UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Influences of habitat interspersion on habitat use by Columbian black-tailed deer Kremsater, Laurie Lynn 1989

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

Item Metadata

Download

Media
831-UBC_1989_A6_7 K73.pdf [ 5.94MB ]
Metadata
JSON: 831-1.0097526.json
JSON-LD: 831-1.0097526-ld.json
RDF/XML (Pretty): 831-1.0097526-rdf.xml
RDF/JSON: 831-1.0097526-rdf.json
Turtle: 831-1.0097526-turtle.txt
N-Triples: 831-1.0097526-rdf-ntriples.txt
Original Record: 831-1.0097526-source.json
Full Text
831-1.0097526-fulltext.txt
Citation
831-1.0097526.ris

Full Text

INFLUENCES OF HABITAT INTERS PERSION ON HABITAT USE BY COLUMBIAN BLACK-TAILED DEER by LAURIE LYNN KREMSATER B . S . F . , University of Br i t i sh Columbia, 1985 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (DEPARTMENT OF FORESTRY) We accept this thesis as conforming to the required standard UNIVERSITY OF BRITISH COLUMBIA APRIL 1989 Laurie Lynn Kremsater, 1989 In presenting this thesis in partial fulfilment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Department of Pores fry The University of British Columbia Vancouver, Canada Date Ajonl 3.1 \W DE-6 (2/88) ABSTRACT Use of forage, cover, and border habitat by Columbian black-tai led deer (Odocoileus hemionus columblanus (Richardson)) was examined at two levels of selection: within home ranges and during home range establishment. Patterns of habitat use were evaluated in re lat ion to changing seasons, different migratory behaviours, and areas of intensive deer use (defined by concentrations of radio locations) . Relative use did not d i f fer from relat ive a v a i l a b i l i t y for forage, cover, and border habitats. A v a i l a b i l i t y of those habitats, however, changed seasonally as deer home ranges changed or different intensit ies of deer use were examined. Cover and border habitats, part icu lar ly borders between old-growth and second-growth forests, were more available in winter than, in summer home ranges. Areas receiving intensive deer use were characterized by more border and cover habitat than areas of less intensive use. Because use was d irect ly proportional to a v a i l a b i l i t y , changing a v a i l a b i l i t y suggested that habitat selection occurred as home ranges were established. Comparisons of forage, cover, and border composition in actual home ranges and areas where home ranges potential ly could have been located suggested preference for cover and border habitats. These comparisons, however, did not indicate disproportionately high use of interspersed habitats, perhaps because of the high degree of habitat interspersion in the study area. TABLE OF CONTENTS Page ABSTRACT i i LIST OF TABLES V LIST OF FIGURES v i LIST OF APPENDICES X ACKNOWLEDGMENTS x i GENERAL INTRODUCTION 1 USE OF FORAGE, COVER, AND BORDER HABITATS IN DEER HOME RANGES INTRODUCTION 3 STUDY AREAS 5 METHODS Deer locations 10 Mapping . . 12 Tests for independence. 13 A v a i l a b i l i t y and use of forage and cover habitats . . .15 A v a i l a b i l i t y and use of border habitats 16 Deer distributions around borders . . .18 RESULTS Independence tests . . .21 A v a i l a b i l i t y and use of forage and cover habitats . . .21 A v a i l a b i l i t y and use of border habitat 27 Deer distributions around borders 34 DISCUSSION 44 A v a i l a b i l i t y and use of forage and cover habitats. . .44 A v a i l a b i l i t y and use of border habitats 48 LITERATURE CITED 53 i V INFLUENCES OF FORAGE, COVER, AND BORDER HABITATS ON HOME RANGE ESTABLISHMENT INTRODUCTION 57 STUDY AREA .59 METHODS Deer locations 60 Mapping 60 Actual home ranges 61 Potential home ranges 61 Habitat composition in actual and potential home ranges 62 Deer distributions around borders . . . . . . 6 5 RESULTS Actual home ranges.. 67 Habitat composition in actual and potential home ranges 67 Forage and cover habitats 67 Border habitats ..70 Deer distributions around borders 72 DISCUSSION 77 Available forage, cover, and border habitat 77 Deer distributions around borders 78 LITERATURE CITED 82 SUMMARY . .84 Within home ranges 84 During home range establishment. 86 Management implications 87 APPENDIX I ..88 APPENDIX II 90 V LIST OF TABLES Page Table 1. Table 2. Table 3. Table 4. Table 5. Table 6 Table 7. Characteristics of the Nanaimo River and Nimpkish Valley study areas Results of use-avai labi l i ty comparisons using Chi-square tests with Bonferroni confidence intervals (a=o.05), for forage (clearcut or meadow) and cover (second -growth and old-growth forest) habitats for 164 black-tai led deer home ranges at Nanaimo River, B.C 26 Regression equations of use on ava i lab i l i ty of forage (clearcut or meadow) and cover (second-growth and old-growth forest) habitats in 164 black-tai led deer home ranges at Nanaimo River, B.C 28 Relative use and relat ive ava i l ab i l i t y of 90-m border habitat during summer, spring, and winter for 50% and 90% home ranges of 7 black-tai led deer at Nimpkish Valley, B.C 32 Results of use-avai labi l i ty comparisons using Chi-square tests with Bonferroni confidence intervals ( =0.05) for 90-m border habitat in 50% and 90% black-tai led deer home ranges for summer and winter at Nanaimo River, B.C 35 Regression equations for use on ava i lab i l i ty of 90-m border habitats for 81 summer and 83 winter 50% black-tai led deer home ranges at Nanaimo River, B . C . . . 37 Results of Bonferroni confidence intervals (a=o.05) for comparisons of actual and randomly-distributed locations of black - ta i l ed deer around borders. Locations were grouped for a l l deer . . . . . 74 V i Figure 1. Figure 2. Figure 3. Figure 4. Figure 5.(a) (b) Figure 6. Figure 7. LIST OF FIGURES Location of Nanaimo River and Nimpkish Valley study areas Page Mosaic of successional stages at Nanaimo River created by forest harvesting and f i re (from McNay and Doyle 1987) Forested and clearcut areas of the Nimpkish Valley study area. Shaded areas are logged, unshaded areas are forested (after Harestad 1979) Summer (a) and winter (b) black-tai led deer home range composition at Nanaimo River, B.C. Home ranges were composed of clearcut or meadow, second-growth forest, and old-growth forest. Points lying towards the or ig in , off the l ine connecting 100% clearcut to 100% second growth, indicate percentages of old growth 22 Average summer o and winter m ava i lab i l i t y of cover habitats in 50%, 90%, and 100% black-tai led deer home ranges at Nanaimo River, B.C. 'SG' indicates second-growth forest; 'OG' indicates old-growth forest 24 Average summer o , spring 12a and winter ® a v a i l a b i l i t y of cover (old growth) habitats in 90% black-tai led deer home ranges at Nimpkish Valley, B.C Examples of regression of use on ava i l ab i l i t y for forage (clearcut or meadow) habitats during summer in 50% and 90% black-tai led deer home ranges at Nanaimo River, B.C 24 ,29 Changes in percent of border habitat in 164 black-tai led deer home ranges during summer o and winter ess at Nanaimo River, B.C. with: (a) changing border width in 50% home ranges, and (b) changing intensity of deer use for 90 m border width 30 V I I Figure 8 Figure 9. Average percent of 90-m border habitat in 50% black-tai led deer home ranges during summer ( o ) and winter ( BS) at Nanaimo River, B.C. 'CC/SG' indicates clearcut habitat next to second-growth forest; 'SG/CC 1 indicates second-growth forest next to clearcut habitat; 'CC/OG indicates clearcut habitat next to old-growth forest; 'OG/CC indicates old-growth forest next to clearcut habitat; 'SG/OG' indicates second-growth forest next to old-growth forest; 'OG/SG' indicates old -growth forest next to second-growth forest. An average of 68% of the home range was within 90-m of a border during summer; an average of 70% was within 90-m of a border during winter 33 Examples of regression of use on ava i lab i l i t y for 90-m clearcut borders next to second-growth habitat in 50% and 90% black-tai led deer home ranges during summer at Nanaimo River, B.C 38 Figure 10, Comparison of the dis tr ibut ion of actual o black-tai led deer locations to the dis tr ibut ion of locations expected from the habitat mosaic m around clearcut (a), second-growth (b), and old-growth (c) borders at Nanaimo River, B.C. 'Clearcut' borders indicate locations in clearcut next to second growth or old growth; 'second-growth' borders indicate locations in second growth next to clearcuts or old growth; 'old-growth' borders indicates locations in old growth next to clearcuts or second growth. Locations for 28 deer are grouped. The * indicates s ignif icant differences between actual and expected using Bonferroni's confidence intervals (.E'0.05). Chi-square c r i t i c a l values are 11.07 (a=0.05) or 9.24 ( £ = 0 . 1 0 ) . Chi -squared tests used frequencies of tota l locations 40 F . .. n Comparison of the distr ibut ion of figure 1 1 . actual • black-tai led deer locations to the distr ibut ion of locations expected from the habitat mosaic m around clearcut and old-growth borders during summer (a), spring (b), and winter (c) at Nimpkish Valley, B.C. Locations are grouped for both sides of the border for 7 deer. The * indicates s ignif icant differences between actual and expected using Bonferroni's confidence intervals (£=0.05). Chi -square c r i t i c a l values are 11.07 (£=0.05) or 9.24 (a=0.10). Chi-squared tests used frequenies of tota l locations 41 Figure 12 (a). Absolute distances moved between consecutive, 2-hour locations in clearcuts versus distance from forest borders at Nanaimo River, B.C 43 (b). Probability of successive locations being in the same or closer distance category from a border given the distance category of the previous location (intensive data collected during mild weather, during 1983 at Nanaimo River, B.C.) 43 Figure 13. An example of 3 potential b lack-ta i led deer home ranges placed within the area considered to be 'available' habitat. =^3 represents the actual home range; ii n represents potential home ranges (of equal area as the actual home range). • ' represents the boundary encompassing 'available' habitat defined by the 1.7 km non-dispersive movement distance around the actual home range 63 i X Figure 14, An example of the three different areas considered 'available' for comparing distributions of actual black-tai led deer locations around borders with those expected from the habitat mosaic. C^D represents actual home ranges; '.".*""-" represents the boundary encompassing the area 'available' for placement of random locations. For a deer with actual home range ' A ' , (b) i l lus trates habitat available within the 1.7 km distance, (c) i l lus trates habitat available within the union of a l l areas generated using the 1.7 km distance, and (d) i l lus trates habitat available within the entire study area 66 Figure 15, Forage-cover combinations during summer a and winter E 3 in 28 composite (a) and 164 yearly (b) black-tai led deer home ranges at Nanaimo River, B.C. 'SG' indicates second- growth forest; 'OG' indicates old-growth forest 68 Figure 16, Figure 17, Proportions of border habitat available during summer o and winter ss in 28 composite (a) and 164 yearly (b) 50% black - ta i l ed deer home ranges at Nanaimo River, B.C 69 Comparisons of distributions of actual CD and randomly-generated locations around clearcut (a), second-growth (b), and old -growth (c) borders during summer and winter, GE indicates random locations placed within 1.7 km radius of a home range; m indicates random locations placed within the union of a l l '1.7 km areas' around every deer's home range; aa indicates random locations placed throughout study area. Locations were grouped for a l l d e e r . . . . 73 X LIST OF APPENDICES Appendix I Calculation of a habitat interspersion index for Nanaimo River and Nimpkish Valley study areas 88 Appendix II An approach for describing deer act iv i ty centers 90 ACKNOWLEDGMENTS The project was funded by the National Science and Engineering Research Council of Canada, which provided both support for data analyses and personal expenses. Data were supplied by Drs. Harestad and Willms, and the Integrated Wildlife-Intensive Forestry Research program. A dig i t ized map of the Nanaimo River study area, created by the efforts of many people associated with the IWIFR program, was obtained from PAMAP with assistance from P. Archibald. Sincere gratitude i s extended to K.A. Scoullar for providing invaluable assistance by programming the computer analyses necessary to complete this thesis . I thank A. Harestad, R.S. McNay, R. Page, and L . Giguere for patiently providing background information on the data sets. Dr. N. Heckman provided s t a t i s t i c a l advice for most of the analyses. A. Rahme helped with data analyses. Thanks to Drs. D. Ta i t , A. Harestad, J . Rochelle, and F. Bunnell who provided advice and direction throughout the development of this thesis . Dr. Bunnell provoked many of the concepts explored herein. In addition, Fred Bunnell has my special thanks for providing learning opportunities beyond those associated with my thesis. I thank my committee members again for their guidance and patience. X i i Fina l ly , I thank the Header House crew for stimulating discussions and valued friendships. 1 GENERAL INTRODUCTION The importance of spatial interspersion of habitats for deer has become conventional "wildlife wisdom." The importance of maintaining mixes of forage and cover and border habitat has been argued largely from theoretical standpoints (Thomas (tech. ed.) 1979, Thomas et a l . 1979, Brown 1985). Research into appropriate forage-cover combinations or amount of border habitat for deer has produced equivocal results (e.g. , Reynolds 1962, 1966; Willms 1971, Hanley 1983, Kirchhoff and Schoen 1983). Kirchhoff and Schoen (1983) and Lyon et a l . (1987) suggested that the value of border habitats for deer cannot be assumed without local evaluation. This thesis evaluates the importance of interspersed habitats, as reflected by borders and forage-cover combinations, to Columbian black-tai led deer fOdocoileus hemionus columbianus) on Vancouver Island, B . C . , and discusses implications of the results to integrated management of forests and black-tai led deer. Radio-location and pellet-group data from two areas of Vancouver Island were used to examine deer use of forage, cover, and border habitat during two levels of habitat selection: within home ranges and during home range establishment. Hypotheses on habitat composition of home ranges and use and ava i lab i l i t y of forage, cover, and border habitats within home ranges were tested. Habitats used 2 within home ranges had been influenced by the i n i t i a l establishment of the home range; examination of both levels of selection was crucia l to evaluating the importance of interspersed habitats to deer. 3 USE OF FORAGE, COVER, AND BORDER HABITATS IN DEER HOME RANGES INTRODUCTION Leopold (1933:131) stated that "wildlife is a phenomenon of edges" and occurs where the types of food and cover i t needs come together. Since that time, the importance of habitat interspersion, the arrangement of habitats in space and time, has become an important concept in w i ld l i f e research and has been used to guide management (e.g. , Thomas (tech. ed.) 1979, Brown 1985, Forman 1986, Lyon et a l . 1987). Several indices have been developed as surrogates for attributes of w i ld l i f e habitat interspersion.^ Among the most widely used are forage-cover ratios and measures of border habitat. These indices are attractive because they are easi ly calculated. In North America, most published information relat ing forage-cover ratios and measures of border habitat to w i ld l i f e concerns areas east of the Cascades (review of Thomas (tech. ed.) 1979). Intensive forest management practices also influence habitat interspersion in coastal forests; they affect characterist ics and amounts of forage, cover, and border habitats. This paper presents data on use of forage, cover, and border habitats by Columbian black-ta i l ed deer (Odocoileus hemionus columbianus (Richardson)) on Vancouver Island and discusses implications to integrated management of forests and black-tai led deer. The broad 4 objective was to evaluate whether or not deer responded to interspersed habitats within home ranges, as reflected by forage-cover combinations and border habitat. Specific objectives included: 1) to test i f relat ive use of either forage, cover, or border habitats was different from relat ive ava i l ab i l i t y of those habitats within deer home ranges; 2) to test i f the dis tr ibut ion of deer locations around borders was different from that dis tr ibut ion expected from the habitat mosaic i f deer showed no a f f in i ty for border habitat. (In the absence of disparate habitat use, the l ikel ihood that any deer location is near a border i s influenced by the degree of habitat interspersion within a home range.); and 3) to document effects of seasonal changes, deer migratory behaviour, and changing intensity of deer use (as defined by proportions of locations) on use of forage, cover, and border habitats. 5 STUDY AREAS Three data sets from two areas of Vancouver Island were used. The largest data set was collected within the Integrated Wildl i fe Intensive Forestry Research program (McNay and Doyle 1987) near the southernmost tributary of the Nanaimo River, 20 km southwest of Nanaimo, B.C. (Fig. 1). The other two data sets were collected near Woss in the Nimpkish Valley, in the mountainous area of northern Vancouver Island. Differences in habitat interspersion and climate between the two study areas were important considerations in evaluating deer responses to spat ial arrangement of habitats (Table 1). Forest harvesting influenced the amount of habitat interspersion in both areas. Logging began at Nanaimo River during the 1940s and was widespread. As a result , the study area is a mosaic of re la t ive ly small patches of different successlonal stages (Fig. 2). Logging in the Nimpkish Valley began during 1915 near Beaver Cove, and advanced by railway both north and south around Nimpkish Lake. Forest harvesting near Woss began during 1948 in the val ley bottom. When data col lect ion from this area began, almost 40% of the forest below 800 m was logged (Fig. 3). 6 Figure 1. Location of Nanaimo River and Nimpkish Valley study areas. 7 Table 1. Characteristics of the Nanaimo River and Nimpkish Valley study areas. Attribute Nanaimo River Nimpkish Valley elevation dominant overstory vegetation 305-1400 m Pseudotsuga menzeisii  Tsuga heterophylla  Thuja pl icata  Abies amabilis 200-1600 m Tsuga heterophylla  Thuja pl icata  Abies amabilis  Pseudotsuga menzeisii serai a/ stages temperature 44% clearcut (0-10 yrs) 28% second-growth (10-40 yrs) 28% old-growth (>120 yrs) -19° to 36°C a/ mean precipitation 1099 mm/yr mean snowfall Nov. 5 cm Jan. 7 cm Apr. 0 cm 23% clearcut (0-27 yrs) 77% old-growth (>200 yrs) - 2 0 ° t o 37*C b / 1980 mm/yr 8 cm c / 39 cm 3 cm Forage:cover interspersion index 0.70 -3.18 d/ / The serai stage composition was recorded during 1980 for Nanaimo River (McNay and Doyle 1987) and during 1975 for Nimpkish Valley (Harestad 1979). Temperature and precipitation data were collected between 1954 and 1973 (pers. comm. Environment Canada Atmospheric services). Snow data was collected from Nanaimo River between 1982 and 1986 (McNay pers.comm.) and from Woss between 1954 and 1975 (Willms 1971 and Harestad 1979). l ' Forage:cover interspersion indices were calculated from Unwin 1981:135 (Approach in App. I . ) . An interspersion index of 0.70 implies forage and cover areas distributed approximately randomly. An interspersion index of -3.18 implies a dominance of cover areas with relat ively few interspersed forage areas. 8 Figure 2. Mosaic of successional stages at Nanaimo River created by forest harvesting and fire (from McNay and Doyle 1987). Figure 3. Forested and clearcut areas of the Nimpkish Valley study area. Shaded areas are logged, unshaded areas are forested (after Harestad 1979). 10 METHODS Deer locations Data on deer habitat use at Nanaimo River were collected between 1982 and 1986 using radio-transmitters on 28 female deer (McNay and Doyle 1987, Page in prep.) . Most locations were taken during daylight hours, but periods of intensive monitoring included night locations. Radio-collared deer were monitored weekly during winter and biweekly during summer. For non-migratory deer, seasons were defined as: 1) summer, from A p r i l 1 through October 31; and 2) winter, from November 1 through March 31. Seasons for deer that migrated were defined using each deer's migration dates. Most migratory deer in the Nimpkish Valley had spat ia l ly separate spring, summer, and winter home ranges. At Nanaimo River, however, deer used the same areas during spring as during winter. Data on deer dis tr ibut ion in the Nimpkish Valley were collected by Willms (1971) and Harestad (1979). Willms (1971) conducted pellet-group counts in 1968 and 1969 in five ecotone study sites to examine patterns of deer dis tr ibut ion between recent clearcuts and mature forests. Pel let group and vegetation sample plots were arranged systematically in para l l e l transects perpendicular to forest-clearcut borders. Harestad (1979) collected radio locations from female deer during 1975 and 1976. Most 11 locations were taken during daylight hours, but crepuscular and night locations were also included. Locations usually were made every two days. Seasons were defined using migration dates of migratory deer. Data from seven deer having the greatest number of relocations are treated here. Both telemetry and pel let group data have associated problems and biases (Collins 1981, White and Garrott 1986, Harestad and Bunnell 1987, Loft and Kie 1988). My analyses focused on telemetry data. Pel let group counts and radio locations were used to evaluate deer use patterns. Vegetation samples, associated with pel let group counts, were used to infer reasons for those observed patterns. Radio locations were used to construct home ranges for 100%, 90%, and 50% of each deer's known locations during summer and winter (Harestad 1981). A 100% home range used a l l of an animal's locations to describe i t s home range; a 90% home range excluded 10% of the most distant locations and so included 90% of the locations that were closest together; a 50% home range used 50% of the locations that were closest together to identify the area used most intensively ( i . e . , an area having a high density of locations). The percentages of tota l locations were used to define areas receiving different intensit ies of deer use (as defined by density of deer locations). A 50% home range was expected to contain greater proportions of preferred habitat 12 than a 90% home range; a 90% home range was expected to contain greater proportions of preferred habitat than a 100% home range. For Nanaimo River, 164 seasonal home ranges were constructed using weekly and biweekly data. Two-hour location data were not used to construct home ranges because of dependence of successive locations and the short time interval represented by each intensive monitoring session. The home range method (Scoullar and Kremsater in prep, see App. II) superimposes probabil ity density functions created from locations to produce contours of animal ac t iv i ty . It i s s imilar to the harmonic mean home range method (Dixon and Chapman 1980) but avoids the shifts in isopleth shape, c r i t i c i z e d by Spencer and Barrett (1984), inherent in the harmonic mean method when more than one ac t iv i ty centre exists . The probabil i ty density function method could not be applied for Nimpkish Valley data because there was no d ig i t i zed map. There, the 19 seasonal home ranges were constructed using the minimum convex polygon method (Harestad 1981). Mapping Forage, cover, and border measurements were taken from existing maps of forest cover and biogeoclimatic ecosystems. 13 These maps (1:20,000) provided detailed information for Nanaimo River. For the Nimpkish Valley, habitat assessments (Harestad 1979) supplemented information from forest cover maps. Detailed plot descriptions described the pel le t group study sites (Willms 1971). Three serai stages were defined: clearcuts (0-10 years old) , second-growth (11-40 years old), and old-growth (>120 years old) . Meadow areas and large water bodies also were differentiated. Finer def init ion of serai stages produced too many types of borders to be considered effect ively . No serai stage contained exclusively forage or cover habitat, and no serai stage was devoid of forage. Clearcuts offered the best forage in spring and summer. Second-growth forest usually offered better forage than old-growth forest during mild winters (Nyberg et a l . in prep). In old-growth forests, l ichen l i t t e r f a l l and unburied ground forage provided more available forage than was available in second-growth during deep snow accumulation (review of Bunnell and Jones 1984). For the purposes of this thesis, clearcuts were considered forage habitat and second-growth and old-growth forests were considered cover habitat. Tests for independence Independence between successive observations i s necessary for most techniques estimating home range size or comparing use and ava i l ab i l i t y of habitats. Independence 14 results when an animal's position at time (t+1) i s not a function of i t s position at time (t). Spec i f ica l ly , when evaluating how deer are distributed around borders between forage and cover habitats, independence results when a deer's distance relat ive to a border at time (t+1) i s not a function of i t s distance relat ive to a border at time (t) . Data used in this paper include 2-week, 1-week, 2-day, and 2-hour intervals between successive locations. The 2-hourly, weekly, and biweekly data were tested for independence using a Chi-square s t a t i s t i c . Insufficient 2-day locations existed to do this test . The nul l hypothesis was that the probabil i ty of being at distance (j) re lat ive to a border at time (t+1), given being at distance (i) at time (t) , did not depend on being at distance (i) at time (t) (N. Heckman pers. comm.). The Chi-square s t a t i s t i c was calculated as: with (s-1)^ degrees of freedom where s= number of distance classes erected around borders, n^ = number of times a deer i s in distance class i , 2 N 15 rij= number of times a deer is in distance class j , n^j= number of times a deer goes from distance class i to distance class j , and N= total number of locations for a deer. The chi-square s ta t i s t i c was calculated for each deer. A v a i l a b i l i t y and use of forage and cover habitats A v a i l a b i l i t y of forage and cover habitats was determined by calculating the areas of clearcut or meadow, second-growth forest, and old-growth forest inside each deer's home range. Patterns of changing ava i l ab i l i t y among seasonal home ranges and among home ranges representing different intensit ies of use were tested using Student's t -test . Because proportions were tested, an angular transformation was used to increase normality and to prevent the variance being a function of the mean. For cases where variances of the two samples were not homogeneous, the Mann-Whitney U-test was used. Because habitat use of individual deer was important, and not just the mean response, the number of deer exhibiting specif ic habitat ava i l ab i l i t y patterns was also reported. Use of forage and cover areas was determined as the proportion of a deer's locations that f e l l within each available habitat type. Use and ava i lab i l i ty of forage and cover habitats were compared using Pearson's chi-squared 16 with Bonferroni's approach to simultaneous confidence l imits (Neu et a l . 1974) for each deer and each season. Regression analyses evaluated use versus ava i lab i l i t y for a l l deer together and by migratory behaviour types. Three behaviours were identi f ied: resident deer, obligate migrators, and facultative migrators. Resident deer did not have c learly separated seasonal home ranges. Obligate migrators moved to different seasonal home ranges regardless of the amount of snow accumulation. Movements of facultative migrators followed patterns of snow accumulation and ablation (McNay and Doyle 1987). A v a i l a b i l i t y and use of border habitats The amount of border habitat in a home range depends on the mix of habitats in that home range. The amount of border habitat and the degree of habitat interspersion are interdependent; borders do not exist without mixes of habitats, mixes of habitats do not exist without borders. The amount of border habitat i s an easi ly calculated index of the degree of habitat interspersion. Deer responses to border habitat are documented as an index to deer response to habitat interspersion. Border habitats were separated into three types a p r i o r i : borders between old-growth forests and clearcuts, second-growth forests and clearcuts, and old-growth forests and second-growth forests. The designation 'clearcut' 17 included meadow habitat. Analyses examined each side of a border separately. For example, borders between clearcuts and old-growth forests consist of a side in clearcut and a side in old growth. The side of the border in clearcut was designated as "clearcut/old growth"; the side of the border in old growth was designated as "old growth/clearcut". "Clearcut/old growth" should be interpreted as: "clearcut areas next to old-growth forests"; "old growth/clearcut" should be interpreted as "old-growth forests next to clearcut areas". Three border widths were delineated 30 m, 60 m, and 90 m (on each side). Ava i lab i l i t y of border habitat was the proportion of the home range within each border width. Use of border habitat was the proportion of locations within each border width. Patterns of changing ava i lab i l i ty of border habitat were described by reporting mean ava i lab i l i t y in different seasons and in home ranges representing different intensit ies of use. Because the response of individual deer was of interest, the number of home ranges having re lat ive ly high a v a i l a b i l i t y of borders was also reported. Use and ava i lab i l i t y comparisons were made using Pearson's chi-square with Bonferroni confidence intervals for each deer and each season. Regression analyses evaluated use versus ava i lab i l i t y for each migratory behaviour type and for a l l deer grouped. 18 Deer distributions around borders Distributions of deer locations around borders were compared with those distributions expected i f deer showed no af f in i ty for border habitat. The degree of habitat interspersion and total length of border within a home range influence the l ikel ihood that any deer location i s near a border. This l ikel ihood, imposed by the habitat mosaic, was determined in Nanaimo River by locating points at random within a deer's home range. UNIRAN, University of Br i t i sh Columbia's random number generator, was used to produce random locations within each deer's home range. Twice as many random locations as actual locations were produced for each deer, resulting in over 2000 random locations in to ta l . For the Nimpkish Valley data set, the pattern imposed by the habitat mosaic was determined manually. A 0.5 by 0.5 cm (80 by 80 m actual dimensions) grid was superimposed over each home range and the distance from each vertex to the nearest border was measured. This resulted in over 500 manual locations in t o t a l . For each border type, six distance-from-border categories were delineated to maximize the degrees of freedom for the Chi-square tests while maintaining adequate sample sizes within each distance category. Smaller widths were used close to borders than were used farther away because the distr ibution of locations close to borders was 19 of part icular interest. Also, the number of locations decreased dramatically with increasing distance from the border so that increasing distance category widths were necessary to include sufficient locations in each distance class (Roscoe and Byars 1971, Fienberg 1980, and Aldredge and Ratt i 1986). The observed frequency distr ibution of radio-locations around borders was compared to the frequency distr ibut ion imposed by the habitat mosaic using Pearson's Chi-square test . The probabil ity of a type I error was set at 0.05. Confidence intervals were calculated using Bonferroni's approach to simultaneous confidence intervals (Neu et a l . 1974). Comparisons between the distr ibution of deer locations and the distr ibution imposed by the habitat were made for each deer, for each migratory behaviour, and f i n a l l y , across a l l deer, for summer and winter separately and for the entire year. Locations made every two hours were used to describe deer movements around borders. The data were treated only descriptively. The probabil ity of a deer moving towards a border given i t s distance from a border; and the absolute distances moved between locations in relat ion to distance from the border are presented. If borders were preferred habitat then one would expect that deer located farther away from borders would tend to move towards borders and that 20 deer located near borders would tend to remain near borders. Also, i f borders were preferred habitat then one would expect that movements within border habitat would tend to be less direct ional than movements farther away from borders, result ing in smaller straight l ine distances moved between consecutive locations when deer were close to borders than when they were farther away. 21 RESULTS Independence tests Results of the independence tests indicated weekly and biweekly deer observations from Nanaimo River were independent of distance from borders for most deer (26 of 28 deer had X 2 values less than the X 2 c r i t i c a l of 44.31, df=25, a.=0.01). Twelve of 28 deer from Nanaimo River had at least one intensive monitoring period (re-locations made approximately every 2 hours) that resulted in successive locations that were not independent with respect to distance from borders. Intensive data were treated only descriptively. A v a i l a b i l i t y and use of forage and cover habitats If deer preferred part icular mixes of forage and cover (as defined by clearcut and forest areas) then home ranges would be expected to contain a narrow range of forage and cover combinations. Seasonal deer home ranges, however, were composed of a wide range of combinations of the three serai stages (Fig. 4). Most home ranges had between 20-77% second growth, 11-69% clearcut or meadow, and 0-28% old growth. A s l ight shi f t towards more cover habitat (second-growth and old-growth forest) occurred during winter. A v a i l a b i l i t y of forage and cover changed as home range boundaries were altered to ref lect different intensit ies of deer use. Examining changing patterns of ava i l ab i l i t y among 22 (a) summer 100 i Percent of home range in clearcut or meadow 75 50 25 25 50 —I— 75 100 (b)winter • 100 Percent of home range in second growth forest Percent of home range in clearcut or meadow 75 50 25 25 — i — 50 75 100 Percent of home range in second growth forest Figure 4. Summer (a) and winter (b) black-tai led deer home range composition at Nanaimo River, B.C. Home ranges were composed of clearcut or meadow, second-growth forest, and old-growth forest. Points lying towards the or ig in , off the l ine connecting 100% clearcut to 100% second growth, indicate percentages of old growth. 23 100%, 90%, and 50% seasonal home ranges indicated important habitats, as areas used most intensively by deer were expected to contain greater proportions of preferred habitat than those home ranges representing areas of less intensive use. Areas of intensive deer use (50% home ranges) were composed of more cover habitat than areas of less intensive deer use, represented by 90% home ranges. The 90% home ranges, in turn, were composed of more cover than 100% home ranges (Fig. 5a). Among 164 50% home ranges, 116 contained greater proportions of cover than did their 100% home range counterparts. Examining changing patterns of a v a i l a b i l i t y among seasons indicated important seasonal habitats. For a l l intensit ies of use evaluated, the mean proportion of the home range constituted by cover habitats was greater during winter than during summer (Fig. 5a). For example, at Nanaimo River, 100% home ranges averaged 49% cover during summer and 56% cover during winter (after angular transform t=2.215, df=162, p_<0.01); 90% home ranges averaged 53% cover during summer and 60% cover during winter (Mann-Whitney U=2608.5, E=0.013); and 50% home ranges averaged 58% cover during summer and 64% during winter (after angular transform t=-1.20, df=l62, p_=0.11). Twenty-one of twenty-seven deer had greater amounts of cover available in winter than in summer 100% home ranges. Data from the Nimpkish indicated that spring home ranges had the smallest proportions of cover 24 Figure 5.(a) Average summer o and winter m ava i l ab i l i t y of cover habitats in 50%, 90%, and 100% black-tai led deer home ranges at Nanaimo River, B.C. •SG' indicates second-growth forest; 'OG' indicates old-growth forest, (b) Average summer o , spring ea , and winter m a v a i l a b i l i t y of cover (old growth) habitats in 90% black-tai led deer home ranges at Nimpkish Valley, B.C. 25 habitat (Fig. 5b). Five of 7 deer had more cover habitat in summer than in winter home ranges. A l l winter home ranges at Nimpkish Valley, however, had more cover than spring home ranges. The greater ava i lab i l i ty of cover habitat may have reflected the dominance of daytime locations in the data from which home ranges were constructed. Observations made every 2 h for periods of up to 3 days during a variety of weather events at Nanaimo River, however, indicated that although deer use of clearcuts increased at night, cover habitats continued to be used more than clearcuts, even at night (Of 1431 night time locations, 987 were in cover habitats, and 444 were in clearcuts. Of 1067 day time locations, 864 were in cover habitats and 203 were in clearcuts) . If deer preferential ly used either forage or cover areas within their home ranges, then use of some habitats may be greater than ava i l ab i l i t y . Interpretation of "use ava i labi l i ty" comparisons, however, must consider patterns of changing a v a i l a b i l i t y . Comparisons of use and a v a i l a b i l i t y using Chi-square s ta t i s t i c s and Bonferroni confidence intervals indicated that use of forage and cover habitats rarely differed from ava i lab i l i t y during both summer and winter (83 departures of 1476 potential departures) for a l l intensit ies of use (Table 2). Of the 83 26 Table 2. Results of use-availabil i ty comparisons using Chi -square tests with Bonferroni confidence intervals (d=0.05), for forage (clearcut or meadow) and cover (second-growth and old-growth forest) habitats for 164 black-tailed deer home ranges at Nanaimo River, B.C. Habitat Home range type definition %locations season n a / Bonferroni use = a v a i l b / confidence use < a v a i l 0 / intervals use > a v a i l 0 / Clearcut 50% summer 81 80 1 0 or winter 83 82 1 0 Meadow 90% summer 81 77 4 0 winter 83 81 2 0 100% summer 81 74 7 0 winter 83 73 10 0 Second- 50% summer 81 81 0 0 growth winter 83 82 0 1 forest 90% summer 81 75 2 4 winter 83 82 0 1 100% summer 81 76 2 3 winter 83 77 1 5 o ld- 50% summer 81 80 1 0 growth winter 83 83 0 0 forest 90% summer 81 75 6 0 winter 83 79 4 0 100% summer 81 69 12 0 winter 83 77 16 0 * / n=number of home ranges tested. b / number of times Bonferroni confidence intervals indicate use=availability. c / number of times Bonferroni confidence intervals indicate use less than ava i lab i l i ty . ° / number of times Bonferroni confidence intervals indicate use greater than ava i lab i l i ty . 27 departures from "use=availability", only 4 occurred in 50% home ranges. Cases for which use was greater than ava i l ab i l i t y occurred only for second-growth habitats, even though second growth was highly available. Cases for which use was less than ava i lab i l i ty occurred for a l l habitats, but predominantly for old-growth habitats, even though old growth was available in only small amounts. Slopes of the regressions of use on ava i lab i l i t y (Table 3) ranged from 0.83 and 1.12 and were not s ignif icant ly different from 1.0 in 9 of 18 cases ( £ = 0 . 0 5 ) , indicating that, in general, deer use was d irect ly proportional to a v a i l a b i l i t y . Values of r 2 were large, part icular ly for home ranges ref lect ing areas of intensive deer use (0.81<r2<0.98 for 50% home ranges, Table 3, and F i g . 6). A v a i l a b i l i t y and use of border habitat I f deer preferred mixes of broad habitats, then one would expect home ranges to contain re la t ive ly large proportions of border habitats. The a v a i l a b i l i t y of border habitat reflected the degree of interspersion within a home range and was influenced by the width assigned to borders, seasonal changes in home ranges, and changes in home ranges to ref lect different intensit ies of deer use. As the width of the area considered to be border increased, the proportion of home ranges comprised by border habitat increased, but dif ferently in different areas or seasons. For example, at Nanaimo River during summer, 40% of the area 28 Table 3. Regression equations of use on avai labi l i ty of forage (clearcut or meadow) and cover (second-growth and old -growth forest) habitats in 164 black-tailed deer home ranges at Nanaimo River, B.C. Habitat Home range type definition %locations season n a / r 2 -y'x Clearcut 50% summer 81 0.99 0.88 9.73 or winter 83 0.97 0.93 6.56 Meadow 90% summer 81 0.94 0.72 12.22 winter 83 0.92 0.82 9.17 100% summer 81 0.89 0.63 13.93 winter 83 0.87 0.67 12.01 Second- 50% summer 81 0.98 0.94 7.30 growth winter 83 0.99 0.94 6.70 forest 90% summer 81 1.04 0.83 10.64 winter 83 1.03 0.85 9.45 100% summer 81 1.06 0.69 13.45 winter 83 1.10 0.71 12.52 old- 50% summer 81 0.93 0.81 6.96 growth winter 83 0.93 0.98 2.35 forest 90% summer 81 1.12 0.59 10.40 winter 83 0.83 0.72 7.24 100% summer 81 1.16 0.43 12.00 winter 83 0.87 0.58 8.85 a / n is the number of home ranges sampled. *>/ coefficient of equation y s s b ^ x ; where y is the percent of locations fa l l ing in a habitat type, and x is the percent of the home range fa l l ing in a habitat type. 29 (a) 50% home range u 100 Percent of locations in forage habitat • « "100 (b) 90% home range Percent of home range in forage habitat 100 Percent of locations in forage habitat 100 Percent of home range in forage habitat Figure 6. Examples of regression of use on ava i lab i l i ty for forage (clearcut or meadow) habitats during summer in 50% and 90% black-tailed deer home ranges at Nanaimo River, B.C. 30 50* 90% KOZ kittnilty of us* Figure 7. Changes in percent of border habitat in 164 black Z+t l e d , d e e r home ranges during summer Q and winter ES ?! J S a K ° R l v e r v B.C. with: (a) changing border width in 50% home ranges, and (b) changing intensity of deer use for 90 m border width. 31 of 50% home ranges was within 30 m of a border, 54% was within 60 m, and 68% was within 90 m (Fig. 7a)). In the Nimpkish Valley study, an average of 17% of the 50% spring home range was within 30 m of a border, 33% within 60 m of a border, and 42% within 90 m. For both Nanaimo River and the Nimpkish Valley, areas of intensive deer use (which are expected to contain preferred habitats) had, on average, greater proportions of border habitat than areas of less intensive use (Figs. 7b and Table 4). In the Nimpkish Valley, 11 of 19 deer home ranges had more border habitat in 50% than in 90% home ranges. At Nanaimo River 50% home ranges had s igni f icant ly more border habitat than 90% home ranges during summer ( 17 of 26 deer, sign test t=1.8, df=25, E = ° « 0 4 ) , but not during winter (14 of 26 deer, sign test t=0.6, df=25, p=0.48). The importance of part icular habitat mixes may change seasonally. In the Nimpkish Valley 6 of 7 deer had more border habitat in spring or winter home ranges than in summer home ranges. At Nanaimo River, however, only 15 of 25 deer had more border habitat in winter than in summer home ranges. Most of the increase in the proportion of border habitat available to Nanaimo River deer during winter resulted largely from an increase in a v a i l a b i l i t y o f second-growth/old-growth and old-growth/second-growth borders in home ranges (Fig. 8). Most of the border habitat in Nanaimo \ 32 Table 4. Relative use and relat ive ava i lab i l i ty of 90-m border habitat during summer, spring, and winter for 50% and 90% home ranges of 7 black-tai led deer at Nimpkish Valley, B.C. 50% home range 90% home range Deer Season use avai l u / a a ' use avai l u/a OFL58 s b / 64 98 < 70 85 < SP — — — 0 0 = w 0 42 < 43 12 > OFL60 s 0 0 = 8 4 > SP 100 100 = 54 78 < w 43 97 < 60 64 < w 67 84 < 47 21 > OFL61 s 26 25 > 18 4 > SP 65 60 > 72 86 < w 0 0 34 24 > OFL62 s 33 52 < 50 30 > SP 47 49 < 82 15 > w 50 31 > 63 39 > OFL67 s —— — 0 0 SP 0 0 = 9 21 < w 53 52 > 38 13 > OFL68 s 38 55 < 41 31 > w 80 100 < 22 53 < OFL71 s 54 62 < 92 82 > SP 0 0 = 93 70 > W 40 58 < 65 85 < / use i s the percent of locations in 90-m border; avai l i s the percent of home range in 90-m border habitat u/a i s the use /avai labi l i ty rat io: < indicates use less than a v a i l a b i l i t y 1 > indicates use greater than a v a i l a b i l i t y 1 = indicates use equal to a v a i l a b i l i t y ' — indicates no seasonal home range. Use was not different from ava i lab i l i t y for any case when tested with X 2 (a =0.05). '/ seasons: S = summer, SP = spring, W = winter. 33 Figure 8. Average percent of 90-m border habitat in 50% black - t a i l e d deer home ranges during summer ( o ) and winter (@) at Nanaimo River, B .C. 'CC/SG 1 indicates clearcut habitat next to second-growth forest; •SG/CC indicates second-growth forest next to clearcut habitat; •CC/OG' indicates clearcut habitat next to old-growth forest; 'OG/CC' indicates old-growth forest next to clearcut habitat; 'SG/OG* indicates second -growth forest next to old-growth forest; 'OG/SG' indicates old-growth forest next to second-growth forest. An average of 68% of the home range was within 90-m of a border during summer; an average of 70% was within 90-m during winter. 34 River deer home ranges occurred between clearcuts and second-growth forest. Home ranges in the Nimpkish Valley consisted of borders occurring between clearcuts and old-growth forest. Use and ava i lab i l i t y comparisons using Chi-squared with Bonferroni simultaneous confidence intervals indicated that re lat ive use of border habitats did not d i f f er from relat ive a v a i l a b i l i t y in greater than 90% of comparisons for Nanaimo River (Table 5). Regression analysis indicated that use was generally d irect ly proportional to ava i l ab i l i t y for Nanaimo River (Table 6 and F ig . 9). Values of r 2 were greater than 0.75 in 13 of 14 cases and slopes did not d i f f er from 1.0 in 10 of 14 cases (o=o.05). In the Nimpkish Valley re lat ive use of borders did not d i f fer s ignif icant ly from relat ive a v a i l a b i l i t y of borders in 50% home ranges; in 90% home ranges, use of border habitats was greater than expected from a v a i l a b i l i t y for 12 of 21 cases (Table 4). Regressions of use versus ava i l ab i l i t y in the Nimpkish Valley produced variable relationships. Regressions of use on ava i l ab i l i t y of border habitat for 50% spring home ranges produced high £ 2 (0.97 < r 2 < 0.99). Regressions for a l l other home ranges generally produced poor relationships. Deer distributions around borders I f deer preferred border habitats then one would expect most deer locations to be near borders. Although deer were 35 Table 5. Results of use-availabil ity comparisons using Chi -square tests with Bonferroni confidence intervals (a=o.05), for 90-m border habitat in 50% and 90% black - ta i led deer home ranges for summer and winter at Nanaimo River, B.C. Border Home range type definition %locations season n a/ Bonferroni confidence intervals use = use < use > avail**/ a v a i l 0 / a v a i l d / CC/SG e/ CC/OG SG/CC SG/OG OG/CC OG/SG 50% summer 81 75 4 2 winter 83 79 3 1 90% summer 81 73 6 2 winter 83 76 5 2 50% summer 81 77 4 0 winter 83 81 2 0 90% summer 81 67 14 0 winter 83 63 20 0 50% summer 81 71 7 3 winter 83 82 0 1 90% summer 81 75 5 1 winter 83 80 0 3 50% summer 81 78 2 1 winter 83 80 3 ,0 90% summer 81 68 13 0 winter 83 67 15 1 50% summer 81 78 3 0 winter 83 80 3 0 90% summer 81 67 14 0 winter 83 71 12 0 50% summer 81 78 3 0 winter 83 81 2 0 90% summer 81 69 12 0 winter 83 74 8 1 a / n=number of home ranges tested. / number of times Bonferroni confidence intervals indicate use=availability. 3 6 c / number of times Bonferroni confidence intervals indicate use less than avai labi l i ty . ° / number of times Bonferroni confidence intervals indicate use grester than avai labi l i ty . e / cc=clearcut, SG=second-growth forest, OG=old-growth forest. Table 6. Regression equations for use on avai labi l i ty of 90-m border habitats for 81 summer and 83 winter 50% black - ta i led deer home ranges at Nanaimo River, B.C. Border type Season r 2 A l l summer 0.99 0.82 13. 09 borders winter 1. 00 0.93 7.38 CC/SG b / summer 1.00 0.91 10. 04 winter 0.97 0.90 8.80 CC/OG summer 0.79 0.77 6.46 winter 1.04 0.95 3.90 SG/CC summer 0.95 0.83 13.54 winter 1.00 0.90 8.61 SG/OG summer 0.95 0.94 4.49 winter 0.96 0.97 4.19 OG/CC summer 0.91 0.47 11.89 winter 0.88 0.88 3.63 OG/SG summer 1.06 0.86 5.89 winter 1.00 0.96 3.93 a / coefficient of the equation y^b^x; where y is the percent of locations fa l l ing in each border type, and x is the percent of the home range fa l l ing in each border type. b / CC/SG = clearcut habitat (next to second-growth forest), CC/OG = clearcut habitat (next to old-growth forest), SG/CC «= second-growth forest (next to clearcut habitat) , SG/OG = second-growth forest (next to old-growth forest), OG/CC = old-growth forest (next to clearcut habitat), OG/SG = old-growth forest (next to second-growth forest). 38 (a) 50% home range 100 Percent of locations in border habitat 100 (b) 90% home range 100 Percent of home range in border habitat Percent of locations in border habitat 100 Percent of home range in border habitat Figure 9. Examples of regression of use on avai labi l i ty for 90-m clearcut borders next to second-growth habitat in 50% and 90% black-tailed deer home ranges during summer at Nanaimo River. B.C. 39 located more frequently near borders than away from borders, when the distr ibut ion of deer locations around borders was compared with the distr ibut ion expected from the habitat mosaic at Nanaimo River, the distributions were not s igni f icant ly different. The observation i s consistent with results of use ava i lab i l i ty comparisons within deer home ranges. Results for individual deer, each migratory behaviour group, and a l l deer together, were consistent -the dis tr ibut ion of deer locations around borders was not s igni f icant ly different from the dis tr ibut ion expected from the habitat mosaic (Fig. 10). Bonferroni confidence intervals indicated s ignif icant differences between actual and expected for only 2 comparisons. The dis tr ibut ion of deer locations around borders in the Nimpkish Valley was different from the dis tr ibut ion expected from the habitat mosaic (Fig. 11). Deer were located closer to borders than expected. Locations for a l l deer vere grouped and tested. Deer were not tested individual ly because sample sizes were too small. I f borders were preferred habitat one would expect deer to remain close to border habitat. Deer located large distances from borders would be expected move towards borders. Deer located at borders would be expected to remain close to borders. Although deer might move at a constant rate, movements within border habitat would tend to be less 40 Figure 10. Comparison of the distr ibution of actual CD black - t a i l e d deer locations to the dis tr ibut ion of locations expected from the habitat mosaic m around clearcut (a), second growth (b), and old growth (c) borders at Nanaimo River, B.C. 'Clearcut' borders indicate locations in clearcuts next to second growth or old growth; 'second-growth' borders indicate locations in second growth next to clearcuts or old growth; 'old-growth' borders indicate locations in old growth next to clearcuts or second growth. Locations for 28 deer are grouped. The * indicates s ignif icant differences between actual and expected using Bonferroni's confidence intervals (q=0.05). Chi-square c r i t i c a l values are 11.07 (o=0.05) or 9.24 (£•=0.10) . Chi-squared tests used frequencies of to ta l locations. 4 0 a (a) clearcut borders X^=16.69 90 100 CO 250 350 >JS0 (m) (b) second-growth borders X2=9.52 90 B 0 B 0 250 350 >3S0 winter X2=2.99 Ch 50 BO S 0 250 350 >350 (cj old-growth borders 50 DO CO 250 . 350 >350 winter X 2 « 1 8 . 9 2 SO B O BO 250 350 >350 41 (a) summer (b) spring so no o  no 350 .>aw Atone* trg#f> bore* (m) (c) winter SO WO ©0 » 0 U O >iiO Figure 11 . Comparison of the distr ibut ion of actual • black - t a i l e d deer locations to the d is tr ibut ion of locations expected from the habitat mosaic m around clearcut and old-growth borders during summer (a) , spring (b), and Winter (c) at Nimpkish Valley, B.C. Locations are grouped for both sides of the border for 7 deer. The * indicates s ignif icant differences between actual and expected distributions using Bonferroni's confidence interval 's (£*=0.05). Chi-square c r i t i c a l values are 11.07 (a«o .05) or 9.24 ( £ = 0 . 1 0 ) . Chi-squared tests used frequencies of to ta l locations. 42 direct ional than movements large distances away from borders, resulting in smaller absolute distances between successive locations. Analyses of locations made every two hours during a variety of weather events at the Nimpkish Valley and Nanaimo River study areas did not indicate that deer had shorter distances between successive locations when they were near border habitat than when they were farther away from borders (Fig. 12a). The probabil ity of moving towards borders increased only s l ight ly as deer were located further from borders (Fig. 12b). These tests show no strong preference for border habitat. 43 (a) 1 5 0 0 Distance moved (m) between consecutive locations 100 300 distance from border (m) 0 J 1 1 1 I i 1—1 , L_l , l_J , l_J ,— 50 100 CO 250 350 >350 dhtane* from border (m) Figure 12 (a) Absolute distances moved between consecutive, 2 -hour locations in clearcuts versus distance from forest borders at Nanaimo River, B .C. (b) Probability of successive locations being in the same or closer distance category from a border given the distance category of the previous location (intensive data collected during mild weather, during 1983 at Nanaimo River, B . C . ) . 44 DISCUSSION Johnson ( 1 9 8 0 ) defined four orders of selection. The f i r s t order, selection of the physical or geographic range of the species, i s a result of natural selection. The remaining three orders: second order - selection of home ranges, th ird order - selection of habitats within a home range, and fourth order - selection of forage items within habitats, involve choice by the animal. This paper has examined selection of habitats by deer within their home ranges, th ird order selection. Those habitats available within the home range have been influenced by second order selection - selection of the home range. Effects of season, migratory behaviour, winter severity and differences in degree of habitat interspersion on habitat use were considered. Effects of roads, predation levels , and hunting pressure on habitat use were not evaluated. A v a i l a b i l i t y and use of forage and cover habitats Within home ranges, a wide range of combinations of forage and cover habitats were available to deer (Fig. 4); suggesting that black-tai led deer can meet their needs from a wide mix of broad habitat types. Chi-sguare, Bonferroni confidence intervals , and regression analyses a l l indicated that re lat ive use of forage and cover habitats was not 45 different from relat ive ava i lab i l i ty of those habitats (Tables 2 and 3). The increase in r 2 values for the regression of use on ava i lab i l i ty with increasing intensity of deer use indicates, however, that core areas ref lect habitats that deer prefer. Patterns emerging from the few cases (83 of 1476 potential departures, Table 2) where use differed from ava i lab i l i t y could have resulted simply from random error, but also may have indicated s l ight habitat preference. Only 4 of the 83 departures that indicated use was different than expected from ava i lab i l i t y occurred in 50% home ranges, implying that core areas are habitats that satisfy deer. Cases where use was greater than a v a i l a b i l i t y occurred only for second-growth forests, although second growth was highly available, implying some preference for second growth. Cases where use was less than a v a i l a b i l i t y occurred for a l l habitats, but primarily for old-growth forests (39 departures), although old growth was often scarcely available. Most departures from use being equal to a v a i l a b i l i t y for old growth, however, occurred in 100% home ranges which may encompass habitats seldom, i f ever, used. In the absence of a severe winter at Nanaimo River, the role of old growth cannot be evaluated. During mild winters, old growth does not appear to be more important than second growth. Relative ava i lab i l i t y of forage and cover habitats changed seasonally as home ranges changed. Seventy-eight 46 percent of deer at Nanaimo River had re lat ive ly more cover in winter than in summer home ranges. Greater use of cover by deer during winter has been a common finding (Brown 1961, Harestad 1979, review of Bunnell and Jones 1984). Kearney and Gilbert (197 6) and Staines (1976) found that deer remained in sheltered habitats rather than foraged in exposed ones during severe winters. Winter severity should influence types of habitats used as cover by deer. Data from Nanaimo River were collected during mild winters when both second-growth and old-growth forest could act as cover. Bloom (1978) also found that deer used immature and mature forests during mild winters in south-east Alaska. In Nimpkish Valley, where winters were more severe (Table 1), winter home ranges had consistently more cover than spring home ranges but 5 of 7 deer had more cover in summer home ranges than in winter home ranges. The greater amount of cover during summer resulted, in part, from a l t i tud ina l migration to areas where more forest cover was available. The use of old-growth during summer in Nimpkish Valley may, however, also suggest that summer thermal cover i s important for deer (e.g. , Parker 1988). As wel l , use of old-growth during summer suggests that i t contained adequate forage, as the review of Bunnell and Jones (1984) suggested. The dominance of clearcut areas in spring home ranges, however, suggests that open areas offer preferred forage at that time of year. Jones (1975) reported that deer ate more forbs (which are more abundant in open areas than in forests) and 47 less shrubs during spring. At Nanaimo River, none of the 28 deer had more old-growth forest available during summer than during winter (McNay and Doyle 1987) , suggesting that open areas and second growth provided preferred combinations of forage and cover. Relative ava i lab i l i t y of forage and cover changed as home ranges were altered to ref lect areas of intensive deer use (Fig 5a). Areas of intensive deer use were characterized by re la t ive ly more cover habitat during both summer and winter than areas of less intensive deer use. Locations made every two hours indicated that although the re lat ive use of clearcut areas increased at night, cover habitats were used more than clearcuts, even at night. Thus, the location of core areas of use in cover habitats was not merely a function of home ranges being constructed primarily from day-time locations. Observations that areas of intensive use were composed of mainly cover habitats supports some foraging theory (e.g. , Barnard 1980, Jenkins 1980, Thompson 1983) which incorporates the amount of energy expended and the degree of r i sk associated with searching for food. The theory predicts that as distance from cover increases, so should cumulative energetic costs of searching and r i sk of predation. If forage were available in cover habitats (as i t was in second-growth forests at Nanaimo River and i n o ld-growth forests in Nimpkish Valley) then deer would choose to 48 stay close to cover, only venturing far into clearcuts to obtain abundant, highly preferred forage. Observations that use remained d irect ly related to a v a i l a b i l i t y (Tables 2 and 3, and F ig . 6) while ava i lab i l i ty changed seasonally and with intensity of deer use, suggested that any selection of forage or cover habitats occurred during the establishment of seasonal home ranges. Cover habitats were more available in core areas and during more severe weather conditions, implying that adequate cover i s important to deer. These results and foraging theory suggest forage areas close to cover or f inely interspersed within cover should be preferred. A v a i l a b i l i t y and use of border habitats Studies of deer use of border habitats have produced variable results . Reynolds (1962), Sweeney et a l . (1984), and Kirchhoff and Schoen (1983) found l i t t l e evidence of increased deer use near borders. Reynolds (1966), Blymyer and Mosby (1977), Short et a l . (1977), Hanley (1983), found variable response to borders, with l i t t l e evidence of increased use near borders in some instances. McCaffrey and Creed (1969), Willms (1971), and Wetzel et a l . (1975) found that use increased near borders and decreased into clearcut areas. These variable responses of deer to borders are, in part, a function of differences in adjacent habitats and characterist ics of the border i t s e l f . Studies finding l i t t l e 49 response of deer to borders tended to be in areas that had a high degree of interspersion of forage and cover areas (e.g. , Reynolds 1962, Hanley 1983, Sweeney et a l . 1984), or had fine-grained interspersion, with forage and cover being available in the same habitat (e.g. , Short et a l . 1977, Kirchhoff and Schoen 1983). Studies finding an apparent preference for border habitat tended to be in areas that had l i t t l e interspersion of forage and cover areas (e.g. , Willms 1971, Wetzel et a l 1975), or had c learly separated forage and cover habitats (e.g. , Blymyer and Mosby 1977). At Nanaimo River, re lat ive use of borders did not d i f f er from relat ive ava i lab i l i t y (Tables 5 and 6, and F ig . 9). Patterns emerged, however, from the few cases where use dif fered. Cases where use was greater than a v a i l a b i l i t y occurred only for borders containing second-growth forest. Cases where use was less than ava i l ab i l i t y occurred for a l l border types, but primarily for borders between old growth and clearcuts during summer and winter, and between old growth and second growth during summer. The trends imply that old growth at Nanaimo River was not preferred summer habitat and mixes of second growth and old growth offered better winter habitat than mixes of clearcuts and old growth during low snowfall winters. In the Nimpkish Valley, re lat ive use of borders was usually greater than their relat ive a v a i l a b i l i t y in 90% home 50 ranges (Table 4). This difference between the two study areas may be explained by the characteristics of the border habitat. The Nimpkish Valley had less interspersion of habitats than Nanaimo River (Table 1), which increased the probabil i ty of observing use of borders being greater than a v a i l a b i l i t y . During winter, snow accumulation in the Nimpkish Valley also could have increased the importance of border habitat, part icular ly forested sides of borders that offered increased forage production due to increased l ight penetration from adjacent openings. Relative ava i l ab i l i t y of border habitat changed seasonally as home ranges changed. Home ranges contained, on average, more border habitat during winter than during summer. That was not consistent across a l l animals (Fig.7a and b, and Table 4). Borders between second-growth and o ld-growth forests seemed to be part icu lar ly important during winter at Nanaimo River (Fig 8). These borders provided easy access to second growth, which offered good forage and cover during the mild winters of this study, and to old growth which would have offered superior forage and cover i f winter conditions became more severe. Open habitats may be avoided during winter because of bur ia l of forage and high costs of locomotion. S imi lar ly , borders between openings and forests may accumulate snow (review of Bunnell et a l . 1985) and so be less used by deer than borders between more forested habitats. 51 Relative ava i lab i l i ty of border habitat changed as home ranges were altered to ref lect areas of intensive deer use (Fig 7b). In the Nimpkish Valley, areas receiving intensive use had more border habitat than areas receiving less intense use. Similarly , at Nanaimo River during summer, areas receiving intensive use had more border habitat than areas receiving less intensive use; during winter, however, the difference was not s ignif icant . Because re lat ive use of borders remained d irec t ly proportional to relat ive ava i lab i l i t y , despite changing a v a i l a b i l i t y seasonally and in areas receiving different intensit ies of deer use, any selection for habitat interspersion probably occurred during the establishment of home ranges. Border habitat i s generally considered important for one of two reasons: 1) the greater richness of border vegetation, and 2) simultaneous access to more than one habitat type. Habitat assessments indicated that border habitat was, at least in the Nimpkish, r icher in species composition (a diversity) than either of the adjacent habitats (data of Willms 1971). Certainly, borders ref lect a mix of habitats, and often offered deer simultaneous access to quite different habitat requirements, usually forage and cover. Analyses done in this paper could not separate the 52 relat ive importance of those biological reasons for the significance of border habitat. Trends from this study and the l i terature suggest borders are important to deer because they indicate a mix of habitats and offer simultaneous access to forage and cover. The importance of borders or habitat interspersion decreases when forage and cover are available within the same habitat (habitats having f ine-grained interspersion). In summary, forage, cover, and border habitats were used largely as expected from their a v a i l a b i l i t y within deer home ranges. Trends from the few cases where use differed from a v a i l a b i l i t y , and changing ava i lab i l i t y seasonally and in areas of intensive use, suggested importance of cover areas and interspersion of forage and cover. Interspersed habitats appeared more important in areas where forage and cover habitats were re lat ive ly d i s t inct and less important where forage was interspersed within cover habitats. Because forage, cover, and border habitats were used as expected from their ava i l ab i l i t y within deer home ranges, changing a v a i l a b i l i t y indicated that any selection for these habitats occurred during i n i t i a l establishment of seasonal home ranges. Therefore, before beneficial spat ial arrangement of habitats can be prescribed for an area, one must examine not only how deer use habitats within their home ranges, but also how habitat interspersion affects where deer establish the ir home ranges. 53 LITERATURE CITED Aldredge, J . R . , and J . T . Rat t i . 1986. Comparison of some s t a t i s t i c a l techniques for analysis of resource selection. J . Wi ld l . Manage. 50: 157-165. Barnard, C . J . 1980. Flock feeding and time budgets in the house sparrow (Passer domesticus L.) Anim. Behav. 28: 295-309. Bloom, A.M. 1978. Sitka black-tai led deer winter range in the Kadashan Bay area, southeast Alaska. J . Wi ld l . Manage. 42:108-112. Blymyer, M . J . and H.S. Mosby. 1977. Deer u t i l i z a t i o n of clearcuts in southwestern V i r g i n i a . Southern J . Appl. For. 1: 10-13. Brown E.R. 1961. The black-tai led deer of western Washington. Wash. State Game Dept. Bio l B u l l . No. 13. 124 pp. Brown, E.R. 1985. Management of Wildl i fe and Fish Habitats in Forests of Western Oregon and Washington. U. S. Dep. A g r i c , For. Serv. Pac. Northwest Region. 332 pp. Bunnell, F . L . and G.W. Jones. 1984. Black-tai led deer and old-growth forests: a synthesis. Pp. 411-420 In W.R. Meehan, T.R. Merre l l , J r . , and T.A. Hanley (technical editors) . Proc. Symp. on Fish and Wildl i fe Relationships in Old-growth Forests. Bookmasters, Ashland, Ohio. Bunnell, F . L . , R.S. McNay, and C.C. Shank. 1985. Snow and trees: deposition of snow on the ground - a review and quantitative synthesis. Research, Ministries of Environment and Forests, IWIFR 17, V i c t o r i a , B.C. 440 pp. Co l l ins , W.B. 1981. Habitat preference of mule deer as rated by pellet-group distr ibutions. J . Wi ld l . Manage. 45: 969-972. Dixon, K . R . , and J . A . Chapman. 1980. Harmonic mean measure of animal ac t iv i ty areas. Ecology 61: 1040-1044. Fienberg, S .E. 1980. The Analysis of Cross-class i f ied Categorical Data. M.I .T . Press, Cambridge, Mass. 198 pp. Forman, R . T . T . 1986. Emerging directions in landscape ecology and applications in natural resources management. Conference on Science in the National Parks. George Wright Society. Pp 59-87. 54 Hanley, T .A. 1983. Black-tai led deer, elk, and forest edge in a western cascades watershed. J . Wi ld l . Manage. 47:237-242. Harestad, A.S . 1979. Seasonal movements of black-tai led deer on Northern Vancouver Island. Ph.D Thesis, Univ. Br i t i sh Columbia, Vancouver, B.C. 184 pp. Harestad, A.S . 1981. Computer analysis of home range data. B.C. Ministry of Environment. Fish and Wildl i fe Bul let in No. B - l l . 25 pp. Harestad, A . S . , and F . L . Bunnell. 1987. Persistence of black-tai led deer fecal pel lets in coastal habitats. J . Wi ld l . Manage. 51: 33-37. Jenkins, S.H. 1980. A size-distance relat ion in food selection by beavers. Ecology 61: 740-746. Johnson, H.D. 1980. The comparison of usage and a v a i l a b i l i t y measurements for evaluating resource preference. Ecology 61:65-71. Jones, G.W. 1975. Aspects of the winter ecology of black-ta i l ed deer (Odocoileus hemoinus columbianus Richardson) on northern Vancouver Island. M.Sc. Thesis, Univ. Br i t i sh Columbia, Vancouver, B.C. 78pp. Kearney, S.R. , and F . F . Gi lbert . 1976. Habitat use by white-ta i l ed deer and moose on sympatric range. J . W i l d l . Manage. 40: 645-657. Kirchhoff, M.D. and J.W. Schoen. 1983. Black-tai led deer use in re lat ion to forest clearcut edges in southeastern Alaska. J . W i l d l . Manage. 47: 497-501. Leopold, A. 1933. Game Management. Charles Scribner's Sons, New York, N.Y. 481 pp. Loft , E . R . , and J . G . Kie. 1988. Comparison of pellet-group and radio-triangulation methods for assessing deer habitat use. J . Wi ld l . Manage. 52: 524-527. Lyon, J . G . , A.M. Asce, J . T . Heinen, R.A. Mead, and N.E .G. Rol ler . 1987. Spatial data for modelling wi ld l i f e habitats. J . of Surveying Engineering 133(2): 88-100. McCaffrey, K . R . , and W.A. Creed. 1969. Significance of forest openings to deer in northern Wisconsin. Wise. Dept. Nat. Res., Tec. B u l l . 44. 104 pp. 55 McNay, R . S . , and D.D. Doyle, 1987. Winter habitat selection by black-tai led deer on Vancouver Island: a job completion report. Research, B.C. Ministry of Environment and Parks and B.C. Ministry of Forests and Lands. IWIFR-34. V ic tor ia , B.C. 89 pp. Neu, CW. , C R . Byers, and J . M . Peek. 1974. A technique for analysis of u t i l i z a t i o n - a v a i l a b i l i t y data. J . Wi ld l . Manage. 38: 541-545. Nyberg, J . B . , L . Peterson, L . A . Stordeur, and R.S. McNay. 1988. Deer use of old-growth and immature forsets following snowfalls on southern Vancouver Island. IWIFR job completion report, E .P . 934.02. in prep. 39 pp. Page," R. Computer analysis of radio-telemetry data in the f i e l d , in prep. Parker, K . L . 1988. Effects of heat, cold, and rain on coastal black-tai led deer. Can. J . Zool. 66: 2475-2483. Reynolds, H.G. 1962. Use of natural openings in a ponderosa pine forest by deer, elk, and catt le . U.S. Dep. A g r i c , For. Serv. Res. Note. RM-78. 4 pp. Reynolds, H.G. 1966. Use of openings in a spruce-f ir forests of Arizona by deer, elk, and catt le . U.S. Dep. A g r i c , For. Serv. Res. Note. RM-66. 4 pp. Roscoe, J . T . and J . A . Byars. 1971. An inspection of the restraints with respect to the sample size commonly imposed on the use of the Chi-square s t a t i s t i c . J . Am. Stat. Assoc. 66: 755-759. Scoullar, K.A. and L . L . Kremsater. An approach for describing animal ac t iv i ty centres, in prep. Short, H . L . , W.E. Evans, and E . L . Boeker. 1977. The use of natural and modified pinyon pine-juniper woodlands by deer and elk. J . Wi ld l . Manage. 41: 543-559. Spencer W.D., and R.H. Barett. 1984. An evaluation of the harmonic mean measure for defining carnivore ac t iv i ty areas. Acta Zool. Fennica 171: 255-259. Staines/ B.W. 1976. The use of natural shelter by red deer in re lat ion to weather in north-east Scotland. J . Zoo l . , Lond. 180: 1-8. Sweeney, J . M . , M.E. Garner, and R.P. Burkert. 1984. Analysis of white-tailed deer use of forest clearcuts. J . Wi ld l . Manage. 48: 652-655. 56 Thomas, J.W. tech. ed. 1979. Wildl i fe Habitats in Managed Forests, the Blue Mountains of Oregon and Washington. U. S. Dep. A g r i c , For. Serv. Handbook 553. 512 pp. Thomas, J .W. , C. Maser, and J . E . Rodick. 1979. Wi ld l i fe habitats in managed rangelands the great basin in southeast Oregon: edges. U. S. Dep. A g r i c , For. Serv. P.N.W.-85, 17pp. Thompson, D.B.A. 1983. Prey assessment by plovers (Charadriidae): net rate of energy intake and vulnerabi l i ty to kleptoparasites. Anim. Behav. 31: 1226-1236. Unwin, D. 1981. Introductory Spatial Analysis. Methuen and Co. Ltd. New York. 212 pp. Wetzel, J . F . , J .R . Wambaugh, and J .M. Peek. 1975. Appraisal of white-tailed deer winter habitats in northern Minnesota. J . W i l d l . Manage. 39: 59-66. White, G . C , and R.A. Garrott. 1986. Effects of biotelemetry triangulation error on detecting habitat selection. J . Wi ld l . Manage. 50: 509-513. Willms. W.D. 1971. The influence of forest edge, elevation, aspect, s i te index, and roads on deer use of logged and mature forest, northern Vancouver Island. M.Sc. Thesis. Univ. B r i t i s h Columbia, Vancouver, B.C. 184 pp. 57 INFLUENCES OF FORAGE, COVER, AND BORDER HABITATS ON HOME RANGE ESTABLISHMENT INTRODUCTION Research on deer use of interspersed habitats, as indexed by use of forage, cover, and border habitat, has produced equivocal results . Part of the v a r i a b i l i t y in results i s attributable to inconsistent def init ion of habitat deemed available. Results of use-avai labi l i ty analyses, which are commonly used to assess importance of part icular habitats, are c r i t i c a l l y dependent on what habitats are considered available to each animal (Johnson 1980). Many studies l imi t available habitat to areas within an animal's home range, areas that an animal can sample and access (e.g. , Hanley 1983, Loft et a l . 1984, Servheen 1983). Selection of habitats within home ranges, however, i s affected by selection of the home range i t s e l f . Few researchers examine selection of home ranges. Some studies, however, consider areas broader than home ranges as available to animals without e x p l i c i t l y recognizing implications to use-avai labi l i ty analyses (e .g. , Witmer and de Calesta 1983, McCorquodale et a l . 1986, Ordway and Krausman 1986). This paper's broad objective i s to examine how forage, cover, and border habitats influence the location of black-ta i l ed deer home ranges. Specific objectives include: 58 1) to compare ava i lab i l i ty of forage, cover, and border habitats between actual home ranges and equivalent size home ranges placed in areas that were potential ly available to each animal; 2) to compare distributions of deer locations around borders with distributions around borders in areas where home ranges potential ly could have been established; and 3) to examine effects of changing the area considered available for potential home ranges. I discuss implications of my results to management of forests and black-tai led deer and suggest implications to interpretation of use-avai labi l i ty studies. 59 STUDY AREA Data were collected within the Integrated Wi ld l i fe Intensive Forestry Research Program (McNay and Doyle 1987) near the southernmost tributary of Nanaimo River, 20 km southwest of Nanaimo, B.C. The study area encompasses 225 km2 of mountainous terrain , ranging in elevation from 305 m to 1400 m. Most of the study area l i e s within the Coastal Western Hemlock Biogeoclimatic Zone (Krajina 1965). Of that area, 25% is dominated by Tsuqa heterophylla (Raf.) Sarg. and Abies amabalis (Dougl.) Forbes and i s c lass i f i ed as CWHb4, the remainder is dominated by Pseudotsuga menzeisii (Mirb.) Franco and Thuja p l icata Donn and i s c lass i f i ed as CWHaj^  (Krajina 1965). A small portion of the study area l i e s in the Mountain Hemlock Zone. During 1982 to 1986, temperatures at Nanaimo River ranged from -19* to 36* C; precipi tat ion averaged 1271 mm/year; and snow accumulations averaged 5.0 cm in November, 10.0 cm in December, and 7.7 cm in January. Logging began in the study area during the 1940s and was widespread. As a result , the study area i s a mosaic of successional stages. Clearcuts (<10 years old) cover 44% of the area, second-growth forest (11 to 40 years old) and o ld-growth forests (>120 years old) compose equal portions of the remaining area (Fig. 1) . 60 METHODS Deer locations Data on deer distr ibution were collected between 1982 and 1986 using radio-transmitters on female deer (McNay and Doyle 1987). Most locations were made during daylight hours. Radio-collared deer were monitored weekly during winter and biweekly during summer. For non-migratory deer, seasons were defined as: 1) summer, from A p r i l 1 through October 31; and 2) winter, from November 1 through March 31. Seasons for deer that migrated to seasonal home ranges were defined using each deer's migration dates. Winter and spring home ranges were not spat ia l ly d i s t inct . Deer locations were determined by triangulation using 3 to 5 bearings from known locations (McNay and Doyle 1987, Page in prep.) . Locations were independent with respect to distance from forage and cover borders (p. 21). Mapping To evaluate influences of forage, cover, and border habitat on location of home ranges, 3 broad habitats were defined by serai stage: clearcuts (0-10 years o ld) , second growth (11-40 years old) , and old-growth (>120 years old) . Meadow and water habitats were also differentiated. Finer def ini t ion of serai stages produced too many types of border habitat to be considered effect ively. 61 Although no habitat contained exclusively forage or cover, for the purposes of this thesis clearcut and meadow habitats were considered forage; second-growth and o ld-growth forests were considered cover. Actual home ranges Home ranges were estimated for summer and winter for 28 deer using radio locations collected between 1982 and 1986. Home ranges were constructed using a method developed by Scoullar and Kremsater (in prep, see App. II) that superimposes deer locations on the study area map to produce a probabil i ty density function. Actual home ranges were created from 100%, 90%, and 50% of a deer's known locations (Harestad 1981). A 100% home range uses a l l of an animal's locations to describe i t s home range; a 90% home range excludes 10% of the most distant locations and so uses the 90% of locations that are closest together; a 50% home range uses 50% of the locations that are closest together to identify an area having a re lat ive ly high density (number of locations per area) of deer locations. Potential home ranges Deer are assumed to establish home ranges within an area that they have sampled. Determining the area sampled is d i f f i c u l t and subjective. Brown (1961) and Bunnell and Harestad (1983) reported average non-dispersive movements by female deer of 1.7 km from their b ir th place. I considered 62 habitats within 1.7 km of a deer's composite home range to have been sampled by the animal and 'available' at the time i t established i t s home range. Twenty c ircular home ranges of size equal to the deer's actual home range were placed randomly within the area considered to be available habitat for each deer (e.g. , F ig . 13). These c ircular home ranges represent potential home ranges that deer could have established given the area deemed available. Twenty potential home ranges were created for summer and 2 0 for winter home ranges for each of 28 deer. Habitat composition in actual and potential home ranges A v a i l a b i l i t y of forage, cover, and border habitats in actual and potential home ranges was determined from the area of each home range in clearcut or meadow, second-growth forest, and old-growth forest. Clearcut or meadow areas were considered forage habitat and are termed 'clearcut' hereafter; second-growth and old-growth forests were considered cover habitat. Three border habitats were defined: between old-growth forests and clearcuts, between second-growth forests and clearcuts, and between old-growth forests and second-growth forests. Analyses examined each side of the border separately. For example, borders between clearcuts and old-growth forests consist of a side in clearcut and a side in old growth. The side in clearcut is 6 3 Figure 13. An example of 3 potential black-tai led deer home ranges placed within the area considered to be •available' h a b i t a t . § = § represents the actual home range; LTD represents potential home ranges (of equal area as the actual home range). ' ' represents the boundary encompassing 'available' habitat defined by the 1.7 km non-dispersive movement distance around the actual home range. 64 designated as "clearcut/old growth"; the side in old growth is designated as "old growth/clearcut". "Clearcut/old growth" should be interpreted as "clearcut habitat next to old growth"; "old growth/clearcut" should be interpreted as "old growth habitat next to clearcut". Three border widths were delineated: 30 m, 60 m, and 90 m (on each side) . A v a i l a b i l i t y of border habitat was the area of the home range within each border width. A v a i l a b i l i t y of forage, cover, and border habitat in actual home ranges was compared with ava i l ab i l i t y of those habitats in potential home ranges. The actual home range provided a single observation of the proportion of forage, cover, or border habitat available to a deer. This single observation was compared with the mean of the sample of 2 0 potential home ranges to determine i f i t was l i k e l y that the actual home range and potential home ranges came from the same population (Sokal and Rohlf 1981:231). Deer were then grouped by migratory behaviour type and the d is tr ibut ion of forage, cover, and border habitats available in actual home ranges compared with the distr ibut ion of those habitats available in potential home ranges for each migratory behaviour. Results from use and ava i l ab i l i t y comparisons within home ranges (pp. 21-31) implied that 50% home ranges reflected habitats preferred by deer. A v a i l a b i l i t y of forage, cover, and border habitat, therefore, was expected 65 to d i f fer more among 50% actual and potential home ranges than among 90% or 100% actual and potential home ranges. Deer distributions around borders Within home ranges at Nanaimo River the distr ibut ion of locations around borders did not d i f fer s igni f icant ly from the distr ibut ion expected from the habitat mosaic in the absence of any preference for border habitat (Fig. 10). Here, deer locations around borders are compared with distributions of locations around borders expected from the habitat mosaic in areas of potential home ranges. To estimate the distr ibut ion expected from the habitat mosaic in the absence of preference for border habitat, locations were placed at random within three areas considered available for potential home range establishment: within a 1.7 km radius of each deer's home range, within the union of a l l areas generated from the 1.7 km radius around every deer's home range, and over the entire study area (e.g. , F ig . 14). Actual distributions of locations were tested against each of the three different potential distr ibutions. Pearson's Chi-sguare test with Bonferroni confidence intervals were computed for summer and winter for each deer, for each migratory behaviour, and for a l l deer grouped. 66 (a) Figure 14. An example of the three d i f f e r e n t areas considered 'available' for comparing d i s t r i b u t i o n s of actual b l a c k - t a i l e d deer locations around borders with those expected from the habitat mosaic. <~"T> represents actual home ranges; r''/S' represents the boundary encompassing the area 'available' for placement of random lo c a t i o n s . For a deer with actual home range 'A'; (b) i l l u s t r a t e s habitat available within the 1.7 km distance, (c) i l l u s t r a t e s habitat available within the union of a l l areas generated using the 1.7 km distance, and (d) i l l u s t r a t e s habitat available within the entir e study area. 67 RESULTS Actual home ranges Home ranges were estimated from locations collected over the 5 years of the study. These composite home ranges contained less extreme combinations of forage and cover than did home ranges created from each year's data. Composite home ranges usually included greater proportions of o ld-growth habitat than did yearly home ranges because including up to 5 years of locations often expanded the home range to include more old-growth forest. The shif t to cover habitats during winter i s evident in both yearly and composite home ranges (Fig. 15). The shift to cover in areas of intensive deer use (defined by proportion of locations) i s also evident in both composite and yearly home ranges (Fig. 15). Composite home ranges contained similar proportions of border habitat as their yearly counterparts (Fig. 16). Because composite and yearly home ranges were s imilar in composition, and because there were too many yearly home ranges to consider, comparisons were made between composite and potential home ranges. Habitat composition in actual and potential home ranges. Forage and cover habitats A v a i l a b i l i t y of forage and cover habitats in actual and potential home ranges was compared for each deer and for 68 Figure 15. Forage-cover combinations during summer • and winter BE in 28 composite (a) and 164 yearly (b) black-tai led deer home ranges at Nanaimo River, B.C. *SG' indicates second-growth forest; 'OG' indicates old -growth forest. 69 (a) composite 30 m 60 m 90 m border width (b) yearly eo-> 30 m 60 m 90 m border width Figure 16. Proportions of border habitat available during summer • and winter ® in 28 composite (a) and 164 yearly (b) 50% black-tai led deer home ranges at Nanaimo River, B.C. 70 each migratory behaviour group. If deer preferred particular mixes of forage or cover, then some differences would be expected among forage and cover composition in actual and potential home ranges. Differences from potential home ranges, however, were few. Ava i lab i l i t y was greater than expected (p <0.10) from potential home ranges in only 8 of 504 (28 deer * 3 home range intensit ies * 2 seasons * 3 habitats) comparisons; i t was less than expected in only 2 of 504 comparisons. Clearcuts were the only areas less available than expected (2 comparisons). Three of the 4 cases when ava i l ab i l i t y of second growth was greater than expected occurred during summer. A l l 3 cases when a v a i l a b i l i t y of old growth was greater than expected occurred during winter. Actual home ranges encompassing 100% of a l l deer locations were more similar to potential home ranges in forage and cover composition than were actual home ranges representing areas of more intensive deer use (made up of 90% and 50% of a l l known locations). Only 2 of the 10 departures from expected occurred in tests using 100% home ranges. These tests show no selection for forage and cover combinations by deer. Border habitats A v a i l a b i l i t y of habitat within 3 border widths was compared in seasonal actual and potential home ranges estimated for the three intensit ies of use for individual deer. If deer preferred part icular habitat mixes, then some 71 differences between border composition among actual and potential home ranges would be expected. A v a i l a b i l i t y of 90-m border habitat was greater than expected from potential home ranges for only 52 of 117 6 (28 deer * 2 seasons * 7 border types * 3 home range intensities) comparisons by 16 deer. (The seven border types were "clearcut/second growth, c learcut/old growth, second growth/clearcut, second growth/old growth, old growth/clearcut, old growth/second growth, and tota l border"). A v a i l a b i l i t y was less than expected for 4 comparisons, a l l occurred during summer. Actual home ranges encompassing 100% of a l l deer locations again showed fewer departures from potential home ranges than did actual home ranges representing areas of more intensive use. The 11 facultative migrators showed fewer departures from expected (15 of 462 comparisons by 7 deer) than did either of the groups showing more fixed migratory behaviour - obligate migrators and residents. Of the 14 resident deer, 8 showed departures (34 of 588 comparisons); 1 of the 3 obligate migrators showed departures (8 of 126 comparisons). Departures among facultative migrators generally (11 of 462 comparisons) exhibited greater use of border habitat than expected with no apparent discrimination of any specif ic border type. Departures among resident deer were consistently for greater use of border habitat; 23/34 for border between clearcut and forests, 8 for borders between 72 second growth and old growth. Departures for the obligate migrator were for greater use of borders between second growth and old growth. Results for 60-m border habitat were similar to those for 90-m border habitat. Deer distributions around borders When deer were tested individual ly , the dis tr ibut ion of deer locations around borders seldom differed from that d is tr ibut ion expected from locations placed at random within a 1.7 km radius of the home range. Although most deer showed patterns of increasing proportion of locations close to borders, rarely was that increase s igni f icant ly greater than that expected from the habitat mosaic. When deer were grouped by migratory behaviour, trends of increasing proportions of locations near borders remained consistent and s ignif icant differences from expected increased, largely due to increased sample sizes. When a l l deer were grouped differences continued to become more s ignif icant suggesting a common underlying trend (Fig. 17). As the area considered to be potential habitat increased from the area swept by the 1.7 km radius around a specif ic home range, to the union of those areas, to the entire study area, s ignif icant differences among the distributions of actual and randomly-located locations 73 Figure 17. Comparisons of distributions of actual • and randomly-generated locations around clearcut (a), second-growth (b), and old-growth (c) borders during summer and winter, m indicates random locations placed within 1.7 km radius of a home range, ra indicates random locations placed within the union of a l l '1.7 km areas' around every deer's home range, oo indicates random locations placed throughout study area. Locations were grouped for a l l deer. 73a (a) clearcut borders OWava kOT> bordv (m) (Wgnea tram borttar (m) (b) second-growth borders >550 (c) old-growth borders increased (Fig. 17 and Table 7). Of the 43 departures of actual locations from expected, 11 occurred for comparisons with random locations throughout the '1.7 km area'; 13 occurred with random locations throughout the union of a l l '1.7 km areas'; and 19 occurred with random locations throughout the study area. Table 7 summarizes general trends. Actual locations were distributed closer to borders than expected. In the f i r s t distance category near borders, 7 of 10 departures indicated a greater proportion of actual locations than expected (Table 7). This proportion declined with increasing distance; at distances >250 m from the border s ignif icant departures indicated less use than expected. 7 5 Table 7. Results of Bonferroni confidence intervals (a=o.05) for comparisons of actual and randomly -distributed locations of black-tailed deer around borders. Locations were grouped for a l l deer. Border Season13/ Area for Total distance category from border0/ typea / random N 1 2 3 4 5 6 locations CC S 1.7 km 861 -d/ union 2067 Btudy 1805 + — cc W 1.7 km 1173 + _ union 2067 + - -study 1805 + — - -SG S 1.7 km 765 _ + union 1758 + study 1078 - + SG W 1.7 tan 939 _ + + union 1758 + -study 1078 — + -OG S 1.7 km 292 union 460 + -study 918 + -OG W 1.7 km 338 union 460 + - - -study 918 + - - - -number of departures greater than expected 2 £ 2 2 fi fi t o t a l significant departures 10 7 6 11 8 1 a/ Border types: CC- indicates locations i n clearcut next to second-growth or old-growth forest; SG» indicates locations i n second-growth forests next to clearcut or old-growth forests: OG« indicates locations i n old-growth forests next to clearcut or second-growth forests. Seasons:. S«*summer; W-winter. C/Distance categories around border: l«0-50m, 2-51-100m, 3«=101-150m, 4«151-250m, 5«=251-350m, 6->350m. 76 A - ( - ) indicates that Bonferroni's confidence intervals documented signi f i c a n t l y more (fewer) actual locations in that distance category than expected from randomly distributed locations. 77 DISCUSSION I explored influences of habitat composition on areas chosen by deer for home ranges. Speci f ica l ly , I tested for: 1 ) Differences in amounts of forage, cover, and border habitats between actual deer home ranges and areas where deer could have potential ly located their home ranges; and 2) Differences in distributions of deer locations around borders between actual locations and locations placed randomly within areas potential ly available during home range establishment. Available forage, cover, and border habitat Few differences in forage and cover composition existed between actual and potential home ranges when deer were tested indiv idual ly . These few differences, however, suggest trends. Of the 8 cases when ava i l ab i l i t y was greater than expected, 7 occurred for cover habitats (second growth or old growth). Of the 4 cases when second growth was more available than expected, 3 occurred during summer. A l l 3 cases when old growth was more available than expected occurred during winter. These observations suggest that, although selection for broad forage and cover habitats was weak, cover may be preferred habitat, with old growth preferred in winter. Home ranges ref lect ing areas of intensive deer use were expected to contain habitats preferred by deer. Among the 3 intensit ies of use tested, 6 78 of 10 departures from expected occurred in 50% home ranges; 5 indicated that cover was more available than expected. Tests of individual deer also revealed few differences in re lat ive amounts of border habitat within actual and potential home ranges, but again, trends were suggested. Of the 56 departures from expected, 52 (93%) indicated greater a v a i l a b i l i t y of borders than expected, suggesting preference for border habitat or habitat mixes. A v a i l a b i l i t y of border habitat was never less than expected in winter. Areas of intensive use, represented by 50% home ranges, accounted for 28 of the 52 cases when ava i lab i l i t y was greater than expected, again suggesting that borders were preferred. Only 8 of the 52 cases when ava i lab i l i ty was greater than expected occurred in 100% home ranges. Again, the more intensively used portion of the home range revealed the greatest departures from habitat potential ly available. Deer distributions around borders The dis tr ibut ion of actual deer locations around borders was compared to the distr ibut ion of locations placed within each of 3 areas representing different amounts of available habitat. The distr ibut ion of actual locations indicated that most locations were near borders. When deer were tested individual ly , however, the dis tr ibut ion of actual locations was not often different from that expected from random, often due to small numbers of locations. Because trends of increased numbers of locations near borders were general across deer, when deer were grouped, differences became significant (Table 7). Distance categories near borders had more locations than expected; . distance categories farther away from the border had fewer locations than expected, indicating that borders were preferred. Differences between actual and expected distributions increased as the area deemed available increased. To the extent that decisions of ava i l ab i l i t y are arbitrary, so w i l l be the conclusions drawn (Johnson 1980). To determine what habitats are available to a deer one would need to know where that deer had travel led during i t s l i f e , how well i t remembered where i t had been, and how behavioural and social characterist ics influenced which habitats were considered available. It i s unlikely that any deer had travel led the entire area defined by the 1.7 km radius. It i s even more unlikely that habitats over the entire study area were available to each animal, yet some studies have made that assumption (e.g. , Witmer and de Calesta 1983, McCorquodale et a l . 1986, and Ordway and Krausman 1986). In this study, increasing the area considered available had l i t t l e impact on the expected dis tr ibut ion of locations around borders because of the high degree of interspersion in the study area. 80 Most studies of deer use of border habitat have not compared patterns of deer locations with patterns of locations imposed by the habitat mosaic in the absence of disparate habitat use (Reynolds 1962,1966; Willms 1971, Blymyer and Mosby 1977, Sweeney et a l . 1984). Results of such studies must be interpreted with caution. For example, in this study, when testing the distr ibut ion of actual deer locations around borders at Nanaimo River, one would have suspected a s ignif icant 'preference 1 for border habitat. When comparing that actual distr ibut ion to the dis tr ibut ion of randomly located locations, however, i t i s clear that much of the apparent 'preference' for border habitat i s imposed by the habitat mosaic. Conventional wi ld l i f e wisdom holds that border habitat confers benefits to w i ld l i f e , including deer (Leopold 1933, Thomas (tech. ed.) 1979, Brown 1985). Tradit ional management prescriptions also hold that mixes of open and forested areas provide optimum habitat for deer (review of Brown 1985). Recent research has provided results that shed doubt on the be l ie f that border habitat, or edge, confers large benefits to deer (Kirchhoff and Schoen 1983, Hanley 1983, Lyon and Jensen 1980). Results from my study indicate that few individual deer establish home ranges containing different amounts of forage, or border habitat than expected, although some preference for cover and habitat interspersion is suggested. The apparent importance of habitat interspersion, as indexed by forage/cover 81 combinations or borders, i s in part a function of the degree of habitat interspersion in the study area. Interspersion of broad habitats is less important in areas where forage and cover are available within one habitat (fine-grained interspersion). As well , the importance of habitat interspersion becomes more d i f f i c u l t to detect in areas having a mosaic of broad habitat types. Research of deer use of borders or habitat mosaics should document the degree of habitat interspersion within the study area to allow for useful comparisons between studies. Results of this study suggest that borders or habitat interspersion are more beneficial to deer in areas where forage and cover are found in separate broad habitats than in areas containing fine-grained interspersion of forage and cover. Fine-grained interspersion of forage within cover areas i s the ideal s i tuation. In some areas, however, c l imatic conditions or local f l o r i s t i c s preclude providing adequate cover and forage in the same area (McNay pers. comm.). In such areas, creation of interspersed habitats would be part icular ly benef ic ia l . This research did not evaluate adequate or optimal patch sizes for cover or forage areas. Evidence from deer distributions around borders, however, indicated that use of both forage and cover areas decreased beyond 100 m from the border. 82 LITERATURE CITED Blymyer, M.J . and H.S. Mosby. 1977. Deer u t i l i z a t i o n of clearcuts in southwestern V i r g i n i a . Southern J . Appl. For. 1: 10-13. Brown E.R. 1961. The black-tai led deer of western Washington. Wash. State Game Dept. B i o l . B u l l . No. 13. 124 pp. Brown E.R. 1985. Management of Wildl i fe and Fish Habitats in Forests of Western Oregon and Washington. U. S. Dep. A g r i c , For. Serv. P a c Northwest Region. 332 pp. Bunnell, F . L . , and A.S. Harestad. 1983. Dispersal and dispersion of black-tai led deer: models and observations. J . Mamm. 64:201-209. Hanley, T .A. 1983. Black-tai led deer, elk, and forest edge in a western Cascades watershed. J . Wi ld l . Manage. 47:237-242. Harestad, A.S . 1979. Seasonal movements of b lack-ta i led deer on northern Vancouver Island. Ph. D. Thesis, Univ. B r i t i s h Columbia, Vancouver, B.C. 184 pp. Johnson, H.D. 1980. The comparison of usage and a v a i l a b i l i t y measurements for evaluating resource preference. Ecology 61: 65-71. Kirchhoff, M.D. and J.W. Schoen. 1983. Black-tai led deer use in re lat ion to forest clearcut edges in southeastern Alaska. J . Wi ld l . Manage. 47: 497-501. Krajina, V . J . 1965. Biogeoclimatic zones and c lass i f i ca t ion of B r i t i s h Columbia. Eco l . Western N. Am. 1:1-17. Leoplod, A. 1933. Game Management. Charles Scribner's Sons, New York, N.Y. 481 pp. Loft , E . R . , E.W. Menke, and T .S . Burton, 1984. Seasonal movement and summer habitats of female black-ta i led deer. J . Wi ld l . Manage. 48: 1317-1325. Lyon, L . J . and C . E . Jensen. 1980. Management implications of elk and deer use of clear-cuts in Montana. J . Wi ld l . Manage. 44: 352-362. McCorguodale, S .M. , K . J . Raedeke, and R.D. Taber. 1986. Elk habitat use patterns in the shrub steppe of Washington. J . Wi ld l . Manage. 50: 664-669. 83 McNay, R . S . , and D.D. Doyle, 1987. Winter habitat selection by black-tai led deer on Vancouver Island: a job completion report. Research, B.C. Ministry of Environment and Parks and B.C. Ministry of Forests and Lands. IWIFR-34. V i c t o r i a , B.C. 89 pp. Ordway, L . L . , and P.R. Krausman. 1986. Habitat use by desert mule deer. J . Wi ld l . Manage. 50: 677-683. Page, R. Computer analysis of radio-telemetry data in the f i e l d , in prep. Reynolds, H.G. 1962. Use of natural openings in a ponderosa pine forest by deer, elk, and catt le . U.S. Dep. A g r i c , For. Serv. Res. Note. RM-78. 4 pp. Reynolds, H.G. 1966. Use of openings in a spruce-f ir forests of Arizona by deer, elk, and catt le . U.S. Dep. A g r i c , For. Serv. Res. Note. RM-66. 4 pp. Scoullar, K.A. and L . L . Kremsater. An approach for describing animal act iv i ty centres, in prep. Servheen, C. 1983. Grizzly bear food habits, movements, and habitat selection in the Missoula Mountains, Montana. J . Wi ld l . Manage. 47: 1026-1035. Sokal, R.R. and F . J . Rohlf. 1981. Biometry: the Principles and Practice of Stat i s t ics in Biological Research. 2 n d ed. , W.H. Freeman and Co. , New York. 859 pp. Sweeney, J . M . , M.E. Garner, and R.P. Burkert. 1984. Analysis of white-tailed deer use of forest clearcuts. J . Wi ld l . Manage. 48: 652-655. Thomas, J.W. tech. ed. 1979. Wildl i fe Habitats in Managed Forests, the Blue Mountains of Oregon and Washington. U. S. Dep. A g r i c , For. Serv. Handbook 553. 512 pp. Willms. W.D. 1971. The influence of forest edge, elevation, aspect, s i te index, and roads on deer use of logged and mature forest, northern Vancouver Island. M.Sc. Thesis. Univ. B r i t i s h Columbia, Vancouver, B.C. 184 pp. Witmer, G.W., and D.S. de Calesta. 1983. Habitat use by female Roosevelt elk in the Oregon coast range. J . Wi ld l . Manage. 47: 933-939. 84 SUMMARY Use of forage, cover, and border habitats by black ta i l ed deer was examined at two levels of selection: within home ranges, and during home range establishment. Within home ranges Black-tai led deer used a wide range of combinations of forage and cover habitats. Relative use of forage and cover areas did not d i f f er from relat ive ava i l ab i l i t y of those areas within home ranges. Relative a v a i l a b i l i t y , however, changed as home ranges changed seasonally, and changed as defined home ranges were altered to ref lect areas of intensive deer use. A s l ight shif t in home range composition to serai stages offering more cover habitat was evident in winter. Areas of intensive deer use were characterized by re la t ive ly more cover than areas receiving less intensive use. Although part icular forage-cover combinations, as defined by serai stages, did not appear important, changing a v a i l a b i l i t y suggested preference for cover habitat during winter and in areas used intensively. Relative use of border habitat did not d i f f er from re lat ive ava i lab i l i t y of border habitat within home ranges at the Nanaimo River area. At the Nimpkish Valley study area, however, re lat ive use of border habitats was usually greater than relat ive ava i l ab i l i t y . High use of border 85 habitat re lat ive to i t s ava i lab i l i ty was more evident in the study area having less interspersion of habitats. For both study areas, relat ive ava i lab i l i t y of border habitat was greater on average during winter than during summer and greater in areas receiving intensive use than in areas receiving less intensive use. The shift towards border habitats during winter and in areas used intensively suggests that borders are preferred habitats. The importance of borders or habitat mixes appears to be a function of the nature of adjacent habitats. Habitat interspersion appears more important where forage and cover are spat ia l ly separate and less important where forage and cover are f inely mixed. Research on deer use of borders or habitat mixes should document the nature of adjacent habitats and the degree of interspersion in the study areas to allow for useful comparisons among studies. Analyses of use and ava i l ab i l i t y of forage, cover, and border habitats within home ranges indicated that use did not often d i f f er from ava i lab i l i t y but a v a i l a b i l i t y changed as home ranges changed seasonally or were altered to ref lect areas receiving intensive use. Changing ava i l ab i l i t y suggested that most selection of forage, cover, or border habitats occurred as home ranges were located, not after their establishment. 86 During home range establishment A v a i l a b i l i t y of forage, cover, and border habitats in home ranges at Nanaimo River was compared to ava i l ab i l i t y of those habitats in areas where home ranges potential ly could have been located. Few deer had different amounts of available forage, cover, or border habitats than expected from composition of potential home ranges, perhaps because of the re la t ive ly high degree of habitat interspersion in that study area. Trends from the few cases where a v a i l a b i l i t y was different from expected suggested preference for cover habitats and borders. Distributions of deer locations around borders indicated that deer were located closer to old-growth borders than expected from the dis tr ibut ion of locations expected from the habitat mosaic. As the area deemed available for potential home ranges increased, differences between observed and expected increased. That increase, although not dramatic due to the high degree of interspersion in the study area, i l lus tra ted how assumptions that delineate what habitats are considered 'available' can have important impacts on results of use-avai labi l i ty analyses. 87 Management implications Habitats having finely interspersed forage and cover are ideal for black-tai led deer. Many areas of Vancouver Island, however, experience cl imatic conditions or local f l o r i s t i c s that preclude providing adequate cover and forage within one broad habitat (McNay pers. comm.). In such areas, habitat interspersion would be important. Mixes of old growth and second growth appear part icular ly beneficial winter habitats. Although this research did not evaluate adequate or optimal patch sizes or shapes, evidence from distributions of deer locations around borders suggested that deer use becomes less than expected beyond 100 m from borders. 88 APPENDIX I: Calculation of a habitat interspersion index for Nanaimo River and Nimpkish Valley study areas. A simple index of habitat interspersion for Nanaimo River and Nimpkish Valley study areas was calculated following a method of Unwin (1981:135-145). The index was based on forage and cover distributions in each study area. To create a forage-cover map for each study area, a 1-km grid was placed over each study area map and gr id ce l l s were c lass i f i ed as forage or cover. Clearcut and subalpine habitats were considered forage, second-growth and o ld -growth forests were considered cover. Spatial autocorrelation of forage and cover was calculated. If l ike habitats tended to be near each other, the result was posit ive spat ial autocorrelation. If d iss imi lar habitats were near each other, negative spatial autocorrelation resulted. To measure spatial autocorrelation, the number of •joins' among grid ce l l s of a part icular type; forage to forage (FF), cover to cover (CC), and cover to forage or forage to cover (FC), were counted. These counted numbers of joins were compared to expected numbers of joins using a Z s t a t i s t i c . The expected number of joins was determined by f i r s t postulating an independent random process, then using probabil i ty theory to predict what value of jo in counts would be expected in the long run. 89 Unwin (1981) showed that an independent random process gives the following results: expected number of FF joins JF F=kp^ expected number of CC joins J c c = k q 2 expected number of FC joins JF C=2kpq where k i s the tota l number of joins on map p is the probabil ity of a grid being coded forage q i s the probabil ity of a grid being coded cover At Nanaimo River and Nimpkish Valley p and q were not known independently of the data set and had to be estimated from the same data as: p= number of forage areas/total number of areas q= 1-p Standard deviations of expected values were (Unwin 1981): standard deviation of J F F = (kp2+2mp;*-(k+2m)p*) ?* ^ standard deviation of J c c = ( k q 2 + 2 m q 3 - ( k + 2 m ) q 4 l ° * 5 standard deviation of J F C=(2(k+m)pq-4(k+2m)p 2q 2) 0 , 5 n where m is 0.5 J i f j ^ - l ) and j ^ i s the number of joins to the i t h area; n i s the to ta l number of areas. The probabil i ty of any one specified number of joins of a specif ic type, given an independent random process was found using the Z s t a t i s t i c : Z = observed number of joins - expected number of joins standard deviation of expected values Unwin, D. 1981. Introductory Spatial Analysis. Methuen and Co. Ltd . New York. 212 pp. 90 Appendix II . An approach for describing deer act iv i ty centers A summary of a new method for determining deer home ranges and patterns of deer act iv i ty within home ranges is presented (Scoullar and Kremsater in prep). The method calculates a probabil i ty density function for each location then superimposes these to describe contours and centres of animal ac t iv i ty . The calculation of the home range i s demonstrated using data collected on Vancouver Island for Columbian black-tai led deer (Odocoileus hemionus  columbianus). Comparisons are drawn with the minimum convex polygon method commonly used for deer. Burt(1943) 2 / defined an animal's home range as "the area traversed by the individual in i t s normal ac t iv i t i e s of food gathering, mating, and caring for young". He believed that occasional movements outside the area should not be included in the home range. Most researchers accept and use Burt's def ini t ion (e.g. , Dixon and Chapman 1980 3 / , Harestad 1979 4 / , Jennrich and Turner 1969 5 / , and Laundre and Keller 1984 6 / ) . There are, however, several methods for identifying and excluding extreme locations. Estimates of home range size and shape vary, therefore, with the method chosen to identify extreme locations. 91 The home range method most commonly used for deer is the minimum convex polygon method which identi f ies outlying locations as ones furthest from the arithmetic mean center of a l l locations. The arithmetic mean center i s found by taking the average of the locations in the 'x' direct ion and the average in the 'y' direct ion. The center of ac t iv i ty calculated using the arithmetic mean is the center of the u t i l i z a t i o n dis tr ibut ion, but may have no bio logica l significance. It does not necessarily correspond to the area of most intense use in the home range. It may, in fact, f a l l ent irely outside the home range. Most s t a t i s t i c a l methods of calculating home ranges (e.g. , Jennrich and Turner 1969) u t i l i z e the arithmetic mean center (review of Dixon and Chapman 1980). The minimum convex polygon method f a i l s to accurately identify areas of greater ac t iv i ty in a home range because i t excludes locations based on their distance from the arithmetic mean center which does not ref lect the b io log ica l center of ac t iv i ty . The minimum convex polygon method has a further weakness in that i t may include areas in the home range that are not used or travel led by the deer. The home range method presented here attempts to estimate the true biological center of deer ac t iv i ty and identify patterns of intensity of deer use. Motivation to produce this new home range method stemmed from concerns about weaknesses of the minimum convex polygon method. 92 The method estimates deer home ranges from telemetry data. F i r s t , the distance of the nearest neighbour for each location is calculated. Locations having the furthest nearest neighbour may be dropped from the home range depending on the def init ion of home range desired. A 100% home range would not exclude any points; a 90% home range would exclude the 10% of locations having the furthest nearest neighbours; a 50% home range would exclude the 50% of points having the furthest nearest neighbours (Fig. 1). Thus extreme locations are eliminated based on the ir distance from their closest neighbour and the percent home range def ini t ion desired. Once the points defining the home range are determined, a probabil i ty density function i s used to describe each location. Radio locations are not accurate point estimates of a deer's location as they are measured with error and are only point samples in time. Rather, for the probabil i ty density function home range method, radio locations are considered to represent areas of habitat used by that deer. The spread of the probabil i ty density function i s defined as the distance between locations in the home range that have the furthest nearest neighbour. That distance w i l l be cal led the 'spread distance', and can be interpreted as the maximum distance around a single location which w i l l be considered habitat used by the animal. Spread distances change as the locations used in a home range are changed to ref lect different intensit ies of use. The probabil i ty density function i s generated by assigning a value of 100 to the grid c e l l that contains the point location. Grid ce l l s at progressively larger distances from the point location are given progressively lower values u n t i l gr id ce l l s at the maximum distance (as defined by the spread distance) have a value of 1 (Fig. 2). Superimposing the probabil i ty density functions for a l l locations and adding the values for each grid c e l l produces a home range gr id . The values in the home range grid are truncated to the nearest 100 to allow easier perception of ac t iv i ty centers (Fig. 3). Areas of intense deer use within the home range are indicated by grid ce l l s having large assigned values. Figures 4 and 5 i l lu s t ra te differences in home range shapes produced by the minimum convex polygon and the probabil i ty density function home range methods for seasonal deer home ranges from Nanaimo River. For some home ranges, both methods produce similar results . For home ranges that have bimodal distr ibutions, or an irregular spat ial d is tr ibut ion of points, however, the methods can produce quite different home range shapes. For the 7 deer used from the Nimpkish Valley, the distr ibut ion of deer locations were neither bimodal nor irregular. Thus the minimum convex polygon method and the probabil i ty density function approach would have resulted in similar home range boundaries. 94 Scoullar K . A . , and L . L . Kremsater. An approach for describing animal act iv i ty centres, in prep. 2 / Burt, W.H. 1943. T e r r i t o r i a l i t y and home range concepts as applied to mammals. J . Mamm. 24: 346-352 3 / Dixon, K . R . , and J . A . Chapman. 1980. Harmonic mean measure of animal act iv i ty areas. Ecology 61:1040-1044. 4 / Harestad, A.S . 1979. Seasonal movements of black-tai led deer on northern Vancouver Island. Ph.D. Thesis. Univ. B r i t i s h Columbia, Vancouver, B.C. 184 pp. 5 / Jennrich, R . J . , and F .B . Turner. 1969. Measurement of non-circular home range. J . Theoret. B i o l . 22:227-237. 6 / Laundre, J .W. , and B . L . Kel ler . 1984. Home-range size of coyotes: a c r i t i c a l review. J . Wi ld l . Manage. 48:127-139. 95 a) 100% x X x * x X X b) 90% X x x X X © X © X c) 50% x X x X Figure 1. An example of elimination of extreme locations for home ranges of different intensi t ies , "x" indicates a location, c i rc l ed locations are excluded from the home range. A 100% home range does not exclude any locations (a); a 90% excludes 10% of the locations that have the farthest nearest neighbour (b): a 50% home range excludes 50% of the locations having the farthest nearest neighbours. 96 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Ci 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Ci 0 1 1 1 2 2 1 1 1 0 0 0 0 0 0 0 Ci 0 0 n 6 9 11 12 14 16 14 12 11 9 6 2 0 0 0 0 0 0 0 0 6 12 17 21 23 27 30 27 23 21 17 12 6 0 0 0 0 0 0 0 1 9 17 25 31 35 40 44 40 35 31 25 17 9 1 0 0 0 0 0 Ci 1 11 21 31 41 47 53 SB 53 47 41 31 21 11 1 0 0 0 0 0 0 1 12 23 35 47 59 66 72 66 59 47 35 23 12 1 0 0 0 0 0 0 2 14 27 40 53 66 79 B6 79 66 S3 40 27 14 2 0 0 0 0 0 0 2 16 30 44 5B 72 86100 86 72 58 44 30 16 2 0 0 0 Cl 0 0 2 14 27 40 53 66 79 B6 79 66 53 40 27 14 2 0 0 0 0 0 0 1 12 23 35 47 59 66 72 66 59 47 35 23 12 1 0 0 0 0 0 0 1 11 21 31 41 47 53 58 53 47 41 31 21 11 1 0 0 0 0 0 0 1 9 17 25 31 35 40 44 40 35 31 25 17 9 1 0 0 0 0 0 0 0 6 12 17 21 23 27 30 27 23 21 17 12 6 0 0 0 0 0 0 0 0 2 6 9 11 12 14 16 14 12 11 9 6 2 0 0 0 0 0 0 0 0 0 0 1 1 1 2 2 2 1 1 1 Ci 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 o 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Figure 2. An example of a probabil i ty density function for 1 location having i t s nearest neighbour 215 m away. Each grid represents 30 m. The grid c e l l containing the point location i s given a value of 100. Grid ce l l s at progressively larger distances from the point location are given lower values (linear decrease) u n t i l ce l l s 215 m away have a value of 1. 97 a) 100% a im i t t i i I I I ! niii i s : : : b) 90% C ) 50% Figure 3. Examples of 100% (a), 90% (b), and 50% (c) home range grids. "99" indicates grid ce l l s containing point locations. Grid ce l l s having large values represent areas of more intensive use. Locations excluded from the home range are designated by 8 8 " • 9 8 a) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 199 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 X 1 1 X 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 X X X X 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 X X X X X 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 X 1 1 1 0 0 0 X X X X X X 0 0 0 0 0 0 0 0 1 1 X 2 2 2 2 2 2 2 2 1 1 1 1 X 1 X X X X X X 0 0 0 0 0 0 0 0 X 1 1 2 2 3 3 3 3 3 2 2 2 2 1 1 X 1 X X X X X X 0 0 0 0 0 0 0 0 X 1 2 2 3 3 3 3 3 3 3 3 2 2 2 1 X 1 X X X X X X 0 0 0 0 0 0 0 1 X 2 3 3 4 4 4 4 4 4 4 3 3 2 2 X X 1 X X X X X 0 0 0 0 0 0 0 1 1 2 3 3 4 4 5 5 5 5 4 4 4 3 3 2 2 X X X99 X 0 0 0 0 0 0 0 0 1 2 2 3 4 4 5 5 S 6 5 5 5 4 4 3 3 2 X 1 X X X 0 0 0 0 0 0 0 0 1 1 2 3 3 4 5 599 6999999 599 4 3 3 2 X X X 0 0 0 0 0 0 0 0 0 1 2 2 3 4 4 5 6 6 6 6 6 6 5 4 4 3 3 2 X 1 X X 0 0 0 0 0 0 0 0 X X 2 2 3 4 5 5 6 6 6 6 6 6 5 4 4 3 3 2 X 1 X 0 0 0 0 0 0 0 0 0 1 1 2 2 3 4 5 5 6 6 6 6 6 6 5 4 4 3 2 2 X 1 0 0 0 0 0 0 0 0 0 0 1 1 2 3 3 4 5 6 6 699 6 6 5 5 4 3 3 2 1 X 0 0 0 0 0 0 0 0 0 0 0 X 1 2 3 4 4 5 6 6 6 6 6 6 5 4 4 3 2 2 1 X 0 0 0 0 0 0 0 0 0 0 0 1 1 2 3 4 5 5 6 6 6 6 6 5 5 4 3 3 2 X 1 0 0 0 0 0 0 0 0 0 0 0 0 X 2 2 3 49999 5 6 6 6 5 S 4 4 3 2 2 1 0 0 0 0 0 0 0 0 0 0 0 0 X 1 2 2 3 4 4 5 5 5 5 5 5 5 4 3 3 2 1 1 0 0 0 0 0 0 0 0 0 0 0 0 1 1 2 2 3 4 4 4 5 5 5 5 5 4 4 3 2 2 1 1 0 0 0 0 0 0 0 0 0 0 0 0 X 1 2 2 3 3 4 4 4 5 5 5 4 4 3 3 2 1 X 1 0 0 0 0 0 0 0 0 0 0 0 0 1 1 2 2 3 3 4 4 4 499 499 4 3 3 2 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 1 1 2 2 3 3 4 4 4 4 4 4 4 4 3 3 2 2 1 1 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 2 2 3 3 4 4 4 4 4 4 4 3 3 2 2 X 1 X 0 0 0 0 0 0 0 0 0 0 0 0 1 1 2 2 3 399 4 4 4 4 4 3 3 3 2 2 1 1 X X 0 0 0 0 0 0 0 0 0 0 0 1 1 X 2 2 3 3 3 3 3 3 3 3 3 3 2 2 2 1 1 X 0 0 0 0 0 0 0 0 0 0 0 0 1 X 2 2 2 3 3 3 3 3 3 3 3 3 2 2 2 1 X X X 0 0 0 0 0 0 0 0 0 0 0 0 1 X 2 2 2 3 3 3 3 3 3 3 3 2 2 2 1 X X X 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 2 2 2 2 3 3 3 399 3 2 2 299 X X X 0 0 0 0 0 0 0 0 0 0 0 0 0 1 X 1 2 2 2 2 2 3 3 3 2 2 2 2 1 X X X 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 2 2 2 2 2 2 2 2 2 1 1 X X X 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 X 1 1 1 2 2 299 2 2 X X 1 X X X 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 X 1 1 X 1 1 1 1 1 1 1 X X 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 X X 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 X 1 1 1 1 X 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 X 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 n n n n 0 0 0 0 0 0 0 0 D fl 0 0 0 p 0 0 0 0 0 0 Q 0 0 D b) Figure 4. Comparison of a 100% home range generated by the probabil i ty density function (a) and minimum convex polygon (b) methods. 99 a) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 199 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 088 0 0 0 0 0 0 0 0 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 199 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 088 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 088 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 199 1 1 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 0 0 0 0 0 0 0 1 1 1 2 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 199 1 1 1 1 0 0 0 0 0 199 299 2 2 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 0 0 0 0 0 1 1 2 2 299 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 088 0 0 0 0 0 0 0 0 1 199 0 0 0 0 0 0 1 2 299 2 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 088 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 X X. Figure 5. Comparison of a 50% home range generated by the probabil i ty density function (a) and minimum convex polygon (b) methods. 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0097526/manifest

Comment

Related Items