, it was assumed that predators were controlling their prey. If the objective had been to test whether predators were really controlling prey, the experiments might have been planned differently. An example of a proper design is the experiment on selected British islands to test whether populations of two species of game birds were limited by the small mammals (voles) that preyed on their nests (Marcstrom and others 1988). For five years in a row predators were removed from one area but not from a second, non-removal area. For the next four years, the removal and non-removal areas were reversed. The experiment showed that predators were, indeed, limiting game bird populations on these islands. The conclusion is strong because variation in time was controlled by doing the experiment simultaneously in two places and variation between areas was controlled by using each area for its own comparison through time. Replication, that is repetition of this design on more islands, would have yielded data on the generality of the results but would not have strengthened the conclusion. Using the notation of Campbell and Stanley (1966), this experimental design can be represented as: population 1 X0 X0 X0 X0 X0 0 0 0 0 population 2 0 0 0 0 0 X0 X0 X0 X0   time →   where X represents the annual removal of predators, and 0 represents observation of the game bird population. The rows represent sets of observations through time and the columns represent simultaneous observations at a time period. It is a design leading to fairly strong causal inference because nearly all possibly confounding effects are accounted for. If the game birds had been randomly placed on different islands the design would have been a true experimental design. As it is, it has to be called a quasi-experimental design (James and McCulloch 1995). Most field experiments are in fact quasi-experiments in this sense. Using this notation, the design of the experiment in wolf control and bear management for the Delta caribou herd in Game Management Unit 20A is:

OCR for page 90
Wolves, Bears, and their Prey in Alaska: Biological and Social Challenges in Wildlife Management Population 1 0 0 0 0 X0 X0 X0 X0 0 0 0 0 0 Population 2 0             0           Population 3 0       0       0         Population 4 0             0           Population 5 0       0       0           time →   This is a multiple time series design in which the treatment (wolf removal) was applied for four consecutive years and the caribou population (0s) was monitored for a series of years both before and after the treatment in the treatment area, as well as sporadically in the four nontreatment areas. With more observations in the nontreatment areas this would be a multiple time series design, another quasi-experimental design. It allows weaker causal inference than the above design of Campbell and Stanley's study of predation on game birds, because initial differences among sites are possibly confounding. In the Delta caribou herd experiment, the results suggest that wolf removal did result in an increase in the caribou herd. The design for most of the other experiments in wolf control and bear management are either one time case studies X 0, or one group pretest-posttest designs 0 X 0, or static group comparisons X 0 0, X 0, all of which are inadequate for causal inference, because even though the management effort may have caused the result, the design of the experiment was inadequate for causal inference. In the first case, there was no nontreatment group and no comparison. In the second case, one group was studied before and after the treatment. This design uses the group for making the comparison, but there may have been other causes that occurred at the same time as X. In the third case there is no assurance that the initial groups did not differ. The danger is that of falsely concluding a cause and effect relationship. None of the above designs incorporate more than one species of predator or prey. If predators switch from one prey to another, these simple designs may be able to test the hypothesis of the effects of control but modeling will be required to characterize the system satisfactorily.

OCR for page 90
Wolves, Bears, and their Prey in Alaska: Biological and Social Challenges in Wildlife Management To understand the degree to which predation by both wolves and brown bears limit a prey population, an experimental design like the following would be needed: population 1 0 0 0 X 0 0 0 population 2 0 0 0 Y 0 0 0 population 3 0 0 0 XY 0 0 0 population 4 0 0 0 0 0 0 0   time →   where X represents wolf removal, Y represents bear removal, and 0 represents observations of the prey population. The GMU 20A experiment does not reveal whether weather initiates population declines in general because it seems to have done just that in the Fortymile caribou herd as wolves switched from eating moose to eating caribou, and adult cows did not breed in some years with severe winters (Boertje and Gardner 1996). In the Porcupine and Denali herds calf mortality is apparently nutrition related (Whitten and others 1992; Adams and others 1995, respectively). All these ideas could be checked by studies designed by the criteria above. In its analyses of the Delta herd (appendix C), the committee found that although calf survival was correlated with annual snowfall, incorporation of snowfall into the model did not improve its explanatory power. Modeling the dynamics of population regulation to estimate what factors are behaving in a density dependent way involves assuming a framework for how things are happening and then estimating the parameters. When based on a carefully designed and controlled experiment, one can assess cause and effect relationships. The same kind of modeling based on purely observational studies cannot distinguish between factors that regulate population densities from those that are merely correlated with them. Prey: predator ratios (Gasaway and others 1983; Keith 1983) are more readily obtained than are full-scale area-specific studies of density, but they are subject to several major confounding variables. If prey populations are low, predators might eat more of the carcass and the prey: predator ratio will become lower. Hayes (1995) observed that wolf packs tend to kill prey at similar rates, regardless of pack size—suggesting that more of the prey carcass may be eaten when wolf packs are large. Preying on alternative prey species and prey switching may affect the results. If prey populations are close to their nutrient/climate ceiling, predation may be primarily compensatory rather than additive (Theberge 1990). Even calf:cow ratios can be misleading because they are affected by changes in mortality rates of both calves and cows. A better alternative to using ratios as response variables is to monitor the density of the population by age class and sex.

OCR for page 90
Wolves, Bears, and their Prey in Alaska: Biological and Social Challenges in Wildlife Management MAKING A DECISION TO INITIATE A PREDATOR CONTROL ACTION The committee's evaluation of the predator control experiments suggest that predator control is likely to achieve its intended goals only if a decision to initiate control is made after a thorough evaluation of the existing situation. The questions to be asked are 1) Under what circumstances is predator control justified? 2) Are there other management options than predator control available that will bring about the desired results? 3) Has an adequate analysis been made of the biologically, economically, and sociologically relevant information? and 4) Is it possible to design a control program that will include the ability to monitor the outcome and assess the degree of success? Even if killing predators can be reasonably expected to cause an increase in ungulate numbers, it is worthwhile to consider if there are alternative, nonlethal means available to reduce the impact of predation. Under certain specific circumstances, they may be more cost-effective, and commonly, they are more socially acceptable (Boertje and others 1995). Perhaps habitat can be manipulated (burned, logged, or otherwise set back successionally) to improve its structure, quality, and distribution. A fire policy integrated with other agencies may be critical for maintaining optimal long-term habitat that will support relatively higher ungulate numbers. Also, the distribution of quality habitat, not just its quantity, may have significant positive effects on ungulate populations. Translocation of wolves or bears is frequently only a temporary solution unless individuals are translocated to sites far enough from the study area so they do not return. However, it has been shown to be temporarily effective in some studies, and it may be applicable in special circumstances—such as in the Kenai Peninsula, where development in the Anchorage area creates a partial barrier to migration. Sterilization of wolves to offset their high reproductive rate is currently under investigation as a effective technique to keep wolf numbers at lower levels. Its effectiveness may be compromised by the tendency of wolf packs to kill at similar rates, regardless of pack size (Hayes 1995), and by the frequency with which unrelated wolves join packs, particularly in Alaska (for example, Meier and others 1995). Diversionary feeding is likely to be more effective for bears than for wolves because the period of bear predation on ungulates is restricted to a much shorter time period. Saving calves from wolves during June is of little value if they kill calves during the remainder of the year. However, diversionary feeding of bears in combination with wolf reductions may be an effective strategy. A decision to reduce predator numbers dramatically over a relatively short period of time will have consequences for the predators themselves. The time it takes for a species to recover to pre-control numbers or more depends not only on their inherent rates of reproduction, but also the magnitude of reduction of human-caused mortality. Wolves breed much more rapidly than bears, and thus

OCR for page 90
Wolves, Bears, and their Prey in Alaska: Biological and Social Challenges in Wildlife Management their potential rate of increase is higher, and they can withstand higher mortality rates. Also, the proximity of other wolf or bear populations that might provide immigrants will affect the rates of local population increase following control. In comparison to bears, wolves can more easily repopulate an area through immigration, because both males and females disperse from natal ranges. Consequently, an assessment of the geographic scale of removal efforts is important not only with regard to ungulate populations of concern, but also to the effectiveness with which the predator population can be held at necessary low levels. Finally, wolf control may provide more food per capita for bears, but the numerical response of bears to changes in food abundance are also much lower than for wolves. Thus bears are more susceptible to control efforts, and will take much longer to rebuild their population. An additional but certainly important concern is what the larger, unintended ecological consequences of such control will be. Even in a qualitative sense, the potential changes on other plant and animals species and systems must be considered. If ungulate numbers increase, competition with other herbivores will likely increase, and perhaps the numbers of those competitors, such as beavers, will go down. With reduced wolf and/or bear numbers, predation on alternate or less common prey species will decrease. One can imagine that in some areas, sheep numbers might increase to or above their carrying capacity, with a subsequent deterioration in habitat and a major population decline. Although this is a ''what if …?" kind of assessment, it points out that potential unintended consequences can ultimately have large repercussions, biologically and politically. It is clear from the experiments already done that killing a substantial proportion of the population in the target area is essential if wolf control is implemented to increase ungulate numbers. A wolf population probably should be reduced so that it is no greater than 40% of its original numbers, and then maintained there for at least 3, and probably 5 years, depending on the response of the ungulate populations. Bears, on the other hand, may need to be removed at lower rates for shorter periods of time to achieve notable results. However, since wolves have a greater per capita effect on ungulates than do bears, and bear predation behavior may be more unpredictable, results of any bear control efforts have less chance of being reliably predicted. Another concern is which individuals in the population are going to be killed. Usually, wolves are killed by a combination of private and government efforts from the air and/or ground that targets whole packs. Focusing on specific geographic areas increases the chances that most or all of identified packs might be removed by ground-based efforts, and aerial hunting particularly if one or more pack members have been radio-marked, increases the chances of removing an entire pack (e.g., Hayes and others 1991). This is important because the removal of 60% of the members of each pack, for example, will probably have less of an effect on ungulate population change than will removal of 60% of the

OCR for page 90
Wolves, Bears, and their Prey in Alaska: Biological and Social Challenges in Wildlife Management packs (for example, Walters and others 1981); this is particularly likely if experienced individuals are the ones left in partially affected packs. Logistical constraints likely limit the opportunities to remove entire packs, and good data comparing the consequences of different removal regimes need to be collected to identify costs and benefits of each. Seasonal changes in predator-prey ecology are another area of concern. Reducing pack size in late autumn has little effect on winter predation rate (Hayes et al. 1991, Hayes 1995), but it should have a greater effect in summer when pack cohesion disintegrates during denning, subadults roam alone or in small groups, and many more moose or caribou calves are available than in winter. Reducing the number of wolves in summer reduces the number of independent hunting units, which in turn should reduce the predation rate of the whole population. Predation will be little changed by reducing the number of wolves in winter because the pack acts as a single unit, and kill rates of various size packs are similar (Hayes 1995). The economic and public relations cost of predator control is high, and consequently, detailed planning is necessary to ensure a successful program. A cost effective decision process involves a series of steps beginning with the least expensive and progressing to those that are more costly. Guidelines for Decision-making The following is a set of guidelines, which, if followed, should increase the probability of success from implementing a management option. The committee emphasizes, however, that even if proper procedures are followed, success cannot be guaranteed, and presenting guidelines does not constitute endorsement by the committee of specific control or management activities. The first step in deciding whether or not to reduce predators is to identify the reasons for wanting more ungulates because not all reasons are equally important. Some reasons for wanting to increase prey numbers include: I. Reasons for increasing ungulate populations. A. Biological emergency. It has been suggested that predator control be considered if there is a "biological emergency"; that is, when a local ungulate population is at risk of extirpation. Local extirpations, and recolonizations have likely occurred frequently in the past, but recolonization may be less likely today in areas where people have substantially altered habitats. If it is determined that a biological emergency exists, the management agency should proceed to III. B. Subsistence emergency. In portions of Alaska where indigenous people rely largely on ungulates, extirpation or greatly reduced populations of moose and caribou would cause significant hardship. In the past, indigenous peoples moved during periods of scarcity to where game were more abundant, but human

OCR for page 90
Wolves, Bears, and their Prey in Alaska: Biological and Social Challenges in Wildlife Management migrations are difficult or impossible in most areas today. If predator management to maintain a satisfactory abundance of ungulates to support an indigenous subsistence lifestyle is a high priority, proceed to II. C. Lifestyle and recreational demands. Abundant ungulate numbers are desired by people whose lifestyle includes recreational hunting. If game numbers and resulting hunting success is low, other protein source are available. Predator reduction to support higher ungulate numbers to satisfy recreational demands is currently a lower priority than for biological or subsistence emergencies. D. Viewing and tourism demands. Abundant ungulate numbers are of value to tourists and the tourist industry. Predator reductions to stimulate an abundance of ungulates to increase tourist satisfaction is also currently a lower priority than for biological or subsistence emergencies. II. Quantifying demand and investigative modeling. Once a compelling reason for predator reduction has been identified, the demand for an increase in ungulate populations should be quantified. Combining the use of questionnaires, personal interviews, historic hunter success records, and comparisons among areas should quantify the unmet demand and indicate the degree that ungulate numbers must increase to meet the demand. Simple population projection models and benefit/cost models should be used to determine the level and duration of predator removal that may be necessary to increase ungulate numbers to where the demand would be met, and to estimate the costs of predator removal. If, after quantifying the demand and investigating the costs of control, predator reduction might be effective, then the agency should proceed to section III. III. Ecological investigations. Ecological investigations are expensive but are needed to assess the likelihood that a predator management activity would achieve its desired goals. The needed data include: A. Historic trends. Are there fewer ungulates now than before, how reliable are the data, and what has changed? Could changes be explained by weather patterns, habitat changes, or human harvest? B. Current trend. Is the population currently decreasing? This is an expensive but important step that is best estimated by monitoring adult survival using telemetry and/or a precise series of censuses. C. Emigration. Do adjacent areas have increased ungulate populations that indicate possible emigration from the area of interest? D. Habitat condition. Are the prey at or close to current environmental carrying capacity? Combining winter severity indices with body condition indexes such as body weight, body fat content, growth rates, pregnancy rates over a series of winters should indicate if habitat is limiting. Direct habitat monitoring should also be used, recognizing that relationships between food and cover quantity, quality, and distribution are complex.

OCR for page 90
Wolves, Bears, and their Prey in Alaska: Biological and Social Challenges in Wildlife Management E. Predator ecology. What are the densities of wolves and bears in the area and what are their major seasonal foods? Which predator species might be depressing ungulate numbers, and how difficult and expensive would reduction programs for each species be in the particular area? F. Limiting factors. An expensive yet critical step is determining the relative importance of the various limiting factors. G. Ecological consequences. Would predator control cause other negative ecological consequences, such as disrupting scavenger populations or other predator species who might depend on prey killed by wolves or bears? IV. Management Options. If the availability of prey is significantly less than the demand, the following options should be investigated. A. Habitat manipulation. Is it feasible to increase ungulate reproduction or decrease predation rates by improving the quantity, quality, or distribution of habitats? B. Nonlethal methods. What is the potential for diversionary feeding, sterilization, and translocation? C. Selective removal. Would selectively removing individual bears or individual wolf packs be effective? D. Timing of removal. Are there certain times of the year when removal of predators would be most effective? E. Removal methods. Assess removal methods to determine how recreational and economic benefits (hunting and trapping) might be realized while encouraging the most humane and efficient methods that are politically acceptable. F. Removal Locations. Can predator control be concentrated in the most critical areas to maximize effectiveness while minimizing effects on the predator population? V. Monitoring Predator Reductions. Most past predator management programs have resulted in unclear results. Control actions have sometimes been of insufficient magnitude, duration, or geographic extent to show clear results. Additionally, pre and post-treatment monitoring have sometimes been insufficient, non-experimental areas have not been maintained, and climatic conditions have often been poorly measured. Wherever possible, predator control programs should be incorporated into a reviewed experimental design to ensure that knowledge is one of the benefits of the reduction program.

OCR for page 90
Wolves, Bears, and their Prey in Alaska: Biological and Social Challenges in Wildlife Management REFERENCES Adams LG, BW Dale, and LD Mech. 1995. Wolf predation on caribou calves in Denali National Park, Alaska. In LN Carbyn, SH Fritts, and DR Seip, Eds. Ecology and conservation of wolves in a changing world. Canadian Circumpolar Inst, Univ Alberta, Edmonton, 1995. Archibald WR, D Janz, and K Atkinson. 1991. Wolf control: a management dilemma . Transactions of the North American Wildlife and Natural Resources Conference 56:497–511. Ballard WB, TH Spraker, and KP Taylor. 1981. Causes of moose calf mortality in southcentral Alaska. J Wildl Manage 45:335–342. Ballard WB. 1991. Management of predators and their prey: the Alaskan experience. Transactions of the North American Wildlife and Natural Resources Conference 56:527–538. Ballard WB and SD Miller. 1990. Effects of reducing brown bear density on moose calf survival in south-central Alaska. Alces 26:9–13. Beaulieu R. 1984. Moose calf mortality study. Saskatchewan Parks and Renewable Resources, Wildlife Population Management Information Base, 84-WPM-8. 5 Pp. Bergerud AT and JP Elliott. 1986. Dynamics of caribou and wolves in northern British Columbia. Can J Zool 64:1515–1529. Bergerud AT and WB Ballard. 1988. Wolf predation on caribou: the Nelchina herd case history, a different interpretation. J Wildl Manage 52:344–357. Bergerud AT and WB Ballard. 1989. Wolf predation on Nelchina caribou: a reply. J Wildl Manage 42:344–357. Bergerud AT. 1990. A review of the dynamics of ungulates and wolves in Northeastern British Columbia 1976–1990. Draft Copy. Bergerud and Associates, Fulford Harbour, BC. Boertje RD, WC Gasaway, DV Grangaard, and DG Kellyhouse. 1988. Predation on moose and caribou by radio-collared grizzly bears in east central Alaska. Can J Zoology. 66:2492–2499. Boertje RD, DG Kellyhouse, and RD Hayes. 1995. Methods for reducing natural predation on moose in Alaska and Yukon: an evaluation. Pp. 505–513 in LN Carbyn, SH Fritts, and DR Seip, Eds. Ecology and conservation of wolves in a changing world. Canadian Circumpolar Institute, Occasional Publication No. 35, Edmonton. Boertje RD, P Valkenburg, and ME McNay. 1996. Increases in moose, caribou, and wolves following wolf control in Alaska. J Wildl Manage 60:474–489. Boutin S. 1992. Predation and moose population dynamics: a critique. J Wildl Manage 56:116–127. Campbell DT and JC Stanley. 1966. Experimental and quasi-experimental designs for research. Rand McNally, Chicago. Chichowski D. 1994. Summary—BC Parks Skeena wildlife surveys, March 1994. BC Parks Branch, Smithers, BC. Crête M and H Jolicoeur. 1987. Impact of wolf and black bear removal on cow: calf ratio and moose density in southwestern Québec. Alces 23:61–87. Eberhardt LL and KW Pitcher. 1992. A further analysis of the Nelchina caribou and wolf data. Wildl Soc Bulletin 20:385–395. Elliott JP. 1986a. Muskwa wolf management project of northeastern British Columbia: 1985–86 Annual Report. Wildl. Work. Rep. No. WR-21. Wildlife Branch. Ft. St. John, BC. 15 Pp. Elliott JP. 1986b. Kechika enhancement project of northeastern British Columbia wolf/ungulate management: 1985–86 Annual Report. Wildl Work Rep No. WR-20. Wildlife Branch, Ft. St. John, BC. 17 Pp. Elliott JP. 1989. Wolves and ungulates in British Columbia's northeast. In Wolf-prey dynamics and management. Proceedings. May 10–11, University of British Columbia. Vancouver, Canada. Wildl Working Report No. WR-40. BC Ministry of Environment, Victoria, BC. Farnell R and RD Hayes. 1992. Results of wolf removal on wolves and caribou in the Finlayson study area, Yukon, 1983–92. Dep Ren Res Whitehorse, YK. Farnell R, RD Hayes, and N Barichello. In preparation. Effects of reducing the number of wolves and human hunting on a population of caribou in the east-central Yukon from 1983–91.

OCR for page 90
Wolves, Bears, and their Prey in Alaska: Biological and Social Challenges in Wildlife Management Gasaway WC, RO Stephenson, JL Davis, PEK Shepherd, and OE Burris. 1983. Interrelationships of wolves, prey and man in interior Alaska. Wildlife Monographs 84:1–50. Gasaway WC, RD Boertje, DV Grangaard, DG Kellyhouse, RO Stephenson, and DG Larsen. 1992. The role of predation in limiting moose at low densities in Alaska and Yukon and implications for conservation. Wildl Monogr 120:1–59. Hatter DF. 1987. Perspectives on inventory of caribou in British Columbia. Spatsizi Association for Biological Research Rep No. 4. Wildlife Report No. R-14. BC Ministry of Environment, Victoria, BC. 93 Pp. Hatter IW. 1984. Effects of wolf predation on recruitment of black-tailed deer on northeastern Vancouver Island. MS thesis, University of Idaho, Moscow. 156 Pp. Hayes RD, AM Baer, and DG Larsen. 1991. Population dynamics and prey relationships of an exploited and recovering wolf population in the southern Yukon. Yukon Fish and Wildl Br Final Rep. TR-91-1. Whitehorse. 67 Pp. Hayes RD. 1992. An experimental design to test wolf regulation of ungulates in the Aishihik area, southwest Yukon. Yukon Fish and Wildlife Branch. Whitehorse. Hayes RD. 1995. Numerical and functional response of wolves, and regulation of moose in the Yukon. M.Sc. thesis., Simon Fraser University, Burnaby, British Columbia. James FC and CE McCulloch. 1995. The strength of inferences about causes of trends in populations. Pp. 40–51 in TE Martin and DM Finch, Eds. Ecology and management of neotropical migratory birds: a synthesis and review of critical issues. Oxford University Press. Jones G. 1985. Management of caribou in Spatsizi Plateau Wilderness park. In R Page, Ed. Caribou research and management in British Columbia. Proceedings of a workshop. BC Ministry of Forests, Research Branch, WHR-27, Victoria, BC. Southwest Yukon. Keith LB. 1983. Population dynamics of wolves. Pp. 66–67 in LN Carbyn, Ed. Wolves in Canada and Alaska, Canadian Wildlife Service Report Series 45 . 135 Pp. Kish L. 1987. Statistical design for research. Wiley, New York. Larson DG, DA Gauthier, and RL Markel. 1989a. Causes and rate of moose mortality in the southwest Yukon. J Wildl Manage 53:548–557. Larson DG, DA Gauthier, RL Markel, and RD Hayes. 1989b. Limiting factors on moose population growth in the southwest Yukon. Yukon Fish and Wildl. Br. Rep. Whitehorse. 105 Pp. Larsen DG and RMP Ward. 1995. Moose population characteristics in the Frances Lake and North Canol areas. 1991. Yukon Fish and Wildlife Branch PR-95-1. Marcstrom V, RE Kenward, and E Engren. 1988. The impact of predation on boreal tetraonids during vole cycles: an experimental study. J Anim Ecol 57:859-72. Meier TJ, LD Burch, D Mech, and L Adams. 1995. Pack structure and genetic relatedness among wolf packs in a naturally regulated population. Pp. 293–302 in LN Carbyn, SH Fritts, and DR Seip, Eds. Ecology and conservation of wolves in a changing world. Canadian Circumpolar Institute, Occasional Publication No. 35, Edmonton. Messier F. 1985. Social organization, spatial distribution and population density of wolves in relation to moose density. Can J Zool 63:1068–1077. Miller SD, and WB Ballard. 1992. Analysis of an effort to increase moose calf survivorship by increased hunting of brown bears on south-central Alaska. Wildlife Society Bulletin 20:445–454. Moses LE and F Mosteller. 1995. Experimentation: Just do it! Manuscript. Oldemeyer JL, AW Franzmann, AL Brundage, PD Arneson, and A Flynn. 1977. Browse quality and the Kenai moose population. J Wildl Manage 41:533–542. Peterson RO and JD Woolington. 1982. The apparent extirpation and reappearance of wolves on the Kenai Peninsula, Alaska. Pp. 334–344 in FH Harrington and PC Paquet, Eds. Wolves of the world. Noyes Publ, Park Ridge, NJ. Peterson RO, JD Woolington, and TN Bailey. 1984. Wolves on the Kenai Peninsula, Alaska. Wildl Monogr 88:1–52

OCR for page 90
Wolves, Bears, and their Prey in Alaska: Biological and Social Challenges in Wildlife Management Reid R and D Janz. 1995. Economic evaluation of Vancouver Island wolf control. Pp. 515–521 in LN Carbyn, SH Fritts, and DR Seip, Eds. Ecology and conservation of wolves in a changing world. Canadian Circumpolar Institute, Occasional Publication No. 35, Edmonton. Schwartz CC and AW Franzmann. 1991. Interrelationship of black bears to moose and forest succession in the northern coniferous forest. Wildl Monogr 88:1–52 Spencer DL and JB Hakala. 1964. Moose and forests on the Kenai. Tall Timbers Fire Ecology Conf 3:11–33. Stewart RR, EH Kowal, R Beaulieu, and TW Rock. 1985. The impact of black bear removal on moose calf survival in east-central Saskatchewan. Alces 21:403–418. Theberge JB. 1990. Potentials for misinterpreting impacts of wolf predation through prey: predator ratios. Wildl Soc Bulletin 18:188–192. Theberge JB and DA Gauthier. 1985. Models of wolf-ungulate relationships: when is wolf control justified? Wildl Soc Bulletin. 13:449–458. Van Ballenberghe V. 1981. Population dynamics of wolves in the Nelchina basin, south-central Alaska. Pp. 1246–1258 in JA Chapman and D Pursley, Eds. Proceedings of the Worldwide Furbearer Conference, Worldwide Furbearer Conference, Inc., Frostburg, MD. Van Ballenberghe V. 1985. Wolf predation on caribou: the Nelchina herd case history. J Wildl Manage 49:711–720 Van Ballenberghe V. 1989. Wolf predation on the Nelchina caribou herd: a comment. J Wildl Manage 53:243–250 Van Ballenberghe V. 1991. Forty years of wolf management in the Nelchina basin, south-central Alaska: a critical review. Transactions of the North American Wildlife and Natural Resources Conference 56:561–566. Walters CJ, M Stocker, and GC Haber. 1981. Simulation and optimization models for a wolf-ungulate system. Pp. 317–337 in CW Fowler and TD Smith, Eds. Dynamics of large mammal populations. John Wiley and Sons, NY. Whitten KR, GW Garner, FJ Mauer, and RB Harris. 1992. Productivity and early calf survival in the porcupine caribou herd. J Wildl Manage 56:201–212. Yukon Fish and Wildlife Branch. 1994. An evaluation of calf survival in the Aishihik caribou herd, Southwestern Yukon. Yukon Fish and Wildlife Branch. 1996. Summary of wildlife information Aishihik-Kluane caribou recovery program.