UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The influence of wolves on the ecology of mountain caribou Allison, Bradley Armstrong 1998

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1998-0339.pdf [ 2.55MB ]
Metadata
JSON: 831-1.0088482.json
JSON-LD: 831-1.0088482-ld.json
RDF/XML (Pretty): 831-1.0088482-rdf.xml
RDF/JSON: 831-1.0088482-rdf.json
Turtle: 831-1.0088482-turtle.txt
N-Triples: 831-1.0088482-rdf-ntriples.txt
Original Record: 831-1.0088482-source.json
Full Text
831-1.0088482-fulltext.txt
Citation
831-1.0088482.ris

Full Text

THE INFLUENCE OF WOLVES ON THE ECOLOGY OF MOUNTAIN CARIBOU by B R A D L E Y A R M S T R O N G A L L I S O N H.B .O.R . , Lakehead University, 1991 B . S c , Lakehead University, 1991 A THESIS S U B M I T T E D I N P A R T I A L F U L F I L L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F M A S T E R O F S C I E N C E in T H E F A C U L T Y O F G R A D U A T E S T U D I E S (Department of Animal Science) We accept this thesis as conforming tp the required standard T H E U N I V E R S I T Y O F B R I T I S H C O L U M B I A M a y 1998 © Bradley Armstrong Al l i son , 1998 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department The University of British Columbia Vancouver, Canada DE-6 (2/88) ABSTRACT The w o l f (Canis lupus)/mountam caribou (Rangifer tarandus caribou) spatial separation model (Seip 1992a) was examined in highland and mountainous areas of east central and southeastern British Columbia to determine the influence of wolves on mountain caribou ecology. Three key elements investigated were: the importance of wol f predation as a mountain caribou mortality factor, the elevational relationships of mountain caribou, wolves, and moose (Alces alces), and the seasonal dietary importance of moose to wolves. Mountain caribou mortality data from the Columbia Mountains and Quesnel Lake supported the hypothesis that w o l f predation is greater in highland than in mountainous areas. W o l f predation was the main mortality factor of caribou in the highlands around Quesnel Lake, but was a minor factor in the other three study areas. W o l f predation at Quesnel Lake occurred primarily during summer/fall at low elevations. Mountain caribou, w o l f and moose radio-telemetry data suggested that wolves in both highland and mountainous areas are more closely associated elevationally with moose than caribou throughout the year. Caribou in highland areas appeared as adept as those in rugged mountains at spacing elevationally away from wolves during all seasons. Significant elevational overlap between wolves and caribou occurred only in the rugged Columbia Mountains, and then, only during summer/fall. W o l f scats from the Columbia Mountains indicated that mountain caribou were of lesser dietary importance than moose to wolves throughout the year. Moose, particularly in winter, were the main diet item of wolves. Beaver was an important dietary item of wolves during summer/fall. Elevational separation appears inadequate in explaining the variation in w o l f predation on mountain caribou in the highland and mountainous study areas. Differences in w o l f densities and the relative densities of moose and caribou may be the main factors influencing the susceptibility of mountain caribou to wol f predation. It is recommended that studies be conducted in both topography types across a wide range of caribou and moose densities to better explain the influence of wolves on mountain caribou ecology. In addition, it w i l l be necessary to compare the geographic distribution of mountain caribou and wolves to assess i f geographic rather than elevational separation occurs. Determining how timber harvesting impacts the numerical and spatial responses of wolves, mountain caribou and moose is also recommended. i i i TABLE OF CONTENTS Abstract ii Table of Contents iv List of Tables I vi List of Figures vii Acknowledgements viii Chapter 1. General Introduction 1 Study Areas 2 Chapter 2. Wolves: Their role in mountain caribou mortality Introduction 6 Methods 7 Results 10 Discussion 14 Chapter 3. Elevational relationships of mountain caribou, wolves and moose Introduction 18 Methods 19 Results 20 Discussion 24 Chapter 4. The seasonal dietary importance of moose to wolves in the Columbia Mountains Introduction 27 Methods 27 Results 28 Discussion 30 iv Chapter 5. General conclusion and management recommendations Literature Cited LIST OF TABLES Table 1. Level of mountain caribou monitoring and annual mortality rate (%)± (sd) within study areas. 11 Table 2. Causes of radio-collared mountain caribou mortalities within study areas. 12 Table 3. Seasonal causes of radio-collared mountain caribou mortalities across all study areas (Columbia Mountains 1981-85, 1992-96; Wells Gray Provincial Park 1986-89; Quesnel Lake 1984-89; Quesnel Highlands 1993-96). Length of season (days) in brackets. 12 Table 4. Type and frequency of mortalities occurring above or below the median seasonal elevation of radio-collared mountain caribou (Columbia Mountains 1981-85, 1992-96; Wells Gray Provincial Park 1986-89; Quesnel Lake 1984-89; Quesnel Highlands 1993-96). 13 Table 5. Mountain caribou density estimates based on mean annual distribution of populations. 14 Table 6. Mean w o l f density estimates within pack territories. 14 Table 7. Wolf/mountain caribou and wolf/moose seasonal intermedian distances (m) within study areas. 24 Table 8. Frequency of occurrence and percent biomass of wo l f diet items by season in the Columbia Mountains. 29 vi LIST OF FIGURES Figure 1. Location of study areas: Columbia Mountains, Wells Gray Provincial Park, Quesnel Lake and Quesnel Highlands, British Columbia. 3 Figure 2. Columbia Mountains radio-collared mountain caribou, wo l f and moose median monthly elevations and interquartile ranges (1992-96 caribou and wolf; 1984-85 moose). 21 Figure 3. Wells Gray Provincial Park radio-collared mountain caribou and wol f median monthly elevations and interquartile ranges (1986-89). 21 Figure 4. Quesnel Lake radio-collared mountain caribou, wo l f and moose median monthly elevations and interquartile ranges (1984-89). 22 Figure 5. Quesnel Highlands radio-collared mountain caribou and w o l f median monthly elevations and interquartile ranges (1993-96). 22 v i i ACKNOWLEDGEMENTS "We are made to exaggerate the importance of what work we do; and yet how much is not done by us!" (Thoreau 1854) From the outset of my search for knowledge and understanding of wolves, mountain caribou and science in general, many individuals graciously helped me move toward attaining this goal. A s always, I recognize my path was directed by God and that He sustained me faithfully throughout this adventure. I thank my wife, Sue, for providing patient support and steadfast encouragement throughout all the various phases of this thesis, in particular when I was absent from home and out tracking wolves in the Columbia Mountains. I greatly appreciate my family members and friends for cheering me onward and providing necessary moments of levity. B . N . McLe l l an , my research supervisor guided me throughout this research, and most notably provided sage advice, fund-raising support, good coffee and occasionally a floor to crash on. D . Shackleton, my academic supervisor, tried diligently to keep me focused whenever I was within arms length of "the hut". M y other committee members: D . Seip, M . Pitt, and T. Sullivan, assisted freely, whenever asked. R. Ramcharita deserves a special note of thanks for volunteering countless hours in my initial efforts to radio-collar wolves. In addition to almost always keeping the snowmobile tanked up, he shared many ideas which undoubtedly increased the quality of this research. J. Flaa assisted in the field, and together with M . Super and D . Mai r spent many hours airborne, locating radio-collared caribou and wolves in the Columbia Mountains. J. Townley helped in backtracking wolves during the winter, while the assistance of D . Lewis, J. Krebs, K . Atkinson and A . Lay also eased my work in the field. G . Galz i , S. Wilson and S. Leung offered advice and assistance with various database queries. I. Clemens assisted with computer programming and data analysis. K . Al l i son created a map of the study areas. Individuals that generously provided data for analysis were: B . N . McLe l l an and J. Flaa, K . Simpson and G . Woods (Columbia Mountains caribou data); D . Seip (Quesnel Lake caribou/wolf/moose data; Wells Gray caribou/wolf data); J. Young and P. Dielman (Quesnel Highlands caribou/wolf data); K . Simpson (Columbia Mountains moose data); and J. Woods (Banff National Park ungulate weight data). Financial support for this study was provided by the B . C . Ministry of Forests, Mount Revelstoke and Glacier National Parks, Columbia Basin Fish and Wildl i fe Compensation Program, Canadian Wildlife Service - University Research Support Fund, The Friends of Mount Revelstoke and Glacier, Revelstoke Community Forest Corporation, Canadian Wildl i fe Federation - Orville Erickson Memorial Scholarship, University o f British Columbia - James Robert Thompson Fellowship, E . Al l i son , E . and W . Bauman and P. Spaetzel. Although they w i l l never read this, I am grateful for the wolves, caribou and moose that through direct and indirect observation, allowed me a brief glimpse into their lives and provided many enjoyable moments afield. Ultimately, I realize that together with the help of many fine individuals, my burdens were lightened, progress was made, and I was truly allowed to enjoy "one of the best times of my life" ( B . N . McLe l l an , personal communication). CHAPTER 1 GENERAL INTRODUCTION Mountain caribou (Rangifer tarandus caribou) are on the blue or vulnerable list o f species in British Columbia (Harper et al. 1994). Their low numbers (Simpson et al. 1997), low rate of reproduction (Bergerud 1978) and reliance on old growth forests for food and shelter which conflicts with timber harvesting needs (Simpson et al. 1996) collectively resulted in this designation. The distribution and abundance of mountain caribou began to decrease across southern British Columbia in the early 1900s (Munro 1947). Explanations for the decline have included habitat degradation, overhunting and predation (Bergerud 1974; Bloomfield 1980; Edwards 1954; Munro 1947), but wo l f (Canis lupus) predation has recently been suggested as the most significant limiting factor across southern British Columbia (Bergerud 1988a, 1996; Seip 1991, ( 1992a; Seip and Cichowski 1996). Bergerud and Elliot (1986) and Seip (1992a) suggested that w o l f populations within mountain caribou range are sustained primarily by moose (Alces alces), but in areas of gentle topography, wolves were also a significant predator of mountain caribou. Ultimately, it was concluded that mountain caribou in more rolling, highland areas experience more w o l f predation than caribou in rugged mountainous areas due to their inadequate degree of spatial separation with wolves and moose (Seip 1992a). To determine the influence of wolves on the ecology of mountain caribou, I investigated three elements of the spatial separation model (Seip 1992a): the importance of w o l f predation as a mountain caribou mortality factor, the elevational relationships of caribou, wolves and moose, and the seasonal dietary importance of moose to wolves. Chapter 1 includes this introduction to the research and a description of the four study areas. In Chapter 2,1 use mountain caribou mortality data collected during five previous studies ( B . N . 1 McLe l l an and J. Flaa, unpublished data; Seip 1990, 1992b; Simpson and Woods 1987; J .A. Young and P. Dielman, unpublished data) to compare the rates, causes, seasons and elevations of mountain caribou mortality among four study areas, and in particular, the impact of w o l f predation on animals in highland and mountainous areas. In Chapter 3,1 use radio-telemetry data ( B . N . McLe l l an , J. Flaa, and B . Al l i son , unpublished data; Seip 1990, 1992b; K . Simpson, unpublished data; J .A. Young, unpublished data) from the four study areas to examine the elevational relationships of mountain caribou, wolves and moose in both highland and mountainous areas. In Chapter 4,1 estimate the seasonal dietary importance of moose to wolves in the Columbia Mountains using wol f scats I collected from that area. The final chapter summarizes my findings and provides some management recommendations. STUDY AREAS The four study areas are located in east central and southeastern British Columbia. These areas are: the Columbia Mountains north of Revelstoke (51°30'N, 118°15'W), Wells Gray Provincial Park (52°15'N, 120°W), the Quesnel Lake area (52°45'N, 120°45'W) and the Quesnel Highlands (52°45'N, 121°W) near Horsefly (Fig. 1). The Quesnel Highlands study area encompasses the Quesnel Lake study area, but since the data examined come from two different time periods, and the areas are of considerably different size, they are referred to as separate studies. Although there are some moderate slopes and plateaus, the topography of the Columbia Mountains and the majority of Wells Gray Provincial Park is typical of rugged terrain. Valleys are narrow and steep with mountains rising to massive rock outcrops and glaciers. Elevations range from 575 to 3200 m in the Columbia Mountains and 630 to 2850 m in Wells Gray Provincial Park. A n estimated 375 caribou inhabit the Columbia Mountains study area (McLel lan 2 Fig . 1. Location of study areas: Columbia Mountains, Wells Gray Provincial Park, Quesnel Lake, and Quesnel Highlands, British Columbia. 3 et al. 1994b), while 275 and 302 were estimated within and adjacent to Wells Gray Provincial Park in 1987 and 1989 respectively (Seip 1990). The Quesnel Lake and Quesnel Highlands study areas contain some rugged mountains but generally consist of highlands with rounded, vegetated mountains and moderate slopes (Holland 1976). Large rock outcrops are rare and glaciers are absent. Elevations range from 728 m to almost 2600 m. The Quesnel Lake study area was estimated to have 220 mountain caribou in 1986, but dropped to 94 in 1989 and 95 in 1997 (Seip 1992b; J .A. Young, unpublished data). In 1997, 358 mountain caribou (J.A. Young, unpublished data) were estimated to reside in the Quesnel Highlands area. A portion of the Quesnel Highlands study area is rugged with mountains reaching 2800 m. The majority of the valley bottoms within the four study areas are in the Interior-Cedar-Hemlock (ICH) biogeoclimatic zone which occurs up to 1250 - 1350 m. Forests in the I C H zone are dominated by closed canopies of western red cedar {Thuja plicata) and western hemlock {Tsuga heterophylla) (Ketcheson et al. 1991). L o w elevations in the northern and western portions of the Quesnel Highlands contained the Sub-Boreal Spruce (SBS) zone and are dominated by hybrid white spruce (Picea engelmannii x glauca) and subalpine fir {Abies lasiocarpa) (Meidinger et al. 1991). M i d slopes in the four study areas are occupied by the Engelmann Spruce-Subalpine F i r (ESSF) biogeoclimatic zone, which extends to 1800 - 2000 m. Lower portions of the E S S F are dominated by closed canopy forests of hybrid white spruce and subalpine fir (Coupe et al. 1991) while subalpine fir increases in abundance at higher elevations and near timberline, grows in small clumps, forming the more open subalpine parkland (Coupe et al. 1991). The Alpine Tundra (AT) zone occurs at the highest elevations in the four study areas. Snowpacks in the mountainous portions of the study area are high and average 1.3 m in valley bottoms and 2.8 m at snow stations between 1620 and 1920 m (B.C. Ministry of Environment 4 1985). Snowfall accumulations in the Quesnel Highlands are lower than those of the mountains and average 1.0 m in valley bottoms and 1.9 m at stations between 1570 and 1900 m during the late winter (B .C . Ministry of Environment 1985). Ungulates in all four study areas include: woodland caribou, moose, mule deer (Odocoileus hemionus), white-tailed deer (O. virginianus), and mountain goats (Oreamnos americanus). Wolves, grizzly bears (Ursus arctos), black bears (U. americanus), and wolverines (Gulo gulo) are the main predators within the study areas. 5 C H A P T E R 2 WOLVES: THEIR ROLE IN MOUNTAIN CARIBOU MORTALITY INTRODUCTION Since the latter part of the nineteenth century, woodland caribou populations have declined across a large portion of their historic North American range (Anderson 1938). In southern British Columbia, the mountain ecotype began decreasing around 1900 and disappeared from some areas (Munro 1947). Currently there are about 2500 mountain caribou and almost all are found in British Columbia (Simpson et al. 1997). Mountain caribou feed largely on arboreal lichen during the winter (Antifeau 1987, Edwards and Ritcey 1960, Rominger and Oldemeyer 1990, Servheen and L y o n 1989; Simpson et al. 1997; Terry 1994), and because these lichens are most abundant on old trees, caribou management conflicts directly with forest harvesting. Maintaining present mountain caribou populations and their requisite habitats is a major conservation challenge in British Columbia (Simpson et al. 1997, Stevenson and Hatler 1985). Several hypotheses have been proposed to explain the decline of mountain caribou. Overharvesting by native (Munro 1947) and recreational hunters due to liberal hunting regulations, combined with improved access, was suspected (Bergerud 1974; Bloomfield 1980; Seip and Cichowski 1996). Habitat degradation due to fires (Edwards 1954) and timber harvesting was thought to reduce caribou winter food, decrease shelter and disrupt travel corridors (Bloomfield 1980). In portions of southern British Columbia, cougar (Puma concolor) predation was believed to have caused the extirpation of caribou in one area (Munro 1947), while predation by an increasing wol f population during the 1930s and 1940s was suggested in other areas (Bergerud 1974; Munro 1947). Although more pluralistic hypotheses have been proposed to explain the widespread decline in mountain caribou (Bloomfield 1980; Stevenson and Hatler 6 1985), w o l f predation is currently emphasized as the primary factor involved (Bergerud 1974, 1988a, 1996; Seip 1991, 1992a; Seip and Cichowski 1996). Bergerud and Ell iot (1986) reported that in areas where wolves are abundant, caribou populations decline, and Seip (1992a) proposed that mountain caribou in highland areas with less ability to space away from wolves and moose, experience more wol f predation particularly in summer, than caribou in rugged mountains. The objectives of this chapter are to: 1) compare rates, causes, seasons and elevations of mountain caribou mortality among four study areas, 2) compare wol f and caribou densities among those study areas, and 3) test whether w o l f predation on caribou is greater in highland areas than in mountainous areas. METHODS Capture and monitoring The mountain caribou mortality data from the four study areas (Columbia Mountains; Wells Gray Provincial Park; Quesnel Lake; Quesnel Highlands) were collected during five previous studies ( B . N . McLe l l an and J. Flaa, unpublished data; Seip 1990,1992b; Simpson and Woods 1987; J .A. Young and P. Dielman, unpublished data). Most mountain caribou in the four study areas were captured by net-gunning from a helicopter in subalpine habitats during late winter (McLel lan et al. 1994a; Seip 1990, 1992b; Young et al. 1996). Other caribou were captured by darting with immobilization drugs during the 1981-85 Columbia Mountains study (Simpson and Woods 1987) and at the start of the 1984-89 Quesnel Lake study (Seip 1992b). A l l caribou were fitted with mortality-motion sensitive radio collars. Radio-collared animals were relocated from aircraft two to four times per month and location and elevation were recorded for each relocation. 7 Mortality factors and seasons When a signal from a motion-sensitive radio collar indicated that a caribou was dead, the site was investigated and i f possible, the caribou was necropsied to determine cause o f death. Mortality causes were first classified as predation, not predation, and unknown. Cases of predation were then divided into wolf, bear, wolverine or unknown. Non-predation deaths were divided into accident, avalanche, malnutrition or unknown. Criteria used to attribute predation mortalities to wolf, bear or wolverine were: presence of predator, tracks, scats, or hair at the mortality site; extensive blood on carcass, ground or snow; distribution of the carcass, and site conditions ( B . N . McLe l l an and J. Flaa, unpublished data; D .R. Seip, unpublished data; J .A. Young, unpublished data). Malnutrition deaths in all four study areas were determined from bone marrow fat content ( B . N . McLe l l an and J. Flaa, unpublished data; D .R. Seip, unpublished data; K . Simpson, personal communication; J .A. Young, unpublished data), with renal and rump fat measures also used in the Columbia Mountains. Mortalities were classified as unknown when insufficient carcass remnants were available at the mortality site to assign a mortality cause. Mortalities were also classed as not predation but unknown when necropsies on relatively intact carcasses failed to reveal a mortality agent. A l l caribou mortalities were assumed to be classified to the correct mortality factor. Date o f mortality was assumed to fall halfway between the date the caribou was last relocated alive and the date when the mortality signal was first heard (Mayfield 1975). I classified season of mortality as spring (May 1 - M a y 28), summer/fall (May 29 - Nov 7), early winter (Nov 8 -Dec 7) or late winter (Dec 8 - Apr 30). These seasons denote shifts in elevation and habitat use by mountain caribou and correspond to the conventional seasons of other mountain caribou studies (Antifeau 1987; Simpson and Woods 1987; Terry et al. 1996). Seasonal delineation for 8 each study area was determined by plotting mean caribou elevation by Julian week (McLel lan et al. 1994a). Boundaries between seasons were then identified when the weekly change between mean caribou elevation was greatest, indicating a major elevational shift. Seasonal boundaries were then averaged among study areas to attain the four standardized seasons. Mountain caribou and wolf density estimates The 95% adaptive kernel method (Seaman and Powell 1991) was used to estimate the annual distribution o f each caribou population and wol f pack from the relocations o f radio-collared animals. Density estimates for each population were obtained by dividing the mean annual distribution by the estimated number of individuals in each caribou population or wo l f pack. The mean of all w o l f pack density estimates in each study area was used for inter-study comparisons. Estimates of caribou population size in each study area were obtained from censuses conducted during late winter as described by McLel lan et al. (1994b), Seip (1990, 1992b) and J .A. Young (personal communication). Estimates of winter wol f pack sizes (Fuller and Snow 1988) were obtained during relocation flights, while the number of packs in each study area were estimated from incidental observations and wol f radio-telemetry data ( B . N . McLel lan , J. Flaa, and B . Al l i son , unpublished data; Seip 1990, 1992b; J .A. Young, unpublished data). Statistical analyses Annual mortality rates were estimated by dividing the total number of caribou mortalities by the total number of caribou years (one caribou monitored for one year is a caribou-year), based on methods described by White and Garrott (1990) for survival data. Mult iple male and female caribou were radio-collared in the Columbia Mountains so tests for differences in male and 9 female caribou mortality rates were made. I f no difference between rates was found, the mortality data of both sexes were pooled. Chi-square tests of independence (Siegel and Castellan 1988) were used to test for differences in mortality rates among genders and areas. I f different rates were found among areas, the contingency table was partitioned into subtables following the methods of Siegel and Castellan (1988) to determine the cause of the difference. Fisher's extact test (Siegel and Castellan 1988) was used to test whether wo l f predation on caribou is higher in highland areas than in mountainous areas using mortality from study areas with adequate sample sizes. Levels of significance for all tests were set atp < 0.05. RESULTS Mortality rates In total, 155 mountain caribou (132 adult females; 11 adult males; 9 juvenile females; 3 juvenile males) from the four study areas were radio-collared and monitored for a total o f 410 caribou-years between 1981 and 1996 (Table 1). Different annual mortality rates were found among the four study areas = 15.56, df = 3,p = 0.002). Partitioning the contingency table showed no difference between rates in Wells Gray Provincial Park, the Columbia Mountains and the Quesnel Highlands = 0.18, df = 1, p = 0.75), but the mortality rate o f caribou at Quesnel Lake was higher than the other study areas (% - 14.56, df = 1, p < 0.001). 10 Table 1. Level of mountain caribou monitoring and annual mortality rate (%)± (sd) within study areas. Study Area No. of caribou- No. (age - sex) of Annual years radio-collared animals mortality of monitoring rate A d F A d M JuvF Juv M Total Columbia Mountains 162 41 9 7 3 60 123 + 26" 1981-85 & 1992-96 Wells Gray Prov. Park 75 27 1 2 - 30 8.0 ± 3 . 1 1986-89 Quesnel Lake 81 31 1 - - 32 28.4 ± 5 . 0 1984-89 Quesnel Highlands 92 33 - - - 33 13.0 ± 3 . 5 1993-96 " Combined mortality rate of male and female caribou, as their mortality rates were not significantly different ( x 2 =0.02, df= l,/? = 0.90). Causes, seasons and elevations of mountain caribou mortalities Mountain caribou in the four study areas died from a variety of factors. The proportion of w o l f to non-wolf caribou deaths differed between the highland area, Quesnel Lake and the Columbia Mountains (Fisher's exact test, p = 0.0002). W o l f predation was the main mortality factor at Quesnel Lake, accounting for 10 (56%) of the known deaths, but was a minor factor in the other three study areas (Table 2). Bear predation was the most prominent mortality factor in Wells Gray Park, causing three (75%) of the known deaths. Avalanches were the greatest single factor contributing to caribou deaths in the Columbia Mountains, causing six (35%) known mortalities. Mountain caribou in the Quesnel Highlands, though, died from a variety o f predation (43%) and non-predation (57%) factors. The highest seasonal mortality rate occurred during spring, but most mountain caribou died during the summer/fall or late winter (Table 3). Thirteen (56.5%) caribou deaths at Quesnel Lake occurred during the summer/fall, eight of which were due to wo l f predation. The summer/fall accounted for five (83.3%) caribou deaths in Wells Gray Park, with bear predation causing three 11 Table 2. Causes o f radio-collared mounta in car ibou mortal i t ies w i th in study areas. Study Area Predation Non-predation Unk. Total mortalities Wolf Bear Wolverine Malnutrition Avalanche Accident Unk. Columbia Mountains - 2 3 1 6 3 2 3 20 1981-85 & 1992-96 Wells Gray Prov. Park 1 3 - - - 2 6 1986-89 Quesnel Lake 10 3 2 1 2 - 5 23 1984-89 Quesnel Highlands 2 1 1 1 2 - 5 12 1993-96 Total 13 8 4 4 8 7 2 15 61 o f these. T e n ( 5 0 % ) deaths occurred dur ing late winter i n the C o l u m b i a Mounta ins w i t h avalanches caus ing f ive o f these and wolver ine predat ion three. In the Quesne l H igh l ands , four (33 .3% ) deaths occurred i n both summer/fal l and late winter, but no clear seasonal morta l i ty factor was observed. Deaths result ing f rom accidents and malnutr i t ion d i d not exhib i t strong seasonal aff init ies i n any study area. Tab le 3. Seasonal causes o f radio-collared mounta in car ibou mortal i t ies across a l l study areas ( C o l u m b i a Mounta ins 1981-85, 1992-96; We l l s G ray P rov inc i a l Park 1986-89; Quesne l L ake 1984-89; Quesne l H igh lands 1993-96). Length o f season (days) i n brackets. Season Predation Non-predation Unk. Total Wolf Bear Wolverine Malnutrition Avalanche Accident Unk. Spring (28) - 2 1 2 3 8 Summer/Fall (163) 10 6 1 1 2 2 7 29 Early Winter (30) - - 1 1 1 2 5 Late Winter (144) 3 4 1 6 2 - 3 19 In study areas where w o l f and bear predat ion occurred, most car ibou were k i l l e d b y these predators at l o w elevations (Table 4). Wo lve r ine predation, however , occurred at both h igh and l o w elevations. Ca r i bou were most l i ke l y to be k i l l e d b y avalanches wh i l e at h igh altitudes, but 12 mortalities caused by malnutrition and accidents occurred with the greatest frequency at low elevations. Table 4. Type and frequency of mortalities occurring above or below the median seasonal elevation of radio-collared mountain caribou across all study areas (Columbia Mountains 1981-85, 1992-96; Wells Gray Provincial Park 1986-89; Quesnel Lake 1984-89; Quesnel Highlands 1993-96). Elevation Predation Non-predation Unk. Total Wolf Bear Wolverine Malnutrition Avalanche Accident Unk. Above Median 2 1 2 6 2 1 4 18 Below Median 11 7 2 4 1 5 1 7 38 Mountain caribou and wolf densities Among the four study areas, the highest estimated caribou density (11.4 caribou/100 km 2) occurred in the highlands around Quesnel Lake in 1986 (Table 5), but in 1989, the estimated density fell to 4.9 caribou/100 km 2 . This was within the range of caribou densities (3.4 to 5.3 caribou/100 km 2 ) observed in the other highland (Quesnel Highlands) and mountainous (Columbia Mountains; Wells Gray Provincial Park) study areas. The two highland study areas (Quesnel Lake; Quesnel Highlands) had the highest estimated w o l f densities and observed winter pack sizes, but wo l f density was not estimated for Wells Gray Park as the number of wolves/pack was unknown (Table 6). Wells Gray Provincial Park, the Columbia Mountains and the Quesnel Lake study areas each had two known w o l f packs, while the Quesnel Highlands study area had four. 13 Table 5. Mountain caribou density estimates based on mean annual distribution o f populations. Study Year of population estimate Number of caribou (95% C.I.) Mean annual distribution (km2) Density (caribou/100 km2) Columbia Mountains 1994 375 (337 -413) 7018 5.3 (1992-96) Wells Gray Prov. Park 1987 275 (220 - 297) 8206 3.4 (1986-89) 1989 302 (258 - 346) It 3.7 Quesnel Lake 1986 220 (197 - 243) 1931 11.4 (1984-89) 1989 94 (86 - 102) It 4.9 Quesnel Highlands 1997 358 (294 - 422) 9398 3.8 (1993-96) Table 6. Mean w o l f density estimates within pack territories. Study Number of Number of Mean density known wolf wolves/radio- (wolves/100 km 2 packs/study area collared pack (winter) of pack territory) Columbia Mountains (1992-96) 2 3-4 0.61 Wells Gray Prov. Park (1986-89) 2 Unknown Unknown Quesnel Lake (1984-89) 2 6-8 0.84 Quesnel Highlands (1993-96) 4 5-8 0.81 DISCUSSION Variation in wolf predation among highland and mountainous study areas Mortality data from the four study areas showed that mountain caribou at Quesnel Lake had a much higher mortality rate than in the other three study areas, as the caribou density dropped from 11.4 to 4.9 caribou/100 k m between 1986 and 1989. The results support the hypothesis that w o l f predation is greater in highland areas than in mountainous areas as suggested by Seip (1992a). W o l f predation was the main cause of caribou deaths at Quesnel Lake (Seip 1992a, 1992b) but heavy wol f predation was not reported in any other study area. Non-wolf mortality 14 factors were more common in rugged mountainous terrain. It appeared that bears were the major predator of caribou in Wells Gray Park (Seip 1990). Avalanches were the single most important mortality factor o f caribou in the Columbia Mountains, but both wolverine and bear predation were also notable ( B . N . McLel lan and J. Flaa, unpublished data; Simpson and Woods 1987). Another study, in the southern Selkirk Mountains, found cougar predation was the cause of most caribou mortalities (Compton et al. 1995). Seip (1992a) attributed differences in wol f predation on mountain caribou in highland and mountainous areas to differences in the degree of elevational and geographic separation between caribou and wolves. Upon examination, elevational separation (Seip 1992a) does not appear to adequately explain the large variation in wol f predation among the four study areas. Except for summer/fall in the Columbia Mountains, caribou in both highland and mountainous areas were found to maintain elevational and hence geographic separation from wolves (see chapter 3). The separation of mountain caribou and wolves in geographically different areas during summer/fall in the Columbia Mountains although not examined in this thesis, may as reported elsewhere (Bergerud 1985; Bergerud et al. 1990; Cumming et al. 1996; Seip 1990, 1992a) explain why wol f predation on caribou was not observed in this study area. Two other factors may account for the variation in wol f predation on mountain caribou among the highland and mountainous study areas. Firstly, w o l f density differences (Bergerud and Ell iot 1986) may explain why few caribou were killed by wolves in the mountainous study areas (Columbia Mountains; Wells Gray Provincial Park), while caribou at Quesnel Lake frequently died from w o l f predation. Bergerud and Elliot (1986) reported that caribou populations decline in areas where w o l f densities exceed 0.65/100 k m 2 . W o l f density within pack territory in the 15 Columbia Mountains (0.61/100 km 2 ) did not reach this critical level and it appears unlikely that wolves in Wells Gray Provincial Park exceeded it either. Wolves in the highlands around Quesnel Lake, however, with more wolves per pack, and smaller territories than wolves in the Columbia Mountains, had densities (0.84/100 km 2 ) above the critical threshold. This may have lead to more w o l f predation on Quesnel Lake caribou. I f wo l f abundance is ultimately determined by the availability of prey biomass (Boertje and Stephenson 1992; Dale et al. 1994; Hayes 1995; Messier 1985, 1994; Packard and Mech 1980), and alternate ungulate prey remain sparse in the wet, rugged mountains of mountain caribou range (Terry et al. 1996), caribou in those areas should continue to be less susceptible to wo l f predation. Secondly, the relative densities of moose and caribou (Cumming et al. 1996; Seip and Cichowski 1996) may also be a factor influencing the susceptibility of mountain caribou to wol f predation. Optimal foraging theory (Krebs and McCleery 1984) predicts that wolves should hunt in the most optimal areas available to them, selecting the most energetically profitable prey (MacArthur and Pianka 1966). Assuming wolves in the four study areas hunted optimally, one can infer from the indirect evidence (wolf scat contents and/or elevations used by wolves) that at the prey densities wolves encountered, moose were considered the more profitable prey. I f the density o f moose decreased and/or caribou density increased significantly, wolves would be expected to switch predation efforts (functional response) from moose to caribou (Farnell et al. 1996). This behaviour would continue until once again it became more profitable to prey on moose (Messier 1995). Perhaps this explains why caribou numbers plunged at Quesnel Lake between 1986 and 1989 (Seip 1992a), but have since stabilized (J.A. Young, personal communication), even though wol f density remained high. The relatively high caribou density (11.4/100 k m ) in 1986 may have been sufficient to alter wolves from preying largely on moose 16 to also prey on caribou as this density is above the equilibrium density of 3.0-4.0 mountain caribou/100 k m 2 as suggested by Seip and Cichowski (1996). Seasons and elevations of mountain caribou mortality factors It has been reported that mountain caribou are most vulnerable to wolf, bear, and cougar predation during summer (Compton et al. 1995; Seip 1990; Seip 1992b) especially at lower elevations (Seip 1992a). Mortality data from the Quesnel Highlands and Columbia Mountains studies support these findings. Unlike other carnivores, wolverines ki l led radio-collared caribou only during the late winter at both valley bottom and subalpine elevations ( B . N . M c L e l l a n and J. Flaa, unpublished data; J .A. Young, unpublished data). Mountain caribou may be most susceptible to predation by this relatively small carnivore in late winter because the caribou have depleted most of their fat reserves and may be in poor condition. Malnutrition deaths did not exhibit this biological pattern in the four areas examined, but this is l ikely due to a relatively small number of mountain caribou mortalities. 17 CHAPTER 3 E L E V A T I O N A L R E L A T I O N S H I P S O F M O U N T A I N C A R I B O U , W O L V E S A N D M O O S E I N T R O D U C T I O N Wolves that prey primarily on moose, but sometimes switch to woodland caribou, have been implicated as a major cause of both past and present declines in mountain caribou populations (Bergerud 1974, 1988a, 1996; Seip 1991, 1992b, Seip and Cichowski 1996). It was suggested that woodland caribou could only exist on ranges shared with high densities of moose or deer and wolves i f a habitat feature allowed caribou to avoid wolves and their alternate ungulate prey (Bergerud 1974). Elevation appears to provide caribou with a unique means to space away from wolves and their alternate ungulate prey. In northern B . C . , the northern ecotype of woodland caribou used rugged mountainous terrain and intra-specific dispersion to provide spatial separation from wolves and moose (Bergerud et al. 1984; Bergerud and Page 1987). In the Quesnel Lake area of east central B . C . , wolves were sustained primarily by moose (Seip 1992b), but during the summer, their distribution overlapped with mountain caribou and they appeared to k i l l enough caribou to cause a rapid decline (Seip 1992a). Meanwhile, caribou that were spatially separated from moose and wolves in the nearby rugged mountains of Wells Gray Provincial Park experienced little wo l f predation (Seip 1990). Seip (1992a) proposed that caribou l iving in highland habitats were more available and thus more prone to w o l f predation than caribou in more mountainous terrain where they were separated from wolves. The spatial separation model (Seip 1992a) of which elevation is a component, may apply throughout mountain caribou range. The objective of this chapter is to test the following hypotheses: 1) wolves show more elevational overlap with moose than with mountain caribou, and 2) mountain caribou in highland 18 areas show less elevational separation from wolves than those living in rugged mountains. METHODS Capture and elevational monitoring of mountain caribou, wolves and moose Radio telemetry data on mountain caribou and wolves from two study areas (Columbia Mountains 1992-96; Quesnel Highlands 1993-96), and moose radio-telemetry data from one study area (Columbia Mountains 1984-85), were used to evaluate the two hypotheses. Seip (1992a) did not statistically test the mountain caribou and wol f radio-telemetry data (Wells Gray Provincial Park 1986-89; Quesnel Lake 1984-89) nor the moose radio-telemetry data (Quesnel Lake 1984-89) so these data were also examined. Captured animals were fitted with radio collars and relocated from aircraft two to four times each month. Ground telemetry was also used to locate wolves in winter during the 1992-96 Columbia Mountains study. The location and elevation were recorded for each relocation. Statistical Analyses Elevation of radio locations were often asymmetrically distributed, thus two-tailed robust rank-order tests (Siegel and Castellan 1988) were used to test whether wolves and caribou, and wolves and moose used different elevations during each season in the highland and mountainous study areas. Frequency of use analyses were not used due to inadequate sample sizes in each study area during some seasons. Seasonal intermedian distances of wolves and caribou, and wolves and moose in highland and mountainous areas were also compared as another measure o f the hypotheses. Seasons were based on elevational movements of caribou (McLel lan et al. 1994a) and were classified as spring, summer/fall, early winter and late winter (see chapter 2). The level of significance for all statistical tests was set atp < 0.05. 19 RESULTS Seasonal use of elevations by mountain caribou, wolves and moose In spring, caribou in the highland (Quesnel Lake; Quesnel Highlands) and mountainous (Columbia Mountains; Wells Gray Provincial Park) study areas used lower elevations than during late winter, however, they remained separate from wolves which occupied still lower elevations (allp < 0.001) (Figures 2,3,4 and 5). In contrast, no difference was observed in the A elevations used by wolves and moose in the Columbia Mountains (U [19,10] = 1.69, p = 0.091) A or in the highlands around Quesnel Lake (17 [6,14] = 0.608, p = 0.543) during spring (Figs. 2 and 4). During summer/fall, wolves in all study areas increased their use of higher elevations (Figs. 2,3,4 and 5), particularly in the rugged Columbia Mountains where elevations used by wolves and caribou did not differ (U [38,1314] = 1.46,p = 0.143), nor did those used by wolves and A moose (U [38,230] = 0.396,p = 0.692) (Fig. 2). Wolves and caribou during summer/fall in Wells Gray Provincial Park, and in both highland areas (Quesnel Lake; Quesnel Highlands) remained elevationally separate (all p < 0.001), although wolves at Quesnel Lake showed some use of high elevations during June (Fig. 4). In the Quesnel Lake area, moose and wolves used different elevations in the summer/fall (U [92,64] = -2.34,p = 0.019) with moose using higher elevations (Fig. 4). During early winter, wolves and caribou in highland and mountainous areas again used lower elevations, but still little altitudinal overlap occurred (Figs. 2,3,4 and 5) (p < 0.01 in all areas). Alternately, wolves and moose were found at similar low elevations in both the Columbia A A Mountains (U [15,6] = 0.134,;? = 0.893) (Fig. 2) and the Quesnel Lake study areas (U [15,15] 20 •—Caribou —•—Wolf —*—Moose • 238 224 351 211 412 316 195 261 165 219 271 322 • 13 17 13 6 21 10 7 8 6 3 9 18 4 19 34 17 1 10 21 60 64 41 39 11 0 2100 i Late I Spring I Summer/Fall | EanV | Winter w i n t e r Month Fig. 2. Columbia Mountains radio-collared mountain caribou, wolf and moose median monthly elevations and interquartile ranges (1992-96 caribou and wolf; 1984-85 moose). -•-Caribou -•—Wolf • 113 85 127 122 123 122 141 159 105 110 125 63 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Late | Spring | Summer/Fall | Early | Winter w i n t e r Month Fig. 3. Wells Gray Provincial Park radio-collared mountain caribou and wolf median monthly elevations and interquartile ranges (1986-89). g -»—• > m LU H 110 • 14 4 18 2100 1900 1700 1500 1300 1100 900 700 500 Caribou Wolf —*— Moose 78 12 3 7 197 7 4 6 190 7 2 3 193 14 221 13 15 191 27 150 18 2 0 120 16 11 112 16 10 110 108 13 9 7 31 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Late Winter Spring Summer/Fall | Early | Winter Month Fig. 4. Quesnel Lake radio-collared mountain caribou, wolf and moose median monthly elevations and interquartile ranges (1984-89). • 83 • 12 48 123 14 19 31 7 Caribou Wolf 48 12 3 7 48 25 6 36 40 12 7 142 184 25 22 Feb Mar Late Winter May Spring Jul Aug Sep Summer/Fall Month Nov Dec Early Winter Fig. 5. Quesnel Highlands radio-collared mountain caribou and wolf median monthly elevations and interquartile ranges (1993-96). 22 = 1.14,/? = 0.255) (Fig. 4) during the early winter. During late winter, the elevations used by wolves in the highlands and mountains were generally lowest (Figs. 2,3,4 and 5) with wolves separate from caribou in each study area (all/? < 0.001). Wolves and moose in both the Columbia Mountains and the Quesnel Lake area used lower elevations than caribou during the late winter (Fig. 2 and 4), yet the elevations used by wolves and moose in the Columbia Mountains did not differ (U [59,71] = 0.113,/? = 0.91), A whereas those of wolves and moose at Quesnel Lake did (U [47,147] = 2.72, /? = 0.007). Seasonal intermedian distances of mountain caribou, wolves and moose In all seasons, in both the rugged Columbia Mountains and the highlands around Quesnel Lake, the median elevation of wolves was more closely associated with the median elevation of moose than that o f caribou (Table 7). Seasonal intermedian distances of wolves and moose ranged from 46 to 305 m in the Columbia Mountains, while those of wolves and caribou ranged between 198 and 975 m. A t Quesnel Lake, the wolf/moose seasonal intermedian distance ranged between 35 and 213 m, compared to a range of 347 to 747 m for wolves and caribou. Caribou and wolves in highland areas did not always have seasonal intermedian distances lower than those of caribou and wolves in the mountainous study areas. Only during spring and late winter was the Quesnel Lake (highland area) wolf/caribou intermedian distance less than that of wolves and caribou in Wells Gray Provincial Park (Table 7). A similar comparison between wolves and caribou in the Quesnel Highlands and the Columbia Mountains showed that only in late winter did wolves and caribou in the Quesnel Highlands have a seasonal intermedian distance less than that of wolves and caribou in the rugged Columbia Mountains (Table 7). 23 Table 7. Wolf/mountain caribou and wolf/moose seasonal intermedian distances (m) within study areas. Season Mountains Highlands Columbia Wells Gray Quesnel Quesnel Mountains Prov. Park Lake Highlands wolf-moose wolf-caribou wolf-caribou wolf-moose wolf-caribou wolf-caribou Spring 305 579 600 35 347 778 Summer/Fall 76 198 400 213 510 746 Early Winter 46 366 200 68 492 427 Late Winter 61 975 850 137 747 736 DISCUSSION The hypothesis that wolves show more elevational overlap with moose than with mountain caribou was supported in both highland and rugged mountains. W o l f scat analyses from these two areas further support these conclusions as the main diet item o f wolves throughout the year in both the Quesnel Lake area (Seip 1992b) and the Columbia Mountains (see chapter 4) was moose. Similar elevational relationships between wolves and moose were reported in northern British Columbia during the summer (Bergerud et al. 1984; Bergerud and Page 1987) and throughout the year in Wells Gray Provincial Park and the Quesnel Lake area (Seip 1992a). The hypothesis that mountain caribou in highland areas show less elevational separation from wolves than those l iving in rugged mountains was not clearly supported. Caribou and wolves in the highlands (Quesnel Lake; Quesnel Highlands) were as likely to use different elevations during each season as were caribou and wolves in rugged mountainous areas (Columbia Mountains; Wells Gray Provincial Park). Even during summer/fall, when wolves were most elevationally mobile and likely to prey upon mountain caribou (Seip 1992a), caribou in the rolling highland areas appeared to space further away from wolves than did caribou in the more mountainous areas. Despite the advantage of greater elevational range to space away from wolves, mountain caribou in the Columbia Mountains were the only population to exhibit 24 significant elevational overlap with wolves, and then, only during summer/fall. The Columbia Mountains caribou population has not declined and no w o l f predation o f radio-collared caribou has been observed to date (B .N . McLel lan , personal communication). This may indicate that caribou and wolves occupied geographically different areas during summer/fall, similar to caribou and wolves in Wells Gray Park (Seip 1990, 1992a). Wolves were reported as the main cause of caribou deaths in the highlands near Quesnel Lake (Seip 1992a); however, the elevations used by the two species differed during all four seasons. A brief overlap in elevational use by wolves and caribou may have occurred during the calving period (Bergerud et al. 1984; Bergerud and Ell iot 1986; Bergerud and Page 1987) and influenced the predation rate. This does not seem likely though as few adults at Quesnel Lake were kil led by wolves during this period (Seip 1992b) and wol f predation on calves was reported to be low in June (Seip 1992a). It appears that without significant elevational overlap, wolves caused a dramatic decline in the caribou population at Quesnel Lake as 80% (8/10) of the w o l f predation occurred at elevations below the median elevations of radio-collared caribou (see chapter 2). M y results suggest that elevational separation is inadequate in explaining the large variation in mountain caribou deaths attributed to wo l f predation in the highland and mountainous study areas examined (see chapter 2). Differences in wo l f densities (Bergerud and Ell iot 1986) and the relative densities of moose and caribou (Cumming et al. 1996; Seip and Cichowski 1996) may be the primary factors influencing the susceptibility o f mountain caribou to w o l f predation within highland and mountainous areas in any given season (see chapter 2). It would be useful i f future studies were conducted in both highland and mountainous areas to examine how different mountain caribou, moose and wol f densities influence the elevational relationships of these species. Additionally, where elevational separation between caribou and wolves does not occur, the geographic locations of caribou and wolves should be compared to determine i f any spatial 25 separation exists. These studies should also be of longer duration than those conducted thus far in order to ascertain the seasonal and annual variability within mountain caribou, w o l f and moose spatial interactions. 26 C H A P T E R 4 THE SEASONAL DIETARY IMPORTANCE OF MOOSE TO WOLVES IN THE COLUMBIA MOUNTAINS INTRODUCTION Early this century, the distribution and abundance of moose increased rapidly in parts of British Columbia (Edwards 1954; Hatter 1949; Spalding 1990). This was possibly due to the clearing and burning of forests for land settlement and railway development (Hatter 1949). Bergerud (1974) and Bergerud and Elliot (1986) suggested that the increased availability o f moose stimulated an increase in wolves that also preyed upon and reduced or eliminated caribou populations lacking an adequate antipredator strategy. With an ample alternate ungulate prey source available, a decrease in caribou would not necessitate reduced w o l f numbers (Bergerud 1988b; Seip 1992b). Wolves in the Quesnel Lake area were reported to be sustained primarily by moose (Seip 1992b); however, studies of the importance of moose in w o l f diet in other areas of mountain caribou range are scarce. The objective of this chapter is to test the hypothesis that wolves in mountain caribou range are sustained primarily by moose throughout the year. METHODS The seasonal importance of moose for two packs of radio-collared wolves in the Columbia Mountains was estimated from 79 wol f scats collected between December 1993 and November 1995. Scats were classified as either summer (May 1 -Nov 7) or winter (Nov 8 - A p r 30). In summer, scats were collected at den and rendezvous sites, and also monthly along an extensive network of logging roads. During winter, scats were collected at kill/scavenge sites, along roads, and while backtracking wolves in snow. Logging roads in the study area were checked just after snowmelt in the spring for winter-deposited scats (Forbes and Theberge 1992). Since coyotes 27 (Canis latrans) inhabited this study area, only scats > 30 mm in diameter (Weaver and Fritts 1979) or smaller scats associated with a wo l f den, rendezvous site or wo l f tracks were collected (Fuller and Kei th 1980). Hair from scats was identified to species using the methods of Adorjan and Kolenosky (1969), Kennedy and Carbyn (1981), and Spilborghs (1996). White-tailed and mule deer hair could not be separated. Frequently, multiple scats collected at the same k i l l or rendezvous site contain hair from the same k i l l and are not independent samples (Huggard 1991). In an attempt to minimize this bias, the effective sample size for these sites was taken as the number of prey species represented in the scats. The regression equation proposed by Weaver (1993) was used to estimate the relative weight of each prey species consumed by wolves, based on the frequency of occurrence of each prey species in scats. Paired t-tests ( S Y S T A T 1992) were used to determine whether moose was the primary diet item of wolves in each season. In testing the wol f diet hypothesis statistically, "primary diet item" was defined as being >50% of the biomass consumed by wolves. Levels of significance were set &tp < 0.05. RESULTS Frequency of occurrence of wolf diet items Moose occurred approximately twice as often in winter scats (50%) than in summer scats (22.0%), whereas the frequency of caribou remains in wol f scats differed little between summer and winter seasons, with each containing 14.0% and 18.2% respectively (Table 8). Deer remains were found only in winter wo l f scats and then, occurred infrequently (6.8%). Beaver, a common food item of wolves in the Columbia Mountains in the summer, comprised 42.0% o f the total diet. Although several other species were found in both summer and winter w o l f scats, each occurred infrequently (Table 8). 28 Table 8. Frequency of occurrence and percent biomass of wo l f diet items by season in the Columbia Mountains. Wolf diet Summer Winter items Frequency of % Biomass Frequency of % Biomass occurrence occurrence Caribou 7 18.9 8 12.7 Moose (adult) 7 42.0 22 77.5 Moose (calf) 4 5.7 - -Deer - - 3 3.4 Beaver 21 21.8 - -Other* 11 11.6 11 6.4 No. of scats 44 35 No. of items 50 44 Note: Percent biomass is based on the regression equation: y = 0.439 + 0.008x, where x is animal mass (Weaver 1993). Animal masses used were as follows: caribou, 124.7 kg (based on a ratio of 22 bulls : 78 cows : 21 calves (McLellan et al 1994b) and D.C. Thomas, unpublished data); adult moose, 343.4 kg (J. Woods, Banff NP unpublished data); moose calf in July, 40.0 kg (Banfield 1974); adult deer (mule and white-tailed combined), 74.2 kg (J. Woods, Banff NP unpublished data); beaver, 14.0 kg (Leege & Williams 1967). *Includes wolf, hoary marmot, red-backed vole, red squirrel, grouse, snowshoe hare, abattoir scraps, and deer mouse. Percent biomass of wolf diet items During winter, wolves in the Columbia Mountains were sustained primarily by moose (t-test = 4.12, df = 34, p < 0.001), with moose clearly the most important species (77.5%; Table 8). A l l five wo l f ki l ls found while backtracking wolves during the winter of 1994-95 were moose, confirming this result. Caribou, the second most prominent diet item, contributed 12.7% of the biomass consumed by wolves. Deer, the only other ungulate found in winter scats, constituted a mere 3.4% of w o l f diet during this season. Wolves in the Columbia Mountains were not sustained primarily by moose in the summer (t-test = -0.25, df = 43, p = 0.403), yet moose was still the most important species in the summer diet. Three prey items accounted for 88.4% of the biomass consumed. In order of importance these were: moose (47.7%; 42% adult, 5.7% calves), beaver (21.8%) and caribou (18.9%). 29 DISCUSSION Bergerud (1974) and Bergerud and Elliot (1986) suggested that in areas with both moose and caribou, wolves would prey primarily on moose. The estimate of winter w o l f diet items in the Columbia Mountains supported this hypothesis, but the estimate of summer wol f diet items did not. St i l l , moose was the main diet item of Columbia Mountains wolves throughout the year and as reported at Quesnel Lake (Seip 1992b), moose were clearly of greater dietary importance to wolves than caribou. Unfortunately, moose census data corrected for sightability were not available for the Columbia Mountains, so testing whether wolves used moose and caribou equal to their availability was not possible. Uncorrected moose census data did indicate that moose were common in the Columbia Mountains during the study (J. Krebs, personal communication). Wolves in the Columbia Mountains relied to a greater extent on beaver during the summer, and less on moose than wolves in the Quesnel Lake area (Seip 1992b). Differences in the relative availability of beaver (Theberge et al 1978) in the two areas may explain this variation, as wolves often prey heavily on beaver when available (Peterson 1977; Potvin et al. 1992; Voigt et al. 1976). It is also possible that the high occurrence of beaver in the sample o f summer w o l f scats is an artifact of collecting a disproportionate number (95%) of wo l f scats at lower elevations during this season. During summer, radio-collared wolves in the Columbia Mountains were located 25% of the time in roadless, high-elevation areas, and often these areas were not visited as the probability of finding w o l f scats at or near the bi-monthly relocation site of a radio-collared w o l f was low. Assuming wolves consumed diet items at elevations proportionally equal to those of their relocations, and that scats would be left in proximity to k i l l or scavenge sites (Huggard 1991), the summer w o l f scat sample may have been biased towards lower elevation prey species (i.e., beaver), and against higher elevation prey species (i.e., caribou and hoary marmots). Therefore, • 30 caution should be used when interpreting the estimates of wo l f summer diet in the Columbia Mountains. This estimate of w o l f diet in the Columbia Mountains and the Quesnel Lake w o l f diet estimate (Seip 1992b) support the view that wolves inhabiting mountain caribou range are largely sustained by moose. Both studies, however, lack sufficient data to test for annual and pack level variations in wol f diet across a range of moose and caribou densities. Future mountain caribou studies in which wolves are also radio-collared, should seek to address this inadequacy. 31 C H A P T E R 5 GENERAL CONCLUSION AND MANAGEMENT RECOMMENDATIONS Wolves were suggested to be the primary factor limiting mountain caribou although they were sustained largely by moose (Bergerud and Elliot 1986; Seip 1992a). Mountain caribou in areas of highland topography were suggested to be most susceptible to w o l f predation as, unlike those in more rugged mountains, they had less ability to space away from wolves and moose (Seip 1992a). The importance of moose in the seasonal diet of wolves was one element of the spatial separation model (Seip 1992a) investigated. W o l f scat contents from the Columbia Mountains supported the hypothesis that wolves in mountain caribou range are sustained primarily by moose. This was particularly evident during winter when few alternate prey species were available to wolves. Similar to wolves at Quesnel Lake (Seip 1992b), mountain caribou were a minor part of Columbia Mountains wol f diet. Since moose appear to be the main prey o f wolves in southern B . C . and w o l f abundance may be strongly associated with the density of moose, attempts should be made to maintain moose densities at their present levels, particularly in highland areas. The importance of wo l f predation as a mountain caribou mortality factor was another element of Seip's (1992a) model that was examined. The caribou mortality data indicated that w o l f predation on mountain caribou was higher at Quesnel Lake than in the Columbia Mountains, supporting the hypothesis that wo l f predation on mountain caribou is greater in highland areas than in mountainous areas. There was no evidence though indicating that caribou in highland areas continually experience high levels of wo l f predation. Possibly wolves prey on mountain caribou in a density-dependent manner (Seip and Cichowski 1996). Mountain caribou in rugged mountains experience limited wol f predation while avalanches, bear, wolverine, and cougar 32 predation appear as more common mortality factors. Acquiring a more comprehensive understanding of the importance o f wo l f predation on mountain caribou populations in both highland and mountainous areas w i l l require more caribou mortality data than are presently available. Radio-collaring more caribou in and o f itself, w i l l not necessarily increase the sample size of caribou deaths with known causes. Increasing the frequency of successive relocation flights (Heisey and Fuller 1985) and reducing response times to caribou mortality sites should help in differentiating among possible mortality factors, lessen the number of caribou deaths of unknown cause and increase sample sizes required to further test the w o l f predation hypothesis. The remaining element of the spatial separation model (Seip 1992a) investigated was the degree of elevational separation between wolves and moose, and wolves and mountain caribou in highland and mountainous areas. The radio-telemetry data supported the hypothesis that wolves show more elevational overlap with moose than with mountain caribou. N o support, however, was demonstrated for the hypothesis that mountain caribou in highland areas show less elevational separation from wolves than did caribou in rugged mountains. Differences in wol f density (Bergerud and Ell iot 1986) and the relative densities of moose and caribou (Cumming et al. 1996; Seip and Cichowski 1996) may explain the large variation in the number of mountain caribou mortalities attributed to wol f predation in the highland and mountainous study areas examined to date. Although the wolf/caribou radio-telemetry data used in this thesis incorporates all the current spatial data simultaneously collected on these species, it explains the elevational relationships of wolves and caribou within a relatively short period of time. To determine the annual and seasonal variation in the elevational interactions of mountain caribou, wolves, and moose it is recommended that future studies addressing this issue be of longer duration. Further, to fully test 33 the spatial separation model (Seip 1992a), the elevational and geographic distributions of mountain caribou, wolves and moose should both be examined. Additionally, the degree to which habitat alteration (timber harvesting and fires) influences wolf, mountain caribou, and moose densities and ultimately, their interactions is still largely unknown. Research directed towards understanding how wolves and their prey respond numerically and spatially to different landscape changes in both highland and rugged mountainous areas should be a primary component of future wolf/caribou/moose radio-telemetry studies. 34 LITERATURE CITED Adorjan, A . S . , and Kolenosky, G . B . 1969. A manual for the identification of hairs of selected Ontario mammals. Ont. Dept. of Lands and Forests. Toronto, Ont. Anderson, R . M . 1938. The present status and distribution of the big game mammals of Canada. Transactions of the 3rd North American W i l d l . Conf. 3: 390-406. Antifeau, T .D . 1987. The significance of snow and arboreal lichen in the winter ecology of mountain caribou {Rangifer tarandus caribou) in the north Thompson watershed of Brit ish Columbia. M S c . thesis, University of British Columbia, Vancouver. Banfield, A . W . F . 1974. The mammals of Canada. University of Toronto Press, Toronto. Bergerud, A . T . 1974. Decline of caribou in North America following settlement. J. W i l d l . Manage. 38: 757-770. Bergerud, A . T . 1978. Caribou. In B i g game of North America, ecology and management. Edited by J .L . Schmidt, and D . L . Gilbert. Stackpole Press. Harrisburg, P A . pp. 83-101. Bergerud, A . T . 1985. Antipredator strategies of caribou: dispersion along shorelines. Can. J. Zool . 63: 1324-1329. Bergerud, A . T . 1988a. Caribou declines in central and southern British Columbia. In Caribou Research and Management in B . C . : Proceedings of a Workshop, Kamloops, B . C . , November 6 and 7, 1985. Edited by R. Page. B . C . Minist . o f Forests, pp. 201-225. Bergerud, A . T . 1988b. Caribou, wolves and man. Trends in Ecology and Evolution. 3: 68-72. Bergerud, A . T . 1996. Evolving perspectives on caribou population dynamics, have we got it right yet? Rangifer, Special Issue 9: 95-116. Bergerud, A . T . , Butler, H . E . , and Mil le r , D .R. 1984. Antipredator tactics of calving caribou: dispersion in mountains. Can. J. Zool . 62: 1566-1575. Bergerud, A . T . , and Elliot, J.P. 1986. Dynamics of caribou and wolves in northern British Columbia. Can. J. Zool . 64: 1515-1529. Bergerud, A . T . , Ferguson, R., and Butler, H . E . 1990. Spring migration and dispersion o f woodland caribou at calving. A n i m . Behav. 39: 360-368. Bergerud, A . T . , and Page, R . E . 1987. Displacement and dispersion of parturient caribou at calving as antipredator tactics. Can. J. Zool . 65: 1597-1606. 35 Bloomfield, M . 1980. The impact of development, settlement and associated activities on mountain caribou in central British Columbia, Canada. In Proceedings of the 2nd International Reindeer/Caribou Symposium. Edited by E . Reimers, E . Gaare and S. Skjenneberg. Direktoratet for vilt og ferskvannsfisk, Trondheim, Norway, pp. 705-715. Boertje, R . D . , and Stephenson, R .O . 1992. Effects of ungulate availability on w o l f reproductive potential in Alaska. Can. J. Zool . 70: 2441-2443. B . C . Ministry of Environment. 1985. Summary of snow survey measurements in British Columbia 1935-1985. Water Mgt. Branch. Victoria, B . C . Compton, B . B . , Zager, P., and Servheen, G . 1995. Survival and mortality of translocated woodland caribou. W i l d l . Soc. B u l l . 23: 490-496. Coupe, R., Stewart, A . C , and Wikeem, B . M . 1991. Engelmann spruce-subalpine fir zone. In Ecosystems o f British Columbia. Edited by D . Meidinger and J. Pojar. Special Report Series N o . 6. B . C . Minist . Forests. Victoria, B . C . pp. 223-236. Cumming, H . G . , Beange, D . B . , and Lavoie, G . 1996. Habitat partitioning between woodland caribou and moose in Ontario: the potential role of shared predation risk. Rangifer, Special Issue 9: 81-94. Dale, B . W . , Adams, L . G . , and Bowyer, R .T . 1994. Functional response o f wolves preying on barren-ground caribou in a multiple-prey ecosystem. J. Animal Ecology. 63: 644-652. Edwards, R . Y . 1954. Fire and the decline of a mountain caribou herd. J. W i l d l . Manage. 18: 521-526. Edwards, R . Y . , and Ritcey, R . W . 1960. Foods of caribou in Wells Gray Park, British Columbia. Can. Field-Nat. 74: 3-7. Farnell, R. , Barichello, N , Eg l i , K . , and Kuzyk, G . 1996. Population ecology of two woodland caribou herds in the southern Yukon. Rangifer, Special Issue 9: 63-72. Forbes, G.J . , and Theberge, J .B. 1992. Importance of scavenging on moose by wolves in Algonquin Park, Ontario. Alces. 28: 235-241. Fuller, T . K . , and Keith, L . B . 1980. W o l f population dynamics and prey relationships in northeastern Alberta. J. W i l d l . Manage. 44: 583-602. Fuller, T . K . , and Snow, W.J . 1988. Estimating winter w o l f densities using radiotelemetry data. W i l d l . Soc. B u l l . 16: 367-370. Harper, B . , Cannings, S., Fraser, D . , and Munro, W.T. 1994. Provincial lists of species at risk. In Biodiversity in British Columbia. Edited by L . E . Harding and E . M c C u l l u m . Environment Canada/Canadian Wildlife Service. Victoria, B . C . pp. 16-23. 36 Hatter, J. 1949. The status of moose in North America. Transactions o f the 14th North American W i l d l . Conf. 14:492-501. Hayes, R . D . 1995. Numerical and functional responses o f wolves, and regulation of moose in the Yukon. Msc . thesis, Simon Fraser University, Burnaby. Heisey, D . M . , and Fuller, T . K . 1985. Evaluation of survival and cause-specific mortality rates using telemetry data. J. W i l d l . Manage. 49: 668-674. Holland, S.S. 1976. Landforms of British Columbia: a physiographic outline. B u l l . 48. B . C . Dep. Mines and Petroleum Resour. Victoria, B . C . Huggard, D.J . 1991. Prey selectivity of wolves in Banff National Park. M S c . thesis, University of British Columbia, Vancouver. Kennedy, A . J . , and Carbyn, L . N . 1981. Identification of w o l f prey using hair and feather remains with special reference to western Canadian national parks. Canadian Wildl i fe Service, Edmonton, Al ta . Ketcheson, M . V . , Braumandl, T.F. , Meidinger, D . , Utzig, G . , Demarchi, D . A . , and Wikeem, B . M . 1991. Interior cedar-hemlock zone. In Ecosystems of Brit ish Columbia. Edited by D . Meidinger and J. Pojar. Special Report Series No . 6. B . C . Minist . Forests. Victoria, B . C . pp. 167-181. Krebs, J.R., and McCleery, R . H . 1984. Optimization in behavioural ecology. In Behavioural ecology: an evolutionary approach. Edited by J.R. Krebs and N . B . Davies. 2nd ed. Blackwel l Scientific Publications, Oxford, pp. 91-121. Leege, T . A . , and Will iams, R . M . 1967. Beaver productivity in Idaho. J. W i l d l . Manage. 31: 326-332. MacArthur, R . H . , and Pianka, E .R. 1966. On optimal use of a patchy environment. The Amer. Naturalist. 100: 603-609. McLel lan , B . N . , Flaa, J., and Super, M . 1994a. Habitats used by mountain caribou in the North Columbia Mountains. Prelim. Rep. Year 2. B . C . Minist . Forests. Revelstoke, B . C . McLe l l an , B . N . , Flaa, J., and Woods, J. 1994b. Mountain caribou censuses in the North Columbia Mountains. B . C . Minist . of Forests. Revelstoke, B . C . Mayfield, H . 1975. Suggestions for calculating nest success. Wilson B u l l . 87: 456-466. Meidinger, D . , Pojar, J., and Harper, W . L . 1991. Sub-Boreal Spruce Zone. In Ecosystems of Brit ish Columbia. Edited by D . Meidinger and J. Pojar. Special Report Series No . 6. B . C . Minist . Forests. Victoria, B . C . pp. 209-221. 37 Messier, F . 1985. Social organization, spatial distribution, and population density of wolves in relation to moose density. Can. J. Zool . 63: 1068-1077. Messier, F . 1994. Ungulate population models with predation: a case study with the North American moose. Ecology. 75: 478-488. Messier, F . 1995. On the functional and numerical responses of wolves to changing prey density. In Ecology and conservation of wolves in a changing world. Edited by L . Carbyn, S .H. Fritts, and D .R . Seip. Canadian Circumpolar Institute, University of Alberta, Edmonton, pp. 187-197. Munro, J .A. 1947. Observation of birds and mammals in central British Columbia. Occasional Papers of the British Columbia Museum. No . 6. Packard, J . M . , and Mech, L . D . 1980. Population regulation in wolves. In Biosocial mechanisms of population regulation. Edited by M . N . Cohen, R.S . Malpass, and H . G . Kle in . Yale University Press. Peterson, R . O . 1977. W o l f ecology and prey relationships on Isle Royale. U . S . Natl . Park Serv. Sci . Monogr. Ser. No. 11. Potvin, F. , Breton, L . , Pi lon, C , and Macquart, M . 1992. Impact of an experimental w o l f reduction on beaver in Papineau-Labelle Reserve, Quebec. Can. J. Zool . 70: 180-183. Rominger, E . M . , and Oldemeyer, J .L. 1990. Early-winter diet of woodland caribou in relation to snow accumulation, Selkirk Mountains, British Columbia, Canada. Can. J. Zool . 68: 2691-2694. Seaman, D . E . , and Powell , R . A . 1991. Kernel home range estimation program. North Carolina State University. Seip, D .R . 1990. Ecology of woodland caribou in Wells Gray Provincial Park. B . C . Minist . Environ. W i l d . B u l l . No . B-68. Seip, D .R. 1991. Predation and caribou populations. Rangifer, Special Issue 7: 46-52. Seip, D .R. 1992a. Factors limiting woodland caribou populations and their interrelationships with wolves and moose in southeastern British Columbia. Can. J. Zool . 70: 1494-1503. Seip, D .R . 1992b. Habitat use and population status of woodland caribou in the Quesnel Highlands, British Columbia. B . C . Minist . Environ. W i l d . B u l l . No . B-71. Seip, D.R. , and Cichowski, D . B . 1996. Population ecology of caribou in British Columbia. Rangifer, Special Issue 9: 73-80. Servheen, G . , and Lyon, L . J . 1989. Habitat use by woodland caribou in the Selkirk Mountains. J. W i l d l . Manage. 53: 230-237. 38 Siegel, S., and Castellan, N J . 1988. Nonparametric statistics for the behavioural sciences. 2nd ed. M c G r a w - H i l l Book Co. , New York. Simpson, K . , Kelsal l , J.P., and Leung, M . 1996. Integrated management of mountain caribou and forestry in southern British Columbia. Rangifer, Special Issue 9: 153-158. Simpson, K . , Terry, E . , and Hamilton, D . 1997. Toward a mountain caribou management strategy for Brit ish Columbia - habitat requirements and sub-population status. B . C . Minist . Environ. W i l d l . Working Rep. WR-90. Simpson, K . , and Woods, G.P. 1987. Movements and habitats of caribou in the mountains of southern British Columbia. B . C . Minist . Environ. W i l d l . B u l l . No . B-57. Simpson, K . , Woods, G.P. , and Hebert, K . B . 1987. Critical habitats of caribou {Rangifer tarandus caribou) in the mountains of southern British Columbia. In Proceedings of Second North American Caribou Workshop, V a l Mor in , Quebec, October 17-20, 1984. Edited by T.C . Meredith and A . M . Martell. Centre for Northern Studies and Research, M c G i l l University, Montreal, pp. 177-191. Spalding, D.J . 1990. The early history of moose {Alces alces): distribution and relative abundance in British Columbia. Contributions to Natural Science, Royal B . C . Museum. Victoria, B . C . Spilborghs, S. A . 1996. Analysis of cuticular scale casting techniques and identification key of seven species of west coast B . C . mammals. B.Sc. thesis, University of British Columbia, Vancouver. Stevenson, S.K. , and Hatler, D . H . 1985. Woodland caribou and their habitat in southern and central Brit ish Columbia. B . C . Minist . Forests. Victoria, B . C . S Y S T A T . 1992. S Y S T A T for windows: statistics. Version 5. S Y S T A T , Evanston, 111. Terry, E X . 1994. Winter habitat selection and foraging patterns of mountain caribou. Msc . thesis, University of British Columbia, Vancouver. Terry, E . , McLe l l an , B . , Watts, G . , and Flaa, J. 1996. Early winter habitat use by mountain caribou in the North Cariboo and Columbia Mountains, British Columbia. Rangifer. Special Issue 9: 133-140. Theberge, J .B. , Oosenbrug, S . M . , and Pimlott, D . H . 1978. Site and seasonal variations in food o f wolves, Algonquin Park, Ontario. Can. Field-Nat. 92: 91-94. Thoreau, H . D . 1854. Walden; or, life in the woods. Ticknor and Fields, Boston. Voigt , D.R. , Kolenosky, G .B . , and Pimlott, D . H . 1976. Changes in summer foods of wolves in central Ontario. J. W i l d l . Manage. 40: 663-668. 39 Weaver, J .L. 1993. Refining the equation for interpreting prey occurrence in gray w o l f scats. J. W i l d l . Manage. 57: 534-538. Weaver, J .L. , and Fritts, S.H. 1979. Comparison of coyote and wol f scat diameters. J. W i l d l . Manage. 43: 787-788. White, G .C . , and Garrott, R . A . 1990. Analysis of wildlife radio-telemetry data. Academic Press. San Diego. Young, J .A. , VanSpall , K . , and Dielman, P .W. 1996. Towards integrated management solutions: The Quesnel Highlands caribou project. Progress Rep. 1993-96. B . C . Minist . Environ. Wil l iams Lake, B . C . 40 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.831.1-0088482/manifest

Comment

Related Items