Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Relationships between coyote ecology and sheep management in the Lower Fraser Valley, B.C. Atkinson, Knut Thomas 1985

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

Item Metadata


831-UBC_1985_A6_7 A84.pdf [ 3.85MB ]
JSON: 831-1.0095964.json
JSON-LD: 831-1.0095964-ld.json
RDF/XML (Pretty): 831-1.0095964-rdf.xml
RDF/JSON: 831-1.0095964-rdf.json
Turtle: 831-1.0095964-turtle.txt
N-Triples: 831-1.0095964-rdf-ntriples.txt
Original Record: 831-1.0095964-source.json
Full Text

Full Text

RELATIONSHIPS BETWEEN COYOTE ECOLOGY AND SHEEP MANAGEMENT IN THE LOWER FRASER VALLEY, B.C. BY KNUT THOMAS ATKINSON B.Sc, The University of British Columbia, 1979 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Department of Animal Science) We accept this thesis as conforming to the required/lstandard THE UNIVERSITY OF BRITISH COLUMBIA FEBRUARY 1985 © Knut Thomas Atkinson, 1985 In presenting t h i s thesis i n p a r t i a l f u l f i l m e n t of the requirements for an advanced degree at the University of B r i t i s h Columbia, I agree that the Library s h a l l make i t f r e e l y available for reference and study. I further agree that permission for extensive copying of t h i s thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. I t i s understood that copying or publication of t h i s thesis for f i n a n c i a l gain s h a l l not be allowed without my written permission. Department of Animal Science  The University of B r i t i s h Columbia 1956 Main Mall Vancouver, Canada V6T 1Y3 Date February 21 1985 DE-6 (.3/81) i i ABSTRACT Domestic sheep farmers in the lower Fraser Valley (L.F.V.) had reported increasing losses of sheep to coyote (Canis latrans) and dog (C. familiaris) predation. The objectives of this study were: (1) to determine if manage-ment and geographic factors predispose sheep farms to coyote and dog preda-tion; (2) to assess the relative impact of coyote and dog predation on the L.F.V. sheep population; (3) to record basic attributes of coyote biology (taxonomy, reproduction, food habits, home range, movements, activity patterns, and predatory behaviour); (4) to provide practical and economical recommendations to reduce or prevent coyote and dog predation on sheep in the L.F.V. One hundred and twelve sheep farmers were interviewed over three years, 1979 to 1981. Farms which lost sheep to coyotes characteristically had relatively large flocks (>50 ewes) on large fields (4+ ha), did not confine sheep at night, and either buried or left sheep carcasses exposed. There were no common factors among farms which lost sheep to dogs. Predation accounted for 28.2% of all mortality and 2.4% of the total population sampled. Coyotes killed 69.7% and 74.7% of all ewes and lambs lost to predators. An average of 24.3% of the farms lost sheep to coyotes and dogs each year. However, 55.2% of the farms which lost sheep to coyotes did so in two or three consecutive years compared to 17.4% of farms which lost sheep to dogs. Coyotes in the L.F.V. were similar in most biological aspects studied to other coyote populations in North America. The only exception was that small rodents, primarily Microtus townsendi composed over 70% (scat volume) i i i of their diet, a proportion higher than in other areas. Domestic livestock (mostly poultry carrion) comprised only 4.3% of the diet, sheep only 0.2%. I concluded that in the rural-urban L.F.V. interface, prevention of coyote predation (and secondarily dog predation) on hobby farms is largely a matter of management. The most effective and economical solution is to provide predator-proof enclosures for night confinement of sheep because coyotes were most active at night. This method could be further enhanced by removing livestock carcasses off the farm or by burying and liming them to avoid attracting coyotes to the farm vicinity. i v TABLE OF CONTENTS Page Abs t r a c t i i Table of Contents i v L i s t o f Tables v i i L i s t of Figures v i i i Acknowledgements x I n t r o d u c t i o n 1 Study Area 3 Methods 6 Farm C h a r a c t e r i s t i c s 6 Animal V a r i a b l e s 6 Management V a r i a b l e s 7 Geophysical and Ve g e t a t i o n a l V a r i a b l e s 8 A n a l y s i s 8 Causes of Sheep M o r t a l i t y 11 Coyote Biology 11 Taxonomy 11 Reproduction 12 Food Habits 12 Telemetry 13 Home Range 14 A c t i v i t y Patterns and Movements 15 Pre d a t i o n by R a d i o - c o l l a r e d Coyotes 15 V Page Results 16 Farm Characteristics 16 Sample Size of Farms 16 Coyotes and Dogs, 1979 16 Coyotes, 1980 17 Dogs, 1980 17 Coyotes, 1981 17 Dogs, 1981 19 Mortality 21 Ewe Mortality 21 Lamb Mortality 22 Coyote Biology 27 Reproduction 27 Food Habits 30 Home Range 33 Activity Patterns and Movement 40 Predation by Radio-collared Coyotes 41 Discussion 42 Farm Characteristics Associated with Predator Losses 42 Coyote Predation 42 Dog Predation 44 Coyote and Dog Predation: Summary 45 Mortality 46 vi Page Coyote Biology 49 Taxonomic Status 49 Reproduction 50 Food Habits 51 Food Habits and Livestock Predation 54 Home Range and Movement from Home Range 56 Activity Patterns and Movement 57 Predation by Radio-collared Coyotes 59 Predation and Sheep 59 Control of Coyote Predation 60 Control of Dog Predation 65 Conclusions 66 Recommendations 67 Literature Cited 68 Appendix i 77 Appendix i i 79 v i i LIST OF TABLES Page 1. Factors contributing to coyote predation on sheep in the Lower Fraser Valley 1979-1981. 18 2. Factors contributing to dog predation on sheep in the Lower Fraser Valley 1979-1981. 20 3. Percent loss (by factor) of sheep mortality in the Lower Fraser Valley 1979-1981. 23 4. Percent of total sheep sample lost to mortality factors in the Lower Fraser Valley 1979-1981. 24 5. Radiotelemetry status of Lower Fraser Valley coyotes. 34 v i i i LIST OF FIGURES Page 1. Map of study area 5 2A. Proportional mortality factors among ewes 1979-1981 (n=327). 26 2B. Proportional mortality factors among lambs 1979-1981 (n=977). 26 3A. Monthly frequency of occurrence of predation upon ewes by coyotes (n=79) and dogs (n=30) in the Lower Fraser Valley 1979-1981. Totals are less than mortality reported in Table 4 as some predation loss dates were unknown. 29 3B. Monthly frequency of occurrence of predation upon lambs by coyotes (n=227) and dogs (n=60) in the Lower Fraser Valley 1979-1981. Totals are less than mortality reported in Table 4 as some predation loss dates were unknown. 29 4A. Composite diet of coyotes (percent scat volume) in the Lower Fraser Valley 1980-1981 (n=862). The "mammals" class consists of unidentified mammals, and identified wild mammals other than rabbits and small rodents. See text for species. 32 4B. Seasonal diet of coyotes (percent scat volume) in the Lower Fraser Valley 1980-1981 (n=862). 32 5A. Area utilized by transient male M3. Numerals refer to plotted points of every third radio relocation. 36 5B. Home ranges of females Fl (transient) and F3 (resident-transient). 36 5C. Home ranges of females F2 and F4 (residents), F5 (transient), and male M6 (resident). 38 ix LIST OF FIGURES (Cont'd) Page 5D. Home ranges of males Ml and M2 (resident-transients), M4, M5, M7 (residents), and M8 (transient). The home ranges designated as M8 and M8! are the areas he consecutively utilized. 38 X ACKNOWLEDGEMENTS With his resourcefulness i n obtaining funding and his often pung-ent comments, Dr. D. Shackleton kept this project going. Thank you very much Shack, I w i l l always appreciate your help and guidance. Project funding was provided by grants to Dr. D.M. Shackleton from the B.C. Science Council, the B.C. Ministry of Agriculture and Food, and A g r i c u l t u r e Canada. Personal funding was provided by a B.C. Science Council G.R.E.A.T. award and through the e f f o r t s of T. Burgess and R. Forbes of the B.C. W i l d l i f e Branch. The cooperation of the B.C. W i l d l i f e Branch was greatly appreciated. D. Pemble and D. Nagy of Conservation O f f i c e r Services, were invaluable instuctors i n the art of trapping, and in novel approaches to f i e l d research. B. Hunt ( p i l o t ) , S. Chestnutt, B. Wong, and A. Tiplady ( a s s i s t a n t s ) , Dr. A. Harestad and M. Gillingham (computers), Drs. A. Anderson and I. White ( v e t e r i n a r i a n s ) , Dr. F. Messier ( g a d f l y ) , E. Brown ( d r a f t i n g ) , and B. Foster ( e d i t i n g ) are a l l thanked for t h e i r contributions. Mrs. J . Tiplady i s e s p e c i a l l y thanked for her many long hours hunched over the word processor. The farmers who granted me interviews and trapping r i g h t s , and p a r t i c u l a r l y L t . Cmdr. P. Johnston (Canadian Forces Base. Aldergrove), a l l gave e s s e n t i a l aid to the study. Mrs. M-L. White, Mrs. J . Tiplady, Mrs. E. Shackleton, Mrs. A. Foster, and my l a t e grandmother, Mrs. K. Anderson, kindly provided many meals to supplement my d i e t . x i ACKNOWLEDGEMENTS (Cont'd) F i n a l l y , thanks to my parents, S a l l i e and Murray Atkinson, not only for the i r support though the project, but e s p e c i a l l y for lending me enough money to st a r t my pr o j e c t . And of course, K i l i (meow) for not complaining when I spent my days o f f for three years working on my thesis i n Vancouver. 1 INTRODUCTION Coyotes (Canis latrans) are found throughout most of North America (Young and Jackson 1951) with their food habits (Bergeron and Demers 1981; Gier 1968; Hawthorne 1972; Hilton 1976) and social organization (Althoff and Gipson 1981; Bowen 1981; Messier and Berrette 1979) varying greatly through-out their range. It is not surprising then that their prey includes domes-tic animals which may be locally more available or vulnerable than native species (Kauffeld 1977), thus bringing coyotes into conflict with farmers. Except for some early descriptive studies (Bond 1939; Criddle et al. 1923; Murie 1935; Sperry 1939), Sperry's 1941 study of food habits appears to be the first quantitative research into coyote biology and livestock predations. The banning of predacides in the United States in combination with the publication of the Cain Report (Cain et a l . 1972) stimulated much interest in coyote research. Since 1972 studies of coyote predation on livestock (Meduna 1977) and wild prey (Messier and Barrette 1979), food habits (Johnson and Hansen 1979), reproduction (Kennelly et al. 1977), popu-lation dynamics (Todd et a l . 1981), behaviour (Camenzind 1978), and predator control (Connolly and Longhurst 1975), have proceeded to gather data on which management decisions could be based. In response to public opinion, predator control research has emphasized evaluation of nonlethal methods. These include guard dogs (reviewed by Green et al. 1984), taste aversion (Bourne and Dorrance 1982, Gustavson et al. 1982), other repellents (Lehner et al. 1976), predator-proof fencing (Dorrance and Bourne 1980; Linhart et al. 1982), and husbandry practices (Jones and Woolf 1983; Robel et a l . 1981; Todd and Keith 1976). Alterations in husbandry practices were advocated by Dorrance (1983) as being easiest to implement on small farm flocks where coyote predation may 2 be easier to prevent than on large ranges. Significantly, two reports which intensively studied husbandry practices and coyote predation (Jones and Woolf 1983, Meduna 1977) were both involved with livestock on small farms. In the Lower Fraser Valley (L.F.V.) of southwestern B.C. sheep produc-tion is largely restricted to hobby farms; farms on two to four ha, typi-cally with 20 to 30 ewes. Predation on small sheep flocks in a rural-residential setting, factors which may predispose farms to coyote predation, and the food habits and movements of coyotes in such an area, are a combina-tion that have not been intensively studied in Canada. Published results of coyote research in North America are from different ecological zones where coyotes are usually long established members of the local fauna. Coyotes have only been found in the L.F.V. since the 1930's (Young and Jackson 1951). Aspects of their biology and impacts on domestic livestock are rela-tively unknown. Coyotes also interact with domestic dogs (C_. familiaris) in this rural residential area. The extent of breeding between coyotes and dogs and the extent of dog predation on sheep are also essentially unknown in the L.F.V. Because of the lack of understanding of the interrelationships of coyote predation, food habits and movements, and their interactions with sheep farm management practices and geographic factors, the following study was undertaken. Its objectives were to determine: (1) i f management and geographic factors predispose sheep farms to coyote and dog predation; (2) the relative impact of coyote and dog predation on the L.F.V. sheep popula-tion; (3) basic attributes of coyote biology (taxonomy, reproduction, food habits, home range, movements, activity patterns, and predatory behaviour) in the study area; (4) practical and economical recommendations to reduce or prevent coyote and dog predation on sheep in the L.F.V. 3 STUDY AREA The study was conducted between January 1979 and October 1981 in an area located on both sides of the Fraser River in southwestern British Columbia, between Deroche in the east, and the Strait of Georgia in the west, but excluding the metropolitan area- surrounding Vancouver (Figure 1). This area is termed the Fraser Low-land and occurs within the coastal trough of the western Canadian Cordillera (Holland 1976). It is characterized geologically by flat and gently dipping sedementary rocks dating to the Cretaceous Age, and topographically by flat-bottomed valleys up to 5 km wide, separated by extensive hills of low relief, ranging in elevation from 15 to 305 m above sea level (Armstrong 1967, cited in Holland 1976). The study area lies in the wetter subzone of the coastal Douglas Fir Zone, characterized by Douglas f i r (Pseudostuga menziesii), western red cedar (Thuja plicata), hemlock (Tsuja heterophylla) and red alder (Alnus  rubra) (Krajina 1969), but much land has been cleared for agriculture and urban expansion. The soils of upper elevation forested areas are character-ized by humo-ferric podzols, and the flood plains by humic gleyzols (Valen-tine et al. 1978). The climate is described by Krajina (1969) as meso-thermal marine. The mean annual temperature averages 8.9°C, ranging from mean monthly tempera-tures of 1.1°C in January, to 16.1°C in July and August. Annual precipita-tion averages 1643 mm, 5.8% of which is composed of wet snow (B.C. Ministry of Agriculture 1976). The human population is concentrated in the metropolis of Vancouver and adjacent municipalities to the west. Eastern municipalities have progres-sively cleared and utilized more land for agricultural purposes and support sparser human densities. 4 Figure 1. Map of Study Area 6 METHODS Farm Characteristics Data on farm characteristics were obtained by contacting and personally interviewing every second sheep farmer listed in the 1978 edition of the Membership Directory of the Lower Mainland Sheep Producers Association (L.M.S.P.A.) [n=187], and every producer within the home range of a radio-collared coyote (see Telemetry later). In both 1980 and 1981, information on sheep losses and any differences in management practices (between years) was obtained for the previous and current year (at farms visited for the first time). Interviews took place between June and September each year, after lambing and during the period of greatest predator losses. Informa-tion was collected on 28 variables using the questionnaire shown in the Appendix I. Animal Variables [No. animals, No. poultry, No. dogs, No. ewes, No. lambs, breed of sheep, ewe losses, lamb losses, flock size, and sheep density]. All estimates of animal numbers were obtained from farmers or from their records. Numbers of animals (farm animals other than sheep, dogs, or poultry) and poultry were collected once and assumed constant for the dura-tion of the study. Sheep numbers were obtained each year. The breed was defined as that of the majority of the ewe flock. Flock size was derived by summing ewe and lamb numbers. Density was obtained by dividing flock size by hectare (ha) of sheep pasture. 7 Management Variables [Month of lambing, confinement, l i g h t s , b e l l s , working dog, guard dog, carcass di s p o s a l , fence type, and fence condition.] Management practices were determined through conversation with the producer and personal observation. The month when producers lambed most of t h e i r sheep was used as the peak month of lambing. Sheep confinement at night was defined as t o t a l confinement i n a paddock or building every night. If sheep were confined only i n t e r m i t t e n t l y , seasonally, or a f t e r predator attacks, they were not considered confined for a n a l y t i c a l pur-poses. Thus, confinement was defined on a yes/no basis, as were the use of l i g h t s , b e l l s , working dogs, and guard dogs. Separation of working and guard dogs from pets was based on the producers b e l i e f i n the dogs perfor-mance as such, not by the breed of dog. Fence type was c l a s s i f i e d into f i v e general categories (pagewire, board, barbed-wire, no fence, and other). Fence condition was evaluated on whether I considered i t a b a r r i e r to canid entrance. Condition was gauged by walking along the fence perimeter and examining fence continuity, height and tension of the lowest wire above the ground, instances of coyotes or dogs digging and crawling under the fence, and whether ditches or g u l l i e s crossed by the fence were blocked by fencing. As coyotes seldom jump fences (D. deCalesta, pers. comm.), and dogs were only found to crawl under fences (pers. observ.), fence height was not considered. 8 Geophysical and V e g e t a t i o n a l V a r i a b l e s [No. ha of sheep pasture, water on property, type of cover, height of cover, pasture to cover d i s t a n c e , minimum and maximum distance to pasture, topography, and distance to settlement] These v a r i a b l e s were recorded through i n t e r v i e w s , personal o b s e r v a t i o n , or map a n a l y s i s . Water was defined as the presence or absence of a perman-ent body of water on or a b u t t i n g the farm. The type and height of f o r e s t cover on or adjacent to the farm was recorded at the time of i n i t i a l con-t a c t , and assumed constant between years. The c l a s s i f i c a t i o n of cover type i s described i n Appendix i i . Cover was defined as the nearest 1.5 ha or l a r g e r treed area. The distance to cover was paced out or d r i v e n to by car and the odometer reading recorded. Minimum distance to pasture was the paced distance from the nearest farm b u i l d i n g to the c l o s e s t point of the sheep pasture. Conversely, maximum distance to pasture was the paced d i s -tance to the f u r t h e s t point of the pasture. Topography was based on an a r b i t r a r y s c a l e of 1 to 5, f l a t to h i l l y . The distance to the nearest sub-d i v i s i o n , v i l l a g e or any other l a r g e urban settlement was measured on 1974 G e o l o g i c a l Survey of Canada 1:50,000 s c a l e topographic maps, which were p e r s o n a l l y updated to show new s u b d i v i s i o n s . A n a l y s i s Although the 1981 data c o l l e c t i o n period ended on September 30 i t was tr e a t e d as a complete year, because by t h i s time almost a l l predator r e l a t e d deaths would have occurred (Robel et a l . 1981). A l l continuous v a r i a b l e s (e.g. d i s t a n c e to cover) were p l o t t e d on scattergrams to determine t h e i r d i s t r i b u t i o n , then categorized according to t h e i r d i s t r i b u t i o n , and included 9 w i t h the o r i g i n a l c a t e g o r i c a l v a r i a b l e s (e.g. fence type) f o r a n a l y s i s . F i n a l a n a l y s i s was performed using the BMDP Stepwise L o g i s t i c Regression package (Engelman 1981) w i t h the p r o b a b i l i t y of a farm l o s i n g sheep to coyotes or dogs as the dependent v a r i a b l e . Predation was treated as a bina r y (success: f a i l u r e ) v a r i a b l e . Thus, magnitude of losses to coyotes or dogs was not the dependent v a r i a b l e , and a farm l o s i n g 20 sheep was t r e a t e d the same as a farm l o s i n g one sheep to coyotes. Each year was treated independently, and coyote and dog predation analyzed s e p a r a t e l y . The p r o b a b i l i t y of an event (predation) was found using a m u l t i v a r i a t e l o g i s t i c r e g r e s s i o n package (Engleman 1981) u t i l i z i n g the f u n c t i o n where and "a" and "b±" are unknown parmeters estimated from the data and are c a t e g o r i c a l v a r i a b l e s from the data. This technique assumes m u l t i v a r i a t e normality of samples (with or without sheep losses to predators) w i t h equal covariance matrices ( H a l p e r i n et a l . 1971). In p r a c t i c e , these are r a r e l y found and f o r dichotomous v a r i a b l e s , the assumption of m u l t i v a r i a t e normal-i t y i s impossible ( T r u e t t et a l . 1967), thus negating the use of d i s c r i m i n -ant f u n c t i o n a n a l y s i s (Press and Wilson 1978). The advantage of l o g i s t i c r e g r e s s i o n i s that i t i s a very robust t e s t and assumes that some v a r i a b l e s are m u l t i v a r i a t e normal and some dichotomous. I t al s o assumes that the explanatory v a r i a b l e s are independent and dichotomous or that some are dichotomous and some m u l t i v a r i a t e normal (Press and Wilson 1978). Thus, many types of underlying assumptions y i e l d the same l o g i s t i c f o r m u l a t i o n . Engleman (1981), Fienberg (1977) and H a l p e r i n et a l . (1971) a l l use the r a t i o of the l o g i s t i c c o e f f i c i e n t and i t s standard e r r o r as an e m p i r i c a l t e s t of s i g n i f i c a n c e . In a l l cases, i f the absolute value of the r a t i o i s P (Predation) = e 1 + e a + £ b i x i 10 l e s s than 2.0 (approximate value of the parametric t - t e s t w i t h an alpha of 0.05 and an i n f i n i t e number of degrees of freedom) i t i s not s i g n i f i c a n t . Some studies (e.g. Pozen et a l . 1984) have used l o g i s t i c r e g r e s s i o n as a p r e d i c t i v e t o o l . A more common use, however, i s as a p r e l i m i n a r y t e c h n i -que to provide a p r e d i c t i v e framework f o r subsequent hypotheses, because the c o e f f i c i e n t / s t a n d a r d e r r o r r a t i o ("t") i s not s t r i c t l y a s t a t i s t i c a l t e s t ( J . Petkau, pers. comm.). I have used chi-square a n a l y s i s on those v a r i a b l e s where at l e a s t one category w i t h i n a v a r i a b l e has a " t " >2.0. This should e l i m i n a t e any var -i a b l e s included by chance or due to anomalies provided by the small number of farms subjected to predation each year. My r e s u l t s should however be viewed only as management g u i d e l i n e s , rather than as proofs derived from c o n t r o l l e d experiments. In Tables 1 and 2, r e g r e s s i o n c o e f f i c i e n t s , stan-dard e r r o r s and t h e i r r a t i o s ("t") are from the l o g i s t i c r e g r e s s i o n r e s u l t s . The remaining columns are the a c t u a l data. The e m p i r i c a l value of a c o e f f i c i e n t i s assessed by comparing the r e l a -t i v e magnitude of category c o e f f i c i e n t s w i t h i n each v a r i a b l e . The greater the p o s i t i v e value the higher r i s k f a c t o r of predation i m p l i e d by that cate-gory. The i n f e r r e d values may be confirmed by noting the percent l o s s of sheep i n each category and how these compare w i t h c o e f f i c i e n t magnitude. The magnitude of the c o e f f i c i e n t i s not comparable between v a r i a b l e s , nor can a r e l a t i v e s c a l e of r i s k be constructed between v a r i a b l e s . I d i d not t e s t f o r i n t e r a c t i o n s due to the l i m i t e d data. I f necessary, categories were pooled w i t h i n v a r i a b l e s f o r chi-square analyses. The prob-a b i l i t y l e v e l f o r chi-square and a l l other t e s t s was 0.05. As the regres-s i o n package only accepts cases w i t h complete data, the number of farms analyzed each year i s l e s s than the number of farms a c t u a l l y v i s i t e d . 11 Causes of Sheep M o r t a l i t y Numbers, and causes of sheep mortality were determined when possible from the farmer's records, or unfortunately, from th e i r memory of m o r t a l i -t i e s . The month of k i l l was recorded for predator-caused deaths only. Losses were divided into categories of: lambing, disease and old age, coyote predation, dog predation, accidental death, and unknown. Ewe losses and lamb losses were treated separately, and lambs were defined as sheep less than one year o l d . Lambing losses were those lambs born a l i v e and dying i n the f i r s t 24 h of l i f e , and ewes that died giving b i r t h or subsequently from complications. Disease related deaths were those of b a c t e r i a l , v i r a l , para-s i t i c , or metabolic o r i g i n , as determined by the producer. Old age did not include those ewes cu l l e d and sold for slaughter. Coyote and dog predation were d i f f e r e n t i a t e d by producers on the basis of information disseminated by the L.M.S.P.A., and other media. In case of doubt, the verd i c t of a govern-ment W i l d l i f e Control O f f i c e r (W.C.O.) was used for i d e n t i f y i n g the predator responsible. Accidental deaths were from causes such as drowning, exposure, or elec-t r o c u t i o n . Unknown deaths were those i n which a carcass was never recovered or where the carcass had decomposed to such an extent that cause of death could not be determined. Coyote Biology Taxonomy Canid carcasses were c o l l e c t e d from the W.C.O.'s, farmers, road k i l l s and trap m o r t a l i t i e s . These were i n i t i a l l y c l a s s i f i e d as either coyotes, dogs, or hybrids, based on pelage c h a r a c t e r i s t i c s (Mahan et a l . 1978; 12 Freeman and Shaw 1979), physical peculiarities (Mengel 1977) or general mor-phology that deviated from the coyote "type". B. Wong performed a morphome-tric analysis on cleaned skulls to determine if there were hybrids in the sample. Her results are in Wong (1984) and are discussed later. Reproduction Reproductive tracts were removed from female coyote carcasses, the number of feti or placental scars noted and the tract preserved in 10% form-alin. Feti large enough to be aged with the Kennelly et al. (1977) scale were subsequently measured to estimate date of conception. Feti too small to be accurately aged (less than one month, Kennelly et a l . 1977), were dated to the appropriate four week conception period. Food Habits Scats were cleared from sample areas before collection commenced, to obtain fresh, accurately dated scats. Fox (Vulpes vulpes) are rare and domestic dogs were not seen hunting, nor did known dog scats found at trap sights resemble coyote scats. Thus, I only collected coyote scats as any scats of doubtful origin were discarded. Collection areas varied with seasonal vegetational growth, land-use by farmers, and areas of use by radiocollared coyotes. Scats were stored in small plastic bags and frozen until analysed. Thawed scats were oven-dried at 60° C for 24 h and steri-lized at 100°-110° C for 3 to 5 h. They were then placed in cheese cloth bags, washed until the rinse water ran clear, and dried for examination. Food items were identified with the use of a reference collection developed 13 by Chestnutt (1981), A. Tiplady, and myself, and with published keys (Adorjan and Kolenosky 1969; Moore et a l . 1974). A l l white feathers i n scats were i d e n t i f i e d as chicken (Gallus g a l l u s ) because a l l chicken remains examined i n coyote stomachs ( c o l l e c t e d from W.C.O. k i l l s ) , at chicken carcass dumps and i n manure, had white feathers. A l l small coloured feathers i n scats were i d e n t i f i e d as of passerine o r i g i n as no white passer-ines were found i n coyote stomachs nor are any white passerines resident i n the Lower Fraser V a l l e y . Food items were estimated by sight into percent volume. For analysis of food habits, seasons were defined as spring (March-May), summer (June-August), f a l l (September-November), winter (December-February) . Telemetry Coyotes were trapped with #2 Victor c o i l - s p r i n g traps attached to log drags and immobilized by hand or with forked s t i c k s . Mouth and legs were then taped, and i f necessary a t r a n q u i l i z e r was administered, using 1.0 to 1.5 ml of a 1:1 mixture of ketamine hydrochloride (Ketaset) and xylazine hydrochloride (Rompun). Coyotes were aged by tooth wear (Gier 1968) in t o classes of j u v e n i l e (less than 1 year), and adult (greater than 1 year). A numbered p l a s t i c sheep tag was placed i n each ear. Thirteen of 14 adult and juve n i l e coyotes captured were equipped with r a d i o - c o l l a r s (AVM, Champaign, I l l i n o i s ) , which transmitted on the 150 to 151 MHz band, and included an a c t i v i t y option. P r i o r to release, coyotes were injec t e d with two to three ml of procaine p e n i c i l l i n ; any wounds were sprayed with an a n t i s e p t i c spray. Tranquilized coyotes were released into shaded areas when they regained consciousness, and were checked p e r i o d i c a l l y u n t i l they departed. 14 Radio-collar signals were received on a TRX-24A receiver ( W i l d l i f e Materials, Carbondale, I l l i n o i s ) using a handheld, 3-element yagi antenna. Approximate locations of coyotes were found by d r i v i n g along roads u n t i l a s i g n a l was received on a roof-mounted whip antenna. Multiple bearings were then taken with the yagi antenna. If a coyote could not be located from the ground, a e r i a l searches were made u t i l i z i n g two 3-element yagis mounted at a 45° downward angle from the wing on the wingstruts of a Cessna 172. Coyotes found i n this manner were subsequently located d i r e c t l y from the ground. Locations were usually made at least once a week, mostly i n the morning and early afternoon. Accuracy of bearings was tested by loc a t i n g activated c o l l a r s and by v i s u a l confirmations of lo c a t i o n s . Home Range Burt (1943) defined home range as the area t r a v e l l e d by an animal i n the course of normal a c t i v i t i e s (food gathering, reproduction and care of young) usually around a home s i t e . Home ranges were plotted using the mini-mum convex polygon method (Mohr 1947) u t i l i z e d i n the program HOME (Harestad 1981). Coyote residency was determined by an a p o s t e r i o r i judgement based s o l e l y on the continuity of an i n d i v i d u a l ' s r a d i o l o c a t i o n s . For example, i f a coyote stayed within an area where i t could always be found for at least three months, i t was c a l l e d a resident. If i t could always be found i n one small area for at least three months, but abruptly moved from that area, and was never relocated there or at a l l , i t was c a l l e d a resid e n t - t r a n s i e n t . Coyotes which were found i n no area co n s i s t e n t l y or which were located for too short a period to f u l f i l l the d e f i n i t i o n of a resident were c a l l e d t r a n s i e n t s . 15 A c t i v i t y Patterns and Movements Coyotes were determined to be active or i n a c t i v e based upon varying si g n a l emissions transmitted from t h e i r r a d i o - c o l l a r . The accuracy of the a c t i v i t y option was i n i t i a l l y tested by placing a r a d i o - c o l l a r on a dog and confirming by sight radio signal v a r i a t i o n correlated with movement. These signals were monitored over a four to f i v e minute i n t e r v a l each hour of a 24 h sampling period. Minimum d a i l y distances moved were obtained by p l o t t i n g the hourly locations and measuring st r a i g h t l i n e distances between consecu-t i v e l o c a t i o n s . For analysis of seasonal movements the year was divided into the same seasons used i n food habits. Daily a c t i v i t y was divided into dawn (one h either side of sunrise for a 2 h t o t a l period), day (the period between dawn and dusk), dusk (one h either side of sunset for a 2 h t o t a l period), and night (the period between dusk and dawn). Predation by Radio-collared Coyotes Once a coyote was r a d i o - c o l l a r e d , a l l known sheep farmers i n the area were informed of the presence of a r a d i o - c o l l a r e d coyote i n the area. They were then given cards with phone numbers at which I could be contacted and asked to inform me immediately of any coyote predation. Upon being informed of a coyote k i l l a 24 h monitoring session was i n i t i a t e d to determine i f any r a d i o - c o l l a r e d coyote r e v i s i t e d the sheep k i l l s i t e . The coyote was moni-tored every 30 minutes when movement commenced. 16 RESULTS Farm Characteristics Sample Size of Farms A total of 112 different sheep producers were visited over three years of the study, representing 60% of the L.M.S.P.A.'s membership. However, some farmers interviewed were not members of the L.M.S.P.A., and the sample is probably closer to 50% of the sheep producers in the L.F.V. Total farms sampled each year were 55 in 1979, 108 in 1980, and 109 in 1981. Thirty-nine farms were visited each of the three years, 67 farms for two consecu-tive years and six farms only once. From the 1981 interviews a total of 6581 sheep (ewes and lambs only) were recorded as exposed to predation, representing 51% of the sheep in the L.F.V. (based on 1981 B.C. Ministry of Agriculture livestock records of 12,985 ewes, lambs and rams; H.H. Bryce, pers. comm). Coyotes and Dogs, 1979 Of 34 farms analyzed, ten (29.4%) suffered losses to coyotes. A common contributory variable was the number of ewes in the flock. The coefficients indicate that the larger the ewe flock, the greater the vulnerability, but none of the coefficients pass the "t" test (Table 1). Three farms analyzed lost sheep to dogs, but no predisposing factors were identified in the limited sample. r 17 Coyotes 1980 Eighty-nine farms were analyzed, of which 16 (18.0%) lost sheep to coyotes in 1980. Three variables were found to be important: method of carcass disposal, size of sheep pasture, and number of ewes (Table 1). Leaving sheep carcasses exposed to scavenging was the important factor in the carcass disposal category. Removal of carcasses or burying them decreased the probability of losses. The percent loss of sheep in each category followed the inferred values of the coefficients, i.e. as coeffi-cient magnitude increased, the percent loss of sheep increased (Table 1: "X=25.95, df=2). Larger sheep pastures (>8.1 ha) were associated with increased coyote i predation (X=107.22, df=3). As in 1979, farms with large flocks (>50 ewes) appeared most susceptible to coyote predation, with percent loss increasing as the probability of losses increased (Table 1: =72.23, df=2). Dogs 1980 Of 89 farms, eight (9.0%) lost sheep to dogs in 1980. Common among them was the use of guard dogs (Table 2). Farms which used guard dogs were more vulnerable to dog predation. Coyotes 1981 Thirteen (14.4%) of 90 farms lost sheep to coyotes in 1981. Farms where sheep were not confined at night ( X=45.54, df=l) or where bells and working dogs were used, were most likely to lose sheep to coyotes (Table 1). Percent losses followed the pattern of coefficient magnitude, although 18 Table 1. Factors Contributing to Coyote Predation On Sheep In the Lower Fraser Valley 1979-1981 Variable Category Log. "t»l % Tot.No Mean Flock Size No. of Coeff Loss Sheep x * S.D. Farms No. Ewes 0-20 -4.94 -0.18 1.19 503 22.8* 12.5 22 1979 21-50 -2.63 -0.09 4.14 532 66.5* 28.9 8 >50 7.57 0.13 6.30 1000 250.0* 89.8 4 No. Ewes 0-20 -1.22 -1.71 0.98 1222 23.9* 12.5 51 1980* 3 21-50 -1.19 -1.65 1.30 2300 79.3* 25.9 29 >50 2.41 2.26 4.95 1938 215.3* 74.3 9 Carcass Bury -0.03 0.05 3.32 3008 52.7* 56.8 57 Disposal Haul -2.82 -2.66 1.33 2326 80.2* 76.1 29 1980* 3 Leave out 2.79 2.76 5.56 126 42.0* 25.3 3 Pasture <2 -1.04 -1.38 0.69 1307 31.8* 23.8 41 Size (ha) 2.1-4.1 -1.15 -1.65 0.46 1722 55.5* 42.0 31 1980* 3 4.2-8.1 -1.01 -1.03 4.83 1429 129.9± 93.4 11 >8.1 3.20 2.64 5.19 1002 167.0* 96.4 6 Confine. yes -1.67 -2.52 0.08 2601 57.8* 58.6 45 1981* 2,3 no 1.67 2.52 1.95 2613 58.0* 56.1 45 Bells yes 1.24 2.24 1.05 1715 100.2* 93.4 17 1981* no -1.24 -2.24 1.00 3499 47.9* 39.1 73 Workdog yes 1.35 1.91 1.13 1147 114.7*102.2 10 1981 no -1.35 -1.91 0.98 4067 47.9*248.0 80 Min.Dist. <10 0.76 0.36 1.12 2235 47.5* 39.8 47 to 11-50 0.83 0.39 0.96 2404 66.7* 57.3 36 Pasture(m) 51-100 7.03 0.00 4.03 124 31.0* 9.1 4 1981 * >100 -8.63 -2.11 0.00 451 150.3*180.3 3 No. of 0 0.90 0.39 0.37 1872 74.8* 80.8 25 Animals 1-10 -1.89 -0.77 0.47 2345 47.8* 43.0 49 1981 11-25 -0.02 -0.01 3.60 250 35.7* 32.9 7 26-50 -6.83 -0.78 0.00 211 52.7* 25.4 4 51-100 3.94 1.41 2.57 311 103.6* 63.7 3 >100 3.90 0.00 8.00 225 112.5* 17.6 2 * Variables passing the "t" test (Engelman 1981) 1 Engelman (1981) % Night Confinement J Significant chi-square P<0.05 19 the difference was not significant for bells ( X=0.03. df=l). Farms, where the minimum distance to pasture lay between 50 and 100 m, appeared to have a higher susceptibility to coyote predation than farms where the minimum distance was less than 50 m or greater than 100m. The percent loss distribution followed the pattern of regression coefficents, but was not -.2. significant (/\.=0.43, df=2). Where the number of animals was greater than 50, farms were predisposed to predation compared to farms with fewer animals. The percent loss di s t r i -bution was not in complete accord with the value of coefficients (Table 1). Dogs 1981 Of 90 farms, 10 (11.1%) lost sheep to dogs in 1981. Two factors were related to a farm's susceptibility to dog predation: month of lambing and minimum distance to pasture from house or barn. Lambing from January to March, and especially in June, increased the risk of dog predation through the rest of the year; percent losses agreed with the coefficient pattern but not significantly (/C=3.39, df=l, Table 2). Farms where the minimum distance to pasture was greater than 50 m were more liable to suffer dog predation than farms where the distance was less. However, the distribution of percent losses between categories was not in accord with the coefficient magnitude occurring in these categories nor was i t significant (*)£=0.95, df=l, Table 2). 20 Table 2. Factors Contributing to Dog Predation on Sheep in the Lower Fraser Valley 1979-1981 Variable Category Log. ••t»l % Loss Tot.No Mean Flock Size No. of Coeff. Sheep x ± S.D. Farms Guard Dog yes 0.62 1.56 0.46 658 43.8± 42.3 15 1980 no -0.62 -1.56 0.35 4802 64.8± 67.7 74 Month of January -1.82 -0.31 0.28 1071 63.0i 41.6 17 Lambing February 2.81 0.00 0.73 1097 45.7± 43.8 24 1981 * March 3.32 3.93 0.58 1722 63.7± 61.1 27 April -5.74 -0.14 0.00 948 72.9± 94.3 13 May -6.33 -0.08 0.00 58 19.3± 14.6 3 June 14.80 -0.15 6.45 31 31.0± 0.0 1 November 5.62 0.00 0.00 58 58.0± 0.0 1 December -5.06 -0.06 0.00 229 57.2± 19.6 4 Minimum(m) < 10 -0.50 -1.09 0.76 2235 47.5± 39.8 47 Distance 11-50 -2.05 -2.17 0.08 2404 66.7± 57.3 36 to Pasture 51-100 1.52 1.17 1.61 124 31.0* 9.1 4 1981 * > 100 4.09 -4.90 0.44 451 150.3*180.3 3 * Variables passing the "t" test (Engelman 1981) Mortality The proportion of farms losing sheep to coyotes (1979: 23.6%, n=55; 1980: 17.6%, n=108; 1981: 15.6%, n=109), to dogs (1979: 5.5%, n=55; 1980: 9.3%, n=108; 1981: 11.9%, n=109), or to both combined (1979: 25.5%, n=55; 1980: 24.1%, n=108; 1981: 23.9%, n=109) did not vary significantly among years (coyotes: =1.60, df=2; dogs: "\_ =1.80, df=2; combined: =0.05, df=2). Of the farms which lost sheep to coyotes, 44.8% [n=29] lost sheep in only one year, while 55.2% lost sheep in two or three consecutive years. In comparison, 82.6% [n=23] of farms that lost sheep to dogs, lost in only one year, while 17.4% lost sheep in two years. Eight (18.2%) of the 44 farms which lost sheep to predators, lost sheep to both coyotes and dogs. Predation accounted for 368 sheep (119 ewes, 249 lambs) lost during the study, representing 28.2% of a l l losses and 2.4% of the total population sampled (Tables 3 and 4). Coyotes killed 69.7% and 74.7% of all ewes and lambs lost to predators. The relative proportions of deaths due to coyote and dog predation varied significantly (.\,=15.51, df=2) between years, with coyotes killing 87.0%, 66.9% and 66.7% of predator caused losses from 1979 t 1981. There was a significant variation from an equal lamb:ewe ratio in bot coyote (^=39.44, df=l) and dog ("^ =7.36, df=l) k i l l s . Coyote kills were composed of 69.1% (n=269) lambs, and dog kil l s of 63.6% (n=99) lambs. Ewe Mortality Deaths due to disease or old age was the major mortality factor (39.8%) among ewes during the study (Figure 2A) and increased significantly (^=7.31, df=l) thoughout the study (Table 3). Coyote predation was the second most important mortality factor, followed by lambing, dog predation 22 and accidental deaths (Figure 2A). All factors except combined dog predation and accidental deaths varied significantly (dogs and accidental: *X. =2.06, df=2; lambing: "^=12.50, df=2; coyotes:")(_ =41.74, df=2) among years. The proportion of the ewe population lost to coyotes decreased signifi-cantly ( >\_=62.78, df=2) through the study (Table 4). Losses to lambing ()(_=21.21, df=2), and to disease and old age (X.= 1 2« 7 8» df=2) also varied significantly among years, while combined losses to dog predation and acci-dents did not (*)(=2.33, df=2). Coyote and dog predation were characterized by abrupt seasonal fluctua-tions. Coyote and dog predation on ewes occurred primarily from spring to early autumn; 82.3% of coyote predation took place from June to September, and 93.3% of dog predation from April to September (Figure 3A). Lamb M o r t a l i t y Almost half of the lamb m o r t a l i t y (Figure 2B) was the r e s u l t of prob-lems which occurred within 24 hours of b i r t h (e.g. d y s t o c i a , poor mothering, trampling, e t c . ) , and this l e v e l did not vary among years ("^=2.72, df=2, Table 3). Coyote predation was the second highest mortality f a c t o r , followed by deaths due to unknown causes, accidental deaths, disease, and dog predation (Figure 2B). Relative losses to coyotes, disease, accidental deaths, and deaths from unknown causes varied s i g n i f i c a n t l y among years ( c o y o t e s : " ^ =40.42, df=2; disease: "^=18.52, df=2; accidents: ~)(_ =17.34, -v2 df=2; unknown: y =21.69, df=2), with coyote predation decreasing, and the remaining losses increasing (Table 3). Dog predation did not vary s i g n i f i c a n t l y (^j=4.04, df=2). 23 Table 3. Percent Loss (by factor) of Sheep Mortality in the Lower Fraser Valley, 1979 to 1981 Percent of Losses 1979 1980 1981 Total Ewes n = 80 178 69 327 Lambing* 6.3 24.7 17.4 18.7 Disease and Old Age* 27.5 42.1 47.8 39.8 Coyotes* 52.5 18.0 13.0 25.4 Dogs ^ 10.0 10.1 14.5 11.0 Accidents^ 3.7 5.1 7.2 5.2 Lambs n = 158 457 364 979 Lambing 53.2 48.1 45.3 47.9 Disease* 3.2 8.1 9.3 7.8 Coyotes* 36.7 17.3 13.5 19.0 Dogs 4.4 8.1 5.2 6.4 Accidents* 2.5 8.3 13.7 9.4 Unknown* 0.0 10.1 12.9 9.5 Significant variation among years P<0.05 Dogs and accidental deaths combined for chi-square analysis. 24 Table 4. Percent of Total Sheep Sample Lost to Mortality Factors in the Lower Fraser Valley, 1979 to 1981 1979 1980 1981 No. Lost % No. Lost % No. Lost % Ewes n = 1286 2954 2640 Lambing* 5 0.39 44 1.49 12 0.45 Disease 22 1.71 75 2.54 33 1.25 and Old Age* Coyotes* 42 3.27 32 1.08 9 0.34 Dogs^ 8 0.62 18 0.61 10 0.38 Accident^ 3 0.23 9 0.30 5 0.19 Lambs 1309 3864 3524 Lambing* 84 6.42 220 5.69 165 4.68 Disease 5 0.38 37 0.96 34 0.96 Coyotes* 58 4.43 79 2.04 49 1.39 Dogs 7 0.53 37 0.96 19 0.54 Accidents* 4 0.31 38 0.98 50 1.42 Unknown* 0 0.00 46 1.19 47 1.33 * Significant variation among years P<0.05 2 Dogs and accidental deaths combined for chi-square analysis. 25 Figure 2A. Proportional mortality factors among ewes 1979-1981 (n=327) Figure 2B. Proportional mortality factors among lambs 1979-1981 (n=979) 26 Accidents 5.2% 6 . 4 % 27 As a proportion of total population losses, coyote predation varied significantly among years (Table 4; A_=42.42, df=2) as did population loss levels to lambing (^=6.85, df=2), accidents ("^  = 11.61, df=2) and unknown causes (^ ==17.01, df=2). Loss levels to disease (^=4.28, df=2) and dogs (^=4.93, df=2) did not vary significantly among years. Coyote predation on lambs (Figure 3B) was clearly seasonal; highest levels of predation occurr-ing between April and September, when 91.7% of all lamb losses to coyotes occurred. Seasonal variation in dog predation on lambs (Figure 3B) was similar to the pattern of dog predation on ewes with 66.7% of predation losses occurring from April to October. The apparent high frequency of loss in December was the result of an attack by two dogs on one farm on one night. Coyote Biology Reproduction Twenty estimates of litter size were derived from placental scar or fet counts and two from free ranging litters. Mean litter size over three breed ing seasons was 5.4 ± 1.8. One female, tagged as a pup (May 30, 1981), was shot with two pups she had been nursing on July 24, 1982 (D. Nagy, pers. comm), indicating that some females do breed as yearlings in the L.F.V. Most courting and pair-bonding would have taken place in December and early January because the mating season, based on conception dates estimated from fetal measurements in 10 litters, extended from early January to April. Thirty percent (n=3) of the aged litters were conceived in January, 50% (n=5 in February, 10% (n=l) in either February or March; and 10% (n=l) in the thi week of April. Parturition should have occurred 60 to 63 days following 28 Figure 3A. Monthly frequency of occurrence of predation upon ewes by coyotes (n=79) and dogs (n=30) in the Lower Fraser Valley 1979-1981. Totals are less than mortality reported in Table 4 as some predation loss dates were unknown. Figure 3B. Monthly frequency of occurrence of predation upon lambs by coyotes (n=227) and dogs (n=60) i n the Lower Fraser V a l l e y 1979-1981. Totals are less than mortality reported i n Table 4 as some predation loss dates were unknown. PERCENT -p. —« —» K) KJ OJ tn o Cn O Cn O —1— 1 1 1 1 p o r>i£j O O fl> o 3 FREQUENCY _ i M K ) Cn O cn o cn O —1 1 1 1 1 1 ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ ^ ^ ^ ^ sWWWWN O O I Q S§3 30 conception (Kennelly 1978), thus placing the peak of births in late April and early May. Food Habits To allow chi-square analysis of seasonal variation in diet, the nine identified food items were further grouped into four categories: small rodents, eastern cottontail rabbits (Sylvilagus floridanus) and possibly mountain cottontails (S. nuttalli: F. Bunnell, pers. comm), fruit and grass, and "various". The latter category involved pooling of unidentified birds and mammals, domestic livestock, identified wild mammals (see below for species), and miscellaneous (insects and garbage). There was a significant seasonal variation in the relative proportions of food items in the diet (/(j=150.15, df=18, Figure 4B), probably due to high seasonal consumption of fruit. Throughout the scat collection period small rodents, of which >95% were Microtus townsendi, were consistently the major food item, averaging 70.2% by volume of all scats (n=862, Figure 4A). The second most important food type was rabbit, averaging 8.1%. Not surprisingly, fruits and seeds were found seasonally, with high volumes found in scats only during summer and fal l (Figure 4B). Fruits and seeds consumed by coyotes included cherries and plums (Prunus spp.), apples (Malus sp.), and small quantities of holly (Ilex sp.) and broom (Cytisus scoparius). Scats composed entirely of grass (Gramminae spp.) were found throughout the year. Use of unidentifiable birds (probably passerines like those found in coyote stomachs examined) and mammals, and of identifiable mammals (raccoon, Procyon lotor; oppossum, Didelphis marsupialis; muskrat, Odontra zibethica; 31 Figure 4A. Composite diet of coyotes (percent scat volume) in the Lower Fraser Valley 1980-1981 (n=862). The mammals class consists of unidentified and identified wild mammals other than rabbits and small rodents. See text for identified species. Figure 4B. Seasonal d i e t of coyotes (percent scat volume) i n the Lower Fraser V a l l e y 1980-1981 (n=862). 3 2 Mitcatlan«ou8 0.3% N: 51 9 4 1 3 0 to a to 100-90-80-70-60-3 50" o > o DC 40-30-20-10-1 2 7 2 3 7 1 6 4 5 9 B Legend EE PLANT ES MAMMAL D PASSERINE SS DOMESTIC eZ) RABBIT • MICROTINE • i MISCELLANEOUS SPR SUM FAL WIN SPR SUM FAL 1980 1981 33 and blacktail deer, Odocoileus hemionius columbianus) varied in volume through the year (Figure 4B), but they were of limited overall importance to the coyote. The percent volume of domestic livestock (cattle, Bos taurus; sheep; pig, Sus scrofa; and chicken), in scats appeared to vary seasonally, but was also of limited importance in the diet (Figures 4A and 4B). Insects were found in scats and though not identified to species, included beetles (Coleoptera), dragonflies (Odonata) and grasshoppers (Orthoptera). Garbage, including cloth, rubber, paper, and plastic were occasionally found in scats. Neither insects nor miscellaneous items appeared to vary seasonally in volume (Figure 4B). Home Range Twenty-six coyotes were captured 29 times; eight males and five females were radio-collared and monitored for periods between 41 and 437 days (Table 5) Eleven coyotes, (one adult male, one female pup, and nine male pups, including M7 who was eventually radio-collared), were eartagged and released. Three coyotes (one adult male, one adult female,and one male pup) were killed because of trapping injuries. Due to a small sample size and limited monitoring time of some females, I did not compare statistically home range sizes between male and female coyotes. The mean size of the 95% home range (containing 95% of an animals locations) was 31.5 ± 61.3 km^  for males and 12.4 ± 16.0 km^  for females. When M3 (transient) and M8 (small number of locations) are deleted, the mean size of male 95% home ranges decreased to 7.7 ± 4.2km2. When Fl and F5 are deleted (small number of locations) the mean 95% home range size is 17.0 ± 20.7 km^  for females. Overall 95% home range size was 24.2 * 48.7 km^  and 34 Table 5. Radiotelemetry Status of Lower Fraser Valley Coyotes Coyote Trapped Last Potential Number of 95% Comments On Located Monitoring Locations Home On Days Range Females F 1 19/1/80 31/3/80 41 26 8.5 c,f 2 26/6/80 27/7/81 396 75 6.8 a,f 3 1/7/80 7/12/80 160 52 40.8 b,f 4 10/1/81 18/8/81 220 53 3.4 a,d,f 5 20/1/81 11/3/81 50 23 2.7 c,f Males M 1 18/3/80 22/11/80 249 87 11.5 b,f 2 6/4/80 7/11/80 184 85 11.5 b,f 3 5/6/80 16/8/81 437 68 182.4 c,f 4 28/6/80 30/6/81 367 91 11.1 a,f 5 9/8/80 11/3/81 215 60 5.5 a,f 6 21/8/80 20/8/81 364 58 1.6 a,f 7 13/9/80 26/8/81 348 80 5.1 a,e 8 6/11/80 18/3/81 132 20 23.3 c,f a. resident b. resident-transient c. transient d. radio-collar failed to transmit until 9/5/81 when the magnet fe l l off, thus activating the collar. e. juvenile (less than 1 year old when radio-collared). f. adult (older than one year when radio-collared). 35 Figure 5A. Area u t i l i z e d by transient male M3. Numerals r e f e r to plotted point of every t h i r d radio r e l o c a t i o n . Figure 5B. Home ranges of females F l (transient) and F3 ( r e s i d e n t - t r a n s i e n t ) . 3 6 37 Figure 5C. Home ranges of females F2 and F4 (residents), F5 (transient) and male M6 (resident). Figure 5D. Home ranges of males Ml and M2 (resident-transients), M4, M5, M7 (residents) and M8 (transient). The home ranges designated M8 and M8! are the areas it consecu-tively utilized. 38 39 when the above animals are deleted this decreases to 10.8 ± 11.2 km2. Of the 13 coyotes radio-collared, six were classified as residents (two females, four males), three as resident-transients (one female, two males) and four as transients (two females, two males). Over half (53.8%) of the coyotes thus did not occupy a permanent home range during the study. The home ranges in Figures 5B, 5C and 5D include only the ini t i a l areas used by resident-tran-sients and transients, e.g. Figure 5D does not include the 4 relocations of Ml after i t left its original range. M3 was the only coyote for which continual monitoring of transient move-ment was obtained. It travelled through a known area of 219.7 km2 over 437 days (Figure 5A), but its erratic and almost constant movement allowed only 68 locations. Every third location ,is plotted in Figure 5A. Although M8 is called a transient it could have had a non-contiguous home range, i.e. two areas of use separated by an area of low or no utilization (Hibler 1977). However, i t was found in neither area (M8 and M8!, Figure 5D) long enough to qualify as a resident before it was killed, nor did it move back to its first range from M8!. Radio-collared or eartagged coyotes were recovered or last located at distances of 0.5 to 13.3 km [x=8.9 "i 4.9] away from their capture point. In eight of nine instances, recoveries of carcasses or last radio locations were obtained north, north-west, or north-east of the capture point. Ml was last located southwest of its capture point and aside from M3 was the only coyote known to move south. I twice attempted to relocate coyotes which might have moved into Washington state but no coyotes were found during either flight. In some cases home range overlap was only spatial (Figure 5B) while in others i t was both, temporal and spatial (Figures 5C and 5D). The extent of home range overlap among males on Canadian Forces Base Aldergrove (C.F.B.A., 40 Figure 5D) in d i c a t e s an absence of t e r r i t o r i a l i t y and possibly some temporal s o c i a l relatedness, e.g. possible non-dispersal of the previous years l i t t e r from t h e i r natal home range. This could r e s u l t i n high l o c a l densities of coyotes. Natal philopatry was documented for a ye a r l i n g female shot 0.5 km from where she was o r i g i n a l l y caught, eight months e a r l i e r . Other than the t r a n s i t o r y residence of F5 within the home range of F2 (Figure 5C), no temporal overlap was found among ra d i o - c o l l a r e d females. However, two weeks a f t e r F4's release, i n an attempt to r e t r i e v e F4 and i t s then non-functioning c o l l a r , a pregnant female coyote was shot in F4's home range,indicating possible home range overlap for two pregnant females. Home range overlap between males and females was assumed on C.F.B.A., (Figure 5D) and found between F2 and M6 (Figure 5C), and i n the home range of F4 (Figure 5C). No females were caught on C.F.B.A., but capture of pups and presence of den s i t e s indicated the existence of at least one female. F4 was i n proestrus when captured, and an attendant coyote, presumably male, barked repeatedly from the surrounding bush while she was processed. I assumed she-mated as pups were l a t e r seen near her radio p o s i t i o n . A c t i v i t y Patterns and Movement Extensive radio interference and loss of external antennae from radio-c o l l a r s resulted i n poor monitoring conditions and severely c u r t a i l e d the number of monitoring sessions. Twenty-four sessions were completed for 24 h (19 male and f i v e female sessions) and eight were terminated when equipment f a i l e d , or the signal was l o s t . Because only f i v e complete female sessions were obtained, the data were pooled for a n a l y s i s . Data from incomplete 41 sessions were incorporated into the analysis of d a i l y movement where por-tions of d a i l y segments were complete. There was no s i g n i f i c a n t difference i n seasonal movement (Kruskal-Wallis t e s t , H=3.69, df=3) although mean spring and summer d a i l y movements (km/day) averaged higher than f a l l and winter movements (spring: 3.8 ± 2.6, n=6; summer: 4.8 ± 3.6, n=7; f a l l : 2.2 ± 1.5, n=5; winter: 2.0 ± 1.4, n=6). Nocturnal t r a v e l , the major movement pattern, was s i g n i f i c a n t l y greater (H=15.12, df=3) than t r a v e l at dusk, at dawn and through the day (dawn: 0.6 ± 1.0 km, n=28; day: 0j7_ ± 1.0, n=28; dusk: Oj/i ± 0.5, n=28; night: 1A_ ± 1.30, n=24). The mean minimum distance moved by L.F.V. coyotes over 24 h was 3.3 ± 2.7 km (range = 0.0 to 9.9 km, n=24). Predation by Radio-collared Coyotes In 1980, three lambs were k i l l e d within three months i n M2's home range. I was n o t i f i e d immediately of only one instance and monitored M2 that night, but he did not move a l l night. The farmer who shot the attack-ing coyote (a f t e r i t had k i l l e d the lamb) did not see a r a d i o - c o l l a r on the coyote, and when M2 was k i l l e d and necropsied f i v e months l a t e r , no shotgun p e l l e t s were found. In three other cases when sheep losses occurred within the home ranges of other r a d i o - c o l l a r e d coyotes, I was not informed of the losses f o r several weeks. 42 DISCUSSION Farm Characteristics Associated with Predator Losses Coyote Predation From the three years of analyzed data, nine variables were found to affect the probability of farms losing sheep to coyotes. Of these var-iables, three (ewe numbers in 1979, use of a working dog, and number of animals on the farm) do not meet the criterion of the "t" test (Engelman 1981, p.337) while six do (ewe numbers in 1980, method of carcass disposal, pasture size, night confinement, use of bells, and minimum distance to pasture). Little confidence may be placed upon the first three variables because they may be artifacts of using a large number of variables in the regression. However, on the premise that they are of some possible signifi-cance, they are included in this discussion. A number of the factors studied appear to be covariables and thus not directly related to sheep losses. Working dogs were usually a management option most often employed with large flocks. Therefore, it was probably the size of the flock and not the presence of a work dog which was related to higher losses (see below). Similarly, that high number of animals pre-disposed sheep farms to coyote predation seems unlikely. High numbers of animals possibly increase the number of available prey (e.g. calves), but no other relationship is readily apparent. Although ewe numbers in 1979 did not meet the "t" criterion, the distribution of losses and pattern of coefficients were similar to ewe numbers in 1980, where the "t" standard was met. In both years, as ewe flock size increased, average flock size increased, percent loss to preda-tion increased, and the coefficient value also increased. The flocks most vulnerable had 50 or more ewes. This relationship between flock size and 43 predation losses has been found i n other areas, i n d i c a t i n g numbers and hence a v a i l a b i l i t y of sheep may encourage predation (Boggess et a l . 1978). Robel et a l . (1981) found t o t a l losses to predators were greater for large than for small f l o c k s , although the rate of loss to both coyotes and dogs decreased as f l o c k size increased. S i m i l a r l y , Dorrance and Roy (1976) reported large f l o c k s were more susceptible to predation, and Jones and Woolf (1983) found high numbers of young swine were associated with increas-ed coyote predation. Not s u r p r i s i n g l y , carcass disposal proved to be an important v a r i a b l e . In the present study, although b u r i a l of carcasses decreased the p r o b a b i l i t y of losses, removal of carcasses was the most e f f e c t i v e method of di s p o s a l . Leaving carcasses i n the f i e l d was the worst method of d i s p o s a l . In the L.F.V., proper carcass disposal i s often compromised by W.C.O.'s who use sheep carcasses for trapping, either as shallowly buried or hidden baits (e.g. i n blackberry bushes (Rubus l a c i n i a t u s ) or log p i l e s ) . The benefit of eliminating offending coyotes must be weighed against the p o s s i b i l i t y of drawing coyotes into an area of available carrion and perhaps causing preda-t i o n (Lehner 1976). Robel et a l . (1981) and Jones and Woolf (1983) found that b u r i a l of carcasses s i g n i f i c a n t l y reduced l i v e s t o c k losses to coyotes. Todd and Keith (1976) also found removal of carrion resulted i n a decreased coyote population. Transient animals moved from that area and increased l o c a l populations i n adjacent areas where carrion was a v a i l a b l e . There was an increase i n both the regression c o e f f i c i e n t and the per-cent losses with increased pasture s i z e , e s p e c i a l l y for pastures over 8 ha. As high percent losses were found among flocks on pastures of 4 to 8 ha and over 8 ha, perhaps the trend should be noted rather than using actual area 44 as a delineating feature. Robel et al. (1981) found that large flocks on large pastures were more susceptible to coyote predation in Kansas. Henne (1975: cited in Roy and Dorrance 1976), reported that most coyote predation in western Montana occurred just prior to dawn. In the L.F.V. coyotes were most active and travelled greater distances at night. The value of restricting coyote access to sheep at such times is shown with lower percent losses of confined sheep at night, in the L.F.V. and elsewhere (Meduna 1977). Robel et al. (1981) found farms with one or more belled sheep suffered significantly fewer losses to coyotes but did not consider i t important. In this study, the indicated susceptibility of using bells may be due to the fact that belled flocks (a bell on one or more sheep) were on average twice as large as flocks without bells. Bells may also aid farmers in locating sheep on large pastures or act as an alarm, particularly at night. The significance of bells is most likely a covariable with large flocks which are more susceptible to predation. The predictive value of the minimum distance to pasture appears low for coyote predation, although percent losses correspond with the pattern of logistic coefficients. The lack of a trend, however, negates serious considertion of this variable. Dog Predation Of three variables found to predispose farms to dog predation, one (use of guard dogs) did not meet the "t" test and the other two (month of lambing and minimum distance to pasture) met only marginally. All three could have 45 been s e l e c t e d by chance, as sample s i z e s of farms l o s i n g sheep to dogs were lower than i n coyote p r e d a t i o n . Of t r a d i t i o n a l l y used breeds of guard dogs, only three dogs (two Komon-dors, one Great Pyrenees) were used i n my sample. Robel et a l , (1981) found that dog predation was higher on those farms where dogs were kept o u t s i d e , but t h i s was not apparent i n the L.F.V. Seventy-five percent of the L.F.V. sheep were i n f l o c k s that lambed from January to March. Robel et a l . (1981) found the same lambing period r e s u l t e d i n high rates of l o s s to dogs, versus October to December or year round lambing. January to March lambing occurs i n the L.F.V. to take advan-tage of market c o n d i t i o n s (M. T a i t , pers. comm.) and i s therefore u n l i k e l y to a l t e r , e s p e c i a l l y as dog predation a f f e c t s very few producers. Robel et a l . (1981) found that as distance from the house to the centre of the pasture incre a s e d , so d i d the rate of sheep l o s s to dogs. Dog preda-t i o n o f t e n r e s u l t s i n m u l t i p l e k i l l s and can f o l l o w extensive chasing and harassment of sheep (Roy and Dorrance 1976), thus pastures d i s t a n t from the house were perhaps more conducive f o r such p u r s u i t , but not s i g n i f i c a n t l y so i n the L.F.V. . As p r e v i o u s l y stated i n the m o r t a l i t y s e c t i o n , dog predation appeared to be an almost random event. The low incidence of dog predation thus accounts f o r the l a c k of p r e d i c t i v e f a c t o r s among farms with l o s s e s . Coyote and Dog Pr e d a t i o n : Summary Except f o r the s i z e of ewe f l o c k s i n 1979 and 1980, no v a r i a b l e s were annually a s s o c i a t e d w i t h coyote and dog pr e d a t i o n . The apparent randomness of dog predation and small sample s i z e account f o r the l a c k of 46 predictability of dog attacks. Although sample sizes of farms losing sheep to coyotes are larger than those losing sheep to dogs, they were also small enough to be easily affected by the inclusion of different farms among the farms losing sheep each year. It is logical to expect an annual repetition of predisposing factors, but the variability among years between samples and probable interactions with coyote biology (e.g. death or movement of offend-ing coyotes or cessation of killing by offending coyotes) could preclude such repetition. Chance could also result in some variables being included. However, those factors found to be predisposing (ewe numbers, night confinement, pasture size, and carcass disposal) corroborate the results of other similar studies (Robel et al. 1981; Jones and Woolf 1983; Todd and Keith 1976). Support is thus given to the recommendation of agri-cultural extension workers that a preventative philosophy be employed (Dorrance 1983) and alterations of husbandry practices (e.g. carrion dis-posal, predator proof fencing, night confinement) be used to reduce coyote (and dog) predation (Anon. 1982; Boggess et a l . 1980; Dorrance 1983; P.P.W.M.C. Undated). Mortality Between 74.5% (1979) and 76.1% (1981) of the sheep farmers sampled in each year had no losses to either coyotes or dogs during the study. These are higher than levels of 39 to 59% reported elsewhere (Balser 1974, Dorrance and Roy 1976, Robel et a l . 1981, Schaefer et a l . 1981), and show, as in most areas, that a small minority of the farms suffered the majority of losses. Recurrent coyote predation on individual farms is known (Ogden et al. 1978) especially for sheep or goats pastured on range (Nass 1977; 47 O'Gara et al. 1983; Wade and Connolly 1980), but except for Schaefer et al. (1981) no figures on recurrent coyote and dog predation among a sample of farms are available for comparison to the L.F.V. Over a three year period in Iowa, recurrent coyote predation occurred more frequently (37%) than did dog predation (20%; Schaefer et al. 1981). The loss of 1.2% of the ewes and 2.1% of the lambs to coyotes in the L.F.V. was comparable to other areas, which ranged from 0.7% to 2.5% and 0.7% to 8% respectively (Dorrance and Roy 1976; Gee et a l . 1977; Robel et al. 1981; Schaefer et al. 1981). Losses to dogs of 0.5% of the ewes and 0.7% of the lambs in the L.F.V. also f e l l in the range of other reported values of 0.2% to 1.1% and 0.1% to 0.8% (Dorrance and Roy 1976; Robel et al. 1981; Schaefer et a l . 1981). deCalesta (1978) found dog predation was relatively higher in western Oregon (16.8% of all losses) than in eastern Oregon (8.1%), and ascribed this to higher dog densities in western Oregon. However, although Boggess et al. (1978) found no correlation between the number of sheep killed and the number of dogs licensed per county, they did find one between the number of sheep killed by dogs and the number of sheep produced per county. Availability of sheep, rather than predator density, may be the determinant factor (Boggess et al. 1978) for dog predation to occur. Boggess et al. (1978) found 49% of sheep killed by predators in Iowa were attributed to dogs and only 36% to coyotes. Generally though, as in the L.F.V., coyotes k i l l more sheep than any other predator (deCalesta 1978; Dorrance and Roy 1976; Klebenow and McAdoo 1976; Robel et a l . 1981; Schaefer et al. 1981). However, most studies of coyote predation (e.g. Gee et al. 1977; Klebenow and McAdoo 1976; McAdoo and Klebenow 1978; Nass 1977) have involved sheep grazed on rangeland (where dogs would be a minor component of 48 the predator population) reflecting the distribution of commercial sheep production in the U.S.A. (U.S.F.W.S. 1978). Coyotes usually k i l l more lambs than adult sheep (deCalesta 1978; Dorrance and Roy 1976; Gee and Magleby 1977; Schaefer et a l . 1981; this study). The reverse may be true with dogs (Dorrance and Roy 1976; Schaefer et a l . 1981), although Robel et al. (1981) found equal percent losses of ewes and lambs, and in the L.F.V. more lambs than ewes were killed by dogs. Lambs may be preferred by coyotes, possibly due to their greater vulnerabil-ity, and numbers in the flock (Connolly et al. 1976; Gluesing 1977). Reasons for the variation in prey selection by dogs among areas are not known. The major causes of deaths of ewes in the L.F.V., old age or disease, and of lambs, problems associated with lambing, were similar to findings reported in Kansas (Robel et a l . 1981), Iowa (Schaefer et a l . 1981), and Alberta (Dorrance and Roy 1976). Reasons for the observed variation in magnitude between years in the L.F.V. among the mortality factors are largely unknown. Changing from pagewire fencing to electric fencing reduced coyote caused losses from over 50 to 0 sheep in 1980 on one farm, and adding a pagewire apron to a pagewire fence reduced coyote caused losses from 3 to 0 on another farm in 1980. Such isolated events would have influenced the trend of coyote predation in the L.F.V. which decreased from 1979 to 1981. It is unlikely that an observer effect would be responsible for the decrease in coyote predation; i.e. that farmers had previously exaggerated losses and did not do so when questioned. A farmer desiring increased coyote control would more likely exaggerate than disclaim losses. Whenever possible I verified losses with the W.C.O.'s, and I detected one occasion when losses were ascribed to coyote rather than dog predation. Deaths due 49 to unknown causes varied because the farms which had lost many lambs without recovering the carcasses were not in the 1979 sample. The seasonality of coyote predation on sheep is a common feature of canid predators which often rely heavily on young prey (Dorrance and Roy 1976; Hatter 1984; Lamprecht 1978; Pimlott et al. 1969; Robel et al. 1981). The peak of lamb kil l s usually occurs during the spring and summer when lambs are on pasture or range, and may involve a synchronization with the coyotes' reproductive cycle. Upon the removal of coyote litters from dens, T i l l and Knowlton (1983) found coyote predation on sheep ceased within three days. The decrease in coyote predation in f a l l and winter in the L.F.V. may have been due to the sale of most lambs by that time, closer confinement of ewes during lambing (January to March), cessation of predation on sheep by adult coyotes as pups become independent or some combination of these factors. The frequency of dog predation also declined in f a l l and winter, probably a function of sheep numbers and availability (Boggess et al. 1978). The apparent inhibition of coyotes to entering barns in pursuit of sheep, does not, however, prevent dogs from entering and killing sheep in such confined quarters, thus resulting in dog predation throughout most of the year (Figure 3). Blair and Townsend (1983) found sheep confinement had no effect on dog predation. Coyote Biology Taxonomic Status Taxonomic identification of suspected crosses among field collected canids followed Lawrence and Bossert (1967) in the use of multivariate 50 statistical analysis (Freeman and Shaw 1979; Gipson et al. 1974; Mahan et al. 1978). Wong (1984) used similar analysis on the L.F.V. skull sample and found no coydogs, but qualified her conclusion because of lack of a known hybrid sample to include in her analysis. A factor possibly limiting the occurrence of coydogs in the L.F.V. is the apparent absence of a feral dog population (D. Pemble, D. Nagy, pers. comm.). No dogs trapped by myself or by W.C.O.'s could be classed as feral, as all dogs were easily managed when while in traps. Reproduction The mean litter size for L.F.V. coyotes is comparable to those from other areas of North America (Hilton 1976; Knowlton 1972; Nellis and Keith 1976; Todd et a l . 1981), as is the temporal pattern of the breeding season in the L.F.V. (Gier 1968; Gipson et al. 1975; Kennelly 1978). Kennelly and Johns (1976) recorded a female in estrus during the first week of April, but as in the L.F.V., i t was considered uncommon. The reproductive cycle of coyotes is probably an evolutionary reflec-tion of prey availability ensuring that the young are raised when the most prey are available. In most coyote habitats, lagomorphs would probably be the major prey species during this period, although young ungulates can be important. Reichel (1976) found pronghorn (Antilocapra americana) fawn mor-tality was correlated with proximity to coyote dens. However, the arrival of sheep in North America provided an important alternative food supply, and as in other areas (Kauffeld 1977; Robel et al. 1981; Schaefer et al. 1981; T i l l and Knowlton 1983), the highest levels of coyote predation on sheep in 51 the L.F.V. coincided with the period of maximum coyote pup growth from May to September (Barnum et a l . 1979). Food Habits The catholic nature of the coyote diet in the L.F.V. is demonstrated in the food types eaten. In general, food items may be placed into categories of minor or major importance. Of minor importance were those food items found infrequently in the diet and probably eaten opportunistically. These items included insects, garbage, wild birds, deer, raccoon, oppossum, muskrat, and domestic livestock. Neonatal cervids may form an important seasonal prey item in some areas (Bowen 1981), especially if the young are born or hidden in habitat types frequented by foraging coyotes (Reichel 1976). However, a l l scats containing deer hair were collected in the same general area and time when scavenging by coyotes on road killed adult deer was noticed. Adult deer are usually killed only by packs of coyotes and then mainly in winter when deep snow impedes locomotion (Bowen 1981; Messier and Barrette 1979). Neither of these conditions occurred in the L.F.V. Raccoons and muskrat probably do not occur in sufficient numbers to serve as staple items in the coyote diet, but are again utilized opportunistically. Oppossums are abundant, but in one instance where an oppossum was found killed by a coyote it was not eaten (S. Janson, pers. comm.). Domestic livestock, if available, are usually found in coyote diets (Bergeron and Demers 1981; Danner and Smith 1980; Fichter et al. 1955; Gier 1968; Gipson 1974; Hawthorne 1972; Kauffeld 1977; Sperry 1941). However, in most cases, as with any prey species, livestock remains in scats cannot be 52 readily separated into cases of carrion feeding and actual predation. Studies of coyote stomach contents have classified livestock as carrion i f maggots (Diptera spp.) were found in the stomach (Sperry 1941; pers. observ.) or when the coyotes died from feeding on poison bait stations (Sperry 1941). The proportions of livestock predation and livestock carrion in the coyote diet usually vary seasonally, with carrion use high in winter and predation high from spring to f a l l (Gier 1968; Korschgen 1957). Circumstantial evidence in the present study indicates that much of the poultry found in scats was scavenged by coyotes. For example, stomach contents composed entirely of chicken feet were found several times, and such items were often found when chicken manure, mixed with carcass parts, was sprayed on fields as fertilizer. Chicken was found in scats collected in F4's home range, where a poultry farmer deliberately threw out chicken carcasses for coyotes to consume. Two radio-collared coyotes (M2, M8) were shot at similar chicken dumps, and this method of chicken disposal is common in the L.F.V. Most sheep detected in scats were probably the result of predation. Sheep carcasses were generally disposed of by farmers or used as bait by the W.C.O. However, coyotes were observed to have dug up shallowly buried carcasses more than one month after burial. Predation on calving cows and newly born calves by coyotes occcurred infrequently during the study (D. Pemble, D. Nagy; pers. comm.). Two cattle dumps to which farmers hauled carcasses for disposal by coyotes were found (pers. observ.; N. Barichello, pers. comm.), and scavenging on cows that died calving was also recorded. It is possible that sheep are underestimated in the L.F.V. coyote diet, i f , prior to scat elimination, offending coyotes are killed. However, W.C.O.'s seldom caught coyotes within 48 h of sheep kills (D. Pemble, pers. 53 comm.), by which time most ingesta should have passed through the coyote's digestive system (Gier 1968). Because scat collection areas were found with no prior knowledge of coyote predation on surrounding farms, a biased esti-mate of sheep in the diet is unlikely. Rather, it appears more plausible that if only a small proportion of coyotes were killing sheep in the L.F.V., their scats would not be found simply because of the area involved. Thus, the contrasting importance of sheep in the coyote diet and coyote predation upon sheep in the L.F.V. is probably the result of extensive utilization of sheep (or other livestock) by a minority of coyotes, as found elsewhere (Althoff and Gipson 1981; Andelt and Gipson 1979; Criddle et a l . 1923; Gier 1968; Sperry 1939; Young and Jackson 1951). Three items compose the major portion of the coyote diet in the L.F.V.; vegetation (fruit and grass), rabbits, and small rodents. The seasonal use of fruit by coyotes is common (Gier 1968; Gipson 1974). Meinzer et al. (1975) reported an extreme case of fruit utilization where 46% by volume of the coyote's annual diet was fruit, with the type of fruit eaten paralleling the successive ripening of the different fruit species. Lagomorphs and small rodents form the major portion (50% to 60% by volume) of the coyote's diet throughout North America (Criddle et al. 1923; Ellis 1959; Gier 1968; Hamilton 1974; Hawthorne 1972; Korschgen 1957; Murie 1935; Reichel 1976; Sperry 1941; Weaver 1977). Coyotes generally appear to prefer lagomorphs over small rodents as a major food source. Eastern cottontails were introduced into the L.F.V. around 1952 (Cowan and Guiguet 1956) and although common are not abundant. Another species, mountain cottontails, is rare (F. Bunnell, pers. com). Cowan and Guiguet (1956) mention snowshoe hares (Lepus americanus) as present in the study area, but hares were never seen or caught in traps. The proportion of rabbits in the 54 coyote diet elsewhere is higher than found in the L.F.V., possibly due to a comparative scarcity of rabbits and higher microtine densities (Hoffman 1979) in the L.F.V. compared to other areas. Taitt and Krebs (1983) found vole populations fluctuated seasonally and annually on Westham Island in the L.F.V. The assumption that vole popula-tion dynamics on Westham Island are similar to those in the rest of the L.F.V. is probably valid (C. J. Krebs, pers. comm.). Vole densities on Westham Island have reached over 1000 ha-*, but the mean annual density from 1979 to 1982 was approximately 200 ha-*. Even if lower vole densities were found in the L.F.V. than those found on Westham Island in spring 1980 (approximately 120 voles ha--'-), voles must have been plentiful enough that alternative buffer-food sources were unnecessary. Food Habits and Livestock Predation Kauffeld (1977) studied the availability of natural prey and its rela-tionship to coyote predation on sheep pastured on range. He found that sheep varied seasonally between three and eight percent by volume in the coyote diet, and were used as a buffer species by coyotes in areas of low natural prey biomass. Seasonal changes in coyote food requirements were thought responsible for varying intensities of coyote predation during lamb-ing, spring trail to range, and late summer (Kauffeld 1977). Similarly, in the L.F.V., sheep as part of the livestock component in the diet (Figure 4B) occurred in scats only in Summer 1980, Spring 1981 and Summer 1981, corres-ponding to periods of highest sheep numbers and availability (due to lambing), and highest levels of coyote predation upon sheep (Figures 3A and 3B). 55 Livestock and small rodent utilization both peaked during the same periods (Winter 1980-81, Spring 1981). Thus, it appears that livestock were not utilized in preference to voles or as a buffer food. In contrast, fruit was eaten extensively in Fall 1980 and 1981, when voles were utilized at relatively lower levels by coyotes. However, this is probably a result of high fruit availability over a short period rather than of low vole avail-ability at the time. Coyote impact on the sheep population is not totally reflected in the diet because they sometimes k i l l two or more sheep during an attack and perhaps feed on only one. Additionally, because of the high surface area to volume ratio of small prey compared to large prey and the fact that mamma-lian prey are composed of much relatively indigestable material (hair), small prey may be over-represented in terms of weight and under-represented in terms of numbers (Floyd et a l . 1978). Thus, on a weight consumption basis, the relative importance of sheep and cattle in the coyote diet would increase slightly over small rodents and rabbits in this study. However, the extremely episodic occurrence of livestock predation, plus the fact that probable chicken carrion was the dominant domestic prey type, suggest that cattle and sheep are of minimal importance in the diet of L.F.V. coyotes. Sheep are probably utilized by individual coyotes when circumstances permit, as in the case of more easily captured young animals or carrion and in periods of nutrional stress, e.g. when pups are being reared (Till and Knowlton 1983). It is possible, however, that if i t were not for the high densities of M. townsendi, sheep farmers in the L.F.V. would experience far more losses to coyote predation. This could be the result of a greater need for an alternative food, or because coyote numbers might be less. If the feeding of pups is the primary stimulus for coyote predation on sheep, then 56 coyote density is probably not the proximal factor. Boggess et al. (1978) found no correlation between the number of coyotes claimed for bounty and the number of of sheep killed in the same county. Home Range and Movement From Home Range Trends towards small home ranges and limited tenure within home ranges were major features of radio-collared coyotes in the L.F.V. The average home range size of males (excluding M3 and M8) was the smallest reported in the literature (Andelt and Gipson 1979; Berg and Chesness 1978; Bowen 1982; Danner and Smith 1980; Gipson and Sealander 1972; Hibler 1977), and I believe female coyote home ranges are similar in size (see review in Laundre and Keller 1984). In the L.F.V., where voles are an abundant year-round food supply, home range size need not be as large as in other areas, if food supply determines home range size in coyotes. The reasons for large scale movements (M3) and the high frequency of transitory residence of other coyotes (Ml, M2, M8, Fl, F3, F5) are impos-sible to determine from the present study. Bowen (1982) found that the male coyote with the largest home range in his study (65 km2) was solitary. Bowen (1982) also had over 50% of his coyotes disperse from their home ranges, similar to the L.F.V. Territorality was not evident among the coyotes studied but, as I have no direct observations of social behaviour, social structure could not be quantified. Dispersal involves an animal leaving one home range, moving to an another area, and establishing a new home range (Hibler 1977). Such disper-sal was not documented in the L.F.V., although two areas of consecutive use by M8 were found. Bowen (1982) found that coyotes up to three years of age dispersed from their home ranges, so although young animals may be more 57 prone to such movements, age is not the only determinant. Two studies on coyotes in different habitats (Althoff 1978; Laundre 1979; both cited in Laundre and Keller 1984) suggested that mate selection may be a factor in dispersal. All but one of the radio-collared or tagged coyotes killed had moved from their original capture site. The mean distance moved was comparable to distances moved by coyotes in California (Robinson and Cummings 1951; Hawthorne 1971; both cited in Berg and Chesness 1978), but lower than dis-tances recorded in Alberta (Nellis and Keith 1976) and Utah (Hibler 1977). Activity Patterns and Movement Seasonal variation in coyote movements and activity may be due to sea-sonal changes in their reproductive cycle (Andelt and Gipson 1979; Laundre and Keller 1981) and to seasonal and diel cycles in ambient temperature (Laundre and Keller 1981). The low point of seasonal movement in the L.F.V. occurred in fa l l and winter, possibly reflecting an increase in prey availability (Todd et a l . 1981). Springer (1982) found coyotes increased their movements in winter, probably responding to a decrease in prey (insects, rodents and lagomorphs). Seasonal variation in coyote movement (although not significant) in the L.F.V. would appear to coincide with fluc-tuations in vole densities induced by cover dieoff and winter flooding (Taitt and Krebs 1983). Thus, predation by coyotes is facilitated during f a l l and winter when low cover and concentration of voles in restricted areas due to flooding results in a decrease in coyote foraging time and effort. Reichel (1976) found that coyotes selected for habitats with higher microtine populations. Lower densities of voles in spring and summer, 58 resulting from increased habitat availability together with decreased avail-ability of voles due to an increase in cover height, should result in an increase in coyote foraging movements. These relationships may be confoun-ded by the increased food demands of reproduction and pup rearing. Laundre and Keller (1981) found female coyotes decreased daily movements, while male coyotes increased daily movements during the reproductive season. Unfortun-ately, I was unable to examine this for L.F.V. coyotes due to pup mortali-ties and a transmitter malfunction. Coyotes in the L.F.V. travelled more at night, the time period when farmers lost most sheep (Henne 1975; cited in Dorrance and Roy 1976; pers. observ.). Other investigators (Andelt and Gipson 1979; Laundre and Keller 1981; Woodruff and Keller 1982) noted major peaks of activity at dawn and dusk, often with extensive movement throughout the night but with l i t t l e activity during the day. Coyotes in the L.F.V. exhibited minimal movement at dusk. Fox (Vulpes vulpes), marten (Martes americana) and kestrel (Falco  tinnunculus) activity has been found to correspond with the activity cycles of their prey (Abies 1969; Rijnsdorp et al. 1981; Zielinski et al. 1983). Voles are on average more nocturnal than diurnal (Calhoun 1945; Davis 1933; Van Home 1983). Coyote activity in the L.F.V. may therefore be related to the activity and availability of microtines, their dominant prey. Conse-quently, coyotes can take advantage of other prey if their temporal availa-bility coincides with coyote foraging periods. Leaving sheep unconfined at night therefore increases the risk of coyote predation. Estimates of total distance travelled per day can increase with the frequency of sampling (Laundre and Keller 1981). Although an increased frequency of sampling could presumably increase the total movement measured, food availability in the L.F.V. may have inversely affected movements 59 (Springer 1982; Todd et al. 1981). The major prey species for the L.F.V. coyotes were microtines, particularly M. townsendi, which maintains the highest mean densities of any North American microtine (Taitt and Krebs, in press). Such densities partially explain the limited movements of coyotes in the L.F.V. Predation by Radio-collared Coyotes None of the radio-collared coyotes were implicated in sheep losses, at least while they were within their home ranges. This may have been due to a combination of the limited tenure within home ranges, the small sample size of coyotes, or slow communication from farmers. Andelt and Gipson (1979) found one mated pair of coyotes were responsible for a l l domestic turkey (Meleagris gallopavo) losses at one farm, and Althoff and Gipson (1981) found three of 19 known coyotes were responsible for subsequent losses at the same farm. In terms of coyotes as a whole, individuals which regularly prey on livestock appear uncommon, although Connolly et al. (1976) found the majority of coyotes under experimental conditions may attack and k i l l sheep. Predation and Sheep Sheep have been bred for docility and easy management, and most of their antipredator strategies have not been selected for by breeders. Some ewes have been known to defend themselves and their lambs by stamping their front feet and rushing at coyotes (Connolly et a l . 1976; Jansen 1974), but most frequently their response is to flee. Mountain sheep (CJ. canadensis) 60 usually flee when attacked by canids, but generally run to escape terrain (Geist 1971), such as rock cliffs or steep banks which are not often avail-able to domestic sheep under pasture conditions. In addition, and of greater significance, is the fact that domestic sheep in a non-predator proof fenced pasture are essentially captive prey (Dorrance and Roy 1976). Domestic sheep are basically defenceless against predator attack and farmers should attempt to forestall a l l possible avenues of interaction between predators and sheep. Control of Coyote Predation Before presenting the more useful control methods, I will now discuss and eliminate possible techniques for reducing sheep losses to canid preda-tors, in light of coyote biology, government regulations, and sheep farming practices in the L.F.V. Coyote control techniques for possible use in the L.F.V. may be divided into lethal and nonlethal methods. Lethal control methods include poison collars, poison baits, coyote getters (a spring-loaded cyanide ejection device), shooting, denning (killing pups at densite), and trapping. The first three methods cannot currently be used in the L.F.V. because use of poison, poison bait or poison application devices for predator control within 1 km of an inhabited dwelling are illegal (Kobylnyk 1983). Large scale control by shooting poses a similar safety problem in the L.F.V. Coyotes are seldom seen in the L.F.V., also limiting the value of shooting as a control technique. Denning has proven very successful in limiting coyote predation ( T i l l and Knowlton 1983). However, denning is confounded in two ways in the 61 L.F.V. Due to the small farm size in the L.F.V., if a coyote den can be found i t is usually on a neighbour's wooded lot rather than on the farmer's cleared field. Secondly, neighbours may not allow the killing of coyotes on their property. Consequently, denning is not a government control method in the L.F.V. Trapping is the predominant control method currently used by W.C.O.'s in the L.F.V. Time consuming and expensive, it is also unselective, with many non-target species being killed or injured (pers. observ.). Arthur (1981) in a public survey found that trapping was judged as the cruelest method of predator control, a viewpoint echoed by many people in the L.F.V. (pers. observ.). A public anti-trapping attitude may conceivably limit trapping's further use as a predator control technique. Trapping by W.C.O.'s is only permitted, by government policy, after sheep or other large livestock (not poultry), have been killed or harrassed by coyotes. This of course means that the producer usually has lost sheep before coyote preda-tion can be controlled. With lethal predator control, losses may decrease for a period, but more coyotes can move into the area and losses may then resume. The L.F.V. coyote population was found to contain many transient individuals who could establish themselves in a vacant area. Costs would therefore be contin-ually incurred by the farmer, the taxpayer, and the consumer. A variety of nonlethal control methods have been tested in recent years with varying degrees of success. They include coyote reproductive inhibi-tors, olfactory or auditory aversion agents, emetics, guard dogs, predator-proof fencing, and night confinement of sheep. Reproductive inhibitors (Balser 1964) are poor agents for coyote population suppression because of a low bait pick-up by coyotes, primarily due to bait pick-up by other species 62 (Linhart et. al. 1968: cited in Sterner and Shumake 1978). Chemical aver-sion agents are unlikely to prevent coyote predation. Lehner et al. (1976), found no chemical which would repel coyotes or dogs consistently without also adversely affecting sheep. Equivocal results were found in other studies (Swanson et al. 1975; Swanson and Scott 1973), and Sterner and Shumake (1978) noted that coyote responses to offensive odours or sounds may only be neophobic. Lithium chloride (LiCl) is used as an emetic for inducing a taste aver-sion to sheep in coyotes to prevent subsequent coyote predation (Gustavson et a l . 1974). The use of LiCl therefore depends upon the maintenance of a resident coyote population (Burns 1983). However, in the L.F.V., over 50% of my radiocollared coyotes were or became transients, while additional coyotes were killed by W.C.O.'s, creating dispersal sinks (Lidicker 1975). The high coyote mobility in the L.F.V. therefore precludes the use of LiCl. In addition, Bourne and Durrance (1982) suggest that a large area has to be baited i f LiCl is to be effective. This would be very difficult to imple-ment in the L.F.V. because not only would bait pick-up be low (due to oppossums, raccoons, and domestic animals), but the diversity of land owner-ship and utilization would hamper any such government program. Finally, LiCl aversion in coyotes has yet to be shown to be universal (reviewed in Burns 1983). The following nonlethal control methods are suited to the L.F.V., because they are relatively easy to implement and, in contrast to previously discussed controls, they are not an a. posteriori solution. If properly applied they prevent the predator-prey interaction and thus reduce losses even further. 63 Guard dogs have been found to effectively reduce or eliminate coyote predation on sheep (Linhart et a l . 1979), although effectiveness varies between individual dogs (U.S.F.W.S. 1978). Green et al. (1984) found that properly trained guard dogs were economically viable if they saved enough sheep annually to defray a cost of $1075 (CDN.) and 119 h labour their first year and $369 plus labour each year thereafter. Once established this is equivalent to an annual loss of four lambs in the L.F.V. Such loss rates were found on individual farms in the L.F.V. Thus, guard dogs are an economic and viable option to reduce sheep losses in the L.F.V. Fencing dingos (C. f_. dingoe) out of sheep pastures has long been prac-ticed in Australia (Pearse 1945), and installation of predator proof fencing is suggested as a primary anti-coyote strategy in the L.F.V. The cheapest of the available fence types is the electric fence and if the design sugges-tions of Armstrong et al. (1981), Dorrance and Bourne (1980), Gates et al. (1981), Linhart et al. (1982), and Porter (1983) are followed, i t offers the most effective barrier to both coyotes and dogs. The value of fencing was obscured in the analysis conducted here because very few farms had good fences; those which did, did not lose sheep to coyotes. For example, one L.F.V. farm consistently lost sheep to coyotes until an electric, predator-proof fence was erected. Predation losses ceased until a fire destroyed the house and generator, after which coyotes resumed killing sheep. With a new generator, losses again ceased. Predator-proof fencing may thus have a high prophylactic value which would be obscured in this analysis and should be evaluated for inclusion in any sheep management plan. Many producers with otherwise sound fences may hesitate before install-ing an electric predator-proof fence. Economical alternatives available are: installation of an electric tripwire around the fence perimeter; 64 staking the lowest fence strand to the ground in conjunction with blocking ditches and streams crossed by the fence, and staking the bottom fence wire to prevent coyotes or dogs crawling under the fence. An added advantage of new fencing however, is that it may be claimed against land taxes. The major method for preventing coyote predation, the easiest to imple-ment, and that used by half the farmers surveyed in 1981 is night confine-ment of sheep in a predator-proof enclosure. The length and expense of fencing a predator-proof corral is obviously less than fencing an entire pasture. Alternatively, as most farmers already have a lambing barn, sheep can be housed in pre-existing buildings every night at no additional expense, as long as they are predator-proof. In addition to the above suggestions, it should be obvious that proper disposal of livestock carcasses should be practiced. Not the least reason being that coyotes feeding initially only on sheep carrion may associate i t with live sheep and subsequently k i l l sheep themselves (Lehner 1976). Thus, in contrast to the extensive use of W.C.O.'s by some sheep producers, sagacious employment of management practices to prevent coyote predation is an economical and effective substitute. Coyotes will always be present, but the cost of coyote control to the taxpayer will decrease, costs of veterinary bills and replacement stock for the producer will be reduced, and ultimately the cost to the consumer may be moderated. As no option is foolproof, predator control may s t i l l have to be used to some extent, but the need can be substantially decreased and at less cost than present methods. 65 Control of Dog Predation Due to the unpredictable nature of dog predation, s p e c i f i c methods of preventing dog predation are not given. However, exclusion management to reduce sheep a v a i l a b i l i t y to coyotes should also s i g n i f i c a n t l y reduce or prevent dog predation. 66 CONCLUSIONS 1. Annual sheep losses from coyote predation on Individual farms indicate geographic (size of pasture) and management factors (size of ewe flock, night confinement, and carcass disposal) influence susceptibility of specific farms to coyote predation. 2. Dog predation is an almost random event with few farms suffering annual attacks. No factors predisposing farms to dog predation could be found. 3. Predation was the second largest cause of death among ewes and lambs. Coyotes killed the majority of sheep lost to predators. 4. We could not identify any coyote x dog hybrids in the sample analyzed. 5. Home ranges and movements of coyotes tended to be smaller than in other areas of North America, which are probably a function of high microtine populations in the L.F.V. 6. Coyotes were most active at night when most sheep losses occurred. 7. Food habits of coyotes in the L.F.V. were characteristically diverse. Small rodents comprised 70% of the diet (scat volume) while domestic stock (mostly chicken) constituted only 4.3%. Sheep represented only 0.2%. 8. Effective, economical solutions for reducing predator losses are avail-able and suitable for L.F.V. sheep producers. 67 RECOMMENDATIONS 1. Sheep should be confined at night in a secure, predator-proof paddock or building. 2. Sheep carcasses and those of other livestock should preferably be removed from the farm or buried under at least 60 cm of earth and covered with lime. If the carcass is to be used for trapping by W.C.O.'s, remove the carcass after trapping and bury (rebury) as above. Good rapport with neighbours may discourage carrion dumps. 3. Large sheep flocks appear to be more susceptible to predation and on large pastures losses may go undetected. Ewes and lambs should be checked daily if sheep are not confined at night. Missing animals can then be noticed and searched for immediately. Weekly (minimum) patrols of fencelines will reduce predator access problems. 4. Electric fencing appears to be a cost efficient method of eliminating predation and should be encouraged as a primary method of controlling coyote access to sheep, especially when starting a new sheep farm. 5. Farmers with property adjacent to large areas of bush should search the area (given permission by the landowner) for coyote dens in the spring. Elimination of pups will reduce future coyote predation. 6. If necessary, a continual assessment of study recommendations and effectiveness of controls can be obtained by having W.C.O.'s document the size of the sheep flock, pasture size, method of carcass disposal, and if night confinement is practiced on each farm where a coyote damage complaint is made. 68 LITERATURE CITED Abies, E. D. 1969. Activity studies of red foxes in southern Wisconsin, J. Wildl. Manage., 33:145-153. Adorjan, A.S. and G.B. Kolenosky. 1969. A manual for the identification of hair of selected Ontario mammals. Res. Rpt. No. 90, Ontario Min. Nat. Res., 64 pp. Althoff, D.P. and P.S. Gipson. 1981. Coyote family spacial relationships with reference to poultry losses. J. Wildl. Manage., 45:641-649. Andelt, W.F. and P.S. Gipson. 1979. Home range, activity and daily move ments of coyotes. J. Wildl. Manage., 43:944-951. Anonymous. 1982. Recognize and control predator attacks on your livestock. United Grain Growers, Winnipeg. 4 pp. Armstrong, R.H., CH. Banks and M.P. G i l l . 1981. A guide to electic fenc-ing relating to principles, installation and safety. H i l l Farming Research Organization, Midlothian, U.K. 41 pp. Arthur, L.M. 1981. Coyote control: The public response. J. Range. Manage., 34:14-15. Balser, D.S. 1964. Management of predator populations with antifertility agents. J. Wildl. Manage., 28:352-358. Balser, D.S. 1974. An overview of predator-livestock problems with emphasis on livestock losses. Trans. 39th N. Amer. Wildl. and Nat. Res. Conf., pp. 292-300. Barnum, D.A., J.S. Green, J.T. Flinders and N.L. Gates. 1979. Nutritional levels and growth rates of hand-reared coyote pups. J. Mamm., 60: 820-923. Berg, W.E. and R.A. Chesness. 1978. Ecology of coyotes in northern Minne-sota, 1978. pp. 229-248, In, M. Bekoff (ed.) "Coyotes: Biology, behaviour and management." Academic Press, New York. 384 pp. Bergeron, J.M. and P. Demers. 1981. Le regime alimentaire du coyote et du chien errant dans le sud du Quebec. Can. Field-Nat., 95:172-177. Blair, B.J. and T.W. Townsend. 1983. Dog predation of domestic sheep in Ohio. J. Range Manage., 36:527-528. Boggess, E.K., R.D. Andrews and R.A. Bishop. 1978. Domestic animal losses to coyotes and dogs in Iowa. J. Wildl. Manage., 42:362-372. Boggess, E.K., F.R. Henderson and C.W. Spaeth. 1980. Managing predator problems: Practices and procedures for preventing and reducing livestock losses. Coop. Ext. Serv., Kansas State Univ. 19 pp. 69 Bond, R.M. 1939: Coyote food habits on the Lava Beds National Monument. J. Wildl. Manage., 3:180-198. Bourne, J. and M.J. Dorrance. 1982. A field test of lithium chloride aversion to decrease coyote predation on domestic sheep. J. Wildl. Manage., 46:235-239. Bowen, W.D. 1981. Variations in coyote social organization: The influence of prey size. Can. J. Zool., 59:639-652. Bowen, W.D. 1982. Home range and spacial organization of coyotes in Jasper National Park, Alberta. J. Wildl. Manage., 46:201-216. British Columbia Ministry of Agriculture. 1976. Climate of British Columbia. Tables of Temperature, Precipitation, and Sunshine. B.C. Min. of Agric. 82 pp. Burns, R.J. 1983. Coyote predation aversion with lithium chloride: Management implications and comments. Wildl. Soc. Bull., 11:128-133. Burt, W.H. 1943, Territoriality and home range concepts as applied to mammals. J. Mamm., 24:346-352. Cain, S.A., J.A. Kadlec, D.L. Allen, R.A. Cooley, M.G. Hornocker, A.S. Leopold and F.H. Wagner. 1972. Predator control - 1971. Rep. to Counc. Environ. Qual., Dept. Inter, by Advis. Comm. Predator Control. Univ. of Michigan Press, Ann Arbor. 207 pp. Calhoun, J.B. 1945. Diel activity and rhythms of the rodents Microtus  ochrogaster and Sigmoden lispidus hispidus. Ecol., 26:251-273. Camenzind, F.J. 1978. Behavioral ecology of coyotes on the National Elk Refuge, Jackson, Wyoming. pp. 267-296, In, M. Bekoff (ed.) "Coyotes: Biology, behavior and management. Academic Press, New York. 384 pp. Chesnutt, S. 1981. Preliminary food habits of coyotes in a mixed farming region. Unpubl. B.Sc. thesis., Univ. of B.C., Vancouver, B.C. 42 pp. Connolly, G.E. 1979. The toxic collar for selective removal of coyotes that attack sheep. Fourth, Fifth and Sixth progress reports. U.S.F.W.S. Denver, Colo. 47 pp. Connolly, G.E. and W.M. Longhurst. 1975. The effects of control on coyote populations. Univ. of Calif., Davis, Div. Agric. Sci. Bull. No. 1872. 37 pp. Connolly, G.E., R.M. Timm, W.E. Howard and W.M. Longhurst. 1976. Sheep killing behavior of captive coyotes. J. Wildl. Manage., 40:400-407. 70 Criddle, N., E. Criddle and S. Criddle. 1923. The coyote in Manitoba. N Can. Field-Nat., 37:41-45. Cowan, I.M., and C.J. Guiget. 1956. The mammals of British Columbia. B.C. Prov. Museum, 414 pp. Danner, D.A., and N.S. Smith. 1980. Coyote home range, movement and relative abundance near a cattle feedyard. J. Wildl. Manage., 44:484-487. Davis, D.H. 1933. Rhythmic activity in the short-tailed vole, Microtus. J. Anim. Ecol., 2:232-238. DeCalesta, D.S. 1978. Documentation of livestock losses to predators in Oregon. Spec. Rept. No. 501, Oregon State Univ. Ext. Serv. 20 pp. Dorrance, M.J. 1983. A philosophy of problem wildlife management. Wildl Soc. Bull., 11:319-324. Dorrance, M.J. and J. Bourne. 1980. An evaluation of anticoyote electric fencing. J. Range Manage., 33:385-387. Dorrance, M.J. and L.D. Roy. 1976. Predation losses of domestic sheep in Alberta. J. Range Manage., 29:457-460. E l l i s , R.J. 1959. Foodhabits and control of coyotes in northcentral Oklahoma. Oklahoma State Univ. Publ. 56:1-31. Engelman, L. 1981. Stepwise logistic regression, pp. 330-344. In, W.J. Dixon (ed.) "BMDP Statistical Software." Univ. of Calif. Press, Berkeley, 726 pp. Fichter, E., G. Schildman and J.H. Sather. 1955. Some feeding patterns of coyotes in Nebraska. Ecol. Monographs 25:1-37. Fienberg, S.E. 1977. The analysis of cross-classified categorical data. M.I.T. Press, Cambridge. 317 pp. Floyd, T.J., L.D. Mech and P.A. Jordan. 1978. Relating wolf scat content to prey consumed. J. Wildl. Manage., 42:528-532. Freeman, R.C. and J.H. Shaw. 1979. Hybridization in Canis (Canidae) in Oklahoma. Southwestern Nat. 24:485-499. Gates, N.L., J.F. Rich, D.D. Godtel and C.V. Hulet. 1978. Development and evaluation of anticoyote electric fencing. J. Range Manage., 31: 151-153. Gee, C.K., R.S. Magleby, W.R. Bailey, R.L. Gum and L.M. Arthur. 1977. Sheep and lamb losses to predators and other causes in the western United States. U.S. Agric. Econ. Rep. No. 369: 41 pp. 71 Geist, V. 1971. Mountain sheep: A study in behavior and evolution. Univ. of Chicago Press, Chicago. 383 pp. Gier, H.T. 1968. Coyotes in Kansas. Kansas State Univ., 118 pp. Gipson, P.S. 1974. Food habits of coyotes in Arkansas. J. Wildl. Manage., 38:848-853. Gipson, P.S., and J.A. Sealander. 1972. Home range and activity of the coyote (Canis latrans frustror) in Arkansas. Proc. Southeast Assoc. Game and Fish Comm., 26:82-95. Gipson, P.S., J.A. Sealander, and J.E. Dunn. 1974. The taxonomic status of wild Canis in Arkansas. Syst. Zool., 23:1-11. Gipson, P.S., I.K. Gipson, and J.A. Sealander. 1975. Reproductive biology of wild Canis in Arkansas. J. Mamm., 56:605-612. Gluesing, E.A. 1977. Sheep behavior and vulnerability to coyote preda-tion. Unpub. Ph.D. thesis, Utah State Univ., Logan, Utah. 121 pp. Green, J.S., R.A. Woodruff and T.T. Tueller. 1984. Livestock guarding dogs for predator control: Costs, benefits and practicality. Wildl. Soc. Bull., 12:44-50. Gustavson, C.R., J. Garcia, W.G. Hankins and K.W. Rusiniak. 1974. Coyote predation control by aversive conditioning. Science., 184:581-583. Gustavson, C.R., J.R. Jowsey and D.N. Milligan. 1982. A 3-year evalua-tion of taste aversion coyote control in Saskatchewan. J. Range Manage., 35: 57-59. Halperin, M., W.C. Blackwelder and J.I. Verter. 1971. Estimation of the multivariate logistic risk function: A comparison of the discrimi-nant function and the maximum likelihood approaches. J. Chron. Dis. 24:125-158. Hamilton, W.J. Jr. 1974. Food habits of the coyote in the Adirondacks. N.Y. Fish and Game. J., 21:177-181. Harestad, A.S. 1981. Computer analysis of home range data. Fish and Wildl. Br., Bull. No. B-ll., B.C. Min. Envir. 26 pp. Hatter, I.W. 1984. Effects of wolf predation on recruitment of black-tailed deer on northeastern Vancouver Island. Unpub. M.Sc. thesis, Univ. Idaho, Moscow, Idaho. 156 pp. Hawthorne, V.M. 1972. Coyote food habits in Sagehen Creek Basin, N.E. California. Calif. Fish and Game., 58:4-12. Hibler, S.J. 1977. Coyote movement patterns with emphasis on home range characteristics. Unpub. M.Sc. thesis, Utah State Univ., Logan, Utah. 67 pp. 72 Hilton, H.H. 1976. The physical characteristics, taxonomic status and food habits of the eastern coyote in Maine. Unpub. M.Sc. thesis, Univ. of Maine, Orono, Maine. 67 pp. Hoffman, S.W. 1979. Coyote prey relationships in Curlew Valley during a period of low jackrabbit density. Unpub. M.Sc. thesis, Utah State Univ., Logan, Utah. 139 pp. Holland, S.S. 1976. Landforms of British Columbia. A physiographic out-line. Bull. No. 48. B.C. Dept. Mines and Petroleum Res. 138 pp. Jansen, CD. 1974. Behavior patterns observed in coyote-sheep inter-actions. Unpub. M.Sc. thesis, Colorado State Univ., Denver, Colorado, 57 pp. Johnson, M.K. and R.M. Hansen. 1979. Coyote food habits on the Idaho National Engineering Laboratory. J. Wildl. Manage., 43:951-955. Jones, J.M. and A. Woolf. 1983. Relationship between husbandry practices and coyote use of swine in western central Illinois. Wildl. Soc. Bull., 11:133-135. Kauffeld, J.D. 1977. Availability of natural prey and its relationship to coyote predation on domestic sheep. Unpub. M.Sc. thesis. Univ. Nevada, Los Vegas, Nevada, 144 pp. Kennelly, J.J. 1978. Coyote reproduction, pp. 73-96, In, M. Bekoff, (ed.) "Coyotes: Biology, behavior and management." Academic Press, New York. 384 pp. Kennelly, J.J. and B.E. Johns. 1976. The estrous cycle of coyotes. J. Wildl. Manage., 40:746-750. Kennelly, J.J., B.E. Johns, CP. Breidenstein and J.D. Roberts. 1977. Predicting female coyote breeding dates from fetal measurements. J. Wildl. Manage., 41: 746-750. Klebenow, D.A. and K. McAdoo. 1976. Predation on domestic sheep in north eastern Nevada. J. Range. Manage., 19:96-100. Knowlton, F.F. 1972. Preliminary interpretation of coyote population mechanics with some management implications. J. Wildl. Manage., 30:369-382. Kobylnyk, R.W. 1983. Permit No. 125-23* SPL-83/87. Pesticide Control act. Special use and restricted permit for use of sodium mono-fluoroacetate (Compound 1080) predacide (P.CP. Act. Reg. No. 17664) for reactive control of coyotes and wolves for protection of domestic livestock and human safety, selected wildlife management units within British Columbia. 4 pp. Korschgen, L.J. 1957. Food habits of the coyote in Missouri. J. Wildl. Manage. , 21:424-435. 73 Krajina, V.J. 1969. Ecology of forest trees in British Columbia. Ecology of Western North America. 1:1-146. Lamprecht, J. 1978. On diet, foraging behaviour and interspecific food competition of jackals in the Serengeti National Park, East Africa. Z. Saugetierk., 43:210-222. Laundre, J.W. and B.L. Keller. 1981. Home range use by coyotes in Idaho. Anim. Behav., 29:449-461. Laundre, J.W. and B.L. Keller. 1984. Home range size of coyotes: A critical review. J. Wildl. Manage. 48:127-139. Lawrence, B, and W.H. Bossert. 1967. Multiple character analysis of Canis lupus, latrans and familiaris with a discussion of the relationship of Canis niger. Amer. Zool., 7:223-231. Lehner, P.N. 1976. Coyote behavior: Implications for management. Wildl. Soc. Bull., 4:120-126. Lehner, P.N, R. Krumm and A.T. Cringan. 1976. Test for olfactory repellants for coyotes and dogs. J. Wildl. Manage., 40:145-150. Lidicker, W.Z., Jr. 1975. The role of dispersal in the demography of small mammals. pp. 103-128, In, F.B. Golley, K. Petrusewisz and L. Ryszkowski (eds.) "Small Mammals: Their productivity and popula-tion dynamics." Int. Biol. Progr. Vol. 5., Cambridge Univ. Press, London. 491 pp. Linhart, S.B., R.T. Sterner, T.C. Carrigan, and D.R. Henne. 1979. Komondor guard dogs reduce sheep losses to coyotes: A preliminary evaluation. J. Range Manage., 32:238-241. Linhart, S.B., J.D. Roberts, and G.J. Dasch. 1982. Electic fencing reduces coyote predation on pastured sheep. J. Range Manage., 35:276-281. McAdoo, J.K. and D.A. Klebenow. 1978. Predation on range sheep with no predator control. J. Range Manage., 31:111-114. Mahan, B.R., P.S. Gipson and R.M. Case. 1978. Characteristics and distri-bution of coyote x dog hybrids collected in Nebraska. Am. Midi Nat., 100:408-415. Meduna, R.L. 1977. Relationships between sheep management and coyote predation. Unpub. M.Sc. thesis, Kansas State Univ., Kansas, 140 pp. Meinzer, W.P., D.N. Ueckert and J.T. Flinders. 1975. Foodniche of coyotes in the rolling plains of Texas. J. Range Manage., 28:22-28. Mengel, R.M. 1977. A study of dog-coyote hybrids and implications concerning hybridization in Canis. J. Mamm., 52:318-335. 74 Messier, F., and C. Barrette. 1979. Etude de la predation du cerf de virginie par le coyote dans le Ravage d'Armstrong, Beauce Sud. Ministere du Tourisme, de la Chasse et de la Peche, Quebec. 167 pp. Mohr, C. 1947. Table of equivalent populations of North American small mammals. Am. Midi. Nat., 37:223-249. Moore, T.D., L.E. Spence and C.E. Dugnolle. 1974. Identification of the dorsal guard hairs of some mammals of Wyoming. .Bull. No. 14., Wyoming Fish and Game Dept. 177 pp. Murie, O.J. 1935. Food habits of the coyote in Jackson Hole, Wyoming. U.S. Dept. of Agric, 24 pp. Nass, R.D. 1977. Mortality associated with sheep operations in Idaho J. Range Manage., 30:253-258. Nellis, C.H. and L.B. Keith. 1976. Population dynamics of coyotes in central Alberta, 1964-1968. J. Wild. Manage., 40:389-399. O'Gara, B.W., K.C. Brawley, J.R. Munoz and D.R. Henne. 1983. Predation on domestic sheep on a western Montana ranch. Wildl. Soc. Bull., 11: 253-264. Ogden, R., M.H. Wallace, C.J. Kaiser, J.M. Lewis, F.C. Hinds and M.E. Mansfield. 1978. Coyote predation in sheep at the Dixon Springs Agric. Centre., D.S.A.C., 6:215-218. Pearse, E.H. 1945. Sheep, farm and station management. Pastoral Review, Pty. Ltd., Melbourne. 88 pp. Pimlott, S.H., J. Shannon and G.B. Kolenosky. 1969. The ecology of the timber wolf in Algonquin Provincial Park. Ont. Dept. Lands and Forests. 92 pp. Pozen, M.W., R.W. D'Agostino, H.P. Selker, P.A. Sytkowski and W.B. Wood, Jr. 1984. A predictive instrument to improve coronary-care-unit admission practices in acute eschemic heart disease. N. Eng. J. Med., 310:1273-1278. Porter, G. 1983. Electric fencing manual: A practical aid to constructing permanent electric fences. B.C. Min. of Agric. and Food, Victoria B.C. 30 pp. Press, J. and S. Wilson. 1978. Choosing between logistic regression and discriminant analysis. J. Amer. Stat. Assoc. 73:699-705. Provincial Problem Wildlife Management Committee. Undated. Predator Problems? B.C. Min. of Agric. 4 pp. 75 Reichel, J.D. 1976. Coyote prey relationships in the National Bison Range. Unpub. M.Sc. thesis, Univ. Montana, Misoula, Montana. 83 pp. Rijnsdorp, A, S. Daan, and C. Dijkstra. 1981. Hunting in the kestrel (Falco tinnunculus), and the adaptive significance of daily habits. Oecologia 50:391-406. Robel, R.J., A.D. Dayton, F.R. Henderson, R.L. Meduna and C.W. Spaeth. 1981. Relationships between husbandry methods and sheep losses to canine predators. J. Wildl. Manage., 45: 894-911. Roy, L.D. and J. Dorrance. 1976. Methods of investigating predation of domestic livestock. Alta. Agric, 54 pp. Schaefer, J.M., R.D. Andrews and J.H. Dinsmore. 1981. An assessment of coyote and dog predation on sheep in southern Iowa. J. Wildl. Manage., 45:883-893. Sperry, C.C. 1939. Foodhabits of pegleg coyotes. J. Mamm., 20:191-194. Sperry, C.C. 1941. Food habits of the coyote. U.S. Dept. of Int., Wildl. Res. Bull. 4. 70 pp. Springer, J.T. 1982. Movement patterns of coyotes in southcentral Wash-ington. J. Wildl. Manage., 46:191-200. Sterner, R.T. and S.A. Shumake. 1978. Coyote damage-control research: A revue and analysis. pp. 297-326, In, M. Bekoff (ed.) "Coyotes: Biology, behavior and management." Academic Press, New York. 384 pp. Swanson, U.B., D.A. Cramer and M.J. Jankousky. 1975. Protection of sheep with olfactory predator repellants. Research highlights of the Anim. Sci. Dept., Colo. State Univ. Exper. Stat. Gen. Ser. 948, pp. 8-9. Swanson, U.B. and G.E. Scott. 1973. Livestock protectors for sheep pre-dator control. Proc. West. Sec. Amer. Soc. Anim. Sc., 24:34-36. Taitt, M.J. and C.J. Krebs. 1983. Predation, cover and food manipulations during a spring decline of Microtus townsendii. J. Anim. Ecol., 52:837-848. Taitt, M.J. and C.J. Krebs. (in press). Population dynamics and cycles. In, R.H. Tamanin (ed.) "Biology of New World Microtus." T i l l , J.A. and F.F. Knowlton. 1983. Efficacy of denning in alleviating coyote depredations upon domestic sheep. J. Wildl. Manage., 47: 1018-1025. 76 Todd, A.W. and L.B. Keith. 1976. Responses of coyotes to winter reduc-tions in agricultural carrion. Alta. Fish and Wildl. Div., Wildl. Tech. Bull. 5:32 pp. Todd, A.W., L.B. Keith and C.A. Fischer. 1981. Population ecology of coyotes during a fluctuation of snowshoe hares. J. Wildl. Manage., 45:629-640. Truett, J., J. Cornfield and W. Kannell. 1967. A multivariate analysis of the risk of coronary heart disease in Farmingham. J. Chron. Dis. 20:511-524. U.S.F.W.S. 1978. Predator damage in the west. A study of coyote manage-ment alternatives. U.S. Dept. Int., 168 pp. Valentine, K.W.B., P.N. Sprout, T.E. Baker, and L.M. Lavkulich. 1978. The soil landscapes of British Columbia. Resource Analysis Br., Min. Envir., Victoria, B.C. 197 pp. Van Horne, B. 1983. Summertime activity pattern of the long-tail vole in Alaska. Holarctic Ecol., 6:199-203. Wade, D.A. and G.E. Connolly. 1980. Coyote predation on a Texas goat ranch. Texas Agric. Progress., 26:12-16. Weaver, J.L. 1977. Coyote-food-base relationships in Jackson Hole, Wyom-ing. Unpub. M.Sc. thesis, Utah State Univ., Logan, Utah. 88 pp. Wong, B. 1984. Use of discriminant function in taxonomic and sex deter-mination of Canids from the Okanagan and Fraser Valley. Unpub. B.Sc. thesis, Univ. of B.C., Vancouver, B.C. 49 pp. Woodruff, R.A. and B.L. Keller. 1982. Dispersal, daily activity and home range of coyotes in southeastern Idaho. North-West. Sc., 56: 199-207. Young, S.P. and H.H.T. Jackson. 1951. The clever coyote. The Stackpole Co., Harrisburg. 411 pp. Zielinski, W.J., W.D. Spencer and R.H. Barrett. 1983. Relationships between food habits and activity patterns of pine martens. J. Mamm., 64:387-396. 77 APPENDIX I QUESTIONNAIRE Name: Address: Phone number: Number of acres Number of acres uncleared Number of acres i n sheep use Livestock Number of: goats poultry c a t t l e pigs horses dogs other Number of ewes , lambs ewes , lambs Breed of sheep Sheep Information Peak month of lambing Sheep confined at night: yes no Night l i g h t s on sheep: yes no Be l l s on sheep? Use of working dogs or guard dogs for sheep Sheep and lambs l o s t i n 1980 (weather, disease, old age, lambing, predators-type) Ewes Lambs To 78 Sheep and lambs lost in 1981 Ewes Lambs To Method of carcass disposal: Sketch of Farm: Include house, barns, fence, size of pastures, which pastures are used, night pasture, streams, ponds, roads, and areas of pradator loss. Type and height of undergrowth Minimum distance to cover from pasture Minimum distance to pasture Maximum distance to pasture Topography: 1-flat 5-hilly Type of fence Condition of fence: P / G 79 APPENDIX II C l a s s i f i c a t i o n of Cover Types Class Name 1 Dense Bush 2 Open Coniferous 3 Dense Coniferous Description Dense stands of red alder (Alnus rubra), salmon-berry (Rubus s p e c t a b i l i s ) and red huckleberry (Vacinnium parvifolium). Douglas f i r (Pseudotsuga menziesii) stands with minimal undergrowth. Red cedar (Thuj a p l i c a t a ) and broadleafed maple (Acer macrophyllum) with heavy growths of swordf em (Polystichum munitum), bracken fern (Pterldium aquilinum), stinging n e t t l e s ( U r t i c a  l y a l l i i ) and vine maple (Acer circinatum). 4 5 Open Alder Open Poplar Red alder, northern black cottonwood (Populus  trichocarpa) and grass (Gramineae spp.). Northern black cottonwood, willow ( S a l i x spp.) and grass (Gramineae spp.). 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



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"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items