Mule Deer Winter Habitat Model Morice and Lakes Forest Districts IFPA Prepared for: Morice and Lakes IFPA Prepared by: Anne-Marie Roberts Smithers, BC DRAFT March 2004 Mule Deer Winter Habitat Model - Morice and Lakes Forest Districts IFPA Executive Summary Species – Habitat models are used to evaluate the potential in the Morice and the Lakes forest districts to provide suitable habitat for wildlife species that were selected by the Ecosystem group of the Morice and Lakes Innovative Forest Practices Agreement (ML-IFPA). The models generally define habitat suitability based on the provision of certain habitat attributes required for living and/or reproduction. Unchanging environmental conditions (such as Biogeoclimatic subzone), location of infrastructure and development, and projected forest conditions (from the rules defined in individual scenarios), supply much of the basic information that can be used in the habitat supply models. There are other habitat attributes that are not directly provided by the available data layers that describe forest cover in terms of species composition and age. These habitat attributes are derived from information provided in the forest cover dataset and from data provided in the Predictive Ecosystem Mapping (PEM) using mathematical models and/or beliefs expressed in the Netica conditional probability tables (Habitat Modeling report #1, in prep). Empirical relationships, scientific literature, and professional expertise are incorporated into these equations and/or tables to describe the changes in the state (e.g. abundance, density) of these habitat attributes through changes in forest succession and disturbance. This report describes the development of the mule deer winter habitat suitability model. Page i 2175 Millar Rd. Smithers, BC V0J 2N6 Ph. (250) 877-6705FAX (250) 877-6805 Mule Deer Winter Habitat Model - Morice and Lakes Forest Districts IFPA Table of Contents Executive Summary ................................................................................................................................ i INTRODUCTION....................................................................................................................................... 5 SPECIES ACCOUNT AND WINTER HABITAT USE INFORMATION........................................... 5 DISTRIBUTION ........................................................................................................................................... 5 Provincial Range .................................................................................................................................. 5 Elevation Range .................................................................................................................................... 5 Provincial Context ................................................................................................................................ 5 ECOLOGY AND KEY WINTER HABITAT REQUIREMENTS .......................................................................... 5 General ................................................................................................................................................. 5 WINTER HABITAT USE – LIFE REQUISITES ............................................................................................... 6 Feeding Habitat .................................................................................................................................... 6 Thermal/Snow Interception Cover Habitat ........................................................................................... 7 MULE DEER MODEL .............................................................................................................................. 8 APPLICATION OF MODEL .......................................................................................................................... 8 Late Winter Habitat Model Assumptions .............................................................................................. 8 MULE DEER WINTER HABITAT MODEL DESCRIPTION ............................................................................. 8 Description of Network Nodes .............................................................................................................. 9 Mule Deer Winter Habitat Suitability ............................................................................................... 9 Thermal Cover ................................................................................................................................ 11 Accessible Forage ........................................................................................................................... 11 Thermal /Snow Interception Value ................................................................................................. 12 Late Winter Forage Abundance ...................................................................................................... 13 Effective Snow Depth ..................................................................................................................... 14 Crown Closure Class....................................................................................................................... 14 Forest Type ..................................................................................................................................... 15 Aspect ............................................................................................................................................. 15 Elevation ......................................................................................................................................... 15 Snow Depth..................................................................................................................................... 15 Litterfall .......................................................................................................................................... 15 Lichen Litterfall .............................................................................................................................. 16 Conifer Litterfall ............................................................................................................................. 17 Arboreal Lichen Abundance ........................................................................................................... 17 Forest Type ..................................................................................................................................... 17 Winter Ground Forage .................................................................................................................... 18 Leading Species and Secondary Species ........................................................................................ 18 Age Class ........................................................................................................................................ 18 ALR................................................................................................................................................. 18 Island ............................................................................................................................................... 18 COMPUTING OUTPUT FOR HABITAT VALUE ........................................................................................... 18 SENSITIVITY ANALYSIS ........................................................................................................................... 19 TESTING AND VALIDATION ..................................................................................................................... 19 RESEARCH NEEDS FOR MODEL VERIFICATION ....................................................................................... 19 IMPLICATIONS TO FOREST MANAGEMENT .............................................................................................. 19 List of Tables Page iii 2175 Millar Rd. Smithers, BC V0J 2N6 Ph. (250) 877-6705FAX (250) 877-6805 Mule Deer Winter Habitat Model - Morice and Lakes Forest Districts IFPA Table 1. Important mule deer winter foods. ................................................................................................. 7 Table 2. Conditional Probability table showing the mule deer winter habitat suitability rating based on the influences of accessible forage, forested thermal cover, and whether the site is located on an island.10 Table 3. Conditional Probability Table showing the probability of late winter forage abundance for nonALR lands. ........................................................................................................................................... 13 Table 4. Conditional probabilities used within Netica to model effective snow depth from snow depth by biogeoclimatic subzone, modified by aspect and elevation. ............................................................... 14 Table 5. Conditional Probability Table predicting total lichen and conifer litterfall. .................................... 16 List of Figures Figure 1. Habitat variables and ecological relationships used to build the mule deer winter habitat suitability Bayesian belief model in the Netica© program..................................................................... 9 Figure 2. Forested thermal cover habitat value based on aspect and the ability of forest cover attributes to provide thermal cover and the interception of snow. ...................................................................... 11 Figure 3. Accessibility of forage in late winter for sites rated for moderate and high abundance forage and modified by effective snow depth. ....................................................................................................... 12 Figure 4. Probable thermal/snow interception value based on forest type and crown closure classes. ... 13 Figure 5. Probability of lichen litterfall associated with stands of nil, low, moderate, or high arboreal lichen abundance. ......................................................................................................................................... 16 Figure 6. Probability of conifer litterfall associated with “forest type” class poor to very good by age class groupings 4, 5-6, and 7+. .................................................................................................................... 17 Page iv 2175 Millar Rd. Smithers, BC V0J 2N6 Ph. (250) 877-6705FAX (250) 877-6805 Mule Deer Habitat Model - Morice and Lakes Forest Districts IFPA INTRODUCTION This report describes the mule deer winter habitat model developed for the Morice and Lakes Innovative Forest Practices Agreement (ML-IFPA). The following document includes a species account for mule deer (Odocoileus hemionus hemionus), outlines the logic used and assumptions made in the preparation of the model, describes the model structure and relationships used to build the model, and outlines testing of model sensitivity and the current level of validation. SPECIES ACCOUNT AND WINTER HABITAT USE INFORMATION Common Name: Mule Deer Scientific Name: Odocoileus hemionus hemionus Species Code: M-ODHE Status: The mule deer is a Yellow (Ym) listed species by the Provincial Tracking Lists of the British Columbia Conservation Data Centre and is managed for hunting purposes in the province of British Columbia (Ministry of Environment, Lands and Parks 1994). Distribution Provincial Range There are three sub-species of mule deer (Odecoileus hemionus) in British Columbia. Columbian blacktailed deer (Odocoileus hemionus columbianus) are found on the southern coast and Vancouver Island; Sitka black-tailed deer (Odocoileus hemionus sitkensis) occur in the Queen Charlotte Islands and the northern mainland coast; and mule deer (Odocoileus hemionus hemionus) are found in the interior portions of the province (Bunnell 1990). Mule deer (O. h. hemionus) are found throughout the noncoastal mainland portion of the province in all biogeoclimatic zones except the Mountain Hemlock biogeoclimatic zone (Stevens 1995). Elevation Range Mule deer (O. h. hemionus) occur in forested habitats from sea level to sub-alpine elevations, and may use alpine areas in summer (Stevens 1995). Provincial Context Mule deer occur commonly within the forested areas of the province. The provincial population of mule deer was estimated at 170,000 in 1996 and is thought to be stable (R. Marshall, pers. comm.). A study of mule deer (O. h. hemionus) near Williams Lake (Armleder et al. 1986) has contributed much to the knowledge of winter habitat use in the interior of BC. The mule deer (O. h. hemionus) in the study area are likely near the northern and western extent of their range (Cowan and Guiget 1978). Ecology and Key Winter Habitat Requirements The information used in this species account is based on research on both interior and coastal subspecies of Odocoileus hemionus. Information was selected from the literature that is appropriate for the mule deer sub-species O. h. hemionus found in the Morice and Lakes Forest Districts. General The mule deer is a medium sized member of the family Cervidae that prefers habitat diversity including conifer forests, mixed forests, meadows and riparian areas. Coniferous forests and riparian areas receive the most use because these habitat types provide security and thermal cover as well as abundant forage (Carson and Peek 1987). Distribution of deer is often along ecotones and deer make use of edge habitats that allow them to access different habitat types for feeding and cover requirements (Kufeld et al. 1988). Mule deer feed on various herbs, grasses, shrubs and trees and their diet varies seasonally depending on factors such as food availability, preference and plant quality (Willms et al. 1976). Winter diets shift from Page 5 2175 Millar Rd. Smithers, BC V0J 2N6 Ph. (250) 877-6705FAX (250) 877-6805 Mule Deer Habitat Model - Morice and Lakes Forest Districts IFPA low-lying grasses and forbs in the early portion of winter to higher shrubs as snow levels increase in late winter (Carpenter et al. 1979, Bunnell 1990). In general, early and intermediate structural stages with low tree canopy closure and abundant herbaceous vegetation and shrubs are preferred as forage habitats by mule deer (Deshamp et al. 1979, Collins and Urness 1983, Loft and Menke 1984, Bunnell 1990). In the later portions of winter, snow levels in open areas may be greater than mule deer are able to accommodate and they are forced to forage in forested areas where trees have intercepted the snow. Mature and old-growth forests are important winter forage areas as the low snow levels resulting from the closed canopy allow better access to available forage species (Kirchhoff and Schoen 1987, Bunnell 1990, Armleder et al. 1994). Thermal cover from cold temperatures and wind are also critical factors in mule deer winter habitat selection (Bunnell 1990, Armleder and Dawson 1992). This requirement is best obtained from forested stands that are multi-layered and have trees with large crowns (Kirchhoff and Schoen 1987, Armleder et al. 1994). In areas of high snowfall, mule deer will occupy separate ranges in summer and winter and are termed migratory, while deer that remain within the same area all year round are termed resident (Ihsle Pac et al. 1988, Bunnell 1990). In some cases, transitional ranges may be present, and are usually adjacent to winter range. Long distance movements between summer and winter ranges have been documented to vary from 19 km (Brown 1992) to a maximum of 148 km (Zalunardo 1965). The movement by mule deer from summer to winter ranges has been found to be a response to the effects of changing daylight lengths rather than changes in snow levels found at higher elevations (Garrot et al. 1987). Mule deer movement from winter to summer range has been suggested to be in response to spring increases in temperature, insect activity and the maturation of vegetation (Loft et al. 1989). Distinct movement corridors between seasonal use areas have also been documented, with deer using the corridors on an annual basis (Thomas and Irby 1990). Numerous studies have found that mule deer display a high degree of fidelity to traditional seasonal ranges (Zalunardo 1965, Ihsle Pac et al. 1988, Kufeld et al. 1989, Loft et al. 1989). This site fidelity could be a function of the topography and vegetation (Ihsle Pac et al. 1988) as well as the traditions of matriarchal groups (Loft et al. 1989). Kufeld et al. (1989) working in Colorado, reported winter home range sizes that averaged 211 ha (range = 172 to 292 ha) for five female deer and 226 ha (range = 117 to 323 ha) for 17 female deer (Kufeld et al. 1989). In Montana, Ihsle Pac et al. (1988) found that the winter range of ten female mule deer averaged 15.9km 2. In any case, winter range must be of sufficient forage quality so that deer can maintain their energy balance and thereby slow the rate of weight loss during winter (Carpenter et al. 1979, Torbit et al. 1985). Winter Habitat Use – Life Requisites Winter is the critical season for mule deer in northern areas because deep snow results in reduced forage availability and an increase in energy expenditure during movement. Areas with deep snow preclude use by mule deer (Gilbert et al. 1970, Parker et al. 1984, Bunnell 1990) and can cause population declines if alternate low-snow habitats are not available (Gilbert et al. 1970). In winter, deer move to southerly aspects, lower elevations and exposed ridges where snow depths are lower (Gilbert et al. 1970, Wilms et al. 1976). Feeding Habitat In winter forage opportunities are reduced by snowfall, which either covers forage plants or limits the mobility of deer. Many researchers have documented the slow decrease of mule deer fat reserves and body weight during the winter due to reduced digestibility of winter forage (Carpenter et al. 1979, Wallmo and Schoen 1980, Torbit et al. 1985). To slow this trend, deer spend more time resting and less time feeding in winter than in summer, to conserve energy and allow their rumen to extract the most from their food (Kufeld et al. 1988, Bunnell 1990). During winter, a variety of shrubs are the preferred forage for mule deer. Mule Deer also forage on conifer needles and arboreal lichens from litterfall (Willms et al. 1976, Armleder et al.1986, Armleder and Dawson 1992). Although shrubs are abundant in open areas, burial by snow and the increased energy cost of travel in deep snow reduces their true availability (Armleder and Dawson 1992). Subalpine fir (Abies lasicarpa), huckleberry (Vaccinium Spp.), bunchberry (Cornus canadensis), five-leaved bramble (Rubus pedatus) and arboreal lichens (Alectoria, Bryoria and Page 6 2175 Millar Rd. Smithers, BC V0J 2N6 Ph. (250) 877-6705FAX (250) 877-6805 Mule Deer Habitat Model - Morice and Lakes Forest Districts IFPA Usnea spp.) are all important winter foods (Willms et al. 1976, Nyberg et al. 1989, Bunnell 1990). Foods offering a minimum of protein content in summer are almost valueless in mid-winter (Einarsen 1946). Use of mature forests by mule deer in winter has been documented by a number of authors, especially in coastal areas where snow accumulations can be high (Harestad 1979, Wallmo and Shoen 1980, Harestad et al. 1982, Armleder et al. 1986, Kirchhoff and Shoen 1987). Use of patchy forests has been documented, with the mosaic of open and closed canopies providing a mixture of forage shrubs and snow interception (Bloom 1978). Carson and Peek (1987) found that coniferous and riparian forest types provided the best conditions of high forage availability and quality combined with thermal cover. It has also been suggested that uneven-aged mature and old-growth forests provide higher quality forage than that found in openings and even-aged second growth, as they have a greater abundance and diversity of understory species (Shoen et al. 1985). Although in summer mule deer may use the cover and forage offered simultaneously by edge habitats, high snow accumulation may negate their use in winter (Kirchoff and Shoen 1983). A summary of foods used by mule deer in winter is provided in Table 1 (Willms et al. 1976, Blower, 1982, Nyberg et al. 1989, Bunnell 1990). Table 1. Important mule deer winter foods. Type of food Plant Species Trees Pseudotsuga menziesii Pinus contorta Abies lasiocarpa Thuja plicata Tsuga heterophylla Shrubs Vaccinium spp. Arctostaphylos uva-ursi Ribes spp. Betula glandulosa Menziesia ferruginea Prunus pensylvanica Rosa spp. Rubus pedatus Amelanchier alnifolia Cornus stolonifera Sorbus spp. Paxistima myrsinites Linnaea borealis Salix spp. Acer glabrum Lonicera involucrate Rubus spectabilis Gaultheria shallon Herbs grass Rubus pedatus Cornus Canadensis Carex spp. Arboreal Lichens Alectoria spp. Cetraria spp. Bryoria spp. Usnea spp. Thermal/Snow Interception Cover Habitat Snow interception cover is likely the most important cover requirement for deer in winter, as deer movement through 5 cm of snow increases energy expenditures by 10% while snow depths of 50 cm increase energy expenditure by up to 400% (Parker et al. 1984). Snow depth is so critical that Bunnell (1990) suggested that there are no suitable habitats for mule deer in areas experiencing snow depths greater than 60 cm that persist for more than two to three weeks. Gilbert et al. (1970) concluded that snow depths greater than 46 cm preclude the use of an area by deer. Predicting good snow interception cover is related to canopy closure, but net timber volume can be used as a more reliable indicator of reduced snow accumulation (Kirchhoff and Shoen 1987). Snow interception cover is best obtained from forest stands that are multi-layered and have trees with large crowns (Kirchhoff and Schoen 1987, Armleder et al. 1994). Armleder et al. (1994) found that mule deer preferred the low snow conditions of old-growth coniferous stands, which displayed conditions of high crown closure (65%) and that the use of these habitats increased with increasing snow depth. Mule deer use a great deal of energy to maintain their body temperature under conditions of cold temperatures. The energy cost of maintaining a constant body temperature is compounded by exposure to wind (Armleder and Dawson 1992). Thermal cover is best provided by multi-layered stands that reduce air movement at ground level and minimize radiation of heat to the open sky (Kirchhoff and Schoen 1987, Armleder et al. 1994). Topographic features can also provide a windbreak and reduce the energy costs of thermoregulation (Nyberg et al. 1989). Page 7 2175 Millar Rd. Smithers, BC V0J 2N6 Ph. (250) 877-6705FAX (250) 877-6805 Mule Deer Habitat Model - Morice and Lakes Forest Districts IFPA MULE DEER MODEL Application of Model Season: Winter (thermal and foraging) Habitat Areas: All landscape units in the Morice and the Lakes forest districts in central British Columbia. Model Output: The model will produce a most probable suitability value for mule deer winter habitat. Verification Level: Verification of the model involved testing the belief net to ensure that the output is consistent with our expected output. Late Winter Habitat Model Assumptions The following section describes the logic and assumptions used to translate habitat element information for mule deer to the variables and equations used in the models. 1. Suitable winter range is critical to mule deer, and is largely a function of accessible forage and thermal/snow cover. 2. Relatively low canopy closures are assumed to support greater amounts of shrubs and herbaceous vegetation and hence forage for mule deer, but availability and accessibility will be reduced due to higher snow levels. 3. High canopy closures are assumed to result in greater snow interception and snow load retention than lower canopy closures, thereby enhancing mule deer habitat cover values (thermal, security, and snow interception). It is assumed that mule deer do not winter in areas of deep and persistent snow packs. 4. Desirable structural attributes required for thermal cover and snow interception , such as a multilayered forest canopy and large trees, are assumed to be correlated with higher structural stage. 5. The ESSFmc, ESSFmcp and AT biogeoclimatic subzones and/or elevations above 1000 m are assumed to have no value to mule deer as winter habitat in the project area due to deep (>60 cm) and persistent (for more than two weeks) snow. 6. Mule deer forage on arboreal lichens as litterfall on the forest floor. It is assumed that the more abundant the arboreal lichens, the more litterfall there will be. A similar assumption is made for conifer litterfall. 7. Late winter forage abundance is a function of winter ground forage, which is derived from the PEM, and both arboreal lichen and conifer litterfall, which are derived from forest cover attribute data. 8. Islands are assumed to have no habitat value for mule deer. 9. ALR lands are assumed to have a habitat values of “Low” rather than “Nil” as indicated for other PEM polygons identified as “development”. 12. Edges or ecotones between areas that provide high habitat cover values and high forage values are recognized to be highly favoured by mule deer because they minimize energy expenditures between accessing high forage sites and good thermal/snow interception habitat. Mule Deer Winter Habitat Model Description Suitable winter range is assumed to be critical to mule deer. Winter range is highly defined by snow depth at both the landscape and the stand scale. Mule deer will often use edges or ecotones between areas that provide high cover values (thermal, security, and snow interception) and high forage values. Page 8 2175 Millar Rd. Smithers, BC V0J 2N6 Ph. (250) 877-6705FAX (250) 877-6805 Mule Deer Habitat Model - Morice and Lakes Forest Districts IFPA The mule deer winter habitat suitability model is developed as a Bayesian belief influence network using the Netica© program (See Figure 1). The model is made up of a series of network nodes, some of which require data manipulation to generate a useful value. The data from these nodes is applied to the model in a logical sequence in order arrive at a predictive value for winter habitat suitability. The relevance and logic of each node is explained in the following section. The output of the mule deer winter habitat model predicts habitat suitability based on cover (thermal and snow interception) value and foraging value. Forest Type Coniferous 25.0 Mixed 25.0 Deciduous 25.0 Non forested 25.0 0.438 ± 0.43 Crown Closure 0 to 1 25.0 1 to 3 25.0 3 to 6 25.0 >= 6 25.0 Aspect Cool 33.3 Warm 33.3 None 33.3 0.533 ± 0.37 Snow Depth Shallow 33.3 Moderate 33.3 Deep 33.3 Elevation < 1000 50.0 >= 1000 50.0 Thermal/Snow Interception Value Nil 36.3 Poor 31.3 Moderate 19.4 Good 13.1 Snow Very Shallow Shallow Moderate Deep Forested Thermal Cover Nil 18.1 Low 39.5 Moderate 26.5 High 15.9 0.328 ± 0.29 Yes No Depth 8.33 26.7 31.7 33.3 Island? 50.0 50.0 Accessible Forage Nil 25.0 Low 46.1 Moderate 20.9 High 7.96 Arboreal Lichen Abundance Nil 25.0 Low 25.0 Moderate 25.0 High 25.0 Leading Species Fd 16.7 S 16.7 B 16.7 P 16.7 H 16.7 Other 16.7 Lichen Litterfall Nil 25.0 Low 37.5 Moderate 25.0 High 12.5 Forest Type Poor 18.1 Moderate 44.4 Good 24.3 Very Good 13.2 Litterfall Nil 9.78 Low 45.1 Moderate 33.9 High 11.3 Conifer Litterfall Nil 20.0 Low 46.4 Moderate 22.9 High 10.7 Late Winter Forage Abundance Nil 25.0 Low 11.7 Moderate 39.8 High 23.5 Mule Deer Winter Habitat Value Nil 66.2 Low 23.2 Moderate 7.54 High 3.12 Secondary Species Fd 16.7 S 16.7 B 16.7 P 16.7 H 16.7 Other 16.7 0 to 1 1 to 4 4 to 5 5 to 7 >= 7 Age Class 20.0 20.0 20.0 20.0 20.0 Winter Ground Forage Nil 25.0 Low 25.0 Moderate 25.0 High 25.0 ALR Non ALR? 50.0 50.0 Figure 1. Habitat variables and ecological relationships used to build the mule deer winter habitat suitability Bayesian belief model in the Netica© program. C Description of Network Nodes Mule Deer Winter Habitat Suitability The mule deer winter habitat suitability node is the final output node of the network model The value of a particular habitat to mule deer is a measure of the thermal cover value (including snow interception cover in winter) and the forage value of that habitat, which is modified by its proximity to other habitat types (See Table 2). For example, an open wetland or meadow may have high forage value, but no cover value. However, the value of such a habitat is increased if it is within 100 m of cover habitat (i.e. high canopy closure, mature coniferous forest). Note that if the polygon is located on an island the resulting winter habitat value is reduced to nil. Example Relationship: Conditional Probability Table Where: MDWHV = Mule deer winter habitat value MDTCV = Mule deer forested thermal cover value MDFAV = Mule deer accessible forage value ISL = Island habitat Page 9 2175 Millar Rd. Smithers, BC V0J 2N6 Ph. (250) 877-6705FAX (250) 877-6805 B Mule Deer Habitat Model - Morice and Lakes Forest Districts IFPA Table 2. Conditional Probability table showing the mule deer winter habitat suitability rating based on the influences of accessible forage, forested thermal cover, and whether the site is located on an island. Mule Deer Winter Habitat Suitability Island? Forested Thermal Cover Accessible Forage Nil Low Moderate High No Nil Nil 100 0 0 0 No Nil Low 100 0 0 0 No Nil Moderate 100 0 0 0 No Nil High 100 0 0 0 No Low Nil 100 0 0 0 No Low Low 0 100 0 0 No Low Moderate 0 60 40 0 No Low High 0 35 50 15 No Moderate Nil 100 0 0 0 No Moderate Low 0 90 10 0 No Moderate Moderate 0 0 100 0 No Moderate High 0 0 20 80 No High Nil 100 0 0 0 No High Low 0 70 30 0 No High Moderate 0 0 20 80 No High High 0 0 0 100 Yes Nil Nil 100 0 0 0 Yes Nil Low 100 0 0 0 Yes Nil Moderate 100 0 0 0 Yes Nil High 100 0 0 0 Yes Low Nil 100 0 0 0 Yes Low Low 100 0 0 0 Yes Low Moderate 100 0 0 0 Yes Low High 85 15 0 0 Yes Moderate Nil 100 0 0 0 Yes Moderate Low 100 0 0 0 Yes Moderate Moderate 100 0 0 0 Yes Moderate High 20 80 0 0 Yes High Nil 100 0 0 0 Yes High Low 100 0 0 0 Yes High Moderate 20 80 0 0 Yes High High 0 100 0 0 Page 10 2175 Millar Rd. Smithers, BC V0J 2N6 Ph. (250) 877-6705FAX (250) 877-6805 Mule Deer Habitat Model - Morice and Lakes Forest Districts IFPA Thermal Cover The thermal cover value of a habitat is a function of forest cover and aspect, as well as snow interception cover. Habitats with good thermal/snow interception cover consist mainly of closed canopy mature forests that are multi-layered and have large trees with deep spreading crowns for optimal snow loading. Figure 2 describes the probable forested thermal cover value as a function of aspect and the ability of forest cover attributes to provide both thermal cover and the interception of snow. One further qualifier to this node is required for Agricultural Land Reserve (ALR) lands. PEM requires forest cover data to predict a site series. ALR Lands often do not have forest cover mapping, so the PEM includes these lands under the broad heading of “development”. The “development” type also includes municipalities and roads, but is clearly not the same type of habitat for mule deer in the winter as are agricultural lands. To accommodate this mapping artefact in the habitat model, an ALR layer was used to identify lands under the development label that are agricultural. In the model “development” site types result in a “Nil” value for thermal cover ;except where they intercept with ALR lands, where they are assigned a “low” value. The ALR lands may have even greater habitat value (than “low”), but at least this permeation in the habitat model enables these site types to be considered as potentially contributing to mule deer winter range. Example Relationship: Conditional Probability Table Where: MDTCV = Mule deer forested thermal cover value CCV = Thermal/snow interception value ALR = Land in the agricultural land reserve ASPECT = Aspect: Warm, cool and none (for slope<10) Forested Therm al Cover for Cool or Flat Aspect, Non-ALR areas 100% Forested Therm al Cover for Warm Aspect, Non-ALR areas 100% Moderate 50% oo d G M od er at e G N oo d 0% M od er at e 0% Po or 25% Po or Nil 25% Therm al/Snow Interception Low il Nil High 75% N Low 50% Probability Moderate il Probability High 75% Therm al/Snow Interception Figure 2. Forested thermal cover habitat value based on aspect and the ability of forest cover attributes to provide thermal cover and the interception of snow. Accessible Forage In winter mule deer eat a variety of shrubs, some herbs and litterfall of arboreal lichens and conifer needles; the proportions of which depend on forage accessibility. Forage accessibility is dependent on both forage abundance and snow depth. Forage abundance is based on litterfall estimates and on winter forage vegetation cover by PEM derived site series and structural stage. Snow depth is evaluated using biogeoclimatic subzones, elevation, and aspect. Page 11 2175 Millar Rd. Smithers, BC V0J 2N6 Ph. (250) 877-6705FAX (250) 877-6805 Mule Deer Habitat Model - Morice and Lakes Forest Districts IFPA For sites with “Nil” forage abundance, the accessible forage is “Nil” and for “Low” forage abundance, accessible forage is “Low” for all snow depths. Accessible forage for “moderate” and “good” late winter forage abundance sites is dependent on effective snow depth (See Figure 3). Example Relationship: Conditional Probability Table Where: MDFAV = Accessible forage by mule deer in winter FH = Late winter forage abundance Accessible Forage for Site w ith Moderate Forage Abundance 100% High 75% Moderate Low 50% Nil 25% 100% High 75% Moderate Low 50% Nil 25% Snow Depth ee p D M od er at e Ve ry Sh al lo w Sh al lo w ee p D Sh al lo w 0% M od er at e Ve ry Sh al lo w 0% Accessible Forage for Site w ith High Forage Abundance Probability Probability SD = Snow depth (as defined by BEC, aspect, and elevation) Snow Depth Figure 3. Accessibility of forage in late winter for sites rated for moderate and high abundance forage and modified by effective snow depth. Thermal /Snow Interception Value In winter, habitat that is non-forested or has an open canopy is generally poor habitat for mule deer due to reduced mobility in the accumulated snow. By contrast, closed or multi-strata forests provide thermal cover by acting as a break to the winter and creating a microclimate under the forest canopy; and, they provide snow interception cover by collecting snow in the forest canopy rather than on the forest floor Thermal/snow interception values are predicted based on forest type and crown closure attributes of the forest cover mapping as illustrated in Good Moderate Poor Nil Figure 4. Example Relationship: Conditional Probability Table Where: CCV = Thermal/snow interception value CRNCL_CL= Crown closure class FT1 = Forest Type Page 12 2175 Millar Rd. Smithers, BC V0J 2N6 Ph. (250) 877-6705FAX (250) 877-6805 Mule Deer Habitat Model - Morice and Lakes Forest Districts IFPA Therm al/Snow Interception For Forest Types w ith Crow n Closure Class 3 to 5 100% 25% 25% Forest Type C N Forest Type Good Moderate s on _f or es te d on i fe ro us s on _f or es te d ec id uo u D N N C M ix ed 0% on i fe ro us s on _f or es te d ec id uo u D on i fe ro us M ix ed 0% 50% ec id uo u 0% 50% 75% D 25% 75% M ix ed 50% Therm al/Snow Interception For Forest Types w ith Crow n Closure Class 6+ 100% Probability Probability 75% C Probability Therm al/Snow Interception For Forest Types w ith Crow n Closure Class 1 and 2 100% Poor Forest Type Nil Figure 4. Probable thermal/snow interception value based on forest type and crown closure classes. Late Winter Forage Abundance The late winter forage abundance node evaluates the total abundance of forage from all sources (arboreal and conifer litterfall and ground vascular vegetation). Note that ALR lands are automatically assigned a late winter forage value of “low” (rather than “nil” as other “development” polygon are). Example Relationship: Conditional Probability Table Where: CCV = Thermal/snow interception value Litterfall = Litterfall MD_W_FOR = Winter ground forage ALR = ALR (Agricultural Land Reserve) Table 3. Conditional Probability Table showing the probability of late winter forage abundance for nonALR lands. Late Winter Forage Abundance Winter Ground Forage Litterfall Nil Low Moderate High Nil Nil 100 0 0 0 Nil Low 100 0 0 0 Nil Moderate 100 0 0 0 Nil High 100 0 0 0 Low Nil 0 30 70 0 Low Low 0 0 100 0 Page 13 2175 Millar Rd. Smithers, BC V0J 2N6 Ph. (250) 877-6705FAX (250) 877-6805 Mule Deer Habitat Model - Morice and Lakes Forest Districts IFPA Low Moderate 0 40 60 0 Low High 0 5 35 60 Moderate Nil 0 20 80 0 Moderate Low 0 0 100 0 Moderate Moderate 0 0 100 0 Moderate High 0 0 20 80 High Nil 0 10 40 50 High Low 0 0 30 70 High Moderate 0 0 10 90 High High 0 0 0 100 Effective Snow Depth Effective snow depth is predicted based on biogeoclimatic subzone and modified by elevation and aspect. Warm aspects and lower elevation habitats (< 1000 m) tend to have lower snow accumulation than other sites (see Table 4). There are no subzones on in the study area that are considered “Moderate” snow depth for mule deer. Example Relationship: Conditional Probability Table Where: ESD = Effective snow depth ASPECT= Aspect SNOW = Snow depth by biogeoclimatic subzone ELE = Elevation (<1000m or >=1000m) Table 4. Conditional probabilities used within Netica to model effective snow depth from snow depth by biogeoclimatic subzone, modified by aspect and elevation. Effective Snow Depth Snow Depth Elevation Aspect Very Shallow Shallow Moderate Deep Shallow <1000 Cool 0 100 0 0 Shallow <1000 Warm 100 0 0 0 Shallow <1000 None 0 100 0 0 Shallow >=1000 Cool 0 100 0 0 Shallow >=1000 Warm 50 50 0 0 Shallow >=1000 None 0 100 0 0 Deep <1000 Cool 0 0 0 100 Deep <1000 Warm 0 0 0 100 Deep <1000 None 0 0 0 100 Deep >=1000 Cool 0 0 0 100 Deep >=1000 Warm 0 0 0 100 Deep >=1000 None 0 0 0 100 Page 14 2175 Millar Rd. Smithers, BC V0J 2N6 Ph. (250) 877-6705FAX (250) 877-6805 Mule Deer Habitat Model - Morice and Lakes Forest Districts IFPA Crown Closure Class Crown closure affects thermal cover value in winter by reducing wind chill by minimizing air movement, and by reducing radiation to the open sky during low temperatures. High crown closure can be provided by mature and old growth forests that have large trees with deep, spreading crowns and a multi-layered canopy and by dense stands of pole/sapling stage forest. The crown closure also affects how much snow interception there is. The higher the crown closure, the better the snow interception. Better snow interception results in reduced snow depth under the canopy. As well, when snow accumulates on the canopy and then falls to the ground, the snow tends to pack, which can make travel relatively easier than in open areas. Crown closure classes are summarized in this node as: class 0 (no canopy), class 1 and 2 (low canopy), class 3 to 5 (moderate canopy), and class 6+ (high canopy). Variables: CRNCL_CL = Crown Closure Class Forest Type The forest type node is summarized from the forest cover data into four types: coniferous (>=80% coniferous species), mixed (20%> coniferous species <80%), deciduous (>=80% deciduous species), non-forested (no tree species). Variables: FT1 = Forest Type Aspect Aspect is simplified into three states: warm (136° to 270°), cool (271° to 135°), and “none” (sites that have slope less than 10°). One assumption is that snow depth or condition is not affected by aspect on sites with slopes less than 10°. Another assumption is that cool aspects tend to maintain a relatively higher snow depth than do warm aspects. Warm aspects are assumed to be favourable to mule deer because of higher snowmelt due to insolation. Variables: ASPECT = Aspect classes Elevation Elevations greater than or equal to 1000m are considered to have too great a snow load for mule deer in winter. In the study area, this elevation band tends to be located near the upper boundary of the SBSmc2 or the SBSwk3 subzones. Variables: ELE = Elevation Snow Depth Snow depth is derived from the biogeoclimatic subzone line mapping. Three categories of snow depth are used: shallow, moderate, and deep. These categories were delineated based on mule deer response to snow loads. The SBPSmc and SBSdk subzones are considered to have shallow snow depths, less than 60 cm. All other subzones are considered to have deep snow depths, greater than 60cm persisting for more than 2 weeks, and are considered to be unsuitable. Variables: SNOW = Snow Depth by biogeoclimatic zone Litterfall Litterfall provides a measure of the total potential of the stand to provide litterfall forage from either arboreal lichens or from conifer tree species. The respective litterfall ratings are effectively submodels. Page 15 2175 Millar Rd. Smithers, BC V0J 2N6 Ph. (250) 877-6705FAX (250) 877-6805 Mule Deer Habitat Model - Morice and Lakes Forest Districts IFPA Table 5 describes how the values for arboreal lichen and conifer literfall are combined to predict total litterfall. Example Relationship: Conditional Probability Table Where: Litterfall = Litterfall (total forage) LL = Lichen litterfall CL = Conifer litterfall Table 5. Conditional Probability Table predicting total lichen and conifer litterfall. Lichen Litterfall Conifer Litterfall Litterfall Total Nil Low Moderate High Nil Nil 100 0 0 0 Nil Low 25 75 0 0 Nil Moderate 0 50 50 0 Nil High 0 25 50 25 Low Nil 25 75 0 0 Low Low 0 100 0 0 Low Moderate 0 40 60 0 Low High 0 0 40 60 Moderate Nil 0 40 60 0 Moderate Low 0 25 75 0 Moderate Moderate 0 0 100 0 Moderate High 0 0 25 75 High Nil 0 0 50 50 High Low 0 25 50 25 High Moderate 0 0 25 75 High High 0 0 0 100 Lichen Litterfall Lichen litterfall is estimated based on the results of the arboreal lichen abundance model which is derived from forest cover. The assumption is that the greater the arboreal lichen abundance the greater the probability that there will be lichen litterfall (See Figure 5). Example Relationship: Conditional Probability Table Where: LL = Lichen litterfall ARB = Arboreal lichen abundance Page 16 2175 Millar Rd. Smithers, BC V0J 2N6 Ph. (250) 877-6705FAX (250) 877-6805 Mule Deer Habitat Model - Morice and Lakes Forest Districts IFPA Lichen Litterfall Probability 100% High 75% Moderate 50% Low 25% Nil 0% Nil Low Moderate High Arboreal Lichen Abundance Figure 5. Probability of lichen litterfall associated with stands of nil, low, moderate, or high arboreal lichen abundance. Conifer Litterfall The probability of conifer litterfall is based on the “forest type” and the age class of the stand. Note that in this model “forest type” is a measure of the leading and secondary tree species’ ability to produce conifer forage, from “poor” to “very good”. For age class 0, the conifer litterfall is “nil” and for age classes 1, 2, and 3, the conifer litterfall is low, regardless of “forest type”. Only once a stand reaches age class 4 and older, is there conifer litterfall (See High Moderate Low Nil Figure 6). Example Relationship: Conditional Probability Table Where: CL = Conifer litterfall AGE_CLS = Age Class FT = forest type Conifer Litterfall for Age Class 4 Stands Conifer Litterfall for Age Class 5 and 6 Stands Conifer Litterfall for Age Class 7+ Stands 75% 75% 75% 50% 50% 25% 25% 0% 0% Poor Moderate Good Forest Type Very Good High Probability 100% Probability 100% Probability 100% 50% 25% 0% Poor Moderate Good Forest Type Moderate Low Very Good Poor Moderate Good Forest Type Nil Page 17 2175 Millar Rd. Smithers, BC V0J 2N6 Ph. (250) 877-6705FAX (250) 877-6805 Very Good Mule Deer Habitat Model - Morice and Lakes Forest Districts IFPA Figure 6. Probability of conifer litterfall associated with “forest type” class poor to very good by age class groupings 4, 5-6, and 7+. Arboreal Lichen Abundance The arboreal lichen abundance node is a parent node whose values are predicted in the arboreal lichen Netica model. This node has the values of high, moderate, low, and nil potential for arboreal lichen abundance. Variables: ARB = Arboreal lichen abundance Forest Type “Forest type” looks at leading and secondary tree species and provides a rating of their combined value as preferable conifer forage species for mule deer. The classes are: poor, moderate, good, and very good. For example, Douglas-fir stands are highly preferable, with other conifers such as spruce, subalpine fir, and hemlock less preferable. Example Relationship: Conditional Probability Table Where: FT = Forest type SPC_1 = Leading species SPC_2 = Secondary species Winter Ground Forage Winter ground forage is derived from the PEM. This is a parent node that is read in from a lookup table where site series and structural stage groupings are rated for their abundance of preferred ground forage. A preferred winter vascular forage vegetation list was compiled from the literature. Vegetation on this list was summarized for presence and abundance by site series. These presence and abundance values were then summarized into ratings of Nil, Low, Moderate, or High abundance. Variables: MD_W_FOR = Winter Ground Forage Leading Species and Secondary Species The leading and secondary species nodes are read in from the forest cover data. This information is used to describe the “forest type” rating of conifer forage species, which combined with age class will give an indication of conifer litterfall. Variables: SPC_1 = Leading species SPC_2 = Secondary species Age Class Age class is from the projected age class in the forest cover data. This is a parent node and is grouped into 5 class categories: 0, 1 to 3, 4, 5 to 6, and 7+. Variables: AGE_CLS = Age Class Page 18 2175 Millar Rd. Smithers, BC V0J 2N6 Ph. (250) 877-6705FAX (250) 877-6805 Mule Deer Habitat Model - Morice and Lakes Forest Districts IFPA ALR Agricultural land reserve (ALR) is a parent node that modifies late winter forage abundance and thermal cover values. ALR Lands often do not have forest cover mapping, so the PEM includes these lands under the broad heading of “development”. The “development” type also includes municipalities and roads, but is clearly not the same type of habitat for mule deer in winter as are agricultural lands. To accommodate this mapping artefact in the habitat model, an ALR layer was used to identify lands under the development label that are agricultural. Instead of receiving a “nil” values for late winter forage abundance and thermal cover as for other polygons with the “development” site type, ALR lands are assigned values of “low” for these model nodes. Variables: ALR = ALR (Agricultural Land Reserve) Island Islands are not used by mule deer in winter. So, while other habitat features may point to a higher habitat suitability rating for a particular polygon, location on an island r5esults in a downgrade to “nil” value in the final winter habitat suitability node. Variables: ISL = Island Computing Output for Habitat Value Non-spatial mule deer foraging habitat suitability will be the output from the model for each polygon. Sensitivity Analysis Testing and Validation The models should be viewed as hypotheses of species-habitat relationships rather than statements of proven cause and effect relationships. Their value is to serve as a basis for improved decision making and increased understanding of habitat relationships because they specify hypotheses of habitat relationships that can be tested and improved. There are several levels at which the models should be validated. The first is an ongoing process during model development during which we have tested the model to ensure that it is acting in a manner that we want it to. The second level is to test the model assumptions and output through field-testing or verification. The mule deer winter model has been tested within Ardea. A working review of the data and mapped output was done with MWLAP representatives from Smithers, BC. Research Needs for Model Verification The performance of a model should be tested against population data, preferably estimates of density or reproductive success, to translate the perception of habitat quality or suitability into differential use of habitat (Brooks 1997). The spatial configuration of habitat will affect the spatial spread and use of a population in a heterogeneous environment (Söndgerath and Schröeder 2002). The output of this model provides a habitat suitability value; however, the pattern and amount of the habitat with respect to caribou use is not evaluated. The importance of the landscape structure varies according to the demographic characteristics of the population (Söndgerath and Schröeder 2002, With and King 1999). Implications to Forest Management Management levers exist within the mule deer model. In winter, mule deer are highly associated with low elevation warm aspect sites that support forage vegetation and are either in close proximity to forested Page 19 2175 Millar Rd. Smithers, BC V0J 2N6 Ph. (250) 877-6705FAX (250) 877-6805 Mule Deer Habitat Model - Morice and Lakes Forest Districts IFPA cover or have forested cover. Forest age, canopy closure, tree species, abundance of forage species (such as shrub species) and configuration of forested habitats adjacent to open south facing slopes are all attributes that can be managed to some extend within mule deer winter range. Page 20 2175 Millar Rd. Smithers, BC V0J 2N6 Ph. (250) 877-6705FAX (250) 877-6805 Mule Deer Habitat Model - Morice and Lakes Forest Districts IFPA MULE DEER MODEL REFERENCES Armleder, H.M. and R.J. Dawson. 1992. Logging on mule deer winter range: an integrated management approach. Forestry Chron. 68(1):132-137. Armleder, H.M., R.J. Dawson and R.N. Thomson. 1986. Handbook for timber and mule deer management coordination on winter ranges in the Cariboo Forest Region. Min. of For., Victoria, BC. Land Management Handbook No. 13. 98pp. Armleder, H.M., M.J. Waterhouse, D.G. Keisker and R.J. Dawson. 1994. Winter habitat use by mule deer in the central interior of British Columbia. Can. J. Zool. 72:1721-1725. Bloom, A. 1978. Sitka black-tailed deer winter range in the Kadashan Bay area, southeast Alaska. J. Wildl. Manage. 42(1):108-112. Blower, D. 1982. Memorandum on key winter forage plants for BC ungulates. Min. Env., Lands and Parks. 1 pp. Brown, C.G. 1992. Movement and migration patterns of mule deer in southeastern Idaho. J. Wildl. Manage. 56(2):246-253. Bunnell, F.L. 1990. Ecology of black-tailed deer. Pages 31-63 in. J.B. Nyberg and D.W. Janz, eds. Deer and elk habitats in coastal forests of southern British Columbia. Min. of For. Victoria, BC. Special Report Series No.5. 310pp. Carpenter, L.H, O.C. Wallmo and R.B. Gill. 1979. Forage diversity and dietary selection by winter mule deer. J. Range Manage. 32(3):226-229. Carson, R.G. and J.M. Peek. 1987. Mule deer habitat selection patterns in northcentral Washington. J. Wildl. Manage. 51(1):46-51. Collins, W.B. and P.J. Urness. 1983. Feeding behaviour and habitat selection of mule deer and elk on northern Utah summer range. J. Wildl. Manage 47(3): 646-663. Cowan, I.McT. and C.G. Guiget. 1978. The mammals of British Columbia. BC. Prov. Mus., Victoria, BC. Handb. No. 11. 7th printing. 414pp. Deschamp, J.A., P.J. Urness and D.D. Austin. 1979. Summer diets of mule deer from lodgepole pine habitats. J. Wildl. Manage. 43(1):154-161. Einarsen, A.S. 1946. Management of black-tailed deer. J. Wildl. Manage. 10(1):54-59. Garrot, R.A., G.C. White, R.M. Bartmann, L.H. Carpenter, and A.W. Alldredge. 1987. Movements of female mule deer in northwest Colorado. J. Wildl. Manage. 51(3):634-643. Gilbert, P.F., O.C. Wallmo and R.B. Gill. 1970. Effect of snow depth on mule deer in Middle Park, Colorado. J. Wildl. Manage. 34(1):15-23 Harestad, A.S. 1979. Seasonal movements of black-tailed deer on northern Vancouver Island. Min. of Env. Fish and Wildl. Rep. No. R-3. 99pp. Harestad, A.S., J.A. Rochelle and FL. Bunnell. 1982. Old-growth forests and black-tailed deer on Vancouver Island. Trans. N. Amer. Wildl. Nat. Res. Conf. 47:343-352. Ihsle Pac, I.H., W.F. Kasworm, L.R. Irby and R.J. Mackie. 1988. Ecology of the mule deer, Odocoileus hemionus, along the east front of the Rocky Mountains, Montana. Can. Field. Nat. 102:227-236. Kirchhoff, M.D. and J. W. Schoen. 1983. Black-tailed deer use in relation to forest clear-cut edges in southeastern Alaska. J. Wildl. Manage. 47(2):497-501. Kirchhoff, M.D. and J. W. Schoen. 1987. Forest cover and snow: Implications for deer habitat in southeast Alaska. J. Wildl. Manage. 51(1):28-33. Kufeld, R.C., D.C. Bowden and D.L. Schrupp. 1988. Habitat selection and activity patterns of female mule deer in the Front Range, Colorado. J. Range. Manage. 41(6):515-522. Page 21 2175 Millar Rd. Smithers, BC V0J 2N6 Ph. (250) 877-6705FAX (250) 877-6805 Mule Deer Habitat Model - Morice and Lakes Forest Districts IFPA Kufeld, R.C., D.C. Bowden and D.L. Schrupp. 1989. Distribution and movements of female mule deer in the Rocky Mountain foothills. J. Wildl. Manage. 53(4):871-877. Loft, E.R. and J.W. Menke. 1984. Deer use and habitat characteristics of transmission-line corridors in a Douglas-fir forest. J. Wildl. Manage. 48(4) 1311-1315. Loft, E.R., R.C. Bertram and D.L. Bowman. 1989. Migration patterns of mule deer in the central Sierra Nevada. Calif. Fish and Game 75(1):11-19. Ministry of Environment, Lands and Parks. 1994. Birds, mammals, reptiles and amphibians not at risk in British Columbia: the revised yellow list, 1994. Habitat Protection Branch. Wildlife Working Rep. No. WR-66. Victoria, BC. 72pp. Nyberg, B.J, S.R. McNay, M.D. Kirchoff, R.D. Forbes, F.L. Bunnell and E.L. Richardson. 1989. Integrated management of timber and deer: coastal forests of British Columbia and Alaska. US Dept. of Agriculture. PNW-GTR-226. 21 pp + append. Parker, K.L., C.T. Robbins and T.A. Hanley. 1984. Energy expenditures for locomotion by mule deer and elk. J. Wildl. Manage. 48(2):474-488. Schoen, J.W., M.D. Kirchhoff and M.H. Thomas. 1985. Seasonal distribution and habitat use by Sitka black-tailed deer in southeastern Alaska. Final report, Federal Aid in Wildlife Restoration Projects. Juneau, Alaska. 44pp. Stevens, V. 1995. Wildlife diversity in British Columbia: Distribution and habitat use of amphibians, reptiles, birds, and mammals in biogeoclimatic zones. Res. Br., BC Min. For. Wildl. Br. BC Min. Environ, Lands and Parks., Victoria, BC. Working Pap. 04/1995. 288pp. Thomas, T.R. and L.R. Irby. 1990. Habitat use and movement patterns by migrating mule deer in southeastern Idaho. Northwest Science. 64(1)19-27. Torbit, S.C., L.H. Carpenter, D.M. Swift, and A.W. Alldredge. 1985. Differential loss of fat and protein by mule deer during winter. J. Wildl. Manage. 49(1):80-85. Wallmo, O.C. and J.W. Schoen. 1980. Response of deer to secondary forest succession in southeast Alaska. Forest Sci. Vol 26, No. 3 pp.448-462. Willms, W., A. McLean and R. Ritcey. 1976. Feeding habits of mule deer on fall, winter, and spring ranges near Kamloops, British Columbia. Can. J. Anim. Sci. 56:531-542. Zalunardo, R.A. 1965. The seasonal distribution of a migratory mule deer herd. J. Wildl. Manage. 29(2):345-351. Page 22 2175 Millar Rd. Smithers, BC V0J 2N6 Ph. (250) 877-6705FAX (250) 877-6805